Pursuant to Section 13 or 15(d) of the Securities Exchange Act of 1934
| Federally Chartered Corporation | 000-50231 | 52-0883107 | ||
| (State or other jurisdiction of incorporation) |
(Commission File Number) |
(I.R.S. Employer Identification No.) |
| 3900 Wisconsin Avenue, NW, Washington, | ||
| District of Columbia | 20016 | |
| (Address of principal executive offices) | (Zip Code) |
| o | Written communications pursuant to Rule 425 under the Securities Act (17 CFR 230.425) |
| o | Soliciting material pursuant to Rule 14a-12 under the Exchange Act (17 CFR 240.14a-12) |
| o | Pre-commencement communications pursuant to Rule 14d-2(b) under the Exchange Act (17 CFR 240.14d-2(b)) |
| o | Pre-commencement communications pursuant to Rule 13e-4(c) under the Exchange Act (17 CFR 240.13e-4(c)) |
| FEDERAL NATIONAL MORTGAGE ASSOCIATION |
||||
| By: | /s/ Beth A. Wilkinson | |||
| Beth A. Wilkinson | ||||
| Executive Vice President - General Counsel and Corporate Secretary | ||||
| Exhibit No. | Description of Exhibit | |
99.1
|
Executive Summary to Report to the Special Review Committee of the Board of Directors of Fannie Mae, dated February 23, 2006, from Paul, Weiss, Rifkin, Wharton & Garrison and Huron Consulting Group, Inc. (the “Report”) | |
99.2
|
Report, excluding Executive Summary to the Report | |
99.3
|
Press release, dated February 23, 2006, setting forth Statement by Stephen B. Ashley, Chairman of the Board of Directors of Fannie Mae | |
99.4
|
Press release, dated February 23, 2006, setting forth Statement by Daniel H. Mudd, President and Chief Executive Officer of Fannie Mae |
EXHIBIT 99.1 A REPORT TO THE SPECIAL REVIEW COMMITTEE OF THE BOARD OF DIRECTORS OF FANNIE MAE EXECUTIVE SUMMARY PAUL, WEISS, RIFKIND, WHARTON & GARRISON LLP Warren B. Rudman Alex Young K. Oh Robert P. Parker Daniel J. Kramer HURON CONSULTING GROUP INC. George E. Massaro Jeffrey H. Ellis February 23,2006
TABLE OF CONTENTS I. INTRODUCTION........................................................................................ 1 II. SUMMARY OF FINDINGS................................................................................ 3 A. Fannie Mae's Application of FAS 91......................................................... 5 1. 1998 FAS 91 Adjustment............................................................ 6 2. Management's Development and Implementation of the Amortization Policy............ 7 B. Fannie Mae's Application of FAS 133........................................................ 8 C. Conclusions About Other Accounting Issues.................................................. 10 1. Accounting for the Allowance for Loan Losses..................................... 10 2. Accounting for Dollar Rolls...................................................... 11 3. Accounting for Forward Commitments............................................... 12 4. Classification of Securities Held in Portfolio................................... 13 5. Recognition of Interest Expense and Income....................................... 14 6. Accounting for Other-Than-Temporary Impairment of Manufactured Housing Bonds and Aircraft Asset-Backed Securities....................................... 14 7. Accounting for Investments in Interest-Only Mortgage-Backed Securities ("IO MBS")...................................................................... 15 8. Securitization of Wholly-Owned MBS............................................... 15 9. Accounting for Income Tax Reserves and Certain Tax-Advantaged Transactions....... 17 10. Accounting for Insurance Products................................................ 18 11. Accounting for Out-of-Portfolio Securitization ("Portfolio Pooling System")...... 18 12. The Debt Repurchase ("Buyback") Program.......................................... 19 13. Accounting for the Amortization of Callable Debt Expenses........................ 19 14. Minority Lending Initiative...................................................... 20 15. Accounting for Realignments and the Security Master Project...................... 21 16. Accounting for Investments in Affordable Housing Partnerships.................... 21 III. CORPORATE GOVERNANCE AND INTERNAL CONTROLS........................................................ 22 A. Board of Directors........................................................................ 22 B. Office of the Chairman.................................................................... 23 C. Internal Audit............................................................................ 24 D. Ethics and Compliance Functions........................................................... 24 E. The Controller's Office................................................................... 26 IV. EXECUTIVE COMPENSATION............................................................................ 26 -i-
V. FANNIE MAE'S INVESTIGATION OF ROGER BARNES'S ALLEGATIONS.......................................... 27 VI. MANAGEMENT'S CONDUCT DURING OFHEO'S SPECIAL EXAMINATION........................................... 28 VII. OTHER ALLEGATIONS................................................................................. 30 A. Issues Raised by Current and Former Fannie Mae Employees................................... 30 B. Fannie Mae's Equity Investments in Gulf Bank............................................... 31 -ii-
EXECUTIVE SUMMARY I. INTRODUCTION The Office of Federal Housing Enterprise Oversight ("OFHEO") began a special examination of Federal National Mortgage Association ("Fannie Mae" or the "Company") in November 2003 (the "Special Examination"). Almost one year later, OFHEO issued a report of its findings to date as of September 17, 2004 (the "OFHEO Report"). Among other things, the OFHEO Report found that the Company's accounting in various respects was not consistent with generally accepted accounting principles ("GAAP") and was motivated by management's desire to portray Fannie Mae "as a consistent generator of stable and growing earnings," and by an "executive compensation structure that rewarded management for meeting goals tied to earnings-per-share, a metric subject to manipulation by management."(1) OFHEO also concluded in its Report that the Company had "dysfunctional accounting policy development, key person dependencies, and poor segregation of duties" that contributed to accounting failures and safety and soundness problems.(2) In September 2004, the Special Review Committee of the Board of Directors of Fannie Mae (the "SRC") engaged former Senator Warren B. Rudman and Paul, Weiss, Rifkind, Wharton & Garrison LLP (collectively, "Paul, Weiss") to conduct an independent investigation of, among other things, the issues that were raised in the OFHEO Report and to report our findings and conclusions to the SRC. This Executive Summary highlights the key findings and conclusions of the Paul, Weiss investigation. The full findings and conclusions are contained in a Report, which we also publish today.(3) The scope of our investigation was initially defined by an agreement dated September 27, 2004 between the Board of Directors of Fannie Mae (the "Board") and OFHEO, which was supplemented by an agreement dated March 7, 2005, between OFHEO and the Board (collectively, "OFHEO Agreements"). The issues raised in the OFHEO Agreements primarily concerned the Company's accounting, internal controls, and corporate governance and structure. The scope of our investigation, however, was not limited to the issues in the OFHEO Agreements. In fact, the SRC did not place any - -------------- (1) OFHEO Report of Findings to Date in the Special Examination of Fannie Mae, dated Sept. 17, 2004, available at http://www.ofheo.gov/media/pdf/ FNMfindingstodate17sept04.pdf., Executive Summary at i. (2) Id. at viii. (3) See "A Report to the Special Review Committee of the Board of Directors of Fannie Mae" (the "PW Report" or the "Report"). The three-volume Appendix to the PW Report includes sample documents of interest that are discussed in the PW Report, and certain submissions that Paul, Weiss received from attorneys who represent former Company officers. 1
limitations on our inquiry and instructed us to follow whatever leads we discovered during the course of our investigation.(4) We received the full support of the SRC and the Board during the course of our review, and the SRC instructed the Company to cooperate fully with our investigation. Pursuant to the OFHEO Agreements, and with the approval of both the SRC and OFHEO, we retained the forensic accounting services of Huron Consulting Group Inc. ("Huron") to assist in our investigation. During the course of the investigation, Paul, Weiss and Huron collectively reviewed more than four million pages of hardcopy and electronic documents and conducted more than 240 interviews.(5) The accounting opinions expressed in this Report are Huron's. Neither Paul, Weiss nor Huron, however, conducted an audit of the Company's financial statements. The task of preparing restated financial statements remains that of the Company, and the task of auditing those financial statements remains that of the Company's independent auditor, Deloitte & Touche LLP. Our engagement was unusual in that the OFHEO Agreements required the Company, contemporaneously with our investigation, to undertake prompt remedial measures with respect to Fannie Mae's accounting processes and procedures and corporate governance. As we detail in the Chapter of our Report addressing Corporate Governance and Internal Controls, the Board and the Company, with our input, have diligently pursued their obligations under the OFHEO Agreements and many remedial measures are already underway. As a result, many recommendations that we would have made are already in the process of being implemented. Accordingly, while we document in the Report many of the significant corrective measures the Company has taken, we do not make significant additional recommendations. Our factual findings and conclusions focus on management's intent and motive with respect to the transactions we reviewed. Paul, Weiss's mandate, however, did not include determining whether any of the conduct we reviewed constituted a violation of law or breach of professional standards or whether the Company may properly assert legal claims against any individuals or entities.(6) We leave to others the task of determining the consequences that should flow from our factual findings. - ---------------- (4) The SRC also specifically asked Paul, Weiss to review allegations made by a former-employee, Roger Barnes, including how the Company addressed Barnes's allegations, and any other matters raised anonymously by employees and former employees. (5) As detailed further in Chapter II of the Report, our document review is ongoing. As recently as February 16, 2006, the Company brought to our attention the existence of new materials that could be relevant to our investigation. If necessary after reviewing all of the materials produced by the Company, we will supplement our findings and conclusions in this Report. (6) For example, while the SRC was initially formed in January 2004 in response to a shareholder demand letter, we were not retained until September 2004 and were not 2
As with any private investigation, we relied on the voluntary cooperation of the Company, its employees (both current and former), and its agents. We did not have the power to compel testimony or production of documents. While we received good cooperation from the Company and its current employees, counsel and auditors, we were not able to interview certain former employees. Most significantly, Timothy Howard, a former Vice Chairman, Chief Financial Officer, and member of Fannie Mae's Board, declined our repeated requests for interviews. Similarly, Leanne G. Spencer, a former Senior Vice President and Controller, cooperated with our investigation during its early stages but declined further interviews after we became aware of a critical document in her files, which Spencer had failed to produce in response to Paul, Weiss's document requests to the Company. Finally, under the SRC's direction, we cooperated fully with the United States Department of Justice ("DOJ"), the Securities and Exchange Commission ("SEC"), OFHEO, and the Public Company Accounting Oversight Board ("PCAOB"). Almost immediately after our retention, beginning in October 2004, we met with and regularly briefed the regulatory agencies on the progress of our investigation. II. SUMMARY OF FINDINGS The Company, under the Board's direction and with OFHEO's input, has undergone an extensive transformation both in personnel and structure since September 2004. Since that time, as we observe in the Chapter describing Corporate Governance and Internal Controls, there has been a dramatic shift in both the "tone at the top" and the Company's internal organization. During the course of our investigation, we communicated our findings to the SRC and the full Board, and the Company has not waited for the issuance of the PW Report before making necessary changes. As a result, (1) the Company has disclosed the principal problematic accounting issues that are the subject of this Report,(7) (2) no member of management who we found knowingly participated in improper conduct continues to be an employee of the Company, and (3) as noted above, our suggestions for changes in corporate governance either have been implemented or are underway. We summarize below our principal conclusions about the Company's accounting practices, internal controls, and corporate governance and structure prior to 2005. We next summarize in detail each of the accounting issues and the related findings - ------------ asked to address the demand letter. We understand that the SRC and the Board are ably represented by other counsel in connection with the demand letter and with respect to pending civil actions, and it was not our role to advise the SRC or the Board in such matters. (7) As noted above, management and its current outside auditor are engaged in a restatement effort that involves a detailed review of all of the Company's accounting policies and practices. This process could result in additional matters being identified that are not addressed in this Report. 3
and conclusions contained in the Report. Due to the complexity of both the accounting and factual issues addressed in the PW Report, however, no summary can serve as an adequate substitute for reading the chapters that contain a full exposition of both the facts and our analyses. Our principal conclusions with respect to Fannie Mae's historical accounting practices, internal controls, corporate governance, and structure prior to 2005, are as follows: First, management's accounting practices in virtually all of the areas that we reviewed were not consistent with GAAP, and, in many instances, management was aware of the departures from GAAP. Management often justified departures from GAAP based upon materiality assessments that were not comprehensive, the need to accommodate systems inadequacies, the unique nature of Fannie Mae's business, or "substance over form" arguments. For example, management unjustifiably departed from GAAP with respect to: (1) its implementation of Statement of Financial Accounting Standards ("FAS") 133 in order to minimize earnings volatility and to avoid having to make investments in new systems to accommodate the standard; (2) its application of FAS 91, because compliance with FAS 91 would have resulted in greater earnings volatility than management had wanted; and (3) its approach to accounting for interest-only securities in combination with other securities to avoid impairment write-downs that would have been required under GAAP for the interest-only securities. Second, except for one instance in connection with the 1998 financial statements, we did not find evidence supporting the conclusion that management's departures from GAAP were motivated by a desire to maximize bonuses in a given period. We did, however, find evidence amply supporting the conclusion that management's adoption of certain accounting policies and financial reporting procedures was motivated by a desire to show stable earnings growth, achieve forecasted earnings, and avoid income statement volatility. For example, management's strategic execution of debt buybacks, purchase of finite risk insurance products, and acceleration of certain expenses related to corporate-owned life insurance, among other strategies, helped the Company to show a trend of stable earnings growth from 2001 to 2004. Similarly, management did not alter its accounting practice for the allowance for loan losses, even though management was aware that the allowance was overstated, because the reduction of the allowance would have generated a "spike" in income. Third, employees who occupied critical accounting, financial reporting, and audit functions at the Company were either unqualified for their positions, did not understand their roles, or failed to carry out their roles properly. This deficiency was most clearly manifested by employees who occupied senior positions in the Office of the Controller ("Controller's Office") and the Office of Auditing ("Internal Audit"). In addition, the resources devoted to accounting, financial reporting, and audit functions were not sufficient to address the needs of an institution as large and complex as Fannie Mae. This was apparent, for example, in our review of the Company's implementation of 4
FAS 149 (concerning the accounting for forward commitments), in which resource constraints led to a haphazard adoption of the standard. Fourth, the information that management provided to the Board of Directors with respect to accounting, financial reporting, and internal audit issues generally was incomplete and, at times, misleading. Management tightly controlled the information flow to the Board generally, and Howard, in particular, filtered the accounting and financial information the Board received. For example, management provided incomplete or misleading information in connection with (1) presentations regarding the 1998 amortization expense calculation; (2) briefings requested by the Board concerning Freddie Mac's restatement announcement in 2003, and whether Fannie Mae had any similar accounting issues; and (3) a presentation regarding the Special Examination in 2004, where the Board was left with the incorrect impression that the Company's accounting under FAS 91 and FAS 133 was justifiable and defensible, and that no restatement would be required. Fifth, the Company's accounting systems were grossly inadequate. This fact became apparent in our review of several areas - most notably our review of the Company's accounting for premium and discount amortization under FAS 91, but also in connection with the Company's accounting under FAS 133 and FAS 149. The accounting for the Company's investments in affordable housing partnerships also was affected by systems limitations. Finally, we conclude that Howard, the former CFO, and Leanne Spencer, the former Controller, were primarily responsible for adopting or implementing accounting practices that departed from GAAP, and that they put undue emphasis on avoiding earnings volatility and meeting EPS targets and growth expectations. As for former Chairman and CEO Franklin D. Raines, we did not find that he knew that the Company's accounting practices departed from GAAP in significant ways. We did find, however, that Raines contributed to a culture that improperly stressed stable earnings growth and that, as the Chairman and CEO of the Company from 1999 through 2004, he was ultimately responsible for the failures that occurred on his watch. * * * * A. Fannie Mae's Application of FAS 91 We reviewed two primary issues with respect to management's application of FAS 91: first, we looked at management's support and motivation for its decision to record only $240 million in additional premium expense in the fourth quarter of 1998 when the Company's own analysis indicated it should have recorded $439 million in additional premium expense; second, we reviewed management's development and implementation of a purchase premium and discount amortization policy in 2000 (the "Amortization Policy" or "Policy") that included a "precision threshold" within which management retained substantial discretion not to make adjustments that were required under FAS 91. 5
1. 1998 FAS 91 Adjustment The Company was required under FAS 91 to amortize premiums and discounts on its loans and mortgage-backed securities ("MBS") using the "level yield" method. The application of the level yield method resulted in periodic adjustments to increase or decrease interest income to reflect, among other things, the effect on the cumulative amortization of premium and discount of differences in actual and estimated prepayments as a result of interest rate movements (this adjustment is referred to at Fannie Mae as "catch-up"). The amount generated for this catch-up adjustment for the fourth quarter of 1998 was $439 million in expense, which, under FAS 91, the Company was required to recognize in the same period. Howard and Spencer, however, recommended to the Office of the Chairman that the Company recognize only $240 million out of $439 million in expense and defer $199 million in expense to future periods. The reduction would be accomplished through periodic "on-top" entries. Recognizing the full $439 million in expense would have caused the Company to miss then forecasted earnings per share ("EPS") for 1998 of $3.22 per share, and also would have resulted in falling below the EPS-based threshold for triggering employees' bonus payments. We conclude that deferral of the $199 million in catch-up expense violated GAAP. Howard and Spencer also recommended and recorded other adjustments that had the effect of making up for the shortfall from forecasted EPS caused by recognizing the $240 million in catch-up expense for 1998. First, Howard and Spencer accelerated a planned change from a non-GAAP to GAAP method of accounting for the tax credits received in connection with the Company's investment in low income housing tax partnerships, which resulted in recognizing an extra year's worth of credits in 1998. The net after-tax effect of the change in accounting for investments in low income housing partnerships was $108 million in income. We conclude that the accounting method adopted by management was in accordance with GAAP, but that management's motive for accelerating the method change was to offset the EPS shortfall created by recording the $240 million in amortization expense. The second adjustment made by management was the reversal of $3.9 million in "aged balances" from a suspense account. The adjustment, which had no support, was recorded as "miscellaneous income" and served to incrementally increase EPS to $3.2309, which triggered maximum employee bonuses for 1998. Because the Company had already exceeded published analyst expectations of $3.22 for 1998 through management's other 1998 actions - i.e., recording only $240 million of catch-up expense and accelerating the recognition of tax credits - we infer that this unsupported $3.9 million entry can be explained only by a motive to increase EPS results from $3.22 to $3.2309, the minimum amount of EPS needed to trigger the maximum bonuses. Howard and Spencer then made incomplete and misleading disclosures to the Board about these entries in their reports on the 1998 financial results. For example, at the January 19, 1999 meeting of the Board of Directors, Howard's presentation omitted the fact that the Company's systems indicated that $439 million in catch-up expense 6
should be recorded, and misleadingly suggested that the $240 million in catch-up expense was recorded at management's option because there was "room" created by the recording of two years' worth of tax credits. Similarly, Spencer failed to inform the Audit Committee in February 1999 that the full catch-up expense adjustment should have been $439 million, and that, in fact, the Company's auditor had noted an audit difference for the unrecorded $199 million expense. Further, Spencer misleadingly described the reason for the ability to record an extra year's tax credits as due to improvements in the Company's systems and controls, rather than management's correction of historical accounting methodology that had not been in accordance with GAAP. Spencer did inform the Company's outside auditor of the decision to reduce interest and guaranty fee income by only $240 million in expense as opposed to the calculated $439 million in expense, and of the accounting change for the tax credits. The outside auditor noted an audit difference for the deferred catch-up expense amount of $199 million, and certified the Company's financial statements for 1998 without qualification. Also, the auditor's 1998 workpapers showed that the outside auditor reviewed the activity in the account from which the $3.9 million was reversed into income and noted an audit difference for the remaining balance in the suspense account. 2. Management's Development and Implementation of the Amortization Policy After noting an audit difference for the $199 million unrecorded amortization adjustment for 1998, the Company's outside auditor asked management to develop and formalize a policy concerning its FAS 91 calculations. Management, under the direction of Howard, developed an Amortization Policy and implemented it in December 2000. We conclude that the Policy was developed for the purpose of avoiding audit differences with the outside auditor, rather than for the purpose of complying with GAAP. For example, the Policy contained provisions that were inconsistent with GAAP, such as the provision creating a "precision threshold" within which management did not have to recognize adjustments that were otherwise required under FAS 91 on the grounds that all amounts within the threshold were the "functional equivalent of zero." Management, however, did review the significant terms of the Policy with the Company's auditor at the time of its adoption and we did not see any evidence that the auditor disagreed with its terms. Significantly, management disregarded the terms of the Policy when it did not suit its purpose. The most obvious example of management's disregard for its own policy was the recognition of catch-up that fell within the calculated range, which was supposedly the "functional equivalent of zero." Howard and Spencer also misled the Board about the purpose of the Policy and how it was implemented. Spencer made a presentation about the Amortization Policy to the Audit Committee in November 2003 in which she failed to disclose the fact that management's implementation of the Policy was not consistent with 7
the Policy's terms. Further, Spencer and Howard were both present at a July 19, 2004 joint meeting of the Audit and the Special Review Committees of the Board where one of the critical issues under discussion was OFHEO's potential allegation that management engaged in earnings management by inconsistently applying the Amortization Policy. Neither Spencer nor Howard disclosed at this meeting the fact that management had, in fact, applied the Policy inconsistently, and that OFHEO's allegations would find support in the facts. B. Fannie Mae's Application of FAS 133 The Company's outstanding debt grew dramatically during the 1990s (commensurate with the growth in its portfolio). The ability to hedge that debt against interest-rate risk was a substantial component of the Company's risk management strategy. The Company used derivatives to hedge the interest-rate risk associated with its debt, and the notional amount of its derivative portfolio also grew tremendously during the 1990s and into the 2000s. FAS 133, which was issued in 1998 and was adopted by the Company on January 1, 2001, required companies to recognize derivatives at fair value, with changes in fair value recognized in income. Companies could avoid the earnings volatility associated with FAS 133 by entering into transactions that qualified for hedge accounting. FAS 133 refers to this as "special accounting." We recognize that there has been substantial criticism of FAS 133 and, in particular, that some hold the view that FAS 133 injects inappropriate volatility into earnings. We are also aware that, in the wake of the SEC's announcement concerning errors in Fannie Mae's accounting under FAS 133, a number of companies have announced that they would restate their FAS 133 accounting. With respect to Fannie Mae's application of FAS 133, we conclude that management did not engage in mere innocuous practical interpretations or modest deviations from a strict reading of the standard. Rather, management's implementation of FAS 133 was motivated not only by a desire to avoid earnings volatility, but also by a desire to avoid substantial changes to the Company's business methods, and/or the need to develop the new and complex accounting systems that would be required to satisfy FAS 133 standards. These considerations led management, with Howard's support and with the knowledge of senior managers in the Controller's Office, to adopt an approach to hedge accounting that deviated from the standard's clear requirements in numerous and important respects. For example, management adopted the so-called "shortcut" method of hedge accounting for many of its hedge transactions, even when the derivatives in those transactions did not have a fair value equal to zero and the terms of the derivatives and the hedged instrument were not "exactly the same," as FAS 133 requires. The Company also disregarded amendments to FAS 133 that the FASB adopted over a year before the standard took effect that foreclosed management's approach to the accounting for transactions the Company referred to as "term-outs," which were an important element of 8
the Company's hedge strategies. The Company's accounting policy regarding anticipated debt issuances also violated FAS 133 requirements by not specifying a single, proper methodology to assess a hedge's effectiveness, and by treating those transactions as perfectly effective based on a "duration matching" methodology that was inconsistent with FAS 133 requirements. Finally, the Company's hedge documentation was insufficient and in most cases incorrect; for example, the Company's documentation posited that the critical terms of the hedged instrument and the derivative were "identical," which was not the case. It appears that senior accountants in the Controller's Office were of the view that any deviations from a "strict application" of FAS 133 were immaterial. However, management did not conduct a systematic or comprehensive test to support that proposition, and the tests that it did conduct provided inadequate support for that view. The record also shows that management took steps throughout the FAS 133 implementation process to keep the Company's outside auditor informed of its decisions. Management engaged the auditor to review the Company's new hedge accounting policies (the "Derivatives Accounting Guidelines") prior to the effective date of FAS 133, to ensure that the principal features of the Company's implementation program complied with GAAP. The audit workpapers reveal that the auditor knew of, and accepted, Fannie Mae's major accounting policies concerning FAS 133 on the grounds that any deviations from GAAP reflected in Fannie Mae's policies were immaterial. In April 2000, moreover, the auditor described to the Board's Audit Committee its planned involvement in the FAS 133 implementation effort and prior to OFHEO's Special Examination, the auditor did not raise any concerns to the Audit Committee or the full Board regarding the Company's approach to hedge accounting. In addition, the Company's Derivatives Accounting Guidelines were available to, and were reviewed by, OFHEO examination staff. As late as June 2002, when OFHEO issued its report on Fannie Mae's operations in 2001, OFHEO reported that the Company's implementation of FAS 133 had a sound basis. Howard set the tone for the FAS 133 implementation effort and, from the outset and throughout the process, he focused the implementation team's efforts on avoiding the volatility associated with FAS 133 while not changing the Company's business practices to any significant degree. However, we did not find any evidence that Howard directed anyone to violate GAAP. Raines's involvement in the implementation effort was minimal. While he was familiar with the Company's goal of avoiding income statement volatility and the complex systems development effort associated with complex hedge accounting under FAS 133, we saw no indication that he knew that the Company's application of FAS 133 contained substantial departures from GAAP. Finally, the Board received assurances from management on several occasions (as well as from the Company's auditor and OFHEO) that the Company's implementation of FAS 133 was appropriate. Prior to the OFHEO Special Examination, 9
the Board did not have any indication that the Company's application of FAS 133 contained substantial departures from GAAP. The SEC's Office of the Chief Accountant announced in December 2004 that the Company's historical application of FAS 133 did not comply with GAAP, and that the Company was disqualified from applying hedge accounting from FAS 133's effective date. The Company is restating its financial statements with respect to its hedge accounting. C. Conclusions About Other Accounting Issues In addition to management's application of FAS 91 and FAS 133, we reviewed management's application of numerous other accounting issues, most of which were identified by OFHEO in the February 11, 2005 Letter, and by the Company in a November 2005 SEC Form 12b-25 filing. We summarize below our findings with respect to those issues. 1. Accounting for the Allowance for Loan Losses We reviewed the Company's accounting for the allowance for losses on loans in its mortgage portfolio and the liability for losses associated with its guaranty of mortgage-backed securities (collectively referred to as the "Allowance"). From 1997 through 2003, the Allowance was essentially unchanged at roughly $800 million despite improved credit quality and improved credit administration. For example, credit losses as a percentage of the average book of business declined from 0.027% in 1998 to 0.006% in 2003, which caused the number of years of losses covered by the reserve to increase from 3.3 years of losses in 1998 to 7.2 years of losses in 2003. The methodology the Company used for setting the Allowance before 2004 (roughly from 1997 through 2003) did not comply with GAAP because it was not based upon a detailed and documented assessment of the loss exposure inherent in the portfolio as required by GAAP. Management, along with the Company's auditor, recognized its departure from GAAP as early as 1998, but did not make any changes to the methodology or the accounting until 2002. Management's methodology for setting the Allowance also did not incorporate its improved credit performance, which should have been a factor in the analysis for setting the level of the Allowance. For example, the Company's forecasted loan losses over the period were consistently in excess of actual loan losses incurred, yet the Allowance was never adjusted to reflect the actual results. We did not find any evidence that management actually used the Allowance to manipulate earnings or to offset unrelated one-time expenses in a given period. We did find, however, that certain members of management - particularly, Spencer - viewed the Allowance as a "war chest" that could be drawn down to offset unrelated one-time events. While, there is no evidence that Spencer used the Allowance in this way, the evidence, at a minimum, reflected her awareness that the Allowance was overstated. In addition, the overstated Allowance made it easier for management to meet year over year earnings targets in subsequent years. Had the excess reserve been 10
reversed when management first became aware that the Allowance was overstated, this "non-recurring" income would have made the subsequent year's earnings growth goals that much more difficult to achieve. 2. Accounting for Dollar Rolls A typical "dollar roll" transaction at Fannie Mae involved a transaction in which the Company borrowed funds from a counterparty for a specified period of time, using a security from the Company's portfolio as collateral. To effect a dollar roll transaction, Fannie Mae would "sell" to the counterparty a security from its portfolio as collateral and simultaneously enter into an agreement to "purchase" a similar security at a future date. Assuming that the relevant accounting standards currently set forth in FAS 140 were satisfied, the Company was required to account for the arrangement as a financing (i.e., a short-term loan) rather than as a sale and a purchase. Failure to comply with the relevant accounting standards had two potential consequences: (1) Fannie Mae would have to account for the transfer of collateral as a sale, with consequent recognition of gain or loss; and (2) as the collateral for dollar rolls were MBS held in the Company's "held-to-maturity" portfolio, the treatment of the transfer of the collateral as a sale would have resulted in the "tainting" of the portfolio (that is, the Company's held-to-maturity securities portfolio would be reclassified as available-for-sale, with significant accounting consequences). Fannie Mae's accounting for dollar roll transactions did not comply with GAAP for a significant portion of the time period covered by this Report. Although FAS 140 became effective in 2000, and the accounting requirements for treating dollar rolls as financings were set forth in previous authoritative literature, the Company did not have an accounting policy that addressed all of the relevant requirements until 2003. In addition, coordination among the offices responsible for dollar roll transactions - particularly between Financial Standards in the Controller's Office, the Securities Trading Operations group in the Treasurer's Office, and Portfolio - was weak. Consequently, there were significant gaps in the Company's processes for addressing the accounting requirements for dollar rolls. The processes failed to address the FAS 140 requirement that the collateral returned to Fannie Mae be "substantially the same" as the securities that Fannie Mae "rolled out." There also was no evidence that, prior to about 2002, and possibly thereafter, the Company satisfied the FAS 140 requirement that the value of the collateral be adequate to reacquire the security. Accordingly, we conclude that management lacked a basis for reaching the conclusion that any given dollar roll transaction properly should be accounted for as a financing. Although we noted significant gaps in the Company's accounting for dollar rolls as financings, we conclude that the failure to follow GAAP in this instance was not intentional or motivated by an effort to achieve forecasted earnings. Rather, the failure stemmed from a lack of rigor in the Company's accounting. We understand that, as part of its restatement effort, the Company is reviewing its dollar roll transactions to determine whether individual transactions did, in fact, comply with the accounting 11
standards, and which transactions should properly have been accounted for as sales and purchases. 3. Accounting for Forward Commitments FAS 149, which had an effective date of July 1, 2003, amended FAS 133 to clarify that firm commitments to purchase mortgage loans or purchase and sell certain MBS should be treated as derivatives. Accordingly, FAS 149 required that these firm commitments (like other derivatives covered by FAS 133) be recorded on the Company's balance sheet at fair value and subsequently marked to fair value at the end of each reporting period until the settlement date. Changes in the fair value of the commitments would be reflected in the Company's earnings unless the derivative qualified as part of a hedging relationship. Fannie Mae designated many of its firm commitments as hedges of the risk resulting from changes in the price of the mortgage loans or MBS the Company would acquire or deliver when the commitment settled. Under FAS 149, hedge accounting would have been appropriate only if the provisions of FAS 133 - and specifically the provisions regarding hedges of forecasted or anticipated transactions - were met. Consequently, as FAS 133 specifies in these circumstances, the Company was required to document the hedged transaction with sufficient specificity so as to identify when that transaction occurred. We reviewed the history of management's implementation of FAS 149 and the policies and procedures that were adopted with regard to hedged transactions involving firm commitments. We found that the effort to implement the standard stretched the Company's resources in both of the departments that were most immediately affected by the new standard: Financial Standards and Portfolio. The resources were strained by a lack of systems and staffing to the point that it became difficult, if not impossible, for the Company to implement the standard correctly and in a timely fashion. The Company did not adopt a final accounting policy regarding FAS 149 until October 2003, nearly four months after the standard's effective date. Likewise, the Company's procedures to address several of the important issues raised by FAS 149 were not complete until months after the standard took effect. The Company revised its hedge documentation several times after the standard's effective date, and, as late as mid-2004, the specifications for the systems necessary to account properly for the hedge transactions were still in the discussion stage. The policies and procedures the Company adopted to implement FAS 149 did not comply with GAAP. For example, the Company's hedge documentation did not describe a hedged forecasted transaction with sufficient specificity such that one could identify whether a transaction that occurred was the hedged transaction. These departures from GAAP resulted from three related factors: (1) the lack of advance preparation for the changes that FAS 149 required; (2) the incorrect assumption at the outset of the implementation that, with only minor exceptions, all commitments would be eligible for hedge accounting; and (3) the unexpected complexity 12
involved in the application of FAS 149 to the wide variety of Fannie Mae's commitments and forward trade transactions. We did not find that the failures in this area resulted from an effort to manipulate the Company's financial results. Weakness in the Company's implementation of FAS 149 became apparent the first time the Company closed its books after FAS 149's effective date, which resulted in a $1 billion error on the Company's balance sheet. Although the error was immediately brought to the Board's attention, rather than explaining the problems associated with the implementation effort, Spencer informed the Board that the implementation process was well in hand. This omission was especially significant as the balance sheet error triggered an examination by OFHEO of Fannie Mae's FAS 149 implementation process and its "end-user" accounting systems. The Board thus lacked relevant information relating to an issue that the Company's principal regulator deemed particularly significant. 4. Classification of Securities Held in Portfolio Our investigation included an assessment of the Company's interpretation and application of FAS 115, which specifies the accounting for a security depending on its classification as either: (1) held to maturity ("HTM"), (2) available-for-sale ("AFS"), or (3) trading. Once a security is classified as HTM, the security may be reclassified only in narrow, specified circumstances. FAS 115 states: "At acquisition, an enterprise shall classify debt and equity securities into one of three categories: held-to-maturity, available-for-sale, or trading." Management's accounting policy did not require the classification of a security to be determined on the date of acquisition, as required by FAS 115. Instead, management interpreted the phrase "at acquisition" to mean "at the end of the month of acquisition." We found no support or justification for such an interpretation. In practice, the Company classified securities "at acquisition," but that classification was subject to change. When a Company trader executed a trade, he or she either would select a classification or the system would classify the security as HTM by default. Near the end of the month, management determined whether HTM securities should be reclassified to AFS. That practice violated GAAP. The Company's approach violated an unambiguous accounting rule regarding the classification of securities as HTM, AFS, or trading. Although the Company's procedure for determining a security's final classification involved the consideration of factors such as "balance sheet effects" and "economic opportunities," we saw no evidence that management intentionally used this mechanism to manipulate its net income. Moreover, at least as of 2003, the Company's auditor was aware of management's practices in this area and did not raise an objection. We also found no evidence that management discussed this issue with the Board prior to the OFHEO Special Examination. 13
5. Recognition of Interest Expense and Income Until early in 2003, the Company's liquid investment portfolio ("LIP") and debt accounting systems (known as ORION and STAR, respectively) calculated interest expense and income on certain investments and debt instruments as if there were 30.4 days in each month, even if the instrument's terms required interest payments on an "actual/365" or "actual/360" basis. As a result of this practice, management avoided the fluctuations in interest income and expense that would result from the fact that the twelve months of the year and the four calendar quarters do not have the same number of days. Management also periodically accrued additional interest expense through "on-top" entries, and these entries were then amortized over the remainder of the year. Management should have accounted for these investments and borrowings by recognizing interest income and expense in accordance with the legal terms of those arrangements, regardless of the fact that such treatment would generate fluctuations in the recognition of income and expense from month to month and from quarter to quarter. Management discontinued these practices in the second quarter of 2003, at which time it began to recognize interest income and expense in accordance with the actual terms of the instruments. Spencer and other officers in the Controller's Office knew or should have known that the Company's practices did not comply with GAAP. The audit workpapers indicated that the Company's outside auditor was aware of this practice at least as of July 2003. We saw no indication that this issue was ever brought to the attention of the Board. 6. Accounting for Other-Than-Temporary Impairment of Manufactured Housing Bonds and Aircraft Asset-Backed Securities In April 2004, OFHEO raised concerns about the Company's accounting for other-than-temporary impairment ("OTTI") of investments in manufactured housing bonds ("MH bonds") and aircraft asset-backed securities ("Aircraft ABS"). We found that the Company did not have a formal process for monitoring investments for OTTI until mid-2003, when it formed an Impairment Committee, and also did not evaluate all of its HTM or AFS investments for OTTI, as required by GAAP, prior or subsequent to formation of the Impairment Committee. Management's failure to monitor all HTM and AFS investments for OTTI represents a control weakness, and suggests the possibility that the Company underreported OTTI on investments that it did not monitor. In addition, prior to 2004, Fannie Mae relied primarily on internally developed discounted cash flow ("DCF") models to measure impairment on MH bonds and Aircraft ABS and, thus, to determine the OTTI amounts it recognized, even though bid/ask dealer pricing was available. While we did not find that management chose to rely on DCF modeling in order to achieve particular earnings goals, we note that the DCF model included assumptions that were subject to management discretion and data errors that impacted both the timing and the amount of OTTI the Company recorded. 14
After OFHEO raised its concerns regarding impairment on MH bonds and Aircraft ABS in 2004, Fannie Mae discussed its policies for measuring and recognizing OTTI with the SEC, and ultimately worked with OFHEO to implement a new policy in April 2004. 7. Accounting for Investments in Interest-Only Mortgage-Backed Securities ("IO MBS") Beginning in 1995, management combined its IO MBS investments with other securities (specifically, MBS and REMIC securities) for accounting purposes, and treated the IO MBS as an increase in the premium or reduction in the discount on the other security.(8) Management initially consulted the Company's outside auditor for this accounting treatment for the IO MBS investments, and the auditor did not object to the Company's approach. We believe that the Company's account for IO MBS investments violated GAAP. EITF 90-2, which addressed an analogous situation and should have been applied by management to the accounting for its IO MBS investments, required that an exchange transaction take place before the accounting for the individual interest-only and principal-only securities can change. Furthermore, management's primary motive for engaging in this accounting treatment was to avoid recognizing impairment charges on the IO MBS. Management did not fully disclose its motivation or all of the material facts relating to its IO MBS accounting to the outside auditor. While management did consult with the auditor for its accounting treatment of IO MBS investments in 1995, by 1998, management intentionally withheld from the auditor its impairment analysis of the IO MBS. Management did so apparently fearing that the new audit team might disagree with the old audit team and require management to change its accounting for IO MBS, which could have resulted in the Company being required to recognize impairment losses. Management failed to inform the Board of the issues relating to its accounting for the IO MBS investments until OFHEO raised questions about these practices in April 2004. In particular, Freddie Mac's restatement raised nearly identical issues, but management, in its presentation to the Board about the Freddie Mac restatement, failed to disclose the existence of its own problematic "synthetic" IO MBS combinations. 8. Securitization of Wholly-Owned MBS In the normal course of its business, Fannie Mae issues guarantees to holders of securities backed by pools of mortgage loans. In a majority of these transactions, lenders transfer pools of mortgage loans meeting certain criteria to Fannie - ----------- (8) These IO MBS are also referred to as "synthetic REMICS" in the PW Report. 15
Mae, which transfers those loans to trusts that Fannie Mae establishes, and for which it serves as trustee. The lenders usually receive a certificate (i.e., MBS) evidencing the right to receive cash flows from the underlying loans (less a guaranty fee). Fannie Mae acquires interests in the MBS it guarantees as an investment, and at times has acquired 100 percent of the MBS from a particular trust. With the adoption of FIN 46 in 2003, the Company was required to determine whether it needed to consolidate any of those trusts onto its balance sheet. However, under FIN 46 a party (other than the transferor) with a variable interest in a Qualifying Special Purpose Entity ("QSPE") is not required to consolidate that entity, as long as it does not have the unilateral right to dissolve the trust or change the entity so it no longer meets the definition of a QSPE. An entity is a QSPE if the transferor does not have the unilateral right to dissolve the trust, and either (1) third parties hold more than ten percent of the beneficial interests in the entity, or (2) the transaction is a guaranteed mortgage securitization ("GMS"). Management treats the securitization of pools of loans as GMSs and the trusts as QSPEs. Under FIN 46, management was required to evaluate its trust portfolio to determine whether it should consolidate those trusts in which it owned 100 percent of the beneficial interests because it had the unilateral ability to dissolve the trust. To avoid the need to evaluate thousands of trusts and the requirement to consolidate those trusts (which would have required the Company to recognize the loans held in the trust rather than the MBS on its balance sheet), management developed a structure in which it would transfer wholly-owned MBS to a new trust, called a Mega, and sell one percent of the beneficial interest in each Mega to a third party. According to management's initial analysis of the relevant accounting standards in 2003, this approach would allow it to avoid consolidation of the trusts because the Company would no longer have the unilateral ability to dissolve them. During a discussion between an accountant in Financial Standards and members of the FASB staff in 2004 regarding another transaction, the accountant raised the issue of whether structures like Megas - which are securitizations of securities, not securitizations of loans - qualified as a GMS. The FASB staff did not disagree with her conclusion that the answer was no. Accordingly, management reevaluated its accounting for Megas; essentially, management concluded that it should consolidate the Megas and account for the transfer of the one percent interest as a secured financing. Management's initial accounting policy in this area was incorrect, but we conclude that this was the result of an inadvertent misinterpretation of the applicable accounting literature.(9) We have not found any evidence suggesting it was motivated by a - ------------ (9) As we discuss in our Report, management also took an alternative approach to certain pools in which it acquired a 100 percent interest in the MBS. Rather than consolidate the loans or include the trust in a Mega, management reclassified the MBS from AFS to HTM. This approach also was inconsistent with relevant accounting standards. The policy that supported this approach was reversed in 2004 as well, and our conclusions regarding the bases for the error and management's intent apply to these circumstances as well. 16
desire to manipulate the Company's financial statements. In fact, the Company reversed course after Financial Standards learned of its mistake and corrected the Company's accounting policy. Accountants in Securities Accounting then assessed the impact of the error on the financial statements and determined it to be immaterial. We understand that management is reviewing these transactions, including the Company's approach to the consolidation of trusts under its new policy, as part of its restatement effort. 9. Accounting for Income Tax Reserves and Certain Tax-Advantaged Transactions We reviewed the Company's process for establishing reserves related to tax credits the Company received as a result of its investments in synfuel partnerships and in connection with certain tax-advantaged transactions known at Fannie Mae as Short Term Interest Securities ("STIS"). We also considered management's reporting of the tax benefits from its synfuels investments and STIS transactions on its financial statements. The process that management used to determine its tax reserves appears reasonable. However, we are not able to form any conclusions as to whether specific tax reserve levels were appropriate and represented known tax liabilities because the Company did not maintain documentation adequate to explain the rationale for its decisions with respect to the establishment and amount of individual tax reserves. Several documents, however, indicate that Spencer and others in Financial Reporting, in some instances, may have recorded amounts to the Company's tax reserve that were not connected to known tax liabilities, but instead were booked for inappropriate earnings management purposes. With respect to the synfuels partnerships, management established a reserve percentage for the purpose of calculating the Company's tax reserve. However, it appears the Company also recorded an additional unsupported "excess" amount in the reserve at year-end 2002 that it did not release to earnings until the third quarter of 2003. The remainder of the reserve was released in the fourth quarter of 2003. Interviewees were unable to explain why the amounts were released over two quarters and we have seen no documents that offer a reason. Management opted to obtain only a draft "should" level opinion from outside counsel for its STIS transactions, even though it expected the IRS to examine the transactions. Apparently, the Company sought to avoid the additional expense associated with issuance of a final opinion, and it believed that the draft opinion provided adequate support for the Company's position. Although certain documents from the Company's files may be read as questioning whether the STIS transactions had an adequate business 17
justification, interviewees stated their belief that the STIS transactions had genuine economic benefits distinct from the tax benefits, and, based on evidence that the transactions were expected to (and did in fact) generate a profit, we have no reason to dispute that assessment. 10. Accounting for Insurance Products Fannie Mae purchases mortgage insurance in order to mitigate its exposure to credit losses on loans, to comply with the Charter Act (i.e., the Company is required to have credit enhancement for loans with a loan to value ("LTV") ratio equal to or greater than eighty percent), and as a broader risk mitigation strategy. Beginning in 2001, however, management, under Raines's direction, began considering finite risk insurance products as a method for accomplishing earnings-related goals in addition to mitigating the Company's exposure to losses. Several of these contemplated transactions were motivated either by a desire to shift income between periods (in particular from 2001 and 2002, into 2003 and 2004), or to offset the impact of other actions that were expected to result in a sharp increase in earnings. In January 2002, management executed a policy with Radian that absorbed a portion of a deductible on an existing insurance policy covering certain high-risk loans, and which had a large premium in 2002 with predictable returns in future periods (the "Radian Transaction"). In November 2005, the Company announced that the Radian Transaction had not been accounted for in accordance with GAAP and that it had to be restated because the policy "did not transfer sufficient underlying risk of economic loss to the insurer" to qualify for the insurance accounting treatment it was given. We agree with this assessment, and also conclude that the Radian Transaction was entered into for the primary purpose of "shifting" income out of 2002 into 2003 and 2004, to demonstrate stable earnings growth. 11. Accounting for Out-of-Portfolio Securitization ("Portfolio Pooling System") In its February 11, 2005 letter to Stephen B. Ashley, OFHEO reported that an error in Fannie Mae's Portfolio Pooling System ("PPS") had led to the misclassification of loans that the Company held in its portfolio. We determined that the error appeared in the interface between PPS, which the Company uses to securitize the loans that it acquires, and LASER, the Company's system of record for loans that it holds in its portfolio. The error resulted in loans destined for securitization at a future date being erroneously classified as held-for-investment ("HFI") rather than held-for-sale ("HFS"). Our inquiry focused on why the error had gone undetected since the program was implemented in the 1980s. We determined that the error in classification would have been relevant to the Controller's Office, as the accounting for loans in the Company's portfolio differs depending on whether they are classified as HFI or HFS. We concluded that the Controller's Office received information regarding the classification of the loans from the PPS system before the system error resulted in an 18
erroneous classification. Moreover, under the Company's accounting policy at the time, loans that were designated within a given month for securitization were accounted for as securities rather than as loans. Because most of the loans that flowed through the PPS system were accounted for as securities under this policy, the number of loans that the Company accounted for as HFS was relatively small. Thus, any discrepancies that resulted from the classification error would have been difficult to detect. 12. The Debt Repurchase ("Buyback") Program We reviewed the Company's debt buybacks for the period from 2000 to 2004, including the motive, accounting, and disclosures for the buybacks. We conclude that buyback transactions were accounted for and reported in accordance with GAAP. Management disclosed the extent of the debt buybacks and the resulting losses in Fannie Mae's public disclosures. We do not dispute that management had legitimate business purposes for executing debt buybacks during the period, including a desire to manage interest-rate risk. However, we conclude that management's execution of buybacks suffered from several deficiencies. First, management's motivation for executing buybacks was primarily earnings driven. Management used debt buybacks to depress income in 2001 through 2003, in order to show stable earnings growth; and management also focused primarily on the present period EPS impact of the buybacks in determining the size of the buybacks. Management never discussed its motivation with the Board, including at a meeting of the Assets & Liabilities Policy Committee of the Board in 2004, where management presented an after-the-fact view of the buybacks conducted in prior years. As a result, the Board was not able to assess the impact of the buyback transactions for awarding bonuses, which were tied to achieving stable EPS growth. Second, buybacks were executed with little or no formal contemporaneous documentation of the economic benefit to the Company, and no clear policy or procedures for the approvals required for the transactions. While Huron's analysis did not identify any clearly non-economic buyback transactions, the absence of any documentation supporting buyback decisions or procedures represents a control weakness. 13. Accounting for the Amortization of Callable Debt Expenses Fannie Mae issued both callable and noncallable debt to finance its activities. When Fannie Mae issued debt, it incurred various expenses such as commissions, legal fees, and similar costs. In addition, any difference between the face amount of the debt and the proceeds from issuing the debt gave rise to a premium or discount on the debt issuance. Relevant accounting literature requires that callable debt expense be amortized over the life of the debt, regardless of a possible call of the debt prior to 19
maturity. Any unamortized expense must be recorded in the accounting period in which the call occurs and the debt is extinguished. Management established a policy, however, of amortizing callable debt expense over the estimated life of the debt - that is, the period between issuance of the debt and the expected call date. In addition, the Controller's Office implemented "amortization end date changes" to reflect new expected call dates, resulting in a change in the amount of expense recorded in future periods. Neither the initial amortization of the expense over the estimated life of the debt, nor the implementation of amortization end date changes, was consistent with GAAP. Moreover, the Company did not apply its approach to the accounting for callable debt expense - and particularly the amortization end date changes - in a consistent fashion. On at least one occasion, for the third quarter of 2002, Financial Reporting made a late on-top entry that was inconsistent with its past practices. The purpose of entry was to offset an unrelated entry by recognizing additional interest income, and thereby bringing net interest income back in line with the Company's expectations. The Company's accounting policy regarding the amortization of callable debt expense appears to reflect Financial Standards' long-standing misinterpretation of the applicable accounting rules, rather than a deliberate disregard of them. However, the evidence concerning the periodic adjustments, and particularly the adjustment in the third quarter of 2002, leads to the conclusion that the Company used these adjustments to meet earnings expectations. Spencer and other Financial Reporting personnel played a key role in recording that adjustment. The Company's outside auditor was aware of the Company's accounting policy regarding callable debt expense, and of the on-top adjustment in the third quarter of 2002, but may not have been fully informed of the nature of, or reasons for, this adjustment. 14. Minority Lending Initiative Fannie Mae implemented a Minority Lending Initiative ("MLI") program in 2002 to increase the Company's financial support for mortgages to African-American homeowners. The initiative was considered an important component of the Company's overall mission and was viewed by some as a means of securing loans that would meet the guidelines set by the Department of Housing and Urban Development ("HUD"). A Company employee raised concerns regarding the initiative in response to an e-mail broadcast to all employees by the Chair of the SRC. The employee was concerned that the Company appeared to be paying an excessive price for loans that were underperforming and that the MLI program might have been devised to meet corporate targets. We saw nothing to indicate that the MLI program had an improper purpose. We did identify one issue concerning the accounting for payments in 2003 that the Company made to Resource Bancshares Mortgage Group, Inc. ("RBMG"), the 20
mortgage lender that originated a majority of the loans acquired under the MLI program. Because direct acquisition of the loans from RBMG may have violated Fannie Mae's Charter, the Company arranged for RBMG to sell the loans to a third party, Self Help. Self Help then sold the loans to Fannie Mae on terms that management concluded were Charter-compliant. The payments at issue were intended to compensate RBMG for the difference between the price the Company had committed to pay RBMG, and the price paid to RBMG by Self Help. The Company capitalized these payments as part of the cost of the acquired loans when they should have been expensed. The aggregate amount of the payments we have been able to identify was approximately $35.5 million. 15. Accounting for Realignments and the Security Master Project We reviewed management's accounting for differences generated in the process of identifying and correcting errors and mismatches between its amortization database and loan and securities databases. The process of adjusting the amortization database to match the loan and securities databases was known as "realignments." Realignments were essentially corrections of errors, and as such, management should have analyzed and accounted for their impact in accordance with APB 20. Management failed to do so. We found that management did not account for realignment impacts properly under GAAP. For the most part, management deferred the recognition of these differences by recording them to balance sheet accounts and amortizing them over multiple years. On other occasions, in addition to deferring the recognition of these differences and amortizing them over time, management included the cumulative deferred realignment amounts and estimates of future realignments in its calculation of catch-up; and on still other occasions, management recognized the realignment impacts into income in the period they were identified. No one we interviewed could explain why management failed to apply APB 20 to realignment impacts or the basis for the inconsistent accounting treatment of such impacts. At a minimum, this demonstrates that the Company did not have adequate accounting policies or procedures to ensure that its personnel complied with GAAP in this area. Furthermore, the decision to capitalize and defer realignment impacts over time smoothed out the errors' impact on income in any one period. With respect to the inclusion of realignments and estimates of realignments in the catch-up calculation in 2003, we conclude that management was motivated, in part, to avoid recording or to reduce the amount of the catch-up adjustment required under the Company's amortization policy. 16. Accounting for Investments in Affordable Housing Partnerships Our inquiry regarding affordable housing partnerships focused on three issues: (1) the Company's accounting for its capital contributions to the partnerships; (2) the methodology used to account for low income housing tax credits ("LIHTC") and net operating losses associated with the partnerships; and (3) the Company's policy and 21
practice regarding the accounting for possible impairment of these investments. In each of these areas, we conclude that the Company's accounting policy and its financial reporting was inconsistent with GAAP. Fannie Mae's accounting for investments in affordable housing partnerships violated GAAP in several respects. This was particularly true in the first half of the 1990s when the Company used an inappropriate accounting methodology to calculate its portion of the net operating losses in the partnerships. In addition, the accounting for the partnerships' net operating losses was incorrect when the Company had obligations with respect to future capital contributions. Management also did not have a formal policy regarding the assessment of impairment in its partnership investments until 2000. The policy the Company developed at that time required that it recognize impairment of each investment only in the tenth year; prior to the tenth year, management did not assess individual investments for impairment as the accounting literature requires. Excluding the events surrounding the accounting for net operating losses and tax credits in 1998 (discussed in Part A.1. above), we have not seen any evidence that the Company's accounting or reporting regarding affordable housing partnerships was done with the intent to affect earnings in any period. Rather, the problems associated with the Company's accounting in this area appear to stem from misinterpretations of relevant accounting standards, and a lack of resources, particularly in the systems area, prior to the late 1990s. The Company is reviewing the accounting for these partnerships during the past several years as part of its restatement effort. III. CORPORATE GOVERNANCE AND INTERNAL CONTROLS Our conclusions with respect to the Company's corporate governance and structure prior to 2005 are organized into the following areas: the Board of Directors; the Office of the Chairman and other key elements of senior management; the Company's ethics and compliance functions; Internal Audit; and the Office of the Controller. In addition to our findings and conclusions, we also describe the substantial changes that have taken place since September 2004. A. Board of Directors With respect to the conduct of the Board prior to September 2004, we conclude that the Board endeavored to operate in a manner consistent with its fiduciary obligations and evolving corporate governance standards. The Board was open to examination by third parties and responsive to outside commentary, and it generally received high marks from outside observers. The Board sought, received, and relied on support and assurances from Company management, internal and external auditors, and regulators. Management shared its accounting policies and practices with its outside auditors and with OFHEO during the relevant period. As a result, both were generally aware of many of the accounting and financial reporting matters and related judgments 22
discussed in this Report. Prior to the release of the OFHEO Report, however, the Board was not notified of any substantial concerns, and received assurances from internal and external sources that the Company was complying with applicable rules and regulations, and with best practices in the industry. The Board, and in particular the Audit Committee, was sensitive to matters relating to accounting and financial reporting. The Audit Committee requested and received briefings regarding the Company's critical accounting policies, and was regularly assured that Fannie Mae was acting in accordance with relevant standards. For example, the Board reacted quickly to the release of Freddie Mac's announcement in 2003 about its accounting issues. Fannie Mae's Board requested reports from management and the Company's outside auditors on whether Fannie Mae might have accounting problems similar to the ones discovered at Freddie Mac. In response, management provided the Audit Committee with a misleading report that identified only minor and immaterial issues at Fannie Mae. The Board also responded appropriately when it received indications that there were significant issues at the Company. The Board has made considerable effort to examine and improve its structure, composition, policies, and practices. The separation of the Chairman and CEO positions, the creation of the Risk Policy and Capital Committee to oversee financial and operational risk management, and the transformation of the Compliance Committee into a permanent committee with broad oversight of compliance matters, are all positive developments. B. Office of the Chairman Through the end of 2004, management did not fully inform the Board of the Company's accounting issues, internal control deficiencies, or the inadequacies of its internal systems. Further, although management paid lip service to a culture of openness, intellectual honesty, and transparency, the actual corporate culture suffered from an attitude of arrogance (both internally and externally) and an absence of cross-enterprise teamwork (with a "siloing" of information), and discouraged dissenting views, criticism, and bad news. Finally, the Company lacked appropriate structure and personnel for adequate risk management across risk areas (with an extremely broad collection of functions and authorities residing in the CFO), and lacked a genuine cross-enterprise approach to operational risk management. Since the end of 2004, the new management team led by CEO Daniel H. Mudd, with the active engagement of the Board, has made a concerted effort to reform the management structure and the "tone at the top." These changes include: (1) redefining management committees and lines of reporting with a view to improving internal controls, management of risks, and horizontal and vertical information flow; (2) adopting a management style that seeks to be more open, collaborative, and humble; (3) establishing a Chief Risk Officer position (with an independent Risk organization); (4) revamping the CFO position with a set of responsibilities more appropriate for the position; (5) eliminating the Law and Policy group, with the movement of core compliance functions to a new, independent Office of Compliance, Ethics & 23
Investigations; (6) integrating the Company's businesses (including the Mortgage Portfolio business, which historically reported to the CFO) under a new Chief Business Officer position; and (7) shifting the Company's external relations toward a more cooperative relationship with OFHEO, Congress, and customers (with a substantial reduction in the size and aggressiveness of Fannie Mae's lobbying and grass roots activities). In sum, as of the date of this Report, the new senior management team is in the process of undertaking meaningful substantive and tonal changes. These changes have improved the functioning of the Company both internally and externally. C. Internal Audit Prior to release of the OFHEO Report in September 2004, the head of Internal Audit lacked the requisite expertise and experience to lead the internal audit operation at an organization as large and complex as Fannie Mae. Moreover, on more than one occasion, the head of Internal Audit took steps that suggested he did not fully appreciate his organization's role within the Company or his proper relationship with senior management. Internal Audit also did not possess a sufficient number of auditors with the requisite mix of technical accounting expertise and auditing experience to carry out its responsibilities related to Fannie Mae's increasingly complex business. Although Internal Audit's workload increased substantially in the years prior to 2005, Internal Audit requested only modest increases in headcount. In addition, the department's training programs were inadequate to compensate for these deficiencies. Internal Audit's communications with the Board and management were deficient and, at times, inaccurate. On a number of occasions, Internal Audit provided assurances to the Audit Committee that Internal Audit's staffing was adequate in terms of quantity and quality (when it had told management otherwise) and that it had audited Fannie Mae's accounting for compliance with GAAP (when it actually audited only for compliance with Fannie Mae policies interpreting GAAP). In addition, Internal Audit's reporting of its audit issues to the Audit Committee (and to members of senior management) lacked clarity and did not succinctly prioritize the findings or the subsequent remediation. The Audit Committee and senior management have acted to address many of these deficiencies. They have taken steps to replace Internal Audit's leadership, restructure its organization, focus its responsibilities on its core audit mission, and reform its processes and procedures. Substantial progress is underway in each of these areas. D. Ethics and Compliance Functions For more than a decade, Fannie Mae has maintained a Code of Business Conduct, provided Code-related training to employees, and investigated violations of the Code and other corporate policies. The Company also has a longstanding and 24
experienced investigative unit to handle employee complaints. Moreover, at the beginning of 2003, Fannie Mae acted to enhance its ethics and compliance program, by (1) pulling together ethics and compliance functions within the Legal Department; (2) creating the Office of Corporate Compliance ("OCC") to develop and monitor business unit compliance plans, administer employee training, and otherwise provide central management of ethics and compliance matters; (3) appointing a Chief Compliance Officer to oversee the existing investigative unit (the Office of Corporate Justice ("OCJ")) and the OCC; and (4) replacing the old Business Conduct Committee (which had been chaired by the head of Human Resources) with a new management-level compliance committee chaired by the General Counsel. Although these accomplishments are worthy of note, and the ethics and compliance functions contained many well-meaning and dedicated professionals, the Company's ethics and compliance program as of late 2004 continued to suffer from the following deficiencies: - Management devoted too few resources to Fannie Mae's ethics and compliance functions (and especially the OCC). - Management undermined the perceived independence and impartiality of the Company's ethics and compliance functions by housing them within a litigation section of the Legal Department, headed by a Chief Compliance Officer who also served as the head of the employment practices litigation group responsible for defending the Company against employee complaints. - Management failed to invest appropriate responsibilities and stature in its Chief Compliance Officer, who did not hold a dedicated position; did not report to the Board of Directors; and had no discernable compliance responsibilities other than to supervise the activities of the OCC and the OCJ. - Without an active management-level oversight committee, and with an under-resourced and relatively low-stature OCC, the Company lacked an effective mechanism for coordinating compliance matters across the enterprise. Since September 2004, Fannie Mae has taken important steps to rectify deficiencies in its ethics and compliance functions. Most notably, it has created a new Office of Compliance, Ethics & Investigations ("OCEI"), which (1) is independent of the Legal Department, (2) reports directly to the CEO and the Compliance Committee, (3) is led by a new Chief Compliance Officer who is committed full-time to ethics and compliance functions, and (4) will not only absorb the functions and resources of the OCC and the OCJ, but will also have a dedicated ethics unit. Moreover, management now provides the Board with detailed written reports on ethics and compliance programs and activities. 25
E. The Controller's Office Prior to September 2004, the Controller's Office suffered from significant resource deficiencies. The headcount of the Controller's Office increased only modestly in the years prior to 2005, even as that office experienced dramatic increases in workload stemming from the introduction of new and complex accounting standards, the Company's decision to become an SEC registrant, and the growth of Fannie Mae's business. The Controller's Office leadership lacked adequate staffing, sufficient accounting and financial reporting expertise, and experience for a financial services company as complex as Fannie Mae. In addition, the Controller's Office relied to a substantial degree on inadequate systems that required considerable manual effort, further straining the already overburdened staff. For example, the closing process was manually intensive and unduly susceptible to human error. The relevant computer systems were not integrated and, consequently, the process of preparing the Company's monthly financial information required significant manual processes, including numerous manual journal entries to the general ledger. In addition, prior to the middle of 2004, the Controller's Office lacked formal written procedures regarding journal entries and account reconciliations, did not have standardized documentation to support journal entries, and permitted employees to sign off on journal entries for other employees. Since the release of the OFHEO Report, Fannie Mae has made changes to the structure and personnel of the Controller's Office, and to the Company's approach to the development of accounting policy. The Controller's Office, with active support from senior management and considerable reliance on outside expertise, has made significant efforts to augment its resources and the procedures and systems used in the development and oversight of accounting policies and financial reporting. IV. EXECUTIVE COMPENSATION In the September 2004 Agreement, Fannie Mae agreed to report on the Company's "compensation regime and its relation to strategic plans and their impact on accounting and transaction decisions and any revisions to avoid inappropriate incentives." In accordance with this undertaking, the SRC initiated a two-part review: (1) a historical analysis of Fannie Mae's executive compensation structure and its relationship to efforts to meet financial goals (such as EPS targets); and (2) a prospective assessment of the Company's compensation structure and recommendations for revisions to that structure. The SRC asked Paul, Weiss to review Fannie Mae's compensation programs and to assess the role of EPS or other financial indicators as a compensation trigger. Paul, Weiss was not asked to review or analyze employment contract issues or any individual compensation issues.(10) - ----------- (10) The SRC engaged Semler Brossy Consulting Group ("Semler Brossy") to evaluate the Company's current compensation structure and to make recommendations on revisions. Semler Brossy presented its report and recommendations to OFHEO on February 24, 2005. 26
Historically, the Company's target compensation levels consistently lagged behind those of the Company's "comparator corporations." Therefore, to facilitate payment of market-competitive compensation for executives, Fannie Mae intentionally set its "maximum" EPS target at levels that the Company expected to achieve. Because the expected EPS number was not an aggressive goal, the Company regularly exceeded it and triggered maximum bonus, stock, and stock option awards. This resulted in executive compensation at (but not above) the target compensation level. Beginning in 2002, the Company attempted to correct this situation and to align EPS targets and target bonuses in accordance with Fannie Mae's written compensation philosophy (that is, executive compensation would have been consistent with the Company's philosophy if the Company met the "target" EPS, rather than the "maximum" EPS). However, due to unanticipated shifts in market compensation, even under its new program Fannie Mae's executive compensation continued to lag behind market levels, and Fannie Mae executives received total compensation at market levels only if the Company met maximum EPS bonus targets. Non-financial corporate performance goals played a part in executives' long-term executive compensation through the Company's PSP. These goals were set, and performance against them was assessed, by the Compensation Committee of Fannie Mae's Board of Directors based on a report prepared by management. We found that management consistently tendered excessively positive reports to the Compensation Committee. During the course of our review, OFHEO requested, and the SRC agreed, that we also review the role that the Legal Department played in compensation decisions. OFHEO's request stemmed from two anonymous letters that accused attorneys in Fannie Mae's Legal Department of excessive and inappropriate involvement in compensation decisions and, specifically, of improperly attempting to "cloak" compensation decisions with confidentiality under the guise of the attorney-client privilege. We found no evidence to support these allegations or that the Legal Department was inappropriately involved in executive compensation decisions. V. FANNIE MAE'S INVESTIGATION OF ROGER BARNES'S ALLEGATIONS In August 2003, Roger Barnes, then a manager in the Controller's Office, raised allegations of accounting impropriety at Fannie Mae, including potential noncompliance with FAS 91. Barnes also alleged that Controller's Office management was not receptive to employee concerns regarding Fannie Mae's accounting, and, shortly thereafter, he alleged that he had been discriminated against on the basis of race and gender. Approximately three months later, after Internal Audit and the Legal Department had conducted three investigations into Barnes's allegations, and Barnes had threatened to bring a lawsuit against Fannie Mae, Barnes and Fannie Mae executed a settlement 27
agreement. In the agreement, Barnes relinquished all legal claims against Fannie Mae in exchange for monetary consideration. The agreement also required Barnes to cooperate with investigations into matters relating to his allegations. Barnes subsequently submitted written testimony to Congress, and he participated in an interview by OFHEO. His testimony and interview raised additional accounting issues and included other allegations against Fannie Mae. In light of these events, the SRC asked Paul, Weiss to determine: (1) whether the Company's investigations into Barnes's accounting allegations were conducted appropriately, and (2) whether the Company entered into the settlement agreement with Barnes for an improper purpose, such as to prevent him from pressing his allegations of accounting impropriety. As the substance of Barnes's allegations concerned the accounting for premium/discount amortization under FAS 91, we also inquired into the substance of those allegations. We conclude that the Company's response to Barnes's allegations was flawed in several respects. The Controller's Office did not communicate appropriately with Barnes regarding either accounting or personnel matters. The remedial measures Fannie Mae directed the Controller's Office to undertake following the investigations into Barnes's allegations were not effective in improving the reporting environment within the Controller's Office. The Company's investigation into Barnes's allegations also suffered from conflicts-of-interest and inappropriate pressure to complete the investigations in an unreasonable time frame due to looming CEO/CFO certification deadlines. In addition, the Legal Department assigned to Internal Audit the task of assessing whether the accounting practices Barnes identified violated GAAP, but Internal Audit was not equipped to render such determinations. As for Barnes's underlying allegations of accounting problems, we conclude that some of his allegations had merit. For example, we address his claims regarding FAS 91 in a separate section and we also conclude that management's practice of editing certain conditional prepayment rates ("CPRs") was inappropriate because, among other things, management could not identify a consistent rationale for changing the CPRs, the Controller's Office made the changes without consulting the economists who developed the CPRs, and the changes were not applied consistently to all areas of Fannie Mae. As for the Company's decision to reach a settlement of threatened litigation with Barnes, we conclude that the decision was based on an appropriate analysis of the Company's litigation risk and was not motivated by a desire to conceal misconduct by Fannie Mae or its employees or officers or by a desire to silence Barnes. VI. MANAGEMENT'S CONDUCT DURING OFHEO'S SPECIAL EXAMINATION We reviewed management's conduct during the OFHEO Special Examination through the issuance of the OFHEO Report in September 2004. In particular, we focused on the adequacy of the Company's document production in 28
response to OFHEO's requests, and on the conduct of the Legal Department and its advisors during the examination, including the accuracy of the information they provided to the Board. With respect to the Company's response to OFHEO's document requests, we found no evidence that anyone at the Company, or its counsel, intended to obstruct or impede OFHEO's Special Examination, or that anyone directed others to destroy evidence or not to cooperate fully with OFHEO. We do find, however, that the Company's Legal Department did not initially undertake a sufficiently comprehensive search for documents in response to OFHEO requests. Many of the documents that were responsive to OFHEO requests did not turn up until 2005, when the Company's lawyers abandoned their approach of allowing employees to search their own files and adopted a new approach of having attorneys review all files in employees' offices for responsive documents. We also found that the Company's outside counsel, which was charged with the task of conducting responsiveness and privilege reviews of the documents collected by the Legal Department, construed OFHEO's requests very narrowly. While we believe that those decisions were made in good faith, it is clear to us that the Company would have been better served by a less restrictive approach by its lawyers to collect and produce documents in response to OFHEO's requests. First and foremost, a more expansive document collection approach would have provided the attorneys with documents that would have enabled them to have a more complete understanding of the facts and be in a better position to recognize the many problems with the Company's accounting practices. Second, a fuller document production may have staunched the increasingly hostile relationship between the Company and OFHEO during the Special Examination. We also reviewed the information and advice that the Board received during the Special Examination, including from the Company's outside counsel and its accounting expert. OFHEO added this issue to our review after it raised questions about whether the Company's lawyers shielded certain documents from OFHEO through an overly aggressive use of privilege during the course of the Special Examination, and whether any lawyers "lied to" or "misled" the Board in connection with the Special Examination. With respect to the Company's assertion of privilege during the Special Examination, while there were instances where documents that the Company had identified as privileged were later determined not to be privileged, we did not find any evidence that lawyers made aggressive privilege determinations in order to shield relevant information from OFHEO. We found that lawyers - both in-house and outside - sought to make good faith determinations of privilege in the fast-moving examination, and did not find any instance where critical documents were placed on privilege logs without any basis for a claim of privilege simply to prevent their production to OFHEO. As for the advice the Board received, our interviews of Board members revealed that they mistakenly believed that the forensic accounting firm outside counsel 29
had retained to assist it in the Special Examination had been engaged to validate the Company's accounting practices, and had opined that those practices complied with GAAP. This misconception arose because outside counsel did not clearly explain to the Board the accountant's limited role throughout the course of the Special Examination. As a result, Board members took significant but unwarranted comfort in the belief that the Company's accounting practices were supported by two major accounting firms: the Company's outside auditors, and the forensic accounting firm hired by outside counsel for the Special Examination. We saw no evidence that would call into question the good faith of the Company's lawyers, or their experts, who were undoubtedly taking directions directly from management about the overall strategy to take with respect to defending the Company in the Special Examination. However, neither management nor the Company's lawyers provided the Board with sufficient information about the issues raised in OFHEO's Special Examination to allow the Board to weigh the risks and make an informed decision about the best course for the Company. Management and the Company's counsel focused unduly on OFHEO's motives in conducting the Special Examination, and they incorrectly dismissed numerous accounting issues as "OFHEO's arguments" and "disagreements." On the one occasion when management and Company attorneys gave the Board a substantive presentation about the issues under review during the Special Examination, they understated the problems, telling the Board of possible "OFHEO arguments" or "disagreements" accompanied by ready assurances that such practices were reasonable and defensible, and did not give the Board a sufficient indication that OFHEO's "arguments" may be well founded. It turned out, of course, that management and the Company's lawyers were wrong about the accounting issues raised by OFHEO. The Company would have been better served if management and the Company's lawyers had informed the Board of all of the material facts and analyzed and discussed the risks arising from those facts in a more dispassionate fashion, and we recommend that the Company's lawyers make a concerted effort to give more balanced and comprehensive presentations to the Board in the future VII. OTHER ALLEGATIONS A. Issues Raised by Current and Former Fannie Mae Employees The Chairperson of the SRC caused a "broadcast" message to be sent to all Fannie Mae employees on November 29, 2004, which encouraged Company employees to contact Paul, Weiss directly with any information or knowledge they might have about "any unusual or atypical transactions in the past five years." In response, a number of Fannie Mae employees contacted us. Several of the issues that were raised were incapable of further review due to the unavailability of the employee to provide specific factual information, and we referred one issue to the Company for further resolution. Some of the information we received related to topics already within our scope from the OFHEO Agreements and other issues. One contact led to our review of the Minority Lending Initiatives and another led to our investigation of the Company's consideration of certain insurance policies. 30
B. Fannie Mae's Equity Investments in Gulf Bank Finally, we investigated the allegations raised by an anonymous former employee in a letter submitted to OFHEO and to the Chairman of the Board concerning the Company's investment in a bank in South Florida, Gulf Bank. The anonymous letter raised questions about the basis for, and the motive behind, an $800,000 equity investment in the minority-owned bank that the Company made as part of its Community Development Financial Institution ("CDFI") program. The author of the letter also made specific allegations that a senior Fannie Mae officer received inappropriate gifts from the Chairman of Gulf Bank, Salvador Bonilla-Mathe. We concluded that the allegations against the officer were unfounded. He received two gifts from Bonilla of minimal value (a bag of coffee and a book about a charity with which Bonilla was associated). Bonilla also sent him a chess set which, following the advice of counsel, he returned to Bonilla. We have found no evidence of misconduct on his part in this respect, or in any other aspect of the Gulf Bank transaction. We also did not find that the decision to invest in Gulf Bank was inappropriate at the time the investment in Gulf Bank closed, Fannie Mae was aware that the bank was under some scrutiny by the Federal Reserve Bank, but it does not appear that the extent of the scrutiny was known; moreover, the decision to proceed with the investment was made with the advice of outside counsel. Finally, although Fannie Mae's investment in Gulf Bank was not written off immediately, it was written off about eighteen months after the investment took place following an outside firm's valuation of all CDFI investments. 31
EXHIBIT 99.2 A REPORT TO THE SPECIAL REVIEW COMMITTEE OF THE BOARD OF DIRECTORS OF FANNIE MAE PAUL, WEISS, RIFKIND, WHARTON & GARRISON LLP Warren B. Rudman Alex Young K. Oh Robert P. Parker Daniel J. Kramer HURON CONSULTING GROUP INC. George E. Massaro Jeffrey H. Ellis February 23,2006
TABLE OF CONTENTS PAGE CHAPTER I: EXECUTIVE SUMMARY ............................................................ (Separate Volume) CHAPTER II: INVESTIGATION METHODOLOGY .................................................... 2 I. INTRODUCTION ..................................................................... 2 A. The OFHEO Report ......................................................... 2 1. FAS 91 ................................................................ 3 2. FAS 133 ............................................................... 4 3. Other Chapters ........................................................ 4 B. Issues to be Reviewed by Agreement with OFHEO ............................ 6 1. September 2004 Agreement .............................................. 6 2. The February 2005 OFHEO Letter and Supplemental Agreement ............. 8 C. Other Issues Brought to Paul, Weiss's Attention .......................... 9 1. Directive to Employees to Come Forward ................................ 9 2. Monitoring Consultants's Activities ................................... 9 3. The Company's November 10, 2005 12b-25 Filing ......................... 12 D. Engagement of Huron ...................................................... 12 E. Documents Reviewed ....................................................... 13 F. Interviews Conducted ..................................................... 14 G. Independent Collection and Analysis of Electronic Materials .............. 15 H. Assessment as to the Company's Cooperation ............................... 17 1. The Board, the SRC and Management ..................................... 17 2. The Leanne Spencer Document Incident .................................. 18 CHAPTER III: BACKGROUND ................................................................... 21 I. INTRODUCTION ..................................................................... 21 II. THE EVOLUTION OF FANNIE MAE ...................................................... 21 III. FANNIE MAE'S ROLE TODAY .......................................................... 24 IV. FANNIE MAE'S PRIMARY BUSINESS ACTIVITIES ......................................... 25 A. Portfolio Investment ..................................................... 25 B. Credit Guaranty .......................................................... 26 V. REGULATORY REGIME ................................................................ 27 A. HUD ...................................................................... 28 B. OFHEO .................................................................... 29 1. Capital Requirements .................................................. 29 2. Corporate Governance .................................................. 30 CHAPTER IV: ACCOUNTING FOR THE AMORTIZATION OF PREMIUM/DISCOUNT ON LOANS AND MORTGAGE - BACKED SECURITIES ................................................. 33 I. INTRODUCTION ..................................................................... 33 II. BACKGROUND ....................................................................... 34 i
PAGE A. Applicable Accounting Standard ....................................................... 34 B. Relevant Amortization Accounting Systems ............................................. 36 C. Accounting Practices Under FAS 91 Prior to 1998 ...................................... 37 1. Management Never Recorded Catch-up Until 1998 ..................................... 37 2. Management Did Not Model Its Non-Core Book In Accordance With GAAP ................ 38 III. 1998 CATCH-UP CALCULATION AND ADJUSTMENT ..................................................... 42 A. The Catch-Up Calculation for 1998 .................................................... 42 B. Evidence Reflecting Management's Focus on Meeting EPS Targets ........................ 43 C. Other Adjustments Made to Mitigate Impact of Catch-Up Expense on 1998 EPS ............ 50 1. Changes in Accounting for LIHTC ................................................... 50 2. $3.9 Million in Miscellaneous Income from Account 162200 .......................... 54 D. Management's Disclosures About the 1998 Adjustments .................................. 57 1. Disclosures to KPMG ............................................................... 57 2. Disclosures Made to the Board of Directors ........................................ 59 3. Disclosures Made to the Public .................................................... 62 E. Management's Reasons for Recording Only $240 Million ................................. 63 1. Interest Rates Volatility Contributed to the Large Catch-Up ....................... 64 2. Management's View That Interest Rates Would Rise in the Future .................... 65 3. Imprecision of the Modeling for the Non-Core Book ................................. 67 4. Inadequacy of the Modeling Systems for the Core Book .............................. 70 F. Findings Regarding the 1998 Catch-Up Decision ........................................ 71 IV. FANNIE MAE'S PREMIUM AND DISCOUNT AMORTIZATION POLICY ........................................ 73 A. Development of the Policy ............................................................ 73 1. Practice Prior to 1999 ............................................................ 73 2. Development of the Amortization Policy ............................................ 74 3. Evidence Reflecting Management's Purpose for the Amortization Policy .............. 75 4. KPMG's Involvement in the Development of the Amortization Policy .................. 81 5. The Final Amortization Policy Adopted in December 2000 ............................ 81 B. The Application of the Amortization Policy ........................................... 84 1. Management Used a Variety of Tools To Manage Catch-Up ............................. 85 2. Catch-Up Was Initially Calculated for Year-End .................................... 88 3. Management Made Adjustments Even When Catch- Up Fell Within the Predetermined Threshold ......................................................................... 89 ii
PAGE 4. KPMG's Awareness of the Manner in Which the Amortization Policy Was Implemented ................................................................... 90 C. Inadequate Disclosures to the Board About the Amortization Policy and Its Application .................................................................. 92 D. 2004 Revisions to the Amortization Policy ........................................ 93 E. Findings Regarding the Amortization Policy ....................................... 94 CHAPTER IV APPENDIX: FANNIE MAE'S FAS 91-RELATED SYSTEMS ........................................... 96 CHAPTER V: ACCOUNT FOR DERIVATIVE INSTRUMENTS AND HEDGING ACTIVITIES ............................ 101 I. INTRODUCTION ............................................................................ 101 II. BACKGROUND .............................................................................. 103 A. Fannie Mae's Debt Policies and Practices ......................................... 103 B. Derivatives and Hedging .......................................................... 105 C. Significant Hedge Transactions ................................................... 108 III. AN OVERVIEW OF FAS 133 .................................................................. 111 1. Requirements for Hedge Accounting ............................................. 112 2. Accounting for Ineffectiveness ................................................ 114 3. Term-Out Transactions and the "Fannie Mae Carve - Out" ........................ 117 IV. FINDINGS REGARDING FANNIE MAE'S IMPLEMENTATION OF FAS 133 ............................... 118 A. The Three Tenets ................................................................. 119 B. Development Efforts During 1999 and 2000 ......................................... 128 C. FAS 138 .......................................................................... 132 D. Post-Effective Date Developments ................................................. 133 V. FINDINGS REGARDING FANNIE MAE'S ACCOUNTING FOR HEDGE TRANSACTIONS ............................. 135 A. Hedge Accounting Methodology ..................................................... 136 B. Significant Aspects of Fannie Mae's Hedge Accounting ............................. 140 1. Fair Value Equal to Zero ...................................................... 140 2. Seven-Day Tolerance for Matching Repricing Dates .............................. 143 3. Forecasted Transactions ....................................................... 145 4. The De Minimis Test ........................................................... 149 5. Identification and Probability of Forecasted Transactions ..................... 152 6. Term-Outs ..................................................................... 155 C. Hedge Documentation .............................................................. 156 D. Other Hedge Accounting Policies .................................................. 160 VI. FINDINGS REGARDING OVERSIGHT OF FANNIE MAE'S HEDGE ACCOUNTING ................................ 162 A. Senior Management ............................................................. 163 B. The Board and the Audit Committee ............................................. 163 C. FASB .......................................................................... 165 D. KPMG .......................................................................... 167 E. OFHEO ......................................................................... 171 iii
PAGE F. Internal Audit ................................................................... 172 CHAPTER VI: OTHER ACCOUNTING ISSUES .................................................................. 174 PART A: ACCOUNTING FOR THE ALLOWANCE FOR LOAN LOSSES .......................................... 174 I. INTRODUCTION .......................................................................... 174 II. BACKGROUND ............................................................................ 175 A. Identification of the Allowance as a Potential Issue ............................. 175 B. Relevant Accounting Principles ................................................... 178 III. FINDINGS CONCERNING FANNIE MAE'S ACCOUNTING FOR THE ALLOWANCE ......................... 179 A. Fannie Mae's Historical Practice for Setting the Allowance ....................... 179 1. Overview of the Components of Fannie Mae's Allowance .......................... 179 2. Management's Process for Setting the Allowance ................................ 180 B. Evidence Reflecting Management's View of the Allowance ........................... 189 1. Management's Awareness that the Allowance May Be Overstated ................... 189 2. Evidence Reflecting Management's Consideration of Using the Allowance to Offset Unrelated Expenses ..................................................... 191 3. Management's Safety and Soundness Perspective on the Allowance ................ 192 C. Transparency of the Allowance Setting Process to OFHEO ........................... 194 D. Conclusions Regarding Fannie Mae's Accounting for the Allowance .................. 196 PART B: ACCOUNTING FOR DOLLAR ROLLS ........................................................... 197 I. INTRODUCTION .......................................................................... 197 II. BACKGROUND ............................................................................ 199 A. FAS 140 and Other Accounting Literature .......................................... 199 B. OFHEO'S Concerns ................................................................. 201 C. Fannie Mae's Accounting Policy ................................................... 202 D. Fannie Mae's Dollar Roll Transactions ............................................ 205 1. Mechanics of the Dollar Roll Transaction ...................................... 205 2. Dollar Roll Fails ............................................................. 205 3. Return of Substantially the Same Collateral ................................... 206 III. FINDINGS .............................................................................. 214 PART C: ACCOUNTING FOR FORWARD COMMITMENTS .................................................... 214 I. INTRODUCTION .......................................................................... 214 II. BACKGROUND ON FAS 149 ................................................................. 217 A. Commitments under FAS 133 Prior to FAS 149 ....................................... 217 B. Adoption of FAS 149 .............................................................. 218 III. BACKGROUND ON FANNIE MAE'S FIRM COMMITMENT PRACTICES .................................. 219 IV. FINDINGS REGARDING FANNIE MAE'S IMPLEMENTATION OF FAS 149 ............................. 221 iv
PAGE V. FINDINGS REGARDING FANNIE MAE'S ACCOUNTING FOR FIRM COMMITMENTS ............................... 226 A. The Company's FAS 149 Policy ............................................................... 226 B. Documentation of Hedged Transactions ....................................................... 228 C. The Probability Assessment ................................................................. 231 VI. FINDINGS REGARDING SENIOR MANAGEMENT, OFHEO, AND THE BOARD .................................... 235 PART D: CLASSIFICATION OF SECURITIES HELD IN PORTFOLIO ................................................ 238 I. INTRODUCTION .................................................................................. 238 II. BACKGROUND .................................................................................... 239 A. Applicable Accounting Principles ........................................................... 239 B. Fannie Mae's Accounting Policy ............................................................. 241 III. FINDINGS ...................................................................................... 241 A. Fannie Mae's Classification of Debt Securities ............................................. 241 B. Practical Application of the Company's Policies ............................................ 245 C. KPMG Review ................................................................................ 248 D. Executive Involvement and the Report to the Audit Committee ................................ 249 IV. CONCLUSIONS ................................................................................... 249 PART E: RECOGNITION OF INTEREST EXPENSE AND INCOME .................................................... 250 I. INTRODUCTION .................................................................................. 250 II. BACKGROUND .................................................................................... 251 III. FINDINGS REGARDING THE RECOGNITION OF INTEREST EXPENSE AND INCOME ............................. 257 PART F: ACCOUNTING FOR OTHER-THAN-TEMPORARY IMPAIRMENT OF MANUFACTURED HOUSING BONDS AND AIRCRAFT ASSET-BACKED SECURITIES .............................................................. 257 I. INTRODUCTION .................................................................................. 257 II. BACKGROUND .................................................................................... 258 A. Applicable Accounting Standards ............................................................ 258 B. Fannie Mae's Practice and Policy ........................................................... 260 1. OTTI Recorded ........................................................................... 260 2. Types of Investments .................................................................... 262 3. OTTI Accounting and Monitoring Methodologies ............................................ 264 III. FINDINGS ...................................................................................... 269 A. Accounting Policy and Monitoring ........................................................... 269 B. Management Modification of the MH Bond DCF Modeling ........................................ 270 C. Fair Value Determinations .................................................................. 273 1. MH Bonds ................................................................................ 273 2. Aircraft ABS ............................................................................ 274 D. Timing of Recognition ...................................................................... 276 PART G: ACCOUNTING FOR INVESTMENTS IN INTEREST-ONLY MORTGAGE-BACKED SECURITIES ........................ 277 I. INTRODUCTION .................................................................................. 277 II. BACKGROUND .................................................................................... 278 A. Applicable Accounting Standards ............................................................ 278 v
PAGE B. Fannie Mae's Accounting for Investments in IO MBS ....................................... 278 1. Motivation for the Decision to Combine IO MBS with Other Securities for Accounting Purposes .............................................................. 279 2. Discussions with KPMG ................................................................ 281 3. The Internal Policy for the Acquisition of IO MBS .................................... 283 4. Subsequent Accounting for Synthetic MBS .............................................. 285 III. FINDINGS ...................................................................................... 287 A. Management's Disclosures About Accounting for IO MBS to the Board of Directors or its Committees ....................................................................... 287 B. Discussions with the SEC Regarding Accounting for IO MBS ................................ 289 IV. CONCLUSIONS ................................................................................... 290 PART H: SECURITIZATION OF WHOLLY-OWNED MBS ............................................................ 291 I. INTRODUCTION .................................................................................. 292 II. BACKGROUND .................................................................................... 292 A. Accounting Literature on Consolidation .................................................. 292 B. OFHEO's Concerns ........................................................................ 293 C. The Company's Approach to Consolidation ................................................. 294 III. CONCLUSIONS ................................................................................... 299 PART I: ACCOUNTING FOR INCOME TAX RESERVES AND CERTAIN TAX-ADVANTAGED TRANSACTIONS .................... 300 I. INTRODUCTION .................................................................................. 300 II. BACKGROUND .................................................................................... 300 A. Applicable Authoritative Literature and Principles ...................................... 300 B. Fannie Mae's Accounting for Income Tax Reserves and Tax-Advantaged Transactions ......... 301 1. Tax Reserves ......................................................................... 301 2. Fannie Mae's Recognition of Synfuel Tax Credits ...................................... 304 3. Issues Related to the STIS Transactions .............................................. 308 III. FINDINGS ...................................................................................... 312 PART J: ACCOUNTING FOR INSURANCE PRODUCTS ............................................................. 313 I. INTRODUCTION .................................................................................. 313 II. BACKGROUND .................................................................................... 314 A. Role of Credit Policy, Generally ........................................................ 314 B. Fannie Mae's Consideration of Finite Risk Insurance Policies ............................ 316 1. MMC Proposal ......................................................................... 316 2. Radian Transaction ................................................................... 318 3. ACE Proposal ......................................................................... 323 4. Corporate Owned Life Insurance ....................................................... 326 III. FINDINGS REGARDING INSURANCE PRODUCTS ......................................................... 328 PART K: ACCOUNTING FOR OUT-OF-PORTFOLIO SECURITIZATION ("PORTFOLIO POOLING SYSTEM") ................... 331 I. INTRODUCTION .................................................................................. 331 II. BACKGROUND .................................................................................... 332 A. PPS and the Coding Error ................................................................ 332 vi
PAGE B. Accounting for HFS Loans ................................................................ 334 III. FINDINGS REGARDING THE PORTFOLIO POOLING SYSTEM ............................................. 335 PART L: THE DEBT REPURCHASE ("BUYBACK") PROGRAM ..................................................... 336 I. INTRODUCTION ................................................................................ 336 II. BACKGROUND .................................................................................. 337 III. FINDINGS REGARDING FANNIE MAE'S BUYBACK PROGRAM ............................................. 340 A. Earnings-Related Strategy ............................................................... 341 B. Documentation for Buyback Decisions ..................................................... 344 C. Economics of Fannie Mae Debt Buybacks ................................................... 345 D. Disclosures ............................................................................. 347 1. Public Disclosure .................................................................... 347 2. Board Disclosure ..................................................................... 350 IV. RECOMMENDATIONS ............................................................................. 352 PART M: ACCOUNTING FOR THE AMORTIZATION OF CALLABLE DEBT EXPENSES ................................... 353 I. INTRODUCTION ................................................................................ 353 II. BACKGROUND .................................................................................. 354 A. Applicable Accounting Principles ........................................................ 354 B. Fannie Mae's Application of the Literature .............................................. 355 C. Amortization End-Date Changes ........................................................... 357 D. The September 2002 End-Date Adjustment .................................................. 360 III. FINDINGS .................................................................................... 364 PART N: MINORITY LENDING INITIATIVE ................................................................. 365 I. INTRODUCTION ................................................................................ 365 II. BACKGROUND .................................................................................. 365 A. The MLI Program ......................................................................... 365 B. Accounting for the Payments to RBMG ..................................................... 368 III. FINDINGS REGARDING THE MLI PROGRAM .......................................................... 369 PART O: ACCOUNTING FOR REALIGNMENTS AND THE SECURITY MASTER PROJECT ................................. 369 I. INTRODUCTION ................................................................................ 369 II. BACKGROUND .................................................................................. 370 III. FINDINGS REGARDING THE ACCOUNTING TREATMENT OF REALIGNMENT IMPACTS .......................... 371 A. Management's Deferral of Realignment Impacts ............................................ 371 1. Deferral of STATS Realignment Impacts ................................................ 372 2. Deferral of LASER Realignments Prior to 1998 ......................................... 375 B. Some Realignment Impacts Were Recorded into Income in the Period in Which They Were Identified .............................................................................. 375 C. Inclusion of Realignment Impacts in the Analysis of Catch - Up .......................... 376 1. Security Master Project .............................................................. 376 vii
PAGE 2. Management's Inclusion of Realignment Impacts in Catch-up Calculation Was Not Consistent with GAAP ......................................................... 384 D. Findings Regarding Fannie Mae's Belated Proposal of a Realignment Policy that Was Inadequate Under GAAP .................. 385 PART P: ACCOUNTING FOR INVESTMENTS IN AFFORDABLE HOUSING PARTNERSHIPS ................ 387 I. INTRODUCTION ................................................................. 387 II. BACKGROUND ................................................................... 387 A. Applicable Accounting Standards .................................... 387 B. Fannie Mae's Investments in Affordable Housing Partnerships ........ 389 C. Fannie Mae's Partnership Accounting ................................ 390 1. Accounting for Capital Contributions ......................... 390 2. The Company's Accounting Methodologies ....................... 392 3. Impairment Analysis .......................................... 397 III. FINDINGS .................................................................... 399 CHAPTER VII: CORPORATE GOVERNANCE AND INTERNAL CONTROLS ................................... 401 I. BOARD OF DIRECTORS ........................................................... 401 A. Summary ............................................................ 401 B. Overview Prior to September 2004 ................................... 402 1. Overview of the Board ........................................ 402 2. Committees of the Board ...................................... 406 3. Board Oversight of Accounting and Financial Reporting In Practice .................................................. 413 C. Findings ........................................................... 416 1. The Board Sought to Meet Evolving Corporate Governance Standards .................................................... 416 2. The Board Was Engaged on Accounting and Financial Reporting Issues ....................................................... 418 D. Subsequent Developments ............................................ 418 1. Separation of Chairman and CEO Functions ..................... 418 2. Changes to Board Committee Structure ......................... 420 3. Changes to Policies and Practices ............................ 422 II. OFFICE OF THE CHAIRMAN ....................................................... 424 A. Summary ............................................................ 424 B. Overview Prior to September 2004 ................................... 425 1. Chairman and Chief Executive Officer ......................... 425 2. Vice Chairmen of the Board ................................... 427 3. Other Members of the Office of the Chairman .................. 430 4. Management Committees ........................................ 431 C. Findings ........................................................... 437 1. Management Did Not Fully Inform the Board of Accounting Issues or Internal Control Deficiencies ...................... 437 viii
PAGE 2. Management Created an Environment that was not Conducive to Open Discussion and Exchange of Views .............................. 438 3. Inadequate Management of Corporate Risks ........................... 446 D. Subsequent Developments ................................................... 449 1. Personnel and Structural Changes to Senior Management .............. 450 2. Changes to Corporate Culture ....................................... 454 III. OFFICE OF AUDITING .............................................................. 456 A. Summary ................................................................... 456 B. Overview Prior to September 2004 .......................................... 457 1. Functions and Responsibilities ..................................... 457 2. Organization and Structure ......................................... 458 3. Past Evaluations of the Office of Auditing ......................... 463 C. Findings .................................................................. 464 1. Leadership ......................................................... 464 2. Relationship with Management ....................................... 465 3. Resource Deficiencies .............................................. 468 4. Inadequate Communications .......................................... 472 5. Deficiencies in the Office of Auditing's SOX 404 Work .............. 475 D. Subsequent Developments ................................................... 476 1. Changes in Leadership and Reporting Line ........................... 476 2. Focus on Core Audit Mission ........................................ 477 3. Internal Structure and Staffing .................................... 477 4. Changes to Audit Reporting ......................................... 478 IV. ETHICS AND COMPLIANCE FUNCTIONS ................................................. 478 A. Summary ................................................................... 478 B. Overview Prior to September 2004 .......................................... 479 1. Code of Business Conduct ........................................... 479 2. Structure of Board Oversight ....................................... 481 3. Management Oversight and Activities ................................ 483 4. Office of Corporate Justice ........................................ 489 5. Office of Corporate Compliance ..................................... 495 C. Findings .................................................................. 498 1. Unstructured Information Flow to the Board ......................... 499 2. Insufficient Resources ............................................. 499 3. Inappropriate Placement of the Ethics and Compliance Functions .......................................................... 500 4. Weakness of the Chief Compliance Officer Position .................. 500 5. Insufficient Cross-Enterprise Coordination ......................... 501 D. Subsequent Developments ................................................... 501 1. Creation of an Independent Ethics, Compliance & Investigations Organization ....................................................... 501 2. Expanded Role of Compliance Committee .............................. 503 3. Enhanced Reporting to the Board .................................... 504 ix
PAGE V. OFFICE OF THE CONTROLLER ........................................................ 505 A. Summary ................................................................... 505 B. Overview Prior to September 2004 .......................................... 506 1. Range of Responsibilities ........................................... 506 2. Senior Vice President and Controller ................................ 506 3. Key Units within Controller's Office ................................ 507 C. Findings .................................................................. 510 1. Inadequate Resources ................................................ 511 2. Deficiencies in the Closing Process ................................. 514 D. Subsequent Developments ................................................... 516 1. Change in Leadership ................................................ 516 2. Structural and Related Personnel Changes ............................ 516 3. Staffing Levels ..................................................... 519 4. Changes to the Closing Process ...................................... 519 CHAPTER VIII: EXECUTIVE COMPENSATION ....................................................... 520 I. INTRODUCTION .................................................................... 520 II. PAUL, WEISS'S INVESTIGATION ..................................................... 522 A. Document Review ........................................................... 522 B. Interviews ................................................................ 522 III. EXECUTIVE COMPENSATION AT FANNIE MAE ............................................ 523 A. Executive Compensation Overview ........................................... 523 1. Fannie Mae's Charter ................................................ 523 2. Fannie Mae's Compensation Philosophy ................................ 523 3. Components of Compensation .......................................... 525 B. The Annual Incentive Plan ................................................. 526 1. Calculation of the AIP Pool and Individual Awards ................... 526 2. Maximizing AIP Funding Pools ........................................ 528 3. The Role of the EPS Incentive ....................................... 533 4. Reducing the Influence of EPS on Compensation ....................... 535 C. Performance Share Plan .................................................... 537 1. Quantitative (EPS) Component ........................................ 538 2. Qualitative (Subjective) Component .................................. 539 IV. LEGAL DEPARTMENT INVOLVEMENT IN COMPENSATION DECISIONS .......................... 544 CHAPTER IX: FANNIE MAE'S INVESTIGATION OF ROGER BARNES'S ALLEGATIONS ........................ 546 I. INTRODUCTION .................................................................... 546 II. BACKGROUND ...................................................................... 547 A. Barnes's Work History ..................................................... 547 B. Barnes's Allegations and the Company's Investigations ..................... 548 III. FINDINGS REGARDING THE INVESTIGATION INTO BARNES'S ACCOUNTING ALLEGATIONS ....... 552 A. August 4, 2003 ............................................................ 552 B. August 5, 2003 ............................................................ 553 C. August 6-7, 2003 .......................................................... 555 x
PAGE D. August 8, 2003 ......................................................... 556 E. August 9-12, 2003 ...................................................... 558 IV. FINDINGS REGARDING ROGER BARNES'S FAS 91-RELATED ALLEGATIONS ................ 559 A. Factor Anomalies ....................................................... 559 B. Manual Factor Changes .................................................. 561 C. CPR Edits .............................................................. 562 V. FINDINGS REGARDING THE INVESTIGATION INTO BARNES'S CULTURAL COMPLAINTS ...... 563 VI. FINDINGS REGARDING THE INVESTIGATION INTO BARNES'S DISCRIMINATION ALLEGATIONS ................................................................. 564 VII. FINDINGS REGARDING THE RESULTS OF THE OCC INVESTIGATIONS .................... 564 A. OCC Decision Part A: Internal Audit's Analysis ......................... 564 B. OCC Decision Part B: Controller's Office Culture ....................... 565 VIII. FINDINGS REGARDING THE RESULTS OF THE OCJ INVESTIGATION ..................... 566 IX. FINDINGS REGARDING BARNES'S RESIGNATION AND SETTLEMENT OF THREATENED LITIGATION .................................................................. 567 X. FINDINGS REGARDING KPMG'S REVIEW AND FINAL REPORT ........................... 569 XI. FINDINGS REGARDING BARNES'S TESTIMONY TO OFHEO AND CONGRESS ................. 570 CHAPTER X: MANAGEMENT'S CONDUCT DURING OFHEO'S SPECIAL EXAMINATION ................... 571 I. INTRODUCTION ................................................................ 571 II. BACKGROUND .................................................................. 571 A. Freddie Mac's Announcement ............................................. 571 B. OFHEO's Announcement to Examine Fannie Mae ............................. 571 III. FANNIE MAE'S RESPONSE TO OFHEO REQUESTS ..................................... 572 A. Document Preservation Efforts .......................................... 572 B. E-mail Review .......................................................... 575 C. Hardcopy Documents ..................................................... 577 D. Interviews ............................................................. 578 E. Findings Regarding the Company's Response to OFHEO Requests ............ 578 IV. CONDUCT OF THE LEGAL DEPARTMENT AND ITS ADVISORS DURING THE SPECIAL EXAMINATION ................................................................. 582 A. There is No Evidence That Privilege Assertions Were Made in Bad Faith to "Cloak" Relevant Documents from OFHEO ......................... 582 B. The Board Was Advised By Wilmer and E&Y That the Company's Accounting Was Reasonable and Defensible ............................... 582 CHAPTER XI: OTHER ALLEGATIONS ......................................................... 589 xi
PAGE PART A: ISSUES RAISED BY CURRENT OR FORMER FANNIE MAE EMPLOYEES .......................... 589 I. INTRODUCTION ..................................................................... 589 A. HTM/AFS Redesignation ....................................................... 591 B. Non-Specific Anonymous Call ................................................. 591 C. Home Solutions (Minority Lending Initiative) ................................ 591 D. Contract and Procurement Systems ("CAPS") ................................... 592 E. Anticipatory Hedge Transaction .............................................. 593 F. Human Resources ............................................................. 593 G. Radian Credit Enhancement Product ........................................... 593 H. Partnership Office .......................................................... 594 I. Fannie Mae Loan Servicing Litigation ...................................... 594 PART B: FANNIE MAE'S EQUITY INVESTMENTS IN GULF BANK ..................................... 595 I. INTRODUCTION ..................................................................... 595 II. BACKGROUND ON THE CDFI PROGRAM ................................................... 597 A. The CDFI Program ............................................................ 597 B. The "Typical" CDFI Transaction Life Cycle ................................... 597 C. The Gulf Bank Transaction ................................................... 598 1. Genesis of Fannie Mae's Relationship with Gulf Bank ..................... 598 2. The Request for an Equity Investment .................................... 599 3. The CDFI Investment Committee ........................................... 600 4. The Transaction Loses Momentum .......................................... 603 5. Gulf Bank's Disclosure Regarding the FRB Examination .................... 605 6. Fannie Mae's Evaluation of the Gulf Bank Disclosure ..................... 606 7. The Closing ............................................................. 608 8. The Cease and Desist Order .............................................. 609 9. Renewal of the CD and the Write-Off ..................................... 610 10. Subsequent Events ...................................................... 611 III. FINDINGS ......................................................................... 611 A. Due Diligence ............................................................... 611 B. Valuation ................................................................... 611 C. Policy Against Having Both CD and Equity Investment ......................... 612 D. Allegations of Wrongdoing Against Rob Levin ................................. 612 E. Levin's Relationship with Salvador Bonilla .................................. 613 F. Levin's Receipt of Gifts .................................................... 613 G. Franchise Value and the CDFI Program Generally .............................. 614 1. Franchise Value ......................................................... 614 2. Mission Value ........................................................... 614 3. Economic Value .......................................................... 615 xii
CHAPTER I: EXECUTIVE SUMMARY (Separate Volume) 1
CHAPTER II: INVESTIGATION METHODOLOGY I. INTRODUCTION This Chapter details the scope of Paul, Weiss's investigation and its investigation methodologies. Paul, Weiss was retained by the Special Review Committee of the Board of Directors of Fannie Mae ("SRC") in September 2004 after OFHEO issued its report, in which OFHEO found, among other things, that several of the Company's critical accounting policies were not consistent with GAAP.(1) The scope of Paul, Weiss's review was initially defined by an agreement executed between the Board of Directors of Fannie Mae and OFHEO (the "September 2004 Agreement"),(2) which agreement was supplemented in March 2005 (the "Supplemental Agreement").(3) Other matters came to Paul, Weiss's attention during the review, which were incorporated into the scope. Pursuant to the above agreements, and with the authorization of both the SRC and OFHEO, Paul, Weiss retained Huron Consulting Group ("Huron"), a forensic consulting firm, to provide expert assistance in the investigation. The facts set forth in this report are based on our review of all of the documentary and other information made available to us by Fannie Mae and its agents, and on our interviews of certain former and current employees of the Company and its agents. In addition, Huron conducted independent and forensic collection and analysis of electronic data directly from Fannie Mae. A. The OFHEO Report In September 2004, almost a year after issuing its first document request to Fannie Mae, OFHEO issued its Report, which set forth OFHEO findings to date from the Special Examination. This section briefly summarizes some of the critical findings of the OFHEO Report, which were incorporated into the scope of Paul, Weiss's review. - ---------- (1) OFHEO Report, available at http://www.ofheo.gov/media/pdf/FNMfindingstodate17 sept04.pdf. (2) Agreement between the SRC and OFHEO, dated Sept. 27, 2004, P VI.2.(a) (hereinafter "September 2004 Agreement"), available at: http://www.ofheo.gov/media/pdf/fnmagreement92704.pdf. This and other examples of documents relevant to the discussion in this Chapter can be found in the accompanying, "A Report to the Special Review Committee of the Board of Directors of Fannie Mae: Appendix" (hereinafter "Appendix"), at Tab A. (3) Supplemental agreement between the SRC and OFHEO, dated Mar. 7, 2005, (hereinafter "Supplemental Agreement"), available at: http://www.ofheo.gov/media/pdf/fannie05agreement.pdf. 2
1. FAS 91 OFHEO concluded in its Report that the Company's accounting for purchase premium and discount amortization and amortization of other deferred charges pursuant to FAS 91 was not in accordance with GAAP. With regard to the Company's FAS 91 accounting for 1998 in particular, OFHEO found that "the Enterprise has not provided OFHEO with any credible analysis which supports the recording of only half of the calculated" amortization expense amount,(4) and that management's deferral of approximately $200 million of estimated amortization expense, "enabled management of the Enterprise to receive 100% of their annual bonus compensation."(5) With respect to the Company's amortization policy, which contained a predetermined materiality threshold within which management could choose not to make adjustments required under FAS 91,(6) OFHEO concluded that this policy was "inconsistent with GAAP, and [was] designed to provide earnings flexibility and minimize earnings volatility."(7) OFHEO further concluded that management inconsistently applied the policy, generated multiple estimates of amortization with varying assumptions so that it could select the estimates with optimal results, and engaged in other efforts the result of which was to create a "cookie jar" reserve to manipulate earnings.(8) OFHEO also concluded that there were "a number of significant control weaknesses in the process of accounting for amortization" used by Fannie Mae that call into question the accuracy of the Company's premium and discount amortization.(9) Specifically, OFHEO found that management: (1) did not appropriately segregate duties and allowed the formation of key person dependencies; (2) poorly documented its accounting practices and results, and did not create or maintain adequate audit trails; (3) used bad data, including illogical or anomalous amortization factors; and (4) modeled multiple alternatives to produce desired results.(10) In a related area, OFHEO deemed inadequate the investigation by the Office of Auditing ("Internal Audit") into the allegations raised in 2003 by then - ---------- (4) OFHEO Report at 8 (footnote omitted). (5) Id. at 1. (6) Id. at 19. (7) Id. at 13 (emphasis omitted). (8) Id at 1-2. (9) Id. at 65. (10) Id. at 66. 3
employee, Roger Barnes. Barnes's allegations related to purported irregularities in management's amortization accounting and the modeling process, including manual factor changes. OFHEO concluded that "the OA [Office of Auditing] analysis... was flawed for the simple reason that their procedures were insufficient to asses the breadth of the issues or their quantitative impact on the catch-up analysis."(11) And furthermore, "[t]he lack of diligence on behalf of the OA in the matter of the manual factor change is inconsistent with their responsibility to exercise due professional care."(12) 2. FAS 133 With respect to FAS 133, OFHEO concluded in its Report that Fannie Mae's implementation of FAS 133 was designed primarily to minimize earning volatility and simplify operations, and that it "misapplied GAAP... in pursuit of these objectives."(13) OFHEO highlighted several specific deficiencies in the Company's approach to FAS 133, including its: (1) failure to perform adequate assessments of hedge effectiveness by inappropriately treating the vast majority of its hedges as "perfectly effective"; (2) improper hedge accounting for derivatives with offsetting terms; (3) misapplication of the "short-cut" or "matched terms" method; (4) improper accounting for changes in both the time and intrinsic value components of purchased interest rate caps; and (5) ambiguous and incomplete documentation of hedging relationships.(14) In addition, OFHEO made a number of observations relating to Fannie Mae personnel and the overall environment within which the Company developed its derivative accounting policies, including: (1) the Company relied on strained Financial Standards resources for advice on the application of accounting policy; (2) accounting and treasury operations personnel lacked an adequate understanding of FAS 133's requirements; and (3) Fannie Mae was willing, on at least one occasion, to proceed with a desired accounting treatment even though they were unable to persuade the Financial Accounting Standards Board to incorporate that accounting into FAS 133.(15) 3. Other Chapters The OFHEO Report also drew several conclusions relating to the various controls and processes surrounding Fannie Mae's accounting and financial reporting, including accounting policy development and review, segregation of duties, and key - ---------- (11) Id. at 74. (12) Id. at 78 (emphasis omitted). (13) Id. at 82. (14) Id. at 90-91. (15) Id. at 87-89. 4
person dependencies. OFHEO focused on the roles and responsibilities of the CFO and executives in the Controller's Office.(16) OFHEO found, among other things, that the Company failed to properly develop and review accounting policies. According to OFHEO, this was partially because the Company relied on too few individuals to make key accounting policy decisions. OFHEO found that "the accounting policy development process within the Controller's Department lends itself to the formation of accounting policies that are aggressive in nature and which do not comport with a strict interpretation of GAAP."(17) OFHEO also identified "critical resource shortages and a lack of technical accounting expertise within the Controller's Office that resulted in key person dependencies."(18) According to OFHEO, both of these circumstances "resulted in an environment that impeded independent thinking and encouraged an inadequate level of policy and procedure development and documentation."(19) Moreover, OFHEO found that the lack of a formal process for developing accounting policy resulted in incomplete disclosures on critical accounting policies by the CFO to the Audit Committee of the Board.(20) The OFHEO Report also criticized a lack of proper segregation of duties in several areas. OFHEO noted that the reporting relationship between the head of Internal Audit and the CFO, which gave the CFO the authority to evaluate and affect compensation decisions for the head of Internal Audit, "critically impairs the independence of [Internal Audit]."(21) OFHEO found that the head of Internal Audit should be independent from the CFO and should report directly to the Audit Committee of the Board.(22) OFHEO also stated that the fact that the head of Internal Audit had been the prior Controller also presented "a major conflict of interest."(23) OFHEO additionally criticized the aggregation of responsibilities in the CFO position at Fannie Mae.(24) In addition to overseeing the head of Internal Audit, the CFO's responsibilities included "risk management, the retained portfolio, the accounting - ---------- (16) See id. at 146-69. (17) Id. at 146. (18) Id. (19) Id. at 147. (20) Id. at 151-52. (21) Id. at 161. (22) Id. (23) Id. at 162. (24) Id. at 157. 5
function, investor relations, treasury, and financial reporting."(25) OFHEO similarly concluded that the Senior Vice President--Financial Reporting and Planning inappropriately oversaw both forecasting the financial statements and financial reporting: "[h]aving the dual responsibility of modeling the amortization and reporting amortization results to the financial statements under the authority of a single individual is a major control weakness that undermines the integrity of the financial reporting process."(26) Similarly, OFHEO criticized as a control weakness the fact that a Director-Financial Reporting was responsible for both modeling the purchase premium and discount amortization amount as well as recording that amortization amount to the financial statements.(27) In sum, OFHEO concluded that "[a] combination of heavy workload, weak technical skills and a weak review environment contributed to the development of key person dependencies" and inadequate segregation of duties.(28) B. Issues to be Reviewed by Agreement with OFHEO 1. September 2004 Agreement In the aftermath of the OFHEO Report, on September 27, 2004, OFHEO and the Board of Fannie Mae entered into the September 2004 Agreement.(29) The September 2004 Agreement required the Board to retain an independent counsel and accounting firm to conduct reviews of several areas, including: (1) Fannie Mae's internal controls relating to accounting; (2) Fannie Mae's accounting and other corporate policies; (3) Fannie Mae's organizational structure and staffing, particularly related to its CFO, Controller's Office, accounting, audit, financial reporting, business planning and - ---------- (25) Id. at 156. (26) Id. at 159. (27) Id. at 159-60. (28) Id. at 168, 169. Subsequent to the OFHEO Report, and at Fannie Mae's request, the SEC's Office of the Chief Accountant ("OCA") reviewed the propriety of the Company's historical accounting practices under FAS 91 and FAS 133 as represented to the OCA by Fannie Mae. Letter from Jonathan Boyles to Stephen M. Cutler and Paul Berger, dated Nov. 3, 2004, FNMSEC 151-59; Letter from J. Boyles to S. Cutler and P Berger, dated Oct. 19, 2004, FMSEC 2215-64. On December 15, 2004, after reviewing Fannie Mae's submissions, the OCA concluded that Fannie Mae's accounting under both FAS 91 and FAS 133 were not in accordance with GAAP. SEC press release, dated Dec. 15, 2004, available at http://www.sec.gov/news/press/2004-172.htm. (29) See supra note 2; see also OFHEO press release, dated Sept. 27, 2004, available at http://www.ofheo.gov/News.asp?FormMode=Releases&ID=187. 6
forecasting, and modeling and financial standards functions; (4) Fannie Mae's compensation regime as related to strategic planning, and its impact on accounting and transaction decisions; and (5) Fannie Mae's governance procedures.(30) The terms of the September 2004 Agreement required Fannie Mae to undertake immediate steps in certain areas to remediate the concerns raised by OFHEO. Other provisions of the Agreement required Fannie Mae to retain third party experts to review and recommend various changes consistent with OFHEO's recommendations. As discussed in greater detail below, Paul, Weiss reviewed some of the work that was undertaken by the Company to implement the September 2004 Agreement, but such matters were not incorporated into Paul, Weiss's independent review. OFHEO approved of the Board's selection of Paul, Weiss as independent counsel on October 5, 2004.(31) Paul, Weiss's mandate under the September 2004 Agreement was to, among other things, conduct a "comprehensive review of the company's accounting policies and practices to ensure they accurately reflect the Board's objectives in complying with law and regulation and in overseeing the operations of Fannie Mae."(32) On October 19, 2004, the Board resolved that the SRC, comprised of independent directors, would oversee the work of Paul, Weiss.(33) The SRC would "oversee the independent reviews conducted by independent counsel and its consultant(s) as required in the agreement between the corporation and [OFHEO] dated September 27, 2004."(34) In its plan for implementing the September 2004 Agreement, the Board stated that the SRC "has instructed the Company to provide Paul, Weiss and the accounting firm the full support and resources of the Company. Paul, Weiss will also have access to all consultants hired in connection with this Agreement."(35) Furthermore, on October 12, 2004, the SRC caused a memorandum to be sent to Fannie Mae's Senior Leadership Team directing that Paul, Weiss and its forensic consulting firm receive "complete access - ---------- (30) September 2004 Agreement P VI.2.(a). (31) Letter from Armando Falcon, Jr. to Ann M. Korologos, dated Oct. 5, 2004. (32) September 2004 Agreement P VI.2.(a). (33) Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Oct. 19, 2004, FMSE 504466-85 at FMSE 504480-82. (34) Id. at FMSE 504481. (35) Plan of Fannie Mae for Implementing the September 27, 2004 Agreement between OFHEO and Fannie Mae, dated Oct. 2004, FMSE-IR 547280-307, at FMSE-IR 547304. 7
to all Fannie Mae personnel and records as they deem necessary during the course of the review."(36) In October 2004, Huron and Paul, Weiss jointly prepared a work plan for the various accounting and other topics consistent with the directives contained in the September 2004 Agreement. The work plan was submitted to, and approved by, both the SRC and OFHEO. 2. The February 2005 OFHEO Letter and Supplemental Agreement OFHEO's Special Examination continued after the issuance of the OFHEO Report. In a letter dated February 11, 2005, OFHEO notified Fannie Mae of preliminary conclusions concerning additional accounting issues that OFHEO had been reviewing as part of its Special Examination.(37) The letter stated that OFHEO had "identified polices that do not appear to be consistent with GAAP" under FAS 65, FAS 115, FAS 140, FAS 149, and FIN 46.(38) OFHEO also expressed concern over the Company's apparent development and application of accounting practices to smooth the recognition of certain income and expense amounts.(39) In addition to these accounting policies, OFHEO raised issues about the internal controls and/or limitations pertaining to Fannie Mae's journal entries, amortization systems, and database modifications as they relate to the amounts of amortization of deferred price adjustments recorded pursuant to FAS 91.(40) The Company announced on February 23, 2005 that the Board and Paul, Weiss would address the issues raised in the February 11, 2005 letter.(41) Pursuant to the February 11, 2005 letter, on March 7, 2005, OFHEO and Fannie Mae entered into the Supplemental Agreement.(42) The Supplemental Agreement called for a number of further internal reviews to be conducted, including reviews of: (1) the procedures relating to "the preparing, revising, validating, authorizing and recording of journal entries"; - ---------- (36) Mem. from Michael J. Williams to Senior Leadership Team, dated Oct. 12, 2004. (37) Letter from Christopher H. Dickerson to Stephen B. Ashley, dated Feb. 11, 2005, FMSE-IR 547318-30 (hereinafter "February 11, 2005 OFHEO Letter"). (38) Id. at FMSE-IR 547318-26. (39) Id. at FMSE-IR 547326-27. (40) Id. at FMSE-IR 547327-29. OFHEO also expressed concern over factor array manipulation. Id. at FMSE-IR 547329-30. (41) Fannie Mae press release, dated Feb. 23, 2005, available at http://www.sec. gov/Archives/edgar/data/310522/000095013305000679/w06038exv99w2.htm. (42) See supra note 3. 8
(2) Fannie Mae's "internal controls relating to modifications of databases supporting the general ledger"; (3) "Fannie Mae's legal and regulatory compliance structures," including recommended "changes to organizational structures, responsibilities and personnel"; and (4) Fannie Mae's current "bylaws and codes of conduct to assure that they support legal and regulatory compliance" and to suggest needed changes.(43) Furthermore, the Supplemental Agreement required the Company to conduct a reaudit and restatement for prior earnings periods, as necessary, and to review its accounting for several accounting standards in addition to FAS 91 and FAS 133, such as FAS 115, FAS 140, FAS 65, FAS 149, and FIN 46.(44) After the SRC requested that Paul, Weiss incorporate the additional accounting policy areas of concern raised in the February 2005 OFHEO letter, Huron and Paul, Weiss jointly prepared a supplemental work plan to address the new accounting areas, which was also submitted to the SRC and OFHEO for approval. C. Other Issues Brought to Paul, Weiss's Attention 1. Directive to Employees to Come Forward As set forth more fully in Chapter XI, Section A, over the course of Paul, Weiss's investigation, additional information was brought to Paul, Weiss's attention, including from anonymous sources and former employees. These additional issues, to the extent feasible, were added to Paul, Weiss's scope. On November 29, 2004, then General Counsel Ann M. Kappler sent an e-mail to all Fannie Mae employees containing the Chairperson of the SRC's instruction that "all employees who have any information or knowledge whatsoever about any unusual or atypical transactions in the past five years to provide Paul, Weiss" with information about such transactions directly.(45) Several employees contacted Paul, Weiss with information, some on the condition of anonymity. To the extent that sufficient information or questions were raised, we reviewed the issues raised and will address them later in this Report. 2. Monitoring Consultants's Activities As noted above, in addition to the tasks assigned to us in the September 2004 Agreement and the Supplemental Agreement, we were asked to monitor the efforts of the following consultants retained by Fannie Mae, Semler Brossy Consulting Group LLC ("Semler Brossy"), Mercer Oliver Wyman ("Mercer"), and Hildebrandt - ---------- (43) Supplemental Agreement PP II.A.1., II.A.4., II.C.1., II.C.5. (44) Id. P II.D.2. (45) E-mail from Ann M. Kappler to all Fannie Mae employees, dated Nov. 29, 2004. 9
International, Inc. ("Hildebrandt"). Semler Brossy, Mercer, and Hildebrandt were engaged directly by Fannie Mae, pursuant to the September 2004 Agreement, in order to provide prospective guidance on executive compensation, corporate governance, and regulatory compliance issues. (a) Semler Brossy The Compliance Committee of Fannie Mae's Board of Directors first engaged Semler Brossy in 2003 to consult on senior management compensation issues and the draft employment contracts for Franklin D. Raines, Daniel H. Mudd, and Timothy Howard. The Committee subsequently engaged Semler Brossy to address executive compensation issues in the September 2004 Agreement. Section IV of the September 2004 Agreement states, in relevant part: "The Board shall cause to be prepared a report to OFHEO on the compensation regime and its relation to strategic plans and their impact on accounting and transaction decisions and any revisions to avoid inappropriate incentives."(46) The Committee tasked Semler Brossy with assessing current executive compensation plans and developing future plans as required by Section IV of the September 2004 Agreement between OFHEO and Fannie Mae; the Committee did not ask Semler Brossy to include any forensic analysis of past executive compensation issues in its review. Semler Brossy provided its report to OFHEO on February 24, 2005, analyzing the metrics used by the Company to determine executive compensation levels and recommending certain changes to those metrics, as well as revisions to the overall executive compensation scheme. We understand that these recommendations may not be implemented until the restatement of earnings is complete, due to uncertainty regarding staffing changes and financial data that will affect the Company's executive compensation plans going forward. We have had a cooperative relationship with Semler Brossy during the course of our review. We met and conferred with Roger Brossy of Semler Brossy on several occasions, particularly as Semler Brossy was undertaking the engagement, developing a workplan, and preparing its final report. In addition, Brossy, as a representative of Semler Brossy, was available to us as a resource for information regarding particular executive compensation contracts and practices at Fannie Mae. Further, in light of Semler Brossy's engagement by Fannie Mae in 2003, Paul, Weiss interviewed Roger Brossy as part of our inquiry into past executive compensation practices. (b) Mercer The Compliance Committee also engaged Mercer in 2004 to work on organizational redesign pursuant to the September 2004 Agreement. Section I.1 of the September 2004 Agreement required that Fannie Mae separate "the function of business planning and forecasting from the controller's function" and "modeling and accounting - ---------- (46) September 2004 Agreement P IV. 10
functions" from each other.(47) Section II.B of the Supplemental Agreement also required the Board of Directors to make certain organizational and staffing changes.(48) To assist in these changes, Fannie Mae expanded Mercer's engagement. Accordingly, Mercer's work has been focused on (1) substantially reorganizing the Finance operation (including the Controller's Office), and (2) forming and building out the new Risk operation, followed by the new Compliance and Ethics operation. In each area, Mercer's work has involved structural recommendations, followed by staffing recommendations (levels, nature of expertise needed, position descriptions, etc.). Mercer has made recommendations to Fannie Mae regarding organizational design and continues to work with Fannie Mae on an advisory capacity. Throughout its work at Fannie Mae, Mercer has provided Paul, Weiss with informal status updates on a regular basis and has requested our review and comments on their preliminary workplans, findings, and recommendations from time to time. On our part, we asked questions, suggested additional areas for their analysis, and reviewed their tentative findings. (c) Hildebrandt Fannie Mae engaged Hildebrandt in response to Section II.C. of the Supplemental Agreement, which requires the Board to "cause to be conducted a review of Fannie Mae's legal and regulatory compliance structures."(49) Hildebrandt's assessment of the Company's legal and regulatory compliance operations has included (1) assessing past division and effectiveness of compliance operations within the Company, (2) recommendations for division of compliance responsibilities between, and compliance coordination among, the new Compliance and Ethics operation, the new Risk operation, the Legal Department, and Internal Audit (including the new Compliance and Ethics operation), and (3) recommendations for the role of the Legal Department going forward. Hildebrandt has sought our views on their tentative workplan and tentative findings and conclusions. As of the date of this writing, Hildebrandt had not completed its review, but is anticipated to do so within the next several months. As we have done with Mercer, we engaged in substantive interactions with Hildebrandt. Hildebrandt has been receptive to our comments and aided our understanding and analysis of regulatory issues. - ---------- (47) Id. P I.1.(a)-(d). (48) Supplemental Agreement P II.B. (49) Id. P II.C.1. 11
3. The Company's November 10, 2005 12b-25 Filing Finally, on November 10, 2005, Fannie Mae filed a Form 12b-25, Notification of Late Filing, with the SEC, in which it described, inter alia, "certain accounting matters that may significantly impact the results of operations for third quarter 2005 and the results of operations for third quarter 2004 that Fannie Mae ultimately reports."(50) Certain of the items had been identified as topics for review as part of the September 2004 Agreement, such as Fannie Mae's accounting practices under FAS 91 and FAS 133.(51) Other items were new. In particular, Fannie Mae announced that it had: - Failed to periodically assess for impairment guaranty fee assets resulting from "buy-up" payments made by Fannie Mae to lenders that adjusted the guaranty fee paid by the lender in order to adjust the pass-through interest rate on mortgage-backed securities;(52) - Incorrectly used the effective yield method instead of the equity method for, or consolidated investments in, Low Income Housing Tax Credit investment partnerships and three synthetic fuel investments;(53) and - Incorrectly gave insurance accounting treatment to a putative mortgage insurance transaction that Fannie Mae subsequently determined did not sufficiently transfer risk to warrant such treatment.(54) Paul, Weiss was asked by the "[t]he Board . . . to review the circumstances surrounding this [insurance] policy and . . . other insurance arrangements."(55) D) Engagement of Huron As noted above, the September 2004 Agreement specifically contemplated the retention of "an independent accounting firm" to conduct the reviews called for in the - ---------- (50) Fannie Mae, Notification of Late Filing (Form 12b-25), at 6 (Nov. 10, 2005) (hereinafter "Form 12b-25"), available at http://www.fanniemae.com/media/pdf/newsreleases/f12b25111005.pdf. (51) September 2004 Agreement P I.2-3. (52) Form 12b-25 at 7. (53) Id. (54) Id. at 8. (55) Id. 12
agreement.(56) Pursuant to the agreement, and with the approval of the SRC and OFHEO as required under the September 2004 Agreement, Paul, Weiss engaged Huron, a consulting firm specializing in the areas of forensic accounting and investigative services. The Huron team was led by George Massaro, the Vice Chairman of Huron, who has more than thirty-two years of audit and accounting experience, and Jeffrey Ellis, a Managing Director with more than eighteen years of public accounting experience, including several as a member of the national office accounting advisory groups of the firms with which he was associated, and has been a member of various working groups assisting the Emerging Issues Task Force of the FASB. Other senior members of the Huron team included: Paul Charnetzki, a Managing Director who has more than twenty-five years of experience as a consultant and expert in litigation matters involving, among other issues, the application of accounting and auditing principles; Joseph Pope, a Managing Director with fourteen years of experience providing financial and forensic accounting services to clients in litigation and complex business disputes and with industry specialization relating to financial markets, securities and derivatives; Tamika L. Tremaglio, a Managing Director with twelve years of experience in forensic accounting and investigations covering a broad spectrum of industries; and John Sullivan, a Managing Director with eighteen years of public accounting experience that included audit clients in a variety of industries. The scope of the Huron team's work ranged from the forensic collection of electronic data, including imaging hard drives of the computers used by individuals who were deemed to be of interest to our investigation, restoring backup tapes, examining documents collected by the Company from employees and assisting in witness interviews. E. Documents Reviewed Since our engagement on October 5, 2004, we have made over 400 document requests of the Company and its agents. As of the date of this writing, Fannie Mae had produced approximately 721,139 documents in response to our requests. During the course of the investigation to date, Paul, Weiss and Huron professionals reviewed over 2 million pages of hardcopy documents and approximately 2 million electronic documents. In addition, Paul, Weiss and Huron professionals conducted on-site inspection of various accounting and other systems and hardcopy documents made available by the Company. In addition to the documents produced in response to Paul, Weiss's requests, we reviewed a number of documents that the Company had produced in the course of the OFHEO Special Examination. As set forth more fully in Chapter X, prior to Paul, Weiss's engagement on October 5, 2004, Fannie Mae had produced approximately 427,467 documents to OFHEO in the course of the Special Examination. Fannie Mae also had produced 32,458 pages directly to the SEC. Since that date, Fannie - ---------- (56) September 2004 Agreement P VI.2.(a). 13
Mae has produced roughly 450,000 pages in response to OFHEO's and the SEC's requests. Paul, Weiss and Huron also reviewed all of these documents in both hardcopy and electronic format. We also continued to receive copies of all documents that Fannie Mae produced to OFHEO and to the SEC pursuant to the respective regulators' requests or subpoenas after the issuance of the OFHEO Report in September 2004. In addition to Fannie Mae's documents, we also reviewed documents from various Fannie Mae agents, including Wilmer and E&Y. Further, we requested and received access to information from KPMG, Fannie Mae's former auditor. Paul, Weiss and Huron professionals also inspected systems and documents directly at Fannie Mae, including HomeSite, the SOX database (which contains Sarbanes-Oxley documentation), Financial Dashboard, and ALEX. Fannie Mae also provided us with an Excel model that used historical PDAMS logic to run FAS 91-related calculations and historical input files from year-end 1998. We also conducted on-site examinations of files kept at Fannie Mae's Office of Corporate Justice ("OCJ") and others regarding selected topics. Paul, Weiss and Huron had access to all documents irrespective of any claim of privilege the Company may have over the materials.(57) F. Interviews Conducted As of the date of this writing, we conducted 241 interviews of 148 current and former employees, and third parties. Our interviews were generally organized by topic, which required multiple interviews of certain individuals. Because our investigation was not based on any judicial process, we lacked the power to compel the testimony of any witnesses who refused to appear, and therefore relied on the voluntary cooperation of witnesses. The interviews were not transcribed. A handful of employees declined our requests for an interview. Most significantly, Timothy Howard, the former Chief Financial Officer, Vice Chairman, and member of the Board of Directors of Fannie Mae, declined repeated requests for an interview. While his counsel in December 2005 made a submission focused on certain limited topics, we have not had the opportunity to inquire directly of Howard his bases for the various accounting decisions that were made under his leadership. - ---------- (57) As of this writing, we learned that the Company discovered in a warehouse approximately twenty-two computers that may have belonged to former or current executives of the Company, which computers had never been imaged or produced before to Paul, Weiss. While we have no reason to believe that the information contained in such computers will contain any new issues, we will nevertheless review these and other materials even after the issuance of this Report, and if any information is discovered that causes us to change or correct any findings in this Report, we will do so. 14
We interviewed Leanne G. Spencer, the former Senior Vice President and Controller of the Company, on four occasions. However, beginning in or about June 2005, Spencer declined further interviews following the late discovery in her files of certain documents that were highly relevant to the 1998 catch-up decision. Jeffrey Juliane, the former Director--Accounting and Audit, appeared for three interviews, but declined further interviews beginning in or about June 2005, when he separated from the Company. Several other former employees declined to be interviewed, including Joe Amato, Vice President--Portfolio Strategies, Debbie Cohen, Vice President--Portfolio Management, William E. Einstein, Vice President--Corporate Tax, Lan Nguyen, Director--Financial Reporting, and Ilan Sussan, Senior Project Manager--Financial Standards. G. Independent Collection and Analysis of Electronic Materials In addition to materials collected and produced by Fannie Mae in response to our requests, we conducted an independent forensic analysis and collection of electronic documents at Fannie Mae. This forensic analysis began with interviews of Fannie Mae information technology ("IT") staff responsible for maintaining and backing up Fannie Mae's information systems. Through these interviews, we identified potential sources of additional data that may be of interest to our review from the following sources: (a) backup tapes maintained by the Company as part of its normal IT operations for disaster recovery purposes, which contained both shared network spaces devoted to particular business units, network spaces of individual employees, and e-mail; (b) documents maintained by individuals on their individual hard drives, and in hardcopy format in file cabinets, etc.; (c) old hard drives and images of hard drives formerly used by Fannie Mae employees who were of interest to our investigation, especially old systems of Fannie Mae executives; and (d) other e-mails that the Company collected but deemed non-responsive and did not produce to OFHEO.(58) For each of the potential data sources, we carried out three tasks: (1) verified what the Company already had collected and produced from these sources; (2) analyzed whether additional relevant information was available from these sources; and (3) tested and collected any additional information that was available. - ---------- (58) We were notified on February 16, 2006 that the Company had located in a warehouse approximately twenty-two additional computers that belonged to former and current employees of the Company. The Company is imaging the hard drives of these computers and will provide the images to Paul, Weiss for review. At this time, we do not have any reasons to believe that there are new or additional information contained in these computers; however, we do intend to review these hard drives and if any new information is discovered in these computers that affect the findings or conclusions in this Report, we will supplement them accordingly. 15
We restored forty-seven backup "sessions" for twelve Novell network servers containing electronic documents organized by individual users and departments. Each "session" corresponds to a point-in-time snapshot of the data contained on a specific server. The Novell and ADSM backup sessions we restored were created between 2002 and 2004, but contain data stored on the server at the time of the backup. We restored five backup sessions of the Netscape e-mail server. The sessions were created between 1998 and 2004. As of the date of this writing, we have imaged 127 hard drives associated with forty non-attorney users and nine attorney users. We imaged an additional nineteen hard drives for which at the time of imaging no specific custodian information was available, but which appeared likely to contain relevant material. Our imaging process included extraction of user-created files in the recycle bins of each hard drive, as well as a review of the unallocated spaces - where deleted files are sometimes recoverable - of the hard drives of the users who may have information relevant to our review. We imaged sets of CDs and DVDs that contained backups of senior management hard drives, as maintained by the Company as part of its normal IT support practices.(59) We also received from Fannie Mae copies of five forensic images of hard drives used by members of the Controller's Office which were created at the request of Fannie Mae's Legal Department in connection with the investigation into Barnes's allegations in 2003. As part of our hard drive imaging program, we also copied images of the hard drives of executive users that were made by the Executive Support Team, which provides technical assistance to senior management, as a regular part of PC maintenance. We also imaged the old systems of executives that Fannie Mae maintained rather than donated to charities pursuant to a company-wide policy of donating computers that had been in use for three or more years and were therefore out of warranty coverage. Additionally, we imaged those hard drives that we believed were associated with executives as well as "ghost images," and similar data created by the Executive Support Team in connection with servicing senior executives' computers. Apart from Fannie Mae executives, our efforts to locate and image other employees' former systems were less successful. Despite our efforts and those of Fannie Mae's IT staff, we were unable to locate approximately two systems that we had identified as being issued previously to Fannie Mae employees, which hard drives may contain information relevant to our investigation. Nonetheless, we were successful in imaging other systems that had been used by each of those users. Additionally, we were unable to image one computer because it was corrupted, but since then we made an image of the data Unisys recovered from it at Fannie Mae's behest. - ---------- (59) A backup, which captures live files, will generally contain less information than an image, which will also capture any information in the unallocated space of a hard drive. 16
We also were unable to image three hard drives that were donated. In particular, a hard drive used by Spencer was donated to Phi Psi on December 15, 2004, a hard drive used by Richard Stawarz was donated to Vanessa Ali Ministries on December 3, 2004, and a hard drive used by Mona Patel was donated to Children's Health Project in November 2004. Additionally, a hard drive used by a Treasurer's Office employee was salvaged on February 25, 2004. Although the Company does not have any documentation regarding why this computer needed to be salvaged, we have been told that it was Fannie Mae's practice to salvage computers that were non-functional. Finally, our review also included materials previously determined by Company counsel in the course of the Special Examination as non-responsive. We obtained and reviewed over 600,000 electronic items in this regard, which had been reviewed during the course of responding to the Special Examination but determined to be non-responsive to OFHEO's document requests. Our review revealed that many of these documents were of interest to the regulators, and we produced the relevant documents. H. Assessment as to the Company's Cooperation 1. The Board, the SRC and Management The SRC and the Board of Directors of Fannie Mae cooperated fully in our investigation. The SRC, in particular, not only directed management and all employees to provide full cooperation to the investigation, but instructed all employees to come forward with any information the employee may have concerning any questionable or irregular transactions, among other things. As set forth more fully in Chapter XI, Section A, this directive from the SRC resulted in several employees coming forward with information. In addition, the SRC instructed the Company and its outside counsel to provide to Paul, Weiss and Huron full access to all information requested.(60) As for cooperation from management, in late 2004, management appeared to view Paul, Weiss's role as that of an adversary, rather than as the SRC's counsel. We experienced some initial difficulties and delays in obtaining access to all relevant information. For example, management initially declined to produce information that was responsive to Paul, Weiss's request or sought to disclose it only on the condition that Paul, Weiss would not share it further. Moreover, in the beginning of our engagement, management through counsel continued to make the best "arguments" in defense of accounting practices criticized by OFHEO in its Report, rather than disclosing all information in its possession concerning the accounting practices at issue. Finally, on one occasion, we learned that an employee's counsel had access through the Company's attorneys to important documents that had not been provided to Paul, Weiss, despite the fact that the documents were responsive to pending Paul, Weiss requests. While the - ---------- (60) Mem. from M. Williams to Senior Leadership Team, dated Oct. 12, 2004. 17
relevant documents were promptly produced once the Company was notified of this situation, we were concerned in the beginning about the level of cooperation being provided by then management. Beginning in late 2004 and early 2005, with the creation of a new team within Fannie Mae that was directed to respond directly to Paul, Weiss's requests, we began to receive full cooperation and full access to information requested. The new team reported not to the Legal Department or incumbent management but directly to Michael J. Williams, then Senior Vice President--e-Business, who had been instructed by the Board to oversee the implementation of the September 2004 Agreement. 2. The Leanne Spencer Document Incident The only exception to the generally good cooperation we received from management and employees throughout 2005 was Leanne Spencer. In May 2005, we became aware of a document in Spencer's files that provided highly-relevant information concerning management's motive and intent behind the decision to record only part of the catch-up expense in 1998 ("the Spencer Document").(61) The Spencer Document had not been produced voluntarily by Spencer, even though it was responsive to Paul, Weiss's document requests to Fannie Mae dating back to October 2004. Also around this time, we learned that, in connection with her office move in January 2005, Spencer brought home two boxes of documents that were never reviewed for responsiveness to the regulators' or Paul, Weiss's requests, and that contained responsive documents. The Company discovered the Spencer Document on April 27, 2005. The Spencer Document was discovered in connection with Fannie Mae's efforts, begun in March of 2005, to ensure that all responsive materials from Spencer were collected and reviewed. The Company began this effort after discovering three boxes from Spencer's former office after she moved into a new office that had not been collected or reviewed previously, and that contained certain responsive materials. As a result, the individual in Michael Williams's group who was overseeing the Company's response to Paul, Weiss requests, sent a team to go into Spencer's new office without warning while Spencer was not present, and pack up all documents and photocopied them. The Spencer Document was discovered in the course of reviewing these documents. Unquestionably, a number of different attorneys from Wilmer and Fannie Mae's Legal Department, as well as people working on responding to Paul, Weiss's requests, had met with Spencer on several occasion to collect responsive documents, including those documents relating to the 1998 event. Spencer herself was interviewed several times by OFHEO, by Paul, Weiss, and by the Company lawyers about 1998 - ---------- (61) Risk Review with CFO, dated Sept. 22, 1998, FMSE-IR 321564-68. 18
events, and therefore plainly understood the relevance of the Spencer Document to the 1998 events.(62) Nevertheless, Spencer did not voluntarily produce these documents. In early January 2006, Spencer's counsel submitted a letter describing the circumstances behind Spencer's failure to identify the Spencer Document to Company counsel.(63) According to the submission, the Spencer Document was "not identified earlier only because of the procedures employed during the company's document collection process."(64) But, instead of identifying what it was about the process that led to her failure to produce this document, Spencer stated only that she apparently did not identify the document as responsive to OFHEO's requests because the label of the file in which the document was kept was named "Buy Up, Buy Down," as opposed to "FAS 91," and the document itself did not reference "a meeting in 1998."(65) Spencer's submission did not address why she did not produce the document in response to Paul, Weiss's requests or whether she had reviewed the files for responsive documents in response to Paul, Weiss's requests. Given that the Spencer Document flatly contradicted Spencer's previous assertions to OFHEO and to Paul, Weiss that the decision to record only $240 million in catch-up expense in 1998 was based on her belief that it was the best estimate and that the decision had nothing to do with EPS, it is difficult to believe that Spencer's failure to produce the document was a mere oversight. Indeed, based on the statements made by Spencer in her Paul, Weiss interviews before the document was discovered, we think that it is entirely possible that Spencer had reviewed the document as late as January 2005 in preparation for the Paul, Weiss interview, but never produced the document.(66) With respect to the documents discovered at Spencer's home, it is unclear whether anyone from Wilmer or the Legal Department inquired of Spencer as to whether Spencer might have responsive documents at her home. However, we have learned that Spencer took documents home after the Special Examination began, and therefore Spencer clearly - ---------- (62) Indeed, in October 2004 alone, in the aftermath of the OFHEO Report and in connection with preparing Franklin D. Raines's and Timothy Howard's congressional testimony, the Company's in-house and outside lawyers met with Leanne G. Spencer at least three times to go over the 1998 catch-up and to ask her for any relevant documents. (63) Letter from David S. Krakoff to Senator Warren B. Rudman, dated Jan. 5, 2006. (64) Id. at 3. (65) Id. (66) Spencer described the attendees and agenda of the September 22, 1998 "Risk Review with CFO" with precision at her January 13, 2005 interview with Paul, Weiss, yet did not describe any documents relating to the meeting; Spencer also conveniently omitted any discussions about the focus on EPS management that was reflected in the agenda for the meeting. 19
knew that even documents at home may be responsive and should have produced them earlier. 20
CHAPTER III: BACKGROUND I. INTRODUCTION This Chapter provides a brief overview of Fannie Mae's history, its role today, its primary areas of business, and the regulatory regime under which it operates. II. THE EVOLUTION OF FANNIE MAE The Federal Housing Administration ("FHA") created the Federal National Mortgage Association ("Fannie Mae") in 1938(67) to purchase and sell mortgages in order to increase the availability of funds to mortgage borrowers.(68) Fannie Mae was to borrow money for such purposes through the issuance of notes, bonds, debentures, and other obligations.(69) While Fannie Mae was initially authorized only to purchase FHA-insured mortgages,(70) Congress approved a mortgage insurance program in 1944 that enabled Fannie Mae to purchase loans guaranteed by the Veterans Administration ("VA") as well.(71) In 1954, Congress rechartered Fannie Mae to expand its functions (as amended from time to time, the "Charter Act").(72) Fannie Mae at that time was intended - ---------- (67) Understanding Fannie Mae: Our History, http://www.fanniemae.com/aboutfm/understanding/index.jhtml?p=About+Fannie+ Mae&s=Understanding+Fannie+Mae (hereinafter "Fannie Mae History") (last visited Feb. 17, 2006). The FHA created the "National Mortgage Association of Washington" on February 10, 1938 (changing its name later that same year to the "Federal National Mortgage Association"), acting under Title III of the National Housing Act, ch. 847, 48 Stat. 1246, 1252 (1934). (68) National Housing Act, ch. 847, 48 Stat. 1246, 1252 (1934). (69) Id. (70) See Gen. Accounting Office Report to the Chairman of the Subcomm. on Capital Markets, Insurance, and Government Sponsored Enterprises of the H. Comm. on Financial Servs. on the Federal Housing Enterprises: HUD's Mission Oversight Needs to Be Strengthened 2 n.6, dated July 1998 (hereinafter "July 1998 GAO Report"), available at http://www.gao.gov/archive/1998/gg98173.pdf. (71) See SERVICEMEN'S READJUSTMENT ACT OF 1944, ch. 268, Section 501, 58 Stat. 284, 292 (repealed and reenacted by Pub. L. No. 85-857, 72 Stat. 1105 (1958)). (72) The "Federal National Mortgage Association Charter Act" was enacted as part of the Housing Act of 1954, Pub. L. No. 560, Section 301, 68 Stat. 590, 612 (codified as amended at 12 U.S.C. Section 1716 (2000)). 21
to become a secondary market facility for home mortgages financed to the maximum degree feasible by private capital.(73) The 1954 Charter Act also initiated the transition of Fannie Mae from a government-owned entity to private company. Under the Act, the government no longer backed any debt that Fannie Mae used to purchase mortgages, but continued to provide protection in other forms, including certain tax benefits and the support of the Federal Reserve Banks.(74) The Charter Act also established a means of replacing government ownership with private stockholders: Fannie Mae would gradually sell common stock and use the proceeds to retire Treasury-owned preferred stock.(75) Fannie Mae retired the last of its government stock on September 30, 1968.(76) In 1965, the Department of Housing and Urban Development ("HUD") was created as a cabinet-level agency to administer U.S. government programs that provide housing assistance and support for the development of American communities.(77) Three years later, Congress assigned HUD the task of serving as Fannie Mae's regulator.(78) The government also amended the Charter Act in 1968, further transitioning Fannie Mae away from the government sector.(79) The amended Act carved out special assistance programs from Fannie Mae's mandate and made them the responsibility of a new government enterprise, the Governmental National Mortgage Association ("Ginnie Mae").(80) Fannie Mae retained responsibility for the secondary - ---------- (73) See id. (74) See Charter Act, Pub. L. No. 560, Section 309, 68 Stat. 590, 621 (1954). (75) Understanding Fannie Mae: Our Charter, http://www.fanniemae.com/aboutfm/understanding/charter.jhtml?p=About+ Fannie+Mae&s=Understanding+Fannie+Mae&t=Our+Charter (last visited Feb. 17, 2006). (76) See id. (77) See Department of Housing and Urban Development Act of 1965, Pub. L. No. 89-174, 79 Stat. 667 (codified as amended at 42 U.S.C. Section 3531 (2000)). (78) See July 1998 GAO Report at 3. The subsequent creation of the Office of Federal Housing Enterprise Oversight ("OFHEO") to supervise the safety and soundness of Fannie Mae and Freddie Mac is discussed in Subsection V. below. (79) See generally Housing and Urban Development Act of 1968, Pub. L. No. 90-448 Sections 802-06, 82 Stat. 476, 536-44 (codified as amended at 12 U.S.C. Section 1716(b) (2000)) (hereinafter "HUD Act of 1968"). (80) See id. Section 801, 82 Stat. at 536. Ginnie Mae is an instrumentality of the United States contained within HUD. The partitioning of Fannie Mae into the two separate bodies 22
mortgage market, and was designated a "[g]overnment-sponsored private corporation."(81) In addition, the amended Act gave Fannie Mae the authority to issue mortgage-backed securities ("MBS").(82) In 1970, Congress authorized Fannie Mae to buy and sell conventional mortgages.(83) That same year, Congress created another entity, the Federal Home Loan Mortgage Corporation ("Freddie Mac"), to increase liquidity in the mortgage market through the purchase of conventional mortgages.(84) While HUD remained responsible for overseeing Fannie Mae, the Federal Home Loan Bank Board regulated the newly created Freddie Mac.(85) Fannie Mae first listed its stock on the New York Stock Exchange ("NYSE") on August 31, 1970.(86) By 1976, Fannie Mae was purchasing more conventional mortgages than FHA and VA loans.(87) In the early 1980s, Fannie Mae continued a shift away from FHA and VA loans and began to purchase adjustable-rate mortgages and second mortgages.(88) In addition, Fannie Mae introduced its MBS - ---------- became effective on September 1, 1968. See Charter Act, 12 U.S.C. Section 1717(a)(2) (2000). (81) HUD Act of 1968 Section 801, 82 Stat. at 536. Fannie Mae and Freddie Mac are often now referred to as government-sponsored enterprises or "GSEs." See 12 U.S.C. Section 4501 (2000). (82) See HUD Act of 1968 Section 804(a), 82 Stat. at 542. In particular, the statute authorized Fannie Mae to issue debt obligations or trust certificates for beneficial interests in pools of mortgages, and could guarantee the timely payment of principal and interest. See 12 U.S.C. Section 1719(d) (2000). (83) See Emergency Home Finance Act of 1970, Pub. L. No. 91-351 Section 201, 84 Stat. 450-51. The legislation required a one-time approval by HUD before Fannie Mae could initiate its conventional mortgage program. Conventional mortgages are defined as mortgages that are not insured or guaranteed by the FHA or VA. See id. (84) See Federal Home Loan Mortgage Corporation Act, Pub. L. No. 91-351 Sections 301-10, 84 Stat. 451-58 (1970) (hereinafter the "Corporation Act"). (85) See id. Section 303, 84 Stat. at 452. (86) New York Stock Exchange, Fannie Mae, http://www.nyse.com/about/listed/fnm.html (last visited Feb. 17, 2006). (87) Fannie Mae History, supra note 67. (88) See Secondary Mortgage Market Enhancement Act of 1984, Pub. L. No. 98-440, Section 203, 98 Stat. 1689, 1693 (authorizing Fannie Mae to buy and sell subordinate lien mortgages). 23
business in 1981.(89) In 1992, Fannie Mae surpassed Ginnie Mae and Freddie Mac as the largest issuer and guarantor of MBS.(90) III. FANNIE MAE'S ROLE TODAY Fannie Mae helps maintain liquidity in the primary mortgage market through the purchase of primary mortgages from entities such as mortgage companies, savings and loan associations, commercial banks, and state and local housing finance agencies.(91) Fannie Mae currently represents the largest source of funds to mortgage lenders in the nation.(92) In 2004, it purchased or guaranteed mortgages valued at $725 billion, and issued new MBS valued at $552 billion.(93) While Fannie Mae serves as both a broad purchaser of mortgages and a creator of MBS products, the Company also encourages equal housing access through programs targeted at particular populations. In its 1994 Trillion Dollar Commitment, Fannie Mae committed to provide $1 trillion in financing to ten million traditionally underserved families.(94) After achieving its goal in 1999, Fannie Mae launched its - ---------- (89) Fannie Mae History, supra note 67. (90) Fannie Mae History, supra note 67. (91) See Jason T. Strickland, The Proposed Regulatory Changes to Fannie Mae and Freddie Mac: An Analysis, 8 N.C. BANKING INST. 267, 270-72 (2004) (providing, in part, a history of Fannie Mae and outlining the functions and importance of the enterprise to the mortgage market); see also Fannie Mae, 2003 Annual Report (Form 10-K), at 4 (Mar. 15, 2004) (hereinafter "2003 Form 10-K"), available at http://www.fanniemae.com/ir/pdf/sec/2004/f10k03152004.pdf. (92) 2003 Form 10-K at 1. (93) 2005 OFHEO Report to Congress, dated June 15, 2005, at 2 (hereinafter "2005 OFHEO Report to Congress"), available at http://www.ofheo.gov/media/pdf/2005reporttocongress.pdf. As a comparison, Freddie Mac in 2004 purchased or guaranteed mortgages valued at $495 billion, and issued new MBS valued at $365 billion. Id. (94) 2003 Form 10-K at 1; see also Regulations of Housing Finance Industry: Hearing Before the Subcomm. on Capital Markets, Securities, and Government Sponsored Enterprises of the H. Comm. on Financial Servs., 106th Cong. 10 (2000) (statement of Franklin D. Raines, Chairman and CEO, Fannie Mae) (testifying that the Trillion Dollar Commitment goals were met eight months early) (hereinafter "Raines May 2000 Congressional Testimony"), available at http://www.fanniemae.com/global/pdf/media/issues/archive/2000/051600.pdf. 24
American Dream Commitment in 2000, through which it planned to provide $2 trillion to eighteen million underserved households by 2010.(95) The American Dream Commitment makes funds available to underserved Americans through lender and community partnerships established by the Company, including collaborative relationships with the National Association of Home Builders, the National Urban League, and the National Association of Hispanic Real Estate Professionals.(96) In January 2004, Fannie Mae expanded its American Dream Commitment by pledging to help six million families become first-time homeowners within a decade, including 1.8 million minority families.(97) IV. FANNIE MAE'S PRIMARY BUSINESS ACTIVITIES Fannie Mae's business falls primarily into two segments: the Portfolio Investment business and the Credit Guaranty business.(98) A. Portfolio Investment The Portfolio Investment business engages in essentially two activities: mortgage portfolio investments and liquid investments. The mortgage investment portfolio consists largely of conventional, single-family fixed- or adjustable-rate first lien mortgage loans, as well as mortgage-related securities acquired from lenders, securities dealers, and investors.(99) Fannie Mae's liquid investments consist largely of short-term, non-mortgage assets, which can be used to generate cash to meet Fannie Mae's liquidity needs.(100) Fannie Mae funds the purchase of its mortgage-related and liquid assets from equity capital and through the sale of debt securities to investors worldwide.(101) Fannie - ---------- (95) 2003 Form 10-K at 1; see also Raines May 2000 Congressional Testimony at 10 (testifying that the American Dream Commitment would prioritize "increasing homeownership among minorities, young families, women-headed families, new immigrants, and others whose homeownership rates lag the general population"). (96) See Fannie Mae's Commitment to Minority Homeownership, http://www.fanniemae.com/initiatives/minority/index.jhtml?p=Initiatives&s= Minority+Homeownership (last visited Feb. 17, 2006). (97) Fannie Mae press release, dated Jan. 27, 2004, http://www.fanniemae.com/initiatives/adc/expansion.jhtml?p=Initiatives&s= Expanding+the+American+Dream+Commitment (last visited Feb. 17, 2006). (98) 2003 Form 10-K at 2. (99) Id. at 9. (100) Id. at 2, 7. Fannie Mae has "set a goal to maintain liquid assets equal to at least [five] percent of total on-balance sheet assets." Id. at 10. (101) Id. at 2. 25
Mae derives income from the difference between the yield it earns on portfolio investments and the interest it pays on Company debt.(102) B. Credit Guaranty In its Credit Guaranty business, Fannie Mae pools mortgages originated by lenders and returns MBS to those lenders.(103) With respect to a majority of Fannie Mae's MBS, the process begins when a primary mortgage lender transfers a pool of similar loans (e.g., similar interest rate or maturity features) to a trust that assumes legal ownership,(104) with Fannie Mae as trustee.(105) The trust certificate holders (usually investment banks, commercial banks, thrifts, and funds) receive monthly distributions funded by the principal and interest payments on the underlying loans.(106) The Credit Guaranty business obtains its primary stream of income from charging fees for the risks it assumes and the services it provides. For example, the Company guarantees the payment of principal and interest due on the mortgages underlying MBS, assuring the MBS holders of steady income (aside from prepayment risk) and no risk of borrower default.(107) In return, Fannie Mae charges a guaranty fee based on the credit risk it assumes as well as the costs of administering the MBS.(108) Fannie Mae receives monthly payments of the guaranty fee by retaining a portion of the interest payments received on the underlying mortgage loans.(109) - ---------- (102) Id. at 7. (103) Id. at 11. The original lenders can hold the MBS for investment or sell them to other investors. Id. at 2. (104) Id.; see also Task Force on Mortgage-Backed Securities Disclosure Staff Report on Enhancing Disclosure in the Mortgage-Backed Securities Markets, dated Jan. 2003 (providing a basic description of the MBS pass-through security and REMIC structures), available at http://www.ofheo.gov/Media/Archive/docs/press/mbsdisclosure.pdf. (105) 2003 Form 10-K at 11. (106) Id.; see also May 2003 Cong. Budget Office Report 3, available at http://www.cbo.gov/ftpdocs/41xx/doc4199/05-06-03-GSEs.pdf. (107) 2003 Form 10-K at 13. (108) 2003 Form 10-K at 2; see also May 2001 Cong. Budget Office Report 9, available at http://www.cbo.gov/ftpdocs/28xx/doc2841/GSEs.pdf. (109) 2003 Form 10-K at 11. 26
V. REGULATORY REGIME In 1989, Congress abandoned the separate regulatory regimes for Fannie Mae and Freddie Mac and brought Freddie Mac under the regulatory authority of HUD.(110) Following a Congressionally mandated study conducted by the Comptroller General,(111) Congress altered the regulatory scheme again, creating a new agency to oversee the safety and soundness of both enterprises.(112) Specifically, the Federal Housing Enterprises Financial Safety and Soundness Act of 1992 ("Safety and Soundness Act") amended the Charter Act to create the Office of Federal Housing Enterprise Oversight ("OFHEO"), an agency within HUD, to "ensure that [Fannie Mae and Freddie Mac] are adequately capitalized and operating safely."(113) In addition to granting safety and soundness oversight powers to OFHEO, this legislation modified the oversight responsibilities retained by HUD with respect to the GSEs' housing-related missions. Among other things, the Safety and Soundness Act clarified the procedures that HUD must follow when reviewing and approving new mortgage programs to be offered by Fannie Mae and Freddie Mac.(114) In 2003, Fannie Mae voluntarily registered its common stock with the Securities and Exchange Commission ("SEC") under Section 12(g) of the Securities Exchange Act of 1934, bringing Fannie Mae under SEC regulatory and enforcement authority for issues pertaining to accounting, financial disclosures, and corporate governance.(115) - ---------- (110) Financial Institutions Reform, Recovery, and Enforcement Act of 1989, Pub. L. No. 101-73, Section 731, 103 Stat. 183, 431 (codified as amended 12 U.S.C. Section 1811 (2000)). (111) Id. Section 1004, 103 Stat. at 509-11 (mandating a study on GSE risk exposure); see also GAO Report on Government-Sponsored Enterprises: The Government's Exposure to Risks, dated Aug. 1990, available at http://archive.gao.gov/d23t8/142017.pdf. (112) Federal Housing Enterprises Financial Safety and Soundness Act of 1992, Pub. L. No. 102-550, Section 1311, 106 Stat. 3941, 3944 (codified as amended 12 U.S.C. Section 4501 (2000)). (113) See id. Section 1313, 106 Stat. at 3945. (114) See id. Section 1322, 106 Stat. at 3953-54. (115) Fannie Mae, Registration of Securities (Form 10-12G) (Mar. 31, 2003) (hereinafter "Form 10-12G"), available at http://www.fanniemae.com/ir/sec/index.jhtml?s=SEC+Filings (follow "All SEC Filings" hyperlink; then follow "Registration Statements" hyperlink for PDF format); see also Standard & Poor's Corporate Governance Score: Fannie Mae, dated Jan. 30, 2003, at 2-3, available at http://www2.standardandpoors.com/spf/pdf/products/Fannie%20MaeFINAL1.pdf. 27
A. HUD Under the Safety and Soundness Act, HUD has the authority to ensure that Fannie Mae is accomplishing its mission of expanding housing opportunities in traditionally underserved communities.(116) Except for matters falling under the purview of OFHEO, HUD has general regulatory power over Fannie Mae and is directed by Congress to make rules and regulations that accomplish the goals of the Charter Act and the Corporation Act.(117) As noted, Fannie Mae must obtain HUD approval prior to the introduction of any new mortgage program.(118) In addition, HUD enforces fair housing rules by administrating anti-discrimination regulations and requiring that each enterprise investigate whether mortgage lenders comply with the Fair Housing Act and the Equal Credit Opportunity Act.(119) Finally, the Safety and Soundness Act requires HUD to set specific housing goals for Fannie Mae each year.(120) The Company must dedicate a specified percentage of new business to fulfilling these goals.(121) By statute, HUD must submit an annual report to Congress detailing the extent to which each GSE is achieving its annual housing goals and fulfilling its statutory mission.(122) The most recent HUD report, issued in January 2005, concluded that Fannie Mae met its housing goals each year from 1996 through 2003.(123) - ---------- (116) See 12 U.S.C. Section 4541 (2000). (117) See id. (118) See id. Section 4542(a). (119) See id. Section 4545. (120) See id. Section 4561(a). The three annual housing goals consist of (1) a low- and moderate-income housing goal, (2) a special affordable housing goal, and (3) an underserved areas housing goal. See id. Sections 4562-4564. (121) See Sections 4562-4564. (122) See 12 U.S.C. Section 4544 (2000). (123) See HUD Report Overview of the GSE's Housing Goal Performance, 1996-2003, dated Jan. 2005, available at http://www.huduser.org/Datasets/GSE/gse2003.pdf. 28
B. OFHEO As noted above, OFHEO functions as an independent agency within HUD. Its mission is to ensure that Fannie Mae and Freddie Mac are adequately capitalized and operate safely and soundly in compliance with applicable laws, rules, and regulations.(124) OFHEO conducts an annual examination to assess Fannie Mae's financial safety and soundness.(125) In 2004, OFHEO's annual examination found weaknesses in management, documentation of policies, internal controls, and information flow. Among other things, OFHEO found deficiencies in the standardization of written policies, information flow to the Board, the operations of the Controller's Department, operational risk management, the functioning of the Office of Auditing, and liquidity management.(126) OFHEO's 2005 Annual Report to Congress noted, however, that Fannie Mae has devoted significant resources to correcting the deficiencies identified in the 2004 annual examination and in OFHEO's Special Examination.(127) The following paragraphs outline some of the primary areas in which OFHEO regulates Fannie Mae's activities. 1. Capital Requirements OFHEO makes quarterly findings on the adequacy of Fannie Mae's capitalization.(128) OFHEO classifies the Company as adequately capitalized, undercapitalized, significantly undercapitalized, or critically undercapitalized.(129) Fannie Mae must meet both a minimum capital standard and a risk-based capital standard to be classified as adequately capitalized.(130) In December 2004, OFHEO announced that it had classified Fannie Mae as significantly undercapitalized as of September 30, 2004.(131) In May 2005, OFHEO - ---------- (124) 12 C.F.R. Section 1700.1(a) (2005). (125) See, e.g., 2005 OFHEO Report to Congress at 10. (126) Id. at 10-11. (127) Id. at 12-13. (128) About OFHEO, http://www.ofheo.gov/Mission.asp (last visited Feb. 17, 2006). (129) OFHEO press release, dated Sept. 28, 2005, at 1, available at http://www.ofheo.gov/media/pdf/capclass92805.pdf. (130) Id. (131) OFHEO press release, dated Dec. 21, 2004, at 1, available at http://www.ofheo.gov/media/pdf/capclass122104.pdf. 29
announced that, as a consequence of an increased capital surplus that was "sufficient to absorb uncertainties in the estimated impact to capital of the accounting errors," it found Fannie Mae to be adequately capitalized as of March 31, 2005.(132) More recently, OFHEO announced in September 2005 that, as of June 30, 2005, Fannie Mae remained adequately capitalized, due once again to a capital surplus that was sufficient to absorb uncertainties in the estimated impact of accounting errors on the Company's capital.(133) In addition, OFHEO announced in early November 2005 that Fannie Mae had achieved, by an imposed September 30, 2005 deadline, a thirty percent capital surplus required by the Company's September 2004 and March 2005 agreements with OFHEO.(134) 2. Corporate Governance In June 2002, OFHEO issued corporate governance regulations applicable to Fannie Mae and Freddie Mac.(135) The regulations set forth a number of requirements focused on the board of directors and board committees. Among other things, these regulations directed Fannie Mae and Freddie Mac to select and follow the corporate governance practices and procedures of a particular body of law, so long as consistent with federal law and the safe and sound operation of the enterprise.(136) In addition, the regulations required each enterprise to incorporate in its bylaws both an audit committee and a compensation committee of the board of directors. The regulation provided that the Company's audit committee had to comply with the audit committee rules of the NYSE, while the compensation committee - ---------- (132) Id. (133) OFHEO press release, dated Sept. 28, 2005, at 2, available at http://www.ofheo.gov/media/pdf/capclass92805.pdf. (134) OFHEO press release, dated Nov. 1, 2005, available at http://www.ofheo.gov/News.asp?FormMode=Releases&ID=250 (last visited Feb. 17, 2006). The September 2004 agreement and the March 2005 supplemental agreement are discussed in Chapter II. (135) See 67 Fed. Reg. 38,361, 38,370-71 (June 4, 2002) (codified at 12 C.F.R. Sections 1710.1-1710.20 (2002)). While the chartering acts of Fannie Mae and Freddie Mac contained some corporate governance provisions, they lacked others commonly associated with state-chartered corporations. When issuing these regulations, OFHEO noted the increased focus on sound corporate governance practices by regulators, corporations, investors, and stock exchanges. Id. at 38,362. (136) 12 C.F.R. Section 1710.10 (2002). Fannie Mae elected to follow Delaware law. See Fannie Mae Bylaws, art. I, Section 1.05, available at http://www.fanniemae.com/governance/bylaws/article1.jhtml?p=Corporate+ Governance&s=Bylaws&t=Article+1:+General+Provisions (last visited Feb. 17, 2006). 30
had to include three or more independent board members.(137) The regulations also set forth certain minimum standards for board conduct and responsibilities, including a requirement to stay reasonably informed of the relevant enterprise's condition, activities, and operations.(138) In April 2004, OFHEO issued proposed amendments to its corporate governance regulations.(139) After a comment period, OFHEO issued final amendments in April 2005, effective June 6, 2005.(140) The amended regulations impose board member term limits, require the majority of seated board members to be independent as defined under NYSE rules, and mandate board meetings no less than eight times a year (with at least one meeting per calendar quarter).(141) The board of directors is tasked with oversight of legal and regulatory compliance, as well as corporate performance.(142) Furthermore, the audit, compensation, and nominating/corporate governance committees are required to comply with new Sarbanes-Oxley Act requirements and the NYSE rules.(143) Aside from board-directed changes, the 2005 amendments require rotation of the outside audit firm's lead or coordinating audit partner at least every five years.(144) In addition, Fannie Mae must maintain both a compliance program designed to ensure that it follows all applicable laws, regulations, and internal controls, and a risk management program designed to manage operational risk; the head of each program must report directly to the board of directors or an appropriate board committee.(145) The amended regulations also require each enterprise to draft and administer a code of - ---------- (137) 12 C.F.R. Section 1710.12 (2002). (138) See id. Section 1710.15. (139) 69 Fed. Reg. 19,126 (proposed Apr. 12, 2004). (140) 70 Fed. Reg. at 17,310-12 (codified as 12 C.F.R. Sections 1710.1-.30 (2005)). The final version of the amendments did not include several of the proposed provisions. For example, a provision to prohibit a GSE chairman of the board from also serving as CEO was excluded after Fannie Mae and Freddie Mac each separately agreed with OFHEO to divide the positions. 70 Fed. Reg. at 17,305. (141) 12 C.F.R. Section 1710.11 (2005). (142) See id. Section 1710.15. (143) See id. Section 1710.12. (144) See id. Section 1710.18. (145) See id. Section 1710.19. 31
conduct and ethics that includes the standards required under Section 406 of the Sarbanes-Oxley Act.(146) - ---------- (146) See id. Section 1710.14. 32
CHAPTER IV: ACCOUNTING FOR THE AMORTIZATION OF PREMIUM/DISCOUNT ON LOANS AND MORTGAGE-BACKED SECURITIES I. INTRODUCTION We address in this Chapter Fannie Mae's accounting for the amortization of premium and discount on loans and other securities pursuant to Statement of Financial Accounting Standards No. 91, Accounting for Nonrefundable Fees and Costs Associated with Originating or Acquiring Loans and Initial Direct Costs of Leases ("FAS 91" or the "Statement"). In particular, we focused on two areas: (1) Fannie Mae's decision to book $240 million in net unamortized premium expense in the fourth quarter of 1998 and to defer the recognition of approximately $199 million in net unamortized premium expense that had been calculated through the use of the Company's then existing systems; and (2) management's creation and implementation of a purchase premium and discount amortization policy (the "Amortization Policy" or the "Policy") in 2000 that included, among other things, a "precision threshold" within which management retained substantial discretion to not make adjustments that were required under FAS 91. Based on all of the facts gathered, we have formed the following conclusions about management's 1998 catch-up decision: - The decision to record only $240 million of the $439 million in calculated catch-up expense was motivated by a desire to meet earnings per share ("EPS") and Fannie Mae's Annual Incentive Plan ("AIP") targets and was not consistent with GAAP; - Management made other adjustments for 1998 to compensate for the EPS shortfall created by recording $240 million in catch-up expense: first, management accelerated a planned accounting method change for recording low income housing tax credits, which resulted in the recognition of two years' worth of tax credits in 1998; second, management reversed $3.9 million of credit balances from an account carrying aged suspense items into income as "miscellaneous income"; - KPMG was aware of both the full catch-up expense, the decision to record only $240 million in expense, and the other adjustments and proposed an audit difference for the unrecorded catch-up calculation; however, KPMG did not qualify its opinion on the Company's financial statements; and - Management's disclosure concerning both the catch-up and the other adjustments to the Fannie Mae Board of Directors, and its disclosure regarding the other adjustments to the public, was incomplete and misleading. 33
With respect to the development and implementation of the Amortization Policy, we conclude that: - The Policy was developed primarily to avoid audit differences with KPMG, and to provide substantial discretion to management to avoid recording catch-up adjustments, rather than to comply with GAAP; - Various provisions of the Policy were not, in fact, consistent with GAAP; - Management implemented the Policy inconsistently and recorded catch-up in a manner designed to reduce income statement volatility; and - Management's disclosures to the Board about the Policy and its implementation were misleading. II. BACKGROUND A. Applicable Accounting Standard FAS 91 sets forth the accounting for nonrefundable fees and costs associated with originations and purchases of loans. FAS 91 applies to various types of assets on Fannie Mae's books, including loans, mortgage-backed securities ("MBS"), and real estate mortgage investment conduits ("REMICs") (collectively, "loans"). For purposes of this report, we will refer to purchase premiums, discounts, and other deferred price adjustments relevant to FAS 91 collectively as "deferred price adjustments." FAS 91 requires the use of the interest method, which recognizes interest income net of deferred price adjustments at a constant effective yield (also referred to as "effective" or "level" yield). Under the interest method, deferred price adjustments should be recognized over the life of the loans as an adjustment to the loans' yield. Under FAS 91, if certain specified criteria (discussed in more detail below) are met, then anticipated prepayments may also be considered in determining the period over which the deferred price adjustments are recognized. As stated in FAS 91: If the enterprise holds a large number of similar loans for which prepayments are probable and the timing and amount of prepayments can be reasonably estimated, the enterprise may consider estimates of future principal prepayments in 34
the calculation of the constant effective yield necessary to apply the interest method.(147) Therefore, a company such as Fannie Mae may estimate future prepayments in its calculation of the constant effective yield if the following conditions are met: (1) the company holds a large number of similar loans; (2) the prepayment of those loans is probable; (3) the timing of the prepayments can be reasonably estimated; and (4) the amount of the prepayments can be reasonably estimated. Fannie Mae claimed to have satisfied these conditions and therefore elected to include estimates of future prepayments in its level yield calculation. Once an enterprise elects to anticipate prepayments in making its FAS 91 calculation, two situations may arise that may impact the timing of the amortization of the deferred price adjustments. The first is a difference between the estimated prepayments and actual prepayments. The second is when the company periodically revises its estimate of future prepayments. These two situations can occur simultaneously. When one or both of these situations occur, the company must recalculate the effective yield to reflect actual prepayments and the revised estimate of future prepayments. As set forth in FAS 91: If the enterprise anticipates prepayments in applying the interest method and a difference arises between the prepayments anticipated and actual prepayments received, the enterprise shall recalculate the effective yield to reflect actual payments to date and anticipated future payments. The net investment in the loans shall be adjusted to the amount that would have existed had the new effective yield been applied since the acquisition of the loans. The investment in the loans shall be adjusted to the new balance with a corresponding charge or credit to interest income.(148) When Fannie Mae adopted FAS 91, it elected to include estimates of future prepayments in its calculations of the constant effective yield.(149) Accordingly, when Fannie Mae's actual prepayments differed from its prior estimates, and/or when Fannie Mae revised its estimate of future prepayments, the Company was required under - ---------- (147) ACCOUNTING FOR NONREFUNDABLE FEES AND COSTS ASSOCIATED WITH ORIGINATING OR ACQUIRING LOANS AND INITIAL DIRECT COSTS OF LEASES, Statement of Fin. Accounting Standards No. 91, P 19 (Fin. Accounting Standards Bd. 1982). (148) Id. (149) Letter from Jonathan Boyles to Stephen M. Cutler and Paul Berger, dated Nov. 3, 2004, FNMSEC 151-59, at FNMSEC 152 (hereinafter "OCA Letter"); Fannie Mae, 2002 Annual Report, at 52 (2003) available at http://fanniemae.com/ir/pdf/ annualreport/2002/2002annualreport.pdf. This and other examples of documents relevant to the discussion in this Chapter can be found in the accompanying Appendix at Tab B. 35
FAS 91 to recalculate the constant effective yield to reflect the difference and immediately recognize in income any cumulative adjustment required in the amortization of its deferred price adjustments. Fannie Mae referred to this cumulative adjustment as the "catch-up." B. Relevant Amortization Accounting Systems In order to better understand the Company's stated rationale for the amount of the 1998 catch-up it recorded and the Amortization Policy generally, as well as the investigation's findings relating to these issues, a brief summary of the Company's FAS 91 amortization accounting systems is useful.(150) The systems that Fannie Mae used to calculate FAS 91 amortization and amortization catch-up can be separated into three groups: source, modeling, and amortization systems. The source systems have two main functions for FAS 91 amortization accounting. First, they provide loan and security data such as the coupon rate, time to maturity, unpaid principal balance, and original premium and discount amounts to the modeling systems. In addition, the source systems provide original premium and discount data to the amortization systems. The modeling systems,(151) in turn, aggregate loan and security level data into pools based on certain characteristics including asset type (referred to at the Company as the "FAS 91 type"), month and year of acquisition, and coupon rate. This information and the selected interest rate path is then applied to prepayment models to generate estimates of conditional prepayment rates ("CPRs"), cashflows, and updated amortization factors that are ultimately used in the calculation of the catch-up estimate. Based on the catch-up estimate, and when management deemed it appropriate, the Company would adjust the actual amortization factors that are provided to the amortization system. The amortization system - known as PDI(152) - applies amortization factors received from the modeling systems to original premium and discount in order to calculate amortization expense or income and post amortization entries to the general ledger. - ---------- (150) For a more detailed description of these accounting systems, we refer the reader to the Appendix that discusses the Company's FAS 91-related systems at the end of this Chapter. (151) Fannie Mae's modeling system in 1998 was known as PDAMS. PDAMS was replaced in October 2001 by AIMS. Tr. of Aug. 31, 2004 OFHEO Dep. of Jeffrey Juliane, at 217:23-219:6 (hereinafter "Juliane OFHEO Testimony"). (152) PDI was subsequently replaced in 2003 by iPDI. 36
The Company's FAS 91 calculations, therefore, depended in part on certain subjective criteria, including the Company's views on interest rates and on prepayment estimates. The Company recognized this and disclosed FAS 91 as one of the critical accounting estimates in its financial statements.(153) The reliability and accuracy of the modeling systems were important components of the Company's catch-up calculation. C. Accounting Practices Under FAS 91 Prior to 1998 1. Management Never Recorded Catch-up Until 1998 Despite the fact that Fannie Mae adopted FAS 91 in 1987, management elected not to record any associated catch-up in accordance with FAS 91 during the periods prior to 1998. No legitimate explanation was offered as to why management failed to comply with GAAP in those periods.(154) It appears that KPMG was fully aware of this situation. KPMG noted an audit difference for unrecorded catch-up as early as 1996, in the amount of - ---------- (153) Fannie Mae, 2003 Annual Report (Form 10-K), at 42 (Mar. 15, 2004), available at http://www.fanniemae.com/ir/pdf/sec/2004/f10k03152004.pdf. (154) In a submission to Senator Warren B. Rudman dated December 8, 2005, counsel for former Chief Financial Officer Timothy Howard asserted that Fannie Mae never recorded catch-up amounts pre-1998 because "changes in the catch-up its model calculated from quarter to quarter could just as easily be caused by known modeling limitations and inherent estimation uncertainties as by any actual differences between previously anticipated future prepayment rates and currently anticipated future prepayment rates." Mem. from Zuckerman Spaeder LLP to Senator W. Rudman, dated Dec. 7, 2005, at 3-4 (hereinafter "Howard Submission"). A copy of the Howard Submission can be found in the accompanying Appendix at Tab G. We reject this rationale for two reasons. First, putting to one side the merits of the claim about the "modeling limitations and inherent estimation uncertainties," which we address infra in Subsections III.E.3 and III.E.4, Howard's assertion ignores FAS 91's requirement to recognize the difference between actual prepayments received and the previously estimated prepayments. Given that there is no modeling needed to compare actual prepayments received to previously estimated prepayments, we do not find Howard's rationale persuasive. Second, if management genuinely was unsure about the amount of catch-up to record due to a question about the reliability of its modeling systems, then it should have sought to improve its modeling systems. We have seen no evidence that management sought to improve its modeling system beyond the capabilities of PDAMS in the eleven-year period during which they ignored the requirements of FAS 91. 37
$103 million.(155) As explained in greater detail below, because the Company did not consistently analyze, and did not report to KPMG, the catch-up associated with its "non-core"(156) book of assets until 1998, the $103 million audit difference only represented a portion of the catch-up that should have been recorded under FAS 91.(157) Management's failure to record any catch-up adjustments in periods prior to 1998 contributed to the accumulation of Fannie Mae's catch-up in 1998. In addition, documents showed that management likely understated catch-up expense in periods before 1998 because it had used interest rate paths to estimate prepayments that were higher than the prevailing market interest rates, notwithstanding the general decline in interest rates in the same time period.(158) 2. Management Did Not Model Its Non-Core Book In Accordance With GAAP Another factor contributing to the large catch-up in 1998 was the fact that management historically did not model amortization on REMICs and "synthetic - ---------- (155) See, e.g., Fannie Mae Audit Differences, dated Dec. 31, 1996, FMSE-IR 22059; Fannie Mae Audit Differences Dec. 31, 1997 - Restated. Leanne G. Spencer said that there had been an unrecorded catch-up balance at Fannie Mae since the Company's adoption of FAS 91, and that KPMG had noted an audit difference about the unrecorded catch-up every year during that period, prior to 2000. (156) In the context of FAS 91 amortization, Fannie Mae refers to premium and discount on whole loans and MBS (excluding REMICs), as well as buy-ups and buy-downs on guaranty fees, as its "core" book and refers to premium and discount on its REMICs and "synthetic REMICs" as its "non-core" book. See, e.g., Fannie Mae Catch-Up Summary, dated Jan. 7, 1999, FMSE-IR 182475. (157) Fannie Mae Accounting Issues Discussion Dec. 31, 1998, dated Jan. 8, 1999 (hereinafter "Accounting Issues Discussion"). (158) Fannie Mae PDA Catch-Up, dated Apr. 10, 2000, FMSE 104860. This document showed that in October 1997, management used a thirty-year fixed rate mortgage ("FRM") of 8.00% even though the actual thirty-year FRM had declined to 7.29% by that time. The catch-up calculated using this rate was $70.0 million, which appears to be the year-end catch-up amount that the Company communicated to KPMG. Summary of Catch-Up as of Oct. 1997, FMSE-IR 34923. These documents also indicated that once the Company dropped the rate it used to 7.40% approximately one month later, the catch-up for the same assets more than doubled, to $148.8 million. Fannie Mae PDA Catch-Up, dated Apr. 10, 2000, FMSE 104860. Accordingly, it appears that management underestimated the 1997 year-end catch-up amount that was communicated to KPMG. In the first three quarters of 1998, management continued to use an interest rate that was higher than the current average FRM as published by Freddie Mac. Id. 38
REMICs"(159) (collectively, the "non-core" book) in accordance with FAS 91. Prior to 1998, the Company did not include REMICs in its level yield calculation or its calculation of catch-up. Rather than amortizing deferred price adjustments for REMICs at a constant effective yield, as is required by FAS 91, management used sum-of-the-years digits ("SYD") and/or straight-line amortization methods as a proxy for the interest method for these securities (the "proxy amortization methods").(160) Several reasons were offered by then management for using a proxy amortization method for REMICs and synthetic REMICs historically, including the inadequacy of the Company's modeling system to model these securities and the then relatively small size of the non-core book.(161) These reasons notwithstanding, as early as 1995, employees in Financial Standards recognized that Fannie Mae's accounting for REMICs did not conform with GAAP, but nothing was done primarily based on the belief that REMICs were difficult to model.(162) More significantly, it does not appear that anyone assessed the modeling of REMICs to determine whether the proxy amortization methods, in fact, approximated the level yield method.(163) Documents also showed that the Company's catch-up related to synthetic REMICs had been growing since at least 1996.(164) As of June 1998, catch-up related to - ---------- (159) Synthetic REMICs were combinations of interest-only securities and other securities, such as REMICs, MBS, and principal-only securities, that were treated as a mortgage-backed security in order to avoid accounting for possible impairments in the value of the interest-only securities. We address the propriety under GAAP of the Company's accounting for "synthetic REMICs" in Chapter V below. (160) Mem. from J. Boyles to Distribution, dated Oct. 26, 1995, FMSE-IR 213989-91, at FMSE-IR 213989; IO/REMIC Package Briefing For Tim Howard, dated Mar. 2, 1998, FM SRC 224143-47, at FM SRC 224145. (161) Mem. from Robert Crane to Richard DePetris, J. Boyles, et al., dated Jan. 12, 1994, FMSE-IR 213974-75, at FMSE-IR 213974. (162) Mem. from J. Boyles to Distribution, dated Oct. 26, 1995, FMSE-IR 213989-91, at FMSE-IR 213990; E-mail dated Mar. 2, 1998, FM SRC 224142-47, at FM SRC 224145. (163) A memorandum from Boyles, then a Director in Financial Standards, acknowledged that "[t]o date, there has been no recalculation of the effective yield to determine how closely [the proxy] method has reflected actual and projected cash flows. Because of the difficulties in getting historical cash flow information for the securities, this method will continue to be used for the existing REMIC portfolio." Mem. from J. Boyles to Distribution, dated Oct. 26, 1995, FMSE-IR 213989-91, at FMSE-IR 213990. (164) IO/REMIC Package Briefing, dated Mar. 2, 1998. This document showed that the synthetic REMICs were evaluated in the fall of 1996 and that Fannie Mae had 39
the non-core book was approximately $186 million.(165) Documents showed that management did not disclose to KPMG the catch-up balance associated with its non-core book until January 8, 1999.(166) In fact, a document that management provided to KPMG on January 8 appeared to have been redacted in ways to suggest that there was no catch-up balance related to Fannie Mae's non-core book prior to year-end 1998.(167) While KPMG became aware of Fannie Mae's non-GAAP accounting treatment of the non-core book by year-end 1998,(168) we have found no evidence showing that KPMG was aware of this treatment prior to 1998. We also saw no evidence that KPMG took steps to quantify the effect of the error on prior period financial statements when it learned of the Company's treatment. By 1998, the size of Fannie Mae's non-core book had grown substantially. The Company's 1998 financial statements note that the net principal balance associated with the non-core book more than doubled that year, increasing from $35 billion as of December 31, 1997, to $77 billion as of December 31, 1998.(169) Perhaps recognizing that the "small size of the book" rationale no longer justified the proxy amortization methods, management began a project in early 1998 - which was known as the "Quick Strike Task Force" - to convert its amortization for the non-core book from the proxy amortization methods to the effective yield method.(170) By the end of 1998, management had been able - ---------- "determined that [that it] had under-amortized premium [or had not recorded enough expense] into earnings by $28 million." (165) Risk Review with CFO, dated Sept. 22, 1998, FMSE-IR 308237-48, at FMSE-IR 308244-45. (166) Accounting Issues Discussion. (167) Compare id., with Fannie Mae Catch-Up Summary, dated Jan. 7, 1999, FMSE-IR 182475. The latter document, which was used internally at the Company, identified catch-up related to the non-core book for 1996, 1997, and prior to October 1998, and included these amounts in the "Corporate Total" catch-up balance. In contrast, the former document, a "Fannie Mae Catch-Up Summary" dated January 8, 1999 - which was found in KPMG's workpapers - was similar in format, but not only had the catch-up balances related to the non-core book prior to October 1998 been removed, but the "Corporate Total" calculations prior to October 1998 omitted the catch-up numbers associated with the non-core book in the January 7, 1999 document. (168) Accounting Issues Discussion. (169) Fannie Mae, 1998 Annual Report, at 46 (1999) (hereinafter "1998 Annual Report"), available at http://www.fanniemae.com/global/pdf/ir/annualreport/1998/fullreport.pdf. (170) Undated Notes, FM SRC 310342-44, at FM SRC 310343; Howard Submission at 5. 40
to model approximately fifty-eight percent of the Company's REMICs using the effective yield method.(171) Other documents indicate that the Company may have modeled a higher percentage of synthetic REMICs using the effective yield method by the end of 1998.(172) The Company converted essentially all of its amortization for the non-core book to the level yield method sometime in 2000.(173) Significantly, the level yield modeling of a portion of these non-core book securities in 1998 showed that the Company needed to record a catch-up expense of nearly $298 million.(174) Under paragraph 13 of Accounting Principles Board Opinion No. 20, Accounting Changes ("APB 20"), "[a] change from an accounting principle that is not generally accepted to one that is generally accepted is a correction of an error for purposes of applying this Opinion."(175) Since the Company switched from a non-GAAP proxy amortization method to a GAAP amortization method for its non-core book, management should have considered the change to be the correction of an error under - ---------- (171) Undated Unmodeled REMIC Premium/Discount as of May 28, 1998, FMSE-IR 28020; Mem. from Jeffrey Juliane and Rene LeRouzes to Janet L. Pennewell, et al., dated June 16, 1999, FMSE-IR 25223-24, at FMSE-IR 25223. These documents indicated that as of May 28, 1998, the Company had modeled fifty-eight percent of its REMICs using the level yield method and that this percentage remained approximately the same as of June 1999. Since we have not been provided with any analysis that shows the percentage of the non-core book that was modeled as of December 31, 1998, and the next analysis we have seen that showed an increased percentage of REMICs modeled is dated June 16, 1999, we assume that only approximately fifty-eight percent of the book was modeled by the end of 1998. (172) An analysis performed as of May 28, 1998 indicates that synthetic REMICs were not part of the fifty-eight percent of REMICs modeled at that time. See Draft of March 1998 REMIC Premium and Discount, dated May 28, 1998, FMSE-IR 26684-85. Several documents indicate that a higher percentage of synthetic REMICs had been modeled using the effective yield method by the end of 1998. See, e.g., 1998 IO Bundle Catch-up, FMSE-IR 23163-64 FMSE-IR 23168-69. (173) See Mem. from J. Pennewell to T. Howard, dated Sept. 15, 2000, FMSE 217491-92, at FMSE 217491. This document, which has the subject "Amortization Policy" states: "Subsequent to that analysis taking place [in June 2000], we achieved a major deliverable of converting our REMIC portfolio to level yield accounting. Not only does this allow us to account for these securities correctly (and consistent with how we are modeling them), it also allows us to now apply the on-top amortization that we have been recording to the general ledger but not been able apply to the individual securities." Id. (174) Fannie Mae PDA Catch-Up, dated Apr. 10, 2000, FMSE 104860. (175) ACCOUNTING CHANGES, Accounting Principles Board Opinion 20, P 13 (Accounting Principles Bd. 1971) (hereinafter "APB 20"). 41
APB 20, and the impact of the correction on prior periods should have been assessed.(176) There is no evidence that anyone at the Company did so either in 1998 or in subsequent periods as an increased percentage of the non-core book was modeled. III. 1998 CATCH-UP CALCULATION AND ADJUSTMENT With this background, we now turn to 1998 and discuss in this Section the facts relating to the decision to record only $240 million of the $439 million in calculated catch-up expense for 1998. We conclude that the evidence strongly suggests that management's decision was motivated by the desire to meet published EPS and internal bonus (AIP) targets. A. The Catch-Up Calculation for 1998 Based on the Company's then existing systems, modeling, and estimation methods, the total catch-up estimated as of year-end 1998 was $439.7 million, of which $141.7 million was attributable to the Company's core book, and $298.0 million was attributable to its non-core book.(177) To derive this number, management applied the base interest rate path of 6.75% for the remaining life of its core and non-core book.(178) This base interest rate path was the rate path used at the time for other financial planning and accounting purposes throughout the Company.(179) Despite the total calculated amount of $439 million in catch-up expense, management recorded only $240 million in catch-up expense as of December 31, 1998 ($180 million was charged to net interest income and $60 million was charged to guaranty fees).(180) Had management recorded the entire $439 million catch-up expense, the Company would not only have missed Wall Street analyst EPS expectations of $3.22 per share, but it would also have missed even the minimum AIP target, resulting in no AIP bonuses for 1998.(181) As is discussed infra in Subsection III.B., management's - ---------- (176) Any subsequent change in the catch-up generated by the increase in the percentage of REMICs modeled using the constant effective yield method also should have been considered as the correction of an error and analyzed under APB 20. (177) Fannie Mae Catch-Up Summary, dated Jan. 7, 1999, FMSE-IR 182475. (178) Id. (179) Undated Q4 Earnings Forecast-Scenario Comparison, FMSE-IR 23789; Catchup Summary: Scenario 2 as of Dec. 31, 1998, dated Jan. 13, 1999, FNMSEC 2678; Accounting Issues Discussion. (180) Journal Entries, dated Jan. 9, 1999, FNMSEC 2710, 2712. (181) Assuming a thirty-five percent statutory tax rate, the after-tax impact of recording the additional $199.7 million of catch-up expense would have been $129.8 million. When divided by the Company's 1,037.4 million diluted average shares outstanding 42
reasons for recording only part of the total catch-up number included interest rate volatility and modeling imprecision, which purportedly rendered the $439 million to be an unreliable estimate. Management thus deferred the recognition of the remaining $199 million catch-up expense to subsequent financial reporting periods, and planned to reduce this amount, in part, through scheduled on-top adjustments in 1999 of roughly $8 million per month.(182) As for how management came to select the $240 million number as the appropriate catch-up expense, then Controller, Leanne G. Spencer, conceded that there is no financial analysis that would show the calculation from which the $240 million expense was derived. According to Spencer, she and Timothy Howard, then Chief Financial Officer, simply became "comfortable" that the $240 million was the "best estimate" to record. As demonstrated below, documents obtained from the Company's files showed that management chose $240 million because it was the greatest amount of catch-up expense the Company could afford to record, and, by making certain other offsetting adjustments, still exceed then Wall Street expectations of $3.22 per share for 1998. B. Evidence Reflecting Management's Focus on Meeting EPS Targets Documents showed that management at Fannie Mae was focused on the growing catch-up and its potential impact on the Company's ability to meet EPS targets for 1998. Further, documents showed that management was concerned about employee morale should the Company fail to meet maximum bonus targets. Management at the highest levels was focused in 1998 on the growing catch-up and its potential impact on the financial statements. For example, in a memorandum dated July 13, 1998 from Spencer, then Vice President--Financial Reporting, to Lawrence M. Small, then President and Chief Operating Officer, with copies to Howard and the Controller, Spencer responded to Small's question, which was to determine "whether it would be practicable to create an allowance linked to our purchased premium and discount amortization."(183) Spencer responded that this was not a good "approach" because, among other reasons, a reserve was not permitted by GAAP and in any event, creating such a reserve would result in Fannie Mae having to "take a - ---------- in 1998, the after-tax impact of recording the entire $439 million expense would have reduced the 1998 EPS to $3.1058, which was below the Company's minimum AIP target of $3.1300. Materials for Jan. 19, 1999 Meeting of the Compensation Committee of the Board of Directors of Fannie Mae, FMSE-IR 276226-73, at FMSE-IR 276233. (182) Monthly Summary of Amortization Adjustments, dated July 14, 1999, FMSE-SP 3102. (183) Mem. from L. Spencer to Lawrence M. Small, dated July 13, 1998, FMSE-SP 85799. 43
charge to income that would put us essentially in the same place as we are without a reserve."(184) Documents also showed that senior management was concerned not only about meeting EPS targets, but also about meeting maximum bonus, or AIP, targets. First, a draft memorandum from Spencer and Janet L. Pennewell, then Director--Business Planning, addressed to Small, Howard and others, dated July 24, 1998, highlighted management's focus on reaching Wall Street analyst EPS expectations for 1998.(185) The memorandum states: Despite the $61.7 billion increase in business volume and 2.1 percent increase in our overall book of business, our EPS remains flat to plan for 1998 and has fallen in each of the out years. We will be working closely with Tim and Rob to develop strategies for reaching a minimum of $3.20 for 1998, which is the analysts' consensus for the year.(186) A few weeks later, in a memorandum dated August 10, 1998, Small explained to then CEO-Designate Franklin D. Raines the significance of reaching not only Wall Street EPS expectations of $3.21 for 1998, but also of reaching $3.23 to achieve maximum AIP bonuses for employee morale.(187) In his memorandum, Small observed that management's then forecast for 1998 was approximately two cents short of analysts' expectations, and wrote: For 1998, I'm reasonably confident there's enough in the "non-recurring earnings piggy bank" to get us to $3.21. While that number should satisfy investors, you should be aware that last year the AIP paid out just short of the maximum. This year, the maximum is $3.23, so at $3.21, the bonus pool will be noticeably lower than in 1997, a fact - ---------- (184) Id. In the memorandum, Spencer articulated two additional reasons why the Company should not create a FAS 91 reserve: (1) "there is no accounting concept of an allowance in the context of purchased premium/discount amortization"; and (2) "the SEC has stated in the past that they are against companies setting up general `corporate allowances' because they are viewed as piggy banks for the company and do not have any basis in accounting theory." Id. (185) Mem. from L. Spencer and J. Pennewell to L. Small, Jamie S. Gorelick, T. Howard and Robert J. Levin, dated July 24, 1998, Zantaz Document 1104455. (186) Id. at 1. (187) Mem. from L. Small to Franklin D. Raines, dated Aug. 10, 1998, FMSE-IR 331263-67, at FMSE-IR 331263-64. 44
which will, of course, be rapidly observed by officers and directors come January.(188) While he did not recall writing this memorandum specifically, when asked to explain the reference to AIP goals, Small stated that as the COO of Fannie Mae, he felt obliged to consider the effect of the Company's earnings plan on the morale of the hundreds of people that benefited from the AIP.(189) Additional documents showed that management's primary focus as it considered the growing catch-up in 1998 was to manage it to a number that did not cause the Company to miss EPS targets. In particular, a document suggested that the large catch-up expense was discussed at an internal meeting - called "Risk Review with CFO" - that took place on September 22, 1998.(190) The document also suggested that the following participants attended the meeting: Howard, Spencer, Thomas A. Lawler, then Senior Vice President--Portfolio Analytics, and Jonathan Boyles, then Director of Financial Standards, among others.(191) Among the items on the agenda for this Risk Review with CFO were "Amortization of Core Book (PDAMS)" and "Amortization of REMICs."(192) While Spencer declined to be interviewed about this memorandum, notwithstanding its discovery in her files, Pennewell identified the pages that followed - ---------- (188) Id. at FMSE-IR 331264. (189) When asked to explain the phrase "non-recurring earnings piggy bank," Small stated that it was an unfortunate choice of words and what he likely meant was that Fannie Mae could increase its earnings through unexpected revenue or one-time events, like the gain on a sale of property, unexpected payment of a loan, or finding savings in the budget. (190) Risk Review with CFO, dated Sept. 22, 1998, FMSE-IR 321564-78, at FMSE-IR 321571-72 (hereinafter "Risk Review with CFO"). As described more fully in Chapter X, this document was discovered in Spencer's office in May 2005 when Company employees took the proactive step of seizing and copying every file from Spencer's office while she was absent. The document - which was responsive to Paul, Weiss's request dating back to November 2004 - was never produced voluntarily by Spencer. After the document was discovered, Spencer declined any further Paul, Weiss requests for interviews. (191) Id. at FMSE-IR 321564. Spencer's calendar contained an entry on September 22, 1998 for a "PDAMS Project" meeting. L. Spencer's Weekly Calendar Appointments for the Week of Sept. 7, 1998 to Oct. 2, 1998, dated Jan. 22, 1999, FMSE-IR 228752. In an earlier interview, before this document came to light, Spencer described portions of this meeting in some detail and noted the same participants as attendees of the meeting. (192) Risk Review with CFO at FMSE-IR 321564. 45
the agenda in this document as notes that typically were prepared by Spencer for important meetings with Howard ("Spencer notes").(193) The Spencer notes stated in no uncertain terms that management's "priority" as of September 1998 was to book amortization expense while still "managing" the Company's EPS to $3.21, which was in accordance with then Wall Street analysts' expectations.(194) In particular, the notes contained a "plan for bringing catch-up into compliance" that included the following "Recommendations": 1. Establish as priority one the goal of making $3.21 per share in 1998. 2. Manage earnings to that target and manage to the net interest margin implied in the Q2 forecast against which we have set expectations. To the extent that we have "surplus" NII [net interest income] or extraordinary NII/g-fee [guaranty fee] adjustments that manifest during the remainder of the year: apply these amounts to increase the topside adjustments. 3. Doing number two will either reduce the catch-up somewhat or keep it from growing further. 4. In the last week of December, the November Laser database will be available. We will run our end of year process off of that database to determine the appropriate - ---------- (193) In addition to these pages of narrative, the "Risk Review with CFO" document included several pages of charts illustrating the Company's catch-up position, as well as charts related specifically to the REMICs and synthetic REMICs catch-up and the Company's buy-up position. Risk Review with CFO at FMSE-IR 321569-78. (194) Risk Review with CFO at FMSE-IR 321567; Mem. from L. Small to F. Raines, dated Aug. 10, 1998, FMSE-IR 331263-65, at FMSE-IR 331263. A submission made on January 5, 2006 by counsel for Spencer explained the focus in this document on reaching an EPS of $3.21 by stating: "In achieving FAS 91 compliance, management continued to recognize the Company's goal of meeting the earnings per share target of $3.21, by applying surplus `NII' and `g-fee' income to reduce the catch-up for the rest of the year." Letter from David S. Krakoff to Senator Rudman, dated Jan. 5, 2006, at 6 (hereinafter "Spencer Submission"). (A copy of the Spencer Submission can be found in the accompanying Appendix, at Tab G.) This assertion fails to comprehend that compliance with FAS 91 does not depend on having "surplus `NII' and `g-fee' income" and is required regardless of impact on EPS. 46
amount of year-end adjustment necessary to bring the catch-up into compliance (e.g. get it under $100 million and show it turning around over the planning horizon). Based on the data in our June database that entry would be in the neighborhood of $53 million. We would source those earnings from a reduction to our loss reserves. 5. We propose bringing the catch-up down to $75 million at the end of the year and then implementing an internal (not public to KPMG) policy whereby we would not eradicate the catch-up in 1999, but would NOT allow it to grow above $75 million. Thus, we would book a topside entry through earnings each month if the in-use factors caused the catch-up to grow. 6. Implement factor changes that increases [sic] speeds in accordance with the rate path we'll assume for Q3, but absolutely no later than for 99 Plan.(195) We find, based on the contents of this document and the circumstances of its discovery, that this document is credible evidence of management's motivation in its handling of the large catch-up expense in 1998. In addition to the express statement that meeting EPS targets was "priority one," the document also reflected management's willingness to utilize other non-GAAP measures - such as timing a reduction of excess loan loss reserves - to offset the impact of the catch-up expense. Finally, a set of schedules prepared in early January 1999 demonstrated further that management selected the $240 million catch-up number by backing into the amount of catch-up expense that could be recognized while still achieving an EPS of approximately $3.225,(196) which would have just exceeded then Wall Street EPS consensus of $3.22.(197) These schedules reflect consideration of three alternative catch-up adjustments for 1998 (the "Earnings Alternatives Schedules"). Each alternative resulted in the target EPS of approximately $3.225, and considered recording increasingly larger amounts of catch-up expense, with correspondingly larger offsetting adjustments that mitigated the impact of the catch-up expense.(198) According to Shaun Ross, then Project - ---------- (195) Risk Review with CFO at FMSE-IR 321567. (196) Fannie Mae Income Statement, dated Jan. 7, 1999, FMSE-IR 21476-79 (hereinafter "Earnings Alternatives Schedules"). (197) Notes for Jan. 14, 1999 Conference Call, FMSE-IR 211000-07, at FMSE-IR 211000; Notes for Jan. 19, 1999 Board of Directors Meeting, FMSE-IR 182240-45, at FMSE-IR 182240. (198) See Earnings Alternatives Schedules. 47
Manager--Business Planning in the Controller's Office, he prepared the schedules in haste on the morning of January 7, 1999 at Spencer's direction.(199) The first alternative ("Alternative I") considered recording only $75 million in catch-up expense and contained no offsetting adjustments, resulting in an EPS of $3.2246.(200) When shown the schedules at an interview, Spencer acknowledged that Alternative I would have been "unacceptable" because it would have recorded too little of the calculated catch-up expense of $439 million. The second alternative ("Alternative II") considered recording $240 million in catch-up expense.(201) This scenario also contemplated an offsetting adjustment resulting from a change in the Company's method for accounting for its investments in partnerships that generated low income housing tax credits ("LIHTC").(202) Based on the LIHTC accounting method change, the result of Alternative II was a nearly identical EPS of $3.2254.(203) Management ultimately selected Alternative II for 1998. The third alternative ("Alternative III") included all of the adjustments in Alternative II; in addition, it contemplated recording an additional $100 million of catch-up expense for a total of $340 million.(204) At the same time, Alternative III considered a - ---------- (199) Spencer stated that she had no recollection about the schedules or how they were used. (200) Earnings Alternatives Schedules, at FMSE-IR 21477. The $75 million adjustment was divided into a $15 million reduction to net interest income and a $60 million reduction to guaranty fee income. (201) Earnings Alternatives Schedules, at FMSE-IR 21478. The $240 million adjustment was divided into a $180 million reduction to net interest income and a $60 million reduction to guaranty fee income, which were the amounts ultimately recorded for 1998. (202) Id. In Alternative II, the recognition of an additional $165 million (the $240 million considered in the second alternative less the $75 million considered in Alternative I) of pre-tax catch-up expense (or $107.3 million after-tax expense) was offset by adjustments related to LIHTC. These adjustments relate to the accrual of the 1998 LIHTC, which generated $123.2 million of after-tax income, as well as the accrual of the related 1998 net operating losses ("NOLs") and the reversal of excess depreciation expense recorded in years prior to 1998, which contributed $23.2 million of pre-tax expense (or $15.1 million after-tax expense). Combined, these LIHTC adjustments had a $108.1 million after-tax income impact, which almost exactly offset the additional $107.3 million after-tax expense impact of increasing the amount of catch-up from Alternative I to $240 million in Alternative II. (203) Id. (204) Id. at FMSE-IR 21479. 48
reversal of the Company's provision for loan losses by $100 million, which reversal was not contemplated in Alternatives I or II.(205) This alternative, therefore, would have allowed the Company to record a total of $340 million of the $439 million catch-up expense while still achieving an EPS of $3.2254.(206) Spencer did recognize the first page of the Earnings Alternatives Schedules - the page that contained no proposed adjustments for catch-up or offsetting adjustments - and stated that something similar to it was used in a meeting with the Office of the Chairman in early 1999. Indeed, the calendars of both Raines, the then new CEO, and Spencer, reflected a meeting - described in Raines' calendar as an "Earnings Alternative Meeting" - at 8 a.m. on Friday, January 8, 1999.(207) The attendees of the meeting, according to Spencer, were Raines, Small, Jamie Gorelick, then Vice Chairman, Howard, Lawler, Sam Rajappa, and Spencer.(208) According to Spencer, at this meeting, she and others reported to the Office of the Chairman on the work that had been done with respect to calculating the catch-up, including the decision to record $240 million in catch-up expense and deferring more than $199 million into future financial reporting periods. Spencer recalled Raines stating, in substance, that if there were any other "elephants under the table," that this was the time to "bring them out," "clean them up," and "take care of them."(209) In other words, according to Spencer, Raines insisted that the Company should book everything that needed to be booked and should not go into 1999 with potential exposure. However, Spencer stated that Raines ultimately became comfortable with her and Howard's recommendation to record only $240 million of catch-up expense. According to Spencer, she and Howard explained at this meeting their rationale for leaving $199 million unrecorded, that is, interest rate volatility and modeling imprecisions. Neither Spencer nor any of the other attendees of this meeting recalled seeing or discussing the three Earnings Alternatives Schedules. - ---------- (205) Id. (206) Id. The $340 million adjustment was divided into a $280 million reduction to net interest income and a $60 million reduction to guaranty fee income. Spencer acknowledged that this alternative was also not acceptable without any supporting work on the loan loss reserves, which she did not recall being done at the time. (207) L. Spencer's Weekly Calendar, dated Jan. 4, 1999 to Jan. 8, 1999, FMSE-IR 228765; F. Raines' Monthly Calendar, dated Jan. 1999, FMSE-IR 27702-13, at FMSE-IR 27702. (208) We have interviewed all of the alleged attendees of this meeting with the exception of Howard, who declined our interview requests. No one other than Spencer recalled the meeting or any discussions that may have occurred at the meeting. (209) Spencer said that the words in quotes were hers, not Raines's. 49
Spencer identified the following three documents as being used at the January 8, 1999 meeting: (1) a document similar to the first page of the Earnings Alternatives Schedules, but containing only the 1998 results with no adjustments for catch-up;(210) (2) a document that showed the core book and non-core book total catch-up numbers for the years 1995 through 2002, and reflected the total 1998 catch-up expense amount of $439.7 million;(211) and (3) a document that illustrated and quantified the change in income for 1998 resulting from the Company's accounting method change for LIHTC.(212) In summary, the evidence shows that at this January 8 meeting, Spencer and Howard briefed the attendees about the full catch-up, the decision to record only $240 million for 1998, with the remaining $199 million to be recorded in the future periods, and the LIHTC accounting change and its financial statement impact. We saw no evidence that Spencer and Howard disclosed to the attendees how they derived the $240 million number from $439 million, or what the Company's obligations were under FAS 91. C. Other Adjustments Made to Mitigate Impact of Catch-Up Expense on 1998 EPS Management made two adjustments that appear to have been motivated by a desire to offset the impact of the $240 million catch-up expense and to allow the Company to meet EPS and AIP targets for 1998. These adjustments were: (1) changing the accounting for LIHTC as of 1998, which had the effect of recognizing two years' worth of LIHTC in one year, and (2) reversing approximately $3.9 million from an account containing aged unreconciled items into income as "miscellaneous income." 1. Changes in Accounting for LIHTC As discussed more fully in Chapter V below, Fannie Mae began investing in partnership entities that own low income housing in 1987.(213) The principal benefit of investing in these partnerships was the ability to defer federal income taxes through the use of accelerated depreciation and to reduce federal income taxes with tax credits. Tax credits are granted to investors in low income housing projects to encourage participation in these investments that would typically result in losses. - ---------- (210) Fannie Mae Income Statement, dated Jan. 7, 1999, FMSE-IR 182474. (211) Fannie Mae Catch-Up Summary, dated Jan. 7, 1999, FMSE-IR 182475. (212) Low Income Housing Tax Credit Investments, dated Jan. 7, 1999, FMSE-IR 182476. (213) Undated Low Income Housing Tax Credit Investments (LIHTC), FMSE-IR 182493-97, at FMSE-IR 182493. 50
Fannie Mae's pre-1998 accounting for LIHTC was not consistent with GAAP. The Company historically accounted for LIHTC on a cash basis, based on information it received from the Schedule K-1 filings of the partnerships in the subsequent year.(214) Documents showed management's awareness that this cash basis accounting was not in accordance with GAAP.(215) Rajappa, who was Controller of the Company from 1994 through 1998, acknowledged that accrual basis accounting was "more consistent" with GAAP than cash basis accounting. Similarly, Spencer stated that when she became Controller in 1998, she advocated the change to GAAP basis accounting because it was "more appropriate." In addition, prior to 1998, the Company improperly recognized its depreciation expense related to LIHTC on an accelerated tax basis rather than on a GAAP basis.(216) No legitimate explanations were offered as to why management was comfortable accounting for LIHTC prior to 1998 in a manner that did not comply with GAAP. Ken Russell from KPMG was aware in 1998 that the Company was taking measures to implement the LIHTC accounting change. Accrual basis accounting for LIHTC required the Company to record the tax benefit in the year that the benefits were earned, and recognize its share of the partnership's operating losses with depreciation expense computed in accordance with GAAP. Given that the Company had been recording the credits in arrears, that is, in the following year when the K-1s were received, effectuating the accounting change was expected to result in the recognition of two years' worth of tax credits in the year that the accounting change was implemented. Although there is some evidence showing that as early as 1996 management considered converting its accounting method for LIHTC to an accrual basis accounting in 1998,(217) documents showed that by 1998, the Company was planning to - ---------- (214) Letter from L. Spencer, dated May 23, 1996, FMSE-IR 306206-23, at FMSE-IR 306206. (215) See, e.g., Undated Low Income Housing Tax Credit Investments (LIHTC), FMSE-IR 182493-97, at FMSE-IR 182495. (216) Mem. from J. Boyles and Brian Harris to Distribution, dated Nov. 18, 1997, FMSE-IR 213784-86, at FMSE-IR 213786; E-mail from R. DePetris to J. Boyles, dated Feb. 3, 1997, FMSE-IR 213773-74, at FMSE-IR 213774. (217) A note from Spencer dated May 23, 1996 reads: The attached schedule details the latest outlook (per Q1 forecast) for the income projected from the equity investments in affordable housing projects. The projections assume, because of the sunset provision, that we will do no new deals after 12/31/97. These projections are cash basis in the forecast, i.e., 1996 recognition of [sic] Fannie Mae's books is the result of our estimation of what will be received on the 1995 K-1 return. In the 51
implement the accounting change for 1999.(218) A document dated November 18, 1998, which reflected comments from Spencer and Pennewell on Small's planned statement at an officers' meeting, contained the following note from Spencer and Pennewell about the LIHTC accounting change: Where you are talking about the accounting change. Nothing that you state is incorrect. However, we would like to soften it a little. Technically, if you "know" about a[n] accounting change you are supposed to book it. We haven't informed KPMG that we intend to implement this next year and our preference would be to not talk to them about it prior to year-end 1998 so they don't say `book it' at year-end. We've limited discussion of this to the inner circle, so wouldn't want to broadcast it to the officer group at this point.(219) As this document demonstrates, not only was management planning to implement the LIHTC accounting change for 1999 as of November 1998, but management also had kept information about the LIHTC accounting from KPMG because of the concern that KPMG would force management to make the change in a way that would be inconsistent with management's timing for the change, which was to support EPS growth in 1999.(220) - ---------- period we should [choose] to change to an accrual method, you simply pull forward the projections and in the year of change, you recognize two year's [sic] worth of activity. My current thoughts would be 1998 would be the year to change. Letter from L. Spencer, dated May 23, 1996, FMSE-IR 306206-23, at FMSE-IR 306206. (218) See, e.g., id.; Letter from L. Spencer to T. Howard, et al., dated Dec. 2, 1998, FMSE-IR 182499. (219) Mem. from L. Spencer and J. Pennewell to L. Small, dated Nov. 30, 1998, Zantaz Document 1519161, at 3. (220) See generally Mem. from J. Pennewell to L. Small, dated Nov. 25, 1998, FMSE-IR 182506 (regarding EPS growth in 1999, Pennewell writes "[t]he biggest contributor [for 1999 EPS growth] is the accounting change for the low income tax credit investments (LIHTC) which adds 3.4 percent to EPS growth. This includes changing from cash to accrual basis of accounting and changing our depreciation method."); E-mail from Shaun Ross to T. Howard, L. Spencer, and J. Pennewell, dated Dec. 22, 1998, FMSE-IR 182498 (forwarding to Howard "the 1999 income impact of the changes in the LIHTC partnerships"). 52
Notwithstanding the above documents, sometime in early January 1999, management determined to accelerate the timing of the accounting change and to make the change for 1998.(221) This new timing for the change is reflected for the first time in the Earnings Alternatives Schedules dated January 7, 1999, discussed supra in Subsection III.B., as one of the adjustments made to offset the impact of catch-up in Alternatives II and III.(222) The change allowed Fannie Mae to recognize the LIHTC and net operating losses for both 1997 and 1998 in 1998, and reverse the excess depreciation expense the Company had recognized in prior years.(223) Documents showed that management analyzed the change in the depreciation method under APB 20 and concluded that the change "[was] closer to a correction of an error,"(224) but other documents show that management concluded ultimately that both issues should have been accounted for as a change in accounting principle.(225) In our view, since both the cash-to-accrual and the depreciation method changes essentially were a conversion from a non-GAAP to a GAAP basis of accounting, the changes should have been considered as corrections of errors under APB 20.(226) Management failed to appropriately consider whether the error in its accounting for LIHTC was material to previously issued financial statements; nor did it consider whether the correction of that error in 1998 resulted in a material misstatement of the 1998 financial statements. By accelerating the accounting change for LIHTC by one year earlier, management was able to mitigate the impact of the $240 million catch-up expense adjustment. Spencer claimed that she did not recall what motivated the Company to - ---------- (221) The Company also changed its depreciation method in 1998 to be consistent with GAAP. (222) Earnings Alternatives Schedules at FMSE-IR 21478-79. (223) By the end of 1998, the excess of tax depreciation over book depreciation had reached $75.4 million ($60.9 million of which related to the period 1987-1997). See Cumulative Catch-Up at Dec. 31, 1998, FMSE-IR 52445. In addition, Fannie Mae estimated that NOLs for 1998 would be approximately $97.7 million and the tax credits for 1998 would be approximately $123.2 million. 1998 Q4 Forecast - LIHTC Partnerships - 1998 Cur. Accr., dated Jan. 6, 1999, FMSE-IR 52487. (224) Mem. from Kimberly Rawls to L. Spencer, et al., dated Apr. 14, 1998, FMSE-IR 178260-61, at FMSE-IR 178261. The memorandum concludes that the effect of the correction could be treated like a change in accounting principle because it was not material. Id. (225) Undated Notes on Changes to LMI Accounting, FMSE-IR 178258-59, at FMSE-IR 178258. (226) APB 20 P 13. 53
change LIHTC accounting methods one year ahead of plan. Howard declined to be interviewed,(227) and Rajappa disclaimed any responsibility for the timing of the accounting change, stating that he was transitioning from the Controller's Office function to his new position in the Office of Auditing ("Internal Audit") during this time period. We conclude that the above evidence strongly suggests that management's motive for accelerating the timing of the LIHTC accounting change was to mitigate the impact of the $240 million catch-up expense and to meet Wall Street EPS expectations. The decision to switch from non-GAAP to GAAP-based accounting methods for LIHTC was appropriate. As discussed in greater detail infra in Subsections III.D.2. and III.D.3., however, the disclosures made to the public and the Board of Directors as to the nature and reason for the accounting change were incomplete and misleading. 2. $3.9 Million in Miscellaneous Income from Account 162200 The second adjustment, which was a reversal of $3.9 million in aged credit items to income from a suspense account for aged unreconciled items known at the Company by the number 162200, was recorded on January 9, 1999.(228) This entry, which was made on the day that the Company's books for 1998 closed, allowed management to meet the maximum AIP bonus payouts for 1998.(229) The journal entry reflecting this $3.9 million reversal from the 162200 account described the amount as "miscellaneous income."(230) This vague description, - ---------- (227) In the Howard Submission, counsel for Howard asserted that the decision to make the LIHTC accounting change for 1998 was made only after personnel recommended that the Company record only $240 million of catch-up expense, and therefore the decision was made irrespective of EPS targets. Howard Submission at 11. We have seen no evidence supporting this assertion, and Howard has cited none. In fact, the evidence we have seen is to the contrary: in Howard's own talking points prepared for a meeting of the Board of Directors on January 19, 1999, he wrote that management decided to make the $240 million adjustment in 1998 because there was "room" created by the recognition of two years' worth of LIHTC credits. Undated Notes for Jan. 19, 1999 Meeting of the Board of Directors of Fannie Mae, FMSE-IR 182240-64, at 182240-41. (228) Journal Entry, dated Jan. 9, 1999, FMSE-IR 262764. (229) The final reported 1998 EPS of $3.2309 was just enough to trigger maximum bonus payouts under AIP. Under Fannie Mae's AIP for 1998, minimum bonuses would have been triggered by an EPS of $3.1300, the target bonus would have been triggered by an EPS of $3.1800, and maximum bonus triggered by an EPS of $3.2300. Materials for Jan. 19, 1999 Meeting of the Compensation Committee of the Board of Directors of Fannie Mae, FMSE-IR 276226-73, at FMSE-IR 276233. (230) Journal Entry, dated Jan. 9, 1999, FMSE-IR 262764. 54
along with the timing of the entry, caused us to review in greater detail the activity in the 162200 account, which after the reversal of $3.9 million, contained approximately $22.5 million of credit balances, among other items. (231) We also observed that KPMG had identified credits in the 162200 account on its list of audit differences going back to 1994, that is, KPMG had notified management of its belief that credit balances in the account should be reconciled and taken into income as early as 1994.(232) However, management did not do so. Management's prior refusal to reconcile and reverse the entire credit balance in the 162200 account per KPMG's recommendation made the decision to reverse $3.9 million into income on January 9, 1999 even more questionable. Moreover, evidence shows that management contemplated using the credits in the 162200 account to "resolve earnings shortfalls" in prior years. In particular, in a memo from Spencer to Howard that discusses the forecasted EPS for 1996, Spencer identifies a list of items that are "up [her] sleeve to solve an earnings shortfall," including what appears to be the 162200 account, which she describes as containing a "cushion" available for this purpose.(233) The $3.9 million entry was comprised of five specific credit items from the 162200 account - credit items that were originally booked to this account during the period June 1994 through January 1998.(234) While the journal entry explained the $3.9 - ---------- (231) Fannie Mae Audit Differences Dec. 31, 1998. (232) As early as 1994, KPMG had noted the balance of this account as an audit difference. FNMA 1622-00 Analysis at Dec. 31, 1994. Every year since 1996, KPMG proposed an audit adjustment, which would have reversed the credit balances in Account 162200, but Fannie Mae did not record those proposed adjustments. Fannie Mae Audit Differences Dec. 31, 1998; Undated mem. from Lynden Collier to Lan Nguyen; Fannie Mae Audit Differences Dec. 31, 1997 - restated; FNMA 1622-00 Analysis at Dec. 31, 1994. KPMG appears to have reviewed the activity in this account on a quarterly basis. See Analysis of Accounts Receivable - 162200, dated July 2003. Fannie Mae ultimately reversed the remaining credit balances in Account 162200 in 2003. Journal Entry, dated Nov. 4, 2003. (233) Undated letter from L. Spencer to T. Howard, FMSE-IR 359599-601, at FMSE-IR 359600. While this document does not specifically refer to the 162200 account by account number, it refers to a "bucket account" and states that the "cushion" is mostly related to net interest income. When shown this document, Pennewell stated her belief that the "bucket account" referred to the 162200 account. Further, an analysis of the composition of the 162200 account from 1999 indicated that the vast majority of the items in this account related to net interest income. Analysis of Account 162200, dated Nov. 30, 1999, FMSE-IR 307790. (234) Compare Analysis of Account 1622-00 as of Sept. 30, 1998, dated Oct. 1998. Analysis of Account 1622-00 as of Dec. 31, 1998, dated Dec. 1998. 55
million as "miscellaneous income," we have seen no contemporaneous analysis or support showing that the items had been reconciled and were the only credits in the 162200 account that warranted reversal into income at this time. Our interview of the employee whose name appears on the journal entry as having processed the entry, revealed no further information about the entry because the employee had no memory of the event. No other witnesses interviewed offered any additional information about this entry. The after-tax impact of the $3.9 million entry from the 162200 account on the 1998 EPS was approximately $0.0024.(235) Internal Fannie Mae AIP documents showed that the Company's reported EPS in 1998 was $3.2309.(236) If this entry had not been booked, EPS would have been approximately $3.2285. During our interviews, we learned from former Controller's Office staff that while EPS was typically rounded to two decimal places for public reporting purposes, for purposes of awarding bonuses under the AIP program, the Company carried EPS to four decimal places.(237) Therefore, an EPS of $3.2285 would not have triggered the maximum AIP payout in 1998, which required an EPS of $3.2300. With the $3.9 million of "miscellaneous income," the EPS for 1998 reached $3.2309. In the absence of any supporting justification for the reversal of $3.9 million on January 9, 1999, we can only infer that the $3.9 million entry was passed to trigger the maximum AIP bonuses. Our inference is supported by the fact that even without the $3.9 million in additional "miscellaneous income," the Company had reached $3.2285, which was sufficient to meet Wall Street expectations of $3.22 per share. Our inference is further supported by - ---------- (235) We derived this number by assuming a thirty-five percent statutory tax rate, which resulted in the after-tax income impact of $2.5 million. When divided by the Company's 1,037.4 million diluted average shares outstanding for 1998, the after-tax impact on EPS is calculated to be $0.0024. Fannie Mae, Selected Financial Information as of Dec. 31, 1998, available at http://www.fanniemae.com/ir/ pdf/earnings/1998/iap123198.pdf. (236) Materials for Jan. 19, 1999 Meeting of the Compensation Committee of the Board of Directors of Fannie Mae, FMSE-IR 276226-73, at FMSE-IR 276233. (237) See, e.g., Agenda-Compensation Committee of the Board of Directors-Fannie Mae, dated Jan. 19, 1999, FMSE-IR 15796-874, at FMSE-IR 15802-03. This document indicates that the maximum AIP bonus trigger in 1996 was $2.4800. The EPS for that year was $2.4764, which the Company rounded up to $2.48 for purposes of presentation in the financial statements. Fannie Mae, Fourth Quarter 1996 Investor Analysts Report, at 1 (1997), available at http://www.fanniemae.com/global/pdf/ir/financial/pack/1996/iap96.pdf. However, the AIP in 1996 allowed for only ninety-three percent of the maximum bonus payout because the EPS of $2.4764 did not meet the maximum AIP target of $2.4800. Agenda Compensation Committee of the Board of Directors - Fannie Mae, dated Jan. 19, 1999, FMSE-IR 15796-764, at FMSE-IR 15802. 56
the fact that the Company previously had resisted KPMG's recommendations to resolve and reverse the entire amount in the 162200 account. Finally, our inference is supported by the fact that no one has offered any justification for selecting the five credit items that were recognized as income on the day the Company closed its books for 1998. Based on goals that were set in January 1998, the Compensation Committee of the Board of Directors of Fannie Mae estimated that the AIP bonus pool would be funded at $25.9 million in 1998 if the Company achieved the maximum EPS goal, $17.3 million if the Company reached the target EPS goal, and $8.6 million if the Company achieved the minimum EPS goal.(238) By hitting $3.2309, the Company exceeded the maximum EPS goal for 1998, and the AIP bonuses paid for 1998 totaled $27.1 million.(239) D. Management's Disclosures About the 1998 Adjustments 1. Disclosures to KPMG Management fully disclosed to KPMG the full calculated catch-up amount for 1998, its rationale for recording only $240 million in expense, and the change in accounting for LIHTC. KPMG noted an audit difference for the unrecorded catch-up expense of $199 million, but did not qualify its opinion for the Company's 1998 financial statements. We have not seen any evidence that KPMG considered any qualitative aspects associated with not recording the full catch-up expense (such as the effect on the Company's EPS or AIP goals) in concluding that the failure to record the full adjustment did not have a material impact on the financial statements. In a meeting on January 8, 1999, management disclosed to KPMG that the catch-up expense calculated was approximately $439 million, but that it was recording only $240 million in catch-up expense.(240) According to Spencer, Boyles and James Parks, then Vice President and Multifamily Controller, attended this meeting with KPMG with her. Based on a memorandum summarizing the meeting that was found in KPMG's workpapers, it appears that management explained to KPMG that it was only recording $240 million in catch-up expense because "the calculation and projection of premium/discount amortization is complicated and imprecise."(241) Specifically, - ---------- (238) Annual Incentive Plan 1998 Plan Year, FMSE-IR 14834-36, at FMSE-IR 14836. (239) Agenda-Compensation Committee of the Board of Directors, dated Jan. 19, 1999, FMSE-IR 15796-874, at FMSE-IR 15802. The actual AIP payout number increased to $27.1 million from the estimate of $25.9 million in the beginning of 1998 due to the changes during the year in the number of employees eligible for the AIP or their salaries. Annual Incentive Plan 1998 Plan Year, FMSE-IR 14834-36, at FMSE-IR 14836. (240) Accounting Issues Discussion. (241) Accounting Issues Discussion. 57
management advised KPMG that: (1) interest rates were volatile and likely would move back up, and the catch-up from the 6.75% base interest rate path used would be completely reversed by the end of 1999 if the scenario were changed to an "up" scenario of 7.00% for six months and 7.25% thereafter; and (2) the projection of unamortized premium/discount uses a flat rate projected into the future instead of more appropriately reflecting a rising rate environment, and if a rising rate projection was used in the model, the catch-up estimate would decrease.(242) In response, KPMG advised management that, while its arguments for recording only $240 million of the catch-up expense may have support, the Company's budgeting and planning processes used the same base interest rate path that had been used in the catch-up calculation resulting in the $439 million number, and, therefore, the Company should have recorded the full $439 million in catch-up expense for 1998.(243) Accordingly, KPMG would note an audit difference in 1998 for the unrecognized catch-up expense of approximately $199 million.(244) KPMG then certified Fannie Mae's 1998 financial statements without qualification.(245) It appears that management agreed to reduce the $199 million in expense in 1999 through on-top adjustments. As for the LIHTC accounting change, according to Spencer, management made a presentation to KPMG at the January 8, 1999 meeting concerning the Company's decision to change to an accrual basis accounting for LIHTC.(246) While the LIHTC accounting change was not discussed in KPMG's workpapers that relate to this meeting,(247) other KPMG workpapers did reflect the 1998 year-end LIHTC adjustments.(248) However, these workpapers characterized the adjustments as resulting from "improved information systems which improved the Company's ability to estimate - ---------- (242) Id. Ken Russell could not recall how management arrived at the $240 million figure, and also could not recall management providing their calculations to KPMG. However, Russell did not tell his audit team to look into the issue. (243) Id. (244) Id. (245) 1998 Annual Report at 63. (246) The presentation showed Fannie Mae implementing the LIHTC accounting change for the year 1999 and not 1998. Undated Low Income Housing Tax Credit Investments (LIHTC), FMSE-IR 182493-97, at FMSE-IR 182496. Spencer stated that the purpose of the meeting was to walk KPMG through the proposed adjustments for 1998 and that the document may have shown the adjustment for 1999 to show that they were, in effect, "pulling 1999 activity into 1998." (247) Accounting Issues Discussion. (248) Fannie Mae Completion Memorandum Dec. 31, 1998, dated Mar. 15, 1999. 58
net operating losses and tax credits."(249) The description in KPMG's workpapers did not describe the LIHTC adjustments as a change in accounting methodology or a correction of an error. Russell was aware, however, that the LIHTC accounting change would allow the Company to recognize additional credits. Finally, as noted above, KPMG reviewed the 162200 account on a quarterly basis since at least 1996, but we have seen no evidence that the $3.9 million income entry from account 162200 was discussed with KPMG in January 1999. Based on a review of its workpapers, KPMG advised that it did not see any indication that it had raised any concerns with management about the reversal of the five items totaling $3.9 million into income on January 9, 1999.(250) In short, while management did not highlight the $3.9 million entry to KPMG, there is also no evidence that management concealed this information from KPMG. 2. Disclosures Made to the Board of Directors Management's disclosure about the 1998 adjustments to the Board of Directors, however, was less complete. Management failed to disclose to the Board of Directors the total catch-up expense number that was calculated, failed to disclose the purported reasons for recording only $240 million out of $439 million in expense, failed to disclose the audit difference that was noted by KPMG, and failed to disclose its decision to recognize the $3.9 million "miscellaneous income." Moreover, the disclosure to the Board regarding the LIHTC accounting change was misleading. The Board was first advised of the 1998 results and the catch-up at a meeting of the Board of Directors on January 19, 1999. While no individual Board members had any specific recollections of the presentations made at this particular meeting, some Board members did recall that Howard typically gave presentations on financial results at Board meetings. The Company produced the notes prepared by Howard for use at the January 1999 Board meeting from Howard's files.(251) Howard's notes for the Board meeting omitted the fact that the Company's systems had calculated $439 million in catch-up expense, omitted the fact that management chose to record only $240 million in catch-up expense, and did not disclose - ---------- (249) Id. Similarly, Russell recalled that he had meetings with James Parks prior to the fourth quarter of 1998 during which he was made aware that Fannie Mae had to gather information to change its LIHTC accounting method and was working on developing a process to enable the Company to recognize the credits for financial reporting purposes more quickly. (250) Account 1622-00 Analysis - 12/31/98. (251) Notes for Jan. 19, 1999 Meeting of the Board of Directors of Fannie Mae, FMSE-IR 182240-64. Because Howard declined to be interviewed, we have not confirmed with him how much of these notes he used at this Board meeting. 59
the fact that KPMG had noted an audit difference for the remaining $199 million of unrecognized catch-up expense. Indeed, Howard's notes misleadingly portrayed the $240 million catch-up expense as an adjustment that management "elected" to record in the fourth quarter of 1998:(252) what we did in the fourth quarter [of 1998] is decide to take a large amount of unamortized premium and recognize it all at once, rather than over a longer period. And we did something similar on the guaranty fee line . . . . Just as we elected to make a lump-sum adjustment to purchase premium to reflect[ ] faster actual and expected prepayments, we made a similar adjustment to guaranty fees.(253) Howard's notes stated that management "elected" to record the $240 million catch-up expense in 1998 because: (1) recording the adjustments at that time allowed the Company to "report higher levels of net interest income and guaranty fees over the next year or two," and (2) the Company had "room" to recognize the catch-up expense "while still coming above the analysts' consensus for the year, thanks to the tax credit adjustment and the exceptional strength in net interest income due to extremely rapid portfolio growth in the second half of the year."(254) The notes did not contain management's purported concerns about the reliability of the Company's modeling systems, the expected rise in interest rates, or the impact of modeling even a portion of the REMIC book under the level yield method on the catch-up. Similarly, there was no reference to KPMG's audit difference. As for the accounting method change for LIHTC, Howard's notes did refer to the LIHTC accounting change as creating "room" in which to record the $240 million catch-up expense,(255) and stated that the change allowed the Company to recognize two years' worth of tax credits in 1998.(256) However, Howard's notes did not mention that the two years' worth of tax credits resulted from a change in accounting methodology, that is, it was a non-recurring one-time event, which would have informed the Board about the true quality of the Company's earnings. The notes also did not - ---------- (252) Id. at FMSE-IR 182241. (253) Id. (emphasis added) (254) Id. (255) Id. (256) Id. at FMSE-IR 182240. 60
mention that the Company's historical accounting for LIHTC was non-GAAP, and therefore the change was a correction of an error.(257) On February 16, 1999, Spencer made a presentation to the Audit Committee concerning the Company's 1998 financial results, including the catch-up. Howard was also present at this meeting.(258) Spencer's presentation similarly failed to disclose critical information to the Board. According to the talking points used by Spencer for this meeting, she described the catch-up adjustment as: "$180 million and $60 million," which "represent our recording additional amortization of purchase premium and discount balances associated with our net mortgage portfolio, and a similar adjustment to prepaid or deferred guaranty fees."(259) Spencer failed to advise the Audit Committee of the audit difference for $199 million. The audit difference was reported to the Audit Committee a year later, in February 2000, in connection with KPMG's audit of the Company's 1999 financial statements.(260) Spencer stated that management began reporting all audit differences to the Audit Committee after the issuance of Statement on Auditing Standards No. 90, Audit Committee Communications ("SAS 90"), in December 1999.(261) While the above auditing standards may apply to KPMG, we conclude that it was management's responsibility - particularly, Spencer's and Howard's - to disclose the audit difference, which, if recorded, would have significantly affected the Company's ability to meet EPS targets for 1998, to the Audit Committee. At a minimum, the Compensation Committee of the Board should have been advised of the $199 million catch-up expense and the impact of that decision on bonus calculations under the AIP. - ---------- (257) See id. The minutes of the meeting do not provide any further information as to what Howard said about the LIHTC accounting change. Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 19, 1999, FMSE 1201-18. (258) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 16, 1999, FMSE 14146-60, at FMSE 14146. (259) The Year in Review: The 1998 Annual Report, dated Feb. 16, 1999, FMSE-IR 22077-91, at FMSE-IR 22082. (260) The Year in Review: The 1999 Annual Report, dated Feb. 15, 2000, FMSE 14590-607, at FMSE 14605-07; Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 15, 2000, FMSE 14501-09, at FMSE 14503. (261) We suspect that Spencer was referring to Audit Adjustments, Statement on Auditing Standards No. 89 (American Inst. of Certified Public Accountants 1999) ("SAS 89"), which was adopted on December 1999 and which required KPMG to report all "passed" audit adjustments (also referred to as "unadjusted" audit differences) to the Audit Committee, even if KPMG considered them to be immaterial. 61
Russell and Julie Theobald, both from KPMG, were both present at this February 1999 Audit Committee meeting, and either could have informed the Audit Committee of the KPMG audit difference, but did not. The meeting minutes only reflect that Russell stated that: (1) "KPMG did not identify any areas within the [1998] financial statements that they believe include[d] unreasonable estimates"; (2) "KPMG did not propose any adjustments that would have a significant effect on the 1998 financial statements prepared by management"; (3) "during their 1998 audit, KPMG had no disagreements with management"; and (4) "Fannie Mae's significant accounting policies . . . [were] in accordance with Generally Accepted Accounting Principles."(262) As for the LIHTC accounting change, based on Spencer's notes, her statements to the Board were misleading. Spencer's notes indicate that she made the following statement to the Audit Committee: "Based on the utilization of these improvements to our information and internal controls we were able to refine the timing of recognition of these tax credits in 1998."(263) The talking points failed to inform the Audit Committee that the Company's historical accounting for LIHTC had not been GAAP-compliant, as management knew, and also omitted the fact that the change related to a simple change in accounting methodology. Further, the talking points failed to disclose that the accounting methodology change to LIHTC resulted in a one-time event of recognizing two years' worth of credits in a single year. None of the Audit Committee members whom we interviewed who were at the February 1999 meeting recalled this discussion. 3. Disclosures Made to the Public Finally, management's public disclosures concerning the 1998 catch-up expense and the LIHTC accounting change tracked the disclosures made to the Board, and accordingly, were similarly incomplete and misleading.(264) For example, the MD&A - ---------- (262) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 16, 1999, FMSE 14146-60. Theobald apparently recalled that, in February of 1999, she explained the audit difference to Thomas Gerrity, who was then serving as the head of the Audit Committee. However, as noted above, the minutes of the Audit Committee meeting contradict this recollection, and we have found no documents that support Theobald's recollection. Regardless of whether Theobald discussed the audit difference with Gerrity in February 1999, it was management's responsibility to provide full disclosure to the Board, and we conclude that management failed to do so. (263) The Year in Review: The 1998 Annual Report, dated Feb. 16, 1999, FMSE-IR 22077-91, at FMSE-IR 22082. (264) 1998 Fourth Quarter Earnings Conference Call, dated Jan. 14, 1999, FMSE-IR 210990-99; Notes for Jan. 14, 1999 Conference Call, FMSE-IR 211000-07. These are Howard's notes for, and the transcript of, a January 14, 1999 analyst conference call discussing Fannie Mae's 1998 financial results. 62
section of Fannie Mae's 1998 Annual Report noted only that management recorded additional amortization in the fourth quarter, which reflected "management's estimate of the effects of the decline in interest rates and the high level of 1998 refinancing activity on the unamortized premium and discount balances associated with the net mortgage portfolio."(265) The MD&A section also contained the following statements about the LIHTC accounting change: The reduction in federal income tax expense and the effective federal income tax rate for 1998 reflect the recording of additional low-income housing tax credits in the fourth quarter of 1998. The additional tax credits were a result of the corporation using improved systems and information to refine the timing of the recognition of the tax benefits associated with investments qualifying for low-income housing tax credits.(266) This disclosure did not accurately reflect that the true reason for the LIHTC accounting change was a change in accounting methodology and also failed to inform the reader that Fannie Mae recognized two years of tax credits in 1998. E. Management's Reasons for Recording Only $240 Million This section considers the legitimacy of several reasons offered by Howard and Spencer for recording only $240 million of catch-up expense for 1998, which reasons are contained in their statements to OFHEO and to Paul, Weiss during our interviews of Spencer, and in submissions made to Senator Rudman through counsel.(267) Both of these submissions in their entirety are attached as exhibits behind Tab G in the Appendix to this Report. Essentially, Howard and Spencer collectively offer the following reasons for recording only $240 million in catch-up expense in 1998: - The large 1998 catch-up was caused by the fact that interest rates had dropped by the largest amount in any three month period in the preceding nine years, which management believed would not occur;(268) - ---------- (265) 1998 Annual Report at 21. (266) Id. at 22. (267) See generally Spencer Submission at 4-7; Howard Submission. (268) Howard Submission at 5-6, 8. 63
- Management expected interest rates to rise, and the Company's exposure to interest rate volatility was not "symmetrical," that is, even a small increase in interest rates was expected to dramatically reduce the catch-up expense, whereas a further decline in interest rates would not cause the catch-up to grow materially; - The catch-up associated with the non-core book, $298 million, was imprecise and believed to be overestimated;(269) and - The $439 million number was imprecise because, among other things, Fannie Mae's then existing models were imprecise.(270) As a preliminary matter, we note that these explanations, if credited, may explain why management doubted the accuracy of the full $439 million in catch-up expense for 1998; however, the explanations do not support management's basis for selecting $240 million. In other words, there is still no explanation as to why $240 million in expense was considered by management as the "best estimate." In our view, evidence outlined above strongly supports the conclusion that management selected $240 million because it was the most catch-up expense it could record, and, considering the other adjustments, still achieve the EPS target of $3.23. Moving to the substance, our view is that the reasons proffered by Howard and Spencer fail to support their decision to record only $240 million in catch-up expense for 1998. We address each argument in turn. 1. Interest Rates Volatility Contributed to the Large Catch-Up First, management's contention that the large catch-up number in 1998 was due to faster prepayments resulting from the unusually large drop in interest rates at the end of 1998,(271) is inconsistent with our review of historic interest rates. Our review of the average monthly rates for thirty-year fixed-rate mortgages (referred to as the "thirty-year FRM"), shows that the thirty-year FRM dropped from 7.00% in June 1998 to 6.72% by September 1998, a drop of twenty-eight basis points.(272) However, the thirty- - ---------- (269) Id. at 6-9; Spencer Submission at 7. (270) Tr. of June 21, 2004 OFHEO Dep. of L. Spencer, at 45:4-22 (hereinafter "Spencer OFHEO Testimony"). (271) Howard Submission at 5. Howard's counsel stated that "[i]n the quarter ending September 30, 1998, interest rates dropped by the largest amount in any three-month period in nine years." See also Spencer Submission at 5. (272) Based on KPMG workpapers, it appears that these rates are the "30-year FRM" rates identified on catch-up schedules created by Fannie Mae. Mem. from Harry Argires to Fannie Mae Working Papers file, datede Apr. 11, 2000; Fannie Mae PDA Catch-Up, dated Apr. 10, 2000, FMSE 104860; 30-Year Fixed Rate Mortgages 64
year FRM dropped from 8.23% in September 1996 to 7.60% in December 1996, a drop of sixty-three basis points.(273) Moreover, the acceleration of prepayments due to declining interest rates was only part of the reason for the large catch-up number in the fourth quarter of 1998. As discussed supra in Subsection II.C.1., management had never recorded catch-up prior to 1998, and documents indicated that management probably had understated catch-up expense in the periods before 1998 because it had used interest rate paths to estimate prepayments that were higher than the prevailing market interest rates in a generally declining rate environment. Also, as discussed supra in Subsection II.C.2., an even bigger component of the $439 million ($298 million) was catch-up expense related to the level yield modeling of a portion of the non-core book which had only begun in 1998.(274) Accordingly, the large catch-up as of the end of 1998 was not just a result of a decline in interest rates during the year. 2. Management's View That Interest Rates Would Rise in the Future Management also contends that based on what happened in 1999 - that is, interest rates did rise and the negative catch-up did flip to a positive catch-up - they were justified in believing the $439 million number was too high.(275) However, the change to a positive catch-up in 1999 was not just based on a rise in interest rates, but was mostly a result of the $158 million of catch-up expense that management recorded through manual on-top entries.(276) Had management not recorded these on-top adjustments in 1999, the catch-up in 1999 would have remained in a deferred expense position. Management told KPMG in January 1999 that the 6.75% flat rate path used in the 1998 catch-up calculation was not the most appropriate rate path.(277) Management told KPMG that rates were volatile and likely to increase over the next year, and therefore the catch-up generated from the 6.75% flat rate path - which was the base interest rate path used at that time for other purposes at Fannie Mae - was not the best - ---------- Since 1971, available at http://www.freddiemac.com/pmms/pmms30.htm (last visited Feb. 12, 2006). (273) 30-Year Fixed Rate Mortgages Since 1971, available at http://www.freddiemac.com/ pmms/pmms30.htm. (274) Fannie Mae Catch-Up Summary, dated Jan. 7, 1999, FMSE-IR 182475. (275) Howard Submission at 9. (276) Fannie Mae PDA Catch-Up, dated Apr. 10, 2000, FMSE 104860; Fannie Mae 1999 Additional Amortization Booked On-Top Rate Schedule, FMSE-SP 29570; 1999 Monthly Summary of Amortization Adjustments, FMSE-IR 176231. (277) See Accounting Issues Discussion. 65
estimate.(278) Management therefore calculated the catch-up in the fourth quarter of 1998 using a rate path of 7.00% for six months and 7.25% thereafter.(279) Management considered these rate paths to be the "up" scenario.(280) In this scenario, the unrecognized portion of catch-up expense in December 1998 would have been reduced from $199.7 million to $65.9 million, and eventually would have become an unrecognized income position by December 1999.(281) In addition, management created a "down" scenario for interest rates using a rate path of 6.25% for six months and 6.75% thereafter.(282) In this scenario, the unrecognized portion of catch-up expense in December 1998 would have increased from $199.7 million to $234.7 million.(283) Therefore, the "down" scenario caused only a $35.0 million increase to unrecognized catch-up expense, as opposed to the corresponding "up" scenario that caused an asymmetric $133.8 million decrease to unrecognized catch-up expense.(284) We note that a straight-forward approach to test the asymmetrical exposure theory would have been to create "up" and "down" interest rate scenarios that were symmetric around the base interest rate path. Instead, the adjustments management made to interest rates in the "up" and "down" scenarios in its analysis were not symmetric, as noted above. In the "up" scenario, management assumed that interest rate would increase twenty-five basis points (from 6.75% to 7.00%) for the subsequent six-month period and then increase another twenty-five basis points (to 7.25%) and remain at that rate indefinitely. In the "down" scenario, management assumed that interest rates would decrease by fifty basis points (from 6.75% to 6.25%) for the next six months and then increase fifty basis points (to 6.75%) and remain at the then current base interest rate path of 6.75% indefinitely. Accordingly, we question the validity of management's quantification of the asymmetric exposure. Other documents indicate that management did create additional interest rate scenarios, including a "down" scenario that was the symmetric counterpart to the - ---------- (278) Accounting Issues Discussion; Tr. of June 21, 2004 OFHEO Deposition of L. Spencer at 44:25-45:22; Draft Letter from J. Boyles to SEC, dated Oct. 8, 2004, FMSE-IR 380348-49, at FMSE-IR. (279) Accounting Issues Discussion. (280) Id. (281) Id. (282) Id. (283) Id. (284) Id. 66
"up" scenario described above.(285) This "down" scenario assumed that interest rates would decrease twenty-five basis points (from 6.75% to 6.50%) for the next six months and then decrease by another twenty-five basis points (from 6.50% to 6.25%) and remain at that rate indefinitely.(286) The calculation of the catch-up estimate based on this "down" scenario indicated that the catch-up would increase more in the "down" scenario than it would decrease in the "up" scenario.(287) In other words, based on this analysis, management's aforementioned conclusions regarding the asymmetric exposure of the 1998 catch-up would appear erroneous. Even aside from the validity of this analysis, the claim that the decision to record only $240 million of catch-up expense was supported by the "up" scenario is not accurate because in the "up" scenario, the calculated catch-up was $305.9 million, not $240 million.(288) If management genuinely believed that the "up" scenario accurately reflected future prepayments, according to its own logic, management still under-recorded catch-up expense based on this scenario by $65.9 million.(289) In fact, management's decision to record $240 million of catch-up expense is consistent with none of the three interest rate scenarios. No reasons were offered as to why management did not record catch-up in line with its own analysis. 3. Imprecision of the Modeling for the Non-Core Book Another reason management offered for its belief that the $439 million catch-up expense was too high was that the Company's FAS 91 modeling and other - ---------- (285) See Catch-Up Summary: Scenario 2 as of Dec. 31, 1998, dated Jan. 13, 1999, FNMSEC 2678. (286) See id. (287) See id. The analysis of this additional "down" scenario only contemplated core book (or PDAMS) catch-up. In this scenario, the unrecognized portion of catch-up expense in December 1998 would have increased from $81.7 million to $274.3 million. Therefore, this "down" scenario caused a $192.6 million increase to unrecognized catch-up expense, as opposed to the corresponding "up" scenario that only caused an asymmetric $66.8 million decrease to unrecognized catch-up expense. (288) See Howard Submission at 8; Accounting Issues Discussion. (289) In addition, the Spencer Submission states that the results of the "up" and "down" scenarios created a "band within which management had confidence that prepayments would be estimable and probable of occurring." Spencer Submission at 7. 67
systems were inadequate.(290) At his OFHEO interview and in the Howard Submission, Howard offered several reasons for the belief that the catch-up calculation for the non-core book as of the fourth quarter of 1998, in particular, was an overestimate.(291) Howard further testified before OFHEO that this belief was validated when Fannie Mae "trued up" the non-core book in 1999 or 2000.(292) Howard may have been referring to the fact that when the Company increased the percentage of the REMIC book that was modeled under the level yield method in May 2000 (when approximately ninety percent of the REMIC book was being modeled), and October 2000 (when essentially all of the REMIC book was modeled), it resulted in approximately $83 million(293) of additional catch-up income. However, this assertion ignores what happened - ---------- (290) Spencer Submission at 4; Howard Submission at 2-3; Tr. of Aug. 5, 2004 OFHEO Dep. of T. Howard, at 211:14-213:4 (hereinafter "Howard OFHEO Testimony"); Spencer OFHEO Testimony at 45:4-22. (291) Howard OFHEO Testimony at 212:12-213:25; Howard Submission at 9. Howard stated that his belief was based on the fact that: (1) not all of the REMICs and synthetic REMICs had been modeled at this time and many of the securities not yet modeled were believed to be "net discounts," which management believed, if modeled, would have reduced the catch-up associated with the non-core book; and (2) management had been conservative in its estimation of prepayment sensitivities of certain REMIC tranches since it did not have data on the cash flow characteristics of all tranches of the REMIC securities it owned. Id. (292) In particular, during his August 2004 OFHEO testimony, Howard stated: When we were building up to this $400 million number, it contained an estimate of the amount of expense we thought we needed to take to book discount expense on the REMIC book. That dollar amount, the negative, which I don't recall, turned out once we got actual detail [in 1999 or 2000] to have been overstated as a component of the $400 million. I don't recall by how much, but it was enough to be certainly noticeable. Howard OFHEO Testimony at 212:16-25. (293) In May 2000, the REMIC-related catch-up changed by $69 million from a deferred expense of $48 million to a deferred income of $21 million, as a result of modeling an increased percentage of the non-core book using the level yield method and a correction of an INTEX data error. The INTEX data error accounted for $58 million of the $69 million change. Mem. From J. Juliane and R. LeRouzes to Distribution, dated Apr. 12, 2000, FMSE-SP 303-04; Mem. from J. Juliane and R. LeRouzes to Distribution, dated May 23, 2000, FMSE-SP 299-300. Undated REMICs Catchup, FMSE-IR 55534; Mem. From Harry Argires to Fannie Mae Working Papers File, dated July 11, 2000. (A full description of INTEX is contained infra in the Appendix to this Chapter that discusses the Company's FAS 68
in 1999: when the percentage of the REMIC book that was modeled increased to sixty-four percent as of August 1999, it resulted in approximately $75 million of additional catch-up expense.(294) Moreover, Howard's statement to OFHEO ignores the fact that the interest rate environment in 2000 was significantly different from the interest rate environment in 1998. Interest rates increased from 6.74% at the end of 1998 to 8.21% at the beginning of 2000.(295) Comparing the results of a catch-up calculation prepared in October 2000 to the results of a catch-up calculation prepared in December 1998, given the differences in the interest rate environments between the two periods, is like comparing apples and oranges.(296) In addition, Howard's assertion ignores the fact that the 2000 catch-up calculation incorporates post-1998 acquisitions and excludes REMICs that were part of the portfolio in 1998 but had been liquidated prior to this subsequent modeling. We conclude that the appropriate manner in which to analyze the estimated impact of the unmodeled REMIC book in 1998 is to analyze the REMIC book as of 1998. The Howard Submission's contention that the unmodeled portion of the 1998 year-end REMIC book was "believed to be" securities that were acquired at a discount(297) - meaning that they would have reduced the catch-up expense if modeled - is contradicted by documents. First, we have found no documents that support Howard's contention, and Howard has cited none. Second, documents show that as of November 30, 1998, the REMIC book contained $584 million of unamortized premium and $167 - ---------- 91-related systems.) In October 2000, the REMIC-related deferred catch-up income increased to $74 million as a result of modeling an increased percentage of the non-core book using the level yield method. Mem. from J. Juliane and R. LeRouzes to Distribution, dated Sept. 22, 2000, FMSE-IR 42932-33; Mem. from R. LeRouzes to Distribution, dated Oct. 17, 2000, FMSE-SP 511-12. When considered together, these changes increased the Company's catch-up income in 2000 by $83 million. (294) Mem. from J. Juliane and R. LeRouzes to Distribution, dated Aug. 11, 1999, FMSE-SP 615-16. (295) 30-Year Fixed Rate Mortgages Since 1971, available at http://www.freddiemac.com/ pmms/pmms30.htm. (296) Even following Howard's logic, the reduction in catch-up expense of approximately $83 million in 2000 was almost completely offset by the increase in catch-up expense of $75 million in 1999, leading to the conclusion that the $298 million catch-up expense calculated as of December 1998 could not have been materially misstated. (297) Howard Submission at 9. 69
million of unamortized discount.(298) Also, the following table breaks down the percentages of the premium and discount on REMICs that had been modeled as of May 1998 and as of August 1999:(299) Premium Discount ------- -------- May 1998 59.5% 57.3% August 1999 52.0% 46.0% Given the large disparity between unamortized premium and discount amounts and the relatively consistent percentages of premium and discount modeled in May 1998 and August 1999, it is unlikely that the unmodeled portion of the REMIC book was in a "net discount" position, and that additional modeling would have resulted in a significant offsetting adjustment to year-end 1998 catch-up. Witnesses also contradict the assertion in the Howard Submission that the large catch-up amount was the result of a "conservative" estimate of the prepayment sensitivities of certain REMIC tranches made by management because it did not have data on the cash flow characteristics of all tranches of the REMIC securities it owned. A manager in Portfolio Strategy who was a member of the "Quick Strike Task Force" that endeavored to model the Company's non-core book using the level yield method stated that management did not include any non-core book securities for which it did not know cash flow characteristics in the calculation of the non-core book catch-up. In other words, management did not include any estimate, conservative or otherwise, of catch-up associated with the non-core book securities that were not yet modeled. 4. Inadequacy of the Modeling Systems for the Core Book In addition to imprecision in the non-core book modeling, management also has pointed to weaknesses in the PDAMS model as a rationale for its belief that the $439 million in catch-up expense was imprecise.(300) This assertion does not appear to be consistent with the facts. Despite management's claims, we have not seen any contemporaneous documentation from 1998 (or associated with the 1998 year-end procedures in early 1999) that supports the contention that management had significant - ---------- (298) Acquired Premium / Discount Balances as of Nov. 30, 1998, FMSE-SP 103389. (299) Draft of March 1998 REMIC Premium and Discount, dated May 28, 1998, FMSE-IR 26684-85; Mem. from J. Juliane and R. LeRouzes, dated Aug. 11, 1999, FMSE-SP 615-16. (300) Spencer Submission at 4; Howard Submission at 2-3; Howard OFHEO Testimony at 211:14-23:4. 70
concerns at that time about PDAMS' ability to provide reasonably accurate and reliable amortization results.(301) F. FINDINGS REGARDING THE 1998 CATCH-UP DECISION We conclude, based on this evidence, that the decision to record only $240 million of the $439.7 million catch-up expense for 1998 was not in accordance with GAAP, and was motivated by management's desire to meet 1998 EPS and AIP targets. While interest rate volatility and system imprecision may have played some role in creating doubt in management's mind about the accuracy of the $439 million number, the evidence surrounding how management selected $240 million out of $439 million plainly demonstrated that the primary focus was on achieving 1998 EPS targets. We also conclude that Howard and Spencer were the individuals responsible for the decision to recognize only $240 million of the $439 million in catch-up expense, and that they made the recommendation to the members of the Company's Office of the Chairman at the January 8, 1999 meeting. We conclude that while the attendees of this meeting were provided information by Howard and Spencer about the full $439 million in catch-up expense, we saw no evidence that the attendees were given the Earnings Alternatives Schedules, or any information that would put into question the accuracy of Howard's and Spencer's presentation at the meeting. Our conclusion is supported by, among other things, the Earnings Alternatives Schedules, which were created at Spencer's direction shortly before the January 8, 1999 meeting, the documents Spencer identified as having been used at this meeting, and by Spencer's own statements - ---------- (301) A document in KPMG's workpapers dated July 6, 1998 that appears to relate to a meeting between certain members of management and KPMG regarding PDAMS states that "[m]odels don't predict peaks and valleys that accurately." PDAMS & Prepayment Rates, dated July 6, 1998. The document notes that Lawler, Spencer and Rajappa attended this meeting on behalf of Fannie Mae and Eric Smith, Theobald, and Russell attended from KPMG. Another document from KPMG's workpapers relating to a December 7, 1998 meeting between certain members of management and KPMG states that "Pool weighted average cots [sic] and weighted average maturities become issues that PDAMS doesn't take into account. For example, PDAMS rounds to two didgets [sic] to calculate WAC which leads to jumps in amortization when the WAC does change." Fannie Mae Purchase/Discount Amortization Discussion, dated Dec. 7, 1998. The document notes that Spencer and Boyles attended this meeting from Fannie Mae, and Theobald and Smith attended from KPMG. However, neither document necessarily supports a conclusion that the models were incapable of producing a reliable estimate of catch-up under normal circumstances. In fact, the December 1998 document specifically notes that "Jonathan Boyles has done a lot of recalculations of pools and amortization to ensure PDAMs is calculating everything appropriately." Id. 71
that it was she and Howard who ultimately became "comfortable" that $240 million was the right number, and that she and Howard persuaded Raines at the meeting in January 1999 to record only $240 million.(302) In addition, we conclude that Howard and Spencer's decision to defer the $199 million in unrecognized catch-up expense with a plan to reduce it prospectively through "on-top" adjustments was not in accordance with GAAP. FAS 91 contains no exceptions to the requirement that catch-up be recognized in the period it is calculated.(303) Moreover, management reduced the $199 million catch-up by recording $158 million of additional amortization expense during 1999, even though interest rates did increase in 1999 and models were generating positive catch-up. Documents show that management recorded these additional amounts as a "reserve" against future interest rate changes.(304) Recording expense to create a "reserve" or for "protection" against future events that may or may not occur is not in accordance with GAAP. We conclude that management failed to comply with FAS 91 prior to 1998 by failing to record any catch-up and failing to apply the interest method for REMICs and synthetic REMICs. Management also failed to comply with GAAP by not performing any tests to determine if the results of the proxy amortization methods rendered materially different results from the interest method. We also conclude that the decision to change the LIHTC accounting method from cash to accrual was appropriate, but find that management's motivation for instituting the accounting change for 1998 was to make up for the earnings shortfall caused by recognizing the $240 million catch-up expense for 1998. Furthermore, we conclude that while management disclosed its full catch-up amount and its decision to book only part of it to KPMG, it failed to disclose the same information to the Board. Management also failed to disclose KPMG's audit difference to the Board until 2000. In addition, management's disclosures to the Board and to the public regarding the LIHTC accounting change were misleading because they did not disclose the true reason for the accounting method change, which was simply going from a non-GAAP accounting method to a method that complied with GAAP. We also find - ---------- (302) Earnings Alternatives Schedules; Fannie Mae Income Statement, dated Jan. 7, 1999, FMSE-IR 182474; Fannie Mae Catch-Up Summary, dated Jan. 7, 1999, FMSE-IR 182475; Low Income Housing Tax Credit Investments, dated Jan. 7, 1999, FMSE-IR 182476. (303) In addition, the decision to recognize any portion the $199 million, much less $158 million of the total, through on-tops in 1999 undercuts management's assertion that the $439 million in catch-up was an unreliable number. It also undercuts management's assertion that the systems were unreliable. (304) See, e.g., 1999 Monthly Summary of Amortization Adjustments, dated Jan. 28, 2000, FMSE-IR 176231-32. 72
that management should have highlighted to the public that this resulted in a one-time event of recognizing two years' worth of credits for 1998, which information would have permitted investors to assess the quality of the Company's earnings. Finally, we find that there is no evidence that management disclosed to the Board the fact that the 1998 EPS was increased by the unsupported decision to recognize $3.9 million in aged credits from the 16220 suspense account in income. IV. FANNIE MAE'S PREMIUM AND DISCOUNT AMORTIZATION POLICY This section addresses the Company's development and implementation of the Amortization Policy. In January 1999, after noting an audit difference for the unrecognized catch-up expense of $199 million, KPMG requested management to develop a policy formalizing the Company's FAS 91 accounting practice. Management agreed, and nearly two years later, in December 2000, implemented the Amortization Policy. Based on our review of the development process, the Amortization Policy itself, and the manner in which the Policy was implemented, we have formed the following conclusions: - - the Amortization Policy was not consistent with GAAP; - - the Amortization Policy was developed by the Controller's Office under the oversight of Howard, and was developed for the primary purpose of managing the income statement volatility that may result from recognizing catch-up adjustments on a regular basis, and to avoid audit differences; - - management implemented the Amortization Policy inconsistently with the letter of the Policy, and at times recorded catch-up adjustments even though the calculation yielded an amount within the Policy threshold, which supposedly represented the "functional equivalent of zero"; and - - management failed to fully disclose the Policy or its manner of implementation to the Board of Directors. A. DEVELOPMENT OF THE POLICY 1. Practice Prior to 1999 As noted above, prior to 1998, management did not recognize catch-up under FAS 91. It appears that in the same period, management had an informal understanding that KPMG would not require management to record catch-up if the amount fell within a threshold of plus or minus $100 million, even though it may treat the 73
unrecognized catch-up as an audit difference.(305) In addition, management considered recording catch-up "prospectively," that is, to estimate and recognize small catch-up amounts in periods subsequent to the measurement of catch-up in order to avoid having to recognize a large catch-up amount in any one period.(306) Documents showed that as early as November 1995, management considered using a threshold or "de minimis catch-up level" for catch-up, and considered recognizing catch-up prospectively.(307) A document discovered in Financial Standards' files stated: "Since true FAS-91 would introduce excessive variance in income recognition from quarter to quarter, we have adopted a `prospective method.'"(308) Documents found in Spencer's files reflected a belief that, as late as 1998, catch-up below $100 million was considered desirable because KPMG would not require Fannie Mae to book such an amount,(309) but management understood that KPMG would propose an audit difference even if the catch-up was less than $100 million. Russell from KPMG did not recall being aware of any informal policy at Fannie Mae pursuant to which management did not recognize catch-up if the amount was within a $100 million threshold. However, documents showed that KPMG was aware that management did not book any significant catch-up adjustments prior to 1998 and noted audit differences in 1996 and 1997 for unrecorded catch-up.(310) 2. Development of the Amortization Policy In early 1999, after management explained its rationale for recording only $240 million in calculated catch-up expense for 1998, KPMG asked management to develop a formal policy for accounting for premium and discount amortization under FAS 91. Pennewell in the Controller's Office led the process, and others who participated in the development of the Policy included Spencer, Lawler, Juliane and Rene - ---------- (305) See, e.g., Undated Proposed Amortization Policy, FMSE-IR 34793-94; Mem. from J. Pennewell to T. Howard, dated Sept. 23, 1999, FMSE-IR 34834-37; Suggested Amortization Strategy, dated Apr. 2000, FMSE-IR 34843-44. (306) Suggest Amortization Strategy, dated Apr. 2000, FMSE-IR 34843. (307) PDAMS Notes, dated Nov. 27, 1995, FMSE-IR 214738-40, at FMSE-IR 214739. (308) PDAM Analysis and Recommendation, dated Nov. 7, 1995, FMSE-IR 214741-43, at FMSE-IR 214741. See, e.g., PDAMS Notes, dated Nov. 27, 1995, FMSE-IR 214738-40, at FMSE-IR 214739. (309) Risk Review with CFO at FMSE-IR 321567. (310) Fannie Mae Audit Differences Dec. 31, 1997, FMSE-IR 22058; Fannie Mae Audit Differences Dec. 31, 1996, FMSE-IR 22059. 74
LeRouzes. In addition, as described in more detail infra in Subsection IV.A.3., Howard was very much involved in the development of the Policy, expounded some of its key components, and had the overall approval authority over the Policy. Individuals from KPMG were consulted periodically in the course of developing the Policy. Unlike other accounting policies at Fannie Mae, it was the Controller's Office, and not Financial Standards, that took the lead in the development of the Amortization Policy. Spencer and other witnesses have tried to explain this anomaly by characterizing the Amortization Policy as an "operational framework," not an "accounting policy."(311) Boyles confirmed that Financial Standards, the department at Fannie Mae that was charged with issuing accounting policies, was not required to review and approve the Amortization Policy because it was not an official "policy." Although Boyles is listed as a carbon copy recipient on certain early drafts of the Amortization Policy and may have been consulted from time to time by those drafting the Policy, we have seen no evidence that Financial Standards or Boyles authorized the Policy that was ultimately adopted. 3. Evidence Reflecting Management's Purpose for the Amortization Policy Drafts of the Policy, from 1999 through the final draft in December 2000, suggested that management was focused primarily on articulating its "practice" with respect to catch-up calculation that could be presented to KPMG as evidence of an objective approach.(312) Most of the early drafts of the Policy contain the following themes: - minimizing "earnings volatility"; - using an average of multiple interest rate scenarios around the Company's base interest rate forecast to determine the catch-up for a given period; - ---------- (311) Spencer could not provide a legitimate explanation of the differences between an "operational framework," and an "accounting policy," or offer any other incidents where the Controller's Office drafted and utilized an "operational framework" separate and apart from Financial Standards. (312) Mem. from L. Spencer and J. Pennewell to T. Howard, dated Mar. 2, 1999, FMSE 217570-71; Mem. from J. Pennewell to T. Howard, dated Sept. 23, 1999, FMSE-IR 34834-37; Purchase Premium and Discount Amortization Policy, dated Dec. 2000, FMSE-IR 55470-71 (hereinafter "Amortization Policy"). 75
- developing a predetermined threshold, targeted value or "cushion," below which the Company would consider the catch-up to be "immaterial"; - managing the catch-up so that it would not exceed that targeted value; and - reducing the catch-up balance over a period in the future if it exceeds a certain amount.(313) None of the above concepts is based on FAS 91. Rather, the concepts reflect the twin goals of reducing earnings volatility and giving management as much discretion as possible in managing the impact of catch-up on the income statement. Documents also showed that management was well aware of the requirements of FAS 91 as it developed the Policy. For example, a draft policy dated September 23, 1999 that was prepared by Pennewell showed that management was aware that FAS 91 required catch-up to be recorded in the current period, but nevertheless believed that a "materiality threshold" could be a rationale for avoiding this requirement: Accounting policy requires us to record any catch-up to adjust prior period amortization in the current period. However, we believe it may be appropriate to adjust the catch-up over time if the amount is not material.(314) In the same memorandum, Pennewell discussed the requirements of Staff Accounting Bulletin No. 99, Materiality, (August 12, 1999) ("SAB 99"), and noted that the SAB reflects "the SEC's concern about the tendency of companies, with the acquiescence of their auditors, to manage earnings by designating certain transactions or events below a certain percentage threshold as `immaterial.'"(315) In our view, this memorandum supports the conclusion that management considered the prospective treatment of catch-up, notwithstanding its awareness that it did not conform with GAAP. Two additional concepts that management considered in certain drafts were treating catch-up that resulted in income ("positive catch-up") differently than catch-up that resulted in expense ("negative catch-up"); and targeting a specific catch-up balance in order to provide a "cushion" or "reserve" in the event of future interest rate changes that could require a catch-up adjustment to be recorded in the future. - ---------- (313) See, e.g., Mem. from J. Pennewell and J. Juliane to T. Howard, dated Dec. 16, 1999, FMSE-IR 34848-50; see also Amortization Policy, FMSE-IR 55470-71. (314) Mem from J. Pennewell to T. Howard, dated Sep. 23, 1999, FMSE-IR 34834-37, at FMSE-IR 34836. (315) Id. 76
For example, the September 23, 1999 draft policy contemplated that, if the catch-up amount was negative (expense), the Company should recognize the expense prospectively over a four-year period.(316) The same draft recommended taking no action if the catch-up was positive (income) and management's assumption was that interest rates would fall in the future.(317) Another draft policy contained the following rationale for the differing treatment of negative and positive catch-up: Given the asymmetric nature of our catchup - the catchup usually becomes more negative when rates go down [than] positive when rates go up, following the asymmetric response of prepayments to changes in interest rates - this recommendation would imply that our target catchup level will usually be somewhat positive.(318) The concept of targeting a specific catch-up balance was suggested by Howard to Pennewell, in a memorandum dated October 27, 1999. Howard stated, as an item for Pennewell to consider as she finalized the policy: ". . . `where would we like our [catch-up target] to be' in order to minimize the amount and frequency of PDA [purchase/discount amortization] adjustments we would have to make as interest rates moved [sic] (as we know they will)."(319) This memorandum further discussed targeting a particular catch-up balance that would shield the Company from having to record adjustments in the event of interest rate changes.(320) In other words, management would target a particular catch-up balance in anticipation of interest rate volatility, which targeted balance would serve as a "cushion" or "reserve" against future volatility. (321) Additional evidence of management's perspective on recording catch-up is reflected in the documents relating to the implementation of BancWare. In 1999, at Juliane's suggestion, management considered replacing PDAMS with BancWare,(322) - ---------- (316) Id. at FMSE-IR 34835. (317) Id. (318) Suggested Amortization Strategy, dated Apr. 2000, FMSE-IR 34843-44, at FMSE-IR 34843. (319) Mem. from T. Howard to J. Pennewell, dated Oct. 27, 1999, FMSE-IR 34847. (320) Id. (321) See, e.g., id.; Suggested and J. Juliane Amortization Strategy, dated Apr. 2000, FMSE-IR 34843-46; Mem. from J. Pennewell to T. Howard, dated Dec. 16, 1999, FMSE-IR 34848-50. (322) A full description of BancWare is contained infra in the Appendix to this Chapter that discusses the Company's FAS 91-related systems. 77
which management believed would allow it to "manipulate" factors to produce an "array of recognition streams" and thus "`bleed' the catch-up within a specified time period."(323) Juliane stated that BancWare had the capacity to target a specific catch-up amount and to plan its reduction over a period of time. According to a memorandum written by Juliane dated March 12, 1999, these capabilities "strengthen[ ] the earnings management that is necessary when dealing with a volatile book of business."(324) BancWare was in operation for a brief period as a supplement to PDAMS, but, ultimately, it was not fully implemented by the Company because of implementation difficulties and management's belief that it could build a better system.(325) Witness statements and documents also showed that management wanted a Policy that would help avoid "artificial volatility" in the Company's financial statements and permit the recording of catch-up into income prospectively over a period.(326) The latter practice appears to have been accepted by KPMG. Documents found in KPMG's workpapers showed that Howard discussed the Company's then growing catch-up number with Russell of KPMG, who was then the lead engagement partner, during an August 11, 1998 meeting.(327) A memorandum in KPMG's workpapers about the meeting reflected the following advice from Russell to Howard: I told Tim that in regard to the FAS 91 amortization of premiums and discount relating to its loan portfolio, that we agree that significant quarterly adjustments would not be the best way to react to changes in prepay rates, however, possibly more frequent prospective adjustments should be made by Fannie Mae to keep the FAS 91 difference at a lesser significant level. He appeared to agree that there should be some level of materiality established and the Company should look into how they could stay within that level. The increase [sic] frequency seems to be a consideration that he found to be acceptable. He was going - ---------- (323) Mem. from J. Juliane to T. Lawler, L. Spencer, et al., dated Mar. 12, 1999, FMSE-SP 364-67, at FMSE-SP 365. (324) Id. (325) Juliane OFHEO Testimony at 217-18. (326) See, e.g., Mem. from T. Howard to J. Pennewell, dated Oct. 27, 1999, FMSE-IR 34847. (327) Mem. from K. Russell to J. Theobald and E. Smith, dated Aug. 25, 1998. 78
to check with Leanne as to the possibilities of establishing some policy like that.(328) Thus, it appears Howard consulted at least Russell in 1998, and Russell agreed with a prospective treatment of catch-up, provided that management establish a "level of materiality" to limit the amount of current period unrecorded catch-up. We saw no evidence that Howard had any subsequent discussions about this issue with any subsequent engagement partners at KPMG. On the concept of avoiding "artificial volatility," this directive was most clearly reflected in Howard's October 27, 1999 memorandum to Pennewell. In this memo, Howard explained how he thought management could use certain practices to "help with our goal of minimizing unnecessary net income volatility," and suggested: to make our corporate sensitivity to these rate changes symmetric, I figured that as long as our PDA sensitivity looked like this we should aim for a PDA catchup position of a positive $85 million. (That way, if rates moved up by 2 standard deviations the catchup would rise by $80 million to a positive $165 million; if rates fell by 2 standard deviations our catchup would decline by $250 million, to a negative $165 million. That's symmetric.) . . . . . . . [G]iven our current projected PDA sensitivity we want to be at $85 million; our probability-weighted expectation is that we will be at $160 million over the next three years, so to get where we want to be we should amortize the difference ($75 million) evenly into income over the next three years (at $25 million per year).(329) Howard thus suggested to Pennewell that in a situation where management expected its catch-up to increase from its target balance of $85 million to $160 million in three years, the Company would manage its target balance by recording catch-up adjustments of $25 million per year, which would help the Company avoid having to recognize a large catch-up adjustment all at once. Another e-mail drafted by Howard to Pennewell, copying Spencer and Juliane, on or about November 25, 2000, further demonstrated his critical role in developing the Policy.(330) The attachment to the e-mail was created by Howard in - ---------- (328) Id. (329) Mem. from T. Howard to J. Pennewell, dated Oct. 27, 1999, FMSE-IR 34847. (330) E-mail from T. Howard to J. Pennewell, L. Spencer and J. Juliane, dated Nov. 25, 2000, FMSE-IR 660837-39, at FMSE-IR 660837. 79
advance of a meeting with KPMG, with the purpose of summarizing his "objectives and point of view on this project."(331) The notes outlined three provisions that were ultimately included in the final Policy: (1) calculating the catch-up using the average of five scenarios; (2) a range within which the catch-up was to be considered the "functional equivalent of zero"; and (3) a time horizon over which the Company could bring the catch-up into income.(332) The notes further proposed that the Company could prospectively "get a leg up" on getting the catch-up below the predetermined threshold for future years by making prospective on-top adjustments.(333) With regard to justifying the predetermined range, Howard stated "I'm pretty comfortable that we can defend the range around our target catch-up by linking it to the uncertainties about the assumptions used to calculate the catch-up figure."(334) Pennewell stated that she and others, including Spencer and Boyles, were not comfortable that a Policy that allowed the Company to bring the amount of catch-up in excess of the threshold into income over a long period of time was consistent with GAAP, and that she and others communicated their concerns to Howard. While Pennewell could not recall any specific discussions she had with Howard, she generally recalled telling Howard that KPMG would not be comfortable with the idea. A draft policy prepared by Pennewell supported her recollection that she did, in fact, express to Howard that KPMG would not approve of using an extended time horizon to bring certain levels of catch-up into income.(335) Notwithstanding Pennewell's and others - ---------- (331) Id. at FMSE-IR 660838-39. (332) Id. (333) Id. at FMSE-IR 660839. (334) Id. (335) Undated Proposed Amortization Policy, FMSE-IR 463599-600. The draft states "we expect them [KPMG] to like the zero, but not the time horizon. They would prefer a time horizon of the current year, with something like a +/- $100 million as an acceptable range." Pennewell's recollection of discussing her concerns with Howard was further corroborated by Howard's critical view of Pennewell's contributions to the development of the Policy. In a performance evaluation he wrote regarding Spencer in 2000, Howard said: Purchase discount has taken a very long time to come to the point of a concrete recommendation, and I thought we went down some unproductive avenues to get where we now are. You'll recall that I wasn't totally convinced of the merits of the recommendation Janet made on this issue a couple of months ago, but was willing to agree to it provided she accepted the condition that if in a future instance her policy resulted in a challenge for us it was clearly her responsibility. 80
concerns, the concept of reducing the catch-up over a three-year horizon was incorporated into the final Policy.(336) According to Pennewell, however, Fannie Mae never put this part of the Policy into practice. 4. KPMG's Involvement in the Development of the Amortization Policy Management met with KPMG prior to finalizing the Amortization Policy. While we have not seen a copy of the final Amortization Policy in KPMG's files, we have seen an undated draft of the Policy.(337) An undated document prepared by Pennewell showed management's expectation that KPMG would challenge certain aspects of the policy, and included defenses to these potential challenges.(338) For example, the document reflected management's anticipation that KPMG would not approve of using as wide a range of interest rate scenarios as management wanted to use (120 basis points) and that KPMG would want management to reduce the catch-up, to the extent it was outside of the plus or minus $100 million threshold, in the current year as opposed to over a three-year horizon.(339) As discussed in more detail infra in Subsection IV.B.4, we find that KPMG was aware of and did not disagree with the Policy's provisions allowing management to use the average of five interest rate scenarios for calculating catch-up and a predetermined threshold in which no catch-up would be recorded. 5. The Final Amortization Policy Adopted in December 2000 Management finalized a written purchase premium and amortization policy in December 2000.(340) The December 2000 Policy contained three main sections that addressed the following goals: (1) to calculate the current catch-up position; (2) to determine a "target range" for the catch-up that reflects "existing assumption risk"; and - ---------- Mem. from T. Howard to L. Spencer, dated Dec. 9, 2000, FMSE-IR 28779-81, at FMSE-IR 28780. (336) Amortization Policy, FMSE-IR 55470-71. (337) Undated Purchase Premium and Discount Amortization Policy. (338) Undated Proposed Amortization Policy, FMSE-IR 34793-94. This document was produced from the files of Pennewell on December 13, 2004. It was part of binder entitled "FAS 91 Binders - Amortization Results 2000 - Current (12/13/04)." (339) Id. at FMSE-IR 34793. (340) Amortization Policy. Spencer, Boyles, Pennewell and Juliane identified this document as the Policy. Tr. of June 15, 2004 OFHEO Dep. of J. Pennewell at 33:11-34:7. 81
(3) to "manage the catch-up over time."(341) By "existing assumption risk," management meant modeling imprecision due to the uncertainty related to forecasting interest rates and the imprecision of estimating prepayments based on its interest rate forecast, as well as other modeling "compromises" that existed due to systems limitations.(342) The Policy also outlined when audit differences would and would not be posted.(343) (a) Average of Five Interest Rate Scenarios The first provision of the Policy stated that the average of five interest rate scenarios are to be used to calculate the catch-up, namely "the base rate together with the four sensitivity runs that encompass a 95% confidence interval around that base rate."(344) In the November 3, 2004 letter to the SEC's Office of the Chief Accountant ("OCA") explaining the Amortization Policy, the Company stated that multiple rate paths were used to capture the asymmetries in how the catch-up might react to interest rate changes and that five interest rate scenarios were used to capture the likely interest rate fluctuations consistent with a ninety-five percent confidence interval.(345) Our investigation has not uncovered, and witnesses were unable to point to, any contemporaneous analysis that supported the Company's determination that the estimate of the five interest rate scenarios would capture the future interest rate path with ninety-five percent confidence. (b) Precision Thresholds The second provision of the Policy set the target catch-up at plus or minus one percent of related revenue for net interest income and plus or minus two percent for guaranty fees, and stated that catch-up in this target range or "precision threshold" is "the functional equivalent of zero," meaning, all numbers within the range can be viewed as zero, and no catch-up would be required to be recorded.(346) The Company's November 3, 2004 letter to the OCA stated that management developed these ranges by measuring the imprecision in its prepayment projections by comparing the impact on its catch-up estimates to a range of prepayment forecasts from Wall Street firms.(347) Again, we found - ---------- (341) Amortization Policy. (342) OCA Letter at FNMSEC 153-55. (343) Amortization Policy at FMSE-IR 55470. (344) Id. (345) OCA Letter at FNMSEC 154. (346) Amortization Policy at FMSE-IR 55470. (347) OCA Letter at FNMSEC 155. 82
no contemporaneous evidence to show that management did, in fact, conduct such an analysis. Further, the OCA letter stated that the precision threshold was designed with three objectives in mind: (i) "to deal with the degree of imprecision associated with forecasting prepayments"; (ii) "to objectively limit the amount of imprecision that the company would tolerate based on pre-established thresholds"; and (iii) "to ensure that the estimation process was done in a consistent and objective fashion."(348) We do not believe that such objectives are consistent with FAS 91, and in any event, the evidence showed that management's implementation of the Policy plainly did not follow such stated objectives.(349) Finally, if the catch-up fell within the precision threshold, then "no audit difference would be posted."(350) If the catch-up for either net interest income or guaranty fees exceeded the respective precision threshold, then an audit difference for that line item would be posted for the amount in excess of the target.(351) - ---------- (348) Id. at 154. Notably, the Amortization Policy accounted for the imprecision in forecasting prepayment speeds twice - first when calculating the catch-up using an average of five interest rate scenarios and again when considering the "precision threshold" to determine whether or not the calculated amount needed to be recorded. (349) What's more, the stated rationale for such a "precision threshold," that prepayment forecasts were "imprecise," was contradicted by employees outside the Controller's Office who had expertise in such a matter. For example, Boyles stated that it was his belief that management not only had the ability to reasonably estimate prepayments, but that Fannie Mae considers its ability to predict prepayments to be a "core competence" of the Company. Additionally, an employee who was responsible for the prepayment modeling systems in Portfolio Analytics characterized Fannie Mae's ability to model CPRs as "state of the art." (350) Amortization Policy at FMSE-IR 55470. The Policy also described a "de minimis amount" beyond the predetermined thresholds within which no audit differences would be posted. This "de minimis amount" appeared to be an additional measure of materiality; however, we have seen no indication that management considered the "de minimis amount" in practice. The Policy defined the de minimis amount to be "the lesser of an adjustment: that impacts pretax Earnings Per Share by 1 cent, an adjustment that causes the net interest margin to change by 1 basis point or an adjustment that causes the Guaranty Fee Rate to change by [0].1 basis point, as applicable." Id. (351) Id. 83
(c) Managing Catch-up Over a Three-Year Period The third section of the Policy outlined a method for managing the catch-up over a three-year time period.(352) The policy stated that when catch-up exceeded the target threshold, management should aim for catch-up within the range based on future forecasts so that the catch-up will be within the "functional equivalent of zero" range within the three years.(353) The Policy suggested that management would achieve these goals by making on-top adjustments, and that the on-tops would generally be front-loaded, but would be made on no specific timetable other than three years.(354) Controller's Office personnel stated that, notwithstanding the written document, the third provision of the Policy was not implemented. Both Spencer and Pennewell conceded that the third section of the Policy had no support in GAAP, and that the only reason it was in the Policy was to appease Howard.(355) * * * * In summary, as discussed above, the Policy merely outlined the numerous ways - in calculating catch-up, in assessing what amount should be recognized, and in delaying the recognition of catch-up over a period of three years - through which management could retain significant discretion to avoid recording catch-up adjustments, which would mitigate the impact of recording catch-up on the Company's financial statements and avoid audit differences. B. The Application of the Amortization Policy The evidence of the manner in which management implemented the Amortization Policy further demonstrated that, notwithstanding the Policy, management's primary goal was to manage the catch-up's impact on the Company's income statements. - ---------- (352) Id. (353) Id. at FMSE-IR 55471. (354) Id. (355) Juliane said that KPMG told management in the first quarter of 2001 that it disapproved of this provision. Juliane also said that management always reduced the catch-up so that it was within the threshold by year-end. Boyles stated that he was not aware of this provision of the Policy until OFHEO began the Special Examination, at which point he told Spencer that this provision was not in accordance with GAAP. Boyles said that Spencer told him that he should not be concerned because this aspect of the Policy had never been implemented. 84
Management did use five interest rate scenarios for calculating catch-up, and also applied the "precision thresholds" to not record catch-up from time to time. However, evidence showed that management also recorded catch-up in ways that deviated from the terms of the Policy. In particular, management used a variety of other tools to manage the catch-up balance, including booking on-top adjustments, changing interest rate paths, and editing CPRs that fed into the modeling system. Significantly, management also recognized catch-up that fell within the "precision threshold" through 2002, even though by the Policy's own terms, any amount within the range was the "functional equivalent of zero." Finally, until 2003, management failed to calculate catch-up as of the interim balance sheet date. Instead, it used an estimate of year-end catch-up to determine whether the catch-up would fall within the "precision threshold" by year-end. If the results of that analysis indicated that catch-up would not be within the threshold, management made periodic on-top adjustments to bring the catch-up within the threshold.(356) 1. Management Used a Variety of Tools To Manage Catch-Up Management used a variety of methods not articulated in the Amortization Policy to manage Fannie Mae's catch-up position. For example, on occasion, management changed the base interest rate path used to calculate the catch-up. Juliane described at least two occasions, once in the second quarter of 2003 and once in the first quarter of 2004, when management used a base interest rate path to calculate catch-up other than the base interest rate path employed throughout the Company. In the OCA letter, the Company asserted that using a different interest rate was necessary "at times when interest rates ha[d] moved materially between the date when interest rates [were] set for the corporate business forecast and the end of the quarter."(357) The Policy, however, did not provide for such an exception. Management also booked on-top adjustments for reasons other than those described in the Policy. Juliane stated that management booked on-top adjustments when it appeared that the catch-up number was going to be outside of the plus or minus one percent range at year-end. For example, Juliane recalled that management booked a $10 million on-top adjustment in November 2002 after conducting an interim run of its year-end catch-up position and discovering that it was significantly out of compliance with the - ---------- (356) Management also engaged in a variety of other unrelated practices to mitigate the impact of the catch-up amount on the financial statements, such as including in the catch-up calculation an estimate that was generated in connection with system realignments, and adjusting the amortization end-date for its callable debt, which resulted in an adjustment to the amortization of deferred debt issue costs that substantially offset the calculated catch-up amount. We address these practices in Chapter VI, Parts M, V, and O below. (357) OCA Letter at FNMSEC 154. 85
precision threshold.(358) According to Juliane, Spencer and Pennewell were responsible for the decisions to book these on-top adjustments. We have also seen instances when management made edits to CPRs that fed into the modeling system, which in turn affected the catch-up calculation. Juliane explained that he would propose CPR edits from time to time based on his market analysis. The Controller's Office would obtain prepayment data from the Bloomberg website and create a CPR range representing the high and low market estimates based on that data. The Controller's Office would then compare the internally-generated CPRs from Portfolio Analytics to this market range and make adjustments to the selected CPRs based on what Controller's Office staff deemed appropriate.(359) Juliane stated that he would suggest CPR edits to Lawler, who would sign off on the adjustment.(360) A document dated March 15, 2003 demonstrated CPR edits that were made in this fashion.(361) Such CPR edits were only made for the use of the Controller's Office in calculating catch-up, and were neither fed back to Portfolio Analytics, nor applied elsewhere in the Company.(362) Finally, an incident that occurred in the third quarter of 2002 clearly demonstrated management's primary motivation behind managing the catch-up. At this time, management faced a positive catch-up of $80 million in order to remain in compliance with the Policy. Rather than recording the amount, management changed the methodology by which it had previously calculated catch-up, which had the effect of reducing the amount of catch-up from $80 million to $43 million. To reach this result, management included catch-up related to current quarter loan acquisitions, which it had never done in the past. Management's previous position had been that the calculation of catch-up on current quarter acquisitions was unnecessary because the rate environment during the quarter remained relatively stable. However, in the third quarter of 2002, - ---------- (358) As discussed infra in Subsection IV.B.3, we have confirmed that management booked a $10 million on-top adjustment in November 2002. However, the catch-up position by year-end was only $64.8 million, which was within the policy range of $93.3 million. Undated Catch-Up Analysis (Including Realignments), FMSE-IR 265826; 2002 Q4 Catch-up Sensitivities, FNMSEC 2869; Journal Entry, dated Dec. 9, 2002, FNMSEC 2870. (359) Juliane OFHEO Testimony at 45:20-46:6. (360) Id. at 45:20-46:10. Lawler claimed that he was not the final decision-maker as to the CPRs that were used to calculate catch-up. Lawler stated that he was never asked "what to book," only whether he believed a particular prepayment rate looked reasonable. (361) CPR Analysis, Mar. 15, 2003, FMSE-SP 2936. (362) A more detailed discussion of and our conclusions regarding management's practice of editing CPRs can be found in Chapter IX. 86
citing the large amount of premiums purchased in the quarter and a volatile rate environment, management changed its methodology to include current quarter acquisitions in its catch-up calculation.(363) Management claimed that this change in methodology was necessary in order to "provide a complete picture of the impact of interest rate/prepayment speed volatility on recorded amortization."(364) While we agree that the change would appear to make the catch-up calculation more accurate since it allowed the Company to take into account current quarter acquisitions, we conclude that management was motivated to do so primarily by the desire to avoid recording the $80 million catch-up adjustment for the third quarter of 2002. This conclusion is based on, among other things, the timing of this change. In other words, while this was not the first quarter in which interest rates were volatile, it was the first time since the implementation of the Policy in December 2000 that management faced having to record a significant amount of catch-up.(365) Further, management resisted recording even the reduced $43 million of catch-up income in the third quarter of 2002. Management stated to KPMG that it would reduce the catch-up to be in compliance with the Policy by year-end. According to then KPMG engagement partner, Mark Serock, he took up this issue with both Pennewell and Spencer, and also advised Howard that if management did not adjust interest income to be in accordance with the Policy by quarter-end, the difference would be included in the audit difference workpapers that would be reviewed with the Audit Committee. After Serock so advised Howard, he was informed by Spencer and Pennewell that management planned to record a $43 million increase to interest income, but that it would also be making an amortization end-date change for the Company's callable debt that would increase interest expense by approximately $35 million,(366) which would, in effect, offset a significant portion of the $43 million of income. In summary, rather than record the $80 million adjustment calculated by its own Policy, management first changed its methodology for calculating catch-up, resisted KPMG's recommendation to record $43 million, and, ultimately, made an end-date change for certain callable debt that resulted in an offsetting adjustment. The sum total of these efforts by management was the elimination of almost the entire $80 million - ---------- (363) FAS 91 Workpaper, dated Oct. 2002. (364) Id. (365) Undated Catch-Up Analysis (Including Realignments), FMSE-IR 265826. KPMG required that Fannie Mae, going forward, continue to include catch-up related to current quarter acquisitions in its catch-up calculation "at least until prevailing market conditions [were] stabilized," FAS 91 Workpaper, dated Oct. 2002, and evidence showed that management did so, at least through the second quarter of 2004. Q2-2004 Forecast Update. (366) See E-mail from Cheryl DeFlorimonte, dated Oct. 10, 2002, FMSE-IR 437919-76. 87
in calculated catch-up income for the third quarter of 2002. This example clearly demonstrates that management's primary motivation with respect to catch-up was to avoid or to minimize its impact on the financial statements.(367) 2. Catch-Up Was Initially Calculated for Year-End From December 2000, when the Policy was first implemented, through the end of 2002, management used an estimate of what the year-end catch-up balance would be to determine whether to record adjustments to its interim period earnings to bring the projected year-end catch-up back within the precision thresholds.(368) In the OCA letter, the Company acknowledged that calculating the catch-up in this manner was not in accordance with GAAP.(369) According to management, it changed this practice in 2003 based on KPMG's instruction and began calculating the catch-up for each interim reporting period.(370) No explanations were offered by the Controller's Office staff as to the reason for this practice, and the recollections also differed as to when and whether KPMG had been aware of this practice. According to Pennewell, Serock advised in 2003 that booking entries as of year-end was no longer appropriate because Enron and Sarbanes-Oxley had changed the environment, and accounting standards were tightening. According to Spencer, at some point in 2002, KPMG told management that it would be a "better practice" to use a quarter-end measurement, but KPMG did not say that the year-end method was not GAAP compliant. Serock stated that he was not aware that management had been using year-end income estimates to determine whether it needed to make adjustments at each quarterly period. Serock stated that he was aware that management was calculating the catch-up on a quarterly basis, but was unaware that it was not recording the catch-up amount at each interim balance sheet date. In addition, Serock could not recall whether anyone at KPMG informed management that it needed to calculate the catch-up as of each balance sheet date and was unaware of any modification management made to the Policy in this regard in the first quarter of 2003. - ---------- (367) Management's reluctance to book catch-up adjustments is also evidenced by their actions in the first three quarters of 2003, when they included in their catch-up analysis the unrecognized portion of the cumulative realignment differences and estimates of future realignments differences, which had the effect of eliminating or reducing the size of the on-top adjustments booked in these quarters. An in-depth discussion of this treatment is discussed in Chapter VI, Part O. (368) OCA Letter at FNMSEC 155. (369) Id. (370) Id. 88
3. Management Made Adjustments Even When Catch-Up Fell Within the Predetermined Threshold The fact that casts the greatest doubt on management's good faith and motive in establishing the Amortization Policy was that management recorded catch-up adjustments even when it was not required to do so under the Policy. On several occasions, management recorded catch-up adjustments despite the fact that the calculated catch-up amount fell within the precision thresholds, which was the "functional equivalent of zero." For example, in 2001, management booked an additional $10 million of catch-up income even though the catch-up was within the predetermined range.(371) An April 12, 2001 memorandum from Juliane and LeRouzes stated that Juliane and LeRouzes determined that the $10 million on-top should be booked "to hedge our exposure if rates continue to decline."(372) In addition, Juliane recalled recording catch-up within the range on at least three other occasions, once in the amount of $19.8 million, which occurred during the fourth quarter of 2001.(373) Juliane understood that this on-top was also booked as a "hedge" and that he was told that the decision to book this amount was made due to a disproportionate exposure in the down-rate environment and in anticipation of future declining rates. The $19.8 million entry was made on January 6, 2002, shortly before Fannie Mae closed its books. Including this entry, Fannie Mae's EPS for the fourth quarter of 2001 was $1.40, which exceeded then Wall Street expectations of $1.39.(374) Fannie Mae's actual EPS was $1.3955.(375) The after-tax impact of the $19.8 million entry on the quarterly EPS for 2001 was approximately $0.0128.(376) In the absence of any other adjustments, had Fannie Mae not booked the $19.8 million on-top adjustment, the fourth - ---------- (371) Mem. from J. Juliane and R. LeRouzes, dated Apr. 12, 2001, FMSE-IR 227051-54, at FMSE-IR 227051; 2001 Q2 Catch-up Sensitivities, FNMSEC 2826; Journal Entry, dated June 6, 2001, FNMSEC 2827. (372) Mem. from J. Juliane and R. LeRouzes, dated Apr. 12, 2001, FMSE-IR 227051-55, at FMSE-IR 227051. (373) See Q4 Catch-up Sensitivities, FNMSEC 2848; Journal Entry, dated Jan. 6, 2002, FNMSEC 2854. (374) Fannie Mae 4th Quarter Profit Rises 23% on Lending Boom, Bloomberg News, Jan. 14, 2002, at 2. (375) Fannie Mae, Selected Financial Information as of Dec. 31, 2001, available at http://www.fanniemae.com/global/pdf/ir/financial/earnings/2001/ iap123101.pdf. (376) Assuming a thirty-five percent statutory tax rate, the after-tax income impact of the $19.8 million adjustment was $12.9 million. When divided by the Company's 1,005.2 million diluted average shares outstanding in the fourth quarter of 2001, the after-tax impact on EPS is calculated to be $0.0128. Id. 89
quarter EPS for 2001 would have been $1.3827 ($1.3955 minus $0.0128), meaning that Fannie Mae would have missed analysts' expectations.(377) Evidence showed that KPMG was aware of the $19.8 million adjustment, and while it noted that the entry was not required under the Amortization Policy, KPMG concluded that the amount was immaterial. Serock stated that KPMG considered SAB 99 in making this determination, but he acknowledged that he did not recall the specific reason for the adjustment. There are additional instances of management making adjustments when the calculated catch-up was within the targeted range. In November 2002, a $10 million on-top adjustment was recorded even though the catch-up position by year-end was only $64.8 million, which was within the Policy range of $93.3 million.(378) Also, there were numerous occurrences of adjustments related to guaranty fees being booked even though the catch-up was within the range through the first quarter of 2002.(379) We have found no credible contemporaneous analysis or explanations that would support management's decisions to record catch-up amounts within the Policy range. According to Pennewell, management stopped recording catch-up that was within the precision threshold in the third quarter of 2002 because KPMG advised that management should record adjustments only to bring the catch-up number to the outside of the range, and not record anything if the catch-up calculation was within the range. Notwithstanding this advice from KPMG, the Company's OCA letter acknowledged that management made yet another $10 million adjustment in the fourth quarter of 2002, even though the catch-up calculation was within the range.(380) 4. KPMG's Awareness of the Manner in Which the Amortization Policy Was Implemented As reflected above, at least certain individuals at KPMG were aware of the ways in which management implemented the Policy. Pennewell stated that the Company - ---------- (377) We note that the $19.8 million adjustment did not seem to affect the Company's ability to reach AIP goals since Fannie Mae comfortably exceeded its maximum AIP goal of 14.9% annual EPS growth by achieving a 21.2% annual EPS growth for 2001 and would have reached that EPS growth without this adjustment. Annual Incentive Plan 2001 Plan Year, FMSE-IR 17500. (378) Undated Catch-Up Analysis (Including Realignments), FMSE-IR 265826; 2002 Q4 Catch-up Sensitivities, FNMSEC 2869; Journal Entry, dated Dec. 9, 2002, FNMSEC 2870. (379) Undated Catch-Up Analysis (Including Realignments), FMSE-IR 265827. (380) OCA Letter at FNMSEC 155. Neither Serock nor Argires from KPMG were aware prior to the OCA letter that management made this adjustment. 90
sent to KPMG documents that showed quarterly catch-up sensitivities that summarized the catch-up results and indicated management's year-end projections, and that management met with KPMG to explain these analyses. KPMG's workpapers indicate, and KPMG staff have confirmed, that Howard, Spencer, Pennewell and Juliane met with Russell, Theobald and Harry Argires of KPMG on December 12, 2000 to discuss the Amortization Policy.(381) The KPMG workpapers noted that at this meeting, management "articulated" its planned purchase premium and discount amortization policy, but there was no indication that management showed KPMG a copy of the written Policy.(382) The KPMG workpapers indicated that management wanted to implement the Policy "because of the sensitivity to PSA rates (which would all be in a band of reasonableness)," and to "allow[__________] for the level of imprecision within the marketplace."(383) The document described the first two provisions of the Policy - the use of an average of five interest rate scenarios and a "dollar threshold of precision" - in some detail; however, the KPMG workpapers contained no reference to reducing catch-up balances in excess of the "dollar threshold of precision" over a longer period.(384) Evidence also showed that KPMG was somewhat familiar with the manner in which the Amortization Policy was implemented. For example, as described supra in Subsection IV.B.1., Serock demonstrated that he was familiar with the Policy's precision threshold when he advised management in the third quarter of 2002 that the Company needed to make an adjustment to interest income to be in compliance with the plus or minus one percent provision of the Policy. Serock and Argires were also aware that management made adjustments even when the catch-up fell within the range - including the $19.8 million adjustment to income made in the fourth quarter of 2001 - but determined that the adjustments were immaterial. On the other hand, Serock stated that the first time he learned that management was applying the Policy by looking only to year-end catch-up estimates, was when he reviewed drafts of the OCA letter in the Fall of 2004. Serock also stated that he was unaware that the Company had been using various tools, such as changing CPRs, that affected the catch-up calculation. In addition, it was not until late 2004 that David C. Britt, KPMG's then new engagement partner, became aware of the practices of recording catch-up within the range and looking to year-end estimates to determine quarterly catch-up numbers. Apparently, Britt learned of these practices only through - ---------- (381) Fannie Mae Accounting for FAS 91, dated Dec. 31, 2000. (382) Id. (383) Id. (384) Id. 91
reviewing the drafts of the OCA letter and promptly informed management that KPMG did not agree with these practices.(385) C. Inadequate Disclosures to the Board About the Amortization Policy and Its Application We have not seen any evidence that management disclosed the Policy or its application to the Board prior to late 2003.(386) In November 2003, Spencer made a presentation entitled "Critical Accounting Policy Deferred Price Adjustments" to the Audit Committee.(387) The minutes of the November 2003 Audit Committee meeting, at which Serock, Britt and Argires of KPMG were present, reflect that, in response to a question from Anne M. Mulcahy, a then member of the Audit Committee, about the Company's FAS 91 policy, "Spencer responded that management could not pick any interest rate path to predict prepayments but instead used an average of rates that is standardized in many processes across the company."(388) Spencer's statement was misleading - we are aware of no other process employed at Fannie Mae at the time that used an average of interest rates or an average of results derived from these interest rates. To the extent that Spencer meant that the interest rates are used in other areas across Fannie Mae, that statement too was inaccurate, as management did not always use the base interest rate that was used in other areas of the Company for catch-up calculation purposes. We conclude that in response to Mulcahy's direct question that reflected her concern that management's discretion was properly constrained by objective measures, Spencer provided a misleading answer that was entirely inconsistent with management's motive for developing the Policy in the first place, and inconsistent with the actual implementation of the Policy. - ---------- (385) While Britt objected to management's practices of booking within the range and looking to year-end estimates to determine quarterly catch-up numbers, he only required the Company to assess the impact on prior periods of the latter practice. See Fannie Mae, Notification of Late Filing (Form 12b-25), at 4 (Nov. 15, 2004), available at http://www.fanniemae.com/ir/sec/index.jhtml?s=SEC+Filings (follow "All SEC Filings" hyperlink; then follow "NT 10-Q, Nov. 15, 2004" hyperlink for PDF format). (386) Spencer acknowledged that prior to November 2003, the Amortization Policy was not disclosed to the Board. (387) Audit Committee Update, Critical Accounting Policy, Deferred Price Adjustments, dated Nov. 17, 2003, FMSE-IR 55456-71; Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Nov. 17, 2003, FMSE 221342-51, at FMSE 221346. (388) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Nov. 17, 2003, FMSE 221342-51, at FMSE 221347. 92
The minutes also reflected that "Ms. Spencer and Fannie Mae management present at the meeting agreed that they would inform the Board of any future changes made to the deferred price adjustment methodology or events where there was a failure to make an adjustment."(389) We saw no evidence in the minutes of any Board Meeting that Spencer followed through on her promise to the Audit Committee.(390) In short, at the November 2003 Audit Committee meeting, Spencer misleadingly painted a picture for the Board that suggested that the Company's Policy was, in fact, objective and that management did not exercise undue discretion with respect to calculating and recording catch-up. Finally, both Spencer and Howard were present at the July 2004 joint Audit and Special Review Committee where the accountants retained by the lawyers for the Company for the Special Examination discussed the propriety of the Company's FAS 91 policy and what OFHEO may allege with respect to the Policy. One of the anticipated criticisms of OFHEO discussed by the accountants was whether the Policy had been implemented inconsistently and to manage earnings.(391) Neither Spencer nor Howard spoke up at this meeting to inform the Board that OFHEO's allegations would be well-founded in the facts, that is, the Policy had been adopted precisely to give maximum discretion to management in managing catch-up, and that since the adoption of the Policy, the Controller's Office staff had routinely applied the Policy inconsistently, to avoid the catch-up's impact on the Company's financial statements. D. 2004 Revisions to the Amortization Policy In August 2004, prior to the issuance of the OFHEO Report, management undertook to change its practices for accounting under FAS 91 and updated its written Amortization Policy. Pennewell was primarily responsible for updating the policy and received assistance from Juliane, among others. Like the December 2000 Policy, the August 2004 Policy permitted calculating the catch-up using an average of the results of the catch-up calculation with five interest rate scenarios and a threshold within which the - ---------- (389) Id. (390) The only other presentation made to the Board relating to the Policy that we reviewed was for a Board meeting in November 2004, which was after the issuance of the OFHEO Report. Financial Performance Update, Fannie Mae Board of Directors, dated Nov. 16, 2004. This presentation indicates that management may have told the Board about a change in the Policy related to the threshold range used to determine whether adjustments needed to be recorded for the guaranty fee component of the catch-up. See id. However, the minutes from this meeting do not reflect that management made this disclosure. Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Nov. 16, 2004, FMSE 504488-94. (391) The discussions that took place at the July 2004 joint Audit and Special Review Committee meeting are discussed further in Chapter X below. 93
catch-up would be considered to be the functional equivalent of zero.(392) The threshold in the August 2004 version of the Policy for both net interest income and guaranty fees was plus or minus one percent.(393) The August 2004 Policy did not contain a provision that allowed management to reduce the catch-up over a three-year period and, instead, explicitly required management to bring the catch-up within the range by quarter-end.(394) The August 2004 policy was silent about whether management could use its judgment to make adjustments when the catch-up was within the range. The Company is in the process of establishing a new Amortization Policy as part of the restatement process.(395) E. Findings Regarding the Amortization Policy Our conclusions concerning the Amortization Policy are as follows: - - The Amortization Policy was designed to give management the maximum amount of discretion to mitigate the income statement effect of recognizing catch-up. Such a purpose has no support in FAS 91. - - The Amortization Policy was also created with the goal of documenting the Company's practice to avoid audit differences, which would have highlighted this - ---------- (392) Amortization Operational Policy, dated Aug. 31, 2004, FMSE-IR 405894-97 at FMSE-IR 405895. (393) Id. (394) Id. (395) After the issuance of the OFHEO Report and prior to the commencement of the restatement process, Janet Pennewell began a SAB 99 analysis to see whether certain ways that management applied the Policy historically - including its use of the precision threshold and year-end estimates, rather than estimates as of interim balance sheet dates, to calculate the catch-up - had a material impact on the Company's financial statements for any given period. However, Pennewell stated that she never completed the analysis because she was not able to recreate all of the data, and thus the analysis was unreliable. In addition, Pennewell stated that the Company had been unable to restore all of the prior PDAMS runs. Huron has sought to re-create the PDAMS runs for 1998, without success. Moreover, we could not determine the interest rate paths management actually used to book amortization because the amortization factors used by the Company to record amortization were not based exclusively on a given interest rate path, but instead also reflected other adjustments made by management. For this reason, and the fact that the Company is currently working with Deloitte & Touch LLP to restate prior period financial results, which will include an assessment of the Company's accounting pursuant to FAS 91, we did not pursue this task. 94
practice to the Audit Committee. An accounting policy such as this should have focused on accounting for purchase premium and discount amortization in accordance with GAAP, rather than focusing on avoiding audit differences. - - We also conclude that management continued to exercise discretion for recognizing catch-up, and disregarded the terms of the Policy when it did not suit its purpose. There is no legitimate explanation for, among other things, management's recording adjustments when the catch-up fell within its predetermined threshold. Management's exercise of discretion to book amounts as a reserve against future interest rate changes was not consistent with FAS 91 and also contradicted management's stated rationale for the Amortization Policy. - Finally, we conclude that Spencer and Howard concealed the true purpose of the Amortization Policy and the manner in which it was implemented from the Board. 95
CHAPTER IV APPENDIX: FANNIE MAE'S FAS 91-RELATED SYSTEMS The systems that Fannie Mae used to calculate FAS 91 amortization and amortization catch-up can be separated into three groups: source, modeling, and amortization systems.(396) I. SOURCE SYSTEMS The source systems have two main functions for FAS 91 amortization accounting. First, they provide loan and security data such as the coupon rate, time to maturity, unpaid principal balance, and original premium and discount amounts to the modeling systems for the purposes of modeling prepayment speeds and ultimately generating amortization factors. In addition, the source systems provide original premium and discount data to the amortization systems for the purpose of calculating amortization income or expense. The five principal source systems used at Fannie Mae during the relevant period through the present are known as "Delivery,"(397) "LASER,"(398) "Strategy,"(399) "STAMPS"(400) and "STATS."(401) - ---------- (396) See Undated AIMS: Data Flow and Aggregation and iPDI: Context Diagram for comprehensive system flowcharts that diagram the amortization process. (397) Delivery is a client-server system used by lenders to submit loans to Fannie Mae. Delivery transfers the data related to Fannie Mae acquisitions to the LASER and Strategy systems. In addition, Delivery also provides single-family original premium and discount data to the amortization system. Data Analysis Specification, AIMS Project, dated Oct. 8, 2003, FMSE 193105-42, at FMSE 193112; Fannie Mae FAS 91 Amortization Process (AIMS Data Aggregation), dated Nov. 12, 2004, FMSE-IR 9370-72. (398) LASER is a mainframe system that tracks single-family loans in the Fannie Mae portfolio. LASER receives data from the Delivery system and tracks loan-level data such as the coupon rate, time to maturity, unpaid principal balance, and premium and discount amounts. For amortization purposes, LASER passes this data to the modeling systems. Data Analysis Specification, AIMS Project, dated Oct. 8, 2003, FMSE 193105-42, at FMSE 193113. (399) Strategy is the system that tracks multifamily loans in the Fannie Mae portfolio. Strategy receives data from the Delivery system and tracks loan-level data similar to that of LASER. Fannie Mae FAS 91 Amortization Process (AIMS Data Aggregation), dated Nov. 12, 2004, FMSE-IR 9370-72. For amortization purposes, Strategy passes this data to the modeling systems and provides multifamily original premium and discount data to the amortization system. Data Analysis Specification, AIMS Project, dated Oct. 8, 2003, FMSE 193105-42, at FMSE 193113. 96
II. MODELING SYSTEMS The modeling systems aggregate loan and security level data into pools based on certain characteristics including asset type (referred to at the Company as "the FAS 91 type"), month, and year of acquisition and coupon rate. This information and the selected interest rate path is then applied to prepayment models to generate estimates of conditional prepayment rates ("CPRs"),(402) cashflows, and updated amortization factors(403) that are ultimately used in the calculation of what Fannie Mae referred to as the "catch-up" estimate. Data received from the source systems is aggregated at two levels. Currently, data is first aggregated by FMIS(404) and then by ALEX(405) before model-ready records - that is, data for similar loans aggregated into a series of FAS 91 "buckets" - ---------- (400) STAMPS, or the Securities Trading and MBS Processing System, is the system that records MBS transactions related to the Fannie Mae portfolio. For amortization purposes, among others, STAMPS provides security data to the STATS system. Id. (401) STATS, or the Securities Trading Accounting Transactional System, is the system that tracks all securities acquired by Fannie Mae. STATS receives data from the STAMPS system and tracks securities at the CUSIP number level. For amortization purposes, STATS provides data to both the modeling and amortization systems. Id. (402) Conditional Prepayment Rates (or "CPRs") are measurements of prepayments expressed as a percentage of the remaining principal balance at the beginning of the measurement period (hence, they are "conditional" on the principal balance). Linda Lowell, Mortgage Pass-Through Securities, in THE HANDBOOK OF MORTGAGE-BACKED SECURITIES 25, 41 (Frank J. Fabozzi ed., 5th ed. 2001). (403) Amortization factors are percentages applied to original premium and discount amounts in order to calculate amortization. (404) FMIS, or the Financial Management Information System, is a "data warehouse" currently used by Fannie Mae, that receives and aggregates information downloaded from the LASER, Strategy, and STATS. Data Analysis Specification, AIMS Project, dated Oct. 8, 2003, FMSE 193105-42, at FMSE 193114. (405) ALEX, or the Asset Liability Extractor, is the system currently used by Fannie Mae that provides data for similar loans that have been aggregated into a series of FAS 91 buckets reflecting the characteristics of those similar loans for all portfolio analytical systems, and includes data such as CUSIP, FAS 91 Type, Acquisition Date, Original Unamortized balance, Current Unamortized balance, Original Face, and Current Face. Data Analysis Specification, AIMS Project, dated Oct. 8, 2003, FMSE 193105-42, at FMSE 193113-15. 97
containing loans with similar characteristics - are passed to AIMS.(406) AIMS then passes this data to Valuation Net ("VN"),(407) which applies an interest rate path to its library of prepayment models in order to calculate CPRs.(408) Portfolio Strategy creates interest rate paths based on the forward interest rates derived from the yield curve. These rates, along with "shocks" to the interest rate paths, are compiled into an interest rate projection file, which contains the interest rate information used throughout Fannie Mae, and is made available within VN. The CPRs generated by VN are then passed back to AIMS, which, in turn, calculates the constant effective yield and creates updated amortization factors. AIMS then applies the updated amortization factors to the original unamortized premium and discount amounts to determine how much should have been amortized based on the current data, assumptions, and modeling. AIMS compares this amount to the amount of amortization actually recorded to the financial statements, which it receives from the amortization system. The resulting difference is the catch-up estimate. Based on the catch-up estimate, and when deemed appropriate, the Company adjusts the actual amortization factors that are then provided to the amortization system. In 1998, the modeling system used by the Company was PDAMS.(409) In 1999, Fannie Mae supplemented PDAMS with a software called BancWare Convergence - ---------- (406) AIMS, or the Amortization Integration Modeling System, is the application currently used by Fannie Mae that is responsible for the calculation of the constant effective yields and amortization factors used to amortize the premiums, discounts, and other deferred price adjustments into income. Data Analysis Specification, AIMS Project, dated Oct. 8, 2003, FMSE 193105-42, at FMSE 193105-08. (407) VN is a portfolio analytics system. It houses two data libraries, the Mortgage Cash Flow Library (MCFL) and the Portfolio Analytics Library (PAL), which work together to calculate CPRs, which AIMS ultimately uses to generate amortization factors. For REMIC amortization, VN interacts with a third party library called INTEX, which provides historical and prospective cash flows for all REMICs. Fannie Mae began using INTEX when it started modeling REMICs in 1998. Mem. from Jeffrey Juliane to File, dated Apr. 5, 2004, FMSE-E 147666-69. (408) At present, there is an added step for modeling REMICs since cash flows require allocation among the different tranches of the REMIC. The process for modeling REMICs involves the same systems described above in the same capacity, with the addition of INTEX. INTEX is a data library licensed by a third party that contains tranche-level REMIC characteristics. INTEX applies these characteristics to estimated cash flow data, for the underlying collateral for the REMICs, received from VN to generate tranche-level cash flows. These tranche-level cash flows are then passed back to VN. Fannie Mae began using INTEX when it started modeling REMICs in 1998. Id. 98
("BancWare").(410) The primary benefit of BancWare appears to have been its ability to run different scenarios more quickly than PDAMS.(411) Management ultimately discontinued BancWare and sought instead to develop a new modeling system to replace PDAMS.(412) PDAMS was ultimately replaced by AIMS in the fourth quarter of 2001.(413) The transition to AIMS appears to have improved the process of generating amortization factors by consolidating system responsibilities, modeling all assets in Fannie Mae's "core" and "non-core" books,(414) incorporating a potentially unlimited number of rate environments, increasing the stratification of generated cash flows, grouping CPRs by month as opposed to longer intervals, and explicitly modeling guaranty fee collateral.(415) - ---------- (409) PDAMS, or the Purchase Discount Amortization Management System, was the application responsible for the calculation of the constant effective yields and amortization factors used to amortize the premiums, discounts, and other deferred price adjustments into income. PDAMS User Documentation, dated Apr. 7, 1994, FMSE-IR 2833-48. PDAMS was used from 1994 to 2001. Tr. of Aug. 31, 2004 OFHEO Dep. of J. Juliane at 216:11-218:23. (410) Mem. from J. Juliane to Distribution, dated Mar. 12, 1999, FMSE-SP 364-67, at FMSE-SP 364. (411) Id. at FMSE-SP 364. (412) Tr. of Aug. 31, 2004 OFHEO Dep. of J. Juliane at 216:11-218:23. (413) Mem. from J. Juliane and Rene LeRouzes to Distribution, dated Oct. 28, 2001, FMSE 217674-76, at FMSE 217675; Mem. from R. LeRouzes to File, dated Apr. 9, 2004, FMSE-E 1851058-59, at FMSE-E 1851058. (414) In the context of FAS 91 amortization, Fannie Mae refers to premium and discount on whole loans and MBS (excluding REMICs) as well as buy-ups and buy-downs on guaranty fees as its "core" book and refers to premium and discount on its REMICs and synthetic REMICs as its "non-core" book. (415) AIMS Requirements Documentation, dated Dec. 22, 2000, FMSE-E 1166285-300; Mem. from J. Juliane to File, dated Oct. 9, 2001, FMSE 56578-80, at FMSE 56578. Documents and witnesses indicate that in 2004, management decided to develop a new amortization system known as MARS (Modeling, Amortization, and Reporting System). Agenda for AIMS Re-engineering Status Meeting, dated June 9, 2004, FMSE-E 155732-33. The primary objective of MARS was to replicate the functionality of AIMS while providing updates for emerging accounting standards, improving audit, control, and reporting responsibilities in addition to reducing processing time. See MARS System Requirements Specifications, dated June 26, 2004, FMSE-E 2128276-302, at FMSE-E 2128279-80. We understand that the 99
From 1989 until 1996, Rick DePetris, then Director--Financial Standards, was responsible for the operation of the PDAMS system. In 1997, Jonathan Boyles replaced DePetris as head of Financial Standards and took over the responsibilities for PDAMS. In 1999, Leanne G. Spencer, then Senior Vice President and Controller, transferred the PDAMS responsibility out of Financial Standards and into the Financial Planning area of the Controller's Office. In 1999, Janet L. Pennewell put Jeffrey Juliane, a Manager in the Controller's Office, in charge of operating PDAMS. Juliane was responsible for PDAMS and AIMS, as well as BancWare, until his departure in the summer of 2005. III. AMORTIZATION SYSTEM From 1998 until 2003, the Premium and Discount Integration ("PDI") system was the amortization system that applied amortization factors received from the modeling systems to original premium and discount in order to calculate amortization expense or income and post amortization entries to the general ledger. In 2003, the PDI system was upgraded and the new system is referred to as the Improved Premium and Discount Integration ("iPDI") system.(416) To calculate amortization income or expense, iPDI applies the amortization factors it receives from the modeling system to the original premium and discount balances reflected in various source systems. iPDI then generates amortization journal entries that are posted to the general ledger. In addition, as described above, iPDI also provides original and current unamortized balances to AIMS for the purpose of calculating the catch-up. Roger Barnes, who was the Manager of Premium/Discount Amortization, was responsible for running the PDI and iPDI systems until July 2003. - ---------- MARS system was never implemented and the project to develop the system was abandoned. Currently, in connection with its restatement, the Company is developing a system referred to as the "Amortization Engine" which will perform premium and discount amortization. (416) The transition from iPDI to PDI was a system upgrade, not a system replacement. iPDI maintains the functionality of PDI, while improving the efficiency of the process. 100
CHAPTER V: ACCOUNT FOR DERIVATIVE INSTRUMENTS AND HEDGING ACTIVITIES In this Chapter, we set forth our findings on Fannie Mae's implementation and application of FAS 133,(417) which governs accounting for derivative instruments and hedge transactions. Hedge transactions were an important part of Fannie Mae's portfolio management strategy, and, of the issues raised by the OFHEO Report, those involving FAS 133 were the most significant in terms of their impact on the Company's financial statements. We begin with a summary of our conclusions regarding FAS 133. Then, in order to place our discussion of the FAS 133 issues into context, we discuss Fannie Mae's debt portfolio and relevant concepts concerning FAS 133 and hedge accounting. Finally, we discuss the Company's implementation of FAS 133 and the objectives that motivated that effort, and our findings. I. INTRODUCTION As noted above, hedge transactions were an important part of Fannie Mae's portfolio management strategy. The Company's outstanding debt grew tremendously during the 1990s (commensurate with the growth in its portfolio), and the ability to hedge that debt against interest-rate risk was a substantial component of the Company's overall risk-management strategy. The Company used derivatives to hedge the interest-rate risk associated with its debt, and the notional amount of the Company's derivative portfolio also grew tremendously during the 1990s and into the 2000s. FAS 133 was issued in 1998 and required that companies record changes in the fair value of their derivative portfolio through earnings. Companies could avoid the earnings volatility associated with FAS 133 by entering into "effective hedges." FAS 133 refers to this as "special accounting," and the rules contain explicit guidelines on how to achieve and account for effective hedges. Fannie Mae undertook a two-year effort to implement FAS 133 prior to the standard's effective date on January 1, 2001. There is substantial evidence that the Company's effort was focused on avoiding the earnings volatility associated with the implementation of FAS 133 by entering into what the Company deemed to be "perfect hedges," or perfectly effective hedges. We are aware that there has been substantial criticism of FAS 133; in particular, we are aware of the view that FAS 133 injects inappropriate volatility into earnings. Apparently, some companies took the view, as Fannie Mae did, that FAS 133 would not be interpreted literally, and that FAS 133 was open to practical interpretation - ---------- (417) ACCOUNTING FOR DERIVATIVE INSTRUMENTS AND HEDGING ACTIVITIES, Statement of Fin. Accounting Standards No. 133 (Fin. Accounting Standards Bd. 1998) (hereinafter "FAS 133"). 101
and application based on the principles involved, even if the interpretation or application deviated from a strict reading of the standard. We are also aware that, in the wake of the SEC's announcement concerning errors in Fannie Mae's accounting under FAS 133 in November 2004, other companies have announced restatements of their own for FAS 133. However, in our view, Fannie Mae did not engage in innocuous practical interpretations or modest deviations from a strict reading of the standard. The Company's implementation of FAS 133 was motivated, not only by a desire to avoid earnings volatility, but also by a desire to avoid the substantial changes to the Company's business methods, and/or the development of new and complex accounting systems, that would have been required to satisfy FAS 133's special accounting requirements. These considerations led the Company, with the knowledge of senior managers in the Controller's Office, to adopt an approach to hedge accounting that relied too heavily on the assumption that its hedges were "perfectly effective." The Company's approach deviated from FAS 133 requirements in numerous and important respects. Indeed, the record of our review shows that the Company's method of hedge accounting conflicted with clear and specific provisions of FAS 133. As one example, the Company disregarded amendments to FAS 133 that the FASB had adopted over a year before the standard took effect. The FASB's amendments were directed specifically to limit the application of hedge accounting with respect to a type of transaction that Fannie Mae had raised with the FASB, and which was referred to within the Company as the "Fannie Mae carve-out." Notwithstanding the amendment, the Company did not change its approach to hedge accounting for these transactions. The Company's accountants relied on an assessment that any deviations from a "strict application" of FAS 133 that were embodied in the Company's approach to hedge accounting were immaterial. However, the fact that the Company even undertook such assessments is inconsistent with its blanket assumption that its hedge transactions were perfectly effective. Moreover, the Company did not engage in a systematic or comprehensive effort to test the proposition that its deviations from a strict reading of the standard were immaterial, and the tests that it did undertake provided little support for its conclusions. The Company's implementation of FAS 133, however, was undertaken in consultation with Fannie Mae's outside auditor, KPMG. The record shows that the Company took steps throughout the FAS 133 implementation process to keep its outside auditor informed of the important decisions. Management engaged KPMG to perform a special review of its Derivatives Accounting Guidelines prior to the effective date of FAS 133 to ensure the principal features of its guidelines complied with the standard. Statements in KPMG's audit workpapers reveal that the auditors knew of and accepted the Company's major accounting policies in this area on the grounds that deviations from GAAP reflected in Fannie Mae's hedge accounting policies were immaterial. In addition, there is evidence that the Company's hedge accounting guidelines were open to and were reviewed by OFHEO in 2001, the year in which FAS 102
133 took effect. As late as June 2002, when OFHEO issued its reports on Fannie Mae's first year of operation under FAS 133, OFHEO reported to Congress and the Company that Fannie Mae's implementation of FAS 133 had a sound footing. There is evidence that Timothy Howard set the tone for the FAS 133 implementation effort, and the evidence indicates that, from the outset and throughout the process, he focused the implementation team's efforts on avoiding the volatility associated with FAS 133 and not changing the Company's business practices to a significant degree. There is no indication, however, that he specifically directed anyone to violate GAAP or to take any specific measures in that direction. Frank Raines's involvement in the implementation effort was minimal. Although he was familiar with the Company's interest in avoiding the income statement volatility and complex systems-development effort associated with hedge accounting under FAS 133, we see no indication that Raines was aware of the Company's departure from GAAP. The Board received assurances on several occasions that the Company's implementation of FAS 133 was correct. In addition, the Board was aware of, and relied on, assurance from KPMG and in OFHEO's report described above. II. BACKGROUND A. Fannie Mae's Debt Policies and Practices Fannie Mae finances its mortgage and securities acquisitions in part by raising funds in the debt market.(418) Fannie Mae issues debt with a variety of terms and features, including fixed and variable interest rates, call options, and terms ranging from a few days to ten years or more. According to Fannie Mae, the most common debt instruments it issues are:(419) NONCALLABLE BENCHMARK NOTES -- Noncallable benchmark notes are securities with maturities ranging from two to ten years. Noncallable benchmark notes are issued on a schedule set at the beginning of each year. This debt may specify payment of interest at a fixed rate or at a variable rate that resets periodically. - ---------- (418) Derivatives Accounting Guidelines, dated Jan. 2001, FMSE-IR 12825-3295, at FMSE-IR 12873 (hereinafter "DAG"). (419) This information is drawn from Fannie Mae's website. See Fannie Mae, Understanding Fannie Mae's Debt, at http://www.fanniemae.com/markets/debt/understanding_fm_debt/funding_ programs.jhtml. It is consistent with other documents we have reviewed. See, e.g., DAG at FMSE-IR 12874-75 103
CALLABLE BENCHMARK NOTES -- Callable benchmark notes are securities with characteristics similar to noncallable benchmark notes, with two important differences. First, they are issued in response to market demands and portfolio considerations, rather than on a pre-determined schedule. Second, callable benchmark notes allow Fannie Mae to exercise a call option and pre-pay the debt. Like noncallable debt, callable debt may have fixed or variable interest rates. DISCOUNT NOTES/BENCHMARK BILLS -- Discount notes and benchmark bills are short-term borrowings with maturities ranging from overnight to one year. Benchmark bills and discount notes with maturities of three and six months are issued at a weekly auction, while benchmark bills and discount notes with maturities of one year are issued at a monthly auction. Discount notes of other maturities are issued through a dealer network. Often, Fannie Mae will repay maturing discount notes or benchmark bills using the proceeds from newly issued discount notes or benchmark bills (referred to herein as "re-issuing" or "rolling over" the maturing discount notes or benchmark bills). When the Company rolls over its debt in this fashion, these streams of debt collectively mimic longer-term callable debt having a variable interest rate.(420) The amount and characteristics of the debt that Fannie Mae issues are important elements in the Company's overall commercial strategy. Generally, the Company's objective is to maintain a spread between the interest it receives from its mortgage and securities portfolios and the interest it pays on its debt. The Portfolio Investment Committee sets the Company's debt strategy on a weekly basis with an eye toward maintaining this spread. The Treasurer's Office is responsible for the issuance and management of the Company's debt. The Controller's Office is responsible for debt accounting. The Company's outstanding debt increased substantially during the period relevant to our review, commensurate with the increase in the size of the Company's retained mortgage portfolio. Fannie Mae reported total debt of $123.4 billion at year-end 1990.(421) By 2000, the Company's total debt had more than quintupled, and it continued to increase thereafter: - ---------- (420) DAG at FMSE-IR 012874. (421) OFHEO Report to Congress, dated June 15, 2004, Appendix at 8, available at http://www.ofheo.gov/media/pdf/finalreport61504.pdf (hereinafter "2004 OFHEO Report to Congress"). 104
Outstanding Total Assets Debt Period(422) ($ in millions) ($ in millions) - ----------- --------------- --------------- 2000 675,224 642,682 2001 799,948 763,467 2002 887,515 850,982 2003 1,009,569 961,732 Fannie Mae's use of short-term discount notes, benchmark bills, and long-term noncallable benchmark notes to finance the acquisition of generally fixed-rate mortgage assets exposes it to substantial interest-rate risk. As a general principle, an increase in interest rates will result in an increase in the interest costs associated with variable rate debt and with the rollover of discount notes and short-term benchmark bills. Because Fannie Mae's mortgage and securities portfolios consist primarily of fixed-rate mortgages, the spread between portfolio income and debt costs will decrease as interest costs increase and interest revenues remain relatively constant. A decrease in interest rates also can affect Fannie Mae's spread. As interest rates decrease, prepayments of mortgages with higher fixed interest rates usually accelerate. Because higher-yielding mortgages are replaced by mortgages with lower interest rates, prepayments reduce the spread between the Company's interest income and fixed debt payments. In addition, interest income on adjustable-rate mortgages that the Company holds in its portfolio will fall as those mortgages reprice to the lower rate. The net impact of these developments, other things being equal, is a decrease in the spread. B. Derivatives and Hedging In order to reduce its vulnerability to interest rate fluctuations, Fannie Mae devotes considerable effort to managing its interest-rate risk. For present purposes, the most important aspect of the Company's program is its use of derivatives to hedge the interest-rate risk associated with its debt.(423) A derivative is an agreement between two parties to exchange a payment based on one set of variables for a payment based on another set of related variables. Broadly speaking, Fannie Mae buys derivative - ---------- (422) Id. (423) As noted in its 2001 Annual Report, Fannie Mae "typically uses derivative instruments . . . to hedge against the impact of interest rate movements on its debt costs to preserve its mortgage-to-debt spreads." Fannie Mae 2001 Annual Report at 64, available at http://www.fanniemae.com/global/pdf/ir/annualreport/2001/fullreport.pdf. 105
instruments that will offset the impact of interest rate movements on Fannie Mae's debt obligations and thereby maintain its spread. As defined in FAS 133, a derivative is a financial instrument that meets all of the following criteria: - It has one or more "underlyings" (i.e., a specified index, such as an interest rate, security or commodity price, or currency) and one or more notional amounts (i.e., a specified number, such as currency units, shares, bushels, or other quantity) used to determine the amount of any payment to be exchanged between the parties to the contract. - It requires no (or a nominal) initial net investment. - Its terms require or permit net settlement, or it can be readily settled net by a means outside the contract (i.e., by entering into an offsetting position on an exchange that results in the party being relieved of its rights and obligations under the contract), or it requires the delivery of an asset that is readily convertible to cash.(424) Most of the derivatives relevant to Fannie Mae's FAS 133 accounting are: (1) interest-rate swaps (pay-fixed, receive-variable swaps; pay-variable, receive-fixed swaps; basis swaps; and forward-starting swaps); (2) options on interest-rate swaps (also known as "swaptions"); and (3) interest-rate caps (agreements in which payments begin if interest rates cross a certain threshold).(425) These derivatives counter the changes in either the cash flows on, or the fair value of, debt that result from changes in interest rates.(426) Given Fannie Mae's use of derivatives to hedge the interest-rate risk associated with its debt, the Company's derivative portfolio is dominated by instruments in which the notional amounts and the underlyings relate to the face amounts and interest-rate features of Fannie Mae's debt.(427) In a typical hedge transaction, Fannie Mae will acquire a derivative (in the following example, a pay-fixed, receive-variable interest-rate swap) that requires it to make a fixed monthly payment based on the application of a rate of interest (e.g., six - ---------- (424) FAS 133 PP 6-9; see also DAG at FMSE-IR 12847. (425) DAG at FMSE-IR 12876. (426) Id. at FMSE-IR 12877. (427) See id. at FMSE-IR 12876. Fannie Mae also maintains a relatively small portfolio of derivatives that hedge foreign exchange rate risk. We do not address these derivatives in our report. 106
percent) to a notional amount (e.g., $100 million); in exchange, the counterparty agrees to make a variable payment based on the application of the London Inter-Bank Offered Rate ("LIBOR") for a specified term (e.g., three-month LIBOR) to the same notional amount and designates that interest-rate swap as a hedge of its variable rate debt. Fannie Mae makes fixed payments on the derivative and uses the variable payments it receives from the counterparty to offset payments it is required to make to holders of its variable-rate debt. To the extent that the interest-rate indices, maturities, repricing dates (dates on which the variable interest rates reset), and other relevant features of the debt and the derivative are the same, the payments received from the counterparty on the derivative would be expected to offset the interest-rate variability associated with the debt. Economically, Fannie Mae has converted its variable-rate debt into fixed-rate debt for the term of the swap. Because this "pay-fixed swap" affects the variability in Fannie Mae's future debt payments on a dollar for dollar basis, this type of hedge is called a "cash-flow hedge."(428) A significant amount of Fannie Mae's derivative portfolio has the reverse effect: economically converting the Company's fixed-rate debt into variable-rate obligations. In a "receive-fixed" swap, the Company agrees to make payments based on the application of a variable interest rate to a notional amount; the counterparty agrees to pay Fannie Mae an amount based on a fixed interest rate applied to the same notional amount. To the extent that the debt and the derivative have appropriately corresponding features, the Company has hedged against the effect of a decline in interest rates on its net interest margin. Because a decline in interest rates results in a decline in the fair value of fixed-rate debt, this type of hedge is referred to as a "fair-value hedge."(429) A basis swap involves a derivative that has the effect of converting the interest rate on a debt instrument from one index to another. For example, Fannie Mae might enter into cash-flow hedge of a LIBOR-based borrowing by linking the debt to a pay-fixed swap under which it receives payments based on LIBOR and makes fixed payments. If Fannie Mae subsequently replaces that LIBOR-based borrowing with debt indexed to the prime rate, the original interest-rate swap would not by itself be effective as a hedge against changes in the variable rate because the debt and the interest-rate swap have different bases. Accordingly, Fannie Mae might enter into a second interest-rate swap under which it would receive a payment based on the prime rate and make a payment based on LIBOR. The Company could then designate the two interest-rate swaps in combination as a cash-flow hedge of the new prime-rate borrowing. - ---------- (428) Paragraph 28 of FAS 133 provides: "An entity may designate a derivative instrument as hedging the exposure to variability in expected future cash flows that is attributable to a particular risk." (429) Paragraph 20 of FAS 133 states: "An entity may designate a derivative instrument as hedging the exposure to changes in the fair value of an asset or a liability or an identified portion thereof . . . that is attributable to a particular risk." 107
C. Significant Hedge Transactions The notional amount of Fannie Mae's derivative portfolio has expanded tremendously in recent years. In 1990, the total outstanding notional amount of Fannie Mae's derivative portfolio was just over $6.5 billion.(430) In 1995, that figure had grown to $128 billion.(431) The growth continued thereafter, and passed the $1 trillion mark in 2003: Notional Amount of Financial Derivatives Outstanding ($ in Period(432) millions) - ----------- --------------------- 1998 187,557 1999 274,970 2000 324,740 2001 533,139 2002 656,595 2003 1,079,088 Fannie Mae enters into a number of different hedging transactions using derivatives in different combinations and in association with different types of debt. Over time, Fannie Mae's "menu" of hedge transactions has increased to approximately seventy different combinations of debt and derivatives. Five transactions are particularly significant for present purposes. Although this list is not comprehensive, it is illustrative, and many of the more complex transactions are combinations of, or variations on, these basic models.(433) SWAPS ON MEDIUM- AND LONG-TERM DEBT -- As described above, Fannie Mae issues medium- and long-term debt. If this is variable-rate debt, Fannie Mae hedges the variability in its future interest payments by entering into a pay-fixed swap - ---------- (430) 2004 OFHEO Report to Congress, Appendix at 16. (431) Id. (432) Id. (433) This list does not include the important concept of "all-in-one hedges" associated with commitments to purchase or sell mortgages or mortgage-backed securities. This concept is discussed in Chapter VI, Part C of our Report, dealing with FAS 149. 108
(i.e., a cash-flow hedge). With respect to fixed-rate debt, Fannie Mae hedges the changes in the fair value of that debt by entering into a receive-fixed swap (i.e., a fair-value hedge). HEDGES OF DISCOUNT NOTE ROLLOVERS -- Fannie Mae hedges the risk of variability in interest rates associated with the rollover of discount notes. To the extent that the Company anticipates rolling over discount notes at periodic intervals, then this program is similar to the issuance of variable-rate debt with repricing dates at intervals equal to the maturity of the discount notes. A suitable pay-fixed swap, therefore, operates as a cash-flow hedge against the variability in interest rates at the rollover dates.(434) SWAPTIONS -- Fannie Mae has accumulated a significant volume of swaptions - that is, options to enter into an interest-rate swap at a future date.(435) The significance of these instruments for present purposes is associated with their treatment under FAS 133, which made it difficult for a company to include changes in the time-value component of an option's fair value in its assessment of the effectiveness of the hedge (discussed below). This difficulty led most companies to include the change in the time-value element of an option's fair value in earnings.(436) Fannie Mae lobbied the FASB to change this aspect of FAS 133 but was not successful. Since 2001, therefore, Fannie Mae has included in its financial statements two measures of earnings: GAAP earnings, which includes changes in the fair value of the option's time value, and "core earnings," which differed from GAAP earnings only in that the changes in the fair value of the option's time value was excluded, with the amount paid for the option amortized to earnings on a straight-line basis over the term of the option. TERM-OUTS -- Fannie Mae dubs certain transactions "term-outs." A term-out transaction involves multiple instruments that are intended in the aggregate to result in a hedging relationship.(437) The following example illustrates a typical term-out - ---------- (434) A diagram from the DAG showing a hedged transaction involving the rollover of discount notes may be found at Tab C of the Appendix to this Report. (435) DAG at FMSE-IR 12879. (436) The FASB staff, in Statement 133 Implementation Issue G20, agreed that, with respect to options meeting certain conditions, changes in the time-value element of an option's fair value could be deferred on a company's balance sheet in other comprehensive income prior to exercise of the option. See ASSESSING AND MEASURING THE EFFECTIVENESS OF A PURCHASED OPTION USED IN A CASH FLOW HEDGE, Statement 133 Implementation Issue No. G20 (Fin. Accounting Standards Bd. 2001). Because the majority of Fannie Mae's swaptions were for use in fair-value hedges, however, the guidance in Implementation Issue G20 was not of significant help to Company management in attempting to reduce earnings volatility. (437) See, e.g., DAG at FMSE-IR 12882. 109
transaction: Fannie Mae issues 90-day discount notes; Fannie Mae anticipates rolling over the notes for five years, and therefore enters into a five year, pay-fixed, receive-LIBOR interest-rate swap with a notional amount equal to the face value of the discount note. The derivative operates as a cash-flow hedge of the interest-rate risk associated with the rollover of the discount notes. After two years, however, the Company discontinues the rollover of the discount notes and instead issues a fixed-rate, two-year note in the same principal amount. The Company therefore has two pay-fixed obligations: the original pay-fixed swap, and the new debt. The Company hedges the fair value of the fixed-rate note by entering into a two-year, pay-LIBOR, receive-fixed interest-rate swap. The Company then designates the original pay-fixed, receive-LIBOR interest-rate swap, in combination with the new pay-LIBOR, receive-fixed interest-rate swap, as a cash-flow hedge of the anticipated issuance of discount notes covering the remainder of the original interest-rate swap. As the payment provisions on the derivatives cancel out, the Company's obligation remains fixed.(438) ANTICIPATORY OR FORECASTED DEBT ISSUANCE -- Fannie Mae developed a program for hedging its exposure to interest-rate changes that are associated with debt that the Company expected to issue in the future. Specifically, the Company sought to mitigate the risk associated with interest-rate changes between the date on which the Company established its plan to issue the debt (and determined the expected interest rate on the debt) and the date on which it actually issued the debt. Depending on the nature of the debt, changes in the interest rate would affect either Fannie Mae's debt payments in the future or (if the debt was a discount instrument) the proceeds it would receive on the debt. To hedge that risk, the Company generally would enter into a cash-settled "short sale" of a financial asset (such as U.S. Treasury securities).(439) Changes in the fair value of these short-sale contracts would be expected to counter any change in the fair value of the anticipated debt issuance as a result of changes in interest rates.(440) Hedging activity associated with forecasted or anticipated debt transactions arose in two different contexts. Fannie Mae's Hedge Desk was responsible for hedge transactions involving the issuance of large debt instruments, such as benchmark notes. Fannie Mae's Funding Desk was responsible for the issuance of discount notes; to the extent that Fannie Mae's strategy involved hedging the rollover of discount notes, that strategy also involved the hedging of an anticipated debt issuance. - ---------- (438) A diagram showing a term-out transaction is included in the Appendix under Tab C ("Term-Out Transaction #4"). (439) DAG at FMSE-IR 12878. (440) Such arrangements, however, would not hedge against changes in the credit component of Fannie Mae's debt costs. 110
III. AN OVERVIEW OF FAS 133 The FASB first considered the development of accounting principles for derivatives in 1983. Following over a decade of study, the issuance of several preliminary studies and reports, public hearings, and an exposure draft in 1996, the FASB issued FAS 133 in June 1998. Under the original provisions of the standard, the Company would have been required to comply with FAS 133 by January 1, 2000. The FASB amended FAS 133 twice before its effective date. In FAS 137, issued in June 1999, the FASB extended the effective date of FAS 133 by one year, to January 1, 2001, for companies (like Fannie Mae) that reported their financial results on a calendar-year basis.(441) In FAS 138, issued in June 2000, the FASB made substantive changes to FAS 133 that are discussed below.(442) In April 2003, following FAS 133's effective date, the FASB issued FAS 149, which clarified that firm commitments to purchase mortgage loans and purchase or sell mortgage-backed securities should be treated as derivatives under FAS 133. FAS 149 is addressed in a separate section of our report. Prior to FAS 133, the accounting literature covering hedge transactions was a patchwork. For example, FAS 80 addressed futures contracts and, by analogy, forward contracts; FAS 52 concerned foreign currency transactions; AICPA Issues Paper 86-2 dealt with options; and a variety of EITF consensuses covered interest-rate swaps. For interest-rate swaps, the general pre-FAS 133 principle was known as "synthetic instrument accounting," under which an interest-rate swap was linked (for accounting purposes) to a debt instrument, with the combined arrangement treated as if the debt had the interest-rate characteristics of the interest-rate swap. For example, a company with a pay-fixed, receive-variable interest-rate swap designated as a hedge of a variable-rate debt instrument would account for the arrangement as a fixed-rate debt instrument. To the extent the interest-rate swap involved in a hedge relationship qualified for hedge accounting, it was not separately recognized as an asset or liability on a company's balance sheet, and changes in its fair value had no direct impact on earnings. FAS 133 introduced three fundamental principles into accounting for derivatives. First, under FAS 133, derivatives "represent rights or obligations that meet the definitions of assets or liabilities and should be reported in the financial statements."(443) Second, "[f]air value is . . . the only relevant measure for derivative - ---------- (441) ACCOUNTING FOR DERIVATIVE INSTRUMENTS AND HEDGING ACTIVITIES - DEFERRAL OF EFFECTIVE DATE OF FASB STATEMENT NO. 133, Statement of Fin. Accounting Standards No. 137 PP 7-8 (Fin. Standards Accounting Bd. 1999) (hereinafter "FAS 137"). (442) ACCOUNTING FOR CERTAIN DERIVATIVE INSTRUMENTS AND CERTAIN HEDGING ACTIVITIES, Statement of Fin. Accounting Standards No. 138 (Fin. Standards Accounting Bd. 2000) (hereinafter "FAS 138"). (443) FAS 133 P 3(a). 111
instruments," and therefore derivatives "should be measured at fair value, and adjustments to the carrying amount of hedged items should reflect changes in their fair value (that is, gains or losses) that are attributable to the risk being hedged and that arise while the hedge is in effect."(444) Third, "[s]pecial accounting for items designated as being hedged should be provided only for qualifying items."(445) In sum, in the absence of a hedge relationship, changes in fair value of a derivative should be recognized in income. Indeed, FAS 133 can be seen as a relatively brief statement requiring that changes in the fair value of derivatives be reflected in earnings, followed by a detailed set of requirements for limiting the impact of that principle by means of hedge accounting, with the caveat that hedge accounting is "special accounting" that may be used only in specified circumstances. Given this new standard, hedge accounting acquired even greater importance for Fannie Mae and other public companies with large derivative portfolios. The FAS 133 requirement that derivatives be recorded at their fair value as assets or liabilities, and that changes in fair value be recognized in earnings to the extent they are not linked in a completely effective hedge, had the potential to inject tremendous volatility into a company's financial statements. Fannie Mae was a vocal adherent of the popular view that this volatility was artificial and could be misleading to investors. During the months leading up to FAS 133's effective date, business journals and periodicals carried articles critical of FAS 133's treatment of derivatives; the articles highlighted the standard's complexity and its implications for injecting volatility into financial reporting.(446) Despite this criticism, FAS 133 took effect on January 1, 2001. 1. Requirements for Hedge Accounting Under FAS 133, "[o]ne aspect of qualification [for hedge accounting] should be an assessment of effective offsetting changes in fair values or cash flows during - ---------- (444) Id. P 3(b). (445) Id. P 3(d). (446) See, e.g., Simon Boughey, U.S. Issuers in Euromarket Are Snakebit by FAS 133, INVESTMENT DEALERS DIGEST, Feb. 22, 1999, at 13 ("Although it's too soon to tell exactly how FASB will interpret the new rules, it is already clear that FAS 133 is a major headache for American issuers in the international debt markets, and that some companies would rather pay greater funding costs than face balance sheet and income volatilities."); Paula Froelich, U.S. Companies Find New Accounting Rule Costly, Inefficient, DOW JONES NEWS SERVICE, Mar. 2, 1999 ("[A] lot of clients have been grumbling about the increased volatility - which could have increasingly negative consequences for the industry."); Joe Niedzielski, Firms Alter Derivatives Use As Rule Change Looms, WALL STREET JOURNAL EUROPE, Nov. 30, 2000, at 20 ("Some big U.S. corporations are altering their use of derivatives in the run-up to a new accounting rule that goes on the books next year."). 112
the term of the hedge for the risk being hedged."(447) The concept of "effectiveness," therefore, is critical to an understanding of hedge accounting and Fannie Mae's application of the standard. "Effectiveness" for these purposes means the extent to which changes in the fair value (or cash flows) associated with a derivative offset the changes in the fair value (or cash flows) associated with the hedged item (e.g., a debt instrument). If the offset is complete, then the hedge transaction is completely effective, and the derivative and hedged item can be said to form a "perfect hedge." The extent to which the changes in the fair value (or cash flows) of the derivative do not offset the changes in the fair value (or cash flows) of the hedged item is referred to as the "ineffectiveness" of the hedge. With this concept in mind, FAS 133 established criteria that must be satisfied to be eligible for the "special accounting" exception for qualifying hedge transactions. The rules regarding fair value and cash-flow hedges differ slightly, but in each case three elements are of particular importance: DOCUMENTATION -- At the inception of the hedge, a company must document the hedge relationship. The documentation must identify "the entity's risk management objective and strategy for undertaking the hedge, including the identification of the hedging instrument, the hedged item, the nature of the risk being hedged, and how the hedging instrument's effectiveness in offsetting the exposure to changes in the hedged item's fair value [or, in the case of a cash flow hedge, variability in cash flows] attributable to the hedged risk will be assessed."(448) It is generally understood that the documentation may be in any form (that is, electronic or paper media), and need not be in one location, so long as it is generated contemporaneously at the inception of the hedge relationship and includes all of the necessary elements.(449) ASSESSMENT OF EFFECTIVENESS -- Both at the inception of the hedge and on an ongoing basis, the hedging relationship must be determined to be "highly effective" in offsetting changes in the fair value (or cash flows) of the hedged instrument. The periodic assessments are required "whenever financial statements or earnings are - ---------- (447) Id. (emphasis added). (448) Id. PP 20 (fair-value hedge), 28 (cash-flow hedge). (449) Other accounting literature elaborates on this requirement. DOCUMENTATION OF THE METHOD USED TO MEASURE HEDGE INEFFECTIVENESS UNDER FASB STATEMENT NO. 133, EITF Topic No. D-102 (Fin. Accounting Standards Bd. Nov. 14-15, 2001), lists five elements that a company must include in its hedge documentation: (1) The hedging instrument; (2) the hedged item or transaction; (3) the nature of the risk being hedged; (4) the method that will be used to assess retrospectively and prospectively the hedging instrument's effectiveness; and (5) the method that will be used to measure hedge ineffectiveness. 113
reported, and at least every three months."(450) FAS 133 does not define "highly effective," but acknowledges that the "highly effective" requirement is satisfied if the similar, pre-FAS 133 "high correlation" requirement hedge accounting was met.(451) Under this principle, a hedge is "highly effective" when changes in the fair value (or cash flows) of the derivative and the fair value (or cash flows) of the hedged instrument offset within a range of 80 to 125 percent. MEASUREMENT AND RECOGNITION OF INEFFECTIVENESS -- As noted above, the documentation required at the inception of a hedge relationship must include the methodology by which the effectiveness of the hedge will be measured. The methodology specified at the outset of the hedge must be used for the duration of the hedge to determine whether the hedge relationship remains highly effective prospectively, and to measure any ineffectiveness associated with the hedge retrospectively. If the entity establishing a hedge relationship identifies an improved method for assessing or measuring effectiveness and wants to apply that method in the future, then it must discontinue the existing hedge relationship and designate a new relationship with new documentation.(452) Additional rules apply to accounting for anticipatory or forecasted transactions. Two rules of particular importance to these kinds of transactions are that the forecasted transaction be probable of occurring and that the documentation identify the hedged transaction with sufficient specificity so it is readily apparent when the hedged transaction has occurred.(453) 2. Accounting for Ineffectiveness As noted above, FAS 133 requires documentation at the outset of a hedge relationship that includes identification of (1) the method that will be used to assess that the hedge remains highly effective, and (2) the method that will be used to measure any ineffectiveness in the hedge transaction. Hedge accounting under FAS 133 is permitted only to the extent of the hedge's effectiveness. Ineffectiveness (measured as the excess of the cumulative change in the fair value of the derivative over the cumulative change in the fair value of the hedged item) is reflected in earnings. The accounting for the effective portion of a hedge depends on the type of hedge involved. In a fair-value hedge, the changes in the fair value of the derivative and the hedged item are both recognized in income: to the extent the hedge is perfectly effective, the changes in the fair values of the derivative and the hedged item will offset - ---------- (450) FAS 133 P 20(b). (451) Id. P 389. (452) Id. P 62. (453) Id. PP 29(b), 28(a)(2). 114
each other. Any ineffectiveness in the hedge - that is, the extent to which the changes in fair values do not offset each other - will result in an increase or decrease in reported income for the relevant period. In a cash-flow hedge, the accounting involves a deferral of changes in the fair value of the derivative for the effective portion of the hedge. To the extent that the hedge is effective, the change in the fair value of the derivative is recognized as a component of other comprehensive income ("OCI"). As with hedge accounting for fair-value hedges, any ineffectiveness is reported in earnings. The process of assessing the effectiveness of a hedge relationship, measuring the ineffectiveness of a hedge transaction, and recognizing any ineffectiveness in income is referred to popularly (but not in the text of FAS 133 itself) as the "long-haul" method of hedge accounting. To the extent that long-haul accounting results in either (1) disqualification of hedge transactions for further hedge accounting because they no longer satisfy the "highly effective" test, or (2) ineffectiveness, regardless of whether hedge accounting continues, then it will inject volatility into a company's income statements from accounting period to accounting period. This volatility can be eliminated entirely only if proper analysis supports the conclusion that the transactions are perfect hedges. It is widely acknowledged that the long-haul method of hedge accounting is cumbersome and complex. FAS 133 permits two alternatives to the long-haul method that may be used to account for certain types of hedge transactions, and these alternatives can simplify the accounting for these transactions. These methods permit an assumption that, in one case, a hedge will be highly effective at the outset (but subject to periodic re-assessment), and in the other case, that a hedge will be perfectly effective throughout its life. These two methods thus permit a company to avoid both the burden and complexity of long-haul accounting and the income-statement volatility associated with the recognition of ineffectiveness. The first alternative is referred to popularly as the "matched-terms method" of hedge accounting, although once again this term does not appear in the text of FAS 133 itself. This method, described in paragraph 65 of the standard, is based on the recognition that "[w]hether a hedge relationship qualifies as highly effective sometimes will be easy to assess, and there will be no ineffectiveness to recognize in earnings during the term of the hedge."(454) The matched-terms method is permitted when "the critical terms of the hedging instrument and of the entire hedged asset or liability . . . or hedged forecasted transaction are the same"; in that circumstance, the Company "could conclude that changes in fair value or cash flows attributable to the risk being hedged are expected to completely offset at inception and on an ongoing basis."(455) As an example of a transaction that would qualify for this matched-terms approach, FAS 133 describes - ---------- (454) Id. P 65. (455) Id. (emphasis added). 115
a hedge of a forecasted purchase of a commodity forward purchase contract that is for the same quantity of the same commodity as the forecasted purchase and that has a fair value of zero at inception. FAS 133 does not define "critical terms" for purposes of implementing the matched-terms method of hedge accounting. It does state, however, that ineffectiveness will result from differences between the derivative and the hedged item in such terms as notional amounts, maturities, quantity, location, or delivery dates.(456) The language of the standard leaves no doubt that this list was intended to be exemplary of a larger set of possible critical terms and was not meant to be all-inclusive. To the extent that the terms of a derivative and a hedged item are not "the same," the transaction does not qualify for matched-terms accounting. The matched-terms method is based on the recognition that ineffectiveness is likely to be minimal or nonexistent when the terms of a derivative and a hedged item are "the same." That observation, however, does not excuse the entity from the hedge documentation requirements described above or from monitoring the hedge relationship for ineffectiveness on a periodic basis. The standard requires, for example, that an entity assess on at least a quarterly basis that there has been no change to the criteria on which the initial matched-terms assessment was made; that is, the entity must "continue to assess whether the hedge meets the effectiveness test and also would measure any ineffectiveness during the hedge period."(457) If at any time the terms of the instruments in the hedge relationship are no longer "the same," then the matched-terms method no longer applies. Further, even if a company employs the matched-terms method, it still is required to specify in its documentation at the outset of the hedge how it will measure the ineffectiveness in a hedge transaction, should that be necessary.(458) The second alternative to long-haul accounting is known popularly as the "shortcut" method. This exception, described in paragraph 68 of FAS 133, applies only to hedge transactions involving interest-rate swaps and recognized assets or liabilities (that is, it applies only to hedges of existing instruments, and not to hedges of anticipated or forecasted debt issuances). To qualify for the shortcut method, among other criteria, the notional amount of the swap must "match" the principal amount of the hedged item, the expiration date of the swap must "match" the maturity date of the hedged item (for fair-value hedges), and the repricing dates on the interest-rate swap and the hedged asset or liability must "match" (for cash-flow hedges). FAS 133 Implementation Issue E4 confirms that the word "match" for purposes of the shortcut method means "exactly the - ---------- (456) Id. P 66. (457) Id. P 67. (458) Id. PP 28, 30, 67; see also CASH FLOW HEDGES: MEASURING THE INEFFECTIVENESS OF A CASH FLOW HEDGE UNDER PARAGRAPH 30(B) WHEN THE SHORTCUT METHOD IS NOT APPLIED, Statement 133 Implementation Issue No. G7 (Fin. Accounting Standards Bd. May 17, 2000) (revised July 11, 2000). 116
same."(459) If all of the requirements of the shortcut method are satisfied, then the entity is not required to assess effectiveness or measure ineffectiveness during the life of the hedge transaction.(460) Like the matched-terms approach, the shortcut method does not affect the requirement that the entity prepare appropriate documentation at the outset of the hedge. Both the matched-terms method and the shortcut method are available only when the fair value of the derivative is zero at the outset of the hedge relationship. This concept is important with respect to term-outs because the hedge transaction involves an existing derivative that is unlikely to have a fair value equal to zero at the inception of that hedge relationship. The concept is also important for derivatives that hedge callable debt: after debt is called, a derivative may be re-linked in another hedge transaction, but its fair value would not be expected to be equal to zero. 3. Term-Out Transactions and the "Fannie Mae Carve-Out" FAS 133 includes a number of examples that illustrate the proper application of hedge accounting principles. One example relates directly to Fannie Mae's term-out transactions. In fact, this example was included in FAS 133 in response to Fannie Mae's comment letter on the standard's exposure draft.(461) The example, therefore, has been referred to within Fannie Mae as "the Fannie Mae carve-out." The Fannie Mae term-out transactions are addressed in Example 8, "Changes in a Cash Flow Hedge of Forecasted Interest Payments with an Interest Rate Swap."(462) In the example, derived from a Fannie Mae term-out transaction similar to that described above, the hypothetical entity links the rollover of discount notes to a pay-fixed, receive-variable swap. Both the discount notes and the variable leg of the swap are set at LIBOR. The dates on which the interest rates are established for the variable leg of the swap and for the discount notes (the repricing dates) are the same. The example concludes that the entity may assume that there will be no ineffectiveness in this first hedge relationship. At the end of the second year, the entity discontinues its discount note program and "terms out" the rollover of discount notes by issuing a three-year, $5 million - ---------- (459) APPLICATION OF THE SHORTCUT METHOD, Statement 133 Implementation Issue No. E4 (Fin. Accounting Standards Bd. 1999) (revised 2003). (460) These descriptions provide a general overview of the recognized hedge accounting methodologies. Both the matched-terms and shortcut methodology have other requirements that are discussed below. (461) Letter from Sam. Rajappa to Director of Research and Technical Activities, Fin. Accounting Standards Bd., dated Oct. 6, 1997, FMSE-IR 217015-26. (462) FAS 133 PP 153-161. 117
debt instrument with interest payable every ninety days at the prime interest rate. The entity is no longer entitled to assume no ineffectiveness because the interest rate indices of the derivative and the debt instrument differ. The entity then enters into a basis swap with a $5 million notional amount and a three-year term, in which it pays LIBOR and receives interest payments at the prime rate. The entity documents a new hedging relationship, designating the two swaps in combination as hedging payments on the newly-issued debt. According to FAS 133, this transaction qualifies for hedge accounting: "Together, the cash flow from the two derivatives are effective at offsetting changes in the interest payments on the three-year note."(463) As initially released in 1998, FAS 133 also stated that this transaction qualified for the shortcut method of hedge accounting. However, the FASB deleted that sentence when it issued FAS 138. In its explanation for this change, the FASB stated that "the hedging instrument in Example 8 does not meet the criterion in paragraph 68(b) to qualify for the shortcut method. The hedging instrument does not have a fair value of zero at the inception of the hedge relationship."(464) The implication of this amendment is that the transaction would not qualify for matched-terms accounting either, as that method also requires that the instruments in the hedge relationship have a fair value equal to zero at the outset. Because the linked instruments would be effective at offsetting interest payments, the transaction would qualify for hedge accounting, but only under the long-haul method. IV. FINDINGS REGARDING FANNIE MAE'S IMPLEMENTATION OF FAS 133 During the years 1998 through 2000, Fannie Mae devoted considerable time and resources to the analysis of FAS 133 and to the formulation of policies and systems to implement the standard. Ultimately, the undertaking involved personnel from Financial Standards and Financial Reporting within the Controller's Office, as well as representatives of the Treasurer's Office, Portfolio, and others. The implementation activities did not end on the January 1, 2001 effective date, and Fannie Mae addressed a number of accounting and operational issues in 2002 and 2003 (separate from the implementation of FAS 149 in 2003). In 2003, for example, the Company considered a substantial overhaul of its derivative accounting systems, with the ultimate objective of developing the capability to account for hedge transactions under the long-haul method. That effort, however, was postponed with the commencement of OFHEO's Special Examination. The substantial record that we have developed of the Company's FAS 133 efforts leads to several significant conclusions regarding Fannie Mae's implementation of FAS 133. The objective, or at least a principal focus, of the FAS 133 implementation - ---------- (463) Id. P 161. (464) FAS 138 P 38. 118
project was to treat Fannie Mae's hedge transactions as "perfect hedges," and thereby to utilize the shortcut method of hedge accounting so as to avoid both the volatility in earnings that would result from the recognition of ineffectiveness and the need to implement complex changes either to the Company's business model and its accounting and financial reporting systems. In order to achieve this result, the Company adopted policies that deviated from the requirements of FAS 133. These policies were established with the knowledge and, in some cases, active involvement of Howard, Leanne Spencer, Jonathan Boyles, and others. A. The Three Tenets As discussed in more detail below, Boyles would recount in 2003 that Fannie Mae's implementation of FAS 133 had been driven by three tenets: avoiding earnings volatility; leveraging existing accounting systems; and ensuring that Fannie Mae's operating earnings were simple and understandable. (465) While these objectives were not per se extraordinary, and almost certainly were shared by other companies with substantial derivative portfolios, we believe the Company took excessive measures to achieve them. In this section, we discuss the development of these tenets, particularly the first. Because Fannie Mae tracked FASB activities routinely, the Company was aware of FAS 133 in concept at least as early as 1996. In 1997, Fannie Mae sent a letter to the FASB commenting on certain aspects of the FASB's approach to hedge accounting,(466) and in 1999 sent another letter requesting an extension of the effective date of January 1, 2000.(467) In requesting the extension, Fannie Mae cited the complexity of the accounting changes FAS 133 would require and the fact that the effective date would require implementation of new hedge accounting systems at the same time that companies were addressing Y2K issues.(468) Fannie Mae's submission was one of many that the FASB received raising similar concerns. The effort to implement FAS 133 began soon after the FASB's release of the final standard in June 1998. Initially, in mid-1998, a small group of individuals, including individuals from the Controller's Office and the Treasurer's Office, participated in an off-site meeting to study the standard and develop an initial understanding of how it would affect Fannie Mae's transactions, accounting, and - ---------- (465) Mem. from Boyles to Distribution, dated Mar. 2, 2003, FMSE 78540-42 (hereinafter "Boyles March 2003 Memo"). (466) Letter from S. Rajappa to Director of Research and Technical Activities, Fin. Accounting Standards Bd., dated Oct. 6, 1997, FMSE-IR 217015-26. (467) Letter from Timothy Howard to Edmund Jenkins, dated Apr. 12, 1999, FMSE-IR 193011-13. (468) Id. at FMSE-IR 193012. 119
systems. The process has been described as an exchange in which Treasurer's Office personnel explained the Company's debt and hedge transactions to the accountants, and the accountants considered the application and implications of FAS 133 to those transactions. Boyles described the off-site meetings in a draft memorandum he prepared early in the process.(469) After describing the meeting, Boyles explained the benefits that the Company would enjoy if its hedge transactions qualified as "perfect hedges": The final standard includes language regarding the concept of a `perfect hedge' that we believe will prove beneficial in the adoption of the standard with a minimum of core accounting system changes. If Fannie Mae's derivative transactions can qualify under the `perfect hedge' criteria, then we should be able to maintain our current `accrual accounting' systems and there would be little impact on our income statement. In instances where we do not qualify for `perfect hedge' accounting treatment there would be some income volatility in earnings from the new standard. The basic premise I am using to evaluate the standard's impact on Fannie Mae is the effect on income.(470) Boyles then reported that Fannie Mae uses "several `term-out' transactions that don't fit neatly into the carve-out the FASB gave us back in December 1997."(471) On this issue, Boyles proposed returning to the FASB for more guidance or altering Fannie Mae's term-out transactions to match the guidance in FAS 133 more closely.(472) Boyles also noted concerns about the treatment of anticipated debt issuances and reported that FAS 133 "could have a significant effect on earnings, particularly in periods where spreads are volatile."(473) Boyles concluded, among other things, that "[w]e anticipate a major systems effort to strengthen the controls associated with this standard so that we meet the necessary criteria to apply the `perfect hedge' rules for our derivative instruments."(474) - ---------- (469) Mem. from Boyles to Rajappa, Linda Knight, Leanne Spencer, Mary Lewers, Mehmood Nathani, Donald Sinclair, and KPMG, et al., dated Sept. 14, 1998, FMSE-IR 217361-64. (470) Id. at FMSE-IR 217361. (471) Id. (472) See id. at FMSE-IR 217363. (473) Id. (474) Id. at FMSE-IR 212364. 120
These early efforts to address FAS 133 were discussed at a meeting with Howard in early December 1998, as reflected in a December 4, 1998 memorandum and associated talking points prepared by Spencer.(475) According to the talking points, Spencer's main objectives at the meeting were: - To provide Howard with a tutorial on basic FAS 133 concepts; - To describe for Howard the work completed up to that time; - To explain that the Company had identified certain transactions that would not qualify for the "perfect hedge" exception; that the term-out transactions "are the culprit" in this respect; and that "we either must adjust our funding on the transactions or get out prior to implementation in order to eliminate earnings volatility"; - To explain further that, with respect to anticipated debt issuances, the "inefficient [ineffective] portion of the hedge will run through current earnings," which will "result in much more earnings volatility"; - To describe for Howard the implementation group's work plan, which included a general description of systems development efforts, along with a detailed project plan to be completed by January 31, 1999.(476) In describing the work completed up to that time, the talking points noted that: - The working group had "developed a working draft describing all affected transactions and what the accounting should be"; - The Company had met with KPMG and had described each transaction and what the proposed accounting would be; and - "The key tenet we are working towards is that we want to fit all of our transactions into the FASB's `perfect hedge' exception."(477) Spencer's memorandum documenting this meeting describes Howard's reactions on these points and his instructions for going forward. According to Spencer, - ---------- (475) Mem. from L. Spencer to T. Howard, et al., dated Dec. 4, 1998, FMSE-IR 192988-992. (476) See id. at FMSE-IR 192991-92. (477) Id. at FMSE-IR 192990. 121
Howard's "first point was that we have not stopped doing some of the transactions for which we have not found an accounting solution that would keep the income statement neutral."(478) Howard also observed that "the new standard will affect so many different aspects of our business" and that Spencer should "make sure [she] got other departments involved in the implementation group."(479) A separate set of talking points, which Spencer appears to have used for her briefing to her colleagues following the meeting with Howard, refers to Howard's "vision that all departments should be working together and that all will be held responsible for ensuring that we adopt the standard."(480) Based on these principles, in 1999 and 2000 the Company engaged in an extensive and systematic effort to implement FAS 133. The inter-departmental working group assigned to the project (per Howard's instructions) considered a number of technical issues, reported periodically to senior management (at the senior vice president level, including Spencer and Knight, the Company's Treasurer), and consulted with Howard (albeit less frequently). A common thread that runs throughout this work, as during the planning meetings in 1998, is the objective to avoid, or at least limit, earnings volatility by fitting all of Fannie Mae's hedge transactions into the "perfect hedge" category. For example, notes handwritten on stationery from a Fannie Mae officers' meeting held on December 8-11, 1998, state, under the heading "FAS 133," "`Perfect hedge' is the goal for FNM accounting."(481) An outline describing the Company's implementation effort prepared in 1999 by Kimberly Stone,(482) one of the principal participants involved in the implementation effort, stresses the involvement of senior officers in the project and states that "one of our main goals is to implement this very complex standard and to continue our robust hedging program but at the same time minimize our earnings and equity volatility."(483) - ---------- (478) Id. at FMSE-IR 192989. (479) Id. (480) Undated Leanne Talking Points, FMSE-IR 193003. (481) Undated handwritten notes, FMSE-SP 70111. No one we had the opportunity to interview recognized the document or recalled the meeting, and it is possible that the document was prepared at a later date. A document with similar handwriting is signed "Tim," raising at least the possibility that the document was prepared by Howard. (482) Kimberly Stone is referred to in some documents as Kimberly Rawls, her maiden name. (483) Undated outline, FMSE-SP 84562-66, at FMSE-SP 84566. 122
The team involved in the FAS 133 implementation effort regularly briefed senior executives. A recurring topic at these meetings was earnings volatility. A topic on the agenda at the June 23, 1999 meeting, for example, was "Ownership of Volatility."(484) An "Update on FAS 133 Project," dated July 1, 2000, lists under "Critical Next Steps": "Determine threshold for earnings and equity volatility (As Soon As Possible)."(485) At the July 26, 2000 meeting, the discussion concerned the education of investors and others on FAS 133's impact; a handwritten note by that agenda items reads: "Tim thinks we could explain $30 mm per qtr of earnings volatility."(486) In an e-mail dated July 21, 2000, Mehmood Nathani, a participant in the FAS 133 efforts from the Treasurer's Office, reported that he and others had met with Howard to discuss the FAS 133 implementation. Nathani announced: We will continue to explore delta hedging as a potentially important way to reduce FAS 133 related earnings volatility. Tim is more interested in Delta Hedging now because of the market's adverse and unreasonable reaction to our latest earnings release. He can no longer feel even somewhat assured that the analysts and external folks will not misinterpret FAS 133's impact when we go live. Therefore, reduction of earnings volatility is still our numero uno objective.(487) - ---------- (484) Agenda, dated June 23, 1999, FMSE-SP 70416. (485) Update on FAS 133 Project, dated July 1, 2000, FMSE 436304. (486) Agenda, dated July 26, 2000, FMSE 86508. (487) E-mail from M. Nathani to Nadine Bates, dated July 21, 2000, FMSE 436303. It appears from the e-mail that the other attendees included Knight, Spencer, and Boyles. Nathani's reference to "the market's adverse and unreasonable reaction" appears to refer to the decrease in the Company's stock price in mid-2000. An earnings release dated July 13, 2000, just eight days before Nathani's e-mail, included the following statement: "Howard also said that the Company had a $46.3 million reduction in fee and other income, net compared with the first quarter, resulting largely from the recognition of a loss on a Benchmark Note hedge early in the second quarter. Howard noted that the dislocation in the agency debt markets that caused the hedge to go out of correlation and triggered immediate recognition of the loss also presented opportunities to repurchase debt on highly attractive terms. Such debt repurchases led to an extraordinary gain of $50.3 million ($32.7 million after tax) in the second quarter." Fannie Mae press release, dated July 13, 2000, Zantaz document 3785175, at 2. 123
Nathani concluded: "Separately, I believe we need to examine all possible ways to reducing [sic] earnings volatility."(488) During the implementation process, Fannie Mae adopted a procedure for the approval of transactions that would not meet the Company's "perfect hedge" guidelines. Any transaction that would not qualify as a "perfect hedge" required the approval of three Company Senior Vice Presidents: the Treasurer; the Senior Vice President--Portfolio Strategy; and the Senior Vice President--Portfolio Management. The signatures acknowledged the officers' "acceptance of accountability for earnings volatility" associated with long-haul accounting.(489) Several interviewees explained that the concern underlying the form was not the earnings volatility per se, but the burdens associated with long-haul calculations. The language of the form does not comport with that explanation, and a memorandum accompanying a similar form (relating to transactions that would not qualify for hedge accounting at all) explains that the approving official "must assume responsibility for any changes in the market value that will flow through earnings."(490) In any event, the proposition that senior officers had to acknowledge, and take responsibility for, the burdens associated with a transaction that did not meet the Company's "perfect hedge" criteria reflects Fannie Mae's overall approach to implementation of the standard, i.e., that Fannie Mae sought to avoid transactions that, in its view, required any assessment or recognition of ineffectiveness. On September 5, 2000, the team responsible for the implementation of FAS 133 made a presentation to the Office of the Chairman.(491) After a brief overview of FAS 133's basic principles, the presentation focused on "business options under FAS 133." This topic included several possible actions to address the earnings volatility associated with FAS 133, such as reducing the Company's use of derivatives, explaining earnings volatility to investors and Wall Street analysts, or using different portfolio strategies to minimize volatility.(492) Under the heading "Challenges Presented by FAS 133," the presentation raises five points: (1) the time value of options would be marked to fair value through earnings; (2) significant volatility in earnings would result from the large volume of cash-flow hedges in the Company's portfolio; (3) any hedge ineffectiveness would be recorded to earnings; (4) some cash-flow hedges still would not - ---------- (488) E-mail from M. Nathani to N. Bates, dated July 21, 2000, FMSE 436303. (489) Derivatives Accounting Guidelines, Dec. 2003, FMSE 112657-113142, at FMSE 112936. (490) Mem. from J. Boyles to Distribution, dated June 21, 2001, FMSE 79004. (491) See FAS 133 Briefing, dated Sept. 5, 2000, FMSE-IR 195328-46. (492) Id. It appears that, by this time, the Company recognized that FAS 133 would result in some volatility in earnings as a consequence of the Company's inability to include changes in the time value of options in measuring and assessing the effectiveness of a hedge relationship. 124
qualify for hedge accounting; and (5) there would be a large, one-time earnings impact in the first quarter of 2001 due to the change in accounting principle.(493) According to a report on this meeting that Stone distributed two days later, the proposals offered at the meeting to address the volatility issue were not well-received. "It appears that a number of the strategies that we were pursuing did not sit will with the executives (i.e., large AFS portfolio, large swings in equity and political risk, breaking out operating versus other earnings, etc.)."(494) Ultimately, Fannie Mae adopted the two-pronged approach to its financial reporting mentioned in Stone's report. Beginning in 2001, Fannie Mae reported GAAP earnings based on its application of FAS 133 and a separate earnings measure called "core earnings," which was the GAAP earnings measure adjusted to eliminate the impact of the change in the time value of purchased options and include amortization of the cost of the purchased options. Although analysts accepted Fannie Mae's alternative reporting approach and discounted swings in earnings that resulted from FAS 133, it is noteworthy, for reasons discussed below, that as late as September 2000 senior management was concerned that this approach would not be well-received. The Company's ability to adopt FAS 133 with minimal interference to the Company's business operations or impact on earnings was important to individual employees involved in the process. Several documents - including a summary, prepared by Stone with Boyles's assistance, of the process by which Fannie Mae implemented FAS 133 - note that the implementation was incorporated into the personal bonus plans of the Fannie Mae executives. In October 2000, several senior executives involved in the FAS 133 project - including Knight and Spencer - nominated a number of the other individuals involved in the FAS 133 implementation effort for a special Chairman's Award. The executives stated: "Never before has the adoption of a new accounting standard had such a significant impact on Fannie Mae's business practices. Had it not been for the hard work and ingenuity of this team, FAS 133 would have created billions of dollars in annual earnings volatility. As a result of their extraordinary efforts, Fannie Mae is now in a position to adopt FAS 133 on January 1, 2001 with a significantly reduced impact to its bottom line."(495) The major goals in the implementation effort were summarized in a 2003 retrospective that Boyles prepared as part of a more general review of the Company's hedge strategies. In the course of (or in anticipation of) that review, Boyles wrote a draft - ---------- (493) See id. at FMSE-IR 195335. (494) E-mail from K. Stone to M. Nathani, dated Sept. 7, 2000, FMSE 436284. Interviewees differed on their interpretation of "political risk," although the predominant view was that the swings in Fannie Mae's reported equity from period to period would draw unwelcome attention to, and possibly misinterpretation of, Fannie Mae's financial statements by the Company's critics in Congress. (495) Undated Chairman's Award Nomination Form, FMSE-SP 89437. 125
memorandum in which he described the Company's implementation of FAS 133. He explained that the implementation effort was guided by the three "tenets" mentioned previously: (1) "Earnings volatility was to be minimized and if there were earnings volatility it should be as predictable as possible"; (2) "We were to leverage off existing systems as much as possible"; and (3) "Operating earnings needed to be simple and easily understood."(496) Boyles explained that the Company "minimize[d] the potential for earnings volatility" by "divest[ing] itself of any derivative that either did not qualify for the `short-cut' method or was not in a hedging relationship where we could assume no ineffectiveness."(497) Further, this principle "was deemed to be very important because by adopting this policy we were able to maintain our existing systems for accruing interest income and interest expense."(498) Boyles noted that "[t]he cornerstone of [the] system was that only transactions where there would be no ineffectiveness would flow into the system. This decision was key because the most important piece of the systems implementation effort was that we would continue to use our existing systems to record interest income and expense."(499) Remarking that the Company already had controls in place with regard to the recording of interest income and expense, he stated: "Had we allowed for ineffectiveness to be included in income/expense we would have severely complicated the implementation effort and the month end close of our financial records."(500) Although it is not clear that Boyles ever completed or distributed his memorandum, the salient points were incorporated into a presentation to the Office of the Chair in May 2003.(501) One of the slides included in the presentation materials repeats the three goals that the Company followed when it implemented FAS 133: "Minimize earnings volatility"; "Leverage existing systems"; "Keep operating earnings simple."(502) According to the presentation, one of the issues on the table for discussion was "Could we design effective hedge strategies that would deliver stable and predictable performance and would allow us to move to a GAAP only presentation?"(503) - ---------- (496) Boyles March 2003 Memo at FMSE 78540. (497) Id. (498) Id. (499) Id. (500) Id. at FMSE 78541 (501) See Examining Our Hedging Strategies - Post FAS 133, A Progress Report, dated May 9, 2003, FMSE 27242-65. (502) Id. at FMSE 27243. (503) Id. at FMSE 27244. 126
Boyles's 2003 memorandum includes the observation that the "three tenets caused the implementation group to make several key decisions that, given today's environment, we may not have made with 20-20 hindsight."(504) Boyles explained during an interview that he believes the Company would have accepted more volatility into earnings had it appreciated in advance how readily the market would accept the Company's "core earnings" as an alternative to the more volatile GAAP income measure that included the volatility associated with options. That explanation, however, does not address the impact of the tenets on those involved in the implementation process prior to 2001. Prior to 2001, the Company did not know whether investors and analysts would accept the Company's core earnings measure and disregard the volatility that remained in the Company's reported GAAP earnings as a result of FAS 133. In our view, the record strongly supports the conclusion that the Company's interpretation of FAS 133 - which, as we explain below, disregarded or "interpreted away" unambiguous language in the statement - was designed to avoid earnings volatility. The goals established at the outset of the process, and confirmed at various times from the Spencer-Howard meeting in December 1998 up to and including Boyles's 2003 memorandum, resulted in more than one departure from the requirements of FAS 133. No one that we interviewed acknowledged this point explicitly, although several acknowledged the three tenets, at least in concept. Even those who recognized the three tenets, however, could not point to an express instruction from a senior officer to follow them.(505) Because Howard declined to cooperate with our investigation, we did not have the opportunity to question him regarding these documents or his views on FAS 133 implementation generally. Several interviewees reported that Howard's overarching instruction was "Get it right," and there is evidence that Howard intended that all of those involved would `be held responsible for ensuring that we adopt the standard."(506) That directive was belied (or at least colored), however, by other guidance regarding the approach he apparently expected the Company to take when implementing the standard (for example, Nathani's instruction after a meeting with Howard that "reduction of earnings volatility is still our numero uno objective").(507) In addition, several documents reflect Howard's personal view that the concepts underlying FAS 133 were ill-conceived. In a July 4, 2002 e-mail, for example, Howard explained: "These wild and unpredictable swings in our GAAP net income-- - ---------- (504) Boyles March 2003 Memo at FMSE 78540. (505) Boyles suggested that the tenets formed in his mind through trial and error. He explained that, as certain ideas were considered and rejected for particular reasons, he developed a sense of what was unacceptable and what was expected. (506) Undated Leanne Talking Points, FMSE-IR 193003. (507) E-mail from M. Nathani to N. Bates, dated July 21, 2005, FMSE 436303. 127
that are completely artificial and convey no real information about the economic substance of our business-- make our GAAP net income useless to investors and security analysts. Analysts and investors uniformly agree that this is true."(508) In his talking points for remarks to the Fannie Mae Board in February 2001, Howard explained that FAS 133's accounting requirements are "not particularly meaningful" because "net income and EPS numbers that move around this much based on changes in the market values of our purchased options - and nothing else on the balance sheet - don't give an accurate picture of our financial performance."(509) To the extent that this opinion of the accounting standard was conveyed to the individuals who were responsible for implementation of the standard, it undoubtedly affected any message that might have been conveyed about "getting it right."(510) In sum, the record indicates that Fannie Mae's principal objective in its implementation of FAS 133 was to avoid earnings volatility (other than with respect to the time value of purchased options) by ensuring that all of the hedge transactions it entered into after the standard's effective date qualified as "perfect hedges." There is substantial evidence as well that Howard established this goal and that many of the individuals responsible for the project, including at least Spencer and Boyles, were aware of Howard's concerns. B. Development Efforts During 1999 and 2000 By 1999, following Howard's December 1998 instructions to Spencer, the group of Company employees who were spending a substantial amount of their time on FAS 133 issues had expanded significantly to include additional personnel from the principal offices affected, as well as representatives of other disciplines (legal, tax, etc.). The principal Controller's Office personnel involved in the development of Fannie Mae's accounting policies under FAS 133 were Boyles and Stone from Financial Standards and Lewers from Financial Reporting. Nathani was the principal representative of the Treasurer's Office and participated primarily by describing the affected transactions and performing analyses. (Others involved in particular aspects of the FAS 133 implementation effort are identified below in connection with the developments in which - ---------- (508) E-mail from T. Howard to L. Spencer, dated July 4, 2002, FMSE-SP 35056-57. (509) E-mail from T. Howard to L. Spencer, dated Feb. 20, 2001, FMSE 429280-86, at FMSE 429281. (510) Certain statements - such as Spencer's December 1998 talking points in which she referred to Howard's instruction that "all will be held responsible for ensuring that we adopt the standard" - might be read to support a conclusion that Howard said "get it right"; overall, however, these statements are too cryptic and infrequent (and belied by other evidence, such as Spencer's point from December 1998 that "[t]he key tenet we are working towards is . . . the FASB's `perfect hedge' exception") to support the conclusion that strict compliance with GAAP was Howard's main concern. 128
they were involved.) Although Stone and Nathani left the Company for personal reasons in 2001 and 2002, respectively, we interviewed them and the other principal participants in the effort during this period. Fannie Mae's FAS 133 working group met regularly throughout 1999 and 2000 to discuss various issues and to make decisions on accounting, financial reporting, and systems matters. The documentary record of our review is replete with agendas, action item lists, internal correspondence (e-mails and memoranda) and similar documents reflecting the time and attention that the personnel involved devoted to this effort. The implementation team addressed a variety of issues at a considerable level of detail. Although the working group regularly briefed senior officers, including Spencer and Knight, interviewees involved in the process were unable to identify a decision-maker with respect to the Company's hedge accounting policies, except perhaps Boyles. Beginning in 1999, the Company began the detailed development of its FAS 133 accounting policies. This resulted in the preparation of a volume, drafted largely by Stone under Boyles's direction, known as the Derivatives Accounting Guidelines (the "DAG").(511) The DAG contains a statement of the Company's approach to (among other things) hedge documentation, assessment of hedge effectiveness at the outset of the transaction and on an ongoing basis, and other aspects of the accounting for hedge transactions in Fannie Mae's books and records. The DAG also contains a "menu" of permitted hedge transactions and the accounting associated with the transactions, many of which were deemed to qualify for treatment as "perfect hedges" under the "shortcut" method. As noted above, any transaction that the Company entered into that did not qualify for shortcut treatment per the DAG required the approval of senior officers. The DAG, and not the text of FAS 133 itself, became the standard for hedge accounting at Fannie Mae. Interviewees routinely stated that Company personnel looked to the DAG, rather than FAS 133, for guidance on hedge accounting matters. Even Controller's Office personnel who were accountants or who were otherwise familiar with FAS 133 stated that they did not consider the language of FAS 133 in determining the appropriate accounting for Fannie Mae's hedge transactions. They explained that Financial Standards had adopted the DAG as an interpretation of FAS 133 for Fannie Mae's purposes, and the DAG therefore was authoritative within the Company. They also noted that the DAG had been "approved" by KPMG. - ---------- (511) DAG at FMSE-IR 12825-13295. The DAG was revised as of December 2003. See Derivatives Accounting Guidelines, dated Dec. 2003, FMSE 112657-3142. All citations in this section of our Report are to the 2001 edition of the DAG. 129
The Company implemented changes in its accounting systems. A document labeled "FAS 133 Accounting Project Kick-Off Meeting," dated January 1999, describes this activity.(512) A central paragraph states: Qualifying for one of FASB's hedge designations is paramount for Fannie Mae in applying the accounting standard. A critical component in qualifying for a hedge designation is demonstrating a linkage between the derivative and its related debt. The present accounting systems do not provide a sufficiently rigorous link. A key component of the FAS133 [sic] project will be to supplement STAR [an existing accounting system that tracked debt instruments and derivatives] data with the required linkage information. The linkage will be derived using data provided by a Fannie Mae Treasury maintained management information system.(513) Essentially, the process as outlined would: (1) establish a system to specify the debt and derivative instruments to be linked in a hedge relationship; (2) identify each link as establishing a cash flow or fair-value hedge (or identify a transaction as "does not qualify" for hedge accounting); (3) calculate fair values of the relevant instruments; and (4) create subledgers, journal entries and analytic reports. The Company devised two systems to implement its FAS 133 accounting and reporting functions. An existing system maintained in the Treasurer's Office, called DEBTS, was modified to establish an automatic link between a derivative and a debt instrument. DEBTS tracked benchmark notes and longer-term debt instruments. (The Short-Term Notes ("STN") application tracked discount notes, and the information in STN was transmitted nightly to DEBTS.) The DEBTS system would establish linkages between the derivatives the Company acquired and its portfolio of debt instruments to form hedge relationships. Initial linkages usually were established when the Company acquired the derivative. However, the DEBTS system would relink debt and derivatives if an initial link were severed (for example, by the discontinuance of discount note rollovers or the call of a debt instrument). The system contained a protocol for assigning derivatives to debt by which term-out transactions and long-term debt would be linked to available derivatives before discount notes. On a monthly basis, information regarding the linkages in the DEBTS system would be delivered to a system known as the FAS 133 Accounting System, which was created for the purpose of complying with FAS 133. Maintained in the Controller's - ---------- (512) See FAS 133 Accounting Project Kick-Off Meeting, dated Jan. 1999, FMSE-E 415603-05. Although this document is labeled a draft, its contents are reflected in other documents and are consistent with interviewees' statements. (513) Id. at FMSE-E 415603. 130
Office, the FAS 133 Accounting System would corroborate the linkage established in DEBTS and identify the hedge accounting designation. The FAS 133 system would draw the designation from the menu of permitted transactions listed in the DAG. For example, a pay-fixed swap linked to the anticipated rollover of discount notes would be designated by the system as a "Cash Flow #1" - the title of that transaction in the DAG. The system would post necessary entries to the Company's general ledger and issue a report on any improper linkages, which would be addressed manually. Separately, procedures were developed for the Company's Hedge Desk, which was responsible for transactions involving hedges of the anticipated issuances of debt (excluding the rollover of discount notes). We have uncovered no evidence that Fannie Mae seriously considered during this period developing and implementing a system capable of long-haul accounting for hedge transactions. Several interviewees confirmed that no serious consideration was ever given to this idea. At one point, the Company did investigate the possibility of obtaining "off-the-shelf" hedge accounting software but abandoned that effort when it became apparent that no such software was available. Instead, the evidence indicates that Fannie Mae was concerned about preserving "perfect hedges." A document dated March 30, 2001, captioned "FAS 133 Implementation," provides a retrospective outline of the FAS 133 implementation project: it states under the heading "Accounting Operations" that "Systems [were] built to implement the short-cut method."(514) A similar document, titled "Derivative Accounting," reports under the heading "Operational Principles" that: "[a]ccounting process [was] designed to avoid any earnings volatility (except for time value pertaining to options)."(515) These points are consistent with two of the basic tenets discussed in the previous section, i.e., minimizing volatility by ensuring that all transactions were treated as "perfect hedges" and utilizing the Company's existing accounting systems (which did not have a long-haul capability) to the extent possible. In 1998, the target date for the implementation of FAS 133 accounting was January 1, 2000. In May 1999, the FASB released an exposure draft of FAS 137, which proposed deferring the effective date of FAS 133 by one year. The FASB acknowledged the challenges that FAS 133 presented and the fact that adherence to the initial effective date (January 1, 2000) would require companies to implement FAS 133 systems at the same time they were addressing any operational issues associated with "Y2K." The addition of a year to prepare, however, did not alter the Company's approach to FAS 133 in any significant way. Lewers, in fact, recalled that the FAS 133 working group made a conscious decision not to use the additional year to refine its approach to FAS 133 by improving its systems capabilities or otherwise. - ---------- (514) See FAS 133 Implementation, dated Mar. 30, 2001, FMSE 415107-09. (515) See Undated Derivative Accounting, FMSE 415372. 131
C. FAS 138 One of the most significant developments during the FAS 133 implementation period was the release in March 2000 of an exposure draft of FAS 138, in which the FASB proposed to amend FAS 133 prior to its effective date. In FAS 138, the FASB addressed a series of issues that had been raised in connection with FAS 133 and made certain clarifying amendments. One of those amendments affected the so-called "Fannie Mae carve-out" for term-out transactions. The FASB removed from FAS 133 a sentence that had provided that the term-out transaction could be treated under the shortcut method of paragraph 68. In its explanation for the change, the FASB stated: "The hedging instrument does not have a fair value of zero at inception of the hedging relationship" and thus "the hedging instrument . . . does not meet the criterion in paragraph 68(b) to qualify for the shortcut method."(516) The FASB also addressed in FAS 138 the treatment of another common Fannie Mae transaction: the rollover of discount notes hedged by a pay-fixed swap. The FASB explained that this transaction, too, did not qualify for hedge accounting using the shortcut method: The swap is designated in a hedge of a series of forecasted interest payments, only one of which relates to a recognized interest-bearing liability [that is, the initial discount note]; the remainder relate to a forecasted borrowing. The shortcut method is limited to either a fair value or cash flow hedging relationship of interest rate risk involving an existing recognized interest-bearing asset or liability and an interest rate swap. Thus, a cash flow hedge of the variability in interest on a probable forecasted lending or borrowing [that is, the future rollover of the discount notes] is not eligible for the shortcut method.(517) Fannie Mae recognized the importance of these changes. Stone's handwritten notes captioned "133 Amendment Issues" stated, among other things, "P 161 [part of the Fannie Mae carve-out in Example 8] - No longer qualify for shortcut method" and "Takes away our argument to apply shortcut method in other places."(518) In a memorandum Stone prepared dated March 8, 2000, she stated: "Another surprise in the Exposure Draft release was that the FASB deleted the ability to qualify for the shortcut - ---------- (516) FAS 138 P 38. (517) Id. P 39. (518) Undated handwritten notes, FMSE 366653. Boyles told us that the amendment did not affect the accounting for term-outs because Fannie Mae was not applying the shortcut method. We address that issue below. 132
method in the Fannie Mae 'carve-out' in the standard. Jonathan will be following up with Leanne re: the implications this may cause."(519) Notwithstanding these observations, Financial Standards concluded that there would be no implications associated with these changes. Stone's notes state, "as long as critical terms are the same, assume no ineffectiveness."(520) The issuance of FAS 138, in sum, did not alter the Company's approach to FAS 133 insofar as the treatment of term-out transactions or hedging the rollover of discount notes were concerned. D. Post-Effective Date Developments Fannie Mae's policies with respect to the implementation of FAS 133 remained relatively static during 2001 and 2002. Several developments in 2003 and 2004, however, are worthy of note. In 2003, at Raines's direction, Fannie Mae undertook an examination of its hedging practices and its overall approach to interest-rate risk. Although this effort did not focus on hedge accounting issues, or FAS 133 specifically, Boyles took this opportunity to review the Company's implementation of FAS 133 during 1998-2000. Apparently, this project prompted Boyles to write the draft March 2003 memorandum that described the three tenets that the Company followed in its implementation of FAS 133 and to incorporate these points into the May 2003 presentation used during a meeting with Raines and others. Both Raines and Boyles stated during interviews that Raines objected to the proposition that the Company was reviewing its hedging policies at that time in an effort to avoid earnings volatility. According to both men, Raines stated that his focus was on effective economic hedging and the maximum avoidance of interest-rate risk as an economic matter rather than for accounting purposes. It does not appear, however, that senior management made those views known prior to 2001, when Howard and others had substantial concerns about volatility in GAAP earnings and the extent to which investors and analysts would accept the Company's alternative "core earnings" measure of performance. Moreover, according to Stone's report of the September 5, 2000 meeting with the Office of the Chair, senior management was concerned that investors and market analysts would disregard the Company's alternative measure of income, and focus instead on the volatile GAAP income statistics.(521) Although Raines may not have been concerned about volatility in Fannie Mae's GAAP earnings in 2003 (by which time the market had accepted the - ---------- (519) Mem. from K. Rawls to Yolanda Green, dated Mar. 8, 2000, FMSE 436300. (520) Undated handwritten notes, FMSE 366653. (521) E-mail from K. Stone to M. Nathani, dated Sept. 7, 2000, FMSE 436284. 133
Company's core earnings alternative), it appears that senior executives were concerned about volatility, and expressed those concerns, right up to FAS 133's effective date.(522) Also in 2003, the Company initiated a program to upgrade its hedge accounting systems with a view toward incorporating into the systems an automated capability to perform long-haul accounting. This effort was prompted by concerns, apparently voiced in the Treasurer's Office, that the menu of transactions approved for hedge accounting because they resulted in "perfect hedges" was too restrictive. This project was not completed, as Fannie Mae's resources later were devoted to the OFHEO Special Examination and other matters. In the first half of 2004, Fannie Mae engaged KPMG to conduct a review of the Company's FAS 133 policies. In the course of this engagement, KPMG reviewed the Company's accounting for fourteen different types of hedge transactions, which account for over ninety-five percent of the Company's hedge transactions. Most of the review was conducted by KPMG's audit team, although certain questions were referred to KPMG's Department of Professional Practice. KPMG confirmed the correctness of the Company's accounting for all of the transactions but one. The Company raised that transaction with the FASB. In a telephone call, which was later confirmed in a letter, Fannie Mae described the term-out transaction at issue.(523) According to the letter confirming the discussion, the FASB agreed that the transaction was eligible for hedge accounting. Notably, the letter does not disclose whether the parties discussed the type of accounting that would be required - that is, whether the Company could rely on the shortcut method for this transaction, or whether it would be required to apply the long-haul method of accounting.(524) Finally, in early 2004, Boyles issued an accounting policy memorandum amending certain aspects of Fannie Mae's FAS 133 accounting policies.(525) The two most significant changes were a prohibition on further use of a duration matching test for the hedging of anticipated debt issuances, and the elimination of a de minimis test to - ---------- (522) Raines did not recall attending the September 5, 2000 briefing. (523) In the transaction, Fannie Mae issued a callable, fixed-rate medium-term note to replace the rollover of discount notes and entered into a receive-fixed, pay-floating interest-rate swap that was cancelable by the counterparty. Fannie Mae then designated the new interest-rate swap in combination with an existing pay-fixed, receive-floating interest-rate swap as a hedge of the issuance of discount notes beyond the maturity date of the medium-term note. The issue Fannie Mae discussed with the FASB staff related to whether the cancelable interest-rate swap could qualify as a hedge of the callable medium-term note because of the option embedded in the interest-rate swap. (524) Letter from J. Boyles to Robert Wilkins, dated July 21, 2004, FMSE 216625-28. (525) Mem. from J. Boyles to File, dated Mar. 13, 2004, FMSE 113686-90. 134
determine whether hedges are expected to be highly effective. As a result of these changes, the Company discontinued its hedging program with respect to anticipated debt issuances (except for the anticipated rollover of discount notes). These amendments are addressed in greater detail below in our discussion of Fannie Mae's hedge accounting policies. V. FINDINGS REGARDING FANNIE MAE'S ACCOUNTING FOR HEDGE TRANSACTIONS The SEC announced on December 15, 2004 that Fannie Mae's hedge accounting policies and procedures did not comply with GAAP. The SEC's press release stated as follows: Regarding Statement No. 133, one of the principles underlying the statement is that derivative instruments are to be recorded at their fair value with changes in fair value reported in earnings. If certain hedge criteria are met, however, Statement No. 133 affords special accounting for the hedge relationship. If the specific hedging requirements are not met, then special hedge accounting is not appropriate. Fannie Mae internally developed its own unique methodology to assess whether hedge accounting was appropriate. Fannie Mae's methodology, however, did not qualify for hedge accounting because of deficiencies in its application of Statement No. 133. Among other things, Fannie Mae's methodology of assessing, measuring, and documenting hedge ineffectiveness was inadequate and was not supported by the Statement.(526) The SEC's decision followed Fannie Mae's submission of a letter describing and defending its hedge accounting policies, and we assume that the SEC's conclusions were based on the presentation in that letter.(527) Certain aspects of Fannie Mae's accounting under FAS 133 discussed in the letter apply broadly to all hedge transactions, and certain aspects apply only to a more or less narrow subset of those transactions. The broadest issues addressed in the Company's letter, and which are - ---------- (526) Office of the Chief Accountant, U.S. Securities and Exchange Commission, press release, dated Dec. 15, 2004, http://www.sec.gov/news/press/2004-172.htm (last visited Feb. 17, 2006). (527) Letter from J. Boyles to Stephen Cutler and Paul Berger, dated Oct. 19, 2004, FMSEC 2215-64 (hereinafter "October 2004 SEC Letter"). Company representatives also met with the SEC, and KPMG provided a separate submission supporting the Company's accounting. 135
discussed in the SEC's press release, concern the type of hedge accounting the Company was using, the requirement that all hedge transactions be documented at inception, and that the documentation clearly identify certain aspects of the hedge relationship. Other aspects of FAS 133 addressed in the SEC letter were especially significant to Fannie Mae given its business practices, including the hedging of the rollover of discount notes, the hedging of forecasted debt issuance, and term-out transactions. Our review focused on these areas and on other issues that we believed would provide the most direct and extensive insight into the Company's accounting policies and procedures. We address below: (1) the hedge accounting methodology Fannie Mae adopted when it implemented FAS 133; (2) aspects of Fannie Mae's accounting that are inconsistent with this chosen methodology; (3) the Company's documentation; and (4) other aspects of the Company's accounting under FAS 133 that we believe reflect on the Company's approach to the standard. In our view, this evidence, viewed as a whole, demonstrates that Fannie Mae departed from clear FAS 133 requirements in order to avoid the income statement volatility and administrative complexity associated with strict adherence to the standard. A. Hedge Accounting Methodology It bears repeating that FAS 133 describes certain permitted methods of hedge accounting that differ in the extent to which they require the periodic calculation and recognition of ineffectiveness. Those methods are referred to in practice as the long-haul, matched-terms, and shortcut methods. In this section, we consider the methods that Fannie Mae applied. The question of whether Fannie Mae intended to follow a particular method of hedge accounting described in FAS 133 is complicated by what appears to be a change in the Company's nomenclature over the period in question and some conflict among those involved in the implementation of FAS 133 as to what the Company's intentions were in this regard. The Company's 2004 submission to the SEC refers to the Company's use of a concept that is "similar to the matched terms method in paragraph 65 of FAS 133 and DIG Issue G9."(528) Individuals involved in the development of the DAG and the implementation of the Company's hedge accounting stated that Fannie Mae's intention was to use the matched-terms method from the outset. Certain references in some documents to "critical terms" - - a phrase used in paragraph 65 of FAS 133 - lend some support to this view. The strong weight of the evidence, however, is that the Company intended to rely on a version of the shortcut method. That is, Fannie Mae would treat certain transactions as perfect hedges, retrospectively and prospectively, if certain "critical - ---------- (528) October 2004 SEC Letter at FMSEC 2234. Another portion of the letter includes the statement: "Our FAS 133 accounting policies relied heavily on ensuring that key terms (notional amounts, indices, repricing dates, maturity, and currencies) of the derivative and the hedge item matched." Id. at FMSEC 2226. 136
terms" of the derivative and the hedged item were aligned in accordance with parameters that the Company devised and prescribed in the DAG and incorporated into its FAS 133 Accounting System. Although some Company employees whom we interviewed referred to this approach as the "Fannie Mae shortcut," rather than as a method derived directly from the text of paragraph 68 of FAS 133, the overall approach was to enter into hedge transactions that, in the Company's view, would result in "no or inconsequential ineffectiveness," with no assessment or measurement of ineffectiveness required.(529) Beginning in 2001, discussions of the Company's hedge transactions referred to the Company's reliance on the shortcut method. The DAG itself refers to the application of the shortcut method for many transactions.(530) The DAG also states that "[t]he term sheet will be the source of documentation on the hedging relationship."(531) Term sheets that the Company issued when it entered into derivative transactions at that time stated that the transaction "will qualify for the shortcut method." Because these documents reflect managements's intentions both at the time it established its policies and when it entered into individual derivative transactions, the repeated reference to the shortcut method in these documents appears to be deliberate and is persuasive evidence of the management's intent.(532) - ---------- (529) Id. at FMSEC 2229. (530) The DAG states: Fannie Mae can use a shortcut method to apply FAS 133 provisions to an interest rate swap in a fair value hedge when it is assumed that there will be no hedge ineffectiveness. Assuming no ineffectiveness is especially important in a hedging relationship involving an interest-bearing hedged item and an interest rate swap because it significantly simplifies the computations necessary for making the hedge accounting journal entries required under FAS 133. DAG at FMSE-IR 13014-15. The DAG contains virtually identical language with respect to cash-flow hedges. See id. at FMSE-IR 12907-08. Other sections of the DAG explicitly refer to applying "the shortcut method." Id. at FMSE-IR 12924. (531) Id. at FMSE-IR 12965, 13023. (532) Sample form term sheets are included in the Appendix at Tab C. The Company would later take the position that the term sheets were not a part of the Company's documentation for its hedge transactions. 137
E-mails, memoranda, and correspondence drawn from Fannie Mae's files during the period 1998 through 2004 also refer repeatedly to the shortcut method.(533) These documents include correspondence - such as correspondence between Fannie Mae and the FASB - in which the receiving party would have understood the phrase "shortcut method" to refer to the method spelled out in paragraph 68 of FAS 133. Stone's notes and memorandum following the release of the exposure draft of FAS 138 mention the deletion of the reference to the shortcut method in the "Fannie Mae carve-out" and consider the implications of that amendment for Fannie Mae's hedge accounting. The authorization forms for transactions that required long-haul treatment required that the forms be completed when a transaction did not qualify for the "shortcut method." Other internal documents, including memoranda prepared by Financial Standards, describe Fannie Mae's hedge accounting by reference to paragraph 68 and the shortcut method.(534) An Internal Audit document prepared in October 2000, for example, states that "Fannie Mae has determined that funding swaps will meet the FAS 133 short cut method criteria to qualify for hedge accounting."(535) Also significant are documents that refer to possible changes to the shortcut method and discuss the implications for Fannie Mae. On September 24, 2002, for example, Boyles sent an e-mail reporting that the FASB was considering an amendment to FAS 133 that would make certain transactions "ineligible for the shortcut."(536) According to Boyles, "[t]his rule does not mean that they would not qualify for hedge accounting but only that they must go through the long-haul method that could create some ineffectiveness."(537) It is not credible that Fannie Mae would have been concerned about changes to FAS 133 that affected the shortcut method if the Company had not intended to apply that method as its documentation indicated. In December 2003, Financial Standards issued a directive that Company personnel should no longer refer to the method of hedge accounting described in the DAG as the shortcut method. The e-mail states: "Please note that we are no longer referring to use of the 'shortcut method' in our FAS 133 documentation (i.e. - ---------- (533) We understand that individuals at the Company sometimes referred to "a" shortcut method, suggesting that it was not referring to the method described in paragraph 68. Given the plethora of other references to "the" shortcut method, that interpretation is unconvincing. (534) See, e.g., Mem. from P. Salfi and I. Sussan to File, dated Oct. 15, 2003, FMSE-IR 650223-26, at FMSE-IR 650224 n.3. (535) Undated Fannie Mae Office of Auditing, FAS 133 - Test 3A - Funding Swaps, Review Performed During October 2000, Zantaz document 2913273. (536) E-mail from J. Boyles to P. Salfi, I. Sussan, and L. Spencer, et al., dated Sept. 24, 2002, FMSE-E 1049251. (537) Id. 138
termsheets/hsm [hedge strategy memoranda]/reports). Instead, consistent with the revised term sheets . . . we state that we 'assume perfect effectiveness' or 'assume no hedge ineffectiveness.'"(538) Boyles explained during an interview that he believed the use of the term "shortcut" was an incorrect characterization of Fannie Mae's methodology and that he thought it necessary to clarify the Company's nomenclature. It is not clear, however, whether the new nomenclature actually reflected a change in methodology. Indeed, an October 2003 memorandum prepared by members of Financial Standards concerning the accounting treatment of certain term-out transactions refers to the requirements of paragraph 68 - the paragraph that describes the conditions to applying the shortcut method - as the basis for the Company's treatment of the hedge as "perfectly effective."(539) At least until late 2003, therefore, the Company's internal documentation and justification for its hedge accounting practices rested on the shortcut method and paragraph 68. Two documents that refer to Fannie Mae's use of the shortcut method are both exemplary and significant because they bracket the period in which Fannie Mae was using hedge accounting and because of the audiences to which they were directed. First, a letter from Fannie Mae to the FASB in April 2000 states, "Fannie Mae, along with many other entities, intends to rely on the 'shortcut' method provided for in the Statement to ease the implementation effort of adopting FAS 133. The shortcut method not only makes implementation easier but also reduces income volatility associated with common hedge relationships."(540) Second, talking points prepared in connection with a July 2004 presentation to the Audit Committee of Fannie Mae's Board of Directors refer to the Company's use of the shortcut method, citing paragraph 68 of FAS 133 specifically.(541) These documents were directed to entities in a context in which one presumes that the authors would have been especially concerned about precision in language and accuracy in description. Accordingly, these references to the shortcut method cannot credibly be - ---------- (538) E-mail from I. Sussan to C. DeFlorimonte, M. Lewers, and M. Patel, dated Dec. 18, 2003, FMSE-E 473762. (539) Mem. from P. Salfi and I. Sussan to Distribution, dated July 31, 2003, FMSE 51390. The memorandum addresses a particular transaction in which the maturities of the relevant instruments did not match. Notwithstanding this "stub period," the authors concluded, "this transaction qualifies for the shortcut method of accounting after the stub period ends." Id. The memorandum refers to an interpretation of FAS 133, Implementation Issue No. E12, which discusses the treatment of stub periods under the shortcut method of paragraph 68. See HOW PARAGRAPH 68(C) APPLIES TO AN INTEREST RATE SWAP THAT TRADES AT AN INTERIM DATE, Statement 133 Implementation Issue E12 (Fin. Accounting Standards Bd. Dec. 6, 2000). (540) Unsigned letter from T. Howard to Director of Research and Technical Activities, Fin. Accounting Standards Bd., dated Apr. 3, 2000, FMSE 48027-30. (541) E-mail from D. Luigs to A. Kappler, copy to R. Bruemmer, dated July 15, 2004, FMSE-E 2239079-114. 139
explained (as some interviewees suggested) as a company-specific shorthand for another acceptable approach.(542) B. Significant Aspects of Fannie Mae's Hedge Accounting 1. Fair Value Equal to Zero FAS 133 provides that the shortcut (and matched-terms) method of hedge accounting can be used only if the fair value of the derivative at the inception of the hedge relationship is equal to zero. Thus, hedge accounting for a transaction in which the derivative has a fair value other than zero at inception requires the long-haul method. Fannie Mae has acknowledged it did not meet this requirement for one of the exceptions to long-haul accounting and that it took another approach.(543) One cause of the repeated violations of the "zero fair value" rule was the automatic relinkage by the Company's DEBTS system of derivatives to debt instruments other than the original hedged item.(544) As noted above, one of Fannie Mae's objectives when it implemented FAS 133 was to leverage its existing accounting systems as much as possible. The DEBTS system, which was the responsibility of the Treasurer's Office, determined which derivatives would be linked to which debt instruments. The process can be summarized as having in essence three steps: (1) When the Treasurer's Office entered into a hedge transaction, Treasurer's personnel would enter the terms of the derivative into the DEBTS system, which would electronically "link" the derivative to a particular debt instrument (in some circumstances, a single derivative, because of its notional amount, would be linked to several different debt instruments); (2) The system would then produce a term sheet describing the transaction, including the derivative and the debt being hedged; and - ---------- (542) The confusion over nomenclature may result from the fact that Fannie Mae adopted an approach to hedge accounting in which it would use a form of the shortcut methodology if the "critical terms" of the transaction matched. A workpaper from the files of KPMG states, for example, that for certain transactions, "If critical terms match, shortcut method can be used if it is an interest-rate swap." KPMG, Process Analysis Document, dated Aug. 2002 (hereinafter "KPMG Process Analysis Document"). (543) October 2004 SEC Letter at FMSEC 002236. (544) The DEBTS system was used for hedge transactions involving issued debt, including discount notes (which were tracked by the STN application), medium-term notes and long-term debt. It was not used for the less frequent, but very important, practice of hedging the anticipated issuance of debt. 140
(3) On a monthly basis, the FAS 133 Accounting System, which was the responsibility of the Controller's Office, would receive linkage information from the DEBTS system and confirm that the linkages met the criteria contained in the DAG for hedge accounting. Parts, but not all, of the Company's guidance in the DAG was incorporated into DEBTS and the FAS 133 Accounting System; however, only the FAS 133 Accounting System incorporated the DAG's numbering system for identifying a particular hedge transaction. Any linkages that did not meet the DAG requirements for the shortcut method would generate an error report in the FAS 133 Accounting System that could be analyzed and addressed by linking the derivative to a different debt instrument that would allow the relationship to qualify for the use of the shortcut method as defined in the DAG. One feature embedded into the DEBTS system was a protocol that would reassess the derivative-debt linkages periodically. Under the protocol, for example, a derivative linked to the rollover of discount notes might be relinked on its repricing date to another derivative in a term-out transaction or to a medium- or long-term debt instrument. Other than for new term-out transactions, it does not appear that the system created term sheets for new linkages, raising a question as to whether the documentation for these transactions existed at all or, if it did, whether it was appropriate.(545) Regardless of whether the new hedge relationships were appropriately documented, these new relationships resulted in new hedge relationships at times when it would not be expected that the derivatives would have a fair value equal to zero.(546) These transactions would therefore require hedge accounting under the long-haul method. The Company did not apply the long-haul method to these transactions and it did not reflect on its income statements any ineffectiveness associated with these transactions. Therefore, Fannie Mae did not account for these transactions properly. The evidence does not permit a definitive conclusion on whether the accountants in Financial Standards or elsewhere in the Controller's Office understood that the DEBTS system routinely relinked derivatives, creating a different hedge relationship than existed previously. Interviewees, including Boyles, claimed that they - ---------- (545) This question depends on whether the documentation for the Company's hedge transactions was contained in the term sheets that were prepared at the outset of a new hedge transaction or was contained in the DEBTS system itself. The answer to that question is not clear: interviewees suggested that DEBTS and the FAS 133 Accounting System, combined with the DAG, comprised the documentation; the DAG, however, refers to the term sheets as the relevant documentation. (546) This conclusion does not apply to circumstances in which the DEBTS system relinked a derivative that was entered into as a hedge of the rollover of discount notes to another discount note. Those "relinkages," as long as they continued unbroken, would be considered continuations of the original hedge relationship. 141
were not familiar with the operation of DEBTS and the documentary evidence on this issue is scanty.(547) The FAS 133 Accounting System, which did reside in the Controller's Office, merely determined whether linkages existing at the end of the month were of the type permitted by the DAG. The fact that these transactions might reflect relinkages, therefore, would not have been apparent through this system because it was not set up to check whether a linkage established in DEBTS resulted in a new hedge relationship. In our view, it does not appear that the lack of such a check was part of a conscious decision to bypass key provisions of FAS 133; instead, it was a failure of communication and follow-through between those responsible for compliance with GAAP and those with a detailed understanding of how the systems operated. Other relinkages, however, were intentional and not embedded within systems. Relinkages would arise, for example, when the Company called outstanding debt that had been part of a hedge relationship. Because the debt instrument no longer existed, the derivative would be designated in a new hedge relationship. Although the derivative would likely have a fair value other than zero, which would preclude the new hedge relationship from qualifying for the shortcut (or matched-terms) method, Fannie Mae continued to apply the shortcut method for the new hedge relationship. Evidence compiled during our review does address the problem of accounting for intentional relinkages of derivatives and hedged instruments. Notably, discussions between the Company and its outside auditor, KPMG, addressed the issue. According to internal e-mail exchanges, the issue raised was the amount of time that could elapse between the end of one hedge transaction and the relinkage of the derivative in another hedge and still have the derivative qualify for the use of the shortcut method. After some discussion in which the Company sought approval for a seven-day period between the two hedge transactions, Fannie Mae adopted a policy (1) that the Company would have three days to link the derivative to the new hedged item in DEBTs, but (2) that it would be required to document its intention to create such a link within one day of the termination of the previous hedge transaction.(548) The available documentation does not reflect any consideration of whether the relinkages should be accounted for under the long-haul method. - ---------- (547) One document that does appear to address this issue is an e-mail, dated December 12, 2000, sent to Stone, with a copy to Boyles and others. The e-mail states (among other things) that "a single Debt Swap may be linked, automatically, to different funding instruments for different periods during its life." Stone expressed concern "about documentation issues." E-mail from Prasad Chintamaneni to K. Stone, dated Dec. 12, 2000, FMSE-E 1691116-18. The response noted that "[t]here were questions regarding documenting these changed links that are yet to be addressed." E-mail from K. Stone to P. Chintamaneni, dated Jan. 2, 2001, FMSE-E 1691115. (548) See E-mail from M. Patel to P. Salfi, et al., dated July 15, 2003, FMSE-E 166732-33; E-mail from P. Salfi to Katarina Skladony, et al., dated Aug. 21, 2003, FMSE-E 166455-56. 142
The same issue arose in connection with the Company's term-out transactions.(549) To the extent that a term-out involves the relinkage of a derivative in a new hedge relationship, FAS 133, as amended by FAS 138, requires that the fair value of the derivative be assessed. Assuming the fair value is other than zero, the Company would be required to measure and recognize any ineffectiveness in the transaction under the long-haul method. FAS 133 thus precluded hedge accounting for these transactions under either the shortcut or matched-terms method.(550) We received various explanations for the Company's position on this issue. The most common explanation, offered by Boyles, was that the Company believed that the transaction would not result in ineffectiveness or that the ineffectiveness would be inconsequential, even if the fair value of the derivative were not equal to zero.(551) However, there is no evidence that the Company conducted any tests prior to the summer of 2004 (discussed below) to determine whether its assumption that the ineffectiveness would be inconsequential was accurate, and even the tests it did conduct during 2004 were not comprehensive. 2. Seven-Day Tolerance for Matching Repricing Dates Under the shortcut method of hedge accounting for cash-flow hedges, the "repricing dates [of the variable rate leg of the derivative] must match those of the variable rate asset or liability."(552) To the extent that the meaning of "match" was unclear, Implementation Issue E4, an interpretation of paragraph 68 issued on July 28, 1999, states that "[t]he verb match is used in the specified conditions in paragraph 68 to mean be exactly the same or correspond exactly."(553) More than a year before the effective date - ---------- (549) It does not appear that the Company raised this issue with KPMG specifically. (550) Further, in a majority of Fannie Mae's term-out transactions, the maturity dates of the offsetting derivatives were not the same; the example in FAS 133 that originally permitted the application of the shortcut method involved derivatives with matching maturity dates. Paragraph 68, in fact, mentions differences in maturity dates as a source of ineffectiveness in a hedge transaction. (551) Boyles also explained that he believed the fair-value provision would only apply to those companies that structured their derivative transactions to contain a financing element (i.e., a fair value of other than zero at inception). (552) FAS 133 P 68(k). (553) APPLICATION OF THE SHORTCUT METHOD, Statement 133 Implementation Issue No. E4 (Fin. Accounting Standards Bd. 1999) (revised Mar. 26, 2003). 143
of FAS 133, therefore, the fact that the shortcut method permitted no flexibility on the maturing of repricing dates was amply confirmed.(554) Nonetheless, Fannie Mae adopted a policy that permitted the repricing dates of variable-rate instruments (or the rollover dates of discount notes) to differ from the repricing dates on the interest-rate swap by up to seven days. That is, the DAG required that "the repricing dates match, within 7 calendar days (plus or minus), the rollover dates of the Discount Notes or the repricing dates of the variable-rate asset or liability."(555) This reflects a deviation from the requirement in FAS 133 that long-haul accounting be applied to any transaction in which the repricing dates did not coincide.(556) The Company's explanation for this policy rested on practicality and materiality. As to practicality, for reasons of market efficiency, Fannie Mae issues benchmark bills on Wednesdays. Fannie Mae did not wish to change this practice and concluded it would not be feasible, in the ordinary course, to obtain derivatives to hedge the debt instruments issued on a given day so that all of the derivatives reprice on the same date. As to materiality, although the Company took the position that any ineffectiveness associated with this practice was inconsequential, we have seen no evidence that the Company tested that proposition prior to the implementation of the rule.(557) The analysis that the Company did conduct before 2001 did not test the - ---------- (554) Similarly, under the matched-terms method of hedge accounting, the Company could conclude that "changes in fair value or cash flows attributable to the risk being hedged are expected to completely offset at inception and on an ongoing basis" if "the critical terms of the hedging instrument and of the entire hedged asset or liability (as opposed to selected cash flows) or hedged forecasted transaction are the same." FAS 133 P 65. (555) DAG at FMSE-IR 12909. (556) Initially, the permitted deviation was plus-or-minus five business days. This formulation, however, proved difficult for the systems involved in the hedge transactions to accommodate because the effect of holidays could not be programmed into the system. Holidays in the United States and other markets, for example, could affect whether the repricing dates were within five business days of each other. The Company therefore amended the policy to permit a deviation in repricing dates of plus-or-minus seven calendar days, making it easier for Fannie Mae to apply its guidelines. (557) One interviewee noted that a seven-calendar-day rule covered the entire week between Wednesday debt offerings, and therefore permitted the Company to link all derivatives to some debt no matter when the derivatives were acquired. This may be overstated, but the rule permitted the Company a certain flexibility to enter into derivative transactions around the time of the debt issuances. 144
ineffectiveness associated with the seven-day repricing date mismatch in a way that would have mirrored the results of applying the long-haul method. According to documents describing the analysis: To estimate the income statement impact of applying this policy (e.g., the ineffectiveness we do not book because the repricing dates "match"), we performed two sets of quantitative analysis looking at the daily movements in 3M LIBOR since 1993 and applied these average differences to our average pay-fixed swap book for the last three years. The first set of quantitative analysis ("average approach") incorporates the idea that there is offset between movements in LIBOR associated with pay-fixed swap funding needs which occur before a Benchmark Bill issuance and movements in LIBOR associated with pay-fixed swap funding needs which occur after a Benchmark Bill issuance. The second set of quantitative analysis ("absolute value approach") does not incorporate this idea of offset and assumes that changes in LIBOR always move against us, which we believe represents a "worst case" scenario and is highly unlikely.(558) The Company's analysis showed that, on average, the mismatch was less than seven days and was clustered fairly evenly around the date of the debt issuance. It quantified the amount of ineffectiveness at between $0.4 million and $5.4 million for each $1 billion of notional amount. However, this range may be understated because the analysis did not compute the ineffectiveness on each hedge relationship outstanding during the period and may have resulted in netting under-hedges (that would not result in the recognition of ineffectiveness) with over-hedges (that would result in the recognition of ineffectiveness). Finally, this issue once again draws the Company's approach into question: by offering that the ineffectiveness associated with this aspect of the Company's hedge accounting was immaterial, the Company presumes to have considered this issue from the perspective of a long-haul methodology, but that perspective was not reflected in the Company's hedge accounting policies or practices at any time. 3. Forecasted Transactions As explained above, special rules apply to the hedging of forecasted or anticipated debt issuances. These transactions include the Company's anticipated rollover of discount notes and the issuance of longer-term debt instruments. Forecasted transactions do not qualify for shortcut treatment. - ---------- (558) Mem. from I. Sussan to File, dated Dec. 17, 2003, FMSE 32657-58, at FMSE 32657. 145
(a) Duration Matching A hedge of a forecasted transaction, like other hedges, has two parts. The first part is the hedged item, or the debt instrument to be issued at some anticipated future date on certain terms. The second part is the derivative designed to offset the impact of changes in market conditions on the proceeds received on the date the debt is issued (or the interest rate on that debt). If, for example, the Company were to forecast that it will issue medium-term debt in fourteen days, the Company might enter into a short sale of another instrument (typically U.S. Treasury instruments) to hedge changes in the cash flows on the anticipated debt issuance attributable to changes in the risk-free interest rate over the fourteen-day period. Any change in interest rates would affect the forecasted debt issuance and the short sale in opposite directions, although not necessarily in the same magnitude. If the critical terms of the hedging instrument and the forecasted transaction are the same, FAS 133 (specifically, the matched-terms method) permits a company to assume there will be no ineffectiveness.(559) If the critical terms are not the same, long-haul accounting would be required. The DAG permitted a shortcut approach to forecasted debt transactions when certain critical terms of the derivative and the debt instrument "matched."(560) The criteria established for applying this shortcut method were: (1) that the risk being hedged be the risk-free (U.S. Treasury) rate, the LIBOR swap rate, or the agency rate; and (2) that the duration of the derivative fall within a specified range of the hedged debt.(561) The policy did not require that the durations match; instead, the hedge qualified for this duration shortcut approach if the durations fell within the ranges shown in the following table: - ---------- (559) Forecasted transactions may be accounted for using the matched-terms method described in paragraph 65, if "[t]he critical terms of the hedging instrument and of the... hedged forecasted transaction are the same." Paragraph 66 of FAS 133 makes clear, however, that "hedge ineffectiveness would result from the following circumstances, among others: ... b. Differences in critical terms of the hedging instrument and hedged item or hedged transaction, such as differences in... maturities..." (560) See DAG at FMSE-IR 12924. Fannie Mae referred to this approach as the "duration shortcut" of hedge accounting for anticipated debt issuances. (561) Id. at FMSE-IR 12925. For these purposes, "duration" refers to the sensitivity of an instrument's price to changes in interest rates. Although duration is measured in months, it is not the same as the maturity of the instrument. 146
Hedged Debt Maturity Duration Range - -------------------- -------------- 0-36 months +/- 6 months 37-60 months +/- 9 months 61-120 months +/- 12 months 121+ months +/- 18 months Although FAS 133 states specifically that ineffectiveness would result from mismatched maturities, Fannie Mae's policy did not require that the maturities of the instruments linked in a hedge transaction match. Therefore, to the extent the maturity dates of the hedged anticipated transaction and the hedging instrument did not match, the Company's forecasted transactions did not qualify for hedge accounting under the matched-terms method. Consequently, these transactions should have been accounted for under the long-haul method.(562) In a March 2004 memorandum discontinuing the duration-matching rule, Boyles acknowledged that this policy "has been a known departure from strict compliance with GAAP," but that the Company permitted the approach "because, from an economic standpoint, the correlation from matching of durations between the anticipated debt and the actual debt to be issued is better than just matching notional and payment/reset dates."(563) Interviewees explained that the Company adopted the duration-matching rule for two reasons. First, the Treasurer's Office personnel responsible for hedge transactions focused on duration, not on maturity, as an important feature of the debt that office issued. Second, the Company took the view that matching durations resulted in a more effective hedge than matching maturities and notional amounts.(564) - ---------- (562) The training materials published by the FASB after the issuance of FAS 133 illustrate the use of a duration approach to hedging. (563) Mem. from J. Boyles to File, dated Mar. 13, 2004, FMSE 113686-90, at FMSE 113686. (564) In addition to the duration rule that was applied to the hedge of a forecasted transaction, the Company had a duration rule that provided flexibility in the terms of the debt to be issued. A memo from Stone, dated February 18, 2000, states, "[t]o add managed flexibility to the type of debt that is issued versus the type of debt that is hedged, Fannie Mae developed the duration rule." Memorandum from K. Rawls to Distribution, dated Feb. 18, 2000, Zantaz document 1925375, at 1. The memo goes on to note, "as market conditions change, the Treasurer's Office can sometimes obtain better execution (lower costs) by substituting (cross-delivering) a type of debt other than what we [sic] being hedged. The current duration rule, however, does not allow adequate flexibility in cross-delivering debt issuances on the long end of the yield curve." The memo then concludes: 147
Contemporaneous documents reflect a somewhat different motive for adopting a policy of permitting the application of the shortcut method when durations were within the defined ranges. The duration range appeared in a November 22, 2000 memorandum from the Hedging Desk to Nathani, Boyles, and Stone.(565) The memorandum states: The proposed duration windows take into account the probability that Fannie Mae will need to hedge both new issue and reopened debt with current or reopened securities which will result at times in larger than normal mismatch of duration. If market conditions change, there will be a need to re-asses [sic] these duration windows. Finally, given the fact that hedging Fannie Mae debt with Fannie Mae benchmark securities results in as perfect a hedge as we can achieve, we suggest using the short cut treatment for such agency hedges that fall within the proposed duration windows.(566) The following month, Stone announced that "KPMG has agreed to our proposal to use a 'shortcut' method on anticipatory debt issuances" within the specified duration parameters.(567) To support the duration-matching rule, the Company prepared an analysis that purported to compare the calculated ineffectiveness in a hedge transaction involving closely-matching durations (but different notional and principal amounts) to the amount - ---------- If the original duration of the anticipated debt is 84 months or less, the Hedge Desk can cross-deliver that does not differ by more than 18 months from the original duration of the anticipated debt. Alternatively, if the original duration of the anticipated debt is greater than 84 months, the Hedge Desk can cross-deliver debt that does not differ by more than 30 months from the original duration of the anticipated debt. This approach calls into question the Company's hedge accounting for forecasted debt issuances on other grounds - specifically, that the documentation did not specify the hedged item and the forecasted transaction described in the documentation may not have been probable. (565) Mem. from U. O'Donnell to Distribution, dated Nov. 22, 2000, FMSE 365840-43. (566) Id. at FMSE 365840. (567) E-mail from K. Stone to Mary Trzeciak, dated Dec. 20, 2000, FMSE-E 845330. 148
of ineffectiveness in a hedge transaction involving identical dollar amounts. The results of this analysis were that the use of the duration method might result in less ineffectiveness than the ineffectiveness that arises when the amounts match.(568) This analysis is not persuasive. According to Ursula Schaeffer, who drafted the analysis, the Company did not test the duration-matching rule in differing market conditions or at various points within the duration window. The Company's analysis was, therefore, a series of examples and not a proof or demonstration that duration matching is necessarily a "better" approach other than for reasons of operational convenience. At KPMG's insistence, Fannie Mae monitored the impact of the duration-matching policy. At the end of 2001 and 2002, Financial Standards selected a sample of the relevant transactions and Treasurer's Office personnel measured the ineffectiveness associated with those transactions. Based on this analysis, the Company concluded in each year that the total ineffectiveness associated with the duration matching rule was immaterial (less than $3 million). At the end of 2003, Financial Standards conducted a more comprehensive assessment and determined that the ineffectiveness associated with these transactions was approximately $12 million ($21 million pre-tax). At that time, the Company recorded the $12 million in earnings and discontinued hedge accounting for anticipated debt issuances. 4. The De Minimis Test To qualify for hedge accounting, it must be established that the hedge of a forecasted debt issuance is highly effective. Accounting literature prior to FAS 133 described a "dollar offset" test in which a company would determine the fair values of the two instruments; if the change in the fair value of one instrument offset the change in the fair value of the other instrument within a band of 80-125 percent, the derivative was deemed sufficiently effective. Although FAS 133 expressly permits this approach, it allows an entity to adopt another suitable "statistical analysis." FAS 133 requires that the methodology used for assessing effectiveness be consistent. Paragraph 62 of FAS 133 requires that the Company designate at the outset of the transaction "the method it will use to assess the hedge's effectiveness . . . ." Moreover, "[o]rdinarily, . . . an entity should assess effectiveness for similar hedges in a similar manner; use of different methods for similar hedges should be justified." Fannie Mae adopted an approach to the assessment of hedge effectiveness for forecasted transactions that was not consistent with this requirement. The Company adopted the dollar-offset methodology, but supplemented it with an alternative methodology. Not only did this approach violate the consistency requirement, but the alternative test was not an acceptable statistical approach. - ---------- (568) Mem. from U. O'Donnell to Distribution, dated Nov. 22, 2000, FMSE 365840-43; Undated analysis, FMSE 365877-79. 149
This alternative method for testing effectiveness - known as the "de minimis" method - would permit hedge accounting if the dollar-offset criterion was not met as long as the interest rate on the issued debt, adjusted for the results of the hedge, was within fifteen basis points ("bps") of the target rate established on the date of the original forecast. Given the magnitude of Fannie Mae's debt issuances, and because the amount of ineffectiveness required to cause the actual rate to differ by more than fifteen bps from the target rate would be significant, the de minimis test allowed the Company to treat derivatives as hedges even when the extent of the dollar-offset was significantly outside the 80-125 percent range. The de minimis test was employed at Fannie Mae prior to FAS 133, and appears to have been incorporated into the DAG without a systematic analysis regarding the consistency of the test with FAS 133's requirements. The DAG introduces the de minimis test as follows: Dollar correlation alone has proven to be an unsatisfactory measure of hedge performance when there is very little interest rate movement over a hedge period. The Deminimus [sic] rule was developed as an alternative means of evaluating the effectiveness of a hedge. To overcome such inequities in the correlation measurement, we have established an alternative for measuring the effectiveness of a hedge.(569) The rule, therefore, appears to be a Company-originated alternative to the dollar-offset approach recognized in FAS 133 and prior accounting literature. Interviewees, including Boyles, acknowledged that the de minimis test was not a statistical method. Moreover, when asked how this approach was appropriate given FAS 133's reference to "the method" for assessing effectiveness, and the requirement that the methodology be consistent for similar transactions, Boyles responded that the Company's approach consisted of a single method with two parts. That formulation, however, is not consistent with other documents on this subject.(570) In addition, this test has not been applied consistently. Prior to the adoption of FAS 133, the Company's de minimis test permitted a window between the actual and target interest rates of ten bps. At the time the Company was developing its approach to FAS 133, individuals in the Treasurer's Office were concerned that increased - ---------- (569) DAG at FMSE-IR 12930. (570) An undated presentation entitled "FASB Statement No. 133: Accounting for Derivative Investments and Hedging Activities - Fannie Mae: Policies and Procedures for FAS 133," FMSE-IR 534549-534697, states: "The Hedge Desk employs two acceptable method of measuring hedge effectiveness: 1. Correlation[;] 2. Deminimus [sic]." Id. at FMSE-IR 534559. 150
market volatility would push hedge transactions outside the permitted ten bps margin. The Company conducted an analysis of the historic volatility in its debt issuances, and concluded (among other things) that Fannie Mae's five- and ten-year debt, which traditionally had experienced spreads of approximately thirty bps, had more recently experienced spreads of approximately sixty-five bps: "The percentage volatility in spreads between time periods virtually doubled going from 26% to 50% in fives and from 15% to 32% in tens."(571) Based on this analysis, the Company considered (and, it appears, encouraged its outside auditor to approve) an increase in the range covered by the de minimis rule to twenty bps. Ultimately, the Company established a new de minimis test with a margin of fifteen bps. The de minimis rule was incorporated into the DAG with this change. Internal correspondence confirms that the rule was expanded to enhance the Treasurer's Office's flexibility to enter into transactions that might not have qualified for hedge accounting under the old, ten bps rule. In an e-mail dated November 24, 1999, from Nathani to Howard, Spencer, and Boyles, et al., Nathani reported "that the de minimis parameters have been extended to fifteen bps from the current ten bps rule."(572) He recounted that, at an earlier planning meeting, "we discussed the importance of getting more flexibility around the de minimis rule. . . . We presented statistics around increased spread volatility and requested twenty bps. Jonathan [Boyles] has agreed to fifteen bps." Nathani concluded by noting that the new rule "allows us to weather more volatile times without having to mark to market through earnings," but he indicated that Fannie Mae would "be careful not to use the fifteen bps as a license for greater hedge inefficiencies - we will continue to attempt to get as high dollar correlation as we can." Howard's response to Nathani is telling as to his view of the hedge accounting rules, "That's great news and excellent work. Congratulations to all involved. I agree with your conclusion not to take the expanded de minimis latitude as a basis for increasing our hedge risk - we should instead view it as an added margin of safety to ensure the applicability of hedge accounting in volatile markets."(573) The accounting policy memorandum that announced the change in the de minimis rule formally linked the change to the Company's convenience and business interests: "Market volatility has increased dramatically since the implementation of the original de minimus measure . . . Meanwhile, there has been no adjustment to the gauges - ---------- (571) De Minimus Research Highlights, dated Oct. 26, 1999, FMSE 514327. (572) E-mail from M. Nathani to T. Howard, L. Spencer, and J. Boyles, et al., dated Nov. 24, 1999, FMSE 514324. (573) Id. A separate e-mail conveys the substance of an exchange with a KPMG partner, Ken Russell, in which Russell expresses the concern that the need to extend the de minimis parameters might reflect the fact that the hedges are simply not effective in the more volatile environment. E-mail from K. Rawls to M. Trzeciak, final e-mail in chain dated Jan. 6, 2000, FMSE 365019. As noted, a subsequent e-mail reports on KPMG's "approval" of the change. 151
used in measuring the effectiveness of hedges. This raises the likelihood that a hedge may lose hedge accounting due to an extreme, although temporary, move in spreads."(574) The policy concludes, "the current ten basis point de minimus requirement is no longer an appropriate measure of a hedge's effectiveness given the historical change in volatilities based on a review of spread volatility . . .. we determined that a 15 basis point requirement is a more appropriate measure of a hedge's effectiveness." In connection with its efforts to increase the de minimis threshold, the Company prepared an analysis showing the percentage of the forecasted debt issuance transactions in which the Company relied on the de minimis approach to qualify the transaction for hedge accounting. In 1998, nearly half of the Company's transactions (forty-eight percent) would not have qualified for hedge accounting under the dollar-offset method, but were accounted for as hedges under the de minimis rule.(575) In 1999 that proportion increased to sixty-six percent. Assuming that these data reflect the tendency in Fannie Mae's hedge transactions after FAS 133's effective date, then at least half of the Company's hedge transactions involving forecasted debt issuances would not have qualified for hedge accounting. The Company discontinued its use of the de minimis test as of January 1, 2004, following an SEC staff speech emphasizing the importance of using only one method to test whether a transaction would be highly effective.(576) The Company did not consider whether its approach had been proper prior to that date. The speech does not alter the conclusion, however, that Fannie Mae's use of the de minimis rule was inappropriate and designed to inject greater flexibility into the Company's approach to hedge accounting so as to avoid the consequences of more volatile markets. 5. Identification and Probability of Forecasted Transactions In discussing hedges of forecasted transactions, FAS 133 states that the "hedged forecasted transaction shall be described with sufficient specificity so that when a transaction occurs, it is clear whether that transaction is or is not the hedged transaction."(577) Further, FAS 133 makes clear that the forecasted transaction must be probable of occurring.(578) The issues of identification and probability raise concerns with - ---------- (574) Undated Mem. from K. Stone to Distribution, FMSE-E 1044982-83. (575) E-mail from M. Trzeciak to K. Rawls, dated Jan. 6, 2000, FMSE 365019. (576) The speech did not mention the de minimis test per se. Accordingly, we have no way of knowing whether the hedge transactions at issue would in fact have qualified for hedge accounting under one acceptable test, which Fannie Mae's did not. (577) FAS 133 P 28(a)(2). (578) FAS 133 P 29(b). 152
respect to Fannie Mae's hedges of its rollover of discount notes and its issuance of benchmark notes. (a) Forecasted Issuance of Discount Notes The Company's documentation of its hedges of the forecasted issuance of discount notes states "Fannie Mae expects to issue Discount Notes for a period of (insert term of swap). As a result, Fannie Mae is exposed to variability in cash flows related to the forecasted issuance of the Discount Notes."(579) Similarly, the hedge relationship diagrammed in the DAG notes that the hedged item is "Discount Note Rollover 90-day."(580) During our investigation, interviewees acknowledged that the Company had a strategy of replacing its discount notes with longer-term debt (e.g., in term-out transactions). As stated in the Company's October 19, 2004, letter to the SEC staff, "Fannie Mae often will choose to 'term-out' the Discount Note issuances described above by entering into a fixed-rate, medium term note to raise the necessary funding. . . . "(581) Because the Company had a pattern of redesignating its interest-rate swaps as a hedge of longer-term debt, it does not appear that it had a basis for asserting that its forecasted transactions (i.e., the rollover of discount notes) were probable, calling into question the Company's basis for applying hedge accounting in these instances.(582) Moreover, the protocol in the DEBTS system gave priority to term-out transactions and unswapped variable-rate long-term debt before considering a linkage with discount notes. It is likely, therefore, that many interest-rate swaps originally designated as hedging the rollover of discount notes were redesignated as a hedge of another instrument. (b) Forecasted Issuance of Benchmark Notes Similar issues arise in connection with the Company's treatment of the anticipated issuances of benchmark notes and long-term debt. FAS 133 requires that the hedged forecasted transaction be described with sufficient specificity so it is clear when the hedged transaction occurs.(583) If the terms of the debt issued differ significantly from - ---------- (579) DAG at FMSE-IR 13120. (580) DAG at FMSE-IR 13168. (581) October 2004 SEC Letter at FMSEC 2222. (582) As the FASB noted in FAS 133, "A pattern of determining that hedged forecasted transactions probably will not occur would call into question both an entity's ability to accurately predict forecasted transactions and the propriety of using hedge accounting in the future for similar forecasted transactions." (583) FAS 133 P 28(a)(2). 153
the terms of the forecasted transaction described in the Company's documentation, the application of hedge accounting would not be appropriate. It appears that the Company at least considered adopting documentation that would disguise its intent with respect to forecasted transaction hedges so as to avoid the need to measure and record ineffectiveness. In a memorandum dated April 25, 2000, Stone discussed Fannie Mae's preference for funding its mortgage purchases with ten-year debt.(584) According to Stone, "Portfolio and Treasurer's is [sic] exploring alternatives to provide ten-year funding when needed as well as reducing the Hedge Desk's exposure to earnings volatility from shifts in the market when anticipated debt issuances are hedged. One alternative being explored is the purchase of a ten-year funding swap that is funded with Discount Notes as product is delivered and unwind the swap when the ten-year Benchmark is issued." In other words, rather than hedge the anticipated debt issuance, the Company would acquire a ten-year swap, link that swap to the rollover of discount notes pending issuance of a ten-year note, and then (following issuance of the note) re-link the derivative in a new hedge relationship. A handwritten note at the top of the memo states "Documentation not to mention (expect to end relationship in 2 months) can end relationship in 2 months & fund w/ ten year Benchmark."(585) Stone explained that the contemplated transaction had the benefit of allowing Fannie Mae to "lock in a ten-year rate as mortgage products are purchased and it does not have to hedge the issuance of an anticipated Benchmark note. This reduces the Hedge Desk's exposure to earnings volatility arising from hedging instruments not correlating with the anticipated debt issuance when markets are volatile."(586) It appears the handwritten note addresses a concern that hedge accounting for the ten-year interest-rate swap would not be appropriate because it was not probable that the initial hedged anticipated transaction (i.e., the rollover of discount notes for a ten-year period) would occur. In each of these instances, the Company adopted an approach to FAS 133's requirements for forecasted debt issuances that departed from principles set forth in the standard. The evidence is persuasive that the Company took this approach because it offered greater flexibility, while the approach that the standard required in these circumstances - long-haul accounting - was incompatible with the Company's business objectives. - ---------- (584) Mem. from K. Rawls to Distribution, dated Apr. 25, 2000, FMSE 365266-68, at FMSE 365266. (585) Id. (586) Id. 154
6. Term-Outs As noted above, the Company's term-outs did not qualify for treatment under the shortcut method or the matched-terms method because there was no demonstration or likelihood that the fair value of the pre-existing derivative was zero. Moreover, the shortcut method was foreclosed by the amendments in FAS 138. The Company has taken the view, however, that its treatment of term-outs was permissible. In support of this proposition, the Company has pointed to its exchanges with the FASB regarding the term-out transactions, including the correspondence that led to the "Fannie Mae carve-out."(587) Although we acknowledge that FAS 133 could be clearer in certain respects, we do not believe that Fannie Mae's position was well-grounded. As a factual matter, term-out transactions begin with the issuance and rollover of discount notes hedged by a receive-fixed swap. As an anticipatory hedge transaction, the rollover of discount notes does not qualify for the shortcut method. FAS 133 does raise the possibility that the rollover of discount notes might qualify for treatment under the matched-terms method, if the critical terms of the discount notes and the derivative are the same. Fannie Mae's correspondence with the FASB on this subject proposes such a transaction and specifies in particular that the reset dates of the derivative and the rollover of the discount notes were the same.(588) The adoption of the "+/- seven-day rule," however, was inconsistent with this statement and, so far as we are aware, was not otherwise disclosed to the FASB. Any prediction as to how the FASB would have responded had it been aware of Fannie Mae's intention to allow for differences in repricing dates would be speculative. We recognize that Fannie Mae wanted to present a specific question to the FASB regarding the rollover of discount notes, and therefore presented a simple scenario; it would be extending this exchange too far, however, to conclude that the FASB approved Fannie Mae's accounting for these transactions. Fannie Mae's correspondence with the FASB regarding the term-out portion of the discount notes is likewise not comprehensive and therefore does not support Fannie Mae's position. The correspondence posited a situation in which the debt instrument, the original derivative, and the new derivative all end on the same date. That is, measured from the date of the term-out, all of the instruments involved in the transaction have the same maturity date. In fact, in many of Fannie Mae's term-out transactions, both by design and in practice, the maturities of the instruments differed substantially. Differences in the maturities could mean that changes in market conditions would affect the fair values of the instruments differently - that is, a change in interest rates might affect the fair value of a four-year instrument differently than it affects the fair value of a two- or three-year instrument. Paragraphs 66 and 68 of FAS 133, moreover, make plain that differences in maturity yield ineffectiveness. - ---------- (587) Letter from S. Rajappa to Director of Research and Technical Activities, Fin. Accounting Standards Bd., dated Oct. 6, 1997, FMSE-IR 217015-26. (588) Letter from J. Boyles to R. Wilkins, dated Jan. 20, 1999, FMSE-IR 192985-86.. 155
Insofar as the specific issues related to term-outs is concerned, Boyles pointed to the language in Example 8 of FAS 133 that remained after the FASB (in FAS 138) had deleted the reference to shortcut accounting: "Together, the cash flows from the two derivatives are effective at offsetting changes in the interest payments on the three-year note. Changes in fair values of the two swaps are recognized in other comprehensive income and are reclassified to earnings when the hedged forecasted transaction . . . affect[s] earnings[.]"(589) Boyles's view was that the FASB recognized that the amount of ineffectiveness would not be significant in the fact pattern used in the example and therefore did not indicate that ineffectiveness should be recognized in earnings. The acknowledgment that the derivatives are "effective at offsetting changes in the interest payments," however, merely restates a requirement for special accounting of all hedge transactions. In our view, this language does not mean that the shortcut (or matched-terms) methodology is appropriate in accounting for Fannie Mae's term-out transactions. Finally, there is no contemporaneous documentation to suggest that the Company considered this issue prior to our review; on the contrary, contemporaneous documents (notably Stone's notes and memorandum discussed above) reflect a recognition that the FASB had removed the authorization to treat term-out transactions under the shortcut method but that the Company did not wish to reconsider its accounting. C. Hedge Documentation For the reasons described above, a significant number of Fannie Mae's hedge transactions did not qualify for the hedge accounting under the shortcut (or matched-terms) method. It bears repeating that, for many of the Company's hedge transactions, the documentation stated that "the critical terms of [the derivative] and [the debt instrument] are the same (as defined in Fannie Mae's Derivative Accounting Guidelines)," and the Company therefore concluded "both at inception and on an ongoing basis . . . that the hedging relationship is expected to be highly effective in achieving offsetting cash flows attributable to changes in the benchmark interest rate . . . and will qualify for the shortcut method."(590) The Company's hedge documentation is insufficient to meet the requirements of FAS 133 and, in some instances, is incorrect. The form of hedge documentation that the Company used beginning in January 2001 posits that "[t]he critical terms of the swap and the debt obligation are identical. . . . Therefore, both at inception and on an ongoing basis, Fannie Mae concludes that the hedging relationship is expected to be highly effective in achieving offsetting changes in fair value attributable to the hedged risk and will qualify for the shortcut method."(591) The hedge transactions - ---------- (589) FAS 133 P 161. (590) DAG at FMSE-IR 13120. This language is taken from the template for documentation of a hedge transaction involving the rollover of discount notes. (591) Mem. Cash Flow Hedge 2001-XX, dated Jan. 1, 2001, PW-PR 816172-78. 156
themselves, however, did not satisfy the conditions described in the documentation or as prescribed in FAS 133: The terms of the derivative and hedged instrument were not "identical." Accordingly, the documentation did not accurately describe the transaction it purported to address or offer a proper basis for hedge accounting. Moreover, there is no evidence that the Company undertook to meet the other requirements of FAS 133, such as the assessment of ineffectiveness on a periodic basis and the designation of a methodology for measuring and recording any ineffectiveness that might arise in its hedge transactions. Transactions that qualified for neither the shortcut nor the matched-terms method require a company to assess, on at least a quarterly basis, the effectiveness of the hedge (both retrospectively and prospectively) and to measure and recognize any ineffectiveness in earnings. Fannie Mae did not have systems in place to assess whether a hedge met the effectiveness test on an ongoing basis. Although the systems "matched" the terms of the hedged item with the terms of the derivative within the band of tolerance specified in the DAG - which appears to have been the mechanism on which the Company relied - that would not be sufficient because the Company did not measure the effects of certain factors, such as the ineffectiveness caused by a derivative with a fair value other than zero at the inception of the hedge relationship. Further, with the exception of a very few transactions,(592) the Company would not accept the earnings volatility associated with ineffectiveness, nor did - ---------- (592) Periodically, the Company would come across a transaction that would not qualify for hedge accounting because of an error in the execution of the transaction. In some cases, the Company decided to treat these transactions as "DNQ" - meaning that they do not qualify for hedge accounting at all. A meeting report dated March 25, 1999 from Nathani to Spencer, Knight, Boyles et al., states that one of the objectives for a subsequent meeting is "a final list of DNQ transactions/impact and recommended solutions." FMSE-IR 193008. A 2004 report on hedge accounting activity relates that "[t]he Treasury Department attempts to manage the DNQ position such that there is no P&L exposure resulting from the position in changing market conditions." E-mail from P. Barbera to L. Spencer, J. Pennewell, and M. Lewers, dated Apr. 29, 2004, FMSE-E 37590-93, at FMSE-E 37591. In other cases, the Company undertook efforts to avoid either the DNQ approach or long-haul accounting. For example, in one case, the Company booked a $5.6 million loss on a transaction because (under the Company's criteria) the transaction did not qualify for hedge accounting based on the fair value resulting from the use of the yield curve at the end of the day. The Company later decided that the transaction would qualify for hedge accounting if it determined the derivatives' fair value using the intra-day yield curve at the time of the transaction itself. The Company reversed the $5.6 million loss the following month. At the same time, the Company decided as a matter of policy that it would apply the yield curve at the end of the day. When the director of Financial Reporting in the Controller's Office responsible for booking these entries noted the inconsistency of this treatment, her concerns were dismissed. 157
it design or install systems to implement that portion of the matched-terms methodology that admits ineffectiveness into the accounting.(593) The fact that the FAS 133 Accounting System tested the linkages between the derivatives and the hedged items on a monthly basis is not the same as measuring the amount of hedge ineffectiveness. The FAS 133 Accounting System was designed only to corroborate on a monthly basis that the linkages between the debt instrument and the derivative were still valid in accordance with the DAG, not to measure ineffectiveness.(594) The Company's hedge documentation rests almost entirely on an assumption that any ineffectiveness associated with its hedge transactions would be inconsequential. Not only does this approach presuppose that the Company was prepared to identify ineffectiveness, but the documentation failed to address how any ineffectiveness in the hedge relationship would be measured if the ineffectiveness were no longer inconsequential. Although the DAG indicates that Fannie Mae would rely on the "Hypothetical Derivative" method to measure and record ineffectiveness when the critical terms of the derivative and the hedged item no longer are the same, its documentation (whether it be the term sheets or the diagrams in the Appendix to the DAG) states that the hedge will be perfectly effective. However, because the critical terms of the derivative and the hedged item were not the same, management did not have adequate documentation of the hedge relationship and therefore was prohibited from applying hedge accounting for those transactions.(595) Moreover, management did not establish a foundation for the assumption that any ineffectiveness in its portfolio of hedge transactions was inconsequential. On the contrary, Fannie Mae's assessments of ineffectiveness in a population of hedge transactions points to the opposite conclusion. For example, according to a February 2001 analysis prepared by Sinclair and sent to Stone, "[t]he average ineffectiveness for - ---------- (593) Even if we were to draw the conclusion that Fannie Mae's intention was to implement the matched-terms method, its hedge accounting still would not have complied with FAS 133 for a number of reasons, including the fact that the critical terms were not the same and the Company's failure to provide the proper documentation regarding the measurement of ineffectiveness on an ongoing basis. (594) Because the original linkages, as well as relinkages, were created in the DEBTS system, corroboration was a necessary control to confirm that these linkages were consistent with the criteria for the "shortcut method" contained in the DAG. (595) It should be noted that the DAG indicates the Company would use the hypothetical derivative method for measuring and recognizing ineffectiveness when the terms no longer match. However, its documentation and policies state that, in most cases, the transaction was either perfectly effective (using the DAG as the Company's documentation) or qualified for the shortcut method (using the termsheets as the Company's documentation). 158
most scenarios is no more than $20,000 per period for each $100 million notional of swap. This implies that on average a long haul book of $1 billion would have ineffectiveness of less than $200,000."(596) In an e-mail exchange referring to the analysis, Sinclair confirmed Stone's assumption that the "period" referred to is a month.(597) Although the analysis concerned transactions that would not qualify for the shortcut method under the DAG,(598) the results show that the ineffectiveness associated with Fannie Mae's hedges might have been as high as $2.4 million per year on a $1 billion hedge. Applying this average ineffectiveness to the notional amount Fannie Mae's interest-rate swap portfolio as of December 31, 2003, and assuming that the fair values of pay-fixed and receive-fixed swaps would move in opposite directions, the ineffectiveness might have reached more than $10 million per month. That amount might be understated. Fannie Mae undertook a second analysis of ineffectiveness in a portion of its portfolio in 2004. At that time, at KPMG's insistence, the Company tested the ineffectiveness associated with a series of term-out transactions. As a result of that analysis, Fannie Mae booked $67 million of ineffectiveness. (Concurrently, Fannie Mae also attempted to evaluate the total ineffectiveness it would have booked during prior years under the long-haul method. The effort proved too complex and was abandoned.)(599) - ---------- (596) Undated Analysis of Ineffectiveness in Long Haul Swaps, FMSE 365970-73, at FMSE 365972. (597) Id. (598) The hedge instruments were presumed to reprice outside of the plus-or-minus seven-day window around the repricing date of the debt instrument. Although the analysis concerned transactions for which long-haul accounting would be required under Fannie Mae's system, it shows the magnitude of ineffectiveness that can be generated when the terms of the instruments linked in a hedge do not match, and it demonstrates the extent to which Fannie Mae was willing to tolerate ineffectiveness as immaterial. (599) The Company's analysis of the $67 million ineffectiveness in its term-out transactions was not a substitute for ongoing ineffectiveness assessments, nor are the results necessarily reflective of ineffectiveness throughout this period. The analysis was based on a computation of ineffectiveness as of a point in time for a series of randomly selected derivatives. The results of that random selection were applied to the remaining portfolio of "similar" transactions. However, it is unclear how the Company determined that the selected derivatives were representative of those that were not selected, which it would have been required to do in order to extrapolate the results of its testing of the sample to the population. Further, it appears that the Company adjusted the results for one transaction, concluding that the selected derivative was not representative of the remaining population. That adjustment lowered the amount of the ineffectiveness when the results of the sample were extrapolated to the population. 159
Finally, Fannie Mae's periodic assessments of the ineffectiveness associated with its duration-matching approach to hedges of forecasted debt issuances also uncovered ineffectiveness. As noted above, according to the Company's analysis, the ineffectiveness associated with that relatively small portion of Fannie Mae's derivative portfolio reached $12 million ($21 million pre-tax). Fannie Mae booked that ineffectiveness and discontinued its hedging program for those transactions. Based on this evidence, and given the size of Fannie Mae's derivative portfolio overall, it appears that the ineffectiveness in Fannie Mae's portfolio was substantial. Management's own analyses - even when directed to only a portion of the portfolio - undermine the Company's view that the ineffectiveness at any given time necessarily or predictably would have been inconsequential. Nonetheless, Fannie Mae's hedge accounting documentation rested on this premise. We see no evidence, therefore, that Fannie Mae had established the foundation for the hedge accounting described in its documentation. D. Other Hedge Accounting Policies Our overall views regarding the Company's implementation of FAS 133 are influenced as well by decisions made on questions that arose after the standard's effective date. These decisions reflect an intention to inject flexibility into the Company's procedures to avoid strict application of accounting standards based on an assumption, often without analysis, that the deviations had an immaterial effect on the Company's financial statements. The first example concerns the situation in which a swap used in a term-out had a stub period that extended beyond the term of the debt instrument. Stone drafted a memorandum addressing the treatment of this issue in which she posited two alternative resolutions: the first alternative involved the adoption of the long-haul method, at least for the stub period; the second alternative was to "[t]reat the basis swap as a DNQ [does not qualify for hedge accounting] with the changes in fair value (from the last date the basis swap qualified for the shortcut method) running through earnings."(600) In her draft, Stone concluded that the second option "is the most viable solution as it is unlikely that option 1 will qualify for the long haul method." Two notes appear on the draft in Stone's handwriting. One says "M - operationally [sic] policy." The second says: Jonathan, Please review. Mary [Lewers] would like to know if we could apply an immateriality argument for these type[s] of transaction. Right now, we know of 1 w/ a mkt value of ~ 400K. Is this a policy you want to make? - ---------- (600) Mem. from K. Stone to Distribution, dated Apr. 13, 2001, FMSE 366485. 160
Kim(601) A second draft of the memorandum, with the same date as the original draft, includes the language noted above regarding the two resolutions, and the preference for the second option.(602) However, the memorandum includes a new concluding paragraph which states: Because the term of the stub period is always less than three months, the amount of ineffectiveness generated from applying the procedures documented above are [sic] likely to be immaterial. Therefore, we will continue to apply the shortcut method in situations where the basis swap has a stub and the basis swap is used in a cash flow hedge termout. Below this new paragraph appears the handwritten note: "Jonathan - Are you OK w/ this statement? K" The final version of the memorandum which was dated May 15, 2001 and issued as a new accounting policy, includes this final paragraph.(603) A similar series of events arose in connection with the analysis of an EITF Issue(604) that concerned, in part, the proper accounting for hedge gains or losses deferred on cash-flow hedges when the underlying debt has been extinguished and those deferred gains or losses are required to be reclassified into earnings. A draft memorandum prepared by Stone and dated April 12, 2001, contains the handwritten note, "Version 1 Follow the EITF [Emerging Issues Task Force]."(605) The memorandum addresses three issues raised by the EITF and concludes in two instances that certain amounts previously recorded as extraordinary items "should be recognized immediately in the Fee and Other Income (Expense) line of the income statement[.]" In the final version of the memorandum, dated May 7, 2001,(606) the memorandum repeats the conclusions regarding the accounting that would apply if the Company were to follow the EITF's guidance; the text of the memorandum was amended, however, to state that these conclusions would result if one were "Applying the EITF" or "Following the EITF." As to those issues, the memorandum contains a new conclusion that reverses the intention to follow or apply the EITF guidance: "However, due to the immateriality of these amounts, Fannie Mae's - ---------- (601) Id. (602) Mem. from K. Stone to Distribution, dated Apr. 13, 2001, FMSE 366483. (603) See Mem. from K. Stone to Distribution, dated May 15, 2001, FMSE 366575. (604) CLASSIFICATION OF A GAIN OR LOSS FROM A HEDGE OF DEBT THAT IS EXTINGUISHED, EITF Issue No. 00-9 (Fin. Accounting Standards Bd. 2000). (605) Mem. from K. Stone to Distribution, dated Apr. 12, 2001, FMSE 365130. (606) Mem. from K. Stone to Distribution, dated May 7, 2001, FMSE 365121. 161
policy will be to continue to record these hedge gains and losses as extraordinary items." The memorandum ends: "Fannie Mae will readdress the policies outlined above if the amounts become material. Another example of Fannie Mae's adoption of rules that bypass recognized accounting precepts is reflected in an e-mail exchange concerning a discrepancy between the instruments linked in a hedge transaction.(607) The e-mail reads: OK. We met with Peter at 4 and Jonathan joined us. I am sure much to his regret. It became apparent that between our 9:30 meeting with Jonathan and our 4, Jonathan had given additional consideration to the 6 mos rule on the forward starting RF swaps. For so long as we put on forward starting receive fixed swaps on an occasional basis, his 6 mos. guidance works. The key here is on an occasional basis. Jonathan readily admits he made up the 6 mos. rule. If we were to do this on a programmatic basis - for example 15 to 20 billion in a week for rebalancing purposes, he is NOT comfortable with the 6 mos. rule he devised. His view is there must be substance behind linking derivatives needed to hedge a piece of debt. He did not intend to make a hard and fast rule out of the 6 months. The e-mail concludes: "By the way, today the desk did 3 trades where there is 6 mos. or great [sic] payer mismatch (what we thought we knew at 9:30). Ursula is checking to see if this qualifies as 'occasional' or it is part of a program. I believe her rationale is we put this under this am's guidance." Boyles did not recall this incident. Although these examples stand out, we believe they reflect and confirm an overall tone in Fannie Mae's implementation of FAS 133 that rested heavily on assumptions of immateriality to justify avoidance of applicable accounting rules in deference to minimizing earnings volatility, reliance on existing accounting systems, and avoiding the need to alter the Company's business practices. VI. FINDINGS REGARDING OVERSIGHT OF FANNIE MAE'S HEDGE ACCOUNTING Company records and interviewees' recollections show that five different groups were involved to a greater or lesser degree in Fannie Mae's implementation of FAS 133. These were: Fannie Mae's senior management, its Board of Directors, the FASB, KPMG, OFHEO and Fannie Mae's Office of Audit. Each played a different role and had a different level of involvement in the process. - ---------- (607) E-mail from Laura Simmons to K. Skladony, dated Mar. 8, 2004, FMSE-E 520178. 162
A. Senior Management We have incorporated into our discussion the involvement in the Company's FAS 133 implementation process of Howard, Spencer, and Boyles. In our view, each of these individuals knew or should have known the concepts that motivated the Company's approach to hedge accounting. Spencer and Boyles, as the senior accountants involved, knew or should have appreciated that the Company's accounting did not meet the requirements of FAS 133 in important respects. As will be seen, however, for the most part, they took steps to ensure that KPMG was informed of the Company's policies, and they relied on KPMG's acceptance of the Company's approach. The September 2000 briefing to the Office of the Chair is the only documented occasion prior to 2001 in which Raines was briefed on FAS 133 issues insofar as they affected Fannie Mae.(608) There is no report of any other occasion in which FAS 133 issues were discussed with Raines prior to 2003, and no other interviewee reported any involvement by Raines in these matters. During his interview, Raines disclaimed any involvement with the implementation of FAS 133. Raines's disclaimer in this respect is credible and supported by the lack of significant documentary or anecdotal evidence linking him to these accounting decisions. B. The Board and the Audit Committee Management's reports to the Board on FAS 133 developments were sporadic and, for the most part, management withheld significant information regarding the Company's implementation efforts and objectives. In November 1998, the Audit Committee received a briefing paper on new accounting developments that noted the FASB's release of its "long awaited standard on the accounting for derivative financial instruments."(609) The Company described the standard in five sentences and reported that "Fannie Mae does not expect the adoption of the new standard to significantly affect earnings, however the new standard will require a large portion of our derivatives to be marked to market through equity."(610) The report describes management's emphasis on the fact that "a number of operations and systems changes will need to be made."(611) - ---------- (608) A document prepared by Boyles, dated June 23, 2000, and a draft transmittal memorandum to Raines captioned "Preliminary Report on How FAS 133 May Affect Our Customers," address the possible impact of FAS 133 on third parties (primarily mortgage bankers). See Undated Preliminary Report on How FAS 133 May Affect Our Customers, Zantaz document 2207964. Raines was also involved in efforts to lobby the FASB and the SEC to eliminate impact of the accounting for the time value of options on income statements. (609) FMSE-SP 87043-047. (610) Id. (611) Id. 163
A year later, in November 1999, management reported to the Audit Committee on the impact of FAS 133. Management stated, among other things: Fannie Mae has completed its project to identify and classify all our derivatives according to the criteria specified in the new standard. In addition, we have substantially completed the development of a system that will capture and record our derivatives and hedging activities in our financial statements. We have been conducting shadow processing activities for these activities since the first quarter of 1999 to evaluate how our derivatives behave in different environments and to prepare operationally for the implementation of the standard. We plan to use the information gathered in our shadow processing phase to develop strategies, if necessary, to offset earnings or equity volatility.(612) This report appears to confirm that, although the FASB had extended the effective date of FAS 133 by one year, the Company did not take advantage of the extension to improve its processes and instead viewed its implementation effort as substantially complete by the end of 1999.(613) The minutes of the Audit Committee's meetings in November 1999 and April 2000 also reflect these status updates. In November, according to the minutes of an Audit Committee meeting, "Ms. Spencer reported that, even though the effective date for FAS 133 . . . has been deferred to January 1, 2001, Fannie Mae has made it a priority to develop the systems necessary to implement and analyze the impact of the new standard, particularly its effect on volatility in shareholder equity."(614) The minutes of the April meeting reflect the comments of a KPMG partner, who "noted that KPMG would perform a significant amount of testing of financial reporting and planning during the - ---------- (612) Id. at FMSE 367238. The report expressed some concern for Fannie Mae's customers: "[S]ince it is almost impossible to obtain a perfect hedge of loans or mortgage servicing rights, mortgage bankers will likely experience significant volatility in earnings." Id. (613) In a document entitled "Audit Committee Talking Points" dated November 16, 1999, there is reference to the extension and a statement that "[t]his will give us enough time to adequately develop the necessary accounting and reporting systems." Zantaz document 694338, at 1. As noted above, it does not appear that substantial efforts were made in this regard during 2000. The talking points also report that management "expect[s] this standard to cause significant volatility in our equity but that it should have only a minor impact on our earnings." Id. (614) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Nov. 16, 1999, FMSE-IR 322676-82, at FMSE-IR 322679. 164
year 2000. She stated that KPMG's test work would emphasize the system and accounting changes required to account for hedges under the new FAS 133 standard."(615) Howard's report to the Board in February 2001, following the effective date of FAS 133, was critical of the standard itself, and focused on the impact of the standard on the Company's financial statements. At least insofar as the minutes of this meeting and Howard's talking points reflect, at no time did management report to the Board on the tenets that the Company had used to implement FAS 133 or the critical decisions that the Company had made with respect to systems development or accounting policy.(616) C. FASB The Company's interactions with the FASB focused on narrow questions. The most important of these related to (1) the time value of options; (2) the term-out transactions and the Fannie Mae "carve-out"; and (3) transactions involving hedges of the risk-free rate. On the first issue, the FASB did not accommodate Fannie Mae's concerns. On the second issue, the FASB initially considered the shortcut method a possibility for term-outs, but unequivocally rejected that approach in FAS 138. As to the third issue, Fannie Mae prevailed on FASB to change the standard. Based on our interviews and the documentary record, it would be incorrect to assume that the FASB had a significant appreciation of the facts behind the Company's accounting policies surrounding FAS 133, or that the FASB in any way approved the Company's accounting. On the contrary, in several significant respects, the FASB's views regarding particular transactions were based on insufficient information regarding the Company's approach to hedge accounting, and therefore should not be considered probative. The Company's first letter to the FASB in October 1997 addressed both basis swaps and term-outs.(617) The letter described these transactions in general terms. The letter asked the FASB to assume that there would be minimal ineffectiveness in the term-out transactions and that the transactions would therefore qualify for hedge accounting. As to the basis swaps, it simply asked the FASB to "allow basis swaps to qualify for hedge accounting and to be marked to market through equity subject to the transaction meeting the same qualifying criteria as a cash flow hedge."(618) As to the latter - ---------- (615) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 18, 2000, FMSE-IR 396847-53, at FMSE-IR 396851. (616) Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Feb. 20, 2001, FMSE 504046-058, at FMSE 504054-55. (617) Letter from S. Rajappa to Director of Research and Technical Activities, Fin. Accounting Standards Bd., dated Oct. 6, 1997, FMSE-IR 217015-26. (618) Id. at FMSE-IR 217017. 165
transaction, the Company suggested that the transaction is "perfectly effective."(619) It did not discuss any differences in the terms of the instruments that might have injected ineffectiveness into the transaction. Moreover, the FASB did not agree with the Company's suggestion that changes in the fair value of the offsetting interest-rate swaps be included in equity, requiring instead that the fair value of the offsetting swaps be recognized in income. The Company sent a letter to the FASB in January 1999 addressing an issue concerning differences between the effective interest rate on Discount Notes and LIBOR.(620) The letter specified that the "reset dates on the floating leg of the swap match the rollover date of the Discount Notes." The letter did not indicate that, according to Company policy, the reset dates could fall within seven calendar days of each other. Also in 1999, the Company drafted a letter to the FASB regarding the FASB staff's tentative decision that would have precluded the use of the shortcut method under paragraph 68 of FAS 133 for derivatives existing at the date of adoption of the standard. The letter stated: "We have already begun our systems development under the assumption that the majority of our swaps qualify for the shortcut method and we have plans to terminate most of the hedging relationships that do not qualify for the shortcut method prior to our adoption of FAS 133."(621) In a letter dated April 3, 2000, when the Company's accounting policies and systems were largely in place, the Company reported to the FASB that it "intends to rely on the 'shortcut' method provided for in the Statement to ease the implementation effort of adopting FAS 133."(622) Yet, when the FASB, in its adoption of FAS 138, rejected reliance on the shortcut method for term-out transactions, Fannie Mae did not change its accounting treatment. Fannie Mae approached the FASB again in 2004 regarding a specific term-out transaction that KPMG questioned; the FASB confirmed that the transaction could qualify for hedge accounting generally but, again, there is no indication that Fannie Mae explained its accounting for the transaction in any detail. Based on the information that Fannie Mae provided to the FASB, we have no basis to conclude that the FASB was sufficiently familiar with Fannie Mae's approach to FAS 133 such that the FASB can be held to have approved the Company's accounting. The correspondence suggests that Fannie Mae's presentations to the FASB were directed - ---------- (619) Id. at FMSE-IR 217018. (620) Letter from J. Boyles to R. Wilkins, dated Jan. 20, 1999, FMSE 57580-81. (621) Letter from T. Howard to Derivatives Implementation Group, dated Sept. 29, 1999, FMSE 366044-45, at FMSE 366044. (622) Unsigned letter from T. Howard to Director of Research and Technical Activities, Fin. Accounting Standards Bd., dated Apr. 3, 2000, FMSE 48027-30. Talking points dated March 29, 2000, prepared for a conference call with the FASB, use similar language. FMSE 61130. 166
to narrow questions, and did not address important issues (such as differences in the repricing dates) to which the FASB might have reacted negatively. The FASB's responses to Fannie Mae's letters go only so far as the information in the Company's inquiries allow. In our view, reliance on the FASB's views to support Fannie Mae's accounting treatment, except as to narrow issues, has no basis in the record.(623) D. KPMG Fannie Mae encouraged KPMG's involvement in the implementation of FAS 133 and provided KPMG with relevant information. The record shows that KPMG had access to Fannie Mae's written hedge accounting policies, including the DAG, and did not raise significant objections to those policies. Both documents and recollections of those involved confirm that KPMG was a part of the implementation process from the outset. A number of Fannie Mae documents refer to the importance the Company placed on receiving KPMG's approval of the Company's approach to hedge accounting even before FAS 133's effective date. Several interviewees referred to the Company's "special engagement" of KPMG to review the DAG and other issues relating to hedge accounting. Additional evidence appears to confirm that the Company paid KPMG an additional fee in 2000 for the work associated with the FAS 133 implementation effort. We have no reason to doubt the statement in Spencer's December 1998 memorandum that those involved in the FAS 133 implementation process had "met with KPMG and walked them through each type of transaction and what our recommended accounting treatment would be. We explained how we didn't want to start systems work without KPMG's sign-off on our accounting proposals."(624) As noted above, KPMG reported to the Audit Committee in early 2000 that it intended to pay significant attention to Fannie Mae's implementation of FAS 133, and KPMG's audit workpapers for 2000 discuss many of the Company's most significant policies in this area.(625) - ---------- (623) A representative of Huron approached the FASB with a request that they discuss with us the FASB's involvement with Fannie Mae's accounting generally, and its implementation of FAS 133 specifically. The FASB declined to participate. (624) Mem. from L. Spencer to T. Howard et al., dated Dec. 4, 1998, FMSE-IR 192988-992, at FMSE-IR 192988. (625) The one significant exception we have identified relates to a feature of the DEBTS system, which links the derivatives to debt. A feature of DEBTS is that these links are not necessarily maintained, at least for Discount Notes. Rather, pay-fixed interest-rate swaps are relinked automatically to hedge (in order of priority) term-out transactions, medium- or long-term variable rate debt, and the rollover of discount notes. As a consequence, a pay-fixed swap linked to the rollover of Discount Notes might automatically be relinked at a later point to hedge a term-out transaction. It appears that this aspect of the DEBTS system was not well 167
This is not to say that KPMG necessarily viewed the Company's application of FAS 133 as correct in all respects. A preponderance of the evidence suggests that KPMG did not object to the Company's application of FAS 133 in several important respects because the auditors came to believe that any departures from a proper application of the standard would be immaterial. KPMG workpapers reveal a significant amount of attention to various aspects of the Company's FAS 133 accounting policies and systems in the months leading up to the effective date.(626) Among other things, the workpapers confirm that KPMG "[r]eviewed the 67 transactions identified in the [Derivatives Accounting Guidelines] and determined that the accounting treatment for the transactions . . . are [sic] reasonable based on the information presented."(627) A KPMG workpaper prepared in 2002 describes the audit work conducted during that year.(628) The document states: "It is important for Fannie Mae to receive hedge accounting treatment because hedge accounting treatment results in only the ineffective portion of a hedge being reflected in earnings. Otherwise, the entire gain or loss related to the derivative is reflected in earnings, resulting in earnings volatility."(629) The workpaper then describes and discusses various aspects of Fannie Mae's hedge accounting. The workpaper discusses the hedge of anticipated issuances of benchmark notes and the rollover of discount notes together. As to the anticipated debt issuance, the workpaper states: Under FAS 133, if the critical terms of the hedging instrument and the hedged item are the same, an entity can assume that a hedge will be perfectly effective at offsetting the hedged transaction's variability in cash flows, which is referred to as the "short-cut" method. Additional criteria necessary for applying the short-cut method include eligible risk and duration parameter.(630) - ---------- understood within the Company, and neither Financial Standards nor KPMG appears to have been aware of it at the time the Company adopted FAS 133. (626) KPMG, Derivatives and Hedging Process and FAS 133 Audit Program, dated Dec. 31, 2000. (627) KPMG, Completion Memorandum, dated Dec. 31, 2000. (628) KPMG, Process Analysis Document. (629) Id. Similar language also appears in the DAG at FMSE-IR 12847. (630) Id. 168
The workpaper then describes an acceptable "duration range" for these purposes - -- that is, the extent to which the duration of the hedging and hedged instrument must match to qualify for Fannie Mae's shortcut treatment. The report states further that Fannie Mae "prove[s] out each year . . . that using the shortcut method based on this prescribed duration range is proper by selecting a sample of anticipatory hedge transactions and measuring hedge ineffectiveness."(631) A handwritten note in the workpaper reports that "KPMG requested that Fannie Mae complete long-haul calculations for selected transactions associated with employing the duration check to verify that any ineffectiveness would be immaterial."(632) KPMG's workpapers also acknowledge that "Fannie Mae can use a shortcut method to apply FAS 133 provisions to an interest rate swap in a cash flow hedge when it is assumed that there will be no hedge ineffectiveness."(633) A handwritten note following that statement refers to the rollover of discount notes: Fannie Mae Derivative Accounting Guidelines Section IV (pages .20 and .21) discuss the policy for assuming no ineffectiveness for the issuance and rollover of discount notes. Discount notes are considered forecasted fixed rate debt instruments, and Fannie Mae hedges the risk of changes in interest payments attributable to changes in the benchmark rate related to the forecasted issuance of discount notes. Fannie Mae assumes no ineffectiveness and applies the shortcut method if the rollover date of the disc. notes and the reset on the interest rate swap are w/in +/- 7 days of each other. (A maximum of five business days.) Although the policy is not in strict compliance with FAS 133 and DIG, KPMG reviewed and agreed to this policy at the time when Fannie Mae was preparing their accounting guidelines and implementing FAS 133. Any ineffectiveness would result from near-term movements in rates as the yield curve at 90 days vs. 95 days (5 business day difference) or longer periods would - ---------- (631) Id. (632) Id. A separate, handwritten statement in that same document appears to say: "These parameters are included in Fannie Mae Derivative Accounting Guidelines and calculations are prepared to prove out effectiveness and ineffectiveness employing a long-haul approach." Id. (633) Id. 169
not be different. Any ineffectiveness would flow through the financial statements within a given quarter.(634) At another part of the workpaper, KPMG states: "if the critical terms of the Discount Note and the swap are the same, no hedge ineffectiveness is assumed using short-cut method."(635) The KPMG workpapers also address the Company's use of the de minimis test for assessing hedge effectiveness. KPMG stated in its workpaper that it had "indicated we would approve a ten bps threshold" and that the use of the test "will still result in earnings impact because the market values of the hedging instrument and the hedged item will still be compared using the dollar offset method."(636) A handwritten note in the paper states: "The 'de minimus'[sic] measure is used only for anticipated issuance of debt. The time being hedged is minimal [illegible - possibly a discussion of the short time period between the planned debt issuance and the actual debt issuance] and therefore it is reasonable to conclude that given the short term of the hedge [illegible] and the historical movements in rates w/in a couple day period --> It is reasonable to conclude that the hedge will be highly effective and qualify for hedge accounting."(637) With respect to existing debt obligations, the workpaper notes: "Fannie Mae assumes no hedge ineffectiveness if the critical terms of the debt obligation and the swap are the same." For these transactions (as well as for anticipated debt issuances and discount not rollovers), the workpaper provides that "if the critical terms of the hedging instrument and of the hedged item do not match, the Hedge Desk [or Funding Desk as the case may be] must assess the effectiveness of the hedges via the long haul method."(638) These workpapers establish that KPMG had access to Fannie Mae's principal hedge accounting policies and was aware of the principal features of Fannie Mae's approach to hedge accounting. KPMG accepted certain aspects of Fannie Mae's approach that were not consistent with a strict reading of FAS 133, and that acceptance was based largely on an understanding that the impact of those deviations would be immaterial. - ---------- (634) Id. (635) Id. (636) Id. (637) Id. (638) Id. (emphasis in original). 170
E. OFHEO Interviewees indicated that Fannie Mae's implementation of FAS 133 was open to OFHEO as well. Several interviewees recalled at least one briefing during which they explained Fannie Mae's approach to hedge accounting to OFHEO personnel and provided OFHEO with copies of relevant materials, including the DAG. In its annual report to Congress in 2002, regarding examination work conducted during calendar year 2001, OFHEO reported that it had "proctor[ed] the implementation of FAS 133 and ongoing compliance."(639) Although OFHEO was careful to avoid a categorical statement to the effect that it had conducted an audit of Fannie Mae's FAS 133 accounting, it acknowledged in its report that it had "evaluated the appropriateness of GAAP accounting for derivatives." In this respect, OFHEO reported that "FAS 133 implementation has been deliberate and well-documented with the necessary investments made to provide systems needed to ensure ongoing compliance."(640) This report appears to confirm that OFHEO had access to Fannie Mae's accounting policies and other information relevant to the implementation of FAS 133, and that OFHEO initially found no weaknesses in Fannie Mae's approach. In our conversations with OFHEO regarding this report, agency officials acknowledged that it had reported no deficiencies in Fannie Mae's application of FAS 133 as of 2002. The officials noted, however, that they are not the Company's auditor and they (like the Company's Board) are entitled to rely on the auditor for a thorough assessment of the Company's accounting. OFHEO also was very candid in its assessment of the weaknesses in its own examination processes prior to 2003.(641) - ---------- (639) OFHEO Report to Congress, dated June 14, 2002, at 25, available at http://www.ofheo.gov/Media/Archive/reports/ar61402.pdf. (640) Id. at 29. (641) Among other things, OFHEO pointed us to the July 2003 testimony of its then-Director Armando Falcon, before the Senate Banking, Housing and Urban Affairs Committee. In response to a question from Senator Sarbanes as to "how we get at the fact that it [referring to Freddie Mac's accounting problems] was all happening unbeknownst to the regulator," Falcon stated: "That's a good question, Senator Sarbanes. And what we are working to do at OFHEO is to build an in-house team of accountants that can take a look at issues like this." Falcon suggested that OFHEO "should have a lower threshold of what gets included in this annual report." In response to a question from Senator Sununu, Falcon stated: "In addition, it has been clear for a long time now to me that the Agency needed to build the depth in at least the examination program, and we are working to try to do that." See Hearing Before the Committee on Banking, Housing and Urban Affairs, S. Hrg. 108-833, July 17, 2003, at 18, 26. 171
F. Internal Audit Fannie Mae's Internal Audit personnel conducted an audit of the FAS 133 implementation effort in the fall of 2000. According to the December 2000 report of that effort, the audit "was conducted to determine whether internal controls over derivatives transactions are adequate and functioning effectively to ensure compliance with FAS 133."(642) The report states that Internal Audit had (among other things) "validated FAS 133 system criteria for classifying derivatives as cash flow, fair value and DNQ (does not qualify) hedges" and "verified that processes and controls had been implemented to meet the FAS 133 documentation requirements for hedging relationships." Internal Audit had no criticisms to report, and concluded that "transactions are accounted for and reported in accordance with FAS 133 requirements." Interviews with several of the audit personnel involved in the FAS 133 audit confirm that, in fact, this statement was incorrect. Internal Audit did not conduct an audit of the Company's compliance with the requirements of FAS 133, as its conclusion might suggest. Internal Audit conducted a controls audit of the Company's systems for implementing hedge accounting. In addition, the controls audit was not based on FAS 133, but on the Company's implementation of FAS 133 as reflected in the Derivative Accounting Guidelines. * * * In sum, several entities were involved in the development of Fannie Mae's FAS 133 accounting policies as they took shape in 1999 and 2000. Although Fannie Mae would later come to rely on its communications with the FASB to support its positions, we find little in the record to support the view that the FASB had a sufficiently clear or complete picture to approve the Company's hedge accounting, or that it did more than respond to the narrow questions that Fannie Mae posed based on hypothetical scenarios that did not explain aspects of the Company's transactions that the FASB might have found troubling. Both OFHEO and the Company's internal auditors had the opportunity to review Fannie Mae's FAS 133 program in detail and both issued reports that praised the Company's implementation efforts without qualification. Indeed, both the OFHEO and Internal Audit reports contain language that is so broad it could have misled readers into believing that the conclusions say more than was intended, particularly with respect to the Company's compliance with GAAP. OFHEO's review is particularly noteworthy given the fact that, a year after OFHEO issued its 2002 report, it would begin its Special Examination of Fannie Mae and conclude that the errors in Fannie Mae's hedge accounting were numerous and fundamental. - ---------- (642) Draft Audit Report, Office of Auditing, FAS 133 Audit, dated Dec. 22, 2000, FMSE-E 47997-8000, at FMSE-E 47999-8000. 172
KPMG had a comprehensive view of the most significant features of the Company's FAS 133 policies. It accepted the proposition that FAS 133 was subject to a materiality standard, and maintained throughout that any departure from FAS 133's requirements reflected in the Company's policies would not distort the Company's financial statements. Having reported to the Board that it would review the Company's implementation efforts in 2000, it then reported that the Company's financial statements were fairly stated. It did not deviate from that view. Finally, the Board was entitled to rely -- and by all accounts did rely -- on the representations and conclusions it received. Management's consistent message to the Board was that the Company was prepared for any challenges that FAS 133 presented, and the reports that the Board received from KPMG, OFHEO and Internal Audit until 2004 confirmed that view. 173
CHAPTER VI: OTHER ACCOUNTING ISSUES PART A: ACCOUNTING FOR THE ALLOWANCE FOR LOAN LOSSES I. INTRODUCTION In this Part, we address the propriety of the Company's accounting for the allowance for losses on loans in its mortgage portfolio and the liability for losses associated with its guaranty of the loans underlying its MBS (collectively referred to as the "Allowance"). The facts showed that management's methodology for setting the Allowance in the period from 1997 through 2003 was not based upon a detailed assessment of the loss exposure inherent in the portfolio, as required by GAAP, and did not reflect the continued trend of improved credit quality that the Company reported in its public statements. Accordingly, all evidence would indicate that the Allowance was overstated as early as 1997. However, we saw no evidence that management intentionally utilized the Allowance to manipulate earnings or to offset unrelated one-time events. Of course, by not reducing the reserve when they knew, or should have known, that it was overstated, management had a potential "war chest" for meeting earnings targets, and the evidence showed that management periodically did consider drawing down the reserve to offset unrelated expenses. Furthermore, had the excess reserve been reversed when management first knew, or should have known, the reserves were overstated, the reduction in income and its "non-recurring" nature would have made the subsequent year's earnings growth goals that much more difficult to achieve. Finally, a downturn in the economy or other adverse changes in credit quality would have required the Company to add to the reserve, further reducing earnings and making it difficult to meet year over year earnings targets in a given period. In summary, we have the following conclusions about management's historical accounting for the Allowance: - - As early as 1997, management believed that the Allowance was in excess of what was required and viewed such excess as a "war chest" to mitigate the impact of unrelated financial events; however, we saw no evidence that management did, in fact, use the Allowance to offset unrelated expenses or to manage earnings to meet EPS targets in a given period. - - The Allowance was essentially unchanged at approximately $800 million throughout the seven-year period from 1997 to 2003, despite improved credit quality and improved credit administration by the Company. Credit losses(643) as a - ---------- (643) Credit losses or "credit-related losses" are defined by Fannie Mae as the sum of foreclosed property expenses plus charge-offs. Fannie Mae 2002 Annual Report (Form 10-K), at 126 (Mar. 31, 2003) (hereinafter "2002 Form 10-K"), available at http://fanniemae.com/ir/pdf/sec/2003/f10k3312003.pdf. 174
percentage of the average book of business declined from 0.027% in 1998 to 0.006% in 2003; because management did not adjust the Allowance based on this experience, the number of years of losses covered by the Allowance increased from 3.3 years of losses in 1998 to 7.2 years of losses in 2003. - - The methodologies used by management to set the Allowance during the period from 1997 through 2003 were not based upon a detailed assessment of the loss exposure inherent in the portfolio, as required by GAAP, and did not measure or reflect the effects of improved credit quality that the Company reported during the period. Management essentially derived the Allowance level by setting the provision for losses (i.e., expense) equal to charge-offs. GAAP requires the Allowance to be assessed and adjusted to reflect management's documented estimate of losses inherent in the portfolio on a periodic basis (e.g., quarterly) with any necessary adjustment to the level of the Allowance reflected as an increase or decrease to the provision for losses. - - During the period, management failed to develop a detailed bottom-up methodology to estimate losses inherent in the portfolio as of the balance sheet date as required by GAAP. SEC Staff Accounting Bulletin No. 102 ("SAB 102") applied to Fannie Mae's Allowance accounting practices as of December 31, 2002, when Fannie Mae became a registrant. Management did not begin working on a detailed Allowance model until some time in 2002, despite KPMG's inquiries as early as 1998 about a detailed Allowance model to support the level of the Allowance, and the model was not deemed complete until the third quarter of 2003.(644) Management has reduced the Allowance by approximately $300 million in 2004, and will restate historical periods as part of the restatement. II. BACKGROUND A. Identification of the Allowance as a Potential Issue Our investigation into Fannie Mae's accounting for the Allowance was the result of three specific areas of work. First, during our review of the 1998 catch-up adjustment,(645) we found a 1999 planning document from the files of Shaun Ross, then a Project Manager in the Controller's Office, which had significant implications for - ---------- (644) Allowance for Loan Losses Audit Report, dated Dec. 15, 2003, Zantaz document 2229258, at 1, 2; Single Family (SF) Allowances for Loan Losses (ALL) Process During Restatement Period 2001 - through 2004, dated Aug. 1, 2005, Zantaz document 2228409. (645) See discussion supra Chapter IV. 175
management's use of the Allowance.(646) As detailed above, the document was a set of schedules dated January 7, 1999 and labeled "Earnings Alternatives," and presented three alternative scenarios for recognizing the 1998 catch-up expense.(647) This document was prepared by Ross at the request of Leanne G. Spencer, then Controller, in connection with the finalization of year-end results for 1998. The third alternative of this schedule appeared to reflect management's consideration of an adjustment to the fourth quarter 1998 results to decrease the provision for losses by $100 million, which would have had the effect of reducing the Allowance by the same amount. The document suggested, at a minimum, the possibility that then management believed that there was a surplus in the Allowance sufficient to absorb a $100 million reduction while maintaining adequate protection against losses.(648) Second, during our review of KPMG's audit workpapers for fiscal years 1998 through 2002, we noted an absence of support prepared by management for the "build up" of the Allowance. Additionally, we found an internal KPMG memorandum that described a meeting on August 11, 1998 with Timothy Howard, then Fannie Mae's CFO, which reflected Howard's expectation that an allowance model would be completed by the end of 1998. The memorandum also stated that Howard "hopes to have the analysis that will provide a solid basis to determine if the [A]llowance needs to be increased, decreased or is at an approprite [sic] level."(649) Ken Russell, the author of this KPMG memorandum, wrote in the memorandum that he did not expect that an increase to the Allowance would be required. (650) Shortly following the date of this memorandum, a document that contained KPMG's notes from a December 31, 1998 Fannie Mae conference call stated the following: "Allowance may be conservative. Are waiting to adjust when they have a sense on long term trends."(651) While the KPMG audit workpapers for 2003 referred to and included a Fannie Mae model supporting a majority of its Allowance, we could not locate similar information in the prior years' audit workpapers. We also noted in the KPMG workpapers other documents that suggested that the components of the 2003 Allowance - ---------- (646) Fannie Mae 1999 Income Statement schedules, dated Jan. 7, 1999, FMSE-IR 21476-79, at FMSE-IR 21477-79. (647) See discussion supra Chapter IV. (648) Compare Fannie Mae 1999 Income Statement schedules, dated Jan. 7, 1999, FMSE-IR 21476, at FMSE-IR 21479, with Fannie Mae 2000 Annual Report, at 43, available at http://www.fanniemae.com/global/pdf/ir/annualreport/2000/financials.pdf. (649) Mem. from Ken Russell to Julie Theobald and Eric Smith, dated Aug. 25, 1998. (650) Id. (651) Fannie Mae Conference Call Notes December 31, 1998, dated Jan. 14, 1999. 176
continued to contain a general reserve category known as "other projected risks," which contained amounts not associated with modeled risks or any particular pool of loans.(652) We subsequently learned that management's process of setting the Allowance balance in the late 1990s through 2003 consisted of quarterly meetings whereby single-family and multifamily loan activity and related credit loss metrics were reviewed. The result of these meetings was to set the provision for loan losses equal to charge-offs.(653) This practice was in contrast to GAAP, which requires the level of the Allowance to reflect an estimate of losses inherent in the portfolio as of the balance sheet date with any required adjustment flowing through the income statement. The third and final reason for our focus on the Allowance was a review of Fannie Mae's published financial statements. We observed that the Allowance balance was essentially unchanged at approximately $800 million throughout the seven-year period from 1997 through 2003. We also noted the following disclosure regarding the Allowance from 1998 to 2002 in Fannie Mae's 2002 Form 10-K: "Over the past five years, our combined allowance for loan losses and guaranty liability for MBS have remained relatively stable although our book of business has expanded. This trend reflects improvements in the credit performance of our book of business."(654) This statement suggested that while the risk grew along with the book of business, the Allowance remained constant due to improved credit performance. However, our own analysis showed that credit performance improved significantly more than the increase in size of the book of business. That is, the credit losses as a percentage of the average book of business declined from 0.027% in 1998 to 0.006% in 2003, which equated to a seventy-eight percent decrease in the percentage of credit losses. In contrast, the size of the book grew from approximately $1.1 trillion in - ---------- (652) Draft Fannie Mae Allowance for Loan Loss Methodology, dated Dec. 29, 2003. The 2003 Allowance contained $51.3 million for "other projected risks"; in 2002, the amount was $149.8 million. The total Allowance balance in 2002 and 2003 was $808.2 million and $796.9 million, respectively. See Q4 2004 Allowance for Loan Losses Summary, dated Jan. 26, 2005, FMSE-IR 283353. (653) Although Fannie Mae's documentation for the Allowance in 2003 does not explicitly indicate a policy of setting the provision for loan losses equal to charge-offs, the Allowance balance continued to remain static throughout 2003. The Allowance did not change significantly from the fourth quarter of 1997 through the fourth quarter of 2002. See, e.g., Undated Notes to the File Quarterly Loss Review Fourth Quarter 1997 Review, FMSE-IR 280137-39, at FMSE-IR 280139; Undated Notes to the File Quarterly Loss Review Fourth Quarter 2002 Review, FMSE-IR 281079-80, at FMSE-IR 281080. (654) 2002 Form 10-K at 56. 177
1998 to approximately $2.2 trillion in 2003.(655) The combined effect of declining credit losses and the policy of provision that equaled charge-offs, led to a substantial increase in the number of years of loss coverage provided for in the Allowance:(656) TABLE 1. DOLLARS IN MILLIONS 1998 1999 2000 2001 2002 2003 2004 - ------------------- ------- ------- ------- ------ ------- ------- ------- Allowance $ 799.0 $ 801.0 $ 806.0 $803.0 $ 808.0 $ 797.0 $ 510.4 Charge-offs (245.0) (149.0) (117.0) (97.0) (123.0) (111.0) (290.9) Coverage Ratio 3.3 5.4 6.9 8.3 6.6 7.2 1.8 ======= ======= ======= ====== ======= ======= ======= These facts raised several questions about the propriety of the Company's historical accounting for the Allowance, including: (1) whether management intentionally maintained an excessive Allowance to manage current and/or future earnings volatility; (2) what was the Company's methodology for supporting its Allowance balance, and how was the methodology documented; (3) what were the reasons for the Company's failure to adjust the Allowance; and (4) whether the Allowance was at an adequate but not excessive level to cover such losses for the period between 1998 and 2004. We determined that these questions merited further investigation. B. Relevant Accounting Principles The accounting principles that apply to an allowance for loan loss is well summarized in remarks by Lynn E. Turner, then Chief Accountant of the SEC, in a speech dated November 6, 1998:(657) GAAP for loans and the allowance for loan losses can be found in FASB Statements No. 5 and 114 and in the AICPA Audit Guide for Banks and Savings Institutions (Audit Guide). The basic premise in this guidance is best summarized in paragraph 7.29 of the Audit Guide which - ---------- (655) Loss Accounting Summary, dated April 2004, Zantaz document 2226776, at 3; Q4 2004 Allowance for Loan Losses Summary, dated Jan. 26, 2005; 2002 Form 10-K at 56; Fannie Mae, 2003 Annual Report (Form 10-K), at 41 (Mar. 15, 2004) (hereinafter "2003 Form 10-K"), available at http://www.fanniemae.com/ir/pdf/sec/2004/f10k03152004.pdf. (656) 2002 Form 10-K at 56; 2003 Form 10-K at 41; Q4 2004 Allowance for Loan Losses Summary, dated Jan. 26, 2005, FMSE-IR 283353; Loss Accounting Summary, dated May 2005, Zantaz document 1945386, at 3. (657) Lynn E. Turner, Chief Accountant, SEC, Remarks to the AICPA Bank and Savings Institution Annual Conference: Continuing High Traditions (Nov. 6, 1998). 178
states: "The allowance for loan losses should be adequate to cover probable credit losses related to specifically identified loans as well as probable credit losses inherent in the remainder of the loan portfolio that have been incurred as of the balance sheet date" . . . GAAP does not permit providing losses for future events. Guidance in this area is [also] provided by the banking regulators. For example, OCC ADVISORY LETTER 97-8 states that "Unallocated reserves . . . must not be used to obfuscate the determination of the overall allowance adequacy, mask significant deteriorating trends in asset quality, or 'manage' earnings." Additionally, Financial Reporting Release ("FRR") No. 28 requires the application of a systematic methodology and rationale in determining the allowance for loan losses. The GAAP principles that Turner described apply to both public and non-public companies. In addition, SAB 102, applied to public companies as of the 2002 financial statements. SAB 102 did not create a new standard, but only served to reinforce the existing accounting literature and GAAP requirements for determining and documenting the amount of the allowance for loan losses.(658) III. FINDINGS CONCERNING FANNIE MAE'S ACCOUNTING FOR THE ALLOWANCE A. Fannie Mae's Historical Practice for Setting the Allowance 1. Overview of the Components of Fannie Mae's Allowance Fannie Mae's Allowance covers both single-family and multifamily loans held in its loan portfolio. The guaranty liability relates to Fannie Mae's guaranty of MBS held by Fannie Mae in its portfolio as well as outstanding MBS held by third party investors. Because management determined that the credit risks are identical, management used the same methodology to set the Allowance for loan losses and guaranty liability for MBS.(659) - ---------- (658) SELECTED LOAN LOSS ALLOWANCE METHODOLOGY AND DOCUMENTATION ISSUES, SEC Staff Accounting Bulletin No. 102, at 1, 3 (July 6, 2001), "draws upon existing guidance, in Commission rules and interpretations, generally accepted accounting principles, and generally accepted auditing standards, and explains certain views of the staff in applying existing guidance related to loan loss allowance methodologies and supporting documentation." (659) 2002 Form 10-K at 55; 2003 Form 10-K at 40. 179
The following diagram illustrates the components of Fannie Mae's Allowance: (FLOW CHART) According to the Company's financial statements, the Allowance components were calculated based on the credit quality of the portfolio.(660) Management analyzed several risk factors when assessing credit quality. As depicted in the diagram above, these risks were analyzed based on the single-family and multifamily loan types. Risk factors that the Company analyzed for both single-family and multifamily loans included delinquency levels, historical loss experience, current economic conditions, and mortgage characteristics.(661) 2. Management's Process for Setting the Allowance Management did not employ a model or a well-documented detailed process to estimate the Allowance from 1997 through 2002. During this period, both the single-family and multifamily portions of the Allowance were based on the Company's "provision equal to charge-offs" policy, which resulted in the Allowance balance being maintained at a relatively static level. Even after implementing a more analytical approach for setting the Allowance in 2003, management maintained the Allowance balance at the existing level. (a) Early 1990s At one point in the early to mid-1990s, the Company apparently had maintained the Allowance at a level approximating two to three times trailing charge- - ---------- (660) Id. (661) Id. 180
offs. We did not find a document written contemporaneously that reflected this approach, but instead, a document from 2000, after the policy changed, that reflected it.(662) Jonathan Boyles described the policy from that time period as being "a laundry list of things to consider."(663) J. Morgan Whitacre, then Director-Business Planning, became responsible from sometime in late 1997 or early 1998 through the fourth quarter of 2003 for creating the presentations that the Controller's Office would use in connection with the Allowance setting process. Whitacre stated that when he first became responsible for this process, he used a "spreadsheet" he had "inherited" from his predecessor to calculate a range of two to three times trailing charge-offs that he would give to Spencer and Janet L. Pennewell for their presentation.(664) We have not located any contemporaneous information or specific documents that management relied upon to set the Allowance level during this period. (b) 1997 through 2002 From at least 1997 to 2002, the provision was set to be equal to charge-offs during a series of quarterly loss review meetings.(665) Information about the single-family and multifamily businesses was gathered and compiled by the Controller's Office, which was responsible for recommending an Allowance level. The collected information was discussed by management teams including the Controller's Office, Credit Policy, Single-family, and Multifamily. The process culminated with a CFO level meeting and a subsequent Chairman's loss meeting involving the CEO.(666) The CFO and CEO level meetings were typically attended by senior management including the Chief Operating Officer, Chief Credit Officer, Controller, as well as Single-family and Multifamily executives. However, attendees recalled that the - ---------- (662) Recommendation to Change Loss Accounting Methodology, dated Nov. 2000, FMSE-E 44694-714, at FMSE-E 44707. (663) Accord Allowance for Losses, dated July 1996, FMSE 31523-31. (664) At some point between 1998 and 2000, he created a "trend model," the output of which he also handed off to Janet L. Pennewell and Leanne G. Spencer. (665) See discussion supra at note 653. (666) These meetings were commonly referred to as either the "Review with Chief Financial Officer" or the "Loss Allowance Review Meeting with Chair" or "Chairman's Loss Meeting." 181
discussion at such meetings focused on credit losses and portfolio performance, rather than discussing specific Allowance requirements and related provision amounts.(667) The documents related to these CFO and CEO meetings confirm that the focus was primarily on credit losses.(668) Documents also show that management discussed data and statistics such as current delinquency levels, historical loss experience, changes in economic conditions, payment performance in areas of geographic concentration, loan purpose, product type, original loan-to-value ratio, and loan age or "seasoning." (669) Whitacre and others acknowledged that the Company did not keep documentation of the Allowance setting process other than the "Chairman's package," prior to 2002. (1) Change in Policy in 1997 At the end of 1997, management changed the Allowance policy to simply provide an amount equal to charge-offs, from the previous trailing two to three years of trailing charge-offs.(670) While not conclusive, this change in provisioning policy may be related to a memorandum written by Michael Quinn to Howard, among others, in late 1997, in which Quinn suggested that the provision be calculated at a level that "will approximate charge-offs."(671) In addition, other contemporaneous documents reflect a recommendation in 1998 to maintain the Allowance at a constant level.(672) - ---------- (667) Adolfo Marzol (then Chief Credit Officer), Robert J. Levin (then Head of Housing and Community Development), and Michael Quinn (then Senior Vice President for the Single-family Business), regularly attended the Chairman's Loss Meetings during this time. (668) Quarterly Loss Review documentation from 1998 through 2002, FMSE-IR 280194-96; FMSE-IR 280159-61; FMSE-IR 280317-19; FMSE-IR 280377-79; FMSE-IR 280424-26; FMSE-IR 280476-47; FMSE-IR 280535-36; FMSE-IR 280579-80; FMSE-IR 280636-37; FMSE-IR 280673-74; Zantaz document 3913800; FMSE-IR 280747-48; FMSE-IR 280783-84; FMSE-IR 280839-40; FMSE-IR 280904-95; FMSE-IR 280960-61; FMSE-IR 281019-20; FMSE-IR 281046-47; FMSE-IR 281079-80. (669) Id. (670) Undated Notes to File Quarterly Loss Review Fourth Quarter 1997 Review, FMSE-IR 280137-39, at FMSE-IR 280139. (671) Mem. from M. Quinn to Timothy Howard, R. Levin, Ann Logan, and Sam Rajappa, dated Dec. 5, 1997, FMSE-IR 338619-21, at FMSE-IR 338620. This and other examples of documents relevant to the discussion in Chapter VI, Part A can be found in the accompanying Appendix, at Tab A.1. (672) 1998 Loss Provisioning Strategies, Zantaz document 1512035, at 1. 182
Management highlighted this change to the Board and to the investing public in 1998 and in 1999.(673) In particular, Howard stated that: "At the present time our intent is to continue to have our quarterly loan loss provision be roughly in line with charge-offs, so as to maintain our loss allowance approximately level. We recognize that our $800 million loss allowance looks conservative in the current environment. But we believe it is prudent to hold it there, at least until we can get a better sense as to where the longer-term loss trends may settle out."(674) The decision to hold the Allowance constant, and to change the provisioning policy to equal charge-offs, also may be related to the fact that beginning sometime in 1997, the Company began experiencing charge-off gains, instead of losses on foreclosures.(675) By late 2000, management was aware that its accounting for "charge-offs" was not "optimal."(676) In a presentation to Howard, Pennewell and Whitacre proposed a change in how Fannie Mae would recognize losses and gains on foreclosures.(677) Under the proposed methodology, which was referred to as "waterfall accounting,"(678) the Company would allocate gains "on a loan by loan basis to first - ---------- (673) Notes for January 1998 Board of Directors Meeting, Zantaz document 1505640, at 6 ("Our projected provision for losses tracks our expected chargeoffs. Our intention this year is to have our provision equal our chargeoffs, so if we continue to do as well on chargeoffs as we did in the second half of last year our provision would be even lower."); Notes for January 14, 1998 Conference Call, Zantaz document 1349789, at 3 ("Our current intent for this year is to have our loss provisions equal our chargeoffs - with only small differences each quarter such as we had last quarter. That would keep our loss allowance at around $800 million. Once we get a little more experience with where our chargeoffs are setting out - and a sense of what may drive them in the future on a trend basis - we will revisit this provisioning approach."); Tr. Notes for Jan.14, 1999 Conference Call, FMSE-IR 278342-49, at FMSE-IR 278345. (674) Tr. Notes for Jan. 14, 1999 Conference Call, FMSE-IR 278342-49, at FMSE-IR 278345. (675) Recommendation to Change Loss Accounting Methodology, dated Nov. 2000, FMSE-E 44694-714, at FMSE-E 44698; accord Mem. from M. Quinn to T. Howard, R. Levin, A. Logan, and S. Rajappa, dated Dec. 5, 1997, FMSE-IR 338619-21, at FMSE-IR 338620. (676) E-mail from J. Morgan Whitacre to J. Pennewell, dated Dec. 28, 2000, FMSE-E 1819010-11, at FMSE-E 1819010. (677) Recommendation to Change Loss Accounting Methodology, dated Nov. 2000, FMSE-E 44694-714. (678) See also E-mail from J. Pennewell to L. Spencer, Jonathan Boyles, Paul Salfi, et al., dated Dec. 2, 2002, Zantaz documents 792470, 792476, at Zantaz document 792476, at 4. 183
recover UPB [unpaid principal balance], then recover accrued interest, and then cover other expenses."(679) However, management did not implement this "waterfall accounting" until 2002. The delay may have been due to Howard's reluctance to implement the change. Howard described his concerns in Spencer's Annual Review for 2000: [M]y main concern on this issue is that we not get into a situation where, when losses start rising, our proposed policy results in our needing to add large amounts to our loss allowance in addition to covering our credit losses. Yet the proposal [Janet Pennewell] presented to me a few weeks ago would, in my view, have done exactly that.(680) Nearly two years later, in June 2002, the Controller's Office made another presentation to Howard stressing the "[n]eed to [c]hange" rather than merely proposing a change.(681) The following month, Boyles, then head of Financial Standards, described the issue in a "white paper" as follows: We do not believe our current method of accounting [for the Allowance] is the preferable method . . . . Under our current methodology we are increasing the [Allowance] by recording gains from the disposition of properties and reducing it through negative provisions.(682) Boyles continued in the "white paper" that the current methodology made it "increasingly difficult for us to produce analytical support for our current allowance balance."(683) Further, Boyles added: With the current public debate regarding whether or not we should register with the SEC we believe it has become - ---------- (679) Recommendation to Change Loss Accounting Methodology, dated Nov. 2000, FMSE-E 44694-714, at FMSE-E 44698. (680) Mem. from T. Howard to L. Spencer, dated Dec. 9, 2000, FMSE-IR 28779-81, at FMSE-IR 28780-81. (681) E-mail from J.M. Whitacre to A. Marzol, Brian Graham, and Laura Kenney, dated Aug. 19, 2002, FMSE-E 1091499-513, at FMSE-E 1091501. (682) E-mail from J. Boyles to L. Spencer, J. Pennewell, P. Salfi, and J.M. Whitacre, dated July 1, 2002, FMSE 413416-19, at FMSE 413417. (683) Id. at FMSE 413417-18. 184
imperative that we begin to put into place the analysis and support for the balance in our allowance for loan losses.(684) Plainly, during the period from 2000 through 2002, management recognized that the Company's methodology made it increasingly difficult to produce analytical support for the allowance balance in accordance with GAAP and needed to be changed. Yet, the person who appeared to have had the authority to do so, Howard, did not act upon the recommendation until late 2002. Boyles's concern in the "white paper" was consistent with concerns raised by KPMG. Mark Serock, who became the engagement partner for KPMG in 2001, stated that the Company's Allowance practices prior to 2002 were "not preferable." The issue, according to Serock, was the "recording [of] recoveries in excess of the unpaid principal balance written off," which the Company committed to change by year-end 2002. Serock said that KPMG had felt that Fannie Mae's Allowance during these years was adequate, but that Fannie Mae needed to improve its documentation that showed how management determined that the level of the Allowance was reasonable. Ultimately, management implemented its "waterfall accounting" at year-end 2002.(685) (2) Introduction of a Model into the Allowance Setting Process(686) In 2002, management began developing the Loss Allowance Model ("LAM") to estimate the allowance attributable to the single-family book of business. The timing and other evidence suggest that the decision was prompted by the Company's decision to become an SEC registrant and the need to comply with SAB 102.(687) - ---------- (684) Id. at FMSE 413418. (685) Mem. from P. Salfi to L. Spencer, dated Jan. 31, 2003, FMSE 26642-44; 2002 Form 10-K at 30; see also E-mail from J. Boyles to L. Spencer, J. Pennewell, P. Salfi, and J.M. Whitacre, dated July 1, 2002, FMSE 413416-19, at FMSE 413417-19; E-mail from J.M. Whitacre to A. Marzol, B. Graham, and L. Kenney, dated Aug. 19, 2002, FMSE-E 1091499-513. (686) While Controller's Office staff used various spreadsheets in its calculation of Allowance recommendations prior to 2002, it did not consider those spreadsheets to be a modeled approach. Allowance for Loan Losses (ALL) & Liability for MBS Guaranty - Briefing to OFHEO, dated June 14, 2005, FMSE-IR 438739-44, at FMSE-IR 438742-43; SF Allowance for Loan Losses - Status Update and Report on Interim Findings, dated Aug. 25, 2005, Zantaz document 2290112, at 4. (687) Credit Loss Accounting Update, dated June 2002, Zantaz document 272788, at 2 ("Must Implement in 2002 SEC Guidance SAB 102 . . . requires detailed analytical 185
The LAM was implemented during the fourth quarter of 2002.(688) However, management chose to retain discretion over the Allowance rather than follow the results of the model. That is, according to Adolfo Marzol, then head of Credit Policy, at the quarterly Chairman's Loss Review meeting, management decided that the difference between the previous quarter's Allowance and the amount the LAM calculated, which was approximately $100 million, would be described as the "unallocated portion" of the Allowance.(689) Richard DePetris, then Controller for Multifamily, confirmed that an unallocated amount was added to the LAM modeled amount for the multifamily portion of the Allowance in the fourth quarter of 2002 to arrive at a level consistent with the previous quarter.(690) We have found no evidence that management sought to substantiate the amount to be set aside for the "unallocated portion." (c) Policy in 2003 The process for setting the Allowance in 2003 continued to involve a series of quarterly loss review meetings, but now the Allowance was calculated using, in part, the results of the LAM. However, it was not until the third quarter of 2003 when a version of the model was determined by Fannie Mae and its internal auditors to be - ---------- support behind loan loss allowance methods and application, including procedures to adjust methodology to reduce any differences between estimated and actual losses that occur. Most companies adopted upon issuance in 2001: we need to comply by year end."); see also E-mail from J. Boyles to L. Spencer, J. Pennewell, P. Salfi, and J.M. Whitacre, dated July 1, 2002, FMSE 413416-19. (688) Although it had been implemented, Fannie Mae did not consider the LAM complete until later, in the third quarter of 2003. Allowance for Loan Losses Audit Report, dated Dec. 15, 2003, Zantaz document 2229258, at 1, 2; Single Family (SF) Allowances for Loan Losses (ALL) Process During Restatement Period 2001 - through 2004 mem., dated Aug. 1, 2005, Zantaz document 2228409. (689) See also Undated Proposed Loss Allowance Methodology: Model Allocated with Unallocated Judgment, FMSE-IR 549499-512, at FMSE-IR 549511. (690) Richard DePetris stated his belief that GAAP allows for an unallocated reserve up to twenty-five percent of the prior period's total balance. In support, DePetris cited SAB 102 and an AICPA exposure draft of a proposed Statement of Position ("SOP"), "Allowance for Credit Losses," but subsequently acknowledged that he could not recall ever reading either SAB 102 or the AICPA's proposed SOP. Whitacre stated that the LAM was seen as confirmation of management's judgment rather than as an analytical tool that would calculate the Allowance in place of judgment. 186
completed.(691) It appears that, throughout 2003, the LAM results continued to be discussed in the context of various loss statistics and a forecasted credit outlook, and not as the result that was modeled for setting the Allowance. Accordingly, notwithstanding the implementation of the LAM and the waterfall accounting, the Allowance was kept static by management at approximately $800 million throughout 2003. (d) Changes in 2004 In 2004, significant changes were made to the single-family portion of the allowance,(692) which resulted in a reduction to the total Allowance balance as shown in the following table:(693) TABLE 2. INCREASE/ DOLLARS IN MILLIONS 2003 2004 (DECREASE) ------------------- ------ ------ ---------- Analysis of Combined Allowance: Single-family Modeled $600.8 $370.1 ($230.7) Other Projected Risks 36.8 (41.9) (78.7) ------ ------ ------ Total Single-family Portion of Combined Allowance $637.6 $328.2 ($309.4) ------ ------ ------ Multifamily Specific $ 20.5 $ 16.3 ($4.2) Modeled 124.3 151.8 27.5 Other Projected Risks 14.5 14.1 (0.4) ------ ------ ------ - ---------- (691) Allowance for Loan Losses Audit Report, dated Dec. 15, 2003, Zantaz document 2229258, at 1, 2; Single Family (SF) Allowances for Loan Losses (ALL) Process During Restatement Period 2001 - through 2004, dated Aug. 1, 2005, Zantaz document 2228409. (692) A similar review of the Multifamily portion of the Allowance did not occur until later, in 2005. Allowance for Loan Losses (ALL) & Liability for MBS Guaranty - Briefing to OFHEO, dated June 14, 2005, FMSE-IR 438739-44; Mem. from Gregory Garner to David Hisey and Carol Connelly, dated Mar. 30, 2005, FMSE-IR 612149-54. (693) Q4 2004 Allowance for Loan Losses Summary, dated Jan. 26, 2005, FMSE-IR 283353. The "Specific" multifamily portion refers to loans individually evaluated by Fannie Mae that are considered impaired. Fannie Mae applies Financial Accounting Standard No. 114, Accounting by Creditors for Impairment of a Loan ("FAS 114"), to estimate the amount of impairment. See 2003 Form 10-K at 41; Allowance for Loan Losses (ALL) & Liability for MBS Guaranty - Briefing to OFHEO, dated June 14, 2005 at FMSE-IR 438742. 187
Total Multifamily Portion of Combined Allowance $159.3 $182.2 $ 22.9 ------ ------ ------- Total Combined Allowance Specific $ 20.5 $ 16.3 ($4.2) Modeled 725.1 521.9 (203.2) Other Projected Risks 51.3 (27.8) (79.1) ------ ------ ------- Total Combined Allowance $796.9 $510.4 ($286.5) ------ ------ ------- Management's reason for reducing the Allowance in 2004 by $286.5 million was "because prior estimates are determined to be beyond the range of reasonableness."(694) Among other things, the work done on the single-family portion of the Allowance in 2004, which resulted in the material reduction, included: improving the documentation for the Allowance in the first quarter of 2004;(695) "discovering" more information about lender make-whole recoveries in the second and third quarters of 2004, which were causing Fannie Mae's losses to be lower than forecasted; and reducing the number of years of historical severity data included in the LAM calculation in the fourth quarter of 2004.(696) Additionally, at Marzol's behest, Fannie Mae increased the number of defaults predicted by the LAM.(697) Based principally on the net effect of these changes, management reduced the total Allowance by approximately $231 million in the fourth quarter of 2004 alone. - ---------- (694) Based on documents reviewed, the single-family portion of the Allowance will be subject to restatement for 2001 through 2004. The multifamily portion of the Allowance does not appear to be included in the pending restatement. See Allowance for Loan Losses (ALL) & Liability for MBS Guaranty - Briefing to OFHEO, dated June 14, 2005, FMSE-IR 438739-44, at FMSE-IR 438742; SF Allowance for Loan Losses - Status Update and Report on Interim Findings, dated Aug. 25, 2005, Zantaz document 2290112. (695) In the first quarter of 2004, Shaun Ross took over responsibility for the LAM from Whitacre. (696) Mem. from S. Ross, Hemant Pradhan and Shelley Klein to File, dated May 4, 2005, Zantaz document 2291383, at 4. This change eliminated from the calculation the high severities Fannie Mae experienced in 1990 through 1992 and made the calculation more responsive to recent experiences. Allowance for Loan Losses (ALL) & Liability for MBS Guaranty - Briefing to OFHEO, dated June 14, 2005, FMSE-IR 438739-44, at FMSE-IR 438744. (697) See also E-mail from A. Marzol to S. Ross, dated Mar. 22, 2004, Zantaz document 107166, at 1. Allowance for Loan Losses, dated Jan. 25, 2005, FMSE-IR 283388-99, at FMSE-IR 283393. The increased number of defaults was derived from the Loss Forecast Model ("LFM"), which Fannie Mae considered to be a "more robust" econometric model created by Credit Policy. Id. 188
According to Ross, the changes in the first three quarters of 2004 were motivated in large part by the change in staffing and his attempt to improve the process. But the changes in the fourth quarter of 2004 were the result of observations that OFHEO had made in the fall of 2004, which were echoed by KPMG in November of 2004. Among other things, OFHEO observed that the Allowance did not "behave directionally consistent with the book of business" and questioned whether the Company was including in its severity estimate too much historical data.(698) Management later determined that the historical loss severity data dating back to 1990 used in the LAM was up to ten times higher than the Company's current experience.(699) In January 2005, under the direction of Robert J. Levin, Fannie Mae's Chief Business Officer and Interim CFO,(700) Fannie Mae determined that its Allowance had been set at too high a level. Interim Chief Risk Officer Marzol stated that he and other senior accounting employees returned to "first principles" concerning the Allowance in order to establish a supportable Allowance that the new Accounting Policy staff would accept under GAAP. The fourth quarter reduction was discussed with the Fannie Mae Audit Committee.(701) B. Evidence Reflecting Management's View of the Allowance As we stated at the outset, we have found no evidence showing that management used the Allowance to offset unrelated expenses or to manage earnings to meet EPS targets in a given period. However, documents showed management's awareness that the Allowance may be overstated, or not in accordance with GAAP, and showed that certain members of management viewed the Allowance as a "war chest" for use in connection with unrelated or unexpected costs. 1. Management's Awareness that the Allowance May Be Overstated Documents showed awareness within management that the Allowance was set at a level in excess of credit loss experience. In Spencer's draft notes to Howard for a September 1997 presentation regarding the Accounting Policies for Foreclosed Assets,(702) - ---------- (698) See also Mem. from S. Ross, H. Pradhan, and S. Klein to File, dated Mar. 15, 2005, FMSE-IR 283444-47. (699) Allowance for Loan Losses (ALL) & Liability for MBS Guaranty - Briefing to OFHEO, dated June 14, 2005, FMSE-IR 438739-44, at FMSE-IR 438743. (700) Levin stated that once he learned the requirements, it was "clear to him" that Fannie Mae's Allowance was set at too high a level. (701) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Jan. 17, 2005, Zantaz document 555388, at 6-7. (702) Draft Notes to T. Howard, dated Sept. 24, 1997, FMSE-IR 338605-612. 189
she wrote regarding the implementation of FAS 114: "A Big Pro: It gives us a vehicle to draw down our loss allowance" and "Consider drawing down provision allowance?"(703) In the same document, under the heading "Prepare an IR Story," the document states: Likewise, we feel a more appropriate measure to assess the adequacy of our reserves would be to insure that we had an allowance sufficient to cover two years of TOTAL CREDIT LOSSES. Using this more conservative approach, we are still overprovided. As such, we will be bringing down our provision for a time until we approach our desired ratio.(704) In addition, in a December 5, 1997 memorandum from Quinn, then Senior Vice President--Credit Loss Management, to Howard, Levin, Ann Logan, and Sam Rajappa, Quinn wrote: At the end of 1999 when most of the portfolio is past its peak default period and if the credit picture continues to improve or stay stable, I believe we would be justified in booking a one-time reversal of the reserve of $300 million.(705) In addition, James A. Johnson, the CEO of Fannie Mae in 1998, recalled being aware of a dialogue about the Company being over-reserved in 1998 because of improvement in credit risk. Although Johnson said he thought the potential over-reserved amount was immaterial, he did vaguely recall KPMG suggesting that the Company reduce its loan loss reserve. Marzol, then head of Credit Policy, stated that he had opposed a reduction of the Allowance based solely on the Company's improved credit experience. Although Marzol disclaimed any participation in setting the Allowance prior to 2004, he nonetheless stated that he considered it overly simplistic for the Allowance to vary based on improved loss experience alone. Marzol believed that other factors, such as the Company's rapid book of business growth and increased participation in sub-prime business all should be considered when setting the Allowance. - ---------- (703) Id. at FMSE-IR 338607. (704) Id. at FMSE-IR 338611. (705) Mem. from M. Quinn to T. Howard, dated Dec. 5, 1997, FMSE-IR 338619-21, at FMSE-IR 338620. 190
As late as 2001, management considered reducing the Allowance and mitigating the one-time positive impact to income created by such a reduction through a finite risk insurance product. In April 2001, personnel from Credit Portfolio Strategies met with personnel from certain Marsh & McLennan companies ("MMC") to discuss the use of finite risk insurance to replace a portion of the Allowance. According to Robert Schaefer, then Director--Credit Portfolio Strategies, Brian Graham, who was then Senior Vice President--Credit Portfolio Strategies, outlined the following objectives for the MMC meeting: Controller's Office personnel were looking to reduce the Allowance, and if such a reduction were to occur, Graham wanted an insurance policy that would (1) allow the Company to expense the premium and thus offset the spike in income that would result from reducing the Allowance, and (2) have a high degree of certainty of "paying off" in future periods (i.e., that Fannie Mae was certain of recovering its premium payment through future claims reimbursements). Documents relating to these discussions with MMC show that management was considering reducing the Allowance by as much as $200 million at the time.(706) As discussed in further detail in Chapter VI, Part J, infra, Credit Portfolio Strategies discussed a number of potential structures with MMC to accomplish this objective, but ultimately, was unable to obtain Financial Standards' agreement that the structures would qualify for accounting as insurance arrangements.(707) 2. Evidence Reflecting Management's Consideration of Using the Allowance to Offset Unrelated Expenses Documents also showed that management considered using the excess amount in the Allowance to offset unrelated events. An example was Spencer's notes from a September 22, 1998 meeting described as a "Risk Review with CFO."(708) The Spencer notes suggested that in order to offset the impact of recording a large year-end catch-up expense, "We would source those earnings from a reduction to our loss reserves."(709) However, we saw no evidence that Spencer did, in fact, source the earnings from the Allowance. - ---------- (706) Handwritten Notes from Robert Schaefer, dated May 19, 2001, Zantaz document 2207005 ("$800M current reserve take it down to $600 . . . transfer $200 m to XL . . . for $300 m in coverage"). (707) Schaefer stated that Financial Standards' stated that the proposal would not receive insurance accounting because the timing of cash flows were set by the contract. KPMG workpapers reflect that it opined that the transaction would not qualify for insurance accounting. Mem. from Harry Argires to Fannie Mae File, dated Mar. 2002. (708) Risk Review with CFO, dated Sept. 22, 1998, FMSE-IR 321564-68. (709) See discussion supra Chapter IV. 191
Another example was the January 7, 1999 earnings alternatives schedules that were prepared under Spencer's direction.(710) The "Alternative III" scenario from these schedules reflected that management considered offsetting the large amount of negative catch-up expense that would be recorded for 1998 by taking $100 million from the provision for losses.(711) Management ultimately did not choose this alternative.(712) In yet another example, Spencer wrote to Howard concerning a 1997 presentation regarding the Accounting Policies for Foreclosed Assets: "Consider drawing down provision allowance? Over the planning horizon, we would have 'war chest' to draw down if necessary. Could do a one time draw, a big event, to cover ideas as restructuring the portfolio, infusing capital into Foundation, settling a tax event (if one occurred), other ideas . . . . This would be linked to an even[t] that we would publicly talk about and the analysts would view positively."(713) In addition to acknowledging an excess Allowance, these notes suggested Spencer's belief of the Allowance as a "war chest." Nevertheless, we saw no evidence that Spencer did draw down the Allowance at this time for the stated purpose. Finally, in what appears to be a September 2000 PowerPoint presentation - which was found in Ross's files and identified then Controller Spencer as the presenter - the "Loss Allowance" was listed as an item under "Possible Ways to Fill the Shortfall" in the EPS. Another slide in the same presentation, titled "Loss Allowance," read: "Potential to draw down a portion of the allowance for loan losses (currently at $809 million), resulting in income recognition."(714) Again, notwithstanding this document, our investigation has not uncovered any instances of an actual use of the Allowance to offset unrelated expenses or to fill earnings shortfalls in a given period. 3. Management's Safety and Soundness Perspective on the Allowance Pennewell stated that some members of senior management, particularly Howard, agreed with many bank regulators that it was prudent to build up the loan loss reserves in case credit performance deteriorated. Indeed, we note that in the 1990s there - ---------- (710) See id. (711) The fourth quarter 1998 "Adjustments" for "Net interest income" and "Guaranty fees" were negative amounts, or reductions of $280 million and $60 million, respectively. The sum of these contemplated adjustments totals $340 million. Fannie Mae 1999 Income Statement schedules, dated Jan. 7, 1999, FMSE-IR 21476-79, at FMSE-IR 21479. (712) See discussion supra Chapter IV. (713) Draft Notes to T. Howard, dated Sept. 24, 1997, FMSE-IR 338605-612, at FMSE-IR 338607. (714) Getting Back on Track Agenda, dated Sept. 2000, Zantaz 1512665, at 25. 192
was a general debate among banking regulators, the SEC, and the accounting profession about the proper level of a loan loss reserve. The bank and thrift failures of the 1980s and early 1990s influenced the bank regulators' "safety and soundness" view, which emphasized a conservative approach towards recording loan loss reserves at levels management can reasonably justify within the confines of GAAP. This view was perceived by many in the industry to be in conflict with the SEC's primary mandate of investor protection and concerns about earnings management. This led to the November 1998 press release announcing the formation of the Joint Interagency project to "better ensure the consistent application of loan loss accounting policy and to improve the transparency of financial statements,"(715) and resulted in congressional hearings and a number of speeches, testimony, and articles by the SEC, the regulators, and the accounting profession expressing their positions on loan loss accounting. The agencies agreed to the following aspects of loan loss allowance practices in the July 1999 Joint Interagency Letter to Financial Institutions: - Arriving at an appropriate allowance involves a high degree of management judgment and results in a range of estimated losses; - Prudent, conservative, but not excessive, loan loss allowances that fall within an acceptable range of estimated losses are appropriate. In accordance with GAAP, an institution should record its best estimate within the range of credit losses, including when management's best estimate is at the high end of the range; - Determining the allowance for loan losses is inevitably imprecise, and an appropriate allowance falls within a range of estimated losses; - An "unallocated" loan loss allowance is appropriate when it reflects an estimate of probable losses, determined in accordance with GAAP, and is properly supported; - Allowance estimates should be based on a comprehensive, well-documented, and consistently applied analysis of the loan portfolio; and - The loan loss allowance should take into consideration all available information existing as of the financial statement date, - ---------- (715) November 24, 1998 Joint Interagency Statement by SEC, Federal Deposit Insurance Corporation ("FDIC"), Federal Reserve Board, Office of the Comptroller of the Currency ("OCC"), and Office of Thrift Supervision. 193
including environmental factors such as industry, geographical, economic, and political factors.(716) Then, in July 2001, the SEC issued SAB 102, which provided further guidance for determining and documenting the level of an allowance for loan losses in accordance with GAAP. Certain members of Fannie Mae management may have believed that the "safety and soundness view" or conservatism "trumped" a literal application of GAAP in this area.(717) Because Howard declined our interview requests, we have not confirmed whether Howard subscribed to this view, as suggested by witnesses and documents. Boyles acknowledged such a point of view in his 2002 white paper, when he wrote, "[w]e recognize that there is a natural tension between the need for an allowance large enough to ensure safety and soundness and the limiting nature of GAAP. We will keep this in mind and will try to balance the two as we move forward during the year putting the framework in place."(718) Additionally, OFHEO reviewed but did not criticize the Allowance methodology or balance from 2000 through 2003. C. Transparency of the Allowance Setting Process to OFHEO Since 2000, OFHEO examined Fannie Mae's Allowance (referred to by OFHEO as the "Loan Loss Allowance") on at least three occasions, issuing reports on December 31, 2001 (the "2001 review"), December 8, 2003 (the "2003 review"), and March 30, 2005 (the "2004 review").(719) Although OFHEO did not complete an examination in 2002, the Company did brief OFHEO about the accounting and methodology changes that occurred during that year.(720) - ---------- (716) Joint Interagency Letter to Financial Institutions by the SEC, FDIC, Federal Reserve Board, OCC, and the Office of Thrift Supervision, dated July 12, 1999. (717) According to Pennewell, Howard believed it was prudent to build up loan loss reserves in conjunction with the safety and soundness view. Pennewell, however, shared the view expressed by the SEC and AICPA that an allowance should move in tandem with credit losses. (718) E-mail from J. Boyles to L. Spencer, J. Pennewell, P. Salfi, J.M. Whitacre, dated July 1, 2002, FMSE 413416-19, at FMSE 413419. (719) OFHEO Examination Worksheet Observation and Follow-Up, dated Dec. 31, 2001, Zantaz document 1282365; OFHEO Examination Worksheet Observation and Follow-Up, dated Dec. 8, 2003, Zantaz document 1282369; Mem. from G. Grarner to D. Hisey and C. Connelly, dated Mar. 30, 2005, FMSE-IR 612149-54 at FMSE-IR 612150. (720) OFHEO Briefing: Loss Allowance, dated Dec. 31, 2002, Zantaz document 3543416. 194
In the 2001 review, although opportunities for improvement were cited, OFHEO noted that Fannie Mae's allowance process was effective and that "Management complies with accounting standards and Fannie Mae's allowance policy."(721) Once again, in the 2003 review, OFHEO noted that Fannie Mae's reserve determination process was acceptable and that "[t]he allowance determination process is in line with professional standards."(722) OFHEO concluded in both of these reviews that the Company's processes "exceed[ed] safety and soundness standards."(723) In August 2004, OFHEO began its exam of Fannie Mae's Allowance. As already discussed, Fannie Mae made changes during the fourth quarter of 2004 that were meant to address OFHEO's comments. In January 2005, Fannie Mae made a presentation to OFHEO describing how its methodology changes addressed OFHEO's commentary regarding directional consistency, responsiveness, back testing and third party review.(724) Also included in this presentation was a response to KPMG's comments regarding severity and defaults, as well as the impact on the Company's Allowance at year-end 2004 as a result of its methodology changes.(725) Thereafter, on March 9, 2005, OFHEO issued a report related to a Sarbanes-Oxley examination, which identified internal control deficiencies and "poor methodologies for determining the loan loss allowance levels."(726) OFHEO issued its 2004 review, but lowered its overall rating from "exceeds" safety and soundness standards to "meets."(727) - ---------- (721) OFHEO Examination Worksheet Observation and Follow-Up, dated Dec. 31, 2001, Zantaz document 1282365, at 4. (722) OFHEO Examination Worksheet Observation and Follow-Up, dated Dec. 8, 2003, Zantaz document 1282369, at 8. (723) OFHEO Examination Worksheet Observation and Follow-Up, dated Dec. 31, 2001, Zantaz document 1282365, at 6; OFHEO Examination Worksheet Observation and Follow-Up, dated Dec. 8, 2003, Zantaz document 1282369, at 1. (724) Allowance for Loan Losses Discussion with OFHEO, dated Jan. 28, 2005, Zantaz document 1651744, at 5. (725) Id. at 7-10. (726) Mem. from Daniel Berkland to A. Marzol, dated Mar. 9, 2005, Zantaz document 2291541, at 3. (727) Misc. Communication to an Enterprise, dated Mar. 30, 2005, Zantaz document 1931394, at 1. 195
D. Conclusions Regarding Fannie Mae's Accounting for the Allowance The Company did not maintain a consistent or an appropriate accounting policy concerning the Allowance, and also failed to develop in a timely manner a methodology to support the Allowance balance under GAAP. We have not seen any documented analysis of the Allowance to support the accounting decisions to calculate an Allowance that approximates two to three times trailing charge-offs pre-1997, or to support a change in methodology to the provision for losses equal to charge-offs beginning in 1997.(728) In spite of the acknowledgement of the need to as early as 1998, Fannie Mae apparently did not begin working on a detailed Allowance model until fiscal year 2002. And, it was not until the third quarter of 2003 when a version of the model was determined by Fannie Mae and its internal auditors to be completed.(729) The development of the model appears to have been driven by Fannie Mae's voluntary decision to become an SEC registrant in 2002. Beyond that, however, there does not appear to have been any urgency in bringing the Company's accounting methodology for the Allowance in compliance with GAAP.(730) Also, Boyles's white paper suggested that the Company was aware that it was not in compliance and had not made much progress in complying with SAB 102 as of that time.(731) Documents showed that management was either complacent or ignorant - ---------- (728) According to Pennewell, Whitacre, and contemporaneous documents, Howard exercised decision-making authority over the Allowance methodology from 1998 onward, apparently until he left the Company. See, e.g., Year 2000 Performance Evaluation, from T. Howard to L. Spencer, dated Dec. 9, 2000, FMSE-IR 28779-81; Mem. from K. Russell to J. Theobald and E. Smith, dated Aug. 25, 1998; E-mail from J.M. Whitacre to A. Marzol, B. Graham, L. Kenney, dated Aug. 19, 2002, FMSE-E 1091499-513, at FMSE-E 1091499. Further, we have been unable to interview Howard about his rationale for these accounting judgments. Similarly, Spencer declined further interviews after the discovery of a file of documents in which, among other things, she described using the Allowance to offset a catch-up adjustment in a note to Howard for a Sept. 22, 1998 "Risk Review Meeting with CFO" meeting. Buy-up File, FMSE-IR 321541-78, at FMSE-IR 321567. (729) Allowance for Loan Losses Audit - Audit Report, dated Dec. 15, 2003. (730) We note that Fannie Mae began developing the LAM more than three years after KPMG first requested that the Company do so. Mem. from K. Russell to J. Theobald and E. Smith, dated August 25, 1998. (731) E-mail from J. Boyles to L. Spencer, J. Pennewell, P. Salfi, and J.M. Whitacre, dated July 1, 2002, FMSE 413416-19, at FMSE 413417-18. 196
about the need to comply with GAAP. For instance, a document from Mary Lewers's files, titled "Q2 2003 Allowance for Loan Losses Summary," contained the following handwritten notations on the margins about the unallocated portion of the Allowance: "Why? There is a permissible range - but is there a floor or a ceiling on the unallocated? W/O the better modeling, we are reluctant to [change] the 808 [total allowance balance] - but we are working toward the better model by the end of 2003."(732) Even after implementing a model as of 2003, based on the fact that the Allowance remained constant until 2004, it does not appear that the changes in methodology or policy seriously affected management's decision-making until 2004. Moreover, several employees - namely, DePetris, Marzol, and Whitacre - stated their belief that it was permissible to have an "unallocated" portion in the Allowance, which required no further analysis or support so long as the amount did not exceed a certain percentage of the total Allowance amount.(733) There was no support for such belief in the authoritative literature, and these employees have not pointed to any.(734) In summary, Fannie Mae's policy prior to 2004, which was based on a provision for losses equal to charge-offs, was not consistent with GAAP and Fannie Mae's methodology did not adequately support the Allowance. PART B: ACCOUNTING FOR DOLLAR ROLLS I. INTRODUCTION The February 11, 2005 OFHEO Letter identified concerns about Fannie Mae's accounting treatment for transactions known as dollar rolls.(735) In light of OFHEO's concerns, we undertook a review of Fannie Mae's dollar roll policies and procedures. In a typical dollar roll transaction, Fannie Mae borrowed funds from a counterparty for a specified period of time, using a security from its portfolio as collateral. From an execution perspective, the borrowing was effected by Fannie Mae's "selling" the collateral and simultaneously entering into an agreement to repurchase a similar security at a future date. Assuming that certain criteria in the accounting standards were met, Fannie Mae would account for the simultaneous sale/repurchase - ---------- (732) Undated Key Allowance Accounting Concepts, FMSE 60753-58, at FMSE 60755. (733) Loss Allowance Methodology, dated November 2002, Zantaz document 2225939, at 3 ("KPMG - SEC rule of thumb that unallocated allowance cannot be more than 26% of total allowance"); Loss Accounting Change and New Loss Allowance Methodology, dated March 18, 2003, FMSE-IR 549627-41, at FMSE-IR 549636. (734) See Supplementary Information - Background to SAB 102, dated July 6, 2001. (735) See February 11, 2005 OFHEO Letter at FMSE-IR 547320-22. 197
arrangement as a financing (i.e., a short-term loan) rather than as a sale of the collateral and a purchase of a new security. If the counterparty did not meet its commitment to deliver a similar security at the maturity date, the dollar roll would be treated as a "fail." If the Company then could not obtain a similar security in the market, the "fail" would result in accounting for the transfer of the collateral as a sale. Dollar rolls offered Fannie Mae a ready source of short-term liquidity, similar to its benchmark bills or discount notes. Execution of a dollar roll that did not comply with accounting requirements, or that resulted in a fail for which the Company could not purchase a substantially similar security, however, had two potential consequences for the Company. First, Fannie Mae would have to account for the transfer of collateral as a sale and the receipt of the security as a repurchase, with consequent recognition of gain or loss. Second, because Fannie Mae used securities from its held-to-maturity ("HTM") portfolio as collateral, treatment of the transfer of the collateral as a sale would have resulted in the tainting of the Company's HTM portfolio.(736) We conclude that Fannie Mae's accounting for dollar roll transactions did not follow GAAP for a significant portion of the time period covered by this Report. The Company's policies did not address all of the requirements set forth in the accounting standards and there were significant gaps in the Company's systems for assessing whether treatment of the transactions as financings was appropriate. Coordination among the offices responsible for dollar roll transactions - particularly Financial Standards in the Controller's Office, the Securities Trading Operations ("STO") group in the Treasurer's Office, and Portfolio - was weak.(737) Based on all of the evidence, we further conclude that the failure to follow GAAP in this instance was not intentional or motivated by an effort to achieve a pre-determined financial outcome. Rather, this situation stemmed from a lack of rigor in Fannie Mae's accounting policies and systems, including a failure to communicate clearly what each group's responsibilities were with respect to these transactions. In our view, this circumstance reflects primarily on the senior management in the Controller's Office, including Leanne G. Spencer, her predecessors, and those in charge of Financial Standards. - ---------- (736) Chapter VI, Part D of this Report discusses the designation of securities as HTM and the consequences of a subsequent sale (or change of designation) of that security. (737) We did not determine whether individual transactions were or were not properly accounted for as financings. We understand that this issue is being addressed as part of the restatement process. 198
We have seen no evidence that KPMG undertook a review of the Company's dollar roll policies or practices. According to KPMG, it did not view dollar rolls as a significant aspect of Fannie Mae's business.(738) II. BACKGROUND A. FAS 140 and Other Accounting Literature FAS 140 addresses accounting for the transfer of financial assets.(739) Under FAS 140, a simultaneous sale/repurchase agreement may qualify as a financing if the transaction satisfies all of the following conditions: 1. The assets to be repurchased or redeemed are the same or substantially the same as those transferred. 2. The transferor is able to repurchase or redeem them on substantially the agreed terms, even in the event of default by the transferee. 3. The agreement is to repurchase or redeem them before maturity, at a fixed or determinable price. 4. The agreement is entered into concurrently with the transfer.(740) In order to satisfy the requirement that the transferor be able to redeem the security (subparagraph (b)), FAS 140 requires that the transferor (borrower) "must at all times during the contract term have obtained cash or other collateral sufficient to fund substantially all of the cost of purchasing replacement assets from others."(741) In effect, this provision requires that either the transferor establish a mechanism to monitor the value of the collateral or that there be market practices that function to ensure the adequacy of collateral. - ---------- (738) Based on data contained in the Company's general ledger accounts available on its Homesite system, Huron determined that the average end-of-month dollar roll balance in 2003 was $2.8 billion. In October 2003, the balance reached $7.8 billion. See Homesite General Ledger for account 211021. (739) See ACCOUNTING FOR TRANSFERS AND SERVICING OF FINANCIAL ASSETS AND EXTINGUISHMENTS OF LIABILITIES (A REPLACEMENT OF FAS STATEMENT NO. 125), Statement of Fin. Accounting Standards No. 140 (Fin. Accounting Standards Bd. 2000) (hereinafter "FAS 140"). (740) See id. P 47. (741) See id. P 49. 199
If the parties' agreement does not require the delivery of the same security to the transferor, then the returned security must be "substantially the same" as the transferred security.(742) FAS 140 provides that the collateral returned is "substantially the same" as the transferred collateral if the securities share all of the following characteristics: 1. The same primary obligor (except for debt guaranteed by a sovereign government, central bank, government-sponsored enterprise or agency thereof, in which case the guarantor and the terms of the guarantee must be the same); 2. Identical form and type so as to provide the same risks and rights; 3. The same maturity (or in the case of mortgage-backed pass-through and pay-through securities, similar remaining weighted-average maturities that result in approximately the same market yield); 4. Identical contractual interest rates; 5. Similar assets as collateral; [and] 6. The same aggregate unpaid principal amount or principal amounts within accepted "good delivery" standards for the type of security involved.(743) FAS 140 does not address the duration of a dollar roll, except to note that financing treatment would not be appropriate if the transferor (borrower) does not have the right to "repurchase or redeem [the collateral] before maturity."(744) Nothing in the standard precludes extension of the dollar roll during the term of the collateral prior to maturity, so long as the other requirements of the standard are satisfied. Because FAS 140 permits the transaction to be treated as a financing rather than a sale, nothing in the literature precludes the use of HTM securities as dollar roll collateral. However, as noted above, by using HTM securities as collateral, the borrower risks a taint of its portfolio if a transaction is treated as a sale. FAS 140 was not the first accounting standard to address dollar rolls. SOP 90-3, which is referenced in FAS 140, originally set forth the requirement that the returned collateral be substantially the same as the transferred collateral, and described - ---------- (742) See id. P 47. (743) See id. P 48. (744) See id. P 47(c). 200
the parameters for making that assessment.(745) FAS 125, which was issued in June 1996, specified substantially-the-same test in the language that would be incorporated into FAS 140.(746) FAS 125 also introduced the requirement that the transferor "have obtained cash or other collateral sufficient to fund substantially all of the cost of purchasing replacement assets from others."(747) Thus, all of the critical requirements for treatment of dollar rolls as financings have been in place at least since 1996. B. OFHEO'S Concerns The February 11 OFHEO Letter questioned whether Fannie Mae was adequately testing to ensure that the security that Fannie Mae received at the end of a dollar roll was substantially the same as the security Fannie Mae transferred to the counterparty.(748) OFHEO expressed doubt that Fannie Mae was complying with the requirement that the weighted average maturity ("WAM") and yield of the returned collateral were similar to the securities that Fannie Mae "rolled out."(749) The agency also challenged the Company's practice of testing the WAM at the level of an overall trade, rather than at the individual security level; the Company's assumption that if the WAMs were similar, the yields would also be similar; and the Company's failure to perform WAM and yield tests for any period prior to 2003.(750) In addition, OFHEO raised a number of operational issues associated with Fannie Mae's treatment of dollar rolls. OFHEO stated that Fannie Mae did not appear to have a monitoring system in place to ensure that it was adequately collateralized at all times during the term of the repurchase agreement.(751) Finally, OFHEO questioned whether Fannie Mae's policy for dealing with dollar roll fails was overly flexible. According to OFHEO, Fannie Mae would not - ---------- (745) See id. P 48 n.18; see also DEFINITION OF THE TERM SUBSTANTIALLY THE SAME FOR HOLDERS OF DEBT INSTRUMENTS, AS USED IN CERTAIN AUDIT GUIDES AND A STATEMENT OF POSITION, Statement of Position No. 90-3 (Am. Inst. of Certified Pub. Accountants 1990). (746) See ACCOUNTING FOR TRANSFERS AND SERVICING OF FINANCIAL ASSETS AND EXTINGUISHMENTS OF LIABILITIES, Statement of Fin. Accounting Standards No. 125 (Fin. Accounting Standards Bd. 1996), at PP 27-28 (hereinafter "FAS 125"). (747) See id. P 28. (748) See February 11, 2005 OFHEO Letter at FMSE-IR 547321. (749) Id. (750) Id. (751) Id. 201
consider a dollar roll to have failed unless Fannie Mae had not received, or purchased in the marketplace, a substantially similar security within four months, whereas market convention requires that a securities trade be considered a fail if delivery is late by even one day.(752) OFHEO noted that, because Fannie Mae's dollar rolls involved securities classified as HTM, a failed dollar roll could have resulted in a tainting event that triggered the reclassification of all securities classified as HTM to available-for-sale ("AFS").(753) C. Fannie Mae's Accounting Policy Fannie Mae adopted an accounting policy for dollar rolls in the 1994 and 1996 versions of its Financial Accounting Policy Manuals.(754) These policies reflected the Company's interpretation of the criteria in SOP 90-3 for determining whether returned collateral was substantially the same as the transferred collateral. Fannie Mae did not revise this policy for, or otherwise incorporate into its internal accounting policies, the additional requirements in FAS 125 and FAS 140 regarding dollar rolls until 2003.(755) Specifically, the policy did not mention the need to monitor the collateral to ensure that the Company received cash or other collateral sufficient to buy it back if the lender failed to deliver. Thus, for several years after that requirement initially became effective, the Company's policy did not reflect the applicable accounting literature. Jonathan Boyles stated that revamping the Company's accounting policies generally (and not just with respect to dollar rolls) had been on his "to do" list, but other obligations had required that he put that effort off until 2003. The 2003 dollar roll policy generally tracked the language of FAS 140 insofar as the conditions for treating dollar rolls as financings are concerned. In language drawn from the standard, the Company's policy required (among other things) that "assets to be repurchased or redeemed are the same or 'substantially the same' as those transferred," and listed specific criteria for making that assessment.(756) It also specified - ---------- (752) Id. (753) Id. at FMSE-IR 547320-21. (754) Reverse Repurchase Agreements and Dollar Rolls, Financial Accounting Policy Manual, dated Oct. 20, 1994, FMSE-SP 79123-239, at FMSE-SP 79185-87 (hereinafter "1994 Policy"); Reverse Repurchase Agreements and Dollar Rolls, Financial Accounting Policy Manual, dated July, 1996, Zantaz document 420510, at 51-53 (hereinafter "1996 Policy"). (755) See Reverse Repurchase Agreements and Dollar Rolls, Financial Accounting Policy Guide, dated Oct. 2003, FMSE-SP 78953-9118, at FMSE-SP 79041-46 (hereinafter "2003 Policy"). (756) See 2003 Policy at FMSE-SP 79041-42. 202
that: "Fannie Mae, at all times during the contract term, must have obtained cash or other collateral sufficient to fund substantially all of the cost of repurchasing replacement assets from others."(757) Several features of the Company's policy apparently sought to fill in certain gaps or ambiguities in the authoritative literature. For example, FAS 140 is silent on the threshold WAM and yield for determining whether the returned security is "substantially the same." Fannie Mae's policy, in addition to requiring that the WAMs of the two securities "be similar," specified that the difference in the WAM of the transferred and returned collateral be within twenty-four months.(758) Likewise, whereas FAS 140 stated only that the returned collateral must have "approximately the same market yield," Fannie Mae's policy specified a difference in yield of up to ten basis points.(759) Although the standard did not provide specific guidance on whether these parameters should be evaluated on the basis of the entire bundle of collateral returned or on a security-by-security basis, Fannie Mae's policy required that the WAM and yield calculations be performed on a weighted-average basis on an entire trade, rather than on the individual securities within that trade.(760) In each of these instances, the 2003 policy draws on earlier Company policies without amendment.(761) As noted, the Company's 2003 policy also addressed "fails" - i.e., the counterparty's failure to deliver the required collateral on the contractual delivery date. The policy stated that, "[i]f the counterparty to a dollar roll transaction fails to redeliver the security(ies) within 3 months of the end of the original contract date, this transaction is considered to have failed."(762) Under the policy, Fannie Mae still would not account for the transaction as a sale, however, so long as, within one additional month, the Company could "use the cash proceeds originally received in the secured financing to repurchase a security from a third party that meets the 'substantially similar' criteria," or to "receive the collateral from the original counterparty."(763) If Fannie Mae was unable to acquire the security from the counterparty or the market within that four month span, then the Company was to record the gain or loss resulting from the sale.(764) - ---------- (757) Id. (758) Id. at FMSE-SP 79042. (759) Id. (760) Id. (761) See 1994 Policy at FMSE-SP 79185; 1996 Policy, Zantaz document 420510, at 51. (762) See 2003 Policy at FMSE-SP 79043. (763) Id. (764) Id. at FMSE-SP 79044. 203
Fannie Mae's policy stated, without explanation, that treating a dollar roll as a sale would not taint the Company's portfolio, even if the collateral involved was an HTM security.(765) The policy did prohibit securities with characteristics similar to those of securities involved in the fail from being used for dollar rolls in the future, however, on the ground that, if re-acquisition of appropriate collateral was impossible, then Fannie Mae was incorrect in its original assumption that a substantially similar security could be acquired from a third party.(766) Our questions regarding Fannie Mae's policy related to timing and substance. As to timing, we have received no clear explanation for why the Company's policy prior to 2003 did not reflect the requirement that the Company obtain and maintain adequate cash to repurchase collateral - seven years after that requirement appeared in FAS 125 and three years after it was incorporated in FAS 140. As to substance, there were several areas in which the 2003 Policy did not track the FAS 140 requirements. In particular, the 2003 Policy considered the securities to be substantially the same if their WAMs were within twenty-four months; while a twenty-four month WAM difference might not result in substantially different yields when comparing two securities with remaining WAMs of 340 and 316 months, the analysis might well yield different results if the mortgage pools had a short remaining term to maturity (that is, if the twenty-four months reflected a greater portion of the loans' remaining terms). Although the criteria in the Company's policy that the yield of the two securities be within ten basis points might have mitigated any concerns with the WAM threshold, the Company did not test the yields. Fannie Mae is not currently engaging in dollar roll transactions. Interviewees stated that this is a result of both changed market conditions and continuing uncertainty over Fannie Mae's dollar roll policy and procedures. - ---------- (765) Id. at FMSE-SP 79043. (766) See also Mem. from Jonathan Boyles to File, FMSE 54164-67, at FMSE 54166. We asked Boyles to explain why a sale resulting from a failed HTM dollar roll would not be considered a portfolio tainting event. Boyles stated that he would not consider a dollar roll fail to be an "intentional" sale from the HTM portfolio. Without an intent to sell out of HTM, Boyles contended, it made no sense to recognize a tainting event. Paul Salfi disagreed with this characterization. He reasoned that sales out of the HTM portfolio were permissible under FAS 115 only in limited circumstances, and, in his view, a failed dollar roll transaction would not qualify as a permissible HTM sale. Salfi questioned whether the "intent" of the sale made any difference for purposes of determining whether portfolio tainting had occurred. 204
D. Fannie Mae's Dollar Roll Transactions. 1. Mechanics of the Dollar Roll Transaction Functionally, dollar rolls were processed much like any other MBS trade. A Fannie Mae trader and the counterparty would agree to terms by which the counterparty would purchase a security from Fannie Mae's portfolio. At the same time, the trader and the counterparty would enter into a purchase agreement under which Fannie Mae would acquire a security that was substantially the same as the security "sold" from the counterparty at a specified future date. The decision to dollar roll a particular security in this fashion would be made by the trading desk personnel. The master agreement under which Fannie Mae and its counterparties structured a dollar roll, in fact, was the same agreement that applied to any other securities trade. The responsibility for carrying out the different tasks associated with a dollar roll transaction fell on different organizations within the Company. As noted, traders within the Treasurer's Office executed dollar roll transactions. STO monitored the settlement date for the return of the collateral and received the collateral back into the Company's portfolio. A group within Portfolio was responsible for maintaining an inventory of securities eligible for use as dollar roll collateral. The Company had no automated systems that were dedicated to testing the returned collateral in accordance with either the criteria in the authoritative accounting literature, or the Company's own policy on dollar rolls. 2. Dollar Roll Fails Although the Company's dollar roll accounting policy specifies a three-month grace period in which a counterparty is required to return adequate collateral and that the transaction will be treated as a "fail" one month later, the accounting literature does not impose such deadlines. These deadlines, therefore, appear to constitute internal guidelines, rather than requirements mandated by GAAP. Boyles stated that he included these provisions in the policy to impose discipline on the dollar roll process. He said that it would be rare or exceptional for Fannie Mae not to have received its security back eventually, and that, in any event, the assets that Fannie Mae used for dollar rolls were standard, highly liquid MBS, so a true fail was highly unlikely. Al Barbieri, the Vice President responsible for the STO, confirmed that he knew of no dollar roll transaction in which the Company ultimately did not reacquire collateral from the counterparty or purchase replacement securities from a third party. Barbieri described a process in which, at or around the Company's designated "fail" date, the STO would contact the counterparty in an effort to resolve the fail. If that was unsuccessful, the matter would be referred to the traders for further resolution. Moreover, we were told by employees that the STO was aware of the possibility that a fail could taint the Company's HTM portfolio, and that they were careful not to act in a way that could be construed to cause a taint. For example, Ilene 205
Topper stated that Fannie Mae employees were always trying to follow the policy of getting collateral back within ninety days and to avoid tainting the portfolio with a sale out of HTM. She said that individuals involved in the process were sensitized to the HTM and taint issues. In this respect, we do not share OFHEO's concern that the Company's dollar roll policy with respect to extensions was too lenient. The extension of a dollar roll beyond the contractual redelivery date does not preclude treatment of the transaction as a financing, so long as the transaction does not extend to the collateral's maturity date. Accordingly, we do not question this aspect of Fannie Mae's policy or its failure to account for transactions as sales solely because they were extended for more than three months beyond the contractual deadline. 3. Return of Substantially the Same Collateral (a) Implementation of the Company's Policy Even after the adoption of a new accounting policy in 2003, the Company did not have a system in place to ensure that the collateral returned at the end of the dollar roll was "substantially similar" to the collateral the Company had posted. None of our interviewees thought that his or her group was responsible for administering this policy, and there was substantial confusion regarding the need to assess returned collateral, the standards that would be applied, and the group responsible for performing that task. The lack of a mechanism to assess the collateral position during the term of the dollar roll raises concerns as to whether that collateral met the requirements of FAS 140. Typically, only the trading desk personnel at Fannie Mae were aware at the outset that these two transactions (the "sale" and the "repurchase") were related. From the perspective of Fannie Mae's securities tracking system ("STAMPS"), and from the perspective of the back office personnel responsible for trade verification and securities settlement, the two legs of the Company's dollar roll transactions were entirely independent. Although securities trade tickets had a code that identified the trade as a dollar roll, it appears that there was no indication (e.g., a flag, or some other signal) within STAMPS to indicate that the "sale" and "repurchase" were linked; the two transactions could be identified as being part of the same dollar roll only through the "deal number" assigned to the transaction, and only the trading desk personnel would have access to the records that would make this linkage decipherable. As Barbieri noted, STO - the office that received the returned collateral - would have no way of knowing that the transaction was part of a dollar roll as opposed to an outright purchase of the security. Nonetheless, except for Barbieri, all of our interviewees from Portfolio assumed that STO was responsible for this task.(767) The rationale for assigning this task to - ---------- (767) Financial Reporting also believed that Al Barbieri's group played the key role in testing returned collateral. See E-mail from J. Boyles to Janet L. Pennewell, Mary Lewers, and Leanne G. Spencer, dated Jul. 14, 2003, FMSE-E 24256-57. In that e- 206
STO was that this group was the first to process the collateral as it came to Fannie Mae from the counterparty. According to Andrew McCormick, for example, the first check would be to ensure that the counterparty made "good delivery" under Bond Market Association ("BMA") guidelines,(768) and then STO would review the collateral for compliance with the criteria for dollar rolls. Barbieri stated that his group evaluated the WAM of securities returned to the Company for a period of time; however, the task was performed manually, and only when and if it was requested. In fact, Barbieri stated that he had voiced his concern over his group's rejecting a trade based on an accounting standard rather than a market convention, since such criteria were not standard practice in the marketplace or part of the BMA guidelines. Barbieri indicated that he would feel uncomfortable telling a counterparty that their securities "fell out" of a trade because they failed a test that was not a part of normal "good delivery." Barbieri also said that it would have been more logical for Fannie Mae to use the BMA guidelines to test for the adequacy of collateral, rather than the WAM/yield tests reflected in Company policy.(769) While there was some dispute, then, over who was responsible for conducting the "substantially similar" test on returned collateral, it is clear that this task was not performed systematically. This weakness was acknowledged in the Company's Portfolio Transactions Guide: Portfolio Transactions management identified a control weakness surrounding the dollar rolls process. Per the Fannie Mae accounting standards, there should be controls in place to perform characteristics checks on the collateral that is returned back. Currently average WAM - ---------- mail, Lewers writes that "[r]egarding dollar rolls, STO (Al Barbieri's group) is manually controlling the redelivered pools to insure that the securities meet the dollar roll accounting guidance." Id. at FMSE-E 24256. Andrew McCormick suggested that the process was handled by a combination of people, including Stephen Bartolini, Lynda Maggio, and Matthew Douthit from his group, Barbieri from the Treasurer Department's back office, and either Pennewell or Ilene Topper from the Controller's Office. (768) The BMA establishes and publishes the market consensus regarding certain terms and parameters of transactions involving different types of securities, including Fannie Mae MBS (http://www.bondmarkets.com/). (769) Barbieri acknowledged, however, that his team was responsible for evaluating whether a dollar roll transaction extended past the ninety-day "fail" period. Barbieri said that this was a collaborative effort across department lines. According to Barbieri, the ninety-day analysis was a manual query within the STAMPS system used to capture that particular fail information. To Barbieri's knowledge, no systems enhancements were planned to automate that process. 207
yield change tests are not currently enforced by the back-office. Portfolio back-office included implementing functionality to perform these two tests into their 2004 budget. This weakness does not cause any material impact to balance sheet or earnings.(770) Even this information was not well known throughout the Company. Ilene Topper, who was responsible for securities accounting within the Controller's Office, said that she thought that the Company was performing a WAM test, but that she did not know for sure. She thought that it might have started in 2004. Matthew Douthit, Vice President--Portfolio Transactions stated that he was aware of this WAM/yield test requirement as early as 2002. As noted, Barbieri stated that his group would perform the WAM test on an as-requested basis. Because the Company's systems were designed to handle all MBS transactions generally, the systems focused on identification of incoming securities that did not fall within the BMA's "good delivery" guidelines. The Company viewed the BMA good delivery guidelines as "analogous" to the requirements of the "substantially the same" accounting standard.(771) According to analyses performed by the STO, these guidelines require that the following conditions be met: The agency name and program description (pool prefix or suffix) including short name and maturity (FNMA 30 yr) are the same. The same coupon. The same bid or offer price. - ---------- (770) Portfolio Transactions Policy and Procedures Guide, dated May 11, 2004, FMSE 219734-81, at FMSE 219768. With respect to the statement in the Portfolio Transactions Guide that the identified control weakness "does not cause any material impact to balance sheet or earnings," most interviewees were not familiar with this statement and did not know what analysis had been performed to support that conclusion. Among those who recalled the statement, there was concern as to whether Portfolio should be opining about something that was the responsibility of the Controller's Office to determine. McCormick believed that his group worked with STO to assist in some analysis on this point. He specifically remembered that the twenty-four-month WAM test in the Company's policy had been examined. McCormick said that, in his view, the passage in the transactions guide regarding materiality was intended more as an instruction that dollar rolls should not be done until the issue was resolved, rather than as a true opinion of materiality. (771) See The MBS TBA Market, Andrew McCormick, dated Sept. 10, 2002, FMSE 367272-90, at FMSE 367282. 208
The same settlement month.(772) On their face, these good delivery principles do not track all of the requirements of the accounting standard for treating the returned securities as substantially the same as the transferred collateral. Barbieri confirmed that he would not have viewed a WAM test as part of the BMA good delivery guidelines.(773) Among the interviewees, there appears to have been some general awareness that Fannie Mae's dollar roll procedures were inadequate. McCormick, for example, acknowledged that the policy added layers of analysis for which the Company never established adequate procedures. Topper said that she had concerns about the adequacy of the policy back in 2003, and she felt that this area needed tightening. Anthony Lloyd said that it was well understood that the dollar roll system needed improvement. In his view, the extra work that the inadequate system created was causing frustration for many. Douthit expressed the view that the BMA good delivery guidelines and market discipline were an adequate mechanism to ensure that counterparties did not deviate from standard practice. (b) Reactions to Fannie Mae's Dollar Roll Policy The question as to whether Fannie Mae was following FAS 140 requirements arose following the issuance of the Company's accounting policy in 2003. It appears that, around that time, the Company implemented an analysis of its prior transactions to determine whether they could be treated as financings. That effort, however, resulted in tension between Financial Standards and Portfolio, which was responsible for managing these transactions. STO analyzed the practical impact of Fannie Mae's policy on the Company's ability to conduct dollar roll transactions. In a series of undated documents captioned "Dollar Roll Validation by STO" that appear to have been successive iterations of a single analysis,(774) several consequences of the dollar roll policy were identified. The most significant of these appears to be that, "[w]hen we do a dollar roll, the street does not distinguish between a dollar roll and a standard buy/sell TBA. Therefore, the - ---------- (772) Undated Dollar Roll Validation by STO, FMSE 387142-43, at FMSE 387143 (hereinafter "Dollar Roll Validation"). (773) We have not undertaken an independent assessment of the BMA good delivery guidelines except to note the apparent lack of requirements that follow the FAS 140 standard, either in language or in substance. (774) None of the interviewees we spoke to could identify who had prepared the documents or for what specific purpose. Nonetheless, it seems likely that the documents were created around the time that Financial Standards reconsidered the dollar roll policy in 2004. 209
counterparty is going to only redeliver securities based upon standard TBA requirements as defined by the BMA."(775) Nonetheless, the STO analysis reported that Fannie Mae had been evaluating the WAM of returned collateral since January 2003, and that, in that time, no collateral received from a counterparty had a WAM difference in excess of the accounting policy's twenty-four month limitation.(776) The analysis stated that "[m]ost of the WAM are within a few months of each other" and that "[t]he requirement that the yield change must not be greater than plus or minus 10 basis points was decided to be futile because it is such a small range."(777) According to the analysis, the STO had conducted an analysis "to determine which inputs affect yield," and several factors - including the TBA assumptions, the yield curve, and the weighted average coupon - were determined not to be factors.(778) The analysis concluded that "WAM could affect yield if the change in WAM was greater than 15 months, but we have not seen this scenario at all this year - the difference in WAM has only been within a few months."(779) Finally, "this yield requirement has been determined to be fairly difficult to automate because of limitations in STAMPS [the Company's system for recording securities transactions] and it would be a major undertaking - most likely not worth the cost."(780) The STO document that we believe to be the latest version of this analysis ends with a number of recommendations concerning Fannie Mae's dollar roll practices going forward. Most significantly, the analysis recommended that the WAM and yield requirements not be used.(781) The recommendation pointed out that FAS 140 does not specify a particular WAM or yield - only that the WAMs be sufficiently similar to "result in approximately the same market yield."(782) The recommendation then stated that the securities trading desk had determined that a dramatic change in WAM would be required to have a significant impact on yield (e.g., ten basis points).(783) The STO analysis recommended that Fannie Mae continue to monitor WAM and "dk" (i.e., reject) - ---------- (775) Dollar Roll Validation at FMSE 387142-43. (776) Id. at FMSE 387143. (777) Id. (778) Id. (779) Id. (780) Id. (781) Id. at FMSE 387126. (782) Id. (783) Id. at FMSE 387127. 210
any pools "that will cause a change in WAM on the deal level to be greater than 15 months."(784) The analysis pointed out the problems that could be created by "dk'ing" securities that fall outside of this fifteen-month window. The STO stated that the TBA good delivery guidelines "do not take into account the same WAM and yield requirements that we have for dollar rolls."(785) As a result, counterparties may not accept Fannie Mae's rejection of collateral and might force Fannie Mae to accept the security.(786) If that were to occur, the document concluded, Fannie Mae would either have to swap the "bad" security with another one in Fannie Mae's portfolio, or the Company would have to sell the returned collateral and buy a new security as a replacement.(787) The origins of STO's assessment of the dollar roll policy are not clear. It does appear that the release of the 2003 accounting policy instigated a discussion between Financial Standards and Portfolio, later involving KPMG, as to whether it was adequate to assess dollar rolls against market conventions. There was substantial concern among people within Portfolio, however, that the proposed tightening of dollar roll requirements would effectively shut down the dollar roll program.(788) In a series of e-mails from September 2004, Douthit engaged Financial Standards staff concerning the "substantially the same" criteria contained in the dollar roll policy. Douthit noted that the accounting group was likely to "hear significant dissent from business folks if your decision rests primarily on our own policy interpretation of the accounting literature without giving due consideration to whether a significant number of accounting firms at banks permit BMA definition of 'substantially the same.'"(789) Douthit stated that, as of the date of that e-mail, Financial Standards had been unable to determine what industry practice was, and, in his view, this was critical in order to "challenge an otherwise impractical and conservative reading of the accounting rule."(790) - ---------- (784) Id. (785) Id. (786) Id. (787) Id. (788) See, e.g., E-mail from M. Douthit to Mukul Gupta, dated Sept. 17, 2004, FNMSEC-E 279958-59. (789) E-mail from M. Douthit to Gregory Ramsey, dated Sept. 1, 2004, PW-PR-700062 (hereinafter "Sept. 1 Douthit E-Mail"). (790) Id. Douthit stated that he had heard at one point that Financial Standards was considering requiring that the WAM values for the outgoing and incoming collateral be exact. 221
A later e-mail message suggests that KPMG subsequently provided industry practice information.(791) According to the e-mail, industry practice had an even stricter standard than Fannie Mae's.(792) Portfolio representatives stated that they were not aware of other parties' using this standard and that, if the Company adopted it, other parties might no longer be interested in engaging in dollar rolls with Fannie Mae.(793) The e-mail also identified a number of other areas in which Financial Standards was considering tightening the policy, including ensuring that dollar roll counterparties were also accounting for the transaction as financings, rather than as sales; researching whether margin calls would be necessary to account for changes in value of the "rolled" securities; and discussing the level (e.g., CUSIP or product/coupon level) at which a fail should be evaluated.(794) Of particular note on this list was the reference to margin calls; we have seen no indication prior to that communication that Fannie Mae acknowledged the requirement that it monitor its collateral during the dollar roll term. The e-mail concluded that "[t]here have been no known [dollar roll] violations, but since we are reviewing our policies and procedures, we might want to stop doing [dollar roll] transactions temporarily. Higher ups will send out a memo if this decision is made."(795) As noted, Fannie Mae discontinued its dollar roll program soon thereafter. In sum, despite the reference to a "substantially the same" test in Fannie Mae policies as early as 1994, the Company had no mechanism in place for ensuring that the policies were implemented. Fannie Mae did not undertake to consider those requirements of FAS 140 and its own accounting policy until it revised its accounting policy for other reasons in 2003. As a consequence of this failure, the Company never had an adequate means for assuring itself that each piece of returned collateral qualified under the standard, and that the transaction qualified as a financing rather than a sale.(796) - ---------- (791) See E-mail from L. Maggio to S. Bartolini and M. Douthtit, dated Sept. 13, 2004, PW-PR-700065-66. (792) Id. at PW-PR 700065. (793) Id. (794) Id. at PW-PR 700065-66. (795) Id. at PW-PR 700066. (796) The accounting literature does not address whether the "substantially the same" analysis should be conducted with respect to the entire trade or with respect to individual securities included in the collateral. We understand that, as part of its restatement, the Company is reviewing every aspect of its dollar roll procedure to ensure consistency with the relevant standards. 212
(c) Monitoring Collateral FAS 140, like FAS 125 before it, required that the Company at all times have obtained sufficient cash or other assets to acquire substantially similar assets.(797) As noted above, this required a mechanism to determine whether a change in the value of the collateral requires an adjustment to the transaction's terms. There is no evidence that the Company addressed this requirement by identifying a system or practice to ensure that the cash received in the transaction was adequate to fund the repurchase of replacement collateral in the cost the counterparty defaulted. Prior to 2003, the Company's accounting policy did not discuss the issue and the reference to the need to address this issue in the 2004 e-mail discussed above bolsters the conclusion that Fannie Mae did not ensure that it had an appropriate mechanism in place. Interviewees confirm this point. Douthit was not aware of a mechanism. Boyles pointed out that the requirement is specified in the 2003 policy, but he had no insight into how it was implemented.(798) Barbieri in STO was unaware of the requirement. We have seen no documentary reference to a mechanism for dealing with this issue. The question of whether the Company actually complied in part with this provision is more complex and, ultimately, not amenable to resolution as part of our review. In our interview with Barbieri, he noted that the Company has used the trade confirmation services of the Depository Trust & Clearing Corporation ("DTCC").(799) Although the particular DTCC services Fannie Mae used has changed over time, Barbieri reported that, at least beginning sometime after 2000, the DTCC services Fannie Mae used required that the Company and its counterparties make available to the DTCC assets sufficient to cover a party's inability to perform its obligations.(800) According to Barbieri, the DTCC contacted Fannie Mae periodically regarding adjustments to this "margin."(801) - ---------- (797) See FAS 125 P 29; FAS 140 P 49. (798) See 2003 Policy at FMSE-SP 79043. (799) Barbieri recalled that prior to about 2000 or so, the Company used only the DTCC's Electronic Pool Notification service, which involves the transmission of pool information. After that date, the Company used some of the DTCC's other services, such as trade comparison, netting and risk management. As a government-sponsored enterprise, Fannie Mae's transactions are cleared through the Federal Reserve Bank. (800) Id. (801) Id. 213
However, this requirement was not linked in any way to compliance with applicable accounting requirements or, before 2003, the Company's own policies.(802) III. FINDINGS We conclude that the Company's treatment of dollar rolls reflected a lack of specific attention to the accounting requirements established by SOP 90-3, FAS 125, and FAS 140, at least prior to 2003, and a lack of coordination between the various groups within the Company that had a role in the accounting for and implementation of the transactions. Even after Financial Standards issued a revised policy in 2003, it does not appear that the Company adopted procedures to ensure compliance with the requirements for treating the transactions as financings. The result was a lack of appropriate systems and understanding and agreement on what standards should be applied, no allocation of responsibility for applying relevant standards, and a failure to track the transactions adequately. However, we have seen no evidence that the Company's approach to dollar rolls stemmed from an intent to manipulate earnings or other financial results. Rather, the record viewed as a whole appears to confirm the conclusion that the issues raised with respect to the Company's dollar roll transactions are the result of a lack of appropriate systems and resources and, prior to the advent of potential fails in 2003, a lack of consideration of the practical issues surrounding dollar roll accounting and the potential consequences of the Company's failure to account for the transactions properly. PART C: ACCOUNTING FOR FORWARD COMMITMENTS I. INTRODUCTION In this Part, we address Fannie Mae's accounting for firm commitments to purchase or sell mortgage loans or MBS following the FASB's adoption of Statement of Financial Accounting No. 149.(803) FAS 149, which had an effective date of July 1, 2003, amended FAS 133(804) to clarify that firm commitments to purchase or sell mortgage loans and MBS should be treated as derivatives. Accordingly, FAS 149 required that firm commitments (like other derivatives covered by FAS 133) be recorded on the company's - ---------- (802) As noted above, we did not determine whether individual dollar rolls were properly accounted for as financings. That issue is being addressed as part of the Company's restatement. (803) AMENDMENT OF STATEMENT 133 ON DERIVATIVE INSTRUMENTS AND HEDGING ACTIVITY, Statement of Fin. Accounting Standards No. 149 (Fin. Accounting Standards Bd. 2003) (hereinafter "FAS 149"). (804) ACCOUNTING FOR DERIVATIVE INSTRUMENTS AND HEDGING ACTIVITIES, Statement of Fin. Accounting Standards No. 133 (Fin. Accounting Standards Bd. 1998) (hereinafter "FAS 133"). The Company's accounting under FAS 133 is considered in Chapter V of this Report. 214
balance sheet at fair value and subsequently marked to fair value at the end of each reporting period until the settlement date. Changes in the fair value of the commitments would be reflected in the Company's earnings unless the derivative qualified as part of a hedging relationship. Fannie Mae designated many of its firm commitments as hedges of the risk resulting from changes in the price of the mortgage loans or MBS the Company would acquire or deliver when the commitment settled. The rules that govern these hedge relationships are the same rules that govern the hedges of variable-rate debt. Consequently, the provisions of FAS 133 regarding the hedge of a forecasted or anticipated transaction apply, including the requirements that the forecasted transaction be identified in the hedge documentation with sufficient specificity to identify when the hedged transaction occurs and that the forecasted transaction be probable of occurring. The February 11, 2005 OFHEO Letter raised several issues regarding Fannie Mae's implementation of FAS 149.(805) First, OFHEO raised concerns regarding the Company's policy for assessing whether the forecasted transaction (that is, settlement of the commitment) was probable of occurring. Second, OFHEO questioned whether Fannie Mae's documentation of hedges involving commitments satisfied the requirements for hedge accounting. Third, OFHEO noted that "Fannie Mae was unable to fully implement [FAS] 149 in the allotted time, which resulted in changes in accounting for transactions subject to [FAS 149] after adoption."(806) OFHEO raised additional questions regarding the Company's application of other aspects of FAS 133 to hedged transactions involving commitments. In sum, OFHEO "question[ed] whether Fannie Mae has properly and consistently applied [FAS] 149 since it[s] adoption."(807) During our review, we considered the history of the Company's implementation of FAS 149 and the policies and procedures the Company adopted with regard to hedged transactions involving firm commitments. As to the history, the evidence we have compiled, including documents and interviews of those involved in the implementation of FAS 149, shows that the effort to implement the new standard stretched the Company's resources in both of the departments that were most immediately affected by it: Financial Standards and Portfolio. The resources were strained by a lack of systems and staffing to the point where it became difficult, if not impossible, for Fannie Mae to implement the standard correctly and in a timely fashion. Thus, Fannie Mae did not have a final accounting policy regarding FAS 149 in place until October 2003, nearly four months after the standard's effective date. Likewise, the Company's procedures to address several of the important issues raised by FAS 149 were not completed until months after the standard took effect. Hedge documentation was revised several times after the standard's effective date and, as late as mid-2004, the - ---------- (805) See February 11, 2005 OFHEO Letter at FMSE-IR 547324-26. (806) Id. (807) Id. 215
specifications for the systems necessary to account for the hedged transactions were still in the discussion stage. We conclude that the policies and procedures Fannie Mae adopted to implement FAS 149 were not consistent with GAAP. The record overall reveals that Fannie Mae's hedge documentation did not adequately identify the hedged forecasted transaction as FAS 133 requires. Most importantly, the Company's hedge documentation did not describe the hedged transaction with specificity to determine whether a transaction that occurred was the hedged transaction. For similar reasons, Company management did not and could not assess at the outset of the hedge whether the hedged transaction was probable of occurring. It does not appear that the Company was driven by an effort to affect the Company's financial statements (that is, to avoid earnings volatility) to the same extent as we have found with its adoption of FAS 133 (except insofar as the Company sought to treat its commitments as perfectly effective hedges). Based on the evidence compiled during our review, we conclude that the departures from GAAP were the result of three related factors: (1) the lack of advance preparation for the changes that FAS 149 required; (2) the incorrect assumption at the outset of the implementation that, with only minor exceptions, all commitments - including commitments that resulted in "fails to ourselves" - would be eligible for hedge accounting; and (3) the unexpected complexity involved in the application of FAS 149 to the variety of Fannie Mae's commitment and forward trade transactions. Weakness in the Company's implementation of FAS 149 became apparent when the first close of the Company's books after FAS 149's effective date resulted in a $1 billion error on the Company's balance sheet. Although the error was brought to the immediate attention of the Board, Leanne G. Spencer told the Board that the implementation process was well in hand.(808) The Board was not told, then or later, of the problems associated with the implementation effort, including the problems Spencer acknowledged she had in communications with her staff.(809) This omission was especially significant as the balance sheet error triggered special review by OFHEO of Fannie Mae's FAS 149 implementation process and its end-user accounting systems. OFHEO's criticism further strained already tense relations between Fannie Mae and its regulator at an especially important period at the outset of OFHEO's Special Examination. - ---------- (808) Mem. from Franklin D. Raines to the Board of Directors, dated Oct. 29, 2003, Zantaz document 3528788, at 2; Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Oct. 30, 2003, FMSE 10449-51, at FMSE 10450. (809) See, e.g., 2003 Performance Review, Jonathan Boyles, dated Feb. 25, 2004, FMSE-SP 102951-53 (hereinafter "Boyles 2003 Performance Review"). 216
II. BACKGROUND ON FAS 149 A. Commitments under FAS 133 Prior to FAS 149 FAS 133 defines a "firm commitment" as a binding agreement between unrelated parties that (1) specifies all significant terms, including the quantity to be exchanged, the price, and the timing of the transaction and (2) has "a disincentive for nonperformance that is sufficiently large to make performance probable."(810) In certain circumstances, a firm commitment met FAS 133's definition of a derivative even before the FAS 149 amendments took effect.(811) Such a commitment would be subject to FAS 133's accounting rules, including the requirement that the commitment be recorded at fair value, unless it qualified for one of FAS 133's scope exceptions. FAS 133 recognized from the outset that a commitment, like other derivatives, could be linked to another instrument in a hedged transaction. Moreover, the FASB staff concluded in Statement 133 Implementation Issue No. G2, dated March 31, 1999 (more than eighteen months before FAS 133's effective date), that a commitment could serve as both the hedging instrument and the hedged transaction in an "all-in-one hedge."(812) In an "all-in-one" hedge, a commitment is linked in a hedge relationship with the asset underlying the commitment: The commitment to purchase (or sell) the relevant asset on specified terms serves as a hedge of the amount that the purchaser (or seller) will pay (or receive) when the forward contract for that asset settles.(813) FAS 133, as it took effect on January 1, 2001, arguably excluded certain commitments from its scope. The standard provided that contracts for the purchase or - ---------- (810) FAS 133 P 540. A firm commitment is analogous to a futures contract that can settle net - that is, instead of delivering the asset in question (settling "gross"), the party can make a cash payment equal to the change in the value of the asset during the contract term. (811) FAS 133 defines a derivative as an instrument or contract (1) with one or more underlyings and notional amounts, (2) that requires no (or minimal) net investment, and (3) that requires or permits net settlement as opposed to delivery of the underlying asset. See FAS 133 P P 6-9. The definition of a derivative is discussed in Chapter V of this Report. (812) CASH FLOW HEDGES: HEDGED TRANSACTIONS THAT ARISE FROM GROSS SETTLEMENT OF A DERIVATIVE ("ALL-IN-ONE" HEDGES), Statement 133 Implementation Issue No. G2 (Fin. Accounting Standards Bd. 1999). (813) For example, if Fannie Mae enters into a commitment to purchase securities at par with a settlement date in thirty days, the commitment (if appropriately designated as a hedge of the security to be purchased) serves as a hedge of the fluctuations in the price of the security during the period between the commitment date and the date on which the forward contract settles. 217
sale of when-issued securities or other securities that do not yet exist are "excluded from the requirements of this Statement . . . only if (1) there is no other way to purchase or sell that security, [and] (2) delivery of that security will occur within the shortest period possible for that type of security."(814) Concisely stated, FAS 133 contained language to the effect that the standard did not apply to certain commitments to purchase or sell MBS on a "to-be-announced" ("TBA") basis. As clarified by Statement 133 Implementation Issue No. C13, FAS 133 also excluded commitments to acquire mortgage loans from its scope.(815) As originally released, Implementation Issue C13 provided an exemption for "loan commitments that relate to the origination or acquisition of mortgage loans."(816) Consequently, when Fannie Mae adopted its Derivatives Accounting Guidelines (hereinafter "DAG") in 2001, it did not treat commitments associated with TBA trades or the purchase of mortgage loans as derivatives.(817) Fannie Mae's interpretation was not unique, and the published materials surrounding the FASB's consideration and implementation of FAS 149 confirm that the DAG did not specify clearly the treatment of these commitments under FAS 133. B. Adoption of FAS 149 The FASB released an exposure draft of FAS 149 in May 2002.(818) The exposure draft modified FAS 133's exemption for commitments with respect to TBA securities trades. Under the terms of the exposure draft, if there existed the possibility that the commitment would settle net through the payment of a pair-off fee (i.e., a fee paid by a nonperforming party to make the other party whole based on changes in the fair value of the commitment), or if the entity did not prepare the proper documentation to show that the commitment would not settle net, then the commitment would fall within - ---------- (814) FAS 133 P 59a. (815) SCOPE EXCEPTION: WHEN A LOAN COMMITMENT IS INCLUDED IN THE SCOPE OF STATEMENT 133, Statement 133 Implementation Issue No. C13 (Fin. Accounting Standards Bd. 2000) (hereinafter "Implementation Issue C13"). (816) Id. (emphasis added). (817) Derivatives Accounting Guidelines Section II.4-.5, dated Jan. 2001, FMSE-IR 12825-13295, at FMSE-IR 12850-51. The DAG appears to rely on Implementation Issue C13 to justify the exemption for commitments involving both whole loans and TBA trades. (818) PROPOSED STATEMENT OF FINANCIAL ACCOUNTING STANDARDS AMENDMENT OF STATEMENT 133 ON DERIVATIVE INSTRUMENTS AND HEDGING ACTIVITIES, Exposure Draft R24(a) (Fin. Accounting Standards Bd. 2002). 218
the definition of a derivative and would be subject to accounting as such under FAS 133.(819) The initial exposure draft did not address commitments for the purchase of mortgage loans. After further discussion, however, on October 9, 2002, the FASB voted unanimously that "only loan commitments that relate to originating loans with potential borrowers" would qualify for the FAS 133 exemption.(820) As of that date, therefore, it was clear that all of Fannie Mae's forward commitments to purchase mortgage loans would have to be accounted for as derivatives. The implication of FAS 149 was that firm commitments for the purchase of mortgage loans and MBS would be required to comply with all of the FAS 133 requirements for hedge accounting, including those concerning documentation and hedges of forecasted transactions. The 2002 exposure draft stated that the new rules would apply to all commitments entered into as of April 1, 2003. On March 12, 2003, the FASB delayed the effective date by three months, to July 1, 2003.(821) III. BACKGROUND ON FANNIE MAE'S FIRM COMMITMENT PRACTICES In the ordinary course of its business, Fannie Mae purchases and sells mortgage loans and MBS. Many of these transactions entail at the outset a firm commitment by Fannie Mae to buy (or sell) the asset at a future date, followed by settlement of the forward contract and delivery of the underlying asset to (or by) Fannie Mae. - ---------- (819) The exposure draft narrowed the exemption by requiring that, for the exemption to apply, the parties to the firm commitment had to conclude and document that, with respect to each trade, "it is probable at inception and throughout the term of that individual contract that the contract will not settle net and will result in physical delivery of a security when it is issued." Id. Thus, FAS 149 still contains an exemption for commitments that are unlikely to settle net. Fannie Mae did not attempt to take advantage of this provision, instead treating its commitments as hedges. (820) FASB ACTION ALERT NO. 02-39, dated Oct. 16, 2002, http://www.fasb.org/action/aa101602.shtml (last visited Feb. 17, 2006). (821) See FASB DERIVATIVES IMPLEMENTATION GROUP SUMMARY OF MARCH 12, 2003 BOARD MEETING DISCUSSION ON THE PROPOSED AMENDMENT TO STATEMENT 133, http://www.fasb.org/derivatives/bdmtg031203.shtml (last visited Feb. 17, 2006); FASB ACTION ALERT NO. 03-11, (Fin. Accounting Standards Bd. 2003), http://www.fasb.org/action/aa031903.shtml (last visited Feb. 17, 2006). 219
A commitment to purchase or sell MBS might involve a "specified" trade or a TBA trade. In a specified trade, Fannie Mae agrees to purchase or sell an existing and identified security. In a TBA trade, the security does not yet exist and the composition of the underlying mortgage pool is unknown. Instead, the seller anticipates acquiring securities with certain features (for example, Fannie Mae MBS reflecting securitization of a pool of thirty-year, fixed-rate, six percent mortgages); the seller commits to sell, and the buyer commits to purchase, the to-be-created security at a specified price on a specified date. By entering into the commitment, the parties establish the price for and the nature of the asset, regardless of any market fluctuations between the commitment date and the settlement date.(822) A commitment to purchase a pool of mortgage loans operates in a similar fashion. The commitment specifies the parameters of the loans (interest rate, term, etc.) and the price of the pool. At the settlement date, the seller is obligated to deliver the loans or pay (or receive) a pair-off fee. Although Fannie Mae buys and sells MBS, it typically does not sell whole loans due to systems limitations. Often, the parties to a commitment have an alternative: Instead of delivering the asset, they may substitute a cash payment equal in amount to the difference between the price specified in the commitment and the market value of the asset on the settlement date.(823) This payment is referred to as a "pair-off" fee. These transactions often take on additional features that made implementation of FAS 149 more complex. For example, in an "as soon as pooled" ("ASAP") trade, the mortgage lender (seller) has the option to deliver a pool of loans to Fannie Mae in advance of the specified settlement date in exchange for an adjustment to the purchase price for the loans. Fannie Mae might choose to retain these loans in its portfolio or securitize the loans and sell the MBS. In certain cases, Fannie Mae would also assume the lender's forward commitment to sell the MBS created from the pool of loans to a third party. Another transaction that had implications for the Company's implementation of FAS 149 is referred to as a matched-terms buy-sell ("MTBS") trade. In an MTBS trade, Fannie Mae enters into simultaneous commitments to purchase a security from one party and to sell that security to another. Prior to November 2003, however, using an automated process called "sorting and sifting," Fannie Mae could compare the security it received under the "buy" commitment in the MTBS to similar securities the Company acquired from other parties. Based on that comparison, Fannie Mae might decide to keep a security it had originally designated to be used to satisfy an - ---------- (822) See generally Mem. from Paul Salfi to Distribution, dated Oct. 22, 2003, FMSE-IR 482061-82 (establishing Company's financial accounting policy for commitments). (823) These terms are set forth in a standard Master Securities Agreement between Fannie Mae and its counterparties. See Undated Master Securities Forward Transaction Agreement P 7, FMSE-IR 694602-10, at FMSE-IR 694604. 220
MTBS trade and satisfy its MTBS sell obligation instead with a security acquired from another party in settlement of another commitment. In other instances, if the seller failed to deliver the security that Fannie Mae was expecting to deliver under an MTBS trade, the Company typically would deliver a security that had been intended for its portfolio to satisfy the MTBS trade. Because Fannie Mae did not take the security into its portfolio, the Company dubbed that transaction as a "fail to ourselves."(824) Other transactions - such as pair-offs, cancel and corrects,(825) and TBA value swaps - also had implications for Fannie Mae's application of FAS 149. The important aspect of each of these transactions for present purposes is that the commitments did not settle at the time, on the terms, and/or with the counterparty the Company originally had anticipated. IV. FINDINGS REGARDING FANNIE MAE'S IMPLEMENTATION OF FAS 149 Fannie Mae decided early in 2003 that, following the issuance of FAS 149, it would implement a program to designate its firm commitments as hedges of the sale or purchase of TBA securities and mortgage loans.(826) Despite the release of an exposure draft proposing changes in the rules for TBA securities trades in May 2002, and the FASB's vote in October 2002 concerning commitments to acquire mortgage loans, the Company did not begin to prepare for the issuance of FAS 149 until early 2003. Paul Salfi, an accountant in Financial Standards who was involved in the Company's FAS 149 implementation efforts, stated that he raised with Jonathan Boyles the potential that FAS 149 could have a significant effect on Fannie Mae. Boyles either did not expect that FAS 149 would have a significant impact on Fannie Mae's operations, or, as we explain below, he was distracted by other matters. In either event, the implementation effort was delayed. - ---------- (824) The ability to sort-and-sift and deliver securities intended for the portfolio to third parties to fulfill the Company's forward sale commitments was viewed by personnel in Portfolio as providing significant economic value to Fannie Mae. When the ability to sort-and-sift was lost, Portfolio management attempted to quantify the lost economic value. See Undated Accounting Issues and Impact on Portfolio Business, FMSE-E 278831-42. (825) A "cancel and correct" occurs when the original trade is canceled and a new trade, with different terms, is executed. (826) E-mail from J. Boyles to Timothy Howard, dated June 23, 2003, FMSE-E 5639-41 ("FAS 149 amends FAS 133 and reverses the DIG issue regarding purchases held for investment and limits it to origination. As a result, we will start treating these contracts as derivatives effective July 1, 2003."). 221
A document captioned "Timeline describing the development of FAS 149 implementation policy" (the "Timeline") records that the initial meeting among the accountants in Financial Standards to discuss FAS 149 occurred on January 17, 2003.(827) An initial meeting with the managers in Portfolio to "discuss potential population of commitments impacted" occurred ten days later. Additional meetings over the next year, primarily involving Lynda Maggio in the Portfolio Middle Office and Salfi, but including others as necessary, addressed various issues surrounding the interpretation and implementation of FAS 149.(828) According to Salfi, by mid-2003, the FAS 149 implementation effort was occupying a significant portion of his time. According to Boyles and Salfi, implementation of FAS 149 was an iterative or "peel the onion" process. Financial Standards worked with Portfolio representatives to understand the Company's various commitment transactions and formulate the accounting for those transactions. This process took several months, especially as the Financial Standards and Portfolio teams considered more varied and complex transactions (such as ASAP and "fails to ourselves" transactions discussed above). A presentation dated October 31, 2003, listed the steps in the FAS 149 implementation process up to the Company's third-quarter earnings release in 2003 (the first release affected by the new standard) and confirmed Boyles's characterization of the Company's approach.(829) The five steps were: (1) "Identified relevant transactions and inventory states"; (2) "Isolated data and developed reports"; (3) "Long discovery period resulted in multiple amendments to data sets and reports requested from portfolio Middle Office"; (4) "Marked to market (M2M) transactions and inventories"; and (5) "Developed automated M2M proposal for whole loans in CORE [systems] release I."(830) Among the items left to be accomplished were: "Expedite review of transactions and inventory states where accounting guidance is pending"; and "Materiality guidance from Controller's office to prioritize list and work flow."(831) It is apparent that, at the outset, Financial Standards was unfamiliar with many of the commitment transactions about which it was required to provide accounting - ---------- (827) The document was prepared by Salfi in late 2003 in response to OFHEO's request for information concerning the FAS 149 implementation process. Salfi based the timeline on his diaries and records. See Undated timeline, Zantaz document 272705 (hereinafter "Timeline"). (828) E-mail from Lynda Maggio to P. Salfi et al., dated Sept. 24, 2003, FMSE-E 165732-37. (829) FAS 149 Checkpoint, dated Oct. 21, 2003, FMSE-E 38344-69, at FMSE-E 383451. (830) Id. at FMSE-E 383453. (831) Id. 222
advice. A presentation to OFHEO that Fannie Mae prepared in November 2003 regarding the implementation process confirmed that it took "several months to understand all the commitment types and to define [the standard's] scope[.]"(832) The presentation noted the standard's complexity and that the FASB provided "[l]ess than four months from issuance to implementation."(833) That iterative and protracted nature of the implementation process is reflected in other documents as well. In a September 2003 e-mail exchange among Salfi, Janet L. Pennewell, and Matthew Douthit in Portfolio, Pennewell listed the commitments that had yet to be addressed in order of priority.(834) Pennewell said that the commitments for which the Company was not yet prepared to implement FAS 149 would not be treated as hedges and would be marked to fair value through earnings.(835) Salfi concluded the exchange with a two-step plan in which certain aspects of the documentation would be resolved by October 1, with more complete documentation to be phased in thereafter. The record also shows that KPMG was brought into the FAS 149 implementation effort at a relatively late date. According to the Timeline, the first draft of a FAS 149 accounting policy was circulated internally on May 28, 2003, just thirty-three days before the standard's effective date, and KPMG did not receive a copy for review until June 20, just eleven days before FAS 149 became effective.(836) New iterations of the draft policy were prepared, and the Timeline shows a meeting on July 28, 2003, attended by KPMG and representatives from Controller's and Portfolio, "to agree upon accounting policy for FAS 149."(837) Spencer would later reflect in hindsight on the - ---------- (832) FAS 149 Adoption at Fannie Mae: OFHEO Review, dated Nov. 6, 2003, FMSE-IR 352360-87, at FMSE-IR 352365. (833) Id. The reference to a four-month implementation period underscores the extent to which the Company's effort to implement the standard was belated. Although it is correct that the FASB issued the standard in April with an effective date in July, the FASB had originally set the effective date as April 1 when it announced its intentions in 2002. Furthermore, with the FASB having completed its redeliberations on the exposure draft by mid-December 2002, it is clear that the Company had more than four months to react to the FASB's new positions on these matters. FASB ACTION ALERT NO. 02-48, http://www.fasb.org/action/aa122002.shtml (last visited Feb. 17, 2006). (834) E-mail from P. Salfi to Janet L. Pennewell and Matthew Douthit, dated Sept. 29, 2003, FMSE-E 1178554-56. (835) Id. (836) Timeline at 1, 2. According to the Timeline, the fact that KPMG had not reviewed the policy previously was raised at a meeting the following day. (837) Id. 223
need "to drive KPMG to full engagement and not allow them to derail you mid-stream as happened with FAS 149."(838) However, the responsibility for any delay in this regard must rest with the Company, which began the implementation process late and brought KPMG into the effort only a few weeks before the effective date. Several factors further complicated the Company's implementation of FAS 149. There is abundant evidence that, from late 2002 through 2003, Financial Standards was understaffed and preoccupied by other matters. In her evaluation of Boyles, Spencer acknowledged that, from an accounting perspective, 2003 was the "Perfect Storm."(839) She pointed to the adoption of FAS 149, the issuance of other accounting standards that had an impact on Fannie Mae, the Company's SEC registration and its first Form 10-K filing, and Freddie Mac's disclosure that it would restate its financial statements as a result of accounting errors.(840) Spencer noted that Financial Standards had confronted these matters "with a thinner staff than we both would have liked" and focused on the need to build up the Financial Standards staff in the future.(841) Douthit (Maggio's supervisor) remarked to his superior, Andrew McCormick, about the lack of adequate resources in Financial Standards at about this time. Maggio also believed that her Portfolio group, which was responsible for providing the fair-value marks for the commitments, lacked adequate staffing. As early as July 2, 2003, Maggio reported to her superiors in Portfolio and to Spencer that the requirements of FAS 149 "will require additional resources in the portfolio middle office group."(842) To justify her request for two additional full-time employees, Maggio outlined the new tasks that implementation of FAS 149 would require and emphasized that "each of these components represents a process that is unique due to source data systems, output format requirements, timing, or valuation technique."(843) Apparently, Maggio's staffing requests were not met effectively. In an e-mail dated November 20, 2003, she - ---------- (838) Boyles 2003 Performance Review. (839) Id. Apparently this characterization originated with Boyles. Spencer wrote: "Maybe your analogy of weathering the 'Perfect Storm' is apt as well." Id. (840) Id. A draft letter to OFHEO that the Company was preparing to describe its implementation of FAS 149, and that appears to reflect Spencer's input also noted that FAS 149 "hit at the all time peak of mortgage commitment activity with interest rates at 45 year lows and a record commitment pipeline of over $135 billion." Draft Letter to C. Dickerson, dated Mar. 25, 2004, Cataphora document i.h.185b92a6d2db24f5141927ab3d5f6d46, at 2. (841) Boyles 2003 Performance Review at FMSE-SP 102951. (842) Mem. from L. Maggio to Leanne G. Spencer, Andrew McCormick and M. Douthit, dated July 28, 2003, FMSE-E 1116570-74. (843) Id. 224
noted that one of the staff slots she requested had been filled by an existing employee who had not been relieved of his or her previous duties; the other slot was filled by hiring an employee from the Company's Office of Internal Audit who, according to Maggio, was not available to work with her group until January 2004.(844) Finally, the delays in implementation resulted in a delay in the development of the Portfolio systems to mark the commitments to fair value and the Financial Reporting systems to account for the hedged transactions. As late as June 2004, the Company still was drafting requirements specifications for the systems it would need to automate tracking and accounting for the commitment activity captured by FAS 149.(845) The 2004 version of the Company's Portfolio Transactions Procedures Guide states: The existing [FAS 149] process is not fully automated with edit checks to strengthen the review process. Many edit checks are automated, but some are still in spreadsheets. Although the process is not efficient, it is effective. . . . There is still some risk within the project. First, Accounting Policy and Legal are still deciding on when certain transactions need to be marked. Second, the multiple marking methods in the review process are not yet bullet proofed. Transfers of data and summary reports are vulnerable due to the multiple spreadsheets used to hold marks.(846) The Guide notes that system enhancements were planned for 2004.(847) There can be little doubt that the lack of advance preparation and adequate staffing affected the Company's ability to implement the accounting standard effectively. As Spencer later acknowledged, "FAS 149 knocked us off our horse which had begun to canter. We are only slowly building momentum."(848) - ---------- (844) E-mail from L. Maggio to P. Salfi, dated Nov. 22, 2003, FMSE-E 172269-71. (845) Requirements Specification, FAS 149: Fail Resolution Probability Report, Version 1.0, dated June 28, 2004, FMSE-SP 95094-100. (846) Portfolio Transactions Procedures Guide, Asset Execution Guidelines, dated 2004, FMSE 219765-74, at FMSE 219768. (847) Id. (848) Boyles 2003 Performance Review at FMSE-SP 102952. 225
V. FINDINGS REGARDING FANNIE MAE'S ACCOUNTING FOR FIRM COMMITMENTS A. The Company's FAS 149 Policy Financial Standards, as noted, circulated the first draft of a FAS 149 policy in May 2003.(849) After several iterations, the final version of the accounting policy was issued in October 2003 (the "Policy").(850) Under the Policy, most of the Company's commitments were presumed to be derivatives and therefore eligible to be used in a hedged transaction. These categories included commitments for the purchase of mortgage loans (including loans acquired through ASAP arrangements), the purchase or sale of when-issued or TBA securities, and the purchase or sale of specified MBS. The Policy excluded other categories of transactions, such as commitments to purchase or sell assets that the Company classified, or would classify, as trading assets under FAS 115, and "best efforts" commitments (that is, commitments that are not "firm").(851) The Policy appears to reflect the Company's intention to designate forward purchase or sale commitments as "all-in-one" cash flow hedges of changes in the market price of the forecasted purchase or sale of the mortgage or security underlying the commitment. As stated in the Policy, "[w]e will designate forward portfolio purchase/sale commitments as the hedging derivatives in cash flow hedges of the risk associated with the impact of changes in [the prices of the] mortgage assets underlying the commitments/derivatives since these mortgage assets will reside in the portfolio or come from the portfolio."(852) It also appears that the Company intended to use the matched-terms method of hedge accounting for most of its commitments.(853) According to the Policy, the critical terms of the commitments and the assets to be purchased or sold would be the - ---------- (849) Timeline at 1. (850) Mem. from P. Salfi to Distribution, dated Oct. 22, 2003, FMSE-IR 482061-82 (hereinafter "Policy"). (851) Id. at FMSE-IR 482067. The Policy also established a rule regarding commitments that did not close on the settlement date. According to this rule, a commitment would not qualify for hedge accounting at the point in time at which it became probable that the commitment "will not settle with similar assets within 60 days of the original settlement date and, therefore, the forecasted transaction is not probable of occurring." Id. (852) Id. at FMSE-IR 482062 (emphasis added). (853) Id. at FMSE-IR 482079. 226
same and "we will conclude that hedges are perfectly effective."(854) In most instances, these would be cash-flow hedges in which the commitment hedged any change in the price of the asset. According to the Policy, purchase and sale commitments associated with an MTBS trade were not to be linked in a hedge relationship because the changes in the fair value of the two commitments would offset each other in the Company's earnings. However, MTBS sale commitments associated with loans acquired under ASAP trades would be linked as fair-value hedges of loans acquired in ASAP trades; according to the Policy, the fair value of the loans and the commitment "should offset because the life of the hedges is typically very short."(855) That is, any change in the value of the pools would not necessarily offset the change in the value of a sale commitment entered into on an earlier date; however, the Company treated these differences as inconsequential.(856) The Policy also addressed circumstances in which a commitment would not settle. According to the Policy, for example, hedge accounting would be appropriate with respect to a transaction in which the counterparty pays a pair-off fee instead of delivering the assets, if Fannie Mae were to use the pair-off fee to defray the cost of purchasing similar securities in the open market.(857) Similarly, under the Policy, hedge accounting would be appropriate, and the Company would continue to apply hedge accounting in a fails to ourselves situation if Fannie Mae were to consummate the purchase commitment by taking a security from another seller into its portfolio.(858) The Policy supported these positions by treating the assets as "fungible."(859) - ---------- (854) Id. (855) Id. at FMSE-IR 482080. The Policy stated that the maximum weighted average time period between loan acquisition and the forward sale date for any month to that point in 2003 was sixteen days. (856) Id. ("So, we will conclude that the hedges are perfectly effective."). (857) Id. at FMSE-IR 482069. (858) Id. at FMSE-IR 482063, 68-69. (859) Id. at FMSE-IR 482069. Another section of the Policy on the same subject stated: "Our proposal is that TBA pools be treated as generic commodities that are interchangeable amongst each other." Id. at FMSE-IR 482074. In this respect, the Policy was internally inconsistent. The notion that the assets are interchangeable or fungible is belied by the sorting and sifting process itself, which is based on recognizing characteristics of an instrument that would make the instrument more appealing than other, similar instruments. As the Policy also noted, "we have a strategy to 'keep the best, sell the rest.'" Id. Further, sorting and sifting was inconsistent with an "all-in-one" approach to hedge accounting and made it difficult to identify whether the transaction that occurred was the hedged transaction 227
Based on the record compiled during our investigation, we conclude that the approach to hedge accounting for commitments reflected in the October 2003 accounting policy and other documents is inconsistent with GAAP in several respects, regardless of whether the Company's approach to hedge accounting was based on the "all-in-one" cash flow hedge accounting guidance in Implementation Issue G2. These inconsistencies center on the FAS 133 requirement that the Company document the hedged transaction at inception, "with sufficient specificity so that when a transaction occurs, it is clear whether that transaction is or is not the hedged transaction."(860) In addition, the Company's procedures for hedging firm commitments did not meet the FAS 133 requirement "[t]hat the occurrence of the forecasted transaction is probable."(861) We explain below why, in our view, Fannie Mae's application of hedge accounting to its firm commitments did not satisfy these requirements, with attention to both the development of the requirements and the processes the Company adopted. B. Documentation of Hedged Transactions The development of Fannie Mae's documentation for hedged transactions involving forward-trade commitments suffered from the same fits and starts that beset the effort overall. Management never developed a procedure that properly documented the link between a commitment (derivative) and the purchase or sale of the underlying asset (the hedged item). Initially, management sought to adopt a practical approach to FAS 133's documentation requirement in which it would compile a list of purchase and sale commitments at month's end. Those involved in the process apparently expected to rely on this approach up until the night before FAS 149's effective date. On June 30, 2003, however, Salfi sent to Maggio and others in Portfolio and Financial Reporting an e-mail notifying them of a change in procedure: I have not received all of KPMG's comments, but one major one has to do with the frequency of hedge documentation. We must document that the commitments are hedges by the end of the day that [they] are entered. We originally thought we could do it monthly, but KPMG would not accept that practical cut.(862) - ---------- identified at the outset of the hedge, as purchases for the portfolio could not be identified until that process was completed. (860) FAS 133 P 28(a)(2). (861) FAS 133 P 29(b). (862) E-mail from P. Salfi to L. Maggio, dated June 30, 2003, Zantaz document 944907. 228
Salfi proposed corrections to the documentation "that we need to have . . . beginning July 1" - that is, the following day.(863) Other issues surrounding the documentation arose during the following months. On September 29, 2003, for example, Salfi reported that the Company would have to identify the individual commitments it intended to use as hedges: "[G]oing forward, we are going to have to identify the deal numbers in our hedge documentation for portfolio commitments that we document as cash flow hedges, MTBS sale commitments that we document as fair value hedges, and MTBS buy commitments that we will be documenting as cash flow hedges because of [fails to ourselves]." According to Salfi, "this last documentation needs to start occurring on Oct. 1 [two days after the e-mail] using the template I sent around on Friday."(864) Maggio forwarded the e-mail to Douthit with the message, "ANOTHER FAS 149 change!!!!!!!!!!"(865) Douthit forwarded Maggio's e-mail to Pennewell, who prioritized the transactions based on their impact on Fannie Mae's income; she identified documentation of fails to ourselves as "the big problem."(866) The Company's hedge documentation during this period specified that the commitments "will now be used to lock-in the purchase or sale price on forecasted mortgage asset acquisitions or sales and hedge against the impact of changes in market interest rates on the future purchase or sale prices."(867) The documentation went on to state that "[t]he critical terms of these commitments match the forecasted purchases/sales, so we expect that the hedges will be perfectly effective at inception and on an ongoing basis."(868) Fannie Mae's approach was problematic, however, because the documentation did not identify which trade, or group of trades, was the hedged item, and - ---------- (863) Id. (864) E-mail from P. Salfi to J. Pennewell, copy to M. Douthit, dated Sep. 29, 2003, FMSE-E 1178554-56. (865) Id. at FMSE-E 1178555. (866) Id. at FMSE-E 1178554. (867) Id. (unnumbered attachment); see also FAS 149 Adoption at Fannie Mae - OFHEO Review, dated Nov. 6, 2003, FMSE-IR 352360, at FMSE-IR 352376 (hereinafter "Hedge Documentation Process"). (868) Hedge Documentation Process at FMSE-IR 352376. Another version of the hedge documentation appears in the August 2003 e-mail from Salfi to Mary Lewers. E-mail from P. Salfi to M. Lewers, dated Aug. 21, 2003, FMSE-E 117249-50. 229
which was not.(869) The MTBS trades, fails to ourselves, and sorting and sifting illustrate the problem with Fannie Mae's approach to documentation. In an MTBS trade, neither the purchase nor the sale commitment was treated as a hedge, because Fannie Mae did not intend to take the securities into its portfolio; moreover, any changes in the fair values of these commitments presumably would offset each other, so hedge accounting was not necessary to address the effect of the derivatives on the Company's earnings.(870) However, as a result of the sorting and sifting process and its approach to satisfying third-party sale commitments in advance of commitments to its portfolio, the Company could take the security delivered pursuant to the MTBS purchase commitment into its portfolio and substitute a security obtained in settlement of another commitment that had been intended for the portfolio and thus had been documented as a hedge.(871) The procedure, in effect, treated the unhedged transaction as the hedged transaction, and vice-versa. Thus, those involved in the transaction did not - indeed, could not - document at the outset of the hedge which commitment transactions would result in the security delivered to the portfolio and, as such, could not determine which commitments could qualify as hedges. To address this issue, the Company would document the settled MTBS buy commitment "as the replacement cash flow hedge of the forecasted acquisition of assets into the portfolio."(872) The Company's draft requirements specifications for the documentation function in a proposed FAS 149 accounting system summarized this approach as follows: At a high level, our documentation plan is to produce a daily hedge memo that indicates the total amount of new buy and sell commitments for the day that qualify for cash flow hedges, and the outstanding qualifying commitments to date. The text of this memo will explain that the outstanding commitments-to-date number includes [fails to ourselves] and third party pair offs. As a backup to this daily memo, we will produce a file listing the IPOR deals - ---------- (869) As noted above, the Policy appears to approach the hedges of commitments as "all-in-one" hedges. The documentation can be read to be consistent with that approach. Alternatively, it can be viewed as an effort to document hedges involving more than one commitment. FAS 133 permits the latter, subject in all cases to the requirement that, when a transaction occurs, it is clear from the documentation whether the transaction is a hedged transaction or not. (870) Policy at FMSE-IR 482069. (871) As a result of problems with the interaction between sorting and sifting and ASAPs, the sorting and sifting process was suspended in November 2003. E-mail from M. Douthit to A. McCormick, dated Nov. 6, 2003, FMSE-E 1179760. (872) Id. 230
[that is, the commitments to acquire securities for the Company's portfolio] that represent our portfolio position that qualifies for cash flow hedge accounting. This file will also include the list of corresponding external commitments (XPOR [commitments to purchase from or sell to third parties], MTBS) and amounts that total this portfolio position. . . . While these two sets of deals (internal and external) will agree in total, we will not create a deal by deal cross-reference. . . . After settlement, we can report on which specific external commitment was used to fill any given [receipt into the portfolio].(873) In other words, the determination of which commitments resulted in the delivery of a security to the portfolio, and thus qualified for hedge accounting, could be accomplished only by a reconciliation process at month-end.(874) As these documents and the fails to ourselves example illustrate, this approach to FAS 149 hedged transactions did not satisfy the requirement in FAS 133 that the Company's documentation describe the transaction such that it would be clear when the hedged transition occurred. We conclude, therefore, that the Company's documentation did not satisfy the requirements of GAAP. C. The Probability Assessment As noted above, FAS 133 specifies that a forecasted transaction is eligible for hedge accounting only if the transaction is "probable" of occurring. For reasons similar to those stated above, we conclude Fannie Mae did not meet this requirement. The Policy identifies a number of instances in which a forward trade would not occur. These include: The counterparty's (or Fannie Mae's) decision to pay a pair-off fee in lieu of delivery; a cancel and correct; and the failure of the Company's counterparty to perform on its commitment obligations outright.(875) The Policy posits that these instances are relatively rare and even in these circumstances "the hedged forecasted transaction (the purchase of loans or securities) is still probable to occur because the underlying assets are typically obtained within 60 days of the original settlement date."(876) - ---------- (873) Requirements Specification: FAS 149, Cash Flow Hedge Documentation, Version 1.3, dated Mar. 17, 2004, FMSE-SP 095598-611 (emphasis added). (874) It is notable that, during implementation of FAS 133, the Company developed systems (the DEBTS system and the FAS 133 system) to ensure that it could document a link between a specified derivative and hedged instrument. (875) Policy at FMSE-IR 482066. (876) Id. 231
The Policy addressed the probability assessment much as it had the documentation requirement: it considered the transactions on an aggregate, end-of-month basis. According to the Policy, this monthly analysis would corroborate the initial assessment that the hedged transactions were probable on the whole: "On a rolling 12 month basis, we will review loan pair-off activity, securities cancel and corrects, and fails to assess the rate at which commitments did not settle (and the forecasted transaction did not occur) within 60 days of the forecasted settlement date and our upfront conclusion that the purchases are probable to occur."(877) That is, the Company would compare the level of closed commitments to the total volume of commitments to determine the overall level of transactions that did not occur. Although the policy itself does not set a probability threshold, the Company set that threshold at seventy-five percent - that is, a transaction was considered probable if, on an aggregate, rolling, twelve-month basis, no more than twenty-five percent of the forward trades linked to a forward commitment failed.(878) According to the Company's documentation and the Policy, at the end of the month, the Company would "verify . . . whether open purchase/sale commitments did meet the criteria above to qualify for hedge accounting."(879) - ---------- (877) Id. (878) Requirements Specification, FAS 149: Fail Resolution Probability Report, Version 1.0, dated June 28, 2004, FMSE-SP 95094-100, at FMSE-SP 95096. (879) Hedge Documentation Process at FMSE-IR 352376. The Policy appears to have selected the end of the month for assessing the probability that open commitments were still expected to result in delivery of a loan or security because that would be the point at which information on closed commitments would be available: To qualify for hedge accounting, we must also document the derivatives used, the hedged items, the risk being hedged, how we will assess hedge effectiveness, and measure hedge ineffectiveness. Each day, we will establish purchase and sale commitments/derivatives as hedging instruments in hedging relationships that protect against the effect of changes in market interest rates on the cash flows related to the prices of future acquisitions or sales. At month end, we will monitor the timing and probability of forecasted acquisitions or sales based on the status of the open commitments. Policy at FMSE-IR 482065. The reference to month-end monitoring caused confusion. In August 2003, Salfi sent Lewers an e-mail with the comment: "Here is what we are using to document the commitments as hedges on a daily basis." E-mail from P Salfi to M. Lewers, dated Aug. 23, 2003, FMSE-E 117249-50. Lewers noted the reference in the documentation to month-end monitoring and posed several questions: "What does 'We will verify at month-end whether open purchase/sale commitments did meet the criteria above to qualify for hedge 232
It is not clear whether the Company had a system in place to corroborate the probability assessment stated in the Policy. We were told during interviews that a procedure was in place to track the probabilities manually from the outset. A presentation regarding the FAS 149 implementation process - which appears to be from no earlier than the fourth quarter of 2003 - however, contained the following language: PROBABILITY RISK In order to qualify for hedge accounting, our forecasted transactions must be probable to occur within 60 days of intended settlement. Probability is defined as 75% actually occurring over a rolling 12 month period. A number of events count against our probability: - Cash pairoffs - Fails over 60 days - Cancel/Corrects that change the commitment parameters or intended settlement beyond 60 days. We are not currently testing our probability %.(880) The decision to exclude certain transactions entirely from the probability assessment is also questionable. In a number of different scenarios, such as fails to ourselves, the hedged transaction that the Company entered into at the outset did not settle, although another transaction may have. Nonetheless, Financial Standards concluded that these transactions would not be included in the assessment of the probability that the hedged transaction that the Company initially entered into would occur. The requirements-specifications document for a proposed FAS 149 accounting system describes the approach to probability: - ---------- accounting' mean? Specifically, who is 'we' and what is the evidence of this verification process?" Salfi responded that he anticipated Financial Reporting (that is, Lewers's group) would determine whether there were open commitments at month's end that would not qualify for hedge accounting. Id. (880) Undated FAS 149 Risk Assessment, FMSE-SP 95114-20, at FMSE-SP 95120 (emphasis added). The document refers to the fact that the Company was able to defer $2.4 billion in other comprehensive income as a result of "third quarter hedging relationships," suggesting that the document was prepared during the fourth quarter of 2003 at the earliest. Id. at FMSE-SP 95118. We were told that it is possible that the statement regarding probability testing referred to automated testing. 233
In the long term, this [probability] report should evaluate each derivative that was given cash flow treatment to determine its disposition. These derivatives can either settle, be substituted with another derivative (third party pair-off), fail, be net settled for cash, or be substantially modified. If the derivative were substituted, we should show the disposition of the replacement trade. To do this correctly, we need to be able to specifically identify the external commitment that is being given hedge treatment and to be able to accurately trace through the replacement trade relationships. This is difficult in the near term.(881) A specification prepared at a later date stated: "We can substitute one external trade for another as the cash flow hedge without impacting probability or creating a DNQ event [that is, a hedged transaction that does not qualify for hedge accounting], as long as we document the new trade when the substitution occurs. This situation happens with third-party pair offs and [fails to ourselves]."(882) In sum, the procedure excluded hedged transactions that had not occurred from the probability assessment, so long as another hedged transaction that settled was substituted for the original trade and that substitution was documented. Further, the decision to ignore pair-offs and fails in determining whether it was probable the security underlying a commitment would be delivered was inconsistent with the Company's apparent reliance on the "all-in-one" cash flow guidance. As an "all-in-one" cash flow hedge required identifying a specific transaction as the hedged item, once that transaction failed or was paired off, hedge accounting was no longer appropriate because it was probable that the hedged transaction would not occur. As such, gains or losses deferred on that transaction should have been recognized in income. Gains and losses on the derivatives were deferred as long as it was probable that the Company would purchase a replacement security. Our conclusions regarding the implementation of the FAS 149 probability assessment are similar to our conclusions regarding the documentation requirement. There was no mechanism in place to meet the FAS 133 requirement that the Company apply hedged transactions only with respect to transactions that were probable of occurring. For this reason, too, the use of hedge accounting for commitments did not comply with GAAP. - ---------- (881) Requirements Specification: FAS 149 Probability Report, Version 1.0, dated Nov. 20, 2003, Zantaz document 785570, at 3-4. (882) Requirements Specification: FAS 149, Cash Flow Hedge Documentation, Version 1.3, dated Mar. 17, 2004, FMSE-SP 95598-611, at FMSE-SP 95601. 234
VI. FINDINGS REGARDING SENIOR MANAGEMENT, OFHEO, AND THE BOARD The evidence overall reveals that senior managers in the Controller's Office had little involvement in the FAS 149 implementation process. Spencer's retrospective on the events of 2003 refers to Salfi's leadership of the effort within Financial Standards, and Boyles stated that responsibility for the FAS 149 implementation effort fell primarily on Salfi. It appears that Salfi endeavored to implement the standard in good faith, but the number and complexity of the issues raised by the new standard, combined with the short time period in which he had to work, led to significant errors. There is no indication that Timothy Howard was involved in the FAS 149 implementation. On the contrary, an exchange between Howard and Boyles in June 2003 reflects Howard's lack of familiarity with the subject. Howard initiated the exchange with an e-mail concerning the accounting problems that had been noted in a recent Freddie Mac press release. The e-mail included a reference to FAS 149: "The one item I'm not clear on, however, is the accounting for forward commitments. You say that Freddie's forward commitments 'generally qualified as derivatives under GAAP.' Did ours not? Why not? And what is FAS 149, what does it change and when does it go into effect?"(883) Howard went on to say that he "need[s] to get more clarification on this point. I don't in any way want to underinform or mislead the Board."(884) Boyles's reply to Howard offered only a brief overview of the issues that FAS 149 raised for Fannie Mae. Boyles explained why the Company did not treat forward commitments as derivatives prior to FAS 149, noted that FAS 149 requires that commitments be treated as derivatives, and explained that KPMG was in the process of reviewing Fannie Mae's accounting, but "[s]o far they have not indicated to me [Boyles] anything other than they agree with our accounting treatment."(885) It does not appear that any other senior officers in the Company were involved in the implementation of FAS 149. An issue that arose later in 2003 drew senior management and the Board into additional discussions regarding the FAS 149 implementation process. The Company made an error in the preparation of its financial statements for the third quarter of 2003, which was the first quarterly close in which FAS 149 affected the financial statements. The error resulted from the duplication of an adjustment relating to the - ---------- (883) E-mail from J. Boyles to T. Howard, dated June 27, 2003, FMSE-E 5639-41, at FMSE-E 5640. (884) Id. (885) Id. at FMSE-E 5639-40. 235
valuation of the Company's outstanding commitments.(886) The calculations were done mainly on spreadsheets (referred to as "end user systems"), as there was no integrated FAS 149 system to make the necessary calculations. The error resulted in a more than $1 billion understatement of total assets and stockholders' equity on the Company's balance sheet. Before the error came to light, on October 16, 2003, management had issued a press release announcing its quarterly results.(887) Accordingly, on October 29, 2003, management issued a new press release and filed a Form 8-K with the SEC.(888) OFHEO issued a statement that same day noting the error and suggesting that it "underscores the need for the special review OFHEO is about to begin of accounting policies, practices, and internal controls at Fannie Mae."(889) The Company's Senior Vice President for Investor Relations responded the next day with a second press statement: "These were honest mistakes made in a spreadsheet used in the implementation of a new accounting standard. The bottom line is that the correction has no impact on our income statement, but resulted in increases to unrealized gains on securities, accumulated other comprehensive income, and total shareholder equity (of $1.279 billion, $1.136 billion, and $1.136 billion, respectively)."(890) Meanwhile, Franklin D. Raines alerted the Board to the error in a memorandum dated October 29, 2003. He told the Board that, in connection with the FAS 149 calculations, "the company calculated unrealized gains utilizing a faulty - ---------- (886) The adjustment was incorporated into a computation that Portfolio provided to the Controller's Office; the Controller's Office later repeated the adjustment. (887) See Fannie Mae press release, dated Oct. 29, 2003, available at http://www.fanniemae.com/newsreleases/2003/2807.jhtml?p=Media&s=News+Rele ases (last visited Feb. 17, 2006). (888) See Fannie Mae press release, dated Oct. 29, 2003, available at http://www.fanniemae.com/newsreleases/2003/2822.jhtml (last visited Feb. 17, 2006). (889) Statement of OFHEO Director Armando Falcon, Jr. Regarding Fannie Mae Accounting Error, dated Oct. 29, 2003, available at http://www.ofheo.gov/News.asp?FormMode=Releases&ID=113 (last visited Feb. 17, 2006). (890) See Statement by Jayne Shontell, Fannie Mae Senior Vice President, Investor Relations, dated Oct. 30, 2003, available at http://www.fanniemae.com/ir/issues/financial/103003.jhtml (last visited Feb. 17, 2006). 236
spreadsheet formula."(891) Management briefed the Board the following day at a meeting convened by conference call. According to the minutes of the meeting, Raines "reviewed the actions the company had taken since discovery of the error .. . ., including reviews by senior management, interaction with the company's regulators . . . and acting to ensure that the corporation fully understands what occurred and implements the corporation's FAS 149 plan expeditiously."(892) Spencer also addressed the Board. According to minutes of the Board meeting she attended, Spencer "assured the Board that her team is focused on expediting the automated implementation of these new accounting requirements and that the necessary corporate resources had been devoted and would continue to be devoted to implementing appropriate system changes to accommodate the requirements of FAS 149."(893) Raines then "added that all appropriate changes would be completed in stages, with final implementation by the end of first quarter 2004."(894) It does not appear that Spencer alerted the Board to the resource issues that had arisen during the implementation of FAS 149 - or the other problems that the Controller's Office had faced in 2003 - either at the October 30 Board meeting or thereafter. On the contrary, according to the minutes of the February 2004 Audit Committee meeting, Spencer referred to 2003 merely as "an active year in adopting changes to accounting standards."(895) In response to a question from one of the committee members concerning FAS 149 specifically, "Ms. Spencer responded that the automation of the FAS 149 processes are [sic] currently underway with an interim solution in place and a more permanent system being developed."(896) None of the Board members we interviewed recalled Spencer, or anyone else, alerting them to problems stemming from a lack of resources or with respect to the implementation of any accounting standard. The Board and Audit Committee minutes suggest that senior management, and particularly Spencer, were not entirely forthcoming to the Board regarding the - ---------- (891) Mem. from F. Raines to the Board of Directors, dated Oct. 29, 2003, Zantaz document 3528788, at 2. (892) Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Oct. 30, 2003, FMSE 10449-51, at FMSE 10449. (893) Id. at FMSE 10450. (894) Id. (895) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 17, 2004, FMSE 504818-30, at FMSE 504821. (896) Id. 237
implementation of FAS 149 or the problems the Company faced during 2003 - for example, that the Controller's Office was understaffed.(897) In fact, the document in which Spencer discussed the difficulties the Controller's Office had faced in 2003(898) was dated just one week after the February 2004 Audit Committee meeting. We have not seen evidence that Spencer told the Board about 2003 being the "Perfect Storm." Spencer's failure to be fully forthcoming with the Board at the time of the $1 billion computational error about the FAS 149 difficulties was especially significant. OFHEO's press statement about the error made clear that OFHEO intended to treat the error as an important matter. Her similar failure before the Audit Committee in February 2004 was notable as well. On February 24, 2004, the week after the Audit Committee meeting, OFHEO reported to Fannie Mae on the results of its FAS 149 review, including its conclusion that "[t]he lack of a fully automated [FAS] 149 accounting process as well as the total number of other end user computing systems at Fannie Mae raise concerns."(899) Particularly in the circumstances surrounding these two meetings, we believe that management should have taken extra care to inform the Board of problems in the implementation of important accounting standards such as FAS 149. Had that information been conveyed, the Board could have assessed the situation and taken such action as it deemed appropriate, both with respect to the Company's internal affairs and with respect to the Company's relations with OFHEO. We conclude that management's failure to keep the Board fully informed deprived the Board of that opportunity. PART D: CLASSIFICATION OF SECURITIES HELD IN PORTFOLIO I. INTRODUCTION In this Part, we consider Fannie Mae's classification of securities it acquired as held-to-maturity ("HTM") or available-for-sale ("AFS"). In its February 11, 2005 letter, OFHEO raised concerns with respect to Fannie Mae's accounting for the securities it held in its portfolio. In particular, OFHEO noted that the Company did not appear to be following the requirements of FASB Statement No. 115, Accounting for Certain Investments in Debt and Equity Securities ("FAS 115"),(900) with regard to the - ---------- (897) See Boyles 2003 Performance Review at FMSE-SP 102951 (acknowledging that the Controller's Office had a "thinner staff than we both would have liked."). (898) See supra note 838 and accompanying text. (899) Letter from A. Falcon, Jr. to F. Raines, dated Feb. 24, 2004, FMSE-IR 352389-90, at FMSE-IR 352389. (900) ACCOUNTING FOR CERTAIN INVESTMENTS IN DEBT AND EQUITY SECURITIES, Statement of Fin. Accounting Standards No. 115 (Fin. Accounting Standards Bd. 1993) (hereinafter "FAS 115"). 238
timing of the classification of a security as either HTM or AFS. OFHEO stated that the Company classified the securities in the portfolio "for accounting purposes at the end of the month in which the trade settles."(901) We conclude that Fannie Mae's accounting under FAS 115 violated an unambiguous accounting rule regarding the classification of securities as HTM, AFS, or trading. Moreover, the Company's procedure for determining a security's classification involved the consideration of factors that would affect the security's fair value. We saw no evidence, however, that the Company intentionally used this mechanism to manipulate its net income. Moreover, at least as of 2003, KPMG was aware of the Company's practices in this area and raised no objection. II. BACKGROUND A. Applicable Accounting Principles FAS 115 states: "At acquisition, an enterprise shall classify debt and equity securities into one of three categories: held-to-maturity, available-for-sale, or trading."(902) As to the HTM category, FAS 115 states: Investments in debt securities shall be classified as held-to-maturity and measured at amortized cost in the statement of financial position only if the reporting enterprise has the positive intent and ability to hold these securities to maturity.(903) Unlike securities classified as AFS or trading, HTM securities are carried at their amortized cost.(904) Classification of a security as HTM therefore minimizes the volatility in equity and earnings associated with securities classified as AFS and trading, respectively. FAS 115 describes the circumstances in which a company may transfer securities from one category to another. The movement of securities from the HTM category to AFS or trading status is limited to six narrow circumstances (for example, changes in the company's regulatory capital requirements, a significant deterioration in the issuer's creditworthiness, or a change in the tax law that alters the tax-exempt status - ---------- (901) According to OFHEO, "[t]his process allows management to conduct several intra-month economic benefits." February 11, 2005 OFHEO Letter at FMSE-IR 547319. (902) FAS 115 P 6 (emphasis added). (903) Id. P 7. (904) See id. 239
of the interest on the debt security).(905) The sale of a security from the HTM category for reasons other than those six narrow circumstances "taints" the HTM portfolio. The penalty for tainting the HTM portfolio is to require the company to reclassify all HTM securities as either AFS or trading securities and to deprive the company of the right to classify new acquisitions as HTM for up to two years.(906) The principle behind FAS 115 is clear: because an HTM security will be held for its entire term (absent unusual circumstances), the cost of that security should be amortized over the security's term. Thus, in the FASB's view, "a change in fair value, which would reflect a change in the market interest rates or prepayment risk, is not relevant for a security that will be held to maturity."(907) A security that a Company intends to sell, or that it may sell prior to the security's maturity, however, should be marked to fair value to reflect the amount that could be realized from a sale of that security. The requirement that the determination of an intention to hold a security as HTM be made "at acquisition" follows from this principle. As the Board explained when it issued FAS 115: The Board deliberately chose to make the [HTM] category restrictive because it believes that the use of amortized cost must be justified for each investment in a debt security. At acquisition, an enterprise should determine if it has the positive intent and ability to hold a security to maturity, which is distinct from the mere absence of an intent to sell. The Board believes that, if management's intention to hold a debt security to maturity is uncertain, it is not appropriate to carry that investment to amortized cost; amortized cost is relevant only if a security is actually held to maturity.(908) Finally, we note that, as a practical matter, postponement of the classification of a security as HTM would allow the acquiror to take into account any change in the value of the security when it decides the category into which the security will be placed, and thereby manage its financial statements to that extent. - ---------- (905) See id. P 8. (906) A GUIDE TO IMPLEMENTATION OF STATEMENT 115 ON ACCOUNTING FOR CERTAIN INVESTMENTS FOR DEBT AND EQUITY SECURITIES: QUESTIONS AND ANSWERS HELD-TO-MATURITY SECURITIES, Questions 12, 13 (Fin. Accounting Stds. Bd. Nov. 1995); see also Armando Pimentel, Remarks to the Twenty-Fourth Annual National Conference on Current SEC Developments (Dec. 11, 1996). (907) FAS 115 P 62. (908) Id. P 59. 240
B. Fannie Mae's Accounting Policy Fannie Mae's accounting policy does not require that it classify the securities that enter its portfolio "at acquisition." The October 2003 version of Fannie Mae's accounting policy, which reflects prior policy as well, states that the Company "classifies its investments in marketable debt and certain equity securities as either [HTM, trading, or AFS] in the month of acquisition and assess the appropriateness of the classification at each reporting date."(909) The policy notes further that "[p]roper classification is important because the accounting treatment depends on the classification."(910) III. FINDINGS A. Fannie Mae's Classification of Debt Securities Given the nature of Fannie Mae's business and its portfolio management strategies, most of the Company's securities were classified as HTM. A smaller number were classified as AFS, based in part on the Company's preview of its securitization activity (securities classified as AFS were used in the REMICs that Fannie Mae created and sold). Fannie Mae also classified all "private label" (non-Fannie Mae or Freddie Mac) securities as AFS. In practice, Fannie Mae typically brings securities into its portfolio during a month as HTM securities and at the end of the month determines the final classification of those securities. Fannie Mae refers to this practice as the Intra-Month Redesignation, or "IMRD," process. The mechanics of the IMRD process are reflected in a number of documents, including the Company's Portfolio Transactions Procedures Guide.(911) Fannie Mae's trading systems allow traders to put one of four electronic "flags" on each security that the Company acquires at the time the trader enters into a purchase commitment. The four flags are HTM, HTM for dollar rolls (that is, the securities will be used as collateral for dollar rolls), AFS, and trading. The system's default classification - the classification when the trader does not specify another classification - is HTM. - ---------- (909) Financial Accounting Policy Guide: Nonmortgage Investments, dated Oct. 2003, FMSE-SP 79012-25, at FMSE-SP 79012; see also id. at FMSE-SP 78967 ("Fannie Mae's policy is to determine the loan/MBS classification by the end of the month in which the asset is purchased."). (910) Id. (911) Fannie Mae Portfolio Transactions Policy and Procedures Guide, dated May 11, 2004, FMSE 219734-81, at FMSE 219773-74. 241
Prior to the end of each month, an executive in Portfolio reviewed the securities acquired during that month and determined whether they should be maintained in the HTM category or moved into the AFS category. These decisions were relayed at a brief "account designation" meeting at the end of each month.(912) Following the meeting, a manager in Portfolio was responsible for implementing the changes in the security classifications conveyed at the meeting. The classification for each security would be changed in the relevant Fannie Mae systems.(913) According to interviewees, Fannie Mae's systems prohibit the removal of a security from the HTM account following the close of the Company's books for the month in which the security was acquired. At that time, only certain personnel had the authorization to override the system controls and remove a security from the HTM category. During the month of acquisition, however, the security could be moved from the HTM category and, as a result of the IMRD process, a substantial number of securities were moved from the HTM to AFS category in the course of a typical month.(914) We are aware of no accounting literature that would support Fannie Mae's interpretation of the FAS 115 requirement that securities be classified "at acquisition" as meaning "at the end of the month of acquisition," or through a procedure like the IMRD process.(915) On the contrary, the FASB considered and rejected a proposal that would have permitted a procedure similar to Fannie Mae's. According to the FASB, "[s]ome respondents [to the FAS 115 Exposure Draft] indicated that, although they have the intent to hold the vast majority of their investments to maturity, they do not know at acquisition which specific securities will or will not be sold. Having to classify securities upon acquisition does not, in their opinion, provide the desired degree of flexibility to manage their portfolio."(916) In response, the FASB rejected any general approach that would - ---------- (912) These meetings were attended by representatives of various trading groups - such as the fixed-rate MBS and multifamily groups - who were supposed to convey their groups' needs for certain categories of securities (such as dollar rolls) for the following month. (913) At our request, management provided us with a demonstration of the relevant systems. (914) It appears that up to sixteen percent of the securities acquired in a given month might be redesignated from the HTM to AFS category. See E-mail from Renie Grohl to Ann M. Kappler, et al., dated Dec. 19, 2002, FMSE-E 2233590. (915) In our view, the classification of a security as HTM in the Company's systems at acquisition was operative, and Fannie Mae improperly redesignated HTM securities to its AFS account. According to several interviewees, some of those involved in the IMRD process viewed the original "flag" as inconsequential (and they viewed the reference in "IMRD" to "Redesignation" as a misnomer). (916) FAS 115 P 67. 242
permit flexibility in the classification of HTM securities in favor of restricting the removal of securities in the HTM category to the six limited circumstances discussed above.(917) For example, the FASB specifically rejected an approach that would allow a company to specify a percentage of its portfolio as not HTM at the end of each year.(918) Interviewees had different explanations for the Company's application of FAS 115 and its interpretation of the "at acquisition" requirement. Several suggested that the Company's approach was a holdover from its practice prior to the adoption of FAS 115. However, this explanation does not account for the failure to address the issue when the Company adopted its FAS 115 implementation policy in 1994, when it reviewed and revised its accounting procedures in 2003, or when it reviewed its practices in connection with the sale in 2003 of securities classified as HTM (discussed below). A 2003 presentation, for example, states: "Classifications of loans and securities must be made by the end of the month."(919) Jonathan Boyles and other interviewees explained the Company's approach by reference to the fact that the Company closed its books on a monthly basis, in contrast to many banks and other financial institutions that close their books on a daily basis.(920) By this view, the Company should not be deemed to have designated its securities prior to the monthly close, as the securities were not entered formally on the Company's books and records until that time. Any gain or loss on the sale of the security during the month of acquisition, and any unrealized gain or loss on the security, would be recognized when the Company closed its books at the end of the month. There is no documentary evidence, however, that Fannie Mae considered this rationale when it adopted its FAS 115 accounting policy, or later when all accounting policies were reviewed. Moreover, under this rationale, the FAS 115 "at acquisition" requirement depends on how often a company closes its books. We have found no language in FAS 115 to support that view, and the FASB rejected reclassification of securities after acquisition except in limited circumstances.(921) Fannie Mae's practice was inconsistent with this underlying principle. - ---------- (917) Id. PP 68-70. (918) Id. P 68. (919) Undated Loan and Security Classifications, FMSE 93541-57, at FMSE 93550. Interviews place the date of this presentation in early 2003. The final page of the presentation, captioned "Summary," states that "[c]lassifications must be determined at the time of purchase." Id. at FMSE 93557. When asked about this apparent inconsistency, Boyles indicated that the point made on the final page of the presentation was intended to mean "month of purchase." (920) As we discuss below, this point is referenced in a KPMG workpaper. (921) FAS 115 P 69. The FASB considered "whether the standard should permit enterprises to sell without justification some specified amount of the [HTM] 243
Finally, it does not appear that, prior to late 2003, the Company considered the ramifications of the guidance in Emerging Issues Task Force Issue No. 96-11, Accounting for Forward Contracts and Purchased Options to Acquire Securities Covered by FASB Statement No. 115 ("EITF 96-11").(922) EITF 96-11, issued in 1996, addressed circumstances in which a company entered into a commitment to acquire a security that would be subject to FAS 115 (that is, the security would be designated at acquisition as HTM, AFS, or trading).(923) Under EITF 96-11, classification was required at the inception of the commitment, and the subsequent accounting for those commitments would be consistent with the accounting required by FAS 115 (i.e., commitments designated as AFS would be marked to fair value with changes recognized in OCI, and commitments designated as trading would be marked to fair value with changes recognized in earnings).(924) With respect to commitments, therefore, classification as HTM, AFS, or trading was required even before the Company actually acquired the security.(925) Consideration of EITF 96-11, therefore, would have confirmed that classification of a security at month's end was inappropriate. There is no evidence that Fannie Mae considered the implications of EITF 96-11 in connection with its interpretation or application of FAS 115. The Company did consider the implications of EITF 96-11 in late 2003, but only from the perspective of accounting for the fair value of the commitments, and not with respect to the classification of the commitments or the acquired securities. In an analysis of EITF 96-11 prompted by the Freddie Mac restatement, Financial Standards noted: EITF 96-11 requires that entities entering forward purchase contracts or purchase options to acquire securities that will - ---------- securities without calling into question the enterprise's intent to hold other debt securities to maturity. The Board rejected that approach as being inconsistent with the premise underlying the use of amortized cost - that management intends to hold all such securities to maturity." The Board concluded that "if the sale of a held-to-maturity security occurs without justification, the materiality of that contradiction to the enterprise's previously asserted intent must be evaluated." Id. (922) ACCOUNTING FOR FORWARD CONTRACTS AND PURCHASED OPTIONS TO ACQUIRE SECURITIES COVERED BY FASB STATEMENT NO. 115, Emerging Issues Task Force of the Fin. Accounting Standards Bd., Issue No. 96-11 (May 1996) (hereinafter "EITF 96-11"). (923) See id. para. 1. (924) Id. paras. 3-4. (925) Under EITF 96-11, with respect to commitments designated as HTM, "[i]f an entity does not take delivery under the forward contract . . . the entity's intent to hold other debt securities to maturity will be called into question." Id. para. 3. 244
be accounted for under FAS 115 . . . be accounted for similarly to the purchased securities. The forwards or options must be designated, at inception, as [HTM, AFS, or trading] consistent with FAS 115.(926) Financial Standards concluded that it had not complied with EITF 96-11, but only by failing to book changes in the fair value of commitments to purchase mortgage revenue bonds ("MRBs").(927) The effect of not applying the guidance in EITF 96-11 as to AFS securities was deemed inconsequential.(928) It appears that the Company did not consider the impact of the guidance on its HTM portfolio.(929) B. Practical Application of the Company's Policies In the absence of a convincing rationale for the Company's interpretation of the FAS 115 requirement that securities be classified "at acquisition," we looked to the Company's application of its policy to determine whether the practice suggested an inappropriate motivation to manage the Company's financial statements. The evidence reveals that Fannie Mae's practices were driven by considerations relating to the systems the Company had developed and its interest in maintaining flexibility in the management of its portfolio. We have found no evidence that the Company used its FAS 115 policy to manipulate earnings directly, although its approach to FAS 115 could have materially affected its financial statements. The implications of FAS 115 for the Company were set forth in a memorandum from James Parks to Timothy Howard and then Controller Michael Quinn, dated April 15, 1993 (one month before FAS 115's issuance).(930) The memorandum described the impact of the new rule on Fannie Mae's business. It noted that the Company would have to establish "separate portfolios" for securities that it might sell - ---------- (926) Mem. from Jonathan Boyles to Distribution, dated Jan. 14, 2004, FMSE-SP 78679-85, at FMSE-SP 78684. (927) Id. at FMSE-SP 78685; Audit Committee Update: Understanding Freddie Mac's Accounting Restatement as It Affects Fannie Mae, dated Jan. 23, 2004, FMSE-IR 379360-78, at FMSE-IR 379375. (928) Audit Committee Update: Understanding Freddie Mac's Accounting Restatement as It Affects Fannie Mae, dated Jan. 23, 2004, FMSE-IR 379360-78, at FMSE-IR 379375. (929) Under EITF 96-11, if an entity fails to classify as HTM a security for which the commitment was designated HTM, then "the entity's intent to hold other debt securities to maturity will be called into question." EITF 96-11 para. 3. (930) Mem. from James Parks to Timothy Howard and Michael Quinn, dated Apr. 15, 1993, FMSE-IR 362117-19. 245
prior to their maturity date and that securities may not be transferred out of the HTM category.(931) It further noted that "[b]uy/sell activity will have to be done on a specific [security-by-security] identification basis, rather than notionally as it is now, and run through a separate 'held-for-sale' portfolio. This portfolio will have to be managed carefully to avoid earnings volatility from the lower of cost or market accounting methodology."(932) The memorandum further stated: "Our ability to manage earnings through sales out of portfolio will be largely curtailed. It will still be possible to manage earnings to a certain extent, however, by establishing a separate portfolio of MBS 'available for sale.'"(933) Parks explained during an interview that his statement about managing earnings referred to Fannie Mae's practice of selling whole loans to re-balance the average yields in its loan portfolio. Parks clarified that, although he understood that FAS 115 applied to debt securities and not whole loans, he thought there might be analogies between the way that Fannie Mae was accounting for its loans and the new requirements of FAS 115. Parks also stated that he was unaware of any situation in which Fannie Mae engaged in transactions other than for a "valid business purpose" - i.e., solely to reach a particular accounting result. Parks's memorandum did not address the issue of classification "at acquisition." In our interviews, the individuals in Portfolio Management who were responsible for making the decision confirmed that they did have available to them information regarding the securities's gain or loss position, although we were told that these data were not reliable and were not used to make redesignation decisions. The executive responsible for making the decisions on whether a security should be designated as AFS at month's end, Matthew Douthit, acknowledged that he considered the securities' coupon and other features to determine which securities would be held-to-maturity and which might be sold. He stated, for example, that in the IMRD process he considered "economic opportunities." In addition, Douthit noted that he also classified securities as AFS in order to offset some of the volatility in Fannie Mae's equity resulting from implementation of FAS 133. Fannie Mae's emphasis on flexibility in its application of FAS 115 created the potential for errors in the Company's classification of securities. An important event occurred in January 2003 when an employee in the Securities Trading Operations ("STO") group of the Treasurer's Office overrode the Company's controls over securities classified as HTM over a month-end in order to sell a security out of the HTM portfolio - ---------- (931) Id. at FMSE-IR 362118. (932) Id. at FMSE-IR 362119. (933) Id. 246
after the close of the month in which the security was acquired.(934) According to the documentation compiled during the Company's own inquiry into the matter, in December 2002 Fannie Mae entered into a brokered trade in which the Company agreed to acquire certain securities and then sell them to a third party. Because of an error at acquisition, the relevant securities were classified as HTM instead of trading. The misclassification was not recognized until the following month. A manager in STO engaged a systems analyst to override the controls that would have prevented sale of those securities, and the securities were transferred in two trades during January 2003.(935) The Company detected the first trade and investigated the matter. It determined that the trade was the result of the initial classification error and concluded that the sale would not taint the portfolio.(936) Interviewees suggested that KPMG, the Office of Auditing, and Financial Standards all took this event seriously and understood the implications that these mistakes had with respect to a potential taint of the Company's portfolio. We understand that, in the wake of this event, Financial Standards provided additional training on the requirements of FAS 115.(937) This training contains the conflicting statements discussed above about the timing of the characterization of the securities.(938) That proposition does not appear to have been challenged at that time. On the contrary, e-mail correspondence - ---------- (934) Mem. from Paul Jackson to Linda K. Knight and Leanne Spencer, dated Mar. 5, 2003, FMSE-IR 379252-54. (935) Id. at FMSE-IR 379254. (936) Id. According to the report, after the first trade from the HTM portfolio was detected, the matter was referred to Boyles for his consideration. He concluded that the trade was the result of an error and probably would not result in a taint of the Company's portfolio. Hearing this, the STO manager then made the second trade. According to the report, Boyles "was not aware at the time of his discussion that these remaining pools had not settled and would not have approved this subsequent transaction." Id. (937) Nonetheless, there is evidence that a second event occurred later in 2003. Apparently, this incident involved a systems error in which an entry to redesignate securities to AFS was duplicated by the system. Once the issue was identified, the securities were moved back to the HTM account. See E-mail from Mary Lewers to Janet L. Pennewell, dated July 1, 2003, FMSE-E 1162461. (938) Compare Undated Loan and Security Classifications, FMSE 93541-57, at FMSE 93550 ("Classifications of loans and securities must be made by the end of the month.") with id. at FMSE 93557 ("Classifications must be determined at the time of purchase."). 247
involving Portfolio Management in late 2003 conveys the understanding within that group that a security has no classification until the month-end close.(939) Thus, while there is no evidence that Fannie Mae used its securities classification procedures to manage earnings directly, it did use the flexibility provided by its policy to determine during the course of a month which securities it would hold to maturity and which it would hold for potential sale, with the corresponding impact on the Company's financial statements at a later date.(940) C. KPMG Review Fannie Mae's policies regarding FAS 115 were available to its outside auditor and a copy of the policy appears in KPMG's workpapers. Parks' recollection was that there were a number of conversations with KPMG at the time the FAS 115 policy was formulated, but he did not recall any details of those conversations. Specifically, Parks did not recall any discussions regarding the timing of the classification of the Company's securities. Boyles recalled conversations with KPMG on this issue; according to Boyles, KPMG agreed with the Company's policy on the timing of the securities' classification, but would have preferred that Fannie Mae classify the AFS category as the default. A KPMG workpaper dated August 2003 sets forth the Company's approach to the classification of securities. The workpaper states: "This [FAS 115] classification requirement is extended by end-of-month financial reporting conventions; therefore, the final classification must be made within the month the security is - ---------- (939) See, e.g., E-mail from Matthew Douthit to Kenneth Scott, dated July 8, 2004, Zantaz Document 374684 ("We do not routinely move securities from HTM to AFS as part of the month-end closing process. During the third week of each month, we do review securities that have settled for portfolio during the current month (as inventory, such assets are 'undesignated' until month-end) and assign an inventory category effective on the first day of the subsequent month (HTM, AFS, or HFT).") (940) One document indicated that Fannie Mae might have used the IMRD process to manage earnings, but we discount the document's reliability in this respect. HTM Portfolio Redesignation, dated Dec. 30, 2004, Zantaz document 2296663. The document is titled "HTM Portfolio Redesignation" and is captioned as a draft. The document relates to the Company's consideration of a possible move of securities from the HTM portfolio to the AFS portfolio in response to its September 27, 2004 agreement with OFHEO to increase its capital surplus. Id. at 1. The document has a chart showing the steps in the HTM Redesignation Process, and it labels each step either "standard" or "enhanced" for purposes of this one-time reclassification. Id. Two of the steps in the process are labelled "Select Securities & Estimate Gain/Loss" and "Validate Gain Loss Estimates." This flow chart indicates that these steps are "standard" procedures, rather than "enhanced." Id. Interviewees denied that actual gain or loss data were used as part of the IMRD process. 248
acquired." The workpaper discusses the IMRD process and notes, "[d]ecision factors considered by Portfolio Management [in the IMRD process] include balance sheet effects and economic opportunity."(941) D. Executive Involvement and the Report to the Audit Committee As mentioned above, in the wake of revelations concerning accounting irregularities at Freddie Mac, the Company undertook an inquiry to determine whether Fannie Mae's accounting raised similar concerns. In the course of that analysis, Financial Standards concluded that the Company had not complied with EITF 96-11, but only with respect to the acquisition of MRBs.(942) In a memorandum sent from Boyles to Franklin D. Raines and others, Boyles concludes on this issue: Based upon our analysis it appears as though our purchases of MRBs that ultimately were put into AFS classification should have been accounted for under EITF 96-11. The largest quarterly purchase of MRBs was approximately $301 million. Had all the purchase commitments been open at a quarter end, all subsequent deliveries been classified as available for sale and rates moved one point, the change in our balance sheet value would have been $3 million.(943) That information was incorporated into a presentation to the Audit Committee dated January 23, 2004. The presentation states, "Fannie Mae did not apply EITF 96-11 but the result was inconsequential (balance sheet effect less than $3 million). No effect on earnings."(944) We have not seen any other indication prior to 2004 regarding a discussion with the Board or a Board committee on this topic. IV. CONCLUSIONS Fannie Mae's policies regarding the application of FAS 115 were incorrect. Separate and apart from the question of whether (and when) the Company's - ---------- (941) Process Analysis Documentation - Acquisitions of MBS & Other Mortgage-Related Securities, dated Aug. 2003. (942) Mem. From J. Boyles to Distribution, dated Jan. 14, 2004, FMSE-SP 78679-85, at FMSE-SP 78684. (943) Id. at FMSE-SP 78685. (944) Audit Committee Update: Understanding Freddie Mac's Accounting as It Affects Fannie Mae, dated Jan. 23, 2004, FMSE-IR 379360-78, at FMSE-IR 379375. 249
portfolio was tainted, its policy resulted in its accounting for securities as AFS when they had been designated "at acquisition" as HTM. The record we have compiled supports the conclusion that Fannie Mae used the flexibility inherent in its approach to FAS 115 in a fashion that had tangible benefits for the Company. The record does not, however, support the conclusion that Fannie Mae used this approach to manage earnings directly. Ultimately, the characterization of the securities as HTM or AFS had an effect on Fannie Mae's financial statements when the securities in the AFS portfolio were sold. KPMG was aware of the Company's accounting policy and the IMRD procedures, but did not object to either. It also acknowledged, and did not object to, the use of the IMRD process to account for "balance sheet effects" or to take advantage of an "economic opportunity." PART E: RECOGNITION OF INTEREST EXPENSE AND INCOME I. INTRODUCTION In this Part, we address Fannie Mae's practice of "smoothing" the monthly interest expense and income on certain investments and borrowings. Until early in 2003, the Company's liquid investment portfolio ("LIP") and debt accounting systems (known as ORION and STAR, respectively) determined interest expense and income on all investments and debt instruments as if there were 30.4 days in each month (a "30/360" convention), even if the instruments' terms required interest payments on either an "actual/365" or "actual/360" basis. As a result of this practice, Fannie Mae avoided the fluctuations in interest income and expense that would result from the fact that the twelve months of the year (and the four calendar quarters) do not have the same number of days.(945) The Company discontinued this practice in the second quarter of 2003 when it introduced a new version of the STAR system called iSTAR. At that point, the Company began to recognize interest income and expense in accordance with the actual terms of the instruments. Fannie Mae's smoothing of interest income and expense came to our attention during the investigation when we uncovered Cheryl DeFlorimonte's March 3, 2003 memorandum discussed below. This issue subsequently was included as a topic for our review in OFHEO's February 11, 2005 letter to Stephen B. Ashley, in which OFHEO asked "why the Enterprise believed it was permitted to recognize monthly interest - ---------- (945) Because the Company's LIP portfolio is relatively small, our inquiry primarily focused on the recognition of interest expense on the Company's debt portfolio. 250
expense on notes payable, in a monthly ratable manner that is different from the contractually specified interest of the notes."(946) As OFHEO's question suggests, Fannie Mae's accounting for interest income and expense on certain arrangements prior to the second quarter of 2003 was problematic for several reasons. As a general proposition, GAAP requires the accounting for interest income and expense to follow the terms of the relevant instruments. Thus, Fannie Mae should have accounted for these investments and borrowings by recognizing interest income and expense in accordance with the legal terms of those arrangements, regardless of the fact that such treatment would generate fluctuations in the recognition of interest income and expense each month. By smoothing interest income and expense from month to month and from quarter to quarter, Fannie Mae misstated its actual net interest income during each period. The Company's accounting during the period prior to the second quarter of 2003, therefore, violated GAAP. II. BACKGROUND The origins of Fannie Mae's practice of smoothing the recognition of interest income and expense date to the 1970s. According to Richard Stawarz, who joined Financial Reporting at that time, the proposition that Fannie Mae should smooth its interest income expense was raised by the Company's then Chief Economist, who wanted to eliminate calendar effects from the stated financial results. Initially, the smoothing was accomplished manually. When the calculation of interest expense was automated into the Company's STAR system, an algorithm to automate the smoothing process was incorporated into the system. The purpose of the smoothing function was known to those who joined Financial Reporting later. A document prepared in 2001 in connection with proposed revisions to the STAR system states that "[t]his process was implemented to smooth out the bumps in interest expense (income) [sic] created by having varying [sic] number of days in each month."(947) Janet L. Pennewell stated that the practice began when the Company was publishing its results on a monthly basis and wanted to avoid variations in reported interest expense that stemmed only from the different number of days from month-to-month. Another concern appeared to be that reported interest income and expense would fluctuate from period to period, while Fannie Mae's mortgage-based revenues typically followed a 30/360 convention (that is, interest paid to investors in mortgage-backed securities is determined as if each month has 30 days). By following a similar convention with respect to net interest expense, Fannie Mae's interest income and expense would eliminate the fluctuations based on the calendar, aligning the recognition of interest expense with the recognition of interest income on a majority of the Company's mortgage portfolio. - ---------- (946) February 11, 2005 OFHEO Letter at FMSE-IR 547326 (emphasis removed). (947) iSTAR - Operations Requirements as of April 17, 2001, Zantaz document 1456873, at 27. 251
The 30/360 day convention was not required by Fannie Mae's accounting policies, which do not refer to a process for monthly smoothing of interest income or interest expense. Fannie Mae's accounting policy for interest income states, "Fannie Mae recognizes interest income in the period that interest is due and records an offsetting entry to accrued interest receivable or cash."(948) Similarly, the applicable policy for interest expense confirms, "Fannie Mae follows standard accrual accounting practices. Fannie Mae accrues interest on a monthly basis, recording a debit to interest expense and a credit to interest payable."(949) In this respect, neither the ORION nor the STAR systems followed the accounting policy, as they did not accurately accrue and recognize interest income in the period in which interest was due or follow "standard accrual accounting policies." In addition to the automatic smoothing accomplished by the Company's systems, the Company also periodically accrued additional interest expense through an on-top journal entry; the additional expense was then amortized over the remainder of the year. For example, the Company recognized an additional expense of $13.5 million in February 2001.(950) In February 2003, the Company recognized an additional $15.4 million in interest expense "to increase debt expense in February to 30.4 days from 28 days."(951) In May 2003, the Company began to amortize that expense into income.(952) The Company posted an on-top interest expense adjustment in March 2002, rather than February; Stawarz stated that this calculation was especially complex, and therefore was delayed by a month.(953) According to Pennewell, the February adjustments were recorded because the smoothing algorithms in the STAR system were viewed as unreliable. An on-top adjustment would be made if the system's results appeared to be inconsistent with - ---------- (948) Interest Income and Accrued Interest Receivable, Financial Accounting Policy Guide, dated Oct. 2003, FMSE-SP 78953-9118, at FMSE-SP 78980. (949) Id. at FMSE-SP 79049. The 1994 version of the policy is similar. (950) Homesite General Ledger for account 162200, Journal ID JE21125 (entry for "Fractional Day Impact Adj.," effective date Feb. 28, 2001). (951) Undated May 2003 Closing/Activity Notes, FMSE-SP 10854. (952) Id. The Company delayed the amortization of the February interest expense overstatement until May to coincide with the recognition of an adjustment associated with the conversion of the STAR system to the iSTAR system. See Undated March 2003 Closing Activity/Notes, FMSE-SP 10857. (953) We have found no evidence that the Company recorded similar entries to adjust the amount of interest income determined by the ORION system. 252
expectations or rough calculations of what the actual amounts should be.(954) Stawarz confirmed that view. He said that those responsible for preparing Fannie Mae's income statements understood that the STAR system results for February were problematic, presumably because February was such a short month.(955) Stawarz was responsible for evaluating the February results against an estimate and determining the amount of the on-top adjustment for that month. According to Stawarz, he calculated the February on-top adjustment on a handwritten work sheet that he did not keep. Fannie Mae's use of the 30.4-day convention for recognizing interest income and expense resulted in unusual aberrations in the Company's annual reported interest expense. One would expect that the approach the Company adopted would affect intra-year accounting, but would not affect the Company's end-of-year results. That is, the Company's monthly and quarterly results would reflect a greater or lesser amount of interest expense than would be the case if actual expense were recorded, but that any differences would be cancelled out in year-end tabulations. An analysis of results for 2002 prepared by Mona Patel in 2003, however, shows that smoothing had the effect of reducing interest expense (and increasing net income) by approximately $27.5 million in 2002.(956) DeFlorimonte, who prepared the memorandum to which the analysis was attached and who, during the relevant period, was responsible for the Company's debt accounting, could not explain, and may not have been aware of the apparent discrepancy.(957) A review of the smoothing methodology showed that a discrepancy did, in fact, exist. This discrepancy appears to be the result of how the ORION and STAR systems calculated the monthly interest to be recognized. A diagram prepared by personnel in the Debt Accounting Group showed that smoothing was accomplished through the system by: 1. determining the total amount of interest to be earned (incurred) over the life of the security (borrowing); - ---------- (954) E-mail from Janet L. Pennewell to Mary Lewers, Cheryl DeFlorimonte, et al., dated Nov. 4, 2002, FMSE-SP 10799. (955) According to Richard Stawarz, even Timothy Howard knew that the reporting systems were less reliable and yielded unusual results in February. (956) Mem. from C. DeFlorimonte to Distribution, dated Mar. 3, 2003, FMSE 55102-12, at FMSE 55107. (957) A separate document showing the effect of smoothing an interest income from the LIP portfolio reveals the same effect. See Fannie Mae Accrual Example Coupon Expense Smoothing Actual Day Accruals, dated Feb. 21, 2002, Zantaz document 1154886. 253
2. determining the implied coupon on the security (borrowing) as if its contractual terms required the monthly accrual of interest on the basis of 30.4 days; and 3. calculating the amount of deferred interest to be recognized monthly in order to adjust the interest income (expense) accrued on the basis of the contractual terms of the security (borrowing) to the smoothed amounts based on the implied coupon.(958) Over the life of the investment or borrowing, the smoothing process would have no impact on the cumulative income or expense recognized. Nor would smoothing affect amount of interest income or expense recognized on a particular investment or borrowing on a year-to-year basis. However, because the Company effectively smoothed interest income and expense over the life of the instrument and not over the calendar year, the process resulted in discrepancies in the calculation of net interest income and expense at year-end. In sum, if the anniversary date of the acquisition of an investment or issuance of a borrowing were other than January 1, smoothing would result in the recognition of either too much or not enough interest at the end of an annual reporting period. In Stawarz's view, however, the impact on the Company's quarterly or annual results would have been negligible, as a large portion of the Company's debt was short-term debt with terms of less than a year.(959) Documentation relating to revisions of the STAR system showed that the Company considered abandonment of the 30/360 approach as early as 1999. A document dated October 13, 1999 captioned "STAR+ Calculations Re-write Analysis" proposed to "[r]emove smoothing from the accounting cost calculations, except for step-up securities."(960) A 2001 document described the rationale for the proposed change, and emphasized that the smoothing process was more complicated than booking interest in accordance with the terms of each instrument: - ---------- (958) Controller's - Debt Accounting Group: Overview to Paul Weiss, LLP, dated Dec. 19, 2005, FMSE-IR 703580-99, at FMSE-IR 703594. (959) Further, if the Company processed an amortization end-date change at a time when a balance in the deferred interest account existed, that balance would subsequently only be reduced to zero over the remaining estimated life of the borrowing (i.e., it would no longer be reduced to zero at the instrument's anniversary date). Amortization end-date changes are discussed in Chapter VI, Part M of this Report. (960) Star+ Calculations Re-write Analysis, dated Oct. 13, 1999, Zantaz document 92466, at 1. 254
Currently, an implied coupon and deferred interest are calculated in cases where the day count convention is not 30/360.... The implied coupon was used in the subsequent level yield and amortization programs. This practice caused numerous issues in amortization of deferred fees, especially with short-term instruments. There were many instances of over amortization in whole months and under amortization (some times [sic] reverse amortization) in partial months. Since this practice (smoothing) caused many subsequent issues, it has been decided not to implement it in the new STAR operations except for the case of step-up securities.(961) The proposal made clear, however, that the elimination of the smoothing function in STAR would not necessarily foreclose use of the 30.4-day convention: "The requirement to smooth out expense/income bumps due to the numbers of days in a month still exists, but will be done outside the system (except for step-up (down) securities)."(962) Final approval to discontinue the smoothing process was not granted formally until 2003, right before the iSTAR system replaced STAR. In her March 3, 2003 memorandum to members of Controller's Office and in the Portfolio Management group,(963) DeFlorimonte proposed to "[d]iscontinue the calculation of and reporting on smoothing monthly expenses and average balances in our debt statistics." (964) According to DeFlorimonte, the elimination of smoothing would "simplify calculations and allow all reconciliation of debt expenses, balances and costs to be more automated and transparent. It will also eliminate the interest free distortions that result from smoothing."(965) DeFlorimonte concluded by noting, "if smoothing is still desired, this will be accomplished by a monthly, quarterly, or other frequency to be determined, on-top entry." (966) Although the memorandum did not receive a formal response, the proposal to - ---------- (961) iSTAR - Operations Requirements as of April 17, 2001, Zantaz document 1456873, at 27-28. (962) Id. at 28. (963) Mem. from C. DeFlorimonte to J. Pennewell, Jonathan Boyles, et al., dated Mar. 3, 2003, FMSE 55102-03, at FMSE 55102. The memorandum does not appear to have prompted a reaction from Boyles or Paul Salfi regarding the consistency of smoothing with Fannie Mae's policy or with GAAP. (964) Id. (965) Id. (966) Id. 255
eliminate smoothing from the iSTAR system was approved following authorization from Timothy Howard.(967) According to DeFlorimonte, the iSTAR system currently recognizes debt expense as it accrues in accordance with the terms of the debt instruments.(968) Documentation from 2004 shows that the Company considered manual on-top entries to smooth the results of the iSTAR system, but did not post those entries. We have seen no indication that the Company has smoothed its interest expense since the introduction of the iSTAR system in 2003. According to Stawarz, those responsible for the Company's financial reporting, up to and including Howard, were aware that the Company was smoothing interest income and expense. Howard, Leanne G. Spencer and others were aware at least as of 2003 that the Company was discontinuing the practice. We have seen no evidence that other members of senior management were aware of this issue prior to that time. Notably, we see no indication that Financial Standards was asked to set policy on this issue, or that this policy was ever raised with the Board or any of its committees. KPMG was aware of the smoothing process at least as of July 2003. (969) At that time, KPMG reviewed the Company's general ledger account 1022, which included a portion of the accrued interest expense that had not been booked.(970) The workpaper describes the smoothing process in considerable detail.(971) The KPMG partner on the audit team at that time confirmed during our interview that he was aware of the process. - ---------- (967) Leanne G. Spencer, in an e-mail dated May 22, 2003, to Pennewell, Stawarz, and Lewers, notes, "I did inform Tim yesterday that we were officially stopping 'smoothing'. He was fine with the decision. Told him we would revisit - if conditions warranted in the future." E-mail from L. Spencer to J. Pennewell, et al., dated May 22, 2003, Zantaz document 944953, at 1. In Stawarz's view, the process was discontinued because interest rates were so low and much of the Company's debt was "swapped out" in hedge relationships. Accordingly, the process was not worth the effort. (968) Mem. from C. DeFlorimonte to Distribution, dated Oct. 8, 2004, Zantaz document 202983, at 1. (969) Analysis of Accounts Receivable - No. 162200, dated July 2003. (970) Id. (971) Id. The document refers to an adjustment that had been necessary "to appropriately maintain the books on a 30/360 day convention for certain debt securities, including short-term notes." Id. 256
III. FINDINGS REGARDING THE RECOGNITION OF INTEREST EXPENSE AND INCOME In sum, senior accountants in the Controller's Office - including Spencer, Pennewell, and Stawarz - knew at some point that the Company was not reporting interest expense and income accurately. Although the documentary evidence is scant, Howard may have been aware as well. Moreover, the Controller's Office apparently did not question the periodic on-top adjustments or whether the supposed need for the adjustments meant that the smoothing process was incorrect or unreliable generally. This process was not required by Company policy, nor was Financial Standards asked for its views, at least until the system was changed in 2003. Finally, it is notable as well that the practice continued for four years between the time it was proposed to eliminate smoothing to the cessation of the practice in 2003, and that both the removal of the smoothing mechanism and the delay in the removal stemmed from practical considerations rather than a concern for proper accounting or accurate reporting. PART F: ACCOUNTING FOR OTHER-THAN-TEMPORARY IMPAIRMENT OF MANUFACTURED HOUSING BONDS AND AIRCRAFT ASSET-BACKED SECURITIES I. INTRODUCTION In this section, we address Fannie Mae's accounting for other-than-temporary impairment ("OTTI") of investments in manufactured housing bonds ("MH bonds") and aircraft asset-backed securities ("Aircraft ABS"). Under GAAP, Fannie Mae is required to evaluate all of its Held-to-Maturity ("HTM") and Available-for-Sale ("AFS") investments for OTTI. Evidence showed that management failed to consistently monitor and adequately document its review of all HTM and AFS investments for OTTI prior to 2003. Although management did begin to consistently monitor its HTM and AFS investments during 2003 through the formation of an Impairment Committee ("IC"), the IC's review was not inclusive of all HTM and AFS investments. Insufficient monitoring of investments both before and after the formation of the IC represents a control weakness that could indicate underreporting of OTTI. Furthermore, prior to 2004, management relied primarily on discounted cash flow ("DCF") models in place of quoted market prices to measure impairment on MH bonds and Aircraft ABS to determine the OTTI amounts it recognized, even though bid/ask and dealer pricing was available. According to GAAP, however, quoted market prices are the best evidence of fair value, and DCF models should only be used to measure fair value in the absence of quoted market prices. While we did not find that management chose to rely on DCF modeling in order to achieve particular earnings goals, we note that the DCF model results included assumptions that were subject to management discretion and data errors that impacted both the timing and the amount of OTTI the Company recorded. We believe that the Company should have procedures and 257
controls that ensure that the assessment of OTTI is not left to management's discretion without appropriate guidelines. In the second quarter of 2004, after discussions with OFHEO and the SEC, Fannie Mae implemented a new approach to evaluating MH bonds and Aircraft ABS for OTTI. The "4-Step Approach" requires the use of credit ratings and market prices to determine the existence of OTTI, even when Fannie Mae does not expect a loss of principal or interest. The 4-Step Approach also requires Fannie Mae to recognize income subsequent to an impairment write-down if its estimate of future cash flows warrants such recognition. The adoption of the 4-Step Approach caused a significant increase in OTTI recognized, particularly for MH bonds. II. BACKGROUND A. Applicable Accounting Standards FASB Statement No. 115, Accounting for Certain Investments in Debt and Equity Securities ("FAS 115"),(972) requires companies to evaluate HTM and AFS securities for impairment and to write those securities down to fair value if other than temporarily impaired: If it is probable that the investor will be unable to collect all amounts due according to the contractual terms of a debt security not impaired at acquisition, an other-than-temporary impairment shall be considered to have occurred. If the decline in fair value is judged to be other than temporary, the cost basis of the individual security shall be written down to fair value as a new cost basis and the amount of the write-down shall be included in earnings (that is, accounted for as a realized loss).(973) SEC Staff Accounting Bulletin Topic 5-M, Other Than Temporary Impairment of Certain Investments in Debt and Equity Securities ("SAB Topic 5-M"), was issued to clarify the Staff's views on accounting for noncurrent marketable securities, specifically as they relate to determining whether declines in fair value are other than temporary. In doing so, SAB Topic 5-M provides guidance to public companies on the factors that companies should consider in determining whether such securities are other than temporarily impaired, such as "the length of the time and the extent to which the market value has been less than cost," and "the financial condition and near-term prospects of the issuer, including any specific events which may influence - ---------- (972) ACCOUNTING FOR CERTAIN INVESTMENTS IN DEBT AND EQUITY SECURITIES, Statement of Fin. Accounting Standards No. 115 (Fin. Accounting Standards Bd. 1993) (hereinafter "FAS 115"). (973) See id. P 16. 258
the operations of the issuer" or, "the intent and ability of the holder to retain its investment in the issuer for a period of time sufficient to allow for any anticipated recovery in market value."(974) Paragraphs Eight and Ten of FASB Emerging Issues Task Force ("EITF") Issue No. 03-01, The Meaning of Other-Than-Temporary Impairment and Its Application to Certain Investments ("EITF 03-01"), provide extensive guidance on factors to consider in evaluating securities for impairment.(975) Once OTTI has been identified, the investment must be written down to fair value by a charge to earnings in the present period. FAS 115 defines fair value as: The amount at which an asset could be bought or sold in a current transaction between willing parties, that is, other than in a forced or liquidation sale. Quoted market prices in active markets are the best evidence of fair value and should be used as the basis for the measurement, if available. If a quoted market price is not available, the estimate of fair value should be based on the best information available in the circumstances.(976) As reflected in the above definition, a company has some flexibility in determining fair value. Nevertheless, it is clear that fair value based on quoted market prices is preferred.(977) After OTTI has been recorded, GAAP does not allow for the investment to be "written-up" if the fair value of the investment recovers. However, GAAP does - ---------- (974) SEC Staff Accounting Bulletin, Topic 5-M, Other than Temporary Impairment of Certain Investments in Debt and Equity Securities (2004), available at http://www.sec.gov/interps/account/sabcodet5.htm#5m. (975) THE MEANING OF OTHER-THAN-TEMPORARY IMPAIRMENT AND ITS APPLICATION TO CERTAIN INVESTMENTS, Emerging Issues Task Force Issue 03-01 (2004). We note that FASB Staff Position FAS 115-1 and FAS 124-1, THE MEANING OF OTHER-THAN-TEMPORARY IMPAIRMENT AND ITS APPLICATION TO CERTAIN INVESTMENTS, ("FSP FAS 115-1 and FAS 124-1") (Nov. 3, 2005), nullified certain impairment evaluation guidance in EITF 03-01 that was not already contained in FAS 115 or SAB Topic 5-M, but carried forward guidance in EITF 03-01 regarding determination and disclosure. (976) FAS 115 P 137 (as amended by ACCOUNTING FOR DERIVATIVE INSTRUMENTS AND HEDGING ACTIVITIES, Statement of Fin. Accounting Standards No. 133 (Fin. Accounting Standards Bd. 1998)). (977) FAS 115 PP 110-11. 259
provide for subsequent income recognition if the entity's expectation of receiving cash flows has significantly increased. American Institute of Certified Public Accountants ("AICPA") Statement of Position No. 03-3, Accounting for Certain Loans or Debt Securities Acquired in a Transfer ("SOP 03-3"), states: [If, upon subsequent evaluation,] [b]ased on current information and events, it is probable that there is a significant increase in cash flows previously expected to be collected or if actual cash flows are significantly greater than cash flows previously expected, the investor should recalculate the amount of accretable yield for the loan as the excess of the revised cash flows expected to be collected over the sum of (1) the initial investment less (2) cash collected less (3) other-than-temporary impairments plus (4) amount of yield accreted to date.(978) B. Fannie Mae's Practice and Policy 1. OTTI Recorded The majority of the OTTI that Fannie Mae recorded between 2001 and the second quarter of 2004 related to MH bonds and Aircraft ABS. Our investigation has focused on the accounting for those securities during that period because of the significance of their contribution to the total amount of recognized impairments. The following tables summarize the book and fair values of Fannie Mae's HTM and AFS investments, unrealized losses on those investments, and the OTTI recorded during that period (dollars in millions): Table I - Mortgage and Non-Mortgage HTM and AFS Investments(979) DECEMBER 31, 2001 DECEMBER 31, 2002 DECEMBER 31, 2003 JUNE 30, 2004 ----------------- ----------------- ----------------- ----------------- FAS 115 BOOK FAIR BOOK FAIR BOOK FAIR BOOK FAIR CLASSIFICATION VALUE VALUE VALUE VALUE VALUE VALUE VALUE VALUE - -------------- ------- ------- ------- ------- ------- ------- ------- ------- HTM 539,254 546,075 440,431 458,624 481,974 492,916 458,257 460,028 AFS 68,357 68,783 203,947 210,500 239,900 243,466 241,255 242,329 - ---------- (978) ACCOUNTING FOR CERTAIN LOANS OR DEBT SECURITIES ACQUIRED IN A TRANSFER, Statement of Position No. 03-3 P .07(b) (Am. Inst. of Certified Pub. Accountants 2003). (979) See Fannie Mae, 2003 Annual Report (Form 10-K), at 59-60, 64-65 (Mar. 15, 2004), available at http://www.fanniemae.com/ir/pdf/sec/2004/f10k03152004.pdf (hereinafter "2003 Form 10-K"); Fannie Mae, 2004 Quarterly Report (Form 10-Q), at 52-55 (Aug. 9, 2004), available at http://www.fanniemae.com/ir/sec/index.jhtml?s=SEC+Filings (follow "All SEC Filings" hyperlink; then follow "Quarterly Filings" hyperlink for PDF format) (hereinafter "2004 2Q Form 10-Q"). 260
Table II - Unrealized Losses on HTM and AFS Investments(980) DECEMBER DECEMBER DECEMBER JUNE 30, FAS 115 CLASSIFICATION 31, 2001 31, 2002 31, 2003 2004 - ---------------------- -------- -------- -------- -------- HTM (721) (133) (1,461) (4,829) AFS (349) (482) (726) (1,441) ------ ---- ------ ------ TOTAL (1,070) (615) (2,187) (6,270) ====== ==== ====== ====== Table III - OTTI Recorded YEAR ENDED YEAR ENDED YEAR ENDED SIX MONTHS ENDED INVESTMENT TYPE 2001 2002 2003 JUNE 30, 2004 - --------------- ---------- ---------- ---------- ---------------- MH BONDS(981) -- (51) (155) (261) AIRCRAFT ABS(982) -- -- (84) (36) OTHER INVESTMENTS(983) (50) (59) (69) -- --- ---- ---- ---- TOTAL(984) (50) (110) (308) (297) === ==== ==== ==== - ---------- (980) See id. (981) Homesite Journal Entry, Journal ID JE29222, dated Dec. 28, 2002; Homesite Journal Entry, Journal ID JE29244, dated Jan. 28, 2003; Homesite Journal Entry, Journal ID JE29314, dated Mar. 28, 2003; Homesite Journal Entry, Journal ID JE49390, dated Sept. 28, 2003; Homesite Journal Entry, Journal ID JE21633, dated Dec. 28, 2003; Homesite Journal Entry, Journal ID JE49139, Mar. 28, 2004; Homesite Journal Entry, Journal ID JE49160, dated Apr. 28, 2004; Homesite Journal Entry, Journal ID JE49320, dated June 28, 2004. (982) Homesite Journal Entry, Journal ID JE29786, dated June 28, 2003; Homesite Journal Entry, Journal ID JE29479, dated Feb. 28, 2004; Homesite Journal Entry, Journal ID JE29316, dated June 28, 2004. (983) Amounts were computed as the difference between total OTTI recorded and MH/Aircraft ABS OTTI. It appears that most "Other Investments" (briefly described below) are not investments subject to FAS 115, but still require evaluation for impairment per the applicable GAAP. Our review focused on OTTI recorded pursuant to FAS 115, specifically that for MH bonds and Aircraft ABS. (984) 2003 Form 10-K at 45-46; 2004 2Q Form 10-Q at 51-52. 261
2. Types of Investments As shown in Table III above, the majority of the OTTI that Fannie Mae recorded between 2001 and the second quarter of 2004 related to MH bonds and Aircraft ABS. (a) MH Bonds Fannie Mae began investing in MH bonds in the mid-1990s in order to satisfy HUD goals and to provide liquidity for loans extended to low and moderate income borrowers.(985) The MH bonds in which Fannie Mae invests are mortgage-related securities backed by manufactured housing loans issued by entities other than Fannie Mae.(986) The Company's MH bond investments were substantial - the portfolio's book value was $10 billion by the end of 2002.(987) Most of Fannie Mae's MH bonds were issued and serviced by Conseco Finance Corporation ("Conseco").(988) Fannie Mae began to closely monitor Conseco toward the end of 2001 as a result of Conseco's financial troubles and potential for bankruptcy,(989) and management noted in early 2002 that the Company could experience losses if the baseline default rate increased.(990) Despite Conseco's bankruptcy on December 17, 2002,(991) management viewed its exposure to be immaterial primarily because it held senior interests, and because "all but one" of the bonds were still rated as investment grade.(992) Fannie Mae - ---------- (985) E-mail from Adolfo Marzol to Charles V. Greener, dated Feb. 20, 2004, Cataphora document i.1077559617000.m278273054, at 1. (986) 2003 Form 10-K at 60. (987) See id. Fannie Mae notes in its 10-K that approximately $0.6 billion of the $10 billion were securities that were guaranteed by Fannie Mae. (988) See Draft mem. from David Kightlinger, et al., to A. Marzol, et al., with attached appendices, dated Sept. 2002, FMSE 106020-27, at FMSE 106021. Other servicers included Oakwood Capital Management and Greenpoint Capital. Id. (989) See Mem. from A. Marzol to Daniel H. Mudd, et al., dated Dec. 18, 2001, FMSE 107000-01, at FMSE 107000. (990) Mem. from Jef Kinney to Timothy Howard, dated Mar. 27, 2002, FMSE 165026-31, at FMSE 165026-27. (991) E-mail from Jeffrey Wise to Linda K. Knight, et al., dated Dec. 18, 2002, Cataphora document i.1040210632000.748582297, at 1. (992) Talking Points - Conseco Bankruptcy, dated Dec. 20, 2002, FMSE-E 2038484. 262
first recognized OTTI on MH bonds in the fourth quarter of 2002.(993) We have found no evidence that Fannie Mae recognized OTTI on MH bonds in any period prior to that date. (b) Aircraft ABS According to a March 14, 2003 draft memorandum from Paul Salfi in Financial Standards, Fannie Mae began investing in Aircraft ABS in the late 1990s with the intent to hold the assets in its Liquid Investment Portfolio ("LIP") for two to five years.(994) The Aircraft ABS in which Fannie Mae invests are trusts that have legal title to an underlying pool of aircraft. Repayments to the ABS bondholders are primarily the cash flows from leasing the aircraft to carriers or, secondarily, from the sale of the airplane itself.(995) By February 2003, Fannie Mae's book value of Aircraft ABS was approximately $440 million.(996) The March 14, 2003 draft memorandum cited "[t]he events of September 11th coupled with the frail economy, high fuel prices, fear of terrorism and war, and increased security costs" as reasons for the decline of the airline industry in 2002.(997) According to Jonathan Boyles, the potential impairment of the Aircraft ABS was a primary driver for organizing a formal IC in 2003. Fannie Mae first recorded OTTI on its Aircraft ABS in June 2003.(998) (c) Other Investments As shown in Table III, above, although the majority of OTTI recognized between 2001 and the second quarter of 2004 was for MH bonds and Aircraft ABS, Fannie Mae recognized OTTI on other investments during this period as well. These investments appear to have been composed largely of investments not subject to FAS 115, such as investments in low income housing partnerships, investments in alternative energy partnerships, equity investments in various housing and community development - ---------- (993) See Homesite Journal Entry for account 713125, Journal ID JE29222, dated Dec. 28, 2002. (994) Draft mem. from Paul Salfi, et al. to Ken Barnes, et al., dated Mar. 14, 2003, FMSE-IR 672690-99, at FMSE-IR 672690-91. (995) Id. at FMSE-IR 672691. (996) Id. at FMSE-IR 672690. (997) Id. at FMSE-IR 672691. (998) See Homesite Journal Entries for accounts 713127, Journal ID JE29786, dated June 28, 2003. 263
institutions, equity investments in technology companies, and an investment in an aircraft leverage lease.(999) 3. OTTI Accounting and Monitoring Methodologies (a) Prior to the 4-Step Approach Fannie Mae's 2001 Financial Accounting Policy Guide contained a brief mention of OTTI, and contained no guidance on the factors that companies should consider in determining whether such securities are impaired. It stated: If the fair value of a HTM or AFS security is less than its carrying amount and the decline is judged to be other than temporary: - Fannie Mae reduces the carrying amount of the investment to its fair value, which is reflected as the new "cost" basis - The write-down is recorded in earnings as a realized loss - The new costs basis is not adjusted for subsequent recoveries in fair value.(1000) In a memorandum dated September 27, 2002 to Leanne Spencer and Mortgage Portfolio personnel, Ilan Sussan addressed OTTI on AFS securities and sought "to describe indicators that should be considered in assessing whether securities are other-than-temporarily impaired due to a decline in their market values."(1001) The memorandum cited FAS 115, SAB Topic 5-M, and AICPA guidance on OTTI assessment, and stated: "If any securities that meet any of these criteria come to your attention, please inform Jonathan Boyles or myself immediately. We will weigh all of these factors to determine if a security was permanently impaired."(1002) Beyond these two documents, we did not find any formal accounting policy for evaluating or recognizing OTTI. Fannie Mae first formed an IC to evaluate investments for OTTI in mid-2003.(1003) The IC Charter ("Charter"), as drafted by Boyles, called for membership by the - ---------- (999) See Fannie Mae's Impairment Policies and Practices, Assets & Liabilities Policy Committee, dated Apr. 20, 2004, FMSE-IR 501826-34. (1000) Financial Accounting Policy Guide, dated Dec. 2001, Zantaz document 1044810, at 43. (1001) Mem. from Ilan Sussan to Leanne Spencer, dated Sept. 27, 2002. (1002) Id. (1003) See Mem. from Jonathan Boyles to L. Spencer, with attachments, dated May 28, 2003, FMSE 69379-83, at FMSE 69379. 264
Chief Credit Officer, the Controller, the Treasurer, the General Counsel, and a Vice President from both Financial Standards and Corporate Tax.(1004) The Charter appears to be consistent with GAAP requirements for OTTI determination and describes review mechanisms that the IC and the business owners of the investments should use to monitor investments. These mechanisms include a "Watch List" of all investments with fair values at or below ninety-five percent of carrying value and a "Potential Impairment Report" describing each at-risk investment and the nature of the impairment risk.(1005) It appears that the IC only monitored "non-agency" (i.e., excluding securities issued by Fannie Mae, Freddie Mac, and other GSEs) and LIP investments, including Aircraft ABS but excluding MH bonds.(1006) While excluded from review by the IC, MH bonds were analyzed for loss by Credit Policy beginning in 2002, after the effects of a potential Conseco bankruptcy on the MH bond market became apparent.(1007) Credit Policy's analysis used DCF modeling to determine the present value of potential losses on MH bonds that were, in turn, used to determine OTTI.(1008) Even though we found no formally documented criteria for evaluating MH bonds for OTTI specifically, we did find some guidance in a January 8, 2004 e-mail. The e-mail described the criteria as "low mark-to-market (less - ---------- (1004) Undated Impairment Committee, Committee Charter, FMSE 130548-51, at FMSE 130548. (1005) Id. at FMSE 130549. (1006) Mem. from Joanne Collins to L. Spencer, dated Oct. 20, 2003, FMSE 69340-69346, at FMSE 69340. Collins writes: "Attached are the Watch Lists for the Non Agency Securities and the LIP. The Non Agency MBS Watch List includes mostly manufactured housing securities that are monitored for impairment by another committee." Id. This language appears to have been boiler plate for the IC recommendations prior to the implementation of the 4-Step Approach. See, e.g., Mem. from J. Collins to L. Spencer, dated Dec. 22, 2003, FMSE 69328-36 at FMSE 69328; Mem. from I. Sussan to L. Spencer, dated Feb. 2, 2004, FMSE 164887-91, at FMSE 164887. (1007) See Draft mem. from D. Kightlinger, et al., to A. Marzol, et al., with attached appendices, dated Sept. 2002, FMSE 106020-27 (analyzing MH bond exposure); see also Mem. from A. Marzol to D. Mudd, et al., dated Dec. 18, 2001, FMSE 107000-01 (same). (1008) See, e.g., Fannie Mae Exposure to Conseco Manufactured Housing Bonds spreadsheet, dated Mar. 27, 2003, FMSE 106388-89, at FMSE 106388. The total present value of losses listed on the spreadsheet matches the cumulative amount of MH bond OTTI recorded as of March 31, 2003. See Homesite Journal Entries for account 713125, Journal ID JE29222, dated Dec. 28, 2002 ($51 million); Journal ID JE29244, dated Jan. 28, 2003 ($1.15 million); and Journal ID JE29314, dated Mar. 28, 2003 ($66.25 million). 265
than 90 percent of par), projected losses occurring within a reasonable time period (20 years), materiality of projected loss or materiality of change in projection from previous quarter, and likelihood that loss will occur."(1009) Fannie Mae's first comprehensive investment impairment policy, including the protocols described in the Charter, appeared in the Company's December 2003 Financial Accounting Policy Guide.(1010) (b) The 4-Step Approach In 2004, OFHEO reviewed Fannie Mae's OTTI accounting for MH bonds in the course of the Special Examination.(1011) According to Boyles, OFHEO believed that the provisions of EITF 99-20, Recognition of Interest Income and Impairment on Purchased and Retained Beneficial Interests in Securitized Financial Assets, rather than FAS 115, applied to these investments. Fannie Mae disagreed, and brought its disagreement to the SEC. Although the SEC staff did not object to the Company's accounting policy, it suggested that Fannie Mae work with OFHEO to draft a policy upon which both could agree. As a result, Fannie Mae revised its OTTI evaluation of MH bonds and Aircraft ABS effective the second quarter of 2004. Fannie Mae memorialized its new approach, known as the "4-Step Approach," in a June 22, 2004 letter to OFHEO,(1012) and communicated the 4-Step Approach to the SEC in a letter dated August 3, 2004.(1013) - ---------- (1009) E-mail from D. Kightlinger to Kieran Gifford, et al., dated Jan. 8, 2004, Zantaz document 1004217. (1010) See Financial Accounting Policy Guide, Asset Impairment Review, dated Dec. 2003, FMSE-IR 511818-21. (1011) See Talking Points to Board, dated Apr. 20, 2004, FMSE-IR 227241-42, at FMSE-IR 227242. (1012) Letter from J. Boyles to Christopher H. Dickerson, with attachments, dated June 22, 2004, FMSE-SP 42125-34, at FMSE-SP 42125-26. (1013) Letter from J. Boyles to Donald T. Nicolaisen, with attachments, dated Aug. 3, 2004, FMSE-IR 362436-41 at FMSE-IR 362437. In addition to communicating the new 4-Step Approach, the letter refers to an April 28, 2004 letter, and states: [Y]ou [SEC] told us [Fannie Mae] that you believed our accounting policies regarding other-than-temporary impairment were consistent with generally accepted accounting principles. You also stated that you believed that the approach by our regulators had value and that we should work with them and our external auditors to settle 266
Boyles stated that the policy represented a compromise between OFHEO's demands and what the Company viewed as appropriate accounting under GAAP. The 4-Step Approach, as documented in the above memorandum, is divided into "Day One" and "Day Two" accounting.(1014) (1) Day One Accounting Day One accounting addresses the threshold question of whether an investment is OTTI. The steps are as follows: Step 1: Does Fannie Mae's estimate of cash flows show loss of principal or interest? If yes--write-down to fair value If no--move to step 2 Step 2: Has the bond been downgraded to BB or lower? If yes--write-down to fair value If no--move to step 3 Step 3: Has the bond been downgraded to BBB and trading below 90 of adjusted book? If yes--write-down to fair value If no--move to step 4 Step 4: Is the bond is [sic] A-rated or better and trading below 80 of adjusted book value? If yes--write down to fair value If no--mark-to-market through OCI(1015) - ---------- the accounting issues and come back to you with our final agreement. Id. at FMSE-IR 362436; See also letter from J. Boyles to D. Nicolaisen, dated Apr. 28, 2004, FMSE-IR 483271-82. This document can be found in the accompanying Appendix at Tab D.6. Boyles explained that his summary of the SEC's position, above, was limited to the question of whether to apply FAS 115 instead of EITF 99-20. He did not believe that the SEC had given broad approval of Fannie Mae's OTTI accounting practices. (1014) Letter from J. Boyles to C. Dickerson, with attachments, dated June 22, 2004, FMSE-SP 42125-34, at FMSE-SP 42125-26. (1015) Id. (footnotes omitted). 267
Step One encompasses Fannie Mae's prior policy for determining whether an investment was impaired on an other-than-temporary basis. The additional three steps respond to OFHEO's demand that the Company use changes in credit ratings and market prices as primary indicators of OTTI, thereby minimizing management judgment. Fannie Mae stated in its June 22, 2004 letter to OFHEO that it would "rely primarily on quoted market prices to determine the fair value of these securities" and calculate OTTI if determined to exist.(1016) (2) Day Two Accounting Day Two accounting provides for income recognition subsequent to the determination and recognition of impairment.(1017) As support for its Day Two accounting, Fannie Mae looked to paragraph 7(b) of SOP 03-3,(1018) which allows subsequent income accretion by adjusting the yield of a security if "it is probable that there is a significant increase in cash flows previously expected to be collected or if actual cash flows are significantly greater than cash flows previously expected."(1019) Under Day Two accounting, Fannie Mae accretes certain OTTI recognized - less a non-accretable portion called a "holdback" - into income over the remaining life of the security.(1020) The holdback is an estimate of the amount of non-accretable OTTI and is determined using cumulative default rates from the results of a historical study performed by Moody's on corporate bonds.(1021) Later in 2004, Fannie Mae published its "Other-Than-Temporary Impairments of Non-Agency Securities: Policy and Operational Procedures" manual, which effectively codified and operationalized the 4-Step Approach.(1022) The manual applies the 4-Step Approach not only to MH bond and Aircraft ABS (the only categories - ---------- (1016) Id. at FMSE-SP 42126. (1017) Id. (1018) Id. at FMSE-SP 42127. (1019) ACCOUNTING FOR CERTAIN LOANS OR DEBT SECURITIES ACQUIRED IN A TRANSFER, Statement of Position No. 03-3 P .07(b) (Am. Inst. of Certified Pub. Accountants 2003). (1020) See Letter from J. Boyles to C. Dickerson, with attachments, dated June 22, 2004, FMSE-SP 42125-34, at FMSE-SP 42129. (1021) See id. at FMSE-SP 42126-27. (1022) See Other-Than-Temporary Impairments of Non-Agency Securities: Policy and Operational Procedures, dated Feb. 10, 2005, Zantaz document 1927868. 268
of investments that were subject to OFHEO's inquiry), but also to all non-agency securities and certain LIP securities.(1023) III. FINDINGS A. Accounting Policy and Monitoring In our view, Fannie Mae's policies and practices for monitoring OTTI prior to the adoption of the 4-Step Approach were inadequate in several respects. First, because FAS 115 was effective as of 1994, management should have begun formal evaluation of all of its HTM and AFS securities for OTTI as of that time. As noted above, the Company did not have a comprehensive process in place to monitor its investments subject to FAS 115 for OTTI until formation of the IC in mid-2003.(1024) In response to the question of why Fannie Mae did not have an IC prior to 2003, Boyles stated only that, prior to the formation of the IC, Portfolio did not generally encounter OTTI issues because it held only highly rated securities. Without an IC, Fannie Mae's OTTI monitoring was reactive rather than proactive; securities were only evaluated for OTTI when brought to the attention of the Controller's Office. As the Conseco bankruptcy and airline industry decline demonstrated, even highly rated investments can fall sharply in value and experience OTTI. Because we have found no consistent practices for monitoring OTTI prior to 2003, we cannot say whether the Company appropriately recognized OTTI on these or any other investments prior to that date. Second, even after establishing a formal process, it does not appear that the IC monitored all HTM and AFS investments that had unrealized losses, and thus potential OTTI. As an example, the December 2003 IC meeting Watch Lists contained only "non-agency" LIP and investments with fair value that was ninety-five percent or lower than book value.(1025) We have found no documentation showing that management verified that all other investments with unrealized losses were not other-than-temporarily impaired in some other way. Although the IC technically included MH bonds on its Watch Lists, it did not monitor them for OTTI.(1026) Furthermore, while it is true that MH bonds were - ---------- (1023) See id. (1024) See supra text accompanying notes 1003-1006. (1025) See, e.g., Mem. from I. Sussan to L. Spencer, with attachments, dated Feb. 2, 2004, FMSE 164887-91, at FMSE 164888-91. We note that certain other investments not on the Watch Lists subject to FAS 115 appear to have been included in the IC's review as well. (1026) See supra note 1006 & accompanying text. 269
evaluated for losses by Credit Policy beginning in 2002, we found very little documentation of management's MH bond OTTI evaluation (other than the DCF model results). Prior to the adoption of the 4-Step Approach, OTTI recorded for MH bonds appears to have been primarily for those bonds with fair values of ninety percent or less of book value, and whose DCF model showed a loss of principal and/or interest. We saw no evidence that other criteria required by SAB Topic 5-M were considered, such as the length of time that the fair value of a security was below book value. Third, we note that the IC was not, in practice, attended by senior management as contemplated by the Charter. Boyles stated that he drafted the Charter at Spencer's request, but that the document he created was "overkill" and that Spencer intended the IC to serve a more advisory role. Over time, Boyles was the most senior employee to attend IC meetings and, eventually, even he only attended sporadically. B. Management Modification of the MH Bond DCF Modeling In addition to the lack of any comprehensive policies or procedures to assess OTTI for all HTM and AFS securities, we also found several instances in which management's exercise of discretion over the DCF model directly affected the amount of OTTI recorded. Although we would expect a model to be refined over time as new information becomes available, and have found no evidence that management used its discretion with respect to the model to manage earnings to specific targets, management's ability to influence the model results represented another control weakness that could indicate underreporting of OTTI. Fannie Mae's MH bond DCF model was created and maintained by Credit Policy beginning in 2002, with Howard in an oversight role.(1027) Fannie Mae relied on model-generated data for impairment purposes to varying degrees between the time of its first OTTI charge for MH bonds through the adoption of the 4-Step Approach. One instance of management's involvement related to the servicing fee assumption for four Conseco bonds on which OTTI was recognized in December 2002. In that case, it appears from a spreadsheet used in a meeting with Howard, Spencer, and KPMG in December 2002, that the bonds were modeled for loss according to five different servicing fee assumptions.(1028) Ultimately, Fannie Mae recorded $51 million of OTTI on MH bonds in December 2002,(1029) which was the loss corresponding to the 100 - ---------- (1027) See Manufactured Housing Securities Status Report, Presented to: Assets and Liabilities Policy Committee, Fannie Mae Board of Directors, dated Nov. 19, 2002, FMSE-IR 655695-708 at FMSE-IR 655703 (indicating Howard as the lead for "Manufactured Housing Project Initiatives," including "Exposure Assessment . . . [m]odel individual securities' cash flows"). (1028) Selected MH Securities Held in Mortgage Portfolio spreadsheet, dated Dec. 19, 2002, FMSE 130466. (1029) Homesite Journal Entry for 713125, Journal ID JE29222, dated Dec. 28, 2002. 270
basis point servicing fee assumption.(1030) Had the higher servicing fee of 120 basis points been used, OTTI recorded would have been $66.6 million.(1031) While selection by management of the 100 basis points did reflect an expected increase in the servicing fee due to the potential Conseco bankruptcy, we have found no valid rationale for not using the higher servicing fee of 120 basis points given other evidence available at the time. It is evident from a March 2002 memorandum to Howard that management was aware at the time that market rates for servicing were as high as 130 basis points.(1032) In fact, management based its "deep dive" analysis in September 2002 on a servicing fee assumption of 120 basis points.(1033) In addition, management appears to have been aware that the bankruptcy court in the Conseco matter had approved an interim order on December 20, 2002 raising the servicing fee to 125 basis points for a period of thirty days.(1034) Later versions of the DCF model reflected the servicing fee increase.(1035) In March 2003, management exercised discretion over another key assumption in the DCF model: the default rate. Robert Kazdin, Director - Credit Policy, who was responsible for modeling the loss on collateral for the MH bond DCF, stated that in early 2003 he updated his portion of the model to reflect improved data on loan pool activities. Kazdin stated that, in March 2003, management, including Brian Graham, Adolfo Marzol and Howard, met and made minor adjustments to his default rate assumptions to reflect higher expected defaults in the near-term and lower expected defaults in the long-term (i.e., yielding lower losses overall). This updated model was referred to as "Scenario Three" and was used by the Company to model the cash flows on MH bonds from March 2003 until late 2005. Although Kazdin was not certain, he stated that the rationale for adjusting the rate to be lower in later years was to be more reflective of "core business" experience.(1036) While management's assumption differed from - ---------- (1030) Selected MH Securities Held in Mortgage Portfolio spreadsheet, dated Dec. 19, 2002, FMSE 130466. (1031) See id. (1032) Mem. from J. Kinney to T. Howard, dated Mar. 27, 2002, FMSE 165026-31, at FMSE 165029. (1033) Draft mem. from D. Kightlinger, et al., to A. Marzol, et al., dated Sept. 2002, FMSE 106020-27, at FMSE 106021-22. (1034) Manufactured Housing Securities Status Report, Presented to: Assets and Liabilities Policy Committee of the Fannie Mae Board of Directors, dated Jan. 21, 2003, FMSE 106994-99, at FMSE 106998. (1035) Mem. from D. Kightlinger, et al., to Brian Graham and K. Gifford, dated Apr. 10, 2003, FMSE 74974-75, at FMSE 74974. The fee was set at 125 basis points for the first year, and 115 basis points thereafter. Id. (1036) We understand that Kazdin used "core business" to refer to those securities backed by conventional mortgages. 271
Kazdin's original model, Kazdin stated that the default rates were the hardest assumption to model, that he was comfortable that management's change to the assumption was reasonable, and that if he thought otherwise he would have objected. In addition, Kazdin noted that, until recently, actual collateral losses tracked below projected losses (under Scenario Three), prompting changes to the model in late 2005. On another occasion in September 2003, Howard and Spencer gave guidance to Graham and others in Credit Policy to alter the DCF model results used to record MH bond OTTI. After Credit Policy completed its third quarter 2003 OTTI estimate based on the DCF, Howard directed Graham to exclude certain items from the estimate that he felt reflected "normal volatility."(1037) In addition, Spencer directed Graham to exclude certain bonds with losses from the estimate.(1038) The impact of these changes in the third quarter of 2003 was documented by Graham in an October 6, 2003 memorandum to Howard, in which Graham noted the exclusion of "a number of effects that reflect normal volatility" that would have "increase[d] total forecast potential losses by approximately $8 million."(1039) Aside from management discretion to alter the DCF models, documents showed that data used in the models sometimes contained errors that, had they been identified earlier, may have resulted in recording OTTI earlier. A particular instance was documented in an August 2003 memorandum and involved the weighted average coupon rate and Oakwood MH bond servicing fee data.(1040) Although the errors in the data - ---------- (1037) E-mail from B. Graham to D. Kightlinger, et al., dated Sept. 23, 2003, FMSE-E 624263-64, at FMSE-E 624263. (1038) See E-mail from L. Spencer to B. Graham, dated Sept. 30, 2003, FMSE-E 1685148-49, at FMSE-E 1685149. (1039) Mem. from B. Graham to T. Howard, with attachment, dated Oct. 6, 2003, FMSE 74631-33, at FMSE 74631. The exclusions Graham listed were consistent with the Howard and Spencer changes noted above: (1) a decrease in losses due to a change in the discount rate used in the model, (2) an increase in losses resulting from the deterioration of the weighted average coupon ("WAC") rate, and (3) certain bonds with de minimis losses per corporate accounting policy guidance. (1040) Mem. from A. Marzol and B. Graham to T. Howard, dated Aug. 22, 2003, FMSE 106844-45, at FMSE 106845. The WAC error was a result of a previous decision to install an edit in INTEX (a third-party cash flow modeling software) that holds the loan level note rates for fixed rate loans in Conseco MH transactions constant over time, assuming that any swings in the data were errors. However, as Conseco began to modify and/or extend the loans, legitimate changes in the WAC were ignored due to the previous edit. Marzol indicated this change "results in an increase in our projected loss for Conseco bonds of $28 million." Id. at FMSE 106844. The servicing fee error was a result of a renegotiation of the fee (by Oakwood due to its bankruptcy) to 100 basis points and senior in the waterfall, and Marzol notes "[t]he 272
appeared to have been corrected, OTTI was potentially underreported in previous periods. C. Fair Value Determinations 1. MH Bonds Fannie Mae's use of DCF models to estimate impairment loss for its MH bonds for 2002 and most of 2003 was based on the belief that the illiquidity of the MH bond market (due to the Conseco bankruptcy and general MH bond market downturn) caused market data to be unavailable and/or unreliable. This rationale has been consistently cited in interviews and revealed in documents.(1041) KPMG also appears to have been aware of the use of DCF models and did not object to it.(1042) Nevertheless, we conclude that the use of the DCF models was inappropriate. While Fannie Mae is correct that DCF models may be used in the absence of better data on fair value, we have found that market data did, in fact, exist for MH bonds during the period. First, when Fannie Mae transferred all of its MH bonds from HTM to AFS in September 2002 in conjunction with its adoption of OFHEO's Risk Based Capital requirements, the Company received prices from dealers for some of the - ---------- resulting increase in our loss projection is approximately $12 million." Id. at FMSE 106845. (1041) See, e.g., E-mail from Rob Schaefer to A. Marzol, et al., dated Feb. 17, 2004, FMSE-E 2155495-97, at FMSE-E 2155496 (stating that "[i]n 2002, when news of Conseco's collapse was breaking, the market was too illiquid to be relied upon for a good market price, and so our marks at that time were based upon internal estimates using general market pricing information"); see also 2003 Form 10-K at 46 (stating that "[w]hen market quotations are not readily available because of the nature of the security or illiquid market conditions - as occurred during the downturn in the manufactured housing sector - we estimate the fair value based primarily on the present value of future cash flows, adjusted for the quality of the rating of the securities, prepayment assumptions, and other factors, such as credit enhancements . . . "). (1042) See KPMG Conseco Bonds workpaper, dated Jan. 2003. In the KPMG year-end 2002 audit workpaper regarding impairments, Fannie Mae's external auditing firm notes, "Kerry Gifford and Steven Shen led the efforts to model the expected cash flows for purposes of determining management's estimate of a current price. Although these securities are traded, a Bloomberg price may not represent an appropriate valuation given the limited volume of trades and a [sic] given the priority of a given investors [sic] variable interest. Fannie Mae performed a NPV calculation for purposes of determining fair value." Id. 273
most illiquid MH bonds in order to mark these securities to fair value.(1043) Second, our review of Bloomberg's pricing service showed that bid/ask pricing was available for most of the MH bonds on which Fannie Mae recognized impairment during the period it used its DCF models to determine OTTI. We note that the OTTI recorded was higher than it would have been if the prices from Bloomberg were used, but GAAP is clear that using quoted market prices to measure value is preferred.(1044) 2. Aircraft ABS Fannie Mae also used a DCF model to estimate fair value for its Aircraft ABS.(1045) The IC cited the same rationale for using DCF models for Aircraft ABS that was used to estimate OTTI for MH bonds - an illiquid market. But quoted market prices had been obtained at the time, and should have been considered when measuring fair value. Further, management's explanations for the decision to use the model were inconsistent. In a July 7, 2003 memorandum, David Benson, Vice President and Assistant Treasurer - Portfolio, explained the DCF analysis that ultimately supported an impairment charge of $83.5 million (recorded in the second quarter of 2003) on Aircraft ABS. The memorandum indicated that Fannie Mae used the DCF model to determine fair value despite the fact that "[q]uoted indicative (not for trading) prices" showed even lower fair values.(1046) Benson appears to have justified the use of DCF based on a "dearth of transactional activity to validate quoted levels."(1047) Benson further stated that he felt confident in the DCF model because he was familiar with the details, and that the fair value price generated by the DCF model represented the prices at which a willing buyer and a willing seller would enter into a transaction currently. But Benson defined indicative prices in the same way, stating that he had no reason to believe the indicative prices received were unreliable or obtained from unreliable sources. Benson added that, ultimately, the decision to use fair value per the DCF model was not his decision. - ---------- (1043) See E-mail from S. Shen to Janet L. Pennewell, et al., dated Sept. 24, 2002, Zantaz document 1940768, at 1. In the e-mail, Shen notes that "[t]he bonds to focus on are the 1999 and 2000 vintages. We had the majority of those deals specifically mark [sic] by the dealers because they have the leasst [sic] amount of liquidity." Id. (1044) Boyles stated that although he was not involved in generating the market values, he assumed that Fannie Mae did not believe the Bloomberg prices were accurate because there were no trades in the market for these bonds. (1045) Mem. from David Benson to Peter Niculescu, L. Knight, et al., dated July 7, 2003, FMSE-IR 321383-85, at FMSE-IR 321384. (1046) Id. at FMSE-IR 321383. (1047) Id. 274
The IC ultimately recommended use of DCF-generated fair values; the July 9, 2003 IC minutes reflect the $83.5 million write-down (listed as $84 million in the minutes), and note: "If we had used market quotes, we would have another $45M in write-downs. However, we believe market quotes represent a fire sale because the market for aircraft securities is currently not very liquid."(1048) In hindsight, management's own analysis in the first quarter of 2004 showed that the claim of "fire sale" prices did not prove to be valid for at least one of the Company's Aircraft ABS. The July 7, 2003 Benson memorandum shows that the PALS A1 security had a DCF-based fair value of $70.19 and a quoted indicative price of $60.00.(1049) In 2004, the PALS A1 security was written down to a fair value of $48.50, a price that was $11.50 lower than the indicative price of $60.00 that was dismissed in 2003 as representing a "fire sale."(1050) In summary, both MH bond and Aircraft ABS quoted market prices were available during the time that management was using the DCF model results to determine the amount of OTTI to record. Management was required under GAAP to consider these prices in determining fair value when recording OTTI, particularly for those securities with market prices that were lower than fair values derived from the DCF models. We note that the SEC provided guidance in 1986(1051) that companies may not rely on their subjective calculations of fair value versus valuations of external pricing - ---------- (1048) Draft Minutes of the Meeting of the Impairment Committee, dated July 9, 2003, FMSE 69363-65, at FMSE 69363. (1049) Mem. from D. Benson to P Niculescu, L. Knight, et al., dated July 7, 2003, FMSE-IR 321383-85, at FMSE-IR 321384. (1050) Mem. from D. Benson to P Niculescu, L. Knight, et al., dated Mar. 5, 2004, FMSE-E 1386760-62, at FMSE-E 1386760. (1051) See Accounting for Loan Losses by Registrants Engaged in Lending Activities, Exchange Act Release No. 34-23,854, 17 C.F.R. Part 211 *1, *13-14 (Dec. 1, 1986), stating: The Commission will presume that active markets reflect objective measures for current fair values, determined by the beliefs of reasonably informed persons regarding the present and future economic utility of the items being traded and the risks associated therewith. Thus, without independent and objective support for derived valuations that can be demonstrated to more appropriately reflect fair value in particular sets of circumstances, derived valuations exceeding current values in active markets should not be used in cases where fair value accounting is required by GAAP. 275
sources and reiterated that view in 2004 in an enforcement proceeding.(1052) Accordingly, we recommend that the Company promptly adopt procedures that ensure objective evaluation of fair value in accordance with GAAP. D. Timing of Recognition In our view, management waited too long to evaluate OTTI on its Aircraft ABS, and thus may have underreported OTTI on those investments prior to the evaluation by the IC. The weak state of the industry, poor market conditions, and the numerous public companies that recorded impairment charges on their aircraft-related assets in 2002 were well documented by Fannie Mae's own due diligence in early 2003.(1053) Indeed, Fannie Mae was aware that several other companies had recorded impairment related to aircraft investments as early as in 2002.(1054) Boyles (the IC chair) suggested that the delay in monitoring resulted from a simple lack of awareness by the Company. Specifically, Boyles said that he had no knowledge that Fannie Mae owned - ---------- (1052) See In the Matter of Morgan Stanley, 2004 SEC LEXIS 2573 (Nov. 4, 2004). The SEC issued a cease and desist order against Morgan Stanley in regards to Morgan Stanley's valuation of certain aircraft in its leasing portfolio and certain bonds in its high-yield bond portfolio. Regarding Morgan Stanley's high-yield bond portfolio, the SEC stated that: The majority of the overvaluations occurred with bonds in the telecommunications industry. For those bonds, Morgan Stanley believed that market conditions rendered third-party price quotations unreliable. In such market conditions, GAAP required Morgan Stanley to use its best efforts to determine the fair value of the bonds, which is the price at which a willing buyer and a willing seller would enter into a current exchange. Instead of following GAAP to determine the fair value for those bonds, Morgan Stanley valued those bonds by "taking a longer view of the market" and essentially put its subjective opinion about the value of the bonds ahead of prices quoted by external pricing sources. In effect, Morgan Stanley valued its positions at the price at which it thought a willing buyer and seller should enter into an exchange, rather than at the price at which a willing buyer and a willing seller would enter into a current exchange. Id. at *12-13. (1053) See Draft mem. from P Salfi, et al. to K. Barnes, et al., dated Mar. 14, 2003, FMSE-IR 672690-99. (1054) Id. at FMSE-IR 672692. 276
Aircraft ABS prior to 2003 and that, if the IC had been in place prior to 2003, the Company might have recognized impairment on Aircraft ABS earlier. In addition, it appears that management's reliance on DCF models in place of "indicative prices" led to further delays in recognizing impairment on Aircraft ABS. As noted above, management was aware in the second quarter of 2003 that, if it had based its fair value on market quotes, it would have recognized an additional $45 million in OTTI on Aircraft ABS, but dismissed these market quotes as a "fire sale."(1055) Nevertheless, management did not record any further OTTI on its Aircraft ABS during 2003. In the first quarter of 2004, management recognized additional OTTI of approximately $21.7 million on one of its Aircraft ABS, PALS 2001-1A A1. We note that $10.2 million of the $21.7 million(1056) would have been recognized in 2003 had its indicative price of $60 not been dismissed as a "fire sale" price. PART G: ACCOUNTING FOR INVESTMENTS IN INTEREST-ONLY MORTGAGE-BACKED SECURITIES I. INTRODUCTION This section addresses Fannie Mae's accounting for investments in interest-only mortgage-backed securities ("IO MBS"). Beginning in 1995, management combined its IO MBS investments with other securities (specifically, MBS and REMIC securities) for accounting purposes, and treated the IO MBS as an increase in the premium or a reduction in the discount on the other security. Management initially obtained approval from KPMG for this accounting treatment for the IO MBS investments. We believe, however, that this approach violates GAAP. EITF 90-2, which addressed an analogous situation and should have been applied by management to the accounting for its IO MBS investments, required that an exchange transaction take place before the accounting for the individual interest-only and principle-only securities can change. Furthermore, management's motive for engaging in combined accounting was to avoid recognizing impairment charges on the IO MBS, which would have resulted in volatility in the income statement. While management initially consulted with and obtained approval from KPMG, by 1998, management knowingly withheld from KPMG its impairment analysis on the IO MBS, apparently fearing that the new KPMG audit team might change the decision of the old audit team and require management to change its accounting, which could have resulted in the Company being required to recognize impairment losses. - ---------- (1055) See supra text accompanying note 1048. (1056) See Homesite Journal Entry for account 713127, Journal ID JE29479, dated Feb. 28, 2004. 277
Finally, management failed to inform the Board of the issues relating to its accounting for the IO MBS investments until OFHEO raised questions about these practices in April 2004. This is particularly troubling because the Freddie Mac restatement raised nearly identical issues, and management, in its presentation to the Board about the Freddie Mac restatement in January 2004, not only failed to disclose the existence of its own problematic synthetic IO MBS combinations, but led the Board to believe that Fannie Mae did not have similar issues to those raised at Freddie Mac. II. BACKGROUND A. Applicable Accounting Standards An investment in IO MBS entitles an investor to a portion of the interest payments on the MBS. Changes in the fair value of an investment in IO MBS are directly related to changes in interest rates: When interest rates increase, the fair value of IO MBS increase; when interest rates decline, the fair value of IO MBS decline. This results from the fact that, as a general matter, when interest rates increase, the speed at which borrowers prepay or refinance their mortgages declines. Conversely, when interest rates decline, the speed at which borrowers prepay or refinance their mortgages increases. Increased prepayments reduce the total amount of interest received over the life of MBS; therefore, the cash flow to the holder of an investment in IO MBS also will decline. For that reason, the Emerging Issues Task Force (the "EITF") concluded, in EITF 93-18, that IO MBS are high risk investments.(1057) A decrease in interest rates that is accompanied by an increase in prepayments could cause an investor to lose some or all of his investment. B. Fannie Mae's Accounting for Investments in IO MBS Fannie Mae began acquiring IO MBS as investments in 1994. The Company's IO MBS portfolio increased from $125 million in July 1994 to $524 million - ---------- (1057) RECOGNITION OF IMPAIRMENT FOR AN INVESTMENT IN A COLLATERALIZED MORTGAGE OBLIGATION INSTRUMENT OR IN A MORTGAGE-BACKED INTEREST-ONLY CERTIFICATE, Emerging Issues Task Force of the Fin. Accounting Standards Bd., Issue No. 93-18 (Mar. 1994) (superceded by RECOGNITION OF INTEREST INCOME AND IMPAIRMENT ON PURCHASED AND RETAINED BENEFICIAL INTERESTS IN SECURITIZED FINANCIAL ASSETS, Emerging Issues Task Force of the Fin. Accounting Standards Bd., Issue No. 99-20 (July 2000)) (hereinafter "EITF 93-18"). EITF 93-18 requires an investor in IO MBS to measure impairment for the difference between the carrying amount of the investment and the current fair value of that investment. Id. EITF 93-18 indicates that impairment should be recognized when the present value (determined using a risk-free rate) of the investor's best estimate of the future cash flows to be received on the investment is less than the carrying amount of the investment. Id. The investor's best estimate of the future cash flows should incorporate the investor's most current estimate of future prepayments. Id. 278
as of April 30, 1995.(1058) Fannie Mae acquired its IO MBS investments when interest rates were high;(1059) as a result, Fannie Mae was exposed to the risk that it would have to recognize impairment of those investments if interest rates declined and borrowers refinanced their mortgages (which did occur beginning in 1995). It appears that management determined the effective yield on the IO MBS at the time of purchase as required by EITF 89-4, but did not recalculate the yield subsequently as EITF 89-4 also required.(1060) 1. Motivation for the Decision to Combine IO MBS with Other Securities for Accounting Purposes In late 1995, management, with KPMG's concurrence, combined its IO MBS with REMIC securities to create "Synthetic REMICs." The combination of the two securities was done only for accounting purposes; in other words, the Company did not enter into any transactions that would have been expected to change the characteristics of either the IO or the REMIC security, "which would have been one permissible reason for combining the securities."(1061) Instead, Fannie Mae's decision appears to have been - ---------- (1058) See Mem. from Sam Rajappa to Timothy Howard, Thomas A. Lawler, Joseph Amato, Deborah Cohen, James Parks, Leanne G. Spencer, Richard DePetris, and Jonathan Boyles, dated June 22, 1995, FMSE-IR 213735-36, at FMSE-IR 213735 (hereinafter "Rajappa Memo"). (1059) See Undated IO/REMIC Package Briefing for Tim Howard, FMSE-SP 86636-40, at FMSE-SP 86636 (hereinafter "IO/REMIC Briefing"). This and other examples of documents relevant to the discussion in Chapter VI, Part G can be found in the accompanying Appendix at Tab D.7. (1060) See Mem. from J. Boyles to Distribution, dated Oct. 26, 1995, FMSE-IR 213989-4050, at FMSE-IR 213990. In this memorandum, Boyles wrote: "IO securities yields are currently being calculated and recorded according to EITF 89-4 at the time of purchase, however, subsequent recalculation of the effective yield with current prepayment assumptions have [sic] only been performed on a portion of the portfolio and is extremely time consuming." Id. (1061) See, e.g., EXCHANGE OF INTEREST-ONLY AND PRINCIPAL-ONLY SECURITIES FOR A MORTGAGE-BACKED SECURITY, Emerging Issues Task Force of the Fin. Accounting Standards Bd. Issue No. 90-2 (Mar. 1990) (nullified by ACCOUNTING FOR TRANSFERS AND SERVICING OF FINANCIAL ASSETS AND EXTINGUISHMENTS OF LIABILITIES, Statement of Fin. Accounting Standards No. 125 (Fin. Accounting Standards Bd. Jun. 1996)) (addressing accounting by an investor who exchanges IO and PO MBS with issuing trust for MBS); see also TRANSACTIONS INVOLVING SPECIAL-PURPOSE ENTITIES, Fin. Accounting Standards Bd. Staff Position Topic D-14 (Fin. Accounting Standards Bd. 1989) (setting forth conditions that must be met in order to allow transferor to avoid consolidating trust; consolidation of trust would require transferor to account for transferred financial assets in same manner as it did before transfer) (nullified by CONSOLIDATION OF VARIABLE INTEREST ENTITIES - AN 279
partially motivated by a desire to avoid recognizing impairments on its IO MBS, a requirement that would have resulted from increased mortgage prepayments.(1062) Contemporaneous documentation indicates that Fannie Mae considered approaching KPMG as early as September 5, 1995, because of concerns over potential impairment charges.(1063) A document entitled "IO/REMIC Package Briefing for Tim Howard" ("IO/REMIC Briefing"), stated the following: Financial Standards performed impairment testing on the IO portfolio in the fall of 1995 and determined that many of the IO's were either impaired or close to being impaired. To mitigate the potential earnings risk, Fannie Mae linked the IO's with specific REMICs that were purchased at the - ---------- INTERPRETATION OF ARB NO. 51, Fin. Accounting Standards Bd. FIN 46(R) (Fin. Accounting Standards Bd. 2003). (1062) See Rajappa Memo at FMSE-IR 213735. Rajappa noted: As a first pass to estimate the impairment trigger point, Portfolio Management used the groupings of IO securities utilized in their periodic mark-to-market calculations. The expected cash flows of these groupings from the April 30 IO portfolio, discounted at a risk-free interest rate (using a June 14 Treasury yield curve) across multiple interest rate scenarios within ALMS, was $544 million. Although there is a $20 million cushion above the impairment trigger, I believe we need to monitor closely this portfolio in light of the effect that changes in interest rates can have on IO values. In addition, I understand that Portfolio Management plans to value the IO portfolio on an individual security basis beginning this month. Id. (1063) See Mem. from R. DePetris to T. Lawler, et al., dated Sept. 5, 1995, FMSE-IR 213731-32, at FMSE-IR 213731. This memorandum summarized a discussion regarding "IO impairment and yield recalculation." Id. DePetris stated that "Portfolio management will redo the July 31 risk-free rate impairment schedule using projected CPRs that are more consistent with our internal prepayment model." Id. At some point after that update, the group planned to "meet with the appropriate representatives from Peat Marwick [KPMG] to discuss our position that measuring impairment for our IO securities without evaluating changes in value of the assets bought in combination with those IOs does not make sense." Id. 280
same time and were looked at together in the decision to purchase the securities.(1064) Management analyzed its IO MBS investments on an individual security basis later in 1995, using a methodology that appears to have been consistent with the guidance in EITF 93-18.(1065) The analysis suggested that a write-down of $29.2 million was required due to impairment.(1066) Based on our review of the Company's 1995 financial statements, however, it does not appear that management recognized the write-down. In addition to the IO/REMIC combinations created in 1995, management also accounted for seven pairs of IO and principal only ("PO") securities acquired between 1999 and 2002 on a combined basis, using the conclusions reached in 1999 as support for that accounting treatment.(1067) The Company's accounting for combinations of IO MBS and PO MBS was subject to the provisions of FASB Statement No. 125, which requires a transfer of a financial asset as a precondition to changing the accounting for that asset. But, as with the prior combinations, Fannie Mae did not transfer its IO MBS and PO MBS and therefore should not have accounted for the separate securities as if there had been a transfer.(1068) 2. Discussions with KPMG As noted above, management discussed its plan to combine IO MBS with other securities with KPMG in November 1995. KPMG concurred with Fannie Mae's combined IO accounting. Management's rationale for this plan was most clearly set forth in a November 27, 1995, draft memorandum from Richard DePetris, then Director--Financial Standards: The purpose of this memo is to document the discussion at last week's meeting relating to Fannie Mae's purchase of - ---------- (1064) IO/REMIC Briefing at FMSE-SP 86636 (emphasis added). (1065) EITF 93-18 paras. 4-5. (1066) See Interest Income Impact, dated Nov. 9, 1995, FMSE-IR 213694-98, at FMSE-IR 213694. The document analyzed the effect on net interest income of writing down the impaired IO MBS and adjusting the yield as required by EITF 89-4 on the remaining securities. Id. (1067) See Draft letter from J. Boyles to John James, dated Oct. 28, 2004, Zantaz document 796794, at 1. (1068) See ACCOUNTING FOR TRANSFERS AND SERVICING OF FINANCIAL ASSETS AND EXTINGUISHMENTS OF LIABILITIES, Statement of Fin. Accounting Standards No. 125 (Fin. Accounting Standards Bd. 1996) (hereinafter "FAS 125"). 281
interest-only securities in combination with other mortgage-backed securities that, together, have the characteristics of MBS collateral. 1. We explained we do not buy IOs in isolation but as held-to-maturity assets with other securities (e.g., POs, standard MBS, and PAC or sequential REMIC bonds) as a package . . . . 2. We agreed that the only accounting guidance related to recombinations of securities was EITF 90-2 that deals with exchanges of POs and IOs from the same trust and requires recording the recombined security at fair value. No existing accounting literature specifically addresses combining securities from different trusts. 3. Peat Marwick [KPMG] agreed that, in substance, buying separate IO and PO securities, even if from different trusts, was similar to buying a REMIC tranche or security issued through a trust that is backed by the same IO and PO securities, particularly since Fannie Mae has the ability to form a trust with the individual securities as collateral. 4. Because we have purchased and monitor market performance of the securities bought as a package in terms of the combined cash flows, Peat Marwick [KPMG] agreed with us that substance could prevail over form in evaluating and accounting for these securities, whether or not the securities were put in a trust at purchase or subsequently. 5. Because Fannie Mae will recover its investment in the "effective" security produced by combining the cash flows of the IOs and other MBS, it would not fall within the definition of a high-risk security in EITF 89-4 nor the provisions in the Exposure Draft on transfers and securitization of financial assets [FAS 125] dealing with financial assets subject to prepayment. Therefore, the "effective" security may be included in the held-to-maturity classification and the principal balance is not subject to writedown due to prepayments.(1069) - ---------- (1069) Draft mem. from R. DePetris to T. Lawler, S. Rajappa, D. Cohen, J. Parks, L. Spencer, Thanasis Simos and KPMG, dated Nov. 27, 1995, FMSE-IR 213933-34, at FMSE-IR 213933. We disagree with KPMG's advice to Fannie Mae. Acquiring an IO MBS and a PO MBS from different trusts is not the same as purchasing a REMIC security because of the possibility that the mortgages underlying the IO 282
KPMG was copied on this memorandum, and we have confirmed that a copy of this document is in KPMG's files. What is not clear from this memorandum is whether management discussed with KPMG its motive behind the combined accounting, that is, management's desire to avoid impairment write-downs. We understand that Ken Russell of KPMG may have been consulted on this matter in his role as a member of KPMG's Department of Professional Practice. Russell agreed that the accounting for the proposed combinations was a matter of substance over form. 3. The Internal Policy for the Acquisition of IO MBS In March 1996, management drafted a policy to implement its accounting for IO MBS. In a draft memorandum from that month, Sam Rajappa, then Controller, discussed the Company's acquisition of synthetic MBS: Fannie Mae's policy is not to purchase interest-only securities (IOs) as individual investments but purchase them with the intention of achieving a desired yield or other investment characteristic through the combination with other securities to form a "synthetic MBS." In Fall 1995, we obtained KPMG Peat Marwick's concurrence that a trust structure, similar to those for MBS, would not be necessary for these synthetic MBS to receive MBS accounting. In December 1995, all securities purchased in combination with IOs were identified and formed into 56 synthetic MBS. During January 1996, these synthetic MBS were transferred to separate pool types in PITS, and are not accounted for as MBS as opposed to each component in the synthetic MBS receiving separate accounting treatment.(1070) The memorandum also set forth the following policy for purchasing synthetic MBS: Synthetic MBS generally can only be formed using securities that were purchased on the same day as the IO security, however, exceptions can be made if there is documentation explaining the reasons for the securities - ---------- MBS and PO MBS will not prepay at the same rate. Furthermore, although Fannie Mae had the ability to form a trust, it would have been required to sell an interest in the trust to one or more third parties to avoid consolidating the trust and unwinding the desired accounting. (1070) Draft mem. from S. Rajappa to T. Lawler, L. Spencer, et al., dated Mar. 25, 1996, FMSE-IR 213556-58, at FMSE-IR 213557. 283
being purchased on different days and prior approval by the SVP - Portfolio Management. Purchasing IO securities with the intention of matching them with flow business is prohibited. (Note: Settlement dates can vary by up to one month.) . . . Synthetic MBS that have a significant premium run the risk of being treated as a high-risk investment and marked-to-market through equity, or potentially through earnings if impaired. Because of this risk, the formation of premium Synthetic MBS must be approved by the Senior Vice President - Portfolio Management.(1071) Fannie Mae issued the policy as a final document on April 10, 1996.(1072) In a March 26, 1996 e-mail exchange between Boyles and Parks, then Controller--Multifamily (formerly Vice President--Financial Standards), Parks suggested revising the draft policy to "provide a guideline as to what is considered 'significant MBS [premium].'"(1073) In response, Boyles suggested leaving the definition intentionally vague for fear of running afoul of FAS 125: We left the wording about what constitutes 'significant premium' purposely vague because if we were to specify 25 percent or 15 percent as significant premium, someone (peat) may come along and force several of our packages into an AFS category. If you remember, we had several packages with premiums over 50 percent. By leaving it vague and requiring SVP sign-off on any premium securities on new packages, we hope to eliminate the problem going forward.(1074) Boyles did not have any specific recollection of this e-mail exchange, stating only that Fannie Mae had defined "significant premium" to mean a premium in excess of twenty - ---------- (1071) Id. at FMSE-IR 213557-58. (1072) See Mem. from S. Rajappa to Distribution, dated Apr. 10, 1996, FMSE-IR 213554-55. (1073) E-mail from J. Boyles to J. Parks, dated Apr. 3, 1996, FMSE-IR 213559-60, at FMSE-IR 213560. Under paragraph 14 of FAS 125, an investment that could be prepaid so the holder would not recover substantially all of the carrying amount should be classified as an available-for-sale ("AFS") or trading security and recorded at fair value. FAS 125 P 14. (1074) Id. at FMSE-IR 213559 (emphasis added). 284
percent.(1075) Boyles's e-mail showed his recognition that the combined accounting was an aggressive interpretation that might be challenged by KPMG, and also demonstrated his willingness to obscure the facts from KPMG. 4. Subsequent Accounting for Synthetic MBS Fannie Mae accounted for its synthetic MBS in the same manner as it accounted for REMICs. That is, because of system limitations, management used the sum-of-the-years's digits ("SYD") method to amortize premium or discount on its REMIC and synthetic MBS securities, rather than level yield accounting required under FAS 91.(1076) As recounted in the IO/REMIC Briefing: These "packages" [the synthetic MBS] were then given 15 year SYD tables for the amortization of the premium based on our analysis of the projected cash flows at the end of 1995. The 15-SYD table was used because we believed it was a close proxy for level yield accounting. We were unable to use level yield accounting because of systems constraints.(1077) By 1998, the use of the fifteen-year SYD table had resulted in a significant under-amortization of "premium" associated with the synthetic MBS. According to the IO/REMIC Briefing for Howard: We reevaluated the packages in the fall of 1996 and determined we had under-amortized premium into earnings by approximately $28 million. This was because rates had dropped since the time of the original analysis. The working group decided that no action should be taken but that we would continue to monitor the problem and conduct a detailed analysis annually. Over the past several months, we again took a look at how these packages are performing and found that we had under-amortized premium into earnings by over $50 million. In addition, our analysis revealed that several of the REMICs linked to IO's are expected to mature in the near future leaving only an IO security in the package. We also found that several packages now had such significant - ---------- (1075) Boyles also stated that he left the decision on how to define significant premium to Lawler because he felt it was a business decision. (1076) See supra Chapter IV. (1077) IO/REMIC Briefing at FMSE-SP 86636. 285
premiums associated with them that they should be classified as available for sale.(1078) In fact documents show that by December 1998, the under-amortization had increased to approximately $180 million.(1079) As noted above, KPMG generally was aware of the Company's catch-up position (and in fact concluded that Fannie Mae's failure to record all of the catch-up in 1998 was in error, posting the unrecorded catch-up to its summary of unadjusted errors).(1080) Further, KPMG was consulted in 1995 about combining IO and other securities; however, evidence showed that the information about management's accounting for IO MBS as of 1998, and the impaired status of the combinations, were kept from KPMG. As stated in the IO/REMIC Briefing: KPMG has apparently forgotten about these [synthetic MBS] transactions, and we have not brought these issues to their attention. They have experienced significant turnover since we originally adopted the "package" accounting and, as a result, there is currently only one member of the audit team that remaining [sic] from the fall of 1995. The accounting team they currently have on the audit is more technically proficient, and if they stumble across these packages, may not be as easily convinced of the current accounting treatment. We have made every effort to keep our analysis confidential.(1081) Boyles did not recall preparing the IO/REMIC Briefing document, but acknowledged that the electronic user identification associated with the document was his, and that he had no reason to dispute the assertion that he was the author. Although Boyles was emphatic that he did not withhold information from KPMG, both the fact that the document was found on Boyles's hard drive and his inability to explain who else could have prepared this document made it difficult to credit Boyles in this regard. - ---------- (1078) Id. (emphasis added). (1079) See, e.g., Fannie Mae Catch-up Summary, dated Jan. 8, 1999, FNMSEC 2677. Management's error in its accounting for IO MBS contributed materially to the amount of the catch-up computed at the end of 1998, a portion of which was recognized by the Company in its 1998 (1080) Fannie Mae, 1998 Annual Report, at 63 (1999), available at http://www.fanniemae.com/global/pdf/ir/annualreport/1998/fullreport.pdf. (1081) IO/REMIC Briefing at FMSE-SP 86636-37 (emphasis added). 286
Moreover, an e-mail dated January 28, 2002, from Spencer to Howard regarding "KPMG and Freddie," on which Boyles was copied, provided further evidence of management's awareness that information about the manner in which Fannie Mae accounted for IO MBS had been kept from KPMG.(1082) Spencer wrote: "There is at least one thing that we know of where we have a favorable accounting treatment and that is on the IO's we have on our books. Freddie is doing IO accounting and we are not. KPMG has not figured that out - but Jonathan reminded me of this."(1083) Accordingly, we conclude that Boyles did, in fact, withhold information from KPMG with respect to the impairment of IO/REMIC combinations and that Howard and Spencer were aware of this fact. III. FINDINGS A. Management's Disclosures About Accounting for IO MBS to the Board of Directors or its Committees We have found no evidence that management communicated its manner of, or motive for, IO MBS accounting to the Board or any of its committees prior to January 2004. Moreover, when management did provide information about its IO MBS accounting methodology to the Audit Committee in a January 2004 presentation, it failed to disclose the Company's accounting methods fully. The January 2004 presentation to the Audit Committee was in response to the Audit Committee's direction to management to determine whether any of the issues identified in Freddie Mac's restatement were applicable to Fannie Mae. (1084) Financial Standards prepared a matrix that identified the accounting issues discussed in the Freddie Mac restatement, the effect on Freddie Mac of the restatement (both in cumulative earnings impact as well as the restated accounting), whether Fannie Mae had similar issues, and whether the Freddie Mac restatement raised concerns for Fannie Mae's own accounting practices. (1085) In this matrix, management made the following statements about Freddie Mac's accounting for IO securities: Freddie Mac invests in IO securities and during the restatement process they determined that they had not been properly applying impairment accounting for these securities. This error arose because they did not have a - ---------- (1082) See E-mail from L. Spencer to T. Howard, dated Jan. 28, 2002, FMSE 487130. (1083) Id. (emphasis added). (1084) See Audit Committee Update: Understanding Freddie Mac's Accounting Restatement as It Affects Fannie Mae, dated Jan. 23, 2004, FMSE-IR 378611-23. (1085) See Review of Accounting Issues Related to Freddie Mac's Earnings Restatement, dated Jan. 14, 2004, Zantaz document 289451. 287
complete population of IO securities due to the following four issues. . . . Freddie Mac combined IO securities with other securities in a trust and accounted for the combined as a security under FAS 115 even though they owned all or substantially all of the new securitization.(1086) In the matrix, Financial Standards then made the following incomplete representation to the Audit Committee about its own IO MBS accounting: We have also combined an IO security with discount MBS (WAMU) transaction and put them into a security. We sold off 10 percent of the security so that we would not have to consolidate the IO back onto our books. It appears that because Freddie Mac owned all or substantially all of the resulting security they had to consolidate this structure, which was exactly what we were trying to avoid. It appears their accounting agrees with how we thought these deals would work . . . . On occasion we have simultaneously purchased an IO security along with the matching PO security and accounted for them as a combined security similar to [Freddie Mac's] transaction regarding linked swaps described above. It appears [Freddie Mac] [has] reached the same conclusion [that the individual IO and PO securities purchased simultaneously should be accounted for as a combined security] as we have.(1087) Financial Standards' presentation suggested to the Board that Fannie Mae's IO MBS accounting should be of no concern to the Board. The statement in the first paragraph was incomplete because it omitted information about the Company's synthetic MBS dating back to 1995, which would have notified the Board that the Company's accounting was even less defensible than the accounting Freddie Mac was restating. Moreover, the discussion in the second paragraph may have given the Board unwarranted comfort. According to Boyles, Freddie Mac's securitization program is substantially different in structure than Fannie Mae's. Those differences were cited by management as support for differences in how the companies accounted for buy-up fees, but were ignored in this circumstance. - ---------- (1086) Id. at 11. (1087) Id. at 11-13. 288
April 2004, after OFHEO and Deloitte & Touche raised questions about the accounting for synthetic MBS, management finally informed the Audit Committee that its accounting for IO MBS was a potential issue. The second-to-last slide in a presentation to the Audit Committee in April indicated, "OFHEO has questioned our practice of bundling for accounting purposes interest-only (IO) with principal-only (PO) securities or other securities."(1088) The minutes did not provide any further information on the details that may have been presented to the Audit Committee.(1089) B. Discussions with the SEC Regarding Accounting for IO MBS After OFHEO raised the Company's synthetic MBS accounting as an issue in the Special Examination, management chose to address this issue directly with the SEC. As part of its preclearance letter to the SEC regarding Fannie Mae's policy for measuring and recognizing other-than-temporary impairments of investment securities, management asked the SEC: May a company that purchases an IO security and a PO security account for them on a combined basis - if: (1) the IO and PO have the same notional amount; (2) the IO and PO are issued by the same counterparty; (3) the purchaser retains a unilateral right to reconstitute the IO and PO and obtain the underlying MBS collateral; (4) the two were purchased in contemplation of each other; and (5) the economics of owning the IO and PO are identical to owning the collateral that underlies them both?(1090) The fact pattern in this letter, however, presented a "best" case scenario and omitted any information about the actual combinations that were formed by Fannie Mae in 1995. When asked why Fannie Mae chose to use the above fact pattern in its letter to the SEC rather than describing the synthetic MBS arrangements in place since 1995, Boyles explained that the Company wanted to address what it considered to be the "best" case, on the theory that if it could not persuade the SEC that combined accounting was appropriate in the "best" case, it would not be able to convince the SEC of the accounting for the synthetic MBS. - ---------- (1088) Audit Committee Briefing: Impairment Policies and Practices and Update on Special Exam, dated Apr. 19, 2004, Zantaz document 808283, at 14. (1089) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 19, 2004, FMSE 504838-46. (1090) Letter from J. Boyles to Donald T. Nicolaisen, dated Apr. 28, 2004, FMSE-IR 483271-82, at FMSE-IR 483279. 289
Boyles also stated that he likely mentioned the existence of the synthetic MBS in a conference call with the SEC. His recollection is supported by an e-mail dated November 3, 2004, from Boyles to the SEC, stating: I already described what this schedule shows . . . . Keep in mind the Post 1996 IO's are being recombined into MBS and the Pre-1996 IO's are being sold and we've taken the necessary writedowns to their market value (about $7 million) in the third quarter to reflect our intention to sell in the fourth quarter.(1091) Attached to the e-mail was a spreadsheet showing the notional amounts of the pre-1996 and post-1996 IOs, along with average price information and estimates of the book value of the IOs as of December 31, 2001, 2002, and 2003, and September 30, 2004.(1092) The SEC ultimately concluded that Fannie Mae should account for the IO and PO securities described in its preclearance letter on a separate basis until it exchanged those separate securities for a mortgage-backed security.(1093) The SEC apparently did not object to Fannie Mae's proposal to apply the proper accounting to its IO MBS prospectively, that is, the SEC did not require the Company to restate prior periods to correct the error.(1094) The SEC also did not object to Fannie Mae's proposal to determine the carrying amounts for the separate IO and PO components by performing a relative fair value calculation and allocating the carrying amount of the combined instrument based on those relative fair values.(1095) Boyles had advised the SEC that Fannie Mae had to allocate the carrying amount of the combined instrument to the individual components because the Company did not have records that reflected those individual components.(1096) IV. CONCLUSIONS Based on our review of the facts, we conclude that: - ---------- (1091) E-mail from J. Boyles to J. James, dated Nov. 3, 2004, Zantaz document 796557. (1092) See Undated Pre/Post 1996 Notional IOs spreadsheet, Zantaz document 796560, at 1-2. (1093) See Letter from J. Boyles to J. James, dated Sept. 15, 2004, Zantaz document 691098, at 1. (1094) Id. (1095) Id. (1096) Id. 290
Management "aggressively interpreted" GAAP when it began accounting for IO MBS on a combined basis with other securities. In our view, the accounting violated GAAP. Management's motive behind the IO MBS accounting on a combined basis was to avoid complying with EITF 93-18, which would have required recognition of impairment losses. Notwithstanding the views of management at the time, the accounting literature governing when it would be appropriate to combine two legally separate financial instruments was clear. Although EITF 90-2 addressed a fact pattern involving IO and PO securities issued from the same trust, the consensus was clear that an exchange transaction was required to take place before the accounting for the individual IO and PO securities could change. The authoritative literature recognizes that, in some cases, a specific transaction may not be covered by an established accounting principle.(1097) In that case, companies should account for the transaction by reference to an established accounting principle that applies to an analogous transaction. In the current circumstance, EITF 90-2 addressed an analogous situation and should have been applied by Fannie Mae to the accounting for its synthetic MBS. Furthermore, FAS 125 provided explicit guidance on what conditions must be met to derecognize a financial asset. This guidance applied to Fannie Mae's accounting for IO and PO securities that it "combined" and treated as an MBS.(1098) Among other conditions, FAS 125 requires a transfer of the financial assets for a company to be able to derecognize those assets.(1099) As with its combination of IO and REMIC securities, Fannie Mae did not transfer the IO and PO securities and thus did not meet the conditions in FAS 125. Although management initially obtained KPMG's approval for its synthetic MBS accounting, it appears that by 1998, management, including Boyles, had concealed information from KPMG for fear that KPMG might require a change in the accounting for IO MBS, including recognition of impairment losses. Both Howard and Spencer were aware of this fact. Management failed to inform the Board of its methodology for synthetic MBS until OFHEO raised questions in April 2004. PART H: SECURITIZATION OF WHOLLY-OWNED MBS In this section of our Report, we discuss pools of mortgages in which Fannie Mae came to acquire a 100 percent interest. - ---------- (1097) See THE MEANING OF PRESENT FAIRLY IN CONFORMITY WITH GENERALLY ACCEPTED ACCOUNTING PRINCIPLES, FOR NONGOVERNMENTAL ENTITIES, Amendment to Statement on Auditing Standards No. 69 P .09, (Am. Inst. of Certified Pub. Accountants 2005). (1098) See FAS 125 PP 9, 10(a), 11(a)-(b), 12. (1099) Id. 291
I. INTRODUCTION In its February 11, 2005 letter to Stephen B. Ashley, OFHEO raised concerns regarding Fannie Mae's treatment of MBS in which the Company came to own 100 percent of the certificates in a trust.(1100) In that circumstance, Fannie Mae continued to report its investment in the MBS as a security. The issue OFHEO raised was whether Fannie Mae was required to consolidate the trust, thereby recognizing the mortgage loans underlying the security.(1101) We discuss below OFHEO's concerns regarding Fannie Mae's approach to the consolidation issue, the relevant accounting literature, and the facts we uncovered in the course of our review. Our conclusion is that the Company's accounting in this area, although designed to avoid the need to consolidate the loans onto its balance sheet, was the result of an inadvertent misapplication of the accounting literature. The Company's accounting based on the misapplication was reversed after Financial Standards learned of its mistake and corrected the Company's accounting policy. Securities Accounting then assessed the impact of the error on the Company's financial statements and determined that impact to be immaterial. We understand that Fannie Mae is reviewing the latter conclusion as part of its restatement effort. II. BACKGROUND A. Accounting Literature on Consolidation In the normal course of its business, Fannie Mae issues guarantees to holders of securities backed by pools of mortgage loans. In a majority of its securitization transactions (referred to as "lender swap transactions"), lenders transfer pools of mortgage loans meeting certain criteria to trusts that Fannie Mae establishes and for which it serves as trustee. The lenders usually receive a certificate (the MBS) evidencing the right to receive the cash flows from the underlying loans, less a guaranty fee retained by Fannie Mae. If the cash flows from the underlying mortgages are not sufficient to meet scheduled principal and interest payments due to the certificate holder, the Company makes up any deficiency under the terms of its guaranty. Under accounting literature that became effective in 2003, the Company records the fair value of the guaranty liability on its balance sheet as a liability on issuance of the certificate; it subsequently recognizes an additional liability for probable losses it will be required to absorb under its guaranty. - ---------- (1100) See February 11, 2005 OFHEO Letter, FMSE-IR 547318-30. (1101) Id. at FMSE-IR 547329. 292
An interpretation issued by the FASB in January 2003 provides guidance on when consolidation of entities meeting certain conditions is required.(1102) FIN 46 provides a scope exclusion for entities that meet the FAS 140 definition of a Qualifying Special Purpose Entity ("QSPE").(1103) The results of this determination can be significant, as the accounting for loans and securities is not identical in all circumstances. FIN 46 governs the consolidation of the Fannie Mae trusts. Under FIN 46's scope exclusion, Fannie Mae is not required to consolidate the trust and its underlying loans if the trust is a QSPE.(1104) The requirements for QSPE status are addressed in FAS 140 and include a two-part test.(1105) First, the trust "cannot be unilaterally dissolved by the transferor, its affiliates, or its agents."(1106) Second, the trust must satisfy one of two conditions: (a) at least ten percent of the fair value of the trust certificates must be held by third parties, or (b) the transfer of the loans must be a "guaranteed mortgage securitization" ("GMS").(1107) FAS 140 defines a GMS as "a securitization of mortgage loans that . . . includes a substantive guaranty of a third party."(1108) Fannie Mae treats the securitization of pools of loans under the lender swap program as GMSs. To the extent that no party has the unilateral ability to dissolve the trust underlying the MBS, then the trusts are QSPEs (assuming other conditions are met). B. OFHEO's Concerns In the February 11 OFHEO Letter, OFHEO noted that, in certain circumstances, Fannie Mae acquired ownership of 100 percent of the certificates in a - ---------- (1102) See CONSOLIDATION OF VARIABLE INTEREST ENTITIES: AN INTERPRETATION OF ARB NO. 51, FASB Interpretation No. 46 (Fin. Accounting Standards Bd. Jan. 2003) (hereinafter "FIN 46"). (1103) See id. P 4. (1104) See id. (1105) See ACCOUNTING FOR TRANSFERS AND SERVICING OF FINANCIAL ASSETS AND EXTINGUISHMENTS OF LIABILITIES (A REPLACEMENT OF FAS STATEMENT NO. 125), Statement of Fin. Accounting Standards No. 140 (Fin. Accounting Standards Bd. 2000) (hereinafter "FAS 140"). (1106) See id. P 36. (1107) See id. (1108) See id. P 364. 293
mortgage-backed security.(1109) Moreover, according to OFHEO, Fannie Mae's accounting policy stated that ownership of 100 percent of a mortgage-backed security gives Fannie Mae the unilateral right to dissolve the underlying pool.(1110) OFHEO cited two approaches that the Company had taken to address "potential consolidation problems" associated with wholly-owned MBS.(1111) According to OFHEO: "These solutions were to transfer wholly owned pools of securities from AFS ["available-for-sale"] to HTM ["held-to-maturity"] or to sell 1% of the wholly owned pool to a third party."(1112) OFHEO observed that Fannie Mae "also developed a process of forming 'mega pools', the collateral of which is comprised of multiple wholly owned pools, so that the sale of 1% of the 'mega pool' could solve the consolidation issue."(1113) OFHEO raised concerns regarding the Company's determination that it could avoid consolidation of the trust either through the sale of a one percent interest in a mega pool or by reclassifying securities from AFS to HTM.(1114) According to OFHEO, "it appears that the Enterprise has created accounting policies related to its MBS pool activities whose primary purpose is to avoid consolidation."(1115) C. The Company's Approach to Consolidation Fannie Mae considered the effect of FIN 46 and the consolidation issue during the first half of 2003. An undated document entitled "Portfolio AFS Whole Pool Movement Analysis, Draft v.2" identified the two strategies noted in the February 11 OFHEO Letter as strategies that could be used as part of a quarterly FIN 46 "clean-up" process.(1116) The document outlined several strategies and stated that "[s]elling as little as 1% of a pool, including newly formed mega pools, is enough to satisfy the requirement of - ---------- (1109) February 11 OFHEO Letter at FMSE-IR 547329. (1110) Id. (1111) Id. (1112) Id. (1113) Id. (1114) Id. (1115) Id. (1116) See Portfolio AFS Whole Pool Movement Analysis, Draft v.2, dated May 27, 2003, FMSE 81258-59. 294
not having control of the security."(1117) The document also discussed "[m]ovements from AFS to HTM" as another mechanism "to ensure compliance."(1118) A document dated the following month captioned "Whole Pools Update" provides further background on the Company's deliberations.(1119) It stated that a number of different options for addressing the consolidation issue had been considered and "[s]enior management settled on a combination" of three strategies: (1) move pieces of whole pools that had slow prepayment rates from AFS to HTM; (2) "sell 1% of the average prepaying whole pools with balances in AFS in the form of megas"; and (3) sell AFS securities that prepay quickly.(1120) The document further indicated that, by that time, the Company had created and sold interests in four megas; that it had sold "100% of the fast prepaying AFS to the street"; and that the Company planned to redesignate MBS with an aggregate balance of $7.2 billion from AFS to HTM.(1121) The FIN 46 "clean-up" process in future months "may include" additional redesignations of securities from AFS to HTM, sales of AFS securities, and/or the formation of megas.(1122) Fannie Mae finalized its accounting policy with respect to FIN 46 consolidation and the clean-up strategies during the second half of 2003. A draft policy memorandum prepared by Paul Salfi and Angela Hairston of Financial Standards opened with two propositions: That a "wholly-owned MBS trust fall[s] within the scope of FIN 46," and "[a]s owner of 100 percent of the beneficial interest in an MBS pool and guarantor of the trust, Fannie Mae has the unilateral ability to dissolve the MBS pool."(1123) The authors concluded that, in those circumstances, FIN 46 would require consolidation of the trusts.(1124) The draft policy addressed two of the options the Company had chosen for addressing the consolidation issue. These options "involve either a) transferring the securities from AFS to HTM or b) selling a portion of the pool to give up control so that - ---------- (1117) Id. at FMSE 81258. (1118) Id. (1119) See Whole Pools Update, dated Jun. 19, 2003, FMSE 60777-80. (1120) Id. at 60777. (1121) Id. at 60778. (1122) Id. at 60779. (1123) Draft mem. from Paul Salfi and Angela Hairston to Andrew McCormick, dated July 2, 2003, FMSE-E 380523-27, at FMSE-E 380523. (1124) Id. 295
we are not required to consolidate."(1125) Jonathan Boyles stated during an interview that selling a portion of the pool meant selling at least one percent. In his view, the sale of any interest in the megas would preclude the Company's unilateral ability to alter the trusts, but he believed that a one percent minimum threshold was advisable. As to the redesignation of AFS to HTM, a later draft of the same memorandum stated that the redesignation would make consolidation unnecessary.(1126) It reasoned that because the accounting for HTM securities and held-for-investment ("HFI") loans are similar,(1127) the amounts recorded would be the same regardless of whether Fannie Mae consolidated and recorded the value of the underlying HFI loans or recorded the value of the HTM securities.(1128) In addition, the Company "would continue to reflect a guaranty liability for MBS on our balance sheet."(1129) In these circumstances, the memorandum concluded, "consolidation of the loans underlying wholly-owned HTM MBS pools would not affect Fannie Mae's financial position" and therefore "it would not be meaningful to the reader of our financial statements if we reclassed the [HTM] MBS to whole loans."(1130) With regard to the creation of mega pools, this draft memorandum incorporated the conclusion from the previous draft that, upon sale of a portion of the mega pools, Fannie Mae would not have the unilateral ability to dissolve a trust.(1131) The final version of the policy, signed by Hairston and addressed to Leanne G. Spencer, was dated October 6, 2003.(1132) In substance, it was the same as the July 30 draft memorandum. - ---------- (1125) Id. at FMSE-E 380524. The draft did not address the third option of selling outright pools that prepaid quickly, as outlined in the June 19 Whole Pools Update memorandum. (1126) See Draft mem. from A. Hairston to Leanne G. Spencer, dated Jul. 30, 2003, FMSE-IR 710015-18, at FMSE-IR 710016. (1127) HTM securities and HFI loans are carried at amortized historical cost. HFS loans, by contrast, are recorded at the lower of historical cost or market, with adjustments recorded in income. (1128) See Draft mem. from A. Hairston to L. Spencer, dated Jul. 30, 2003, FMSE-IR 710015-18, at FMSE-IR 710016. (1129) Id. (1130) Id. (1131) Id. at FMSE-IR 710017. (1132) See Mem. from A. Hairston to L. Spencer, dated Oct. 6, 2003, FMSE 81253-57. 296
In early 2004, another accountant in Financial Standards, Jennifer Liner, was assigned to work on the accounting for a transaction that also raised issues under FIN 46 and the QSPE scope exception. Liner stated that, in the course of her work on that transaction, she had reviewed Fannie Mae's accounting policies on the subject. She also stated that she discussed certain issues with the FASB staff.(1133) In particular, Liner asked whether the securitization of MBS would be considered a GMS as defined by FAS 140.(1134) The FASB staff responded in the negative.(1135) Liner subsequently described the conversation in a new accounting policy; referring to FAS 140's definition of a GMS as "a securitization of mortgage loans," the FASB staff opined that a GMS "only applies to loans securitized into securities, not resecuritizations of securities," as in the case of a mega.(1136) It appears that Liner presented this reasoning to the FASB staff and the staff did not disagree. Liner reported her discussion to others in the Controller's Office(1137) and began to prepare the new policy regarding wholly-owned MBS. The new policy, dated November 2, 2004, addressed both the mega pools issue and the effect of redesignation of MBS from AFS to HTM on the requirement that the Company consolidate the trusts.(1138) As to the megas, Liner wrote that the Company had "viewed these as guaranteed mortgage securitizations since the underlying securities going into the [mega] are (or were believed to be) guaranteed mortgage securitizations, and the Trust is just a pass-through entity."(1139) She noted the FASB staff's concurrence with her conclusion - ---------- (1133) E-mail from Jennifer Liner to Vickie Lusniak, dated Aug. 12, 2004, Zantaz document 1048922. (1134) Id. (1135) Id. (1136) Mem. from J. Liner to Distribution, dated Nov. 2, 2004, FMSE-IR 357892-902, at FMSE-IR 357893 n.4 (hereinafter "Liner Mem."). (1137) See E-mail from J. Liner to Jonathan Boyles, et al., dated Aug. 13, 2004, Zantaz document 1048922, at 1 ("[W]e confirmed with FASB that a guaranteed mortgage securitization only includes securitization of loans to securities, not securities to securities"). (1138) See Liner Mem. In connection with the discussion of megas, Liner also considered similar issues raised in other types of pools, such as PPS pools and portfolio REMICs. PPS pools are like megas, but rather than receiving and securitizing a pool created by another party, Fannie Mae creates the pool itself from loans that it acquires for its portfolio. A Portfolio REMIC is a similar structure, but the resulting trusts have different tranches; security holders acquire an interest in a particular tranche of the trust. Id. at FMSE-IR 357892-93. (1139) Id. at FMSE-IR 357894. 297
that "[m]egas are resecuritizations of securities (rather than loans), therefore they do not meet the definition of guaranteed mortgage securitization in FAS 140."(1140) In order to be a QSPE and avoid consolidation, therefore, "we must sell at least 10 percent of the fair value of the beneficial interest in the Trust to third parties and 10 percent of the fair value of the beneficial interests in the Trust must be held by third parties throughout the life of the Trust."(1141) Liner concluded that the megas would be subject to consolidation under FIN 46.(1142) As to the redesignation of securities from AFS to HTM, Liner's analysis acknowledged that the accounting for the HTM securities and HFI loans is the same, at least with respect to each being recorded at its amortized cost.(1143) She also acknowledged the fact that Fannie Mae reports its guaranty of the MBS as a liability, which would be inconsistent with consolidation of the loans on the Company's balance sheet (that is, Fannie Mae could not record a liability to itself).(1144) She concluded that FIN 46 required consolidation of the pools and that, following consolidation, the guaranty should be reported as an allowance for loss associated with the consolidated loans.(1145) Securities Accounting prepared an analysis of the impact of the error in the original policy.(1146) The analysis concluded that the transfer of an interest in the mega pools, which had been recorded as a sale, should have been recorded as a secured borrowing, since the assets in the mega pools had been (or should have been) consolidated onto the Company's balance sheet.(1147) According to Boyles, the Company concluded that it was not required to consolidate the individual loans into its balance sheet because the MBS were guaranteed mortgage securitizations and Fannie Mae's transfer of an interest in the megas precluded unilateral dissolution of the trusts. The - ---------- (1140) Id. (1141) Id. at FMSE-IR 357894. (1142) Id. at FMSE-IR 357895. Liner's reasoning with respect to PPS pools differed slightly from her reasoning with respect to megas, but the conclusion was the same. Id. at FMSE-IR 357894-95. (1143) Id. at FMSE-IR 357896. (1144) Id. at FMSE-IR 357897. (1145) Id. (1146) See Mem. from Chris LeBel to Ilene Topper, dated Nov. 11, 2004, FMSE-IR 357873-76. (1147) Id. at FMSE-IR 357873-74. 298
analysis concluded that the cumulative impact of the accounting error on the Company's earnings was $17.9 million.(1148) According to Salfi, KPMG was aware of the Company's deliberations regarding the application of FIN 46 during 2003. III. CONCLUSIONS Based on the evidence available to us, we agree with OFHEO that Fannie Mae's focus during the first half of 2003 was on the development of strategies for avoiding consolidation of its trusts onto its balance sheets. Consolidation would have had some impact on the Company's balance sheet and income statement. Boyles also noted that the "clean-up" and, if necessary, consolidation of the individual mortgage-backed security trusts among the thousands that Fannie Mae had established would be an enormous undertaking. In his view, therefore, the Company's concerns were also administrative, and the creation of mega pools and similar strategies were a means of avoiding that burden. However, the evidence overall leads us to the conclusion that this accounting error in the 2003 policy was an inadvertent misapplication of the accounting literature. As to mega pools, the documentation reveals that Fannie Mae's accounting team was focused almost exclusively on one issue: that Fannie Mae's ownership of 100 percent of a trust would give it the unilateral ability to dissolve the trust, and therefore might preclude treatment of the trust as a QSPE. It does not appear that the authors of the policy gave much thought to the separate issue of whether the mega pools (or similar structures) would meet the second part of the QSPE requirement - particularly, whether the megas would be considered GMSs. As noted above, they simply operated on the assumption that they would, because the MBS included in the mega pools were considered GMSs. The conclusion that the redesignation of securities from AFS to HTM had the same impact on the Company's balance sheets as consolidation of the underlying loans was incorrect because loans and securities are presented separately on the Company's balance sheet. It appears that the authors of the original policy focused on the similar manner in which HFI loans and HTM securities are carried. In addition, their approach was consistent with its treatment of the guaranty as a liability. They do not appear to have considered the approach Liner devised of reclassifying the guaranty liability as a loan loss reserve. Most importantly, Liner addressed the QSPE issue with the FASB staff in 2004 and, following that conversation, Financial Standards reversed its previous policy - ---------- (1148) Id. at FMSE-IR 357874. 299
promptly.(1149) It appears that Liner prepared her new policy memorandum with no objection from others at the Company and without substantial concern for the impact it ultimately would have on the Company's financial statements. PART I: ACCOUNTING FOR INCOME TAX RESERVES AND CERTAIN TAX-ADVANTAGED TRANSACTIONS I. INTRODUCTION In this Part, we discuss issues related to Fannie Mae's accounting for income tax reserves. Our review centered on the reserves related to tax credits the Company received from synfuel partnerships and on the tax benefits associated with certain tax-advantaged transactions known at Fannie Mae as Short Term Interest Securities ("STIS"). We focused on how Fannie Mae reported tax contingencies and benefits on its financial statements. II. BACKGROUND A. Applicable Authoritative Literature and Principles Financial Accounting Standards Board ("FASB") Statement No. 5, Accounting for Contingencies ("FAS 5"), is the governing standard for recognition of a liability for tax-related contingencies.(1150) Under FAS 5, a liability for a tax contingency would not be recognized until it is probable that "a liability had been incurred at the date of the financial statements," and the amount of potential loss is reasonably estimated.(1151) Because certain of Fannie Mae's tax-advantaged transactions created temporary differences (i.e., differences between the book and tax basis of assets and liabilities), the Company was also required to follow the requirements set forth in FASB Statement No. 109, Accounting for Income Taxes ("FAS 109").(1152) FAS 109 sets forth the objectives of accounting for income taxes, which are to recognize: (1) the amount of taxes payable or refundable for the current year; and (2) deferred tax liabilities and assets - ---------- (1149) We do not address the separate question of whether the Company was required to consolidate the trusts in the period in which it acquired a 100 percent interest in the trust (that is, whether any FIN 46 "clean-up" could be prospective only). (1150) ACCOUNTING FOR CONTINGENCIES, Statement of Fin. Accounting Standards No. 5, (Fin. Accounting Standards Bd. 1975). (1151) See id. P 8. A company may not make an accrual for general or unspecified business risks, often referred to as "reserves for general contingencies." See id. P 14. (1152) ACCOUNTING FOR INCOME TAXES, Statement of Fin. Accounting Standards No. 109, (Fin. Accounting Standards Bd. 1992). 300
for the future tax consequences of events that have been recognized in an enterprise's financial statements or tax return. FAS 109 does not define the requisite confidence level that an enterprise must have to recognize financial statement benefits that result from a certain tax position. While practice has been mixed, a company was generally required by its outside auditor to conclude that it was more likely than not that the company's position would be sustained if challenged in order to recognize a tax benefit from a tax-advantaged transaction for financial reporting purposes. B. Fannie Mae's Accounting for Income Tax Reserves and Tax-Advantaged Transactions 1. Tax Reserves (a) Fannie Mae's Process for Determining Amounts in the Tax Reserve Interviewees have explained that Fannie Mae's income tax reserve was reviewed at quarterly meetings. Personnel from Corporate Tax(1153) and the Legal Department participated in these meetings ("quarterly tax reserve meetings") and reviewed the tax reserve by discussing the Company's potential or unresolved tax disputes with the Internal Revenue Service ("IRS"). The group considered the likelihood that the IRS would prevail in a dispute over each issue and would reach a consensus as to the percentage likelihood of such loss. Those percentages were used in turn to compute the amount of a tax reserve for each item. Minutes were not kept of the quarterly tax reserve meetings. Although no written policy existed to guide the decisions made at the quarterly tax reserve meetings, we understand that the participants at these meetings came to a consensus by considering existing precedent and the degree of uncertainty surrounding each issue. Corporate Tax Director Lavina Chawla explained that Legal Department personnel played a pivotal role in the assessment and would consider factors such as tax law guidance, any perceived ambiguity in the law, IRS inquiry trends, previous IRS positions, and the conduct of other companies on similar issues. Carolyn Swift, currently Special Tax Counsel--Corporate Tax,(1154) explained that Fannie Mae tax attorneys typically were already familiar with the issues prior to the quarterly tax reserve meetings, since they likely had discussed them when the Company first considered entering into the transactions. In addition, Fannie Mae occasionally asked outside counsel for advice with respect to potential or unresolved tax disputes. - ---------- (1153) Although there is no formal tax department at Fannie Mae, we refer collectively to the Company's tax accountants as "Corporate Tax." (1154) Carolyn Swift was the head of the tax attorney unit from 1989 to 2002. 301
Since recording the tax reserve was ultimately a matter of financial accounting, after the quarterly tax reserve meetings, the head tax accountant at the time (either a Director or Vice President in Corporate Tax) would meet with the Company's Controller to review the decisions made with respect to the tax reserve at the quarterly tax reserve meetings. We understand that these meetings took place on a quarterly basis while Leanne G. Spencer served as Controller from 1999 through 2005, but during Sam Rajappa's tenure as Controller, prior to 1999, these meetings occurred only when tax reserve amounts were significant. Financial Reporting was responsible for recording the tax reserves after the meeting with the Controller. As noted, we interviewed a number of individuals who participated in the quarterly tax reserve meetings and reviewed documents relating to Fannie Mae's tax reserve. The process the Company used to establish the reserves was reasonable. Given the lack of documentation supporting the rationale for tax reserve determinations, and the fact that the establishment of reserve amounts depended on the exercise of collective judgment at the time the reserve level was set, we have not reached any conclusions as to whether the amounts agreed to at the quarterly tax reserve meetings were reasonable and whether they represented known tax liabilities.(1155) (b) Decisions Regarding the Amount Recorded in the Tax Reserve Documents suggest that management may have recorded amounts to the tax reserve for inappropriate earnings management purposes. A 1996 memorandum written by Spencer, then Vice President--Financial Reporting, to Timothy Howard suggests that Spencer may have contemplated making adjustments to the tax forecasts, and possibly the tax provision or tax reserve, that were based on decisions unrelated to known tax liabilities.(1156) In particular, with regard to the Company's forecast for the third quarter of 1996 quarter, Spencer wrote: We've done the following: . . . reversed the tax entry we booked in September for qtr [sic] management of $2.3 million and assumed that reversal would occur in Dec 96 . . . [a]nd, I have now held nothing back such as the - ---------- (1155) The lack of specific documentation relating to tax reserve amounts does not necessarily indicate that the reserves were inappropriate, but it precludes us from conducting a specific analysis of the issue. Furthermore, we were not able to interview certain individuals who played important roles in Corporate Tax during the period on which our review focused. William E. Einstein (a former Vice President) declined our request for an interview, and we could not locate Christina Immelman (a former Director). Finally, Leanne Spencer stopped cooperating with our investigation and was thus unavailable. (1156) See Undated mem. from "Leanne" to "Tim," FMSE-IR 359599-601, at FMSE-IR 359599. 302
previous earnings management items that I've been plugging to the tax line over the last several forecasts, with the reversing of the tax entry."(1157) Unfortunately, we were unable to question Spencer about this document because it was produced after Spencer announced her intention to decline all future requests for interviews.(1158) Regardless, this memorandum suggests that Spencer may have been over-reserving for tax liabilities that she later reversed to income. A second document raises additional concerns about the recording of the tax reserve amounts. A 1997 document entitled "December Earnings Actions" found in Shaun Ross's files(1159) listed three "Proposed Actions," including "Accrue Interest for Tax Items in Dispute (Misc)" in the amount of $12 million.(1160) When shown the document, Ross said that he neither recognized it nor understood the material represented. Rajappa, who was Controller in 1997, said that the document did not provide enough context for him to understand whether the "proposed actions" were "optional."(1161) Based on our understanding of the process used to determine the tax reserve, if management thought that there were tax items in dispute, accruing interest for such items would not be optional.(1162) A third document is a memorandum from Fannie Mae tax attorney Hal Gann to Howard, entitled "Usefulness of a `Cushion' for Deferred Taxes." Gann wrote that Fannie Mae could better evaluate and manage its tax law exposure by "not booking into current earnings all of the savings projected for various tax-advantaged transactions," - ---------- (1157) Id. (1158) See discussion supra Chapter II. (1159) Shaun Ross was a Financial Analyst in the Controller's Office before Fannie Mae promoted him to Project Manager in 1997. (1160) Undated December Earnings Actions, Zantaz document 1512032. While the document itself is undated, we believe its reference to an achievable EPS amount of $2.8325 indicates its creation year as 1997. The reported EPS for year-end 1997 was $2.83. Fannie Mae Information Statement, dated Mar. 31, 1998, at 19, available at http://www.fanniemae.com/markets/debt/pdf/infostmtmar1998.pdf. (1161) Rajappa speculated that the creator of the document may have intended to illustrate the amount included in the forecast, but labeled it as a proposed action because further work was necessary before the forecasted figure could be finalized. (1162) Rajappa generally agreed with this conclusion, stating that accruing interest is not optional for Fannie Mae when it believes it is "going to lose" an item in dispute. 303
and instead using part of the tax reserve as a "cushion."(1163) Gann noted how Fannie Mae could manage the "cushion" to result in smoother Company earnings: [A] book "cushion" can help you maintain consistent growth in earnings. If we find an idea that can trim $35 million off of our 1999 tax expense by capturing a permanent loss of up to $100 million, should we execute all $100 million and risk depressing 2000 earnings by comparison if we do not find as many good ideas in that year? Should we limit ourselves to $60 million of losses -$21 million of tax savings - just to avoid setting the bar too high for 2000? A "cushion" may offer a compromise that allows us to capture all $100 million of loss on the tax return, but book only $21 million of the savings into 1999 earnings. The additional $14 million of reserved tax savings can be managed, like a reserve for credit losses, to walk that fine line between saving enough for a rainy day and depressing earnings by being overly pessimistic.(1164) While no interviewees recalled Spencer ever disagreeing with the tax reserve amounts reached by consensus at the quarterly tax reserve meetings, the three documents discussed above call into question whether Spencer or other recorded amounts relating to the tax reserve inappropriately. 2. Fannie Mae's Recognition of Synfuel Tax Credits (a) Overview of Synfuel Tax Credits Beginning in 1998, Fannie Mae acquired limited partnership interests in three Internal Revenue Code Section 29 synthetic fuel facilities.(1165) The investments Fannie Mae made in these partnerships essentially allowed the Company to earn tax credits based on the fuel production of these alternative energy facilities.(1166) Before - ---------- (1163) Mem. from Hal Gann to Timothy Howard, dated Sept. 28, 1998, Zantaz document 2254187, at 1. Leanne G. Spencer, Einstein, and Swift are listed as carbon copy recipients of this memorandum. Id. at 2. (1164) Id. at 2. (1165) Mem. from Jonathan Boyles and Sheslie Royster to Executive Officers, dated July 9, 2003, Zantaz document 807185, at 1. Internal Revenue Code Section 29 was renumbered and amended in 2005 as Section 45K. It provides a production tax credit for "liquid, gaseous, or solid synthetic fuels produced from coal." I.R.C. Section 45K (c)(1)(C) (2005). (1166) Mem. from J. Boyles and S. Royster to Executive Officers, dated July 9, 2003, Zantaz document 807185, at 1. 304
investing, Fannie Mae required that each partnership obtain a Private Letter Ruling ("PLR") from the IRS affirming that the proposed synthetic fuel production process used by the partnerships resulted in the creation of a synthetic fuel that would make investors in the partnerships eligible for the tax credit.(1167) Fannie Mae initially purchased interests in three synfuel partnerships, an investment that totaled $53 million.(1168) In addition, Fannie Mae made periodic "tax credit fundings" based on the partnerships' synfuel production levels.(1169) Fannie Mae booked tax credits generated by the investments and claimed tax deductions for the partnership-generated operating losses.(1170) Based on management's sense that the IRS had increased its scrutiny of synfuel investments, it determined that it would not recognize all of the synfuel tax credits earned on its financial statements, and instead reserved a portion of the credits.(1171) The uncertainties increased when the IRS began auditing numerous synfuel syndicators in 2002, including two of the Fannie Mae partnerships.(1172) For purposes of allocating a reserve amount for the synfuel tax credits in 2002, individuals who participated in the quarterly tax reserve meetings apparently reached a consensus that a reserve of twenty-five percent of the total tax benefit would be appropriate.(1173) Although interviewees could not recall why twenty-five percent was chosen as the appropriate level for the tax - ---------- (1167) Id. (1168) Id. (1169) Id. (1170) Id. (1171) See, e.g., Mem. from W. Einstein to T. Howard, dated Sept. 1, 2000, FMSE-IR 375930-37, at FMSE-IR 375934 (expressing concern that the IRS might view some synfuel partnership practices as "spray on" tax credits and treat them as abusive corporate tax shelters). Other evidence suggests that management added to the tax reserve for synfuel tax credits based on increased IRS scrutiny. (1172) KPMG Tax Credits Documentation, dated Dec. 31, 2003; Undated 2002 Tax Strategies PowerPoint Presentation, Zantaz document 482600, at 2. (1173) See Synfuel Reserves, dated Feb. 2004 (listing a twenty-five percent "required reserve" for tax credits received from 1999 through 2002); see also Synfuel Reserve Analysis, dated Oct. 14, 2002, FMSE-IR 347242-57 (hereinafter "Synfuel Reserve Analysis Folder"), at FMSE-IR 347249 (listing the reserve percentage as twenty-five percent for year-end 2002); Undated 2002 Tax Strategies PowerPoint Presentation, Zantaz document 482600, at 2 (recommending the creation of a twenty-five percent reserve for the synfuel tax credits). 305
reserve, a PowerPoint slide entitled "2002 Tax Strategies" states that this amount "would make our risk assessment similar to our largest partner in this [sic] deals...."(1174) (b) Amounts Actually Reserved for the Synfuel Tax Credit Documents indicate that Fannie Mae in fact maintained a reserve that exceeded twenty-five percent of the Company's total synfuel tax credits. A spreadsheet dated October 14, 2002 analyzed the 2002 synfuel tax credit reserve amounts and listed the "reserve %" line-item as twenty-five percent.(1175) It projected Fannie Mae's actual benefit received through year-end to be just under $70 million, and accordingly calculated the "Reserve Requirement" to be just over $17 million. However, a subsequent line-item listed the "current reserve" for the unrecognized synfuel tax credits as approximately $45 million, which would be more than $27 million over the twenty-five percent reserve. The spreadsheet further indicates that this "excess" amount was to be released at year-end.(1176) If management changed its synfuel reserve estimate for the end of 2002 and thought that $28 million of the reserve was "excess," it should have recognized that change by releasing $28 million of the reserve at that time.(1177) However, it does not appear that Fannie Mae released this excess amount at year-end 2002. A Fannie Mae document found in KPMG's workpapers states that Fannie Mae's cumulative synfuel tax credits as of year-end 2002 was $70.4 million and indicated that the then current reserve level was $46.2 million, which was $28.6 million in excess of the twenty-five percent "required reserve" amount.(1178) A September 15, 2003 tax reserve schedule also notes $28 million of excess reserve associated with the - ---------- (1174) Undated 2002 Tax Strategies PowerPoint Presentation, Zantaz document 482600, at 2. (1175) Synfuel Reserve Analysis Folder at FMSE-IR 347249. (1176) Id. (listing the $27,695,719 as a line-item labeled "Release"). (1177) ACCOUNTING CHANGES, Accounting Principles Board Opinion No. 20 (As Amended), P. P. 10, 13, 31, 36 (Fin. Accounting Standards Bd. 1971). Accounting Principles Board Opinion No. 20, Accounting Changes (APB 20") requires management to update estimates and correct errors in previously issued financial statements, including those related to tax reserves. (1178) Synfuel Reserves, dated Feb. 2004. 306
synfuel tax credits.(1179) A handwritten note on this document indicates that the $28 million was "on top of" the twenty-five percent reserve.(1180) No one we interviewed offered any reason why the Company would have reserved an amount above the twenty-five percent level. Nevertheless, we note that, in 2003, KPMG concluded that the tax credits included in Fannie Mae's tax provision, including the synfuel tax credits, were materially correct.(1181) (c) Reversal of Amounts Reserved for Synfuel Tax Credits In July 2003, the IRS further called into question the validity of the synfuel tax credits when it issued Announcement 2003-46, which stated the IRS had reason "to question the scientific validity of test procedures and results that have been presented as evidence that fuel underwent a significant chemical change," and indicated the possibility that the IRS might revoke PLRs that allowed enterprises to recognize tax credits related to investments in synfuel producers that relied on improper procedures and results.(1182) However, in November 2003, the IRS issued Announcement 2003-70, which concluded that the procedures and results relating to the chemical changes in synfuel production were valid if applied "in a consistent and unbiased manner," and thus investments in partnerships that produced synfuel accordingly would be eligible for tax credits.(1183) In addition, the audits of Fannie Mae's synfuel partnerships ended with positive results in 2003, leading Fannie Mae to determine that it no longer needed to - ---------- (1179) Fannie Mae IRS Interest and Tax Reserve Accrual Analysis as of December 31, 2003, dated Sept. 15, 2003, Zantaz document 2146859. This tax reserve schedule indicates that the unbooked synfuel credits were included as "Over-Provided Taxes" in the "Reserve Available." Id. (1180) Id. (1181) KPMG Tax Credits Documentation, dated Dec. 31, 2003. The workpapers indicate that KPMG "agreed the actual (or estimated) credits reported by the operator to the amounts recorded on the Synfuel credit worksheet." Id. KPMG also reconciled "the tax credits reported on the subledgers to the amounts reported in the [general ledger] and tax provision." Id. (1182) IRS Announcement 2003-46, Section 29 - Test Procedures and Significant Chemical Change, dated July 28, 2003, available at http://www.irs.gov/irb/2003-30_IRB/ar13.html (last visited Feb. 17, 2006). (1183) IRS Announcement 2003-70, Section 29 - Test Procedures and Significant Chemical Change, dated Nov. 17, 2003, available at http://www.irs.gov/irb/2003-46_IRB/ar12.html (last visited Feb. 17, 2006). 307
reserve amounts in its tax reserve related to synfuel tax credits.(1184) As a result, Fannie Mae released its reserves for these credits. According to internal Fannie Mae records, the Company released $28,570,379 of its synfuel tax reserve during the third quarter of 2003.(1185) Records show that $30,481,304 was released near the end of the fourth quarter.(1186) Witnesses were unable to explain why the amounts were released over two quarters, and we have seen no documents that offer a reason. Since there is no documentation regarding the rationale for releasing the synfuel reserve over two quarters, we are unable to conclude whether releasing the amounts in two stages during 2003 was appropriate. 3. Issues Related to the STIS Transactions (a) Overview of the STIS Transactions The STIS transactions had two steps. Fannie Mae would first transfer a pool of mortgage loans into a REMIC trust, and then it would sell interests in the trust that were entitled to the first six to eight coupon interest payments.(1187) Fannie Mae retained all other cash flows related to the mortgages, and it accounted for the proceeds received as a borrowing, i.e., the Company recognized no income for book purposes.(1188) The STIS transactions provided a tax benefit in that they allowed Fannie Mae to recognize a loss for tax purposes in the year of sale.(1189) The Company would recognize taxable income in subsequent periods as the discount created from the loss on the sale was accreted for tax purposes.(1190) - ---------- (1184) See E-mail from Lavina Chawla to C. Swift, dated June 1, 2004, Cataphora document i.1086214772000.m789455637, at 3. (1185) See Homesite General Ledger for account 811116, Journal ID JE25313 (setting forth the recognition of the reserve into income); see also Homesite General Ledger for account 811117, Journal ID JE25313 (same). (1186) See Homesite General Ledger for account 811115, Journal ID JE25608 (setting forth the recognition of the reserve into income); see also Homesite General Ledger for account 811117, Journal ID JE25608 (same). (1187) Mem. from C. Swift to Ann M. Kappler, dated Nov. 20, 2003, Zantaz document 558508. (1188) Mem. from Richard Stawarz to S. Rajappa, dated June 16, 1998, FMSE-IR 301919-22, at FMSE-IR 301919. (1189) E-mail from James M. Harrington to J. Boyles, dated Feb. 12, 2002, FMSE-IR 597083. (1190) Id. 308
In this manner, the STIS transactions allowed Fannie Mae to monetize a portion of its loan portfolio and obtain a tax benefit. Fannie Mae received cash up-front, while deferring recognition of the taxable income on the accretion of the discount over the life of the underlying loans. To accomplish this goal, Fannie Mae stripped only a de minimis(1191) amount of interest from each loan, such that the underlying loan could be treated as "unstripped" for tax purposes, and the Company could include all remaining interest payments in income, as if the transaction had not taken place.(1192) KPMG(1193) created a complex model that Fannie Mae used to identify and quantify which loans to strip payments from in order to achieve these goals.(1194) Fannie Mae entered into two STIS transactions, one in 1998 ("STIS 1") and one in 2000 ("STIS 2"). Fannie Mae originally contemplated entering into STIS transactions on a quarterly basis, but ultimately did not due to market conditions. (b) Legal Opinion Documents show, and Swift confirmed, that Fannie Mae obtained a draft "should" opinion letter from an outside law firm prior to entering STIS 1,(1195) but never received a final, signed version of the letter.(1196) Swift explained that Fannie Mae often chose not to incur the expense of obtaining a final legal opinion when deciding whether to enter tax-related transactions.(1197) Similarly, Fannie Mae opted not to obtain a should opinion letter for STIS 2 after outside counsel confirmed that its views with regard to the transaction were the same as for STIS 1.(1198) - ---------- (1191) The de minimis guidelines are set forth in I.R.C. Section 1273(a)(3) (2000). (1192) Mem. from C. Swift to A. Kappler, dated Nov. 20, 2003, Zantaz document 558508. (1193) The KPMG group that proposed the STIS transactions was separate from the KPMG audit group. (1194) See Letter from Steven M. Rosenthal to S. Rajappa, dated Feb. 17, 1998, FMSE-IR 368899-901. (1195) See supra text accompanying note 1210. (1196) E-mail from Michael Rufkahr to C. Swift and Alexis Moore, dated Oct. 29, 2003, Zantaz document 2116486-88, at 1. (1197) In fact, Swift added that Fannie Mae sometimes chose only to receive oral legal opinions as a further less expensive alternative to draft written opinions. (1198) Id. 309
(c) Business and Economic Justifications for the STIS Transactions With respect to tax-advantaged transactions, it is generally understood within the accounting profession that companies must have a legitimate business purpose unrelated to tax benefits to justify recognizing a benefit on tax-advantaged transactions.(1199) Fannie Mae's management was aware that any tax-advantaged transaction that the Company entered into also needed to have business and economic justifications unrelated to the tax benefits.(1200) Documents show that management determined that the STIS transactions had the following non-tax benefits: (1) Accelerated receipt of cash from sale of interest coupons; (2) Protection from the risk of extreme increases in the rate of prepayment speeds on the underlying mortgages; (3) Access to new investors that traditionally had not been willing to invest in Fannie Mae short-term debt; and (4) A positive financial statement impact.(1201) Several documents we reviewed during the investigation led us to focus on the existence and relative weight of the business purposes that management articulated as support for the STIS transactions. For example, a document containing handwritten notes that we understand were written by William E. Einstein, then Vice President--Corporate Tax, read: How critical is it that we have a business purpose? - w/no prepayment risk, we don't reach any new investors.(1202) - ---------- (1199) See Tina Steward Quinn & Tonya K. Flesher, A Weapon From the Past, JOURNAL OF ACCOUNTANCY, July 2002, available at http://www.aicpa.org/PUBS/JOFA/jul2002/quinn.htm (last visited Feb. 17, 2006). (1200) We note that Fannie Mae was not required to have an independent business justification to invest in synfuel partnerships. (1201) Mem. from W. Einstein and S. Rajappa to T. Howard, with attachments, dated Mar. 19, 1998, FMSE-IR 183262-69, at FMSE-IR 183263-64; Mem. from W. Einstein to T. Howard, with attachments, dated Feb. 25, 2000, FMSE-IR 301924-28, at FMSE-IR 301925-26; Accord Short Term Interest Security Additional Discussion Meeting, dated Jan. 7, 2000, FMSE-IR 376037. (1202) Fax cover sheet, from S. Rosenthal to W. Einstein, dated Oct. 29, 1997, FMSE-IR 182795. As noted above, Einstein has left Fannie Mae's employ and was not available to be interviewed. 310
Another document was a note from Rajappa to Howard, attached to a March 1998 memorandum entitled an STIS "Executive Summary."(1203) Rajappa wrote, in part: This is the `business justification' memo on STIS and as such influenced quite heavily by the lawyers (OGC [Office of General Counsel] & [outside counsel]).(1204) Later on the same page, Rajappa continued: The last two schedules depict that these deals are profitable (just barely!) even without considering the economic benefits of postponement of payment of taxes.(1205) Documents indicated that the expected pre-tax profit for STIS 1 was nearly $800,000, inclusive of all up-front costs.(1206) The expected pre-tax profit for STIS 2 was nearly $700,000, assuming no costs from STIS 1 were allocated to the second transaction.(1207) When asked about his first comment, Rajappa explained that his intent was to convey to Howard that both Fannie Mae's internal and outside counsel had reviewed and commented on this memorandum and that Einstein and Rajappa had not created this document alone. Rajappa stated that it was not his intent to convey to Howard that the lawyers had influenced the business justifications. Rajappa explained that his intent with regard to the second handwritten note to Howard was to convey that the deal made economic sense even without the tax benefits, even if the results could be characterized as "not that profitable." Swift and Jonathan Boyles, both of whom had knowledge of the STIS transactions at the time STIS 1 was entered in 1998, stated that they believed that the business and economic justifications were genuine and contemplated by management prior to entering the STIS transactions. - ---------- (1203) Handwritten notes from "Sam" to "Tim," dated Mar. 20, FMSE-IR 183262-69, at FMSE-IR 183268. (1204) Id. (1205) Id. (1206) Mem. from W. Einstein to File, with attachments, dated Nov. 2, 1999, FMSE-IR 376038-41, at FMSE-IR 376039. (1207) Mem. from W. Einstein to H. Gann, et al., with attachments, dated Nov. 3, 1999, FMSE-IR 375944-57, at FMSE-IR 375953 (an attachment estimating the profit based on a transaction with zero over-collateralization). 311
(d) Results of the STIS Transactions As it turned out, due to market conditions, the actual financial statement benefits of the STIS transactions were not as great as the Company had projected. Interest rates began to fall in 1998, causing prepayment rates on loans to increase, which both shortened the period during which Fannie Mae could defer recognizing the discount on the loans and caused the Company to incur prepayment risk. Without these benefits, it became uneconomical for Fannie Mae to engage in STIS transactions and, because of the poor market conditions, the Company decided not to enter these transactions quarterly as originally contemplated. Furthermore, management had planned to allocate the approximately $2.5 million in upfront costs among the quarterly STIS transactions the Company anticipated.(1208) Instead, Fannie Mae recognized all of the upfront costs as a cost of STIS 1.(1209) We understand that management decided to enter into STIS 2 in 2000 because it believed that interest rates would soon rise. III. FINDINGS We conclude that the process management used to determine its tax reserves appears reasonable; however, we were unable to form any conclusions as to whether specific tax reserves were appropriate. We also note that documents indicate that Leanne Spencer and Financial Reporting may have recorded amounts to tax reserves that were not connected to known tax liabilities, but instead were booked for inappropriate earnings management purposes. We conclude that management determined the amount of the reserve related to the Company's synfuel tax credits, but kept an unexplained additional or "excess" amount in the reserve at year-end 2002 that it released to earnings in 2003. In addition, management did not release the reserve related to synfuel tax credits all at once, but instead released a portion of the reserve in the third quarter of 2003 and the rest in the fourth quarter of 2003. Interviewees were unable to explain why the amounts were released over two quarters, and we have seen no documents that offer a reason. With regard to the STIS transactions management opted to obtain only a draft "should" opinion(1210) from outside counsel even though management expected the - ---------- (1208) Mem. from W. Einstein and S. Rajappa to T. Howard, with attachments, dated Mar. 19, 1998, FMSE-IR 183262-69, at FMSE-IR 183267 (STIS Transaction Cost Allocation). (1209) Mem. from W. Einstein to File, with attachments, dated Nov. 2, 1999, FMSE-IR 376038-41, at FMSE-IR 376038. (1210) A "should" opinion generally refers to a legal opinion that states a company's anticipated position "should prevail" upon challenge by a taxing authority. 312
IRS to examine the transactions. Interviewees stated that they believed that that the STIS transactions had genuine economic benefits distinct from the tax benefits. Even though certain documents from the Company's files appear to question whether the transactions had an adequate business justification, we agree with that assessment. PART J: ACCOUNTING FOR INSURANCE PRODUCTS I. INTRODUCTION In this section, we report on management's use and consideration of certain insurance arrangements primarily for the impact such arrangements may have had on the Company's earnings. Fannie Mae purchases mortgage insurance in order to mitigate its exposure to credit losses on loans, to comply with the Charter Act (i.e., the Company is required to have credit enhancement for loans with a loan to value ("LTV") ratio equal to or greater than eighty percent), and also as a broader risk mitigation strategy.(1211) Mortgage insurance may be purchased on an individual loan basis, or to cover losses incurred on a pool of loans. Fannie Mae refers to mortgage insurance that it acquires (as opposed to being provided by the lender or borrower) as a "back-end credit enhancement" ("BCE") because the policy covers loans already acquired by the Company. Fannie Mae typically purchased BCE policies to cover losses incurred on a pool of loans. Beginning in 2001, management began considering finite risk insurance(1212) products and other insurance transactions for earnings-related goals in addition to the goal of mitigating the Company's exposure to losses. Specifically, the insurance products were considered by management either to shift income between periods (in particular from 2001 and 2002, into 2003 and 2004), or to offset the impact of other actions that would result in an increase in earnings that had not been forecasted. In particular: - In early 2001, management discussed but did not execute with MMC Capital ("MMC") an insurance product, the premium for which would offset a significant increase in income attributable to a possible reduction of the allowance for loan losses (the "Allowance")(1213) (the "MMC Proposal"). - ---------- (1211) E-mail to Adolfo Marzol, Timothy Howard, et al., dated July 13, 2002, FMSE-E 2163479-93. (1212) For purposes of this analysis, a policy is considered a "finite risk" insurance arrangement if it does not provide for a true transfer of risk. FASB Staff Update, Insurance Risk Transfer, January 2006. (1213) Fannie Mae, 2002 Annual Report (Form 10-K), at 51 (Mar. 31, 2003), available at http://www.fanniemae.com/ir/pdf/annualreport/2002/2002annualreport.pdf (hereinafter "2002 Form 10-K"). 313
- In January 2002, management executed a policy with Radian, which was expected to result in recognition of premium expense in 2002, with claim reimbursements in 2003 and 2004. The stated purpose of the policy was to cover Fannie Mae's exposure for a portion of a deductible on an existing insurance policy covering the Expanded Approval Timely Payment Rewards (or "EA/TPR") loans (the "Radian Transaction"). Fannie Mae announced in November 2005 that it will restate its accounting for the Radian Transaction on the grounds that it "did not transfer sufficient underlying risk of economic loss to the insurer" for the policy to qualify for the insurance accounting treatment it was given.(1214) - In 2002, management restructured a corporate life insurance ("COLI") policy to accelerate the payment of policy acquisition costs in the current period rather than continuing to pay such costs on an annual basis for the period 2002 through 2007 (the "COLI Restructuring"). - Management also considered executing finite risk/BCE arrangements to accomplish more strategic purposes. That is, in the latter half of 2003 and the beginning of 2004, management contemplated a finite risk insurance transaction with ACE in order to expand the number of insurers in the mortgage insurance market but ultimately did not do so because the policy did not qualify as insurance for accounting purposes. We conclude that management's accounting for the Radian Transaction was not in accordance with GAAP, and that management was, in fact, motivated by the objective to shift income out of 2002 into future years in executing the policy. II. BACKGROUND A. Role of Credit Policy, Generally At all relevant times, Credit Policy's primary function at Fannie Mae was, and is, to manage credit risk within the Company.(1215) As part of its role, Credit Policy - ---------- (1214) Fannie Mae, Notification of Late Filing (Form 12b-25), at 8 (Nov. 10, 2005), available at http://www/fanniemae.com/media/pdf/newsreleases/f1625111005.pdf. (1215) 2003 Fannie Mae Self Assessment Questionnaire, A. Marzol, Executive Vice President--Credit Policy, dated Oct. 2003, FMSE-E 600866-921, at 600866; Corporate Objective Progress Report, dated Oct. 30, 2001, Zantaz document 3912425; Undated 2002 Fannie Mae Self Assessment Questionnaire, A. Marzol, Executive Vice President--Credit Policy, Zantaz document 2734087; 2004 Fannie Mae Self Assessment Questionnaire, A. Marzol, Executive Vice President--Credit Policy, Zantaz document 2318357, at 1. 314
was responsible for, among other things, obtaining credit enhancements on loans, including those with an LTV equal to or greater than eighty percent. Certain members of Credit Policy were very conscious of how the group's work affected the Company's earnings. Brian Graham, a former Senior Vice President, stated that all credit enhancements had the effect of smoothing the impact of losses on the financial statements when the losses were incurred, and that Credit Policy's function was to mitigate the volatility of losses or, more colloquially, to smooth losses.(1216) Graham stated that, while in hindsight it may have been better not to use the term "earnings smoothing" to describe what they were doing, he believed that using insurance policies to mitigate the Company's exposure to risk of loss was, and remains, the proper function of Credit Policy. In addition, evidence showed that management from 2001 through 2004 viewed BCEs as a source of revenue generation in addition to protecting Fannie Mae against credit losses. Indeed, Adolfo Marzol, head of Credit Policy, stated that he viewed revenue growth as part of his responsibilities.(1217) Likewise, contemporaneous correspondence between Graham and members of the Controller's Office reflect a willingness to manage the timing of claim reimbursements in order to manage the timing of income recognition. For example, in a 2001 e-mail, Graham informed Leanne G. Spencer that he had instructed his team to "defer receipt" of payments owed to Fannie Mae on Secondary Market Coverage deals in order to avoid "exaccerbat[ing] [sic] your 4Q [earnings] situation."(1218) Spencer approved Graham's instruction.(1219) As we discuss - ------------------- (1216) Credit Policy on occasions also referred to this function as "smoothing earnings." E.g., E-mail from Robert Schaefer to Brian Graham, Carlos Perez, et al., dated May 14, 2001, FMSE-IN 1475-78, at FMSE-IN 1475 (this and other examples of documents relevant to the discussion in Chapter VI, Part J can be found in the accompanying Appendix, at Tab D.10); see also E-mail from Janet L. Pennewell to Christine Cahn and R. Schaefer, dated Mar. 7, 2001, FMSE-IN 1460; Credit Portfolio Strategies Weekly Status Update, dated June 1, 2001, FMSE-IN 32209-11, at FMSE-IN 32210. (1217) E-mail from A. Marzol to Lawrence M. Small, Franklin D. Raines, Jamie S. Gorelick and Sharon L. Taylor, dated July 10, 1999, FMSE-IR 20690-92, at FMSE-IR 20690, 20692; E-mail from R. Schaefer to B. Graham and C. Perez, et al., dated Nov. 28, 2001, Zantaz document 2206443, at 11 ("Their goals mirrored ours, in that they were also concerned about showing increasing revenue."). (1218) E-mail from Leanne G. Spencer to B. Graham, dated Dec. 12, 2001, FMSE-E 2052293. Graham further wrote that he was "[s]till working on some deals that would have a bigger and positive impact on your 4Q and 2002 challenge." Id. He also stated that Secondary Market Coverage was one area in which Credit Policy was a revenue generator for Fannie Mae by seeking more cost effective insurance for pools of loans. (1219) Id. 315
in greater detail below, Credit Policy also considered insurance policies as a means to shift income from 2001 and 2002, into 2003 and 2004.(1220) B. Fannie Mae's Consideration of Finite Risk Insurance Policies 1. MMC Proposal According to various members of Credit Policy interviewed, there was pressure in early 2001 to reduce the Company's loan loss reserves. Graham stated that the Controller's Office asked him in early 2001 to find an insurance product that would replace the protection provided by the loan loss reserve should Fannie Mae be compelled to draw down its reserve.(1221) Graham then met with representatives of MMC in early 2001 to discuss potential policies for this purpose.(1222) According to Robert Schaefer, Director--Credit Policy, who attended this meeting with Graham, Graham explained to MMC representatives that Fannie Mae was considering reducing its loan loss reserve and was looking for a credit enhancement transaction that would have a one-time charge to the P&L that would offset the positive income effect from the reduction in the loan loss reserve. The other requirements, according to Schaefer, were that the policy have a high chance of "paying off" in future years, and would be treated as insurance for accounting purposes. Documents prepared by MMC representatives to summarize the meeting corroborated Schaefer's recollection, and stated: "We discussed two issues raised by Fannie Mae (1) FM's interest in a structured insurance policy (i.e., finite risk) to replace some portion of the $800mm+ loan loss reserve . . . . FM's objectives would be to spread - ------------------- (1220) E.g., E-mail from A. Marzol to B. Graham, Daniel H. Mudd, Louis Hoyes, Timothy Howard and L. Spencer, et al., dated Oct. 17, 2001, FMSE-IN 2732; E-mail from R. Schaefer to C. Perez and Darren Thompson, dated May 3, 2002, FMSE-IN 1457. Indeed, there appears to have been an "overall corporate effort to move income out of 2002 and into later years." QBR Summaries, dated May 2002, FMSE-IR 565521-27, at FMSE-IR 565521, 56524. (1221) Undated Start Date as an SVP, Zantaz document 2904394; E-mail from B. Graham to C. Perez and R. Schaefer, dated June 25, 2001, FMSE-IN 1512. (1222) In negotiating the MMC Transaction, Fannie Mae Credit Portfolio personnel, including. Schaefer, Graham, Perez, and others had conversations with representatives of two companies under the Marsh & McLennan Companies' umbrella (collectively, "MMC"). They met with James Carey and Nicholas Zerbib of MMC Capital, which was, at that time, the private equity business of MMC. They also met with Joseph Peiser and Paul Sclafani, of MMC Enterprise Risk, the risk management consulting business of MMC. E-mail from J. Carey to B. Graham, R. Schaefer, N. Zerbib, J. Peiser, and P. Sclafani, dated Apr. 19, 2001, FMSE-IN 1463-65. 316
volatility of the loss reserve over time, whether that volatility resulted from accounting pressure to reduce or increase the reserve or from actual experience."(1223) MMC then proposed what it characterized as a finite risk insurance product.(1224) It appears that Financial Standards and Credit Policy recognized early on that such a policy might not qualify for insurance accounting.(1225) Nonetheless, from June through July 2001, Credit Policy assembled a working group of individuals in Financial Standards, Corporate Tax and the Legal Department to vet the MMC proposal.(1226) The working group was initially referred to as the "Loss Smoothing Team," and the project was known as the "Loss Insurance Initiative."(1227) - ---------------- (1223) E-mail from R. Schaefer to B. Graham, et al., dated May 14, 2001, FMSE-IN 1475-78, at FMSE-IN 1477 ("We discussed two issues raised by Fannie Mae: (1) FM's interest in a structured insurance policy (i.e., finite risk) to replace some portion of the $800mm+ loan loss reserve . . . . FM's objectives would be to spread volatility of the loss reserve over time, whether that volatility resulted from accounting pressure to reduce or increase the reserve or from actual experience."); accord Handwritten Notes from R. Schaefer, dated May 19, 2001, Zantaz document 2207005; E-mail from B. Graham to R. Schaefer and C. Perez, dated June 12, 2001, FMSE-IN 1497-500, at FMSE-IN 1497 ("Our goal should not be to keep the loss reserve high. Instead our goals should be: - Accelerate future year expenses into 2001 and 2002 - Potentially enable us to draw down our reserve in 2003 and beyond . . . ."). (1224) E.g., E-mail from J. Carey to B. Graham, R. Schaefer, et al., dated Apr. 19, 2001, FMSE-IN 1463-65, at FMSE-IN 1465; E-mail from R. Schaefer to B. Graham, C. Perez, et al., dated May 14, 2001, FMSE-IN 1475-78, at FMSE-IN 1475. (1225) E-mail from Gregory Johnson to R. Schaefer, Jonathan Boyles, Chip Jordan, et al., dated June 5, 2001, FMSE-SP 80210; E-mail from G. Johnson to R. Schaefer, et al., dated June 6, 2001, FMSE-IN 1496. (1226) Members of the working group included Financial Standards employees Paul Salfi and J. Boyles; Fannie Mae lawyers Michael Daze and Daniel Smith; Tom Gick, an outside attorney with Sutherland Asbill & Brennan LLP; Credit Policy personnel R. Schaefer, C. Perez, Michael Goldberg, and B. Graham; and Controller's Office personnel J. Morgan Whitacre and C. Jordan; David Menn, Director--Corporate Risk and Insurance Management; and G. Johnson, Senior Financial Analyst--Credit Portfolio. E-mail from R. Schaefer to P. Salfi, C. Perez, et al., dated July 12, 2001, FMSE-SP 80205-09. (1227) See, e.g., Credit Portfolio Strategies Weekly Update, dated June 29, 2001, FMSE-IN 33120-122, at FMSE-IN 33121. Schaefer later told the team to "refer to this project as the Loss Insurance Initiative." E-mail from R. Schaefer to P. Salfi, J. Boyles, C. Perez, M. Goldberg, G. Johnson, J.M. Whitacre, M. Daze, B. Graham, C. Jordan, D. 317
Sometime during the second half of 2001, Fannie Mae terminated its consideration of the MMC proposal. According to both Graham and Schaefer, the inability for the transaction to qualify as insurance for accounting purposes was the reason for its termination. According to Schaefer, Financial Standards ultimately opined that the proposal would not receive insurance accounting because the timing of cash flows were set by the contract.(1228) Additionally, Graham stated that the economics of the MMC Proposal were unfavorable due to MMC's failure to understand the mortgage market. 2. Radian Transaction According to Adolfo Marzol, former interim Chief Risk Officer, Franklin D. Raines announced to the Senior Leadership Team at a quarterly business review ("QBR") meeting in approximately May 2001 that Fannie Mae's outlook reflected a revenue strength that required "a mindset shift." Marzol said he understood Raines to be asking the attendees to look for opportunities to make investments for the future instead of looking to maximize profits in the current year. Marzol also recalled Raines specifying that the investments should neither be "fliers" nor create a multi-year drain on profitability. We also note that a May 2001 document suggests that Controller's Office likewise shifted its focus to increasing EPS in 2003 and later years.(1229) Marzol identified the following actions that, to his knowledge, flowed from Raines's guidance: the Radian Transaction, a restructuring of a COLI policy(1230), debt buybacks,(1231) and the sale of Title I loans.(1232) - ------------------- Smith, D. Menn, David Kightlinger, Scott Lesmes, Jennifer M. Osborn, and Marylou Batson, dated July 12, 2001, FMSE-SP 80205-09. (1228) KPMG workpapers reflect that it opined that the transaction would not qualify for insurance accounting. Undated mem. to Fannie Mae File. We have not heard that the tax or other attorneys consulted by Credit Policy expressed any objection to the deal. (1229) E-mail from R. Schaefer to B. Graham, et al., dated May 25, 2001, FMSE-IN 1479-81, at FMSE-IN 1480 ("Morgan W also mentioned that [with] less concern surrounding $6.46, there was less inertia behind an immediate reduction to the loss reserves. If anything, there is more concern about an earnings drop-off after 2003.") Graham and Schaefer clarified that by "less inertia," Schaefer meant less impetus. (1230) See infra Subsection II.B.4. (1231) See infra Chapter VI, Part L. (1232) See discussion infra Chapter VI, Part L. According to Marzol, Title I loans were partially-insured government loans. Fannie Mae's pool of such loans was performing very poorly and the Company took the opportunity to sell the loans. 318
In the summer of 2001, following Raines's guidance, Marzol told his team in Credit Policy that the Company was looking to make investments in the current year for the future and asked Graham to make a list of things that Credit Policy had foregone due to prior budget constraints. The list prepared by Graham in response included, among other things, reducing the size of the deductible on a back-end mortgage insurance policy he had obtained earlier in the year for Fannie Mae's EA/TPR loans.(1233) That item, however, was not immediately authorized. Documents showed that Raines reiterated his guidance to Credit Policy in the fall of 2001. In an e-mail to his reports in November 2001, Marzol wrote: "Dan, Lou and I met last night with FDR [Raines]. He personally reiterated guidance he had given through Dan last week - that he views keeping the credit loss line under control as a top priority. He is more than prepared to trade revenues for losses . . . ."(1234) While Graham did not recall the e-mail in particular, he did recall hearing in the fall of 2001 that management was very concerned with potential volatility in credit losses; Graham said that his team again looked at ways to address the deductible exposure on the EA/TPR product. A structure was bid out to several insurers, but the transaction was ultimately executed with Radian; according to Graham, the Radian bid was attractive, possibly because it was mispriced.(1235) Evidence showed that management chose to insure the EA/TPR loans in part because the potential default characteristics of those loans would result in Fannie Mae receiving a substantial amount of expected claim reimbursements in 2003 and 2004, accomplishing the objective of shifting income to those years. For instance, in a November 20, 2001 e-mail, Graham wrote to Schaefer: "The goal is to have lots of loans that will fail quickly (albeit 18 months out) and in large enough numbers to generate the desired P&L result."(1236) Additionally, in a December 12, 2001 e-mail, Marzol wrote to Mudd that the "goal is to take portions of the portfolio that we are absolutely sure will have losses in 2003-2004 and transfer that exposure for highly cost effective premium levels, with as much premium recognized in 2002 as possible. Target portfolios include - --------------- (1233) When Credit Policy received bids on insurance for the EA/TPR product, Marzol decided that Fannie Mae would be better off because of pricing to purchase insurance with a deductible. Accordingly, in the first transaction insuring the EA/TPR product, Fannie Mae agreed to absorb losses on the portfolio of loans up to one percent of the original principal balance of the loans (i.e., a one percent deductible). (1234) E-mail from C. Perez to R. Schaefer, et al., dated Oct. 17, 2001, FMSE-IN 2732. (1235) E-mail from B. Graham to R. Schaefer, dated Nov. 26, 2001, FMSE-IN 1348. (1236) E-mail from B. Graham to R. Schaefer, dated Nov. 20, 2001, FMSE-IN 3245. 319
the 1% of initial loss not credit enhanced for EA/TPR loans as well as similar ABN-AMRO product we acquired."(1237) The Radian Transaction went through several reviews by various business heads at Fannie Mae before it was approved. First, Credit Policy reviewed the transaction's structure with Financial Standards.(1238) According to interviewees, Financial Standards personnel told Credit Policy employees that its advice on whether the transaction would qualify for accounting as insurance depended on the answers to three questions.(1239) Schaefer recalled that Financial Standards may have done this because they were not very knowledgeable about the insurance area and the business aspect of insurance; with Graham's guidance, Schaefer provided Financial Standards with answers to the three questions.(1240) According to Schaefer, Financial Standards concluded that the transaction qualified as insurance for accounting purposes after reviewing the answers. Financial Standards's conclusion that the arrangement qualified as insurance for accounting purposes was, in part, based on the conclusion that it was reasonably possible that the insurer could incur a significant loss.(1241) Schaefer, who - ------------- (1237) E-mail from A. Marzol to D. Mudd, B. Graham, L. Hoyes, and T. Howard, dated Dec. 12, 2001, Zantaz document 4002092; see also E-mail from B. Graham to L. Spencer, J. Pennewell, A. Marzol, et al., dated Dec. 18, 2001, Zantaz Document 2146743 (identifying the Radian Transaction as a "`Business Management' Strategy" that would have an "Earnings Impact: Insurance premium amortized over time for book and tax; claims payments offset losses which will tend to occur in the 2003-2005 timeframe"). (1238) We have not seen evidence that KPMG specifically vetted the Radian Transaction. (1239) The three questions were: was there "`more than a remote' probability of variability in the amount of the non-present valued cash flows; [was there] `more than a remote' probability of variability in the timing of the cash flows; [and was the] probability greater than ten percent that [PV(MI Premiums) - PV(MI Claim Reimbursement) - PV(G&A expenses)] > 10%*PV(MI Claim Reimbursements)." EA/TPR Loss Insurance Proposal, dated Dec. 18, 2001, FMSE-IN 29761-62. (1240) Id.; E-mail from P. Salfi to J. Boyles, dated Dec. 19, 2001, Cataphora document i.1008778059000.551783345. Additionally, Schaefer stated that at some point in late 2001 he learned that Radian would be accounting for the transaction as a deposit. He said he later told Salfi, and that Salfi had no reaction to the news. Schaefer did not recall whether he told Salfi before or after he had provided Financial Standards with the answers to the three questions, but, he said Financial Standards did not change its conclusion that the Radian Transaction qualified for insurance accounting based on the information. (1241) EA/TPR Loss Insurance Proposal, dated Dec. 18, 2001, FMSE-IN 29761-62; E-mail from P. Salfi to J. Boyles, dated Dec. 19, 2001, Cataphora document i.1008778059000.551783345. 320
provided the underlying information to Financial Standards on which they relied to opine on the Radian Transaction, stated that he understood the accounting criterion to mean there should be a ten percent chance that the insurer could suffer a loss of ten percent or more of the premium. In determining the insurer's possible losses, the Company included not only claim reimbursements, but also the insurer's costs of administering the policy. Our review of the analysis showed that it was only by including an estimate of the insurer's administrative costs that management concluded that it was reasonably possible the insurer would incur a significant loss under the policy. In our view, the inclusion of the administrative costs was not in accordance with GAAP. FASB Statement No. 113, Accounting and Reporting for Reinsurance of Short-Duration and Long-Duration Contracts (hereinafter "FAS 113"), provides guidance on when an arrangement transfers sufficient risk to qualify as insurance for accounting purposes. In particular, FAS 113 states: The ceding enterprise's [insured's] evaluation of whether it is reasonably possible for a reinsurer [insurer] to realize a significant loss from the transaction shall be based on the present value of all cash flows between the ceding [insured] and assuming [insurer] enterprises under reasonably possible outcome. . . .(1242) The guidance in Paragraph 10 is clear that only payments between the insurer and the insured should be considered. To make this point even clearer, however, the FASB staff issued interpretive guidance in 1993 that is directly relevant to the Radian Transaction. Question 16 of "Accounting for Reinsurance: Questions and Answers about Statement 113," and its response state: Q -- In determining the amount of the reinsurer's loss under reasonably possible outcomes, may cash flows directly related to the contract other than those between the ceding and assuming companies, such as taxes and operating expenses of the reinsurer, be considered in the calculation? A -- No. Paragraph 10 states that the evaluation is based on the present value of all cash flows between the ceding and assuming enterprises under reasonably possible outcomes and therefore precludes considering other expenses of the reinsurer in the calculation.(1243) - ------------------- (1242) ACCOUNTING AND REPORTING FOR REINSURANCE OF SHORT-DURATION AND LONG- DURATION CONTRACTS, Statement of Fin. Accounting Standards No. 113, P. 10 (Fin. Accounting Standards Bd. 1992) (emphasis added). (1243) ACCOUNTING FOR REINSURANCE: QUESTIONS AND ANSWERS ABOUT STATEMENT 113, at Question 16 (Fin. Accounting Standards Bd. 1993). 321
While Fannie Mae is not a reinsurer, it was unreasonable for Financial Standards to conclude that the risk transfer test was satisfied with respect to the Radian policy by including administrative costs that the literature plainly dictated were inappropriate for a reinsurer to include. Whether the arrangement is called insurance or reinsurance, the objectives of the insured or ceding parties are the same - to transfer the risk of loss. As such, the same test for making that determination would apply. The Radian Transaction was also reviewed and approved by Timothy Howard in consultation with Spencer and Janet L. Pennewell. Before approving the transaction, Howard asked Spencer and Pennewell for assurance that accounting had vetted and recommended the transaction. After the meeting, Spencer and Pennewell sent Howard the following e-mail: We both would lean slightly in favor of [executing the Radian Transaction]. On the plus side, it helps some with our challenge and gives us a little additional credit protection in a bad environment. On the negative side, we don't think the economics are very compelling and think the payback has a chance of being longer than shown in their base case.(1244) Howard forwarded the e-mail to Graham and Schaefer, stating that he was "fine [with] going ahead with this," but characterizing his approval as "lukewarm."(1245) The Radian Transaction also was presented by Marzol to Daniel H. Mudd and discussed with other Executive Vice Presidents, including Louis Hoyes, Executive Vice President--Single Family Business. While Mudd had no recollection of having reviewed the policy, documents showed that senior executives were copied on an exchange between Hoyes and Marzol about the merits of the policy.(1246) The documents showed that Hoyes objected to the policy because of his belief that the terms of the insurance motivated the Company to foreclose on mortgagees who were poor and/or had blemished credit, a motivation that Hoyes deemed to be contrary to Fannie Mae's - ------------------- (1244) E-mail from T. Howard to B. Graham, R. Schaefer, L. Spencer, J. Pennewell, dated Dec. 21, 2001, FMSE-IN 1397. (1245) Id. (1246) According to Schaefer, Marzol and he made a presentation to Mudd that was similar to the presentation they previously made to Howard. See also Loss Insurance Proposal, FMSE 529289-93 ("Presented to Dan Mudd 1/08/02"); Calendar File, Zantaz document 2132291 ("Mudd, Dan 1/8/02 8:30 10:00 `Adolfo will drop by to talk to you.'"). It is not clear whether Mudd's approval was being sought or whether he was simply being briefed; moreover, it appears based on the calendar entry that Mudd's meeting with Marzol and Schaefer occurred before the Hoyes-Marzol e-mail exchange began debating the policy. 322
mission. The offending term, according to Hoyes, was that the policy paid benefits only on foreclosure.(1247) In the e-mail Hoyes sent to Marzol, which copied Mudd, Robert J. Levin and Howard, Hoyes wrote: "Should we be exposing Fannie Mae to this type of political risk to `move' $40 million of income? I believe not."(1248) When asked to explain the "$40 million of income" comment, Hoyes said he was simply responding to Marzol's message to him, which had made that point. Indeed, Marzol's e-mail to Hoyes explained: The transaction is intended to deliver on a request of FDR and a corporate need to make sure the loss line is well managed and that we do so while revenues are strong. . . . Bottomline [sic] - there is a significant reduction of losses that occurs in 2003-2005 if foreclosure rates are near the average. Otherwise, the reduction in losses occurs from 2003 to 2007. The net income after tax impact is a reduction in NIAT for 2002 and 2003, then turning into positive P/L in either scenario.(1249) Notwithstanding Hoyes's objection, Marzol had already received approval from Howard, and he gave his team approval to move forward with the transaction sometime on January 8, 2002.(1250) The transaction was executed at the end of January 2002. 3. ACE Proposal Another finite risk insurance policy that management considered, but ultimately did not execute, was a proposal offered by ACE. ACE representatives visited Fannie Mae in March or April 2003 to discuss opportunities for doing business with the Company. Fannie Mae had a strategic interest in doing an insurance deal with ACE: there were only seven potential counterparties in the mortgage insurance market, of which Fannie Mae had policies with four, and management wanted to increase the number of potential counterparties.(1251) According to Schaefer, Graham determined that - -------------- (1247) Hoyes said the EA/TPR product was targeted to borrowers at the upper-end of "sub-prime" in order to expand the bounds of "prime." As such, he said his job was to motivate this staff to work with borrowers in default on their mortgages in order to keep them in their homes and he did not want this policy to create a disincentive. (1248) E-mail from L. Hoyes to A. Marzol, D. Mudd, Robert J. Levin, Thomas E. Donilon and T. Howard, dated Jan. 8, 2002, Zantaz document 4159375. (1249) Id. (1250) E-mail from A. Marzol to R. Schaefer, B. Graham, et al., dated Jan. 9, 2002, FMSE-IN 10997-1001, at FMSE-IN 10997. (1251) Schaefer stated that the ACE Proposal was meant to create a new insurance policy on loans with a loan to value ratio of greater than eighty percent that were delivered through the Dedicated Channel. At the time, these loans were insured by PMI under 323
the best opportunity for business was with ACE, a new market entrant who had not participated in the primary risk layers of insurance, and for ACE to cover prime business such as loans delivered to Fannie Mae through the Dedicated Channel.(1252) Fannie Mae and ACE considered several structures over an extended period of time. The first structure that Credit Policy brought to Financial Standards' attention was one that Financial Standards opined would not receive insurance accounting. KPMG, through Financial Standards, also indicated that the accounting would need to be bifurcated so that the payment for coverage over the threshold would be treated as insurance, but the premium for the base coverage would be treated as a deposit.(1253) Schaefer said, aside from the accounting treatment question, the accountants were uncomfortable with the "optics" of the proposal's direct connection between the premium to be paid and claims reimbursements. Credit Policy then considered an alternate structure that would give Fannie Mae the right to cancel the policy, which would create the "risk" of loss to ACE needed for accounting purposes.(1254) Financial Standards opined that the entire premium under the revised structure qualified for insurance accounting treatment.(1255) Additionally, lawyers vetted the revised ACE Proposal for compliance with the Charter Act and potentially other issues. Ultimately, the ACE proposal was not consummated, but the facts surrounding the ultimate rejection of the ACE Proposal bear mention. At the outset, Thomas E. Donilon had expressed concerns based on public policy considerations - ------------------- the pmiSELECT program. He also stated that Fannie Mae wanted an alternative to PMI. Schaefer clarified that the Dedicated Channel loans insured under pmiSELECT were not related to the Home Solutions loans that PMI refused to continue insuring under pmiSELECT. (1252) At the time, these loans were covered under the pmiSELECT program. (1253) E-mail from P. Salfi to Michelle Vaniels, dated Oct. 16, 2003, FMSE-E 1171242-43; Draft mem. from P. Salfi to R. Schaefer, et al., dated Dec. 11, 2003, FM SRC 354186-88. (1254) KPMG asked Fannie Mae whether the cancellation provision was more than a conceptual possibility. Fannie Mae took the position that, although Fannie Mae would continue to have a Charter requirement to maintain a credit enhancement on the loans, conceptually, there could be an economic reason to cancel the transaction with ACE. For instance, Schaefer said, if home prices continued to rise and Fannie Mae performed a mark-to-market analysis that showed a large percentage of the loans having LTVs below eighty percent, it would cancel the policy. (1255) E-mail from P. Salfi to R. Schaefer and Jason N. Barrett, dated Dec. 16, 2003, FMSE-E 177626-29. 324
concerning Fannie Mae's introduction of ACE as a new entrant in the mortgage insurance field. His concerns did not kill the proposal.(1256) Credit Policy continued to push for approval of the ACE Proposal even after Financial Standards and KPMG expressed misgivings.(1257) Boyles expressed his concerns about the proposal to Marzol in mid to late December 2003. And on December 30, 2003, Dan Smith sent Marzol and others a memorandum stating, in relevant part, that KPMG: - advises that the structure satisfies accounting rules necessary to treat our payments to Ace [sic] as insurance premiums, but - counsels us to consider carefully the risks of entering a transaction in which the accounting treatment, based on technical rules, may appear inconsistent with the underlying economics of the business transaction.(1258) Nonetheless, in a January 2, 2004 e-mail from Marzol to Howard, Marzol wrote: "We are moving towards executing, but we still have some time where we can have further dialogue."(1259) Marzol's continued push for the proposal notwithstanding Boyles's and KPMG's concerns reflected a tone that was representative of Fannie Mae's overall culture, that is, a senior business executive did not take accountants's objections to a - -------------- (1256) We were told that Donilon's concerns were addressed. However, no one could tell us what steps were taken to resolve them. (1257) Financial Standards and KPMG provided an opinion that the policy would receive insurance accounting so long as ACE would agree to include a cancellation option in the policy. Without the cancellation option, Fannie Mae would account for the premium and annual adjusted premium in the bifurcated manner described above, which presented great operational challenges. See also Undated Events leading up to the Present, FMSE 54905-11, at FMSE 54905. Nonetheless, regardless of the cancellation option and an understanding that the accounting was supportable under GAAP, there was concern that the accounting may not be consistent with the economics. E-mail from D. Smith to A. Marzol, B. Graham, J. Boyles, P. Salfi, Ann M. Kappler, T. Donilon, et al., dated Dec. 30, 2003, Cataphora document i.1072828593000.1988983619. It is not clear, but it is possible that this concern had to do with whether Fannie Mae's right to cancel the policy was a substantive right given its need for insurance coverage for loans with LTV ratios in excess of eighty percent. (1258) E-mail from D. Smith to A. Marzol, B. Graham, J. Boyles, P. Salfi, A. Kappler, T. Donilon, et al., dated Dec. 30, 2003, Cataphora document i.1072828593000.1988983619. (1259) E-mail from A. Marzol to T. Howard, B. Graham, J. Boyles, and L. Spencer, dated Jan. 2, 2004, Cataphora document i.1073065333000.1693380808. 325
proposed transaction as definitive, and looked to Howard to be the ultimate decisionmaker. According to Schaefer, the ACE Proposal was rejected sometime in 2004 after ACE withdrew its agreement to the cancellation option, and the prior proposed bifurcation accounting was deemed operationally too difficult to implement. According to contemporaneous documents, ACE withdrew its agreement in July 2004.(1260) 4. Corporate Owned Life Insurance During 2001 and 2002, Credit Policy also analyzed whether it could restructure existing COLI policies to accelerate the payment of deferred acquisition costs due from the Company, which the Company had been paying through a reduction in the return on the policy.(1261) According to Marzol, the restructuring was another item that came out of Raines's guidance to invest for the future by incurring expenses in the present period. The Company decided to restructure COLI in this manner because management felt that it had the wherewithal to pay the costs at once in 2001 and 2002.(1262) Fannie Mae restructured one COLI policy it had with MetLife in response to Raines's guidance.(1263) As Graham wrote in a December 18, 2001 e-mail, the earnings - ---------------- (1260) E-mail from R. Schaefer to D. Thompson, et al., dated July 20, 2004, Cataphora document i.1090331892000.32456200, at 5 ("ACE decided that our unrestricted right to cancel the policy was unworkable for them."). (1261) E-mail from B. Graham to L. Spencer, J. Pennewell, A. Marzol, Joseph J. Sakole, dated Dec. 18, 2001, Zantaz Document 2146743 ("Accelerating Amortization of Capitalized Expenses in COLI deals"); Credit Portfolio Strategies Weekly Status Report, dated Mar. 15, 2002, FMSE-IN 33008-10; Corporate Owned Life Insurance, dated Apr. 9, 2002, Zantaz Document 1944984. (1262) Marzol said that he meant Fannie Mae had earnings sufficient to absorb the charge. Graham believed that Fannie Mae wished it had not structured the deals as it had when Fannie Mae entered the deals because it ended up financing the broker commission at the policy earnings rate, which exceeded Fannie Mae's borrowing cost. Fannie Mae took the opportunity in 2001 to restructure the MetLife policy in a way that Fannie Mae would have preferred it had done originally; the restructuring closed in 2002. According to Schaefer, Graham's objective was similar to his objective for the Radian Transaction: to pay as much expense as Fannie Mae could in 2001 and 2002 in order to avoid paying those expenses in outer years. (1263) E-mail from B. Graham to T. Howard, A. Marzol, L. Spencer, Jef Kinney and Jeffrey Cravath, dated Dec. 21, 2001, FMSE-E 994863 ("We're trying to cut similar deals with the other insurers we have COLI out under the assumption that you still want this type of flexibility around the first and second quarter of 2002. Let us know if/when you want to take the hit.") 326
impact would reduce "miscellaneous income in the period it is taken; future earnings would be enhanced since yields would no longer be depressed by amortization expense."(1264) The restructuring of the policy resulted in Fannie Mae recording $11 million in miscellaneous expense during the second quarter of 2002, which represented an acceleration of the deferred acquisition costs and state premium taxes that Fannie Mae was originally scheduled to pay, through reductions in the return on the policy, through December 31, 2007.(1265) The Company paid for the acceleration of the costs through a reduction in the cash surrender value in its policies with MetLife.(1266) Fannie Mae considered restructuring other policies in 2002, pursuant to Pennewell's guidance that "there [was] additional opportunity to accelerate expenses into" 2002 and that the Company was "looking at initiatives that will add income in 2004 and 2005."(1267) Notes on a presentation about the economics of accelerating the deferred expenses indicate that the Company then planned to "[n]ext look[] @ AIG Life 1998" and "New York Life."(1268) However, we have not seen any evidence that Credit Policy restructured additional COLI policies in 2002. The consideration given to restructuring other COLI arrangements was part of the broader efforts described throughout this section to identify transactions that would have a similar result as the restructuring of the COLI policy (i.e., that would result in the recognition of expense in 2002 but would generate a return to the Company in subsequent years). That effort appears to have been directed by Howard with assistance - ------------------- (1264) E-mail from B. Graham to L. Spencer, J. Pennewell, A. Marzol, J. Sakole, dated Dec. 18, 2001, Zantaz document 2146743. (1265) Corporate Owned Life Insurance, dated Apr. 9, 2002, Zantaz document 1944984, at 2, 12; see also Letter Agreement, dated Dec. 23, 1997, FMSE-IR 720841-61; Letter from Metropolitan Life Insurance Company to Darren Thompson, dated Apr. 1, 2002, FMSE-IR 720840. (1266) Corporate Owned Life Insurance, dated Apr. 9, 2002, Zantaz document 1944984, at 2. (1267) E-mail from R. Schaefer to C. Perez and D. Thompson, dated May 3, 2002, FMSE-IN 1457; see also Credit Portfolio Strategies Weekly Status Report, dated May 10, 2002, FMSE-IN 32736-40, at FMSE-IN 32737; E-mail from T. Howard to B. Graham and L. Spencer, dated Dec. 31, 2001, FMSE 644037 ("We're trying to cut similar deals with the other insurers we have COLI out under the assumption that you still want this type of flexibility around the first and second quarter of 2002. Let us know if/when you want to take the hit. Unless we hear different guidance, we will continue to work toward the same goal of taking hits in Q1 and Q2 next year.") (1268) Corporate Owned Life Insurance, dated Apr. 9, 2002, Zantaz document 1944984, at 1. 327
from Spencer. In an e-mail dated December 31, 2001, to Graham and Spencer, Howard responded to Graham's earlier e-mail that described in general the economics of the COLI restructurings and stated Graham's "[g]oal of note was to find out whether an approach like this - if we can make it all work - is consistent with what you are looking for in Q1 and/or Q2."(1269) In his response, Howard stated: Leanne is the right person to take the lead on this. We're trying to build up an inventory of these sorts of transactions. Ideally, we can get a notion of the economics and payback of each without the party that came up with the idea having to do too much work. That won't be the case all of the time, unfortunately. As we get into the quarter, we should have a better sense of which ones we want to do.(1270) Howard's response that "we should have a better sense" of what transactions management wanted to do as "we get into the quarter" seems to echo Marzol's point about the Company deciding to restructure the MetLife COLI policy because it had the "wherewithal to pay the costs" (i.e., sufficient earnings to absorb the charge). In other words, management would decide between transactions that presumably were economically viable based on where they were during the quarter in relation to earnings forecasts, and only those transactions that would not impair the Company's ability to meet analysts' expectations would be executed. III. FINDINGS REGARDING INSURANCE PRODUCTS We conclude that, during the period from 2001 through 2004, Credit Policy considered transactions primarily for earnings-related goals rather than merely to mitigate the Company's exposure to credit losses or to satisfy the requirements of the Charter Act. While we acknowledge that the proper role of Credit Policy is to mitigate losses through the use of mortgage insurance products, and that a well-designed insurance policy will have the effect of smoothing the impact of losses when they occur, the evidence we found reflected management's goals were really to "smooth" income and to use insurance policies as a way to show stable earnings growth.(1271) In particular, we - ------------------- (1269) E-mail from T. Howard to B. Graham and L. Spencer, dated Dec. 31, 2001, FMSE 644037. (1270) Id. (1271) E.g., E-mail from B. Graham to R. Schaefer and C. Perez, dated June 12, 2001, FMSE-IN 1497-500; E-mail from R. Schaefer to B. Graham, et al., dated May 14, 2001, FMSE-IN 1475-78, at FMSE-IN 1477; E-Mail from J. Peiser to R. Schaefer, dated May 28, 2001, FMSE-IN 1482-85, at FMSE-IN 1485 (stating that a benefit of the MMC "Property Catastrophe program," was to "achieve earnings smoothing for first catastrophe - risk transfer for second catastrophe."); Credit Portfolio Strategies Weekly Status Update June 1, 2001, FMSE-IN 32209-11, at FMSE-IN 32210 328
found that the Radian Transaction was one part of an inappropriate overall corporate effort to shift income out of 2001 and 2002, into 2003 and 2004. Fannie Mae has announced that it will be restating its accounting for the Radian Transaction based on its determination that there was insufficient risk transfer to Radian for the transaction to qualify for the insurance accounting treatment that it was given.(1272) We concur that the accounting for the Radian Transaction was not in accordance with GAAP. Apart from the error in accounting for the policy, however, we also question its economics. Viewing Fannie Mae's upfront premium payment on the "insurance" policy as the equivalent of the purchase of any other investment, we question whether management reasonably expected the return on that investment to be equal to, or greater, than the return on any other investment of comparable risk. Management did consider this question when analyzing the terms of the Radian Transaction in November 2001.(1273) Based on estimates of the timing of claim reimbursements and an upfront premium payment of forty-six basis points, Credit Policy determined that the "investment" would result in a yield of 3.25 percent.(1274) At that time, after the Federal Reserve had lowered interest rates, the yield on a three-year Treasury Bill was 3.5 percent.(1275) It appears that the only way to justify the "investment" in the Radian policy as being a better economic investment was if it had a lower risk than the Treasury Bill.(1276) However, even that comparison is not completely valid given that there is no uncertainty as to the timing of cash flows on a Treasury Bill; when the timing of cash flows is uncertain, a prudent investor would demand a higher yield to compensate for the uncertainty. Thus, in our view, the Company accepted a lower yield on its "investment" in the Radian policy solely to achieve the income-shifting objective. Further, while Fannie Mae did not enter into either the MMC Proposal or the ACE Transaction, we also conclude that the two policies were considered by management for an improper purpose. With respect to the MMC Proposal, Credit Policy was acting on a request by the Controller's Office to enter into an insurance transaction that would offset the impact of reducing the loan loss reserve, and not really to address - ---------- (Under the heading, "Loss Smoothing," the document reads: "Had a follow-up call (with Mike, Greg, and Dan Smith) with Joe Peiser @MMC to discuss his proposal for Earnings Smoothing."). (1272) Fannie Mae, Notification of Late Filing (Form 12b-25) at 8 (Nov. 10, 2005). (1273) E-mail from B. Graham to R. Schaefer, M. Goldberg, C. Perez, et al., dated Nov. 26, 2001, FMSE-IN 1348. (1274) Id. (1275) Id. (1276) Id. 329
any loss mitigation concerns.(1277) This reflects, at a minimum, that the Controller's Office was aware that the Allowance was excessive in 2001, but as discussed in detail in Chapter VI, Part A, the Allowance was not reduced until 2004. Further, evidence suggested that management was concerned about the potential spike in income as a result of the reduction in Allowance, and what such an event may mean for the Company's ability to meet Raines's challenge of doubling EPS in five years, or to the Company's goals of showing steady earnings growth.(1278) Accordingly, we find that the primary motivation behind consideration of the MMC Proposal was inappropriate. Further, while Fannie Mae did not enter into either the MMC Proposal or the ACE Transaction, we also conclude that the two policies were considered by management for an improper purpose. With respect to the MMC Proposal, Credit Policy was acting on a request by the Controller's Office to enter into an insurance transaction that would offset the impact of reducing the loan loss reserve, and not really to address any loss mitigation concerns.(1277) This reflects, at a minimum, that the Controller's Office was aware that the Allowance was excessive in 2001, but as discussed in detail in Chapter VI, Part A, the Allowance was not reduced until 2004. Further, evidence suggested that management was concerned about the potential spike in income as a result of the reduction in Allowance, and what such an event may mean for the Company's ability to meet Raines's challenge of doubling EPS in five years, or to the Company's goals of showing steady earnings growth.(1278) Accordingly, we find that the primary motivation behind consideration of the MMC Proposal was inappropriate. As for the ACE Proposal, the evidence reflecting management's discussions about this proposal showed that Credit Policy advocated a transaction notwithstanding the misgivings expressed by KPMG, Financial Standards, and the Law and Policy Group.(1279) Moreover, Credit Policy's decision to embed a right to cancel in the proposal to satisfy the insurance accounting test illustrated management's outcome-oriented approach, rather than a merit-based approach, toward structuring the transaction. Without the right to cancel, the proposal would not have qualified for insurance accounting treatment. However, it is not clear whether the right to cancel was more than a theoretical possibility.(1280) Finally, with respect to management's restructuring the MetLife COLI policy in 2002, again, evidence showed that management's sole motivation for doing so was to manage earnings and incur expenses in 2002 to depress income. In particular, Fannie Mae only accelerated expenses to be incurred through 2007 (instead of all costs due under the policy). We do not find fault with management's accounting in 2002 to - ---------- (1277) E-mail from R. Schaefer to B. Graham, et al., dated May 14, 2001, FMSE-IN 1475- 78, at FMSE-IN 1477 ("We discussed two issues raised by Fannie Mae: (1) FM's interest in a structured insurance policy (i.e. finite risk) to replace some portion of the $800mm+ loan loss reserve . . . . FM's objectives would be to spread volatility of the loss reserve over time, whether that volatility resulted from accounting pressure to reduce or increase the reserve or from actual experience."). (1278) E-mail from R. Schaefer to B. Graham, et al., dated May 25, 2001, FMSE-IN 1479- 81, at FMSE-IN 1480 ("Morgan W also mentioned that with less concern surrounding $6.46, there was less inertia behind an immediate reduction to the loss reserves. If anything, there is more concern about an earnings dropoff [sic] after 2003."). It was possible that the Company was worried that it would not attain $6.46 in 2003 and was looking to increase earnings by reducing the Allowance, but also by purchasing insurance to replace the lost Allowance coverage. Graham said Fannie Mae did not go forward with the MMC Proposal because the proposal was not workable. See also discussion supra Chapter VI, Part A. (1279) See E-mail from A. Marzol to T. Howard, B. Graham, J. Boyles, and L. Spencer, dated December 31, 2003, Cataphora document i.1073065333000.1693380808. (1280) Undated ACE Primary MI Structure, FMSE-IN 7846-47. 330
recognize the acceleration of COLI policy expenses it would incur through 2007, which appears to be consistent with GAAP. We note only that Fannie Mae chose to accelerate its costs when the Company had excess earnings, even though it would have been economical for the Company to have done so previously. PART K: ACCOUNTING FOR OUT-OF-PORTFOLIO SECURITIZATION ("PORTFOLIO POOLING SYSTEM") This section of our report concerns an error in the interface between two systems that resulted in loans destined for securitization being erroneously designated as held for investment. I. INTRODUCTION In its February 11, 2005 letter to Stephen B. Ashley, OFHEO reported that an error in Fannie Mae's Portfolio Pooling System ("PPS") had led to the misclassification of loans that the Company held in its portfolio. The matter referred to Fannie Mae's Out of Portfolio Securitization ("OOPS") program, which the Company uses to securitize loans it acquires. According to OFHEO: During a system upgrade in 2004 for the Portfolio Pooling System (PPS), the Enterprise found a system error whereby all loans acquired during the month intended for pooling were marked as HFI [held for investment] at the end of the month regardless of whether the loans were initially marke[d] as HFS [held for sale] or HFI. In subsequent months, upon securitization and sale of the mortgage securities, the underlying loans were taken off the books and gains or losses were booked. This re-designation process occurred from the time that the OOPs [sic] program was created in 1983 until 2004, when, AFTER 21 YEARS, management detected and corrected the system error.(1281) OFHEO stated its understanding that the Company "did not perform any analysis to evaluate the magnitude of the error or the impact on the financial statements." Finally, OFHEO concluded, "For Fannie Mae to operate a system for two decades that inaccurately classified securities for financial reporting is simply unacceptable, and is indicative of a breakdown in oversight and compliance." (1282) We identified two issues for review in connection with the PPS error. First, what was the nature of the error, and second, why did it go undetected for over twenty years? - ---------- (1281) February 11, 2005 OFHEO Letter at FMSE-IR 547322. (1282) Id. 331
II. BACKGROUND Based on the information available to us, we have no reason to believe that the error was other than a coding mistake in the link between two of the systems involved in the OOPS and loan accounting processes. The Company likely did not notice the error in the interface because the Controller's Office received a report on HFS loans from PPS that was used to record lower of cost or market value ("LOCOM") adjustments before the error had any effect on the systems of record. Moreover, in light of the Company's accounting policies at the time (which permitted the Company to decide the classification of loans at the end of the month), combined with its strategy of holding substantially all mortgage loans as investments, it is understandable how the error could have been missed.(1283) When the Company uncovered the error, it considered the financial impact of the matter and, in view of its accounting policy at the time, concluded that it was inconsequential. We understand that, in the course of its restatement work, the Company is reviewing that assessment based on a re-evaluation of the relevant accounting standards. A. PPS and the Coding Error The coding error that OFHEO described in its letter occurred in the interface between two systems: PPS and LASER, the Company's system of record for the mortgage loans it holds in its portfolio. In order to understand the error and its impact, a review of the process that ends with Fannie Mae securitizing loans from its portfolio is necessary.(1284) - ---------- (1283) Although, FAS 65 is not as clear as FAS 115 regarding when loans should be designated as HFI or HFS. Compare ACCOUNTING FOR CERTAIN MORTGAGE BANKING ACTIVITIES, Statement of Fin. Accounting Standards No. 65, P. 32 (Fin. Accounting Standards Bd. 1982) (hereinafter "FAS 65") with ACCOUNTING FOR CERTAIN INVESTMENTS IN DEBT AND EQUITY SECURITIES, Statement of Fin. Accounting Standards No. 115 (Fin. Accounting Standards Bd. 1993) (hereinafter "FAS 115"). FAS 65 is read to require that a company determine the loan's classification at the time of the transaction. Fannie Mae designated loans as HFI or HFS at the end of the month of acquisition. We understand that Fannie Mae is reconsidering that policy as part of its restatement. Our explanation of how the error could have been missed considers the context in which the Company determined the classification of its loans; that explanation should not be construed as agreement with that accounting policy. (1284) At our request, personnel from Portfolio Management delivered a presentation describing the systems involved in the OOPS process and the nature of the error. We received similar information from personnel currently in Financial Standards. 332
Fannie Mae acquires mortgage loans through its "cash" program. The Company then determines whether it will keep the loans in its portfolio or whether it will securitize the loans (that is, transfer a pool of loans to a trust in exchange for a mortgage-backed security). PPS is used only to process the loans that the Company acquires in the cash program; it is not involved in the Company's far larger "swap" program, in which Fannie Mae exchanges loans for Company-issued MBS and directs those loans into a trust.(1285) Fannie Mae executes commitments to purchase loans each month through its eCommitting facility. When the transaction settles, the loans are processed through the Company's IBM Cash system. The IBM Cash system validates the loans received against the commitment, "funds" the loans, and records the transfer of ownership to Fannie Mae. The loans are then processed for transfer to PPS. The IBM Cash system and PPS sort loans into those that are eligible for pooling and those that are not. For example, Fannie Mae does not have the ability to pool adjustable-rate mortgages ("ARMs"); therefore, ARM loans are sorted into the "ineligible for pooling" category. Other loans also might have features that make them ineligible for pooling. Ineligible loans are passed through to LASER and do not appear in PPS. LASER's default setting for loans is HFI, so ineligible loans are given that designation. Loans that are eligible for pooling enter PPS. At that stage, and throughout the month in which the loans are acquired, a group in Portfolio Transactions reviews the loans and, by the end of the month, divides the loans into three categories: (1) loans that will be pooled and securitized at the end of the month; (2) loans that might be pooled and securitized at a future date but not at the end of that month; (3) and loans that the Company will keep in its portfolio indefinitely. At the end of month, PPS "tags" the loans. Those destined for pooling that month or later are designated as HFS; those that the Company will hold in its portfolio are designated as HFI. PPS transmits that information to LASER through an interface. Loans that are slated to be pooled that month are delivered into a trust. We understand that the system error arose from the difference in the timing of when loans are set-up in LASER and when PPS transmits its information to LASER. Loans are set-up in LASER at the beginning of the month following the month in which they are acquired. Accordingly, when PPS transmits classifications to LASER at the end of the month of acquisition, LASER does not show a record for that loan and the designation is not effective. At the beginning of the next month, when the loans are set up in LASER, that lack of designation results in a default designation of HFI. We also understand that loans designated for a future pool would appear in PPS each month - ---------- (1285) In an analysis of the PPS error that Jonathan Boyles prepared in March 2004, he reported that securitizations of the Company's own portfolio totaled approximately $7 billion, or 0.6% of the Company's total securitization activity. Mem. from J. Boyles to File, dated Mar. 31, 2004, Zantaz document 1944857, at 5. 333
subsequent to acquisition until those loans were either pooled or Portfolio Transactions determined they were not to be pooled. At the end of each of those succeeding months, those loans would reflect an HFS tag in LASER as a result of the designation in PPS until they were either included in a pool for securitization or the loan designation was changed from HFS to HFI (i.e., the loan would not be pooled). As OFHEO noted in its letter, the OOPS process was developed in the early 1980s. We have seen no documentation or other evidence to indicate that the system error was anything more than an oversight regarding the interface between PPS and LASER and the timing of the data exchange between the systems. As OFHEO noted, the principal question in these circumstances is one of controls - - that is, why the Company did not recognize the error during that twenty-year period. B. Accounting for HFS Loans The designation of loans as HFI or HFS was significant to Fannie Mae from an accounting perspective. Loans designated as HFI are carried on Fannie Mae's balance sheet at amortized historical cost. HFS loans are recorded at the acquisition cost, but are subsequently recognized at LOCOM, with adjustments reflected in income.(1286) We therefore considered the perspective of the Controller's Office with respect to the financial reporting of the loans that the Company held for sale. Three aspects of the error were significant. First, the Controller's Office did not make LOCOM adjustments based on the HFS inventory in LASER. Rather, the Controller's Office received a report each month of the inventory in PPS. According to Richard Stawarz, the Controller's Office based its LOCOM adjustments on the PPS reports. A document captioned "FAS 65 LOCOM Accounting Policy and Operational Procedures," although apparently updated recently to reflect more current practice, referred to the Controller's Office's receipt of reports from Portfolio, and from PPS specifically, regarding the inventory of HFS loans.(1287) Because the error did not affect the loans in PPS, the error would not have been apparent to the Controller's Office. Second, the Company's general ledger accounts during the period 2000 through 2005 show entries for LOCOM adjustments on HFS loans, but not consistently. The entries in 2000 were less than $200,000. We found no adjustments during 2001 through March 2004. Periodic LOCOM entries appeared in the general ledger after that - ---------- (1286) FAS 65 P. 4. (1287) FAS 65 LOCOM Accounting Policy and Operational Procedures, dated June 10, 2004, FMSE-IR 632788-91. The document has no indication of when it was first produced. The person identified as the "Process Performer" was Emily Passeri, who joined the Company in December 2003, but it is possible that the document predates her. 334
date. We have not identified the reason(s) for the three-year gap in LOCOM adjustments. However, the earlier LOCOM adjustments are consistent with the statement that the Controller's Office received reports of HFS loan inventory, and explains why an error in the designation process would have been invisible to it. Finally, in light of the Company's accounting practices in this area (i.e., classifying loans at the end of the month), the recorded LOCOM adjustments at least through 2004 were relatively small. This relates to the Company's accounting policy concerning HFS loans. Because the Company did not designate loans until the end of the month, it treated loans acquired and pooled during a month as securities, not as loans. Any decline in the fair value of most of the loans that passed through the PPS system during the month of acquisition, therefore, were small, and discrepancies were difficult to recognize. A change in that accounting treatment in 2004 appears to have originated in a memorandum prepared by Jeffrey Juliane, who noted that, "[d]uring the holding period of OOPS loans, prior to securitization, these assets need to be valued at LOCOM."(1288) He said "any loan that is purchased in the month and pooled in the same month, no matter how long the holding period of these loans are [sic] these assets are not valued at the lower of cost or market prior to delivery to STATs for transformation into a security. This is incorrect." (1289) Juliane reported that at the end of the previous period, the Company "had outstanding on our books $1.4 billion dollars [sic] of OOPs inventory in our security portfolio."(1290) LOCOM adjustments began again in March 2004.(1291) III. FINDINGS REGARDING THE PORTFOLIO POOLING SYSTEM The group within Fannie Mae to which the HFS designation was relevant was the Controller's Office. The PPS-LASER interface error did not affect their work, as they received reports from the PPS system before the error had any effect on the designation of the loans. Moreover, the Company's approach to accounting for HFS loans prior to 2004 meant that the volume of loans subject to LOCOM accounting was small, making detection of any error in the population more difficult. With these points in mind, we can understand how the error in the interface between PPS and LASER could have been missed for as long as it was. - ---------- (1288) Mem. from Jeffrey Juliane to Janet L. Pennewell and Mathew Douthit, dated Jan. 28, 2004, FMSE-IR 521365-69, at FMSE-IR 521365. (1289) Id. (1290) Id. (1291) This information is reflected in the Company's general ledger, which is available on Homesite. 335
PART L: THE DEBT REPURCHASE ("BUYBACK") PROGRAM I. INTRODUCTION This section addresses Fannie Mae's execution of debt repurchase transactions ("buybacks") in the period from 2000 through 2004, including management's motivation for the buybacks, the impact of the buybacks on the Company's reported earnings, and the disclosures internally and to the public concerning the buybacks. Buyback transactions were not raised as an area of concern by OFHEO in its 2004 Report or in its February 11, 2005 letter to Stephen B. Ashley. We focused on buybacks as transactions after discovering e-mails between senior management that appeared to focus on EPS impact of conducting buyback transactions. Based on our review, we conclude that buyback transactions were accounted for and reported in accordance with GAAP. Management also fully disclosed the debt buybacks and the resulting losses in Fannie Mae's published financial statements. We further conclude that there were legitimate business reasons for executing debt buybacks from 2000 to 2004, including management of interest rate risk and other portfolio management-related goals. We have found however, that management's execution of buybacks suffered from the following deficiencies. First, while there was a legitimate business purpose for conducting debt buybacks, management also had the purpose of using buybacks to affect the Company's EPS. Evidence showed that management conducted buybacks to depress income in 2001 through 2003, so that the Company could show stable earnings growth in subsequent years. Moreover, in determining the size of the buybacks on any particular occasion, management appears to have focused primarily on the impact on current period EPS, rather than on any economic analysis. Management's EPS-related purpose for the buybacks was never disclosed to the Board at any time, including at a meeting of the Assets and Liabilities Policy Committee of the Board in January 2004, where management presented, for the first time, an after-the-fact view of the buybacks conducted in prior years. Notwithstanding management's failure to disclose its EPS-related strategy for buybacks, however, the financial press ultimately did understand and focused on this strategy in mid-2004. We believe it was incumbent on management to fully disclose to the Board its EPS-related purpose in executing buybacks, before executing the transactions, especially in view of the fact that management's compensation was tied not only to EPS, but also to its ability to grow EPS by certain percentages on an annual basis. Second, in our view, management executed buybacks in the relevant time period with little or no contemporaneous analysis of the transactions' economic benefits to the Company, and followed no clear policy or procedure before executing such transactions. While we have not found evidence that the buyback transactions were clearly uneconomic, the absence of any procedure or contemporaneous documentation supporting buyback decisions represented, in our view, a control weakness. For example, management who were interviewed about the buybacks conducted from 2001 through 2003 offered two contradictory statements about whether contemporaneous economic 336
analysis existed. On the one hand, Controller's Office staff, such as Janet L. Pennewell, contended that there was a contemporaneous "economic analysis"(1292) supporting Portfolio's recommendations on buybacks; indeed, a document reflecting Franklin D. Raines's request for supporting economic analysis for buybacks seemed to corroborate Pennewell's recollection. On the other hand, Peter Niculescu, Executive Vice President--Portfolio, dismissed the idea of conducting economic analysis before recommending buybacks during the relevant time period as ridiculous, and contended that such analysis would have been so "simple" that it would not have been worth the paper on which it is written. This inconsistency in senior management's recollections about the need or existence of supporting documentation for transactions that caused the Company losses in excess of $2 billion in one year, suggested to us that the Company's policies and procedures relating to such transactions were inadequate. Accordingly, we recommend that the Company develop and implement clear procedures that identify the documentation and approval authorities required for conducting buyback transactions. II. BACKGROUND Fannie Mae is one of the world's largest issuers of debt securities. Fannie Mae has stated in public disclosures that its primary purpose for selling debt instruments is to obtain funds to finance its mortgage purchases and other business activities.(1293) Fannie Mae operates in the secondary mortgage market by purchasing mortgage loans from a variety of lenders, including mortgage companies, savings institutions, credit unions, and commercial banks. Fannie Mae packages these loans into MBS, for which it guarantees payment of principal and interest, or purchases these loans for cash and retains the mortgages in its portfolio.(1294) From time to time, Fannie Mae extinguishes debt by repurchasing previously issued noncallable debt at market rates, or by exercising its option for callable - ---------- (1292) We have not found any contemporaneously prepared analysis that would have permitted management to review and approve the "economics" of buyback transactions prior to their execution. An "economic" analysis would reflect the assumptions made with respect to the terms and duration of the replacement debt and the debt cost savings compared to the incremental cost of the buyback, as well as the effect of the buyback transactions on Fannie Mae's asset/liability mix. See infra Subsection III.B. (1293) See Fannie Mae, Understanding Fannie Mae Debt, available at http://www.fanniemae.com/markets/debt/understanding_fm_debt/ ufmd_role.jhtml? p=Debt+Securities&s=Understanding+Fannie+Mae+Debt (last visited Feb. 17, 2006). (1294) Fannie Mae, 2003 Annual Report (Form 10-K), at 2 (Mar. 15, 2004), available at http://www.fanniemae.com/ir/pdf/sec/2004/f10k03152004.pdf (hereinafter "2003 Form 10-K"). 337
debt securities.(1295) According to Fannie Mae's disclosures, debt extinguishment serves to reduce future debt costs and enable the Company to strategically manage interest rate risk in response to market conditions.(1296) The total amount of debt extinguished appears on the Company's income statement as a gain or loss on the repurchase of debt.(1297) From 2001 to 2003, Fannie Mae recognized cumulative pre-tax losses on debt extinguishments of approximately $3.5 billion.(1298) This figure includes the call and repurchase of debt securities and notional principal of interest rate swaps.(1299) The total debt extinguishment amounts for fiscal years 2001 through 2003 are summarized below:(1300) Table 1: FISCAL YEAR (DOLLARS IN DEBT TOTAL DEBT PRE-TAX GAIN MILLIONS) DEBT CALLED REPURCHASED EXTINGUISHMENT (LOSS) - ------------ ----------- ----------- -------------- ------------ FY 2001 $ 173,000 $ 9,000 $ 182,000 ($ 524) FY 2002 174,000 8,000 182,000 (710) FY 2003 246,000 20,000 266,000 (2,261) ----------- ----------- -------------- ------------ Total $ 593,000 $ 37,000 $ 630,000 ($ 3,495) =========== =========== ============== ============ The substantial increase in debt extinguishment activity during 2003 was driven by several factors. Per Fannie Mae's 2003 Form 10-K: The historically low interest rate environment that persisted through the first half of 2003 made it economical for us to call or redeem high cost debt. In anticipation of increasing liquidations in our mortgage portfolio as a result of heavy refinancing activity, we replaced the higher cost debt with shorter-term, lower-cost debt. This dynamic resulted in a - ---------- (1295) Fannie Mae publishes a list of outstanding callable and noncallable debt on the Company's website. See Fannie Mae, Debt Activity, available at http://www.fanniemae.com/markets/debt/debt_activity/outstanding.jhtml? p=Debt+S ecurities&s=Debt+Activity&t (last visited Feb. 17, 2006). (1296) 2003 Form 10-K at 38. (1297) Id. at 122. (1298) Id. at 38. (1299) Id. (1300) Id.; Fannie Mae, 2002 Annual Report (Form 10-K), at 32 (Mar. 31, 2003), available at http://www.fanniemae.com/ir/pdf/sec/2003/f10k03312003.pdf (hereinafter "2002 Form 10-K"). 338
temporarily elevated spread between the yield on our mortgage assets and our funding cost. We chose to reinvest a portion of the income generated from these temporary circumstances in debt repurchases to reduce our future debt costs.(1301) Additionally, Fannie Mae's 2003 Form 10-K states that: A loss realized on a debt repurchase in the current year results in correspondingly lower debt costs in the future. The debt cost savings will be realized over the period remaining until the scheduled maturity of the debt repurchased. We select individual securities to repurchase based on which securities we believe are trading at attractive prices in the market as well as asset/liability risk management purposes. We expect a significant reduction in the level of debt repurchases in 2004.(1302) The Company refers to the early extinguishment of noncallable debt securities as its "Benchmark Securities Buyback Program," which will be referred to below as the "Benchmark Buyback Program."( 1303) The majority of debt extinguishment losses from 2001 through 2003, shown in Table 1, supra, resulted from the Benchmark Buyback Program. As can be seen in Table 2, infra, below, the total loss incurred by Fannie Mae from this program over this period was approximately $3 billion. Additionally, the Benchmark Buyback Program represented approximately 89% of the Company's debt repurchases during the years 2000 to 2004,(1304) and 86% of the losses - ---------- (1301) 2003 Form 10-K at 38-39. (1302) Id. at 39. (1303) See Fannie Mae, Benchmark Securities - Buyback Program, available at http://www.fanniemae uybacks_overview (last visited Feb. 17, 2006). (1304) During the same period, there were seventy buyback transactions categorized by Fannie Mae as "outside" the Benchmark program. Of this total, all forty-four transactions in 2004 appear to be outside the Benchmark program. See Undated Buyback Total Debt spreadsheet, Zantaz document 1573400, attached to e-mail from Ashley Dyson to David Benson, et al., dated Mar. 16, 2005, Zantaz document 1573396. 339
incurred on buybacks in that period.(1305) Table 2 is a summary of data obtained from Fannie Mae.(1306) Table 2 FISCAL YEAR (DOLLARS IN MILLIONS) FY 2000 FY 2001 FY 2002 FY 2003 FY 2004 TOTAL - ----------------------------- ------- -------- --------- ----------- ------- ---------- Benchmark Securities Buybacks w/Book Gain $ 52.58 $ - $ 8.12 $ 2.10 $ - $ 62.79 Buybacks w/ Book Loss - (357.59) (610.05) (2,104.21) - (3,071.85) ------- -------- --------- ----------- ------- ---------- Benchmark Book Gain (Loss) 52.58 (357.59) (601.93) (2,102.11) - (3,009.06) ------- -------- --------- ----------- ------- ---------- Outside Benchmark Securities - (53.85) - (55.32) 32.14 (77.03) ------- -------- --------- ----------- ------- ---------- Total Book Gain (Loss) $ 52.58 ($411.44) ($ 601.93) ($2,157.43) $ 32.14 ($3,086.08) ======= ======== ========= =========== ======= ========== This section only addresses Fannie Mae's extinguishment of Benchmark debt.(1307) Moreover, given the size of the losses and increases in buyback activity, the primary focus of our investigation was fiscal years 2001 through 2003. III. FINDINGS REGARDING FANNIE MAE'S BUYBACK PROGRAM Our investigation addressed the following primary issues with respect to Fannie Mae's debt buyback program: (A) the use of debt repurchases to meet current EPS targets and depress future EPS growth targets through a reduced EPS base; (B) the possibility that uneconomic debt repurchases were entered into for this purpose; and (C) the adequacy of disclosures relating to the buybacks. - ---------- (1305) See Huron's Analysis of Fannie Mae's March 15, 2005 Debt Repurchase, Benchmark Buyback Program spreadsheet: Debt Buyback Results Percentage of Total Transactions, FY 2000-2004 worksheet. The analysis was derived from Undated Buyback Total Debt spreadsheet, Zantaz document 1573400, attached to e-mail from A. Dyson to D. Benson, et al., dated Mar. 16, 2005, Zantaz document 1573396. (1306) See Undated Buyback Total Debt spreadsheet, Zantaz document 1573400, attached to e-mail from A. Dyson to D. Benson, et al., dated Mar. 16, 2005, Zantaz document 1573396 (documenting buyback transactions in fiscal year 2000 through the first quarter of 2005). The book gain (loss) differences of ($112.6) million, ($108.1) million, and ($103.6) million in 2001, 2002, and 2003, respectively, between Table 1 and Table 2, represent the variance from the reported gain (loss) on debt extinguishment per Fannie Mae's 2003 Form 10-K and the total book gain (loss) of Benchmark buybacks recorded on the Company's transaction spreadsheets. See id. (under the "Book Gain ($mm) Pre-tax" column). (1307) Our review of Fannie Mae's accounting treatment for extinguishment of callable debt will be addressed separately. See discussion infra Chapter VI, Part M. 340
A. Earnings-Related Strategy In its published financial statements, the Company offered two primary reasons for entering into debt buyback transactions: (1) management of the Company's interest rate risk, and (2) the reduction of future debt costs.(1308) In order to achieve its goals related to interest rate risk, management uses debt buybacks to manage the duration gap, as a "portfolio management tool" to control the Company's debt trading mix, and as a means to achieve asset/liability balancing. We have no reason to question the propriety of these stated goals, and further conclude that the Company's financial reporting for these transactions was in compliance with GAAP. However, we found several documents suggesting that management set parameters for the debt buybacks based on current quarter earnings forecasts, and that buyback transactions were executed strategically within those parameters to achieve specific, to-the-penny, EPS figures. These EPS figures, in turn, were in line with analyst expectations and predetermined AIP targets. For example, in a September 24, 2003 e-mail to Spencer and Niculescu, Howard described a conversation that he had had with Franklin D. Raines in which the two contemplated using debt repurchases to control EPS growth for the third quarter of 2003, and specifically to achieve a double-digit increase over the previous year. It appears that the goal of hitting a predetermined EPS target, including a one cent "cushion," was the primary driver in the decision about how much debt to repurchase in this quarter. Howard stated: I was able to get to Frank this evening on the buyback loss for the third quarter. (I also showed him the sheet that had $8.87 EPS in 2004, but told him that it was the result of front-ended income that we did not intend to let flow through.) His thought for the third quarter - which I think is a good one-- was to come in at an EPS number that would be a double-digit increase from the third quarter of 2002. A third quarter EPS of $1.79 would do that - it would be 10.5% above the $1.62 we reported in the third quarter of 2003. If that's what we want to do, doing $400 million buyback tomorrow would cause us to fall short of our objective. Using Leanne's numbers, a $400 million buyback would put us at $1.78. And of course, that's without any cushion. - ---------- (1308) Fannie Mae, 2001 Annual Report, at 26, available at http://www.fanniemae.com/ir/annualreport/index.jhtml?s=Annual+Reports+%26 +Proxy+Statements (last visited Feb. 17, 2006); 2002 Form 10-K at 32; 2003 Form 10-K at 38. 341
So - we need a lower cap. Without doing any analysis, I'd be inclined to say $350. A $375 million cap would give us a penny of cushion ($1.80 versus our $1.79 target). Leanne, how much of a cushion would you like, if we're shooting for double-digit growth from Q3 2003? Your view of the right cushion should determine the maximum loss number we give to Dave.(1309) Another example, a September 5, 2001 e-mail from Howard to Spencer and Pennewell, demonstrated a similar approach to budgeting buyback losses in terms of EPS targets, in this instance for future periods. According to Howard, Linda K. Knight and others believed: [Fannie Mae] could generate losses in the $40 to $50 million range. The current projection has us doing $80 million in losses on buybacks, so - assuming the rest of the projection remains unchanged and we do indeed wish to come in at $1.31 on third quarter EPS - this would leave us with $30 to $40 million to do later on this month in long zeros or 30 years.(1310) Pennewell did not dispute that buyback transactions had been used to meet EPS forecasts. She described a process, similar to the approach demonstrated in the above e-mails, whereby Howard and Raines determined a "budget" for buyback losses based on the Controller's Office earnings forecast, and applied that budget to the transactions recommended by the Treasurer. Pennewell stated her belief that the buyback budget corresponded to the difference between actual earnings and earnings targets. Knight described a similar practice and stated that the Treasurer generally identified buyback opportunities based on favorable pricing and then engaged in conversations with Spencer and/or Howard, who gave approval for certain of those transactions based on specific gain/loss parameters. Niculescu stated that he was not involved in the earnings-strategy component of buyback sizing decisions. Rather, according to Niculescu, he had given Howard in 2003 a "standing" recommendation to execute as many debt repurchase transactions as possible due to the compelling economic environment. Neither Knight nor Niculescu could identify an instance in which the Company missed an EPS target as a - ---------- (1309) E-mail from Timothy Howard to Peter Niculescu and Leanne G. Spencer, dated Sept. 24, 2003, Zantaz document 554934, at 2. This and other examples of documents relevant to the discussion in Chapter VI, Part L can be found in the accompanying Appendix, at Tab D.12. (1310) E-mail from J. Pennewell to Shaun Ross et al., dated Sept. 5, 2001, Zantaz document 1562367, at 1 (forwarding earlier e-mail from T. Howard to L. Spencer, J. Pennewell, et al.). 342
result of executing a buyback transaction. We have similarly not found any evidence of such an instance. Additional documents support the notion that the principal driver behind Fannie Mae's buyback budget related to present-period earnings management. For example, in a December 10, 2003 e-mail to Knight regarding a debt buyback transaction, David Benson, Vice President -- Credit Policy, cited Pennewell as communicating a "need" for $250 to $325 million.(1311) Similarly, in a May 30, 2003 e-mail to Pennewell and Spencer, Benson similarly stated that "[the buyback transaction's] objective would be to meet any quarter end requirements for income and liability management purposes."(1312) And, in an October 16, 2003 e-mail to Spencer and others about not "overshooting" EPS, Howard stated that "[t]he FAS 149 transition adjustment gave us an additional $285 million (pre-tax) in income. To hit the EPS number we'd planned for, we needed to buy back that much more debt."(1313) Other evidence showed that another objective of management's buyback strategy was to diminish earnings in the present period in order to more easily achieve future EPS growth goals. In particular, management strategically used debt buyback transactions as one of several means to achieve the challenge laid down by Raines in 1999 to double EPS in five years. An unidentified speech prepared for a senior management retreat in 1999 made this point clear: So having to double earnings when your base year has very high income is a challenge. Tim [Howard], who has always been more clever than I, on the other hand, used the excess float income to buy back debt at a loss thereby depressing his income in the base year.(1314) In addition to this document, members of management interviewed during this investigation have consistently cited debt buybacks as a method used by Fannie Mae in 2001 and 2002 to "shift" income into future years.(1315) For example, Adolfo Marzol stated that Fannie Mae repurchased debt pursuant to guidance from Raines in May 2001 - ---------- (1311) E-mail from D. Benson to Linda K. Knight, dated Dec. 10, 2003, Zantaz document 1577517. (1312) E-mail from D. Benson to L. Spencer and J. Pennewell, dated May 30, 2003, FMSE-IR 566482-84, at FMSE-IR 566482. (1313) E-mail from T. Howard to L. Spencer et al., dated Oct. 16, 2003, FMSE-E 31897-98, at FMSE-E 31897. (1314) Cape Cod Strategic Meeting, dated June 28, 1999, FMSE-IR 577572-86, at FMSE-IR 577572. (1315) QBR Summaries, dated May 2002, FMSE-IR 565521-27 at FMSE-IR 565524. 343
that the Company seek opportunities to invest in the future instead of maximizing profit in the current year.(1316) Niculescu also stated that management, and Howard in particular, sought buybacks as a means to pursue a "stable pattern of earnings," which he confirmed meant sequential earnings growth.(1317) This EPS-related strategy for buybacks was, in our view, a possible motivation for management because management's bonus plan was tied to final reported EPS without regard for non-operating or non-recurring items. Had the Company not executed buybacks in these years, the Company would have achieved EPS numbers that were even higher than the actual results. Such results, however, would have set a higher bar for showing growth for years 2003 and 2004. By engaging in transactions that depressed the EPS for 2001 and 2002, management lowered the base year for purposes of measuring subsequent years' EPS growth, which was a metric used for setting bonuses.(1318) B. Documentation for Buyback Decisions We have found no contemporaneous documentation of any analysis of the economic benefits of entering into buyback transactions, either by the Controller's Office, Mortgage Portfolio, or senior management, including the CFO. Based on our review of documents and based on witness interviews, we conclude that management did not perform any contemporaneous economic analysis of the benefit of debt buyback transactions. While certain individuals, such as Pennewell, stated the belief that such analyses had been performed by Mortgage Portfolio, none could recall with any specificity or identify any documentation. Knight and Niculescu, furthermore, denied that such analysis was prepared in their areas. Niculescu stated that the calculation of economic benefit to Fannie Mae of buyback transactions was so simple and obvious that it likely would not have been written down. He also stated that in 2003, he had made a standing recommendation to Howard to buy back as much debt as possible. Nevertheless, at least one document we reviewed suggests that at least Raines believed that an economic analysis was relevant to his decision as to whether to authorize a large debt repurchase. Responding to brief descriptions by Howard and Niculescu of the rationale for entering into $1.7 billion of debt repurchase and losses of $290 million, Raines stated: "I am fine with this. The document you showed me didn't seem to have the economics of the buybacks included - ---------- (1316) While not referring to specific guidance from Raines, Robert Schaefer similarly recalled that buybacks were one way that the company sought to increase income in future years. (1317) Louis Hoyes, Executive Vice President - Single Family, also noted that buybacks were one means by which Fannie Mae could achieve stable earnings growth. (1318) See generally infra Chapter VIII, Subsection III.B.2. 344
on it. Did I misread it?"(1319) This e-mail reflected at least Raines's impression that such an analysis was being performed. While the economic advantage of some buyback transactions may have been so obvious or intuitive to Niculescu or Howard given market conditions at the time, the total absence of any documentation requirement or process for conducting these significant financial transactions, demonstrated an ad hoc approach to conducting buybacks that, in our view, was a control weakness. Similarly, while Fannie Mae publicly stated that interest rate risk management was a primary purpose for conducting debt buybacks,(1320) we also did not find any analysis or documentation of this effort. Knight affirmatively stated that Fannie Mae did not, as a matter of course, analyze a link between replaced debt and the new, cheaper debt issued by the Company. Niculescu minimized asset/liability balancing as a driver for buyback transactions and stated that buybacks would only help Fannie Mae's asset/liability management strategy to a very small extent. Niculescu further stated that an analysis regarding the effect of buyback transactions on Fannie Mae's asset/liability balancing would be complex, and would be difficult to execute contemporaneously for each transaction.(1321) Finally, we found no policy or operational document governing Fannie Mae's debt buyback program.(1322) Pennewell, Knight, and Niculescu all stated that Howard retained and exercised his authority both to set the budget for buyback losses, and ultimately to sign-off on only certain of those transactions identified by the Treasurer. Howard's authority over buyback transactions appears to be confirmed by the documents described above. The lack of written guidance on debt buybacks, and Howard's apparently broad discretionary authority without Board of Directors authorization, further cemented our views that there were little or no controls governing buyback decisions. C. Economics of Fannie Mae Debt Buybacks To determine whether the buyback transactions executed by management during the relevant period were, in fact, economically beneficial to the Company, we analyzed over 600 debt securities repurchased between 2000 and 2004 as part of the - ---------- (1319) E-mail from F. Raines to T. Howard, dated Dec. 17, 2003, Zantaz document 872583, at 1. (1320) 2003 Form 10-K at 38. (1321) Id. (1322) See Buyback Process Meeting, dated Jan. 20, 2004, FMSE-IR 685233 (suggesting a lack of formal guidelines for buyback transactions). 345
Benchmark Buyback Program.(1323) Our testing did not identify any transactions that were clearly uneconomic. Our assessment was high-level and based on data provided by the Company.(1324) A full assessment of the economics of buyback transactions would center on the savings associated with replacing high-cost debt with lower-cost debt compared to the incremental cost of the buyback. The assessment would also include the related repricing and interest rate risk of future movement of interest rates at the time of the buyback transactions as well as the use of capital to fund the significant losses incurred. As part of our analysis, we compared the loss on the retired debt to the present value of the interest savings from repaying the high coupon debt and replacing it with lower market rate debt of the same maturity. Use of matched maturity replacement debt assumes that the maturity of the replacement debt is equal to the remaining period of the extinguished debt. This assumption makes the transaction "neutral" from an interest rate risk/asset and liability perspective before consideration of prepayments on mortgage investments. Under this testing, we concluded that 76.7% of the Benchmark transactions had insufficient savings on a matched maturity basis. The present value of the total benchmark buybacks with insufficient savings on a matched maturity basis for the period 2001 through 2004 was approximately $627 million and the calculated net future cost for Benchmark buybacks form 2001 through 2004 totaled approximately $404.4 million.(1325) These findings are summarized in Table 3, below. - ---------- (1323) We did not include transactions identified by the Company as "Not Benchmark" or "Outside Benchmark" in this analysis. (1324) We relied primarily on several versions of Fannie Mae's "Buyback Total Debt" spreadsheets that detail historic buyback transactions. The updated versions contain current and historical data. The most recent spreadsheet was attached to a March 16, 2005 e-mail from Dyson to Niculescu, Benson, Knight, Ross, and others regarding "MTN buyback." The attachment "Buyback Total Debt" contains buyback transaction data from 2000 through the first quarter of 2005, including the "Notional Repurchased ($mm)" amount, the Company "Book Gain ($mm) Pre-tax," and corresponding "Trade Date" information for each buyback. See supra notes 1304 and 1306. (1325) Per Fannie Mae's documentation, thirteen, thirteen, and forty-four buyback transactions in 2001, 2003, and 2004, respectively were categorized as "outside Benchmark securities." See Undated Buyback Total Debt spreadsheet, Zantaz document 1573400, attached to e-mail from Ashley Dyson to David Benson et al., dated Mar. 16, 2005, Zantaz document 1573396. 346
Table 3 FANNIE MAE FY 2001-2004 BUYBACKS -------------------------------------------- Number of % of Debt Buyback Transactions Transactions Amount ($mm) Transactions ------------------------- ------------ ------------ ------------ Buybacks w/ Net Future Savings 122 $ 222.59 23.3% Buybacks w/ Net Future Cost 401 ($ 627.03) 76.7% --- --------- ----- Total Benchmark Buybacks 523 ($ 404.44) 100.0% --- --------- ===== Outside Benchmark Securities 70 ($ 18.91) --- --------- TOTAL 2001 - 2004 BUYBACKS 593 ($ 423.35 === ========= However, to the extent the Company was replacing and shortening the debt maturity in response to an acceleration of prepayments on the mortgage assets held for investment, this matched maturity assumption does not reflect the true potential savings because it does not reflect the lower cost of shorter term debt and therefore is not conclusive. Indeed, while the lack of a full economic analysis by the Company prevents us from fully modeling all factors, our independent analysis of reasonable "break even" rates (i.e., the rates at which the Company could have borrowed in order to recover losses incurred as a result of the buyback transactions) and corresponding maturities during the relevant period suggests that the Company could have recouped the losses incurred in connection with the buybacks. D. Disclosures We next examined whether management fully disclosed all components of its debt buyback strategy to the public and to the Board of Directors. 1. Public Disclosure Fannie Mae accurately disclosed the most important facets of its Benchmark Buyback Program to the investing public. These disclosures included some statements about the role of earnings in buyback decisions. For example, Fannie Mae stated in its Form 10-Q for the quarter ending March 31, 2004: "Because debt repurchases, unlike debt calls, may require the payment of a premium and therefore result in higher extinguishment costs, we generally repurchase high interest rate debt at times (and in amounts) when we believe Fannie Mae has sufficient income available to absorb or offset those higher costs."(1326) - ---------- (1326) Fannie Mae, 2004 Quarterly Report (Form 10-Q), at 11 (Aug. 9, 2004), available at http://www.sec.gov/Archives/edgar/data/310522/ 000095013304003115/w99629e10vq.htm. An earlier draft of similar language in the second quarter 10-Q also contained the following qualifier: "provided the impact of such purchases is consistent with our earnings goals." See Revised Debt Extinguishment Language, 347
By early 2004, market analysts had made the connection between Fannie Mae's Benchmark Buyback Program and the Company's earnings goals. For example the Dow Jones Newswire published an article on April 15, 2004 regarding Fannie Mae's first quarter financial results, entitled "Fannie Mae 1Q EPS Seen On Target, As Debt Buyback Slowed."(1327) The article stated: Fannie Mae (FNM) is expected to post a modest gain in first-quarter profits, largely because the mortgage giant scaled back a deliberate strategy of debt repurchases that weighed down the bottom line in previous periods. . . . "We believe a reduction in debt buyback expenses provided a cushion to earnings," Lehman Brothers analyst Bruce Harting said. Fannie's debt-repurchase strategy was a voluntary practice it used in recent years to "cookie jar" or push out abundant earnings to future quarters. In effect, Fannie Mae paid premiums to buy back high-cost debt, moderating earnings growth and lowering future funding costs. In breaking the habit, Fannie Mae may be signaling that the days of heady earnings growth have waned. "The losses on debt repurchase last year were largely discretionary, designed to transfer income from years of enormous strength, such as the 2003 refi boom, to leaner times, such as years of falling refis, rising ARMS and slowing asset growth," Sanford Bernstein analyst Jonathan E. Gray recently said in a research note.(1328) In an internal e-mail forwarding the above article to Niculescu, Howard did not dispute Jonathan Gray's characterization of the buybacks: FYI - I'll bet that most analysts and investors will view the absence of debt buybacks in the first quarter the same way Jonathan Gray did. If so, we may not even get the question, and if we do get it, some version of his - ---------- dated May 10, 2004, Zantaz document 790842, at 2. The qualifier, however, was removed before the disclosure was finalized. (1327) E-mail from T. Howard to P. Niculescu, dated Apr. 15, 2004, Zantaz document 877604, at 1 (forwarding the Dow Jones Newswire article of same date). (1328) Id. at 1 (emphasis added). 348
interpretation as our answer is what most of our audience will expect to hear.(1329) While Fannie Mae's public disclosures did not state expressly that formal economic analyses supported buyback decisions, management made several statements in disclosure documents and on investor calls that suggested the existence of a contemporaneous analysis of economic benefit, which, as noted above, did not exist. For example, the Company stated in its 2003 Form 10-K that "[t]he historically low interest rate environment that persisted through the first half of 2003 made it economical for us to call or redeem high cost debt."(1330) Howard and Niculescu also engaged in the following discussion during a July 15, 2003 call with investors: PETER NICULESCU, EVP MORTGAGE PORTFOLIO BUSINESS, FANNIE MAE: Let me add just a couple of brief points. The first is that the actual liquidations we're expected to see in the mortgage portfolio in July and August will almost certainly be higher than the numbers we recorded in May and June . . . it takes a couple of months for that to flow through into mortgage liquidations. And you will see higher prepayments, higher liquidations out of the mortgage portfolio in July and August than we had seen previously. The second point is that our opportunities to buy back debt have been relatively advantages to us in the last couple months, really ever since the presses partly as a result of the Freddie Mac restatement. And it may very well be the case as we look forward in the next couple of months that we will continue to see advantageous, economically advantageous opportunities for us to buy back debt as a result of some of these, the current market trading levels. [DAVID HOCHSTIM]: Okay. And then could you just talk again a bit about debt repurchases and the NPV of - I guess prospective purchases and how much more positive it can be if you get these market opportunities that Freddie creates for you and over what period of time could these, I guess, benefits accrue? PETER NICULESCU: . . . It's worth noting that if we - if we buy, let's say, three-year debt and as a result, incur a -- a loss at the time of purchase, what we have done is - ---------- (1329) Id. (1330) 2003 Form 10-K at 38. 349
reduced lower our costs of funds for the next three year, by the future value of that - of that debt by - of cost. So it is an immediate lowering of our - of our cost of funds for that time period. Obviously we're lowering the cost of funds by less on an annual basis but for a longer time period. TIM HOWARD: Peter, it's true that the payback we get is typically more than the present value of the expenditure, that we will typically be able to achieve debt reissue or having trade better for subsequent issues. PETER NICOLESCU [sic]: Absolutely. This is part of what's been helpful to us as a result of some of the market volatility that we've seen recently that we've been able to buy back debt at levels that are generally more available to us than issuing the debt so there's a net benefit in the immediate future that we can execute that. Along the term there's probably a fit to the overall market in our ability to issue and that becomes clear to our investors and our debt that if the debt reaches a level where it's attractive for us to buy it back, we will do so, and that -- that, in turn has a beneficial effect, I believe, on we can issue debt and where the debt program as a whole goes in the future.(1331) Niculescu confirmed that there were no contemporaneous analyses performed to determine "economically advantageous opportunities" or to ensure that "payback" was more than "present value of the expenditure." In fact, in Niculescu's view, such analyses would be so elementary as to not require written documentation. Notwithstanding his views, the above discussion could be read to suggest that management had conducted economic analysis before making buyback decisions, which would not be accurate. 2. Board Disclosure We have found no evidence showing that management disclosed to the Board its earnings-related strategy for buybacks. In fact, we found no statements to the Board or any of its committees prior to 2004 - when the vast majority of the buyback transactions already had been executed - and the statements in 2004 to the Board did not disclose the earnings-related strategy for buybacks. Niculescu gave a presentation on "Debt Repurchase" to the Assets and Liabilities Policy Committee of the Board of Directors of Fannie Mae on January 22, - ---------- (1331) Draft updated disclosure outline attachment to e-mail from J. Lopez to G. Baer et al., dated Sept. 30, 2004, Zantaz document 2920368, at 9-10. 350
2004.(1332) The presentation made brief reference to an earnings component in buyback decisions, stating: "If market conditions provide a compelling opportunity to buy back debt, we will consider the repurchase and factor the earnings and cost impacts into the decision."(1333) Statements elsewhere in the presentation, however, as well as the line quoted above, suggest that decisions on buybacks were based on economics, cost/benefit analysis, and interest rate risk assessments. For example, Niculescu cited the following as factors for "[d]etermining [w]hich [s]ecurities to [r]epurchase": a. Duration management - If the match of assets with liabilities requires less long-term debt, we can reduce long-term debt by buying it back b. Reducing the average future cost of debt - The attractiveness of buying back debt in any particular maturity is relative to where we expect to be able to fund at the same maturity in the future. This involves making judgments on the relative attractiveness of different repurchase candidates.(1334) Taken as a whole, in our view, the presentation suggested to the Board that buyback decisions were based primarily on an analysis of multiple factors relating to the economic benefit to the Company, and did not disclose the fact that earnings management was a principal driver behind management's buyback decisions. In summary, we have not found any evidence that management ever disclosed to the Board its strategic use of buybacks to structure EPS growth or to achieve forecasted EPS. Such information, in our view, was critical information for the Board's ability to assess the quality of the Company's performance; moreover, because bonuses were calculated based on the final reported EPS without regard for non-operating or non-recurring items, we believe that management's strategic use of buybacks to achieve earnings goals should have been made transparent to the Compensation Committee for its consideration in awarding bonuses. - ---------------------- (1332) See Debt Repurchase presentation, Assets and Liabilities Committee, Fannie Mae Board of Directors, dated Jan. 22, 2004, Zantaz document 1371682. (1333) Id. at 4. (1334) Id. at 8. 351
IV. RECOMMENDATIONS Based on these findings, we believe that the Company would benefit from three primary recommendations. First, we recommend that management contemporaneously document the basis for recommending or executing buyback transactions. We do not challenge earnings targets as a factor to consider before executing buyback transactions. However, we question the strategic use of buyback transactions to achieve specific EPS goals, as outlined above, and we find that the disclosure to the Board about management's strategy for the buybacks were inadequate. We believe that accurate documentation of the decision to conduct buyback transactions is necessary to ensure that management's purpose is transparent to all, including the Board. We recommend that Fannie Mae implement procedures and policies for such documentation going forward. Second, in similar vein, the Company should improve its internal controls regarding buyback transactions by creating and documenting a clear execution, approval, and monitoring process. As noted in the April 2005 Office of Audit's "Debt Funding Audit," the Company's controls were inadequate and: "Accounting policy guidelines need to be expanded to better address . . . benchmark repurchase agreements[.]"(1335) In addition, the report noted that the monitoring controls of the Treasurer's Back Office ("TBO") require improvement including more formal review and approval by management.(1336) Management's corrective actions addressing these concerns stated that effective immediately, "TBO will ensure that all Benchmark repurchases are verified by a TBO analyst and reviewed by TBO management, including formal sign-off on the Trade Log Binder."(1337) Finally, the corrective actions stated: "In addition, all benchmark repurchases require approval from the Executive vice [sic] President [Niculescu] prior to execution. Treasury Compliance will monitor all transactions, to ensure that proper authorization is received. Monitoring efforts began March 28, 2005."(1338) We recommend that management implement these corrective actions immediately, to the extent it has not already done so. Additionally, we recommend that the Executive Vice President -- Mortgage Portfolio perform a comprehensive review of the economics of buyback transactions as part of the approval process. The Company also should maintain its documentation of any analysis consistent with its document retention policies. - ------------ (1335) Debt Funding Audit, Audit Report, Office of Auditing, dated Apr. 29, 2005, FMSE-IR 685070-87, at FMSE-IR 685071. (1336) Id. at FMSE-IR 685079. (1337) Id. at FMSE-IR 685081. (1338) Id. 352
Finally, we recommend that management provide regular and comprehensive briefings to the Board of Directors regarding the purpose, nature, and results of buyback transactions before they are executed, rather than after-the-fact. PART M: ACCOUNTING FOR THE AMORTIZATION OF CALLABLE DEBT EXPENSES In this Part, we discuss Fannie Mae's accounting for the costs it incurred when it issued callable debt.(1339) We consider the Company's policy, as well as its application of that policy through "amortization end-date changes." I. INTRODUCTION As described in other sections of this Report, Fannie Mae issued both callable and noncallable debt to finance its activities. When Fannie Mae issued debt, it incurred various expenses such as commissions, legal fees, and similar costs. In addition, any difference between the face amount of the debt and the proceeds from issuance gave rise to a premium or discount on the debt issuance. Fannie Mae's accounting for these debt issue costs was inconsistent with GAAP in several respects. These accounting treatments were not driven by systems limitations, or by an effort to simplify the application of complex accounting rules to a large portfolio. On the contrary, the Company's accounting in this area required cumbersome and time-consuming operations that would have been unnecessary had it properly followed the accounting literature. Moreover, the Company did not apply its approach to the accounting for callable debt expense in a consistent fashion. On at least one occasion, for the quarter ended September 2002, it made a late on-top entry that can be explained only as an effort to offset another entry and thereby bring net interest income in line with the Company's expectations. The Company's approval of the amortization of callable debt expense appears to reflect Financial Standards' long-standing misinterpretation of the applicable accounting rules, rather than a deliberate disregard of them. The evidence concerning the periodic adjustments, and particularly the adjustment in the third quarter of 2002, however, leads to the conclusion that the Company used these adjustments to meet earnings expectations. At the very least, Leanne G. Spencer and Janet L. Pennewell played key roles in recording that adjustment. The evidence also shows that KPMG was aware of the Company's accounting policy regarding callable debt expense, and also was aware of the September 2002 on-top adjustment. It does not appear, however, that KPMG knew all of the circumstances surrounding that adjustment. - -------------- (1339) This issue was not raised by OFHEO in its 2004 Report or in its February 11, 2005 letter to Stephen B. Ashley. 353
II. BACKGROUND A. Applicable Accounting Principles The relevant accounting literature, Accounting Principles Board Opinion No. 21, Interest on Receivables and Payables, ("APB 21"), requires that premium or discount associated with debt be amortized over the life of that debt: With respect to a note that by the provisions of this section requires the imputation of interest, the difference between the present value and the face amount shall be treated as discount or premium and amortized as interest expense or income over the life of the note in such a way as to result in a constant rate of interest when applied to the amount outstanding at the beginning of any given period. This is the interest method. Other methods of amortization may be used if the results obtained are not materially different from those that would result from the interest method.(1340) When the debt is not callable, the life over which the premium or discount and issue costs should be amortized is equal to the time period between issuance and maturity. APB 21 does not distinguish between noncallable and callable debt in this regard. That is, unless and until the call option is exercised, the period over which debt issue costs should be amortized is the period between issuance and contractual maturity. If the debt is called prior to maturity, any unamortized costs would be recognized in the period in which the debt is repaid. - -------------- (1340) INTEREST ON RECEIVABLES AND PAYABLES, Accounting Principles Board Opinion No. 21, P. 15 (Accounting Principles Bd. 1971) ("APB 21"). OMNIBUS OPINION - 1967 AMORTIZATION OF DEBT DISCOUNT AND EXPENSE OR PREMIUM, APB Opinion No. 12 (Accounting Principles Bd. 1967) ("APB 12"), also addresses the interest method, stating: The objective of the interest method is to arrive at a periodic interest cost (including amortization) which will represent a level effective rate on the sum of the face amount of the debt and (plus or minus) the unamortized premium or discount and expense at the beginning of each period. APB 12 makes clear that the period over which premium (or discount) and expense (i.e., debt issue costs) are recognized is the same. 354
B. Fannie Mae's Application of the Literature Fannie Mae's practice for amortizing issue costs on callable debt was to determine the likely date on which it would call the debt and amortize the expense over the period from issuance to that date. Fannie Mae's 1994 accounting policy guidelines regarding debt issue costs stated: The estimated life is used to determine the effective cost of funds for callable debt. The estimated life is reviewed as needed based on interest rate movements and the interest accrual (for "step-up" debt) the amortization period for discount, issue cost and deferred hedging results is adjusted accordingly on a prospective basis.(1341) In other words, the amortization period for callable debt (1) was to be based on the estimated, rather than the actual, life of the debt, and (2) the estimate could be revised on an undefined "as needed" basis. The 1994 policy cites as support the consensus on EITF 86-15, and quotes the following passage from that consensus: The borrower's periodic interest cost should be determined using the interest method based on the outstanding term of the debt. In estimating the term of the debt, the borrower would consider plans, ability and intent to service the debt. Debt issue costs should be amortized over the same period used in the interest cost determination.(1342) The Company reworded its policy regarding the amortization of callable debt costs in the 2003 version of its accounting policy guidelines, although the substance - including the reference to review of the amortization period "as needed" - remained the same: Fannie Mae uses estimated life to determine the effective cost of funds for callable debt. Fannie Mae reviews the estimated life based on interest rate movements on a regular basis. Fannie Mae adjusts the amortization rate for premium/discount and hedge results on a prospective basis. Additionally, Fannie Mae amortizes debt issuance costs over the estimated life of the callable debt using the interest - -------------- (1341) Financial Accounting Policy Manual, dated Oct. 20, 1994, FMSE-SP 79123-239 at FMSE-SP 79193 (emphasis added). (1342) INCREASING-RATE DEBT, Emerging Issues Task Force of the Fin. Accounting Standards Bd., Issue No. 86-15 (May 1986) ("EITF 86-15"). 355
method. The amortization period associated with these debt issuance costs is reviewed as needed based upon interest rate movements and debt extinguishments.(1343) The 2003 policy references APB 21 as the source for this guidance. In applying its accounting policy, Fannie Mae would compute "amortization end-date changes." That is, if a change in interest rates resulted in a new expected call date for the Company's debt, when the Company made an amortization charge, the Company would amortize the relevant expense over the debt's new estimated life. As reflected in the Company's policy, the practice was to determine the estimated life of the Company's callable debt at the outset and to re-evaluate that estimate at varying intervals. In our view, the Company's policy violated GAAP. APB 21 requires that all debt issue costs be amortized "over the life of the note," consistent with the period over which premium or discount is recognized. Fannie Mae's 1994 policy guidelines, and individuals in the Controller's Office during our interviews, referred to EITF 86-15 as the basis for the Company's policy in this area. That guidance, however, does not address callable debt. EITF 86-15 addresses a fact pattern in which a company issues debt that matures in three months, but that could be extended at the issuer's option in three month increments at a higher interest rate, with a maximum possible term of five years. The literal application of APB 21 to that fact pattern would have resulted in the amortization of debt issue costs over the first three months, even though it was likely that the issuer would exercise its option to extend the maturity date. The EITF concluded that the issuer should project the term for which the debt would be outstanding and use that estimated period to determine the effective interest rate. We do not believe it was appropriate for Fannie Mae to analogize to the guidance on extendible debt in determining its accounting for callable debt. Moreover, although the 1994 policy refers to EITF 86-15, the 2003 version of the policy omits that reference in deference to APB 21. However, that change had no impact on the Company's accounting. It does not appear on the record available to us that the Company considered changing its policy during its review of the accounting policy in 2003.(1344) - ----------------- (1343) Financial Accounting Policy Guide, dated Oct. 2003, FMSE-SP 78953-9118 at FMSE-SP 79049. (1344) The treatment of issue costs on callable debt was considered even prior to the date of the revision to the policy. See Mem. from Jonathan Boyles to James Kocornik-Mina, dated May 28, 2002, FMSE-SP 40331. The memorandum states: "If, however, the expected life changes (as commonly happens with callable debt in changing rate environments) a new yield must be computed from the date the expected life changed to the new end of life." 356
Finally, even assuming that it was appropriate to analogize to the consensus reflected in EITF 86-15, that guidance would not permit the Company to change the amortization rate once the initial expected life had been chosen. This feature of Fannie Mae's policy does not have a basis in the accounting literature. Separate from the misinterpretation of the accounting literature is the reference in the Company's accounting policy (since at least 1994) to changes in the amortization end-dates "as needed." The evidence (discussed below) shows that the Company's application of its policy resulted in inconsistent revisions to the amortization end-dates. The Company performed the end-date changes only when it was logistically convenient or otherwise deemed necessary by senior management in the Controller's Office. The Company has not offered a rationale in any accounting standard for this feature of its policy. In sum, the record shows that Fannie Mae's policy regarding the amortization of the debt issue costs associated with callable debt was based on an incorrect interpretation of the accounting rules. The Company's policy, moreover, appears to have invited inconsistent application of the amortization end-date change process. Particularly in view of the latter, we reviewed the history of the Company's application of that process. C. Amortization End-Date Changes By all accounts, as reflected in both interviews and documents, Fannie Mae did not apply its policy regarding the calculation of amortization end-date changes for callable debt with regularity. The officials responsible for the application of the policy - including Janet L. Pennewell and individuals in Financial Reporting responsible for debt accounting - acknowledged that the end-date changes were conducted only episodically.(1345) When an amortization end-date change was undertaken, the process was initiated by a request from the Controller's Office to Portfolio Management for an update on the expected call dates of the Company's debt. Portfolio Management would then determine the expected call dates and prepare an estimate of the income or expense associated with the end-date changes. The Debt Accounting group in the Controller's Office would incorporate the new end-dates into the Company's STAR (and, later, its iSTAR) debt accounting system in a test environment and determine the actual income or expense associated with the end-date changes. The results of the test runs in the Controller's Office were then compared with the estimates received from Portfolio Management, and any discrepancies were evaluated. Senior personnel in the Controller's - ----------- (1345) We did not discuss this issue with Leanne G. Spencer prior to her refusal to participate in further interviews. 357
Office would then decide whether or not to adjust the amortization of issue costs by moving the test into production.(1346) There are indications that the Company was very sensitive even to seemingly minor differences between Portfolio Management's projections and the results of the tests conducted in Financial Reporting. In 2001, Cheryl DeFlorimonte reported a $1.2 million difference between the Portfolio Management estimate and the results of the Financial Reporting calculations. She told her supervisors, "We have spent all day trying to ascertain the reason for the unusually large variance of $1.2 million, to no avail."(1347) She then explained the disruption this had on other activities: "We have been unable to use the STAR production database for the last four work days. Consequently, a back log of new deals, rate resets, and calls have accumulated. We need to make a decision ASAP as to whether we should rely on the results from the test done in STAR Acceptance and proceed with the update to Production, or whether we should again postpone the update due to the $1.2 million variance between STAR and Portfolio Management projections."(1348) Pennewell sent an e-mail to DeFlorimonte and others later that afternoon stating: "just [sic] so you know, I approved this despite the difference from projection. The alternative was deferring and trying again next month, which did not seem desirable since I view the only downside as the risk that we record a little too much expense this year."(1349) The Portfolio Management personnel responsible for calculating the end-date changes did not ordinarily produce a report on this subject, or calculate the income effect of the end-date changes. Rather, that group was able to extract the relevant estimates from its broader effort to manage the Company's portfolios and provide earnings estimates. Supporting the financial reporting function, therefore, required an extra undertaking on the part of Portfolio Strategy. - ---------------- (1346) Documents reflect the role of Financial Reporting on the decision of whether to book an adjustment and in what amount. An e-mail dated January 26, 2000 from Gary Robinson to Ted Belay, for example, concerns the procedure for implementing an amortization end-date change in February. After describing the process (including an initial run in STAR's test mode, and then in its production mode), it states, "Gary will present the results to Financial Reporting's management on the morning of Thursday, Feb. 17th. The results of this anaylyst [sic] may require that we eliminate certain months for amortization end date changes in the source file." E-mail from G. Robinson to T. Belay, dated Jan. 26, 2000, Zantaz document 738103. (1347) E-mail from Cheryl DeFlorimonte to Richard Stawarz and Janet L. Pennewell, et al., dated Mar. 19, 2001, FMSE-IR 684282. (1348) Id. (1349) E-mail from J. Pennewell to C. DeFlorimonte, et al., dated Mar. 19, 2001, FMSE-IR 684281. 358
The Controller's Office, too, had no system for processing an end-date change efficiently, or for calculating the effect on the Company's debt-cost amortization into a normal close of the Company's books at the end of a reporting period. An amortization end-date change, we have been told, required up to four days' effort during which the STAR debt accounting system would be unavailable to process other activity. Accordingly, the step could not be implemented either at the close of the quarter or when other work was deemed more important. For example, in August 2004 (after the Company had converted to iSTAR), the Controller's Office decided not to process an end-date change or to book the resulting change in the amortization of callable debt expense; Financial Reporting "pass[ed] on Aug[ust] implementation due to resource constraints and materiality." The implementation of the end-date change was scheduled for the following November.(1350) Other documents convey a similar message. In November 2003, for example, Financial Reporting notified Portfolio Management of an anticipated end-date change calculation. The notification states: "After several discussions with members from the Financial Reporting group, we have decided to proceed with the amortization end date changes using the November file. During the month of December, we will be tasked with providing numerous estimates, including the impact from debt buyback/derivative termination activity. We prefer not to complicate matters with amortization end date estimates the last month of the quarter/year."(1351) In 2003, Financial Reporting began a process to implement the end-date changes in a more systematic fashion. According to instructions issued by Pennewell, the most significant change in the Company's procedures that resulted from this effort was that Portfolio Management would provide Financial Reporting with a new estimate of end-date changes and the income impact of those changes on a regular basis. At that point, Financial Reporting would "determine whether the impact is material...." Then, "if material," the end-date changes would be implemented in a test mode, evaluated against the Portfolio Management estimates, "and any differences researched. Once we're comfortable with [the] results," the changes would be recorded in the debt accounting system.(1352) - ---------------- (1350) E-mail from Rick Swick to William Quinn, dated Aug. 30, 2004, FMSE-E 1785523. Another e-mail in the same chain, dated August 23, 2004, from DeFlorimonte to Pennewell and Mary Lewers explained that the end-date change was impossible due to deployment of iSTAR Release 2.1. The e-mail stated that the end-date change would be implemented in November: "Our policy is to perform end date changes during the first or second month of the quarter. However, based on current resources, processing end date changes and simultaneously producing quarterly reporting during the month of October is not recommended." Id. (1351) E-mail from C. DeFlorimonte to R. Swick, dated Nov. 6, 2003, FMSE-E 1821333. (1352) E-mail from J. Pennewell to R. Swick and W. Quinn, dated Aug. 11, 2003, FMSE-E 265996-97. 359
A memorandum incorporating these instructions was prepared in October 2004.(1353) It called on Portfolio Management to provide an end-date change analysis for Financial Reporting on a quarterly basis. The updates would be scheduled during the first month of the quarter but, "[i]n the event a conflict with another significant business activity renders an update impracticable during the month of the scheduled update," then (1) if the update had been scheduled for the first month of the quarter, it would be rescheduled for the second month, and (2) if the update was scheduled for the second month, the situation would be presented to senior management for their review; "[d]epending on the materiality of the estimated impact on income/expense that would result from end date changes during that month of the scheduled update, senior management will make a decision whether to defer processing the amortization end date changes . . . or process the update in spite of the conflicting business activity."(1354) Apparently, in light of the issuance of the OFHEO Report in September 2004, there was little, if any, further consideration of these issues. D. The September 2002 End-Date Adjustment The Company recorded $35.3 million of additional amortization expense through an on-top entry for the third quarter of 2002. Contemporaneous documentation associates this entry with a separate $43 million entry to recognize additional interest income as a result of a premium/discount amortization adjustment. Had the Company not recognized the $35.3 million adjustment, its net interest income would have increased by 3.7 percent over net interest income for the second quarter of 2002. The amortization end-date change reduced the change in net interest income from the second quarter to 2.3 percent.(1355) According to one member of Financial Reporting, who recorded the events leading up to that adjustment: Based on discussions with KPMG, Tim and Frank, today Janet requested that we book an on-top estimated amortization expense entry for the September reporting period. Because of the decreasing rate environment, a large income entry for purchase discount amortization has been generated for the September reporting period. These - -------------- (1353) According to DeFlorimonte, Financial Standards delayed preparation of a formal procedure to implement Pennewell's instructions pending the conversion of the STAR system to the iSTAR system. (1354) Draft mem. from C. DeFlorimonte to Distribution, dated Oct. 4, 2004, Zantaz document 202983. (1355) Comments attributed to Controller's Office management in a survey conducted in late 2002 by Marsh Risk Finance indicate that a change in net interest income of more than three percent "would be viewed as significant." 360
decreasing rates have also necessitated that we perform an amortization end date update on our callable debt portfolio. This end date change procedure will increase expense, (pulling up the end dates) and has been scheduled to be implemented for the October reporting period. However, in order to match our mortgages with our liabilities, the 35.3 million estimated expense impact from the amortization end date changes will be recorded as an on top entry on September. This on-top entry will be reversed during the October reporting period and the actual amortization end date chances [sic] will be processed. Management staff attending this discussion included [Pennewell, Richard Stawarz, Mary Lewers, Gary Robinson, Jennifer Wall and DeFlorimonte].(1356) Journal entries dated October 10, 2002, with an effective date of September 28, 2002, reflect the $35.3 million entry.(1357) The suggestion in the e-mail is that the decision to book this adjustment was prompted by a desire to offset most of a discount amortization adjustment. That suggestion, although not based on the first-hand knowledge of the e-mail's author, appears credible for several reasons. First, as the e-mail notes, the amortization end-date change was not scheduled until the following quarter. Therefore, Financial Reporting did not have the opportunity to process its own adjustment in the STAR system, but rather took the unusual step of basing the end-date change entry on a Portfolio Management estimate. Then, in the fourth quarter, the Company reversed the $35.3 million entry and processed the actual end-date change. The decision to book the end-date change ahead of schedule and based on an estimate gives credence to the e-mail's implication that the Company used the entry to offset another unusual entry that period. Second, other evidence shows that the Company was concerned about the volatility associated with these adjustments, even on a monthly basis. In the fourth quarter of 2002, the Company calculated that the amortization end-date change resulted in additional interest expense of $31.5 million, $24.1 million, and $19.3 million for October, November, and December 2002, respectively.(1358) The Company booked the - ----------------- (1356) E-mail from C. DeFlorimonte to C. DeFlorimonte, dated Oct. 10, 2002, FMSE-IR 437919. The e-mail was prepared by DeFlorimonte and sent to herself. She stated that she prepared the e-mail because of the unusual nature of the entries she was asked to book. We have seen no corroboration of DeFlorimonte's suggestion that Frank Raines was involved in the decision to book this entry. (1357) Journal Entry, dated Oct 10, 2002, FMSE-IR 437922. (1358) E-mail from C. DeFlorimonte to J. Pennewell and W. Quinn, dated Nov. 4, 2002, FMSE-SP 10804. 361
$31.5 million in October, at the same time that it reversed the $35.3 million entry reversing the entry from the second quarter.(1359) These October entries therefore largely offset each other. The Company also recorded an additional adjustment of $14.5 million in October, reflecting one-third of the $43.4 million expense associated with the November and December estimates. The $14.5 million entry was reversed by a $7.5 million entry in November and a $7 million entry in December.(1360) This approach was recorded in Stawarz's handwritten note as follows: "Recorded a $14.5 million OOC [out-of-cycle] expense in Oct to smooth remaining expense from the debt end date over Nov and Dec. Reverse $7.5 mil of above in Nov. and remaining $7.0 mil in Dec."(1361) An entry on the Company's December 2002 Closing Notes, which Stawarz also prepared, states: "Reversed $7.0 million end amortization smoothing entry (to spread costs evenly over the quarter) and credited to clean debt expense."(1362) Although Stawarz acknowledged that he prepared the handwritten notes, he professed to have no recollection of the $35.5 million adjustment in the third quarter of 2002, or the other entries in the fourth quarter. Third, it appears that the Company did not fully disclose the circumstances of this adjustment in this matter to its outside auditors. A KPMG audit workpaper dated October 25, 2002, captioned "On-top entry related to the amortization of debt issuance costs" discusses the $35.3 million entry. After discussing the Company's general approach to the amortization of callable debt issue costs, the memorandum states: "Generally, the change in amortization period is run through the STAR system and is reflected in the general ledger at period end. However, in the third quarter, the on-top entry was booked due to timing of when the rate path information was available and the completion of the analysis."(1363) DeFlorimonte confirmed that this was not an accurate - -------------- (1359) Undated October 2002 Closing Notes, FMSE-SP 10807. (1360) Journal entries reflecting these adjustments are attached to an e-mail from Pennewell to DeFlorimonte dated November 4, 2002. E-mail from J. Pennewell to C. DeFlorimonte and W. Quinn, et al., dated Nov. 4, 2002, Zantaz document 1974463. In addition, DeFlorimonte prepared a handwritten chart showing the various entries based on instructions she received. Undated handwritten notes, FMSE-IR 437928. (1361) Id. (1362) December 2002 Closing Notes, dated Dec. 2002, FMSE-SP 10860. Handwriting on another copy of the Pennewell e-mail cited supra note 1360 states: "43.4 /3 = 14.5. On top entry to be booked over the next three months per Dick/Gary." E-mail from C. DeFlorimonte to J. Pennewell and W. Quinn, et al., dated Nov. 4, 2002, Zantaz document 1449228. (1363) Mem. from Jennifer Eng to Third Quarter Workpapers - Quarterly Review Binder, dated Oct. 25, 2002. 362
statement of what had occurred. It should be noted, however, that KPMG's workpaper attributes this information to Portfolio personnel; it does not appear that KPMG discussed the matter with anyone in the Controller's Office.(1364) Fourth, there are other indications that the Company used its amortization of issue costs to offset the impact of entries in other areas. At about the same time as the $35.3 million adjustment was under consideration, an issue arose with respect to a separate adjustment of nearly $19 million associated with the amortization of the expense associated with exercised swaptions. An e-mail from Pennewell states: "Leanne says book the whole $19 million in September. We'll likely offset with a pda [premium/discount amortization] on-top. Both should be OOC."(1365) The impact of the amortization end-date changes on the Company's financial statements is reflected in e-mails exchanged at about this time. In her November 4, 2002 e-mail to Pennewell, DeFlorimonte noted that the increase in expense as a result of the September/October end-date change was $74.9 million.(1366) This is consistent with the Company's journal entries regarding the on-top adjustments (that is, the $31.5 million booked in October (after reversal of the September entry) and the $43.4 million increase in November and December). DeFlorimonte projected an increase in expense of $72.9 million in 2003 and a decrease in expense of $11.9 million in 2004. Pennewell responded that "we may still want to spend some time trying to figure out what is making the 2003 and 2004 numbers so different."(1367) Finally, it appears that the manner in which the end-date change was executed in the third quarter of 2002 was inconsistent with how end-date changes were processed in other periods. For example, in May 2002, the Company did not process - ----------------- (1364) In an interview, Mark Serock, the KPMG audit partner at the time, stated that he had discussed the $43 million premium/discount amortization adjustment with Pennewell, Spencer and Timothy Howard. After Serock concluded his discussion with them, Howard and Spencer approached KPMG to tell them of the amortization end-date adjustment. (1365) E-mail from J. Pennewell to M. Lewers, C. DeFlorimonte, R. Stawarz, and Shaun Ross, dated Nov. 4, 2002, FMSE-SP 10799. No one we spoke to recalled the e-mail or the $19 million entry. Pennewell speculated that DeFlorimonte might have been concerned about the need to book an unexpected $19 million expense, and that the e-mail was Pennewell's effort to resolve the situation. However, when asked whether she was concerned about a $19 million adjustment at that time, DeFlorimonte said she was not. (1366) E-mail from C. DeFlorimonte to J. Pennewell and W. Quinn, dated Nov. 4, 2002, FMSE-SP 10804. (1367) E-mail from J. Pennewell to C. DeFlorimonte, dated Nov. 4, 2002, FMSE-SP 10801. 363
end-date changes for any debt that it anticipated calling in June, July, or August.(1368) However, in September 2002, a significant portion of the $35.3 million adjustment resulted from accelerating amortization of issue costs on debt called in October, November, and December into the third quarter result.(1369) III. FINDINGS The Company's accounting policy regarding debt issue costs was inconsistent with GAAP insofar as it (1) treated callable debt and noncallable debt differently at the outset, and (2) sanctioned end-date amortization charges "as needed." Although one can question both the reference to inapplicable accounting literature in the pre-2003 policy, and the failure to change the policy when the appropriate literature was cited in the 2003 version of the policy, we have seen nothing to suggest that these deviations from GAAP were other than inadvertent misinterpretations of the literature. The Company's application of its policies and procedures regarding the amortization of callable debt issue costs was inconsistent and, on at least one occasion in the third quarter of 2002, deployed in a fashion that was a deviation from the Company's past practice for the purpose of offsetting another entry and avoiding volatility in a key financial measure. Moreover, although it appears that Fannie Mae's outside auditor was aware of the Company's accounting policy as reflected in its policy guidelines, and at least as of 2002 was aware of the adjustment in September 2002, the evidence suggests that KPMG may not have been fully informed as to the nature of, or reasons for, this adjustment.(1370) - ------------------- (1368) E-mail from C. DeFlorimonte to R. Stawarz, dated May 30, 2002, Zantaz document 1974487. (1369) The record contains a chart reflecting the amortization end-date changes and cost of the debt included in the calculation. FMSE-IR 437950-76. An incomplete copy of the chart was also included in KPMG's workpapers. (1370) In April 2005, the Company's Office of Auditing ("Internal Audit") issued an audit report on debt funding. See Office of Auditing, Debt Funding Audit, dated Apr. 29, 2005, Zantaz document 1450686. Its findings are relevant to the issues under consideration. In summary, Internal Audit concluded: "Policies and modelling functions for changing amortization end dates on callable debt lack Financial Standards' guidance and approval. Also, controls need improvement to ensure consistent implementation, effective data validation checks, and adequate documentation." Internal Audit also noted that "[a]ccounting policy guidance needs to be expanded to better address step rate debt, benchmark repurchase agreements and debt issuance costs." 364
PART N: MINORITY LENDING INITIATIVE In this Part, we report on accounting issues that arose in the course of Fannie Mae's Minority Lending Initiative ("MLI"). I. INTRODUCTION Fannie Mae implemented the MLI program in 2002 to increase the Company's financial support for mortgages to African-American homeowners. The initiative was seen as an important component of the Company's overall mission and, by some, as a means of securing loans that would meet the guidelines set by HUD. Concerns regarding the initiative itself were raised to us by a Company employee in response to an e- mail broadcast to all employees by Ann M. Korologos, Chair of the SRC of the Board of Directors.(1371) The employee was concerned that the Company appeared to be paying an excessive price for loans that were underperforming and that the MLI program might have been devised to meet corporate targets. As noted above, we have concluded that the plan was envisaged by some as a means of achieving corporate targets - specifically, the Company's internal mission goals as well as HUD goals - but we saw nothing in the record of the program to indicate that it had an improper purpose. In the course of our review of the record, however, we identified one issue concerning the accounting for payments in 2003 that the Company made to Resource Bancshares Mortgage Group, Inc. ("RBMG"), the mortgage lender that sold Fannie Mae a majority of the loans acquired under the MLI program. The Company capitalized these payments as part of the cost of the acquired loans. Given the nature of the payments, however, we believe the Company should have expensed them in 2003. The aggregate amount of the payments we have been able to identify was approximately $35.5 million. II. BACKGROUND A. The MLI Program The MLI program began early in 2002 when the Company set a goal to acquire approximately 10,000 purchase- money mortgages of African-American homeowners.(1372) The program involved a new product that would be launched through a lending partner involved in the African-American community.(1373) - ------------ (1371) See infra Chapter XI, Part A. (1372) Minority Lending Initiative Review Report, dated Oct. 10, 2003, FMSE-KD 15557-633, at FMSE-KD 15560 (hereafter "MLI Report"); MLI 1 Update, dated Sept. 2, 2003, FMSE-IR 670769-80, at FMSE-IR 670771 (hereinafter "MLI 1 Update"). (1373) MLI 1 Update at FMSE-IR 670770. 365
Fannie Mae identified RBMG as one of its lending partners in the MLI program.(1374) The first phase of the initiative involved an agreement between Fannie Mae and RBMG to sell loans that satisfied the MLI criteria to Fannie Mae. In 2002, RBMG began to offer mortgages to its customers under the terms of the program. RBMG in turn sold these loans to Fannie Mae. Under the terms of the agreement, Fannie Mae paid a premium for the loans - essentially paying a market price for prime loans, but receiving loans with credit characteristics more closely associated with subprime loans.(1375) The MLI program was troubled from the outset in many respects. The most significant issue stemmed from the fact that the loans originated under the MLI program had loan to value ("LTV") ratios of ninety-seven percent or greater.(1376) Fannie Mae understood that it was required under its Charter to have some form of credit protection for the amount of the loans in excess of eighty percent LTV.(1377) Accordingly, Credit Policy arranged for a mortgage insurer to cover the loans under an existing policy.(1378) However, because the loans to be insured had lower credit scores than the loans the policy was originally written to cover, and because the Company only planned to hold the loans for a short time before selling them, Credit Policy provided verbal assurances that the insurer could cancel the coverage after a one-year period.(1379) As Fannie Mae continued to acquire loans from RBMG, however, it was unable to find a buyer for the MLI loans at a price that would not generate a substantial loss. In February 2003, the Company made the decision to classify loans it had acquired to that point, as well as new loans it would acquire under the program, as held for investment.(1380) - ------------ (1374) MLI Report at FMSE-KD 15661. Fannie Mae also had arrangements to acquire MLI loans originated by two other lenders. MLI 1 Update at FMSE-IR 670773. Neither of those lenders appears to have been a party to the arrangement discussed herein. (1375) See E- mail from Renee Schultz to Andrew McCormick, dated Aug. 28, 2003, FMSE-E 1116534-35, at FMSE 1116535 ("We are offering a market rate for sub-prime credit quality, and paying up 110 bps in price for it!!"). (1376) See MLI 1 Update at FMSE-IR 670774. (1377) MLI Report at FMSE-KD 15564. Fannie Mae's initial plan was to accumulate loans originated under the program and sell them. Id. at FMSE-KD 15558. Accordingly, the loans should have been classified as held for sale on the acquisition date. (1378) Id. at FMSE-KD 15565. (1379) Id. at FMSE-KD 15564-65. (1380) Id. at FMSE-KD 15563. 366
The mortgage insurer canceled Fannie Mae's coverage under the policy after the one-year period.(1381) Accordingly, loans in the MLI pipeline that the Company had committed to purchase, as well as the loans that the Company was holding pending possible sale, would not have been in compliance with the Charter. After receiving notice that the mortgage insurer was exercising its right to cancel coverage of the MLI loans, the Company considered several options to address the Charter compliance issue relating to loans that Fannie Mae had committed to acquire.(1382) These strategies revolved primarily around selling some or all of the loans at a loss, or acquiring mortgage insurance at high premiums.(1383) As to loans "in the pipeline" that the Company was obligated to acquire, the Company structured another arrangement. Under this arrangement, another company, Self Help, agreed to purchase the loans that Fannie Mae was committed to purchase from RBMG.(1384) Self Help would then sell the loans to Fannie Mae with a six- month recourse period (that is, Self Help would repurchase any loans that defaulted within six months of the sale). According to Fannie Mae, this credit enhancement to the loans resolved the Charter compliance problem.(1385) In recognition of the fact that the loans were originated at favorable interest rates given the borrowers' credit ratings, however, Self Help acquired the loans from RBMG at a discount to the price Fannie Mae was committed to pay. In order to make RBMG whole, Fannie Mae made a series of payments to RBMG equal to the difference between the price at which Self Help acquired the loans and the Company's commitment price. The MLI program and, in particular, the payments to RBMG reflecting the subsidy that Fannie Mae paid on the loans, resulted in a substantial economic loss for the Company.(1386) In September 2003, the decision was made to cancel the program.(1387) - ------------- (1381) See id. at FMSE-KD 15568. (1382) MLI 1 Update at FMSE-IR 670773. (1383) Id. (1384) E- mail from Rebecca Mattoni to Janet L. Pennewell, dated June 21, 2004, FMSE 543587-90. (1385) Id. at FMSE 543587. The original proposal was for RBMG to execute a six- month repurchase agreement, but the Company concluded that such an agreement was not feasible. MLI 1 Update at FMSE-IR 670773. A transaction with Self Help was considered subsequently. Undated MLI Loan Pipeline: Deal Alternatives, FMSE-E 170959-60, at FMSE-E 170959. (1386) MLI 1 Update at FMSE-IR 670773-74. 367
Based on the information available to us, we do not question the rationale or motives behind the MLI program. We understand that the management issues raised by the MLI experience were examined within the Company in late 2003 and certain recommendations - including the need for more coordination and cooperation - were proposed.(1388) B. Accounting for the Payments to RBMG In late 2003, Financial Standards was asked to consider how the Company should record the payments to RBMG. At one point, Financial Standards concluded that the payments should be recognized as an expense. Financial Standards then reversed that decision and determined that the payments could be capitalized as part of the cost of the loans acquired from Self Help. E- mail exchanges on the subject suggest that this change of views occurred following discussions with KPMG, but the exchange does not permit a definitive conclusion as to the nature of KPMG's advice.(1389) Jonathan Boyles recalled that KPMG approved the capitalization of the payments. According to internal documents, the Company made payments to RBMG in October, November, and December 2003 totaling $35,536,254,42.(1390) We have been unable to determine the rationale for the ultimate decision that the payments should be - ----------- (1387) MLI Report at FMSE-KD 15564. (1388) See id. at FMSE-KD 15570-71. (1389) Consultation with KPMG on this issue is reflected in two e- mails. In an e-mail exchange between Jonathan Boyles and Paul Salfi, Boyles reported that "kpmg [sic] has said that if we buy these and put them into hfi [held for investment] then they would go on at cost. Only a pair off would be an immediate hit. They checked with their national office." E- mail from J. Boyles to P. Salfi, dated Sept. 24, 2003, FMSE-E 171246. This e-mail appears to reflect a view that the payments to RBMG should have been expensed, resulting in an "immediate hit." In a second e- mail, also dated September 24, 2003, from Leanne G. Spencer to Louis Hoyes and Timothy Howard, states: "Jonathan took KPMG through the transaction w/ Self- help [sic]. They opined, after seeing the transaction, that capitalizing the loss from the pairoff was appropriate." E- mail from L. Spencer to L. Hoyes and T. Howard, dated Sept. 24, 2003, FMSE-E 1155661. Boyles, who had been copied on the earlier e- mail, responded that "[t]his treatment is consistent with other pair off transactions we do when we pair off and buy like assets." Id. (1390) E- mail from R. Schultz to Jeffrey Juliane, dated Nov. 21, 2003, Zantaz document 1011100. 368
capitalized. Had Fannie Mae recognized the $35.5 million of additional expense, it would have reduced earnings for 2003 by $.02 per share. III. FINDINGS REGARDING THE MLI PROGRAM In our view, the Company's decision to capitalize the payments to RBMG was not consistent with GAAP. FAS 91 provides that only those fees paid to the party from whom loans are acquired (or amounts received from that party) may be capitalized as part of the purchase price.(1391) In this case, the relevant payments did not go to Self Help. We are unable to determine the basis for the decision to capitalize the expense because we are unable to reach a conclusion as to the content of KPMG's advice. The only documentary sources of that advice - the two e- mails - are ambiguous. Boyles' recollection is that KPMG approved the decision to capitalize the expense and his e- mail indicates that, in his view, the payments to RBMG were similar to a pairoff fee, which the Company historically had treated as a cost of the acquired loan. PART O: ACCOUNTING FOR REALIGNMENTS AND THE SECURITY MASTER PROJECT I. INTRODUCTION This section addresses the Company's accounting for differences generated in the process of identifying and correcting errors and mismatches between its amortization database and loan and securities databases. The process of adjusting the amortization database to match the loan and securities databases was known as a "realignment." We focused on management's accounting for the impact of realignments initially because of the concerns OFHEO raised in the OFHEO Report. OFHEO reported, in substance, that management's accounting for realignment impacts from 2002 through 2004 was not consistent with GAAP and was designed to manage earnings. Specifically, OFHEO took issue with management's practice of including realignment differences in its estimation of catch- up, alleging that "the catch-up results were changed by including adjustments that were incorrect or inconsistently applied."(1392) We concur with OFHEO's assertion that management did not account for realignment impacts properly under GAAP. In certain circumstances, realignments represented corrections of errors, and as such, management should have analyzed and accounted for their impact in accordance with Accounting Principle Board Opinion No. - ------------- (1391) FAS 91 P. 15. (1392) OFHEO Report at 58. 369
20, Accounting Changes ("APB 20"),(1393) effective for fiscal years beginning after July 31, 1971. Management failed to do so. Additionally, management introduced data into iPDI, the amortization system of record, through realignments that it knew, or should have known, was incorrect. Moreover, management's accounting for realignment impacts identified between 2000 and 2004 was not consistent. For the most part, management deferred the recognition of these differences by recording them to balance sheet accounts and amortizing them into income over multiple years. On other occasions, in addition to deferring the recognition of these differences and amortizing them into income over time, management included the cumulative deferred realignment amounts and estimates of future realignments in its calculation of catch-up; and still on other occasions, management recognized the realignment impacts into income in the period they were identified. Employees with responsibility for the realignments whom we interviewed could provide neither a credible reason why management failed to apply APB 20 to realignment impacts nor any coherent rationale for the inconsistent accounting treatment of such impacts. At a minimum, this demonstrates that the Company did not have adequate accounting policies or procedures to ensure compliance with GAAP in this area. Furthermore, the decision to capitalize and defer realignment impacts over time smoothed out the errors' impact on income in any one period. With respect to the inclusion of realignments and estimates of realignments in the catch- up calculation in 2003, we conclude that management was motivated, in part, to avoid recording or to reduce the amount of the catch-up adjustment required under the Company's amortization policy. II. BACKGROUND Fannie Mae's current system of record for premium and discount amortization is iPDI, which was historically known as PDS and PDI (collectively, "iPDI"). iPDI calculates and records premium and discount amortization to the general ledger by applying amortization factors to original premium and discount amounts. The original premium and discount amounts that iPDI relies upon are received monthly, for new acquisitions, from the STAMPS (for MBS(1394)) and Delivery (for whole loans) systems. The original premium and discount data from STAMPS and Delivery is also provided to the STATS (for MBS) and LASER (for whole loans) databases which track the Company's portfolios of MBS and whole loans at a security or loan level. After the original premium and discount data is provided to iPDI, STATS, and LASER, changes to the classification of the loan or security, or amount of original - ------------ (1393) ACCOUNTING CHANGES, Opinion No. 20 (Accounting Principles Bd. 1971). (1394) MBS data includes Real Estate Mortgage Investment Conduits ("REMICs") data. 370
premium and discount, may be made to the data in STATS and LASER.(1395) Changes to the data in STATS and LASER included adjustments to correct securities misclassifications, make post-purchase price adjustments and track liquidations. Because there is no automatic feed from STATS and LASER to iPDI, Fannie Mae periodically conducted manual "realignments" so that the original premium and discount balances in iPDI matched those in STATS ("STATS realignment") and LASER ("LASER realignment").(1396) These realignments caused changes to the original premium and discount balances in iPDI, which resulted in corresponding changes to the accumulated amortization balances in iPDI and ultimately to the balance sheet. Rather than recognize the offsetting amount to earnings, management most often deferred the recognition of these amounts. We review in this section management's various methods for dealing with the impact of these realignment adjustments, none of which was consistent with GAAP. III. FINDINGS REGARDING THE ACCOUNTING TREATMENT OF REALIGNMENT IMPACTS A. Management's Deferral of Realignment Impacts Management's accounting for realignment impacts was, for the most part, inappropriate under GAAP because it allowed the Company to defer over multiple years the recognition of income or expense resulting from errors in amortization. As such, APB 20 governs the treatment of realignment impacts. Because the data necessary to properly classify loans and securities was available at the time of acquisition, we consider realignment differences to arise out of a "misuse of facts." Accordingly, based on the guidance in APB 20, we conclude that the impact of realignments is a correction of an error. APB 20 requires management to determine the impact of such an error on prior periods, as well as the impact of correcting the cumulative error on the current period. If any of these impacts are determined to be material, then the Company must report the error as a prior period adjustment and disclose the error in the current period financial statements.(1397) If all of these impacts are determined to be immaterial, then the - ------------ (1395) Changes to classification are significant because the classification of loans and securities is a key factor in the determination of how quickly the associated premium and discount should be amortized. (1396) We do note that, in one instance, the manner in which Fannie Mae treated collateral used in dollar roll transactions created errors in STATS. We also note that in this case management ultimately "transferred" these errors to iPDI by updating the balances in iPDI with those in STATS rather than identifying the reasons for the differences between STATS and iPDI. (1397) APB 20 P.P. 36-38. 371
Company can simply record the cumulative error in the current period without any disclosure. SEC Staff Accounting Bulletin No. 99 - Materiality ("SAB 99"), issued and effective on August 12, 1999, is the pronouncement that addresses materiality and should be used in determining the accounting treatment in these situations. Some members of the Controller's Office - Janet L. Pennewell, Mary Lewers, and Richard Stawarz - stated that the rationale for management's accounting for realignments was that the data in iPDI was better controlled and hence more reliable than the data in STATS and LASER; however, we do not find this explanation to be persuasive. If the data in iPDI was in fact better controlled and more reliable than the data in STATS and LASER, the Company should not have updated the information in iPDI, which was used in determining the amortization.(1398) 1. Deferral of STATS Realignment Impacts In 2000, and roughly once a year thereafter, the Company performed a realignment analysis, which compared iPDI to STATS and recorded the realignment impact to the balance sheet.(1399) In 2000, the realignment impact was approximately $81.9 million of additional expense. Rather than analyze this impact under APB 20, management deferred the realignment impact in two balance sheet accounts. Approximately $77.1 million of the $81.9 million adjustment was recorded to an account that typically contained amortization on-top amounts that had not yet been applied to certain accumulated amortization accounts (i.e., 1201-07).(1400) Documents and interviews suggest that Jeffrey Juliane may have considered this account to be similar to a reserve and that recording an amount to this account was viewed as the equivalent of recording it to the income statement.(1401) This account typically contained amortization on top - ---------------- (1398) Indeed, Stawarz conceded that not processing a realignment would have been an appropriate option. Stawarz also stated that Fannie Mae intended, ultimately, to move the amounts recorded to the deferred pool account into income once the data issues in STATS had been resolved. We have seen no evidence that amounts recorded to the 1201-01 account were ever reversed and recorded to the appropriate accounts. (1399) See, e.g., Fannie Mae STATS Realignment Project Analysis - Summary, Period Ended October 31, 2000, dated Nov. 30, 2000, FMSE-SP 5009-12; Journal Entry, dated Nov. 30, 2000, FMSE-SP 11624; Fannie Mae STATS to PDI Realignment, Period Ended May 31, 2003, dated May 15, 2003, FMSE 193818-20; Journal Entries, dated May 20, 2003, FNMSEC 4375, 4379, and 4382. (1400) Fannie Mae STATS Realignment Project Analysis - Summary, Period Ended October 31, 2000, dated Nov. 30, 2000, FMSE-SP 5009-12; Journal Entry, dated Nov. 30, 2000, FMSE-SP 11624. (1401) E- mail from Joyce Philip to Jeffrey Juliane, dated July 1, 2003, FMSE-E 2104712. Patricia Wells stated that amounts recorded to the 1201-07 account already would 372
amounts recorded but not yet applied to certain accumulated amortization accounts.(1402) Inexplicably, the remaining $4.8 million was recorded to another balance sheet account.(1403) In August 2001, management identified the differences between STATS and iPDI, but did not record any adjustments attributable to the differences identified in the analysis.(1404) Pennewell stated that she postponed processing the realignment to avoid causing iPDI to reclassify original premium and discount into different acquisition years. According to Pennewell, such a reclassification was not appropriate because acquisition years generally should not change.(1405) Although the data anomalies were never explained or resolved, a year later, management went forward with the realignment in order that the Company's systems not continue to be misaligned.(1406) Pennewell's concerns in 2001 appear to have been reasonable. However, given her explanation, management's decision to record the results of the realignment analysis in 2002 created errors in iPDI that ultimately resulted in errors in amortization. From 2002 to 2003, management's practice was to record the realignment impact to the balance sheet (account 1201-01), and amortize these amounts into income over time.(1407) - ------------ have been charged to the income statement through on-top entries. This explanation, coupled with the mechanics of the realignment entry, in this instance, was consistent with viewing the account as a reserve. In other words, since the balance in the account was established by on-top entries that impacted the income statement, any entries clearing amounts from this account could have been considered already charged to the income statement. (1402) In effect, the Company's treatment of the realignment difference as a reduction of the balance results in using the "reserve" for a purpose other than the one for which it was established. (1403) This account, 1230-90, the "PDI Conversion Clearing" account, contained deferred amortization amounts related to a previous system conversion. (1404) Fannie Mae STATS iPDI Realignment Summary Analysis, Period Ended August 31, 2001, dated Sept. 13, 2002, FMSE 193762-73. (1405) Mem. from Janet L. Pennewell to File, dated Nov. 15, 2004, Zantaz Document 1000371, at 2. (1406) Mem. from J. Pennewell to File, dated Nov. 15, 2004, Zantaz Document 1000371, at 2. (1407) Account 1201-01, also referred to as the deferred pool bucket by Fannie Mae, is an asset account subject to amortization. Amounts recorded to this account are treated as new acquisitions and amortized over a proxy life consistent with a fixed-rate 373
Stawarz stated that the decisions on how to treat realignment impacts were made in multiple meetings at which he, Roger Barnes, Lewers, Pennewell, and Leanne G. Spencer together concluded that the deferral of realignment impacts was appropriate.(1408) Beyond this, employees' recollections were vague as to who had the ultimate responsibility for the decision to defer realignment impacts. We find that the absence of a clear chain of authority or approval for accounting decisions that could have as much as an $81 million impact (i.e., in 2000) on the Company's financial statements was a failure in the Company's internal controls. As discussed above, we conclude that STATS realignments, which were undertaken in large part to correct misclassified securities in iPDI, should have been accounted for as a correction of an error under APB 20.(1409) - ------------ conventional mortgage. Mortgage Portfolio Deferred Price Adjustments Policy and Operational Procedures, Realignment Policy, dated Apr. 30, 2004, FMSE 193585-86; Mortgage Portfolio Deferred Price Adjustments Policy and Operational Procedures P. 4.2.4, dated Aug. 31, 2004, FNMSEC 4169 (these documents, as examples of documents relevant to the discussion in Chapter VI, Part O, can be found in the accompanying Appendix, at Tab D.15.); see also Tr. of June 8, 2004 OFHEO Dep. of J. Juliane at 137:24-138:14 (hereinafter "June Juliane OFHEO Testimony"). (1408) Richard Stawarz was the only individual to acknowledge to us his role in deciding the accounting treatment for the realignment impact. Pennewell claimed that she did not decide how the realignment differences would be recorded; rather, she said Roger Barnes was responsible for determining the appropriate treatment of them. But Barnes claimed that the decisions came down from upper management through his direct supervisor, Mary Lewers. Lewers, for her part, disclaimed any role in instructing or conveying instructions to Barnes and pointed to Pennewell and Stawarz, even though Barnes did not report directly to either of them. Nonetheless, Lewers did say that she agreed with the decision to defer the recognition of realignments. (1409) This also appears to have been KPMG's conclusion following its review of certain realignments. After that review, KPMG noted that the "misclassifications appeared to be an error . . . pursuant to APB 20." KPMG Significant Issues and Decisions KPMG workpaper, dated Dec. 1, 2003; see also FAS 91 - Amortization Catch-up Adjustment for Net Interest Income KPMG workpaper, dated July 30, 2003; see also discussion infra at note 1452. 374
2. Deferral of LASER Realignments Prior to 1998(1410) Management also processed LASER realignments since the early 1990s. The impact of these LASER realignments in the early 1990s were in the magnitude of tens of millions of dollars and, as with STATS realignments performed in 2002 to 2003, were capitalized and amortized over a proxy life.(1411) Management offered rationale similar to that offered for STATS realignment impacts for its treatment of LASER realignment impacts, that is, they had more confidence in the data in iPDI than they did in the data in LASER due to data integrity issues in LASER.(1412) This rationale is no more persuasive for LASER realignments than it was for STATS realignments. We conclude that the LASER realignment impacts also should have been considered corrections of errors under APB 20, and therefore management's treatment of LASER realignment impacts in the early 1990s was not in accordance with GAAP. B. Some Realignment Impacts Were Recorded into Income in the Period in Which They Were Identified Beginning in 1998, management recorded LASER realignment impacts (which involved relatively minor amounts) to the income statement in the period in which the differences were identified.(1413) We have not heard any explanation for treating these - ------------ (1410) Our investigation has not uncovered contemporaneous documentation regarding these pre-1998 LASER realignments. Our knowledge of the treatment of these realignment impacts is based solely on statements made by Fannie Mae employees in the context of describing LASER realignments performed in the "early 1990s." We are not certain that the treatment of LASER realignment impacts in the "early 1990s" is consistent with the treatment of all pre-1998 LASER realignment impacts; however, we note that the treatment described to us differed from the treatment that we observed starting in 1998 and for which we had contemporaneous documentation. (1411) Mem. from J. Juliane to File, dated Oct. 22, 2003, FNMSEC 2892-93. (1412) See Pennewell OFHEO Testimony at 426:24-429:23. (1413) Journal Entry, dated Dec. 3, 2002, FNMSEC 4411 (recording difference to income statement); Journal Entry, dated Mar. 28, 2003, FNMSEC 4399 (same); Journal Entry, dated Dec. 31, 2003, FNMSEC 4348 (same); Journal Entry, dated Mar. 25, 2004, FNMSEC 4272 (same). Both Financial Standards and the Office of Auditing questioned management's methodology and assumptions used to calculate the impact, and have criticized Controller's Office's failure to test its assumptions about the cause of the mismatches between LASER and iPDI. See Undated FAS 91 - Realignment Handwritten Notes, FMSE-IR 289633 (notes from the files of Gregory Ramsey); Mem. from Joyce Philip to L. Spencer, dated Sept. 3, 2004, FMSE-IR 377946-48, at FMSE-IR 377947. We saw no evidence that Controller's Office addressed these questions. 375
impacts differently from the pre-1998 impacts, and have assumed that the difference may be due to the relatively small size of the LASER realignments beginning in 1998. The inconsistent treatment of these impacts suggested that management had unfettered discretion in accounting for realignment impacts, and there was no clear policy or practice at the Company on how to account for realignment impacts. C. Inclusion of Realignment Impacts in the Analysis of Catch-Up Beginning in 2003, management included the cumulative, unrecognized portion of realignment impacts that had been deferred in prior periods ("cumulative realignment deferral") and estimates of the impact that a security reclassification effort, known as Security Master, would have in its FAS 91 amortization calculation.(1414) Management disclosed this information to KPMG, and KPMG did not object.(1415) 1. Security Master Project Beginning in 2002, management undertook what became known as the Security Master Project. The Security Master Project was intended to set rules for the assignment of FAS 91 types to securities, which would reclassify securities in STATS that had been assigned the wrong FAS 91 types, and automate the process of assigning FAS 91 types to securities in STATS by implementing a new rules-based application.(1416) Management anticipated recognizing a realignment impact as a result of implementing Security Master, and included an estimate for the Security Master impact in the FAS 91 amortization calculations in the first quarter of 2003.(1417) In its amortization calculation for the second quarter of 2003, management revised its estimate of the impact of implementing Security Master, and also included the - ------------ (1414) See Q1-2003 Forecast Update, Amortization Runs Sensitivity, December Book, FMSE-IR 55889; FAS 91 - Amortization Catch-up Adjustment for Net Interest Income, dated July, 2003; Mem. from J. Juliane to File, dated July 29, 2003, FMSE-SP 91-93; Mem. from J. Juliane to File, dated Oct. 22, 2003, FNMSEC 2892-93; Mem. from J. Juliane and Rene LeRouzes to File, dated Apr. 27, 2004, FMSE-SP 260-62. (1415) See FAS 91 - Amortization Catch- up Adjustment for Net Interest Income, dated July 30, 2003; Significant Issues and Decisions KPMG workpaper, dated Dec. 1, 2003. (1416) Mem. from J. Juliane to File, dated July 29, 2003, FMSE-SP 91-93, at FMSE-SP 92. (1417) FAS 91 - Amortization Catch-up Adjustment for Net Interest Income, dated July 30, 2003. 376
cumulative realignment deferral for STATS realignments.(1418) Finally, in its amortization calculation for the third quarter of 2003, Pennewell made the decision to include all outstanding cumulative realignment deferrals.(1419) According to Pennewell, KPMG concurred with her decision to include realignments in the catch-up analysis. By including the estimated impacts of Security Master and the cumulative realignment deferral in the catch-up analysis, management was able to reduce the size of, or avoid altogether, the on-top adjustment it would have been required to record under the Company's Amortization Policy(1420) ("Policy") during the first three quarters of 2003. Our review of the facts shows that: - In the first quarter of 2003, the estimated catch-up before the inclusion of the estimated Security Master Project impact was roughly $177.2 million in income. The Policy threshold in this quarter was $99.6 million, and therefore a $77.6 million adjustment to increase interest income would have been required under the Policy to bring the catch- up to the precision threshold.(1421) By including the $118.5 million estimated expense for the Security Master Project, the catch-up number was reduced from $177.2 million to $58.7 million, which fell within the Policy threshold, and thus no adjustment was recorded.(1422) If management had not included the estimated impact of Security Master in the calculation of catch-up, the Policy would have required management to record the $77.6 million of catch- up income, which would have increased Core EPS from the reported amount of $1.84 to $1.89 for the quarter.(1423) Analysts' consensus - ------------ (1418) This included the unamortized amounts related to the November 2002 and May 2003 STATS realignments. Mem. from J. Juliane to File, dated July 29, 2003, FMSE-SP 91-93. (1419) The LASER realignment deferrals relate to realignments performed prior to 1998. Mem. from J. Juliane to File, dated Oct. 22, 2003, FNMSEC 2892-93. (1420) We conclude, in Chapter IV of this Report, that the Amortization Policy was not in accordance with GAAP. (1421) Q1 2003 Forecast Update, Amortization Runs Sensitivity, December Book, dated Apr. 30, 2003, FMSE-IR 55889. (1422) Id.; Q1-2003 Forecast Update, Amortization C-Up Sensitivity, December Book, FMSE-SP 564A. (1423) Assuming a thirty-five percent statutory tax rate, the after-tax income impact of recognizing an additional $77.6 million of interest income would have been $50.4 million. When divided by the Company's 990.2 million diluted average shares outstanding for the first quarter of 2003, the after-tax impact on EPS would have 377
expectation for the quarter were $1.73.(1424) While the Company's results exceeded analysts' expectations, by reducing the amount of the catch- up that would have had to be recognized as additional interest income, the Company was able to avoid a spike in net interest income and net income that may have made it difficult to meet EPS growth targets in future periods. - In the second quarter of 2003, the estimated catch-up before the inclusion of a Security Master estimate and any cumulative realignment deferral was $220.5 million of income. The Policy threshold in this quarter was $105.2 million, that would have required management to recognize an additional $115.3 million of interest income to bring the catch-up to the threshold.(1425) However, management reduced the catch- up adjustment to $126.4 million by including a revised Security Master estimate and cumulative STATS realignment deferral totaling $94.1 million of deferred expense. Accordingly, an adjustment to recognize additional interest income of only $21.2 million was recorded.(1426) If management had not included these realignment adjustments into the calculation of catch-up, it would have had to record an additional $94.1 million of interest income, which would have increased Core EPS from the reported $1.86 to $1.92 for the quarter.(1427) Analysts' consensus expectation for the quarter was $1.87.(1428) - -------------- been an increase of $0.05. Fannie Mae Selected Financial Information as of Dec. 31, 2003, available at http://www.fanniemae.com/ir/pdf/earnings/2003/iap123103.pdf. This analysis assumes management would have deferred the realignment impact as it had previously done. (1424) Fannie Mae First Quarter Net Income Rises 60.5%, BLOOMBERG NEWS, Apr. 14, 2003. (1425) Q2-2003 Forecast Update, Amortization C-Up Sensitivity, dated Aug. 8, 2003, FMSE-IR 35544. (1426) Mem. from J. Juliane to File, dated July 29, 2003, FMSE-SP 91-93, at FMSE-SP 91. (1427) Assuming a thirty-five percent statutory tax rate, the after-tax income impact of the $94.1 million would have been $61.2 million. When divided by the Company's 982.3 million diluted average shares outstanding for the second quarter of 2003, the after-tax impact on EPS would have been an increase of $0.06. Fannie Mae Selected Financial Information as of Dec. 31, 2003, available at http://www.fanniemae.com/ir/pdf/earnings/2003/iap123103.pdf. This analysis 378
- In the third quarter of 2003, the calculated catch-up was $298.1 million of additional interest income. Since the Policy threshold in this quarter was $111 million, an adjustment to recognize an additional $187.1 million of interest income would have been required. However, by including adjustments which totaled a net of $79.6 million of expense,(1429) the calculated catch-up amount was reduced to $218.5 million of income. The new catch-up amount required only the recognition of an additional $107.5 million of interest income.(1430) Had management not included these realignment adjustments, it would have been required to record an additional $79.6 million of interest income under the Policy, which would have increased Core EPS from the reported $1.83 to $1.88 for the quarter.(1431) Analysts' consensus - ------------- assumes management would have deferred the realignment impact as it had previously done. (1428) Fannie 2nd-Qtr Net Income Seen Rising 28%: Outlook, BLOOMBERG NEWS, Jul. 15, 2003. (1429) In addition to including, as adjustments, the cumulative realignment deferral amounts for both STATS and LASER realignments and the "actual impact" of the Security Master Project, management reduced the catch-up amount by an additional $30.4 million to reverse the effects of amortization of discount on four principal-only ("PO") securities sold during the quarter. A document that discusses the additional adjustment stated that "during the third quarter Fannie Mae sold [four] PO's that had a current period catch- up of $30.4 million of income. Since our modeling was done using a June 30 book, we had to adjust the [catch-up] results for any material subsequent events." Mem. from J. Juliane to File, dated Oct. 22, 2003, FNMSEC 2892-93, at FNMSEC 2892. We have seen no evidence that the Company ever adjusted its catch-up to reflect the effect of sold securities prior to this quarter. (1430) Mem. from J. Juliane to File, dated Oct. 22, 2003, FNMSEC 2892-93, at FNMSEC 2892. (1431) Assuming a thirty-five percent statutory tax rate, the after-tax income impact of the $79.6 million would have been $51.7 million. When divided by the Company's 975.9 million diluted average shares outstanding for the third quarter of 2003, the after-tax impact on EPS would have been an increase of $0.05. Fannie Mae Selected Financial Information as of Dec. 31, 2003, available at http://www.fanniemae.com/ir/pdf/earnings/2003/iap123103.pdf. This analysis assumes management would have deferred the realignment impact as it had previously done. 379
expectation for the quarter were $1.75.(1432) As noted previously, by treating realignments as the Company did, management avoided a spike in net interest income and net income for the quarter that may have made it difficult to meet future EPS growth targets. (a) Management Rationale for Incorporating Adjustments into Catch-Up Calculations According to Pennewell, it was KPMG that instructed management to include the estimated impact of implementing Security Master into the catch-up analysis as soon as it could reasonably be estimated. However, a document obtained from KPMG's workpapers for the first quarter of 2003 showed that it was management who proposed to KPMG the idea of including the Security Master estimate into the catch- up framework. The document stated that the Company's calculated catch-up for the first quarter of 2003 fell outside the Policy threshold, and KPMG indicated that management should have recorded the additional income of $77.6 million to comply with its Policy.(1433) According to the workpaper, management declined to do so because it was "in the process of implementing a new application named - Security Master."(1434) According to this document, management chose not to record the $77.6 million amount because it anticipated incurring $118.5 million of expense when Security Master was implemented, which would bring catch- up within the Policy range.(1435) As for including cumulative realignment deferrals in the catch- up analysis, according to Pennewell, the idea was first raised in a conversation that she had with Ann Eilers of the Office of Auditing ("Internal Audit") at the beginning of July 2003.(1436) Pennewell believed that it was appropriate to include cumulative realignment deferrals in - ------------- (1432) Fannie Mae 3rd-Qtr Profit Seen Rising 8%: Outlook, BLOOMBERG NEWS, Oct. 16, 2003. (1433) FAS 91 - Quarterly Review KPMG workpaper, dated Apr. 2003. (1434) Id. (1435) Id. Documents also showed that, from the beginning, the Controller's Office staff sought to spread out over time the financial impact of the Security Master project. In a May 2002 e- mail, Pennewell asked Juliane for available options for recording the effects of the Security Master Project. Specifically, Pennewell asked Juliane when Fannie Mae could start recording an estimate, how Fannie Mae could "get kpmg comfortable," and whether Fannie Mae could "justify booking it over the next 12 months (instead of just through the end of the year?) or instead, could we just bleed it off over time?" E- mail from J. Pennewell to J. Juliane, dated May 16, 2002, FMSE-IR 176623. (1436) E- mail from Ann Eilers to J. Pennewell, dated July 1, 2003, FMSE-E 53219-20; Undated Chart of Realignments and Treatment Used, FMSE-IR 353179. 380
catch-up because realignments identify "inception-to-date" (i.e., cumulative) adjustments to amortization and should be included in the overall analysis of catch-up, which she also considered to be an "inception-to-date" adjustment. We do not believe that an "inception-to-date" adjustment view of realignment impacts justifies the inclusion of these amounts into catch-up. This rationale is inconsistent with management's stated purpose for the "precision" threshold in the Policy. By its terms, the Policy was established to account for fluctuations in catch-up due to the "imprecision associated with forecasting prepayments," and not to deal with security classification errors.(1437) Since cumulative realignment deferrals do not relate to the imprecision in modeling prepayments, they should not have been included in the catch-up framework. In addition, Pennewell's justification for including realignment impacts in the calculation of catch-up is further evidence that management viewed the "precision" threshold, set forth in the Amortization policy, more as a "materiality" threshold for all "inception to date" differences. Juliane also rationalized the treatment of including cumulative realignment deferrals in the catch-up analysis, saying: "When you sit down and you take a hard look at what these things actually represent, these are acquisitions that have no underlying collateral associated with them, which in effect is a paid-off security or paid-off loan. So fully incorporating these into the [catch-up] framework makes complete sense to me."(1438) Juliane, like Pennewell, could not offer a credible reason why a realignment impact that is unrelated to the imprecision in forecasting prepayments should be considered in relation to amounts provided for the "precision" threshold. We conclude that the inclusion of both the estimated impact of Security Master and the cumulative deferred realignment amounts was motivated by management's desire to reduce the calculated catch-up to an amount closer to the Policy threshold and thereby avoid the earnings volatility that would have resulted from recording large catch-up adjustments to interest income.(1439) (b) Calculation of the Security Master Estimate As for how management estimated the impact of Security Master in 2003, Juliane apparently conducted an analysis of the impact that Security Master Project would have on the calculated catch-up. That analysis generated three estimates, all of which would have reduced interest income.(1440) According to Juliane, management - --------- (1437) Letter from Jonathan Boyles to Stephen M. Cutler, dated Nov. 3, 2004, FNMSEC 151-59, at FNMSEC 154; see discussion supra Chapter VI. (1438) Tr. of Aug. 31, 2004 OFHEO Dep. of J. Juliane at 121:9-15 (hereinafter "August Juliane OFHEO Testimony"). (1439) See discussion supra Chapter IV, Subsection IV.B. (1440) Undated MBS Reclass Summary Update, FMSE-SP 155-57. LeRouzes, Juliane's direct report who was responsible for running PDAMS and later, AIMS, claimed 381
elected not to use the larger estimate because it contained mistakes, but never explained what the mistakes were. Rene LeRouzes, who prepared the estimates at Juliane's instruction, stated that he did not know of any mistakes in the estimates. LeRouzes also stated that it was Juliane who chose the final estimate that was used in the first quarter of 2003, but that he did not know the basis for this selection.(1441) While we agree with OFHEO's conclusion that by including the $118.5 million estimate of the Security Master impact in the catch-up analysis the Company was able to avoid recording any catch-up adjustment, we found no evidence to support OFHEO's assertion that management selected the $118.5 million estimate instead of the $393.3 million estimate in order to avoid recording an expense adjustment related to catch-up.(1442) Further, as discussed below, we do not consider the $393.3 million estimate to be the correct estimate of the impact of Security Master, and we also note that the Company could have selected the $393.3 million of additional expense and still exceeded analysts' expectations that quarter. The first of the three estimates prepared by Juliane and LeRouzes(1443) compared actual recorded amortization for the affected securities ($783.8 million of expense) to the amortization that would have been recorded if the securities had not been misclassified ($902.3 million of expense). The result was an additional expense of $118.5 million, which management ultimately chose. While this calculation appears to accurately identify the impact of reclassifying these securities on actual recorded amortization, it ignores the impact it would have on catch-up, since the catch-up generated using the old classification would differ from the catch-up associated with the new classification. In our view, this component should be included in the estimate because the amount was ultimately incorporated into the catch-up framework. The second of the three estimates(1444) compared actual recorded amortization for these affected securities ($783.8 million of expense) to the amortization that should have been recorded if the securities had not been misclassified and if management used its most recent interest rate projections to calculate amortization ($1,177.1 million of expense), resulting in an adjustment that would reduce interest income by $393.3 million. In other words, the additional expense of $393.3 million, which OFHEO concluded was the appropriate estimate of the Security Master impact, - ------------ that he prepared the estimates of Security Master, but that Juliane made further modifications to which he was not privy. See E-mail from KPMG to J. Juliane, dated July 10, 2003, FMSE-E 371464-65; see also E-mail from J. Juliane to KPMG, dated July 9, 2003, FMSE-E 371347-92. (1441) Undated MBS Reclass Summary Update, FMSE-SP 155-57. (1442) OFHEO Report at 60. (1443) Undated MBS Reclass Summary Update, FMSE-SP 155-57, at FMSE-SP 156. (1444) Id. at FMSE-SP 155. 382
incorporated both the effect of the misclassifications on actual recorded amortization and the total amount of catch-up associated with these securities. We find that this calculation too is not correct because it accounts for the total amount of catch-up, rather than the impact on catch-up, associated with reclassifying these securities. In our view, the third estimate,(1445) which was not selected by either management or OFHEO, was the most accurate calculation of the three. It compares the estimate of the amortization that the Company should have recorded using its most recent interest rate projections, both before ($1,110.4 million of expense) and after ($1,177.1 million of expense) the correction of the security misclassifications. Since the $1,110.4 million of expense was presumably already accounted for in either actual recorded amortization or the calculation of catch-up, the impact of correcting the misclassified securities appears to be the difference of $66.7 million of additional expense. We cannot state with certainty what motivated management to select $118.5 million as their best estimate of the impact of Security Master.(1446) Had the Company included the $393.3 million expense estimate in its calculation of catch-up, it would have caused catch-up to change from $177.2 million of income to $216.1 million of expense and the Company would have had to recognize $116.5 million of expense to get within the policy threshold of $99.6 million. Recognizing the $116.5 million of additional expense would have reduced Core EPS from $1.84 to $1.76 for the quarter (as compared to analysts' consensus expectations of $1.73 for the quarter).(1447) Alternatively, had the Company included the $66.7 million expense estimate in its calculation of catch-up, it would have reduced catch-up from $177.2 million of income to $110.5 million of income and the Company would have had to recognize $10.9 million of income to get within the catch-up threshold of $99.6 million. Recognizing the $10.9 million of additional income would have increased Core EPS from $1.84 to $1.85 for the quarter.(1448) Analysts' consensus expectation for the quarter were $1.73.(1449) - ----- (1445) Id. at FMSE-SP 157. (1446) Juliane declined to be interviewed further after he separated from the Company, and we had not covered these areas with him prior to his separation. (1447) Assuming a thirty-five percent statutory tax rate, the after-tax income impact of recording the $116.5 million on-top would have been $75.7 million. When divided by the Company's 990.2 million diluted average shares outstanding for the first quarter of 2003, the after-tax impact on EPS would have been a decrease of $0.08. Fannie Mae Selected Financial Information as of Dec. 31, 2003, available at http://www.fanniemae.com/ir/pdf/earnings/ 2003/iap123103.pdf. (1448) Assuming a thirty-five percent statutory tax rate, the after-tax income impact of recording the $10.9 million on-top would have been $7.1 million. When divided by the Company's 990.2 million diluted average shares outstanding for the first quarter of 2003, the after-tax impact on EPS would have been an increase of $0.01. Fannie 383
KPMG reviewed the estimates relating to Security Master and did not object to the estimates themselves or their inclusion in catch-up.(1450) However, KPMG did note that while management viewed "the change to be a change in estimate resulting from an improved methodology permitted by the implementation of a new computer software accounting system . . an alternate case could be made that the adjustment resulted from an error."( 1451) Nonetheless, KPMG did not believe that it would be necessary to correct previous reporting periods for the Security Master impact.(1452) 2. Management's Inclusion of Realignment Impacts in Catch-up Calculation Was Not Consistent with GAAP Management's decision to include the impact of Security Master and the cumulative realignment deferrals in the calculation of catch-up was inconsistent with the Company's own Amortization Policy, and was also not in accordance with GAAP. In addition, management's justification for including realignment impacts in the calculation - ------------- Mae Selected Financial Information as of Dec. 31, 2003, available at http://www.fanniemae.com/ir/pdf/earnings/ 2003/iap123103.pdf. (1449) Fannie Mae First Quarter Net Income Rises 60.5%, BLOOMBERG NEWS, Apr. 14, 2003. (1450) Security Master - Post Implementation Review, Amortization Catch-Up Analysis Impact, dated Dec. 9, 2003, FNMSEC 4507-08; See FAS 91 - Amortization Catchup Adjustment for Net Interest Income, dated July 2003; Significant Issues and Decisions KPMG workpaper, dated Dec. 2003. (1451) Significant Issues and Decisions KPMG workpaper, dated Dec. 2003. (1452) Id. KPMG concluded that it would not be necessary to correct previous reporting periods for two reasons. First, due to changing amortization speeds over the course of the preceding ten years, it would be impractical or impossible to identify the amount related to prior periods. Second, since the Security Master realignment resulted in additional amortization expense, it was likely that recording its impact in any one quarter would have provided an offset to catch-up, which was in a deferred income position over the past two years. Id. We disagree with KPMG's views. First, APB 20 contemplates a scenario in which the prior period effects are not determinable. "In those rare situations, disclosure will be limited to showing the effect of the change on the results of operations of the period of change (including per share data) and to explaining the reason for omitting accounting for the cumulative effect and disclosure of pro forma amounts for prior years." APB 20 P. 26. Management did not make such disclosures in the 2003 Financial Statements. Second, we conclude, infra Subsection III.C.2., that assessing realignment impacts in relation to the catch-up amounts was not in accordance with GAAP and inconsistent with the stated purpose of the Amortization Policy. 384
of catch-up is further evidence that it viewed the precision threshold as being flexible enough to accommodate items that were unrelated to the imprecision of estimating prepayment speeds. Further, by including the realignment impacts, management did avoid recognizing any catch-up income in the first quarter of 2003, and reduced the amount of catch-up that should have been recorded in the second and third quarters of 2003. Accordingly, we believe that management was motivated by a desire to reduce or eliminate earnings volatility that would result if the full calculated catch-up were recorded. By reducing the amount of additional interest income recorded in those periods, the Company was able to avoid a spike in net income that would have been difficult to replicate in future periods. D. Findings Regarding Fannie Mae's Belated Proposal of a Realignment Policy that Was Inadequate Under GAAP As noted above, we saw no evidence of a clear policy or procedure at the Company on how to account for realignment impacts. Evidence did show that as early as 2001 and again in 2003, Internal Audit had noted as issues the lack of a written policy or procedure for, and inconsistent practices with respect to, realignments.(1453) Staff auditors within Fannie Mae raised the issues to Jonathan Boyles, Eilers and Sam Rajappa.(1454) In early 2003, one staff auditor even noted that she believed a realignment was an inadequate method for resolving STATS differences and expressed concern over a $7.7 million STATS adjustment that was posted to the 1201-07 (on-top) account instead of to income.(1455) However, we saw no indication that any actions were taken to address this concern.(1456) - ---- (1453) E-mail from R. Barnes to J. Philip and P. Wells, dated Jan. 22, 2001, FMSE-SP 12019-20. (1454) We note, however, that Boyles did not recall anyone from Internal Audit raising the issue to him. Tr. of Aug. 24, 2004 OFHEO Deposition of J. Boyles at 95:11-99:16. At least one Internal Audit workpaper questioned whether the accounting treatment given to these amounts was appropriate. Amortization Audit, Mortgage-Related Amortization Audit, dated July 7, 2003, FNMSEC 4484-86. The final audit report did not convey the detailed concerns that the audit found. Instead, the final report conveyed a concern that amortization adjustments arising out of realignments were, in the absence of a written policy, subject to management discretion. Audit Report, Office of Auditing, Amortization Audit, dated July 9, 2003, FNMSEC 4469-76. (1455) E-mail from R. Barnes to M. Lewers, dated Jan. 3, 2003, FMSE-E 1691504. (1456) An Internal Audit witness also stated her belief that the role of Internal Audit did not extend to correcting or recommending substantive changes. 385
It was not until 2004 that management drafted a written policy addressing the accounting treatment for realignment differences.(1457) That document was drafted as part of Fannie Mae's Sarbanes-Oxley documentation and was based on historical practice.(1458) The draft policy was reviewed by Juliane, Pennewell, and Wells before submitting the draft to Internal Audit.(1459) The "Policy and Operational Procedure" required that all "routine" realignments - those that occurred on a quarterly schedule for STATS and an annual LASER realignment - be recognized immediately on the income statement if the "adjustment" was less than $3 million. If the "adjustment" exceeded $3 million, the policy required that it be recorded on the balance sheet, amortized over time, and included in the quarterly catch-up analysis. Only "non-routine" realignments would be subject to analysis by Financial Standards.(1460) None of the purported "reviewers" of this draft policy could identify any support for the policy in GAAP. In fact, Juliane acknowledged that the policy was not consistent with GAAP. According to Juliane, the $3 million threshold came from Financial Standards, but he said that the figure was the only part of the policy that he recalled discussing with them.(1461) Although the policy was not consistent with GAAP because it did not require assessing realignment impacts as errors in accordance with APB 20, it does not appear that the Company implemented the policy prior to the commencement of the - ---- (1457) See Mortgage Portfolio Deferred Price Adjustments Policy and Operational Procedures, Realignment Policy, dated Apr. 30, 2004, FMSE 193585-86; Mortgage Portfolio Deferred Price Adjustments Policy and Operational Procedures P. 4.2.4, dated Aug. 31, 2004, FNMSEC 4169; Pennewell OFHEO Testimony at 427:9-17. (1458) See Mortgage Portfolio Deferred Price Adjustments Policy and Operational Procedures, Realignment Policy, dated Apr. 30, 2004, FMSE 193585-86; Mortgage Portfolio Deferred Price Adjustments Policy and Operational Procedures P. 4.2.4, dated Aug. 31, 2004, FNMSEC 4169. (1459) See E-mail from Emily Passeri to P. Wells and J. Juliane, dated May 3, 2004, FMSE-E 1363919-20. (1460) Mortgage Portfolio Deferred Price Adjustments Policy and Operational Procedures, Realignment Policy, dated Aug. 31, 2004, FNMSEC 4169. (1461) A January 6, 2005 revision to the realignment policy identifies Pennewell and Lewers as approvers. Mortgage Portfolio Deferred Price Adjustments, Policy and Operational Procedures, Realignment Policy, dated Jan. 6, 2005, FMSE-IR 483646-47. Pennewell said that while she was not involved in drafting the realignment policy, she did review it. 386
Special Examination. Fannie Mae currently is focused on improving the accuracy of its data and systems in order to avoid realignments. PART P: ACCOUNTING FOR INVESTMENTS IN AFFORDABLE HOUSING PARTNERSHIPS We address in this section of our Report Fannie Mae's accounting for the tax credits generated by the Company's investments in affordable housing partnerships. I. INTRODUCTION In Chapter IV, we discuss certain issues surrounding the Company's accounting for its investments in affordable housing partnerships with respect to its 1998 financial statements. Our investigation of those issues raised additional questions concerning the Company's accounting in this area. Our inquiry regarding affordable housing partnerships focused on three issues: (1) the Company's accounting for its capital contributions to the partnerships; (2) the methodology used to account for low income housing tax credits ("LIHTC") and net operating losses associated with the partnerships; and (3) the Company's policy and practice regarding the accounting for possible impairment of these investments. In each of these areas, we conclude that the Company's accounting policy and its financial reporting were inconsistent with GAAP. Excluding the events discussed in Chapter IV surrounding the accounting for net operating losses and tax credits in 1998, we have seen no evidence that the Company's accounting or reporting regarding affordable housing partnerships was done with the intent to affect earnings in any period. Instead, in our view, the problems associated with the Company's accounting in this area stemmed from misinterpretations of certain accounting standards and a lack of resources, particularly in the systems area, prior to the late 1990s. II. BACKGROUND A. Applicable Accounting Standards Affordable housing partnerships are entities that invest, directly or indirectly, in low income housing projects that meet certain statutory criteria. Under the Internal Revenue Code, the partnership that sponsors a housing project that meets the criteria is entitled to a tax credit (the LIHTC), which is allocated to the partnership by a state agency. Subject to certain terms and conditions, the investor is entitled to its proportionate share of the partnership's LIHTC allocation. - ----- (1462) This allocation is a direct (1462) Affordable housing partnerships and low income housing tax credits are governed by 26 U.S.C. Section 42 (2000). For background and general information regarding the LIHTC program, see Department of Housing and Urban Development, LIHTC Basics, http://www.hud.gov/offices/cpd/affordablehousing/training/lihtc/basics/. 387
credit to the investor's income tax liability and is unrelated to any net income or losses reported by the partnership to the investors. Under current law, the investor is entitled to the LIHTC for up to ten years.(1463) In certain circumstances, however, the tax credits are subject to recapture if the project does not meet the statutory affordable housing requirements for a fifteen-year "compliance period" (that is, the ten-year term of LIHTC eligibility plus an additional five years).(1464) Prior to 1994, the principal accounting standard for investments in real estate partnerships generally was SOP 78-9.(1465) Under that standard, "investments in noncontrolled real estate general partnerships should be accounted for and reported under the equity method."(1466) Insofar as the application of the "equity method" is concerned, SOP 78-9 refers to a separate standard, APB 18,(1467) as the relevant "guide."(1468) The equity method, as outlined in APB 18, entails the recognition of an investment in the partnership at the investor's cost. At each reporting period, the investor adjusts the carrying amount of the investment to recognize the investor's share of the partnership's net income or loss.(1469) The investor is required to consider whether the value of the investment has been impaired - as by a series of operating losses or other factors - and whether any impairment in the value of the investment should be recognized.(1470) EITF 94-1, which addresses the accounting for investments in affordable housing partnerships, was prompted in part by Congress's 1993 extension of the legislation authorizing tax credits for investments in affordable housing partnerships.(1471) - ----- (1463) 26 U.S.C. Section 42(f). (1464) See 26 U.S.C. Section 42(i). (1465) ACCOUNTING FOR INVESTMENTS IN REAL ESTATE VENTURES, Statement of Position 78-9 (Am. Inst. of Certified Pub. Accountants 1978) (hereinafter "SOP 78-9"). (1466) Id. P. .06. (1467) THE EQUITY METHOD OF ACCOUNTING FOR INVESTMENTS IN COMMON STOCK, Opinion No. 18 (Accounting Principles Bd. 1971) (hereinafter "APB 18"). (1468) SOP 78-9 P. .06. (1469) See APB 18 P. 6(b). (1470) APB 18 P. P. 6(a), 6(b). (1471) ACCOUNTING FOR TAX BENEFITS RESULTING FROM INVESTMENTS IN AFFORDABLE HOUSING PARTNERSHIPS, Emerging Issues Task Force Issue of the Fin. Accounting Standards Bd., Issue No. 94-1 para. 1 (May 1995) (hereinafter "EITF 94-1"). The 388
The consensus offers investors an option, in certain circumstances, to adopt an "effective yield method" of accounting for investments in partnerships that generate LIHTC.(1472) Under the effective yield method, "the investor recognizes tax credits as they are allocated and amortizes the initial cost of the investment to provide a constant effective yield over the period that the tax credits are allocated to the investor."(1473) In determining whether the investment qualifies for the use of the effective yield method, EITF 94-1 directs that only the tax credits be considered.(1474) Net operating losses generated by the partnership do not figure into the determination of the investor's net income or loss.(1475) Under EITF 94-1, investors that choose not to use the effective yield method are referred to SOP 78-9, which "generally requires use of the equity method of accounting for limited partnership investments unless the limited partner's interest is so minor as to give the partner virtually no influence over partnership operating and financial policies."(1476) The consensus further states that "a limited partnership investment in a qualified affordable housing project should be reviewed periodically for impairment."(1477) B. Fannie Mae's Investments in Affordable Housing Partnerships Fannie Mae made its first investment in an affordable housing partnership in 1987.(1478) By 1990, the Company's aggregate equity commitment in these partnerships was $176 million.(1479) The level of investment increased significantly over the next twelve years, reaching $4.0 billion in 2000 and $6.4 billion in 2002.(1480) - ------------- Revenue Reconciliation Act of 1993 retroactively extended and made permanent prior legislation authorizing the tax credit. The prior legislation had expired in 1992. Id. (1472) Id. para. 4. (1473) Id. para. 5. (1474) Id. para. 4(b). (1475) Id. para. 5-6. (1476) Id. para. 7. (1477) Id. para. 9. (1478) Undated History of LIHTC Investment Accounting at Fannie Mae, Zantaz document 1825606 (hereinafter "LIHTC History"). (1479) Id. (1480) Id. 389
According to a Fannie Mae document, the Financial Reporting group of the Controller's Office is responsible for: the accounting for affordable housing partnerships.(1481) However, many of the accounting and financial reporting functions regarding the investments, at least in recent years, have been the responsibility of the LIHTC Financial Management group in the Housing and Community Development ("HCD") Office.(1482) The Multifamily Controller supports this work.(1483) Among other things, the groups in HCD are responsible for the forecasts that Fannie Mae uses to recognize net income or losses on the partnerships during the year; the reconciliations (or "true-ups") between the projections and actual results as reflected on the partnerships' Form K-1s; and for the calculations associated with the recognition of the LIHTC in Fannie Mae's financial statements.(1484) C. Fannie Mae's Partnership Accounting 1. Accounting for Capital Contributions Under EITF 94-1, a company that is unconditionally obligated to make a capital contribution to a partnership in the future should recognize that obligation as a liability on its financial statements.(1485) A contingent obligation should be recognized as a liability when the contingency becomes probable of occurring.(1486) Prior to EITF 94-1, the accounting literature provided that a company should record future capital contributions as a liability if (1) a partnership's losses exceeded the amount of the partner's investment, and (2) the investor was committed to provide additional financial support to the partnership, either by a legal obligation to do so or by some other indication of a commitment (such as past practice or a statement of intent to provide additional support).(1487) Fannie Mae's accounting policies did not provide guidance regarding the proper accounting for the Company's capital contributions. According to James Parks, a - ----- (1481) Undated Accounting Policies, Practices and Controls: LIHTC Investments, Zantaz document 1722005, at 2. (1482) See Low Income Housing Tax Credit Overview, dated Jan. 31, 2004, FMSE 179922-35 (hereinafter "LIHTC Overview") (discussing responsibilities of LIHTC Financial Management). (1483) Prior to 2000, the Multifamily Controller was part of the Controller's Office. (1484) LIHTC Overview at FMSE 179927-33. (1485) EITF 94-1 para. 10. (1486) Id. (1487) SOP 78-9 P. .15. 390
Fannie Mae Vice President who became Multifamily Controller in 1994 (after having been responsible for Financial Standards), Fannie Mae did not follow the principles in SOP 78-9, or EITF 94-1 upon its adoption, with respect to capital contributions the Company was committed to make. Specifically, Parks stated that he was concerned regarding how Fannie Mae accounted for net operating losses of a partnership that were in excess of the recorded investment when the Company was committed to make additional capital contributions in the future. Parks said that Fannie Mae reported partnership losses only to the extent of its net investment (that is, the Company's capital contributions less previously recognized operating losses), even if (1) the losses exceeded the net investment and (2) Fannie Mae was obligated to make additional capital contributions to those partnerships. Parks questioned whether Fannie Mae should have recognized losses when it was obligated to make capital contributions in the future. Fannie Mae's accounting when it was obligated to make future contributions to affordable housing partnerships was inconsistent with the relevant accounting guidance, including SOP 78-9 and EITF 94-1, as discussed above. In addition, it appears that Company officials were aware of this deviation from GAAP at least as of 1997. A memorandum drafted by Jonathan Boyles and Brian Harris addresses the issue and concludes that "[i]f Fannie Mae has a legally binding obligation to make future equity contributions [to an affordable housing partnership], these amounts should be recorded immediately as a liability."(1488) Boyles drafted a virtually identical memorandum in 1999.(1489) That memorandum is attached to a separate memorandum from Boyles to the file, in which he reports on a conversation he had with Leanne G. Spencer, Parks, and Richard DePetris. According to Boyles, all of the attendees agreed with the accounting conclusions reflected in Boyles's memorandum, but "we decided not to gross up Fannie Mae's balance sheet for future equity contributions because the operational problems associated with adopting this accounting method outweighed the benefits."(1490) Boyles reported that recording the future capital contributions "would have added approximately $700 million in both assets and liabilities, and it was decided that this was immaterial to a $500 billion balance sheet."(1491) As will be seen, however, this approach - which understated the Company's obligations to the partnerships - affected more than just the Company's balance sheet. The failure to "gross up" the balance sheet to reflect future capital - ---- (1488) Mem. from Jonathan Boyles and Brian Harris to Distribution, dated Nov. 18, 1997, FMSE 55229-31, at FMSE 55229 (hereinafter "Boyles & Harris Mem."). (1489) Mem. from J. Boyles to Distribution, dated Feb. 18, 1999, FMSE 46568-72 (hereinafter "Boyles Mem."). (1490) Mem. from J. Boyles to File, dated Feb. 18, 1999, FMSE 46567. (1491) Id. 391
contributions Fannie Mae was obligated to make resulted in the Company's failure to recognize losses in the periods in which they were incurred if those losses exceeded the Company's recorded investment. 2. The Company's Accounting Methodologies (a) 1987 - 1995 Fannie Mae first invested in affordable housing partnerships in 1987.(1492) Its method of accounting for these investments has changed several times. The history of the Company's accounting reflects the fact that the Company began with a methodology that was not consistent with GAAP. Prior to 1995, Fannie Mae accounted for its investments in affordable housing partnerships on a cost recovery basis rather than the equity method described in SOP 78-9.(1493) That situation arose because at least some of those involved in the accounting for affordable housing partnerships were unaware of the relevant accounting literature.(1494) According to a document that we believe was drafted in 1998 or 1999, there were "[n]o explicit rules on accounting for LIHTC investment until 1995."(1495) Jim Parks confirmed that, in Fannie Mae's view at the time, there was no applicable accounting literature on affordable housing partnerships until the release of EITF 94-1. Parks said that, given the lack of relevant literature, Fannie Mae's accounting for its investments in affordable housing partnerships under the modified cost recovery method was "not inconsistent" with GAAP. Parks's view was that the differences between applying the modified cost recovery method and the equity method were immaterial, and that the modified cost recovery method was a more conservative approach in any event. The Company's 1994 Financial Accounting Policy Manual confirms that Fannie Mae used a modified cost recovery methodology to account for these investments, but it cites the relevant accounting literature. The policy refers to SOP 78-9 as support for the Company's position and, in a separate comment, states: "Fannie Mae does not consider use of the equity method of accounting or consolidation appropriate for these investments because, while Fannie Mae may have contributed capital significant to the partnership or joint venture, it generally is not entitled to equity profits and does not have - ---- (1492) LIHTC History at 1. (1493) Id. It appears, based on the description in the document, that Fannie Mae was applying a modified cost recovery method under which no income was recognized until the Company recovered its full investment in the partnership. (1494) See Undated LIHTC Impairment Analysis, FMSE-IR 360283-300, at FMSE-IR 360283 (hereinafter "LIHTC Impairment Analysis"). (1495) Id. 392
significant influence over operations."(1496) The policy also states that the modified cost recovery approach was more conservative than the equity method.(1497) Based on the available evidence, we question whether the Company's accounting for its investments in affordable housing partnerships was consistent with GAAP; moreover, given language in the Company's 1994 accounting policy and its reference to SOP 78-9, we conclude that, prior to 1995, the Company made a conscious decision not to use the equity method specified in the applicable accounting literature. As noted, SOP 78-9 required that investors in real estate partnerships use the equity method unless the partner's interest in the partnership was "so minor that the limited partner may have virtually no influence over partnership operating and financial policies."(1498) As the Company's policy acknowledges, Fannie Mae's investments were not "minor." Moreover, under SOP 78-9, an interest in a partnership is "minor" when the investor "is, in substance, in the same position with respect to the investment as an investor that owns a minor stock interest in a corporation . . ."(1499) Although we have not independently reviewed records of the affordable housing partnerships in which the Company invested during this period, we understand that Fannie Mae holds partnership interests in excess of the level that would be considered "minor" under SOP 78-9.(1500) (b) 1995 - 1999 Documents in the Company's files indicated that the Company changed its accounting methodology in 1995 to the equity method.(1501) In addition, DePetris, who joined the Multifamily Controller's group in 1996, recalled using the equity method to account for the partnership investments at that time. The 1996 version of the Financial - ----- (1496) Financial Accounting Policy Manual Sections J.1-J.2, dated Oct. 20, 1994, FMSE-SP 79179-80, at FMSE-SP 79180 (emphasis added). (1497) Id. ("Fannie Mae uses the cost recovery method of accounting because it is a conservative method of revenue recognition and deemed appropriate for these types of investments."). (1498) SOP 78-9 P. .08. (1499) Id. P. .08. (1500) See LIHTC Overview at FMSE 179922. Fannie Mae's 2003 Financial Accounting Policy Guide specifies that an investment in a partnership is "minor" only if the Company's investment is "below 3-5%" of the partnership capital. Financial Accounting Policy Guide Section J.1, dated Dec. 2003 (hereinafter "2003 Financial Accounting Policy Guide"). We note, however, that, because of the transition rules for EITF consensuses, EITF 94-1 effectively "grandfathered" prior accounting treatments. (1501) LIHTC History at 1; see also LIHTC Impairment Analysis at FMSE-IR 360283. 393
Accounting Policy Manual, however, did not reflect any change in the Company's approach; in fact, the 1994 and 1996 versions of the policy are identical in this respect.(1502) The documents that describe the Company's application of the equity method highlight aspects of the Company's approach that were inconsistent with the applicable accounting standards. As an initial matter, the Company applied the equity method "[o]ne year in arrears."(1503) That is, rather than recognize its portion of the partnership's net income or loss in the year that the partnership realized them, Fannie Mae would recognize them in the following year. DePetris recalled that Fannie Mae could not recognize its portion of the partnerships' net income or loss in the same year as the partnerships because the Company could not always obtain financial statements from the partnerships. Thus, the Company would not recognize gain or loss on its investments until it received the K-1 in the following year. Parks recalled recommending that Fannie Mae adopt accrual accounting for its investments (that is, that the Company accrue for its portion of the partnership's net income or loss in the year they are realized by the partnership), but he conditioned his recommendation on the Company's implementing more sophisticated accounting systems. The accounting literature specifies that, if the partnership's financial statements are not timely, then the investor should record its share of earnings or losses based on the most recent financial statements available.(1504) Neither DePetris's nor Parks's explanation for the Company's approach adequately explains the deviation from the requirements of the standard. In light of Fannie Mae's ability to obtain timely financial statements in recent years,(1505) it is not clear why it could not have obtained partnership financial statements on a more timely basis. As to Parks's explanation, the Company began to account for its investments by accruing for its portion of the partnerships' net income or loss in connection with its 1998 statements. However, according to KPMG, the primary determinant of when Fannie Mae began to comply fully with GAAP with respect to its equity method accounting was the improvement of processes surrounding the inputting of data into spreadsheets, and not the implementation of new systems. Moreover, Parks - -------- (1502) Financial Accounting Policy Manual Section J.1-J.2, dated July 1996. (1503) LIHTC History at 1. (1504) APB 18 P. 19(g) states: "If financial statements of an investee are not sufficiently timely for an investor to apply the equity method currently, the investor ordinarily should record its share of the earnings or losses of an investee from the most recent available financial statements. A lag in reporting should be consistent from period to period." (1505) According to DePetris, the decision was made in 2004 to use available financial information from the partnerships, rather than to wait for the K-1s. 394
acknowledged that the primary advantage of the new system was the ease with which it allowed the user to change assumptions. It is not clear why Fannie Mae's prior accounting system, which was based on spreadsheets, required final data, rather than currently available data, to determine the Company's portion of the partnership's net income or loss. Fannie Mae's application of the equity method during this period was inconsistent with the accounting standards in another respect. As an investor in a partnership, Fannie Mae was required to recognize its portion of the partnership's net operating losses ("NOLs"), if any, and deduct those losses in determining the Company's net income.(1506) The NOLs in each period include depreciation of the partnership's assets. Prior to 1998, Fannie Mae was recognizing its share of NOLs using the K-1. The K-1 recognized depreciation expense using a 27.5-year life, which differed from the term used for book purposes.(1507) For accounting purposes, an investor is required to recognize its portion of the GAAP basis net income or loss as reported by the partnership.(1508) The accounting literature applicable to partnership investments requires that the earnings and losses on the investments be determined in accordance with GAAP.(1509) During this period, however, Fannie Mae was applying the equity method by recognizing "Tax basis NOLS."(1510) This issue was raised to Financial Standards in 1997 and the Company concluded that, although GAAP on this point was clear, the difference was immaterial.(1511) Early in 1999, Boyles addressed the issue again and confirmed that "GAAP requires Fannie Mae to record depreciation on a book basis for these investments. Fannie Mae has made the change to book depreciation."(1512) - ----- (1506) See APB 18 P. 19(c). (1507) Mem. from Kimberly Rawls to File, dated Apr. 14, 1998, FMSE-IR 178260-21. (1508) APB 18 P. P. 3(f), 19(c). (1509) Id.; SOP 78-9 P. .06. (1510) LIHTC History at 1. (1511) See Boyles & Harris Mem. at FMSE 55231. The memorandum contains the conclusion that "GAAP requires Fannie Mae to record depreciation on a book basis for these investments," but continues: "however, since the difference is likely to be immaterial to the financial performance of the company, we would not be required to do so. The decision to switch is more of a cost/benefit business decision because the tax department will need to be provided the details of the book/tax differences on a quarterly basis to complete the return." Id. (1512) Boyles Mem. at FMSE 46570. 395
(c) Post-1999 By 2000, Fannie Mae's aggregate equity commitments in affordable housing projects exceeded $4 billion.(1513) Two years later, the aggregate equity commitments reached $6.4 billion.(1514) In 2001, Fannie Mae adopted the effective yield method of accounting for its guaranteed funds.(1515) Documents surrounding the decision focus on the fact that the effective yield method is appropriate for guaranteed investments and that, by using the method, there is a more direct relationship between the returns and the net investment balance.(1516) It was also noted that, compared to the equity method, the effective yield method generally results in higher income recognition in the earlier years of an investment and lower income in later years.(1517) Fannie Mae's application of the effective yield method did not follow the guidance in EITF 94-1 in all respects. The terms of the consensus specify that, under the effective yield method, "the investor recognizes tax credits as they are allocated and amortizes the initial cost of the investment . .. ."(1518) Any other tax benefits, including benefits stemming from recognition of the NOLs, are to be accounted for separately.(1519) Fannie Mae's application of the effective yield method, however, was based on both the guaranteed tax credits and the tax benefits that the Company derived from the partnerships' NOLs. The Company's 2003 Financial Accounting Policy Guide refers to an "internal rate of return considering the cost of the investment and all tax benefits allocated to us."(1520) According to an exchange between Parks and DePetris, the reference to "all" tax benefits includes NOLs: - ---- (1513) LIHTC History at 1. (1514) Id. (1515) Id. (1516) See Undated Low-Income Housing Tax Credits: Change in Accounting Method, FMSE-IR 262406. (1517) Id. (1518) EITF 94-1 para. 5. (1519) "The effective yield is the internal rate of return on the investment, based on the cost of the investment and the guaranteed tax credits allocated to the investor. Any expected residual value of the investment should be excluded from the effective yield calculation." Id. (1520) 2003 Financial Accounting Policy Guide Section J.1 (emphasis added). 396
EITF 94-1 explicitly requires that only tax credits be used in calculating the IRR of a guaranteed LIHTC investment, which is then used to amortize the cost of that investment. The current EYM accounting model uses the tax benefits from both credits and NOLs, which produces a different IRR and, consequently, a different cost amortization pattern. Although our syndicator agreements calculate the guaranteed rate of return by using both credits and NOLs, it's not clear that that is sufficient justification for departing from the explicit requirement in EITF 94-1 of using only credits in the calculation.(1521) This approach was inconsistent with GAAP. 3. Impairment Analysis Both SOP 78-9 and EITF 94-1 require that investors in affordable housing partnerships assess their investments to determine whether there has been an other-than-temporary impairment in the investment's value.(1522) EITF 94-1 describes one method for assessing whether an investment in an affordable housing partnership has been impaired.(1523) At each period, the investor deducts from its net investment its portion of the expected future NOLs. The investor compares that result to the partnership's expected benefits of the investment, including both the LIHTC and any other tax benefits associated with the investment. If the benefits are less than the adjusted basis in the partnership, then the investment is impaired. Prior to 2000, Fannie Mae did not have a formal policy regarding impairment assessments for its investments in affordable housing partnerships. The Company's informal policy was to write off any remaining net investment at the end of the fifteen-year compliance period. That approach, however, foreclosed an assessment of impairment during any prior period, including at the end of the tenth year, when the project was no longer eligible for LIHTC. Given that these credits were the principal economic benefit of the affordable housing partnerships, it is reasonable to assume that many of the Company's investments had been impaired before the fifteen-year period expired. - --- (1521) E-mail from James Parks to Richard DePetris, dated Dec. 2, 2004, FMSE-IR 262405. (1522) SOP 78-9 P. .20 provides: "A loss in value of an investment other than a temporary decline should be recognized." EITF 94-1 states: "The Task Force reached a consensus that a limited partnership investment in a qualified affordable housing project should be reviewed periodically for impairment." EITF 94-1 para. 9. (1523) EITF 94-1, Schedule 4. 397
The Company considered the impairment issue in 2000. Its analysis noted the significant growth in the Company's affordable housing partnership investment portfolio in the late 1990s and the fact that the affordable housing partnership investments are impaired "probably by year 10 when credits expire (possibly earlier)."(1524) The growth in the portfolio was significant because, as various calculations in the analysis showed, recognition of impairments under the pre-2000 method would result in ballooning losses. The fifteen-year policy, for example, was projected to result in $4.1 million in impairment losses in 2002, $31.4 million in 2004, $38.7 million in 2010, and $190 million in 2014 (albeit with year-to-year fluctuations).(1525) The analysis also noted that the Company's new automated accounting system "allows easier partnership level analysis of investments."(1526) The earlier spreadsheet system was "not able to do long-term earnings projections" whereas the new system has "automated capability to do earnings estimates scenarios beyond [the] forecast horizon."(1527) The Company considered several different potential impairment policies and evaluated the impact of each on the Company's partnership investment portfolio.(1528) The options considered included writing off the remaining book value of the investment at year ten or year fifteen, with or without assuming that the partnership had a residual value at that time.(1529) The Company also considered establishing a reserve for future impairments, apparently analogous to a reserve for loan losses.(1530) Fannie Mae chose to implement a policy premised on the impairment of the investment by year ten. According to the 2003 policy: "Fannie Mae regularly evaluates all of its investments for impairment, and any other-than-temporary impairment in the value of such investments is recognized through earnings. Typically, impairment - ----- (1524) LIHTC Impairment Analysis at FMSE-IR 360285. (1525) Id. at FMSE-IR 360300. (1526) Id. at FMSE-IR 360284. (1527) Id. This explanation of the advantages of the new tax system is similar to Parks's assessment discussed above. (1528) See id. at FMSE-IR 360287-300. (1529) See id. at FMSE-IR 360287. (1530) Id. Under "next steps," the LIHTC Impairment Analysis stated: "Decide whether to establish reserve for LIHTC assets (Consider LIHTC reserve while evaluating overall level of Fannie Mae loss allowances)." Id. at FMSE-IR 360294. The various impairment options are reflected in charts that project future impairment losses under each scenario. Id. at FMSE-IR 360288-300. 398
is recognized in early years on tax credit investments because of minimal future cash flow expectations. Additionally, unless a tax credit has projected future cash flows after ten years, any remaining investment balance is written off in year 10."(1531) According to DePetris, in practice all investments were considered impaired at year ten, but not before.(1532) The accounting literature, as well as the Fannie Mae internal policy, requires that impairments be assessed on an individual investment basis.(1533) The impact of the new policy, however, is unclear. As a general matter, the policy of writing off the value of the investments in year ten (versus year fifteen) increases losses in the short run, with a less dramatic increase projected into the future. According to DePetris, however, in 2000, Fannie Mae did not have a significant number of investments that were beyond year ten, so the change in policy did not result in a significant, immediate write-off. A statement in the impairment analysis, under the heading "Next Steps/Decisions" stated: "Decide whether/when to discuss with KPMG."(1534) DePetris could not explain this reference and recalled that the Company had consulted with KPMG on the impairment issue. In fact, DePetris recalled that KPMG audited the ten-year impairment policy. III. FINDINGS Fannie Mae's accounting for investments in affordable housing partnerships violated GAAP in several respects. This was particularly true in the first half of the 1990s when the Company was, among other things, using an inappropriate accounting methodology. We note that the Company is reviewing the past several years' accounting for these partnerships as part of its restatement effort. With the exception of the 1998 accounting changes discussed previously in Chapter IV, we see little evidence that any individual policy, or the policies in the aggregate, were designed or implemented to manipulate Fannie Mae's financial statements. Parks's statement regarding the lack of sophisticated systems, although perhaps not a comprehensive explanation for the Company's accounting in this area, is consistent with views expressed in other areas during the course of our investigation. The spreadsheet system in place prior to about 1997 apparently did restrict the Company's ability to handle more complicated forecasts and projections. Finally, there is - --- (1531) 2003 Financial Accounting Policy Guide Section J.2. (1532) DePetris also stated that the impairment analysis was based on an evaluation of individual investments, but the policy was applied on a portfolio-wide basis. (1533) As the analysis acknowledges, one would expect investments in a number of these partnerships to be impaired prior to year ten. (1534) LIHTC Impairment Analysis at FMSE-IR 360294. 399
no indication of any pressure from senior officers to adopt different policies or procedures to meet a desired outcome. Our conclusion, therefore, is that the problems in this area stemmed from misinterpretations of the applicable standards, a lack of focused attention to the applicable standards, and inadequate resources to tackle some of the more challenging aspects of the accounting for these partnerships as the program grew during the 1990s. 400
CHAPTER VII: CORPORATE GOVERNANCE AND INTERNAL CONTROLS This Chapter of the Report provides an overview of Fannie Mae's Board of Directors, its Office of the Chairman and other elements of senior management, and certain key internal control functions (ethics and compliance, Office of Auditing, and the Office of the Controller) prior to the publication of the OFHEO Report; sets forth our conclusions regarding these entities and functions from the perspectives of corporate governance and internal controls; and discusses the substantial changes that have taken place since the OFHEO Report was released in September 2004. I. BOARD OF DIRECTORS A. Summary With respect to the conduct of the Board prior to September 2004, our findings and conclusions are as follows: 1. The Board endeavored to operate in a manner consistent with its fiduciary obligations and evolving corporate governance standards. The Board was open to examination by third parties and responsive to outside commentary, and it generally received high marks from outside observers. The Board sought and received support and advice from appropriate sources of expertise, including Company management, internal and external auditors, and regulators. Prior to the release of the OFHEO Report, the Board received substantial assurances from internal and external sources that the Company was complying with applicable rules and regulations and with best practices. 2. The Board (and especially the Audit Committee) was sensitive to matters relating to accounting and financial reporting. The Audit Committee requested and received briefings regarding the Company's critical accounting policies, and was assured that Fannie Mae was acting in accordance with relevant standards. The Board reacted quickly to the release of the announcement concerning accounting issues at Freddie Mac, and requested and received updates and assurances from management and KPMG on those issues, as well as advice as to whether there might be similar issues at Fannie Mae. In response, management provided the Audit Committee with a report identifying only minor and immaterial issues at Fannie Mae. 3. The Board responded appropriately when it did receive indications that there were significant issues at the Company. We believe that the Board has made considerable effort to examine and, as warranted, improve its structure, composition, policies, and practices. We believe that the separation of the Chairman and CEO positions, the creation of the Risk Policy and Capital Committee to oversee financial and operational risk management, the transformation of the Compliance Committee into a permanent committee with broad oversight of compliance matters, and the ongoing revamping of delegations of authority, are all positive developments. 401
B. Overview Prior to September 2004 The past five years have seen a substantial transformation in the legal environment in which corporate boards operate and in the benchmarks for best practices in the corporate governance area. Among other developments, Congress passed the Sarbanes-Oxley Act in 2002 ("SOX"), which imposed various governance and internal controls requirements on companies that are SEC registrants.(1535) Also in 2002, the New York Stock Exchange ("NYSE") proposed new corporate governance standards for NYSE-listed companies.(1536) With respect to Fannie Mae specifically, OFHEO, which conducts periodic examinations of Fannie Mae's corporate governance practices, promulgated new corporate governance regulations in 2002,(1537) and amended those regulations in 2005.(1538) As we explain below, the Board and its Committees responded to each of these pronouncements with (as appropriate) new membership criteria, procedures, and responsibilities. The Company looked to outside guidance from law firms and from rating agencies to advise on additional steps and to confirm that the Board's practices were in line with applicable standards. The Board received such confirmation from outside ratings agencies and other appropriate sources, including - through 2003 - from OFHEO. 1. Overview of the Board Although Fannie Mae's corporate governance differs in some respects from that of other corporations due to its status as a government-sponsored enterprise, the role of the Company's Board is typical of other companies: it oversees the management of the Company and its business, and it represents the interests of the Company's stockholders in optimizing the long-term value of the Company.(1539) - ---------- (1535) Sarbanes-Oxley Act of 2002, Pub. L. No. 107-204, 116 Stat. 745 (2002). (1536) NYSE, Inc. Corporate Governance Rule Proposals Reflecting Recommendations from the NYSE Corporate Accountability and Listing Standards Committee As Approved by the NYSE Board of Directors August 1, 2002, available at http://www.nyse.com/pdfs/corp_gov_pro_b.pdf (proposed rules); NYSE, Inc., Final Corporate Governance Rules, dated Nov. 4, 2003, available at http://www.nyse.com/pdfs/finalcorpgovrules.pdf (final rules). (1537) See Corporate Governance Regulations, 12 C.F.R. Sections 1710.1-1710.20 (2005). (1538) See Corporate Governance Regulations, 70 Fed. Reg. 17,310 (Apr. 6, 2005) (to be codified at 12 C.F.R. Section 1710). (1539) See Fannie Mae Corporate Governance Guidelines, dated Apr. 23, 2004, Zantaz document 1454040. More formally, these Guidelines charged the Board with: (1) selecting, compensating, and evaluating the performance of the Chairman/CEO 402
The Charter Act provides that Fannie Mae's Board is to have eighteen members, of whom five are to be appointed by the President.(1540) However, the President has not appointed any directors since May 2004, thereby effectively reducing the size of the Board to thirteen directors.(1541) From 2000 through the end of 2004, three members of Company management sat on the Board, one as Chairman/CEO (Franklin D. Raines) and the other two as Vice Chairmen (at various times, Jamie S. Gorelick, Timothy Howard, and Daniel H. Mudd).(1542) As of September 2004, the executive members of the Board were Raines, Howard, and Mudd.(1543) The Board included ten non-management directors. These directors in 2004 represented a cross-section of interests and backgrounds, including academics, former senior government officials, individuals with experience at other financial institutions, individuals with experience in the mortgage banking industry, an expert in derivatives and financial risk, and the CEO of a public company.(1544) - ---------- of Fannie Mae, and planning for his or her succession; (2) overseeing the election, retention, and compensation of qualified senior executives; (3) reviewing and approving Fannie Mae's strategic and annual operating plans, budget, and corporate performance; (4) advising management on significant issues facing the Company; (5) reviewing and approving significant corporate actions; (6) overseeing the financial reporting process, communications with stockholders, and the corporation's legal and regulatory compliance program; and (7) nominating directors and overseeing effective corporate governance. Id. As is discussed in greater detail below, the Board modified this statement of its responsibilities in 2005 to reflect an expanded role in overseeing management. (1540) 12 U.S.C. Section 1723(b) (2000). (1541) The last Fannie Mae Board meeting attended by Presidential appointees was the meeting of May 10, 2004. See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated May 10, 2004, FMSE 504424-25 at FMSE 504424. (1542) Prior to 2000, Lawrence M. Small, the former President and COO of Fannie Mae, also served as a director. However, he was not a Vice Chairman of the Board. (1543) More detailed information regarding the management directors is provided in Subsection II.B. below. (1544) Since September 2004, new non-management members of the Board have included: an accountant, former public company CFO, and former member of the FASB; a senior executive at another public company with substantial experience in credit management; and a financial services consultant and former senior executive of an investment fund. 403
Fannie Mae's Corporate Governance Guidelines set forth the Board's governance policies. The Board first adopted these Guidelines in 1995. It substantially updated and revised them in January 2003 as an outgrowth of the Nominating and Corporate Governance Committee's Corporate Governance Benchmarking Project (discussed below);(1545) it did so again in early 2004 to reflect final corporate governance standards adopted by the NYSE.(1546) The Board included the Corporate Governance Guidelines in its shareholder proxy statements and published them in a "Corporate Governance" section of Fannie Mae's website.(1547) The Board updated and revised these Guidelines in January 2003. The Board also expanded its conflicts-of-interest policy to create a Code of Business Conduct and Ethics for Directors.(1548) In addition to prescribing a number of conflicts-of-interest policies, the Code required directors to comply with, and to oversee compliance by Fannie Mae officers and other employees with, all laws and regulations. The Code also obligated directors to promote ethical behavior by: (1) encouraging employees to raise ethical concerns with supervisors; (2) encouraging employees to report violations of law, regulation, or Company policy, to appropriate personnel; and (3) emphasizing to employees that they would not permit retaliation against employees who acted in good - ---------- (1545) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504220-21. (1546) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 23, 2004, FMSE 504353-90, at FMSE 504363. (1547) Over time, these Guidelines have: (1) outlined the responsibilities, composition, size, qualifications, and election of the Board; (2) set forth Board policies as to meeting schedules, quorums, attendance, agendas, executive sessions, meeting materials, independence, orientation and education, compensation, and performance evaluations; (3) addressed the Board's committee structure, as well as policies governing committee chairmanships, meeting schedules, agendas, attendance, and reports to the full Board; (4) detailed the policies governing performance reviews of the Chairman/CEO and senior management; (5) set forth policies regarding director access to management and outside advisors, as well as external communications with the Board and with the Audit Committee regarding accounting, internal controls, and auditing matters; and (6) described Fannie Mae's ethics guidelines for directors and for employees. The current version of the Corporate Governance Guidelines can be found at http://www.fanniemae.com/governance/pdf/corpgovguidelines.pdf. (1548) See Fannie Mae Code of Business Conduct and Ethics and Conflict of Interests Policy for Members of the Board of Directors, dated Jan. 21, 2003, FMSE 13542-46; see also Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504221. 404
faith in raising ethical issues.(1549) Each director was required to certify compliance with this new Code of Business Conduct and Ethics on an annual basis.(1550) As of September 2004, the Corporate Governance Guidelines required the Board to meet at least seven times per year.(1551) Between 1999 and 2002, the Board met an average of eight times per year, either in person or by phone. In 2003, in response to ongoing developments (such as the accounting issues at Freddie Mac), the Board met nineteen times. In 2004, again in response to ongoing developments (including OFHEO's Special Examination), the Board met twenty-six times. One meeting each year was a two-day retreat, with time set aside for the Board to focus on a particular topic of interest, such as the market and long-term financial outlook for the Company, the policy and political environment facing the Company, risk management, and strategic planning issues.(1552) The minutes of Board meetings reflect the formal portion of those meetings, with presentations by Committee chairmen and by members of management. The Board also allotted time at each Board meeting for informal discussion. Beginning in January 2003, the Board's Corporate Governance Guidelines required that it schedule time at each meeting for non-management members of the Board to meet in executive - ---------- (1549) See Fannie Mae Code of Business Conduct and Ethics and Conflict of Interests Policy for Members of the Board of Directors, dated Jan. 21, 2003, FMSE 13542-46. (1550) The Nominating and Corporate Governance Committee was and remains responsible for enforcing this certification requirement. The current version of this Code of Business Conduct and Ethics and Conflict of Interests Policy for Members of the Board of Directors can be found at http://www.fanniemae.com/governance/codeofethics/index.jhtml?p= Corporate+Gov ernance&s=Codes+of+Ethics (last visited Feb. 17, 2006). (1551) Fannie Mae Corporate Governance Guidelines, dated Apr. 23, 2004, Zantaz document 1454040, at 54. In 2005, the Board amended the Corporate Governance Guidelines to require the Board to meet at least eight times per year. See Fannie Mae Corporate Governance Guidelines, available at http://www.fanniemae.com/governance/pdf/corpgovguidelines.pdf/. (1552) See Mem. from Thomas E. Donilon to the Board of Directors, dated July 15, 2004, FMSE-KD 48911-12; see also Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 14-15, 2003, FMSE 504286-302, at FMSE 504286; Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 15-16, 2002, FMSE 504166-76, at FMSE 504167; Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 16-17, 2001, FMSE 504074-84, at FMSE 504074. 405
session.(1553) In addition, the Chairman/CEO typically met alone with the non-executive directors prior to Board meetings.(1554) 2. Committees of the Board Fannie Mae's Bylaws authorized the Board to delegate its authority to committees. As of September 2004, Fannie Mae had six standing committees: Nominating and Corporate Governance; Audit; Compensation; Assets and Liabilities Policy; Housing & Community Development; and Executive. Each of these committees had written charters, approved by the full Board, that set forth its responsibilities.(1555) In conducting this part of our review, we focused primarily on the Audit and Nominating and Corporate Governance Committees, which are discussed below. As one would expect, each of these committees has evolved in the changing environment of the past several years. (a) The Audit Committee The Audit Committee has a broad range of responsibilities, including oversight of financial reporting, internal and external auditing, internal controls, and regulatory and legal compliance.(1556) These responsibilities, which are delineated in the Audit Committee Charter, were substantially clarified and expanded in 2003 as an outgrowth of the Nominating and Corporate Governance Committee's Corporate Governance Benchmarking Project. Subsequent amendments were made in February 2004.(1557) - ---------- (1553) Fannie Mae Corporate Governance Guidelines, dated Jan. 20, 2003, FMSE 111395-407, at FMSE 111402. (1554) See, e.g., Moody's Investors Service, Corporate Governance Assessment (Mar. 2004), FMSE 222925-32, at FMSE 222927. (1555) Each of these committees continues to operate today, except for the Assets and Liabilities Policy Committee, which was replaced by the Risk Policy and Capital Committee in February 2005. In October 2004, the Board also established a Compliance Committee to oversee the Company's compliance with the September 2004 Agreement with OFHEO and related matters. The Risk Policy and Capital and Compliance Committees are discussed later in this Subsection. (1556) In November 2005, the Audit Committee's regulatory and legal compliance oversight responsibilities were more narrowly focused on accounting, audit, and tax matters, and the responsibilities of the Compliance Committee were correspondingly expanded to cover regulatory and legal compliance matters. (1557) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Feb. 17, 2004, FMSE 504391-406, at FMSE 504392-97. 406
As of September 2004, the Audit Committee Charter described the purpose of the Committee to be as follows: - oversee (a) the accounting, reporting, and financial practices of the corporation and its subsidiaries, including the integrity of the corporation's financial statements, (b) the corporation's compliance with legal and regulatory requirements, (c) the outside auditor's qualifications and independence, and (d) the performance of the corporation's internal audit function and the corporation's outside auditor; and - prepare the report required by the rules of the Securities and Exchange Commission to be included in the corporation's annual proxy statement. (1558) Beginning in 2000, each member of the Audit Committee was required to meet standards for director independence and experience established by the NYSE.(1559) The Audit Committee thereafter provided written affirmations that its members met these NYSE requirements.(1560) Subsequently, after enactment of SOX, the Company was required to disclose in its annual Form 10-K whether the Audit Committee consisted of at least one "Audit Committee Financial Expert."(1561) In February 2004, the Board designated its first two Audit Committee Financial Experts.(1562) - ---------- (1558) Audit Committee Charter, dated Apr. 23, 2004, Zantaz document 1454040, at 47-49. Beginning in 2004, the Audit Committee also absorbed the functions of the Technology Committee, which the Board dissolved in December 2003. See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 14-15, 2003, FMSE 504286-302, at FMSE 504296, 504298. (1559) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 17-18, 2000, FMSE 503999-4008, at FMSE 504005-06. See also NYSE, Inc., Listed Company Manual Section 303A.07 (2004), available at http://www.nyse.com/lcm. (1560) See, e.g., Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 14-15, 2003, FMSE 504286-302, at FMSE 504298-99; Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 17-18, 2000, FMSE 503999-504008, at FMSE 504005-06. (1561) Sarbanes-Oxley Act of 2002, Pub. L. No. 107-204, Section 407, 116 Stat. 745, 790 (2002); 17 C.F.R. Section 229.401(h) (2005). (1562) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Feb. 17, 2004, FMSE 504391-406, at FMSE 504398. 407
The Audit Committee Charter since January 2003 has required the Committee to meet on at least a quarterly basis.(1563) In practice, Audit Committee meeting minutes show that the Audit Committee held four meetings each year between 1997 and 2001; seven meetings in 2002; nine meetings in 2003; and seventeen meetings in 2004. The Committee met regularly in executive session with representatives of management, the Office of Auditing ("Internal Audit"), and KPMG (the Company's external Auditor prior to December 2004).(1564) Each spring, the Audit Committee held a joint meeting with the Assets and Liabilities Policy Committee to receive an annual examination report from OFHEO. One of the primary duties of the Audit Committee was the appointment, retention, compensation, and oversight of the work of the external auditor.(1565) Beginning in January 2003, the Audit Committee Charter specified the responsibilities of the Audit Committee with respect to its oversight of the external auditor, including oversight of the external auditor's reviews of the Company's financial statements, disclosures, internal controls, and risk exposures.(1566) Also beginning in January 2003, the Committee Charter required the Audit Committee Chairman and the external auditor to review the propriety of earnings releases and other financial disclosures. Moreover, it required the Committee to review with the external auditor and the CEO and CFO the bases for the CEO's and CFO's certifications of the Company's financial statements. Beyond the financial statements and releases, the Charter directed the Committee to discuss with the external auditor, as well as the head of Internal Audit, the adequacy and effectiveness of Fannie Mae's internal controls, including any deficiencies in such controls. In addition, it directed the Committee to discuss with the external auditor Company policies regarding risk assessment and mitigation, and compliance with those policies.(1567) - ---------- (1563) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504228. (1564) See, e.g., Audit Committee Charter, dated Apr. 23, 2004, Zantaz document 1454040, at 49 ("The Committee shall meet separately in executive session, periodically, with each of management, the head of the internal audit department, and the outside auditor."). This remains the requirement today. (1565) See id. This remains the case today. (1566) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504223-30. Previously, the Audit Committee Charter contained a more general statement of the Committee's responsibility to oversee the work of the external auditor. See, e.g., Audit Committee Charter, dated Nov. 18, 2002, FMSE 15166-67. (1567) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504223-30 (containing amended Audit Committee Charter). 408
Audit Committee minutes reflect that KPMG discussed these topics with the Audit Committee during audit plan updates throughout the year, during quarterly meetings to review CEO and CFO financial statement certifications, and during the Committee's consideration of the annual financial statements in February. It also appears that the Audit Committee established a formal "due diligence" process prior to the annual reappointment of KPMG. This process involved KPMG's written response to the Audit Committee's questions and a formal meeting to discuss.(1568) The minutes also indicate that the Committee held executive sessions with KPMG at most meetings. In addition, both Audit Committee Chairman Thomas P. Gerrity and former Committee Chairman Vincent A. Mai stated that they held informal meetings with KPMG once or twice a year to provide an additional opportunity for KPMG to raise any concerns or irregularities outside the presence of management. In addition to overseeing the external auditor, the Audit Committee also was charged with overseeing Fannie Mae's internal audit function.(1569) The Committee Charter required the Committee to: appoint, and when appropriate, replace, the head of Internal Audit; oversee its budget, staffing, and mission; review the scope and performance of the Internal Audit function; review the audit plan and performance against the plan; and discuss with the head of Internal Audit whether there were any restrictions or limitations on Internal Audit's ability to carry out its responsibilities. The Charter did not require the head of Internal Audit to report directly to the Audit Committee. After the publication of the September 2004 OFHEO Report, the Board amended the Audit Committee Charter to state that "the head of the internal audit department shall report directly to the [Audit] Committee.(1570) The Audit Committee was also the primary overseer of the Company's ethics and compliance functions. Historically, the Committee Charter required the Committee: to review the status of the Company's compliance with legal and regulatory developments; to receive periodic reports on the activities of management's Business Conduct Committee, which oversaw ethics and compliance activities; and to review the - ---------- (1568) See, e.g., Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Nov. 17, 2003, FMSE 504796-805, at FMSE 504801; KPMG LLP Presentation to Fannie Mae's Senior Management and Audit Committee Chair, dated Nov. 6, 2003, FMSE-IR 309205-23. (1569) See, e.g., Audit Committee Charter, dated Apr. 23, 2004, Zantaz document 145040, at 49. (1570) Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Oct. 19, 2004, FMSE 504466-85, at FMSE 504477. However, the Board's Corporate Governance Guidelines at times in the past referenced this direct reporting relationship, as did internal organizational charts. See, e.g., Fannie Mae Corporate Governance Guidelines, dated Jan 20, 2003, FMSE 111395-407, at FMSE 111403 ("The corporation's Senior Vice President for Operations Risk, who is responsible for Fannie Mae's internal audit function, reports directly to the Audit Committee."). 409
Code of Business Conduct and monitor compliance with the Code.(1571) In January 2003, the Board amended the Charter to require the Committee to establish procedures for the receipt and handling of complaints regarding accounting, internal accounting controls, and auditing matters, including procedures for confidential, anonymous submission of concerns by employees regarding accounting or auditing matters.(1572) (b) Nominating and Corporate Governance Committee The Nominating and Corporate Governance Committee was (and continues to be) responsible for overseeing both the nomination of the Board's directors and the Board's corporate governance policies. As of September 2004, the Committee's key responsibilities consisted of: (1) identifying and nominating individuals to serve as members of the Board, recommending to the Board tenure and retirement policies for directors, and reviewing memberships on outside boards; (2) overseeing the Company's corporate governance policies and compliance with the Corporate Governance Guidelines and the Code of Business Conduct and Ethics for directors; and (3) evaluating annually the performance of the Board and its committees.(1573) In July 2002, Raines requested that the Committee launch a comprehensive review of Fannie Mae's Board policies and practices with the goal of ensuring that Fannie Mae was "best in class" in the corporate governance area.(1574) The Committee conducted this project (which the Committee dubbed the "Corporate Governance Benchmarking Project") with the assistance of an outside law firm.(1575) The project included an assessment of Fannie Mae's corporate governance policies and practices against the standards contained in SOX, NYSE listing requirements, OFHEO and SEC regulations, and best practices.(1576) - ---------- (1571) See, e.g., Audit Committee Charter, dated Nov. 18, 2002, FMSE 15166-68; Audit Committee Charter, dated July 18, 2000, FMSE 14495-97. (1572) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504227 (containing amended Audit Committee Charter). (1573) See id. at FMSE 504231 (containing amended Nominating and Corporate Governance Committee Charter). (1574) See Minutes of the Meeting of the Nominating and Corporate Governance Committee of the Board of Directors of Fannie Mae, dated July 16, 2002, FMSE 505365-66, at FMSE 505365. (1575) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504220. (1576) See Minutes of the Meeting of the Nominating and Corporate Governance Committee of the Board of Directors of Fannie Mae, dated July 16, 2002, FMSE 410
Based on the results of this project, the Committee recommended a number of updates and additions to Fannie Mae's Bylaws, Board committee charters, and policies. Notable recommendations included: (1) updating the Bylaws to require that the Board maintain Audit and Compensation Committees and to reflect the Company's choice to follow the corporate governance laws of the State of Delaware; (2) updating the Audit Committee Charter to provide for procedures to handle accounting/internal controls/audit complaints; (3) revising the Corporate Governance Guidelines to require regular executive sessions of the Board; (4) expanding the Board's conflicts-of-interest policy to create a code of conduct and ethics for the Board; (5) increasing the frequency of Board meetings; (6) requiring regular evaluations of the Board and Board Committees, and of the Chairman/CEO; (7) addressing director compensation; (8) requiring director consultation with the Nominating and Corporate Governance Committee prior to accepting directorships at other public companies; (9) developing internal procedures to support CEO and CFO certifications of financial statements; (10) establishing a new Housing & Community Development Committee; and (11) publishing Fannie Mae's corporate governance policies on its external web site.(1577) These recommendations were subsequently acted upon by the Board.(1578) Beginning in early 2003, the Committee also began soliciting or supporting assessments of Fannie Mae's corporate governance practices by third party rating agencies, including Standard & Poor's ("S&P"), Governance Metrics International, Moody's Investors Service, and the Corporate Library.(1579) As described by the Company, the purpose of these rating exercises was to validate the Company's actions - ---------- 505365-66, at FMSE 505365; Fannie Mae Corporate Governance Benchmarking Project: Issues for Consideration, dated Aug. 6, 2002, FMSE 12859-87. (1577) See Draft Working Scorecard for Benchmarking Project, dated Jan. 16, 2003, FMSE 13512-16 (detailing the status of recommended corporate governance reforms); Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504232. (1578) See, e.g., Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504220-21. (1579) See, e.g., Standard & Poor's, Corporate Governance Score: Fannie Mae (July 2, 2004), FMSE 224953-67; Moody's Investors Service, Corporate Governance Assessment, dated Mar. 2004, FMSE 222925-32; GovernanceMetrics International, Rating of Fannie Mae, dated June 27, 2003, FMSE-KD 17128-35; Standard & Poor's, Corporate Governance Score: Fannie Mae, dated Jan. 30, 2003, FMSE-KD 17078-96; The Corporate Library, Board Effectiveness Ratings, dated 2003, FMSE-KD 17121-27. 411
resulting from the Corporate Governance Benchmarking Project and to increase the transparency of Fannie Mae's governance policies and practices.(1580) In early 2004, the Committee recommended further changes to the Company's corporate governance policies, mainly to reflect changes to the NYSE listing requirements and SEC rules.(1581) The Committee also received briefings from the Legal Department on a new set of corporate governance regulations that OFHEO had proposed, and the Chairman of the Committee met with OFHEO's Director to discuss corporate governance topics.(1582) Beginning in 2003, as an outgrowth of the Corporate Governance Benchmarking Project, the Nominating and Corporate Governance Committee adopted a formal process for evaluating the performance of the Board and its committees, as well as that of the Chairman/CEO. This included surveys to all directors that requested comments in such areas as: (i) the quality of Board materials, committee reports, and presentations to the Board; (ii) the length, number, and use of time at Board meetings; (iii) the skills, experience, and diversity of the Board; (iv) the balance between the Board and the CEO; and (v) Board access to senior management.(1583) Committee-specific questionnaires were also distributed to members of individual committees.(1584) Committee Chair Ann M. Korologos presented summaries of the evaluations to the Board each year.(1585) Our review of these director evaluations prior to September 2004 found that, overall, they were highly complimentary with respect to the functioning of the Board and its committees. The one criticism that appeared with any consistency is that - ---------- (1580) Fannie Mae press release, dated Jan. 31. 2003, available at http://www.fanniemae.com/newsreleases/2003/2369.jhtml?p=Media&s= News+Rele ases. (1581) Minutes of the Meeting of the Nominating and Corporate Governance Committee of the Board of Directors of Fannie Mae, Mar. 30, 2004, FMSE 505435-37, at FMSE 505435. (1582) 1582 Id. at FMSE 505437. (1583) See, e.g., Board Self-Evaluation Summary, dated 2003, FMSE 13863-84. (1584) See, e.g., id. at FMSE 13870-71. (1585) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 20, 2004, FMSE 504431-37, at FMSE 504436 ("Ms. Korologos reviewed the Board's self-assessment results. All were very positive. She stated that each Committee will review its respective results."); Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 14-15, 2003, FMSE 504286-302, at FMSE 504296. 412
too much time was being consumed at Board meetings by management presentations, with insufficient time for open discussion. 3. Board Oversight of Accounting and Financial Reporting In Practice We discuss elsewhere in this Report the principal accounting and financial reporting issues that we have addressed in the course of our review. We explain in those sections of the Report the extent to which the Board had prior notice of, or involvement in, the decisions made in those areas. We did not find evidence indicating that the Board knew of, or had reason to suspect, the existence of significant accounting problems within the Company prior to the beginning of OFHEO's Special Examination. On the contrary, the Board received assurances in many instances from management, often from KPMG, and in some cases from other advisers as well, that the Company's accounting was in accordance with GAAP.(1586) (a) General Oversight As noted above, the responsibility for the Board's oversight of accounting and financial reporting functions fell primarily on the Audit Committee. The Audit Committee undertook that responsibility by looking to appropriate expertise, including the Company's external auditors. For each year prior to 2004, KPMG advised the Audit Committee that it had conducted an audit in accordance with Generally Accepted Audit Standards and opined that the Company's financial statements complied with GAAP. In addition, there were no concerns noted in the required communications during the period. A number of senior executives were regular participants at Audit Committee meetings. Depending on the point in time, attendees included the members of the Office of the Chairman, as well as the Controller, Leanne G. Spencer, the General Counsel, Ann M. Kappler, and the head of Internal Audit, Sam Rajappa. Spencer made presentations each February regarding the Company's financial statements, as well as periodic presentations (at times accompanied by the head of Financial Standards, Jonathan Boyles) regarding critical accounting policies and emerging accounting issues. Kappler presented periodic updates on litigation and compliance matters. Rajappa, along with Vice President--Audit Ann Eilers, provided updates on Internal Audit's activities. Together with KPMG (which also attended Committee meetings regularly), Rajappa and/or Eilers provided quarterly updates on the status of the Internal Audit/KPMG Joint Audit Plan and, more recently, compliance efforts with respect to Section 404 of SOX ("SOX 404"). Rajappa and/or Eilers also provided annual assessments of the resources and budget available to Internal Audit to accomplish its audit plan. These management presentations provided the Committee with regular assurances that the Company's - ---------- (1586) See, e.g., KPMG's Annual Report to the Audit Committee, dated Feb. 17, 2004, FMSE 222095-104; Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 17, 2004, FMSE 504818-30, at FMSE 504822-23. 413
accounting policies and practices were functioning properly and that critical internal control functions were adequately staffed and resourced (even though, as discussed below, this was not the reality in a number of important respects). As a matter of style, the minutes of the Audit Committee meetings do not include tremendous detail on back-and-forth discussions between the management presenters and the members of the Committee. Committee self-assessments contain a number of comments to the effect that its sessions could have been longer to accommodate more discussion, particularly as the Committee's responsibilities expanded in recent years.(1587) Nonetheless, the minutes reflect the active involvement of Committee members who raised questions, indicated areas of particular interest or concern, and requested follow-up action by management. According to Committee Chairman Gerrity and others, management communicated with the Audit Committee informally separate and apart from the formal Committee meetings (although many of these communications were not documented). (b) The Board's Response to Freddie Mac's Restatement In order to assess the role of the Board in the areas of accounting and financial reporting in a concrete context, we considered, among other things, its conduct in the wake of the accounting issues raised at Freddie Mac. From the outset, the Audit Committee and the full Board received numerous reports on developments at Freddie Mac over a thirteen month period, and paid particular attention to the potential implications of those developments with respect to Fannie Mae's accounting.(1588) Throughout this period, and particularly in a report to the Audit Committee in February 2004, management and KPMG assured the Board that the matters raised in the Freddie Mac situation had no significant implications for Fannie Mae.(1589) - ---------- (1587) In the 2003 Audit Committee self-assessment, two of five respondents "cautioned on the need to ensure sufficient time given the growing agenda." Audit Committee Self Assessment Survey - 2nd QTR 2003, FMSE-IR 629619-25, at FMSE-IR 629622. (1588) Freddie Mac announced that it would restate its financial statements in January 2003. Freddie Mac then engaged the law firm of Baker Botts LLP ("Baker Botts") to conduct an internal review of the facts and circumstances surrounding the accounting errors the Company had uncovered. Baker Botts issued a report in June 2003, which the Company released to the public in July of that year. In November 2003, Freddie Mac issued a press release announcing the results of its restatement and providing additional details on the accounting errors that it had uncovered. See Freddie Mac press release, dated Nov. 21, 2003, available at http://www.freddiemac.com/news/archives/investors/2003/ restatementa_112103.ht ml. (1589) See KPMG's Annual Report to the Audit Committee, dated Feb. 17, 2004, FMSE 222095-104; Minutes of the Meeting of the Audit Committee of the Board of 414
Publicly available reports about Freddie Mac's restatement as of January 2003 concerned two types of transactions: the securitization of a mortgage security and a spread lock. The Audit Committee received a report on these transactions at its February 2003 meeting, and was told that neither of these issues affected Fannie Mae. According to minutes of the meeting, Audit Committee members requested and received confirmation on these points from Howard and from KPMG.(1590) The Board met three times by telephone in June 2003, and the Freddie Mac situation was the major agenda item in the first two meetings,(1591) and led the agenda in the third meeting. During the third meeting, Howard "discussed seven Freddie Mac accounting issues disclosed in Freddie Mac's public statements." The issues fell within four categories: security classification; accounting for derivative instruments; asset transfers and securitizations; and valuation issues. Howard "contrasted apparent Freddie Mac practices with Fannie Mae's practices, concluding that Fannie Mae did not have the same practices and approaches disclosed by Freddie Mac to date."(1592) Howard prepared a summary of his conclusions on the seven issues, which Executive Vice President for Law and Policy Thomas E. Donilon distributed to the Board on July 1, 2003.(1593) Controller Leanne Spencer provided a similar report to the Audit Committee following the release of the Baker Botts report in July 2003. According to the minutes of the Committee meeting, Spencer reported that the Company had "discovered that a small percentage of commitments (less than 2 percent in 2002, and 3.5 percent in 2001) did not settle within the allowable period and therefore should have been marked to market." Spencer concluded that the error "would have had an immaterial impact on financial reporting for any relevant reporting period."(1594) - ---------- Directors of Fannie Mae, dated Feb. 17, 2004, FMSE 504818-30, at FMSE 504822-23. (1590) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 18, 2003, FMSE 504756-62, at FMSE 504760. (1591) Minutes of the Meeting of the Board of Directors of Fannie Mae, dated June 9, 2003, FMSE 504274-75, at FMSE 504274; Minutes of the Meeting of the Board of Directors of Fannie Mae, dated June 13, 2003, FMSE 504276-77, at FMSE 504276. (1592) Minutes of the Meeting of the Board of Directors of Fannie Mae, dated June 27, 2003, FMSE 504278-85, at FMSE 504279-80. (1593) Mem. from T. Donilon to the Board of Directors, dated July 1, 2003, FMSE-IR 244893. (1594) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated July 15, 2003, FMSE 504780-87, at FMSE 504786. Representatives of KPMG are listed as attendees of the meeting. Id. at FMSE 504780. 415
By late November 2003, Fannie Mae was tracking thirty-one accounting issues disclosed in Freddie Mac's restatement. Company management considered whether Fannie Mae might have similar accounting issues. This effort culminated in a presentation to the Audit Committee in February 2004. Spencer's presentation to the Audit Committee in February 2004 highlighted the differences between Freddie Mac and Fannie Mae. She noted "major structural differences in their securitization business" and other significant differences in the companies's "business practices." Spencer reported that, "in consultation with KPMG," Fannie Mae had considered the thirty-one issues identified in connection with Freddie Mac's restatement; that eleven of the issues were inapplicable to Fannie Mae; that eighteen of the issues were applicable to Fannie Mae but raised "no concern"; and that four issues (including two subtopics) required further diligence. As to the four issues that required further diligence (buy-ups, commitments on purchases of AFS securities, out-of-portfolio securitizations, and accounting for loan losses), Spencer reported in each case either that the Company's accounting practice had been reviewed by the SEC or that the impact of any errors appeared to be immaterial. Spencer concluded by informing the Audit Committee that the Company would report to the SEC on the results of its due diligence and would inform OFHEO as to the results of the analysis.(1595) C. Findings 1. The Board Sought to Meet Evolving Corporate Governance Standards Overall, we find that Fannie Mae's Board sought in good faith to respond to evolving legal and other standards in the area of corporate governance; that it sought appropriate internal and external support and advice in meeting these standards; and that its aim was to establish governance policies and practices at least in line with those of leading U.S. corporations and peer institutions. The Corporate Governance Benchmarking Project, conducted with the assistance of an outside law firm, reflects the Board's intent that the Board remain up-to-date in the area of governance, consistent with its obligations and responsibilities. As noted above, we also find that the Corporate Governance Benchmarking Project produced a number of meaningful improvements from a - ---------- (1595) See Audit Committee Update: Understanding Freddie Mac's Accounting Restatement as It Affects Fannie Mae, dated Jan. 23, 2004, FMSE-IR 331074-88, passim. The Company briefed the SEC on these issues in February 2004. See SEC Briefing, dated Feb. 17, 2004, FMSE-IR 599847-62. The Audit Committee was not listed as a recipient of a memorandum that Boyles prepared in which he reported on the results of the briefing, including certain SEC criticisms and his statement that, in one area, "the process we have in place is not exactly what we showed the SEC though it gets very close." Mem. from Jonathan Boyles to Leanne G. Spencer, et al., dated Mar. 13, 2004, FMSE-IR 486916-18, at FMSE-IR 486917. 416
governance perspective. Moreover, based on the advice and reviews that it was receiving from outside advisers, rating agencies, and OFHEO, the Board had every reason to believe, up through release of the OFHEO Report, that it had put in place an effective and highly regarded governance system. In both 2003 and 2004, S&P gave Fannie Mae a score of 9.0 out of 10.0 on its corporate governance practices, stating that such practices were "consistently strong or very strong" across each of S&P's areas of analysis.(1596) In 2003, Governance Metrics International rated the Board an 8.0 out of 10, or "above average."(1597) Also in 2003, The Corporate Library rated Fannie Mae's Board as the "best stakeholder board" it surveyed due to its "explicit commitment to setting the standard for outstanding corporate governance."(1598) In addition, prior to the release of the OFHEO Report in September 2004, OFHEO consistently praised the operations of Fannie Mae's Board. In OFHEO's 2003 Report to Congress, OFHEO assessed the Company's Board as follows: - ---------- The Board of Directors discharges its duties and responsibilities effectively and in accordance with the leading practices that are evolving. The Board is appropriately engaged in the development of a strategic direction for the company, and the Board ensures management appropriately defines the operating parameters and risk tolerances of the Enterprise in a manner consistent with the strategic direction; legal standards; and ethical standards. The Board has an effective process for hiring and maintaining a quality executive management team, and the Board effectively holds the executive management team accountable for achieving the defined goals and objectives. The Board is appropriately informed of the condition, activities and operations of the Enterprise, and has (1596) See Standard & Poor's, Corporate Governance Score: Fannie Mae, dated July 2, 2004, FMSE 224953-67, at FMSE 224953; Standard & Poor's, Corporate Governance Score: Fannie Mae (Jan. 30, 2003), FMSE-KD 17078-96, at FMSE-KD 17078. Following the release of the OFHEO Report in September 2004, S&P lowered Fannie Mae's governance rating twice, and then altogether retracted the rating. See Phyllis Plitch, S&P Pulls Out of Corporate Governance Scoring, ASSOCIATED PRESS, Sept. 9, 2005, available at http://www.mercurynews.com/mld/mercurynews/business/12604850.htm. (1597) GovernanceMetrics International, Rating of Fannie Mae, dated June 27, 2003, FMSE-KD 17128-35, at FMSE-KD 17129. (1598) The Corporate Library, Board Effectiveness Ratings (2003), FMSE-KD 17121-27, at FMSE-KD 17125. 417
sufficient, well-organized time and the necessary resources to carry out its duties and responsibilities.(1599) OFHEO offered the Board similar praise in its earlier reports to Congress.(1600) 2. The Board Was Engaged on Accounting and Financial Reporting Issues The Board and its Audit Committee had put in place systems and support for overseeing the Company's accounting and financial reporting, and reasonably relied upon these systems and support to remain engaged in overseeing management's policies and practices in these areas. The Board, and especially the Audit Committee, was responsive to issues relating to the Company's accounting and financial reporting. The available evidence indicates that the Audit Committee played an active role in the oversight of these areas and received regular reports (and assurances) from management, internal auditors, and outside auditors regarding these matters. The record of Fannie Mae's reaction to the revelation of accounting problems at Freddie Mac confirms that the Board was responsive to potential issues in these areas that were brought to its attention, and that it consistently received assurances regarding Fannie Mae's accounting practices. D. Subsequent Developments Since the end of 2004, the Board made a number of significant changes to its structure, composition, and practices. Some of these steps were taken at the behest of OFHEO, which required certain governance changes as part of its September 2004 and March 2005 agreements with Fannie Mae, and which issued revised corporate governance regulations in April 2005. Other reforms are the product of the Board's own efforts, working with senior management and outside consultants and advisors, to restructure Fannie Mae. As of the beginning of 2006, some changes to the Board are still in the planning stages, while others are only partially complete. The following reflects our current understanding of these changes and their impact. 1. Separation of Chairman and CEO Functions The most significant change to the structure of the Fannie Mae Board of Directors occurred after the resignation of Chairman/CEO Raines on December 21, 2004. Rather than simply replacing Raines with another Chairman/CEO, the Board separated - ---------- (1599) OFHEO Report to Congress, dated June 2003, FMSE-IR 1180-84, at FMSE-IR 1183. (1600) See, e.g., OFHEO Report to Congress, dated June 15, 2002, FMSE-IR 1170-79, at FMSE-IR 1172-73; OFHEO Report to Congress, dated June 15, 2001, FMSE-IR 1166-69, at FMSE-IR 1169. 418
the positions of Chairman and CEO.(1601) The Board amended the Company's Bylaws in January 2005 to give the Board flexibility to split the Chairman and CEO positions.(1602) The Board appointed Stephen B. Ashley as the non-executive Chairman of the Board and Mudd as interim President and CEO of the Company.(1603) In June 2005, after conducting a national search for a new CEO, the Board appointed Mudd to the CEO position.(1604) In splitting the roles of Chairman and CEO, the Board adopted a position that OFHEO had been pressing since its Special Examination of Freddie Mac in the second half of 2003. Whereas Raines stated that he opposed splitting the functions because he believed that a unified Chairman and CEO served as a vital bridge between management and the Board, Ashley stated that the separation of CEO and Chairman allows the current Board to function more effectively because the existence of a non-executive Chairman enables the Board to serve as a stronger check on management. Fannie Mae's most current position description for its non-executive Chairman reflects a "checks and balances" concept: The Chairman of the Board will provide leadership to the Board of Directors, provide for the proper flow of information to and from the Board, oversee closely management's implementation of Board strategic decisions and other significant actions, assist and advise the Chief Executive Officer, assist in promoting effective relations between the company and its external stakeholders (in particular its shareholders, investors, and OFHEO), and promote the highest standards of ethical conduct and integrity by the company and its employees.(1605) Although we recognize that the relationship between the non-executive Chairman and the CEO continues to evolve at Fannie Mae, and that the current - ---------- (1601) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Dec. 21, 2004, FMSE 504507-16, at FMSE 504508. (1602) Unanimous Written Consent of the Fannie Mae Board of Directors, dated Jan. 19, 2005, Zantaz document 1288950, at 1. (1603) Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Dec. 21, 2004, FMSE 504507-16, at FMSE 504508. (1604) Minutes of the Meeting of the Board of Directors of Fannie Mae, dated June 1, 2005, FMSE 510877-78, at FMSE 510878. (1605) Letter from Stephen B. Ashley to Armando Falcon, Jr., dated Apr. 19, 2005, FMSE-IR 314138-45 (attachment entitled "Duties of the Non-Executive Chairman of the Board at Fannie Mae"). 419
arrangement is dictated in part by the temporary circumstances in which the Company finds itself, the separation of these two functions has given the Board a more transparent view of the Company and a more direct role in the oversight of Fannie Mae's management. This appears to have brought substantial benefits in terms of Board oversight of management, information flow between the Board and management, and the tone established and projected at the top of the enterprise. 2. Changes to Board Committee Structure In addition to separating the CEO and Chairman functions, the Board has created two new standing committees. The first is the Compliance Committee, which the Board formed initially to monitor Fannie Mae's compliance with its September 2004 Agreement with OFHEO. The Committee's responsibilities have since expanded significantly. The Board also replaced the Assets and Liabilities Policy Committee with a Risk Policy and Capital Committee in order to provide broader and more effective oversight of the Company's risk management practices. The Compliance Committee, through a combination of diligence in carrying out the Company's commitments and openness in its communications, appears to have improved Fannie Mae's relations with OFHEO.(1606) As a consequence of the Committee's success and the current workload of the Audit Committee, the Board recently gave the Compliance Committee a broader mandate to oversee compliance matters within the Company.(1607) The Compliance Committee now is responsible for "overseeing the corporation's legal and regulatory compliance program, including reviewing and discussing the corporation's Code of Business Conduct and the activities of management's Corporate Compliance Advisory Committee and monitoring compliance with the Code, overseeing the corporation's response to any regulatory examination or other inquiry, and making periodic reports to the Audit Committee regarding legal and regulatory compliance."(1608) - ---------- (1606) This significant progress has been confirmed to us by senior OFHEO officials as well as by Company management. (1607) See Compliance Committee Charter, dated Nov. 15, 2005, FMSE 547749-50, passim. (1608) Id. at FMSE 547749. The Compliance Committee now shares responsibility for oversight of the Code of Business Conduct and of compliance matters with the Audit Committee, but, in practice (as noted above), the latter's focus is now on compliance relating to financial reporting, accounting, and audit and tax matters. See, e.g., Audit Committee Charter, dated Nov. 15, 2005, FMSE 547741-45, at FMSE 547744. The Nominating and Corporate Governance Committee remains responsible for overseeing compliance with the Code of Business Conduct for directors. See, e.g., Fannie Mae Corporate Governance Guidelines, dated Nov. 15, 2005, FMSE 547726-40, at FMSE 547739. 420
The Board has established a new Risk Policy and Capital Committee. This Committee is an outgrowth of the Company's retention of Mercer Oliver Wyman Consulting ("Mercer") to review the organizational design of the Company.(1609) As part of its review, Mercer observed that the Assets and Liabilities Policy Committee's oversight of the Company's risk management policies and practices could be strengthened, and that, in particular, the Board needed a better structure for overseeing operational risk.( ) Mercer recommended revamping the Assets and Liabilities Policy Committee for this purpose.(1610) In February 2005, the Board adopted Mercer's recommendation by reconstituting the Assets and Liabilities Policy Committee as the Risk Policy and Capital Committee.(1611) The Board subsequently approved a charter for the new Committee designed to reflect comprehensive risk oversight responsibilities, including oversight of credit risk, market risk, liquidity risk, and operational risk.(1612) Among other things, the Risk Policy and Capital Committee has worked to develop comprehensive risk policies for the Company, in consultation with management and outside consultants, with a focus during 2005 on policies covering interest rate risk, counterparty risk, credit risk, private label securities, liquidity risk, model validation, derivatives, and the methods of setting risk policies.(1613) A second set of policies is - ---------- (1609) The Mercer engagement fulfilled an obligation set forth in the OFHEO agreement. Paul, Weiss has received periodic briefings from Mercer and the Fannie Mae team with which Mercer has been working, throughout the course of Mercer's organizational design work. (1610) See Mercer Oliver Wyman Consulting, Organization Design Project: Discussion Notes Regarding the Proposed Risk & Capital Committee of the Board, dated Feb. 8, 2005, FMSE-IR 417687-97, at FMSE-IR 417689-90. (1611) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Feb. 17, 2005, FMSE 503339-50, FMSE 503343-44. (1612) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Apr. 19, 2005, FMSE-IR 440463-82, at FMSE-IR 440476-77. Responsibility for overseeing political and reputational risk remains with the full Board. See Minutes of the Board of Directors of Fannie Mae, dated Apr. 19, 2005, FMSE-IR 440463-82, at FMSE-IR 440471. (1613) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 19, 2005, FMSE 519390-400, at FMSE 519394. Although each of these policies has been drafted, not all of them had been approved by the end of 2005. Of those requiring full Board approval, both the credit and interest rate risk policies have been approved. See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Nov. 15, 2005, FMSE 638723-44, at FMSE 638726; see also Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Sept. 20, 2005, FMSE 534653-62, at FMSE 534657. 421
expected to follow that will cover foreign exchange, operational risk, and new product approvals. 3. Changes to Policies and Practices (a) Delegations of Authority At a meeting in May 2005, the Nominating and Corporate Governance Committee revisited the issue of the Board's delegations of authority to management. At that meeting, Board Chairman Ashley expressed the view that reforming current delegations is necessary to "ensure . . . proper management and oversight of such a large and complex organization." General Counsel Kappler also briefed the Committee on steps being taken by management to work "from the lower levels of management upward to ensure proper delegations of signature authority to bind the Company, and other similar internal controls within the business units."(1614) The Board subsequently enlisted Mercer to assist it in restructuring delegated authorities from the Board to the CEO, and from the CEO to management. It is our understanding that Mercer conducted a benchmarking study as part of its work. As of the end of 2005, the Company's work in this area is continuing. (b) Other Changes Fannie Mae's Corporate Governance Guidelines now require, in accordance with the 2005 OFHEO corporate governance regulations, that the Board meet at least eight times annually, and no less than once a calendar quarter.(1615) In practice, the Board continues to meet with greater frequency than is dictated by OFHEO or Board guidelines, and much more frequently than it did prior to the onset of OFHEO's Special Examination. The same basic trend is evident in the Board committees. Whereas the Audit Committee Charter previously required that it meet no less than once a quarter, it now requires that the Committee meet at least six times per year;(1616) and while the charters of the Nominating and Corporate Governance and Assets and Liabilities Policy Committees previously did not set forth a minimum number of meetings, the relevant - ---------- (1614) Minutes of the Meeting of the Nominating and Corporate Governance Committee of the Board of Directors of Fannie Mae, dated May 23, 2005, FMSE 511232-38, at FMSE 511235. (1615) See Fannie Mae Corporate Governance Guidelines, dated Nov. 15, 2005, FMSE 547726-40 at FMSE 547734. (1616) See Audit Committee Charter, dated Nov. 15, 2005, FMSE 547741-45, at FMSE 547744. 422
committee charters now require at least four meetings per year.(1617) In practice, like the full Board itself, these Committees are meeting more frequently than required. In addition, it is our understanding that Board and Board committee meetings now involve, overall, a high level of discussion and debate. In the Board's 2005 self-evaluation, many Board members noted "the candid and open atmosphere that currently exists in the Board and Committee meetings."(1618) Board committee minutes in 2005 reflect directors' asking management probing questions and challenging management's assertions.(1619) For example, Audit Committee minutes in 2005 reflect directors questioning Internal Audit as to audit planning and findings and questioning the Controller's Office as to its accounting practices.(1620) Minutes of the Risk Policy and Capital Committee reflect a similar tone.(1621) Indeed, former interim Chief Risk Officer Adolfo Marzol characterized Risk Policy and Capital Committee meetings as involving "robust" discussions led by Committee Chair Leslie Rahl. We recommend that the Board continue to review its own practices and procedures in the coming months and years, and to solicit candid evaluations from - ---------- (1617) See Nominating and Corporate Governance Committee Charter, dated Nov. 15, 2005, FMSE 547753-54, at FMSE 547754. (1618) Mem. from John K. Wulff to S. Ashley, dated Oct. 16, 2005, FMSE 535798-800, at FMSE 535798 (summarizing comments of Audit Committee members). (1619) Minutes for the Board, by contrast, still exhibit a more minimalist approach to reflecting discussion and debate. (1620) See, e.g., Minutes of the Audit Committee of the Board of Directors of Fannie Mae, dated Jun 20, 2005, FMSE 510748-62, at FMSE 510749 (questioning scope of the Internal Audit Plan for 2005); Minutes of the Audit Committee of the Board of Directors of Fannie Mae, dated May 23, 2005, FMSE 510105-23, at FMSE 510111- 12 (questioning by several Board members of the Vice President -- Audit, Ann Eilers, as to actions taken by Internal Audit to address negative audit findings); Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 17, 2005, FMSE 503025-503034, at 503028 (asking how the Committee could be assured that Fannie Mae has "best in class" implementation of accounting standards). (1621) See, e.g., Minutes of the Meeting of the Risk Policy and Capital Committee of the Board of Directors of Fannie Mae, dated May 19, 2005, Zantaz document 1465929, at 3 (asking Peter Niculescu to discuss procedures and controls used to make certain calculations); id. at 4 (asking that management report on the private label securities portfolio, with specific information on impairment); Minutes of the Meeting of the Risk Policy and Capital Committee of the Board of Directors of Fannie Mae, dated Apr. 18, 2005, FMSE 509916-27, at FMSE 509927 (asking that more work be done before management asks for support for a permanent increase in the portfolio limit). 423
OFHEO, from outside consultants, and from within the Company. We think it is particularly important for the Board to monitor closely ongoing improvements in the areas of the Company's accounting and internal control functions and in related information technology systems. It is important for the Board to complete its ongoing effort to revise delegations of authority to management; to oversee corresponding revisions in the delegations from the CEO to other officers of the Company; and to review how well both sets of delegations are working on a periodic basis. II. OFFICE OF THE CHAIRMAN A. Summary With respect to the function of the Office of the Chairman ("OOC") and other elements of the senior management structure at Fannie Mae through the end of 2004 (when the OOC was dissolved), we conclude that: (1) management did not fully inform the Board of accounting issues, internal control deficiencies, or the inadequacies of internal systems; (2) although Fannie Mae espoused a corporate culture of openness, intellectual honesty, and transparency, the actual environment suffered from (i) an attitude of arrogance (both internally and externally), (ii) an absence of cross-enterprise teamwork (with a "siloing" of information), and (iii) a culture that discouraged dissenting views, criticism, and bad news; and (3) the enterprise lacked appropriate structure and personnel for adequate risk management across risk areas (with an extremely broad collection of functions and authorities residing in the CFO, and the absence of an effective cross-enterprise approach to operational risk management). Since the end of 2004, we believe that Mudd and other senior officers of the Company, with the active engagement of the Board, have made a concerted effort to reform the management structure and the "tone at the top." As discussed above, these changes have included: (1) redefining management committees and lines of reporting with a view to improving internal controls, management of risks, and horizontal and vertical information flow; (2) adopting a management style that seeks to be more open, collaborative, and humble; (3) establishing a Chief Risk Officer position (with an independent Risk organization); (4) revamping the CFO position with a set of responsibilities more appropriate for the position; (5) eliminating the Law and Policy group (with the movement of core compliance functions to a new, independent Office of Compliance, Ethics & Investigations); (6) integrating the Company's businesses (including the Mortgage Portfolio business, which historically reported to the CFO) under a new Chief Business Officer position; and (7) shifting the Company's external relations toward a more cooperative relationship with OFHEO, Congress, and customers (with a substantial reduction in the size and aggressiveness of Fannie Mae's lobbying and grassroots activities). Our sense, as of the date of this Report, is that meaningful, substantive, and tonal changes are well underway at the senior management level of Fannie Mae, and that these changes have improved the functioning of the enterprise both internally and externally. 424
B. Overview Prior to September 2004 The OOC was the center of management power at Fannie Mae. It consisted of Fannie Mae's Chairman/CEO, the various vice chairmen of the Board (who also served as the Company's most senior executives), and the Executive Vice President -- Law and Policy. The OOC was phased out as an organizational unit after December 2004, when Chairman/CEO Raines retired and the positions of Chairman and CEO were separated. 1. Chairman and Chief Executive Officer James A. Johnson served as Chairman and CEO of Fannie Mae from February 1991 until December 1998. Raines succeeded Johnson as Chairman and CEO in January 1999, and served in this capacity through December 2004. Raines first joined Fannie Mae in 1991 as Vice Chairman, reporting to Johnson. From 1991 to 1996, he was part of the OOC and oversaw the Company's finance, legal, and corporate development operations. In September 1996, he left Fannie Mae to become the Director of the Office of Management and Budget. In May 1998, he returned to Fannie Mae as CEO-designate. Upon Johnson's retirement as Chairman/CEO at the end of 1998, Raines assumed the title of Chairman/CEO, and he served in this capacity until he retired on December 21, 2004.(1622) In his capacity as Chairman, Raines presided over meetings of the Board of Directors. He was also authorized to determine agendas for Board meetings, in consultation with the chairs of the Board committees.(1623) At Board meetings, he periodically delivered the Chairman's Report, which addressed a broad range of topics, including strategic initiatives, regulatory and legislative developments, Fannie Mae's charitable activities, and annual meeting proposals contained in proxy statements. He attended meetings of the Audit, Technology, Compensation, Assets and Liabilities Policy, Nominating and Corporate Governance, and Housing & Community Development Committees of the Board. Raines was also responsible for managing the day-to-day affairs of the Company. The Company's Bylaws granted the Chairman/CEO "plenary authority to perform all duties ordinarily incident to the office and such other duties as may be assigned to him from time to time by the Board." He had authority to "prescribe, amend and rescind requirements and procedures governing the manner in which the general - ---------- (1622) See Fannie Mae press release, dated Dec. 21, 2004, available at http://www.fanniemae.com/newsreleases/2004/3405.jhtml. (1623) See Fannie Mae Corporate Governance Guidelines, dated Apr. 23, 2004, Zantaz document 1454040, at 54. 425
business of the corporation shall be conducted." He also had authority to appoint all officers and employees of the Company other than the most senior officers.(1624) Raines acknowledged that Fannie Mae had a significant history of "CEOcentered" decision-making.(1625) Raines and then COO Mudd attempted to broaden decision-making authority in 2000 as part of a corporate reorganization, but subsequent comments from other senior officers reflect ongoing concern about the degree to which Raines delegated authority in practice and his level of involvement in the details of decision-making.(1626) Raines's direct subordinates varied during his tenure as Chairman/CEO. Originally reporting to Raines in 1999 were President and COO Small, Vice Chair Gorelick, Executive Vice President and Chief Credit Officer Marzol, and the Senior Vice Presidents for Corporate Development and Communications. Mudd began reporting to Raines after Mudd replaced Small as COO in 2000.(1627) Also in 2000, CFO Howard began reporting directly to Raines, after having previously reported to Small. In 2003, - ---------- (1624) See Fannie Mae Bylaws, Section 4.08 (2004). With the recent split of the Chairman and CEO positions, the CEO is now limited to the powers historically provided to Fannie Mae's President - namely, such powers "as the Board of Directors may prescribe, or as the Chairman of the Board may delegate to him." Fannie Mae Bylaws, Section 5.06 (2005). (1625) See Remarks at the Mid-Year Officers Meeting: Getting to 2003, dated July 18-19, 2000, FMSE-IR 675181-534, at FMSE-IR 675227-29 ("One of the things I said to the Operating Committee in our retreat was that Fannie Mae for twenty years has been a CEO-centered company. The magic words have always been, `blank wants it this way.' . . . `Frank wants it this way.' End of discussion."). (1626) See, e.g., Leadership Perspectives: Report on Frank Raines, dated Nov. 2003, FNMSEC 44110-156, at FNMSEC 44128 (Chief Credit Officer Adolfo Marzol commented that Raines "does not seem to delegate as well as he should given his responsibilities. Too many issues are decided by him or by group votes of the SLT."); Officer Year-End Leadership Survey: Rater Feedback Summary on Franklin D. Raines, dated Oct. 2002, FNMSEC 44077-109, at FNMSEC 44087 (Senior Vice President -- Missions and Markets Michael Quinn stated that "[m]icromanaging on such issues as budget, hiring and immaterial credit decisions is what Frank's organizational structure is geared for and that is not conducive to accelerating change."); id. at FNMSEC 44107 (Senior Vice President -- Housing and Community Development Robert J. Levin stated that "Frank is almost always the smartest person in the room on virtually any subject. This can cause him to `take over' almost any project in the Company.") (1627) The title of "President" was not used between 2000 and June 2005, when it was combined with the CEO title rather than the COO title (with Mudd's appointment as President and CEO). 426
when Gorelick resigned as Vice Chair, Executive Vice President -- Law and Policy Donilon began reporting directly to Raines. At the time of the publication of the OFHEO Report in September 2004, only Howard, Mudd, and Donilon reported directly to Raines.(1628) Raines participated in a number of management committees. As of 2004, these committees included the Senior Leadership Team, the Operating Committee, the Assets & Liabilities Committee, and the Credit Committee - each of which Raines chaired. He also participated as a member of the External Affairs Committee and the Risk Policy Committee.(1629) These committees are discussed below. 2. Vice Chairmen of the Board Prior to 2000, Fannie Mae's Bylaws authorized the Board to elect a single vice chairman. When Mudd joined the Company in 2000, the Bylaws were changed to provide for one or more vice chairmen, each of whom was to perform such duties as the Board prescribed or as the Chairman/CEO delegated.(1630) During Raines's tenure as Chairman/CEO, the Board usually had two vice chairmen at any given time. In addition to serving as Board directors, the vice chairmen also served as the most senior members of management (other than the Chairman/CEO). Timothy Howard served as Vice Chairman of the Board, as well as Executive Vice President and CFO of the Company, at the time he resigned in December 2004.(1631) Howard joined Fannie Mae as Vice President and Chief Economist in 1982, and he subsequently served as Senior Vice President -- Economics and Planning. He was appointed Executive Vice President -- Economics, Strategic Planning, and Financial Analysis in September 1987, and became Executive Vice President -- Asset Management in January 1988. He was named CFO in February 1990, began reporting directly to - ---------- (1628) See Officers Organization Chart, dated Nov. 19, 1999, FMSE-IR 1199; Officers Organization Chart, dated Nov. 21, 2000, FMSE-IR 1197; Officers Organization Chart, dated Nov. 21, 2000, FMSE-IR 1196; Officers Organization Chart, dated Nov. 18, 2003, FMSE-IR 1192; Officers Organization Chart, dated Oct. 19, 2004, FMSE-IR 1190. (1629) See List of all Officer Meetings and Standing Participants 2004, dated Nov. 8, 2004, FMSE-IR 2821-24. (1630) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Feb. 23, 2000, FMSE 503977-82, at FMSE 503978-79 (amending section 4.08a of the Bylaws). (1631) See Fannie Mae press release, dated May 20, 2003, available at http://www.fanniemae.com/newsreleases/2003/2553.jhtml?p=Media&s=News+Rele ases; http://www.fanniemae.com/newsreleases/2004/3405.jhtml. 427
Raines in 2000, and became a member of the OOC in November 2000. He was named Vice Chairman of the Board in May 2003.(1632) At Board meetings, Howard delivered periodic reports concerning the Company's financial performance and its stock price performance. In January of each year, he reviewed with the Board the Company's annual business plan. On occasion, he also briefed the Board on such topics as the impact of new accounting pronouncements,(1633) risk management,(1634) manufactured housing bonds,(1635) and derivatives.(1636) Howard also attended meetings of the Assets and Liabilities Policy, Audit, Compensation, and Technology Committees of the Board. At meetings of the Assets and Liabilities Policy Committee, Howard provided regular reports on the Company's capital account and its use of derivatives. Howard also regularly made recommendations to that Committee concerning equity issuance and repurchase and the payment of dividends on the Company's preferred and common stock. At Audit Committee meetings, Howard introduced proposed resolutions to appoint the external auditor for the coming year. He also provided occasional briefings on the impact of key accounting changes (e.g., the - ---------- (1632) See E-mail from Lindsay Conway to Steven Fisher, et al., Speaker Bios, dated Apr. 16, 2004, Zantaz documents 1161976, 1161977; Fannie Mae press release, dated May 20, 2003, available at http://www.fanniemae.com/newsreleases/2003/2553.jhtml?p=Media&s=News+Rele ases. (1633) See, e.g., Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Feb. 20, 2001, FMSE 504046-58, at FMSE 504054. (1634) See, e.g., Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 14-15, 2003, FMSE 504286-302; Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 15-16, 2002, FMSE 504166-76, at FMSE 504174. (1635) See, e.g., Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504238. (1636) See, e.g., Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Apr. 16, 2002, FMSE 504147-59, at FMSE 504156. 428
advent of FAS 133(1637)), as well as on the Company's risk assessment and risk management practices.(1638) When Raines became Chairman/CEO in 1999, Howard's direct reports included the Senior Vice Presidents for Financial and Structured Transactions, Financial Information, Portfolio Management, Portfolio Strategy, and Investor Relations, as well as the Treasurer and the Controller. In 2002, the reporting line for Executive Vice President and Chief Credit Officer Marzol shifted from Mudd to Howard, as did that for the head of Internal Audit, Rajappa (on a "dotted line" basis).(1639) As of the fall of 2004, Howard oversaw financial risk policy, corporate financial strategy, credit finance, accounting and budget functions, procurement, and investor relations, as well as the Company's Mortgage Portfolio business (which included portfolio transactions, portfolio strategy, and the Treasurer's Office). He also served on a number of management committees. In addition to being chairman of the Risk Policy Committee, he was a member of the Senior Leadership Team, the Operating Committee, the External Affairs Committee, the Operations, Transactions, and Investments Committee, the Assets & Liabilities Committee, and the Credit Committee.(1640) The functions of these committees are described below. Mudd joined Fannie Mae in 2000 as Vice Chairman and COO (replacing Small). He served in this capacity until December 2004, when he replaced Raines as CEO on an interim basis.(1641) On June 1, 2005, the Board of Directors unanimously approved Mudd's appointment as President and CEO, thereby ending his "interim" status.(1642) - ---------- (1637) See, e.g., Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 20, 2001, FMSE 504659-65, at FMSE 504661-62; Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 16, 1999, FMSE 504603-17, at FMSE 504606-07. (1638) See, e.g., Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 14, 2003, FMSE 504767-75, at FMSE 504771-72. (1639) See Officers Organization Chart, dated Nov. 16, 1999, FMSE-IR 1198; Officers Organization Chart, dated Nov. 19, 2002, FMSE-IR 1193. (1640) See Officers Organization Chart, dated Oct. 19, 2004, FMSE-IR 1190; List of All Officer Meetings and Standing Participants 2004, dated Nov. 8, 2004, FMSE-IR 2821-25, at FMSE-IR 2821. (1641) See Fannie Mae press release, dated Dec. 21, 2004, available at http://www.fanniemae.com/newsreleases/2004/3405.jhtml. (1642) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated June 1, 2005, FMSE 510877-78. 429
At Board meetings through the end of 2004, Mudd periodically presented the COO's Report, which discussed budget and compensation issues and reviewed the achievement of annual corporate objectives. At the first Board meeting of each year, he presented the Corporate Performance Review, which discussed the Company's performance in such areas as earnings and revenues, HUD housing goals, workplace issues, and technology improvements. Mudd also regularly attended meetings of the Assets and Liabilities Policy, Audit, Compensation, Housing & Community Development, and Technology Committees of the Board. When Mudd became COO in 2000, his direct reports included Executive Vice President and Chief Technology Officer Julie St. John, Executive Vice President and Chief Credit Officer Marzol, Executive Vice President -- Single-Family Mortgage Business Louis Hoyes, the Senior Vice President -- Human Resources, and the Vice President -- Diversity.(1643) In addition, the head of Internal Audit, Rajappa, reported to Mudd on a "dotted line" basis. In 2002, Executive Vice President -- Housing and Community Development Robert J. Levin and e-Business President Michael Williams began reporting to Mudd, while (as noted above) Marzol's reporting line and Rajappa's "dotted line" reporting relationship shifted to Howard.(1644) As of the fall of 2004, Mudd oversaw the Single-Family Mortgage, e- Business, and Housing & Community Development businesses, as well as contracts and procurement, facilities management, corporate training and education, organizational development, recruiting, compensation and benefits, and employment practices. He also served on a number of management committees. In addition to chairing the Operations, Transactions, and Investments Committee, he was a member of the Senior Leadership Team, the Operating Committee, the Risk Policy Committee, the External Affairs Committee, the Assets & Liabilities Committee, and the Credit Committee.(1645) These committees are described below. 3. Other Members of the Office of the Chairman Donilon served as Fannie Mae's Executive Vice President -- Law and Policy from May 2000 until April 2005. Previously, from 1999 to 2000, Donilon served as the Senior Vice President and General Counsel of the Company. Initially, Donilon reported to Vice Chair Gorelick; after she left the Company in 2003, Donilon reported directly to Raines. Donilon became part of the OOC in 2003. - ---------- (1643) See Officers Organization Chart, dated Nov. 21, 2000, FMSE-IR 1196-97. (1644) See Officers Organization Chart, dated Nov. 19, 2002, FMSE-IR 1193. (1645) See Officers Organization Chart, dated Oct. 19, 2004, FMSE-IR 1189-90; List of All Officer Meetings and Standing Participants 2004, dated Nov. 8, 2004, FMSE-IR 2821-25, at FMSE-IR 2821. 430
Donilon served as the Corporate Secretary, and was responsible for facilitating, and creating a record of, meetings of the Board. Donilon attended most Board meetings, as well as meetings of all Board committees with the exception of the Housing & Community Development Committee. Donilon periodically briefed the Board on legislative and regulatory developments and corporate governance matters. As head of Law and Policy, Donilon supervised the Legal Department, Regulatory Policy, Government and Industry Relations, and Communications. His principal responsibilities included managing Fannie Mae's relationships with its regulators, Congress, the media, and the housing and mortgage finance industry. He chaired meetings of the External Affairs Committee and served on a number of other management committees, including the Senior Leadership Team, the Operating Committee, the Risk Policy Committee, the Operations, Transactions, and Investments Committee, and the Credit Committee.(1646) The functions of these committees are described below. 4. Management Committees Members of the Office of the Chairman chaired or otherwise participated in a number of senior management committees that served information-sharing, advisory, and/or decision-making functions within the Company. Many of these committees do not appear to have had charters or other formal organizing documents. The structure of these committees changed significantly over time as part of various management reorganizations. As of September 2004, the most important of these committees consisted of the Senior Leadership Team, the Operating Committee, the External Affairs Committee, the Assets & Liabilities Committee, the Disclosure Committee, and several committees devoted to risk management. (a) Senior Leadership Team The Senior Leadership Team ("SLT") was a senior management committee created by Chairman/CEO Raines. As described by Mudd, the SLT was intended to "build alignment, integration and teamwork for the entire enterprise."(1647) The SLT typically met once a week, until its dissolution in December 2004 when Raines and Howard left the Company. Members of the SLT included the members of the OOC, along with each of the Executive Vice Presidents and the President of the Company's e-Business - ---------- (1646) See, e.g., Officers Organization Chart, dated Oct. 19, 2004, FMSE-IR 1189-90; List of All Officer Meetings and Standing Participants 2004, dated Nov. 8, 2004, FMSE-IR 2821-25, at FMSE-IR 2821. (1647) Notes from Dec. 9, 2003 SLT Meeting, dated Dec. 12, 2003, FMSE-IR 273713. (1648) See id.; Organization Charts, 2004, FMSE-IR 1188-201, at FMSE-IR 1193. 431
Non-SLT members (including, for example, General Counsel Kappler and Senior Vice President -- Operations Risk Rajappa) also attended occasionally to brief the SLT on topics related to their particular areas of responsibility. Depending on the issue area, the SLT served an informational, advisory, and/or decision-making role. The SLT discussed a variety of topics at its meetings. It frequently discussed human resource issues (including officer promotions and staff reorganizations, employee performance evaluations, and benefits) and, in 2004, compliance with SOX.(1649) Other periodic topics included the Company's progress in achieving annual HUD housing goals, competition from other financial institutions, planning for quarterly business reviews and year-end officers' meetings, stock price goals, corporate expense and budget targets, and tax issues related to the relationship of the Company with the Fannie Mae Foundation. (b) Operating Committee The Operating Committee was an information-sharing committee designed to address cross-divisional issues and the goals and strategic plans for the Company.(1650) The Committee discussed a broad range of issues, including housing goals, debt market conditions, portfolio projections, political and regulatory developments, and financial and business results. Operating Committee meetings were chaired by Raines and were attended by the other members of the OOC, the Executive Vice Presidents, and select Senior Vice Presidents. The committee met once a week. (c) External Affairs Committee The External Affairs Committee was not a formal committee, but rather a weekly informal session in which senior management discussed Fannie Mae's important external relationships, including those with Congress, the mortgage industry, and OFHEO. Representative topics of discussion included: pending legislation and contacts with legislators and other officials; public relations and industry outreach; developments regarding Freddie Mac and Ginnie Mae; and campaign contributions and other politically-oriented funding activities. - ---------- (1649) The SLT's review of SOX 404 compliance included at least two instances when it met with representatives of KPMG, the Company's outside auditor. See, e.g., Minutes of the Meeting of the Senior Leadership Team, dated Dec 13, 2004, FMSE-IR 140294, at FSME-IR 140294; Minutes of the Meeting of the Senior Leadership Team, dated Sept. 13, 2004, FMSE-IR 264568. (1650) See, e.g., Mem. from Hal Gann to F. Raines, dated Dec. 6, 2002, FMSE-IR 272863-64, at FMSE-IR 272863. 432
Donilon chaired meetings of the External Affairs Committee, succeeding Vice Chair Gorelick when she left the Company in 2003. Regular attendees included the members of the OOC, Executive Vice President -- Housing and Community Development Levin, Senior Vice President -- Investor Relations Jayne Shontell, and select officers from the Law and Policy Division (including General Counsel Ann Kappler and the Senior Vice Presidents for Regulatory Policy, Communications, and Government and Industry Relations).(1651) (d) Assets & Liabilities Committee The Assets & Liabilities Committee served to update senior management as to the Company's overall market position and its business results versus projections. The members of the Committee consisted of Raines (who served as Chairman), Mudd, Howard, the Executive Vice Presidents who headed the Mortgage Portfolio and Single-Family businesses, the Executive Vice President -- Housing & Community Development, the Controller, the Treasurer, and certain other senior officers.(1652) The Committee met once a week. A typical meeting would include reports on (i) the Company's position in the mortgage market and the overall market environment, (ii) the Mortgage Portfolio business (including the mortgage portfolio's duration gap, options purchases, and forecasts of the Company's regulatory capital), (iii) the Single-family and Multifamily businesses (including mortgage acquisitions, progress in achieving regulatory housing goals, and the status of minority lending initiatives), and (iv) the e-Business. (e) Disclosure Committee The Disclosure Committee was established in October 2002, with a key goal of advising the Chairman and CEO and the CFO regarding their certifications as to the accuracy of Fannie Mae's quarterly and annual financial disclosures.(1653) The Committee met several times per quarter. The members of the Disclosure Committee included General Counsel Kappler (who served as Chair), Controller Spencer, Senior Vice President -- Investor Relations Shontell, Senior Vice President -- Operations Risk Rajappa, Senior Vice President -- Communications Charles Greener, Senior Vice - ---------- (1651) See, e.g., E-mail from T. Donilon to William Maloni, dated May 22, 2004, Zantaz document 1698424. (1652) See, e.g., Undated Management Committee Structures and Membership, Zantaz document 553142, at 1; List of All Officer Meetings and Standing Participants 2004, dated Nov. 8, 2004, FMSE-IR 2821-25, passim. (1653) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Oct. 15, 2002, FMSE 504723-29, at FMSE 504724. These certifications were required by SOX. Fannie Mae voluntarily certified its financial statements in 2002, prior to registering its common stock with the SEC in 2003. 433
President and Treasurer Linda K. Knight, and Senior Vice President -- Policy and Standards Robert Englestad.(1654) In January 2004, the Audit Committee approved a charter for the Disclosure Committee.(1655) This charter charged the Disclosure Committee with, among other things: (1) designing, implementing, and monitoring controls and policies sufficient to ensure that information contained in the Company's financial statements was reported accurately and within time frames established by the SEC; (2) reviewing the Company's financial disclosures for accuracy and completeness, and approving these disclosures prior to their release; (3) discussing the preparation of the Company's financial disclosures with the Chairman/CEO, and CFO, and reviewing with the Chairman/CEO and CFO any significant disagreements among management, the outside auditor, internal auditors, or other employees in connection with the financial disclosures; and (4) facilitating open communication about disclosure among management, Internal Audit, the external auditors, the Audit Committee, and key employees.(1656) The due diligence process in preparation for Chairman/CEO and CFO certifications as to Fannie Mae's financial reports flowed up to the Disclosure Committee, with oversight by the Audit Committee. First, Senior Vice Presidents representing key business units provided certifications to Raines and Howard regarding their knowledge of any untrue statements or omissions of material fact in the financial reports. Controller Spencer provided a similar certification as to whether the financial reports fairly presented in all material respects the financial condition of the Company. In addition, as the head of Internal Audit, Rajappa obtained "Internal Control Representation Letters" from the Senior Vice Presidents addressing the adequacy of internal controls over financial reporting and the existence of any significant deficiencies or material weaknesses in those controls that could have a material impact on the Company's financial reporting. Based on these reports, Rajappa certified to Raines and Howard whether the Company's internal controls over financial reporting were adequate. After these certifications had been provided, the Disclosure Committee met to evaluate the financial reports and certifications and to provide its own certification to Raines and Howard.(1657) Raines and Howard then met with members of the Disclosure - ---------- (1654) See Fannie Mae Disclosure Committee Charter, dated Jan. 23, 2004, FMSE-IR 253048-50, at FMSE-IR 253048. (1655) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Jan. 23, 2004, FMSE 504806-15, at FMSE 504811. Prior to January 2004, the Disclosure Committee appears to have operated under a draft charter that had not been approved by the Board. (1656) See Fannie Mae Disclosure Committee Charter, as amended on Jan. 23, 2004, FMSE-IR 253048-50. (1657) See, e.g., Certification of the Members of the Disclosure Committee, dated May 10, 2004, Zantaz document 1920088. 434
Committee to review the procedures undertaken to prepare the financial reports, and to discuss any outstanding concerns. At such meetings, Kappler indicated her assessment as to whether the due diligence process supported the assertions contained in the Chairman/CEO and CFO certifications. Based on these certifications and assessments, Raines and Howard would decide whether to certify the financial reports.(1658) (f) Risk Management Committees Fannie Mae employed an evolving set of management committees to oversee the various risks facing the Company. The Board formally approved a broad reorganization and expansion of the risk management committee structure in October 2000, with the creation of three new management-level committees: the Operations, Transactions, and Investments Committee ("OTIC"); the Portfolios and Capital Committee; and the Credit Risk Policy Committee.(1659) At a January 2004 meeting, the Board approved the dissolution of the Credit Risk Policy Committee and the Portfolios and Capital Committee and consolidation of their functions into a new Risk Policy Committee ("RPC"). The Board also approved the re-chartering of the OTIC and creation of a new top-level Credit Committee to resolve disputes arising at RPC and OTIC meetings.(1660) (1) Risk Policy Committee The RPC was broadly responsible for overseeing the Company's financial risks. Among other things, the RPC was charged with: (1) discussing and deciding key corporate policies relating to the management of interest rate and credit risk; (2) defining the Company's financial risk tolerance, reviewing risk management strategies adopted by the business units for consistency with this tolerance, and resolving any inconsistencies; (3) ensuring reasonableness of assumptions and consistency of structure in major corporate models used to measure, evaluate, and price financial risk; (4) achieving consensus on risk assessments, capitalization, hurdle rates, and pricing targets for all new products or initiatives that could potentially affect the risk profile of the Company; and - ---------- (1658) Audit Committee minutes set forth the due diligence process for each certification. See, e.g., Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Mar. 12, 2004, FMSE 504834-37, at FMSE 504835; Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Nov. 13, 2003, FMSE 504792-95, at FMSE 504793; Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Aug. 12, 2003, FMSE 504788-91, at FMSE 504789. (1659) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Oct. 17, 2000, FMSE 504009-18, at FMSE 504010-14. (1660) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 23, 2004, FMSE 504353-90, at FSME 504371-73. 435
(5) ensuring adequate and efficient capitalization of the Company and its business units.(1661) Howard chaired the RPC. The members of the RPC were selected by the CFO and approved by the SLT. Besides Howard, members of this Committee included the Executive Vice Presidents for the Mortgage Portfolio and Single-Family businesses, Chief Credit Officer Marzol, e-Business President Williams, and several Senior Vice Presidents. Mudd and Donilon served ex officio.( 1662) In fulfilling its mandate, the RPC was supported by Senior Vice President -- Corporate Financial Strategy Thomas A. Lawler and his Risk Policy group. The full RPC exercised direct oversight of interest rate risk. Other risk policy matters were handled by one of three subcommittees of the RPC: the Credit Risk Subcommittee (chaired by Marzol); the Analytics and Capital Subcommittee (chaired by Lawler); and the Corporate Reporting and Standards Subcommittee (chaired by Spencer). (2) Operations, Transactions and Investments Committee As originally established in 2000, the OTIC was charged with evaluating new investments, large transactions, and new business initiatives, as well as "requests for commitments of resources, monies, or investments that involve[d] cross-sectional departments" within the Company.(1663) The mission of the OTIC expanded in January 2004, when it was charged by the Board with "oversee[ing] the management of all operational risk throughout the Company" and "oversee[ing] security and privacy management."( 1664) - ---------- (1661) Risk Policy Committee Charter, dated Feb. 15, 2004, Zantaz document 738653. (1662) See Overview of Current Risk Policy Organization and Recommendations for Interim Steps in Transition to CRO Organization, dated Jan. 5, 2005, Zantaz document 1331424, at 17; Fannie Mae's Risk Oversight Structure, dated July 30, 2004, Zantaz document 1219244. (1663) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated July 17-18, 2000, FMSE 99407-16, at FMSE 99410; Minutes of the Assets and Liabilities Policy Committee of the Board of Directors of Fannie Mae, dated Oct. 17, 2000, FMSE 504917-504924, at FMSE 504917-18. (1664) Minutes of the Assets and Liabilities Policy Committee of the Board of Directors of Fannie Mae, dated Jan. 22, 2004, FMSE 504986-504994, at FSME 504991; see also Operations, Transactions, and Investments Committee Charter, dated Jan. 13, 2004, Zantaz documents 882230, 882233. The re-chartered OTIC remained responsible for acting as a forum to resolve issues related to transactions or investments. 436
The OTIC was chaired by Mudd. Other members of the Committee included Howard, Donilon, Marzol, Levin, Hoyes, Williams, Kappler, Rajappa, Spencer, and St. John.(1665) After the OTIC was re-chartered in January 2004, it formed two working groups: one for operational risk, and another for security and privacy. The main task completed by the operational risk working group in 2004 was a benchmarking project that compared Fannie Mae's operational risk management practices with those of other large financial institutions (discussed below). (3) Credit Committee The main function of the Credit Committee was to resolve disputes over credit, interest rate, and operational risks that had not been resolved in the RPC or OTIC. Raines presided over Credit Committee meetings, which met quarterly or, when necessary, more frequently. The other members of the Credit Committee consisted of Howard, Mudd, Donilon, and Marzol.(1666) C. Findings 1. Management Did Not Fully Inform the Board of Accounting Issues or Internal Control Deficiencies Our investigation found that management's reports to the Board sometimes omitted critical adverse information and thereby left directors with a false sense of reassurance concerning the Company's policies and practices, the level and nature of internal resources, and the degree to which the Company genuinely met "best practices."( 1667) - ---------- (1665) Overview of Current Risk Policy Organization and Recommendations for Interim Steps in Transition to CRO Organization, dated Jan. 5, 2005, Zantaz document 1331424, at 14, 18. (1666) See id.; An Overview of Fannie Mae's Risk Oversight Committee Framework, dated 2004, Zantaz document 2229168, at 2-3. (1667) In the Board's 2003 self-assessments, all fourteen directors rated their access to management, the quality of management presentations to the Board, and the adequacy of pre-meeting information to be "good" or "very good." See Board of Directors Evaluation - Compilation of Results, dated 2003, FMSE 81544-49. However, some comments by Board members reflected a belief that management presentations should have included more financial and operational detail. See Board Self-Evaluation Summary, dated 2003, FMSE 13872-78, at FMSE 13875. Directors' comments in the 2004 self-assessments reflected a similar combination of overall satisfaction with the quality of management presentations and sporadic 437
For example, we can find no evidence indicating that the Board (including Board committees) was briefed in any meaningful fashion concerning: (1) the existence of significant resource deficiencies in Internal Audit, the Controller's Office, or the Office of Corporate Compliance; (2) the fact that the Company's central compliance offices (the Office of Corporate Compliance and the Office of Corporate Justice) were housed in the litigation section of the General Counsel's office, with a Chief Compliance Officer who managed key litigation functions for the Company (including employment practices litigation); (3) the absence of a meaningful cross-enterprise approach to operational risk management; or (4) the broad absence of documented corporate policies and procedures within the Company.(1668) As a consequence, the Board lacked the information necessary to assess the claims of senior management that the Company possessed a first-rate accounting operation and robust internal control and compliance functions. Specific examples of management communications to the Board regarding significant accounting and financial reporting issues are discussed in other Chapters of this Report. 2. Management Created an Environment that was not Conducive to Open Discussion and Exchange of Views The Company espoused a corporate culture that lauded intellectual honesty, openness, and transparency. Corporate culture was a frequent topic of Frank - ---------- suggestions that heightened financial detail would be helpful. See Board Self-Evaluation Summary, dated 2004, FMSE-KD 26204-12. (1668) We also note that, on two occasions, management received, but it did not share with the Board, recommendations to reform the Board's delegations of authority to management to conform to best practices. An outside law firm first presented these recommendations to management in a 2001 memorandum. See Mem. from O'Melveny & Meyers LLP to A. Kappler and Anthony Marra, dated Nov. 30, 2001, Zantaz document 1217433. A Deputy General Counsel in the Legal Department prepared a second memorandum in 2003 and presented it (in draft form) to senior management. See Draft mem. from T. Marra to T. Donilon, A. Kappler, Monica Medina, and Iris Aberbach, dated Nov. 2003, FMSE-KD 17042-50, at FMSE-KD 17044. It appears that management did not forward the 2003 memorandum to the Board. However, it did inform the Nominating and Corporate Governance Committee in August 2004 that it was undertaking a benchmarking project to review the Board's delegation practices. See Minutes of the Meeting of the Nominating and Corporate Governance Committee of Fannie Mae, dated July 14, 2004, FMSE 505440-43, at FMSE 505440-41. This project soon was overtaken by the events that followed the OFHEO Report in September 2004. 438
Raines's speeches to the business community,(1669) as it was a frequent topic of his weekly messages to employees. For example, in a September 4, 2003 message to employees, Raines encouraged employees to maintain an open working environment: Our openness and intellectual honesty with each other helps to make us the quick, creative, fast, agile, and successful company we are. Just as we aim for best-in-class transparency externally, and welcome the toughest questions from all of our stakeholders, we need to be transparent and address the tough questions internally as well.(1670) Fannie Mae's senior management also discussed from time to time the need to encourage internal honesty and openness. For example, the minutes of a 2003 SLT meeting state that: Mr. Raines asked the other SLT members for advice about how he can encourage employees to ask questions, share doubts, and discuss problems with their managers promptly? How can he get managers to create an atmosphere that encourages employees to share their thoughts more promptly?(1671) Fannie Mae's Code of Business Conduct and associated policies also emphasized the importance of openness and honesty. For example, the Code required managers to "[maintain] a workplace environment that encourages open communication about and compliance with the Code."(1672) Similarly, Fannie Mae's Ethical Responsibility Policy stated that "Fannie Mae's policy is to maintain a corporate culture characterized by openness, integrity, responsibility, and accountability."(1673) This policy required: - ---------- (1669) See, e.g., Remarks Inaugurating the Business Roundtable Institute of Business Ethics at the University of Virginia (Jan. 14, 2004), available at http://www.darden.virginia.edu/corporate-ethics/ media-rainesremarks040114.htm/. (1670) HomeSite News: Frank Asks Us to Speak Up in His Sept. 5 Weekly Message, dated Sept. 4, 2003, FMSE 22935-37, at FMSE 22936. (1671) Minutes of the Meeting of the Senior Leadership Team, dated Sept. 15, 2003, FMSE-IR 273609. (1672) Code of Business Conduct, dated May 11, 2004, FMSE-IR 36806-30, at FMSE-IR 36809. (1673) Fannie Mae Ethical Responsibility Policy, dated June 25, 2004, FMSE-IR 36624-26, at FMSE-IR 36624. 439
all Fannie Mae people managers to create and maintain an open working environment that invites and entertains discussion on a wide range of issues relevant to Fannie Mae's operations and compliance with the Law and the Code. Managers are expected to achieve such an environment by leading by example (1674) Measures were also used to test how well these standards were being met in practice. For example, officers were required to complete regular self-assessments in which they explained how they maintained the appropriate "tone at the top" in their business units.(1675) In addition, Human Resources used outside resources to conduct periodic surveys of employee sentiment to gauge the corporate culture.(1676) Based upon the results of these employee surveys (which, as discussed below, were generally poor), management introduced a variety of measures aimed at improving honesty and openness. For example, after a 1999 employee survey revealed concerns about leadership, communication, and culture, Vice Chair Gorelick outlined a series of corrective measures that included training managers on leadership qualities, requiring them to "attend coaching workshops to . . . ensure open and honest dialogue with their staff," and requiring them to hold breakfast and lunch meetings and open office hours with staff.(1677) After the 2001 employee survey failed to demonstrate a significant improvement in employee sentiment,(1678) management promised to introduce a series of measures that were rather similar to those it promised in 1999, including "regular and interactive employee communications with Dan and Tim," brown bag lunches, walk- - ---------- (1674) Id. at FMSE-IR 36625. (1675) See, e.g., Fannie Mae Self Assessment Questionnaire: Legal Department, dated 2003, Zantaz document 4129242 (asking "[h]ow do you establish the appropriate `tone at the top' and ensure it is followed in your department?"). (1676) See, e.g., Culture Scan Preliminary Results: Fannie Mae, dated Nov. 2004, FMSE-IR 208890-952; Hewitt, EPS Survey Results - Overall Corporate Findings, dated Aug. 2001, FMSE-IR 308768-80. (1677) Mem. from Jamie S. Gorelick to the Operating Committee, dated Feb. 28, 2000, Zantaz document 738059. (1678) See Hewitt, Fannie Mae: EPS Survey Results - Overall Corporate Findings, dated Aug. 2001, FMSE-IR 308768-80, at FMSE-IR 308777; 2001 Employee Perspectives Survey, dated Aug. 23, 2001, FMSE-IR 308757. 440
arounds, and "gab sessions" with employees.(1679) Similar measures were proposed yet again in 2004.(1680) While it is fair to say that speeches, statements, and policies issued by senior management over time emphasized the importance of an open and honest atmosphere, it is not clear that this was the message received by employees. Rather, what was commonly described to us was a set of information "silos" within the Company, with skepticism expressed as to how well those silos connected even at the top of the enterprise. (a) A Culture that Discouraged Dissenting Views, Criticism, and Bad News Mudd acknowledged to us that, generally speaking, Fannie Mae's prior leadership was not good at listening to ideas and issues raised by people at the middle and lower levels of the Company, and senior management meetings often were too formal and scripted and did not present enough opportunity for open debate. Mudd's view was echoed by other members of management. For example, a 2004 working group of Senior Vice Presidents urged senior leadership to "encourage a culture of greater directness in communications." The working group noted that "[i]t is great that we are, as a Company, very polite, but when politeness is a mask for passive/aggressive behavior it retards genuine progress."(1681) These types of observations were widespread in employee responses to a 2004 Human Resources survey (known as the "Culture Scan").(1682) Representative responses included comments that the decisions and styles of senior leadership "cannot be criticized"; that "we don't value dissenting opinions and protect dissenters"; and that "[w]e need a more open culture that welcomes honest open critics and a diversity of thought styles."( 1683) A number of employees complained of having to do things the "Fannie Mae Way," which was described as "not publicly disagreeing and working it out - ---------- (1679) See 2001 Corporate EPS Action Plan, dated Jan. 27, 2002, FMSE-IR 308667-68, at FMSE-IR 308667; Undated EPS Divisional Action Plan, FMSE-IR 308831-32. (1680) See Mem. from Rebecca Senhauser to the Office of the Chairman, dated Apr. 14, 2004, Zantaz document 818391 (proposing efforts to address culture and morale that included a town hall meeting, an employee culture survey, and forums with the Office of the Chairman). (1681) Presentation of Senior Vice President Working Group, dated Mar. 16, 2004, FMSE-IR 468570-577, at FMSE-IR 468577. (1682) See Culture Scan Preliminary Results: Fannie Mae, dated Nov. 2004, FMSE-IR 208890-952. (1683) Id. at FMSE-IR 208899, 913, 919. 441
behind the scenes in an indirect way."(1684) One employee in the Legal Department commented that "[t]he culture of `making nice' really exists. Many folks are not willing/comfortable to tell senior management what they don't want to hear."(1685) Other employees cited difficulties in confronting differences or suggested that candor was lacking because management disliked bad news.(1686) (b) Absence of Cross-Enterprise Teamwork Fannie Mae's corporate culture also was characterized by a lack of teamwork at the senior management level. Chairman Ashley observed the issue as one of compartmentalization of information at senior levels.( 1687) Mudd stated that senior managers tended to defend their territories and to integrate vertically rather than across the enterprise. As far back as 2000, Mudd described a "horizontal problem of silos and a lack of cooperation and flow of information across Fannie Mae."(1688) A number of other senior officers stated that management exhibited a "silo" mentality and did not readily share information among its members. In her 2002 evaluation of Fannie Mae's leadership, Spencer urged Raines to "do more to force team work with his team" because "[t]hey still are not operating as a cohesive unit."(1689) Similarly, at the SLT's December 2003 retreat, certain SLT members commented that they needed to "work together more"; to wear their "enterprise" hats (rather than their "function" hats) more often; and to reduce "competition among SLT members."( 1690) Some SLT members said that they were reluctant to give up some of their decision-making autonomy by bringing issues for decision before the SLT, and others - ---------- (1684) Id. at FMSE-IR 208900. (1685) Id. at FMSE-IR 208913. (1686) See id. at FMSE-IR 208904-14. Raines himself observed during the previous year that "employees sometimes sit on a question for months or years before raising it, or raise it with [the Office of] Corporate Justice or the Office of the Chairman before raising it with their own managers." Minutes of Senior Leadership Team Meeting, dated Sept. 15, 2003, FMSE-IR 273609. (1687) See S. Ashley, 2003-CEO Evaluation Form, FMSE 220631-50, at FMSE 220649. (1688) Remarks at the Mid-Year Officers Meeting: Getting to 2003, dated July 18-19, 2000, FMSE-IR 675181-534, at FMSE-IR 675396. (1689) Officer Year-End Leadership Survey: Rater Feedback Summary on Franklin D. Raines, dated Oct. 2002, FNMSEC 44077-109, at FNMSEC 44095. (1690) Handwritten Notes of Jane Hopkins at Dec. 9, 2003 SLT Offsite Meeting, FMSE-IR 273714-22, at FMSE-IR 273712, 273714, 273718. 442
expressed concern that their decision-making authority would be "hijacked" by the SLT.(1691) Although concerns about coordination and information-sharing appear to have existed across the enterprise, a number of senior officers directed specific criticisms at Howard's Finance organization. These criticisms related not only to relations between Finance and other parts of the Company,(1692) but also within Finance itself.(1693) (c) A Culture of Arrogance The 2004 Culture Scan also contained employee comments from almost every department expressing concern that the outside world perceived Fannie Mae to be "arrogant" and inflexible.(1694) Ashley commented during his interview that he had - ---------- (1691) Id. at FMSE-IR 273717, 273720. (1692) See, e.g., Leadership Perspectives: Report on Tim Howard, dated Nov. 2003, FNMSEC 43958-86, at FNMSEC 43975 (Senior Vice President--Investor Relations Shontell noted that Howard was "proud of finance and it shows. (His pride, in turn, supports an arrogance within Finance towards others in the Company.)."); E-mail from Kathy G. Gallo to F. Raines, dated Apr. 15, 2003, Zantaz documents 827452, 827454 (Gallo summarized comments in Howard's 2003 leadership perspectives survey, which included "Tim could benefit by increasing the amount of networking he does with all of Fannie Mae's constituencies"); Officer Year-End Leadership Survey: Rater Feedback Summary on John T. Howard, dated Oct. 2002, FNMSEC 43939-57, at FNMSEC 43944 (Treasurer Linda K. Knight said that "I think Tim could improve his interactions with the other Executive Vice Presidents to ensure Finance is fully in synch with the other departments."); id. at FNMSEC 43945 (Chief Credit Officer Adolfo Marzol said that "Tim has enjoyed the luxury of the `self-containment' of the Finance division. He does not, for example, have to deal with an independent `chief prepayment risk' officer. For these reasons, I think Tim is good at making change within the Finance division but struggles a little bit more when issues get very cross-functional."). (1693) See, e.g., Leadership Perspectives: Report on Tim Howard, dated Nov. 2003, FNMSEC 43958-86, at FNMSEC 43971 (Controller Spencer said that "employees (below rank of officer) would benefit from hearing from [Howard] more The `I' in him holds him back from being an `initiator' of communication or an `articulator' of vision"; and that "Tim is reluctant to get [the] whole team together and deliver communication and messaging that everyone can hear and understand their part and where they fit in"). (1694) See Culture Scan Preliminary Results: Fannie Mae, dated Nov. 2004, FMSE-IR 208890-952, at FMSE-IR 208912 (Multifamily comments); id. at FMSE-IR 208913 (Legal Department comments); id. at FMSE-IR 208915 (Single-Family comments); id. at FMSE-IR 208916 (comments from Controller's Office, Investor Relations, 443
perceived this arrogance when he served as President of the Mortgage Bankers' Association of America prior to joining Fannie Mae's Board.(1695) Duane Duncan, the Senior Vice President--Government and Industry Relations, acknowledged that many in Congress and at OFHEO held this same perception of Fannie Mae, due to the Company's combative stance towards its critics and its aggressive lobbying and grass roots campaigns. Fannie Mae apparently exhibited this same attitude even to consultants that it hired to critique and help reorganize the Company.(1696) In a 2003 letter to Raines, Rajappa described the problem as follows: Finally, we are perceived to be arrogant. Our own surveys have shown that, my former colleagues now at Wells, B of A, and Citi Mortgage tell me that. We briefly discussed this at Homestead, but more needs to be done on a day-in-day-out basis by everyone who has contact with the public We must learn to listen effectively, answer the question asked, not just deliver the message we have prepared.(1697) - ---------- Credit Finance, and Strategy); id. at FMSE-IR 208918 (Law and Policy comments); id. at FMSE-IR 208921 (e-Business comments). (1695) See also D. Mudd, Remarks at the Mid-Year Officers Meeting: Getting to 2003, dated July 18-19, 2000, FMSE-IR 675181-534, at FMSE-IR 675376-77 (". . . the customer is not always referred to in Fannie Mae in the respectful terms that somebody that is paying our revenues and enabling us to meet our payroll should be."). (1696) In a July 2000 speech at an Officer's Retreat, Mudd chided management for being unwilling to accept criticism: If we're such a great employer, why aren't we facing the reality of the hard questions and the poor ratings that come in some of the employee surveys? We make a sport, it seems, of paying consultants a lot of money, bringing them in, baiting them, laughing at them and then kicking them out and bringing a new set in at some point later . . . [A] lot of those outside observers trying to present us with a mirror reflection of ourselves have said the same thing. And we keep saying that there's something wrong with the mirror and writing a check for that honor. Id. at FMSE-IR 675376. (1697) Letter from S. Rajappa to F. Raines, dated Aug. 19, 2003, at FMSE-IR 503417-18, at FMSE-IR 503417. 444
Raines seemed to acknowledge this problem, at least by late 2004. In talking points for his remarks at a November 2004 officers' meeting, Raines said: [w]e need to be candid about the fact that we are not perfect. We may have believed our own PR a little too much. We allowed ourselves to be arrogant. We thought we had a lot to teach and little to learn from others.(1698) When we asked Raines about Fannie Mae's perceived arrogance, he acknowledged that there were some employees who felt it was easier to dictate terms to others than to treat them as individuals with worthwhile views. However, Raines also noted that charges of arrogance could be seen as extensions of critics' frustrations with Fannie Mae's success.(1699) Fannie Mae's perceived arrogance helped to sour its relationship with its safety and soundness regulator. Mudd stated that many members of senior management believed that OFHEO had no right to question Fannie Mae's policies in light of the Company's years of sustained success.(1700) Board member Kenneth Duberstein observed that senior management, and in particular Raines and Howard, did not respect OFHEO's views and treated OFHEO dismissively. According to Duberstein, management's low regard for OFHEO was not lost on OFHEO's then Director, Armando Falcon, Jr., who complained during a spring 2004 meeting with Duberstein about the arrogance exhibited toward OFHEO by Raines and Howard, and about Fannie Mae's unwillingness to cooperate fully in the Special Examination. Duberstein said that he approached Raines about what he heard during this meeting, and that Raines dismissed it out-of-hand. Fannie Mae's perceived arrogance also extended to Congress, where Fannie Mae had a reputation for lecturing lawmakers and for employing a massive lobbying network to impose its views upon them. The Company's Law and Policy division included a sophisticated lobbying organization that was known to be aggressive. The aggressive tactics appear to have included vigorous use of grass roots and media campaigns.(1701) On occasion, Fannie Mae directed this apparatus toward OFHEO as well.(1702) - ---------- (1698) FDR Context Setting: 2004 Officers Meeting, dated Nov. 18, 2004, FMSE-IR 674388-94, at FMSE-IR 674389. (1699) Raines also said that he made an effort to retrain employees to listen effectively and to answer the questions asked of them. (1700) Mudd also said that some in management resented OFHEO's sudden assertiveness, after years of passivity, in the wake of the Freddie Mac scandal. (1701) See, e.g., Jeffrey H. Birnbaum and Annys Shin, Fannie Mae Dissolving Grass-Roots Lobbying Network, WASH. POST, Sept. 16, 2005, at D01 (discussing Fannie Mae's use of its regional partnership offices to conduct grassroots campaigns to lobby 445
Although such tactics often proved effective in protecting the Company against perceived legislative threats in the short run, the view has been expressed inside the Company and elsewhere that they did long-term harm to important Fannie Mae relationships on Capitol Hill.(1703) 3. Inadequate Management of Corporate Risks (a) Interest Rate and Credit Risk Beginning in 2000, management began consolidating its credit and interest rate risk policy functions under Howard. Howard historically was responsible for overseeing interest rate risk policy, insofar as the Mortgage Portfolio business reported to him.(1704) At Raines's request, Howard assumed oversight of the credit policy function as well in 2002, "in order to put all of the Company's financial risk and strategy responsibilities under a single direct report to the Chairman and CEO."(1705) In 2004, Howard also became chairman of the Risk Policy Committee, which (as noted above) replaced the Credit Risk Policy Committee and set both credit and interest rate risk policy. Howard's responsibility for credit risk policy increased further after Marzol resigned as Chief Credit Officer in the summer of 2004. Senior management decided not - ---------- legislators in their home districts); Bethany McLean, The Fall of Fannie Mae, FORTUNE, Jan. 10, 2005, available at http://www.pbs.org/wsw/news/fortunearticle_20050110_01.html (describing a March 2004 television advertisement directed at defeating a GSE reform bill). (1702) See, e.g., E-mail from W. Maloni to T. Donilon, dated Oct. 31, 2003, FMSE-E 2284626-27 (recommending that Fannie Mae contact friends in Congress "suggesting why more money for an agency that never put enough money into exams wouldn't resurrect them. Have a Democrat nail [OFHEO Deputy Director] Blumenthal, suggesting that he is setting up the Company in the coming accounting audit."). (1703) See, e.g., Jeffrey H. Birnbaum and Annys Shin, Fannie Mae Dissolving Grass-Roots Lobbying Network, Wash. Post, Sept. 16, 2005, at D01. (1704) According to former Chief Credit Officer Marzol, Fannie Mae did not traditionally have a separate interest rate risk policy function. (1705) Mem. from T. Howard to All Employees, dated Aug. 26, 2004, Zantaz documents 1237574, 1237575. As noted above, Marzol began reporting directly to Howard in 2002, after having reported previously to COO Mudd. Marzol stated that he asked Raines not to shift his reporting line to Howard, because much of his work as Chief Credit Officer related to the credit guaranty business, which reported to Mudd. Marzol said that Raines denied his request because he wanted to give a single executive officer oversight of both credit and interest rate risk. 446
to replace Marzol,(1706) and instead divided Marzol's responsibilities among various groups, most of which reported to Howard.(1707) By the late summer of 2004, Howard's authority over financial risk had consolidated to the point where he(1708) and Raines(1709) both referred to Howard as Fannie Mae's "Chief Risk Officer." This process reportedly occurred despite disagreement among senior officers as to whether this combination of functions within Howard's portfolio was good for the Company - and despite concern being expressed by Ashley at the time about the failure to replace Marzol as Chief Credit Officer.(1710) According to the Vice President and Assistant to the Chairman Jill Blickstein, Raines's view was that - ---------- (1706) After Marzol announced his intention to resign as Chief Credit Officer, management discussed the possibility of replacing Marzol, or instead replacing the Chief Credit Officer function, with a more robust chief risk officer function. Management ultimately decided not to pursue either option, but instead to wait and see what types of risk management requirements might be imposed by OFHEO. (1707) As explained by Howard in a memorandum to employees announcing Marzol's resignation, Marzol's responsibilities were divided "among the business units, [Howard] as chair of the Risk Policy Committee, a new group in the finance division called Credit Finance, and Thomas A. Lawler's group, which we will rename Risk Policy to reinforce its connection with the Risk Policy Committee." Mem. from T. Howard to All Employees, dated Aug. 26, 2004, Zantaz documents 1237574, 1237575. (1708) See Tr. of OFHEO Dep. of T. Howard, dated Aug. 5, 2004, at 8:7-8 ("I also serve informally as the Company's chief risk officer"); see also Draft mem. from T. Howard to All Employees, dated Aug. 2, 2004, FMSE-IR 681436 (in this draft of Howard's memo to employees announcing the reorganization of risk management functions, he noted that "[i]n November 2002 the credit policy function moved from single-family into finance, putting all of the Company's financial risk assessment and strategy responsibilities under me as the Company's chief risk officer."). (1709) In an August 2004 e-mail exchange between Vice President and Assistant to the Chairman Jill Blickstein and Senior Vice President--Human Resources Rebecca Senhauser, Blickstein relayed a conversation in which Raines dismissed the need to create a new Chief Risk Officer function because he considered Howard to be the Chief Risk Officer. See E-mail from J. Blickstein to R. Senhauser, et al., dated Aug. 13, 2004, Zantaz document 722473. (1710) See E-mail from D. Mudd to F. Raines, dated Aug. 10, 2004, Zantaz document 723716 (in which Mudd recounted a phone call from Ashley in which the latter asked whether, in failing to replace Marzol, the Company was "replacing accountability with succotash," and expressed the view that "we can't be ad hoc" in addressing important credit decisions). 447
Fannie Mae was unlike other companies in that Howard occupied a more senior position than did most CFOs.(1711) Although there is logic in having the Company's financial risk functions report to a single senior executive, that executive should not have been Howard. The CFO, whose primary responsibility was to oversee the Company's financial reporting, should not have had primary responsibility for overseeing its financial risk policies as well. The convergence of functions within the CFO presented inherent conflicts of interest. The critical checks-and-balances function of having a vibrant, independent risk organization was undermined by the convergence of powers and responsibilities in Howard. Indeed, Fannie Mae benchmarked this issue, and reportedly found no other comparable Company that had a CFO who was also its Chief Risk Officer.(1712) (b) Operational Risk Responsibility for overseeing operational risk resided primarily with COO Mudd.(1713) As discussed above, at the management committee level this responsibility was vested in the OTIC, which Mudd chaired.(1714) In practice, however, it appears that the Company did not have an effective cross-enterprise approach to managing operational risk. In 2004, the OTIC benchmarked Fannie Mae's approach to operational risk as part of a reassessment of Fannie Mae's risk management practices. The OTIC concluded that Fannie Mae's existing operational risk management structure was too decentralized and too uneven in coverage among the business units. The OTIC also concluded that risk management practices were not guided by a standard framework to - ---------- (1711) See E-mail from J. Blickstein to R. Senhauser, et al., dated Aug. 13, 2004, Zantaz document 722473. (1712) See id.; see also E-mail from Barbara Ryan to Scott Lesmes and A. Marzol, dated Aug. 11, 2004, Zantaz document 723440. (1713) Responsibility for managing certain operational risk resided in the Senior Vice President--Operations Risk, Rajappa. Rajappa's main function was to lead Internal Audit, but, as reflected by his title, he had broader responsibilities that included ensuring business continuity and generating and tracking Key Performance Indicators for senior management and the Board. As noted above, Internal Audit was also assigned responsibility for design and implementation of the Company's compliance with SOX 404. (1714) As noted above, broad responsibility across the full range of operational risk issues was not formally set forth in the OTIC's Charter until January 2004. 448
identify, measure, and track risks, and no formal mechanism existed to escalate and resolve problems.(1715) In June 2004, the OTIC recommended to the SLT that the Company approve a plan for operational risk management oversight that included (1) direct oversight of operational risk within the OOC; (2) development of forward-looking and uniform operational risk measures linked to business management and performance; (3) creation of a central unit responsible for enterprise-wide operational risk assessment, reporting, and management; and (4) development of common standards and tools with defined escalation triggers, protocols, and loss/incident tracking.(1716) In July 2004, the OTIC further recommended to the SLT the creation of a new operational risk unit reporting to the COO, and led by a new Senior Vice President--Enterprise Operational Risks. D. Subsequent Developments Fannie Mae's top-level management has undergone dramatic changes in the aftermath of the OFHEO Report. As a result of a series of personnel changes and reorganizations (notably including the separation of the CEO and Chairman functions), the OOC no longer exists as a management institution and senior management structure and tone has changed significantly. The effects of this transformation thus far appear to be positive. First, the relationship between the CEO and the Board, and between the CEO and other senior officers, is healthier. The current and planned reorganization of functions and lines of reporting appears designed to enhance internal controls, management of risks, and the flow of information both vertically and horizontally within the Company. In addition, the new management team, led by the new CEO, appears so far to be exhibiting a "tone at the top" that is more open, collaborative, and humble. This new management team seems to be better positioned than its predecessors to achieve the elusive but important goal of transforming the corporate culture of Fannie Mae. - ---------- (1715) See D. Mudd, Presentation to the SLT on Operational Risk at Fannie Mae, dated June 7, 2004, FMSE-IR 272016-40, at FMSE-IR 272028-29. (1716) See id. at FMSE-IR 272030. 449
1. Personnel and Structural Changes to Senior Management (a) Changes to Chief Executive Officer Position Chairman/CEO Raines retired from the Company in December 2004.(1717) As noted above, the Board decided in the aftermath of Raines's retirement to split the roles of Chairman and CEO, and to name Mudd the interim CEO.(1718) Although the parameters of the new CEO position continue to evolve to some degree, a formal job description was prepared in March 2005. This document is interesting not only because there appears to have been no formal description of the CEO position at Fannie Mae prior to this point in time, but also because of the document's emphasis on (i) communications with the Chairman and the Board, (ii) accounting and financial reporting, (iii) internal controls and compliance, (iv) relations with regulators, and (v) "tone at the top."(1719) In practice as well as on paper, it seems clear that the Company intends to depart from the "CEO-centric" organization that characterized Fannie Mae in the past. As discussed above, Ashley appears to have cast the non-executive Chairman position as a strong independent check upon the CEO. (b) Changes to the Chief Financial Officer Position In December 2004, Vice Chairman and CFO Howard resigned from the Company.(1720) To replace Howard, the Board named Executive Vice President--Housing and Community Development Levin to serve as interim CFO. After an extensive search process, the Board announced its selection of Robert Blakely to serve as CFO in November 2005. Blakely had been serving as Executive Vice President and CFO at - ---------- (1717) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Dec. 21, 2004, FMSE 504507-16, at FMSE 504507; Fannie Mae press release, dated Dec. 21, 2004, available at http://www.fanniemae.com/newsreleases/2004/3405.jhtml. (1718) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Dec. 21, 2004, FMSE 504507-16, at FMSE 504508. As also noted above, after serving in an interim capacity for almost six months, Mudd was named President and CEO by the Board in June 2005. See Fannie Mae press release, dated June 1, 2005, available at http://www.fanniemae.com/newsreleases/2005/3530.jhtml?p=Media&s=News+Rele ases. (1719) See Letter from S. Ashley to A. Falcon, Jr., dated Apr. 19, 2005 FMSE-IR 314138-45 (attachment entitled "Duties of the Chief Executive Officer"). (1720) See Fannie Mae press release, dated Dec. 21, 2004, available at http://www.fanniemae.com/newsreleases/2004/3405.jhtml. 450
MCI, where he oversaw MCI's restatement and the restructuring of its finance, accounting, and controls functions.(1721) Under the new Fannie Mae management structure, Blakely's duties are significantly narrower than Howard's were, and are more typical of a CFO. Blakely's primary responsibility will be to ensure the accuracy, integrity, and timeliness of the Company's financial reporting, accounting, and related control functions, and to oversee the development and enhancement of automated financial reporting systems.(1722) He will report directly to the CEO, and his direct reports will consist of the Senior Vice President and Controller, the Senior Vice President--Accounting Policy, and the head of Investor Relations.(1723) Internal Audit will not report on a dotted line basis to Blakely (as it did to Howard).(1724) Moreover, as discussed below, primary responsibility for overseeing interest rate and credit risk policy now lies with the Chief Risk Officer. Finally, the Mortgage Portfolio business, which previously reported to Howard, now reports to Levin, who (with the appointment of Blakely as CFO) has been named the first Chief Business Officer of the Company. (c) Changes to Chief Operating Officer Position Also in November 2005, the Board named Michael Williams, the President of Fannie Mae's e-Business unit, to serve as COO (as successor to Mudd). As COO, Williams is responsible for operations management, including the "administrative functions in the areas of technology systems, business operations, and human resources."(1725) His direct reports consist of Executive Vice President and CIO St. John, Senior Vice President--Human Resources Rebecca Senhauser, and the Vice Presidents for Corporate Facilities Services, Pricing Restatement, and Regulatory Agreements and Restatement.(1726) - ---------- (1721) See Fannie Mae press release, dated Nov. 10, 2005, available at http://www.fanniemae.com/newsreleases/2005/3641.jhtml?p=Media&s=News+Rele ases/. (1722) See id. (1723) See Officers Organization Chart, dated Nov. 15, 2005, FMSE 537763-64, at FMSE 537763. (1724) As discussed below, the new Chief Audit Executive, Jean Hinrichs, reports to the Chairman of the Audit Committee of the Board, with a limited "dotted line" to CEO Mudd for certain administrative purposes. (1725) See Fannie Mae press release, dated Nov. 10, 2005, available at http://www.fanniemae.com/newsreleases/2005/3641.jhtml?p=Media&s=News+Rele ases/. Blakely joined Fannie Mae in January 2006. (1726) See Officers Organization Chart, dated Nov. 15, 2005, FMSE 537763. 451
Under this new structure, the COO no longer oversees the operation of Fannie Mae's Single-Family Mortgage and Housing and Community Development businesses. Instead, this responsibility belongs to Chief Business Officer Levin. In addition, the COO no longer bears responsibility for overseeing operational risk policy; this task, as explained below, now falls to the Chief Risk Officer. (d) Changes to Law and Policy Group In April 2005, Executive Vice President--Law and Policy Donilon resigned from Fannie Mae.(1727) The Company thereafter eliminated Donilon's position, as well as the Law and Policy group itself. The Company decided to reconstitute the key elements of the Law and Policy group - namely, the Legal Department, Government and Industry Relations, and Communications - as independent organizations reporting directly to the CEO. In addition, as is discussed below, the functions of the Office of Corporate Justice and the Office of Corporate Compliance are in the process of being moved from the Legal Department to the new Office of Compliance, Ethics & Investigations, along with most of the functions from Regulatory Policy. Moreover, we understand that the lobbying activities of Government and Industry Relations have been substantially reduced. (e) Establishment of the Chief Business Officer Position In November 2005, Fannie Mae established a Chief Business Officer ("CBO") position and appointed interim CFO Levin to that position.(1728) As CBO, Levin is responsible for overseeing the Single-Family Mortgage, Housing and Community Development, and Mortgage Portfolio businesses.(1729) According to Levin, one of the central missions in his new role is to eliminate the silos that previously encased the business units and to create in their stead an integrated business that is "one Fannie Mae." Levin has taken steps in this regard, including the establishment of new management-level committees designed to coordinate activities across the business units (e.g., activities relating to capital management and mortgage-backed securities). - ---------- (1727) See Fannie Mae press release, dated Apr. 15, 2005, available at http://www.fanniemae.com/newsreleases/2005/3498.jhtml?p=Media&s=News+Rele ases/. (1728) See Fannie Mae press release, dated Nov. 10, 2005, available at http://www.fanniemae.com/newsreleases/2005/3641.jhtml?p=Media&s=News+Rele ases/. (1729) See id.; see also Officers Organization Chart, dated Nov. 15, 2005, FMSE 537764. As discussed above, the Mortgage Portfolio business reported to the CFO (Howard) prior to 2005, while the Single-Family and Housing and Community Development businesses reported to the COO (Mudd). 452
(f) Establishment of the Chief Risk Officer Position As required by its September 2004 Agreement with OFHEO, Fannie Mae established a separate Risk organization in early 2005.(1730) The Chief Risk Officer ("CRO") oversees officers who have responsibility for interest rate and other market risk, credit risk, and operational risk, as well as risk policy, risk capital, model validation, and analytics. In addition, the business unit risk officers, who report to business unit management, also report on a "dotted line" basis to the CRO, as does the new Price Verification unit in the Controller's Office. At the Board level, the CRO reports on risk issues to the Risk Policy and Capital Committee, and also delivers a risk oversight report to a joint meeting of the Risk Policy and Capital and Audit Committees at least once a year.(1731) At the management level, the CRO reports directly to the CEO, chairs the Risk Policy Committee, and participates in a number of other senior management committees. Although the organizational design of the CRO function continues to evolve, we understand that the CRO expects the Risk organization to grow to approximately 115 employees over the coming year. Some of this staffing represents new functions within the Company, while some of it represents a shift in responsibilities from other parts of the Company - for example, the shift of responsibility for Key Performance Indicators, and SOX 404 design/implementation from Internal Audit. In January 2005, Fannie Mae appointed former Chief Credit Officer Marzol as its interim CRO.(1732) Marzol left this position in November 2005 and was replaced by Deputy CRO Mark Winer, who had been Vice President--Portfolio Strategy in the Company's Mortgage Portfolio business (and had previously overseen - ---------- (1730) See Agreement between Fannie Mae and OFHEO, dated Sept. 27, 2004 (requiring the Board to "direct the appointment of a chief risk officer, to be independent of other corporate responsibilities and to have duties crafted in consultation with OFHEO"); see also Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Feb. 17, 2005, FMSE 503339-50, at FMSE 503340-41 (approving implementation plan and job description for the CRO). (1731) See Audit Committee Charter, dated Nov. 15, 2005, FMSE 547741-45, at FMSE 547745; Risk Policy and Capital Committee Charter, dated Nov. 15, 2005, FMSE 547755-56, at FMSE 547756. (1732) See Fannie Mae press release, dated Jan. 4, 2005, available at http://www.fanniemae.com/newsreleases/2005/3415.jhtml?p=Media&s=News+Rele ases. 453
management of interest rate, liquidity, and market risks at PNC Financial Services).(1733) As of the beginning of 2006, the Company had not yet named a permanent CRO. (g) Changes to Management Committee Structure Fannie Mae's management committee structure remains in flux as this Report is written, with ongoing reassessment and restructuring. Although the outcome and impact thus remain unclear, the changes made and contemplated appear designed to enhance cross-enterprise coordination. Among other changes, the SLT has been replaced by the Executive Committee, which now consists of Mudd's direct reports and those Executive Vice Presidents who are not direct reports to Mudd. The Executive Committee meets twice as frequently as the SLT and, as described to us, is less driven by set agendas and presentation materials, with deliberate opportunity for impromptu discussion. In addition, the OTIC has been replaced by the Business Operations Committee ("BOC"), which is now described as a cross-enterprise forum to discuss compliance, risk, and internal controls issues, including audit issues (internal and external), legal and regulatory issues, OFHEO exams and reports, and reports from the CRO.(1734) The BOC, which meets approximately once a month, has a broad cross-enterprise membership that includes the CEO, CFO, CRO, CIO, Chief Audit Executive, General Counsel, Treasurer, business unit Executive Vice Presidents, and Vice President--Regulatory Agreements & Restatement. 2. Changes to Corporate Culture Upon assuming the role of CEO, Mudd resolved to change the corporate culture at Fannie Mae. For example, Mudd has said that he intends to establish "an attitude of service" towards Fannie Mae's customers and investors.(1735) He also has said that he wants the Company to be less combative toward critics and more open to change; as he put it recently, "[w]e need to realize that not every contrary opinion is a direct - ---------- (1733) See Fannie Mae press release, dated Nov. 3, 2005, available at http://www.fanniemae.com/newsreleases/2005/3637.jhtml?p=Media&s=News+Rele ases/. (1734) See Minutes of the Meeting of the Business Operations Committee, dated July 7, 2005, FMSE-IR 699289 ("Business Operations Committee (formerly OTI)"). (1735) Annys Shin, Mudd Vows to Change Fannie Mae Culture, WASH. POST, June 3, 2005, at D04. 454
attack on the Company."(1736) In addition, Mudd has emphasized the need for Fannie Mae to have a more open and candid internal environment. Past efforts to reform Fannie Mae's corporate culture have not received high marks from employees. As Mudd himself noted in launching a reform initiative back in 2000, Fannie Mae has had a tendency to "make a sport" of hiring consultants to make reform proposals and then disregarding recommendations for change.(1737) Comments we have received from Fannie Mae employees, directors, and regulators suggest a widespread perception that the "tone at the top" has improved since Mudd took over as CEO. Mudd is commonly credited with an open and engaging management style, which encourages senior management to engage in more candid exchanges of views and information than it did before. He also appears to be working to improve communications with the Chairman and the rest of the Board; to break down the "silo" mentality; to increase the resources committed to internal control and compliance functions; and to solicit the views and concerns of employees using town hall meetings, visits to employee offices, and other outreach efforts. Mudd also appears to have made progress in improving Fannie Mae's external relations, including with customers, OFHEO, and Congress. Part of this progress appears to be related to changing both the size and aggressiveness of the Company's lobbying and grass roots activities; and part of it appears to be attributable to the Company's willingness to admit mistakes and to work with OFHEO and Congress in - ---------- (1736) Annys Shin, Fannie Mae's Repairman: CEO Daniel H. Mudd Must Fix Problems With Accounting and Culture, WASH. POST, Nov. 16, 2005, at D1 (also quotes Mudd as saying that, while management's prior approach to criticism was to "shut[] the doors and huddl[e] and navel gaz[e]," his approach now is to "open the doors and get out and ask people bluntly, right up front . . . `What have we gotten wrong?'"). (1737) See D. Mudd, Remarks at the Mid-Year Officers Meeting: Getting to 2003, July 18-19, 2000, FMSE-IR 675181-534, at FMSE-IR 675376. 455
addressing problems.(1738) Even former regulators reportedly have expressed appreciation for the changes in the Company's position and attitude under Mudd.(1739) The demands of OFHEO and of the Board have played a major role in creating the structural and cultural changes that appear to have occurred at Fannie Mae so far. As of the date of this Report, meaningful substantive and tonal changes have occurred at the senior management level of Fannie Mae, and these changes have improved the functioning of the enterprise both internally and externally. III. OFFICE OF AUDITING A. Summary With respect to the functioning of the Office of Auditing ("Internal Audit") prior to release of the OFHEO Report in September 2004, our conclusions are as follows: 1. The head of Internal Audit lacked the requisite expertise and experience, including prior professional experience or training as an auditor, to lead Internal Audit at an organization as large and complex as Fannie Mae. Moreover, on more than one occasion, the head of Internal Audit took steps that suggest he did not fully appreciate his organization's role within the Company or his relationship with senior management. 2. Internal Audit did not possess a sufficient number of auditors with the requisite mix of technical accounting expertise and auditing experience to carry out its responsibilities related to Fannie Mae's increasingly complex business. Although Internal Audit's workload increased substantially in the years prior to 2005, Internal Audit requested only modest increases in headcount. In addition, the department's training programs were inadequate to compensate for these deficiencies. - ---------- (1738) We note that Mudd began his tenure as CEO by acknowledging to Congress Fannie Mae's mistakes and by pledging cooperation with Congress and the Company's regulators. See Regulatory Reform of the Housing Government-Sponsored Entities: Hearing Before the S. Comm. on Banking, Housing, and Urban Affairs, 109th Cong.(2005) (statement of D. Mudd) ("Let me say right here at the outset: Fannie Mae understands that we have disappointed a lot of people - people who count on us to get things right . . . people, such as the members of this Committee, who have a right to expect that our books and our internal controls are above reproach."). (1739) For example, former OFHEO Director Armando Falcon, Jr. recently praised Mudd's approach. Annys Shin, Fannie Mae's Repairman: CEO Daniel H. Mudd Must Fix Problems With Accounting and Culture, WASH. POST, Nov. 16, 2005, at D1 (saying of Mudd: "[h]e seems earnest and well intentioned. He's trying hard to do all the right things and turn the Company around."). 456
3. Internal Audit's communications with the Board and management were deficient and, at times, inaccurate. On a number of occasions, Internal Audit provided assurances to the Audit Committee that Internal Audit's staffing was adequate in terms of quantity and quality (when it had told management otherwise) and that it had audited Fannie Mae's accounting for compliance with GAAP (when it actually audited only for compliance with Fannie Mae policies interpreting GAAP). In addition, Internal Audit's reporting of its audit issues to the Audit Committee (and to members of senior management) lacked clarity and did not succinctly prioritize the findings or the subsequent remediation. 4. The assignment of the lead on SOX 404 design/implementation to Internal Audit was poorly conceived and executed. It caused a substantial diversion of Internal Audit resources away from performing audits, and produced results that were largely discarded by the Company's new Risk organization and its new external auditor. The Audit Committee and senior management have acted to address many of these deficiencies. They have taken steps to replace Internal Audit's leadership, restructure its organization, focus its responsibilities on its core audit mission, and reform its processes and procedures. Substantial progress is underway in each of these areas. B. Overview Prior to September 2004 1. Functions and Responsibilities Internal Audit's official functions and responsibilities were set forth in the Internal Audit Manual, which included the Internal Audit Charter.(1740) Internal Audit's mission statement provided that it was to (1) "[c]ontribute to the achievement of Fannie Mae's objectives, by promoting effective management of financial and operational risks and maintenance of an effective internal control environment"; and (2) support the Board of Directors by providing independent and objective audit, assurance, and risk assessment services, and the prompt communication of audit and follow-up results to senior management and the Board.(1741) - ---------- (1740) See Fannie Mae Office of Auditing: Audit Manual, dated 2004, Zantaz document 574585. (1741) See, id. at 8. In addition, Internal Audit was subject to (1) the "Standards for the Professional Practice of Internal Auditing" and the "Code of Ethics" promulgated by the Institute of Internal Auditors; (2) the standards set forth by the Committee on Sponsoring Organizations in its publication, "Internal Control - Integrated Framework"; (3) the standards for information technology control set forth in "Cobit Control Objectives"; and (4) "Generally Accepted Auditing Standards" published by the American Institute of Certified Public Accountants. Id. at 8-10, 19. 457
2. Organization and Structure (a) Senior Vice President--Operations Risk From January 1999 through January 2005, Rajappa headed Internal Audit.(1742) Prior to heading Internal Audit, Rajappa was Fannie Mae's Controller from 1994 through 1998. Rajappa did not have a formal background in auditing or training as an auditor.(1743) According to Rajappa, former COO Small asked him to take the reins of Internal Audit after he made comments to Small regarding a lack of checks and balances at Fannie Mae. Rajappa assumed the title of Senior Vice President--Operations Risk. In addition to Internal Audit, he was also responsible for managing certain elements of operational risk, including Fannie Mae's Y2K preparedness program and its business continuity planning. Under Rajappa's leadership, Internal Audit's responsibility included certain operational risk tasks, such as developing and monitoring the Company's "Key Performance Indicators" ("KPIs") and design/implementation of SOX 404. During his tenure as head of Internal Audit, Rajappa had a direct reporting line to the Audit Committee. He attended Audit Committee meetings and briefed the Committee on such topics as the development and implementation of the annual audit plan and SOX 404 compliance measures (in combination with KPMG); remediation of audit issues; monitoring of KPIs; and the adequacy of Internal Audit's resources. He occasionally provided reports to the Audit Committee on the results of discrete audits, such as the annual Derivatives Control Audit. He also communicated regularly with Audit Committee Chairman Gerrity. Rajappa had a "dotted line" reporting relationship to senior management, initially to the COO (Small followed by Mudd) and subsequently (from 2002 on) to CFO Howard.(1744) Rajappa stated that he met with Howard on a monthly basis, and also met - ---------- (1742) See Fannie Mae press release, dated Nov. 17, 1998, available at http://www.fanniemae.com/newsreleases/1998/0072.jhtml (noting Rajappa's transition to Senior Vice President--Operations Risk, effective on Jan. 1, 1999); Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 18, 2005, FMSE-IR 422501-15, at FMSE-IR 422512 ("Mr. Gerrity reported that the [Audit] Committee had decided to move the current head of Internal Audit out of that position."). (1743) Rajappa holds a bachelor's degree and a master's degree in mechanical engineering and an MBA in finance. Prior to joining Fannie Mae, he worked in the controller's offices of several companies. (1744) Rajappa recalled having reservations about reporting to the CFO from a best practices perspective, and having expressed such reservations to Howard. But he also recalled having an understanding at the time that OFHEO had approved this reporting relationship. 458
with him on an ad hoc basis as necessary to discuss such administrative issues as staffing, vacations, and expense reports. Howard also addressed certain budget and personnel review issues with Rajappa, as Howard reviewed and approved Internal Audit's budget (in combination with Human Resources, the Controller's Office, and the COO), and prepared Rajappa's annual performance review (in consultation with Audit Committee Chairman Gerrity). Rajappa also stated that Howard was involved in the development and implementation of Internal Audit's approach to SOX 404 compliance. According to Rajappa, however, Howard did not exercise substantive oversight of Internal Audit and did not exert influence over the annual audit plan. (b) Vice President--Audit Rajappa's deputy was the Vice President--Audit, Ann Eilers. Eilers assumed the Vice President position in 2003, having served previously as Director of Internal Audit's Finance group for several years.(1745) Eilers's role was described to us as akin to a chief operating officer of Internal Audit. She managed Internal Audit's day-today operations, ensured that the Internal Audit directors properly executed the annual audit plan, and oversaw Internal Audit's SOX 404 compliance efforts. She regularly joined Rajappa at Audit Committee meetings, and they both participated in Internal Audit's monthly meeting with KPMG. (c) Auditing Groups Various Internal Audit directors who reported to Eilers headed groups responsible for executing the annual audit plan. As of September 2004, Internal Audit had five groups responsible for auditing Fannie Mae's businesses: (1) the Finance group (which audited Portfolio and Treasurer's Office operations and performed the annual derivatives audits); (2) the Technology group (which audited business applications and technology infrastructure); (3) the Housing and Community Development group (which audited the Housing and Community Development business); (4) the Single-Family Business group (which audited the guaranty fee business and conducted lender audits); and (5) the Credit and Portfolio Analytics group (which audited financial models, credit risk and portfolio strategy functions, and the capital management area in the Treasurer's Office).(1746) (d) Audit Activities According to employees, Internal Audit policy required that business processes determined to be high risk (i.e., high inherent risk and high control risk) were - ---------- (1745) Eilers joined Fannie Mae in 1997, serving initially as an auditor in Internal Audit's Housing and Community Development group. (1746) A new group, Accounting and Auditing Standards, was established in late 2004. This group was assigned responsibility for auditing the development of the Company's accounting policies. 459
to be audited annually; those determined to be low risk (i.e., low inherent risk and low control risk) were to be audited every three years; and other processes were to be audited between every one and three years. Audit teams had discretion, however, to adjust risk profiles based on past audit findings and their knowledge of risks in their assigned business units. After consultation and coordination with KPMG, an annual audit plan was presented to the Audit Committee jointly with KPMG. The audit plan linked each audit project to one or more of the Company's top risks, and provided an estimate of the staffing and budgetary resources required to execute the plan. Internal Audit would then update the Audit Committee from time to time as to its progress in executing the annual audit plan.(1747) Once an audit was complete, an audit report was prepared and the business unit subject to audit was given an opportunity to comment. If the business unit disputed the audit findings, it could raise its concerns with the auditor-in-charge, with the Director, or with Eilers or Rajappa. In rare cases, disagreements were reportedly elevated to COO Mudd or CFO Howard. However, no Internal Audit employee we interviewed recalled modifying the substance of audit findings in response to business unit concerns. Audit reports were labeled with color codes to communicate the nature of Internal Audit's findings. "Red" indicated deficiencies requiring immediate management attention; "yellow" indicated controls that needed strengthening; and "green" indicated that controls were effective. Internal Audit transmitted copies of red and yellow audit reports to Raines and Howard. After issuance of an audit report, Internal Audit monitored the business unit's effort to address any audit issues and reported the status of such remediation efforts using two reports, both of which went to senior management and the first of which also went to the Audit Committee. The first was a monthly report called the "Audit Tracking List" ("ATL"), which identified and monitored "higher risk" control issues.(1748) The second was a quarterly report called the "Audit Issues Follow-Up" ("AIF"), which tracked the status of all open audit issues. - ---------- (1747) See, e.g., Office of Auditing and KPMG: 2003 November Update, dated Nov. 17, 2003, FMSE 15583-91; Office of Auditing and KPMG: 2003 Mid-Year Status, dated July 15, 2003, FMSE 15944-50. (1748) Audit deficiencies were entered on the ATL if they were (1) likely to cause a financial loss equal to or greater than $1 million; (2) likely to cause harm to Fannie Mae's external customers or negatively impact Fannie Mae's public image; (3) likely to prevent Fannie Mae from meeting corporate goals; (4) believed to violate the Charter, Code of Business Conduct, or laws and regulations; or (5) believed to require the attention of the Audit Committee. Fannie Mae Office of Auditing: Audit Manual, dated 2004, Zantaz document 574585, at 46-47. 460
(e) Other Activities (1) Certification of Internal Controls As discussed above, beginning in the third quarter of 2002, Internal Audit assisted the Chairman/CEO and the CFO in preparing their quarterly certifications as to the effectiveness of Fannie Mae's internal controls. These certifications were required by Section 302 of SOX, and they accompanied Fannie Mae's quarterly and annual financial disclosures. On a quarterly basis, Internal Audit received "Internal Control Representation Letters" from the heads of the business units; it then provided Raines and Howard with a certification as to the adequacy and effectiveness of the Company's internal controls.(1749) (2) Self Assessment Questionnaires Internal Audit sought to monitor the Company's internal control environment throughout the year by means of "Self Assessment Questionnaires" ("SAQs"). SAQs asked business unit managers how they maintained an effective control environment, and asked them to self identify internal control weaknesses of which they were aware. Responses were prepared by business unit managers in coordination with Internal Audit, with Internal Audit investigating any responses that indicated control weakness or high-risk activities. (3) Key Performance Indicators Internal Audit also monitored "significant business or operational activities and trends" throughout the year through so-called "Key Performance Indicators" ("KPIs"). Internal Audit generated KPIs by working with the business units to identify key operations and to establish thresholds against which to measure performance. Internal Audit provided monthly KPI reports to the Audit Committee and senior management. - ---------- (1749) See, e.g., Mem. from S. Rajappa to F. Raines and T. Howard, dated Aug. 2004, Zantaz document 545500; Mem. from S. Rajappa to F. Raines and T. Howard, dated May 2004, Zantaz document 545031; Mem. from S. Rajappa to F. Raines and T. Howard, dated Aug. 2003, FMSE 83158. 461
(4) Sarbanes-Oxley Section 404 Design/Implementation(1750) As noted above, Internal Audit was assigned the task of leading the Company's SOX 404 compliance effort.(1751) According to Rajappa, Internal Audit received this assignment because it already had extensive knowledge of Fannie Mae's internal controls through its audit work, and because it had become familiar with SOX 404 through research it had conducted. To support this substantial assignment, Internal Audit obtained $1 million in supplemental funding, the majority of which was allocated to hiring outside contractors to document controls and business processes, and to generate flow charts of controls.(1752) Nonetheless, much of the SOX 404 work fell on Internal Audit personnel, who also had responsibility for their regular audit workload. Deficiencies identified during testing of the controls were placed on a tracking list known as the "S-Ox Issues Follow-Up" ("SIF"). Rajappa updated the SIF monthly and also periodically updated the Audit Committee on the progress of Internal Audit's SOX 404 work.(1753) Internal Audit also classified internal control deficiencies based on their seriousness and likelihood to occur.(1754) Internal Audit reported "Significant Deficiencies" (i.e., internal control problems that had more than a remote - ---------- (1750) Section 404 requires issuers to provide, as part of their annual report to the SEC on Form 10-K, an assessment by management "of the effectiveness of the internal control structure and procedures of the issuer for financial reporting," with annual testing and attestation by the issuer's external auditors as to both management's methodology and its results. 15 U.S.C. Section 7262 (2002). (1751) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 19, 2004, FMSE 504838-46, at FMSE 504840 ("Mr. Rajappa added that the Office of the Chairman had asked internal audit to lead all project management efforts for the Company in the area of S-Ox 404 implementation."). (1752) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 19, 2004, FMSE 504838-46, at FMSE 504840; Sarbanes-Oxley Section 404 Compliance Project: Project Methodology and Management, dated Mar. 14, 2004, FMSE-IR 467104-270, at FMSE-IR 467110. (1753) See, e.g., Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated July 19, 2004, FMSE 504860-69, at FMSE 504861-62. (1754) See, e.g., Office of Auditing and KPMG, 2004 Mid-Year Status, dated July 19, 2004, FMSE 224862-70, at FMSE 224867-68. 462
likelihood of occurrence and a dollar impact that could be more than inconsequential) to senior management and to the Audit Committee.(1755) (5) Consulting Projects Internal Audit also performed occasional consulting projects for business units.(1756) This was one of Rajappa's initiatives, as he wanted Internal Audit to have "a strong internal consulting role in addition to internal audit role."(1757) Typically, these projects did not result in an audit opinion. 3. Past Evaluations of the Office of Auditing Prior to the end of 2004, Internal Audit regularly received praise for the manner in which it conducted its work. OFHEO's examinations and annual reports to Congress commonly concluded that Internal Audit's practices were "effective" and/or exceeded safety and soundness standards.(1758) Internal Audit also received high marks in - ----------- (1755) See, e.g., S. Rajappa, 404 Implementation Plans: Senior Leadership Team Presentation, dated Oct. 11, 2004, FMSE-IR 1606-77, at FMSE-IR 1610; Office of Auditing and KPMG, 2004 November Update, FMSE-IR 143534-44, at FMSE-IR 143539. (1756) See Fannie Mae Office of Auditing: Audit Manual, dated 2004, Zantaz document 574585, at 159-60. (1757) Mem. from S. Rajappa to D. Mudd, dated Sept. 28, 2000, FMSE-IR 537926-29, at FMSE-IR 537928. (1758) For example, OFHEO's June 2003 Report to Congress stated that: Internal audit appropriately identifies and communicates control deficiencies to management and the Board of Directors. There are established policies and procedures that delineate internal control process and standards for the control environment. Management effectively ensures compliance with established internal controls The audit functions are independent and effective and adhere with the standards that are evolving. The Internal and External Audit functions have the appropriate independence. Auditors performing the work possess appropriate professional proficiency. The scope of internal audit work performed is appropriate, and the internal audit work is complete. The management of the Internal Audit department is effective. The Board of Directors are appropriately involved with and follow up on identified audit issues. The auditor's risk assessment process is 463
a 2002 evaluation by the Investment Training and Consulting Institute.(1759) In addition, the Institute of Internal Auditors cited Internal Audit's SAQ and KPI processes as examples of best practices in 2002.(1760) C. Findings Notwithstanding the praise it received, Internal Audit suffered from a number of deficiencies with respect to its leadership, its relationship with management, its resources, and its communications with management and with the Audit Committee. 1. Leadership As noted above, Fannie Mae appointed Rajappa to head Internal Audit even though he had no experience or formal training as an auditor. Board minutes do not reflect any presentation or discussion concerning Rajappa's qualifications when he was appointed Senior Vice President--Operations Risk,(1761) but both Raines and Small stated that they thought he was qualified to oversee Internal Audit.(1762) Given the importance of the internal audit function in a Company as large, complicated, and heavily regulated as Fannie Mae, the Company would have been better served with leadership by a professional auditor with significant expertise and experience in that field. - ---------- effective. Internal Audit is appropriately involved in new products and new initiatives. OFHEO Report to Congress, dated June 2003, FMSE-IR 1180-84, at FMSE-IR 1183. (1759) The Investment Training and Consulting Institute gave Internal Audit its highest overall rating of "generally conforms," concluding that "Fannie Mae's Internal Audit Team has developed and supports many leading practices throughout their processes." Quality Assessment Review: Fannie Mae Internal Audit Department, dated Sept. 2002, FMSE-IR 467089-103, at FMSE-IR 467091. (1760) See Institute of Internal Auditors Research Foundation, Adding Value: Seven Roads to Success (2002), FMSE-IR 990-97. (1761) See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Nov. 17, 1998, FMSE 6-18, at FMSE 15; see also Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 16, 1999, FMSE 504603-17, at FMSE 504603-04. (1762) Raines stated he was unaware that Rajappa lacked certification as an auditor, but that the absence of such a qualification did not concern him. 464
2. Relationship with Management During our investigation, we have found several instances that raise questions regarding Internal Audit's relationship with senior management and the business units it was responsible for auditing. Although we have not identified evidence that Internal Audit compromised its audit work, there is troubling evidence that Internal Audit's leadership viewed itself as accountable to senior management, rather than the Audit Committee. In addition, Internal Audit fostered an environment of collaboration with the business units, rather than objective assessment. We view these instances, as described below, in the context of Rajappa's relationship to Howard. It is neither unusual nor inherently improper for the head of internal audit to report to a member of senior management on a "dotted line" basis for administrative purposes, and many companies have chosen to have this reporting line go to the CFO.(1763) However, Rajappa had been Howard's direct subordinate for several years while Rajappa was Controller. The change in Rajappa's position, therefore, required that both he and Howard acknowledge their new reporting relationship. It does not appear that Rajappa fully made, or understood, the implications of that transition.Rajappa sought to expand his role beyond the leadership of Internal Audit and into management functions. As Senior Vice President -- Operations Risk, he proposed the establishment of an operations risk oversight committee that would create a framework for the identification and analysis of operation risks; according to Rajappa's proposal, the committee would have a "strong functional reporting to the office of the chairman" in addition to the Audit Committee.(1764) When Rajappa was asked about the possible conflict of interest between his role with the Committee and his role as head of Internal Audit, he responded: "...[N]ot sure; company's call. Ann [Eilers] is the auditor, I am ops. risk, she reports to me, we both report to the AC of the Board. I think there's enough independence."(1765) On another occasion, Howard (with Rajappa's apparent acquiescence) treated the head of Internal Audit as a direct subordinate on matters relating to communications with the Audit Committee. Rajappa was contacted directly by Audit - ---------- (1763) See, e.g., Presentation, Institute of Internal Auditors Global Auditing Information Network, Reporting Lines and Scope of Work (July 6, 2003), available at http://www.theiia.org/?doc_id-4366 (noting that, of 303 chief audit executives surveyed, approximately forty-two percent reported administratively to the CFO, thirty-two percent reported to the CEO/President, and the remainder reported to others). (1764) S. Rajappa, Operations Risk - Oversight, dated Mar. 6, 2000, FMSE-IR 563552-66, at FMSE-IR 563567. (1765) E-mail from S. Rajappa to J. Morgan Whitacre, dated Mar. 4, 2004, FMSE-E 699855. 465
Committee Chairman Gerrity regarding OFHEO's finding in 2004 that the Company had numerous outmoded manual accounting systems. Rajappa responded to Gerrity. Howard heard what Rajappa had done and made the following report to Spencer: I just got off the phone with Sam. I made it "blisteringly clear" to him that on any future calls he gets from the Chairman of our Audit Committee on accounting-related issues he must run the question or issue by you before he or anyone else gets back to Gerrity. He said he got the message and would do so in the future. (He said that he'd agreed to call Tom back before realizing you weren't here on Friday, so he spoke with Janet instead. I responded that going forward that wouldn't be good enough. He had to reach either you or me before responding to Gerrity on any accounting-related question he was asked. He said he understood.)(1766) In our interview with Gerrity, he said that he had no knowledge of any limitation on Rajappa's ability to communicate with him directly. He said that he would have been "shocked" to learn that Rajappa had to clear communications through senior management. Structurally, the Company's bonus compensation program treated Internal Audit as it did any other business unit, with bonuses (for directors and above) linked to earnings per share targets. Although we found no evidence to suggest that management intended this compensation system to affect the aggressiveness of Internal Audit, or that it led auditors to minimize audit issues, this compensation structure was not conducive to maintaining the appearance and fact of Internal Audit's independence and objectivity.(1767) - ---------- (1766) See E-mail from T. Howard to L. Spencer, dated Mar. 3, 2004, FMSE-E 1808128-29. (1767) A 2004 study by compensation consultants, which focused on compensation of the Company's management level employees, recommended (among other things) that bonus compensation for the head of Internal Audit should not be tied to earnings. See Fannie Mae and Semler Brossy Report to the Compensation Committee: Appropriate Compensation Structure and Incentives for Fannie Mae Management, dated Feb. 23, 2005, FMSE-IR 241780-946, at FMSE-IR 241863, 241901. We believe Internal Audit's annual incentive compensation should be based on individual and departmental goals unrelated to corporate earnings, including but not limited to training, achievement of audit plan, and retention. A comprehensive performance-based compensation process would have enabled the Audit Committee to assess Internal Audit more accurately by measuring its performance in relation with the critical functions of Internal Audit. 466
On one occasion, Rajappa emphasized to Internal Audit staff the link between compensation and hitting earnings targets. For example, in early 2000, Rajappa made a presentation to Internal Audit staff entitled "Address To Audit Group On What We Can Do To Help Achieve $6.46 EPS." The talking points for that address included the following: By now every one of you must have 6.46 branded in your brains. You must be able to say it in your sleep, you must be able to recite it forwards and backwards, you must have a raging fire in your belly that burns away all doubts, you must live, breathe and dream 6.46. You must be obsessed on 6.46 . . . After all thanks to Frank we all have a lot of money riding on it.(1768) Although these talking points also instructed Internal Audit staff to "be fair but tough where there are serious control issues" and to "never compromise or dilute your conclusion even if your customers get mad at you," Rajappa's message can, at best, be described as mixed.(1769) Moreover, Rajappa forwarded his presentation to Raines (not Gerrity) with a handwritten note saying "I thought you may be interested in knowing what I told my Audit group about how they can help achieve $6.46."(1770) Although Rajappa and Eilers informed us that Rajappa's relationship to Howard was in fact only administrative in nature,(1771) these incidents suggest a disturbing pattern in which the "dotted line" nature of Rajappa's reporting relationship to Howard, and Internal Audit's relationship to senior management, was not understood or honored. Similarly, Internal Audit's leadership did not try to maintain an appropriate distance from the business units, notwithstanding the critical internal control functions for which they had responsibility. Rajappa stated that he and his staff deliberately pursued good relationships with the business units. Internal Audit employees generally insisted that this cooperative approach did not compromise their independence, but there was broad acknowledgment that it caused Internal Audit to be perceived as "business friendly," perhaps at the expense of objectivity. In a January 2005 review of - ---------- (1768) Address to Audit Group on What We Can Do to Help Achieve $6.46 EPS, dated Feb. 16, 2000, FMSE-IR 265330-32, at FMSE-IR 265331. (1769) Id. at FMSE-IR 265330-32. Rajappa was unable to identify for us any steps he took to ensure that Internal Audit staff did not misconstrue his remarks. (1770) Id. at FMSE-IR 265330. Raines stated that he did not view Rajappa's remarks as inappropriate because they simply informed Internal Audit staff that they have a role to play in advancing corporate goals. (1771) Both Rajappa and Ann Eilers stated that Howard approved such things as Internal Audit's staffing, vacations, and expense reports. 467
Internal Audit's operations, E&Y provided the Audit Committee with a list of recommendations. One recommendation was that "Internal Audit personnel should push back more with management, be tougher and sharper in exchanges." E&Y reported that "there was no evidence of management coercion, but rather it appeared that Internal Audit was working with management in a collaborative approach that could be more independent."(1772) Although there is no benefit to hostile relations between an internal audit department and the business units to be audited, the evidence strongly suggests that Internal Audit's "business friendly" philosophy undercut perceptions of Internal Audit's toughness and effectiveness.(1773) 3. Resource Deficiencies During Rajappa's tenure as Senior Vice President -- Operations Risk, Internal Audit's workload increased significantly as Fannie Mae's business grew larger and more sophisticated. During this period, Internal Audit also faced new demands arising from Fannie Mae's reporting to the SEC and its assumption of responsibility for SOX 404 design/implementation. Internal Audit staffing was not commensurate with these demands, in terms of numbers or qualifications. (a) Headcount For most of Rajappa's tenure, staffing levels in Internal Audit were essentially stagnant. Between 1999 and 2002, authorized headcount increased by only one employee, while actual headcount increased by only three employees.(1774) In 2003, Rajappa sought only a modest increase in authorized headcount (three employees), even while informing management that (i) in the previous five years, virtually no new auditors had been added to accommodate increases in "outputs by our department . . . in audits (120 [vs.] 70) and reporting (Audit Issues Tracking, Dashboards, Risk Assessment Model and Audit Plan, etc.)," and (ii) staffing constraints were forcing Internal Audit to limit the scope of its reviews, to combine, postpone, or cancel audits, to delay department - ---------- (1772) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Jan. 17, 2005, FMSE 516154-65, at FMSE 516162-63. (1773) For example, the director of the Office of Corporate Compliance said that she wished Internal Audit's auditors were more like "junkyard dogs." (1774) During this period, authorized headcount increased from 53 to 54, while actual headcount increased from 49 to 52. See Fannie Mae Office of Auditing: Completion of 1999 Audit Plan, dated Feb. 15, 2000, FMSE 14703-12, at FMSE 14712; Office of Auditing and KPMG: Joint Audit Plan 2002, dated Apr. 16, 2002, FMSE 15305-12, at FMSE 15309; Office of Auditing and KPMG: Joint Audit Plan 2003, dated Apr. 14, 2003, FMSE 16114-20, at FMSE 16117. 468
initiatives such as updates to the audit manual, and to work in some cases excessive overtime.(1775) At the beginning of 2004, despite Internal Audit's recognition that staffing shortages would require a reduction in its audit activities for 2004,(1776) and despite the challenges posed by Internal Audit's new responsibility for SOX 404 design/implementation, the 2004 audit plan called for no further increase in authorized headcount.(1777) Instead, the 2004 audit plan reduced the number of planned audits to half of what was planned for 2003.(1778) It was not until the middle of 2004 that Rajappa requested from senior management another increase in authorized headcount.(1779) Internal Audit's authorized headcount increased from fifty-seven to sixty-five employees during the second half of 2004 (although actual headcount only increased from fifty-three to fifty-eight employees).(1780) Nonetheless, Internal Audit had difficulty completing its tasks for the year. By the end of 2004, Internal Audit had cancelled five audits and postponed or delayed completion of 11 audits (despite its substantially scaled back audit plan for 2004).(1781) Whereas Internal Audit issued ninety-six audit reports and - ---------- (1775) Undated Head Count Request for 3 FTE in 2003 (appears to have been drafted in February 2003), FMSE-IR 284780-82, at FMSE-IR 284781-82. (1776) See Fannie Mae Office of Auditing: 2004-2005 Audit Plan Time Line, dated Oct. 2003, Zantaz document 2912971, at 2 ("When selecting the audits for 2004, keep in mind the number of audits that had to be recast this year [2003] due to resource issues."). (1777) Compare Office of Auditing and KPMG: Joint Audit Plan 2004, dated Apr. 19, 2004, FMSE 223834-45, at FMSE 223842, with Office of Auditing and KPMG: Joint Audit Plan 2003, Apr. 14, 2003, FMSE-IR 646655-62, at FMSE-IR 646658. (1778) Compare Office of Auditing and KPMG: Joint Audit Plan 2003, Apr. 14, 2003, FMSE-IR 646655-62, at FMSE-IR 646656 (stating plans to conduct more than 100 audits), with Office of Auditing and KPMG: Joint Audit Plan 2004, dated Apr. 19, 2004, FMSE 223834-45, at FMSE 223837 (stating plans to conduct fifty-four audits). (1779) See Mem. from S. Rajappa to T. Howard, D. Mudd, L. Spencer, and R. Senhauser, dated June 18, 2004, FMSE-IR 423231-35, at FMSE 423233 (requesting authority to hire six new employees). (1780) See Office of Auditing: Completion of 2004 Internal Audit Work, dated Jan. 17, 2005, FMSE-IR 219782-90, at FMSE-IR 219789. (1781) See id. at FMSE-IR 219788. 469
memoranda in 1999, and more than one hundred in each year from 2000 to 2003,(1782) output declined dramatically in 2004 to forty-four audit reports and memoranda.(1783) (b) Qualifications and Experience Internal Audit not only lacked a sufficient number of employees, but also lacked employees with certain critical skill sets. In particular, while a significant number of Internal Audit employees possessed advanced degrees and professional certifications,(1784) employees inside and outside Internal Audit observed that Internal Audit's accounting expertise was uneven if not inadequate.(1785) Due to the limited number of employees with public accounting background or equivalent experience at other financial institutions, Internal Audit did not have the necessary resources to audit effectively aspects of Fannie Mae's business operations that were heavily influenced by accounting principles and policies.(1786) As noted above, Internal Audit did not have a - ---------- (1782) See Fannie Mae Office of Auditing and KPMG: Completion of the 2003 Audit Plan, dated Jan. 23, 2004, FMSE 221354-61, at FMSE 221356-57; Fannie Mae Office of Auditing and KPMG: Completion of the 2002 Audit Plan, dated Feb. 21, 2003, FMSE 16278-85, at FMSE 16282; Fannie Mae Office of Auditing: Completion of the 2001 Audit Plan, dated Feb. 19, 2002, FMSE 15517-21, at FMSE 15520; Fannie Mae Office of Auditing: Completion of 2000 Audit Plan, dated Feb. 20, 2001, FMSE 15058-63, at FMSE 15061; Fannie Mae Office of Auditing: Completion of 1999 Audit Plan, dated Feb. 15, 2000, FMSE 14703-12, at FMSE 14711. (1783) See Office of Auditing: Completion of 2004 Audit Work, dated Jan. 17, 2005, FMSE-IR 219782-90, at FMSE-IR 219788. (1784) Between 2002 and 2004, the number of Internal Audit employees holding advanced degrees increased from fifty-five percent to fifty-eight percent, and the number holding certified internal auditor, certified public accountant, or certified information systems auditor certifications increased from fifty-one percent to sixty percent. See Office of Auditing and KPMG, Joint Audit Plan 2004, dated Apr. 19, 2004, FMSE 223834-45, at FMSE 2233842; Office of Auditing and KPMG, Joint Audit Plan 2002, dated Apr. 16, 2002, FMSE 15305-12, at FMSE 15312. Eilers and Rajappa reported to us in January 2005 that twenty to thirty percent of Internal Audit's auditors were certified public accountants. (1785) For example, Wall cited poor or average accounting expertise of several audit directors. Karen McBarnette, the Director for the Single-Family business audit group, stated that Internal Audit needs to hire more individuals with accounting expertise. (1786) For example, Leah Malcolm, the Internal Audit Director who oversaw the annual derivatives audits, stated that Internal Audit did not identify problems with the Company's derivatives accounting policies against which it conducted its audits because no one in Internal Audit had sufficient FAS 133 expertise. 470
group with assigned responsibility for auditing the development of the Company's accounting policies until late 2004. Uneven accounting and auditing expertise also created key person dependencies. For example, in June 2004, Rajappa warned senior management that Internal Audit only had one employee with the public accounting background necessary to audit the Financial Reporting function in the Controller's Office, and that public accounting skills "cannot be cross trained given that it takes many years to acquire knowledge.(1787) Rajappa also noted that only one level 5 employee was qualified to audit all aspects of Fannie Mae's derivatives operations.(1788) Moreover, a significant majority of the auditors had no prior audit experience before joining Internal Audit.(1789) Recognizing that workload demands undoubtedly presented challenges to improving training programs, neither the deficiencies in skill sets within Internal Audit nor the changing external expectations for public companies appear to have generated a sense of urgency with respect to ongoing training of Internal Audit staff. Internal Audit directors expressed concern that management had historically placed too little emphasis on audit training, with insufficient allocation of time for this purpose.(1790) In July 2003, when Rajappa received an e-mail from an Internal Audit employee suggesting the need to toughen continuing education requirements for internal auditors, Rajappa responded that he was "not sure" he wanted to "legislate anything such as CE Credits" and that, "[i]f the directors, managers, and the employees all do the right thing, it should happen."(1791) Subsequently, in June 2004, he informed senior management that, whereas internal audit departments in peer institutions provided, on average, sixty hours of training annually (with forty hours generally required to maintain most technical licenses), "we have had to cut back significantly due to workload (currently most staff have less than 20 hours of - ---------- (1787) See Mem. from S. Rajappa to T. Howard, D. Mudd, L. Spencer, and R. Senhauser, dated June 18, 2004, FMSE-IR 423231-35, at FMSE-IR 423232. Rajappa asked for authority only to hire a replacement employee, because the existing employee intended to move to a business unit in 2005. See id. (1788) Id. (1789) A study conducted by E&Y in the spring of 2005 found that sixty-two percent of Fannie Mae's auditors had no prior audit experience. See Ernst & Young, Fannie Mae Office of Auditing: Internal Audit Transformation - Recommendations, dated June 30, 2005, FMSE 510811-33, at FMSE 510822. (1790) Quality Assessment Review: Fannie Mae Internal Audit Department, dated Sept. 2002, FMSE-IR 467089-103, at FMSE-IR 467096. (1791) E-mail from S. Rajappa to Shazia Khurshid, dated July 7, 2003, FMSE-E 572488-91, at FMSE-E 572488. 471
training)."(1792) Nonetheless, by the end of 2004, Internal Audit staff appears to have met this minimum forty hour requirement. Internal Audit reported that its staff had ultimately amassed, on average, forty-two hours of training for the year.(1793) 4. Inadequate Communications (a) Failure to Communicate Staffing Problems to the Audit Committee Although Rajappa had a direct reporting line to the Audit Committee, communications with the Audit Committee and with senior management were inconsistent and, with respect to the Audit Committee, ineffective. Up through August 2004,(1794) there is no evidence that Internal Audit conveyed information to the Audit Committee concerning staffing shortages and attrition problems. On the contrary, Rajappa and Eilers consistently assured the Audit Committee that Internal Audit possessed adequate resources to accomplish its mission.(1795) - ---------- (1792) Mem. from S. Rajappa to T. Howard, D. Mudd, L. Spencer, and R. Senhauser, dated June 18, 2004, FMSE-IR 423231-35, at FMSE-IR 423234. (1793) See Office of Auditing: Completion of 2004 Internal Audit Work, dated Jan. 17, 2005, FMSE-IR 219782-90, at FMSE-IR 219790. (1794) In August 2004, Eilers informed Gerrity that Internal Audit was canceling five audits and postponing six audits due to the demands of SOX 404 work. See E-mail from A. Eilers to T. Gerrity, dated Aug. 23, 2004, FMSE-IR 543704. (1795) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 19, 2004, FMSE 504838-46, at FMSE 504840 ("Mr. Rajappa then briefly described his organization and said that the company had provided him with sufficient resources and budget to complete both audit and the S-Ox 404 work."); Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Jan. 23, 2004, FMSE 504806-15, at FMSE 504807 ("In response to a question from Mr. Gerrity, Mr. Rajappa reported that for 2004, the department headcount was adequate and if more resources were needed this issue would be raised with executive management."); Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 14, 2003, FMSE 504767-75, at FMSE 504768 ("[Eilers] stated that the Company had provided appropriate budget and resources to complete the Internal Audit plan."); Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 19, 2002, FMSE 504690-700, at FMSE 504695 ("Ms. Eilers noted that the Company has provided adequate headcount and funding to conduct an effective internal audit program."); Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 17, 2001, FMSE 504666-71, at FMSE 504668 ("Mr. Rajappa stated that the company had provided the audit 472
From late 2003 on, the difference between the messages conveyed to the Audit Committee and those conveyed to management and Internal Audit staff is striking. In January 2004, Rajappa informed Howard that "[w]e are bleeding from inside and outside recruitment" and identified the "need for a temporary moratorium on hiring from Internal Audit."(1796) Despite the message to Howard (and also despite having identified a need to reduce planned audits in 2004 due to resource issues), Rajappa went on to report to the Audit Committee, in January 2004, that Internal Audit "headcount was adequate."(1797) In April 2004, Rajappa reported to the Audit Committee that he had "sufficient resources and budget to complete both audit and SOX 404 work,"(1798) even though, in February 2004, he had warned Howard that (i) Internal Audit had experienced a twenty-three percent attrition rate during the prior year, including the loss of key staff with critical skills, and (ii) "[g]iven the external pressures created by Sarbanes-Oxley 404, it is imperative that we stabilize staffing levels within the division."(1799) In June 2004, in a headcount request to senior management, Rajappa included summarized results of a benchmarking study relating to staffing. This study found (using 2003 data) that the Company's internal auditors covered "significantly higher revenue and assets than the industry average auditor," and that its "ratio of auditors to total number of Company employees is lower than the industry average." Although Rajappa expressed the view that, "due to our solely domestic operations and core business lines we feel that neither of these statistics is a perfect fit," he concluded that "[o]ur current staff level should be increased to be more in line with the complexity of our business to help us meet current demands."(1800) There is no evidence that Rajappa ever informed the Audit Committee of the results of this benchmarking project - even - ---------- department with sufficient budget and headcount to do the department's work effectively."). (1796) See E-mail from S. Rajappa to T. Howard, dated Oct. 24, 2003, Zantaz document 4260093, at 1. (1797) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Jan. 23, 2004, FMSE 504806-15, at FMSE 504807. (1798) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 19, 2004, FMSE 504838-46, at FMSE 504840. (1799) Mem. from S. Rajappa to T. Howard, dated Feb. 11, 2004, FMSE-IR 509886-87, at FMSE-IR 509886. (1800) See Mem. from S. Rajappa to T. Howard, D. Mudd, L. Spencer, and R. Senhauser, dated Jan. 18, 2004, FMSE-IR 423231-35, at FMSE-IR 423233. 473
though Rajappa had informed the Audit Committee the previous year that Internal Audit's headcount was in line with staffing at comparable institutions.(1801) In this same headcount request, Rajappa warned senior management (but not the Audit Committee) that "[c]ontinued high attrition rates in the audit department (23% for 2003 and 10% YTD for 2004) [were resulting in] fewer experienced auditors, challenges in cross training staff, and difficulties in transition responsibilities to existing staff."(1802) He went on to note that two of Internal Audit's five Directors had left Internal Audit in 2003, and warned that turnover was resulting in inexperience that risked "an increased possibility of errors".(1803) From late 2003 through August 2004, the updates senior management received on resource issues were simply inconsistent with the markedly more positive messages conveyed to the Audit Committee. (b) Failure to Communicate Audit Standards Rajappa and Eilers stated that Internal Audit did not audit Fannie Mae's accounting practices for compliance with GAAP; rather, Internal Audit audited the accounting practices for compliance with the accounting policies developed by the Financial Standards group in the Controller's Office. Rajappa said that, in line with the standards for internal audit functions issued by the Institute of Internal Auditors, responsibility for ensuring compliance with GAAP rested with the Controller's Office and with KPMG. Nonetheless, Internal Audit informed the Audit Committee on a number of occasions that it did, in fact, audit Fannie Mae's accounting practices to determine if they were in accordance with GAAP, including in areas critical to the Company's financial reporting.(1804) For example, in both 2002 and 2003, Rajappa reported to the Audit Committee that Internal Audit's annual derivatives audits had tested transactions for compliance with FAS 133.(1805) - ---------- (1801) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 14, 2003, FMSE 504767-75, at FMSE 504768 ("In response to a question from Mr. Malek, Mr. Rajappa noted that the department headcount was in line with other institutions and provided examples."). (1802) Mem. from S. Rajappa to T. Howard, D. Mudd, L. Spencer, and R. Senhauser, dated June 18, 2004, FMSE-IR 423231-35, at FMSE-IR 423231. (1803) Id. at FMSE-IR 423233. (1804) Rajappa stated, however, that he did not know if the Audit Committee understood that Internal Audit did not audit the Company's accounting practices to GAAP. (1805) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 16, 2002, FMSE 504701-13, at FMSE 504703 (Rajappa reported that, "through data mining techniques, Audit was able to test 100 percent of 474
Similar problems appear in Internal Audit's communications with management, again in areas critical to the Company's financial reporting. For example, during an August 8, 2003 meeting to discuss Internal Audit's investigation into Roger Barnes's accounting-related allegations, Rajappa reported that Internal Audit had concluded that the Company's amortization practices conformed to GAAP.(1806) Similarly, in a December 2003 e-mail to Spencer and Kappler, Eilers informed them that Internal Audit reviewed accounting entries to ensure compliance with FAS 133 and FAS 91.(1807) (c) Inadequate Communication of Audit Findings and Control Deficiencies Although the reports that Internal Audit circulated to the Audit Committee and/or senior management (i.e., the ATL and AIF) identified and tracked audit issues, they did not effectively prioritize findings by importance or severity. Nor did these reports provide a broader perspective - for example, trends in audit reports or common weaknesses identified across business units. 5. Deficiencies in the Office of Auditing's SOX 404 Work By December 2004, Internal Audit reported to the SLT that it had completed documenting and testing 99% of the Company's key financial reporting controls and that 86% of the identified issues had been remediated and retested. Despite the near completion of its work, Internal Audit had not identified any controls as having "Material Weaknesses" (defined by Fannie Mae as internal control problems material to the Company's financial statements with more than a remote likelihood of occurrence). It was not until the SEC concluded that the Company had failed to comply with FAS 133 and FAS 91 that any internal controls were classified as Material Weaknesses. After Fannie Mae agreed to restate its financial statements and terminated its relationship with KPMG in December 2004, the Company suspended Internal Audit's SOX 404 work, and sought external assessments of the reliability of that work. As a result of these assessments, Fannie Mae's management concluded that Internal Audit's - ---------- the derivatives transactions for compliance with FAS 133 and hedging effectiveness"); Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 14, 2003, FMSE 15928-33, at FMSE 15931 (Rajappa reported that "100% of the derivative transactions were tested for compliance with certain elements of FAS 133"). (1806) See Summarized Minutes of the Meeting Regarding Unamortized Balances and Factor Analysis, dated Aug. 8, 2003, FMSE 24417-19. (1807) See E-mail from A. Eilers to A. Kappler and L. Spencer, et al., dated Dec. 10, 2003, FMSE-E 910150-51, at FMSE-E 910150 ("Additionally we review accounting entries to ensure financial reporting is in compliance with Generally Accepted Accounting Principles (i.e. FAS 133, FAS 115, FAS 91, etc.)."). 475
work did not satisfy SOX 404 requirements, and engaged PricewaterhouseCoopers ("PWC") to re-perform the Company's SOX 404 design/implementation project (under the supervision of the Operational Risk unit in the Company's new Risk Organization, which, as noted above, took over responsibility for SOX 404 compliance from Internal Audit).(1808) Not only was the work performed by Internal Audit inadequate, but it also had the effect of significantly reducing Internal Audit's output in other areas. D. Subsequent Developments Since the publication of the September 2004 OFHEO Report, Fannie Mae has taken a number of steps to reorganize Internal Audit, to upgrade the quantity and quality of its staff, and to reform Internal Audit's auditing and audit reporting processes. In 2005, the Company drew upon external expertise in this effort, retaining E&Y to perform an assessment of Internal Audit's functions, structure, and personnel,(1809) and engaging PWC to perform an analysis of Internal Audit's organization, staffing needs, and skill requirements.(1810) 1. Changes in Leadership and Reporting Line The Board hired Jean Hinrichs in June 2005 to replace Rajappa. Unlike Rajappa, Hinrichs has had professional training and experience as an internal auditor. She is certified both as an internal auditor and as a fraud examiner. Prior to joining Fannie Mae, Hinrichs served from 1999-2004 as the Managing Director for Risk at Barclays Global Investors, and from 1997-1999 as Barclays's Managing Director for Internal Audit.(1811) Hinrichs reports directly to the Audit Committee, while also reporting to CEO Mudd for certain administrative purposes. - ---------- (1808) During the course of its work, PWC identified hundreds of potential control problems that Internal Audit had not identified in 2004. See Mem. from Kathryn Rock to R. Levin, dated Aug. 12, 2005, FMSE 519599-616, at FMSE 519606. (1809) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Mar. 28, 2005, FMSE 503042-47, at FMSE 503042-43; see also Ernst & Young, Fannie Mae Office of Auditing: Internal Audit Transformation - Recommendations, dated June 30, 2005, FMSE 510811-83, at FMSE 510813. (1810) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Oct. 17, 2005, FMSE 537768-75, at FMSE 537769. (1811) See Fannie Mae press release, dated June 27, 2005, available at http://www.fanniemae.com/newsreleases/2005/3552.jhtm;?p=Media&s=News+Rele ases. 476
Unlike Rajappa, Hinrichs stated to us that none of her compensation is linked to the Company's annual financial performance. 2. Focus on Core Audit Mission Internal Audit is no longer responsible for such operational risk activities as KPIs, SAQs, and SOX 404 compliance, which have moved to the new Risk organization. In addition, lender audits have been moved from Internal Audit to the Single-Family Mortgage business. 3. Internal Structure and Staffing In June 2005, E&Y recommended to the Audit Committee that Internal Audit revise its internal structure to better align with the organizational structure of Fannie Mae; that it raise the status of the heads of each of Internal Audit's audit groups to Vice President level positions; and that it establish an "Audit Practices" function to develop Internal Audit's audit methodology, its policies and procedures, and its technology support tools, and to oversee quality assurance and auditor training.(1812) In September 2005, Hinrichs discussed with the Audit Committee her goals for a new organizational structure for Internal Audit.(1813) During the course of 2005, and despite the reduced scope of its responsibilities, the authorized headcount for Internal Audit increased from sixty-five to seventy employees, and increased its actual headcount from fifty-eight to sixty employees.(1814) It also retained eight full-time contractors to augment Internal Audit's existing staff.(1815) With respect to training, E&Y recommended not only that Internal Audit create a Vice President for Professional Audit Practices, but also that Internal Audit increase its budget, establish training goals, require that internal auditors receive a minimum number of continuous professional education credits, and develop programs - ---------- (1812) See Ernst & Young, Fannie Mae Office of Auditing: Internal Audit Transformation - Recommendations, dated June 30, 2005, FMSE 510811-33, at FMSE 510823. (1813) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Sept. 19, 2005, FMSE 535007-15, at FMSE 535008-09. (1814) See Org. Chart Internal Audit, dated Nov. 30, 2005, FMSE-IR 699356. (1815) See Mem. from J. Hinrichs to the Members of the Audit Committee, dated Nov. 9, 2005, FMSE 537789-90, at FMSE 537790. Hinrichs stated that she had Audit Committee support to grow as necessary, and that she might need to seek additional staff once she finalized the 2006 audit plan. 477
tailored to each auditor's needs.(1816) Hinrichs stated that she is in the process of implementing E&Y's recommendations. She said that she has already increased Internal Audit's budget for training from $50,000 in 2005 to $250,000 in 2006, and is in the process of developing individualized training programs for Internal Audit staff. 4. Changes to Audit Reporting Hinrichs stated that one of her priorities upon assuming the helm of Internal Audit was to overhaul the systems by which Internal Audit communicated its findings to the Audit Committee, senior management, and the business units. The Audit Committee is now provided a monthly package including a prioritized list of audit issues. In 2005, Internal Audit has worked to improve its system of reporting audit findings in order to provide the Board and management with a clearer and more concise picture of the nature of these findings and their priority. More work remains to be done, but some meaningful changes have already taken place. IV. ETHICS AND COMPLIANCE FUNCTIONS A. Summary In reviewing the Company's ethics and compliance programs as they existed prior to the publication of the September 2004 OFHEO Report, we found programs with a number of positive attributes. For more than a decade, Fannie Mae had maintained a Code of Business Conduct, provided Code-related training to employees, and investigated violations of the Code and other corporate policies. The Company also had a longstanding and experienced investigative unit to handle employee complaints. Moreover, at the beginning of 2003, Fannie Mae acted to enhance the centralization and profile of its ethics and compliance program, by (1) pulling together ethics and compliance functions within the Legal Department; (2) creating the Office of Corporate Compliance ("OCC") to develop and monitor business unit compliance plans, administer employee training, and otherwise provide central management of ethics and compliance matters; (3) appointing a Chief Compliance Officer to oversee the existing investigative unit (the Office of Corporate Justice ("OCJ")) and the new OCC; and (4) replacing the old Business Conduct Committee (which had been chaired by the head of Human Resources) with a new management-level compliance committee chaired by the General Counsel. Although these accomplishments are worthy of note, and although we believe that the ethics and compliance functions contained a number of well-meaning and dedicated professionals, the Company's ethics and compliance program as of late 2004 continued to suffer from the following deficiencies: - ---------- (1816) See Ernst & Young, Fannie Mae Office of Auditing: Internal Audit Transformation - Recommendations, dated June 30, 2005, FMSE 510811-33, at FMSE 510823. 478
1. The level and nature of reporting to the Board on Fannie Mae's ethics and compliance functions were unstructured, with a lack of the type of systematic, written reporting that would permit the Board to assess how well the ethics and compliance offices and related management committees were functioning. 2. Management devoted too few resources to Fannie Mae's ethics and compliance functions (and especially the OCC). 3. Management undermined the perceived independence and impartiality of the Company's ethics and compliance functions by housing them within a litigation section of the Legal Department, headed by a Chief Compliance Officer who also served as the head of the employment practices litigation group responsible for defending the Company against employee complaints. 4. Management failed to invest appropriate responsibilities and stature in its Chief Compliance Officer, who did not hold a dedicated position; was subordinate to the General Counsel; did not report to the Board of Directors; and had no discernable compliance responsibilities other than to supervise the activities of the OCC and the OCJ. 5. Without an active management-level oversight committee, and with an under-resourced and relatively low-stature OCC, the Company lacked an effective mechanism for coordinating compliance matters across the enterprise. Since September 2004, Fannie Mae has taken important steps to rectify deficiencies in, and perceptions of, its ethics and compliance functions. Most notably, it has created a new Office of Compliance, Ethics & Investigations ("OCEI"), which (1) is independent of the Legal Department, (2) reports directly to the CEO and the Compliance Committee, (3) is led by a new Chief Compliance Officer who is committed full-time to ethics and compliance functions, and (4) will not only absorb the functions and resources of the OCC and the OCJ, but will also have a dedicated ethics unit. Moreover, management now provides the Board with detailed written reports on ethics and compliance programs and activities. Fannie Mae has selected the first head of the OCEI and has begun to build-out this new office. We view these structural reforms as very positive, and we encourage the Board and senior management to provide the OCEI with the necessary resources. B. Overview Prior to September 2004 1. Code of Business Conduct The Code was (and remains) the primary policy document that sets forth the ethical obligations of Fannie Mae employees. Fannie Mae first published the Code in 479
1992.(1817) It was substantially reorganized and updated in January 2003 to make it more user-friendly and accessible to employees and to better align it with legal developments over the prior decade (including enactment of SOX).(1818) It was amended again in April 2004 to reflect the role of the OCC as well as new NYSE rules requiring the Company to notify stockholders if provisions of the Code were waived for certain executives.(1819) Despite these amendments, the core principles of the Code have remained largely the same, including the basic duty to conduct Fannie Mae's business affairs in a responsible and ethical fashion. The Code required employees to report any departure from or violation of the Code, and prohibited retaliation against employees who made such reports.(1820) In addition, the Code required employees to be open and trustworthy in Fannie Mae's financial reporting and public communications. Prior to 2003, the Code directed employees who were involved in approving transactions, supplying supporting documentation, or determining account classification to comply with laws and regulations requiring the Company's financial statements, books, and records to accurately reflect all transactions, with prompt and proper recordation of all disbursements and receipts of funds.(1821) As amended in 2003, the Code also directed (1) officers to "make sure that the internal accounting, operational, and disclosure controls and procedures in [their] business area are in place, understood, and followed"; (2) employees involved in preparing Fannie Mae's financial disclosures to make sure that - ---------- (1817) See Mem. from Doug Bibby to the Audit Committee, dated Apr. 9, 1997, FMSE-IR 368-69. (1818) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504743-55, at FMSE 504754. This amendment restated the Code in terms of seven "Key Principles" of ethical behavior: (1) "[b]e accountable"; (2) "[u]phold Fannie Mae's core commitment to diversity"; (3) "[p]rotect Fannie Mae's corporate assets"; (4) "[b]e open and trustworthy in all financial reporting and public communications"; (5) "[a]void conflicts between Fannie Mae interests and personal interests"; (6) "[o]bey the laws and regulations governing [Fannie Mae's] business transactions"; and (7) "[f]oster a safe, healthy and productive workplace." Code of Business Conduct, dated Jan. 21, 2003, FMSE 9782-98. (1819) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 19, 2004, FMSE 504838-46, at FMSE 504844-45. (1820) See, e.g., Code of Business Conduct, dated May 11, 2004, FMSE-IR 36806-30, at FMSE-IR 36809; Code of Business Conduct, dated Jan. 2000, FMSE-IR 36924-78, at FMSE-IR 36931. (1821) See, e.g., Code of Business Conduct, dated Jan. 2000, FMSE-IR 36924-78, at FMSE-IR 36952. 480
they "comply with Fannie Mae's disclosure controls and procedures"; and (3) all officers and employees to cooperate fully with internal and external auditors.(1822) The Code also instructed employees to avoid conflicts of interest, and to abide by various laws and regulations that govern Fannie Mae's business transactions.(1823) In addition, employees were instructed as to how to respond appropriately to government investigations of Fannie Mae.(1824) 2. Structure of Board Oversight Until November 2005, the Audit Committee was the Board committee primarily responsible for overseeing Fannie Mae's ethics and compliance program.(1825) Among other things, its Charter required the Audit Committee to approve all substantive changes to the Code of Business Conduct, and to monitor compliance with the Code.(1826) Prior to 2005, Audit Committee minutes do not tend to reflect discussion by the Committee regarding specific internal investigations of alleged legal or Code violations by employees of the Company.(1827) General Counsel Kappler informed us that such - ---------- (1822) Code of Business Conduct, dated Jan. 21, 2003, FMSE 9781-98, at FMSE 9790. (1823) See, e.g., Code of Business Conduct, dated May 11, 2004, FMSE-IR 36806-30, at FMSE-IR 36816-17, 36819-20; Code of Business Conduct, dated Jan. 2000, FMSE-IR 36924-79, at 36933-40, 36953-58. (1824) Employees were directed to refer to the Legal Department any government investigative request for information, access to files, or an interview, and never to: (1) destroy or alter documents in anticipation of a document request from the government or a court; (2) lie or make misleading statements to government investigators; or (3) cause anyone to give false or misleading information to any government investigator or to destroy or conceal information that a government investigator requests. See, e.g., Code of Business Conduct, dated May 11, 2004, FMSE-IR 36806-30, at FMSE-IR 36820-21; Code of Business Conduct, dated Jan. 2000, FMSE-IR 36924-78, at FMSE-IR 36960. (1825) As discussed above, the Board in November 2005 gave the Compliance Committee primary responsibility for overseeing Fannie Mae's ethics and compliance program. (1826) See, e.g., Audit Committee Charter, dated Apr. 23, 2004, Zantaz document 1454040, at 49; Audit Committee Charter, dated July 18, 2000, FMSE 14495-97, at FMSE 14496. (1827) One exception was the internal investigation of the allegations raised by Roger Barnes in August 2003. See Minutes of the Meeting of the Audit Committee of the Board of Fannie Mae, dated Aug. 12, 2003, FMSE 504788-91, at FMSE 504791. For a discussion of the allegations raised by Barnes and the Company's responses to those allegations, see Chapter IX of this Report. 481
discussions did occur, but that they occurred during executive session where no minutes were kept. The Audit Committee Charter also required the Audit Committee to oversee the activities of management committees related to ethics and compliance, including the work of management's Business Conduct Committee ("BCC") and its successor, the Corporate Compliance Advisory Committee ("CCAC").(1828) The Audit Committee received reports each year from the chairman of the BCC (and in 2003, from the General Counsel) regarding ethics and compliance matters, including Code training and awareness activities, Code certification efforts, and changes and proposed changes to the Code.(1829) The Audit Committee does not appear to have received a similar annual report from the CCAC since its inception in January 2004. After the enactment of SOX, the Audit Committee Charter also required the Committee to "[e]stablish procedures for the receipt, retention and treatment of complaints received by the corporation regarding accounting, internal accounting controls or auditing matters, including procedures for confidential, anonymous submission of concerns by employees regarding accounting or auditing matters."(1830) As noted above, the Committee adopted such procedures in January 2004.(1831) These procedures required management to establish various routes by which employees and members of the public could file complaints with the Audit Committee (including mail, e-mail, and a phone "hot line").(1832) They also charged the General Counsel with processing and conducting - ---------- (1828) See, e.g., Audit Committee Charter, dated Feb. 17, 2004, FMSE 504391-97, at FMSE 504396; Audit Committee Charter, dated July 18, 2000, FMSE 14495-97, at FMSE 14496. (1829) See, e.g., Mem. from the Business Conduct Committee to the Audit Committee, dated July 10, 2003, FMSE 16005-07; Mem. from K. Gallo to the Audit Committee, dated Apr. 9, 2002, FMSE 15391-93. (1830) Audit Committee Charter, dated Jan. 21, 2003, FMSE 504223-28, at FMSE 504227. (1831) See Minutes of the Audit Committee of the Board of Directors, dated Jan. 23, 2004, FMSE 504806-15, at FMSE 504809-10; Fannie Mae Audit Committee, Complaint Procedures for Accounting, Internal Accounting Controls and Auditing Matters, dated Jan. 23, 2004, FMSE 221378-80. When Kappler proposed that the Audit Committee adopt this set of procedures, she noted that most employee complaints in the past had concerned human resources matters that would not have been subject to the new procedures. See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Jan. 23, 2004, FMSE 504806-15, at FMSE 504810. (1832) See Fannie Mae Audit Committee, Complaint Procedures for Accounting, Internal Accounting Controls and Auditing Matters, dated Jan. 23, 2004, FMSE 221378-80, at FMSE 221379. 482
preliminary reviews of such complaints,(1833) but left the Audit Committee responsible for directing and overseeing all subsequent investigations, with assistance from the General Counsel, the head of Internal Audit, or such other persons as the Committee determined to be appropriate. The head of Internal Audit was required to verify that the Audit Committee received all reports or descriptions of complaints related to accounting, internal controls, and auditing matters.(1834) 3. Management Oversight and Activities (a) The Business Conduct Committee and the Corporate Compliance Advisory Committee (1) Business Conduct Committee Prior to 2004, the BCC was the management committee primarily responsible for overseeing the development of the Code and for its enforcement.(1835) Specifically, the BCC was tasked with: (1) recommending to the Audit Committee substantive amendments and revisions to the Code (and, beginning in 1998, approving technical and administrative amendments to the Code); (2) overseeing the publication and distribution of, and the process for certifying compliance with, the Code; (3) overseeing Code training and awareness programs; and (4) reviewing with the OCJ the status of investigations, suspected violations, and other enforcement matters relating to the Code."(1836) The BCC was chaired by the Senior Vice President for Human Resources. - ---------- (1833) Under these procedures, all complaints received by the Audit Committee were to be forwarded to the General Counsel and the head of Internal Audit. After conducting a preliminary review of the Complaint, the General Counsel was required to submit a copy (or a description of the complaint) to the Committee, together with any recommendations as to appropriate handling. The General Counsel was also charged with maintaining a log of all complaints received that tracked their receipt, investigation, and resolution, and with preparing a report to the Committee on at least a semi-annual basis summarizing the contents of this log. See id. (1834) See id. (1835) See, e.g., Business Conduct Committee Charter, dated Jan. 2003, FMSE-IR 36608-09, at FMSE-IR 36608. (1836) See id. at FMSE-IR 36608-09. The BCC also had responsibility for setting policy on appropriate penalty levels for Code violations and addressing "new or complex issues" that might arise from OCJ investigations. See Roles and Responsibility in Fannie Mae's Ethics Compliance Program, dated Feb. 7, 2001, FMSE-IR 36610-14, at FMSE-IR 36613. In addition, when a complaint raised allegations against a Fannie Mae officer, the Office of the Chairman was charged with resolving the matter, but was required to present its proposed resolution to the BCC for its concurrence or alternate recommendation. See Roles and Responsibilities in Fannie 483
The BCC Charter required the Committee to deliver an annual report to the Audit Committee detailing "matters that arise under the Code, including any pertinent issues involved in distributing and certifying compliance with the Code as well as any significant violations."(1837) These reports, which were usually delivered in April, discussed technical amendments and proposed substantive changes to the Code, Code training and awareness activities, annual Code certification updates, and, on some occasions, gave overviews of select OCJ investigations.(1838) The BCC Charter required the BCC to meet at least twice each year.(1839) Based on the BCC minutes, it appears that the Committee met between two and six times each year from 1998 to 2003. (2) Corporate Compliance Advisory Committee In July 2003, the Audit Committee agreed with management's proposal to replace the BCC with the CCAC. The Audit Committee intended the CCAC to assume the BCC's responsibilities for "providing guidance as to adherence to the Code," and to also assume additional responsibility "as to matters related to corporate compliance that are not related to the Code."(1840) The CCAC Charter required the Committee to oversee the activities of the newly-created OCC in addition to the investigative activities of the OCJ.(1841) Among - ---------- Mae's Code Compliance Program, dated Jan. 2003, FMSE-IR 36615-20, at FMSE-IR 36620. (1837) Business Conduct Committee Charter, dated Jan. 2003, FMSE-IR 36608-09, at FMSE-IR 36609. (1838) See, e.g., Mem. from the Business Conduct Committee to the Audit Committee, dated July 10, 2003, FMSE 16005-07; Mem. from K. Gallo to the Audit Committee, dated Apr. 9, 2002, FMSE 15391-93; Mem. from Thomas R. Nides to the Audit Committee, dated Apr. 6, 2001, FMSE 14926-29; Mem. from T. Nides to the Audit Committee, dated Apr.18, 2000, FMSE 14552-56; Mem. from T. Nides to the Audit Committee, dated Apr.14, 1999, FMSE 14195-98, at FMSE 14197 (with overview of OCJ investigation of Code violations by a Vice President). (1839) See Business Conduct Committee Charter, dated Jan. 2003, FMSE-IR 36608-09, at FMSE-IR 36608. (1840) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated July 15, 2003, FMSE 504780-87, at FMSE 504784. (1841) See Undated Corporate Compliance Advisory Committee Charter, FMSE-IR 698349-50, at FMSE-IR 698349. We can find no evidence that the CCAC Charter was ever presented to the Audit Committee. 484
other things, the CCAC was charged with reviewing the OCC's proposed business plan for the coming year and its report of activities in the preceding year (including Code awareness and training programs).(1842) The CCAC was also charged with reviewing annual reports of the past and pending activities of the OCJ, including summaries of OCJ investigations for the prior year.(1843) Unlike the BCC, the CCAC was chaired by the General Counsel (Kappler). In practice, the CCAC did not end up functioning as had been envisioned when it was established. Although the CCAC was established as a standing committee with quarterly meetings, it appears to have met only three times during 2004, with no meetings during the second half of the year. Several planned meetings were reportedly cancelled when no issues were pending before the CCAC.(1844) Finally, although the CCAC Charter specified that the Chair of the CCAC was to provide an annual report to the Audit Committee on compliance matters, including "any significant violations of laws, regulations or the Code, and significant monitoring and enforcement efforts,"( 1845) we found no evidence that such a report was presented to the Audit Committee in 2004 (or in 2005).(1846) - ---------- (1842) Although not mentioned specifically in the Charter, the responsibilities of the OCC in practice included the development of business unit compliance plans, which sought to facilitate business unit compliance with laws and regulations specifically applicable to each such unit. The director of the OCC, Deborah House, stated that she sought guidance from the CCAC as to the OCC's role in developing these business unit compliance plans. (1843) See Undated Corporate Compliance Advisory Committee Charter, FMSE-IR 698349-50, at FMSE-IR 698349. (1844) We understand that the Committee remains in existence, at least formally, and that it met at least once in 2005. See, e.g., Minutes of the Meeting of the Corporate Compliance Advisory Committee, dated Mar. 9, 2005, Zantaz document 1074379. (1845) Undated Corporate Compliance Advisory Committee Charter, FMSE-IR 698349-50, at FMSE-IR 698349. (1846) The CCAC Charter also required a quarterly report to the COO on "any violations of applicable laws, regulations, or the Code of Business Conduct (Code) and the corrective actions that were taken." Id. Kappler's quarterly litigation reports to the Office of the Chairman (which included appendices containing OCJ data) appear to have addressed this reporting requirement. 485
(b) Roles of the General Counsel and the Chief Compliance Officer At the officer level, lead responsibility for Fannie Mae's ethics and compliance program rested with General Counsel Kappler and Deputy General Counsel and Chief Compliance Officer Remy. (1) General Counsel Kappler served as Fannie Mae's General Counsel from May 2000 until December 2005.(1847) In that capacity, she supervised the activities of the OCJ upon its migration to the Legal Department in 2001, as well as the activities of the OCC upon its creation in late 2002. As of 2004, both the OCJ and OCC reported to Kappler through the Chief Compliance Officer. Kappler stated that she received information about employee complaints and OCJ investigations from the director of the OCJ, from Remy, and from OCJ logs that were distributed to her on a periodic basis. According to Kappler, she received copies of final OCJ investigatory reports, but she said that she did not recall reviewing any drafts of those reports.(1848) Kappler provided quarterly litigation reports to members of the Office of the Chairman that contained appendices summarizing the investigatory activities of the OCJ. She also presented information to the Senior Leadership Team from time to time concerning OCJ activities.(1849) - ---------- (1847) In November 2004, Kappler was promoted from Senior Vice President to Executive Vice President, while remaining General Counsel. See Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Nov. 16, 2004, FMSE 504488-94, at FMSE 504493-94. In August 2005, she announced her resignation from Fannie Mae, effective at the end of 2005. See Fannie Mae press release, dated Aug. 24, 2005, available at http://www.fanniemae.com /newsreleases/2005/3589.jhtml?p=Media&s=News+Releases. (1848) Kappler stated that she did review a draft of the OCC's report on the Barnes investigation. (1849) See, e.g., Office of Corporate Justice: Appendix to Litigation Report, dated second quarter 2004, FMSE-IR 209678-79; Office of Corporate Justice: Appendix to Litigation Report, dated fourth quarter 2003, FMSE-IR 209693-96; see also Office of Corporate Justice: 2003 Year-in-Review, dated Mar. 16, 2004, FMSE-IR 264175-97 (presented to the Senior Leadership Team by Kappler and Donald M. Remy). 486
Kappler took the lead in reporting to the Audit Committee on ethics and compliance matters in 2003.(1850) In January 2003, she submitted a revised Code for the approval of the Committee.(1851) In July 2003, she delivered the final report of the BCC and presented plans for its replacement with the CCAC.(1852) In August 2003, she briefed the Audit Committee on an investigation conducted by the OCC and Internal Audit into the accounting and other allegations associated with Roger Barnes.(1853) In the first half of 2004, Kappler presented the Audit Committee with further proposed revisions to the Code(1854) and a new whistleblower complaint policy.(1855) In January 2004, as noted above, she also became responsible for providing the Audit Committee with preliminary reviews of all complaints involving accounting- and audit-related matters, and with preparing a semi-annual report showing the status of all such complaints. As of August 2004, she also was charged with reporting to the Audit Committee any OCJ findings of employee fraud, as well as the initiation of any fraud-related investigation involving a Fannie Mae officer.(1856) With the exception of her August 2003 briefing on the Barnes investigation, Audit Committee minutes prior to September 2004 do not reflect briefings - ------------ (1850) Prior to 2003, the lead effectively rested with the head of the Human Resources department, who, as noted above, chaired the BCC. Kappler did provide annual litigation updates to the Audit Committee prior to 2003. (1851) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504743-55, at FMSE 504754 (Kappler noted that the prior version of the Code had been written ten years earlier and needed to be updated, in part to reflect SOX). (1852) See Minutes of the Audit Committee of the Board of Directors, dated July 15, 2003, FMSE 504780-87, at FMSE 504784. (1853) See Minutes of the Audit Committee of the Board of Directors of Fannie Mae, dated Aug. 12, 2003, FMSE 504788-91, at FMSE 504791. This investigation is discussed in Chapter IX of this Report. (1854) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Apr. 19, 2004, FMSE 504838-46, at FMSE 504844-45. (1855) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Jan. 23, 2004, FMSE 504806-15, at FMSE 504809-10. (1856) See Anti-Fraud Policy, dated Aug. 19, 2004, Zantaz document 1371602, at 4. 487
by Kappler as to the outcome of specific internal investigations of employee complaints.(1857) (2) Chief Compliance Officer As part of Fannie Mae's efforts to update its ethics and compliance program to address amendments to the U.S. Sentencing Guidelines that concerned corporate governance programs,(1858) Fannie Mae created a Chief Compliance Officer position in November 2002 and named Vice President and Deputy General Counsel Donald Remy to that position (simultaneously promoting him to Senior Vice President).(1859) His duties as Chief Compliance Officer supplemented his existing responsibilities as Deputy General Counsel, which included managing the employment law, antitrust, and general insurance practice groups in the Legal Department. Fannie Mae publicly described Remy as being responsible for "the development, implementation, maintenance and enforcement of a comprehensive program for managing Fannie Mae's legal and regulatory compliance risk."(1860) Kappler, however, did not recall any wider role for the Chief Compliance Officer than overseeing the activities of the OCJ and the OCC.(1861) Remy described his oversight of OCJ and - ------------ (1857) Kappler stated that, to the extent that investigations or complaints pertained to accounting improprieties or fraud, she was required to, and did, report such matters to the Audit Committee. She said that she informed the Audit Committee of allegations against senior officers of the Company, and of alleged violations of law. She also said that she tended to present Committee members with oral rather than written briefings on special complaints or investigations. (1858) In 2002, House suggested to Kappler that a Chief Compliance Officer be designated in order to comply with recommendations of the U.S. Sentencing Guidelines. However, she suggested that the General Counsel serve as Fannie Mae's Chief Compliance Officer. See Mem. from D. House to A. Kappler and T. Marra, dated Sept. 9, 2002, FMSE-IR 672285-90, at FMSE-IR 672287. (1859) See Fannie Mae press release, dated Nov. 19, 2002, available at http://www.fanniemae.com/newsreleases/2002/2269.jhtml?p=Media. Remy joined Fannie Mae as Vice President and Deputy General Counsel in 2000 with responsibility for the litigation, antitrust, insurance coverage, and employment matters practice groups. (1860) Id. (1861) Although some employees considered the Chief Compliance Officer to be the Company's "Chief Ethics Officer," Remy himself disagreed with this characterization and acknowledged that no identifiable Chief Ethics Officer existed at the Company. 488
OCC as broad in nature; he stated that he reviewed reports and policies generated by the OCJ and OCC, but he did not necessarily exercise day-to-day control over their activities. Kappler stated that, as Chief Compliance Officer, Remy had authority to go directly to the Audit Committee to report on compliance matters, but she did not recall his ever doing so. Instead, as discussed above, Kappler, rather than Remy, typically reported to the Audit Committee (and, as necessary, the full Board) on compliance matters. 4. Office of Corporate Justice The OCJ was (and continues to be) responsible for investigating and resolving employee complaints of violations of the Code and other Company policies, as well as violations of laws and regulations. The functions of the OCJ have evolved over time, and, at times some of those functions were shared with other offices. (a) Organization and Function In early 2001, Fannie Mae reorganized a pre-existing Human Resources department into two separate units: the OCJ and the Office of Employment Practices ("OEP"). At this time, the OCJ became part of the Legal Department and was granted authority to investigate and resolve not only workplace-related allegations, but also alleged legal violations and violations of the Code and Fannie Mae policies.(1862) Alan Tanenbaum became Director of the new OCJ. As OCJ Director, Tanenbaum supervised a staff of between two and four full-time attorneys.(1863) He reported directly to Remy, who in turn reported to Kappler. Tanenbaum (and/or one of his subordinates) attended meetings of the BCC to report on the activities of the OCJ.(1864) Tanenbaum did not report to the Audit Committee or to the full Board. (b) Overview of Investigative Activities The primary activities of the OCJ have involved the solicitation of complaints from employees, the processing and investigation of complaints, the reporting of investigative findings, and the resolution of complaints (including the formulation of - ------------ (1862) See Mem. from T. Nides to the Audit Committee, dated Apr. 6, 2001, FMSE 14926-29, at FMSE 14928. The OEP remained in Human Resources and had responsibility for drafting Code of Business Conduct policies and amendments, promoting ongoing employee Code awareness and education, and conducting the annual Code certification process until the OCC was created at the end of 2002. (1863) As of September 2004, the OCJ employed four full-time attorneys. (1864) See, e.g., Minutes of the Regular Meeting of the Fannie Mae Business Conduct Committee, dated Mar. 8, 2002, FMSE-IR 209623-24 (attended by Tanenbaum). 489
remedial actions). The following summarizes the policies and procedures by which the OCJ accomplished these activities. (1) Policies for Directing Complaints to the OCJ For a number of years, Fannie Mae's Complaint Process Policy has directed employees to report improper behavior either directly to the OCJ (or its predecessor), or to the employees' managers, their Human Resources representatives (their "HR Talent Team"), or to Fannie Mae officers (all of whom, in turn, were directed to refer such matters to the OCJ). Beginning in 2001, this Policy covered all alleged violations of the Code of Business Conduct.(1865) By 2004, this Policy also encompassed other Company policies beyond those set forth in the Code.(1866) In June 2004, Fannie Mae issued its Ethical Responsibility Policy, which directed employees to report not only violations of the law and the Code, but also improprieties in accounting, internal accounting controls, or auditing matters.(1867) Employees were encouraged to report such violations or improprieties to their supervisors, but they were informed that they could also report their concerns to the OCJ, the OCC, or the Audit Committee. This policy also informed employees that failure to report violations or improprieties could result in remedial or disciplinary action up to and including termination of employment.(1868) - ------------ (1865) See Complaint Process Policy, dated Dec. 10, 2001, FMSE-IR 36634-35; see also Complaint Process Policy, dated Jan. 6, 2004, FMSE-IR 36627-30. (1866) See Complaint Process Policy, dated Jan. 6, 2004, FMSE-IR 36627-30, at FMSE-IR 36627. On the other hand, employees were instructed to work with their managers or HR Talent Team representatives (rather than the OCJ) to resolve matters involving employment performance reviews and ratings; compensation decisions; or "interpersonal, communication, or employee-relations matters that do not involve alleged violations of the Code or Company policy." Id. at FMSE-IR 36628-29. (1867) See Ethical Responsibility Policy, dated June 25, 2004, FMSE-IR 36624-26, at FMSE-IR 36624. Examples of such improprieties included: (1) "fraud, deliberate error, or misrepresentation (oral or written) in the preparation, maintenance, evaluation, review, or audit of any Fannie Mae financial statement, operational results, or financial records"; (2) "deficiencies in or noncompliance with Fannie Mae's internal accounting controls"; (3) "misrepresentations or false statements by a Fannie Mae employee to a senior Officer or to external or internal auditors regarding Fannie Mae's finances or any aspect of Fannie Mae business that they are examining"; or (4) "deviation from full and fair reporting of Fannie Mae's financial condition." Id. (1868) See id. at FMSE-IR 36625. 490
Lastly, as part of the annual Code certification process, employees were required to disclose any activities that they observed or engaged in during the prior year that they believed to violate the Code. These disclosures also were forwarded to the OCJ for investigation.(1869) (2) Means of Processing and Investigating Complaints Fannie Mae offered employees a number of means by which to file complaints with the OCJ. Fannie Mae publicized these means through references in the Code, in OCJ materials, in complaint policies posted on Fannie Mae's intranet, in communications from senior management, and in promotional campaigns.(1870) Employees could raise complaints by: (1) visiting the OCJ's office and filling out a Workplace Dispute/Claim Form (which was the means by which the OCJ historically received most of its complaints);(1871) (2) raising a complaint at the OCJ's annual outreach sessions with Fannie Mae's business units and regional offices;(1872) (3) reporting to their supervisors, to Fannie Mae officers, or to their HR Talent Team representatives (who were then obliged to refer such reports to the OCJ); or (4) contacting OCJ by e-mail, by phone, or by a toll-free phone hotline (the "Hotline"). The Hotline was significantly upgraded in 2004, after the OCJ became concerned with statistics indicating that employees scarcely utilized the existing system.(1873) - ------------ (1869) See Roles and Responsibilities in Fannie Mae's Code Compliance Program, dated Jan. 2003, FMSE-IR 36615-20, at FMSE-IR 36617. (1870) See, e.g., HomeSite News: Compliance Office Outlines Reporting Options, dated Oct. 3, 2003, FMSE 24386. (1871) See Corporate Justice System Process, dated Nov. 11, 2003, Zantaz document 1736411, at 1. (1872) See, e.g., Office of Corporate Justice: 2004 Year-in-Review, dated Feb. 17, 2005, Zantaz document 438055; Office of Corporate Justice: 2003 Year-in-Review, dated Mar. 16, 2004, FMSE-IR 264175-97; Office of Corporate Justice, 2002 Year-in-Review, dated Mar. 2003, FMSE-IR 192850-71, at FMSE-IR 192857. (1873) One member of the OCJ informed us that employees did not use the old Hotline because they did not trust its ability to preserve anonymity. She noted, for example, that the Hotline phone in the OCJ had Caller-ID, and thus could identify the phone number of a caller. In both 2003 and 2004, only seventeen percent of OCJ complaints were received through the Hotline. See Office of Corporate Justice: 2004 Year-in-Review, dated Feb. 17, 2005, Zantaz document 438055, at 20. 491
(3) Policies and Procedures for the Intake and Processing of Complaints The OCJ maintained logs to record incoming complaints and reports of violations and to track their status, including a log of open and pending cases. On this open case log, OCJ attorneys entered the names of the complainant (if known) and the accused, the date of the report or complaint, an assigned case number, and a description of the allegations.(1874) Beginning in the summer of 2004, whenever an employee alleged fraud or an accounting impropriety, the OCJ also entered information about the complaint in a second log.(1875) If fraud was alleged, Fannie Mae's Anti-Fraud Policy required the OCJ to make an initial assessment of the viability of the allegation, and, if determined to be viable, to notify in writing the Chief Compliance Officer, the General Counsel, and the head of Internal Audit.(1876) Once a complaint was received and logged, the OCJ made a determination as to how to address it. If the OCJ chose to act on a complaint,(1877) it decided whether to: (1) provide informal guidance to assist the parties in resolving their dispute on their own; - ------------ (1874) The OCJ also generated a log listing cases as they closed, as well as a log that listed remedial actions directed by the OCJ but not yet carried out by the relevant business units. The latter log was forwarded to the OCC, which tracked implementation of overdue remedial actions in the relevant business units. (1875) This log was distributed on a monthly basis to Chief Compliance Officer Remy, General Counsel Kappler, and OCC Director House. The OCJ also prepared and delivered to the head of Internal Audit and the General Counsel a quarterly report summarizing all complaints and reports of fraud or accounting impropriety that were received and/or closed by the OCJ during that quarter. See Draft Office of Corporate Justice: Policy and Procedures Manual, dated Dec. 7, 2004, Zantaz document 1371031, at 49. (1876) The Anti-Fraud Policy defined "fraud" broadly to include "any intentional act or omission affecting or involving, or potentially affecting or involving, Fannie Mae, that is committed or attempted for the purpose of securing an improper or unlawful gain or benefit for Fannie Mae or any other individual or entity, regardless of whether the gain or benefit is actually realized." Anti-Fraud Policy, dated Aug. 19, 2004, Zantaz document 1371602, at 1. (1877) Under the Complaint Process Policy, the OCJ had discretion to "decide not to address any portion or all of a report, complaint, or referral that it [determined did] not sufficiently allege a violation of Code or policy, or that [was] more suitable for the employee's supervisory chain of command or HR Talent Team to address." Complaint Process Policy, dated Jan. 14, 2004, FMSE-IR 36627-30, at FMSE-IR 36628. 492
(2) offer to mediate the dispute without issuing any formal decision; or (3) initiate an investigation, with interviews, other fact-gathering, and issuance of a written decision that could be accompanied by a directive for remedial action.(1878) Some types of complaints, however, required or almost invariably led to investigations.(1879) The OCJ investigated thirty-four percent of the complaints it received in 2003, and forty-one percent of the complaints it received in 2004.(1880) (4) Policies and Procedures for Investigating Complaints If the OCJ chose to conduct an investigation, Tanenbaum assigned one or more OCJ attorneys to the case. OCJ protocol directed the investigating attorney to notify the accused of the allegations against him or her, to schedule a time to review the allegations with the accused, and to provide the accused with an opportunity to respond to the allegations and to present evidence and documentation in his or her defense. The investigating attorney also was required to notify the management chain of the accused (to the extent that doing so would not compromise the investigation), as well as the relevant practice group attorney and HR Talent Team member.(1881) If the OCJ concluded that it did not have the expertise or resources to conduct the investigation, the OCJ Director could obtain assistance from external sources, such as forensic accounting firms or outside counsel, or from internal departments, such as Internal Audit. Although the OCJ and the OCC (when it briefly conducted investigations in 2003) occasionally utilized Internal Audit for investigative purposes,(1882) Remy stated that Internal Audit was not suited for this role because it did - ------------ (1878) See Corporate Justice System Process, dated Nov. 24, 2003, Zantaz document 1736411, at 1-2. (1879) For example, in the event that the OCJ determined an allegation of fraud to be viable, the Anti-Fraud Policy required the OCJ to conduct an investigation (or engage outside counsel to do so) and/or to "direct prompt action to ensure that Fannie Mae derives no unlawful or improper gain or benefit from the activity." Anti-Fraud Policy, dated Aug. 19, 2004, Zantaz document 1371602, at 3. (1880) See Office of Corporate Justice: 2004 Year-in-Review, dated Feb. 17, 2005, Zantaz document 438055, at 20. We had an opportunity to review the OCJ's files, and we did not encounter accounting issues similar to those addressed elsewhere in the Report. (1881) See Protocol for OCJ Attorneys, dated Aug. 25, 2003, FMSE 24368-85, at FMSE 24370. (1882) For example, as discussed in Chapter IX of this Report, the OCC sought Internal Audit's assistance in investigating Barnes's accounting-related allegations in August 2003. 493
not possess forensic auditing expertise.(1883) In addition, Tanenbaum observed that perhaps Internal Audit was reluctant to become involved in investigative work due to a lack of available staff. The OCJ reported that the average amount of time from receipt to closure for investigated cases was 105 days in 2004, down from 129 days in 2003.(1884) (5) Investigative Findings The OCJ found violations in sixty-six percent of the cases it investigated in 2004, up from forty-six percent in 2003, forty-four percent in 2002, and thirty-two percent in 2001. Most violations concerned misuse of Fannie Mae technology or harassment of employees.(1885) When the OCJ concluded an investigation, it prepared a Decision Memorandum that set forth its investigative findings. A second type of report, called an "Action Memorandum," often accompanied the Decision Memorandum, setting forth any disciplinary or remedial measures that the OCJ prescribed. Action Memoranda most often directed counseling, additional training, or a formal reprimand, but it also directed a variety of other measures, including suspension without pay or termination of employment in some cases.(1886) According to Tanenbaum, no one outside the OCJ provided substantive input into the preparation of Decision Memoranda or Action Memoranda. However, the OCJ occasionally shared draft reports with Remy, who provided editorial comments. According to Remy, he did not typically review draft OCJ reports unless they involved allegations against officers or allegations involving fraud or accounting issues. As noted above, Kappler stated that she did not recall reviewing any drafts of OCJ reports. - --------------- (1883) Nonetheless, Internal Audit was brought in to investigate Roger Barnes's accounting allegations as discussed in Chapter IX of this Report. (1884) See Office of Corporate Justice: 2004 Year-in-Review, dated Feb. 17, 2005, Zantaz document 438055, at 23. (1885) See id. at 27; Office of Corporate Justice: 2003 Year-in-Review, dated Mar. 16, 2004, FMSE-IR 264175-96, at FMSE-IR 264187. (1886) In 2004, for example, of the cases that led to Action Memoranda, the OCJ directed counseling in approximately twenty-seven percent of cases, additional training in approximately twelve percent of cases, formal reprimands in approximately eleven percent of cases, memoranda of concern in approximately nine percent of cases, termination of employment in approximately five percent of cases, suspension without pay in approximately four percent of cases, and miscellaneous actions in other cases. See Office of Corporate Justice: 2004 Year-in-Review, dated Feb. 17, 2005, Zantaz document 438055, at 33. 494
Final copies of OCJ decisions that included an Action Memorandum were distributed to: (1) the management chain of the accused; (2) the HR Talent Team representative of the accused; (3) the employment group attorney assigned to the accused's business unit; and (4) the OCJ Director. The complainant and the accused received final copies of the Decision Memorandum, but not the Action Memorandum.(1887) Remy and Kappler stated that they also received final copies of these OCJ decision documents. In the event that the OCJ directed disciplinary or other remedial action, the OCJ could set a deadline for the action to be taken. If management did not act before the deadline and/or failed to notify the OCJ of its action, the OCJ referred the matter to the OCC for follow-up using tracking reports.(1888) 5. Office of Corporate Compliance In December 2002, Fannie Mae established the OCC as part of a more formal, centralized compliance program.(1889) Initially, the OCC was charged with four main responsibilities: (1) overseeing the development of business unit compliance plans and monitoring adherence to those plans; (2) recommending changes to the Code of Business Conduct; (3) devising and conducting Code-specific and other compliance training for employees; and (4) investigating employee allegations of Code violations and other non-workplace-related complaints. As discussed below, the Company removed the OCC's investigative function in early 2004 and returned it to the OCJ. Deborah House was appointed as the first director of the OCC, reporting to Remy. She came to that position after working as a Vice President and counsel in the Legal Department's multifamily products practice group. As of late 2004, House's staff consisted of four director-level attorneys, three senior managers (one of whom was an attorney), and a paralegal. House did not report to the Audit Committee or the full Board - ------------ (1887) See Protocol for OCJ Attorneys, dated Aug. 25, 2003, FMSE 24368-85, at FMSE 24373. In cases involving alleged fraud, the Anti-Fraud policy required the OCJ to notify the OCC, the Chief Compliance Officer, the General Counsel, and Internal Audit of all findings and any actions directed as a result of the investigation. The General Counsel was then required to report to the Audit Committee any findings of fraud. See Anti-Fraud Policy, dated Aug. 19, 2004, Zantaz document 1371602, at 4. (1888) See, e.g., Draft Office of Corporate Justice: Policies and Procedures Manual, dated Dec. 7, 2004, Zantaz document 1371031, at 38. (1889) See Legal Department: Presentation on Corporate Compliance, dated Nov. 2002, Zantaz document 2298837, at 3. Kappler stated that she wanted to consolidate compliance programs that spanned the Legal Department and Human Resources and place them under the control of the Legal Department. But she said that she hoped eventually to migrate the OCC out of the Legal Department and establish it as an independent office. 495
on the OCC's activities,(1890) but she did, on at least one occasion, update the CCAC on the status of the OCC's efforts to develop business unit compliance plans.(1891) (a) Compliance Activities A primary focus of OCC's efforts in 2004 was the development of business unit compliance plans. The goal of these plans was to enhance compliance throughout the Company by identifying the laws, regulations, Code provisions, and Fannie Mae policies applicable to each business unit and setting forth specific policies and procedures for ensuring compliance. The concept was that business units would "own" their individual compliance plans, but with central oversight, monitoring, and assistance by the OCC. By the end of 2004, OCC attorneys, working with the individual business units, had completed the drafting and approval of compliance plans covering each of the business units (forty plans in total).(1892) Each compliance plan contained a description of the business unit, a schedule of the laws and regulations that specifically affected the business unit, a chart of Code provisions that were of particular concern to the business unit, and a schedule of both Company-wide and customized training that each employee in the business unit had to complete. The plan also listed an inventory of policies or procedures that either existed or needed to be developed and implemented by the business unit to facilitate compliance, and identified any outstanding directives of the OCJ or OCC that remained unfulfilled.(1893) The OCC also worked during 2004 to develop programs to train employees on the laws and regulations listed in the business unit compliance plans. By the end of 2004, the OCC had rolled out an internet-based training program (covering the Code of Business Conduct). The OCC also conducted classroom training sessions in - ------------ (1890) Kappler stated that she provided oral briefings to the Audit Committee regarding the activities of the OCC. Kappler also stated that Internal Audit updated the Audit Committee on the status of the OCC's business compliance plans as part of Internal Audit's periodic SOX 404 updates. (1891) See Corporate Compliance Advisory Committee Agenda, dated Apr. 29, 2004, FMSE-IR 209667 (includes "Status Report on Business Unit Compliance Plans"). (1892) See Mem. from D. Remy and D. House to the Audit Committee, dated June 15, 2005, FMSE 510176-78 (reporting on the activities of the OCC). (1893) See, e.g., Treasurer's Office Business Unit Legal and Regulatory Compliance Plan, dated first quarter 2005, FMSE-IR 228686-719. 496
2004 in certain areas of law, including antitrust, fair employment practices, and fair lending.(1894) The OCC was also responsible for monitoring each business unit's adherence to its compliance plan. The OCC was required to report significant violations "first to senior management and then, absent resolution, to the Audit Committee."(1895) (b) Activities Relating to the Administration of the Code of Business Conduct OCC's administration of the Code involved Code awareness, certification, interpretation, and training. With respect to Code awareness efforts, the OCC published the Code on Fannie Mae's internet and intranet sites. The OCC also gave new employees copies of the Code with their offer letters and required them to sign forms acknowledging their receipt and review of the Code. In addition, the OCC ensured that the Office of the Chairman routinely published messages on the importance of compliance, and was involved in drafting some of the weekly messages from Raines relating to ethics and compliance matters. With respect to Code certification, employees were required to certify annually their compliance with the Code and to disclose any known violations of the Code.(1896) In 2004, the OCC introduced an online version of the annual certification.(1897) With respect to Code interpretation, the OCC administered a database of the Code and Code-related policies, and provided Code-related guidance to employees by e-mail and phone.(1898) The OCC also developed and disseminated new Company policies that expanded upon and interpreted the Code.(1899) - ------------ (1894) See Mem. from D. Remy and D. House to Members of the Board of Directors, dated Feb. 10, 2005, Zantaz document 1074343. (1895) Id. (1896) See, e.g., Code of Business Conduct, dated May 11, 2004, FMSE-IR 36806-30, at FMSE-IR 36824. As part of the certification process, employees were given an opportunity to disclose Code violations. (1897) See Fannie Mae Self-Assessment Questionnaire (SAQ): SAQ Responses - Legal Department, fourth quarter 2004, Zantaz document 1188629, at 4. (1898) See id. (1899) For example, the OCC in 2004 developed the Ethical Responsibility and Anti-Fraud policies. See id. at 5. 497
With respect to Code training, the OCC rolled-out an internet-based version of mandatory Code training in November 2004.(1900) All Fannie Mae employees were required to complete the training by the end of 2004, and the OCC tracked their compliance with this requirement.(1901) (c) Investigative Activities In the first year of its existence, the OCC shared investigatory responsibilities with the OCJ. According to Remy, the OCC was assigned to investigate Code violations and other non-workplace-related complaints, while the OCJ remained responsible for investigating workplace-related complaints (such as alleged discrimination or sexual harassment). The OCC's first investigative effort was the Roger Barnes case, commencing in August 2003. The OCC investigated Barnes' allegations of accounting improprieties, as well as his allegations concerning an adverse working environment in the Controllers' Office; the OCJ investigated Barnes' allegations of discrimination and harassment.(1902) Aside from the multi-pronged Barnes investigation, the OCC conducted only a few minor investigations before the Legal Department, in early 2004, removed the investigative function from the OCC and returned it to the OCJ. According to Remy, the Company removed the OCC's investigative responsibilities because: (1) the OCJ had more resources and expertise than the OCC to commit to investigatory work; and (2) there was too much confusion between the OCJ and OCC as to their respective investigatory responsibilities. Remy also explained that employees traditionally recognized the OCJ as the "cop" of the Company, and that he was concerned about this perception becoming muddled. C. Findings By late 2004, Fannie Mae had established an ethics and compliance program that had a number of positive attributes. For more than a decade, Fannie Mae had maintained a Code of Business Conduct, provided Code-related training to employees, and investigated violations of the Code and other corporate policies. The Company also had a longstanding and experienced investigative unit to handle employee - ------------ (1900) See id. (1901) See id. (1902) The OCC performed another investigation arising out of Barnes's allegations several months later, after KPMG asked the Company to investigate whether other accounting-related concerns had been raised by employees but ignored by Controllers' Office management. Each of these investigations arising out of Barnes's allegations is discussed in Chapter IX of this Report. 498
complaints. Moreover, beginning in late 2002, Fannie Mae acted to enhance the centralization and profile of its ethics and compliance program, by: (1) pulling together ethics and compliance functions within the Legal Department; (2) creating the OCC to develop and monitor business unit compliance plans, administer employee training, and otherwise provide central management of ethics and compliance matters; (3) appointing a Chief Compliance Officer to oversee OCJ and the new OCC; and (4) replacing the BCC with a new management-level compliance committee chaired by the General Counsel (rather than the head of Human Resources) as well as business unit representation. Although these accomplishments are worthy of note, and although we believe that the ethics and compliance functions were overseen by well-meaning and dedicated professionals, the Company's ethics and compliance functions as of late 2004 suffered from: (a) unstructured information flow to the Board; (b) inadequate commitment of resources; (c) inappropriate placement of the ethics and compliance functions within the organization; (d) inappropriate placement of the Chief Compliance Officer position; and (e) insufficient cross-enterprise coordination. 1. Unstructured Information Flow to the Board Prior to 2005, we can find no evidence of systematic, written reporting by management to the Board concerning ethics and compliance matters. Except for the annual high-level report from the BCC to the Audit Committee concerning ethics and compliance activities (which does not appear to have been continued by its successor, the CCAC), information concerning ethics and compliance matters appears to have been conveyed largely in oral briefings that were driven by events (e.g., a particular employee complaint). In particular, we do not find evidence of the type of information flow that would permit the Board to assess how well the ethics and compliance offices and related management committees were functioning, in terms of such core areas as resources, independence, central coordination, trends in reported matters, and strategic planning. Major structural changes, such as the manner in which ethics and compliance functions were centralized in the Legal Department in late 2002, do not appear to have been presented for Board-level review. 2. Insufficient Resources Both the OCJ and the OCC suffered from resource deficiencies. It appears, however, that that the OCJ addressed to a significant degree its resource problems by outsourcing complex or sensitive investigations over the past few years (which had the additional benefit of allowing the OCJ to draw upon outside expertise and independence). The OCC, by contrast, did not have the resources necessary to match its mission, and the goal of creating a vibrant, cross-enterprise overseer of ethics and compliance had not been realized. Although generating forty business unit compliance plans with a handful of attorneys in the space of a year is impressive, the lack of resources resulted in drawing personnel away from training and guidance activities, and 499
does not appear to have promoted a broad willingness on the part of business unit personnel to look to the OCC for support.(1903) 3. Inappropriate Placement of the Ethics and Compliance Functions Management's decision to house the OCC and the OCJ in a litigation section of the Legal Department was inappropriate, in that it appeared to compromise the independence of the ethics and compliance program and limited its status and visibility within the organization. This placement was particularly troublesome for the OCJ. The OCJ's placement undercut employee confidence by raising at least the possibility that OCJ investigations could be tainted by management's litigation strategies. This potential conflict of interest was exacerbated by the fact that Remy, the Chief Compliance Officer and supervisor of OCJ, was also the Deputy General Counsel in charge of the Legal Department's employment practice group - a practice group that was responsible for defending the Company's interests against claims brought by Fannie Mae employees. 4. Weakness of the Chief Compliance Officer Position As noted above, when Fannie Mae announced the appointment of Remy as its first Chief Compliance Officer in late 2002, it described Remy as being responsible for "the development, implementation, maintenance and enforcement of a comprehensive program for managing Fannie Mae's legal and regulatory compliance risk."(1904) In reality, Remy did not have the authority, visibility, or range of functions appropriate to a Chief Compliance Officer. He did not appear as Chief Compliance Officer before the Board or the Audit Committee, nor did he chair the Company's management-level compliance committee. Remy had substantial responsibilities within the Legal Department that were not related to his ethics and compliance responsibilities, including some responsibilities that conflicted with his duties as Chief Compliance Officer. As noted above, a number of employees and Board members did not know that Remy was the Company's Chief Compliance Officer. Although it is not unusual for a company to have a chief compliance officer who is situated in the legal department,(1905) it is both unusual and undesirable for - ------------ (1903) Mercer came to a similar conclusion after surveying a number of business unit managers in 2004. See Mercer Oliver Wyman Consulting, Report on Compliance, Ethics & Investigation, dated June 30, 2005, Zantaz document 1963752, at 7 (noting a "[h]igh degree of business unit frustration with degree and quality of corporate-level compliance support - compliance plans viewed as lists of laws and regulations to be complied with, with limited `how to' support."). (1904) Fannie Mae press release, dated Nov. 19, 2002, available at http://www.fanniemae.com/newsreleases/2002/2269.jhtml?p=Media. (1905) See, e.g., Corporate Executive Board, The Role of the Chief Compliance Officer (May 2004) at 6 (in a survey of chief compliance officer functions at sixteen companies, the chief compliance officer reported to the general counsel in nine 500
a chief compliance officer not to report directly to a company's board on compliance matters;(1906) and it is highly unusual for a chief compliance officer to have responsibility for supervising a company's employment practices litigation. 5. Insufficient Cross-Enterprise Coordination From 1993 through 2003, the BCC recommended revisions to the Code of Business Conduct, oversaw management's compliance training and employee certification programs, reviewed the activities of ethics and compliance personnel, and reported annually on compliance matters to the Audit Committee. Although the CCAC took over responsibility for many of these same functions, it appears to have become largely moribund by the second half of 2004. We cannot identify any meetings in the last six months of 2004, nor can we find any evidence that an annual report was delivered to the Audit Committee in 2004 (or 2005). When we met with Kappler in early 2005, she acknowledged that the CCAC had not functioned as intended, and she noted that management was considering its dissolution. Without an active CCAC, and with an under-resourced and relatively low-stature OCC, the Company lacked an effective mechanism for coordinating compliance matters across the enterprise as of the second half of 2004. D. Subsequent Developments 1. Creation of an Independent Ethics, Compliance & Investigations Organization In the wake of the September 2004 OFHEO Report and subsequent interaction with OFHEO, Fannie Mae's Board agreed to conduct a review of the Company's legal and regulatory compliance structures and to establish a separate Office of Compliance, Ethics & Investigations that would report directly to the CEO and the Board.(1907) The head of this new office was to be hired and removed only with the - ------------ instances; in three other instances, the general counsel was the chief compliance officer). (1906) See id. at 9 (noting in only one of sixteen companies surveyed did the chief compliance officer not report to the Board). (1907) To assist in the process of redesigning Fannie Mae's ethics and compliance program, the Company engaged Mercer, which also worked on redesign efforts for other parts of the Company (and especially the Finance organization). Mercer's report setting forth its design and staffing recommendations for the new ethics and compliance organization was completed and submitted to OFHEO in June 2005. See Mercer Oliver Wyman Consulting, Report on Compliance, Ethics & Investigation, dated June 30, 2005, Zantaz document 1963752; see also Minutes of the Meeting of the Compliance Committee of the Board of Directors of Fannie Mae, dated June 22, 2005, FMSE 538347-48. 501
approval of the Board; to have no other responsibilities at the Company; and to operate independently, with the authority to communicate directly with the Board and OFHEO.(1908) As designed by Mercer (working with Company management), the Company's OCC, OCJ, and Regulatory Reporting functions will be wrapped into the new OCEI with a proposed minimum fifty percent increase in staffing and regular reporting to the Audit and/or Compliance Committee, the CEO, the management-level Business Operations Committee, and OFHEO. The OCEI will be situated outside the Legal Department, with the head of OCEI reporting directly to the CEO and the Board. The head of OCEI will be a dedicated position at the Senior Vice President level, with hiring and removal subject to Board approval. As contemplated, the head of the OCEI would report to the CEO at least monthly, and to the Board at least quarterly, on his or her compliance risk assessment, on investigations conducted, and on identified breaches or deficiencies and corrective measures recommended or taken. The head of the OCEI would also be expected to report to OFHEO on major issues and developments, including whistleblower cases.(1909) On December 20, 2005, the Company announced that William Senhauser had been selected as the Company's new Chief Compliance Officer and the - ------------ In addition, the Board engaged Hildebrandt International ("Hildebrandt") to review Fannie Mae's legal and regulatory compliance structures. Hildebrandt's report concerning the Company's legal and regulatory compliance structures was completed and delivered to OFHEO in September 2005. See Hildebrandt International: Legal and Regulatory Compliance Structures Review: Final Report, dated Sept. 9, 2005. With respect to design and staffing of the new ethics and compliance organization, Hildebrandt largely absorbed the recommendations coming out of Mercer's work. (1908) See Supplemental Agreement between the SRC and OFHEO, dated Mar. 7, 2005, at 4, available at http://www.ofheo.gov/media/pdf/fannie05agreement.pdf. (1909) See Mercer Oliver Wyman Consulting, Report on Compliance, Ethics & Investigation, dated June 30, 2005, Zantaz document 1963752. As designed by Mercer, the OCEI will consist of: (i) a Compliance unit, headed by a Vice President for Compliance, which will absorb most of the functions of the OCC (but with over twice the staffing level); (ii) an Investigations unit, headed by a Vice President for Investigations, which will absorb the functions of the OCJ (and which will have authority to engage external investigators); (iii) a Complaint Evaluation unit (to receive employee complaints and direct them, as appropriate, to the Investigations unit, HR, or elsewhere); (iv) an Ethics unit, headed by a Vice President for Ethics, which would absorb Code-related and other ethics functions from the OCC; and (v) a Regulatory Reporting unit, to oversee reporting to OFHEO and HUD and to coordinate examinations by these agencies. 502
first head of the OCEI, with direct reporting to the CEO and the Compliance Committee.(1910) As of the beginning of 2006, the build-out of the OCEI is just commencing. Thus, while we believe that the contemplated design and staffing for the new organization should represent considerable improvement, it is too early to assess the functioning of the OCEI. We agree with both Mercer and Hildebrandt, which recommended the establishment of a "Control/Coordination Committee," composed of the head of the OCEI, the General Counsel, the CRO, and the Chief Audit Executive, in order to ensure cross-enterprise coordination of legal, compliance, ethics, and risk management programs and activities.(1911) 2. Expanded Role of Compliance Committee In its September 2005 report, Hildebrandt recommended that Fannie Mae consider establishing a permanent Board committee specifically to oversee the legal and regulatory compliance functions of the Company.(1912) The Board accepted this recommendation in November 2005 and shifted primary responsibility for ethics and compliance oversight from the Audit Committee to the Compliance Committee (whose mandate had previously been focused on ensuring implementation of the Company's agreements with OFHEO). The Audit Committee remains responsible for overseeing compliance matters relating to financial reporting, accounting, audit, and tax matters.(1913) This appears to be a suitable means of accomplishing more direct interaction between the Board and management on ethics and compliance matters, especially given the success that the Compliance Committee appears to have had over the past year in rebuilding the Company's relationship with OFHEO. - ------------ (1910) See Fannie Mae press release, dated Dec. 20, 2005, available at http://www.fanniemae.com/newsreleases/2005/3660.jhtml;jsessionid=3ND P2WGQQTTQ1J2FQSISFGI?p=Media&s=News+Releases. (1911) See Mercer Oliver Wyman Consulting, Report on Compliance, Ethics & Investigation, dated June 30, 2005, Zantaz document 1963752, at 40; Hildebrandt International: Legal and Regulatory Compliance Structures Review: Final Report, dated Sept. 9, 2005, at 9. (1912) Hildebrandt International: Legal and Regulatory Compliance Structures Review: Final Report, dated Sept. 9, 2005, at 9. This recommendation is in line with a trend among U.S. public companies, whose audit committees have faced substantially increased demands in recent years. (1913) See Compliance Committee Charter, dated Nov. 15, 2005, FMSE 547749-50; Audit Committee Charter, dated Nov. 15, 2005, FMSE 547741-45, at FMSE 547744. 503
3. Enhanced Reporting to the Board Even prior to the Mercer and Hildebrandt reports and the effort to create an new ethics and compliance organization, the Company took some significant steps to improve the quality and quantity of reporting to the Board on ethics and compliance matters. In a February 2005 memorandum to the Audit Committee, Kappler acknowledged that it would be prudent to adopt a more "formalized reporting mechanism" to report OCJ investigations to the Audit Committee in light of the rising frequency of whistleblower and accounting-related complaints since the publication of the OFHEO Report.(1914) The same month, Remy briefed the Audit Committee for the first time on OCJ matters.(1915) Subsequently, Audit Committee minutes indicate that Kappler briefed the Audit Committee regularly on internal complaints and OCJ investigations. On at least one occasion, Kappler discussed resourcing problems with the Audit Committee.(1916) It also appears that, in 2005, management began to provide systematic reports to the Board on the activities of the OCC. For example, Remy and House - ------------- (1914) Mem. from A. Kappler to the Audit Committee, dated Feb. 11, 2005, FMSE-KD 28309-17, at FMSE-KD 28309. Kappler proposed to provide to the Audit Committee written reports detailing (1) significant violations of the Code, (2) terminations of employment directed by the OCJ, and (3) OCJ investigations involving whistleblower claims and allegations of accounting-related impropriety. See id. (1915) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 17, 2005, FMSE 503025-34, at FMSE 504034 ("Mr. Remy provided highlights of the 2004 Year-in-Review report of the Office of Corporate Justice."). Audit Committee minutes reflect that Remy appeared before the Committee on one other occasion in 2005, to report on the Company's anti-fraud policy and program. See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Oct. 17, 2005, FMSE 537768-75, at FMSE 537773-74. (1916) See Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Oct. 17, 2005, FMSE 537768-75, at FMSE 537774-75 ("Ms. Kappler reviewed activities of the Office of Corporate Justice. She said that some inquiries are lagging due to the need for more resources and that she and Mr. Williams were planning how to address the issue."). 504
provided written reports to the full Board on OCC matters in February and June 2005,(1917) and Kappler briefed the Audit Committee on the activities of the OCC in June 2005.(1918) The February 2005 report from Remy and House outlined the OCC's progress in 2004 in such areas as developing business unit compliance plans, adoption of new corporate policies (the Ethical Responsibility Policy and the Anti-Fraud Policy), and the rolling out of the Company's new computerized Code training program. In addition, the February report attached a sample business unit compliance plan for review, and recommended that the Audit Committee receive a report from management concerning the functioning of the Company's ethics and compliance program at least once a year.(1919) In our view, this represents progress in enabling meaningful Board engagement on ethics and compliance matters. Going forward, we think it is especially important for management reporting in these areas to provide meaningful risk assessments and discussion of perceived resource or other deficiencies. V. OFFICE OF THE CONTROLLER A. Summary With respect to the functioning of the Office of the Controller (the "Controller's Office") prior to the release of the OFHEO Report in September 2004, our conclusions are as follows: 1. The Controller's Office suffered from significant resource deficiencies. The headcount of the Controller's Office increased only modestly during the years prior to 2005, even as the Controller's Office experienced dramatic increases in workload stemming from the introduction of new and complex accounting standards, the Company's decision to become an SEC registrant, and the growth of Fannie Mae's business. The Controller's Office leadership lacked adequate staffing and sufficient accounting and financial reporting expertise and experience for a financial services company as complex as Fannie Mae. In addition, the Controller's Office relied to a substantial degree on inadequate systems that required considerable manual effort, further straining the already overburdened staff. - ------------ (1917) See Mem. from D. Remy and D. House to Members of the Board of Directors, dated Feb. 10, 2005, Zantaz document 1074343; Mem. from D. Remy and D. House to Members of the Board of Directors, dated June 15, 2005, FMSE 510176-78. (1918) See Minutes of the Audit Committee of the Board of Directors of Fannie Mae, dated June 20, 2005, FMSE 510748-62, at FMSE 510760. (1919) See Mem. from D. Remy and D. House to Members of the Board of Directors, dated Feb. 10, 2005, Zantaz document 1074343. 505
2. The closing process was manually intensive and unduly susceptible to human error. The relevant computer systems were not integrated and, consequently, the process of preparing the Company's monthly financial information required significant manual processes, including numerous manual journal entries to the general ledger. In addition, prior to the middle of 2004, the Controller's Office lacked formal written procedures on journal entries and account reconciliations, had no standardization for documentation to support journal entries, and permitted employees to sign off on journal entries for other employees. Since the release of the OFHEO Report, Fannie Mae has made changes to the structure and personnel of the Controller's Office, and to the Company's approach to the development of accounting policy. The Controller's Office, with active support from senior management and considerable reliance on outside expertise, has made significant efforts to augment its resources and the procedures and systems used in the development and oversight of accounting policies and financial reporting. B. Overview Prior to September 2004 1. Range of Responsibilities As of September 2004, the Controller's Office had responsibilities that included the preparation of the Company's financial statements and its external and internal financial reporting. It handled Fannie Mae's accounting, payroll, tax, insurance, and procurement activities, managed the Company's budget and expenses, and performed business planning.(1920) 2. Senior Vice President and Controller From 1999 through the beginning of 2005, Spencer served as Senior Vice President and Controller of Fannie Mae. Spencer joined Fannie Mae in 1991 as a Director in the Financial Information division of the Controller's Office. She became a Director in the Financial Reporting division in 1993, and a Vice President one year later, reporting to former Controller Michael Quinn. She began reporting to Rajappa when he replaced Quinn as Controller in 1994, and, from 1994 to 1999, she was responsible for managing Financial Reporting (as well as Financial Standards for a brief period of time). In 1999, she succeeded Rajappa as Controller. Spencer reported to CFO Howard during her tenure as Controller. Prior to joining Fannie Mae in 1991, Spencer worked at KPMG and as the controller of a professional services firm that provided temporary office help. She holds a bachelor's degree in accounting, but she is not a certified public accountant. As Controller, Spencer appeared regularly before the Audit Committee to discuss the Company's financial statements. She briefed the Audit Committee as to the - ------------ (1920) See, e.g., Undated L. Spencer: Job Description, Zantaz document 126448. 506
structure and content of the Company's annual report on Form 10-K,(1921) and she met with the Audit Committee Chairman at the end of each quarter to review the quarterly earnings reports before their release.(1922) She also participated in periodic presentations to the Audit Committee concerning the critical accounting policies of the Company.(1923) 3. Key Units within Controller's Office As of September 2004, the Controller's Office was divided into six units: (1) Accounting Policy and Corporate Tax ("Financial Standards"); (2) Financial Reporting and Planning ("Financial Reporting"); (3) Operations & Systems; (4) Financial & Procurement Services; (5) Corporate Risk & Insurance Management; and (6) Housing & Community Development.(1924) Of these units, Financial Standards, Financial Reporting, and Operations & Systems are described briefly below. (a) Financial Standards (1) Functions and Responsibilities Aside from its tax functions, Financial Standards was responsible primarily for formulating Fannie Mae's accounting policies by interpreting and applying rules promulgated by the FASB, the Emerging Issues Task Force, and the SEC. Financial Standards also provided accounting training for employees, and helped Financial Reporting draft language for Fannie Mae's periodic financial disclosures (including annual and quarterly reports). - ------------ (1921) See, e.g., Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 17, 2004, FMSE 504818-30; Audit Committee Update: Financial Statement Review, dated Feb. 17, 2004, FMSE 221811-15; Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 18, 2003, FMSE 504756-62. (1922) See, e.g., Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Jan. 23, 2004, FMSE 504806-15, at FMSE 504806. (1923) See, e.g., Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 17, 2004, FMSE 504818-30, at FMSE 504820-21 (Spencer presentation on other than temporary impairment); Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Nov. 17, 2003, FMSE 504796-805, at FMSE 504800 (Spencer presentation on amortization of deferred price adjustments); Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Feb. 18, 2003, FMSE 504756-62, at FMSE 504759 (Spencer presentation on allowance for losses, deferred premium discount and deferred guaranty fees, and fair value of purchased options). (1924) See Controller's Department - Overview, dated Sept. 2004, FMSE-IR 1550. 507
(2) Senior Vice President -- Accounting Policy and Corporate Tax Jonathan Boyles served as the head of Financial Standards from 1997 through the beginning of 2005. He joined Fannie Mae in 1994 as a manager in Financial Standards. He was promoted to Director in 1998; to Vice President in 2000; and to Senior Vice President in May 2004. Prior to joining Fannie Mae, Boyles spent four years as an accountant at Grant Thornton, serving clients in the construction, retail, banking, and not-for-profit sectors. He also worked at KPMG for two years, providing auditing and consulting services to mortgage banks. He holds a bachelor's degree in accounting and is a certified public accountant.(1925) Boyles reviewed all significant accounting policies and generally had final say at Fannie Mae on accounting issues. He provided accounting advice to all areas of Fannie Mae, including advice to the business units on how to account for new types of transactions. In some cases, his advice took the form of a memorandum; in other cases, it was informal and undocumented. (3) Process of Developing Accounting Policies Fannie Mae's accounting policies were set forth in formal memoranda compiled in an accounting policy manual. As a matter of course, Financial Standards sought KPMG's approval prior to finalizing or amending these accounting policies.(1926) Also as a matter of course, Financial Standards briefed Howard and Spencer on significant accounting policies. According to Boyles, Howard and Spencer did not provide formal written approval of these policies; rather, Boyles deemed the policies to be approved unless Howard or Spencer voiced concerns with them.(1927) On occasion, Financial Standards also consulted with the FASB, the SEC, and/or OFHEO prior to finalizing accounting policies. In these instances, Financial Standards typically presented these bodies with an issue raised by a Fannie Mae transaction or practice and sought guidance as to whether the Company's approach conformed to accounting standards. Financial Standards lacked sufficient staffing to be engaged in any systematic way in the implementation of accounting policies by Controller's Office or business unit personnel. As noted above, Financial Standards was available as a resource - ---------- (1925) See Tr. of Aug. 3, 2004 OFHEO Dep. of J. Boyles, at 8:7-12. (1926) See id. at 30-32, 34. (1927) See id. at 20-24. 508
if employees wanted to consult with it on the implementation of its accounting policies (e.g., development of business unit practices and procedures), but it rarely took the initiative in monitoring such implementation, and did not view this as part of its responsibility. (b) Financial Reporting (1) Functions and Responsibilities Financial Reporting oversaw the monthly general ledger closing process and prepared reports and analyses necessary to compile monthly, quarterly, and annual reports for both internal and external reporting. It also participated in the implementation of new accounting policies and served as an accounting liaison for the various portfolio investment and credit guaranty business units.(1928) (2) Senior Vice President -- Financial Reporting and Planning Janet L. Pennewell led Financial Reporting from 1999 through early 2005. Pennewell joined Fannie Mae in 1984 as a financial analyst in Financial Reporting, but she moved to Investor Relations for a number of years before returning to the Controller's Office in 1996. Pennewell became Vice President -- Financial Reporting in 1999, and was promoted to Senior Vice President -- Financial Reporting and Business Planning in May 2004. Pennewell holds a bachelor's degree in accounting and a master's degree in finance and is a certified public accountant.(1929) (3) Closing Process Financial Reporting managed the closing process in support of Fannie Mae's internal financial reports and its external financial statements, quarterly reports, and annual reports. During the closing process, accounting information for a particular month was loaded to, and finalized in, the general ledger system. In general, journal entries were to be prepared by one person (usually the person responsible for the activity) and approved by another person. The level of approval required varied depending on the amount of an entry and the time frame in which it was submitted for approval. - ---------- (1928) See Controller's Business Unit Finance Compliance Plan 2004-2005, dated Dec. 29, 2004, FMSE-IR 227416-59, at FMSE-IR 227417. Financial Standards supported this effort, providing Financial Reporting with inserts for Fannie Mae's various financial reports and with blank templates and tables to populate with financial data. Financial Reporting, in turn, supplied data to Investor Relations for inclusion in the Company's earnings releases. (1929) See Tr. of June 15, 2004 OFHEO Dep. of Janet L. Pennewell, at 6:19-9:14. 509
In addition to the reconciliation of general ledger accounts, Financial Reporting performed ongoing analyses of the income statement, including interest income and expense, guaranty fees, administrative expenses, and debt extinguishment gains and losses. These analytical procedures functioned as reasonableness tests and were designed to ensure completeness and accuracy, proper classification, and the substantiation of account balances.(1930) (c) Operations & Systems (1) Functions and Responsibilities Operations & Systems performed the majority of accounting for the Company's transactions, including debt issuance, derivatives, and securities held in the mortgage and liquid investment portfolios. These accounting responsibilities included application of FAS 91 and FAS 133. (2) Senior Vice President -- Operations & Systems This position was created in May 2004 and had not been filled as of September 2004. Mary Lewers, who became Vice President -- Financial Accounting in May 2004, served as the acting head of Operations & Systems,reporting primarily to Spencer (with reporting on certain matters to Pennewell).(1931) Lewers joined Fannie Mae in 1983 as a financial analyst in the Controller's Office. She became a Director in 1991, and, from 1993 to 2000, she supervised the development of various accounting systems, including debt, premium/discount amortization, derivatives, and liquid investment portfolio systems. In 2002, Lewers assumed responsibility for securities accounting. Lewers holds a master's degree in finance and a bachelor's degree in accounting and is a certified public accountant. C. Findings In performing its critical responsibilities, the Controller's Office was hindered by significant understaffing, a lack of personnel with the requisite experience and expertise to handle the complex accounting tasks required of Fannie Mae's Controller's Office, and poorly integrated computer systems that required significant human intervention. - ------------- (1930) See Financial Reporting, Monthly Financial Analytics to Support Closing Policies and Procedures, dated Aug. 17, 2004, FNMSEC 2392-98, at FNMSEC 2396. (1931) See Organizational Chart, dated Sept. 2004, FMSE-IR 1555. 510
1. Inadequate Resources (a) Working Environment During Spencer's tenure as Controller, the Controller's Office assumed substantial new responsibilities. It absorbed a number of functions from other areas of the Company, including e-business billing, processing of accounts receivable, securities accounting, and procurement. It also became subject to new reporting requirements arising from the Company's registration with the SEC. Moreover, the Controller's Office had to accommodate Fannie Mae's growing business and the introduction of such complex new accounting standards as FAS 133, FAS 149, and others. As Spencer acknowledged in 2004, the consequence of these increasing demands was that the "workload has quadrupled for many of our areas in a very short time."(1932) Prior to 2004, Spencer accommodated this increase in workload with only marginal increases in headcount. Between 2000 and 2003, headcount in the Controller's Office increased by only 13 employees.(1933) The resulting strains on Controller's Office personnel were recognized by Internal Audit.(1934) Spencer acknowledged that she tried to run the Office very efficiently from a resource perspective. She said that she looked closely at requests to increase headcount to determine if needs were permanent or whether they resulted from high-intensity but temporary surges in workload. Although some employees attributed understaffing to management's perception of them as non-revenue-generating "overhead" that should not be increased unless absolutely necessary, Spencer said that no one directed her to maintain lean staffing. It was not until late 2003 - after the Roger Barnes investigation, revelations of accounting problems at Freddie Mac, and the onset of OFHEO's Special Examination of Fannie Mae subjected the Controller's Office to heightened scrutiny - that Spencer acknowledged that the rising tide of work was not a temporary surge and that more staff was necessary to accommodate it.(1935) As Spencer later noted in a 2004 - ---------- (1932) See Communication from L. Spencer to All Controller Staff, dated Apr. 22, 2004, Zantaz document 907239. (1933) See Controllers Headcount 1999-2004, dated Oct. 3, 2004, FMSE-IR 379167. (1934) In a departure message in October 2003, Paul Jackson, the Internal Audit Director for Finance, warned Rajappa that "staffing is way to [sic] low [in the Controller's Office] to support business needs, including internal audit regulatory examinations, etc." See E-mail from P. Jackson to S. Rajappa, dated Oct. 17, 2003, FMSE-E 1031680-81, at FMSE-E 1031681. (1935) See L. Spencer: 2003 Performance - Self Assessment, dated 2003, FMSE-E 22031-38, at FMSE-E 22031 ("The unfolding accounting and control problems at Freddie Mac have been a drain on us during the year.... I have come to accept that [our] 511
memorandum to Howard and Mudd, "[w]e now acknowledge that the world with our regulator has changed and the workload will not abate in the near future."(1936) Spencer met with Human Resources in November 2003 to obtain recruiting assistance; launched a project with HR to recruit candidates for key positions within the Controller's Office,(1937) and committed herself, as one of her 2004 corporate goals, to "[r]eview the organizational structure of the Controller's Department to ensure the company is positioned to handle the increasing complexity of accounting requirements while continuing to meet the needs of the business."(1938) From November 2003 through September 2004, the number of employees in the Controller's Office increased from 206 to 233 (with 56 new employees hired, but approximately half that number lost to attrition). Most of the new hires were placed in Financial Reporting, Financial Standards, and Operations & Systems.(1939) The Controller's Office faced morale issues as well. Employee survey data from as far back as 1999 reveal widespread employee concerns about the culture within the Controller's Office.(1940) Despite adoption of an action plan to address employee concerns, employee ratings of the Controller's Office in 2001 declined from 1999 levels in numerous survey categories.(1941) Another action plan was adopted, similar - ------------ new responsibilities will not lighten up in the near future, so I am adjusting our staffing plans accordingly so that we can address these needs without sacrificing our accounting work."). (1936) See Mem. from L. Spencer to T. Howard and D. Mudd, dated July 9, 2004, FMSE-IR 381407-08, at FMSE-IR 381407. (1937) See Controller's Top Gun Project, dated Nov. 4, 2003, FMSE-IR 375611-12. (1938) E-mail from L. Spencer to J. Pennewell, dated Dec. 16, 2003, Zantaz document 4155147. (1939) See Controller's Organizational Overview, dated Sept. 22, 2004, FMSE-IR 361514-25. (1940) See Focus Group Meeting - Controller's Level 4, dated Nov. 15, 1999, FMSE-IR 522215-17; 2001 Employee Perspectives Survey: Controller's Office, dated Aug. 2001, FMSE-IR 308623-26. (1941) See 2001 Employee Perspectives Survey: Controller's Office, dated Aug. 2001, FMSE-IR 308623-26, at FMSE-IR 308624; see also EPS Follow-up Meetings, FMSE-IR 308743-55. 512
in many respects to that adopted back at the end of 1999 (increased accessibility to Spencer, and steps to encourage information flow and discourage fear of retaliation).(1942) In the late summer of 2003, the OCC investigated the allegations by Roger Barnes of substantial cultural problems in the Controller's Office, including his allegation that he had been ostracized and retaliated against for raising questions about Fannie Mae's accounting practices. In its October 2003 investigative report, the OCC found, among other things, that: (1) "for a significant number of employees in the Controller's Office the Reporting Environment is unacceptable"; (2) while not a universal view within the Office, significant numbers of staff felt that "the environment there is not conducive to the full, open and frank discussion of issues and concerns relating to accounting issues and practices" and "to be successful in the Controller's Office `you are either on the team or not'"; (3) some managers were viewed as lacking "people skills" such that "when issues are raised it is taken as a challenge to the manager's integrity rather than as an opportunity for full and frank discussion"; and (4) "there is not wide communication as to why things are done the way they are done and . . . some employees may work in silos which may unnecessarily create issues and concerns."(1943) As a result of its findings, the OCC prescribed a series of remedies, including full implementation of the 2001 action plan, an "all hands" meeting with Spencer, and training for certain managers within the Controller's Office.(1944) Many of these prescribed steps were ultimately taken.(1945) Spencer provided updates on the status of the plan to the Legal Department, acknowledging that progress was being made.(1946) - ---------- (1942) See Undated Controller EPS Draft Action Plan, FMSE-IR 308741-42; Undated EPS Divisional Action Plan, FMSE-IR 308831-32. (1943) See Office of Corporate Compliance: Decision # 2003-1 (Part B, Reporting Environment), dated Oct. 6, 2003, FMSE 24562-66, at FMSE 24563-64. This OCC report, and other investigatory activities related to Barnes's allegations, are discussed at length in Chapter IX of this Report. (1944) See id. at FMSE 24566. (1945) See Controllers Plan, dated May 5, 2004, FMSE-E 37819-22. (1946) In response to concerns expressed by Chief Compliance Officer Remy that Spencer still had not fully implemented the OCC's remedies several months after the OCC issued its report, Spencer acknowledged that, "with the demands of the various audits and OFHEO reviews, the overspill from the Freddie Mac restatement, and the work we've been supporting with the performance calibration season, I have had reduced time to spend on implementing the plan." Mem. from L. Spencer to D. Remy, dated Feb. 5, 2004, FMSE-E 37823-25, at FMSE-E 37824. 513
(b) Expertise and Experience The Controller's Office in a number of cases employed individuals in senior positions who lacked the requisite expertise to perform or supervise the complex tasks that they were assigned. Although Boyles was the Company's designated expert on accounting policy, he had only limited experience in consulting on technical accounting matters prior to joining Fannie Mae. Boyles was the ultimate approver of accounting policy and his supervisor (Spencer) did not have the requisite accounting expertise to act as a meaningful check on his development of accounting policies. Moreover, Boyles stated that he hired individuals based upon their policy and industry experience rather than their accounting expertise, and that he did not have anyone on his staff with prior hedging or derivatives experience. Similarly, as Vice President -- Financial Accounting, Lewers was responsible for supervising debt accounting, derivatives accounting, securities accounting and portfolio analytics, and for developing systems that supported accounting in these areas. Yet she acknowledged, for example, that she relied heavily upon Financial Standards and her staff to ensure that the accounting for Fannie Mae's derivatives portfolio complied with GAAP.(1947) 2. Deficiencies in the Closing Process (a) Inadequate Policies and Procedures Until 2004, Financial Reporting lacked formal written policies and procedures to support the closing process, analytical review, journal entries, or account reconciliations. It was not until the summer of 2004, as implementation of SOX 404 required the Company to formally document policies and procedures, that Financial Reporting took action to address these deficiencies.(1948) Prior to May 2004, the Controller's Office lacked standard requirements for documentation to support journal entries. As a result, Controller's Office staff employed inconsistent practices as to the amount of supporting detail that they provided.(1949) In some instances, no meaningful supporting documentation (or no supporting documentation at all) was provided for the journal entry.(1950) - ---------- (1947) See Tr. of July 13, 2004 OFHEO Dep. of M. Lewers, at 152:15-154:14, 215:21-216:22, 257:1-259:5. (1948) See Financial Reporting Monthly Financial Analytics to Support Closing Policies and Procedures, dated Aug. 17, 2004, FNMSEC 2392-430, at FNMSEC 2396. (1949) There was also no requirement to store supporting documentation in a central location; instead, these documents remained with the business units. See 514
In addition, up until the end of 2004, individuals other than the listed preparer of a journal entry were permitted to sign off on the preparer's behalf. Members of the Controller's Office acknowledged that there were instances when employees signed journal entry forms on behalf of preparers who were not present in the office. Typically, the person signing indicated that he or she was doing so on behalf of the preparer. However, during the period 1998 through 2002, a number of "on top" adjustments were entered on behalf of an employee who had not prepared them, although her name appears on them. (b) Inadequate Technological Support The Controller's Office failed to support the closing process with adequate technology. The Controller's Office maintained numerous spreadsheets (commonly referred to at Fannie Mae as "End-User Controls" or "EUCs") and ad-hoc reporting tools to compensate for poorly integrated systems.(1951) Extensive reliance on EUCs and the manual processing and analysis of data created a weak control environment conducive to errors. Errors arose not only from manual activity, but also from the challenges of creating adequate levels of employee knowledge across a multitude of EUCs and systems. An example of the resulting control risks occurred in October 2003, when manual entries on an EUC and a lack of familiarity with the procedures associated with a new accounting standard contributed to an approximately $1.1 billion computational error in Fannie Mae's third quarter earnings press release.(1952) - ---------------- BearingPoint Review of Current Controls and Processes Within Fannie Mae's Controller's Office, dated Apr. 5, 2005, FMSE-IR 281271-93, at FMSE-IR 281280. (1950) For example, a January 1999 journal entry in the amount of $3.9 million - an entry that helped Fannie Mae to meet its 1998 EPS target for bonus plan (AIP) purposes - was entered for the stated purpose of "record[ing] miscellaneous income," with no further supporting documentation. This is discussed further in the FAS 91 section of this Report. When BearingPoint reviewed the Company's closing process in the spring of 2005, it found a number of journal entries to be inadequate or could not recreate them based on the backup provided to it. See id. (1951) In its Spring 2005 review of the Company's closing process, BearingPoint found more than fifty EUCs that were deemed to be critical. See id. (1952) In this instance, a Controller's Office employee who normally performed calculations to capture the fair value of certain instruments was on vacation during the September 2003 month-end closing process. Another employee performed the analysis in lieu of the vacationing employee, and adjusted a spreadsheet without realizing that the spreadsheet had already been updated to incorporate the impact of FAS 149. The resulting error was a $1 billion understatement of total assets and stockholders equity in the Company's earnings press release, which the Company 515
D. Subsequent Developments Since the publication of the September 2004 OFHEO Report, Fannie Mae has undertaken significant reforms of the structure, composition, policies, and practices of the Controller's Office. A number of these reforms are the product of recommendations provided by outside consultants retained by the Company - and especially Mercer (with respect to changes to structure and staffing) and BearingPoint (with respect to the closing process). Although substantial personnel and structural changes have taken place, a number of important areas of reform (such as systems upgrades and staffing increases) remain very much in process. 1. Change in Leadership Spencer resigned as Controller on January 17, 2005, and Fannie Mae appointed David Hisey, a certified public accountant, to replace her.(1953) Prior to joining the Company, Hisey was a managing director at BearingPoint and an audit partner at KPMG.(1954) 2. Structural and Related Personnel Changes Following the Company's September 2004 agreement with OFHEO(1955) and the resulting organizational redesign work by Mercer, the Company has undertaken a broad reorganization of the Controller's Office, including the movement of some functions outside the Controller's Office (including accounting policy and business planning and forecasting functions) and restructuring of functions within the Office. (a) Accounting Policy In January 2005, Boyles resigned as head of Financial Standards. Fannie Mae subsequently split Financial Standards into two separate functions: Accounting - ------------ subsequently corrected by filing an amended 8-K with the SEC. See Mem. from J. Wall, dated Dec. 3, 2003, FMSE-E 1090737-40; Fannie Mae, Current Report on Form 8-K, dated Oct. 29, 2003, available at http://www.fanniemae.com/ir/sec/index.jhtml?p=Investor+Relations&s= SEC+Filing s(follow hyperlink to "All SEC Filings," then "Form 8-K)". (1953) See Fannie Mae, Current Report on Form 8-K, filed Jan. 21, 2005, available at http://ccbn.10kwizard.com/cgi/convert/pdf/ FEDERALNATIONAL8K.pdf?pdf=1&repo=tenk&ipage=3208349&num=2&pdf=1&xml= 1&odef=8&dn=2&dn=3. (1954) See id. (1955) See Agreement between the SRC and OFHEO, dated Sept. 27, 2004, available at http://www.ofheo.gov/media/pdf/fnmagreement92704.pdf. 516
Policy (reporting to the CFO) and Corporate Tax (reporting to the Controller). In contrast to Financial Standards, the new Accounting Policy unit is responsible for ensuring consistent understanding and implementation of accounting standards across the Company.(1956) In June 2005, Fannie Mae appointed Scott Blackley as the Senior Vice President -- Accounting Policy. He is now responsible for ensuring consistent understanding and implementation of accounting standards across the Company. Prior to joining Fannie Mae, Blackley worked at America Online, where he was Vice President -- Accounting Policy and Assistant Controller. Previously, he served as a partner at KPMG, and as a professional accounting fellow in the SEC's Office of the Chief Accountant. Blackley is a certified public accountant, and he is a technical expert on derivatives, business combinations, and revenue recognition.(1957) In addition to Blackley, Fannie Mae appointed Eric Schuppenhauer, a certified public accountant, as Senior Vice President -- Accounting Policy in August 2005. Schuppenhauer will have responsibility for evaluating Fannie Mae's prior accounting policies and practices relating to financial restatements as well as implementing new or changed accounting standards. Prior to joining Fannie Mae, Schuppenhauer worked at Credit Suisse First Boston as a managing director for Global Accounting Policy and Assurance. Previously, he was a senior advisor to the Chief Accountant of the SEC, an SEC Professional Accounting Fellow, and a senior manager in KPMG's Department of Professional Practice.(1958) (b) Financial Reporting Pennewell resigned as Senior Vice President for Financial Reporting and Planning in January 2005. Fannie Mae named Paul Noring to replace Pennewell in May 2005. Prior to joining Fannie Mae, Noring was an audit partner at PWC. At PWC, Noring developed considerable experience in derivatives risk management operations, - ------------ (1956) See Mercer Oliver Wyman Consulting, Organizational Design Project - Organization Redesign Discussion Notes - Corporate Senior Vice President Meeting, dated May 4, 2005, FMSE-IR 552781-96, at FMSE-IR 552791. (1957) See Fannie Mae press release, dated May 18, 2005, available at http://www.fanniemae.com/newsreleases/2005/3519.jhtml?p=Media&s=News+Rele ases. (1958) See Fannie Mae press release, dated Aug. 24, 2005, available at http://www.fanniemae.com/newsreleases/2005/3589.jhtml?p=Media&s=News+Rele ases. 517
controls and business process environments surrounding financial instrument accounting, fair-value measurements, and allowance for loan loss methodologies.(1959) (c) Operations & Systems Lewers departed Fannie Mae in 2005. Subsequently, Fannie Mae split Operations & Systems into two separate offices that each report directly to the Controller: Accounting and Financial Controls & Systems. In May 2005, Fannie Mae appointed Greg Kozich to serve as Senior Vice President -- Accounting. Prior to joining Fannie Mae, Kozich was an audit partner at PWC where he managed complex technical accounting, auditing, and operational matters relating to financial instruments, with an emphasis on derivatives and mortgage-backed security accounting.(1960) Financial Controls & Systems is responsible for financial process controls, accounting systems, infrastructure, information technology, and automation.(1961) In June 2005, Fannie Mae appointed Mary Doyle to serve as Senior Vice President -- Financial Controls & Systems. Previously, Doyle was Vice President -- Corporate Accounting at Sallie Mae and a tax partner at Arthur Anderson.(1962) (d) Valuation & Price Verification As part of the reorganization, Fannie Mae has established an independent price verification and valuation function that reports directly to the Controller. This function is intended to provide an independent review of fair values derived by the Mortgage Portfolio business before they are included in the Company's audited financial statements.(1963) - ------------ (1959) See Fannie Mae press release, dated May 18, 2005, available at http://www.fanniemae.com/newsreleases/2005/3519.jhtml?p=Media&s=News+Rele ases. (1960) See id. (1961) See Controllers Business Reporting, dated Dec. 9, 2005, at FMSE-IR 699236-38. (1962) See Fannie Mae press release, dated May 18, 2005, available at http://www.fanniemae.com/newsreleases/2005/3519.jhtml?p=Media&s=News+Rele ases. (1963) See Mercer Oliver Wyman Consulting, Organizational Design Project - Organization Redesign Implementation Supporting Materials and Discussion Notes Prepared for the Compliance Committee, dated Mar. 24, 2005, FMSE-IR 289760-804, at FMSE-IR 289773. 518
3. Staffing Levels As a result of both the organizational redesign build-out and the restatement effort, the Company is in the process of substantially increasing the size of the Controller's Office. As of December 2005, headcount was approximately 260 (versus approximately 230 in September 2004). According to Hisey, headcount is expected to increase by 150 employees in 2006 (of whom approximately sixty percent will be dedicated to accounting operations), with a long-term goal of growing to 450 employees.(1964) 4. Changes to the Closing Process As noted above, Fannie Mae retained BearingPoint in the spring of 2005 to conduct a review of the Company's systems and procedures related to the closing process.(1965) BearingPoint found that Fannie Mae's closing process required excessive reliance on human intervention, lacked appropriate policies governing journal entries, and needed to fill key vacancies.(1966) BearingPoint recommended that the Controller's Office (1) "[i]ncrease automation and develop Company standards for journal entry approval and support retention"; (2) "[a]dd centralized review and develop standard policies for adjustments/reconciliations"; (3) "[f]ill crucial vacancies and formalize training programs"; and (4) "[r]educe use of EUCs and [database] mods, and increase core system capabilities."(1967) The Company hopes to implement the majority of the recommendations by the end of 2006. According to Hisey, the Controller's Office has begun implementing BearingPoint's recommendations. He stated that, as of December 2005, the closing process remained a fairly manual process, but that work was underway to automate most of it by the end of 2006. Hisey also noted ongoing efforts to standardize the journal entry process, and to improve the overall functionality and capability of Controller's Office computing systems. The Company has also initiated a broader multi-year effort to overhaul and fully integrate the various systems within the Finance group (dubbed the "Finance Transformation Project") which currently remains in the planning stages. - ------------ (1964) Update Briefing with David Hisey, dated Dec. 13, 2005. (1965) See BearingPoint, Review of Current Controls and Processes Within Fannie Mae's Controller's Office: Presentation to the Audit Committee, dated Apr. 18, 2005, FMSE 503077-88, at FMSE 503079. (1966) See id. at FMSE 503084-87. (1967) Id. at FMSE 503079. 519
CHAPTER VIII: EXECUTIVE COMPENSATION I. INTRODUCTION In the September 27, 2004 Agreement between Fannie Mae and OFHEO, Fannie Mae agreed to report on the Company's "compensation regime and its relation to strategic plans and their impact on accounting and transaction decisions and any revisions to avoid inappropriate incentives."(1968) In accordance with this undertaking, the SRC initiated a two-part review: (1) a historical analysis of Fannie Mae's executive compensation structure and its relationship to efforts to meet financial goals (such as earnings per share ("EPS") targets); and (2) a prospective assessment of the Company's compensation structure and recommendations for revisions thereto. The SRC assigned to Paul, Weiss the historical analysis; specifically, the SRC asked Paul, Weiss to review Fannie Mae's compensation programs and to assess the role of EPS or other financial indicators as a compensation trigger.(1969) The SRC engaged Semler Brossy Consulting Group ("Semler Brossy") to evaluate the Company's current compensation structure and to make recommendations on revisions thereto.(1970) During the course of our review, OFHEO requested, and the SRC agreed, that we also review the role that the Legal Department played in compensation decisions. OFHEO's request stemmed from two anonymous letters that accused Fannie Mae's Legal Department attorneys of excessive and inappropriate involvement in compensation decisions and, specifically, of improperly attempting to cloak compensation decisions with confidentiality under the guise of the attorney-client privilege.(1971) On the issues assigned to us by the SRC, we reach the following conclusions: - ---------------------- (1968) September 2004 Agreement P. IV. (1969) Paul, Weiss was not asked to review or analyze employment contract issues or individual compensation issues involving Franklin D. Raines or Timothy Howard. Paul, Weiss also did not review compensation plans for employees below the Director level and did not consider individual compensation decisions for employees at any level. For our analysis of the relationship at Fannie Mae between executive compensation and FAS 91 accounting methods in 1998, see supra Chapter IV. (1970) Semler Brossy Report on Appropriate Compensation Structure and Incentives for Fannie Mae Management, dated Feb. 23, 2005, FMSE-KD 25819-985 (hereinafter "Semler Brossy Report"). Semler Brossy presented its report and recommendations to OFHEO on February 24, 2005. (1971) Letter from OFHEO to Alex Young K. Oh, dated Feb. 2, 2005 (requesting that the two anonymous letters remain confidential). 520
- Between 1998 and 2003, Fannie Mae's target compensation levels consistently lagged behind those of the Company's "comparator corporations." Therefore, to facilitate payment of market-competitive compensation for executives, Fannie Mae intentionally set its "maximum" EPS target at levels that the Company expected to achieve. Because the expected EPS number was not an aggressive goal, the Company regularly exceeded it. Consequently, maximum bonus, stock, and stock option awards were triggered, but this did not result in payment of executive compensation in excess of levels specified by the Company's formal compensation philosophy. Beginning in 2002, Fannie Mae attempted to set EPS targets to coincide with target bonuses pursuant to Fannie Mae's written compensation philosophy. However, due to unanticipated shifts in executive compensation generally, Fannie Mae's executive compensation continued to lag behind market levels, and Fannie Mae executives received total compensation at market levels only if the maximum EPS bonus targets were met. - Beginning in 1999, the Company aligned its compensation practices with its business objective to double EPS in five years. The Company's use of EPS as the sole determinant for whether bonuses would be awarded and the size of these bonus awards, as well as for determining its performance-based stock grant plan award ("Performance Share Plan" or "PSP"(1972)), was not a best practice. Fannie Mae should have (and has since) developed additional financial and non-financial measures for its bonus awards and PSP to reduce reliance on EPS in determining compensation. - Non-financial corporate performance goals (as well as EPS) played a part in executives' long-term compensation through the Company's PSP. These goals were set, and performance against them was assessed, by the Compensation Committee of Fannie Mae's Board of Directors (the "Compensation Committee") based on a report from management. Management consistently tendered excessively positive self-analyses to the Compensation Committee in these areas. The failure to present a balanced view of the Company's performance leaves the impression that management did not provide full and complete information to the Compensation Committee. - ---------------------- (1972) Fannie Mae alternately refers to the PSP as the "Performance Share Program." 521
- We found no evidence to support the allegations that the Legal Department was inappropriately involved in executive compensation decisions. II. PAUL, WEISS'S INVESTIGATION A. Document Review Paul, Weiss has reviewed approximately 235,000 pages of documents relating to executive compensation issues, including all executive compensation documents in the Legal Department's files. We reviewed both paper and electronic documents, consisting primarily of (1) correspondence among Fannie Mae management regarding compensation issues, and (2) material prepared by Fannie Mae management and provided to Fannie Mae's Board of Directors and Compensation Committee. We focused our review on the period between and including 1998 and 2004.(1973) In addition, we reviewed documents prepared by Semler Brossy, including its February 23, 2005 report to the SRC on the Company's then present compensation structure and recommendations for reform. B. Interviews We conducted thirteen interviews concerning executive compensation issues. Two of these interviews were with Fannie Mae's compensation consultants: Alan Johnson, compensation consultant to the Company's management, and Roger Brossy, consultant to the Compensation Committee.(1974) The remainder of the interviews were with current or former Fannie Mae employees and directors with responsibilities relating to executive compensation.(1975) The former head of Fannie Mae's Human Resources - ---------------------- (1973) Due to the events surrounding the Special Examination, the Company did not award bonuses to Senior Vice Presidents and above for 2004. Other incentive-based compensation plans were decreased and/or suspended pending the release of Fannie Mae's restatement. Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 18, 2005, FMSE-IR 422500-15, at FMSE-IR 422505. (1974) In addition to these interviews, we met with Roger Brossy and his team from Semler Brossy both at the commencement of Semler Brossy's review and prior to the issuance of the February 23, 2005 report. Paul, Weiss reviewed Semler Brossy's draft report before it was transmitted to the Board of Directors and was satisfied with the modifications to Fannie Mae's executive compensation proposed therein. Paul, Weiss also consulted with Brossy several times by telephone during the course of the investigation. (1975) We interviewed Lorrie Rudin, Emmanuel Bailey, Rebecca Senhauser, Anne M. Mulcahy (former Chair of the Compensation Committee), Thomas E. Donilon, Donald M. Remy, Judith Dunn, Anthony Marra, and Daniel H. Mudd. Rudin is Fannie Mae's Director -- Executive Compensation and Benefits. Paul, Weiss interviewed Rudin on three occasions: Two interviews were conducted at Fannie 522
Department, Kathy G. Gallo, and former CFO, Timothy Howard, declined our requests for an interview. III. EXECUTIVE COMPENSATION AT FANNIE MAE A. Executive Compensation Overview 1. Fannie Mae's Charter Fannie Mae's compensation philosophy and practices were (and continue to be) driven in large part by its Charter, which is set by federal statute. In relevant part, the Charter provides that the Board of Directors: shall have the power . . . to fix and to cause the corporation to pay such compensation [to Fannie Mae employees] for their services, as the board of directors determines reasonable and comparable with compensation for employment in other similar businesses (including other publicly held financial institutions or major financial services companies) involving similar duties and responsibilities, except that a significant portion of potential compensation of all executive officers . . . shall be based on the performance of the corporation.(1976) The Charter thus provided the framework that Fannie Mae used in setting executive compensation programs and levels, including the requirements (1) that an individual officer's compensation be "comparable" to that for an equivalent position at other diversified financial institutions, and (2) that a "significant portion" of such compensation be tied to Fannie Mae's corporate performance. 2. Fannie Mae's Compensation Philosophy Fannie Mae implemented its Charter through a "compensation philosophy" that Fannie Mae management provided to, and that was approved by, the Compensation Committee of the Board of Directors.(1977) Although the Company revised Mae to coordinate document production in response to Paul, Weiss's requests and to review the structure of Fannie Mae's executive compensation programs; the final interview was conducted at Paul, Weiss's offices. Paul, Weiss also spoke several times with a Fannie Mae employee who requested anonymity. (1976) 12 U.S.C. Section. 1723a(d)(2), Fannie Mae Charter Act Section. 309(d)(2) (2000) (emphasis added). (1977) Compensation Philosophy, in EXECUTIVE COMPENSATION PROGRAMS, dated Sept. 2003, FMSE-IR 283130-228, at FMSE-IR 283141 (hereinafter "SEPT. 2003 EXECUTIVE COMPENSATION PROGRAMS"). The Compensation Committee approved 523
the document over the years, the basic philosophy, as summarized in the materials management provided to the Compensation Committee, remained substantially the same during the relevant time period (1998-2004): FANNIE MAE'S PHILOSOPHY OF EXECUTIVE COMPENSATION The central tenets of Fannie Mae's compensation philosophy are pay for performance and comparability. Pay for performance is reflected strongly in the structure of Fannie Mae's compensation programs. It is the core principle underlying the programs. Other than base salary, all major elements of Fannie Mae's compensation program are tied to annual and long-term performance goals. Furthermore, through the use of stock vesting over multi-year terms, Fannie Mae tightly aligns the interest of executives with those of shareholders. Comparability is reflected in both the design of the compensation programs and in the levels of pay delivered to executives.(1978) Fannie Mae implemented this philosophy by engaging third party compensation consultants to gather data on compensation at corporations with which Fannie Mae competed in the employee market (the "comparator group" or "comparator corporations"). The compiled data included information on salary, annual bonus, and long-term incentives. The Company then set its target cash compensation level (salary plus annual bonus) for each executive at the median (fiftieth percentile) of the executive's peer group, and set his/her total compensation target (salary, annual bonus, and long-term incentives) at the peer group's sixty-fifth percentile.(1979) - ----------- the 2003 philosophy in its January 20, 2003 meeting. Minutes of the Meeting of the Compensation Committee of the Board of Directors of Fannie Mae, dated Jan. 20, 2003, FMSE 505241-271, at FMSE 505242-44. (1978) 2003 Executive Compensation Program Overview, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Nov. 17, 2003, FMSE-IR 17540-625 (hereinafter "NOV. 2003 AGENDA"), at FMSE-IR 17595. We chose 2003 as representative of the statements provided to the Board on an annual basis, although some changes had been made prior to that time. (1979) Compensation Philosophy, in SEPT. 2003 EXECUTIVE COMPENSATION PROGRAMS, at FMSE-IR 283141-42; see also Overview of Executive Compensation Recommendations, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Nov. 16, 1998, FMSE-IR 14335-480, at FMSE-IR 14339-40 (hereinafter "NOV. 1998 AGENDA") ("Stock-based incentives are to be used to make up the balance between 50th percentile cash compensation and the 65th percentile total compensation objective."). 524
3. Components of Compensation Fannie Mae's executive compensation structure (for director-level employees and above) traditionally has had two components: cash and "variable long-term awards" (also known at Fannie Mae as "long-term incentives," or "LTI").(1980) The cash component consists of (1) annual salary and (2) a year-end bonus under a program known as the "Annual Incentive Plan" or "AIP." The size of the Company's AIP pool was determined by a formula based exclusively on EPS. As EPS increased from a minimum threshold to a maximum ceiling, the pool of funds available for award increased commensurately. The fund-target was the sum of all of the individual officers' and directors' target bonuses, which in turn were calculated to bring each individual's cash compensation to the fiftieth percentile of the relevant peer group and the sixty-fifth percentile of total compensation for that group. The LTI portion of the compensation package was entirely equity-based. For Senior Vice Presidents and above, the LTI was divided between stock options and what Fannie Mae referred to as "performance shares" under the PSP.(1981) The size of each year's award under the PSP was determined in equal measure by the Company's EPS results and by certain non-financial performance measures. Fannie Mae also offered other long-term compensation programs dependent on the employee's level within the Company. For example, Fannie Mae sponsored an Executive Pension Plan ("EPP") for Executive Vice Presidents and above; a Supplemental Pension Plan for Company officers who did not participate in the EPP; a general deferred compensation plan for directors and above; an Employee Stock Ownership Plan ("ESOP") for all employees except those who participated in the EPP; and an Employee Stock Purchase Plan ("ESPP") for those individuals at the Vice President level or below.(1982) - ---------------------- (1980) Minutes of the Meeting of the Compensation Committee of the Board of Directors of Fannie Mae, dated Jan. 20, 2003, FMSE 505241-71, at FMSE 505243-44. (1981) Fannie Mae Performance Share Program - Summary, in NOV. 2003 AGENDA, at FMSE-IR 17603-04 ("Performance shares are long-term incentive awards with a target payout value equal to a number of Fannie Mae common shares. The number of shares `earned' depends upon Fannie Mae achieving certain measures of long-term performance."). For these executives, the total value of the LTI was divided in half between stock options and performance shares. Performance Shares, in SEPT. 2003 EXECUTIVE COMPENSATION PROGRAMS at FMSE-IR 283205-07. Vice Presidents and Directors did not participate in the PSP. Id. (1982) Employee Stock Purchase Plan, in SEPT. 2003 EXECUTIVE COMPENSATION PROGRAMS, at FMSE-IR 283225-26. 525
The EPP, ESOP, and ESPP generally constituted a very small portion of executive compensation, and we accordingly did not focus our review on these programs.(1983) Instead, we concentrated on the funding for the AIP and PSP awards, each of which was determined by Fannie Mae's EPS results. We also considered certain non-financial performance measures used to analyze the non-EPS-based criteria for half of the PSP award.(1984) B. The Annual Incentive Plan 1. Calculation of the AIP Pool and Individual Awards According to Company policy, "[t]he purpose of the [AIP was] to encourage greater focus on performance among the [management] of Fannie Mae by relating a significant portion of [executives'] total compensation to the achievement of annual financial, strategic or operational objectives."(1985) The "annual objective" to which the size of the AIP award was correlated changed several times after the plan was implemented in 1985.(1986) Initially, AIP was tied to the Company's return on equity - ---------------------- (1983) For example, Rudin recalled the amount of the ESOP contribution rarely exceeded two percent of the aggregate compensation of all eligible employees. See also Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 20, 1998, FMSE-IR 660-75, at FMSE-IR 662 (resolving to set the ESOP pay out at two percent if financial performance is within the minimum and maximum targets set for the year's Annual Incentive Plan); Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 19, 1999, FMSE 503872-89, at FMSE 503877 (same); Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 18, 2000, FMSE 503953-67, at FMSE 503954-55 (same); Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 16, 2001, FMSE 504031-44, at FMSE 504036 (same); Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 15, 2002, FMSE 504117-32, at FMSE 504121 (same); Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 21, 2003, FMSE 504202-40, at FMSE 504210 (same); Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 23, 2004, FMSE 504353-90, at FMSE 504385 (same). The Compensation Committee could recommend to the Board different contributions if the minimum EPS was not achieved or the maximum EPS was exceeded. See, e.g., Minutes of the Meeting of the Board of Directors of Fannie Mae, dated Jan. 20, 1998, FMSE-IR 660-75, at FMSE-IR 662. (1984) We did not identify any concerns regarding the award of stock options, and no one raised any such concerns to us during our investigation. Accordingly, this Chapter does not address them. (1985) Fannie Mae Annual Incentive Plan, as amended through Jan. 16, 2001, FMSE-IR 598523-49, at FMSE-IR 598523. (1986) Annual Incentive Plan, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Jan. 2000, FMSE-IR 16333-392, at FMSE-IR 526
(rather than to EPS).(1987) Then, beginning in 1991, it was tied to net income.(1988) In 1995, the Company moved to an EPS measure "both to bring this measure into conformity with [Fannie Mae's] PSP target (which also is EPS-based) and to provide [the Company] with the flexibility to repurchase shares without penalizing AIP participants."(1989) Since that time, the Company-wide bonus pool from which individual AIP bonus awards were drawn was determined by a formula dependent upon EPS. Fannie Mae's AIP-EPS measure consisted of minimum, target, and maximum EPS figures, which, until 2002, correlated to an AIP pool of 50 percent of the target amount, 100 percent of the target amount, and 150 percent of the target amount, respectively, and, after 2002, to 75 percent, 100 percent, and 125 percent.(1990) The "target" AIP pool was to be the sum of eligible executives' target bonuses (as determined by the fiftieth percentile of their respective peer groups) and was set to correspond to a specific EPS figure. For example, in 1998, Fannie Mae planned to distribute a bonus pool of $17.3 million if the Company hit its EPS target of $3.18.(1991) If EPS hit $3.23 (the maximum), then the bonus pool would increase to $25.9 million. If EPS dropped to the threshold level of $3.13, the bonus pool was expected to be $8.6 million.(1992) Once the size of Fannie Mae's bonus pool was determined, the pool was allocated among the Company's departments; each department head then was responsible for dividing the department's portion of the pool among its eligible officers. Until 2001, executives determined their subordinates' bonus payouts based on a combination of each - ----------- 16340 (hereinafter "JAN. 2000 AGENDA") (noting that the AIP was instituted in 1985). (1987) Id. (1988) Id. (1989) Id. Beginning in 2001, the AIP (like all of the Company's internal goals) switched from being tied to GAAP EPS to the Company's "core" EPS. For discussion of this change, see supra Chapter V. (1990) 2004 Executive Compensation Program Overview, FMSE-IR 18087-99, at FMSE- IR 18091; Annual Incentive Plan, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Jan. 20, 1998, FMSE-IR 14826-74, at FMSE-IR 14836 (hereinafter "JAN. 1998 AGENDA"). (1991) JAN. 1998 AGENDA at FMSE-IR 14836. The estimated $17.3 million was the sum of all eligible officers' and directors' target bonuses. The sum was based on the current number of participants and could change based on changes in the number of participants, their employment levels, or their salaries. (1992) JAN. 1998 AGENDA at FMSE-IR 14836. 527
subordinate's personal performance and the Company's performance overall.(1993) For example, for the 1997 AIP payout, one Vice President in Financial Reporting received a bonus of 45.11% of base salary in recognition of individual performance;(1994) the Company awarded all Vice Presidents an additional bonus of 9.62% of base salary as a result of the Company's EPS results.(1995) Beginning with the 2001 AIP award, the Compensation Committee eliminated individual goals, made the entire AIP pool dependent upon achievement of the EPS targets, and delegated to management the allotment of that pool.(1996) This change was based on management's desire for increased flexibility in awarding bonuses.(1997) Management reasoned that, "[u]nder the current structure, the corporate component of the AIP becomes a floor that all officers get, despite their personal performance By eliminating the corporate/personal AIP split and making the entire bonus pool available for management's discretion, management could better recognize distinctions in both performance and job scope."(1998) As management recognized, however, "AIP pool funding" would continue to be "a function of corporate financial performance."(1999) 2. Maximizing AIP Funding Pools For each year between 1995 and 1999, the Compensation Committee concluded that the Company had exceeded its target EPS goals for determining the AIP pool: - ---------------------- (1993) See, e.g., Annual Incentive Plan, in JAN. 1998 AGENDA, at FMSE-IR 14839. (1994) Id. (1995) Id. (1996) Minutes of the Meeting of the Compensation Committee of the Board of Directors of Fannie Mae, dated Nov. 10, 2000, FMSE 505167-85, at FMSE 505176. The Committee retained authority to determine the AIP payouts for the members of the Office of the Chairman. Cf. Annual Incentive Plan, in FANNIE MAE EXECUTIVE COMPENSATION PROGRAMS, dated Feb. 2004, FMSE-EC 1-100, at FMSE-EC 3, 47- 49 (hereinafter "FEB. 2004 EXECUTIVE COMPENSATION PROGRAMS"). (1997) Proposal for 2001 Annual Incentive Plan, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Nov. 20, 2000, FMSE-IR 15877-16060, at FMSE-IR 15954 (hereinafter "NOV. 2000 AGENDA"). (1998) Id. (1999) Id. 528
Minimum Maximum (Threshold) Target (Ceiling) Actual EPS Achievement Against Year EPS EPS EPS Goals 1995 $ 2.0975 $ 2.1475 $ 2.195 $ 2.1515 8% of range between target and maximum (above target)(2000) 1996 $ 2.380 $ 2.430 $ 2.480 $ 2.4764 93% of range between target and maximum (above target)(2001) 1997 $ 2.740 $ 2.790 $ 2.840 $ 2.8325 85% of range between target and maximum (above target)(2002) 1998 $ 3.130 $ 3.180 $ 3.230 $ 3.2309 Above maximum(2003) 1999 $ 3.590 $ 3.640 $ 3.690 $ 3.7199 Above maximum(2004) The AIP funding pool was funded at greater than the target for each of these years. In 2000, Fannie Mae changed its corporate performance measure from an EPS dollar figure to a percentage representing the "increase in EPS" over the prior year.(2005) As noted below, Fannie Mae continued to exceed its maximum AIP goals following this change, resulting in maximum, or near-maximum, AIP bonus pools. - ------------------ (2000) Annual Incentive Plan, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Oct. 19, 2004, FMSE-IR 18046-102, at FMSE-IR 18090 (hereinafter "OCT. 2004 AGENDA"). (2001) Id. (2002) Id. (2003) Id. (2004) Id. (2005) This change was made to be consistent with Raines's business goal of double-digit earnings growth and doubling EPS over five years. See Annual Incentive Plan, in JAN. 2000 AGENDA, at FMSE-IR 16341; Performance Share Program, in JAN. 2000 AGENDA, at FMSE-IR 16366. 529
Achievement Minimum Target Maximum Actua_EPS Against Year EPS_Growth EPS_Growth EPS_Growth Growth Goals 2000 12.3% 13.6% 14.9% 15.3% Above maximum(2006) 2001 12.3% 13.6% 14.9% 21.2% Above maximum(2007) 2002 15.0% 18.0% 21.0% 21.3% Above maximum(2008) 2003 10.0% 12-13% 16.0% 15.5% 86% of range between target and maximum (above target)(2009) 2004 8.0% 10.0% 12.0% - - (2010) Fannie Mae's record of consistently hitting the maximum bonus level was noted by Alan Johnson when Fannie Mae's management first engaged him to consult on compensation issues.(2011) Johnson believed that the Company's prior compensation consultant erroneously had conflated various distinct executive positions when assessing market compensation. To provide the Company with what he believed was more accurate market data, Johnson redesigned Fannie Mae's compensation system into forty-one "levels" and provided market compensation data for each level.(2012) Each employee - ---------------- (2006) Annual Incentive Plan, in NOV. 2003 AGENDA at FMSE-IR 17596. (2007) Id. (2008) Id. (2009) Annual Incentive Plan, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Jan. 22, 2004, FMSE-IR 18215-343, at FMSE-IR 18231-34 (hereinafter "JAN. 22, 2004 AGENDA"). (2010) These results have not been determined because the Company has not issued financial statements for the 2004 fiscal year. (2011) This issue apparently predated Alan Johnson's engagement. In 1997, the Company found that compensation for members of the Office of the Chairman and Executive Vice Presidents had lagged behind comparator group levels. 1997 Annual Incentive Plan, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Nov. 17, 1997, FMSE-IR 14905-15063, at FMSE-IR 15034 (hereinafter "NOV. 1997 AGENDA"). The Company proposed to address this by expanding bonus ranges for these officers. Id. In 1998, the issue arose again. See Overview of Executive Compensation Recommendations, in NOV. 1998 AGENDA at FMSE-IR 14339. That same year, the Company's then compensation consultant found that target total compensation for SVPs was too low. To close the gap, the Company recommended awarding additional stock grants to the SVPs, rather than recommending long-term or structural revisions. Id.; see also Johnson Associates, Inc., Officer Compensation Review Initial Discussion Outline, dated July 30, 2002, FMSE-EC 7875-912, at FMSE-EC 7901 (demonstrating base salary below market median). (2012) Under Johnson's system, level forty-one corresponded to the CEO, Executive Vice Presidents were assigned to levels in the mid-to-high thirties, Senior Vice Presidents 530
was assigned a level and paid in accordance with that level. Fannie Mae also assigned each officer level a bonus target, which was calculated as a percentage of that level's target salary. As discussed below, however, this leveling process did not succeed in aligning Fannie Mae's compensation targets with the market. It appears that management was aware that target compensation lagged behind the market and used achievable EPS targets as a way of responding to this. In a separate report issued by Johnson as part of his initial analysis, Johnson cited several management interviewees who noted that the Company's executive compensation reached competitive levels only when the Company attained its maximum EPS targets, and thus its maximum bonus payouts.(2013) An e-mail chain among individuals in Fannie Mae management from early January 2002 confirms that management was aware that AIP-EPS targets were set at achievable levels so as to increase the likelihood of maximizing the AIP payout.(2014) According to the e-mail, Franklin D. Raines wanted AIP-EPS bonus targets to be set such that an individual would receive cash compensation (salary and bonus) at the fiftieth percentile of his/her peers in the comparator group if the Company reached its expected EPS figure.(2015) However, the Compensation Committee materials already prepared for the January 2002 Committee meeting had set the expected EPS figure for 2002 as theUnder Johnson's system, level forty-one corresponded to the CEO, Executive Vice Presidents were assigned to levels in the mid-to-high thirties, Senior Vice Presidents to levels twenty-three and higher, Vice Presidents to levels seventeen and higher, etc. maximum AIP-EPS achievement, rather than the target.(2016) Thus, if maximum EPS were achieved or exceeded, then the bonus only would pay employees at the fiftieth percentile - ---------- to levels twenty -three and higher , Vice Presidents to levels seventeen and higher, etc. (2013) See Update on Executive Compensation Study, dated July 1, 2002, FMSE-EC 14739-42, at FMSE-EC 14741 ("Incentive plans have `always' hit max. Acknowledgement by some that easier goals are designed to generate needed compensation."); Johnson Associates, Inc., Compensation Committee Update, dated Nov. 19, 2002, FMSE-EC 3076-100, at FMSE-EC 3080 ("Maximum bonus pay-out needs to reach competitive compensation levels."); Johnson Associates, Inc., Compensation Committee Update, dated Nov. 19, 2002, FMSE-EC 3022-51, at FMSE-EC 3031 ("Actual bonuses competitive; target bonus low vs. market."); see also Johnson Associates, Inc., draft Detailed Compensation Review Information, dated Aug. 9, 2002, FMSE-EC 7952-80, at FMSE-EC 7964 ("No more `lay-up' goals to get to 150% payout."). (2014) E-mail from Beverly Fitzgerald to L. Rudin, dated Jan. 11, 2002, FMSE-IR 633763-64. (2015) Id. at FMSE-IR 633763. (2016) See Annual Incentive Plan - 2002 Plan Year, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Jan. 15, 2002, FMSE-IR 17493-534, at FMSE-IR 17502 (hereinafter "JAN. 2002 AGENDA"). 531
of cash compensation. If less than the maximum EPS were achieved, then executives' cash compensation would be below the fiftieth percentile, contrary to Fannie Mae's compensation philosophy (and Raines's instruction). Given the Committee's schedule, however, the 2002 materials were not changed to align the target EPS with the Company's expected results.(2017) According to Johnson, and as reflected in the e-mail chain discussed above, the Company attempted to change this pattern of aligning competitive compensation with maximum EPS targets by adjusting the target EPS to coincide with the fiftieth percentile of cash compensation for each level among comparator corporations. Each year, however, the success of the mortgage and financial services markets led to increases in comparators' compensation (salaries, bonuses, and long-term incentives). Fannie Mae did not forecast for the increases in its January projections. Thus, although the Company tried to respond to this pattern by increasing compensation, it nonetheless continued to need to award maximum AIP payouts through achievable EPS goals to keep its cash compensation in the competitive fiftieth-percentile range. In January 2004, management recommended that the Compensation Committee increase compensation and change the compensation structure to keep pace with the market.(2018) The Compensation Committee declined to change the structure at that time, choosing instead to raise all compensation by four percent.(2019) We conclude that Fannie Mae knowingly set maximum EPS goals at levels management believed it would achieve, but that this situation originated and persisted because Fannie Mae used flawed data to set compensation for its officers. Specifically, as discussed above, the market data failed to reflect the full effect of increases in comparator group compensation between the setting of compensation targets in January of each year and the determination of the size of the bonus pool at the end of each year. As a consequence, Fannie Mae's target cash pay (as a combination of salary and bonus) consistently was below the fiftieth percentile of cash compensation and the - ---------- (2017) E-mail from B. Fitzgerald to L. Rudin, dated Jan. 11, 2002, FMSE-IR 633763-64, at FMSE-IR 633763. (2018) Minutes of the Compensation Committee of the Board of Directors of Fannie Mae, dated Jan. 22, 2004, FMSE-IR 276449-72, at FMSE-IR 276450 ("The Committee decided not to make changes to the award structure going forward at this time but to review it and the market at mid-year."). (2019) Id. at FMSE-IR 276450-51; cf. 2004 Accomplishments - Lorrie Rudin, FMSE-E_EC 145400-02, at FMSE-E_EC 145400 (referring to executive compensation increase of three percent). According to Johnson, Fannie Mae eventually will experience officer retention problems if it does not keep pace with its peers in the realm of compensation. 532
sixty-fifth percentile of total compensation within the comparator group.(2020) Each year, when the Company recognized flaws in that year's market data, it attempted to adjust for these inaccuracies by setting the following year's maximum EPS goals at achievable levels.(2021) When Fannie Mae's EPS for the year reached the maximum point, executives were paid larger bonuses, such that their overall compensation was in line with comparators' compensation as specified in the Company's compensation philosophy. 3. The Role of the EPS Incentive Under Raines's leadership, the Company developed EPS-driven business goals. Given the link between EPS and the AIP bonus pool, achievement of these business goals was inextricably linked to achieving maximum bonus pools. Moreover, as discussed above, only by reaching the maximum bonus pool could the Company reach the competitive executive compensation levels consistent with its philosophy. However, our review reveals that Fannie Mae determined its business goals in advance of the compensation goals, and the latter then were set to correlate with the former. In May 1999, Fannie Mae announced its goal to double EPS in five years, from $3.23 in 1998 to $6.46 by the end of 2003.(2022) Beginning with the 2000 award year, the Company sought to align the AIP targets with this long-term objective.(2023) The same year, the Company changed the EPS measure from a strict numerical measure to a measure of percentage increase over the previous year,(2024) and set its maximum EPS target at the exact growth level (14.9%) that, if sustained, would result in doubling EPS in five years.(2025) The Company also established an EPS Challenge Option Grant, which - ---------- (2020) See Minutes of the Compensation Committee of the Board of Directors of Fannie Mae, dated Jan. 22, 2004, FMSE-IR 276449-72, at FMSE-IR 276456. (2021) Id. (2022) See Special Stock Option Grant Tied to Earnings Per Share Goals, FMSE-IR 15458-63, at FMSE-IR 15459; see also Remarks by Timothy Howard, Nationsbanc Montgomery Securities Growth Stock Conference, May 12, 1999, FMSE-IR 34614-23, at FMSE-IR 34623; Remarks Delivered by Frank Raines, Merrill Lynch Investor Conference, Sept. 14, 1999, FMSE-IR 34584-602, at FMSE-IR 34584. (2023) Annual Incentive Plan, in JAN. 2000 AGENDA at FMSE-IR 16341. (2024) Minutes of the Meeting of the Compensation Committee of the Board of Directors of Fannie Mae, Jan. 18, 2000, FMSE 505139-48, at FMSE 505140 ("Mr. Raines noted that the goal is in line with the Company's overall performance goals.").Rudin recalled that Howard played a role in the process of seeking approval from the Compensation Committee for the change from a strict numerical measure to a measure of annual percentage growth. 533
vested stock options in all employees upon the Company's reaching the $6.46 EPS level in or before 2003.(2026) This structure did not necessarily create an incentive to maximize the AIP bonus pool each year. First, the size of the bonus pool was on a sliding scale; thus, the actual compensation at stake in hitting maximum EPS goals, versus slightly missing them, could be small.(2027) For each percentage point increase or decrease between the minimum and the maximum, the pool was adjusted accordingly. Thus, if EPS results were halfway between the EPS target and the maximum, the bonus pool would be funded halfway between the payout target and the maximum (i.e., 112.5 percent). Second, as noted above, any individual's bonus was subject in whole (or, prior to 2001, in part) to management discretion.(2028) The existence of a larger bonus pool did not guarantee any individual officer a larger bonus. Finally, as noted by Johnson, Fannie Mae's range of potential bonus payouts (i.e., seventy-five percent to 125 percent of target bonuses) was narrow, and Semler Brossy found that range was narrower than the equivalent ranges at Fannie Mae's peer financial institutions.(2029) Johnson had expressed concerns to the Company that its bonus range was too narrow to provide effective financial incentives for executives to exceed their goals. This same concern regarding overly narrow bonus ranges is echoed throughout many reports and internal documents in which Fannie Mae - ---------- (2025) Annual Incentive Plan, in JAN. 2000 AGENDA at FMSE-IR 16340-41; E-mail from T. Howard to L. Rudin and Thomas R. Nides, dated Jan. 7, 2000, FMSE 560402 (noting 14.9% EPS growth target resulting from "Frank's setting of public goals"); Undated, untitled document labeled "Source: LS FNM Computer," FMSE-IR 379280-83, at FMSE-IR 379281 ("In the AIP program, the rewards tie to the earnings goals set by the company to be on the path to doubling earnings per share by the end of 2003 or 14.9% as max."). The PSP goal also was changed to be "fully consistent with our pledge to maintain double digit earnings growth." Recommendation - PSP16 (2000-2002 Cycle) Performance Criteria, in JAN. 2000 AGENDA at FMSE-IR 16366. (2026) See EPS Challenge Option Grant, in FEB. 2004 EXECUTIVE COMPENSATION PROGRAMS FMSE-EC at 81. The reward for the EPS Challenge Grant was stock options, but, because Fannie Mae's stock price has fallen, these options are currently underwater. (2027) Proposal for 2001 Annual Incentive Plan, in NOV. 2000 AGENDA at FMSE-IR 15955. (2028) The Compensation Committee, rather than management, determined bonus payouts for members of the Office of the Chairman. See FEB. 2004 EXECUTIVE COMPENSATION PROGRAMS at FMSE-EC 3, 47-49. (2029) See Semler Brossy Report at FMSE-KD 25905 ("[T]he payout range associated with these performance levels is narrower than prevailing practice."). 534
executives note the lack of correlation between personal performance and compensation at the Company.(2030) In sum, during the period from 1998 through 2003, the Company's bonus structure placed significant importance on achieving, or coming close to, the EPS that would result in the maximum AIP pool. Fannie Mae management well understood this situation, which was driven by the need to keep executive compensation competitive. 4. Reducing the Influence of EPS on Compensation Toward the end of the period under review, the Company became aware that using EPS as a compensation trigger was being questioned in the marketplace. In or about mid-2003, management, as a result of increasing public commentary on the issue, as well as the implementation of new corporate governance guidelines under the Sarbanes-Oxley Act, asked Johnson to review the Company's use of EPS as the sole performance measure for the AIP. Johnson conducted a market study on this question and issued a report to the Company in October 2003.(2031) Johnson concluded that many financial companies used EPS as a component of their financial performance measures, but few used EPS as their sole performance measure for determining compensation elements.(2032) Johnson also studied alternative performance measures and concluded that, although EPS is an imperfect measure of financial performance, all other financial performance measures also were imprecise.(2033) In addition, he emphasized that, regardless of the performance measure used, the Compensation Committee had complete discretion to adjust awards above or - ---------- (2030) E.g., Johnson Associates, Inc., Detailed Compensation Review Information, draft dated Aug. 9, 2002, FMSE-EC 7952-80, at FMSE-EC 7960 ("Our highest priority should be to motivate contributors and help poor contributors exit - Need process to get there."), FMSE-EC 7962 ("Broader range doesn't exist within levels to reward super performance and penalize low-end: . . . super performer not paid much more than average."), FMSE-EC 7963 ("[L]ittle meaningful distinctions on performance. Few have compensation reduced or receive large increases."); Compensation Committee Discussion, dated Oct. 15, 2002, FMSE-IR 17377-400, at FMSE-IR 17389 ("Target bonus percentages [a]rtificially low vs. market . . . Odd dynamic where `overachieve' needed to be competitive"); Semler Brossy Report at FMSE-KD 25905 ("[T]he payout range associated with these performance levels is narrower than prevailing practice."). (2031) Johnson Associates, Inc., Fannie Mae Performance Metrics study, dated Oct. 9, 2003, FMSE-IR 293968-77. (2032) See id. at FMSE-IR 293973-74. (2033) Id. at FMSE-IR 293977 ("Study of comparator performance measures did not clearly demonstrate other metrics better than those currently available."). 535
below the amount indicated by the Company's numerical results. In theory, this discretion reduced incentives for manipulation because management simply could ask the Compensation Committee for a greater award, or the Compensation Committee could authorize a larger bonus pool on its own. In January 2004, following OFHEO's release of its report on Freddie Mac's accounting and compensation, the Compensation Committee (specifically, then Compensation Committee Chair Anne M. Mulcahy) asked whether Fannie Mae should consider moving away from EPS as the sole determinant of the size of the AIP pool.(2034) The Company provided Mulcahy with a copy of Johnson's earlier EPS study.(2035) Despite Johnson's conclusion that using EPS was appropriate, Mulcahy, after consulting with Brossy, tasked Company management with developing new performance measures(2036) that potentially would include not only an earnings measure (EPS) but also a risk management measure and a mission measure.(2037) The Company did not seek to eliminate the use of EPS all together. As Brossy has noted, "as a for-profit company by definition, Fannie Mae has an obligation to drive shareholder value, and the Company and its Board cannot ignore the importance of earnings performance over time."(2038) In July 2004, after several months of work and consultations with other compensation experts, management and Johnson presented recommendations to the Compensation Committee for changing the AIP measures for the 2005 AIP year to a combination of EPS (weighted fifty percent), a risk management measure (twenty-five percent), and a mission measure (twenty-five percent).(2039) The Committee approved this - ---------- (2034) See Minutes of the Meeting of the Compensation Committee of the Board of Directors of Fannie Mae, dated Jan. 22, 2004, FMSE-IR 276449-72, at FMSE-IR 276458. (2035) See id.; Mem. from Kathy G. Gallo to A. Mulcahy, dated Jan. 13, 2004, FMSE-EC 28982. (2036) See Minutes of the Meeting of the Compensation Committee of the Board of Directors of Fannie Mae, dated Jan. 22, 2004, FMSE-IR 276449-72, at FMSE-IR 276458. (2037) Performance Measures for the AIP, FMSE-EC 29090-98, at FMSE-EC 29090. (2038) Semler Brossy Report at FMSE-KD 25885. (2039) Compensation Committee Update: Review of Current Practices, dated July 2004, FMSE-IR 18154-77, at FMSE-IR 18168. Determination of whether the risk management and mission goals had been met would be undertaken by the Compensation Committee with the assistance of two other Board of Directors' committees: the Assets and Liabilities Policy Committee and the Housing & Community Development Committee, respectively. Id. 536
change; however, because the Company has been unable to issue financial statements, this change has not been implemented.( 2040) The Company's decision to change from a sole-EPS measure to a weighted combination of measures does not reflect a conclusion by Fannie Mae or its regularly engaged consultants that EPS is an inappropriate measure for determining compensation. To the contrary, Brossy and Johnson also agree that the use of EPS as a compensation performance measure is neither inherently problematic nor more problematic than other numerical indicators (even if, as discussed previously, EPS should not be the sole measure).(2041) Brossy and Johnson also agree, based on their experience and references to other institutions, that it is logical and sensible for financial services companies to employ numerical financial targets to trigger compensation elements.(2042) That Fannie Mae used and continues to use EPS as part of a measure to determine whether business financial goals are met for purposes of triggering compensation is not improper. Implementation of Semler Brossy's recommendations to include an EPS target as one of several factors considered in bonus situations will be an improvement over the Company's past practice. C. Performance Share Plan The other executive compensation program that depended on EPS as a compensation trigger was the PSP. Fannie Mae granted performance shares only to officers at the SVP level and above.(2043) The reason for this limitation was Fannie Mae's belief that "SVPs and above are considered to have the greatest direct impact on the long-term objectives of the program."(2044) The PSP ran on a staggered three-year cycle; that is, whether goals were met and the award paid depended on a review of the Company's performance over the - ---------- (2040) Annual Incentive Plan, FMSE-IR 18090-91, at FMSE-IR 18091. Semler Brossy's recent recommendations for changing the AIP employ similar measures. Semler Brossy Report at 111, FMSE-KD 25930. (2041) Semler Brossy Report at 66, FMSE-KD 25885; see also Johnson Associates, Inc., Fannie Mae Performance Metrics study, dated Oct. 9, 2003, FMSE-IR 293968-77 (analyzing various financial performance measures). (2042) See Semler Brossy Report at 66, FMSE-KD 25885; Johnson Associates, Inc., Fannie Mae Performance Metrics study, dated Oct. 9, 2003, FMSE-IR 293968-77. (2043) The PSP did not formally limit grants to this set of officers, but it was Fannie Mae's practice not to award performance shares to officers below the SVP level. Performance Shares, in SEPT. 2003 EXECUTIVE COMPENSATION PROGRAMS at FMSE-IR 283205. (2044) Id. 537
previous three years. For example, the first three-year PSP cycle ran from 1988-1990; the second PSP cycle ran from 1989-1991; the third PSP cycle ran from 1990-1992; and so forth.(2045) At the beginning of each PSP cycle, the Board of Directors would establish the term of the cycle and set the goals for senior management to achieve.(2046) At the end of each cycle, the Compensation Committee determined whether the goals set by the Board had been achieved. Unlike the AIP, the PSP was paid in stock: fifty percent of the earned shares were granted in the year of the award, and fifty percent were granted the following year.(2047) This staggered award was intended to support the purpose of the PSP, which was to "motivate[] executives to focus on the areas we consider critical to our long-term success from both a quantitative and a qualitative perspective."(2048) Since the PSP's inception, the PSP criteria have been (1) achievement of a quantitative measure and (2) achievement of a qualitative measure. Each component was weighted fifty percent in the calculation of shares awarded. 1. Quantitative (EPS) Component For the relevant time period, the quantitative aspect of the PSP was an EPS measure. Prior to 2002, this measure was the average annual percentage increase in EPS over the three years of the plan. Beginning with the 2002-2004 PSP cycle, the goal was changed to annualized compound EPS growth over the three-year period.(2049) As with AIP, the Compensation Committee recommended and the Board set three EPS goals for the PSP: a minimum (or "threshold") level that paid forty percent of the target pool of shares; a target level, which paid 100 percent of the pool; and a - ---------- (2045) Performance Share Program, in JAN. 2000 AGENDA at FMSE-IR 16362. Fannie Mae initiated the PSP in 1982; between 1982 and 1987, each PSP cycle was four years in duration. Id. (2046) Performance Shares, in SEPT. 2003 EXECUTIVE COMPENSATION PROGRAMS at FMSE-IR 283207; Semler Brossy Report at FMSE-KD 25857. (2047) Performance Shares, in SEPT. 2003 EXECUTIVE COMPENSATION PROGRAMS at FMSE-IR 283206. For PSP 15 (1999-2001) and earlier, fifty percent of the total award was paid at the end of the three-year cycle, and the balance paid in equal installments at the beginning of the following two years (i.e., twenty-five percent and twenty-five percent). Id. (2048) Fannie Mae Performance Share Program - Summary, in NOV. 2003 AGENDA at FMSE-IR 17603. (2049) January 15, 2002 Compensation Committee Meeting, FMSE-IR 17487-89, at FMSE-IR 17488. No one we interviewed recalled why this was changed. 538
maximum (or "high") level, which paid 150 percent of the pool. As an example, for the 2000-2002 PSP cycle (PSP 16), the quantitative goals were expressed as the average annual growth in EPS:(2050) Achievement Level EPS% Change PSP Vesting - ----------------- ----------- ----------- High (Maximum) 14.9% 150% Target 13.6% 100% Low (Threshold) 10.0% 40% Also as with AIP, Fannie Mae correlated its maximum PSP targets to the Company's goal of doubling EPS within five years.(2051) The effect of reaching a particular EPS target on the PSP award, however, was tempered by two facts. First, because PSP payouts were determined by three years of financial results, a single year's EPS results would have a relatively smaller impact on the eventual PSP payout than it would have on the AIP pool. Second, the fifty percent weighting of qualitative performance measures in the PSP calculation further diminished the PSP impact of reaching annual EPS targets. Accordingly, EPS results played a significant role in PSP awards as the sole financial measure therein, but the impact of EPS on the PSP was significantly less prominent than its impact on the AIP. 2. Qualitative (Subjective) Component The qualitative or subjective component of the PSP depended on Fannie Mae achieving certain business performance goals.(2052) At the beginning of each PSP three-year cycle, the Board of Directors reviewed the "Corporate Goals and Objectives" prepared by management.(2053) In the Corporate Goals and Objectives, Company management set forth broad corporate goals and tailored business objectives for the forthcoming year. For example, the 2004 Corporate Goals and Objectives set forth six missions: - ---------- (2050) Performance Share Program, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Jan. 16, 2001, FMSE-IR 16699-748, at FMSE-IR 16743 (hereinafter "JAN. 2001 AGENDA"). (2051) See supra note 2025 and accompanying text (discussing Company's target of annual 14.9% increase in EPS). (2052) Johnson considers the corporate performance measure for the PSP to have been a leading practice at the time it was implemented and to be a best practice today. (2053) See, e.g., 2004 Corporate Goals and Objectives, FMSE-E 2161591-620. 539
LONG-TERM CORPORATE GOALS Goal 1: Acknowledged Leadership in Increasing Access to Affordable Housing Goal 2: Leading Presence in the Secondary Market and Partner of Choice Goal 3: Optimal Risk Management Goal 4: Record Financial Performance Goal 5: Corporate Culture to Enhance Strategy Execution Goal 6: E-Commerce Company Infrastructure to Increase Capabilities and Lower Costs(2054) Within each goal was a list of objectives. For example, "Goal 1: Acknowledged Leadership in . . . Affordable Housing," had eight separate objectives: (1) American Dream Commitment; (2) HUD goals; (3) minority lending; (4) first time home buyers; (5) community development; (6) affordable rental housing; (7) addressing homelessness; and (8) outreach.(2055) These Corporate Goals and Objectives usually then were translated by the Board of Directors into Corporate Performance PSP measures.(2056) These measures were detailed but subject to the Compensation Committee's qualitative review.(2057) Twice a year (mid-year and after year-end), management would provide the Board with "Corporate Performance Assessments" (often referred to as "CPAs" or "Report Cards"). For the seven PSP cycles ending between 1998 and 2004, Fannie Mae produced only three three-year assessments.(2058) In other years, only a single year's - ---------- (2054) See id. at FMSE-E 2161592. (2055) Id. at FMSE-E 2161593-97. (2056) See, e.g., Fannie Mae Performance Share Program - Summary, in OCT. 2004 AGENDA at FMSE-IR 18098. (2057) In each of the PSP cycles during the period under review, the Committee determined that management had merited the maximum, or 150%, award for the non-numerical portion of the PSP. Feb. 2004 Executive Compensation Programs at FMSE-EC 76. The only exception was the 2000-2002 cycle, for which the Committee awarded 140%. Id. (2058) Breaking Records!: Corporate Performance Assessment for 1996-1998, FMSE 1162-95; Corporate Performance Assessment 1999-2001: Our Best Year Yet!, FMSE-EC 20945-64; Corporate Performance Assessment 2000-2002, FMSE 10542-60. 540
performance was assessed.(2059) Even the three-year reports focused almost entirely on the last year of the PSP cycle.(2060) Roger Brossy commented that a single-year review (rather than a three-year review) did not ensure a realistic analysis of the performance over the full three-year period, but instead only gave the most recent information to be assessed. Brossy questioned whether this approach ensured compliance with the multi-year plan. The CPAs often were exceedingly positive, primarily reporting on the Company's achievements. For example, the 1996-1998 Corporate Performance Assessment given to the Board had a baseball theme and was entitled "Breaking Records!"(2061) The assessment began with a heading, "Record Shattering Results, Outstanding Performance," and had subheadings such as "Record Volumes," "Record Total Business Outstanding," "Record Earnings and Stock Performance," "Record Credit Performance," and "Record Accomplishments in the Information Technology Arena."(2062) Under each goal, management extolled the Company's virtues and achievements. The report was written in a highly positive manner, using phrases such as "we significantly enhanced," "[o]ur efforts have contributed significantly," and "[w]e are determined to do more and build upon our record."(2063) Our review of the entire report revealed few, if any self-critical statements. Indeed, the report concluded with a "highlights" section that repeated Fannie Mae's "records" for the three-year period.(2064) Similarly, the 2003 CPA began: "For our mission, business and disciplined growth strategy, 2003 was one of Fannie Mae's most successful years ever."(2065) The executive summary then proceeded to tell the Board of Directors of highlights of Fannie Mae's "Mission," Business," "Disciplined Growth Strategy," and - ------------------------- (2059) See e.g., Year-End Corporate Performance Assessment 1999: A Record Year Ends a Record Decade, FMSE 3313-32; Corporate Performance Assessment 2003, FMSE-EC 6040-57. (2060) See Corporate Performance Assessment 1999-2001: Our Best Year Yet!, FMSE-EC 20945-64, at FMSE-EC 20947 ("2001 truly was a watershed year for our company . . . ."); Corporate Performance Assessment 2000-2002, FMSE 10542-60, at FMSE 10543 ("2002 was an extraordinary year for our service to families and communities, and for our business and financial results overall."). (2061) Breaking Records!: Corporate Performance Assessment for 1996-1998, FMSE 1162-95, at FMSE 1162. (2062) Id. at FMSE 1163-64. (2063) Id. at FMSE 1167-68. (2064) Id. at FMSE 1181. (2065) Corporate Performance Assessment 2003, FMSE-EC 6040-57 at FMSE-EC 6041. 541
Policy Challenges."(2066) With respect to its analysis of Goal 1: "Increasing access to Affordable Housing," the CPA began: "In the face of extraordinary market conditions and significant legislative and regulatory issues, Fannie Mae once again met our key mission goals and helped increase homeownership opportunities for millions of American families."(2067) Fannie Mae also listed what it perceived as future "challenges." Most significantly, Fannie Mae identified its biggest challenge as what to do in future years, "[h]aving reached [certain of its] most prominent goals" already.(2068) Fannie Mae repeated this formula in its discussion of each of the remaining areas. Although Fannie Mae did recognize that it faced a single policy challenge in 2003, that challenge was laid squarely at the feet of Freddie Mac and the revelations of its management and accounting problems.(2069) Otherwise, the only self-critical statement in the presentation was that the Company's administrative expenses ran "slightly higher than expected . . . due primarily to higher than planned development and testing costs for the Core Infrastructure technology project."(2070) In tone and presentation, this CPA (like others) reads more like a marketing piece than an internal document intended to provide members of the Board of Directors with a realistic insight into Fannie Mae's performance in the previous calendar year.(2071) When Brossy saw a CPA for the first time, he expressed to Mulcahy his concern about the positive cast of the document.(2072) Brossy recalls that Mulcahy agreed. However, she indicated to us that she believed it would be very difficult to have a meaningful, reflective discourse on corporate performance with management.(2073) Both - ---------------- (2066) Id. at FMSE-EC 6041-42. (2067) Id. at FMSE-EC 6044. (2068) Id. at FMSE-EC 6045. (2069) Id. at FMSE-EC 6051. (2070) Id. at FMSE-EC 6052. (2071) The "marketing" feel of the document may have resulted in part from the Communications team's editorial review, which we understand was a step in the drafting process. See E-mail from Robert J. Levin to Michael Quinn, dated Dec. 6, 2004, at FMSE-E_EC 54175-76. (2072) Brossy was provided a copy only shortly before the meeting at which it was discussed; he was not given an opportunity to comment on it. In addition to Brossy, Thomas A. Lawler and Mudd individually expressed the view that the CPAs were overly positive. E-mail from J. Morgan Whitacre to T. Lawler, dated Dec. 26, 2002, FMSE-E_EC 36647-48 (noting Lawler's response to draft CPA). (2073) This tone may have been intentional: In at least one instance in connection with PSP performance, management sought to limit the information provided to the 542
Brossy and Mulcahy suggested that the positive light of the CPAs was immaterial because there were "not a lot of weaknesses" at the Company at that time. So, although the CPAs emphasized the positive rather than providing a more balanced (or even negative) view of the Company, the perception of the Compensation Committee and others, even without the CPAs, was that the Company was doing well. Although there is no doubt that Fannie Mae appeared to be performing well, the tone of the document raised questions regarding its reliability. A more balanced approach, which included real business challenges and how Fannie Mae overcame them, as well as issues that the Company confronted or would confront in the future, would have lent a greater tone of seriousness to this process for determining a potentially significant portion of a senior executive's compensation. One other criticism surrounding the CPAs was that they, in fact, offered too much information. Johnson recalled that, in presenting the CPAs to the Compensation Committee, the presenter overwhelmed the members of the Compensation Committee with unnecessary details. Johnson suggested that the Compensation Committee thus became mired in minutiae. Mulcahy also indicated that she believed that the Compensation Committee was given too much information generally. Similarly, Rudin recounted that she had heard that sentiment expressed by Compensation Committee members prior to 2000, although she did not recall by whom. In concept, Johnson found Fannie Mae's PSP "to be well-designed and considered it a `best practice' among financial service companies,"(2074) and we have not concluded otherwise except to note these areas of potential weakness. To establish more rigor in the goal setting and achievement analyses, particularly as presented through the CPA, Semler Brossy has recommended that Fannie Mae retain the internally set corporate performance goals but put pressure on the Board to ensure that the goals in fact are met before awards are made.(2075) To do so, Semler Brossy recommends the use of an - ----------- Compensation Committee. For the 2003 PSP17 payout, when asked by a co-worker if management needed to give the Compensation Committee the details of how the payout "would look" with a "CPA grade of less than 100%," Rudin responded that she did "not think it would be `positively viewed' if I added an alternative below 100%, as in `why would you even want to put that possibility into the committee's head.'" E-mail from K. Gallo to L. Rudin, dated Jan. 14, 2004, FMSE-E_EC 152286-87. When asked who would not "positively view[]" the edit, Rudin identified "senior management," including Howard, Raines, and Leanne Spencer. Rudin indicated that she normally gave the Compensation Committee choices only above 100 percent. (2074) Performance Share Program, in AGENDA - COMPENSATION COMMITTEE OF THE BOARD OF DIRECTORS - FANNIE MAE, dated Jan. 20, 2003, FMSE-IR 17752-89, at FMSE-IR 17781. (2075) Semler Brossy Report, FMSE-KD 25886-88. 543
outside firm to make objective assessments of the corporate goals.(2076) These steps will improve the effectiveness of the qualitative assessment and ensure that the Compensation Committee is receiving relevant and understandable information from management. IV. LEGAL DEPARTMENT INVOLVEMENT IN COMPENSATION DECISIONS In anonymous letters received by OFHEO during the course of Paul, Weiss's investigation, individual(s) accused Fannie Mae's Legal Department of improperly reviewing all compensation decisions for purposes of cloaking the decisions in the attorney-client privilege. We found no evidence to support these allegations. To investigate this concern, we requested and reviewed thousands of documents from Fannie Mae's Legal Department related to executive compensation, including documents involving (1) negotiation of employment contracts for Raines, Daniel H. Mudd, and Howard;(2077) (2) general ERISA/benefits issues; (3) disparate impact analyses undertaken to insure against invidious discrimination; and (4) the Company's position with respect to OFHEO proposed regulations on compensation. We also interviewed Donald M. Remy, Judith Dunn, and Anthony Marra from the Legal Department and asked others involved in compensation issues about the role of the Legal Department. We found nothing to support an allegation that the Legal Department was making inappropriate changes to compensation decisions or otherwise reviewing materials for the purpose of cloaking compensation decisions with the attorney-client privilege. The Legal Department's primary involvement in compensation decisions involved a review for any indication of unlawful discrimination. As noted above, certain bonus payments under the AIP program were determined by department heads, who allocated amounts to their subordinates from a pool. The Legal Department reviewed the allocations for indications of discrimination or other impropriety. We found only one instance in which the Legal Department was involved in issues relating to compensation at Fannie Mae in other than what might be considered the ordinary course. In March 2004, Dunn and others in the Legal Department conferred with Professor David Yermack, a professor at New York University, regarding the use of EPS as a measure of financial performance for compensation purposes. According to e-mail correspondence between attorneys in the department, the Legal Department contacted Yermack for purposes of assessing Fannie Mae's position on proposed OFHEO regulations that would have restricted or eliminated the use of EPS in calculating - ---------- (2076) Id. at FMSE-KD 25888. (2077) Only Raines, Howard, and Mudd had negotiated employment contracts with the Company. 544
incentive compensation.(2078) This also followed OFHEO's report regarding Freddie Mac, and our interviews suggested that OFHEO's criticism of Freddie Mac's compensation practices at least partially motivated Fannie Mae's 2004 inquiries. Although Fannie Mae clearly was seeking support for its position that EPS was a suitable measure, Yermack was critical of using EPS as a target for compensation purposes. Yermack warned the Legal Department that "it is relatively easy to smooth earnings in a way most convenient to managers."(2079) Yermack also noted, as Brossy and Johnson had, that exceeding goals year after year suggested that targets were set too low, that managers knew the targets, and that managers performed accordingly. Yermack noted, in particular, that the PSP goals were "much too achievable."(2080) Although Fannie Mae did not, in the end, oppose OFHEO's proposal that compensation should not focus solely on earnings performance, except to note that the language proposed by OFHEO was not consistent with Fannie Mae's Charter, resulting in some ambiguity as to the regulation's effect.(2081) In conclusion, we found no evidence to corroborate the anonymous allegations that the Legal Department was wrongly involved in executive compensation decisions. The most likely genesis of the concern was the Department's proper review of compensation practices and decisions to guard against discrimination. We saw no evidence that the Legal Department's involvement resulted in the improper assertion, or contemplated assertion, of privilege or confidentiality for decisions regarding compensation matters. - ---------- (2078) E-mail from Saadia Mahmud to R. Senhauser, et al., dated Feb. 6, 2004, FMSE-IR 293943-44. (2079) E-mail from J. Dunn to T. Donilon, et al., dated Mar. 5, 2004, FMSE-E_EC 162705-06. (2080) Id. (2081) Letter from Ann M. Kappler to Alfred M. Pollard, dated June 14, 2004, at 11, available at http://www.ofheo.gov/media/PDF/CorpGovAmendFNM.pdf (responding to Proposed Corporate Governance Regulation, RIN 2550-AA24). 545
CHAPTER IX: FANNIE MAE'S INVESTIGATION OF ROGER BARNES'S ALLEGATIONS I. INTRODUCTION In August 2003, Roger Barnes, then a Manager in the Controller's Office, raised allegations of accounting impropriety at Fannie Mae, including potential noncompliance with FAS 91. Barnes also alleged that Controller's Office management was not receptive to employee concerns regarding Fannie Mae's accounting, and, shortly thereafter, he alleged that he had been discriminated against on the basis of race and gender. In the subsequent three months: (1) The Company's Office of Audit ("Internal Audit") and its Legal Department conducted three investigations into Barnes's allegations; (2) Barnes threatened to bring a lawsuit against Fannie Mae; and (3) Barnes and Fannie Mae executed a settlement agreement. In the agreement, Barnes relinquished all legal claims against Fannie Mae in exchange for monetary consideration. The agreement also required Barnes to cooperate with investigations into matters relating to his allegations. Barnes subsequently submitted written testimony to Congress, and he participated in an interview by OFHEO. His testimony and interview raised additional accounting issues and included other allegations against Fannie Mae. In light of these events, the SRC asked Paul, Weiss to determine: (1) whether the Company's investigations into Barnes's accounting allegations were conducted appropriately, and (2) whether the Company entered into the settlement agreement with Barnes to prevent him from pressing his allegations of accounting impropriety. Because the substance of Barnes's initial accounting questions concerned premium/discount adjustments to the Company's mortgage portfolio, we also inquired into the substance of his allegations. As to the issues raised by the events surrounding Barnes's allegations and departure from Fannie Mae, we draw the following conclusions: - Controller's Office management communicated with Barnes poorly, and, in some instances, inappropriately, regarding both accounting and personnel matters. The remedial measures Fannie Mae directed for Controller's Office management following the investigations into Barnes's allegations were not effective in improving the reporting environment within the Controller's Office. - Fannie Mae's investigations into Barnes's allegations were flawed in several respects, including conflicts-of-interest, inappropriate pressure to complete the investigations in an unreasonable time frame, and assigning assessment of proper GAAP accounting to Internal Audit, a department that was neither equipped nor authorized to render such determinations. 546
- Management's explanation of the issues Barnes raised relating to the existence of "anomalous" amortization factors above 100 percent or below zero percent was reasonable. - Although there was an inappropriate lack of documentation regarding the $6.5 million factor change Barnes identified, management's assertion that the amount involved was immaterial was not unreasonable. - Management's practice of editing certain conditional prepayment rates ("CPRs") was inappropriate considering, among other things, that management could not identify a consistent rationale for changing the CPRs, the Controller's Office made the changes without consulting the economists who developed the CPRs, and only the Controller's Office used the changes for calculating catch-up -- the changes were not applied consistently to all areas of Fannie Mae. - Barnes has raised additional allegations regarding Fannie Mae. The Company did not have the opportunity to investigate these allegations while he was with Fannie Mae. - The Company's decision to reach a settlement of threatened litigation by Barnes was based on an appropriate analysis of the Company's litigation risk. It was not intended to conceal misconduct by Fannie Mae or its employees or officers. II. BACKGROUND A. Barnes's Work History Barnes began his employment at Fannie Mae in October 1992 as Assistant Manager of MBS Accounting. Between 1992 and 2003, Barnes generally received top performance ratings and was promoted several times, eventually rising to the position of Manager of Premium/Discount Amortization. In this position, he supervised three analysts and reported to Mary Lewers, then Director -- Financial Accounting.(2082) Barnes's responsibilities primarily included premium/discount amortization accounting via the PDI/iPDI systems and separate cash reconciliation functions. In this role, Barnes regularly worked with Jeffrey Juliane, then Manager of New Products, who reported directly to Janet L. Pennewell. Juliane oversaw the AIMS system that transmitted to PDI/iPDI data used to calculate premium/discount amortization. - ---------- (2082) Lewers reported to Pennewell, Vice President -- Financial Reporting and Planning, and, during the period relevant to our investigation, Pennewell reported to Spencer, who became Senior Vice President and Controller. 547
B. Barnes's Allegations and the Company's Investigations Beginning as early as 1997, Barnes began to draft memoranda and e-mails documenting his concerns regarding both Fannie Mae's accounting practices and the difficulties he experienced in seeking a promotion to a Director-level position within the Controller's Office.(2083) Barnes sent numerous e-mails to Stephen Spivey, a Senior Developer in the systems area with whom Barnes had worked on PDI. Spivey stated that he paid little attention to Barnes's accounting allegations, rarely responded to them, and was unsure why Barnes sent them to him. Barnes also wrote memoranda to himself and to "file"; these memoranda contained allegations similar to those outlined in his e-mails to Spivey. Barnes's e-mails and memoranda outline a broad range of allegations regarding, among other things, perceived mistreatment and racial discrimination by his supervisors. These allegations include the following: - In 1994, Barnes was twice denied promotion to Manager positions in the Controller's Office in favor of white employees.(2084) - In 1998, another employee was moved to a fixed-wall office, despite Barnes's allegation that several minority managers held seniority over her.(2085) - In 1999, Controller Leanne G. Spencer asked an African-American employee to hold her water bottle during a presentation to Controller's Office personnel.(2086) - In 1999, another manager and her subordinates received an internal award that Barnes felt he and his subordinates should have received. Barnes alleged that this employee received the award because Controller's Office management wanted to "showcase" her.(2087) - ---------- (2083) See, e.g., E-mail from R. Barnes to R. Barnes, dated Mar. 20, 1997, FMSE 22104 (regarding being passed over for promotion in favor of white employees); E-mail from R. Barnes to S. Spivey, dated Mar. 4, 2002, FMSE-IR 4248-49 (regarding 2001 performance evaluation); E-mail from R. Barnes to S. Spivey, dated Oct. 5, 2002, FMSE 22885 (listing ten accounting-related concerns). (2084) E-mail from R. Barnes to R. Barnes, dated Mar. 27, 1997, FMSE 22104. (2085) E-mail from R. Barnes to R. Barnes, dated Nov. 9, 1998, FMSE 22105. (2086) E-mail from R. Barnes to R. Barnes, dated Feb. 24, 1999, FMSE 22108. (2087) E-mail from R. Barnes to R. Barnes, dated May 19, 1999, FMSE 22633-34. 548
- In 2000, Juliane began attending biweekly meetings with Pennewell, which Barnes viewed as a sign that Pennewell planned to promote Juliane to Director.(2088) - In 2000, Juliane and Martin Waugh were moved to fixed-wall offices.(2089) - In 2000, Spencer held in her hand, but did not refer to, notes that Barnes had prepared for her use during a meeting.(2090) - In 2001, Juliane was invited to participate in a leadership training program that Barnes perceived as a stepping-stone to a Director-level promotion. Barnes wrote that "[m]anagement was even instructed to deny the existence of the program . . . if employees asked."(2091) - During a 2001 meeting, Juliane made a "caustic comment" about Barnes being "verbose," and, rather than rebuke Juliane, Pennewell laughed. Barnes wrote, "This is as bad as Leanne rubbing my tummy, or resting her hand on my leg at the picnic or always excessively hugging."(2092) - In 2001, Barnes alleged that "senior management" was using on-top adjustments to reduce Fannie Mae's income tax liability.(2093) - In 2001, Barnes was offered a transfer to the Core Infrastructure Project, a transfer that management suggested would offer significant advancement potential. Barnes declined the transfer because his supervisors would not guarantee him a promotion to Director along with or after the transfer. In 2002, Barnes learned that the individual who accepted leadership of the Core project was promoted to Director when he was transferred.(2094) - ---------- (2088) E-mail from R. Barnes to R. Barnes, dated July 20, 2000, FMSE 22138. (2089) E-mail from R. Barnes to R. Barnes, dated Aug. 23, 2000, FMSE 22139. (2090) E-mail from R. Barnes to S. Spivey and R. Barnes, dated Oct. 17, 2000, FMSE 22147. (2091) E-mail from R. Barnes to S. Spivey, dated Feb. 15, 2001, FMSE 22183-84. (2092) E-mail from R. Barnes to C. DeFlorimonte, dated July 19, 2001, FMSE 22196. (2093) E-mail from R. Barnes to S. Spivey, dated Aug. 8, 2001, FMSE 22607. (2094) E-mail from R. Barnes to S. Spivey, dated Mar. 6, 2002, FMSE-IR 4250. 549
- In March 2002, Barnes claimed that the Controller's Office terminated seven employees, one of whom was a Director. All of the employees, according to Barnes, were religious or ethnic minorities.(2095) - In May 2002, Barnes wrote that Fannie Mae was "consciously using the REMICs to reduce short term income." He alleged that Raines and Howard had decided to move income from 2002 via REMIC amortization because the Company had enough income in 2002 to meet that year's "goals."(2096) Later that year, Barnes compiled a list of Fannie Mae's accounting practices that he found troubling, including, among other things, premium/discount amortization using "negative factors" and "intentionally not designing audit trails" for the AIMS system.(2097) - In May 2002, Spencer held a "grief session" for Controller's Office personnel after the suicide of a former Fannie Mae employee. Barnes wrote that this was hypocrisy and "ridiculous" because Spencer had not even informed Controller's Office personnel of the employee's resignation.(2098) - In May 2002, Barnes reported that he mistakenly had believed that eight minority Directors had been terminated earlier in May. Barnes wrote that one of the fired Directors was white and that another white director was offered a package to resign "purely to make it look like race had no impact."(2099) - In 2003, Barnes wrote that, unlike other managers, he had not been informed of Fannie Mae's new guidelines regarding snow and inclement weather.(2100) - ---------- (2095) E-mail from R. Barnes to S. Spivey, dated Mar. 7, 2002, FMSE-IR 4245. (2096) E-mail from R. Barnes to S. Spivey, dated May 7, 2002, FMSE-IR 4241. (2097) E-mail from R. Barnes to S. Spivey, dated Oct. 5, 2002, FMSE 22885. (2098) E-mail from R. Barnes to S. Spivey, dated May 21, 2002, FMSE-IR 4252-56, at FMSE-IR 4252. (2099) E-mail from R. Barnes to S. Spivey, dated May 30, 2002, FMSE-IR 4251. In his e-mail, Barnes refers to an earlier e-mail he says he sent about directors being terminated; we did not locate that e-mail. (2100) E-mail from R. Barnes to R. Barnes, dated Feb. 11, 2003, FMSE 22229. 550
- In October 2003, Donald M. Remy sent an e-mail to a broad group of Controller's Office and Internal Audit managers regarding preparation for the OFHEO Special Examination. Barnes did not receive this e-mail directly but instead was informed of it by a colleague. Barnes claimed that management intentionally excluded him from the team preparing for the OFHEO review.(2101) During the summer of 2003, in addition to continuing to document the perceived discrimination issues listed above, Barnes also began to compile data relating to the Company's FAS 91 systems. As part of this data compilation, it appears that he asked one of his subordinates, Patricia Wells, to create a CD of amortization documents that he added to his file. Barnes told Wells that this CD had been requested by Lewers, but, in reality, Barnes used it to perform analyses of perceived FAS 91 noncompliance. On July 18, 2003, during a regularly scheduled monthly meeting with Lewers, Barnes included as a topic of discussion items regarding Fannie Mae's FAS 91 accounting.(2102) Barnes stated that he told Lewers that there appeared to be a problem with unamortized balances increasing and exceeding historical costs. According to Barnes, because Controller's Office management was not receptive to criticism, he raised these concerns in the form of questions, rather than through direct confrontation.(2103) Barnes and Lewers had a brief (one-minute) conversation regarding the perceived FAS 91 problems, but because Barnes framed his questions as relating to data analysis, rather than to GAAP noncompliance, Lewers was not particularly concerned by these questions. During a meeting on the afternoon of July 29, 2003, Pennewell informed Barnes that Fannie Mae had promoted Juliane to Director.(2104) As a result of this promotion, Juliane became Barnes's immediate supervisor.(2105) Approximately one hour after this announcement, Barnes sent an e-mail to Sam Rajappa, Senior Vice President -- Operations Risk (with responsibility for Internal Audit), requesting a confidential meeting "regarding analysis and research [Barnes had] been conducting for a number of weeks."(2106) Rajappa responded by asking Barnes to meet with him and Ann Eilers, Vice - ---------- (2101) Handwritten notes on e-mail from Debbie D. Milton to L. Spencer, et al., dated Oct. 13, 2003, FMSE 22716-18, at FMSE 22716. (2102) See One on One Meeting July 18, 2003, dated July 18, 2003, FMSE-IR 132789 (listing meeting agenda item of "detailed analysis of reports, factors and balances"). (2103) Barnes stated that Lewers often did not respond to his questions in any way, instead sitting in silence until Barnes moved on to another topic. (2104) Mem. from R. Barnes to File, dated July 29, 2003, FMSE 22265. (2105) See id. (2106) E-mail from R. Barnes to S. Rajappa, dated July 29, 2003, FMSE 22259. 551
President -- Internal Audit, on Monday, August 4.(2107) Barnes did not detail his accounting concerns in the e-mail to Rajappa; as a result, Rajappa did not know beforehand what issues Barnes intended to raise when they met. After receiving Barnes's e-mail, but before the meeting, Rajappa contacted Ann M. Kappler, General Counsel, and informed her that an employee had requested a confidential meeting to discuss accounting issues. Rajappa told Kappler he would follow up with her after the meeting with the employee. It does not appear that Rajappa identified Barnes as the employee in question at that time. III. FINDINGS REGARDING THE INVESTIGATION INTO BARNES'S ACCOUNTING ALLEGATIONS A. August 4, 2003 At the August 4 meeting, Barnes shared with Rajappa and Eilers a binder of material he had prepared containing his analysis of certain "anomalies" within Fannie Mae's premium/discount amortization accounting system. Barnes highlighted concerns relating to: (1) deferred items amortized in excess of 100 percent; (2) deferred items amortized in reverse (also referred to as "negative amortization"); and (3) unusually rapid amortization.(2108) Barnes informed Rajappa and Eilers that he had not raised these concerns to his supervisors because he did not feel comfortable doing so. Barnes explained to Rajappa that Spencer and Pennewell repeatedly had told him that he had "no need to know" about the amortization factors, and that he did not understand them. Furthermore, according to Barnes, Pennewell, Lewers, and Spencer had excluded him from meetings on the topic, and he stated that employees who raised questions about accounting were viewed by management as not being "part of the team." Rajappa encouraged Barnes to inform his management chain of the accounting concerns because Rajappa would be required to do so to investigate Barnes's concerns properly in any event. Rajappa also told Barnes that he would not be mistreated for raising these issues with management. After meeting with Barnes, Rajappa and Eilers immediately met with Kappler. Kappler and Rajappa together decided that Internal Audit would investigate Barnes's accounting allegations and that the Legal Department would investigate Barnes's concerns about the reporting environment in the Controller's Office. They also agreed that Rajappa would inform KPMG and Thomas P. Gerrity, Chair of the Audit Committee, about the allegations. - ---------- (2107) E-mail from S. Rajappa to R. Barnes, dated July 29, 2003, FMSE 22797. (2108) Barnes later raised concerns regarding manual adjustments made to the FAS 91 amortization modeling factors. 552
B. August 5, 2003 During the morning of August 5, Barnes sent an e-mail to Spencer, Pennewell, Lewers, Stawarz, and Juliane listing his concerns that Fannie Mae's FAS 91 accounting was not in compliance with GAAP.(2109) This was the first time he had contended to his supervisors that any Fannie Mae accounting practice relating to amortization of premiums and discounts did not comply with GAAP. He also provided each with a copy of the binder that he had provided to Rajappa.(2110) Given the seriousness of the allegations, the recipients of this e-mail engaged in immediate verbal and written communication regarding Barnes's assertions. Within several hours, Pennewell decided to task Juliane with determining whether Barnes's concerns were justified.(2111) Juliane reacted negatively to Barnes's e-mail, forwarding it to his subordinate, Rene LeRouzes, with a note stating: "look at the [expletive] that I'm dealing with. we need to jump on this. Need to look at the [factors] that are hosed Let's discuss."(2112) Juliane explained to us that his frustration stemmed from the fact that he had explained these factor issues to Barnes previously, and Barnes never had indicated disagreement or asked any follow-up questions before sending his e-mail to the entire management group of the Controller's Office. On August 5, 2003, Rajappa called a meeting with Paul Jackson, an Internal Audit Director, and Joyce Philip, an Internal Audit Manager, to begin the investigation into Barnes's concerns.(2113) Rajappa selected Jackson and Philip because, in July 2003, they had completed an audit of Fannie Mae's amortization system ("July audit"). In the July audit, Internal Audit had found control problems within the system, as well as the need for a $155 million accounting adjustment to correct for MBS incorrectly identified by the wrong FAS 91 type.(2114) - ---------- (2109) E-mail from R. Barnes to M. Lewers, et al., dated Aug. 5, 2003, FMSE-IR 496799. (2110) See id. ("The supporting documentation is too expansive to cover in email. Please see the hard copy provided in your inbox."). (2111) E-mail from J. Pennewell to J. Juliane, et al., dated Aug. 5, 2003, FMSE 23266 ("Roger and Jeff - I'd like to request that you guys get together to discuss Roger's analysis, and [report] back to Dick, Leanne, and me I'd like this to be a high priority."). (2112) E-mail from J. Juliane to R. LeRouzes, dated Aug. 5, 2003, FMSE-E 117364. (2113) The Internal Audit team later enlisted the assistance of Gunes Kulaligil, an Internal Audit staff member specializing in data analysis and validation. (2114) Audit Report: Office of Auditing, Amortization Audit, dated July 9, 2003, FMSE-IR 176943-50. 553
During our interview with Jackson, he noted his surprise that Barnes had not raised any concerns about amortization accounting during the July audit. Jackson particularly was surprised because Barnes generally had been critical of the Controller's Office procedures during the audit. Jackson believed Barnes was given ample opportunity during the July audit to raise these issues, but he had not done so. Barnes explained to us that he did not raise his concerns about the Controller's Office or accounting to Internal Audit during the July audit because he had hoped that Internal Audit would find the problems without his help. He said that he did not complete his analysis of the anomalies until after Internal Audit issued the July audit results. During our interviews, no one in Internal Audit expressed any concerns about using the same auditors who recently had examined the amortization area. Rajappa indicated that he did not believe the auditors had a conflict-of-interest because there was no connection between the previous audit and Barnes's allegations. Philip, however, stated that Internal Audit had become aware during the July audit of at least one of the issues Barnes later raised. Philip also explained that she believed that she and Jackson were the appropriate individuals to conduct the investigation because they were the most familiar of all Internal Audit staff with the issues to be investigated. However, both Kappler and Deborah House of the Legal Department's Office of Corporate Compliance ("OCC") subsequently concluded (in hindsight) that Internal Audit should not have been involved. Although we have no evidence that either the assignment of Internal Audit to this investigation or the conduct of the investigation itself was undertaken to conceal misconduct, Internal Audit had a conflict-of-interest in performing this work because of its previous audit of the same area. Specifically, to substantiate Barnes's allegations, the auditors would have had to find that their own conclusions during the July audit were incorrect. Both Kappler and Rajappa should have recognized this conflict, and, instead of assigning Internal Audit to the investigation, they should have assigned the investigation into Barnes's accounting allegations to someone else. From the time he learned of Barnes's allegations, Rajappa was concerned that any problems with the catch-up calculations implicated by Barnes's complaints might prevent the CEO and CFO from certifying the previous quarter's financial results. The preliminary CEO/CFO certification meeting for that quarter had taken place on August 1, and the overall FAS 91 catch-up process had been discussed at that meeting; final certification tentatively had been scheduled for either August 13 or 14. Thus, Rajappa believed Barnes's subsequent allegations could impede certification. Despite acknowledging this concern, Rajappa maintained that he did not give Jackson and Philip any deadlines to complete their work or urge them to finish their portion of the investigation to allow for certification by the deadline. After meeting with Jackson and Philip, Rajappa separately briefed Howard, Gerrity, and Raines, as well as Mark Serock and Harry Argires of KPMG on Barnes's allegations and Fannie Mae's plan to investigate those allegations. No one 554
whom Rajappa briefed disagreed with Internal Audit's workplan, and each person asked to be kept informed of developments in the investigation.(2115) It is unclear whether the CEO/CFO certification was explicitly mentioned, but all of the relevant parties whom we interviewed noted that they were aware of the approaching certification date and that Barnes's allegations might affect Raines's and Howard's ability to sign their certifications. Kappler assigned the investigation into Barnes's complaints regarding the Controller's Office culture to House. Rajappa and House first met to discuss the investigation on August 5. During that meeting, Rajappa described to House the division of the investigation between Internal Audit and the Legal Department and provided her with general information about Barnes's cultural allegations as well as Barnes's complaints about specific individuals within his reporting chain. Rajappa asked House to keep him apprised of developments in OCC's investigation. Also on August 5, Spencer called a meeting of Controller's Office personnel to discuss Barnes's allegations. It is undisputed that Pennewell became agitated towards Barnes during this meeting. Pennewell explained to us that she reacted angrily because she felt that Barnes was calling her integrity into question by going over her head to Spencer with his concerns. Barnes felt that Pennewell's behavior constituted retaliation for his having approached Internal Audit, and, shortly thereafter, he informed House of the meeting and that he had been confronted by Pennewell, Spencer, and Juliane regarding bringing his concerns to Internal Audit. House telephoned Spencer and told her that the confrontation should not have occurred, particularly in the context of an investigation. Spencer, who felt that Pennewell was demoralized already by the allegations, did not inform Pennewell of the call from House.(2116) It seems that Spencer, Pennewell, and Lewers subsequently determined that the appropriate way to avoid any appearance of retaliation was to have as little contact with Barnes as possible.(2117) C. August 6-7, 2003 On August 6 and 7, the Internal Audit team met with Controller's Office personnel (including Barnes, Juliane, Lewers, Wells, and LeRouzes) and began analyzing - ---------- (2115) Rajappa told Paul, Weiss that KPMG affirmatively agreed with Internal Audit's workplan and did not offer any suggestions of it own. Separately, Rajappa stated that Howard expressed some anger over the fact that Rajappa had briefed Raines before briefing Howard. (2116) House later called Pennewell directly to counsel her against holding meetings with Barnes to discuss his allegations. (2117) Barnes stated that his supervisors ostracized him after he raised his allegations. Barnes interpreted Spencer's, Pennewell's, and Lewers's tactical avoidance as retaliation, whereas Spencer, Pennewell, and Lewers believed avoiding contact would prevent any claims of retaliation. 555
data in relation to Barnes's allegations.(2118) Juliane, in particular, assisted Internal Audit in gathering and analyzing the relevant data.(2119) Jackson characterized Juliane's demeanor during their conversations as "defensive and devastated," primarily because Juliane felt his integrity was being questioned. In fact, Juliane later stated to Jackson that he had not slept for days after the allegations were made. Jackson also kept Barnes informed of the status of the investigation, informing him during this period that Internal Audit had not yet substantiated Barnes's allegations. Witnesses provided inconsistent descriptions of the results of Internal Audit's interviews. Philip indicated that Internal Audit had uncovered at least some of the issues Barnes raised during the July audit and that the Controller's Office viewed these as problems to be corrected. Controller's Office personnel, however, indicated that these issues were not problematic at all; instead they suggested Barnes simply misunderstood the data he had compiled.(2120) D. August 8, 2003 On the morning of August 8, 2003, Rajappa, Eilers, Jackson, Philip, and Gunes Kulaligil of Internal Audit; Spencer, Pennewell, Lewers, Jonathan Boyles, and Juliane of the Controller's Office; Serock and Michael Tascher of KPMG; and House of the Legal Department all met with Barnes to discuss Internal Audit's investigation into Barnes's accounting allegations.(2121) Witnesses have differing recollections as to the purpose of the meeting: Most viewed it as Internal Audit's presentation of its conclusions, while the Internal Audit attendees considered it a meeting to update everyone on the status of Internal Audit's investigation. - ---------- (2118) Internal Audit's workplan had three stages: (1) Verify the mathematical accuracy of Barnes's calculations; (2) determine whether the factor anomalies were isolated occurrences or pervasive throughout Fannie Mae's portfolio; and (3) determine whether the anomalies impacted the analysis of Fannie Mae's catch-up position that was part of the CEO/CFO certification process for the second quarter of 2003. Cf. Fannie Mae Office of Auditing: Amortization Investigation, dated Aug. 2003, FMSE 23295-96. (2119) See Analysis on Factors, dated August 5, 2003, FMSE 23267 ("Analysis performed by Jeff Juliane to explain apparent discrepancies in factors."). (2120) See infra Subsection IV ("Findings Regarding Roger Barnes's FAS 91-Related Allegations"). (2121) Summarized Minutes of the Meeting (8/8/03): Unamortized Balances and Factor Analysis, dated Aug. 8, 2003, FMSE 24411-19, at FMSE 24417-19 (hereinafter "August 8 Minutes"). Rajappa scheduled this meeting as early as August 6 - just two days after meeting with Barnes for the first time. 556
At the meeting, Juliane explained the source of the anomalies Barnes had identified.(2122) No one disagreed that Barnes had identified the anomalies correctly; instead, the only concern was whether the existence of the anomalies meant that Fannie Mae's accounting did not comply with GAAP.(2123) Serock expressed KPMG's initial view that these anomalies did not indicate GAAP noncompliance.(2124) It is undisputed that Barnes said little, if anything, during the meeting. Jackson drafted minutes of the meeting, and Rajappa circulated the draft to all attendees.(2125) The final minutes indicate that Internal Audit, KPMG, and Boyles, then head of Financial Standards, all agreed that Fannie Mae's accounting was in compliance with GAAP (to the extent Barnes had challenged it was not).(2126) Internal Audit inappropriately led the Company to believe that it was investigating the accuracy of Barnes's allegations regarding GAAP noncompliance, despite Internal Audit's having neither the capability nor the intention of investigating GAAP issues. This course of action began when Rajappa and Kappler asked Internal Audit to oversee the accounting investigation, which, on the face of Barnes's allegations, raised issues concerning GAAP compliance. And, it continued when Rajappa made statements during the August 8 meeting that all attendees interpreted as confirming GAAP compliance.(2127) Despite the indications in the August 8 minutes, Rajappa now denies that he opined on GAAP issues.(2128) In his interview with Paul, Weiss, Rajappa stated that Internal Audit's conclusion was that the FAS 91 system was in compliance with the - ---------- (2122) August 8 Minutes at FMSE 24417. (2123) Id. ("Juliane stated that he had reviewed Roger's amortization analysis and this analysis was correct."). (2124) Id. at FMSE 24418. (2125) E-mail from S. Rajappa to L. Spencer, et al., dated Aug. 25, 2004, FMSE 22734 (attaching draft meeting minutes for review and comments). Barnes was an addressee of this e-mail, but he did not receive the draft minutes because he was on leave at the time of their circulation. (2126) August 8 Minutes at FMSE 24418. (2127) Serock stated that he understood the reference in the minutes to being in compliance with GAAP was not to suggest that the Company's review actually tested GAAP compliance, but instead that the process itself was compliant with GAAP and that the allegations were incorrect. (2128) No one else within Internal Audit (Philip, Jackson, and Eilers) identified any inaccuracies in the minutes. 557
internal standards established by Financial Standards rather than in compliance with GAAP. Additionally, he maintained that the conclusion that the factors identified by Barnes were accounted for properly was based upon representations by Juliane and Pennewell. Rajappa had multiple opportunities to revise both the meeting minutes and OCC report, discussed below, that referred to Internal Audit's GAAP determinations, but he did not do so. Indeed, he did not express any disagreement with these documents until his interview with us. Since every party that opined on whether Fannie Mae's amortization practices raised by Barnes were in compliance with GAAP relied on Internal Audit's findings, the reliability of those opinions could be called into question. E. August 9-12, 2003 Following the August 8 meeting, Rajappa individually briefed Raines, Howard, and Gerrity on Internal Audit's conclusions. Rajappa gave these briefings in anticipation of a previously scheduled August 12 Audit Committee meeting set to discuss certification of the previous quarter's financial results. In each of his briefings, Rajappa apparently stated that Internal Audit's testing was complete, that Internal Audit's initial conclusion was that Barnes's allegations had no impact on the second quarter financial statements, and that Rajappa felt comfortable signing his sub-certification. Yet, at the same time, Jackson and Philip still were performing testing of Barnes's allegations regarding manual amortization factor changes, and they continued to do so until approximately the end of September. Rajappa and Kappler also told Gerrity and the Audit Committee that Barnes was satisfied with Internal Audit's investigation.(2129) Barnes, however, never made such a statement; instead, Barnes had told House that he was glad the Company had looked into his concerns. House either mistransmitted this statement to Rajappa or Rajappa misunderstood House's characterization. Regardless, Gerrity reported that he was left with the understanding that Barnes's concerns had been addressed, and that Barnes would not be pressing the issues further. On August 12, 2003, the Audit Committee held a telephonic meeting.(2130) During this meeting, Gerrity, by prior arrangement with Rajappa and Kappler, raised the Barnes matter to the Audit Committee and Kappler provided a report on the Legal - ---------- (2129) E-mail from S. Rajappa to A. Kappler, dated Aug. 11, 2003, FMSE-SP 3400 (stating, in planning for Audit Committee meeting, "I will note that Roger has indicated that he is satisfied"); Audit Committee Meeting: Quarterly Certification, dated August 12, 2003, FMSE 15573-76, at FMSE 15576 ("[T]he employee has expressed satisfaction that his concerns have been addressed."). (2130) The Company certified its quarterly financial results on August 14, 2003. Fannie Mae Form 10-Q, dated Aug. 14, 2003, FMSE-E_EC 113092-309, FMSE-E_EC at 113307, 113309. 558
Department's investigations.(2131) Rajappa then informed the Audit Committee that Internal Audit had investigated the accounting issues and that both the Company and KPMG had found that these issues presented no problems related to the certification.(2132) Rajappa later stated that he did not know what independent analysis, if any, KPMG performed to reach this conclusion. There is no doubt that Internal Audit felt pressured to produce the results of its investigation in an unreasonably short period of time due to the upcoming CEO/CFO certification, and it appears that Rajappa himself was the source for that pressure. Our conclusion is based primarily on the fact that Rajappa provided "conclusions" from the investigation to Raines and Gerrity after only three days of what eventually became a two-month process. It was inappropriate for Rajappa to inform Raines and the Audit Committee that the investigation was complete when, in fact, a substantial amount of test work remained to be performed. IV. FINDINGS REGARDING ROGER BARNES'S FAS 91-RELATED ALLEGATIONS In addition to reviewing the Company's investigation into Barnes's FAS 91 accounting allegations, we also independently and substantively reviewed certain of these allegations themselves.(2133) For example, Barnes alleged that: (1) Management improperly used amortization factors that were less than zero or greater than 100 percent to calculate premium and discount amortization, and (2) improperly "develop[ed] factors to meet desired objectives."(2134) We discuss these two issues below. A. Factor Anomalies To calculate amortization, Fannie Mae applies percentages, known as amortization factors or simply "factors," to original premium and discount amounts. Fannie Mae uses cumulative factors to calculate accumulated, or "inception to date," amortization. In theory, since the premium or discount on any given loan cannot be amortized more than 100 percent or less than zero percent, cumulative factors should always be between zero and 100 percent. However, documents indicate and Fannie Mae personnel have stated that, on occasion, management used cumulative factors that were - ---------- (2131) Minutes of the Meeting of the Audit Committee of the Board of Directors of Fannie Mae, dated Aug. 12, 2003, FMSE 15573-76, at FMSE 15576. (2132) Audit Committee Meeting: Quarterly Certification, dated Aug. 12, 2003, FMSE 23184-85, at FMSE 23185. (2133) Other allegations made by Barnes, including allegations regarding realignments and Fannie Mae's use of Account 1622, are discussed in other chapters of this Report. (2134) See Mem. from Finance Division Manager to Office of Chairman, F. Raines and T. Howard, dated Sept. 23, 2002, FMSE 22882-84, at FMSE 22882. 559
negative or greater than 100 percent to calculate premium and discount amortization. Barnes alleged that because of these "anomalous" factors, the Company's financial statements were not in compliance with GAAP.(2135) Management, specifically Juliane and Lewers, explained that these factor anomalies sometimes occur when premiums and discounts are combined into the same FAS 91 "bucket" in Fannie Mae's amortization systems.(2136) In the case of the amortization of REMIC securities, which were modeled at the Committee on Uniform Securities Identification Procedures ("CUSIP") level, multiple REMIC CUSIPs are combined into a single FAS 91 bucket for financial reporting purposes. When a FAS 91 bucket includes CUSIPs with both premiums and discounts there is a potential to have anomalous factors. Below is a hypothetical example of this situation: ORIGINAL AMORTIZATION ACCUMULATED Premium/(Discount) Factor Amortization Premium $ 11,000,000 70% $ 7,700,000 Discount ($ 10,000,000) 50% ($ 5,000,000) Net $ 1,000,000 270% $ 2,700,000 In this example, a FAS 91 bucket for REMICs contains two CUSIPs, one with a $11 million premium and the other with a $10 million discount. Combined, the two REMICs net to a premium of $1 million. If the cumulative amortization factor for the premium CUSIP were seventy percent, Fannie Mae would have amortized $7.7 million of the premium. If the cumulative amortization factor for the discount CUSIP were fifty percent, Fannie Mae would have amortized $5 million of the discount. When combined, it appears as though Fannie Mae amortized $2.7 million of net premium even though only $1 million of premium was available to be amortized. The cumulative amortization factor in this case would be 270 percent. However, the seemingly anomalous factor would be appropriate since the amortization factors for the individual CUSIPs were between zero and 100 percent, and the anomalous factor resulted only when these CUSIPs were combined for financial reporting purposes. This explanation and real examples were presented to and accepted by Internal Audit, the OCC, and KPMG.(2137) Although we did not investigate every instance - ---------- (2135) Mem. from R. Barnes to File, dated Aug. 5, 2003, FMSE 22899-90, at FMSE 22899. (2136) See Tr. of June 8, 2004 OFHEO Dep. of J. Juliane at 65:6-67:4, 125:10-126:9 (hereinafter "Juliane OFHEO Testimony"); Tr. of July 13, 2004 OFHEO Dep. of M.Lewers at 251:3-258:17. (2137) See Office of Corporate Compliance: Decision # 2003-1 (Part A, Amortization and Documentation Issues), dated Sept. 29, 2003, FMSE 24411-19. 560
in which anomalous factors may have occurred at Fannie Mae, we conclude that management's explanation as to why cumulative factors are sometimes above 100 percent or below zero is reasonable. We also found that anomalous factors existed consistently at least as far back as October 1998.(2138) We conclude that this supports management's contention that anomalous factors were a by-product of the Company's modeling process. B. Manual Factor Changes We understand that Barnes's concern that management manually adjusted amortization factors to manage earnings related to an incident in July 2003, when management manually changed a factor generated by AIMS to elicit a $6.5 million impact on the amortization calculated by iPDI. Juliane explained that this manual factor change was made to correct the amortization performed by iPDI. Juliane said that this occurred as part of a reconciliation process, in which the results produced by iPDI were compared to the predicted values generated by AIMS. In this instance, the amount of amortization expense calculated by AIMS was $6.5 million higher than the amount that was booked by iPDI.(2139) Juliane explained that he discussed this discrepancy with Pennewell, and they concluded that AIMS contained more accurate data than iPDI. Therefore, management made a manual factor change to increase amortization expense in iPDI by $6.5 million.(2140) This explanation was presented to Internal Audit, the OCC, and KPMG. Each of these groups concluded that there was an inappropriate lack of documentation surrounding the $6.5 million factor change. However, each also accepted management's assessment that the amount involved in the factor change was immaterial.(2141) Given the size of the adjustment, its timing (it did not occur at quarter- or year-end),(2142) and the lack of evidence that the adjustment had an improper purpose (such as managing earnings), that conclusion does not seem unreasonable. - ---------- (2138) Purchase Discount Amortization Management System, Catch-up Adjustments, Reporting Period 10/1998, dated Dec. 3, 1998, FMSE-IR 22442-57. (2139) See Juliane OFHEO Testimony at 118:2-119:11. (2140) See id. (2141) See Office of Corporate Compliance: Decision # 2003-1 (Part A, Amortization and Documentation Issues), dated Sept. 29, 2003, FMSE 24411-19 (hereinafter "OCC Part A"). (2142) Had the adjustment occurred at quarter or year-end, it might have raised a more significant issue. 561
C. CPR Edits Although Barnes did not raise this issue with Internal Audit, Barnes separately alleged that Fannie Mae personnel were "[m]anaging income by developing factors to meet desired objectives."(2143) We investigated this issue and learned that management, from time to time, would manually change CPRs produced by Valuation Net ("VN")(2144) before they were inputted to AIMS. These CPR edits had the potential to allow Fannie Mae to manage earnings because CPRs directly affect estimates of future cash flows, indirectly affect amortization factors, and ultimately affect earnings. Pennewell explained that CPR edits were necessary when the CPRs produced by VN varied significantly from Fannie Mae's current outlook of prepayment rates. Juliane explained that he would propose CPR changes from time to time based on market analysis. We understand that Fannie Mae would use prepayment estimates published by Bloomberg to create a CPR range representing high and low market estimates based on that data.(2145) Fannie Mae would then compare the VN produced CPRs to this market range and make adjustments to the CPRs when management deemed the adjustment to be appropriate. Juliane indicated that, based on an analysis of the prepayment data from the Bloomberg website, he would suggest CPR edits to Thomas A. Lawler, who would sign-off on the adjustment.(2146) Although Lawler admitted that he signed off on these CPR edits, he stated that his function was to check the adjustments only to see if they seemed "reasonable," not to determine whether the adjustments were appropriate. Lawler also stated that these CPR edits were made for the use of the Controller's Office only, and not applied globally. When we asked Leonard Mills, the Fannie Mae employee responsible for the modeling in VN that produced CPRs, about these CPR edits, he said that he was unaware that anyone in Financial Reporting, within the Controller's Office, ever had altered the CPR outputs of VN manually. Mills did say that Pennewell had approached him at some point to ask his opinion of the Bloomberg prepayment rate ranges, but he rejected the idea of Fannie Mae using these ranges for a number of reasons, including: (1) The prepayment range was likely very large; (2) the range obtained from Bloomberg was not based on the same assumptions as Fannie Mae's data; (3) the ranges may have been based on projections that were up to a month old; and (4) the results were not analogous to the results of the modeling performed by Fannie Mae. - ---------- (2143) See Mem. from Finance Division Manager to Office of the Chairman, F. Raines and T. Howard, dated Sept. 23, 2002, FMSE 22882-84. (2144) VN is a modeling application, which for FAS 91 amortization purposes, applies an interest rate path to its library of prepayment models to estimate CPRs. (2145) CPR Analysis, dated Mar. 15, 2003, FMSE-SP 2936; CPR Analysis, dated Dec. 15, 2001, FMSE 217689. (2146) See Juliane OFHEO Testimony at 45:20-46:10. 562
For several reasons, we find that it was improper for Juliane, as a Director in Financial Reporting within the Controller's Office, to propose and process CPR edits. First, CPRs are generated by VN through sophisticated models monitored by eight economists and headed by Mills, who considered Fannie Mae's CPR modeling capabilities to be "state of the art." Pennewell admitted that, although she does believe the CPR edits were justified, in hindsight, it would have been preferable from an internal controls perspective to rely solely on the CPRs generated by Portfolio Analytics. Second, the CPR edits were made without consulting Mills, whose input should have been considered critical to any decision regarding the accuracy of the VN outputs. Third, any warranted changes in CPRs should have been applied consistently to all areas of Fannie Mae, not just the premium and discount amortization calculated by the Controller's Office. Fourth, there does not appear to have been any consistent rationale for the CPRs that management chose to alter or for the magnitude to which they were adjusted.(2147) v. FINDINGS REGARDING THE INVESTIGATION INTO BARNES'S CULTURAL COMPLAINTS During early August, as Internal Audit was completing its investigation, OCC began its investigation into Barnes's additional allegations about the Controller's Office culture.(2148) OCC attorneys House, Karl Chen, and Patricia O'Connor interviewed a substantial number of Controller's Office personnel and management over several weeks.(2149) OCC found that some employees in the Pennewell-Lewers reporting chain shared Barnes's concerns about the atmosphere and management's attitude towards employees who raised questions, but most employees in other groups indicated they felt comfortable raising issues to their supervisors.(2150) Based on Barnes's complaints and the - ---------- (2147) Some CPRs that originally fell outside of the market range were adjusted to the approximate edge of the range while others were adjusted to the approximate middle of the range. Still others were adjusted to an amount that was within the range but neither at the edge nor the middle of the range. Fannie Mae even adjusted some CPRs that fell within the market range. In many cases, rather than adjusting the CPRs to be more consistent with the middle of the range, the adjustments forced CPRs closer to an edge of the range. (2148) The OCC investigation commenced on August 5, 2003, but the majority of OCC's interviews occurred after the August 8 meeting. This does not cause concern because the August 8 meeting focused on accounting issues, rather than the cultural issues Barnes had raised. (2149) O'Connor was and is employed by the Office of Corporate Justice; she was tasked to OCC's Barnes investigation because OCC did not have significant investigatory experience at the time. (2150) Office of Corporate Compliance: Decision #2003-1 (Part B, Reporting Environment), dated Oct. 6, 2003, FMSE 24562-66, at FMSE 24563 (hereinafter "OCC Part B") ("Staff in the Controller's office who are not in the same reporting chain as Barnes . . . indicated that they feel very comfortable raising issues."). 563
employee statements made in the interviews, OCC concluded that Spencer, Pennewell, and Lewers were not fostering an open environment in which employees felt comfortable raising concerns.(2151) Our interviews confirmed that the Controller's Office suffered from managerial deficiencies, particularly as to communications regarding personnel decisions and accounting policies. Lewers, for example, inappropriately failed to address Barnes's personnel concerns and his accounting complaints. Both Lewers and Pennewell, by refusing to address accounting matters with him, acted in such a way that Barnes reasonably believed he would receive negative treatment if he directly raised any accounting-related concerns. Proper managerial action in this regard might have allayed Barnes's concerns before he raised them formally. VI. FINDINGS REGARDING THE INVESTIGATION INTO BARNES'S DISCRIMINATION ALLEGATIONS Early in the OCC investigation (on or about August 8), Barnes told O'Connor and House that Controller's Office management repeatedly had discriminated against him on the basis of race and gender. House referred this complaint to Alan Tanenbaum of the Office of Corporate Justice ("OCJ").(2152) Tanenbaum and his staff commenced an investigation into Barnes's discrimination allegations, including document review and interviews. This investigation continued after Barnes left Fannie Mae, and OCJ ultimately concluded that there was no discrimination, but that there were environmental problems within the Controller's Office.(2153) VII. FINDINGS REGARDING THE RESULTS OF THE OCC INVESTIGATIONS A. OCC Decision Part A: Internal Audit's Analysis The OCC issued its Decision 2003-1, Part A, on September 29, 2003; this decision focused on the accounting issues Barnes had raised.(2154) House and Rajappa co-signed the Decision, which incorporated Internal Audit's analysis.(2155) The Decision - ---------------- (2151) OCC Part B at FMSE 24564. During the OCC interviews, many individuals complained about Spencer, Pennewell, and Lewers. One individual said that she even heard Spencer referred to as "the blonde ambition tour." Undated Summary of Interview with a Fannie Mae Employee, PW-PR 283521-23, at PW-PR 283522. (2152) Fannie Mae has tasked OCJ with investigating discrimination complaints. (2153) Mem. from A. Tanenbaum to L. Spencer, dated Jan. 12, 2004, FMSE-IR 508828-40 (hereinafter "OCJ Report"). (2154) See generally OCC Part A. (2155) Id. This decision also incorporated the August 8 Minutes. 564
concluded, in relevant part, that "Internal Audit is satisfied that the amortization issues raised by Barnes can be reasonably explained, [and] are in compliance with GAAP . . . ."(2156) On October 2, 2003, Jackson and Philip met with Barnes to provide him with the results of their investigation.(2157) It appears that none of the participants in the meeting were aware that Internal Audit's formal findings already had been issued. Philip and Jackson each noted that Barnes did not question or complain about the adequacy of Internal Audit's review of his allegations. After the meeting, however, Barnes sent Jackson an e-mail stating that Barnes neither agreed nor disagreed with Internal Audit's conclusions.(2158) B. OCC Decision Part B: Controller's Office Culture On October 6, 2003, OCC issued Decision 2003-1, Part B, which addressed the environmental issues in the Controller's Office.(2159) OCC attorney Chen prepared the first draft of the report, concluding that the Controller's Office culture was not conducive to open discussions regarding accounting issues.(2160) House revised the draft and sent it to Remy for review.(2161) Remy alerted House that he had concerns about the report because it was "not crisp and clear."(2162) After discussion between House and Remy, House drafted and sent to Remy and Kappler a "kinder gentler" version of the report, which, according to House, was "much less harsh and more sympathetic than earlier versions" in its criticism of Controller's Office management.(2163) - ---------- (2156) Id. at FMSE 24413. (2157) See E-mail from R. Barnes to P. Jackson, dated Oct. 2, 2003, FMSE 22719 (thanking Jackson for meeting to discuss Internal Audit's findings). (2158) Id. (2159) See generally OCC Part B. (2160) Untitled draft, dated Sept. 2, 2003, FMSE-IR 463838-49, at FMSE-IR 463847-48. (2161) Remy oversees both OCC and Fannie Mae's litigation group. (2162) Chen had been with OCC only two weeks when the Barnes investigation began; we understand that this was the first draft report of this type to come out of OCC and thus potentially could have been subject to more scrutiny than might otherwise be expected. (2163) E-mail from D. House to A. Kappler, dated Oct. 2, 2003, FMSE-E 2170320 (attaching revised draft and noting that Remy had approved it). 565
The Decision directed certain remedial actions to improve the Controller's Office reporting environment.(2164) Most notably, OCC required that Pennewell, Lewers, and Juliane receive "coaching" to assist them in communicating with their subordinates.(2165) Pennewell, however, informed Lewers that Lewers was being assigned a coach to aid her professional development, rather than noting the relationship of this assignment to the Barnes matter. As a result, we conclude that Lewers had no reason to believe that her behavior towards Barnes had been criticized by OCC; this thwarted any potential beneficial effect of the allegedly remedial measure. Other OCC-directed actions included the conduct of regular employee-management meetings within the Controller's Office. Although all of the meetings were held, employees did not report any improvement in the Controller's Office culture following the OCC Report and Barnes's departure. The OCC also mandated that Spencer attend a workplace-environment seminar.(2166) Spencer did not attend this seminar prior to her departure from Fannie Mae, apparently due to scheduling issues. The Decision concluded that "[t]he Reporting Environment in the Controller's Office is unacceptable."(2167) According to Thomas E. Donilon, this conclusion caused several executives to whom the report was distributed, including Donilon, Howard, and possibly Spencer, to complain that the Decision failed to provide sufficient information to allow the reader to understand the basis for OCC's conclusions. Howard, in particular, voiced concern that the Decision was too vague for him to know how to remedy the problems it described. These concerns were conveyed to Remy and House. House construed the concerns as management's request for a supplemental or amended decision, but she determined that amending a report that already had been signed and issued would set a poor precedent, and she declined to do so. No one made any further requests, and the subject was dropped. VIII. FINDINGS REGARDING THE RESULTS OF THE OCJ INVESTIGATION On January 12, 2004, after a four-month investigation, OCJ issued its report relating to Barnes's allegations of discrimination.(2168) This report concluded that, - ---------- (2164) OCC Part B at FMSE 24566. (2165) Id. (2166) Id. at 24566. (2167) Id. at 24565. (2168) See generally OCJ Report. Although Barnes left Fannie Mae in the middle of the investigation, as discussed below, OCJ decided that its investigation into his discrimination complaints should continue to determine whether there were any institutional problems within the Controller's Office that needed to be addressed. This is why the final decision post-dates Barnes's departure from Fannie Mae. 566
although there was no evidence of race or gender discrimination, Controller's Office management had communicated its personnel decisions to Barnes poorly (or, at times, not at all), leading him to misunderstand the basis for those decisions.(2169) As a result of its investigation, OCJ directed, inter alia, that Spencer limit the ability of those in Pennewell's line of authority to hire their own subordinates and that Pennewell direct Juliane to improve his communication skills.(2170) IX. FINDINGS REGARDING BARNES'S RESIGNATION AND SETTLEMENT OF THREATENED LITIGATION On October 16, 2003, shortly after the OCC issued Part B of its decision but before OCJ had completed its investigation, Barnes's attorney(2171) sent to Fannie Mae (care of Raines) a litigation demand letter, claiming that Fannie Mae had discriminated against Barnes based on age and race and that the Company had retaliated against Barnes as a whistleblower in violation of federal law.(2172) Barnes threatened to file a lawsuit unless Fannie Mae responded by October 27, 2003.(2173) Among his allegations, Barnes claimed that Fannie Mae had: (1) Discriminated against him by denying him a series of promotions for which he had applied or had indicated an interest; and (2) retaliated against him for sending an anonymous interoffice memorandum to Raines and Howard.(2174) The demand letter attached a copy of an anonymous memorandum that Barnes said he had sent to Raines and Howard in September 2002 and that detailed alleged accounting problems within the Controller's Office.(2175) Barnes's counsel further charged that Fannie Mae management - ---------- (2169) OCJ Report at FMSE-IR 508833-40. Portions of the report relating to Barnes's allegations of inappropriate physical contact by Spencer were redacted from versions delivered to Pennewell and Lewers. Id. at FMSE-IR 508831; see also Handwritten notes from A. Tanenbaum to E. Bailey, dated Dec. 15, 2004, FMSE-IR 359602-14 (attaching unredacted version of report). (2170) OCJ Report at FMSE-IR 508828-29. (2171) Barnes stated that he had been seeking to retain counsel since the spring of 2003. From his comments during his interview, we surmise that Barnes must have hired counsel during the summer of 2003, although Barnes would not confirm that information. (2172) Letter from Elaine D. Kaplan to F. Raines, dated Oct. 16, 2003, FMSE 24667-84. (2173) Id. at FMSE 24683. (2174) Id. at FMSE 24667-84. (2175) Id.; Mem. from Finance Division Manager to Office of the Chairman, F. Raines, and T. Howard, dated Sept. 23, 2002, FMSE 22882-84. 567
had treated Barnes in a hostile manner since that time (and thus that Fannie Mae must have known that Barnes had sent the memorandum).(2176) Because the anonymous memorandum appended to the demand letter did not appear on correspondence logs maintained in the Office of the Chairman, Fannie Mae's litigation group, under the direction of Remy, launched an investigation into whether Raines or Howard actually had received this anonymous letter. Remy initially stated that Fannie Mae had located forensic computer evidence indicating that the letter had been drafted well after its alleged date, evidencing that it was not created and sent as Barnes claimed. Remy later withdrew from this position, acknowledging that there was no specific evidence of Barnes creating the letter after the date listed. Tom Foster, the head of Fannie Mae's Information Security Management ("ISM") team that conducted the forensic search, confirmed that the letter was not found on Barnes's work computer.(2177) Nonetheless, Fannie Mae attorneys and executives believed that Barnes did not send the letter as he claimed because it was not found in the correspondence logs of the Office of the Chairman.(2178) Fannie Mae hired outside counsel to investigate and report on the substantive allegations contained in the demand letter. Outside counsel conducted numerous interviews and advised Fannie Mae on the risks and prospects of litigation with Barnes. Based on all the information gathered relating to the allegations contained in the demand letter, Fannie Mae concluded that it was in the Company's interest to settle Barnes's threatened litigation.(2179) During the course of our investigation, we had access to and reviewed privileged material and attorney work-product related to Fannie Mae's decision to settle the threatened litigation. We also interviewed the attorneys who participated in making this decision. Based on all of the information available, we found no evidence that the Company settled the anticipated litigation to prevent Barnes from pressing his allegations of accounting impropriety. Indeed, because the settlement specifically required Barnes to cooperate with any internal, OFHEO, or other regulatory investigation into Fannie Mae, - ---------- (2176) Letter from E. Kaplan to F. Raines, dated Oct. 16, 2003, at FMSE 24675. (2177) At our request, Huron conducted its own forensic analysis of Barnes's hard drive and confirmed ISM's conclusion. (2178) Blickstein, Vice President and Assistant to the Chairman and CEO, stated that such a letter would have been considered "nuclear" and that Fannie Mae's Office of the Chairman would have responded to it immediately. (2179) The parties executed a settlement agreement on November 3, 2003. Settlement Agreement, dated Nov. 3, 2003, FMSE-IR 8207-15. 568
he was (and continues to be) contractually obligated to discuss his concerns with all relevant investigators.(2180) X. FINDINGS REGARDING KPMG'S REVIEW AND FINAL REPORT In January 2004, Argires of KPMG informed Rajappa that Fannie Mae needed to perform additional work to investigate Barnes's environmental allegations; we were told that KPMG asked for this work so it could sign-off on Fannie Mae's year-end financial statements. Argires stated that, because Barnes had alleged that it was difficult for employees to raise issues within the Controller's Office, KPMG's forensic group recommended that the Company perform additional procedures, including examining employees' e-mails to determine if other employees in that group had raised issues internally that had not been addressed properly. In connection with this request, Rajappa and Eilers examined approximately 3,500 e-mails identified by keyword searches on the computers of sixty individuals.(2181) Internal Audit, with assistance from Pamela Verick-Stone of KPMG, crafted the search terms. Rajappa stated that he kept KPMG informed of the process and provided KPMG with a CD containing all of the reviewed e-mails.(2182) After reviewing the e-mails, Rajappa and Eilers concluded that there had been no stifling of discussion within the Controller's Office. In fact, they found that the accounting issues that had been raised had been fully disseminated and discussed. KPMG also determined that Internal Audit and OCJ were "adequately independent" and, therefore, KPMG could rely on their investigation. KPMG did not make any follow-up requests after the e-mail review was complete. As part of the same KPMG-requested review, Fannie Mae imaged certain employees' computer hard drives.(2183) House and Tanenbaum then reviewed the files on these hard drives. Like Rajappa and Eilers, House and Tanenbaum found no evidence of complaints having been raised but not answered. Rajappa then conveyed the results of both reviews to KPMG. Based on this information, KPMG concluded that Barnes's - ---------- (2180) Id. at FMSE-IR 8207-08 ("The Employee agrees that he will fully cooperate with any investigation conducted by Fannie Mae or by any federal, state, or local government authority relating to Fannie Mae."); see id. at FMSE-IR 8212. (2181) Barnes's computer was not examined at this time because it had been searched previously in connection with the decision to settle his discrimination charges. (2182) Rajappa does not know what KPMG did with this CD. (2183) These employees were Spencer, Pennewell, Lewers, Juliane, and Barnes. 569
allegations would not prevent it from signing off on Fannie Mae's 2003 financial statements.(2184) XI. FINDINGS REGARDING BARNES'S TESTIMONY TO OFHEO AND CONGRESS On September 1, 2004, Barnes testified before OFHEO, and, on October 6, 2004, he provided sworn written testimony to the House Subcommittee on Capital Markets, Insurance, and Government Sponsored Enterprises.(2185) In both sets of testimony, Barnes asserted that Fannie Mae's senior management (up to and including Raines) had engaged in inappropriate earnings management and intentional accounting impropriety.(2186) He also alleged that he had mentioned these concerns repeatedly to Controller's Office management and had been punished as a result.(2187) He further claimed that other Fannie Mae personnel who had raised concerns internally had been demoted, reassigned, or fired.(2188) Barnes did not make these allegations during the investigations conducted while he was at Fannie Mae. Therefore, Fannie Mae did not have the opportunity to investigate these claims earlier. - ---------- (2184) Mem. to Administration File from M. Serock, Steve Galotta, and R. Forster, dated Jan. 21, 2004. (2185) Written Testimony of Roger Barnes, Former Manager of Financial Accounting, Deferred Assets in Fannie Mae's Controller Division, dated Oct. 6, 2004 (hereinafter Barnes Congressional Testimony), available at http://financialservices.house.gov/media/pdf/100604rb.pdf; Tr. of Sept. 1, 2004 OFHEO Interview with R. Barnes (hereinafter "Barnes OFHEO Testimony"). (2186) See, e.g., Barnes Congressional Testimony at 11-12; Barnes OFHEO Testimony at 81:21-83:7. (2187) See, e.g., id. at 2, 8-9, 13; Barnes OFHEO Testimony 184:8-12 ("The top-rated manager can't get promoted, but he is the only one asking questions. Something is wrong."). (2188) Id. at 2. 570
CHAPTER X: MANAGEMENT'S CONDUCT DURING OFHEO'S SPECIAL EXAMINATION I. INTRODUCTION This Chapter reviews Fannie Mae's conduct in the OFHEO examination that commenced in late 2003, up through the issuance of the OFHEO Report in September 2004. In particular, we focused on the adequacy of the Company's production of documents in response to OFHEO requests, and on the conduct of the Legal Department and its advisors during the examination, including the accuracy of the information they provided to the Board. II. BACKGROUND A. FREDDIE MAC'S ANNOUNCEMENT On January 22, 2003, Freddie Mac announced that its outside auditor would delay certification of Freddie Mac's year-end 2002 financial statements and that it expected to restate its financial statements for 2001, 2002, and possibly 2000.(2189) Freddie Mac announced the results of its restatement on November 21, 2003.(2190) In December 2003, OFHEO released its "Report of the Special Examination of Freddie Mac" (the "Freddie Report").(2191) The Freddie Report found, among other things, that Freddie Mac "cast aside accounting rules, internal controls, disclosure standards, and the public trust in the pursuit of steady earnings growth."(2192) B. OFHEO's Announcement to Examine Fannie Mae In July 2003, OFHEO announced that it would conduct a special examination of Fannie Mae's accounting policies, internal controls, and financial reporting (the "Special Examination"). On July 17, 2003, Armando Falcon, Jr., then Director of OFHEO, testified before the United States Senate Committee on Banking, Housing, and Urban Affairs. Although the Director's testimony focused on the Freddie Mac restatement process, he added that OFHEO intended to conduct a review of Fannie Mae to "independently - ---------- (2189) Freddie Mac press release, dated Jan. 22, 2003, available at http://www.freddiemac.com/news/archives/investors/2003/4q02.html. (2190) Freddie Mac press release, dated Nov. 21, 2003, available at http://www.freddiemac.com/news/archives/investors/2003/restatement_ 112103.html. (2191) OFHEO Report of the Special Examination of Freddie Mac, dated Dec. 2003, available at http://www.ofheo.gov/media/pdf/specialreport122003.pdf. (2192) Id. at i. 571
evaluate the accounting policies at Fannie and examine whether their implementation is resulting in a high level of conformance to GAAP."(2193) The next day, on July 18, 2003, OFHEO General Counsel Alfred M. Pollard sent a letter to Thomas E. Donilon, then Executive Vice President -- Law and Policy at Fannie Mae, stating that "the Director has indicated that he will undertake a review of Fannie Mae's accounting policies . . . in light of concerns surrounding accounting interpretations over the past few years relating to FAS 133, FAS 125, FIN 45 and other FASB releases and with accounting control functions."(2194) The letter directed the Company to "assure that document retention policies maintain the integrity of written, electronic and other information media in the accounting area and related internal control functions."(2195) As described below, OFHEO issued its first document request to Fannie Mae on November 21, 2003.(2196) III. FANNIE MAE'S RESPONSE TO OFHEO REQUESTS A. Document Preservation Efforts In response to OFHEO's July 2003 letter to Donilon, Fannie Mae's Legal Department compiled a list of fifty employees whom it believed were most likely to have information relevant to OFHEO's inquiry. The list included employees from the Office of the Chairman, Office of the Controller ("Controller's Office"), Office of Auditing ("Internal Audit"), Mortgage Portfolio, and the Legal Department. On July 22, 2003, Ann M. Kappler, Executive Vice President and General Counsel, sent a memorandum to these fifty individuals, asking them to "ensure the retention of all documents and material, including electronic files, e-mail and other media dating back to 1999 and going forward that concern Fannie Mae's accounting policies and related internal accounting control functions, including, specifically, documents or information involving interpretations relating to FAS 133, FAS 125, FIN 45 and any other FASB releases."(2197) - ---------- (2193) Oversight of Freddie Mac Accounting Practices: Hearing Before the S. Comm. on Banking, Housing, and Urban Affairs, 108th Cong. (2003) (statement of Honorable Armando Falcon, Jr., Director, OFHEO), available at http://www.ofheo.gov/News.asp?FormMode=Release&ID=84. (2194) Letter from Alfred M. Pollard to Thomas E. Donilon, dated July 18, 2003. (2195) Id. (2196) See infra text accompanying note 2204. (2197) Mem. from Ann M. Kappler to Distribution, dated July 22, 2003, FMSE-IR 376116-17, at FMSE-IR 376116. 572
As the Company identified other employees as potential custodians of documents in these areas, those employees were provided a copy of the retention memo.(2198) On August 19, 2003, the Company began archiving all incoming and outgoing e-mails for fifty-one employees to ensure the retention of any relevant e-mails. As company counsel at Wilmer explained in a memorandum submitted to Paul, Weiss on November 17, 2005 ("Document Collection Memorandum"),(2199) the Legal Department implemented the following protocol: "employees with operational and management responsibility in the areas identified by OFHEO were added to an `e-mail archive' that automatically captured and stored e-mails received by each employee at the time of receipt, and, on a nightly basis, copied e-mails sent by each employee."(2200) The Legal Department then reminded these employees, via an August 18, 2003 e-mail, that "[t]he process will not preserve electronic documents other than e-mail" and, therefore, they "should not delete any electronic documents or information stored on [their] m-drive, a shared drive, or any other location (including information on [their] calendar)."(2201) The Company added additional names to the list of relevant employees over the course of the Special Examination, ultimately archiving the e-mails of 121 employees.(2202) On October 13 and 14, 2003, after obtaining and reviewing a copy of OFHEO's Request for Proposals ("RFP") from accounting firms to assist OFHEO in conducting its Special Examination, the Legal Department sent e-mails to the original fifty employees to highlight other potential issues raised in the RFP. In addition to the areas specified in Kappler's July 22, 2003 memorandum, the October 14, 2003 e-mail identified for employees several other areas of potential interest to OFHEO, including: (1) policies and transactions relating to the timing of income recognition; (2) policies and transactions relating to derivative financial instruments and securitization; (3) "transactions that do not appear to have a valid business risk management purpose" and/or "that have been executed without proper authorization"; (4) management's process - ---------- (2198) Mem. from Jody A. Kris to Senator Warren B. Rudman, dated Nov. 17, 2005, at 3 (hereinafter "Document Collection Mem."). (2199) Id. This and other examples of documents relevant to the discussion in Chapter X can be found in the accompanying Appendix, at Tab G. (2200) Id. at 3. On account of a new e-mail system at the Company, as of late 2003/early 2004, all e-mails sent by employees were captured instantaneously by the e-mail archive. Id. n.2. (2201) E-mail from Donald M. Remy to A. Kappler, et al., dated Aug. 18, 2003, PW-PR 853405-06, at PW-PR 853405. (2202) Document Collection Mem. at 3. 573
for formulating sensitive accounting estimates; and (5) how estimates and judgments are estimated and monitored.(2203) On November 21, 2003, OFHEO issued its first request to Fannie Mae for hardcopy and electronic documents, including e-mails. The request applied to all Fannie Mae employees and covered materials from January 1999 to the date of the letter, including: (1) copies of all communications to or from Fannie Mae employees, Board members, or agents concerning communications with the SEC; and (2) copies of all documents prepared by Fannie Mae employees, consultants, contractors, or Board members related to FAS 133.(2204) In response, on November 26, 2003, Fannie Mae's Legal Department sent an e-mail to the employees it had identified in July 2003, attaching OFHEO's request and explaining that someone from the Legal Department would contact them to collect all potentially responsive documents. The e-mail stated that employees should continue to retain all documents relevant to Kappler's July 22, 2003 memorandum.(2205) That same day, the Legal Department sent an e-mail to all Fannie Mae officers to inform them of OFHEO's request and direct them to review their files for any responsive documents.(2206) The Legal Department continued to send out additional reminder e-mails periodically.(2207) As part of its document preservation efforts, the Company restored certain employees' e-mails. With respect to those individuals identified by the Company as most likely to have relevant, accounting-related information, the Legal Department instructed - ---------- (2203) E-mail from A. Kappler to Franklin D. Raines and Timothy Howard, et al., dated Oct. 14, 2003, FMSE-IR 376118-19, at FMSE-IR 376118; see also Letter from A. Kappler to A. Pollard, dated Nov. 7, 2003, at 2; E-mail from Debbie D. Milton to Leanne G. Spencer, et al., dated Oct. 13, 2003, FMSE 490446-48. (2204) Letter from A. Falcon, Jr. to F. Raines, dated Nov. 21, 2003, FMSE-IR 376129-32, at FMSE-IR 376130-32. (2205) Id. at FMSE-IR 376129. (2206) E-mail from Jonathan Griffith to #OFHEO 5, dated Nov. 26, 2003, FMSE-IR 376125-28, at FMSE-IR 376125. (2207) See, e.g., E-mail from A. Kappler to all Senior Vice Presidents and Vice Presidents, dated Feb. 18, 2004, FMSE-IR 376135-47 at FMSE-IR 376135. On September 27, 2004, following the release of the OFHEO Report, the Legal Department sent a records retention memorandum to all employees in the Controller's Office, Internal Audit, Portfolio Management, Portfolio Strategy, Treasurer's Office, and others. FMSE-IR 376155-77, at FMSE-IR 376156-57. 574
the Company's information technology personnel to restore such individuals' e-mails directly from backup tapes in order to recreate their mailboxes as of July 19, 2003.(2208) Regarding electronic documents other than e-mail, however, the Company did not initially suspend its ordinary practice of recycling its system-wide disaster recovery backup tapes. We found no evidence that anyone made an affirmative decision not to suspend the disaster recovery backup tapes, and the Company's outside lawyers view this failure to secure the backup tapes as understandable, because attempting to recover responsive electronic documents in this manner would have been extremely costly and inefficient. Indeed, according to the Document Collection Memorandum, "[t]he disaster recovery tapes store information in a format not conducive to easy restoration or searches to identify the tiny fraction of data contained on them that would pertain to Fannie Mae's accounting systems and controls. Moreover, that fraction of data was replicated on Fannie's servers, and Fannie Mae had clearly instructed relevant personnel not to destroy such information."(2209) For a short time period in September 2004, and in response to civil law suits that were filed at the time, the Company did retain daily and weekly backup tapes for the disaster recovery systems.(2210) This practice, however, was discontinued after a few months due to practical constraints outlined above. In responding to OFHEO's initial requests for employee e-mails, Fannie Mae's Legal Department drew upon three additional sources to the restored e-mails previously mentioned.(2211) The first was the archived e-mails from select employees. The second source was a series of annual backup "snapshots" taken of the Company's entire computer system. These snapshots, which were made near the end of the years 2000 through 2003, included all of the e-mails residing on the system on the day that the backup was made. Due to the Company's pre-existing e-mail retention policy, however, these snapshots generally recorded only e-mails from the approximately four to six weeks prior to the date of the snapshot. The third source of recovered e-mail was the e-mail that individual users of the Fannie Mae system had archived independently by saving messages to storage folders on their local hard drives or on their network space. B. E-mail Review The Company had extensive communication with OFHEO about e-mail restoration, review, and production throughout the course of the Special Examination. Most of the communication focused on Fannie Mae's ability to respond in a timely fashion to what it considered to be OFHEO's extremely broad e-mail requests. For example, we understand that in early January 2004, in-house and Wilmer lawyers - ---------- (2208) Document Collection Mem. at 5. (2209) Id. at 5. (2210) Id. (2211) See supra text accompanying note 2208. 575
requested a meeting with OFHEO, at which they expressed the view that substantial time and resources would be required for the Company to respond adequately to OFHEO's November 2003 document request, particularly for conducting the responsiveness and privilege review of all recovered e-mails sent by relevant employees. The lawyers thus suggested utilizing search terms to narrow the pool of potentially responsive e-mails, but did not receive permission to do so. Over the next several months, OFHEO issued additional document requests to Fannie Mae, which included requests for e-mails. On June 28, 2004, OFHEO issued a subpoena duces tecum to Fannie Mae, seeking the production of e-mails from thirty-six employees or former employees.(2212) The subpoena directed that production begin on July 6, 2004 and be completed by July 26, 2004.(2213) At Fannie Mae's request, the Company's lawyers met with OFHEO representatives on July 7, 2004 to discuss the scope of the subpoena. At that meeting, the lawyers expressed the view that OFHEO's timeline for the production of all e-mails was not realistic on account of the volume of material that needed to be reviewed. At this meeting, OFHEO representatives told Fannie Mae that it would not allow the use of search terms in reviewing e-mails. In August 2004, Fannie Mae organized an "all hands" review of e-mail messages, in which attorneys from the Company's divisions throughout the country were brought in to review e-mails, along with dozens of attorneys from Wilmer. E&Y, the accounting firm retained by Wilmer, also assisted with the review under Wilmer's direction. In all, Fannie Mae enlisted over 100 individuals who were provided with certain guidelines for identifying responsive e-mails. In total, during the period from January 2004 up to the issuance of the OFHEO Report,(2214) the Company produced 965,565 pages of e-mail in response to OFHEO's initial requests. Following the release of the OFHEO Report, the Company produced an additional 1,146,565 pages of e-mail through November 2005.(2215) - ---------- (2212) Letter from Charlotte A. Reid to Jodie Kelley, with attached subpoena, dated June 28, 2004. The subpoena required production of each e-mail from the thirty-six individuals "relating to, referring to, reflecting, pertaining to, describing, evidencing, constituting, or commenting on, in whole or in part, with respect to Fannie Mae's accounting policies, procedures and practices, securities valuation, financial reporting, internal audit, internal controls, or related matters." Id. at 3. (2213) Id. at 4. (2214) See OFHEO Report. (2215) Mem. from J. Kris to Senator W. Rudman, dated Dec. 16, 2005, at 2 (hereinafter "Supplemental Document Collection Mem."). 576
C. Hardcopy Documents The Fannie Mae Legal Department maintained responsibility for hardcopy document collection in response to OFHEO's document requests.(2216) Within Fannie Mae's Legal Department, Jodie Kelley, Vice President and Deputy General Counsel, was primarily responsible for the Company's document collection efforts. Kelley reported to Donald M. Remy, Senior Vice President and Deputy General Counsel, who in turn reported to Kappler. Kelley stated that she delegated the document collection tasks to a team of junior attorneys who reported to her. Kelley (and later Remy, as well) met with the team responsible for collecting documents on a regular basis. According to the Document Collection Memorandum, the discussions at these meetings focused on the scope of OFHEO's requests, potential interviewees, document collection efforts to date, and transmission of collected documents to outside counsel for review. As additional requests were received by the Company in the summer of 2004, Wilmer attorneys began participating in these team meetings to help coordinate the Company's response. After Fannie Mae received OFHEO's first document request in November 2003, the junior attorneys, initially accompanied by attorneys from Wilmer, met with the employees identified by the Company as likely to have responsive documents. The attorneys showed the employees a copy of OFHEO's request and asked them to produce all responsive documents from the relevant time period, which was January 1, 1999 to the date of the request. Although the attorneys explained the type of documents that were covered by the request, the responsibility for collecting and identifying responsive materials was left to the employees. This approach applied to hardcopy and electronic documents other than e-mail. By contrast, all of the documents, network spaces, and hard drives of employees who received individual subpoenas were collected and reviewed by attorneys in the Legal Department. In its Document Collection Memorandum, Wilmer challenged the notion that employees were left with discretion to produce responsive documents, stating that "[e]mployees were not left to interpret the requests as they saw fit - they were provided with substantial, detailed guidance by the Legal Department members as to what OFHEO sought and strict instructions to comply fully with those demands."(2217) As detailed more fully in Section E below, notwithstanding the Legal Department's "detailed guidance," numerous documents that were responsive to OFHEO's requests were not produced until after the issuance of the OFHEO Report. - ---------- (2216) Wilmer was not responsible for collecting documents and was involved primarily with reviewing documents for responsiveness and privilege, and for the physical production. Wilmer apparently became more involved in the collection process when subpoenas were directed by OFHEO to individual employees. (2217) Document Collection Mem. at 11. 577
When employees provided documents to lawyers in the Legal Department, Fannie Mae sent those documents to Wilmer for responsiveness and initial privilege review. Fannie Mae's Legal Department then conducted a second-level privilege review and made the final decisions on whether to withhold documents. As for electronic documents, Fannie Mae created a server that allowed employees to download the electronic documents to a shared drive, such that Fannie Mae could eventually download the documents and send them to Wilmer for review and production. D. Interviews From November 2003 to September 2004, OFHEO requested interviews with fifty-seven current or former employees of Fannie Mae. And, after withdrawing certain requests, OFHEO ultimately met with forty-eight employees, some on multiple occasions.(2218) Those interviewed included Timothy Howard, Leanne G. Spencer, Janet L. Pennewell, Jonathan Boyles, Thomas A. Lawler, Jeffrey Juliane, Sam Rajappa, Rene LeRouzes, Roger Barnes, and Mary Lewers. Individuals from Wilmer and the Legal Department typically attended the interviews, as did the interviewee's personal counsel, if counsel had been retained. Also present were individuals from Deloitte & Touche ("Deloitte"), which OFHEO had selected in February 2004 to provide accounting advice in connection with the Special Examination. OFHEO continued to conduct interviews of Fannie Mae employees following the release of its Report. E. Findings Regarding the Company's Response to OFHEO Requests From the period November 2003 through the issuance of the OFHEO Report in September 2004, OFHEO issued to Fannie Mae 340 individual requests on twenty-four occasions. In response, the Company produced to OFHEO approximately 412,670 pages of hardcopy and electronic documents exclusive of e-mail. In contrast, after OFHEO issued its report in September, from October 2004 to November 2005, OFHEO issued 151 individual requests on thirty-eight occasions, and the Company produced in response an additional 744,081 pages of hardcopy and electronic documents exclusive of e-mail.(2219) We have reviewed the pages produced after October 2004, and have determined that a significant number of documents were, in fact, responsive to OFHEO's earlier requests. The failure to produce these documents before October 2004 resulted from a number of factors, including the Company's initial failure to approach the task of collecting documents in a sufficiently comprehensive and methodical manner and the Company's attorneys' unduly narrow construction of a number of the document request. - ---------- (2218) Id. at 1-2. (2219) Supplemental Document Collection Mem. at 2. Wilmer provided these page numbers, which, for purposes of this Report, we have assumed to be accurate. Paul, Weiss did not independently verify through a re-count. 578
Company employees and lawyers were not given clear or standardized instructions to help them identify relevant documents. For example, certain employees recalled being told that they were not required to look for documents in their files if they did not create the document, but merely were copied on the document. The junior attorneys involved in the document collection also did not recall receiving any specific instructions on how to collect responsive documents; and Kelley, who was in charge of the overall production, did not regularly attend document collection interviews. The closest to a routine process that we discovered was a template created by a junior lawyer. The template provided a checklist of items to cover in the interviews, including the interviewees' supervisor and direct reports, OFHEO's individual document requests, other employees who may have relevant knowledge of documents, and whether the interviewee stored any documents outside of the office, including at home.(2220) This junior lawyer initially prepared the template for his own use, so that he could have a list of relevant questions for the document collection interviews he was conducting, and he later shared the template with other attorneys. However, neither this lawyer nor anyone else from the Legal Department conducting the document collection interviews could be certain that they used the template for many of the interviews. Moreover, even after reviewing the templates that have been produced to us, we still do not have a clear record of the instructions that Company employees received, the places that were searched, or the documents that were deemed non-responsive. In short, we are unable to form a view that the Legal Department conducted the document collection and review in a methodical and professional manner.(2221) In the Document Collection Memorandum, Wilmer, on behalf of the Company, notes that almost all of the relevant employees received the template: Fannie Mae . . . identified 93 employees as potential custodians of responsive documents for the November 21, 2003 request Of those 93, four were no longer employed by the company when the interviews took place. Of the remaining 89, there are only 11 employees for whom a template (or associated handwritten notes) has not been located - 3 of whom were lawyers involved in responding to the Special Examination The failure to retain (or even to complete) a template for 8 of the 93 eligible employees does not create a reasonable basis for inferring that the Legal Department failed to instruct those 8 employees to direct the lawyers to potentially responsive documents wherever they might be stored, in light of the - ---------- (2220) See Undated Template, FMSE-IR 365233-37. (2221) See, e.g., Undated Template for Brian Graham, FMSE-IR 430239-42; Undated Template for Thomas A. Lawler, FMSE-IR 430104-08. 579
otherwise consistent approach and shared understanding . . . . Likewise, the fact that some of the templates were not dated and did not have all fields populated with notations does not create a reasonable basis for inferring that the Legal Department somehow failed to search for documents in certain locations within the employee's possession [E]ach of the lawyers has affirmed the shared understanding that the lawyers were tasked with guiding the employees to identify all responsive documents in their custody, wherever located.(2222) Simply noting the existence of templates, however, does not demonstrate that the Legal Department did, in fact, undertake a comprehensive review for documents in response to OFHEO requests. The fact remains that many of the documents that were responsive to OFHEO's requests did not turn up until 2005, when the Company abandoned its initial approach of allowing employees to search their own files and adopted a new approach of having attorneys review documents in employees' offices for responsiveness. Finally, it appears that Wilmer, which was charged with the task of conducting responsiveness and privilege reviews of the documents collected by the Legal Department, construed OFHEO's requests very narrowly. For instance, it is our view that the presentation notes of Timothy Howard for the meeting of the Board of Directors in January 1999,(2223) where the 1998 catch-up was discussed, were responsive to OFHEO's Request 9, which asked for "notes, transcripts, minutes and draft minutes . . . recordings and/or summaries of all Board and Board Committee meetings from January 1999 to date taken or compiled by Fannie Mae employees."(2224) However, the Company did not produce this document until after the issuance of the OFHEO Report. Wilmer stated that it did not deem this document to be responsive to OFHEO's request because: [The document] does not purport to memorialize what was said at the meeting. Therefore, the determination that the - ---------- (2222) Document Collection Mem. at 9-10. (2223) Notes for Jan. 19, 1999 Meeting of the Board of Directors of Fannie Mae, FMSE-IR 182240-64. (2224) Letter from A. Falcon, Jr. to F. Raines, dated Nov. 21, 2003, FMSE-IR 376129-32, at FMSE-IR 376130-32. 580
document was not responsive to Request 9 was reasonable.(2225) Similarly, the Company did not produce a draft Quarterly Business Review ("QBR") summary, which reflected Howard's comments to the effect that the Company was trying to push income out of 2001 into 2003 and 2004.(2226) It is our view that this document was responsive to OFHEO Request 12, which called for "complete unedited agendas, minutes, presentations, reports or recordings . . . prepared with regard to or made in conjunction with the Quarterly Business Reviews from January 1999 to date."(2227) Wilmer offered the following rationale for the decision to designate the draft summary as non-responsive: Fannie Mae . . . understood Request 12 to seek production of a complete set of final documentation from the Quarterly Business Reviews over the five year period in question Lawyers at WCPHD and the Fannie Mae Legal Department reasonably determined that the document was not final, was not a "complete unedited report," and was therefore non-responsive.(2228) To be sure, lawyers often disagree in good faith regarding how to conduct document reviews and respond to document requests. We discuss the examples noted above to highlight what we consider to be unduly narrow constructions of OFHEO document requests and to note that this may have been one reason why certain relevant documents were not initially produced in response to OFHEO's requests. It is important to emphasize, however, that we view Wilmer's decision as those made by a zealous advocate in good faith. It is not our view, and we do not find, that Wilmer intended to hinder or obstruct OFHEO's examination. In the final analysis, the Company would have been better served by a less restrictive approach by its lawyers in handling OFHEO document requests. First and foremost, such an approach would have provided the attorneys with documents that would have enabled them to have a more complete understanding of the facts and be in a better position to recognize the potential flaws in the Company's accounting practices. This would have helped the lawyers to avoid taking factually unsupportable positions in defense of the Special Examination. Second, a fuller document production may have - ---------- (2225) Supplemental Document Collection Mem. at 5. (2226) QBR Summaries, dated May 2002, FMSE-IR 565521-27. (2227) Letter from A. Falcon, Jr. to F. Raines, dated Nov. 21, 2003, FMSE-IR 376129-32, at FMSE-IR 376132. (2228) Supplemental Document Collection Mem. at 5-6. 581
staunched the increasingly adversarial and hostile relationship between the Company and its principal regulator, OFHEO, during the Special Examination. IV. CONDUCT OF THE LEGAL DEPARTMENT AND ITS ADVISORS DURING THE SPECIAL EXAMINATION Our investigative scope also included a review of the information and advice that had been given to the Board during the Special Examination by third parties, including by Wilmer and E&Y. This scope was added after OFHEO raised questions about whether the Company's lawyers shielded certain documents from OFHEO through an overly aggressive use of privilege during the course of the Special Examination, and whether anyone in the Legal Department "lied to" or "misled" the Board in connection with the Special Examination. A. There is No Evidence That Privilege Assertions Were Made in Bad Faith to "Cloak" Relevant Documents from OFHEO With respect to the Company's assertion of privilege during the Special Examination, while instances where documents that had been deemed privileged were, on subsequent review, deemed not legitimately privileged, we did not find any evidence that lawyers did, or were instructed by anyone to, make aggressive privilege calls in order to shield relevant information from OFHEO. We found that lawyers - both in-house and outside - sought to make good faith determinations of privilege in the fast-moving examination, and found no instance where critical documents were placed on privilege logs without any basis for a claim of privilege simply to prevent their production to OFHEO. It is true that documents withheld from OFHEO during the Special Examination and placed on the privilege log prepared by outside counsel were subsequently produced by the Company to OFHEO in 2005. However, the decision to do so was made only after a lengthy discussion between the Board of Directors, Paul, Weiss, and OFHEO, which culminated in the implementation of a Memorandum of Understanding between the Board of Directors and OFHEO ("MOU"), which sought to balance OFHEO's need for access to all information at Fannie Mae to conduct its safety and soundness review, against Fannie Mae's legitimate interest in protecting confidential privileged information against disclosure to any third parties. Accordingly, we do not consider the fact that privileged documents were produced post-MOU to be evidence that the withholding of such information pre-MOU was in bad faith. B. The Board Was Advised By Wilmer and E&Y That the Company's Accounting Was Reasonable and Defensible As for the allegation that anyone in the Legal Department "lied to" or "misled" OFHEO, we again did not find any evidence that any lawyer - in-house or outside - knowingly provided false information to the Board during the course of the Special Examination. 582
Our interviews of Board members revealed that they did have the mistaken belief that E&Y was engaged in a "Substantive Review" of the Company's accounting, and that it would opine on the GAAP compliance of the Company's accounting practices. As it turned out, E&Y never completed the Substantive Review and never opined on the Company's accounting as in accordance with GAAP. While we did not find any evidence that anyone intentionally provided misinformation about the role of E&Y to the Board or to senior management, in our view, the failure to clarify E&Y's limited role from the outset - as well as during a joint meeting of the Audit and the Special Review Committees of the Board on July 19, 2004 ("July 2004 meeting") - created confusion for the Board, and led the Board to take significant but unwarranted comfort in the belief that the Company's accounting practices were supported by E&Y as in accordance with GAAP. There is no dispute that E&Y had been retained by outside counsel to the Company.(2229) However, the retainer agreement that specified E&Y's role was never provided to the Board. Further, based on information provided to the Board, E&Y was mistakenly perceived by several members of the Board and senior management to be performing a "validation" of the Company's accounting practices, rather than providing mere "litigation support" services to Wilmer.(2230) For example, in February 2004, the Audit Committee was informed that E&Y would conduct tests and procedures to validate the GAAP-compliance of the Company's critical accounting practices - including FAS 91 and FAS 133 - and also to determine whether such accounting practices were "best practices."(2231) While this "Substantive Review Plan" was described by E&Y in hindsight as a mere "issue-spotting" exercise consistent with its litigation support function,(2232) we conclude that the information about the Substantive Review Plan further assured the Board that management and its counsel had engaged a second "Big Four" firm to opine on the GAAP-compliance of the Company's accounting practices that were under scrutiny. E&Y emphatically disputed the suggestion that its role was to provide such a "second opinion" on the Company's accounting practices, stating that: Our role was to identify accounting issues and assist the lawyers in evaluating the risks associated with them. In - ---------- (2229) See Letter from Russell J. Bruemmer to Dale Kitchens, with attachments, dated Jan. 26, 2004, FMSE-IR 694352-60. (2230) Letter from Kevin M. Downey to Senator W. Rudman, with attachments, dated Oct. 20, 2005, at 25. (2231) See Substantive Review Plan, dated Feb. 4, 2004, FMSE-IR 694361-63. (2232) See Letter from J. Andrew Heaton to Robert P. Parker, dated Dec. 30, 2005, at 3 (hereinafter "December 2005 E&Y Letter"). 583
this role, we informed our client about the accounting issues we identified. However, we were not engaged to, and did not, conduct an independent review of the Company's accounting or determine whether Fannie Mae's accounting complied with GAAP - nor did we perform the work that would have been necessary for us to reach such a conclusion.(2233) Unfortunately, we saw no evidence that E&Y or Wilmer clearly explained this limitation on E&Y's work to the Board at any time in 2004. Thus, it appears entirely reasonable for certain Board members and senior management to have understood E&Y's comments about the Company's accounting practices as a "second opinion" that the accounting was consistent with GAAP.(2234) As noted above, it is E&Y's view that it never conducted procedures to validate the GAAP-compliance of the Company's accounting policies.(2235) However, in the July 2004 meeting, E&Y and Wilmer representatives both made statements concerning the work they had performed under the Substantive Review Plan that gave further comfort to the Board that the Company had a reasonable basis for its accounting decisions and policies, and that there was no reason to believe that a restatement would be required.(2236) Again, we saw no evidence that either E&Y or Wilmer ever explained to the Board members at this meeting that E&Y was not in a position to opine on the GAAP compliance of the Company's accounting practices. At the July 2004 meeting, Wilmer, E&Y and Boyles briefed the Board on the key accounting issues under review, FAS 91 and FAS 133.(2237) We carefully examined the information and advice that the Board received at the July 2004 meeting because it was the only time, prior to September 2004, when the Board was given a substantive presentation on the Company's accounting practices at issue in the Special - ---------- (2233) Id. at 2 (emphasis added). (2234) Indeed, notwithstanding E&Y's position, as late as December 23, 2005, Wilmer stated that "E&Y was expected to formulate and present its expert views about Fannie Mae's accounting and whether Fannie Mae's policies did not comply with GAAP in the areas subject to examination, and it did exactly that." Letter from William R. McLucas and Charles E. Davidow to Senator W. Rudman, dated Dec. 23, 2005, at 14 (hereinafter "December 23, 2005 Wilmer Letter"). (2235) December 2005 E&Y Letter at 2. (2236) See Minutes of the Audit Committee of the Board of Directors of Fannie Mae, dated July 19, 2004, FMSE 504860-69, at FMSE 504866-68 (hereinafter "July 2004 Audit Committee Minutes"); see also December 23, 2005 Wilmer Letter at 2-3. (2237) July 2004 Audit Committee Minutes at FMSE 504866-68. 584
Examination. The recollections of witnesses who were interviewed about the discussion at this meeting differed substantially. On balance, we conclude that the tone and the content of the briefing by Wilmer and E&Y were such that the briefing provided substantial comfort to the Board that the Company's accounting was reasonable and defensible, and that a restatement would not be necessary.(2238) E&Y contends that it did not advise the Board at this July 2004 meeting that in its view, the "[C]ompany's positions and rationales for its decisions and policies in the areas reviewed were justifiable and reasonable,"(2239) and that, in fact, the discussion at the meeting included the caution that OFHEO may "seek to cause a restatement."(2240) In support, E&Y points to the draft minutes of the July 2004 meeting, which reflected both of the above points,(2241) although neither point was included in the final version of the minutes.(2242) In our view, the most significant point the Board members heard at the July 2004 meeting from Wilmer and E&Y - which is unchallenged by E&Y - was that "neither [Wilmer] nor E&Y had reached a conclusion that a restatement in any period under review was required."(2243) - ---------- (2238) See id.; see also December 23, 2005 Wilmer Letter at 2-3 (summarizing Wilmer's presentation at the July 2004 meeting as containing the point that "Fannie Mae's FAS 91 and FAS 133 accounting was based on defensible business judgments, and neither E&Y nor WilmerHale had concluded that a restatement was necessary based on the work done so far"); December 2005 E&Y Letter at 4 ("It is certainly fair to say that the reports made at the July meeting were not wholly negative. For instance, it would not be surprising if Board members viewed as positive the fact that management had articulated practical reasons and business rationales for its policies"). (2239) July 2004 Audit Committee Minutes at FMSE 504866. (2240) December 2005 E&Y Letter at 3-4. (2241) See Draft excerpts of the Minutes of the Audit Committee of the Board of Directors of Fannie Mae, dated July 19, 2004, FMSE-IR 296116-17, at FMSE-IR 296116. (2242) December 2005 E&Y Letter at 4; see also July 2004 Audit Committee Minutes at FMSE 504866. (2243) July 2004 Audit Committee Minutes at FMSE 504866. We also note that the minutes of the July 2004 meeting were high-level summaries of the discussion that actually took place, and do not capture the tone and the detail of the discussions. Moreover, it is undisputed that E&Y reviewed and signed-off on the talking points that were to be used by Bruemmer at the July 2004 meeting, which contained the conclusion that the Company had put forward justifiable and reasonable reasons for its accounting decisions and its policies. See Executive Summary, dated July 13, 2004, FMSE-IR 694454-60, at FMSE-IR 694454; see also December 23, 2005 585
In addition to the minutes of the meeting, there are two documents that reflect portions of the relevant discussions that occurred at the meeting: the handwritten notes that were prepared by Dale Kitchens of E&Y in advance of the meeting, which Kitchens stated that he followed during the meeting;(2244) and the handwritten notes of David C. Britt, the engagement partner from KPMG who attended the meeting.(2245) Both of these documents reflect that the salient issues and questions raised by OFHEO with respect to both FAS 91 and FAS 133 were explained to the Board. For example, with respect to FAS 91, an E&Y accountant noted that OFHEO may allege that the one to two percent policy was "not GAAP" and was "[d]esigned to manage earnings," and that the 1998 audit difference was "material."(2246) However, the presentation by Wilmer and E&Y at the July 2004 meeting described any risks faced by the Company as mere "OFHEO arguments" or "disagreements," without any explanation of the facts that may form the basis for such OFHEO views. Many attendees of the meeting also recalled E&Y's Michael S. Joseph stating at the meeting, in substance, that had E&Y been Fannie Mae's auditor, they would have signed-off on the financial statements. While such a categorical statement was not supported in any document memorializing the meeting, the Britt notes do reflect the comment from E&Y's Joseph that "E&Y would not have objected to the [FAS 91] policy if they were the auditor;"(2247) and the Britt notes also reflect the comment attributed to Wilmer that there was "[n]o evidence of earnings management."(2248) Thus, notwithstanding the description of the anticipated OFHEO and Deloitte views, the tone of the E&Y and Wilmer presentation was such that Britt himself took substantial comfort and advised the Audit Committee after the presentation in an executive session that the FAS 91 policy was GAAP-compliant if objectively and consistently applied. It does not appear from these notes that E&Y's comments at the July 2004 meeting about the Company's accounting were accompanied by the clarification that E&Y was not undertaking a GAAP-review of the Company's accounting. Also, the - ---------- Wilmer Letter at 6. Accordingly, we do not place much significance on the edits that were made to the final minutes of the July 2004 meeting. (2244) See Undated handwritten notes of D. Kitchens, FMSE-IR 296111-13. (2245) See Undated handwritten notes of David C. Britt. (2246) Id. (2247) Id. Joseph has acknowledged that he did not fully comprehend the Company's FAS 91 policy. Joseph said that he had misunderstood how the Company's FAS 91 policy worked in practice when he spoke at the July 2004 meeting, and that he would not have agreed with the policy had he understood how the policy actually worked. (2248) Id. 586
Board apparently was not advised that E&Y was not performing the tasks reflected in the Substantive Review Plan. According to the final minutes of the July 2004 meeting, at the conclusion of the presentations by Wilmer and E&Y, the Board directed management to continue with the Substantive Review Plan, reflecting the Board members' belief that the Substantive Review Plan was on-going.(2249) Based upon this record, in our view, it was reasonable for the Board to believe that E&Y, in fact, was providing a "second opinion" that the Company's accounting was in accordance with GAAP.(2250) We also note that in July 2004, the Company possessed many documents that reflected management's intent to manage earnings. Nevertheless, the presentation at the July 2004 meeting did not inform the Board of any of these facts, and instead, provided comfort from the Company's outside lawyers and a Big Four accounting firm that the Company's accounting was reasonable and defensible against OFHEO arguments. We have not seen any evidence that would call into question the good faith of the Company's lawyers, or their experts, who were undoubtedly taking directions directly from management about the overall strategy to take with respect to OFHEO and how to defend the Company in the Special Examination. We conclude, however, that the lawyers neglected to provide the Board with sufficient information to allow the Board to weigh the risks and make an informed decision about the best course for the Company. While management may have been updated regularly by the lawyers on the relevant facts about the Company's accounting, the Board did not have regular access to such information. At the one meeting where the Board was given a substantive presentation, they were merely told of possible "OFHEO arguments" or "disagreements" accompanied by ready assurance that such practices were reasonable and defensible, and not provided any factual information that showed that OFHEO's "arguments" may be well founded. Finally, there plainly was an unhealthy relationship between Fannie Mae management and its primary regulator that had built up over the years, and by 2004, management's approach to OFHEO was extremely adversarial. Management and its counsel focused unduly on OFHEO's motives in conducting the Special Examination, which led them to dismiss the accounting issues as merely "OFHEO's arguments" and "disagreements." It turned out, of course, that management and the lawyers were wrong about the accounting issues raised by OFHEO. We conclude that the Company would have been better served if management and its lawyers had objectively informed the Board of all of the facts and the risks arising from those facts, and we recommend that the Company's lawyers make a concerted effort to do so going forward.(2251) - ---------- (2249) See July 2004 Audit Committee Minutes at FMSE 504868. (2250) See supra text accompanying note 2234. (2251) Wilmer's position is that the Board, at the July 2004 meeting, was adequately advised of all of the risks that faced the Company, and that the SEC's decision on FAS 133 was simply an event unforeseen by any of the Company's accountants. 587
- ---------- See December 23, 2005 Wilmer Letter at 17. While that may be, our view remains that going forward, the Board should be fully apprised of all facts relevant to significant regulatory risks in an objective manner. 588
CHAPTER XI: OTHER ALLEGATIONS PART A: ISSUES RAISED BY CURRENT OR FORMER FANNIE MAE EMPLOYEES I. INTRODUCTION On November 29, 2004, Ann M. Korologos, then the Presiding Director on Fannie Mae's Board of Directors, sent a "broadcast" message to all Fannie Mae employees asking them to contact us with any information or knowledge they might have about "any unusual or atypical transactions in the past five years."(2252) Korologos's message defined "unusual or atypical transactions" to include: - Transactions that appear to (or did) significantly accelerate or defer the pattern of income recognition. - Transactions that appear to (or do) enable Fannie Mae to change the character or classification of an asset or liability. - Transactions that do not appear to have been (or were not) executed at prevailing market prices. - Transactions that do not appear to (or did not) have a valid business or risk management purpose or that lack economic substance. - Transactions that appear to have been (or were) executed without appropriate authorization. - Transactions that do not appear to have been (or were not) accurately recorded in our financial statements. - Transactions that appear to (or did) violate generally accepted accounting principles or any Fannie Mae rule or practice. - Transactions that appear to have been (or were) executed so that bonus targets are achieved.(2253) In the course of our investigation, we received several responses to Korologos's message. In addition, we received an unsolicited communication from a mortgage lender regarding his experience with Fannie Mae. - ---------- (2252) E-mail from Ann M. Kappler to all Fannie Mae employees, Nov. 29, 2004 (attaching message from Ann M. Korologos, Presiding Director, Fannie Mae Board of Directors). (2253) Id. The e-mail was prepared in consultation with OFHEO and the SEC. 589
We developed the following process for addressing the issues raised in those responses: 1. When we received a response to the message, we asked the individual to provide any documentation he or she had relating to the issue raised. 2. Whether or not documentation was immediately forthcoming, Paul, Weiss and Huron jointly decided whether the issue raised presented a significant legal and/or accounting issue that was or should be included in the scope of our investigation. Our presumption was that we would treat such contacts as though they did raise a significant issue unless and until we had satisfied ourselves otherwise. If the contact raised an issue that was clearly outside the scope of our review, our practice was to call the employee to discuss his or her response, document the conversation, and close the matter. 3. If the issue warranted further inquiry, we arranged for a phone call or meeting with the respondent. At a minimum, we would contact the respondent by phone in order to: (i) understand that person's job and involvement with the issue raised; (ii) obtain more details on the issue, if possible; (iii) understand why the employee believed that the issue was problematic; and (iv) obtain the names of the individuals at Fannie Mae who are responsible for the matter, as well as other individuals at any level who might have more information. We would then decide whether to proceed with the matter further and, if so, what steps to take. In response to the broadcast message, ten Fannie Mae employees contacted us.(2254) Seven of these contacts took place within the first month following Korologos's message, while the remaining three contacts were made over the course of the following year. Of the ten contacts, one matter was closed with no further action; one matter was closed after interviewing the respondent; two contacts were invited, but so far have failed, to provide additional information concerning their allegations; one matter was the subject of significant, stand-alone review; and the remaining five were folded into other areas of our review. - ---------- (2254) Two employees contacted us to discuss the same issue. 590
A. HTM/AFS Redesignation On November 28, 2004, we were contacted by an employee with a question about Fannie Mae's decision to move a large amount of securities from Fannie Mae's held-to-maturity ("HTM") portfolio to its available-for-sale ("AFS") portfolio. This employee expressed no particular concerns about this transaction, but said that he was coming forward out of an abundance of caution. We interviewed the employee and discussed the issues he had raised. He confirmed that the events he had brought to our attention involved the redesignation of the securities from HTM to AFS in connection with OFHEO's promulgation of new minimum capital requirements. We concluded that this redesignation was appropriate. We took no further action on this matter.(2255) B. Non-Specific Anonymous Call On December 10, 2004, we received a call from a self-described Fannie Mae employee who declined to identify herself to us. The caller claimed to have serious allegations, but said that she would discuss them with us only on the condition that not doing so would result in legal liability for her. The caller asked for advice from us as to whether she had a legal obligation to come forward. We declined to offer legal advice on this question, particularly in the absence of any information regarding the nature of her allegations, but we encouraged the caller to provide the information to us anonymously and/or to discuss the matter with another attorney. The caller said that she would consider doing so, but to the best of our knowledge we have had no further contact with her. C. Home Solutions (Minority Lending Initiative) On December 10, 2004, we received a call from a Fannie Mae employee who raised concerns about a Fannie Mae loan program called Home Solutions. The respondent stated that Home Solutions was a loan program targeted at borrowers living in zip codes with a high percentage of minority residents. The respondent was troubled by the aggressive underwriting on the program and indicated that the mortgage product had been created to meet internal targets - including possibly bonus targets - rather than as an economically viable business line. The respondent provided us with documentation about the Home Solutions program, and we also submitted document requests to the Company about the program (also called the Minority Lending Initiative or "MLI"). That information, as well as information we derived independently, confirmed that the program was targeted - ---------- (2255) This specific allegation was separate from our consideration of the Company's redesignation of securities from HTM to AFS as part of a regular, intra-month process, which we discuss in Chapter VI, Part D. 591
to minority borrowers and that it was implemented as an effort to meet Fannie Mae's internal business goals for reaching minority homeowners. After reviewing the initial documentation, we identified certain potential accounting and related issues in connection with the MLI program. Accordingly, we added the program to the scope of our review of other minority lending programs. A discussion of our review of this program can be found in Chapter VI, Part N of this report. D. Contract and Procurement Systems ("CAPS") On December 14, 2004, we received a communication from a Fannie Mae employee that raised concerns about Fannie Mae's CAPS group - then in the Controller's Office - which oversees the Company's contract management. The letter questioned the Company's controls surrounding its contract management procedures, as well as the relationship between the CAPS group and the Legal Department. We interviewed several Fannie Mae employees on this issue. Although some of these employees denied any serious problems in the CAPS group, the majority of employees that we spoke with expressed the view that the CAPS group did have significant management and controls issues. In particular, they alleged that the CAPS group negotiated significant contracts without any legal supervision or guidance, the CAPS group did not have a means of tracking contracts, significant contracts (communications contracts, systems development contracts, etc.) were allowed to expire, and some contracts were negotiated with such poor terms that they placed Fannie Mae at risk. We also were told that there had been no internal audit of the CAPS group as of the time that our CAPS-related interviews took place. The documentation that we reviewed appeared to confirm that observation. These issues were outside of the scope of our review; in addition, we were aware that the organizational structure of the Controller's Office, including the CAPS group, was under review as part of the work conducted by Mercer Oliver Wyman. Under the circumstances, we brought this matter to the attention of Fannie Mae's Special Review Committee for further action by the Company. The Company engaged The Hackett Group to conduct an evaluation. The Hackett Group issued a report in August 2005 that substantially confirms certain of the issues brought to our attention and makes specific recommendations for improvement in the CAPS group organization.(2256) - ---------- (2256) The Hackett Group, Procurement Diagnostic Assessment & Strategic Transformation Plan: Final Report & Transformation Roadmap, with appendixes, dated Aug. 10, 2005, FMSE-IR 703065-214. We also received information that there were potential problems concerning Fannie Mae's accounting for certain equipment. The individual who made that allegation referred us to a former Company employee. We contacted the former employee, who said that his comment had been based on a misunderstanding. 592
E. Anticipatory Hedge Transaction On December 16, 2004, we received an e-mail contact from a Fannie Mae employee who raised concerns regarding an anticipatory hedge transaction in 1998, which the contact believed did not qualify for hedge accounting. According to the respondent, a substantial loss on the transaction was nonetheless deferred. We addressed anticipatory hedging transactions in several FAS 133-related interviews, and we identified the transaction that concerned the contact. Essentially, a sharp movement of interest rates in the days before the Company closed out the hedge of an anticipated debt issuance resulted in the loss on the derivative exceeding the gain on the hedged debt issuance by $24.7 million. Management made the decision to defer the entire loss, even though the applicable authoritative accounting literature required the excess to be expensed. KPMG identified the issue and included the expense of $24.7 million in their schedule of passed adjustments for the 1998 year-end.(2257) Given the date of the transaction and our conclusions regarding the Company's accounting for anticipatory hedge transactions, discussed in Chapter V, we did not deem it necessary to pursue a separate inquiry into this incident. F. Human Resources We were contacted by an employee who raised concerns about cultural issues in Fannie Mae's Human Resources department. We met with this respondent and discussed primarily the "tone at the top" of Fannie Mae's management, both in the Human Resources department and among the Company's top executives. We incorporate the issues that this respondent raised into our broader review of executive compensation issues. Our findings on this issue are described in Chapter VIII of this Report. G. Radian Credit Enhancement Product We were contacted about a credit enhancement arrangement from 2002 with Radian in the form of a pool mortgage insurance policy on loans Fannie Mae acquired in connection with a program entitled Expanded Approval with Timely Payments Reward Product ("EA/TPR"). The respondent questioned Fannie Mae's decision to pay $34.7 million for $38.6 million of insurance coverage when the Company had already purchased other credit enhancement for that product. The respondent also stated that there was a suggestion that the policy was undertaken to defer income from current periods to future periods. Upon reviewing documentation provided by the employee on the Radian arrangement, we added this topic to the scope of our review. A full description of our review of this topic can be found in Chapter VI, Part J. - ---------- (2257) KPMG workpapers. 593
H. Partnership Office We were contacted regarding one of Fannie Mae's Community Business Centers, formerly known as a Partnership Office. The respondent's concerns related primarily to another Company employee, who was associated with a prominent politician. According to the manager, before giving the employee a fairly severe evaluation, the manager received a call from a senior official in Human Resources who said that the Company keeps track of the ratings of the twenty to twenty-five employees "who can do political harm to the company." The Manager was not instructed to change his evaluation, but soon after he gave the employee his review, he received a call from the office of Thomas E. Donilon (although not from Donilon himself), asking about the review. The respondent also expressed concerns regarding the Company's management of his office. We did not believe that a particular incident involving an employee's evaluation was within the scope of our review, particularly given that there was no change in the employee's evaluation. We also did not believe that the Company's management of an individual Community Business Center fell within our scope. We did ask several interviewees whether they were familiar with a list of "politically sensitive" employees that the Company tracked, but none had heard of such a list. Likewise, we did not see any documentation that referred to such a list. I. Fannie Mae Loan Servicing Litigation We received a letter, dated November 8, 2005, that described a six-year-long litigation between Fannie Mae and a mortgage servicing company. The dispute, which dated back to 1991, involved a breach of claim that the mortgage servicing company brought against Fannie Mae alleging that the Company had breached its mortgage servicing contract. According to information provided with the letter, Fannie Mae prevailed in that lawsuit. The author of the letter stated that "[r]ecent revelations" about Fannie Mae were "a very interesting reminder of several issues raised in my dispute with and litigation against" the Company. He also stated that "much of what we were arguing at the time has now proven to be the very cause and/or effect of some of [Fannie Mae's] problems and abuses today." He suggested that he could "provide some valuable insight into [our] investigation." Although we believed that the issues raised in a fifteen-year-old lawsuit are beyond the scope of our investigation, we contacted the author of the letter and invited him to provide us with a written submission setting out any relevant information that he could share about Fannie Mae. We also invited him to send us any additional documentation that he had to support his claim that Fannie Mae's current problems were in some way related to the issues raised in the prior litigation. On February 15, 2006, the respondent sent us a supplemental letter, including a copy of a letter he had sent to the 594
Company's CEO the week before. Given the timing of the submission, and the fact that the matter has been brought to the Company's attention, we took no further action. PART B: FANNIE MAE'S EQUITY INVESTMENTS IN GULF BANK In January 2005, Stephen B. Ashley, acting Chairman of the Fannie Mae Board of Directors, and Armando Falcon, Jr., then Director of OFHEO, received an anonymous letter, dated January 4, 2005, from a self-described "Former Fannie Mae Employee."(2258) The letter concerned Fannie Mae's $800,000 equity investment in Miami-based Gulf Bank as part of Fannie Mae's Community Development Financial Institution ("CDFI") Initiative. According to the letter, Fannie Mae made the investment at a time that Gulf Bank was under a Cease and Desist Order issued by the Federal Reserve Bank, and that Fannie Mae wrote off the investment "days after it was made."(2259) The letter also asserted that the investment was prompted by a personal relationship between Robert J. Levin, at the time Fannie Mae's Executive Vice President--Housing and Community Development, who maintained ultimate responsibility for CDFI investments, and Salvador Bonilla-Mathe, the Chairman of Gulf Bank and a former Fannie Mae Board member by presidential appointment. Finally, the letter claimed that Levin received inappropriate gifts and other consideration (such as vacations) from Bonilla.(2260) I. INTRODUCTION The charge that Levin accepted inappropriate gifts or other consideration from Bonilla is groundless. The evidence we compiled showed that Levin developed an acquaintance with Bonilla during the Gulf Bank transaction, not before it. In the course of their association, Bonilla and Levin met and, on occasion, socialized in an appropriate fashion. Documents showed, and Levin acknowledged, that he received three gifts from Bonilla. Two of the gifts were tokens of minimal value, including a bag of coffee and a book about a charity with which Bonilla was associated. As to the third gift - a chess set - Levin sought the advice of the Company's General Counsel. Following her advice, - ---------- (2258) Letter from "A Former Fannie Mae Employee" to Stephen B. Ashley, dated Jan. 4, 2005, FM SRC 34840. Falcon was copied on the letter, as were Daniel H. Mudd, CEO of Fannie Mae, and Congressman Richard H. Baker. (2259) Id. (2260) Id. The letter also alleged a hostile work environment and made statements about compensation matters that were unrelated to the Gulf Bank transaction. Id. The latter were unrelated to the executive compensation issues within the scope of our investigation (see Chapter VIII) and we do not address those statements further. By the time we received the letter, the work environment matter had been investigated by the Company's Office of Corporate Justice ("OCJ"); to the extent that issue was within the scope of our review, we found no evidence in the course of our work that would support it. 595
Levin returned the chess set to Bonilla. There is no evidence that Levin accepted anything else of value from Bonilla. Similarly, we have found no evidence to suggest that Levin's actions in connection with the Gulf Bank matter were influenced either by a personal relationship with Bonilla or that Levin brought undue pressure on Fannie Mae employees, either individually or as part of the CDFI Investment Committee, to approve and consummate the Gulf Bank transaction. As a preliminary matter, Bonilla did not approach Levin initially; Bonilla approached Levin's then boss, Lawrence M. Small, who asked Levin to pursue the transaction. In addition, (1) at least some of the communication that referred to Levin's interest in this transaction came from others in the Company who also supported the transaction, and (2) Levin's immediate subordinates - including Senior Vice President--Housing and Community Development Barry Zigas - felt free to express reservations about the investment, but nonetheless gave it their support. Levin maintained that, apart from a few shared meals, his relationship with Bonilla was always strictly professional. Although Levin's level of involvement and interest in the transaction seems out of proportion to the size of the investment, Levin's personal interest in a CDFI transaction of this type was not unprecedented. His explanation - that he respected Bonilla and was keenly interested in helping to improve mortgage lending in the underserved Hispanic neighborhoods in Miami - is credible. Fannie Mae's equity investment in Gulf Bank does not appear to have been extraordinary in the context of Fannie Mae's CDFI program. The transaction was consistent with the purpose of the CDFI program, which was to support potential housing partners in underserved communities and also to generate other "franchise value" for Fannie Mae. With the possible exception of the final two weeks of 2000 when the transaction was presented to the CDFI Investment Committee, there is little evidence that the transaction was given preferential treatment. The CDFI Investment Committee's consideration of the transaction at that time does appear to have been rushed, and the circumstances do suggest that the CDFI staff and Committee were under some pressure to act. That pressure, at least in part, was the result of Bonilla's unrealistic and unrealized request that the transaction close by the end of 2000. And, notwithstanding that request, the transaction did not close for another nine months. Gulf Bank's disclosure to Fannie Mae in 2001 that the Federal Reserve found compliance violations at the bank was addressed by Levin to the Fannie Mae Legal Department and outside law firms. Levin also sought information on Gulf Bank and Bonilla from a private investigation firm. The available evidence is that no one inside or outside Fannie Mae raised concerns about the transaction moving forward. The Company's write-off of the Gulf Bank investment occurred about eighteen months after the investment closed and was part of a larger, portfolio-wide review of CDFI investments. We found nothing unusual about this practice, and the timing of the write-off was dictated by other factors. 596
II. BACKGROUND ON THE CDFI PROGRAM A. The CDFI Program The CDFI Initiative was established in 1994 as part of a broader effort, dubbed the National Partners Project, to provide support to institutions involved in community-based investments. Another goal of the program, at least at the outset, was to generate financial returns on the investments made in the CDFI institutions ("CDFIs").(2261) Fannie Mae's technical staff sometimes would help the institutions that received CDFI investments to establish mortgage lending operations. Although the recipients did not always offer mortgages prior to Fannie Mae's investment, encouraging the CDFIs to enter the mortgage business was one of the program's goals. Fannie Mae's CDFI investments usually took the form of certificates of deposits ("CDs") (generally $100,000, although sometimes larger amounts) or equity investments. The shares purchased in these institutions usually paid no dividends and, given the nature of the institutions involved, were often highly illiquid. Fannie Mae generally did not make equity investments in larger institutions that had ready access to other sources of capital. B. The "Typical" CDFI Transaction Life Cycle According to Susan Weintraub, the CDFI analyst who worked on the Gulf Bank transaction, the investment was "normal" compared to other potential investments she worked on while with the CDFI Initiative. She characterized the Gulf Bank transaction as an "average, typical transaction." Weintraub described a typical CDFI project as having several stages. At the outset, CDFI investment candidates often were identified by Fannie Mae's regional and partnership offices, which would forward these suggestions to the CDFI staff at Fannie Mae's headquarters; the CDFI staff would then undertake the necessary due diligence. As part of this due diligence process, information about the CDFI candidate would be distributed to other groups within Fannie Mae, such as Credit Policy, to evaluate the potential investment. Corporate Finance would assist by valuing the equity that Fannie Mae might acquire in the deal. The CDFI staff also would create an investment package for review by Fannie Mae's CDFI Investment Committee to use in determining whether to approve a candidate for investment. This Committee, whose members were appointed by Levin, - ---------- (2261) Letter from Susan M. Golden to Salvador Bonilla-Mathe, with attachments, dated Dec. 7, 1999, FMSE 481158-65, at FMSE 481158. 597
generally included Fannie Mae officers at the Senior Vice President level and above. According to Levin, he had overall oversight of the program, but at the time of the Gulf Bank transaction, Roger L. Williams was in charge of the program on a day-to-day basis. Levin stated that Williams, a Vice President, reported to Zigas. If the CDFI Investment Committee approved the proposed investment, then CDFI staff would prepare the relevant documents and assist in the closing process. One interviewee suggested that the rejection rate was approximately ten percent, the majority of which were rejected on the basis of curable defects. Ultimately, most (if not all) CDFI projects were approved. C. The Gulf Bank Transaction 1. Genesis of Fannie Mae's Relationship with Gulf Bank On May 20, 1999, Bonilla wrote to Small, then President and Chief Operating Officer at Fannie Mae. Bonilla told Small that he (Bonilla) had met the day before with Nitin J. Dave and Cecilia LaVilla-Travieso at Fannie Mae's partnership office in Miami to discuss the "Minority and Women Owned Lenders Initiative."(2262) Bonilla indicated that, due to its rapid growth, Gulf Bank soon would require "additional capital infusions," and he expressed interest in an investment from Fannie Mae. Bonilla also expressed a desire to become part of Fannie Mae's "deposit program," and he asked for Small's help in "securing the maximum available [deposit] as soon as possible."(2263) On May 25, 1999, Small replied to Bonilla, stating that he would alert his colleagues to Bonilla's "interest in a relationship."(2264) Small then forwarded both his letter and Bonilla's letter to Levin with a handwritten note stating that Bonilla "wants to get into our deposit program for minority banks."(2265) He asked Levin to follow-up on the request and reminded Levin that Bonilla was a former Board member. Levin apparently forwarded this message to a member of the CDFI staff, along with a handwritten note stating: "Here's the item I mentioned yesterday. Pls [sic] give me frequent status reports."(2266) During our interview with Levin, he recalled that Fannie Mae's partnership office was already working on the Gulf Bank transaction before he became - ---------- (2262) Letter from S. Bonilla to Lawrence M. Small, dated May 20, 1999, FMSE 482947. (2263) Id. (2264) Letter from L. Small to S. Bonilla, dated May 25, 1999, FMSE 482946. (2265) Handwritten mem. from L. Small to Robert J. Levin, with attachments, dated May 25, 1999, FMSE 482945-47, at FMSE 482945. (2266) Id. 598
involved; he did not remember that Bonilla had approached Small directly.(2267) Levin did not recall either receiving Small's note about Bonilla's letter or sending the note on to someone else, although the handwritten note did appear to be in his handwriting. Levin speculated that his request for "frequent" updates was probably because of Small's involvement. Within a few weeks of this exchange, Fannie Mae acquired a Gulf Bank CD in the amount of $100,000.(2268) There is no indication that the request for a "capital infusion" was discussed at this time. 2. The Request for an Equity Investment In late October 1999, Bonilla visited with Glenn Austin and Dave at Fannie Mae's Southeast Regional Office ("SERO") to discuss a possible equity infusion from Fannie Mae via the CDFI program and Bonilla's goal of expanding Gulf Bank's mortgage lending business, possibly through "acquisition of a Hispanic mortgage banking firm to give him a jump start."(2269) According to Dave's e-mail recounting the meeting, Bonilla later communicated with Small on this subject.(2270) Dave indicated that Small then wrote to Levin, "requesting him to stay in the loop on it," and that Dave had been asked to "spearhead" Bonilla's request. Dave asked Weintraub to send Bonilla a "CDFI letter."(2271) In an e-mail from around the same time, Dave noted that it would "behoove us to leverage our CDFI initiative to increase our business with the MWOLs, CDFIs, and HFAs, regionally and nationwide, and still extract the fundamental [f]ranchise value from such prudent investments."(2272) - ---------- (2267) Levin's recollection that the partnership office - the Southeast Regional Office - was the source of Bonilla's first contact with Fannie Mae is consistent with Weintraub's recollection that the Gulf Bank opportunity came from that office through the usual channels. (2268) E-mail (within chain) from Susan Weintraub to Nitin J. Dave, final e-mail within chain dated Nov. 22, 1999, FMSE 482900-03, at FMSE 482900 (reminding Dave that Fannie Mae had "purchased a one-year $100,000 CD at Gulf Bank on June 24, 1999"). (2269) Id. (e-mail within chain from N. Dave to S. Weintraub). (2270) Id. Specifically, Bonilla spoke to Small, and then followed up with a letter. Id. We have not seen a copy of the letter. (2271) Id. (2272) E-mail from N. Dave to Glenn Austin, et al., dated July 31, 1999, FMSE 482898-99, at FMSE 482899. 599
Levin wrote to Small on November 22, 1999, forwarding an e-mail from Austin to Dave, which Levin thought would provide Small with "a description of the follow up that's occurred so far as a result of Salvadore's [sic] letter to you."(2273) Levin stated that "Glenn and his team, and the CDFI team are on it."(2274) The forwarded e-mail from Austin stated that Austin had "spoken with Rob about giving Gulf Bank special attention when the equity investment application is complete."(2275) Although Fannie Mae sent a request to Gulf Bank for supporting documents in early December 1999,(2276) the bank's response was slow in coming. Over the course of the next twelve months, members of Fannie Mae's CDFI team exchanged several e-mails soliciting documentation from Gulf Bank needed to proceed with the transaction.(2277) Part of the delay is attributable to Gulf Bank's consideration of an acquisition target to expand into mortgage banking.(2278) We have not determined the cause of the remaining delay, but it appears to have been caused by Gulf Bank, rather than Fannie Mae.(2279) 3. The CDFI Investment Committee Despite the year-long delay in providing the requested information, by the middle of December 2000, Bonilla asked Fannie Mae to approve and close the transaction by the end of Gulf Bank's fiscal year on December 31. During the second - ---------- (2273) E-mail from R. Levin to L. Small, dated Nov. 22, 1999, FMSE 482902-03, at FMSE 482902. Several other Fannie Mae employees were copied on the e-mail. (2274) Id. (2275) E-mail from G. Austin to N. Dave, et al., dated Nov. 22, 1999, FMSE 482902-03, at FMSE 482902. (2276) See Letter from S. Golden to S. Bonilla, with attachments, dated Dec. 7, 1999, FMSE 481158-65. (2277) See, e.g., E-mail from S. Weintraub to N. Dave and S. Golden, dated Apr. 17, 2000, FMSE 486630; E-mail from N. Dave to G. Austin, et al., dated June 25, 2000, FMSE 486631; E-mail chain between Roger L. Williams, N. Dave, and S. Weintraub, dated Sept. 14, 2000, FMSE 486634; and E-mail chain between R. Williams, S. Weintraub, and David Elam, dated Dec. 5, 2000, FMSE 482586. (2278) E-mail from N. Dave to G. Austin, et al., dated June 25, 2000, FMSE 486631. We asked our interviewees whether Gulf Bank had ever actually purchased a mortgage lending institution with the funds it received from Fannie Mae. Dave replied that he and Bonilla went before Gulf Bank's Board of Directors to argue in favor of such a purchase, but the Board declined. (2279) See infra text accompanying note 2283. 600
week of December 2000, Dave sent an e-mail conveying Bonilla's request to Williams, Weintraub, and Tanya P. McInnis. Dave asked that the Gulf Bank transaction be considered at the next CDFI Investment Committee meeting.(2280) Williams replied that the next meeting would not take place until late January 2001. Williams also told Dave that, even with approval in December, the transaction probably would not close until after the first of the new year.(2281) Nonetheless, Williams did set the wheels in motion to obtain CDFI Investment Committee approval before the end of 2000. Around that same time, Dave provided some insight into Gulf Bank's urgency. Dave indicated that the bank's assets had "grown substantially this year (>25%)," and that, if the transaction did not close, the bank would "have to shrink its balance sheet (which Mr. Bonilla prefers not to do if at all possible) in order to stay at a well capitalized levels [sic] at the current [a]sset size."(2282) Handwritten notes from a teleconference that occurred on December 18, 2000, reflected a conversation among various Fannie Mae employees involved in the Gulf Bank transaction.(2283) The notes indicated that Jonathan Roman (who would be involved in valuing the shares that Fannie Mae was to acquire in the transaction) was in New York and would be on vacation for the rest of the year. They also noted that Levin "wants to get [the] deal done - whatever it takes." The notes listed the remaining documentation required to proceed with the planned funding as well as the names of the legal documentation contacts, which included Daniel F. Danello at Fannie Mae and Fannie Mae's outside attorney on the transaction. The notes also stated that Gulf Bank needed the equity investment for "capital reasons."(2284) A package of Gulf Bank-related materials was distributed to the CDFI Investment Committee on December 19, 2000. Williams noted in the cover memorandum: Rob Levin and SERO have asked us to accommodate the timeframe of a potentially strong affordable housing partner, and as such, we request from you an expedited - ---------- (2280) E-mail (within chain) from N. Dave to R. Williams, S. Weintraub, and Tanya P. McInnis, final e-mail within chain dated Dec. 11, 2000, FMSE 482586-89, at FMSE 482587. (2281) E-mail from R. Williams to N. Dave and T. McInnis, dated Dec. 11, 2000, FMSE 482586-89, at FMSE 482587. (2282) E-mail from N. Dave to R. Williams, dated Dec. 12, 2000, FMSE 486654. (2283) Handwritten teleconference notes, dated Dec. 18, 2000, FMSE 483243. The author of these notes is unknown. (2284) Id. 601
review of this transaction by this Thursday, December 21, 2000. We apologize for rushing this transaction and asking you to act on this transaction prior to our next scheduled Committee meeting at the end of January 2001.(2285) The contents of the package generated some concerns about the transaction, which were discussed during a December 21 teleconference. One participant, identified as "BZ" (probably Barry Zigas), expressed concern about an "apparent lack of care/urgency by GB" (probably Gulf Bank).(2286) Zigas also asked about Fannie Mae's "exit strategy" in the event that Gulf Bank did not carry out its plan of acquiring a mortgage company. Another participant, identified as "DE" (probably David Elam), stated that the investment "should stay for franchise reasons."(2287) Notwithstanding these concerns, the CDFI Investment Committee voted unanimously to approve the transaction.(2288) Levin apparently agreed that the transaction should go forward. He told Dave and Williams that he had spoken with Bonilla and told him that "we're giving his situation #1 priority."(2289) A few weeks later, Levin invited Bonilla to contact him if Bonilla sensed "that anything is slipping off course."(2290) During our interview, Levin stated that he viewed these messages as simply talking to a customer to keep him happy. Nonetheless, Levin said that he did view the deal as a priority. - ---------- (2285) Cover mem. from R. Williams to CDFI Investment Committee, dated Dec. 19, 2000, FMSE 481515-53, at FMSE 481515 (emphasis omitted). During our interview, Wilson pointed out that the cover memorandum also stated: "As we agreed at the last Committee meeting, if you have questions or comments on this transaction we will have a phone meeting at 10:00am on Thursday, December 21st." Wilson interpreted this to mean that the Committee had discussed Gulf Bank at a prior meeting, and this memorandum reflected the need to make a final decision on the investment after a two-day expedited review. Nonetheless, Wilson stated that it would be very unusual for the Committee to make a decision regarding an investment within such a short period of time. (2286) Handwritten teleconference notes, dated Dec. 21, 2000, FMSE 483242. The author of these notes is unknown. (2287) Id. (2288) See Signed vote pages, dated Dec. 19, 2000 through Dec. 21, 2000, FMSE 481507-14. (2289) E-mail from R. Levin to N. Dave and R. Williams, dated Dec. 21, 2000, FMSE 482675. (2290) Letter from R. Levin to S. Bonilla, dated Jan. 3, 2001, FMSE-IR 211440. 602
4. The Transaction Loses Momentum Although the Committee approved the transaction in December 2000, efforts to close the deal appear to have been hampered by two factors: a concern over the price that Fannie Mae would pay for Gulf Bank shares and an earthquake in El Salvador that affected Bonilla's family there. With respect to pricing, Dave told Williams and Weintraub that Fannie Mae would have to discuss with Bonilla the share price that had been approved by the Committee, which Dave said was "significantly less than . . . the valuation documentation furnished to us by Gulf Bank."(2291) Then on January 19, 2001, Levin wrote to Bonilla to thank him for Bonilla's gift of a book about Mercy Hospital and of a package of coffee.(2292) A handwritten postscript conveyed Levin's "sympathy to [Bonilla] over the earthquake" and Levin's concern for Bonilla's family.(2293) This reference appears to relate to a major earthquake in El Salvador on January 13, 2001. Later messages indicate that Bonilla's family sustained substantial economic losses as a result of the earthquake, and Bonilla made several trips to El Salvador during the first half of the year to assist there.(2294) Around this time, the first sign of Gulf Bank's later regulatory troubles appeared. Bonilla sent responses to a questionnaire concerning Gulf Bank's use of Fannie Mae's CDFI investment to Joan Wilson, a new Director in the CDFI program.(2295) In response to a particular question regarding the bank's regulatory standing, Bonilla wrote: As a result of a routine examination of the Bank by the Federal Reserve Bank ("FRB") subsequent to December 31, 2000, it was determined that the Bank may have failed to comply with certain rules and regulations relating to the filing of Currency Transaction Reports. The FRB has the ability to impose fines or sanctions against the Bank for this failure. Management believes the Bank has - ---------- (2291) E-mail from N. Dave to R. Williams and S. Weintraub, dated Jan. 2, 2001, FMSE 486664. (2292) Bonilla is listed as a member of the Mercy Foundation, which is associated with Mercy Hospital in Miami. See http://www.mercyfoundationmiami.org/members.html (last visited Feb. 17, 2006). (2293) Letter from R. Levin to S. Bonilla, dated Jan. 19, 2001, FMSE-IR 211442. (2294) See, e.g., E-mail from R. Williams to R. Levin and Barry Zigas, dated Mar. 1, 2001, FMSE 482671. (2295) See Letter from S. Bonilla to Joan Wilson, with attachment, dated Jan. 3, 2001, FMSE 483319-25. 603
taken necessary measures to correct this failure and to assure compliance with these rules and regulations.(2296) We have seen no indication that this initial disclosure of the regulatory examination of Gulf Bank raised concerns at Fannie Mae that the transaction should not proceed. On January 17, 2001, Williams told Levin that Gulf Bank had not yet filed with the State of Florida certain documents required to complete the transaction and that the "transaction appears to have moved from a fast track to a slow track."(2297) Williams stated that there had apparently been a "disconnect" between Bonilla and Gulf Bank's Chief Financial Officer about the urgency of the transaction.(2298) Williams also noted that the valuation of Gulf Bank's shares performed by a member of Roman's group at Fannie Mae suggested a lower price per share ($25-$30) than the independent valuation submitted by Gulf Bank ($37-$45). Williams said that he and Dave would speak to Roman to see how much "flexibility, if any, he may have in his appraisal."(2299) Handwritten notes from that day appear to reflect a discussion concerning the valuation of Gulf Bank's shares and indicate that, according to Roman, "20% discount immediately must be offset by franchise values."(2300) On January 19, 2001, Roman forwarded to Williams his valuation of the transaction.(2301) Roman noted in the cover e-mail that Fannie Mae's decision should be "based on the value of the relationship and other intangible attributes."(2302) Roman noted that Fannie Mae "would have a nega[t]ive mark-to-market were it booked today but still with a reasonable chance of a positive valuation in the future if they can deliver somewhere in line with their projections." He said that he found Gulf Bank's "price expensive but not unreasonably so even before factoring in intangibles."(2303) - ---------- (2296) Id. at 483323. (2297) E-mail from R. Williams to R. Levin, dated Jan. 17, 2001, FMSE 488157. (2298) Id. (2299) Id. (2300) Handwritten notes, dated Jan. 17, 2001, FMSE 488707. The author of these notes is unknown. (2301) See E-mail from J. Roman to R. Williams, with attached memorandum, dated Jan. 19, 2001, FMSE 482633-35. (2302) Id. at FMSE 482633. (2303) Id. Apparently, the previous valuation had been prepared by one of Roman's subordinates, who was reluctant to release the valuation until Roman had approved it. 604
According to Roman's valuation memorandum, "buying Gulf Bank stock at $37 or even $40 is not irresponsible; it is in the realm of a reasonable difference of opinion."(2304) But the memorandum pointed out that "were it marked-to-market today by methods consistent with [those used by Gulf Bank's valuation firm], a $750,000 investment could result in a markdown of around $150,000 on Fannie Mae's books." Roman concluded that "we need to make a comprehensive business call rather than a stock portfolio call. Is the value of the relationship we gain with Gulf Bank and its constituents worth the incremental value?"(2305) On March 1, 2001, Williams wrote to Levin and informed him that Bonilla wanted to close the transaction by the end of March, and that Williams did not think that would present a problem. Bonilla also had indicated that he wanted to "have an event with Governor Jeb Bush to announce the investment."(2306) After further delays caused by misunderstandings over the way that the transaction would be structured, Williams met with Bonilla, and the two of them discussed the possibility of increasing Fannie Mae's investment from $750,000 to $800,000. The bank's capital had grown sufficiently to allow Fannie Mae to "make this larger investment and still remain below the 10% threshold."(2307) On May 4, 2001, Williams sent a request to the members of the CDFI Investment Committee requesting approval of the increase based on the bank's growth as reflected in its unaudited financial statement; the Committee subsequently approved the requested increase.(2308) 5. Gulf Bank's Disclosure Regarding the FRB Examination On July 26, 2001, Williams wrote to Levin to inform him that Gulf Bank had received the required approvals for the transaction from the State of Florida. Williams also relayed Bonilla's hope that Bonilla could meet with Levin for a purpose unknown to Williams.(2309) - ---------- (2304) Id. at FMSE 482635. (2305) Id. (2306) E-mail from R. Williams to R. Levin and B. Zigas, dated Mar. 1, 2001, FMSE 482671. (2307) E-mail (within chain) from J. Wilson to R. Levin, final e-mail within chain dated Apr. 25, 2001, FMSE 487705. (2308) See Signed May 4, 2001 vote pages, dated May 18, 2001 through May 30, 2001, FMSE 481494-501. (2309) E-mail from R. Williams to R. Levin, et al., dated July 26, 2001, FMSE 488540. 605
On July 31, 2001, Levin met with Bonilla in Miami to discuss regulatory issues arising out of a recent examination by the FRB of Atlanta, and the potential effect of these issues on the transaction. It is unclear exactly how much information was disclosed to Levin at this meeting. Levin recalled at our interview that someone else at Fannie Mae - whose identity he could not remember - had told him that Gulf Bank's regulatory problems stemmed from its failure to fill out certain forms. Levin also recalled that Bonilla told him that the bank had hired a consultant to help deal with the problem and that the bank was working to resolve the situation. The day after Levin's meeting with Bonilla, Williams wrote to Wilson and others involved in the transaction to inform them that Levin had placed the transaction "on hold until further notice," and that they should direct further inquiries to Levin.(2310) 6. Fannie Mae's Evaluation of the Gulf Bank Disclosure Upon his return from Miami, Levin set the wheels in motion for Fannie Mae to evaluate the potential effect of Gulf Bank's regulatory problems. On August 1, 2001, Levin wrote to Bonilla and informed him that Levin had "spoken to our senior attorney here about the situation and told her that you will send me material for us to review."(2311) On August 6, 2001, Bonilla wrote a letter to Levin describing the FRB's investigation into whether Gulf Bank had failed to comply with the Bank Secrecy Act.(2312) According to the letter, the FRB had "discovered that our compliance officer failed to file currency transaction reports (CTRs) as required by the Bank Secrecy Act (BSA) and Federal regulations." The letter also noted that the bank had engaged BSA experts and that this team was "conducting a BSA due diligence investigation, rectifying any BSA deficiencies in the Bank and enhancing the Bank's BSA compliance system."(2313) The letter went on to state that, "[t]o date, neither state nor Federal regulators nor other government agencies have imposed any enforcement sanctions or penalties on the Bank," and that "[n]either the Bank nor any of its employees have been charged with any violations of state of federal laws or regulations."(2314) But, the letter advised, "generally the regulators eventually require banks to enter into agreements to ensure future compliance with the applicable rules and regulations," and that "under - ---------- (2310) E-mail from R. Williams to J. Wilson, et al., dated Aug. 1, 2001, FMSE 488561. (2311) E-mail from R. Levin to S. Bonilla, dated Aug. 1, 2001, FMSE 482926. (2312) See Letter from S. Bonilla to R. Levin, dated Aug. 6, 2001, FMSE 483776-91. (2313) Id. at 483777. (2314) Id. 606
applicable state and Federal laws and regulations the Bank may be subject to monetary fines and other sanctions."(2315) There is a handwritten note on the front cover of the letter, addressed to "Ann" (probably Fannie Mae's then General Counsel Ann M. Kappler) stating that "[t]his is the item about which Rob spoke to you last week."(2316) At Levin's request, Fannie Mae's outside counsel were consulted regarding the Gulf Bank situation. The Company asked one of its outside attorneys to look into whether there were any "outstanding regulatory enforcements [sic] actions against Gulf Bank."(2317) Donald M. Remy, a Deputy General Counsel at Fannie Mae, asked another outside attorney for her views on Fannie Mae's moving forward with the transaction in light of the FRB's investigation.(2318) Levin also engaged a private investigation firm, Decision Strategies, to conduct a background check on Bonilla and Charles H. Morley, one of the BSA experts that Gulf Bank had retained to assist it.(2319) The report on Bonilla did not reveal any significant information, and the research on Morley was called off after his professional background was confirmed. There is little documentation regarding the legal advice the Company received from outside counsel. During our interview, Levin could not recall the precise advice that he had received, but he did recall that the tenor of the advice was sufficiently comforting that he decided to move forward with the transaction.(2320) Levin said that he convened a meeting with Zigas, Dana Moore (Levin's Chief Risk Officer for the division, who has since resigned from Fannie Mae), Williams, and possibly others, to discuss the results and decide on a course of action. Levin said that they decided to proceed with the transaction, but that he does not clearly remember this meeting. Wilson recalled that Fannie Mae asked Gulf Bank to certify, prior to closing, that there had been no material change in the bank's position since the signing of - ---------- (2315) Id. at 483778. (2316) Id. at 483776. (2317) E-mail from Outside Counsel to R. Williams, S. Weintraub, and Daniel F. Danello, dated Aug. 8, 2001, FMSE 488568-69, at FMSE 488568. (2318) Letter from Donald M. Remy to Outside Counsel, dated Aug. 15, 2001, FMSE 489320. (2319) See Letter from Michael Dorrler to Gary Owens, dated Aug. 8, 2001, FMSE 489336-37. (2320) Levin's characterization of the tenor of the advice is consistent with information that we obtained in other interviews. 607
the stock purchase agreement. The bank's officers signed the certifications, which Wilson forwarded to Weintraub and Danello, among others. Levin then called Wilson and informed her that he did not see any reason why the equity investment should not proceed; she recalled his statement being something to the effect of "go ahead and do it." We asked several interviewees whether it was unusual for Fannie Mae to close on an investment in a bank that was currently being investigated by the FRB. Only one - Weintraub - stated that it was not unusual. She said that, during her years with the CDFI program, several banks had informed Fannie Mae that they were under investigation, but because they were not subject to any regulatory action, Fannie Mae had made the investment. She could not recall the banks involved in those transactions. 7. The Closing On August 22, 2001, Wilson wrote to Williams and others involved in the transaction to inform them that Levin had "asked that we move forward on funding the $800,000 equity investment in Gulf Bank." Wilson said that she would work with the team to finalize the investment.(2321) During our interview, Wilson did not recall the CDFI Investment Committee's being reconvened to address the change in circumstances of the transaction. She stated that the normal procedure would have been to call or e-mail the committee members with an update, rather than to reconvene them. Wilson noted that, in hindsight, it perhaps was strange that the Committee did not meet by phone at this time, but she could not recall any telephonic committee meetings regarding specific investments during her tenure at the CDFI program. We asked Levin and Wilson about two e-mails exchanged between them at the end of August 2001 concerning a transaction update to be sent to the CDFI Investment Committee members about the Gulf Bank transaction. Levin's message to Wilson instructs her to "[t]ake out the reference to the long term business relationship and the business and franchise value."(2322) Wilson's subsequent e-mail to the Committee contains no reference to these items, but does note that, after "consultations with our legal staff, Rob approved going ahead with the total investment."(2323) Levin said that the blurb in Wilson's e-mail was for purposes of the committee minutes, but that he could not recall the purpose of his own message to Wilson. He said that he had determined already that there was not as much franchise value associated with the transaction as he had first expected. Wilson indicated that these e-mails probably were related to the update she sent to the CDFI Investment Committee members after receiving Levin's feedback. We asked if this update e-mail indicated that the Committee was not consulted regarding Gulf - ---------- (2321) E-mail from J. Wilson to R. Williams, et al., dated Aug. 22, 2001, FMSE 488570. (2322) E-mail from R. Levin to J. Wilson, dated Aug. 28, 2001, FMSE 482927. (2323) E-mail (within chain) from J. Wilson, final e-mail within chain dated Aug. 28, 2001, FMSE 482928. 608
Bank's regulatory problems; Wilson confirmed that the Committee was informed of, but not consulted on, Levin's decision to proceed with the investment. The closing took place on August 29, 2001.(2324) The next day, Levin wrote to congratulate Bonilla on the closing of the transaction.(2325) 8. The Cease and Desist Order A little more than two months after the transaction closed, the FRB and the Florida Banking Department issued a Cease and Desist Order directed to Gulf Bank.(2326) Although the order dealt primarily with the BSA issues that Bonilla had identified to Levin prior to the closing, the order also dealt with a number of other issues that do not appear to have been covered by Bonilla's briefing or written correspondence.(2327) On December 3, 2001, Fannie Mae's outside counsel sent a memorandum to Williams, Wilson, and Danello outlining the order, its implications for Fannie Mae's transaction with Gulf Bank, and Fannie Mae's possible claims for breach of the representations and warranties in the stock purchase agreement.(2328) The memorandum noted that the order referred to a report of examination dated June 7, 2001, and stated that such a report should have provided Gulf Bank with a sufficiently detailed warning that Gulf Bank could have provided Fannie Mae with a fuller disclosure of the pending regulatory issues.(2329) Despite this initial inquiry into Fannie Mae's options in connection with the order, it does not appear that Fannie Mae pursued any possible remedies. Shortly after the first of the new year, Levin wrote to Bonilla, thanking him for the gift of a chess set, but informing Bonilla that he would be returning the set. Levin sent a blind copy of this letter to Kappler.(2330) Levin confirmed that he returned the set. - ---------- (2324) See Community Based Lending CDFIs Disbursement Form, dated Aug. 29, 2001, FMSE 481127-28. (2325) E-mail from R. Levin to S. Bonilla, dated Aug. 30, 2001, FMSE 482930. (2326) Order to Cease and Desist Issued Upon Consent, dated Nov. 28, 2001, FMSE 483380-99. (2327) See id. at FMSE 488840-42. (2328) See id. at FMSE 488839-46. (2329) Id. at FMSE 488843. (2329) id. at FMSE 488843. (2330) Letter from R. Levin to S. Bonilla, dated Jan. 3, 2002, FMSE-IR 211444. 609
9. Renewal of the CD and the Write-Off At the end of 2002, Fannie Mae's CD deposit at Gulf Bank came up for renewal.(2331) In the e-mail that accompanied formal investment renewal documentation that Dave and LaVilla-Travieso prepared, they stated that "the Chairman of the bank, Mr. Bonilla, is very well known to Frank and Rob."(2332) The formal investment renewal documentation also noted that Bonilla's "strong ties with Senator Graham, the House Members for Dade and Broward and Governor Bush will be invaluable in our needs for support at the Hill in DC." The summary concludes: "We strongly believe that Mr. Bonilla and Gulf Bank provide us substantial franchise value."(2333) During our interview, Levin stated that he believed this assessment of Bonilla's importance was overstated, but he said that he did not know the politicians in South Florida and could not assess Bonilla's importance in the community. We asked whether Levin thought this type of analysis was appropriate when considering Fannie Mae's investment decisions. Levin said that it was part of their business to know their partners and what kind of influence they had, but that such information did not make or break a given deal. Sometime in early 2003, Fannie Mae wrote off the entire Gulf Bank investment.(2334) This write-off was the result of a portfolio-wide review of CDFI investments carried out by a third party vendor, Sheshunoff. During our interview, Roman stated that CDFI Initiative staff began to mark its equity portfolio to market annually beginning in 1999. Roman noted that a mark-down might occur soon after an equity investment if that happened to be the time period when the Sheshunoff analysis - ---------- (2331) See E-mail chain between N. Dave and Cecilia LaVilla-Travieso, et al., dated Dec. 9, 2002, FMSE 484118; see also Letter from S. Bonilla to C. LaVilla-Travieso, dated Dec. 31, 2002 (stating a $100,000 CD would mature on February 8, 2003). (2332) E-mail from N. Dave to D. Elam, et al., dated Jan. 6, 2003, with attached documentation, FMSE 482642-48, at FMSE 482642. By this time, Franklin D. Raines had named Bonilla to a Fannie Mae advisory committee. According to Raines, he made those selections based on recommendations from other Company officers. (2333) Undated Recommendation Summary for CDFI Deposit Investment: Gulf Bank, FMSE 482643-45, at FMSE 482643 (part of documentation attached to Jan. 6, 2003 e-mail from Dave cited in note 2331). In January 2004, during a later renewal of the Gulf Bank CD, CDFI Initiative staffer Mark W. Lacey stated that Bonilla had "a strong history with Fannie Mae . . . and a personal relationship with Rob Levin." E-mail from M. Lacey to Frank Quesada, Greg Awad, et al., dated Jan. 5, 2004, FMSE 482663-64. (2334) See Equity Customer Evaluation Grid: Specific Customer Write Down, dated Spring 2003, FMSE 485757. 610
was conducted.(2335) Wilson confirmed that the Gulf Bank write-off occurred in the context of the Sheshunoff portfolio evaluation and that she did not remember any write-downs in 2003 happening outside of the regular process.(2336) 10. Subsequent Events In the Summer of 2003, Levin wrote to Bonilla to thank him for an enjoyable dinner among Levin, Bonilla, and their wives.(2337) The language of the letter implies that Levin's wife had not met Bonilla before that evening. There is no indication that the Levins and Bonillas otherwise socialized together. As Gulf Bank's financial condition deteriorated in the wake of the Cease and Desist Order, its Board of Directors began looking for a buyer for the bank. On February 17, 2004, Gulf Bank entered into a sale agreement with PanAmerican Bank.(2338) III. FINDINGS A. Due Diligence Neither the CDFI Initiative staff nor the Investment Committee members "cut corners" with regard to the due diligence portion of the process or the consideration of the investment. Only Dave suggested that the diligence on Gulf Bank may have been inadequate, but it was his view that the CDFI Investment Committee was "a sham" and performed only token reviews generally. We found no other information to support that allegation. B. Valuation We found a difference of opinion concerning Fannie Mae's valuation of Gulf Bank's stock. Levin stated that he did not know what the reference to "intangible attributes" in Roman's valuation e-mail meant. Levin said that this investment supported Fannie Mae's mission to increase Hispanic lending, and Gulf Bank was seen as important for gaining a potential foothold in that market. We asked Levin whether it would be unusual for Fannie Mae to have entered into a transaction knowing in advance that there - ---------- (2335) Roman stated that the valuation of debt investments was adjusted for changes in market interest rates. (2336) Dave claimed that such mark-downs were unusual and should not occur if a transaction undergoes proper due diligence. He stated that many CDFI investments were not properly scrutinized, as evidenced by the large number of investments in the CDFI portfolio that had mark-downs. (2337) Letter from R. Levin to S. Bonilla, dated July 10, 2003, FMSE-IR 211443. (2338) Letter from S. Bonilla to M. Lacey, dated Mar. 1, 2004, FMSE 483770-71, at FMSE 483770. 611
would be a $150,000 write down. Levin said that valuing these entities was difficult, but he agreed that it would be unusual to go forward anticipating a write down. Williams and Dave both expressed the view that Roman was instructed to value the stock higher than he thought it was worth. Roman's description of the valuation process, however, indicates that Levin did not bring pressure to bear on him, and we have no reason to doubt his first-hand account. Notably, Roman's valuation itself made clear that it took account of both commercial and "franchise" values.(2339) C. Policy Against Having Both CD and Equity Investment Of the six individuals that we interviewed, only Williams seemed to have an inkling that maintaining both a CD and an equity investment in the same institution violated Fannie Mae's CDFI policies. In an e-mail from early in the transaction, Williams noted that it was Fannie Mae's "preference" not to have an equity investment and a deposit in an institution at the same time.(2340) D. Allegations of Wrongdoing Against Rob Levin Although our interviewees suggested that Levin was more directly involved with this transaction than with most CDFI deals, we have found no evidence that Levin exerted inappropriate pressure on CDFI staff in connection with this transaction. Roman stated that Levin called him and asked him to have the valuation analysis performed so that a decision could be made regarding the investment. But Roman stated that he discussed the valuation with Williams several times, and that, although Williams expressed reservations about it, Williams never indicated that Levin was pushing to approve the investment. Roman did say, however, that Gulf Bank was the only transaction that he discussed with Levin directly. Williams described Levin's involvement in the Gulf Bank equity investment as "significant," in that he often checked in with the CDFI team regarding the progress of the investment. Williams characterized this level of involvement by Levin as unusual, and he stated that it created "pressure" to make the transaction happen. Dave felt that, after Levin's return from Miami in 2001, the normal due diligence for CDFI investments was bypassed because of "Rob's clout," which Levin exercised because he thought that Bonilla had a good reputation and would be helpful to Fannie Mae. We asked Williams and Dave whether Levin explicitly pressured people at Fannie Mae to approve the Gulf Bank investment to help Bonilla; both used the same phrase to describe the form that Levin's pressure took: "get it done." - ---------- (2339) See supra notes 2301-2305 and accompanying text. (2340) E-mail from R. Williams to D. Elam and S. Weintraub, dated Dec. 5, 2000, FMSE 482586. 612
Despite these suggestions of heightened involvement by Levin, we did not find any instances of inappropriate conduct on his part. Our interviewees generally disputed that Levin had directed anyone to do "whatever it takes" to make the transaction happen. Levin noted that it was not uncommon for people at Fannie Mae to invoke the names of senior officers in a bid to raise the priority level of an issue, and he thought it possible that this had happened in Gulf Bank's case. It appears, in fact, that Dave, in particular, invoked Levin's name on several occasions. On the whole, however, we have found no evidence that anyone at Fannie Mae, either at the direction of Levin or on their own initiative, knowingly entered into a transaction they believed was inappropriate or violated Fannie Mae's policies and procedures in order to conclude the Gulf Bank deal. E. Levin's Relationship with Salvador Bonilla We found evidence that Levin and Bonilla engaged in a few social activities together, such as the dinner also attended by both men's wives.(2341) We have found no credible evidence, however, that would call into question Levin's assertions that he and Bonilla never had a close personal relationship. At the outset of the transaction, Bonilla contacted Small, not Levin;(2342) Levin appears to have reacted to Small's expression of interest in the matter.(2343) Levin stated that, prior to the Gulf Bank transaction, their relationship was only in the context of Bonilla's Board membership. Levin said that he and Bonilla had not kept up their relationship after the transaction closed. Levin said that they had some meals together, but Levin could not recall who paid for these meals. Levin said that it could have been Bonilla at times, but that Levin may have done so too. With the exception of one message by Mark W. Lacey - which was sent well after the transaction had closed - all of the documents suggesting a close personal relationship between Bonilla and Levin (or between Bonilla and any other senior office at Fannie Mae) are traceable back to Dave. We believe that Dave may have had an exaggerated sense of the closeness of Levin's and Bonilla's relationship. F. Levin's Receipt of Gifts We found no credible evidence to support the allegation that Levin received inappropriate gifts from Bonilla that would have influenced Levin's professional judgment on the Gulf Bank matter. Apart from Levin himself, only Williams recalled any gifts: he had heard of Levin's receiving a chess set from Bonilla. - ---------- (2341) See supra note 2337 and accompanying text. (2342) See supra note 2262 and accompanying text. (2343) See supra note 2265 and accompanying text. 613
G. Franchise Value and the CDFI Program Generally 1. Franchise Value There is substantial evidence that the perceived "franchise value" of the Gulf Bank transaction was probably a significant factor in Fannie Mae's decision to move forward with the deal. "Franchise value," however, had several components. Franchise value included support for Fannie Mae's mission of promoting affordable housing across the country and the goal of achieving minority participation in the mortgage business. Franchise value also included any political value that such investments could bring, e.g., in the form of generating grass-roots support for Fannie Mae or creating ties with influential individuals. Franchise value considerations were an important characteristic of the CDFI program generally, and the Gulf Bank transaction does not appear to have been unique in this regard. Dave said that the Gulf Bank transaction was viewed as a "franchise investment," i.e., an accommodation made in expectation of future political support. Roman agreed that Bonilla's purported political connections entered into the decision-making process regarding the investment, but he noted that this was true of many CDFI investments.(2344) He acknowledged that there is a fair amount of discussion on this topic in the documents, but he said that Bonilla's political connections were not much of a motivator in the Gulf Bank case, and that Fannie Mae had never done transactions just for franchise value. Wilson explained that "franchise value" referred to actions that would support Fannie Mae's mission goals (such as affordable housing) and/or benefit the Company from a public- or government-relations standpoint. When asked whether the term also included political value, Wilson stated that mission goals and political goals generally were connected, although she did not recall if any politicians were personally involved in the Gulf Bank matter. 2. Mission Value We found evidence that Fannie Mae took into account the "mission value" of the Gulf Bank investment in helping the Company to meet its minority lending goals. Levin said that Gulf Bank did not have much of a mortgage origination function, so Gulf Bank was an attractive partner to increase lending penetration in a heavily Hispanic neighborhood. Levin said that the CDFI program was the best way to achieve greater penetration in South Florida and that Bonilla indicated that he wanted to start a real mortgage lending operation. - ---------- (2344) Interviewees were able to identify at least one other CDFI transaction that could be viewed as "political"; it involved a bank whose president had been an officer in an important mortgage banking organization. 614
Wilson recalled that Gulf Bank met the CDFI criteria because it was active in an underserved, low income area; was minority-owned; had established management; and had plans for how to utilize the CDFI funds in a mortgage-lending business.(2345) We asked if it was unusual for Fannie Mae to make an equity investment in a bank that did not already have any kind of mortgage business, and Wilson replied that the pre-existence of such a business was not a criterion for equity investments.(2346) Gulf Bank's failure to pursue such a business actively does not appear to have been a significant concern to Fannie Mae. 3. Economic Value Although there is some evidence that Fannie Mae considered the CDFI program to be, at least in part, a profit-oriented program, the CDFI staff and Levin freely acknowledged that such investments were inherently risky and frequently subject to write-offs. Indeed, it appears that many CDFI transactions, including the Gulf Bank deal, were undertaken with the expectation that they would be written off, at least in part. According to Levin, while there was some intent that the Company achieve minimum investment returns, there were different expectations for the various kinds of investments made under the CDFI program. Equity investments were risky by their nature, and expectations were accordingly lower. On debt investments, Levin said that the goal was to break even at the portfolio level, even if a particular investment did not perform well. Roman stated that, when the CDFI program was created, Fannie Mae intended to receive a return on its CDFI investments. That goal became difficult to meet, however, when nonprofit entities entered the community development field, as these entities did not expect any return on their investments. By 1999, Fannie Mae had determined that its best-case scenario would be to break even on its CDFI program. Accordingly, Roman viewed CDFI investments as budget items, and he analyzed them in terms of whether the potential benefits for Fannie Mae's relationships with investment recipients outweighed the budgetary cost. Weintraub stated that it would not be unusual for Fannie Mae to have an immediate negative mark-to-market on a CDFI investment, considering that the CDFI Initiative dealt with small, community banks with illiquid stock. In sum, although the Gulf Bank transaction might have been unusual in certain respects - including Levin's personal involvement - we found no evidence that - ---------- (2345) Wilson recalled that Dave was tasked to help Gulf Bank develop the technological capability to implement these plans. (2346) Wilson stated that nonprofit entities were subject to a different analysis, which focused on the entity's debt repayment history, information technology infrastructure, and management. 615
anyone involved in the transaction acted improperly. To the extent that those involved considered factors other than economic return when approving the investment, and were not overly concerned that the investment might be subject to a write-down, both were consistent with the nature of the CDFI program. 616
news release
|
 |
Consumer Resource Center: 1-800-732-6643
Contact:
|
Chuck Greener | Brian Faith | ||
| 202-752-2616 | 202-752-6720 | |||
Number:
|
3693 | |||
Date:
|
February 23, 2006 |
Chairman of the Fannie Mae Board of Directors
| • | “The Company, under the Board’s direction and OFHEO’s input, has undergone an extensive transformation in both personnel and structure since September 2004.” | ||
| • | “Since [September 2004] ... there has been a dramatic shift in the ‘tone at the top’ and the Company’s internal organization.” | ||
| • | “[N]o member of management who we found knowingly participated in improper conduct continues to be an employee of the Company.” | ||
| • | “[Paul, Weiss’s] suggestions for changes in corporate governance either have been implemented or are underway.” |
Page Two
| • | “With respect to the conduct of the Board prior to September 2004, we conclude that the Board endeavored to operate in a manner consistent with its fiduciary obligations and evolving corporate governance standards.” |
| • | “[M]anagement’s accounting practices in virtually all of the areas that [Paul, Weiss] reviewed were not consistent with [Generally Accepted Accounting Principles], and, in many instances, management was aware of the departures from GAAP.” | ||
| • | “[E]xcept for one instance in connection with the 1998 financial statements, [Paul, Weiss] did not find evidence supporting the conclusion that management’s departures from GAAP were motivated by a desire to maximize bonuses in a given period. [Paul, Weiss] did, however, find evidence amply supporting the conclusion that management’s adoption of certain accounting policies and financial reporting procedures was motivated by a desire to show stable earnings growth, achieve forecasted earnings, and avoid income statement volatility.” | ||
| • | “[E]mployees who occupied critical accounting, financial reporting and audit functions at the Company were either unqualified for their positions, did not understand their roles, or failed to carry out their roles properly.” | ||
| • | “[T]he information that management provided to the Board of Directors with respect to accounting, financial reporting, and internal audit issues generally was incomplete and, at times, misleading.” | ||
| • | “The Company’s accounting systems were grossly inadequate.” | ||
| • | “[T]he former CFO and ... the former Controller, were primarily responsible for adopting or implementing accounting practices that departed from GAAP, and that they put undue emphasis on avoiding earnings volatility and meeting EPS targets and growth expectations. As for the former Chairman and CEO ... we did not find that he knew that the Company’s accounting practices departed from GAAP in significant ways. We did find, however, that [the former Chairman and CEO] contributed to a culture that improperly stressed stable earnings growth and that, as Chairman and CEO of the Company from 1999 through 2004, he was ultimately responsible for the failures that occurred on his watch.” |
Page Three
| • | “[A]lthough management paid lip service to a culture of openness, intellectual honesty, and transparency, the actual corporate culture suffered from an attitude of arrogance (both internally and externally) and an absence of cross-enterprise teamwork (with a “siloing” of information), and discouraged dissenting views, criticism, and bad news.” |
| • | Separating the roles of CEO and Chairman | ||
| • | Hiring a new senior management team, including President and CEO Dan Mudd, as well as new Chief Financial Officer Bob Blakely, new Controller David Hisey, and new Chief Audit Executive Jean Hinrichs | ||
| • | Eliminating two management Board positions, retaining only one management position on the Board | ||
| • | Replacing the company’s outside auditor with Deloitte & Touche and directing the new auditor to conduct a comprehensive re-audit of the company’s processes, controls, and accounting policies and methodologies | ||
| • | Completely reorganizing the Finance function, including the appointment of a new Controller, a new Chief of Accounting Policy, and eight other new officers and the separation of the Controller’s office and the Portfolio business from the CFO function, with the head of the Portfolio business no longer reporting to the CFO |
Page Four
| • | Reorganizing and strengthening the Internal Audit function, led by a new Chief Audit Executive with a direct line of reporting to the Board’s Audit Committee | ||
| • | Establishing an ongoing Compliance Committee of the Board, led by Patrick Swygert, charged with ensuring Fannie Mae’s fulfillment of all legal and regulatory obligations, including the company’s agreements with its safety and soundness regulator, OFHEO | ||
| • | Establishing a new and separate Compliance, Ethics and Investigations function, led by new Chief Compliance Officer Bill Senhauser, reporting to the CEO and the Board’s Compliance Committee | ||
| • | Reorganizing the Risk function through the restructuring of the Risk Policy and Capital Committee of the Board and the establishment of a new Chief Risk Officer position | ||
| • | Reorganizing and decentralizing the company’s Law and Policy division so that each function – Communications, Government and Industry Relations, and Legal, including the new General Counsel – reports directly and separately to the CEO | ||
| • | Establishing regular interactions between the Chairman and other members of the Board and OFHEO; weekly meetings and status reporting between management and OFHEO | ||
| • | Adopting a new executive compensation structure with broader performance goals | ||
| • | Fulfilling the OFHEO-approved capital restoration plan, including meeting the regulator’s requirement of a 30 percent capital surplus. |
news release
|
|
Consumer Resource Center: 1-800-732-6643
Contact:
|
Chuck Greener | Brian Faith | ||
| 202-752-2616 | 202-752-6720 | |||
Number:
|
3692 | |||
Date:
|
February 23, 2006 |
President and CEO, Fannie Mae
Page Two