The Division of Finance has reviewed your revised draft audit report entitled Audit of FDIC’s Investment Policies (Assignment No. 2004-069) that was performed by PricewaterhouseCoopers LLP (PwC) under a contract with the Office of the Inspector General (OIG).

We were pleased that the draft report concluded that “the investment staff of the Division of Finance generally performed well in managing the FDIC’s investment portfolio in the context of the stated investment strategy, interest rate environment, and assessment of certain insured institutions undergoing financial stress.” We were also pleased to see that in the course of its review, PwC identified no instances of non-compliance with applicable laws and regulations relating to the Federal Deposit Insurance Act, the U.S. Treasury’s Operating Circular that governs the Government Account Series (GAS) Investment Program, and the FDIC’s Board- approved Corporate Investment Policy. Finally, we are gratified that PwC was able to replicate and match the portfolio total return figures generated by PORTIA, the FDIC’s portfolio information management system of record, thus verifying such calculations are being performed and reported on appropriately.

Unfortunately, there are aspects of this draft audit report we find troubling (even after the extensive revisions it has already undergone) and we continue to believe it is fundamentally flawed for two reasons. First, as we have previously indicated, the charge to PwC was overly narrow in that it overemphasized measuring how well the FDIC is achieving the “highest possible return” on the Bank Insurance Fund (BIF) and Savings Association Insurance Fund (SAIF) investment portfolios, given the numerous constraints within which these portfolios must be managed. The Corporate Investment Policy, which was carefully crafted and reviewed at the highest levels of the Corporation, explicitly and deliberately places the achievement of investment returns as a tertiary goal (maintaining liquidity^{[ 1 ]} and controlling deposit insurance fund balance volatility both take precedence). While PwC acknowledged there are other considerations besides achieving the highest possible return that the FDIC must take into account as it develops and implements its investment strategy, it does not appear this fact was given due consideration in formulating the findings and recommendations in the draft audit report.

Second, and more importantly, there seems to be a fundamental misunderstanding of how the unique circumstances in which the FDIC must operate color the development of our investment strategies given basic principles of sound fixed income investment management. The 46-month period that the audit covered was generally characterized by historically low U.S. Treasury market yields—indeed, in some instances during the period, the yields hit 40-year lows. Additionally, throughout much of the period, the market consensus was that such yields were “artificially” low—in part reflecting the Federal Reserve’s very accommodative monetary policy—and that yields were expected to rise rather substantially in the not-too-distant future. All else being equal, given such market expectations, prudent fixed income investment management would dictate shortening the duration of the portfolio in anticipation of investing in longer-term securities once the expected rise in yields has occurred (or at least partially occurred). Combine this basic principle with the fact that the FDIC faces a unique constraint that most other fixed income investors do not—it generally cannot trade out of a longer-term investment once it is made absent a compelling liquidity need—and one can readily understand why FDIC maintained a bias towards shorter-term investments throughout the period studied. For FDIC to have made substantial investments at the longer end of the yield curve, it would have had to believe that the then current (and historically low) U.S. Treasury yields were going to stay at that level for the entire six- to twelve-year holding period of the securities purchased. Such a belief would have been in direct contradiction with the market’s consensus forecasts throughout that period and we believe that few, if any, prudent investors would be willing to make such a wager.

We have carefully considered each of the draft report’s five recommendations. As noted below, we concur with one recommendation, and will give further consideration to a second recommendation as conditions merit. We do not concur with three recommendations as delineated, although we appreciate the spirit of the corporate governance recommendation. Before discussing each of the audit’s five recommendations and our reasons for concurrence, partial concurrence, or non-concurrence, we will first discuss in more detail our specific concerns with the audit’s investment performance finding.

The basic methodology employed during the course of the audit to assess the FDIC’s relative investment performance was to compare different hypothetical investment scenarios against the BIF portfolio’s actual total returns over the period from January 1, 2001, to October 31, 2004. The findings from the BIF portfolio analysis were assumed to apply equally well to the SAIF portfolio, which was not studied in this audit engagement.

In the 10-year Treasuries scenario (Scenario A-3), PwC concludes that the 46-month total return would have been 28.25% (3.65 percentage points higher than the BIF portfolio’s actual total return of 24.60%). Other reported hypothetical scenarios showed similar “superior” returns achieved over the BIF portfolio, albeit to a lesser degree. While we do not dispute the mathematical accuracy of the results of PwC’s modeling exercise, we take great exception to the implication in the report that FDIC’s management of the BIF investment portfolio (and by extension, the SAIF investment portfolio) significantly lagged those of other investment strategies that could have realistically been employed by the FDIC in managing these funds. Specific issues we have with the analysis include the following:

• Many of the PwC constructed investment scenarios represented unrealistic investment strategies that FDIC could not have pursued. For instance, many of the scenarios presented in the draft report ignored the FDIC’s long-run strategic objective (contained within our Corporate Investment Policy) of maintaining a laddered maturity structure.^{[ 2 ]} For example, we estimate that PwC’s Scenario A-3 resulted in a portfolio structure with approximately 65% to 75% of the portfolio being concentrated in securities maturing within seven to ten years as of October 31, 2004. For a number of reasons, maintaining such a maturity structure for a deposit insurance fund would not be prudent and more importantly, would be a clear violation of the maturity limits contained in the FDIC’s Board-approved investment policy. In short, the implication in the report that the FDIC could have earned an additional 365 basis points for the BIF portfolio during the period under study is misleading and inaccurate because it implies pursuit of a strategy we could not and would not pursue.
• The period under study did not represent a full interest rate cycle and therefore, a prudent investment professional can not reasonably assess the merits of a particular investment strategy over this short period. The 46-month period ending October 31, 2004, which was characterized by generally failing interest rates, biases the results of the PwC analysis in favor of pursuing the heavy long-end maturity sector bets like those being placed in Scenario A-3. A longer period of study that included a period of rising interest rates as well would clearly make for a much fairer comparison of the relative merits (and likely relative performance) of the BIF portfolio versus the hypothetical investment scenarios employed by PwC.
• The “passive” portfolios constructed by PwC in conducting its analysis were not truly Passive. PwC retained the very successfully performing BIF Treasury Inflation-Protected Securities (TIPS) purchases in each of its passive investment scenarios and thus unfairly inflated the performance results of many of these reported scenarios. If the hypothetical portfolios created were designed to be passive, we believe they should have been completely passive and not included these TIPS purchases.
• Lastly, the draft audit report does not recognize that during 2002 and 2003, the FDIC was faced with two possibilities that when viewed together would have made investing significant sums in 10-year Treasury securities imprudent at best.^{[ 3 ]} First, the BIF was faced with the real risk of failures of BIF-insured financial institutions. Second, and as elaborated upon earlier, the consensus interest rate forecasts throughout this two-year period were projecting yields on U.S. Treasury securities to rise. Both of these facts are clearly demonstrated in the data presented in Appendices 1 and 2.

  Appendix 1 is a table reflecting the FDIC’s estimate of BIF’s loss exposure that was reported in the footnotes to the BIF’s audited financial statements for the years 2000 through 2004. While the footnotes disclose FDIC’s best estimates of the potential losses that could be experienced in these troubled institutions, the amount of cash required for working capital purposes at the time of resolution would very likely be far greater. To raise funds for working capital purposes, the funds must sell all its investments before borrowing. Hence, all securities, including perhaps longer-term securities purchased when yields were lower, would have to be liquidated to generate cash in volumes sufficient to meet the working capital needs for these resolutions.

  Appendix 2 contains a table reflecting certain average overnight investment rates, 10-year Treasury security yields, and the spread between these two investment alternatives. This appendix also includes the Blue Chip Forecast yields for 10-year U.S. Treasury securities 3 to 6 months forward and 9 to 12 months forward from each calendar quarter end during the period covered by the audit.

  Given the forecasted increases in yields throughout the period and the very real possibility that the FDIC might have to redeem its longer-term securities to fund bank failures, a prudent investor would not have invested in 10-year U.S. Treasury securities that might have to be sold at losses (perhaps at substantial losses given the volatility in the Treasury markets) in as little as a few months after purchase. This is especially true if the losses would be significantly greater than the increase in yields offered by 10-year Treasury securities compared to overnight investments during their holding period.^{[ 4 ]}

Discussion of the Audit’s Five Recommendations

Having addressed our concerns about the draft audit report’s investment performance finding, we will next address each of the draft report’s five recommendations.

Recommendation #1

That the CFO’s office should initiate an internal review to consider.

• establishing a limit on the amount of cash-equivalents;
• formulating a proposal to permit adoption of investment guidelines that would allow investments in currently restricted instruments, especially including Social Security par value specials; and
• determining to what extent the FDTC is willing to assume some additional levels of risk for the potential benefit of maximizing returns given what PwC characterized as “perfect transactional liquidity.”

Management’s Response to Recommendation #1:

The FDIC has carefully considered each facet of this multi-part recommendation and has decided against initiating any additional internal reviews to address the above three items. The Investment Advisory Group (IAG) considered and rejected the idea of establishing a specific upper bound on the amount of cash-equivalents that either BIF or SAIF may hold at its most recent meeting on April 21, 2005. Placing such rigid limits on cash-equivalents would force DOF’s investment staff to make security purchases only on those days when the insurance funds receive cash. Such an “autopilot” approach would preclude “market timing” purchases (for example, buying Treasury securities when their prices dip) that we have used successfully in the past to enhance the long-term investment earnings of the insurance funds. Indeed, the IAG recently reviewed DOF staffs market timing performance for all of 2004 and the first quarter of 2005 and the relevant attribution analysis showed staff added value through its judicious use of market-timed purchases.

With respect to the second part of this three part recommendation—expanding the FDIC’s investment authority—on at least three separate occasions in the recent past, FDIC staff and management have carefully considered and consistently rejected the idea of seeking permission to expand the FDIC’s current investment authority. (An expansion of FDIC’s existing investment authority beyond the GAS program would require legislative action.)

The issue was studied in late 2000 under then FDIC Chairman Donna Tanoue; staff and the IAG’s recommendation at that time was to not pursue such authority. Subsequently, the FDIC commissioned an independent report, dated March 21, 2001, titled Reform of Deposit Insurance, which considered the issue of the FDIC’s pursuing a more flexible investment strategy. The report’s authors, Alan S. Blinder (former Vice Chairman of the Federal Reserve System’s Board of Governors) and Robert Westcott, while acknowledging that a more diversified portfolio might indeed deliver both lower risk and a higher return, nonetheless conclude that “[o]ur instincts are that the FDIC is such an important symbol of consumer protection that it should continue to invest in the safest, most liquid assets available—that is, U.S. Treasuries through the Bureau of Public Debt.” The results of both Blinder and Wescott’s and staffs studies were shared with congressional staff members.

Finally, in January 2003, CFO Steven App reviewed this matter with the FDIC’s Advisory Committee on Banking Policy, and collectively they came to the same conclusion as previous efforts that considered expanding the FDIC’s investment authority—the current authority is sufficient. Again, although staff and management recognized that expanding the FDIC’s investment authority could enhance portfolio returns, the potential for such returns would be accompanied by, among other things, higher portfolio market value volatility, increased transactions costs, additional liquidity concerns, and would introduce the potential for unintended market signaling. Accordingly, as the FDIC has reviewed this matter rather extensively over the recent past, each time concluding not to seek legislative changes to expand the FDIC’s investment authority, initiating an additional internal review is not warranted at this time.

With respect to the third part of the recommendation, we do not concur with the recommendation to seek authority to invest in par value specials. (These securities are currently only available to federal government trust funds and are not available to revolving funds, such as the BIF or the SAIF.) The U.S. Department of Treasury describes par value specials as follows:

“Par value specials are non-marketable securities that bear interest rates established according to statutory formulas, generally based on the average market yield on outstanding Treasury securities with specified maturities. They are purchased and redeemed at par or face value, so there are no premiums or discounts when purchased, and there are no gains or losses when sold. Thus, there is significantly less market risk involved as a result of interest rate shifts with investing in par value specials. Par value specials are only available to Investment Funds when the statute governing a particular fund specifically authorizes them to be issued to the fund and provides the interest rate formula for them.

The Social Security Trust Funds, Medicare Trust Funds, and a limited number of other Investment Funds are authorized by statute to hold par value specials. As of July31, 2000, 21 of the 214 Investment Funds are invested in par value specials. This represents $1.843 trillion of the $2.214 trillion invested on behalf of the Investment Funds. Par value specials eliminate capital losses when securities are redeemed after interest rates have risen, and also eliminate capital gains when securities are redeemed after interest rates have fallen. The legislative authority for an Investment Fund to use par value specials represents a policy decision that the Investment Fund should be protected from capital risk while still receiving a long-term interest rate that generally exceeds the short- term interest rate (because the yield curve is usually upward-sloping).”^{[ 5 ]}

Staff believes that to ask the U.S. Treasury and Congress to in effect subsidize the BIF and the SAIF by allowing the insurance funds to earn long-term U.S. Treasury security market yields without incurring the attendant price risk by allowing it to invest in par value special securities is inconsistent with principles of sound public policy.

Finally, FDIC will continue to evaluate risks and returns of alternative classes and maturities of available U.S. Treasury security investments. FDIC will also continue to invest in longer-term U.S. Treasury securities as market conditions and cash flow projections warrant. The FDIC will not automatically invest in longer-term U.S. Treasury securities when in its collective judgment the risk/reward tradeoff does not justify such investments.

Recommendation #2

To promote complete financial transparency in the reserve ratio and to more accurately communicate the FDIC’s true economic and financial condition as it relates to assessments, that the CFO consider including in financial reports a reserve ratio calculation to reflect a market- value approach for valuing the held-to-maturity portion of the investment portfolio. PwC notes that while the purpose would be to promote transparency, the resulting volatility could be a disruptive factor in FDIC’s operations and this would be better suited for a situation where the reserve ratio is determined based on a range which provides more flexibility in the decisions related to assessments. Given that current pending legislation considers such a range, PwC suggests its recommendation should be given further consideration in connection with passage of that legislation.

Management’s Response to Recommendation #2:

At this time, we are not going to include in our financial reports a reserve ratio calculation to reflect a market-value approach for valuing the held-to-maturity portion of the investment portfolio. We believe the financial reports that senior management currently delivers to the FDIC Board of Directors and the general public accurately and fully communicate the FDIC’s financial condition as it relates to insurance funds investments, assessments, and reserve ratios. The fund balances and related reserve ratios are determined and reported in accordance with generally accepted accounting principles (GAAP) and the FDIC has received clean audit reports from the Government Accountability Office (GAO) for the past 13 years on the financial statements of both the BIF and the SAIF. Our extensive footnote disclosures with respect to investments explicitly detail the book and market value of all categories of securities held in the BIF’s and SAIF’s investment portfolios.

However, if deposit insurance reform legislation is enacted that replaces the Designated Reserve Ratio with an acceptable range in which the reserve ratio could float, we concur that this recommendation would merit further consideration. FDIC staff and the IAG have already discussed this issue and have concluded that this matter would merit further review under such circumstances.

Recommendation #3

That the CFO replace the current goal regarding liquidity with a goal to manage the reported volatility of the value of the BIF and SAIF.

Management’s Response to Recommendation #3:

We concur with this recommendation with the understanding it is principally the Director, DOF, and not the CFO, who determines DOF’s annual performance goals. As currently written, the DOF Annual Performance Goal could be misleading, depending on how one interprets the term “liquidity.” A broad interpretation of the term liquidity is consistent with the draft report’s statement that “[a]s a participant in the GAS program, the FDIC’s investments enjoy perfect ‘transactional liquidity’ through the U.S. Department of Treasury.” However, as mentioned earlier, FDIC investment staff have historically used the term “liquidity,” in the context of the Corporate investment portfolios, to mean those securities that may be readily saleable without triggering potentially adverse accounting consequences. Accordingly, to better align the goal with the accompanying measurement guideline, we agree that the subject goal should be revised, to address minimizing or controlling portfolio price volatility, and Division of Finance staff will make such a revision.

Recommendation #4:

As a guideline for measuring investment returns, that the CFO consider comparing the FDIC’s short-term market forecast at the moment of purchases to subsequent movements in yields; and that the FDIC compares its returns with those of hypothetical portfolios based on rigid investment rules that pre-determine the Treasury’s investments, amounts, and timing of purchases.

Management’s Response to Recommendation #4:

We have carefully considered this recommendation, but do not concur with its recommended actions. For the past few years, extensive analyses have been performed annually as part of a formal, detailed investment performance attribution analysis. Moreover, as part of the quarterly investment review prepared for the IAG, staff discusses its investment security purchases in the context of current and anticipated Treasury market conditions. Finally, more ad-hoc comparisons are performed on a regular basis during each calendar quarter as staff prepares for and executes the actual investment security purchases (analyzing forward curves, tracking consensus Treasury yield forecasts as reported monthly by Bloomberg, etc.).

With respect to the recommendation that the FDIC compare its results with that of a hypothetical investment portfolio, we have considered and do not concur with that recommendation. Several years ago, staff conducted a study of how best to efficiently measure its relative investment performance. This culminated in the selection of the Merrill Lynch 1 - 10 Year Treasury Index (Index) as an investment performance benchmark. We discussed the choice of this benchmark with the PwC audit team extensively, and we continue to be comfortable that this is the best alternative for assessing our relative performance going forward.

We recognize that the Index is not an ideal yardstick for measuring the investment performance of the BIF and SAIF investment portfolios; however, we nonetheless believe that comparing portfolios’ investment returns to the Index provides useful information regarding the FDIC’s investment strategies and tactics, on both short-term (one-year) and longer-term horizons. It is true that the Index does not include TIPS, overnight investments, and 10- to 12-year Treasury security investments that the FDIC can make. However, to the extent that the FDLC strategically invests in such securities, a comparison to the Index will help determine whether such investments add value—either on a short-term or long term basis, as gleaned through the investment performance attribution analysis mentioned above. Being able to strategically invest in these securities (that are not part of a “plain vanilla” intermediate Treasury note and bond index) is important, especially in light of some of the more unique constraints placed on the FDIC, including: 1) the need to maintain, under certain economic or banking industry situations, a large dollar value of securities that can be sold without triggering a requirement to mark-to- market all securities; and 2) the restrictions in the GAS program against actively trading securities. In addition, comparing the performance of the BIF and SAIF’s conventional Treasury holdings to the index helps staff determine whether making particular maturity and duration decisions also add value, especially over the longer run.

Accordingly, we believe that also comparing the BIF and SAIF’s investment performance to hypothetical portfolios on an ongoing basis is not warranted, especially from a cost/benefit perspective. Moreover, implementing such a program is subject to potential problems and criticisms, such as selecting the “appropriate” hypothetical portfolio(s) to be used, determining the beginning and end periods employed for the horizon period, etc. As discussed extensively above, the results of comparison when utilizing such portfolios can be extremely sensitive to those variables.

Recommendation #5:

That the CFO retain an independent investment management firm to provide periodic (at least semi-annually) assessments of the DOF’s investment strategies and performance that are provided to the FDIC’s Board of Directors.

Management’s Response to Recommendation #5:

We fully support the concept of independent review and, as an organization, are always striving to achieve the highest standards of good corporate governance. However, engaging private investment management and consulting firms to semi-annually conduct these types of analyses would be a very expensive endeavor because these firms generally possess the skills requisite to examine complex portfolios that have trading positions with significant value at risk. Given the unique and limited parameters under which the FDIC manages its investment portfolios coupled with its wealth of internal investment expertise, its solid internal controls, and the strength of its management oversight program, FDIC management is confident it is meeting the highest standards of good corporate governance and so does not concur that the added considerable expense of semi-annual independent reviews will be beneficial to the Corporation or its stakeholders. However, given the materiality of the investment portfolios to the funds, we do believe that a limited, periodic review of the Corporate investment program every four to five years makes good sense, although the costs of such reviews should be closely managed given the limited risk inherent in this program.

We appreciate the opportunity to respond to your draft audit report. Please call Louis E. Wright at 202-416-6979 should you have any questions.

Attachment

Footnote 1:   In this context of the Corporate Investment Policy, liquidity is defined as investments that can be sold without triggering the requirement under SFAS No. 115 that all securities be subsequently market-to-market. Such investments include cash equivalents (overnight investments), available-for-sale securities, and held-to-maturity securities maturing within three months.

Footnote 2:   Interestingly, PwC states in its report that “with respect to maturity limits, we agree with the FDIC that it is prudent to maintain a laddered portfolio”.

Footnote 3:   In fact, it would have been a violation of the Department of Treasury’s rules for the GAS program. The Department of the Treasury Operating Circular states in Section 4060(f) that “A program agency for a government investment account shall, to the best of its ability, develop its investment strategy so to select investments having maturities suitable to the cash disbursement needs of the program being financed through the account. The principal amounts and maturities of investments should be selected to coincide approximately with the program agency’s disbursement estimates, so that the investments may be bought and held to their maturities.”

Footnote 4:   The calculus that the FDIC is constantly required to make is whether the extra spread earned during the holding period will exceed the potential capital loss incurred upon sale of the longer-term security purchased. In the steep yield curve environment that existed in 2002 and 2003, the spread was significant. However, the consensus for significant higher yields again was also quite pronounced, which implies the potential capital losses incurred upon a forced sale could have been significantly greater than the potential additional yield earned.

Footnote 5:   Treasury Responsibilities in Investment Fund Administration, Report for the Secretary of Treasury, U.S. Department of the Treasury, November 2000.

*   Contingent liability for banks that are likely to fail within one year of the reporting date, absent some favorable event such as obtaining additional capital or merging, when the liability becomes probable and reasonably estimable.

**   In addition to recorded contingent liabilities, the FDIC identifies additional risk in the financial services industry that could result in a material loss to the BIF should potentially vulnerable financial institutions ultimately fail. This risk is evidenced by the level of problem bank assets and the presence of various high-risk business models that are particularly vulnerable to adverse economic and market conditions. Due to the uncertainty surrounding future economic and market conditions, there are other banks for which the risk of failure is less certain, but still considered reasonably possible.