Corporation Comments for Audit Report No 06-013

This memorandum constitutes the management response from the Division of Insurance and Research (DIR) to the findings and recommendation in the draft Office of the Inspector General (OIG) audit entitled, FDIC Reserve Ratio and Assessment Determinations (Assignment No. 2005-032). The stated objective of the OIG audit was to determine whether the FDIC accurately determines the Bank Insurance Fund (BIF) and the Savings Association Insurance Fund (SAIF) reserve ratios and whether the Corporation has adequate internal controls in place to ensure that the FDIC accurately calculates, projects, and processes assessments for financial institutions. DIR’s response will address the OIG findings and one recommendation related to the calculation of the reserve ratio and assessment determinations and the FDIC’s process for estimating insured deposits. A separate FDIC management response from the Acting Deputy to the Acting Chairman will address the OIG findings and recommendations relating to Board member involvement in this issue. As discussed below, DIR agrees with the OIG recommendation and will implement it in a timely manner. We believe that adopting this recommendation will improve the Corporation’s ability to ensure the validity of key data, estimates and assumptions regarding the reserve ratio.

Beginning in 1989 with the passage of Financial Institutions Reform, Recovery, and Enforcement Act (FIRREA), the FDIC used a methodology for determining the reserve ratio that assumed 100 percent of Oakar deposits were insured. This methodology was used because conditions were such that typically close to 100 percent of deposits acquired by banks from thrifts in the early and mid 1990s were insured, especially in the case of deposits acquired from conservatorships managed by the Resolution Trust Corporation. In recent years, however, SAIF- member thrifts, like their BIF-member bank counterparts, began to rely more on uninsured deposit financing than in earlier years. In addition, beginning in the late 1990s but most noticeably in the past four years, the growth rates of aggregate domestic deposits and insured deposits began to diverge. These growth rates have an effect on the allocation of estimated insured deposits between BIF and SAIF in institutions with Oakar deposits. The developing trends in the composition and growth of deposits raised concerns among FDIC staff and management that the existing methodology for the allocation of insured Oakar deposits between the BIF and SAIF needed to be reexamined.

In 2004, the FDIC initiated a review to determine whether its existing methodology continued to result in allocations that were consistent with the deposit structure of SAIF-member institutions and appropriately represented the levels of insured and uninsured deposits that SAIF- members and BIF-members were actually bringing to Oakar transactions. The FDIC considered a number of possible alternative methodologies, including the assessment base allocation methodology noted in Tables 3 and 4 of the OIG draft report. While none of the methodologies provided a perfect solution for the allocation of estimated insured deposits between BIF and SAIF, FDIC staff ultimately recommended a methodology to the Chairman that it felt most appropriately represented the proportions of BIF and SAIF estimated insured deposits in Oakar transactions and recommended that the FDIC apply this new methodology prospectively. The new methodology became effective in the fourth quarter of 2004. The recent passage of deposit insurance reform legislation merges BIF and SAIF, which resolves the specific allocation issues regarding insured Oakar deposits that are the subject of the audit. Therefore, beginning in first quarter 2006, this allocation issue no longer exists because of the merger of the BIF and SAIF funds.

The OIG Report concluded that while the FDIC accurately calculated the BIF and SAIF reserve ratios and accurately calculated and collected assessments due from financial institutions, the internal controls over the reserve ratio and assessment determination processes need to be strengthened. DIR concurs with this finding and believes that formalizing procedures will strengthen our current verification procedures and ensure timely, comprehensive review of estimates and assumptions supporting deposit insurance calculations.

The Report describes key issues and events that contributed to the OIG’s recommendation related to the Corporation’s ability to ensure the validity of key data, estimates and assumptions regarding the reserve ratio. DIR concurs with the analysis and findings in the report as they are described below:

As discussed above, the FDIC adopted the change to the estimated insured deposit allocation method for Oakar institutions prospectively, beginning with the last quarter of 2004. The remainder of this section reviews the following aspects of that decision. First, there is no single correct method to estimate and allocate insured deposits in calculating the reserve ratios for each fund and there is no scientific way to determine at what precise point in time an existing method becomes insufficiently representative so as to require a change. Second, previous changes in estimating insured deposits have been adopted prospectively. Finally, in this situation, even if the decision had been to depart from past practice and apply the revised method retroactively, the FDIC could not legally impose assessments on banks for prior periods.

1. There is no single correct method to estimate and allocate insured deposits and there is no scientjfic way to determine at what precise point in time an existing method becomes insufficiently representative so as to require a change.

As a starting point, it is important to recognize that the distribution of insured deposits between BIF and SAIF at Oakar institutions in calculating the reserve ratios for each fund is an estimate, not a precise determination. In fact, the exercise of trying to allocate estimated insured deposits based on the actual past distribution of such deposits between the acquired thrift and original bank would be impracticable.^{[ 1 ]}

For this reason, it cannot be assumed that there is a single correct method to allocate estimated insured deposits at Oakar institutions for the reserve ratio. One can only look at industry trends and select an allocation method that is judged to be a reasonable reflection of such behavior at the time the method is adopted. In addition, given that trends can change over time, there is no scientifically reliable way to determine at what precise point in time an existing method for estimating insured deposits or allocating these deposits between the funds becomes insufficiently representative so as to require a change. Congress recognized this inherent uncertainty by using the term “estimated insured deposits.” For that reason, a prospective application of the new method seemed most appropriate in this case.

2. Previous changes in estimating insured deposits have been adopted prospectively.

For example, beginning with the June 30, 2000 reporting cycle, the FDIC changed its algorithm for computing estimated insured deposits. Prior to that date, the FDIC had estimated uninsured deposits for every institution and subtracted this estimate from total domestic deposits to estimate insured deposits. With the change, if a reporting institution provided a voluntary estimate of its uninsured deposits, this amount was used instead of the FDIC-calculated estimate. These voluntary estimates had been reported since 1993. This change in computation methodology resulted in a $47 billion increase in estimated insured deposits. The new method was not applied retroactively to earlier periods.

Additionally, effective with the March 31, 2002 reporting cycle, all insured institutions were required to provide their own estimates of uninsured deposits. Since that date, these estimates have been used in the computation of insured deposits for all institutions. This change resulted in an estimated $56 billion increase in estimated insured deposits. No adjustments were made to prior periods.

Most Oakar transactions through the mid-1990s were BIF-member purchases of SAIF- member thrifts, including many thrifts that were in Resolution Trust Corporation conservatorship. Almost all of the deposits of the acquired SAIF members were insured. In addition, until the end of 1996, the law also permitted acquiring banks to reduce by certain amounts the acquired deposits that SAIF could assess (and increase the amount that BIF could assess). The result was that in many transactions the amount of deposits attributed to SAIF under the method then in effect was less than the amount of insured deposits that thrifts brought to Oakar transactions. This provided further justification to treat acquired SAIF deposits as fully insured. Also, under the original allocation method, an Oakar institution’s insured SAIF deposits were assumed to grow at the same rate as its domestic deposits. This was consistent with growth trends observed in most years after the 1989 FIRREA legislation.

Gradually over time, thrifts have increased their reliance on uninsured deposits (though such reliance is still significantly less than that of banks). Thus, the assumptions underlying the prior method today conform less closely to actual transactions than previously. More importantly, industry domestic deposits have grown in recent years at faster rates than estimated insured deposits. Under the prior method, these diverging growth rates would, in the aggregate, raise SAIF’s share, and decrease BIF’s share, of estimated insured deposits held by Oakar institutions. However, domestic and insured deposit growth rates only began to diverge on a sustained basis in 2001. It would not have been possible to know until later that this divergence in growth rates would be persistent. Thus, the prior method reasonably reflected both actual Oakar transactions and deposit growth patterns until a few years prior to the adoption of the new method. Empirical evidence to justify a change in allocation method did not exist in 1997 or for several years thereafter and the application of the new methodology beginning 1997 would not be appropriate.

3. Even if the decision had been to depart from past practice and apply the revised method retroactively the FDIC could not legally impose assessments on banks for prior periods.

The $96 billion figure cited in the OIG report based on the DIR analysis does not represent a potential adjustment to prior allocations of estimated insured deposits between BIF and SAIF. Rather, the DIR analysis was a tool used by FDIC to determine how the proposed method might affect the allocation of estimated insured deposits in the future absent a merger of the deposit insurance funds. The analysis simulated the effect that the new method might have had over an historical period, between the start of 1997 and the third quarter of 2004. The analysis required staff to make several technical assumptions to handle data limitations. The
$96 billion figure represented an example of the potential magnitude of difference that might result from the application of the proposed method going forward.

As noted above, the OIG recalculated BIF and SAIF reserve ratios using DIR’s analysis and concluded that the new allocation method would have caused the BIF to drop below the 1.25 percent DRR for a
6-quarter period starting in late 2001 and in 2005. This could 1eave the impression that the FDIC, at the time it was considering revising the allocation method, might have faced a decision as to whether to charge banks to recoup assessments for the 6-quarter period. It is important to point out that the FDIC could not legally have imposed assessments on banks for prior assessment periods. Although the methodology became less representative over time, it remained within the broad parameters permitted under the statute. Because the reserve ratio remained above the DRR under a legally valid approach, the law precludes the charging of assessments. Finally, it should be noted that this issue is now moot with the merger of the insurance funds.

DIR concurs with this recommendation. We will develop policies and procedures that will require DIR managers to periodically identify, validate, and approve key assumptions and estimates that support the calculation of the reserve ratio. The policies and procedures will require analysis of industry-wide issues and trends that could affect current or future estimates, as well as other factors such as economic trends, changes in institution accounting practices, etc. The policies and procedures will also focus on issues that affect the integrity of the application data that support reserve ratio calculations. For example, DIR will periodically validate the integrity of bank data systems, and discuss and approve the process and assumptions used to generate the data.

The policies and procedures will include a schedule of periodic meeting to discuss concerns. Managers and staff will adequately document material discussions, decisions, testing, and validation results, including but not limited to, discussions of alternatives consider but not accepted, decisions to accept or change key assumptions and estimates, and the communication of recommendations and decisions to appropriate DIR management.

DATE:	April 10, 2006

TO:	Russell A. Rau
	Assistant Inspector General for Audits

FROM:	Arthur J. Murton
	Director

SUBJECT:	Draft Report Entitled, FDIC Reserve Ratio and Assessment Determinations (Assignment No. 2005-032)

FDIC Reserve Ratio and Assessment Determinations

CORPORATION COMMENTS