Home > News & Events > Speeches and Testimony




Speeches and Testimony

Testimony of
Donna Tanoue
Chairman
Federal Deposit Insurance Corporation
Merger of the Deposit Insurance Funds
Before The
Subcommittee On Financial Institutions And Consumer Credit
Of The
Committee On Banking And Financial Services
U.S. House Of Representatives
10:00 A.M., February 16, 2000
Room 2128 Rayburn House Office Building

Madam Chairwoman and members of the Committee, I appreciate this opportunity to testify on behalf of the Federal Deposit Insurance Corporation (FDIC) regarding the merger of the deposit insurance funds and related issues.

I want to emphasize four points in my comments this morning. First and foremost, the Congress should merge the deposit insurance funds now. A merger is unequivocally in the best interest of the American taxpayer and the timing could not be better. Second, while the FDIC has, in the past, partially offset assessments with assessment credits, it has never rebated funds from the deposit insurance funds and we would urge great caution before embarking upon such a path. This is especially true at this point in the economic cycle and given the rapidly changing technological and competitive landscape facing the banking industry. The test of an insurance fund is not how it does in good times but how it does in bad times, and neither the Bank Insurance Fund (BIF) nor the Savings Association Insurance Fund (SAIF) have been stress-tested since they were recapitalized. Third, we believe the Congress should consider rebates only as part of broader deposit insurance reform that allows risk based premiums to be more forward looking and allows deposit insurance to be priced to provide appropriate economic incentives. Fourth, any rebate program would have to address a myriad of operational complexities and questions of fairness.

RECENT HISTORY AND CURRENT CONDITION OF THE FUNDS

The FDIC was established in 1933 and the Federal Savings and Loan Insurance Corporation (FSLIC) was established in 1934. Throughout its history, the FDIC has insured some savings institutions, notably state-chartered savings banks, but for the most part it has insured commercial banks. The FSLIC insured savings and loan associations (S&Ls). In 1989, in the aftermath of the savings and loan crisis of the 1980s and the insolvency of the FSLIC, the Savings Association Insurance Fund was established to succeed the FSLIC fund, and the FDIC fund was renamed the Bank Insurance Fund. Both funds were put under the management of the FDIC.

The banking industry and the BIF were under extreme financial stress in the early 1990s. At the end of 1991, in part because of a $16 billion loss reserve for anticipated failures, the BIF had a negative net worth of $7 billion. A series of assessment rate increases from 1990 to 1993 more than tripled annual assessment income, which peaked at $5.8 billion in 1993. The fortunes of banks improved with the economy following the recession of 1990-1991, and it became apparent that fewer banks would fail than had been projected in 1991. This permitted a reversing of the loss reserves that had been set aside for anticipated failures. From 1992 through 1994, banks paid insurance premiums totaling $17 billion, and the reversal of loss reserves added nearly $13 billion to the fund's income. In May 1995, the BIF became fully recapitalized at 1.25 percent of insured deposits, with a balance of $24.7 billion.

The SAIF was slower to reach full capitalization. By statute, from its inception in 1989 through 1992, substantially all SAIF assessment revenue was diverted to help pay for the S&L crisis: 44 percent of SAIF assessments went to the Financing Corporation (FICO), 37 percent to the FSLIC Resolution Fund (FRF), and 18 percent to the Resolution Funding Corporation (REFCORP). For the most part, FICO and FRF paid for the obligations of the old FSLIC while REFCORP provided part of the funding for the Resolution Trust Corporation (RTC). The FICO annual draw of $793 million continued to 1996. By the time the BIF reserve ratio reached 1.25 percent in 1995, the SAIF lagged at 0.37 percent. The SAIF did not reach full capitalization of 1.25 percent of insured deposits until September 1996, when it was boosted by a $4.5 billion special assessment on SAIF-assessable deposits pursuant to the Deposit Insurance Funds Act of 1996 (Funds Act). The Funds Act also separated FICO assessments from deposit insurance assessments and expanded the FICO assessment base to include all FDIC-insured institutions.

With both insurance funds on sound footing and the bank and thrift industries benefiting from a strong economy, the FDIC was able to lower assessment rates. By law, when a fund is at or above the Designated Reserve Ratio (DRR), the FDIC is now generally prohibited from charging assessments to well-capitalized, well-managed, institutions. More than 90 percent of insured banks and thrifts qualify for the highest, 1A, rating and pay no insurance premiums. As a result, annual assessment income for the BIF averaged just $27 million from 1997 through 1999 (nine months, annualized), and annual assessment revenue for the SAIF averaged just $14 million in the same period (see Tables 1 and 2). Annual investment earnings for the three-year period averaged $1.6 billion for the BIF and $558 million for the SAIF.

Although BIF revenues covered expenses and deposit growth in 1997 and 1998, insurance losses escalated in 1999 and the fund reported a net loss for the first nine months of the year. As of September 30, 1999, the BIF totaled $29.5 billion. The reserve ratio -- the fund balance as a percentage of insured deposits -- remained 1.38 percent, the same as the reserve ratio at year-end 1997 and 1998. For the SAIF, revenues were slightly more than sufficient to cover expenses and fund growth necessitated by insured-deposit growth. On September 30, 1999, the SAIF balance stood at $10.2 billion and its reserve ratio was 1.44 percent1, compared to $8.9 billion and 1.30 percent at the end of 1996.

Table 1

Bank Insurance Fund, 19911999

(Dollar amounts in millions)

           
 

Interest

Income

Assessment

Income

Loss

Provisions

Fund

Balance

Reserve

Ratio (%)

           

9/99 YTD

$1,293

$25

$917

$29,499

1.38

1998

1,674

22

(38)

29,612

1.38

1997

1,519

25

(495)

28,293

1.38

1996

1,267

73

(325)

26,854

1.34

1995

1,068

2,907

(33)

25,454

1.30

1994

521

5,591

(2,873)

21,848

1.15

1993

165

5,784

(7,677)

13,162

0.69

1992

299

5,588

(2,260)

(101)

(0.01)

1991

471

5,160

15,476

(7,028)

(0.36)

Table 2

Savings Association Insurance Fund, 19911999*

(Dollar amounts in millions)

           
 

Interest

Income

Assessment

Income

Loss

Provisions

Fund

Balance

Reserve

Ratio (%)

           

9/99 YTD

$432

$10

$6

$10,205

1.44

1998

563

15

32

9,840

1.39

1997

535

14

(2)

9,368

1.36

1996

254

5,222

(92)

8,888

1.30

1995

169

970

(321)

3,358

0.47

1994

83

1,132

414

1,937

0.28

1993

25

898

17

1,156

0.17

1992

7

172

(15)

279

0.04

1991

3

94

20

94

0.01

* Fund balance for 9/99 includes the SAIF Special Reserve. Assessment income for 1996 includes the SAIF special assessment of $4.5 billion.

CONGRESS SHOULD MERGE THE FUNDS NOW

After many years of persistent attempts, Congress succeeded in modernizing many of the laws governing the financial services industry in the United States through passage of the Gramm-Leach-Bliley Act of 1999 (GLBA). In many ways, the legislation is forward-looking, creating new opportunities that will benefit the financial services industry, the U.S. economy and consumers well into the new century. GLBA also updated some laws to reflect the current marketplace, eliminating obsolete statutes that had been chipped away by subsequent legislative or regulatory measures or bypassed in whole or in part by innovation. Some of these laws, notably the Glass-Steagall Act, had lingered since the 1930s. However, there is one relic of the statutory framework established after the Great Depression that GLBA did not address -- the existence of separate deposit insurance funds for banks and thrifts.

A Combined Fund Would Be Stronger and More Efficient

A merger of BIF and SAIF would ensure that the risks to the deposit insurance system are as diversified as possible. The more concentrated the risks -- by numbers of institutions, by geography, by types of products -- the more concentrated are the dangers and the greater is the likelihood that trouble in a single institution or in a small group of institutions would seriously impact a fund. We encourage our insured-institutions to diversify, and the same principle applies to the insurance fund.

With ongoing consolidation in the industry and the rise of the "megabank," the FDIC's risk is increasingly located in a few large institutions. From June 1990 to September 1999, the share of SAIF-insured deposits held by the three largest institutions rose from 8.7 percent to 15.7 percent. The BIF had a larger increase in concentration during this period, with the share of its three largest insured institutions rising from 5.0 percent to 14.0 percent. In a combined insured-deposit base, the three largest institutions would hold only 12.7 percent. A combined deposit insurance fund, with a balance of $40 billion and a reserve ratio of 1.4 percent, would be better equipped than either fund alone to address the increased concentration of the industry. A recent paper by an FDIC economist shows that, on the basis of historical data, a combined fund would have a lower probability of insolvency than either fund individually2. This translates to better protection for taxpayers.

A combined fund also would be more efficient than the present structure. In 1995 and 1996, the BIF had recapitalized and the FDIC could lower its assessment rates substantially, while the SAIF remained undercapitalized and was required to maintain higher rates. Thus, identical products were available at different prices. When such a price disparity exists, consumers -- in this case, banks and thrifts that pay deposit insurance assessments -- naturally gravitate to the lower price. Despite moratoriums, exit and entrance fees, and bans on deposit shifting, market forces ultimately prevailed. Institutions wasted time and money trying to circumvent restrictions that prohibited them from purchasing deposit insurance at the lowest price. The Deposit Insurance Funds Act of 1996 (Funds Act) led to the elimination of the disparity in deposit insurance assessment rates that then existed between the BIF and the SAIF, but as long as there are two deposit insurance funds, whose assessment rates are determined independently, the prospect of a premium differential exists. A merged fund would guarantee that such a disparity would not recur in the future. It would have a single assessment rate schedule whose rates would be set solely on the basis of the risks that institutions pose to the single fund.

The FDIC has examined the mechanics of merging the funds, and has found that there are no significant obstacles or expenses in such a merger. Indeed, a merger of the funds would result in lower costs and regulatory burden for approximately 850 institutions that hold both BIF- and SAIF-insured deposits (Oakar deposits) that must be tracked and assessed separately. Although these costs may not be large in absolute dollars, they represent unnecessary expenditures.

The Timing for a Merger Is Optimal

The arguments for a merger of the BIF and the SAIF are persuasive and the timing is optimal. Changes in the bank and thrift industries in recent years -- and in the larger financial services industry -- have been substantial. Many of the statutory differences between bank and thrift charters have been narrowed, bringing them into keener competition with one another. In the 1930s, when the FDIC and the FSLIC were established, S&Ls were, in general, mutual institutions that primarily offered savings accounts and home mortgages for consumers. Because their charters were limited, S&Ls were not allowed to offer checking accounts, consumer loans, or commercial loans. Indeed, their loans were virtually all long-term, fixed-rate residential mortgages. Commercial banks, on the other hand, served mostly commercial customers. More than two-thirds of bank deposits were demand deposits and banks made very few residential mortgages.

Over time, the distinctions between banks' and thrifts' powers have become blurred. Each has entered what was once the other's domain. Both offer essentially an identical array of deposit accounts. In addition, in the aftermath of the Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994, both banks and thrifts can branch nationwide. From the point of view of the insured depositor, there is virtually no difference between banks and thrifts.

In 1996, the Funds Act provided for a merger of the funds on January 1, 1999, if there were no savings association in existence on that date. It was thought at the time that a new charter that was common both to banks and thrifts would be developed, and the thrift charter could be eliminated. This did not occur. However, GLBA addressed what some believed to be an inequity in federal law that had permitted the combination of banking and commerce through unitary savings and loan holding companies. Such combinations were prohibited to bank holding companies. GLBA bans new unitary thrift holding companies from engaging in nonfinancial activities or affiliating with nonfinancial entities. Thus, while there remain separate charters, the elimination of the unitary thrift holding company puts to rest the rationale to condition the merger of the funds on the elimination of the thrift charter.

Not only have the banking and thrift industries become more similar over time, the composition of who holds SAIF-insured deposits has changed as well. The name Savings Association Insurance Fund connotes a fund that insures deposits at savings associations. When it was established in 1989, this was indeed the case. Virtually all SAIF-insured deposits were held by SAIF-member savings associations. However, over the last decade, this changed dramatically. As of September 30, 1999, commercial banks (38 percent) and state-chartered savings banks (8 percent) held over 45 percent of all deposits insured by the SAIF. Indeed, 25 of the 50 largest holders of SAIF-insured deposits are BIF members, including First Union National Bank (ranked second) and Bank of America, N.A. (ranked third). The name Savings Association Insurance Fund has become a misnomer. The SAIF has become a true hybrid fund.

The current health of the bank and thrift industries and of the insurance funds also indicate that now is an ideal time to merge the funds. Despite recent indications of deteriorating credit quality, the condition of the bank and thrift industries reflects the current favorable economic environment, with high, broad-based profitability, sound balance sheets and low numbers of failures. This current industry strength is evident in the deposit insurance funds. With low levels of assets from failed institutions, both funds are highly liquid, with the preponderance of the funds' assets invested in interest-bearing U.S. government securities. The reserve ratios of both funds are similar. As of September 30, 1999, the reserve ratio of the BIF was 1.38 percent, and that of the SAIF was 1.44 percent. A combined fund would have a reserve ratio of 1.40 percent, causing only a minor amount of dilution of the SAIF. While these favorable conditions have existed for several years, economic history indicates such conditions do not persist indefinitely. History also tells us that, when there exists a perception of disparity in the quality of one of the funds, the notion of merging them becomes controversial. Now is an excellent time to merge the funds, rather than when the industry or one or both of the funds come under stress.

REBATES AND DEPOSIT INSURANCE REFORM

Madam Chairwoman, your letter of invitation also asked us to discuss rebates and a cap on a merged insurance fund. I will start this discussion with a brief history of the assessment credits paid by the FDIC from 1950 to 1983 (Appendix A contains a longer discussion of assessment credits). I will then explain why the FDIC believes that great caution should govern any consideration of rebates and a cap on the fund. Finally, I will explain why the FDIC believes that any consideration of rebates should be put in the context of broader deposit insurance reform, outline general principles that we believe should govern that reform, and draw implications of those principles for the risk-based premium system and rebates.

History of Assessment Credits

The idea of rebates for insured institutions originated in the late 1940s when many observers felt that a $1 billion fund was sufficient to cover almost any economic contingency. However, because the FDIC and its insurance fund had not been tested by a major business downturn during this period, the adequacy of the deposit insurance fund was not known. The FDIC was reluctant to support a permanent reduction in the basic assessment rate.

At the time, all institutions paid a statutory flat assessment rate of one-twelfth of one percent (8.3 basis points). Rather than reducing the flat rate for deposit insurance, the Federal Deposit Insurance Act of 1950 created what was commonly referred to as rebates, but which the statute, more accurately, defined as assessment credits3. At the end of each year, the FDIC calculated net assessment income by taking gross assessment income and subtracting operating expenses and insurance losses. In the years when assessments exceeded the FDIC's costs, the assessment credit authority initially permitted the FDIC to keep 40 percent of the net assessment income and required that it credit the balance -- 60 percent of net assessment income -- pro rata to insured banks. In 1961 the assessment credit was increased to two-thirds of net assessment income, and in 1980 it reverted back to 60 percent.

To take an example, in 1978 the assessment base averaged $973 billion, and at an assessment rate of 8.3 basis points, gross assessment income was $811 million. Insurance losses in 1978 were estimated to be $42 million and operating expenses were $103 million. Net assessment income was thus $811 million minus $145 million, or $666 million. Two-thirds of this, or $444 million, was credited back to the industry so that effective premiums totaled $367 million ($811 million minus $444 million). With an assessment base of $973 billion, the effective premium was 3.8 basis points.

Two aspects of the old assessment credit system are significant. First, the assessment credits were a partial offset to mandatory, flat rate, assessments. The FDIC never provided rebates from money in the deposit insurance fund. Indeed, every year that assessment credits were provided, the fund continued to grow. For example, in 1978 the fund increased by $803 million -- $222 million in retained net assessment income and $581 million in investment income.

Second, there were no circumstances under which insured institutions paid an effective premium of zero. As shown in Table 3, the effective premium -- after the assessment credits --from 1950 through 1983 was always above 3.0 basis points. By contrast, in 1998 the average effective premium was less than 0.1 basis points. Of course, given the current risk-based premium system, the current effective premium is a blend. Some institutions pay much more, but for the best-rated institutions-currently over 90 percent-the assessment rate is zero.

Table 3

BIF Effective Premium Rates, 1935 to 1998

Effective

Effective

Assessment

Assessment

Year

Rate (b.p.)

Year

Rate (b.p.)

1998

0.08

1966

3.2

1997

0.08

1965

3.2

1996

0.24

1964

3.2

1995

12.4

1963

3.1

1994

23.6

1962

3.1

1993

24.4

1961

3.2

1992

23.0

1960

3.7

1991

21.3

1959

3.7

1990

12.0

1958

3.7

1989

8.3

1957

3.6

1988

8.3

1956

3.7

1987

8.3

1955

3.7

1986

8.3

1954

3.6

1985

8.3

1953

3.6

1984

8.0

1952

3.7

1983

7.1

1951

3.7

1982

7.7

1950

3.7

1981

7.1

1949

8.3

1980

3.7

1948

8.3

1979

3.3

1947

8.3

1978

3.8

1946

8.3

1977

3.7

1945

8.3

1976

3.7

1944

8.3

1975

3.6

1943

8.3

1974

4.4

1942

8.3

1973

3.9

1941

8.3

1972

3.3

1940

8.3

1971

3.5

1939

8.3

1970

3.6

1938

8.3

1969

3.3

1937

8.3

1968

3.3

1936

8.3

1967

3.3

1935

8.3

No assessment credits were issued between 1984 and 1991 as mounting insurance losses led to expenses in excess of gross assessment income. Assessment credit authority was eliminated in 1991 as part of the Federal Deposit Insurance Corporation Improvement Act of 1991 (FDICIA). FDICIA required the FDIC to implement a risk-based premium system. The Congress appears to have intended that the ability to adjust assessments through a risk-based system would operate in lieu of assessment credits4. For example, rather than charging 8.3 basis points and providing an assessment credit of 4.5 basis points for an effective assessment rate of 3.8 basis points as was done in 1978, the FDIC could simply set premiums at 3.8 basis points.

In total, the banking industry received $6.7 billion in assessment credits between 1950 and 1983. It is interesting to note that in 1991, at the height of the banking crisis, the BIF had a net worth of negative $7.0 billion. Taking account of foregone interest income, the BIF would never have been insolvent if assessment credits had never been provided. It also would have recapitalized faster and the industry would not have had to pay as much in premiums at a time when many banks were struggling to survive.

In 1996, the Funds Act effectively reestablished assessment credit authority for the Bank Insurance Fund under the terminology of "refunds." This authority requires the FDIC at the end of each semiannual period to refund any balance in the BIF that exceeds the amount required to meet the DRR, subject to two limitations. First, the amount of the refund cannot exceed the assessments paid by a member during the semiannual period. Second, refunds cannot be paid to institutions that are not well-capitalized or that "exhibit financial, operational, or compliance weaknesses ranging from moderately severe to unsatisfactory." This condition bars refunds to all but the best-rated institutions, those rated 1A under the current risk-based premium system. Since this group of institutions has had an assessment rate of zero since 1996 and paid no premiums, they are not eligible to receive refunds. The Funds Act had no refund provisions for members of the SAIF.

Madam Chairwoman, your letter of invitation also asked that we discuss the rebate authority of the National Credit Union Share Insurance Fund (NCUSIF). The NCUSIF's rebate authority bears some resemblance to the FDIC's old assessment credit system, although there are enough differences between the funding structures of the FDIC funds and the NCUSIF, that comparisons are difficult. A discussion of the NCUSIF is presented in Appendix B.

Caution Must Govern Any Consideration of Rebates

As happened in the late 1940s, questions are now arising about whether the insurance funds are overcapitalized. These questions are understandable, given the recent low loss experience of the funds. Certainly, if the ratio of insurance funds to insured deposits (i.e., the reserve ratio) was expected to grow indefinitely, a cap would be in order, but history tells us that this is not the case. The funds tend to grow during good times and to fall during bad times. Accordingly, we believe that great caution should govern any consideration of rebates, especially at this stage in the economic cycle.

In this regard, it is important to note that the deposit insurance reforms enacted since the banking crisis of the 1980s have not been tested in an economic downturn. The BIF grew rapidly between 1992 and 1996 as reserves for anticipated losses were reversed. But, with little reserves left to reverse, the BIF reserve ratio has essentially remained unchanged -- even in these exceptional economic times -- since 1997. It stood at 1.38 percent at year-end 1997, year-end 1998, and September 30, 1999. It is reasonable to assume that the reserve ratio will decline in an economic downturn.

Indeed, the BIF suffered a loss of $113 million during the first nine months of 1999. This was primarily because of unexpectedly large losses at just two medium-sized institutions, but it should serve as a cautionary note, especially at this stage of the economic cycle. Recent analysis by FDIC staff indicates that business lending risks are on the rise. Although still low by historical standards, one sign that credit quality in commercial and industrial portfolios is deteriorating is a doubling of domestic commercial and industrial loan losses at banks during the first half of 1999 compared to the same period in 1998. Other signs are slower corporate profit growth and rising corporate bond defaults.

The SAIF reserve ratio has followed a slightly more favorable trend. At year-end 1996, following capitalization of the SAIF, the reserve ratio stood at 1.30 percent. With only one failure since 1996, the fund stood at 1.44 percent as of September 30, 1999 (including the SAIF Special Reserve, eliminated by GLBA). However, the SAIF also is not immune from the economic cycle.

In addition, the banking industry is undergoing rapid change. Recent research by the FDIC shows that, on the basis of historical data, the recent industry consolidation has increased the insolvency risk of the BIF5. Technological advances make it possible for individuals and businesses to move funds around the globe almost instantaneously. Financial engineering is increasing the opportunities but -- when not properly managed -- also the risks faced by the industry.

It is important to note that deposit growth, not just losses, can impact the deposit insurance funds significantly. A recent announcement by a major investment bank illustrates how quickly insured deposits can grow and an insurance fund reserve ratio can change in the information age. According to recent press reports, by June, this investment bank plans to sweep uninvested funds in cash management accounts (CMA) -- brokerage accounts where uninvested funds are swept into uninsured taxable money-market funds -- into insured deposits at two banks that it owns, providing each CMA holder with up to $200,000 in FDIC insurance. Due to an increase in insured deposits with no commensurate increase in insurance funds, the volume anticipated in this case -- $100 billion -- would cause serious dilution of the BIF reserve ratio. On a pro-forma basis, the September 30, 1999, BIF reserve ratio would decline by 6 basis points to 1.32 percent. Similar programs by other companies could easily erode the remainder of the BIF's "cushion" above the DRR of 1.25 percent.

Rebates Should Only Be Considered As Part Of Broader Deposit Insurance Reform

Rebates would represent a significant change in our current risk-based premium system, as the system was designed to replace a fixed-premium system with assessment credits. As such, we believe very strongly that any discussion of rebates should be put in the context of overall deposit insurance reform.

There are three fundamental principles that we believe any deposit insurance reform would have to meet:

  1. The reform should strengthen the banking system. One of the purposes of deposit insurance is to help bring stability to the banking system. To the extent possible, deposit insurance should not put additional stress on banks in difficult economic times.
  2. The reform should strengthen the deposit insurance system. In order for deposit insurance to protect insured depositors without putting taxpayers at risk, the deposit insurance fund must have adequate reserves, and the ability to raise funds during times of need without harming otherwise solvent institutions.
  3. To the extent practicable, reform should not distort banks' economic incentives. A certain amount of moral hazard is inevitable in any deposit insurance scheme6. However, if the insurance is mispriced, it will further alter bank behavior, shifting risk that should be borne by banks' stockholders to the deposit insurance fund and taxpayers.

The aforementioned three principles have a number of implications for reform of the current risk-based deposit insurance system. In particular, risk-based premiums should be forward-looking, and deposit insurance should be priced to provide appropriate economic incentives. Insured institutions benefit if risk is priced appropriately from the start -- not after the fact, when all the FDIC can do is price premiums to cover losses.

The current risk-based premium statute authorizes the FDIC to charge premiums to institutions that "exhibit financial, operational, or compliance weaknesses ranging from moderately severe to unsatisfactory." This language closely tracks the definition of a CAMELS 3, 4, or 5 institution. While the CAMELS rating system has become more forward-looking in recent years, it is still, in large part, a snapshot of an institution's current condition, not a prospective look at the risks an institution is undertaking. To partially address this problem, the FDIC, working with the other banking agencies, has developed off-site screens to identify highly rated institutions that nonetheless exhibit high levels of risk in their operations. The FDIC has implemented the screens, along with procedures for reviewing the risk management practices of insured institutions, and will charge higher premiums if deficiencies are not corrected in a reasonable period of time. This is, however, only a partial solution. The fact remains that the statutory language gives the FDIC only limited ability to separate risk classification from CAMELS ratings and to price risk prospectively.

A related concern is the application of a flat rate, in this case zero, to a large proportion of insured institutions. A private insurer would only charge a flat rate if there were negligible differences in risk-taking across all institutions in the group or if it were too difficult to discern any meaningful differences. Many institutions in the best-rated premium category are strong and well-managed, but not all 9,500 institutions are equally so. During the banking crises in the 1980s, a small percentage of institutions that were well-capitalized with good earnings when the economy was strong, experienced problems, and in some cases failed, when the economy softened. The FDIC should be given greater discretion over risk classifications, so that risk classifications can better distinguish among different levels of risk and be made more forward-looking.

Finally, in order to provide the appropriate economic incentives to insured institutions, the marginal cost of deposit insurance should not be zero. Presumably, the rationale behind a statutory zero premium is that, as long as a fund is above the DRR, it does not need additional funds. However, there is no uniquely correct level of fund reserves. Each level of the fund corresponds to a different insolvency risk, and to a different risk that insured institutions will have to pay premiums if -- or more accurately when -- failures rise in an economic downturn.

In this respect, your letter of invitation asked us to discuss how the 1.25 percent DRR was established. To the best of our knowledge, the 1.25 figure first found its way into the statute, albeit not as the DRR per se, as part of the Depository Institutions Deregulation and Monetary Control Act of 1980. The 1980 Act established a range in which the reserve ratio of the fund was to be maintained. The assessment credit percentage was to be adjusted if the reserve either exceed 1.40 percent or fell below 1.10 percent. At the midpoint of that range, 1.25 percent, the FDIC was authorized, but not required, to reduce the net assessment income by an amount that would result in maintaining the reserve ratio at not less than 1.25 percent. The 1.25 percent DRR was established in 1989 by the Financial Institutions Reform, Recovery, and Enforcement Act, probably as a result of the language in the Monetary Control Act of 1980. We are not aware of any analysis, however, that supported the choice of 1.25 percent as the DRR.

Aside from raising money for the insurance funds, premiums also serve to align economic incentives. When a valuable product is offered at zero cost, it leads to that product being overused, causing distortions throughout the marketplace and, in the case of deposit insurance, exacerbating moral hazard.

Some have argued that in recapitalizing the funds, depository institutions have, in effect, prepaid premiums. However, because prepaid insurance distorts the incentive to avoid risk, it is generally not available except for life insurance policies. In addition, even prepaid life insurance does not provide for coverage to be increased with no limit at no additional cost. Yet, depository institutions can grow their deposits without incurring any additional costs for deposit insurance. New institutions have been chartered without ever paying a cent for deposit insurance. If we want a deposit insurance system that minimizes distortion and moral hazard, then institutions should pay deposit insurance premiums that better reflect the risk they pose to the fund.

With respect to rebates, we believe, as we said earlier, that any proposal should be part of broader deposit insurance reform. However, there are some general conclusions that can be drawn from the three principles we have outlined. For instance, we would be concerned with a rebate that resulted in the FDIC, in effect, paying to provide insurance to a large portion of the industry. This would exacerbate the already bad economic incentives caused by charging zero premiums, and further increase moral hazard.

We would also be concerned about a proposal that capped the fund at the DRR7. Rebating all funds above the DRR would result in all operating costs and insurance expenditures above those that could be funded by earnings on the fund being funded by premiums -- in essence, putting the industry on a pay-as-you-go basis. Since bank failures are likely to increase in an economic downturn, when bank earnings are already coming under pressure, this would result in the industry paying premiums when it can least afford them, which would not serve to strengthen the industry.

Similar concerns would arise if a combined fund were capped at 1.5 percent of insured deposits. At September 30, 1999, 1.5 percent of insured deposits of a combined fund was $42.6 billion. Based on current assessment rates, interest income, insurance losses and deposit growth, it appears unlikely that a merged fund would reach a 1.5 percent reserve ratio in the foreseeable future. If the reserve ratio does surpass 1.5 percent, it is likely to coincide with a strong economy and strong industry earnings. By contrast, under the current premium system, it is a virtual certainty that the fund will, at some point in time when the economy and bank earnings are weak, fall significantly below the DRR. When this happens, the FDIC will be forced by statute to charge premiums of 23 basis points or more. We have misgivings about increasing the premiums that banks will have to pay when earnings are under stress, in order to pay rebates when earnings are strong.

At what level, if any, it might be appropriate to cap the fund is a very complex question. It is extremely difficult to develop a number that would represent the ideal reserve ratio under all conditions. The answer would depend critically on what other deposit insurance reforms would be put in place, and how rebates over the cap would be structured. It also would depend on economic conditions and industry changes, such as consolidation and new uses of technology.

Rebates Would Pose Significant Operational Complexity

Any rebate (or cap) proposal would almost certainly result in operational complexities and questions of fairness. A key question would be upon what basis are rebates to be distributed? The simplest arrangement would be to distribute rebates on the basis of the current assessment base, that is, domestic deposits. While simple, this would almost certainly result in a negative marginal cost for insurance (i.e., each dollar in new deposits would result in the institution getting more funds back from the FDIC). It would also raise some fairness issues that must be addressed.

First is the issue of de novo institutions. Since the respective insurance funds were capitalized, 814 new banks and thrifts have been chartered. For the most part these insured institutions have not contributed to the deposit insurance funds and thus rebates to these institutions would amount to a windfall. Although these institutions account for a relatively small portion of the industry, this group does include a number of institutions with plans for significant growth.

The logic of this concern also applies to institutions that have grown rapidly since the capitalization of the insurance funds. The major investment bank reported by the press, that I referred to earlier, is the most obvious and has the potential to be the most significant of these. With $100 billion in deposits, it could get a larger rebate than all but five depository organizations.

Moreover, recent experience suggests that rapid growth often indicates greater risk to the insurance funds. Last week I testified on three recent failures that had extraordinarily high loss rates. In two of these failures, the banks recorded deposit growth rates in the two years before failure far in excess of the industry average. A rebate system that rewards rapid growth may not be consistent with the principle of appropriate economic incentives.

There are alternatives to using the current assessment base. Rebates could be based on an earlier snapshot in time. The choice would be somewhat arbitrary and it is not clear whether and how such an allocation should change over time. Actual contributions to the fund could serve as the basis; the critical decision here is over what period should such contributions be measured. Looking back five years from today would skew the allocation significantly toward SAIF members, given the special assessment in 1996. To reflect the contributions of BIF members to that fund's capitalization would require perhaps a ten-year "lookback." Basing a future rebate system on a picture of the industry during its most difficult period and before the dramatic consolidation and modernization underway may introduce problems with incentives and operational issues that are difficult to foresee at this point.

CONCLUSION

Madam Chairwoman, Congress should act promptly to merge the BIF and the SAIF. A merger represents a further step in modernization and would both reduce insolvency risks and eliminate the possibility of identical federal deposit insurance coverage once again being offered at two different prices. A merger has been debated for several years. Fortunately, many of the arguments preventing a merger have been resolved or diminished. We should merge the funds now to diversify the risk to the deposit insurance funds and to better protect the taxpayer's interest.

We also would urge the Subcommittee to exercise great caution in considering rebates, especially since key deposit insurance reforms enacted in the last decade have yet to be tested in an economic downturn. For example, under the risk-based premium system, more than 90 percent of the banking industry pays no deposit insurance premium -- with the result that the BIF reserve ratio has not grown and has remained essentially unchanged since 1997, even in these strong economic times. In addition, the FDIC has never provided rebates from money in the deposit insurance funds. Instead, only partial offsets to mandatory flat rate assessments were allowed. The BIF also suffered a loss of $113 million through the first nine months of 1999 and there are indications that business lending risks are on the rise. Finally, if Congress decides to mandate rebates despite these concerns, it should be done in the context of overall deposit insurance reforms that strengthen the banking system, strengthen the deposit insurance system and do not distort banks' economic incentives.

Appendix A: HISTORY OF ASSESSMENT CREDITS

The idea of rebates for insured institutions originated in the late 1940s. By year-end 1946, the deposit insurance fund (the fund we now know as the Bank Insurance Fund) had surpassed $1 billion in reserves. Many observers felt that a $1 billion fund was sufficient to cover almost any economic contingency. While the deposit insurance fund grew from $292 million in 1934 to $1.2 billion in 1949, the ratio of the deposit insurance fund to insured deposits -- the reserve ratio -- remained stable, primarily due to deposit growth. The reserve ratio was 1.61 percent at year-end 1934, and 1.57 percent at year-end 19491.

By 1950, the fund had reached a balance of $1.2 billion. Bankers were voicing concern that the statutory assessment rate was too high. As provided under the Banking Act of 1935, insured banks paid an assessment at a flat annual rate of 8.3 cents per $100 of assessable deposits. However, because the FDIC and its insurance fund had not been tested by a major business downturn during this period, the adequacy of the deposit insurance fund was not known. The FDIC was reluctant to support a permanent reduction in the basic assessment rate.

Rather than reducing the deposit insurance assessment rate, the Federal Deposit Insurance Act of 1950 created an assessment credit system, or what is commonly referred to as a rebate system. The law required the FDIC to calculate net assessment income to determine whether insured banks received assessment credits. Under this original assessment credit system, insured banks continued to pay the 8.3 basis points in assessments. At the end of the year, the FDIC calculated net assessment income by taking gross assessment income and subtracting operating expenses and insurance losses. If the net assessment income was positive, the assessment credit authority required the FDIC to retain 40 percent of the net assessment income and credit the balance -- 60 percent of net assessment income -- pro rata to insured banks. The FDIC applied this credit toward the payment of their deposit insurance assessment for the following semiannual period. Thus, in years in which premiums paid by insured banks were more than sufficient to cover operating expenses and insurance losses, banks received an assessment credit. The FDIC used the remaining portion of net assessment income to continue to build the fund.

This procedure tended to stabilize FDIC earnings during periods of fluctuating loss experience and also allowed insured banks to pay lower effective premium rates during years in which losses were low. From 1950 to 1980, the effective assessment rate stayed in the range of 3.1 cents to 3.9 cents per $100 of assessable deposits, except for a slight blip in 1974 (4.4 cents). Also, the assessment credit formula returned current net assessment income only; there was no rebate of investment income or principal. Despite the implementation of the assessment credit program in 1950, the low insurance-loss experience of the 1950s and 1960s allowed the insurance fund to grow, reaching $4.4 billion at the end of 19702. However, the fund's growth rate still trailed that of insured deposits and the reserve ratio declined to 1.25 percent by the end of 19703 .

It was not until the Depository Institutions Deregulation and Monetary Control Act of 1980 that the assessment system was specifically linked to the deposit insurance fund's reserve ratio. That Act established a range in which the assessment credit percentage was to be adjusted if the reserve ratio either exceeded 1.40 percent or fell below 1.10 percent. The Act also reduced the assessment credit percentage from 66-2/3 percent to 60 percent, the level that had been in effect from 1950 to 1960. Because of mounting losses, reduced assessment credits were paid in 1981 through 1983, and no assessment credits were paid thereafter.

Effective assessment rates grew rapidly as insurance losses mounted throughout the 1980s and early 1990s. When the full statutory rate of 8.3 basis points of $100 in deposits proved too low, Congress mandated an increase to 12 basis points in 1990 and gave the FDIC board of directors more flexibility to raise rates. With losses continuing at record levels, rates were increased twice in 1991, first to 19.5 basis points and then to 23 basis points.

In 1989, the Financial Institutions Reform, Recovery, and Enforcement Act barred the use of assessment credits unless the reserve ratio of the fund was at or exceeded its Designated Reserve Ratio (DRR). With the passage of the Federal Deposit Insurance Corporation Improvement Act of 1991 (FDICIA), the Congress gave the FDIC the flexibility to implement a risk-based premium system and eliminated the FDIC's assessment credit authority. The Congress appears to have intended that the ability to adjust assessments through a risk-based system would operate in lieu of assessment credits4. For example, rather than charging 8 basis points and rebating half, the FDIC could simply set premiums at 4 basis points.

The Deposit Insurance Funds Act of 1996 (Funds Act) effectively reestablished rebate authority for the Bank Insurance Fund under the terminology of "refunds." The Funds Act also provided for the possible merger of the BIF and the SAIF, at which time all insured institutions were to be eligible for refunds, subject to the limitations listed below. Prior to a merger, SAIF-member institutions would not be eligible for refunds.

The Funds Act requires the FDIC at the end of each semiannual period to refund any balance in the BIF that exceeds the amount required to meet the designated DRR, subject to two limitations. First, the amount of the refund cannot exceed the assessments paid by a member during the semiannual period; and second, refunds cannot be paid to institutions that are not well capitalized or that "exhibit financial, operational, or compliance weaknesses ranging from moderately severe to unsatisfactory."

The second condition bars refunds to all but the best-rated institutions, those rated 1A under the current risk-based premium system. However, since the beginning of 1996, the assessment rate for BIF members rated 1A has been zero, so these institutions have paid no premiums and cannot receive a refund because of the restriction limiting refunds to semiannual payments made.

Insurance

Deposits in Insured Banks

Insurance

Reserve

Year

Coverage

Total

Insured

Fund

Ratio (%)

1998

$100,000

$2,996,396

$2,141,268

29,612

1.38

1997

100,000

2,785,990

2,055,874

28,293

1.38

1996

100,000

2,641,797

2,007,042

26,854

1.34

1995

100,000

2,478,888

1,951,963

25,454

1.30

1994

100,000

2,462,650

1,895,258

21,848

1.15

1993

100,000

2,490,816

1,905,245

13,122

0.69

1992

100,000

2,512,278

1,945,550

(101)

(0.01)

1991

100,000

2,520,074

1,957,722

(7,028)

(0.36)

1990

100,000

2,540,930

1,929,612

4,045

0.21

1989

100,000

2,465,922

1,873,837

13,210

0.70

1988

100,000

2,330,768

1,750,259

14,061

0.80

1987

100,000

2,201,549

1,658,802

18,302

1.10

1986

100,000

2,167,596

1,634,302

18,253

1.12

1985

100,000

1,974,512

1,503,393

17,957

1.19

1984

100,000

1,806,520

1,389,874

16,259

1.19

1983

100,000

1,690,576

1,268,332

15,429

1.22

1982

100,000

1,544,697

1,134,221

13,771

1.21

1981

100,000

1,409,322

988,898

12,246

1.24

1980

100,000

1,324,463

948,717

11,020

1.16

1979

40,000

1,226,943

808,555

9,793

1.21

1978

40,000

1,145,835

760,706

8,796

1.16

1977

40,000

1,050,435

692,533

7,993

1.15

1976

40,000

941,923

628,263

7,269

1.16

1975

40,000

875,985

569,101

6,716

1.18

1974

40,000

833,277

520,309

6,124

1.18

1973

20,000

766,509

465,600

5,615

1.21

1972

20,000

697,480

419,756

5,159

1.23

1971

20,000

610,685

374,568

4,740

1.27

1970

20,000

545,198

349,581

4,380

1.25

1969

20,000

495,858

313,085

4,051

1.29

1968

15,000

491,513

296,701

3,749

1.26

1967

15,000

448,709

261,149

3,486

1.33

1966

15,000

401,096

234,150

3,252

1.39

1965

10,000

377,400

209,690

3,036

1.45

1964

10,000

348,981

191,787

2,845

1.48

1963

10,000

313,304

177,381

2,668

1.50

1962

10,000

297,548

170,210

2,502

1.47

1961

10,000

281,304

160,309

2,354

1.47

1960

10,000

260,495

149,684

2,222

1.48

1959

10,000

247,589

142,131

2,090

1.47

1958

10,000

242,445

137,698

1,965

1.43

1957

10,000

225,507

127,055

1,851

1.46

1956

10,000

219,393

121,008

1,742

1.44

Table A-1 (continued)

Insurance

Deposits in Insured Banks

Insurance

Reserve

Year

Coverage

Total

Insured

Fund

Ratio (%)

1955

10,000

212,226

116,380

1,640

1.41

1954

10,000

203,195

110,973

1,543

1.39

1953

10,000

193,466

105,610

1,451

1.37

1952

10,000

188,142

101,841

1,364

1.34

1951

10,000

178,540

96,713

1,282

1.33

1950

10,000

167,818

91,359

1,244

1.36

1949

5,000

156,786

76,589

1,204

1.57

1948

5,000

153,454

75,320

1,066

1.42

1947

5,000

154,096

76,254

1,006

1.32

1946

5,000

148,458

73,759

1,059

1.44

1945

5,000

157,174

67,021

929

1.39

1944

5,000

134,662

56,398

804

1.43

1943

5,000

111,650

48,440

703

1.45

1942

5,000

89,869

32,837

617

1.88

1941

5,000

71,209

28,249

554

1.96

1940

5,000

65,288

26,638

496

1.86

1939

5,000

57,485

24,650

453

1.84

1938

5,000

50,791

23,121

421

1.82

1937

5,000

48,228

22,557

383

1.70

1936

5,000

50,281

22,330

343

1.54

1935

5,000

45,125

20,158

306

1.52

1934

5,000

40,060

18,075

292

1.61

Appendix B: NCUSIF REBATE AUTHORITY

The National Credit Union Share Insurance Fund (NCUSIF), administered by the National Credit Union Administration, protects the deposits (or shares) of credit union members. In some ways, the NCUSIF is similar to the BIF and the SAIF: accounts are insured to $100,000, and the NCUSIF's assets are invested primarily in U.S. government securities. As of October 31, 1999, the NCUSIF had a balance of $4.1 billion, including $3.2 billion in members' contributed capital and $927 million in retained earnings.

However, the NCUSIF's funding structure differs markedly from those of the FDIC's funds. Since 1985, member credit unions have been required to maintain a deposit with the NCUSIF equal to one percent of their insured shares, adjusted annually to reflect share growth (or semi-annually for credit unions with $50 million or more in assets). Each credit union counts its one percent deposit with the NCUSIF as an asset, and as part of the credit union's capital. This accounting treatment is allowed because NCUSIF must refund the one percent share to a credit union if that institution voluntarily terminates NCUSIF membership, such as by converting to another type of charter. FDIC-insured institutions do not carry any part of the BIF or SAIF reserves as an asset nor are they required to make such a one percent adjustment for deposit growth each year.

The Federal Credit Union Act defines the NCUSIF's "normal operating level" as an equity ratio specified by the Board, which shall not be less than 1.2 percent and not more than 1.5 percent. The Board has set the normal operating level at 1.3 percent (This most closely compares to the FDIC's Designated Reserve Ratio (DRR) of 1.25 percent.)

In addition to the one percent share deposit, NCUA may assess a credit union an annual premium for NCUSIF coverage of a flat rate specified "in an amount stated as a percentage of shares (which shall be the same for all insured credit unions)1." Although the NCUA has discretionary authority to assess premiums if the equity ratio falls below 1.3 percent, the NCUA must assess premiums if the equity ratio falls below 1.2 percent. Premiums were last collected in 1992 as the NCUSIF has been at or above its normal operating level of 1.30 since that time. The FDIC has the authority to charge risk-based premiums, although current law generally limits the FDIC's ability to charge a rate of other than zero to well-capitalized and well-managed (1A-rated institutions) so long as the DRR is met.

The NCUSIF's rebate authority requires distribution of the fund's equity if the fund's reserve ratio exceeds the normal operating level, currently set at 1.30 percent. In these circumstances, the distribution can be in the form determined by the NCUA Board and may include a waiver of insurance premiums, premium rebates, and/or distributions from NCUSIF equity.2 The distribution amount must be sufficient to restore the fund to its normal operating level. Distributions based on each institution's contributed capital have been paid by NCUSIF each year from 1995 through 1999. For each of these five years, the credit union industry has paid in more to the NCUSIF in its annual one percent adjustment than the industry has received in dividends, although this is not necessarily true for individual institutions. By contrast, well-capitalized and well-managed BIF and SAIF members do not pay any premiums.

Summary of Testimony of Donna Tanoue Chairman

1The Deposit Insurance Funds Act of 1996 required the FDIC on January 1, 1999, to set aside, in a special reserve, any funds in the SAIF in excess of the designated reserve ratio. The Gramm-Leach-Bliley Act of 1999 repealed this requirement.
2 Oshinsky, Robert, Merging the BIF and the SAIF: Would a Merger Improve the Funds' Viability? Federal Deposit Insurance Corporation, Division of Research and Statistics. Working Paper Series 99-4.
3 Assessment credits only applied to the FDIC fund, the predecessor of the BIF. The FSLIC, the predecessor of the SAIF, never provided assessment credits.
4 Section-by-Section Analysis of S.543, Congressional Record, February 21, 1992, p. S-2073, Discussion of Section 212: in E. Conforming Amendments
5 Oshinsky, Robert. Effects of Bank Consolidation on the Bank Insurance Fund. Federal Deposit Insurance Corporation, Division of Research and Statistics. Working Paper Series 99-3.
6 When applied to deposit insurance, the term moral hazard refers to the incentive for insured banks to engage in riskier behavior than would be feasible in the absence of insurance. 7 Using September 30, 1999 data, capping the fund at 1.25 percent would have capped the fund at $35.5 billion, $4.2 billion below the combined fund balance of $39.7 billion.
1 Table A-1 attached to this testimony shows the reserve ratio-the ratio of the deposit insurance fund to estimated insured deposits for the Bank Insurance Fund from 1934 through 1998.
2 In 1961, the rebate proportion was increased from 60 percent to 66 2/3 percent.
3 A portion of the decline also was attributable to periodic increases in the deposit insurance coverage limit.
4 Section-by-Section Analysis of Section .543, Congressional Record, February 21, 1992, p. S-2073, Discussion of Section 212: in E. Conforming Amendments.
1 12 U.S.C. Section 1782(c)(2).
2 64 Fed. Reg. S6148 (Oct. 18, 1999) to be codified at 12 C.F.R. Section 741.4(e).

****

Last Updated 02/09/2000communications@fdic.gov