Testimony of Chairman Alan Greenspan
Federal deposit insurance reform
Before the Committee on Banking, Housing, and Urban Affairs, U.S. Senate
April 23, 2002
Chairman Sarbanes, Senator Gramm, it is a pleasure to appear before this Committee to present the views of the Board of Governors of the Federal Reserve System on deposit insurance reform. I will be addressing general reform issues, as proposed by the Federal Deposit Insurance Corporation (FDIC) in the spring of 2001, rather than any specific bill. Consequently, I will be expressing the broad views of the Federal Reserve Board on the issues associated with modifications of deposit insurance.
At the outset, I should note that last year's FDIC report highlighted the significant issues and developed an integrated framework for addressing them. In broad terms, while the Board opposes any increase in coverage, we support the framework for other reform issues that the FDIC report constructed.
Benefits and Costs of Deposit Insurance
Despite its initial quite limited intent, the Congress over the intervening decades has raised the maximum amount of coverage five times to its current $100,000 level. The last increase, in 1980, was a more than doubling of the cap level and was clearly designed to let depository institutions, particularly thrifts, offer an insured deposit free of the then prevailing interest rate ceilings on such instruments, which applied only to deposits below $100,000. Insured deposits of exactly $100,000 thus became fully insured instruments in 1980, but were not subject to an interest rate ceiling. The efforts of thrifts to use $100,000 CDs to stem their liquidity outflows resulting from public withdrawals of smaller below-market-rate insured deposits led first to an earnings squeeze and an associated loss of capital and then to a high risk investment strategy that led to failure after failure. Depositors acquiring the new larger denomination insured deposits were aware of the plight of the thrifts but unconcerned about the risk because the principal amount of their $100,000 deposits was fully insured by the United States government. In this way, the 1980 increase in the deposit insurance cap to $100,000 exacerbated the fundamental thrift problem associated with concentration on long-term assets in a high and rising interest rate environment. Indeed, it significantly increased the taxpayer cost of the bail out of the bankrupt thrift deposit insurance fund.
Not withstanding this problematic episode, it is clear that deposit insurance has played a key--at times even critical--role in achieving the stability in banking and financial markets that has characterized the past almost seventy years. Deposit insurance, combined with other components of our banking safety net (the Federal Reserve's discount window and payment system guarantees) has meant that periods of financial stress are no longer characterized by depositor runs on banks and thrifts. Quite the opposite: Asset holders now seek out deposits--both insured and uninsured--as safe havens when they have strong doubts about other financial assets.
Looking beyond the contribution of deposit insurance to overall financial stability, we should not minimize the security it has brought to millions of households and small businesses. Deposit insurance has provided a safe and secure place for those households and small businesses with relatively modest amounts of financial assets to hold their transaction and other balances.
These benefits of deposit insurance, as significant as they are, have not come without cost. The very same process that has ended deposit runs has made insured depositors--as in the 1980s with insolvent, risky thrifts--largely indifferent to the risks taken by their depository institutions because their funds are not at risk if their institution is unable to meet its obligations. As a result, the market discipline to control risks that insured depositors would otherwise have imposed on banks and thrifts has been weakened. Relieved of that discipline, banks and thrifts naturally feel less inhibited from taking on more risk than they would otherwise assume. No other type of private financial institution is able to attract funds from the public without regard to the risk it takes with its creditors' resources. This incentive to take excessive risks is the so-called moral hazard problem of deposit insurance, the inducement to take risk at the expense of the insurer.
Thus, two offsetting implications of deposit insurance must be kept in mind. On the one hand, it appears fairly unambiguous that deposit insurance has contributed to the prevention of bank runs that could have destabilized the financial structure in the short-run. On the other, there are growing concerns that even the current levels of deposit insurance have created such underwriting imbalances at insured depository institutions that future large systemic risks have arguably risen.
Indeed, the reduced market discipline and increased moral hazard, have intensified the need for government supervision to protect the interests of taxpayers and, in essence, substitute for the reduced market discipline. Deposit insurance and other components of the safety net also enable banks and thrifts to attract more resources, at lower costs, than would otherwise be the case. In short, insured banks and thrifts receive a subsidy in the form of a government guarantee that allows them both to attract deposits at lower interest rates than would be necessary without deposit insurance, and to take more risk without the fear of losing their deposit funding. Put another way, deposit insurance misallocates resources by breaking the link between risks and rewards for a select set of market competitors.
From the very beginning, deposit insurance has involved a tradeoff. On the one hand, deposit insurance contributes to overall short-term financial stability and the protection of small depositors. On the other hand, deposit insurance induces higher risk-taking, resulting in a misallocation of resources and larger long-term financial imbalances that increase the need for government supervision to protect the taxpayers' interests. Deposit insurance reforms must balance these tradeoffs. Moreover, any reforms should be aimed primarily at protecting the interest of the economy overall, and not just the profits or market shares of particular businesses.
The Federal Reserve Board believes that deposit insurance reforms should be designed to preserve the benefits of heightened financial stability and the protection of small depositors without at the same time further increasing moral hazard or reducing market discipline. In addition, we urge that the implementing details be kept as straightforward as possible to minimize the risk of unintended consequences that comes with complexity.
Recommendations for Reform
1. Merging BIF and SAIF. The Board supports the FDIC's proposal to merge the BIF and SAIF funds. Because the charters and operations of banks and thrifts have become so similar, it makes no sense to continue the separate funds. Separate funds reflect the past, but neither the present nor the future. Merging the funds would diversify their risks, reduce administrative expenses, and widen the fund base behind an increasingly concentrated banking system. Most importantly, the federally guaranteed insurance coverage provided to the two sets of institutions are identical, and thus the premiums need to be identical as well. Under current arrangements, the premiums could differ significantly if one of the funds fell below the designated reserve ratio of 1.25 percent of insured deposits and the other fund did not. Should that occur, depository institutions would be induced to switch charter to obtain insurance from the fund with the lower premium. This could distort our depository structure. The federal government should not sell an identical service, like deposit insurance, at different prices.
2. Statutory Restrictions on Premiums. Current law requires the FDIC to impose higher premiums on riskier banks and thrifts but prevents it from imposing any premium on well-capitalized and highly-rated institutions whenever the corresponding fund's reserves exceed 1.25 percent of insured deposits. The Board endorses the FDIC recommendations that would eliminate the statutory restrictions on risk-based pricing and allow a premium to be imposed on every insured depository institution, no matter how well capitalized and well rated it may be or how high the fund's reserves.
The current statutory requirement that free deposit insurance be provided to well-capitalized and well-rated banks when FDIC reserves exceed a predetermined ratio maximizes the subsidy provided to these institutions and is inconsistent with efforts to avoid inducing moral hazard. Put differently, the current rule requires the government to give away its valuable guarantee when fund reserves meet some ceiling level. This free guarantee is of value to banks and thrifts even when they themselves are in sound financial condition and when macroeconomic times are good. At the end of last year, 92 percent of banks and thrifts were paying no premium. Included in this group were banks that have never paid any premium for their, in some cases substantial, coverage and fast-growing entities whose past premiums were extraordinarily small relative to their current coverage. We believe that these anomalies were never intended by the framers of the Deposit Insurance Fund Act of 1996 and should be addressed by the Congress.
The Congress did intend that the FDIC impose risk-based premiums, but the 1996 act limits the ability of the FDIC to impose risk-based premiums on well-capitalized and well-rated banks. And these two variables--capital strength and examiner overall rating-- do not capture all the risk that banks and thrifts could create for the insurer. The Board believes the FDIC should be free to establish risk categories based on any well-researched economic variables and to impose premiums commensurate with these risk classifications. Although a robust risk-based premium system would be technically difficult to design, a closer link between insurance premiums and individual bank or thrift risk would reduce moral hazard and the distortions in resource allocation that accompany deposit insurance.
We note, however, that significant benefits in this regard are likely to require a substantial range of premiums but that the FDIC has concluded in its report that premiums for the riskiest banks would probably need to be capped in order to avoid inducing failure at these weaker institutions. We believe that capping premiums may end up costing the insurance fund more in the long run should these weak institutions fail anyway, with the delay increasing the ultimate cost of resolution. The Board has concluded, therefore, that if a cap is required, it should be set quite high so that risk-based premiums can be as effective as possible in deterring excessive risk-taking. In that way we could begin to simulate the deposit insurance pricing that the market would apply and reduce the associated subsidy in deposit insurance.
Nonetheless, we should not delude ourselves that even a wider range in the risk-based premium structure would eliminate the need for a government back-up to the deposit insurance fund, that is, eliminate the government subsidy in deposit insurance. To eliminate the subsidy in deposit insurance--to make deposit insurance a real insurance system--the FDIC average insurance premium would have to be set high enough to cover fully the very small probabilities of very large losses, such as during the Great Depression, and thus the perceived costs of systemic risk. In contrast to life or automobile casualty insurance, each individual insured loss in banking is not independent of other losses. Banking is subject to deposit run contagion, creating a far larger extreme loss tail on the probability distributions from which real insurance premiums would have to be calculated. Indeed, pricing deposit insurance risks to fully fund potential losses--pricing to eliminate subsidies--would require premiums that would discourage most depository institutions from offering broad coverage. Since the Congress has determined that there should be broad coverage, the subsidy in deposit insurance cannot be fully eliminated, although we can and should eliminate as much of the subsidy as we can.
Parenthetically, the difficulties of raising risk-based premiums explain why there is no private insurer substitute for deposit insurance from the government. No private insurer would ever be able to match the actual FDIC premium and cover its risks. A private insurer confronted with the possibility, remote as it may be, of losses that could bankrupt the insuring entity would need to set especially high premiums to protect itself, premiums that few, if any, depository institutions would find attractive. And, if premiums were fully priced by the government or the private sector, the issuing entities would likely lower their offering rates, reducing the amount of insured deposits demanded, and consequently the amount outstanding would decline.
3. Designated Reserve Ratios and Premiums. The current law establishes a designated reserve ratio for BIF and SAIF of 1.25 percent. If that ratio is exceeded, the statute requires that premiums on well-capitalized and well-rated institutions must be discontinued. If the ratio declines below 1.25 percent, the FDIC must develop a set of premiums to restore the reserve ratio to 1.25 percent; if it appears that the fund ratio cannot be restored to its statutorily designated level within twelve months, the law requires that a premium of at least 23 basis points be imposed on all insured entities.
These requirements are clearly procyclical, lowering or eliminating fees in good times when bank credit is readily available and deposit insurance fund reserves should be built up, and abruptly increasing fees sharply in times of weakness when bank credit availability is under pressure and deposit fund resources are drawn down to cover the resolution of failed banks. The FDIC recommends that surcharges or rebates should be used to bring the fund back to the target reserve ratio gradually. The FDIC also recommends the possibility of a target range for the designated reserve ratio, over which the premiums may remain constant, rather than a fixed target reserve ratio and abruptly changing premiums.
We support such increased flexibility and smoothing of premiums. Indeed, we recommend that the FDIC's suggested target reserve range be widened in order to reduce the need to change premiums abruptly. Any floor or ceiling, regardless of its level, could require that premiums be increased at exactly the time when banks and thrifts could be under stress and, similarly, that premiums be reduced at the time that depositories are in the best position to fund an increase in reserves. Building a larger fund in good times and permitting it to decline when necessary are prerequisites to less variability in the premium. This approach stands in favorable contrast to the present system that is designed to stabilize the designated reserve ratios of the funds at the cost of, perhaps wide, premium instability.
In addition to widening the range for the designated reserve ratio, the Board would recommend that the FDIC be given the latitude temporarily to relax floor or ceiling ratios on the basis of current and anticipated banking conditions and expected needs for resources to resolve failing institutions. In short, on stability grounds we prefer less hard-wiring of the rules under which the FDIC operates and more management flexibility for the Board of the FDIC, operating under broad guidelines from the Congress.
4. Rebates. Since its early days, the FDIC has rebated "excess" premiums whenever it felt its reserves were adequate. This procedure was replaced in the1996 law by the requirement that no premium be imposed on well-capitalized and highly rated banks and thrifts when the fund reaches its designated reserve ratio. The FDIC proposals would re-impose a minimum premium on all banks and thrifts and a more risk-sensitive premium structure. These provisions would be coupled with rebates for the stronger entities when the fund approaches what we recommend be a higher upper end of a target range than the FDIC has suggested, and surcharges when the Fund trends below what we suggest be a lower end of a target range.
The FDIC also recommends that the rebates not be uniform for the stronger entities. Rather, the FDIC argues that rebates should be smaller for those banks that have paid premiums for only short periods or that have in the past paid premiums that are not commensurate with their present size and hence FDIC exposure. The devil, of course, is in the details. But this latter proposal makes considerable sense, and the Board endorses it. More than 900 banks--some now quite large--have never paid a premium, and without this modification they would continue to pay virtually nothing, net of rebates, as long as their strong capital and high supervisory ratings were maintained. Such an approach is both competitively inequitable and contributes to moral hazard. It should be addressed.
5. Indexing Insured-Deposit Coverage Ceilings. The FDIC recommends that the current $100,000 ceiling on insured deposits be indexed. The Board does not support this recommendation and believes that the current ceiling should be maintained.
In the Board's judgment, it is unlikely that increased coverage, even by indexing, would add measurably to the stability of the banking system today. Macroeconomic policy and other elements of the safety net, combined with the current, still-significant level of deposit insurance, continue to be an important bulwark against bank runs. Thus, the problem that increased coverage is designed to solve must be related to either the individual depositor, the party originally intended to be protected, or to the individual bank or thrift. Clearly, both groups would prefer higher coverage if it cost them nothing. But Congress needs to be clear about the nature of a specific problem for which increased coverage would be the solution.
Depositors. Our surveys of consumer finances suggest that most depositors have balances well below the current insurance limit of $100,000 and those that do have larger balances have apparently been adept at achieving the level of deposit insurance coverage they desire by opening multiple insured accounts. Such spreading of asset holdings is perfectly consistent with the counsel always given to investors to diversify their assets--whether stocks, bonds, or mutual funds--across different issuers. The cost of diversifying for insured deposits is surely no greater than doing so for other assets. An individual bank would clearly prefer that the depositor maintain all of his or her funds at that bank, and would prefer to eliminate the need for depositor diversification by being able to offer higher deposit insurance coverage. Nonetheless, the depositor appears to have no great difficulty--should he or she want insured deposits--in finding multiple sources of fully insured accounts.
In addition, the singular characteristic of postwar household financial asset holdings has been the increasing diversity of portfolio choices. The share of household financial assets in bank deposits has been declining steadily since World War II as households have taken advantage of innovative attractive financial instruments with market rates of return. There has been no break in that trend that seems related to past increases in insurance ceilings. Indeed, the most dramatic substitution out of deposits in recent years has been from both insured and uninsured deposits to equities and mutual funds holding equities, bonds, and money market assets. It is difficult to believe that a change in ceilings during the 1990s would have made any measurable difference in that shift. Indeed, the data indicate that the weakness in stock prices in recent quarters has been marked by increased flows into bank and thrift deposits.
Depository Institutions. Does the problem to be solved by increased deposit insurance coverage concern the individual depository institution? If so, the problem would seem disproportionately a small bank issue since insured deposits are a much larger proportion of total funding at small banks than at large banks. But smaller banks appear to be doing well. Since the mid-1990s, adjusted for the effects of mergers, assets of the smaller banks, those below the largest 1,000, have grown at an average annual rate of 13.9 percent, almost twice the pace of the largest 1,000 U.S. banks. Uninsured deposits at these smaller institutions have also grown more rapidly than at larger banks--at average annual rates of 22 percent at the small banks versus 11 percent at the large banks, both on the same merger-adjusted basis. Clearly, small banks have a demonstrated skill and ability to compete for uninsured deposits. To be sure, uninsured deposits are more expensive than insured deposits, and bank costs would decline and profits rise if their currently uninsured liabilities received a government guarantee. But that is the issue of whether subsidizing bank profits through deposit insurance serves a national purpose. I might add that throughout the 1990s, small banks' return on equity was well maintained. Indeed, the attractiveness of banking is evidenced by the fact that more than 1,350 banks were chartered during the past decade.
Some small banks argue that they need enhanced deposit insurance coverage to equalize their competition with large banks because depositors prefer to put their uninsured funds in an institution considered too big to fail. As I have noted, however, small banks have more than held their own in the market for uninsured deposits. In addition, the Board rejects the notion that any bank is too big to fail. In 1991, Congress made it clear that the systemic-risk exception to the FDIC's least-cost resolution of a failing bank should be invoked only under the most unusual circumstances. Moreover, the resolution rules under the systemic-risk exception do not require that uninsured depositors and other creditors, much less stockholders, be made whole. Consistent with this view, the market clearly believes that large institutions are not too big for uninsured creditors to take at least some loss. Spreads on large bank subordinated debt are wider than spreads on similar debt of large and highly rated nonbank financial institutions. Indeed, there are no AAA-rated U.S. banking organizations.
To be sure, failure to index deposit insurance ceilings has eroded the real purchasing power value of those ceilings. But there is no evidence of any detrimental effect on depositors or depositary institutions, with the possible exception of a small reduction in those profits that accrue from deposit guarantee subsidies that lower the cost of insured deposits. The current deposit insurance ceiling appears more than adequate to achieve the positive benefits of deposit insurance that I mentioned earlier in my statement, even if its real value were to erode further.
Another argument often raised by smaller banks regarding the need for increased deposit insurance coverage is their inability to match the competition from those large securities firms and bank holding companies with multiple bank affiliates, offering multiple insured accounts through one organization. The Board agrees that such offerings are a misuse of deposit insurance. But, raising the coverage limit for each account is not a remedy since it would also increase the aggregate amount of insurance coverage that large multi-bank organizations would be able to offer. The disparity would remain.
There may come a time when the Board finds that households and businesses with modest resources are having difficulty in placing their funds in safe vehicles and/or that there is reason to be concerned that the level of deposit coverage could endanger financial stability. Should either of those events occur, the Board would call our concerns to the attention of the Congress and support adjustments to the ceiling by indexing or other methods. But today, in our judgment, neither financial stability, nor depositors, nor depositories are being disadvantaged by the current ceiling. Raising the ceiling now would extend the safety net, increase the government subsidy to banking, expand moral hazard, and reduce the incentive for market discipline, without providing any real evident public benefits. With no clear public benefit to increasing deposit insurance, the Board sees no reason to increase the scope of the safety net. Indeed, the Board believes that as our financial system has become ever more complex and exceptionally responsive to the vagaries of economic change, structural distortions induced by government guarantees have risen. We have no way of ascertaining at exactly what point subsidies provoke systemic risk. Nonetheless, prudence suggests that we be exceptionally deliberate when expanding government financial guarantees.