Statement to the Congress.

Statement by Alan Greenspan, Chairman, Board of Governors of the Federal Reserve System, before the Subcommittee on Financial Institutions and Consumer Credit of the Committee on Banking and Financial Services, US. House of Representatives, August 2, 1995

I am pleased to be able to appear here today to offer my thoughts on the status of the Savings Association Insurance Fund (SAIF) and on deposit insurance more generally.

The combination of deposit insurance and a central bank providing discount window credit has made the contagion of bank runs that often characterized the nineteenth century and the first third of the twentieth century an anachronism. The United States has not suffered a financial panic or systemic bank run in the past fifty years. In large part, this reflects the safety net, whose existence, as much as its use, has helped to sustain confidence.

But deposit insurance is not without its costs. By relieving depositors of the consequences of bank failure, government guarantees of bank deposits make depositors relatively indifferent to bank failure and thus encourage banks to have larger, riskier asset portfolios than would be possible in a wholly market-driven intermediation process. Without the safety net, additional risks would have to be reflected in some combination of higher deposit costs, greater liquid asset holding, or a larger capital base, and these, in turn, would constrain risktaking. In the late 1980s and early l990s, the Congress responded to problems at insured depository institutions - and their insurance funds - with legislation designed to induce these entities to be more prudent risk-takers.

Today, we are here to address an evolving competitive imbalance and other implications of two insurance funds with sharply different premiums. But it is critical to underline that even if there were no evolving problem with SAIF, the existing deposit insurance system, with its reliance on two funds, would be inherently unstable.

With deposit insurance, as it is currently administered and funded, depositors do not move their funds from depository institution to depository institution based on the soundness of particular insurance funds. Depositors are generally unaware, and indeed should be unconcerned, about the Bank Insurance Fund (BIF) versus SAIF. In the mind of the typical depositor, the Federal Deposit Insurance Corporation (FDIC) provides the insurance, and the details of one fund versus another receive little attention.

Competitive depository institutions cannot differentiate themselves by the quality of the deposit insurance that is offered because it is the same insurance regardless of whether it is from BIF or SAIF. In either case, it is government-mandated and government-sponsored deposit insurance. For identical insurance, it is rational that depository institutions seek the one available at the lowest cost. If a substantial difference in deposit premiums exists between SAIF and BIF, the institutions paying the higher premium will pursue insurance offered by the other insurance fund unless there is some other reason to remain with their current fund.

Although today we are discussing what to do about SAIF, I want to stress that the problem we are addressing is a general one. If there is no substantial difference between BIF and SAIF insurance and if there is no substantial difference between the advantages granted to BIF institutions or SAIF institutions, then anytime one deposit insurance fund has difficulties that result in substantially higher deposit premiums, members will try to shift to the other deposit insurance fund. In the process, the disadvantaged fund becomes increasingly vulnerable to insolvency as its premium base declines. This, in turn, engenders a still greater incentive to leave the troubled fund or requires the payment of still higher premiums to support it. Short of effective barriers to exit, once initiated the downward spiral does indeed lead to fund insolvency. Thus, having two deposit insurance funds creates a mechanism that is prone to instability now, and probably in the future. Today, the problem is at SAIF; it may, at some date in the future, be at BIF.

The Congress can attempt to legislate barriers that try to stop institutions from shifting deposits, but the history of efforts to legislate against such strong financial incentives is not encouraging. We are, in effect, attempting to use government to enforce two different prices for the same item - namely, government-mandated deposit insurance. Such price differences only create efforts by market participants to arbitrage the difference. In the present case, with SAIF institutions expected to pay at least five times more per year for the same deposit insurance, this arbitrage means that SAIF institutions will pursue all avenues open to them profitably to move deposits from SAIF to BIF.

The difference between paying, say, 24 basis points and paying 4.5 basis points for deposit insurance translates into about $1.4 billion per year in additional premiums paid for SAIF deposits. For SAIF institutions, this equals roughly 18 percent of their 1994 pretax income. Given the large potential financial gains to SAIF institutions if they move deposits to BIF, the current deposit insurance system will impose a large dead-weight loss on the financial system. Many of the political, policy, financial, and legal institutions concerned with banking issues will be preoccupied, for the foreseeable future, with the details of this issue because SAIF institutions will continually strive to move deposits into BIF and BIF institutions will attempt to thwart such movements.

Indeed, BIF institutions suffer under the current system to the extent that SAIF members successfully shift their deposits to BIF. One way for a SAIF institution to minimize its cost under the current system is for that institution either to acquire or to be acquired by a BIF institution. The SAIF institution can be funded from nondeposit sources, while its depositors are encouraged to shift funds to the BIF institution. Current BIF members would almost surely find their premiums higher than otherwise because the new BIF deposits come without the associated insurance fund reserves, requiring older BIF deposits to pay a higher assessment to maintain the required 1.25 percent reserve ratio on both the new and the old deposits.

Using the FDIC's projections of future deposit premiums, a migration of only $40 billion to $50 billion per year of SAIF deposits to BIF deposits might yield higher deposit premiums for existing BIF members than if those members were to participate in any of a number of proffered solutions to the potential SAIF problem, each of which would remove incentives to migrate. Such a shift of deposits seems entirely credible if a large deposit premium difference exists between SAIF and BIF, since $50 billion amounts to only 7 percent of the existing SAIF assessment base. Furthermore, even this relatively small migration suggests that payments of The Financing Corporation (FICO) bond interest funded by SAIF could be put in jeopardy in the very near future. If action is not taken shortly, a future congressional appropriation for interest on FICO bonds might be required, or further increases in SAIF premiums on the smaller SAIF deposit base might be necessary, or possibly even the imposition of higher premiums on both SAIF and BIF deposits might be needed.

Meanwhile, SAIF institutions will be harmed directly by the continuation of a deposit premium higher than that to be assessed on BIF members, and the returns on capital of SAIF members will be driven lower than similarly situated competitors. As I noted, BIF institutions will be harmed by the inflow of new deposits shifted from SAIF institutions requiring BIF members to pay higher premiums. The only winners created by the looming deposit premium difference between SAIF and BIF deposits will be those depositories able to "game" the system and leave SAIF first. The solution to this problem is to end this game and merge SAIF and BIF.

A prerequisite is to put SAIF on a sound basis. This could be accomplished if, as has been recommended, the institutions that hold SAIF deposits pay a special one-time assessment to recapitalize SAIF at the legally mandated 1.25 percent ratio of insured deposits. Such a one-time charge is large: SAIF member institutions would pay as much as $6.6 billion, or 85 to 90 basis points of their deposit base. This assessment seems unlikely, however, to drive healthy SAIF members into insolvency, and weaker SAIF institutions can be allowed a longer pay-in period. The merging of a recapitalized SAIF with a sound BIF would then consolidate the FICO bond obligation of SAIF into the new insurance fund and effectively obligate past BIF members to participate on a pro rata basis.

Most bankers would argue, with some justice, that they should not be responsible for this legacy of the thrift crisis in which they played no role. Many may, nonetheless, conclude that two or two and one-half basis points per year in additional deposit premiums for the FICO interest payments may be a price they would willingly pay to finally remove the incentives of SAIF members to shift to BIF, with the associated increase in the premiums of BIF members.

Even after SAIF is recapitalized, in the years immediately ahead some large savings and loan associations, still suffering from the residue of past difficulties, may continue to represent a risk of relatively large loss to their federal deposit insurer. If SAIF were not merged with BIF, or if that merger were delayed, the risk of such loss would expose a recapitalized SAIF both to a reserve shortfall and to a higher deposit insurance premium to once again rebuild its reserves. An industry that had just paid a large one-time premium to recapitalize its insurance fund would be understandably concerned about that possibility. If such losses were to occur to a merged BIF-SAIF fund, the necessity of reserve-building would be shared among banks and thrift institutions pro rata - implying a larger dollar burden on the larger commercial bank industry. Banks would be understandably concerned about such exposure, especially after accepting a pro rata share of the FICO interest obligation.

Both sets of institutions are thus sensitive to the small probability of a large thrift institution's failure imposing still further costs on them. One way to address these concerns is for the Congress to arrange a catastrophe contingency funding arrangement over, say, the next five years to bridge the period over which this risk exists. It has been suggested, for example, that over such an interval public funds be made available in any year that losses to SAIF, or losses created by current SAIF members to a merged BIF-SAIF, exceed $500 million. If increased budget outlays are, with good reason, not acceptable to the Congress, one possibility is that this catastrophe insurance be financed through a small special insurance fee, paid to the Treasury by SAIF members to cover the potential taxpayer risk exposure.

Discussions about merging BIF with a recapitalized SAIF insurance fund and sharing the FICO interest obligation among the members of both deposit insurance funds raise the question of retaining separate bank and thrift charters. It is difficult to overstate the importance of savings and loan associations in the financing of the residential housing market in the first two decades after World War II. Their continued demonstration to other lenders of the basic credit quality of the lower downpayment, long-term, conventional, amortized, residential mortgage instrument revolutionized housing finance. Their success led public policymakers to look at thrift institutions as innovators operating at the cutting edge of the market. But beginning in the 1970s, market forces and innovations began to erode the original purpose of specialized thrift institutions and, hence, their charter. The development of mortgage-backed securities - along with the technological revolution facilitated by the computer - has lessened the special franchise of thrift institutions by creating a secondary market for most residential mortgages. As a result, the standard residential mortgage no longer requires specialized financial institutions to originate or fund these instruments.

So far in this decade, savings and loan associations and savings banks have originated 25 percent of residential mortgages - compared with 50 percent over the previous twenty years - and hold, on average, only 28 percent of outstanding residential mortgage debt, compared with two-thirds during the earlier period. Currently, only two thrift institutions are among the top fifteen mortgage servicers and none are among the top ten originators. Over the past decade, when thrift institutions' participation in the residential mortgage market receded, the aggregate supply of housing finance was unimpaired and mortgage rates apparently unaffected. Indeed, events over the past decade suggest that market forces and innovations have reduced the relative yield on the standard residential mortgage, while at the same time other market forces have made deposit rates increasingly competitive. In such an environment, significant questions are raised about the economic viability of any institution that by law or regulation is required to place most of its assets in mortgage instruments and fund them in the deposit market.

Two conclusions are clear. First, the nexus between thrift institutions and housing largely has been broken without any detriment to housing finance availability. Second, a public policy that induces - let alone requires - thrift institutions to specialize in mortgage finance threatens the continued viability of many of these entities - particularly those without wide and deep deposit franchises, tight cost controls, and the ability, when necessary, to effectively originate and sell standard mortgages that cannot profitably be held long-term. A broader charter for thrift institutions - such as a commercial bank charter that lets them hold a wider range of assets - thus would seem to be good public policy.

Even if such modifications of the thrift charter are not adopted, but especially if charter changes are made, serious consideration ought to be given to reevaluating tax rules that not only induce mortgage specialization but penalize thrift institutions that try to adopt more diversified portfolios. The special bad debt reserve treatment that provides tax benefits - and, hence, subsidy - to mortgage lending by thrift institutions should be considered for removal going forward. In addition, consideration should be given for grandfathering the reserve buildup from this past tax subsidy in order to remove it as a barrier for entities that wish to diversify. A penalty should not be charged institutions that are striving to respond rationally to market realities.

Charter changes and adjustment of tax policies will not mean that all, or even most, thrift institutions will give up mortgage originations and portfolio holdings. These entities have, over the years, built up special skills that they will continue to use in mortgage finance. Some - those with strong cost controls, greater expertise, and the ability to respond to changing market conditions - will probably continue to be strong, profitable, and viable institutions specializing mainly in mortgage credit. But long-run health for the thrift industry as a whole, I think, requires that most cannot be mortgage specialists to the same degree as in the past, and good public policy must, at a minimum, drop those provisions that penalize diversification for those that choose to do so.

Let me conclude by clarifying why the Federal Reserve is concerned about the SAIF problem and believes it is necessary to resolve it. The Federal Reserve's primary concerns are sustainable economic growth and financial stability. A healthy and competitive financial system is critical for maintaining and promoting economic growth. One key component of a healthy financial system is a sound depository institution system, and an important component of a sound depository institution system is that depository institutions are not given artificial incentives to switch between insurance funds or to abandon an insurance fund to gain competitive advantages. Such "regulatory arbitrage" wastes scarce and valuable resources that could be much more productively employed.

Furthermore, as we know from our experience in the last recession, uncertainties about the resolution of insurance fund failures, and the regulatory policies needed to protect the taxpayer while these uncertainties are resolved, can only inhibit the willingness of depository institutions to lend. While there were many reasons monetary policy encountered strong headwinds during that period, surely the legislative and regulatory reactions to the taxpayer funding of the thrift deposit insurance fund and to the depleted nature of BIF compounded our problems.

Whatever solution is finally adopted, we should not lose sight of first principles. A deposit insurance system that focuses the attention of banks and thrift institutions on the relative status of their funds, and a system that rewards those who can jump ship first, is, to say the least, counterproductive. What is needed is a deposit insurance system whose status is unquestioned, so that the depositories can appropriately focus their attention on the extension and management of credit in our economy. I might also add that a congressional decision to provide a more banklike thrift charter and banklike taxation would be consistent with market trends and stronger depositories, and would not be likely to reduce mortgage credit flows.