House Banking Committee Testimony 2/16/00

Deposit Insurance

My name is Martin Mayer, I am a Guest Scholar at The Brookings Institution, and I write about financial subjects. I speak here not for any institution or indeed anybody but myself.

Deposit insurance is an important subject, and not, I think, well understood. Professor Milton Friedman and Anna Schwartz considered it the most successful program of the New Deal, and the FDIC under Bill Seidman promoted it as a merit good--a costless activity that promoted the public welfare. Because its obvious function is to pay depositors when the bank can’t, it is usually discussed as a benefit to depositors--and, of course, as a preventative against bank runs. To the extent that the insurance does benefit depositors, they pay for it in the difference between the interest they earn on insured deposits and the interest paid by alternative instruments that are not risk-free. And the systemic benefit from the end of bank runs has a price in the moral hazard that permits these institutions to build up large losses that must be borne by others. The everyday benefits accrue to the stockholders of the depositories, who are able to borrow money from the public more easily and more cheaply. The cost borne by those institutions and stockholders is the irritation of intrusive rules, examination and supervision by various government agencies. I note in passing that the big banks opposed deposit insurance in 1933, and their representative at Congress for that purpose was George Moore, then assistant to the president of National City Bank of New York, later the first chairman of Citicorp, whose memoirs I helped to write. The objection to deposit insurance, Moore said, was that the competence of bankers is not an insurable risk.

Seen through the wrong end of the telescope, as an institution among many, deposit insurance is most usefully regarded as a put option that entitles the owners of the depositories, the purchasers of the option, to sell their assets to the government for whatever price may be required to fill a hole between banks’ deposit liabilities and the market value of the assets on their books. In this analysis, it is not unreasonable that any growth in the insurers’ surplus should be taken into revenues by the government, as option-writers generally take into income the receipts from selling them--or that the government should stand ready to advance money beyond the resources of the insurance fund in times of unanticipated volatility. Options-writers are at greater risk than their algorithms define..

The key fact is that the liabilities of depositories are the currency of the country. As Judge Oliver Wendell Holmes, Jr., wrote in upholding state deposit insurance laws in 1911 (in Noble State Bank v. Haskell, 219 US 111), "The primary object of the required assessment. . . is to make the currency of checks secure." People’s confidence in the currency is a function of their belief that there are more than enough assets in the vaults of the issuer to cover the asserted value of the circulating liabilities. Thus for centuries there was a requirement that the issuer of paper money have enough precious metals in his vaults to cover some significant fraction of the face value of the money. In the United States, the requirement that Federal Reserve notes be backed by gold worth 25% of the total issuance (with gold valued at $35 an ounce) survived until 1968, and it was removed from the law--after frantic efforts by William McChesney Martin--with only a two-vote margin in the Senate.

In an economy where most money is expressed by balances in bank accounts, the government must--this is not a question of "too big to fail," as I had once thought--the government must be prepared to cover bank liabilities. The traditional economists’ argument that uninsured deposits provide a buffer for the insurance fund is simply wrong. In the end, all deposits--defined here as bank liabilities that can be cashed by their holders without reference to market prices--will have to be covered because they too are part of the currency. In the words of Andrew Sheng, who was deputy governor of the Hong Kong Monetary Authority when he said them, the losses of a decapitalized banking system are an implicit fiscal deficit. The deposit insurance funds are a way to reserve some of the capital of the banking system to reduce the likelihood that public moneys will have to be committed when the decapitalization is local rather than system-wide.

And for this purpose--as we somewhat bitterly recognized in the early days of the Bush administration, when the losses of the S&Ls were first admitted--there is no difference between banks and thrifts. Especially since Congress decided in 1980 that thrifts could offer transaction accounts and in 1991 that Prompt Corrective Action should be taken against both money-losing banks and money-losing thrifts before their capital disappeared. Because the rationale for deposit insurance applies equally to banks and thrifts, and because there is now enough money in both BIF and SAIF so that no immediate obligations would be incurred, there is no reason today to maintain separate insurance funds.

It is somewhat more difficult to answer the question of whether the funds should be capped at 1.5% of deposits, with rebates to the insured when the funds are above that level. The number would seem to be correct--the New York Clearing House requires a 1.5% deposit against the maximum uncovered entry permitted to a participant in CHIPS, and the clearing house of the Chicago Mercantile Exchange accepts 1.5% as the minimum gross margin for clearing members.

Rebating to the insured is a little tricky, however, because there can be two views of who owns the fund. In modern economic theory, the residual risk-taker is the owner, and is thus entitled to any windfall gains. The government as the residual risk-taker therefore would seem to have the right to recycle for its own benefit--as part of the debt-reduction drive--the interest on its bonds that it would otherwise be asked to pay to the banks and thrifts as rebates.

Nevertheless, I would support a cap and distribution of the earnings above that cap, for fear that a larger insurance fund would encourage extension of the safety net to non-depository activities of the banks and non-banking affiliates in the financial services holding company. Both Paul Volcker and Gerald Corrigan testified in the 1980s of their concern that banks permitted to engage in commerce would be unable to maintain the firewalls separating the insured institutions from the rest of the enterprise. I know the law intends to separate out the larger group of uninsured activities from the banking activities, but the road to forbearance is paved with such intentions. We also have in the new law this obscure phrase about "complementary to finance"--which I hope this committee will revisit ASAP--and thus a heightened risk of major losses outside the normal banking franchise, with the associated risk that supervisors possessed of a very large insurance fund and terrified of possible publicity of their own negligence will find that depositor preference still leaves enough cloth to hide the dirt.

Back in the days when derivatives were new and small, Mr. Seidman implied that the FDIC would stand behind member banks’ swap obligations, and the recent testimony by Mr. Greenspan and Mr. Summers to the Senate Agriculture Committee seems to argue that they are willing to put the safety net under the derivatives activities of our biggest banks, as they did implicitly in September 1998 in the LTCM fiasco. At best, the "legal certainty" they assert for derivatives pays out the derivative contracts first, in violation of depositor preference. And the fact that no supervisor knows the extent of the open interest in fashionable derivatives contracts written behind closed doors escalates the risk to bank insurers if these instruments are treated as though they were on all fours with repurchase agreements.

The threat that absent "legal certainty" in the United States the banks might take their derivatives business to other countries is hollow: legal certainty in the Cayman Islands, or even in Brussels, is worth very little to beneficiaries of financial contracts here. The law should be rigidly written to prohibit the use of deposit insurance funds to pay off non-deposit liabilities inside the United States and deposit liabilities in off-shore branches and affiliates. Obviously, if we are to use "market discipline" as a mechanism in supervising banks, there must be a total prohibition on the use of deposit insurance funds to pay off subordinated debt.

This does not mean that we must follow a sink-or-swim policy with reference to the what the Fed now likes to call Large Complex Banking Organizations; it means we should rethink the roles of the deposit insurer and the central bank in the new century. Let me put out a thought I am in process of developing: that in the modern world the deposit insurance fund is the lender of last resort to the banking system, because of the role of deposit insurance in maintaining the security of the currency; but the central bank is the lender of last resort to the markets, maintaining liquidity when panic regurgitates offers and swallows bids.

In the course of this process, the deposit insurer may have to borrow from the central bank--as indeed the FDIC did from different Federal Reserve District Banks in the Continental reconstruction and the First Republic rescue--but that borrowing should be a matter of public record, requiring the assent of the Secretary of the Treasury, as FDICIA implies.

Similarly, the central bank, the Fed, pouring money into the system through open market operations to assure the liquidity of the markets, will need help from the charterers and examiners who live closer to the banks themselves. This is not a new story. Benjamin Strong in 1916 recalled his experiences as J. P. Morgan’s point man in the Knickerbocker Trust failure of 1907: "I fear we must calculate that the American banker by and large will do in future emergencies just what he has done on similar occasions in the past. He will gather up every dollar of reserve money that he can lay his hands on and lock it up so tight that the Reserve Banks will never get hold of it until the crisis is past. . . Frankly, our bankers are more or less an unorganized mob." Without Jerry Corrigan and Si Keehn to persuade the banks to support the market-makers in 1987, we would not have got out of the 1987 market crash with so little damage.

That was ad hoc, and I suppose all crisis management is in the end ad hoc. But an established sense of who should be responsible for what would be very useful. The work of financial modernization is by no means complete. We can improve our understanding and planning capacity if we accept the division of function between the deposit insurer and the central bank, recognizing the banks as a special case rather than the central concern in an age where mutual funds alone are a larger source of financial intermediation than all the banks. Deposit insurance will perform its role more effectively if the central bank as an institution concentrates its attention more heavily on the markets, an attention that should include more careful and transparent supervision of the creation and sale by the banks of marketable paper and off-the-balance-sheet derivative instruments.