It has been said that experience is a tough teacher -- first you get the test, then you learn the lesson. In the banking crisis of the 1980s and early 1990s, banking regulators were tested, and from their experience they learned lessons. Did we learn the correct lessons?

When I became FDIC Chairman, I initiated a project to find the answer to that question, an answer based on objective analysis. The result is a series of 14 papers we will publish over the coming year. The purpose of this symposium is to discuss the first three papers in this series, which focus on supervisory issues.

This afternoon, I want to focus my remarks on the role of Federal deposit insurance in our banking and financial system. One of the lessons of the 1980s and early 1990s is that deposit insurance was eminently successful in maintaining stability in the banking system during the crisis. A second lesson is that this success came at enormous cost to the insurance funds, to the taxpayer, to the surviving institutions, and to their customers.

Our experience in the crisis reminded us that guaranteeing savings can be a costly business, although it may be necessary to stabilize the banking system in times of stress to prevent runs on individual banks from spreading to become banking panics.

One event toward the end of the most recent banking crisis underscored how quickly public confidence can evaporate -- and the importance of deposit insurance in maintaining public confidence in the banking system.

In early 1991 -- just six years ago -- the New York Times described recent events at the Bank of New England in this way:

"Frantic depositors pulled nearly $1 billion out of the bank in two days; small savers trooped through the lobbies with their money in wallets, bulging envelopes and briefcases, and money managers yanked out multimillion-dollar deposits by remote control with computer and telex orders.

"Some local crooks even tried to get in on the action. The Federal Bureau of Investigation said it foiled a plan by six men who had hoped to rob an armored car they figured would be loaded with cash for all the withdrawals."

The New York Times story concluded: "Yet as soon as Washington stepped in, with the Federal Deposit Insurance Corporation taking over the bank on Sunday, the panic subsided."

Bank of New England customers may have had doubts about their bank -- but their doubts were not contagious. Because Federal deposit insurance maintained public confidence in the banking system, a run on the Bank of New England did not spread to other banks or into a general banking panic, with depositors at other banks demanding their funds, too.

How costly was this protection?

From 1980 to 1994 -- 1,617 banks failed or received financial assistance from the FDIC. These failures severely tested the FDIC insurance fund. During the same period, nearly 1,300 savings and loans failed. These failures more than bankrupted the old savings and loan insurance fund and directly cost the taxpayers of America $125 billion, and billions more in indirect costs.

As a result of the experience of the 1980s and early 1990s, deposit insurance has become part of the continuing debate on how the banking industry should be modernized -- and at the center of the discussion of deposit insurance is the problem of moral hazard.

The problem of moral hazard occurs when insurance induces the insured to take more risk than they would take if they were not insured. Any deposit insurance fund -- any form of insurance, in fact -- faces the problem of moral hazard. With deposit insurance, the insured party is the depositor. Insurance permits insured depositors to ignore the condition of their institutions. Even fundamentally unsound institutions may have little difficulty obtaining funds. Because insured depositors may no longer have an incentive to monitor and discipline their institutions, the managers of those institutions may take more risks than they otherwise would. In short, deposit insurance can create opportunities for managers to make high risk/high return investments, without the market discipline of having to pay creditors to take that risk.

Moreover, as the paper that is the subject of the next panel states: "with respect to the basic trade-off between promoting stability and controlling moral hazard, bank regulators (in the 1980s) showed a preference for solutions that tipped the balance toward stability, a policy that was apparent in the treatment of large-bank failures.

"This contributed to the success of the deposit insurance system in avoiding bank runs and disruptive interruptions in credit flows. . . . (But) By protecting uninsured depositors, the methods used to resolve large-bank failures removed a source of market discipline that could have reinforced supervisory efforts to constrain risk."

To inform the debate over deposit insurance in the context of modernizing the banking charter, today I want to ask a number of fundamental questions, beginning with a question based on our experience in the 1980s and 1990s.

Did the problem of moral hazard created by Federal deposit insurance lead to a large number of failures of insured institutions in the 1980s and early 1990s, causing massive losses in the insurance funds?

To answer that question, moral hazard has to be broken into two components. First, in the case of a solvent institution, deposit insurance may lessen -- or may eliminate -- the incentive for insured creditors to monitor the activities of management and owners. Second, where a banking organization is insolvent or is approaching insolvency, deposit insurance may provide an incentive to bank management to take abnormal risks, thus magnifying losses to the insurance fund.

For the thrift industry in the 1980s, moral hazard contributed to huge losses in the savings and loan insurance fund, to its ultimate failure, and to substantial costs to the taxpayer. What began as an asset/liability mismatch problem, aggravated by rapidly rising interest rates in the beginning of the decade, became an enormous credit problem as real estate markets collapsed.

Weak regulatory oversight and the lack of resources to close insolvent thrifts encouraged some institutions to speculate widely in real estate and other ill-conceived efforts to "grow" out of their problems.

For banks in the 1980s and early 1990s, the role that moral hazard played in the significant losses to the insurance fund is not as clear. Certainly, deposit insurance did remove the incentive for insured creditors to monitor a bank's activities, but its effect is difficult to measure.

Moreover, the moral hazard that arises when banks approached insolvency and owners had less and less at stake was effectively restrained to a much greater extent than was the case with savings and loans by supervision of problem institutions. As will be detailed this afternoon, this restraint is indicated by the dividend, capital, and asset growth behavior of problem banks at that time.

Higher prudential standards for banks and more immediate regulatory attention to serious problems -- as well as a solvent bank insurance fund with the resources to solve problems as they were identified -- accounted for the difference in the experience of banks and thrifts. This difference should inform the debate over the role of deposit insurance in banking's future.

As this debate has developed, two alternatives to the current system have been offered: The first is to privatize the deposit insurance system. The second is to reduce the scope of the current system, and thus rely more on the markets to discipline the banking system. The two alternatives are, of course, not mutually exclusive.

Let's briefly analyze these proposals by seeking answers to four questions: One, what led Congress to make deposit insurance the responsibility of the Federal government? Two, can deposit insurance effectively be provided by another supplier? Three, how much less than the equivalent of a "full faith and credit" pledge by the Federal government will the public accept? Four, would reducing the scope of the deposit insurance system bring positive results?

First, what led Congress to make deposit insurance the responsibility of the Federal government?

Recurring and worsening banking panics marked the history of banking in the United States until the creation of the Federal Deposit Insurance Corporation in 1933. Nine thousand banks suspended operations from 1930 through 1933. The year after the FDIC was created, nine insured banks failed.

Even though the banking crisis of the 1980s and early 1990s resulted in a dramatically high number of bank failures, there was no banking panic -- no contagion that could have threatened sound banks -- and public confidence in the banking system held steady.

Today the banking industry is healthy and the economy is strong. Because the memories in good times can be short, it is important to remember the lessons of history.

It was the historical experience in the 1930s that has led a broad range of economists to conclude that Federal deposit insurance solved a problem that had plagued the banking system -- and the economy -- for more than a century, the problem of maintaining public confidence in a banking system marked by liabilities that were liquid and assets that were illiquid.

For example, in his The Great Crash, 1929, John Kenneth Galbraith observed: "Federal insurance of bank deposits, even to this day, has not been given full credit for the revolution that it has worked in the nation's banking structure. With this one piece of legislation, the fear which operated so efficiently to transmit weakness was dissolved. As a result one grievous defect of the old system, by which failure begot failure, was cured. Rarely has so much been accomplished by a single law."

In their A Monetary History of the United States, 1867-1960, Milton Friedman and Anna J. Schwartz similarly laud the role of deposit insurance in stabilizing the banking system: "Federal insurance of bank deposits was the most important structural change in the banking system to result from the 1933 panic and, indeed in our view, the structural change most conducive to monetary stability since state banknote issues were taxed out of existence immediately after the Civil War."

More recently, Federal Reserve Board Governor Janet Yellen, who has been nominated to become Chairwoman of the Council of Economic Advisors, addressed the issue also by reminding us of history. She said: "Deposit insurance was introduced both to protect individual depositors and to prevent panics surrounding individual banks from spreading throughout the financial system.

"Would we be better off as a country giving that up?" Governor Yellen asked rhetorically. "I don't think it is obvious that we would be. We would have to think through very carefully what implications the reduction or elimination of deposit insurance would have for systemic risk. The Depression taught us a lesson."

These tributes to Federal deposit insurance, however, do not address the question of whether a supplier other than the Federal government can provide essential depositor protection. In answering that question, the experience of private and state insurance providers in the banking crisis of the 1980s and early 1990s should give us some guidance.

As recently as 1982, there were 32 deposit insurance funds in operation. Only eight survived the crisis. Six operate today, three cover state credit unions and three are very limited in scope or are being phased out. Almost all the other funds collapsed because of the failure of one or more institutions. Most of the funds were state-sponsored, although the state did not usually provide any financial guarantees to the fund. These funds typically were mutual insurance funds with a board of directors drawn from the insured institutions.

In response to the failure of state deposit insurance plans in Ohio and Maryland, those states required state-chartered institutions to obtain Federal deposit insurance. Approximately 150 institutions were added to FDIC coverage in 1985 as a result.

Federal Reserve Chairman Alan Greenspan has observed: "Confidence in the stability of the banking and payments system has been the major reason why the United States has not suffered a financial panic or systemic bank run in the last half century."

It is my belief that deposit insurance can help maintain stability in the banking system only if depositors have confidence in the insurance plan. To inspire confidence during a period of turmoil, deposit insurance must be a certainty for the insured depositor.

The experience with private and state-sponsored insurance plans in the 1980s and early 1990s suggests that the limited pool of resources on which they can draw inspires less confidence than does the unlimited pool of resources of the Federal government. Bank failures may come in waves, because the performance of the industry is closely tied to the performance of the economy.

While it may be possible to design private insurance funds that could handle isolated failures successfully, our experience in the 1980s in Ohio and in Maryland suggests that limited plans have difficulty handling failures in waves.

Further, if private insurance is substituted for Federal deposit insurance, a private insurance plan facing depletion of its fund during a crisis would likely have to seek financing from the banking industry or other private sources of funds at the same time that the economy may be weak and the banking industry is having difficulties. Moreover, if the private insurance supplier fails, the Congress may have to act to restore public confidence. That would take time, and based upon the experience during the savings and loan crisis, Congressional action might occur only after serious damage has been done and costs have been significantly increased.

In considering privatizing Federal deposit insurance, therefore, the serious question becomes: How much less than the equivalent of a "full faith and credit" pledge by the Federal government will the public accept -- in other words, how much less would fully protect the banking system in times of crisis? We do not know the answer, but history suggests that we cannot predict the depth or duration of a crisis, and that should make us wary.

The final question I want to ask today is: Would reducing the scope of deposit insurance bring positive results?

In this regard, one observer, former FDIC Chairman Bill Isaac, recently wrote in the American Banker: "What's needed is more private-sector discipline. This will come about once the scope of depositor protection is curtailed sharply, including abandonment of the too big to fail' doctrine. Millions of organizations and sophisticated individuals must be given the incentive to understand, monitor, and control the risks in the financial system."

I agree that market discipline is an important element of a sound deposit insurance system. Our goal is to assure the stability of the banking system in times of great stress, not to eliminate all bank failures. An effort to eliminate all bank failures would involve over-regulation of banks that would lessen their effectiveness in providing financial intermediation in the economy.

The question remains: What has been done to encourage market discipline and what more can be done? I will discuss these issues more in a moment, but first let's consider the issue of the scope of deposit insurance.

In terms of insuring individual deposits, the scope of coverage increased until 1980 and then declined, in terms of today's dollars. Let me explain.

As of January 1, 1934, the FDIC insured deposits up to $2,500. In 1996 dollars, however, that $2,500 is the equivalent of $30,000 today. Six months later, the insurance limit was raised to $5,000, which is almost $60,000 in today's dollars. In 1969, the limit was raised to $20,000, which is about $85,000 in today's dollars. When the limit was raised to $40,000 in 1974, that was the equivalent of $127,000 today.

From its very beginning, deposit insurance covered more than just the average American's "food and rent" money -- is was sufficient to cover some savings.

Moreover, depositors today are insured up to $100,000 -- a limit that has been in place since 1980. The dollars of 1980 are not the dollars of 1996, however -- $100,000 in 1980 was the equivalent of $190,000 today. In this sense, for the individual depositor, the scope of deposit insurance coverage has declined by almost half since 1980.

I am not advocating any change in the level of today's coverage for deposits -- the marketplace has already done that. Of course, the other side of the scope of insurance coverage is uninsured depositors and the so-called too big to fail' doctrine, as Bill Isaac points out. The Federal Deposit Insurance Corporation Improvement Act, however, significantly reduced the authority regulators have to deal with large institutions that are failing. It leaves us with enough flexibility, with appropriate oversight by Congress, to achieve a solution where the failure would present a genuine risk to the system. This can occur, however, only if the Secretary of the Treasury -- in consultation with the President -- determines that there would be "serious adverse effects on economic conditions or financial stability." Such a decision would be undertaken only after written favorable recommendations from both the FDIC Board of Directors and the Board of Governors of the Federal Reserve System, with at least two-thirds of the members of each body voting in favor of the recommendation. That is a high standard, particularly when one considers that the recommendation would have to be defended to the Congress.

Moreover, the FDIC has been required by law since 1991 to accept the proposal from a potential purchaser that is the least costly to the insurance fund of all the proposals we receive. In more than half of the failures in 1992 -- 66 out of 120 -- uninsured depositors received less than 100 cents on each dollar above the $100,000. That was a significant increase in uninsured depositors experiencing losses from 1991, when fewer than 20 percent of the failures involved a loss for uninsured depositors. While the number of bank failures in 1992 was lower than in previous years, the number of uninsured depositors experiencing a loss was significantly greater. Moreover, as the paper that is the subject of the next panel points out, resolution with losses to uninsured depositors have not produced large-scale withdrawals at other institutions -- though, in the years since 1992, with record levels of bank profits, failures slowing to a trickle, and no major bank threatened with failure, the system has not come under stress.

Further, I would ask, can depositors be expected to impose market discipline on banks? After all, it was this approach that led to the recurring banking panics that marked most of our history until 1933. A number of years ago, banking analyst Karen Shaw Petrou concisely described why the Congress created the FDIC to benefit individual Americans: "After the collapse of the early 1930s, it was agreed that individual savers should have a protected right to place a limited amount of money in a financial institution without having to worry that it could be lost. Individual depositors should not have to read a detailed report of a bank's condition before deciding where to deposit their retirement or other savings, since most depositors would be hard pressed to interpret such information. To solve the problem, the government took upon itself the obligation to interpret the financial condition of banks for depositors, and to back up its judgments with limited federal deposit insurance."

In standing in the place of the depositor, banking supervisors seek to mitigate the problem of moral hazard created by Federal deposit insurance through examinations and safety-and- soundness regulations.

The challenge to the regulators is to develop safety-and- soundness regulation that comes as close as possible to market discipline, without imposing inefficient, ineffective regulations on banks, regulations that unduly inhibit the important function of financial intermediation that they perform for the economy. Market discipline, however, does have a critical role in addressing the problem of moral hazard that deposit insurance creates -- that discipline, however, can perhaps more effectively be imposed by large creditors and shareholders of banks.

At least since the least cost test has been imposed on the FDIC, large creditors should understand the potential for losses on their exposures to banks. That was perhaps less true with respect to earlier large-bank resolutions. In addition, over the past few years, we have undertaken two reforms in deposit insurance that give shareholders a greater incentive to curb excessive risk taking at their institutions: one is higher, risk-based, capital standards; the second is risk-related insurance premiums.

Higher risk-based capital standards expose shareholders of an institution to greater loss, and risk-based standards expose shareholders to greater loss as the institution's risks increase. Not incidentally, the regulations that put higher minimum capital standards into effect impose restrictions on dividend payments and other capital distributions if an institution falls below the minimum.

Similarly, risk-based premiums are designed to reduce income in institutions that take on excessive risk, and that reduction in income is aimed at giving shareholders reason to curb the excesses. As you know, in 1993, the FDIC established risk-based deposit insurance premiums. Banks and thrift institutions were divided into nine groups, depending upon the risks they present to their insurance fund.

Part of that risk calculation is based on capital and part on supervisory factors such as asset quality, loan underwriting standards, and management. We are now analyzing whether other factors are relevant to risk -- and whether our current 27-basis point spread is sufficient to price the risks to the insurance fund posed by individual institutions. Making it more costly for banks to take on excessive risk will impose more economic discipline on their judgments.

In conclusion, in the 1980s and early 1990s, deposit insurance helped maintain financial stability, but at great cost, particularly with respect to the savings and loan industry.

We should learn from that experience.

Those lessons could lead us to continue to improve our current Federal deposit insurance system -- as we have begun to do -- to make it more sensitive and responsive to the marketplace -- finding even better regulatory surrogates and incentives for marketplace discipline.

Some say that those lessons should lead us to replace the current system with a privatized approach. But before we take that course, we should agree on the answers to the questions: What would happen if there were no Federal deposit insurance program? Can a supplier other than the Federal government bear the costs necessary to provide deposit insurance coverage sufficient to maintain stability in the banking system in times of extreme stress? How much less than the equivalent of a "full faith and credit" pledge by the Federal government will the public accept?

Without firm answers to those questions, in privatizing Federal deposit insurance we may be putting the banking system at risk. We know Federal deposit insurance works to stabilize the banking system in times of great stress. Can we be sure that another approach will work as well?

Thank you.