Skip to main content

The Too Big to Fail Problem

A proposal to reform deposit insurance

September 1, 1997


Gary H. Stern Former President (1985 - 2009)
The Too Big to Fail Problem

"[M]any governments in recent years have found it necessary or desirable, acting as a lender of last resort or otherwise, to directly intervene in their national banking systems to provide liquidity or capital support in the interest of preserving financial stability. As distinctions between banks and other financial institutions become increasingly blurred and their international character more pronounced, the proper role of governmental safety nets and their implications for institutional behavior need

Paul Volcker, July 1997
From Global Institutions, National Supervision and Systemic Risk
A Group of Thirty report

This warning may seem to come at an unlikely time, at least as far as the U.S. banking industry is concerned. Banks are performing well, along with the US economy, and the industry has been stable in recent years—just six bank failures last year and just six in 1995. But trouble may lurk below the surface. Interstate banking restrictions have been lifted and the barriers between commercial and investment banking are starting to fall; US banks are consolidating in record numbers and the size and complexity of our largest banks are growing. While this consolidation and growth may not, in and of itself, be bad, one thing is clear: The loss of just one of these banks will pose an even greater systemic risk than before. Therefore, as Mr. Volcker suggests, the time is right to re-examine the safety net that protects our banking industry.

The last time Congress re-examined this safety net was shortly after the banking crisis of the 1980s. Recall that bank failures in the United States reached about 1,000 between 1986 and 1990, and the banks' deposit insurance fund teetered on the edge of insolvency. Indeed, the Federal Deposit Insurance Corp. had to request a $30 billion line of credit from the US Treasury to remain solvent. And, of course, the problems in the savings and loan industry were even greater, where the virtual failure of the industry bankrupted the Federal Savings and Loans Insurance fund and cost the American taxpayer over $100 billion.

It took a crisis of such huge dimensions to spur a policy response, and in 1991 Congress passed the Federal Deposit Insurance Corp. Improvement Act (FDICIA). The intent of this law, for the most part, was to redesign the federal safety net and the power of federal banking regulators in the hopes of preventing another S&L-type crisis. Here we argue that while FDICIA was a step in the right direction, it stops short of bringing market forces to bear on our largest banks. If the United States is to avoid another banking crisis, uninsured depositors must in fact be uninsured.

Why is it important that uninsured depositors bear some of the risk in a bank failure? Because having uninsured depositors bear some of the risk is a key ingredient for mitigating a costly side effect with deposit insurance, a costly side effect known as moral hazard: Once a person or institution is insured, the insured has an incentive to take on more risk than otherwise. This is why, for example, fire and auto insurance policies have deductibles. With deductibles, the insured have more incentive to pick up old paint cans and drive cars more carefully.

The moral hazard problem is particularly severe in banking because of the lack of deductibles. Governments often provide 100 percent depositor protection, especially at large banks where a loss could have industry wide repercussions (a practice known as too big to fail—TBTF).

In 1991 the US Congress created an explicit and more stringent TBTF policy. Prior to 1991, there was an unwritten TBTF policy, and the government was much freer to rescue large banks and protect uninsured depositors. Under FDICIA, uninsured depositors cannot receive protection if it raises costs to the FDIC. There is an exception, however, if the failure of the bank poses a systemic risk. The emergency bailout can go forward with the approval of the Secretary of the Treasury (who must consult with the President), the Federal Reserve Board of Governors and the FDIC's board of directors. Therein lies the problem: FDICIA still leaves the door open for moral hazard in the extreme. Uninsured depositors at the very largest US banks still know they are likely to be fully protected and have little reason to monitor how their deposits are invested. And, as a result, large banks have an incentive to take on more risk than they would otherwise.

Congress recognized the problem of moral hazard and tried to make the protection of uninsured depositors less likely in FDICIA by giving regulators new tools to close a troubled bank before large losses ever develop. In theory, these tools should work. Closing banks before they are insolvent eliminates moral hazard because the insured cannot take advantage of the insurance fund. In practice, however, closing banks before they are insolvent is like stopping forest fires. Many sound precautions can be taken, but conflagrations will always occur. For large complex banks, financial fires are likely to be burning for some time before regulators become aware of the fire.

Our recommendation is that when a TBTF bank is rescued, uninsured depositors cannot be fully protected. Specifically, we recommend that uninsured depositors can recover only 80 percent of their deposits or the market value of their deposits, whichever is larger. Because depositors' losses would be capped, this policy would limit the systemic risks. At the same time, it would lift the de facto risk-free curtain that shields large banks and would force depositors to take a good look inside. With up to 20 percent of their deposits over $100,000 at risk, it would behoove depositors to pay attention to what banks are doing with their money, and, likewise, such a plan would cause banks to more carefully conduct business. Banks would not only have to compete on the basis of price and service, but also on safety and soundness.

And this competition would likely spread beyond big banks. By inducing depositors to look past a bank's size to its fundamentals, depositors would probably begin looking at smaller banks with good reputations, thereby spreading this incentive throughout the industry and, in theory, making the whole industry more safe and sound than it would be otherwise.

FDICIA did not go far enough in addressing the moral hazard problem. The recent success of the banking industry should not lull us into a false sense of security. Another banking crisis is possible and this time—with ever larger and more complex banks—the costs could be even greater. The time is right for uninsured depositors to become truly uninsured, and thus reduce the insidious effect of moral hazard in the banking industry.

Related content

Special study: Too Big To Fail