09/30/2013 11:49 am ET Updated Nov 11, 2014

Is the "Too Big to Fail" Problem Too Big to Solve?

There seems to be almost universal consensus that using public funds to protect large institutions from failure, commonly called "bailout," is bad policy. There is nothing like a consensus, however, on what should be done about it, and execution seems to be floundering. Three approaches have emerged, only one of which has much chance of being successful.

Approach 1: Commit That Bailouts Will Never Happen Again
Under this approach, the government adopts a policy that it will never again rescue a major bank faced with impending failure, regardless of what the consequences might be. This seems to be the favored policy position of those who have not thought through all the implications.

The problem with this approach is that so long as we have government agencies and public officials with responsibilities for promoting economic growth, price level stability and high employment, it cannot be implemented. The public officials who made the bailout decisions during 2007-9 were forced to choose between using public funds to bail out an imprudent institution, or allowing the failure of that institution to destroy hundreds of innocent firms and the jobs of thousands of innocent workers. They properly chose the bailout as the lesser evil. If public officials ever have to face that horrible choice again, we will want them to make the same decision.

Approach 2: Prevent Systemically Important Firms From Failing by Imposing High Capital Requirements

This is the Dodd-Frank approach that Federal agencies and legislators are now attempting to implement. Under this scheme, regulators would tag as "systemically important" (SI) every financial firm that is so large and inter-connected with other firms that its failure would destabilize the world's financial system. Since failure results from losses that exceed a firm's capital, SI firms would be subjected to capital requirements high enough to absorb the losses that might occur under the worst circumstances. Just as the ocean liner Titanic was built to be unsinkable, SI firms would be made unfailure-able. The analogy is apt, even if the underlined word doesn't yet exist.

The first step in this approach is to identify SI firms, a process that has already started. Under Dodd-Frank, a super-committee of regulatory agencies has been compiling a list of banks and other major firms that are systemically important. While this requires some tough decisions, particularly as it applies to firms other than banks, it is clearly doable.

What is not doable is using capital requirements to reduce the risk exposure of the SI firms, to the point where these firms could survive any economic shocks to which they might be subjected. The problem is that capital requirements can be gamed by the SI firms subjected to them, meaning that they can respond to capital requirements designed to reduce their risk exposure by increasing their exposure in other ways not barred by the regulations. I will discuss this in more detail next week.

The SI firms have a very strong incentive to game regulations designed to reduce their risk exposure. These are profit-making firms that must compete with foreign firms in multiple markets that involve taking risk, and their competitors may not be subject to the same degree of regulatory oversight.

In principle, regulators can prevent SI firms from gaming the system by closing the loopholes as they arise. But this requires that the regulators be at least as smart, and as strongly motivated, as the SI firms. We have a ton of experience suggesting that this is not the case -- that if regulatory tools that reduce profits can be gamed, they will be gamed. The only way to avoid this is with regulatory tools that can't be gamed.

Approach 3: Adopt a Better Regulatory System That Shifts the Cost of Bailouts to Systemically Important Firms

The major objection to bailouts is less that the firms affected don't deserve it but that public funds are used for the purpose. If we had a way to require SI firms as a group to pay the cost of bailouts, the too-big-to-fail problem becomes manageable.

A regulatory system that can't be gamed by SI firms already exists and has been rigorously tested in other markets. This system could be easily modified to shift the cost of any required bail-out to SI firms. This alternative approach is discussed in the second article in this series.

You can contact the professor at