A little thought reveals that it is not possible to create a set of special regulations for systemically important companies without identifying them. If the administration believes no one in the market will notice, that’s an absurd idea. As soon as it becomes clear that certain companies have been singled out for special treatment, they will be considered too big to fail and the problems of moral hazard will arise, just as they did with Fannie and Freddie. On the other hand, the idea of resolving systemically significant companies through use of a government agency, like the FDIC, seems—in a superficial sense--to avoid the Fannie and Freddie problem, because it is not necessary to identify these companies until they are actually failing. However, once again a little thought reveals that this is another flawed solution. No one can tell the difference between a company that will create a systemic breakdown when it fails, and one that will only cause some temporary disruption in the economy. Thus, the likelihood is that the relevant officials will err on the side of resolving every large financial company as though it were systemically significant. This only makes sense from the standpoint of the government. Why risk being blamed if a real crisis ensues? Moreover, the political pressure to take over large failing companies—as shown by the GM and Chrysler bailouts—will be irresistible once the authority and funding is made available to the government. If anything, this badly conceived idea will produce more moral hazard than designating systemically important companies in advance. The market will simply assume that every large company is too big to fail, and that will actually turn out to be true if the Obama plan is adopted. In effect, we will have created a permanent TARP.

The administration should abandon the idea of trying to gain control over the whole financial system and focus instead on better regulation of commercial banks. That’s where systemic risk really may exist, and where the necessary regulatory resources are already in place.

Policymakers have two basic options for reforming the financial system so it is less vulnerable to the failure of just one or two major institutions. First, they can establish rules that shrink the maximum permitted size of U.S. banks and other financial entities. If Lehmann Brothers and A.I.G. had been one-tenth as large, other market participants as well as public regulators would have viewed their insolvency with much less alarm. Government authorities and the courts could have unwound the companies’ businesses in an orderly way with less risk to the overall financial system and, equally importantly, without bailing out the firms’ shareholders and bondholders. Lehmann Brothers’ and A.I.G. executives and the shareholders and bondholders who backed them would have experienced enormous losses, but the normal operations of U.S. capital markets could have continued with little interruption.

A second option is to increase the capital requirements of large institutions above those of smaller institutions and to subject the biggest institutions to more stringent rules and tighter regulatory scrutiny. We would still have major institutions that are too big to fail. Higher capital requirements would ensure, however, that an institution’s equity holders would absorb a greater percentage of the losses if a bailout were needed. (Reporters, pundits, and ordinary voters sometimes forget that a government bailout usually offers a lifeline to an institution’s debt holders and counter-parties but not to its shareholders.) Moreover, a regime of tougher regulatory oversight could mean that these institutions will be less likely to engage in behavior that places them at grave financial risk.

The more conservative of these two options is the second one, and it is the one the Obama Administration favors. Were it adopted regulatory authorities would not need to oversee a drastic downsizing of the nation’s biggest banks and other highly interconnected financial institutions. Bank of America, Citibank, Wells Fargo, and JP Morgan Chase could retain their current size, but they would need to support their lending activity with a higher ratio of shareholder equity. They would also have to pay, directly and indirectly, for the additional cost of more intense federal oversight and tougher regulatory rules. The problem of moral hazard will remain, because bondholders and bank counterparties will continue to expect the government to bail out big institutions in the event of insolvency. Big banks’ shareholders would, however, be forced to absorb more of the losses that follow insolvency. If this approach is to work, regulatory oversight of the big institutions will have to be much stronger.

People who oppose tight government regulation or who are skeptical public regulators are up to their job should favor restrictions on financial institutions’ size and market share. The tradeoff is that large size may also give financial institutions advantages that flow from sheer economies of scale rather than outsize market power. If we limit the maximum size of banks and other financial institutions we are giving up the potential benefits of scale economies, not only for the banks themselves but also for their depositors and borrowers. 

Whether we adopt option #1 (unlikely) or option #2 (far more likely), it seems plain that the current regulatory regime cannot be left unchanged. The events of the past 18 months show that the risks of the status quo are too great.

I support the idea of creating better resolution mechanisms for banks and nonbanks that would resolve problems associated with allowing them to fail. But the devil is in the details. Some approaches to designing this new resolution mechanism – specifically, those that would vest discretionary resolution authority in the Fed or any other government authority – would likely make the too-big-to-fail problem worse because those government authorities would be correctly perceived as more inclined and able to use public funds to bail out large complex institutions. Furthermore, it would be extremely unwise to ask the Fed to manage resolution policy; making the Fed into a discretionary bankruptcy court would further compromise its independence.

The right approach to reforming the resolution of large financial institutions has two parts. First, amend the bankruptcy code to cure any technical deficiencies that make it hard to apply bankruptcy to financial institutions (e.g., the treatment of derivatives contracts). The bankruptcy court, rather than a regulatory authority, is the right place to handle resolution. It is also important that the resolution rules be strict and not subject to too much judicial discretion. The law should require that shareholders in a failed institution face a complete loss, that long-term debtholders face losses commensurate with the negative net worth of the failing institution, and that any government assistance for the sake of incentivizing a merger should be defensible on the basis of a “least cost resolution” test, meaning that no government resources would be used unless doing so produces concrete and demonstrable savings on the transaction to taxpayers. Placing the responsibility for enforcing these strict standards in a court would increase the chance that resolution would be handled properly by applying the rule of law to a preexisting code, and would minimize the chance that political expediency would create an abuse of discretionary regulatory authority.

Second, it is crucial that regulatory authorities in the U.S. work with those in the U.K., and eventually with those in other countries, to establish effective, pre-specified rules for allocating losses across borders. In the Lehman bankruptcy, significant disputes arose among different countries’ regulatory authorities and courts over which country’s affiliate had the better claim to certain assets within the institution. The difficulty of resolving those cross-border conflicts makes it harder to apply credible market discipline to failed institutions; when bankruptcy is a mess, policy makers want to find an alternative. Regulators and financial institutions should have clearly specified and publicly disclosed plans in place that describe how ownership interests by affiliates will be treated by all regulators so that there is no opportunity for disagreement among the regulatory authorities of the various countries in which affiliates are located. The bankruptcy codes and regulatory rules of the various countries should explicitly recognize and respect those arrangements.