"Too Big to Fail" Again

I wrote a long post for The Atlantic's business blog the other day about the idiocy of the whole "too big to fail, too big to exist" idea, which is sadly catching on in the blogosphere. And it's showing no signs of slowing down. Matt Yglesias wrote this about Goldman Sachs:

So instead of [the issue] being "if they want taxpayers to save them, then they have to take fewer risks and become smaller" it's given that taxpayers will save them, then have to take fewer risks and become smaller.

Oy. First of all, Matt ignores the fact that Goldman subjected itself to stricter regulation when it converted to a bank holding company—including tighter controls on leverage. (In a slightly related point, it amazes me when people who have been paying somewhat close attention to the financial crisis assert that the Obama administration wants to "turn the clock back to 2006" in the financial sector. Honestly, what planet are these people living on? Everything has changed. It wouldn't be possible to go back to the financial sector circa 2006 even if the Obama administration wanted to—and it doesn't.)

Second, Matt falsely assumes that the solution to the too big to fail (TBTF) problem is for Goldman to "take fewer risks and become smaller." As I noted in my Atlantic post, this is an absurd idea. TBTF has always been a misnomer—the issue is whether a financial institution is "too systematically important to fail." Yes, Goldman is too systematically important to fail, but it's silly to suggest that the solution is to force Goldman to "become smaller."

The reason Goldman, JPMorgan and the rest are too systematically important to fail is that the Bankruptcy Code, which governs the resolution of failed bank holding companies (and most other nonbank financial institutions), can't facilitate the orderly resolution of systematically important financial institutions. The fact that bank insolvencies are fundamentally different from regular corporate insolvencies is the whole reason we have a separate insolvency regime for commercial banks and thrifts, which is run by the FDIC rather than bankruptcy courts. If we had an insolvency regime that was capable of resolving large complex financial institutions (LCFIs) without causing a financial crisis, then no LCFI would be "too systematically important to fail." Some LCFIs would still be systematically important, but the point is that with an effective insolvency regime, regulators would no longer fear the broader consequences of letting them fail. Thus, they would be systematically important, but not too systematically important to fail.

How to design an effective insolvency regime for systematically important financial institutions is what the debate in the TBTF literature is all about. There are lots of ideas on how to design an effective insolvency regime for systematically important financial institutions—my favorite being the NewBank concept, which involved the creation of a dormant bank that is available for activation to clear and settle government securities, should one of the two banks that provide such services (Bank of New York and JPMorgan) ever fail. I won't pretend to have all the answers on how to design an effective insolvency regime, but this is fundamentally what the TBTF debate is about. Notice, however, that a new insolvency regime for nonbank financial institutions does away with the need for a definition of "too systematically important to fail" (which is impossible to define ex ante anyway).

But nowhere in the TBTF literature have I ever seen a proposal to force systematically important financial institutions to "become smaller." There's a reason for that: it's a really, really stupid idea.