I’ve been critical of the use of the term “too big to fail” which implies that the problem with banks that we bailed out rather than sending through bankruptcy is that they were too big. In fact, the issue has nothing to do with size. Bankruptcy is problematic for any bank, since bankrupting a bank leads to bank runs. The problem isn’t that we had some banks that were “too big to fail,” the problem is that we had some banks that weren’t eligible for the FDIC resolution process.
Mike Konczal tries to rescue the concept by positing a two-dimensional space of bank problematicness:
Here’s why I don’t think of “too big” as a myth for resolving a firm. In my mind, the farther you are from the origin in that graph, the harder it is for the government to detect problems and properly deter large firms under resolution authority. (This is why I draw our “safe” resolution as a circle, instead of a square.) Holding for a liquidity risk, the larger the firm, the more vicious the effects of having a shadow banking run on the rest of the financial sector and on the real economy.
That seems fine to me. But it still leaves me with the conclusion, from the point of view of a writer, that the phrase “too big to fail” is very misleading. If we eliminated all very large banks but don’t implement a new resolution process, then we’ll be right back where we were before the crisis. The banks, even though they’re much smaller, will still be too big to fail because we’ll still be afraid of shadow bank runs. So the resolution process is vital.
As a separate issue, it does seem plausible that for reasons of political economy and feasibility of problem-detection that smaller banks would be desirable. But it’s important for people not to get the idea that shrinking banks works as an alternative to regulating them and establishing a resolution process.