Seeker Blog

An in-depth and longish analysis by structured finance expert Economics of Contempt. My bottom line is the administration policy has a lot to recommend it. There are important improvements needed — which are very clearly outlined here.

One thing I’ve been noticing is that many commentators on “too big to fail” (TBTF) policy have clearly never read the Obama administration’s financial reform proposals, or at least have an extremely poor understanding of what the administration is proposing to do. This is unfortunate, because TBTF policy is an important, albeit complex, topic. It can’t be addressed in a snappy op-ed, or by simply saying “make them smaller” (as if size alone is the problem). It requires serious thought on a number of related issues.

This post is my attempt to have the beginnings of a serious discussion of TBTF policy. It’s long, since I took about half of my flight to London to write it. But this is a complex issue — there’s no getting around that. If you believe that the Obama administration isn’t proposing to do anything about TBTF, or if you believe, like Joe Stiglitz, that the administration is proposing to create “institutions too big to be resolved,” then I’m sorry, but you’ve been seriously misled. My aim in this post is to explain what the administration is actually proposing to do about TBTF, and also to explain where I think the administration’s proposals have gone wrong, and what I would do differently. If you think you already know what the administration is proposing to do about TBTF, then you can probably skip to the next section on “What I would do differently.” But I’m betting most people never fully read the administration’s proposals.
(…)

What I would do differently

More realistic PCA triggers. First, the prompt corrective action (PCA) regime. The reason I think the PCA regime is the key to TBTF policy is because it makes the new resolution authority for Tier 1 FHCs credible. Some commentators dismiss the resolution authority as a solution to TBTF because they say the market won’t believe that the government will actually use it. They say the market will continue to believe that the government will opt for a bailout rather than the new resolution authority. In that case, the new resolution authority wouldn’t do anything to fix the moral hazard that TBTF induces — the JPMorgans and Goldmans of the world would still be betting that the government will bail them out, and thus would still have an incentive to take excessive risks. Now, I think those commentators are wrong, and that the government would opt for the new resolution authority over a bailout, even without a PCA regime.

But a PCA regime comes as close as possible to guaranteeing that the government will actually use the new resolution authority. Lehman Brothers was essentially allowed to come careening into bankruptcy court — or, as one former Lehmanite put it to me, Lehman went down with “guns blazing.” There was no mechanism to force Lehman to take specific steps to recapitalize itself in the months between Bear’s failure and September 15th. All we had was Hank Paulson and Tim Geithner badgering Dick Fuld about raising capital or finding a buyer, which they apparently did, to no avail.

A PCA regime not only minimizes the chances that a Tier 1 FHC will actually fail, but it also prepares a failing Tier 1 FHC for an orderly resolution. Lehman was still transferring assets in between its various European and North American branches at a furious pace right up until its failure. As a result, a lot of hedge funds were very surprised to discover that their assets were not in segregated accounts, but in fact had been transferred to Lehman Brothers International (Europe) and then rehypothecated. This was a significant source of uncertainty in the days following Lehman’s bankruptcy filing — it was Knightian uncertainty in action. A PCA regime would prevent this kind of thing from happening, as the Fed would have the authority to restrict transactions between affiliates once a Tier 1 FHC becomes significantly undercapitalized.