Bloggers Address Gary Gorton's Paper

One reason I think Gary Gorton's work is so useful is because it reflects his appreciation of how subtle this financial crisis was. If you were totally outside it, or just calling for some unspecified disaster, it is tempting to see the crisis as something really obvious (eg, outlaw greed, hubris, and poorly targeted regulation).

Recently Fed Chief Ben Bernanke recommended people read Gary Gorton's latest paper, probably after reading my blog post praising it (heh).

So, you can watch Gorton's trenchant analysis get transformed into tripe in real time. James Kwak, Ezra Klein, Mike Rorty, Mark Thoma, and Felix Salmon, and Dr. Manhattan jumped in. I'm not saying they are all generating tripe, merely it basically demands a reader have correct prejudices to draw the correct inferences from the union of all these writings.

The debate is generating the focus of any decentralized collective, where everyone distills a different key point. Part of this group dynamic is simply specialization, where people focus on their unique insights because they want to say something new (interesting ideas are important, true and new). But it is very confusing when everyone agrees with the basics of what everyone else says (eg, there were regulatory failures) excepting the priority of the concerns.

First, I would agree with Dr. Manhattan that regulation has a poor batting average in the financial sector, and merely saying our new regulators should be smart, disinterested, and hardworking is like saying our political leaders should be so to. They aren't, and never have been. Highlight a specific regulatory idea, because merely calling for 'more' only ensures that it will be more of the same.

Secondly, Kwak and Klein get it backward when they first describe a repo secured by an 'informationally insensitive' derivative, then talk about a bank run caused by investors afraid that the bank will default. In the repo problem described by Gorton, the repo collateral-- mortgage securities rated AAA -- is what became risky or 'informationally sensitive'. This was the basis of the contraction, the prime mover, which lead to Bear and AIG's collapse, not Bear and AIG leading to mortgage problems. The causation arrow's direction is very important for both diagnosis and potential cures.

Felix Salmon notes that we shouldn't guarantee informationally insenstive assets, as it encourages the belief that investing is riskless. I agree, but I don't think it is possible to think investors will not create some new 'riskless' time bomb in the future. The situation is, like Minsky's financial instability hypothesis, endogenous. Currently investment grade securities are viewed like junk bonds circa 2006. For example, at the 5 year point, A-rated bonds trade at 170 basis points over Treasuries, the same spread BB-rated bonds traded at in June 2006. Historically, A-rated bonds have about a 0.1% annualized default rate, BB a 1.15% annualized default rate, so currently people don't think anything is riskless.

After another 50 years of no defaults, the new AAA asset securities will be of two types. An asset class with an even longer period of no default, like US Treasury debt, or the next generation of mortgage backed securities, which will have explicitly addressed the issues relevant to this current crisis. In the latter case, a reasonable argument can be made that such securities should not include the 2007-8 scenario in their future default rate expectations because there has been a structural change specifically rectifying the mistakes so that the 2007-8 scenario is no longer relevant (eg, all mortgage ABS now includes a 50% collateral value collapse as having a 3% probability, unlike previously).

As a practical matter, worrying about securities with 0.01% annualized default rates is a waste of time, like worrying about World Wars. Sure, they happen, but they are always different, you probably will have no control, and for most people they won't experience such a crisis here their entire working life. So the asset becomes, through no mandate or official diktat, money, and through the money multiplier, a keystone in any future financial collapse. If people generally believe an asset is AAA, it will be treated like money endogenously through the process Gorton outlines. Such assets do not acquire this characteristic merely by a rating agency, people actually believe it is true using all of their insight. They are usually write, but not always.

I think it is much better to focus on the regulations we know work, such as stable property rights, and accept that growing economies will have recessions. Robert Lucas has written about how "the potential for welfare gains from better long-run, supply side policies exceeds by far the potential from further improvements in short-run demand management". The current crisis is short term demand management. The mistakes, with hindsight, are obvious and should be addressed. But the general take-away should not be how to prevent anything like this from happening again, because it is going to happen again in spite or because of anything we do now. Remember, crashes are endogenous, but so are recoveries.