Leverage Cycles and the Anxious Economy (American Economic Review, Vol. 98, No. 4, September 2008)

"This article is again about leverage. The relevant point it makes is that, just as there’s a business cycle with booms and recessions, so there is a leverage cycle. And the two are intimately connected. So, in boom times, leverage tends to be high, and in fact may become too high. By too high I mean that it becomes too easy for banks or other financial institutions to be wiped out by relatively small declines in asset values. Now, high leverage is a calculated risk that an individual bank may be willing to take on. If I’m highly leveraged, I stand to profit greatly if my bet pays off, and so some probability of collapse may be tolerable. But the problem is that my collapse doesn’t stop with me. If I go down, then other banks could well go down too. The Fostel-Geanakoplos paper describes a mechanism by which that happens. In their model, some banks are especially optimistic about assets and so take on high leverage. If assets turn out to be worth less than they thought, they get wiped out. But now – with them gone – there is lower demand for assets, causing a further fall in their value, which then wipes out banks with somewhat lower leverage, and so we have the same kind of downward spiral as in the Kiyotaki-Moore paper. So, these papers are related. They both give explanations for how a relatively small piece of bad news can give rise to a big deterioration – in liquidity, lending, and production. In both cases, the remedy, once such a decline begins, is for the government to step in and provide liquidity. In Fostel-Geanakoplos, furthermore, the government can act beforehand to limit crises (downward spirals) by constraining leverage in the first place. Yes, you do that by regulating leverage or equivalently by limiting liquidity. If leverage is restricted, then the bank’s capital requirements are higher, and so it can’t lend as much. In other words, the bank’s liquidity – how much it can finance productive projects – is reduced. Another way of accomplishing the same thing is for the central bank to raise interest rates. So, monetary policy and leverage/capital requirements are closely linked. Unfortunately, in the time leading up to the current crisis, not only was leverage high, but interest rates were low, both of which encouraged overly risky behaviour. Yes, it was. In fact, Geanakoplos made much the same set of points in a sequence of earlier papers going back to the late 1990s. Yes, or at least understand what’s in them. I think most of the pieces for understanding the current financial mess were in place well before the crisis occurred. If only they hadn’t been ignored. We’re not going to eliminate financial crises altogether, but we can certainly do a better job of preventing and containing them. This interview took place in September 2009, in the midst of the financial crisis."