Anil Kashyap, Raghuram Rajan and Jeremy Stein presented a fascinating paper at the Jackson Hole Symposium. This is my third post related to this symposium and yes, finally, the papers are available at the Kansas Fed website. The Buiter paper that I blogged about a few days ago after reading it on the WSJ website is also now available at the Kansas Fed website.
Kashyap, Rajan and Stein propose an instrument similar to a catastrophe bond which would automatically recapitalize the banking system when system wide losses cross a certain threshold. The starting premise of their paper is that the high level of leverage in banks is a rational response to governance problem of banks. If one accepts this premise, then their proposal for contingent capital makes a lot of sense. It gives banks capital to the banks when it is needed in the crisis, but does not give it to them in the boom when they might squander it.
It is a very clever idea, but I tend to be sceptical of the notion that there is something special about banks that exempt them from the Modigliani Miller arguments about capital structure. Rather I think Friedman was right in his statement in a different context that banks are not regulated because they are different, but banks are different because they are regulated. After all, many non bank financial firms run quite well with lower levels of leverage than banks, and the free banking era was also characterized by relatively low levels of leverage.
Kashyap, Rajan and Stein design their instrument very elegantly to deal with moral hazard problems at the level of the individual banks. My fear is that it will attract moral hazard on the part of the regulators. In fact, I think that their bond is best seen as non voting equity in a deposit insurance corporation. The governance implications of this are a nightmare.