Too Specific to Fail
The FPD Forum opened this morning with a not-so-cheerfully titled plenary The End of the World As We Know It?. The panelists were Daron Acemoglu of MIT, Alan Rosling of Tata Sons, Tim Harford of the Financial Times, and Michael Klein of FPD/World Bank. Perhaps my favorite line from the plenary: Too specific to fail. Acemoglu was discussing the bailout of banks in the U.S., and pointed out that the moral hazard problem isn't so much an issue of "too big to fail" as "too specific." The human capital and knowledge contained in investment banks is highly specific, and it's exactly during a crisis that it's needed. The upshot? It's very difficult to punish the managers of these banks for taking excessive risks in the good times when they're needed to help halt a systemic crisis during the bad times.
Comments (1)
E-mail
Digg
Bookmark
Facebook
I can understand the too big or too systemic to fail part, which is why I have for a long time argued that “a regulation that regulates less, but is more active and trigger-happy, and treats a bank failure as something normal, as it should be, could be a much more effective regulation,” and even suggested a simple progressive tax on size.
But I could not really get a hold of the “too specific too fail” concept.
It is said that “The human capital and knowledge contained in investment banks is highly specific, and it's exactly during a crisis that it's needed,” but which goes a bit against the lack of knowledge and human wisdom that investment bankers displayed and which helped to fuel this crisis.
I am not at all a banker basher, far from it, but out there in the real world there are many suggesting it is better to call the fire brigade than the arsonist even if the latter has more recent expertise.
More than sophisticated financial experts we need back to basic practitioners so as to be able to recover at least a simple starting point that we all, including bank directors, have a reasonable chance to comprehend.
And from there on… please keep it simple. What is transparency worth if it cannot be understood?
Posted by: Per Kurowski | Feb 28, 2009 8:08:35 AM