Which Taleb is Right?

Pablo Triana today wades back in to the Black-Scholes debate, and takes aim at Michael Lewis:

Black-Scholes has been blamed in certain quarters for the subprime crisis. Essentially, the argument is that those blinded by the dictates of the model took too many risks too eagerly (and cheaply)...
In fact, Black-Scholes may not be used that much in the markets to begin with. New research by veteran traders and best-selling authors Nassim Taleb and Espen Haug points in that direction. Clearly, a formula that isn't used can't have much of an effect on markets, let alone cause the massacre that began last summer.

This is much the same point as I was making after the Lewis piece came out. But in a comment on that blog entry, Nassim Taleb himself sides with Lewis. So who's right? Triana's view of what Taleb is saying, or Taleb's own view of what Taleb is saying?

I think there might be less conflict here than meets the eye. Taleb's defense of Lewis's thesis is basically that Lewis uses the term "Black Scholes" not to refer to the Black-Scholes theorem itself, but rather to portfolio theory more generally, and the CAPM model as a whole. In the narrow sense, then, Triana's Taleb is right. And in the broader sense, I think that Taleb and Lewis have a bit of work to do if they're to persuade me that CAPM (and, by extension, Black Scholes) is to blame for the housing bubble, the subprime crisis, and the present credit crunch.

The way I see it, the debt bubble was much more the result of an old-fashioned search for yield than it was the result of a new-fangled search for "alpha". Did bankers and investors fool themselves by plugging low volatility numbers into their models in order to boost their risk-adjusted returns? Undoubtedly. But the connection between those models, on the one hand, and Black-Scholes, on the other, is tenuous at best, as Triana shows. After all, CAPM has been around a long time; it can't have been that disastrous, given the amount that markets have risen over the course of its existence.