
Via edge.org. Benoit Meandelbrot, for whom the hypnotically beautiful Mandelbrot set is named after, jumps on the Black Swan wagon.
Chuck can probably provide more color on this, but it seems to me that Mandelbrot's critique that VaR models is flawed because it depends on a normal distribution is not entirely fair. I assume that the quants at Wall Street have swapped out the normal distribution assumption and put in some sort of a fat-tail distribution. Or at least that's what I hoped they did.
Obviously the results sucked, to say the least. But I'm pretty sure that no one took the normal distribution assumption as something set in stone.
The WSJ OpEd writer is also a bit silly to proclaim that "The good news for VaR and similar models is that the free market alone would not have allowed the bubble of subsidized mortgages". No, but the free market did allow the bubble of unsubsidized mortgages. Indymac, Countrywide, WaMu, Bears Stearns, et al were as free market as they came.
1 comment:
I have no constructive comment on this, need more ejumacation before opening my mouth about risk management.
Post a Comment