Fat Tails

It is worth noting that the losses, which are still slowly becoming public, from the un-pegging of the Swiss franc, are a perfect illustration of the problem with financial pricing models that rely on standard statistical distributions such as the normal distribution. People use the standard distributions for purposes such as option pricing – the Black-Scholes model that cratered LTCM is an example – because they are mathematically tractable and well understood. The problem is that, like the Fed’s DSGE economic models, they don’t work in reality. They would predict that the recent move in the Swiss france would occur once in a few trillion years. Never, in other words. Except it just happened, because in the real world distributions have “fat tails.” That is, extreme events occur much more frequently than would “normally” be expected.

While there is much public whining from the losers, there is silence from the winners, probably including Nassim Taleb and George Soros, at least some of whom surely had been quietly rolling over call  options on the Swissie waiting for the inevitable event. Bad pricing models meant that those options were mispriced and cheap, so the losses while waiting would have been small, very small relative to the eventual payoff.

Both comments and trackbacks are currently closed.
%d bloggers like this: