"Until today or tomorrow, the typical turkey enjoyed a fairly decent life," commented our friend Nassim Taleb, in Zurich yesterday.
"You can understand how fraudulent most economic analysis is," Nassim explained, "just by looking the life of the turkey. The animal is fed for 1000 days...and then it is killed. So, if you plotted out the turkey's life on a chart, it would look great for 1,000 days...each day, the food arrived reliably, and each day, the turkey gained weight. The turkeys would look around and say they were enjoying growth and a bull market.
Momentum investors would see it as an opportunity. The quants would run linear regressions on the data and prove that the risk was minimal. "
Ben Bernanke would describe the turkey's life - with no setbacks - as the product of a "great moderation." Turkey stockbrokers would assure their clients that nothing had ever gone wrong in the turkey's life. Turkey econometricians and theorists would come up with explanations for why the turkeys' growth would continue forever and they'd pat each other on the back for having finally mastered the "turkey cycle." Turkey politicians would run for re-election on the grounds that they had helped create a better world. And turkey economists would project further weight gains...until the turkey was the size of a hippopotamus
Then, come Thanksgiving, and all of a sudden, something goes wrong. Alas,
all the turkeys' theories, models, and conceits were for the birds.
"Rare events can't be modeled," Nassim continued. "Because they are too rare. You can't get a statistically reliable sample. Alan Greenspan recently explained that he 'had never seen anything like this before.'
Well, of course he had never seen it before. It never happened before.
"Because these events are so rare, they are also completely unpredictable...and usually much worse than you can expect. Like Thanksgiving Day for the turkey."