Saturday, November 8, 2008

The Black Swan



Yo Gabba Gabba and Gamma Boy's recent tizzy over the effectiveness of models, financial and otherwise is perfect fodder for a book that I am currently reading called The Black Swan: The Impact of the Highly Improbable, by Nassim Nicholas Taleb.

I'm only on page 70, but an important point he has made so far is summed up in the story of a Thanksgiving turkey. Say a turkey lives for 1,000 days, he gets fed handsomely every day and life is great. If he were to model his expectations for the rest of his life based on what he knows, he would fail to account for his slaughter on day 1,001.

I highly recommend the book. It'll change the way you think, plus he's a good writer and all-around interesting guy. And his hedge funds are up 65% to 115% this year. Incidentally, I think Black Swan thinking needs to be incorporated somehow into the Sensible Party's platform. It has vast implications for public policy as well as business planning.

I'm touting this book to move the discussion along, not to say that all models are bad. I think we need to be judicious in how we use them. In the case of climate science, they represent our best guess based on what we know. However, we shouldn't ignore the chance that a Black Swan event could invalidate the model completely.

4 comments:

Lt. Weinberg said...

I don't see how the turkey analogy could possibly relate to real life, unless you believe that we are oblivious to the will of a supreme being who ultimately decides our fate (i.e. we are turkeys who don't know our planned destiny). The weather, in theory, should be easier to predict than economics since it follows rigid physical laws (whereas the economy is victim to irrational human behavior). If the model doesn't work, you revise the model, you don't give up on modeling...

Al Powell said...

Yeah, my description of the book was way too pithy. You really have to read it. I forgot to mention how he divides the world into areas in which randomness is mild, like weather, and in which it can be extreme, like financial markets. Weather is much more apt to be modeled with more certainty. But, it's still possible to have an asteroid hit or something.

Restless Native said...

Other required reading for the SP will be a book called "Nudge" by Theler, Richard and Sunstein, Cass, both of the U of Chicago. Bear in mind that while Obama taught there so did Milton Friedman.

Talks about "choice architecture" and focuses on eliciting the "desired" result by setting things up such that people make "good" decisions without being forced to do so.

Restless Native said...

BTB, is Lt Weinberg a certain smart guy in cell-type stuff with a wife who went to a large state school with a penchant for burning large piles of combustible cellulose and jack boots?