Now when I heard the title of Nassim Nicholas Taleb’s latest book, The Black Swan: The Impact of the Highly Improbable for the first time, my mind quickly went to the incredible juggling catch by Lynn Swann in Superbowl X over the sprawled out Dallas defender. The impact of that highly improbable catch led to an eventual win for the Steelers. The impact of the highly improbable Black Swans that Taleb mentions here, are far more devastating or exhilarating, except perhaps to Steeler and Cowboy fans.

Complete aside for those of you, who like me, wanted to see some of Lynn Swann’s incredible catches again, it turns out the NFL is rather protective of their rights to their images (I do seem to recall the phrase without the express written consent of the National Football League is strictly prohibited being thrown at me a few times). Anyway, if you go to this NFL.com link, and then search on Lynn Swann, choose the Baryshnikov in Cleats video, you wont be disappointed, unless you grew up a Cowboy fan.

Now for the heart of it, Taleb followed up his previous work Fooled by Randomness (a great book in its own right, which shows that just by pure random chance we should have very successful people, and though their success is largely attributable to luck, people do not see the impact of luck and focus in on other characteristics of that individual) with this book. To Taleb, and true to science there are two types of theories, those that are known to be false, and those that are open to be proven false. An epistemocrat as he defines it is someone who holds their own knowledge in suspect.

So, a bit of history here first to explain the title. For centuries all swans that had ever been observed in the western body of knowledge were white. It was assumed that all swans were white, because all swans that had been observed were white. With the discovery of Australia by the western world came word of Black Swans. They were completely unexpected, and became a bit of a novelty. Taleb uses this now famous example of how one cannot truly extrapolate no matter how many data points, that all swans are white, because a single counter-example will destroy the argument. For Taleb, and for me frankly, to remember the great empirical nature of science is paramount.

Life in general does not act like a Gaussian distribution. Its much more Mandelbrotian in its distribution. That means that most statistical applications which assume a Gaussian distribution are fundamentally flawed, unless they are dealing with events that are Gaussian. For example, height is a typically Gaussian distribution. However wealth is a non Gaussian distribution, as the average is quite low, but multiple occurrences of extremes away from the average that would be so unlikely as not to happen a million lifetimes.

Not only are most things non-Gaussian in nature, particularly as they relate to the larger economy, but there are unforeseen things that can have incredible impact, partly because they are unforeseen. These are what Taleb calls his Black Swans. So that Risk Management as defined by the portfolio managers is really a sham, because they are based on Gaussian distributions.

Furthermore, non-Gaussian distribution implies that we cannot make accurate predictions, since at best we can fit models to the past that really are just fitting a past curve rather than looking into the future. One of the other great points that Taleb drives home is around the failings of economic theory around the concept that man is rational. Man has biases, and will be effected by them to one degree or another. These two concepts together make the markets very unpredictable, and make risk much greater than it otherwise seems.

So what does Taleb do in the face of such problems? Well, he does a couple of things. He first segregates the black swans into two groups, positive black swans and negative swans. Then he tries to minimize his exposure to negative black swans, and maximize his exposure to positive black swans. He does this practically by putting around 85% of his portfolio in conservative US bonds (if the US were to default on their debt, he would be at risk, but would probably have much larger problems as well), and 15% in hyper-aggressive investments (venture capital and similar vehicles).

What does this mean practically? Well a couple of very good directions for me personally. The first is that I will be doing more research into biases, to understand the anchoring, confirmation, survivorship and similar biases, and their role in the larger markets (so called behavioral economics). Also it will help give me a much more critical evaluation of analytics tools and their practical limitations.

The Black Swan at Amazon

Wikipedia on Taleb

http://www.fooledbyrandomness.com/ — Taleb’s personal site, clearly he spends time on things other than his own website.

Malcolm Gladwell on Taleb in the New Yorker