How should we try to predict Black Swans? This is a tautologically difficult thing to do: by definition, we can’t predict unpredictable events. However, perhaps we can prepare for rare events as a general framework. Taleb loves to attack financial engineers most of all people. He sees these people as unethically hiding the possibility of a Black Swan event so that their investment strategies look good, and perform well — until they completely flop. (Taleb wrote this book in 2007, before the most recent recession. I’m sure he’s having a field day now gloating about how stupid financial engineers were being prior to 2008, and had they only listened to him, this mess could have been averted.) Taleb recommends that we invest most of our money in the most secure investments we can find (something like government bonds), and then invest the rest of it in a diverse selection of highly speculative ventures with some chance of huge positive Black Swans.
Taleb’s criticism of financial engineers, quants, and traders is quite legitimate in some ways. These people get big bonuses at the end of the year if they’re successful. However, they don’t get big negative bonuses if they lose heaps of money. So, there’s terrific motivation for these people to bet against Black Swans: a very good year followed by a very bad year is much better for them than two consecutive averagely good years. In fact, anything they can do to push disasters past the end-of-year horizon is personally beneficial to their bonus collections.
So, all that’s well-presented in the book. The problem is that then Taleb insists on going on a tirade against the bell curve. To him, anyone who starts talking about bell curves, Gaussian distributions, normal distributions, Poisson distributions, or other well-behaved statistics, are dishonest thugs who are only capable of shrouding their scams in fancy mathematical terminology.
It’s all very well to say that sometimes event distributions don’t satisfy traditional distributions, but without giving us a better option, it’s unclear how we can deal with them. We use the Gaussian distribution because many events follow it exceptionally well, and because it’s the best explanation we have for how events tend to be distributed. Taleb makes some half-hearted attempt to tell us to replace classical statistics with techniques from chaos theory, even going so far as to say that the latter techniques are more intuitive and easier to use.
Um, what? The problem is that chaos theory does not give us techniques to use to understand event distributions and make predictions. One might say that that’s the whole point of the subject. Very fine details in initial conditions give us diverging outcomes down the line. How did Taleb, who was good friends with Benoît Mandelbrot (one of only two mathematicians Taleb seems to respect, the other being Henri Poincaré), manage to miss this?
Taleb also claims that Black Swan events dominate the trajectory of history. I think that’s sort of true, but not really in the sense that Taleb does. One could say that the development of the Internet was an example of a Black Swan event that thoroughly shook up the world. Okay, probably no one predicted it long before it showed up. On the other hand, we’d have to be pretty silly not to expect that some technological innovations will occur that will shake up the world. We don’t know the details of what they’ll be, but that there will be some is inevitable. Kevin Kelly, author of What technology wants (which I haven’t read yet; that’s coming soon), claims that when technological innovations are ready to break through, they’re already completely inevitable, and that there are generally several groups of people working on the same thing at the same time. So, Kelly would probably debate Taleb’s view on Black Swan technologies.
But even though Taleb knows a lot about Black Swan events, thinks about them a lot, and considers them to be the cause of much of the trajectory of history, he recommends that we avoid being placed in situations in which they’re likely to show up. So, he advises that we find jobs in fields in which growth is steady and predictable (such as being a mathematician, even though he doesn’t like us very much), rather than choosing those in which we might make a fortune, or we might make very little (such as writing books).
That’s good advice for the individual, maybe, but not so good for the progress of society. One of the most memorable segments of a book overflowing with groundbreaking ideas was the anteater dialogue, in which the anteater was seen to be beneficial to an ant colony, while naturally being harmful to each individual member of the colony. I’d be very sorry to live in a world in which everyone follows Taleb’s advice, since innovation is always a Black Swan event for the individual. For society to progress in certain ways, a few individuals have to take major risks.
While I have a fair number of criticisms of Taleb’s book, especially with respect to his disdain for mathematics, I loved reading it. It’s full of fascinating and original ideas, and it just happens that a few of them are flops. That’s inevitable in such an original work. I give Taleb major kudos for having the courage to present his ideas so publicly, subjecting them to heavy scrutiny. He said he is hoping that someone will eventually write a book called The white swan, challenging his ideas.
So, I definitely recommend reading this book, but I’d advise a healthy dose of skepticism. Rather, the reader ought to filter out the good ideas from the bad ones and ignore the latter. Have fun with it!