My review of Taleb’s *The Black Swan*

by on June 13, 2007 at 8:28 pm in Books | Permalink

The book is very stimulating, here is one excerpt from my review on Slate.com:

Another human failing stems from the nature of happiness.  In the short run, people’s happiness is often shaped more by how many "positive events" occur in their day than by the arrival of one important piece of good news.  Winning $100,000 in the lottery feels almost as good as winning $1 million.  We therefore look, consciously or not, for small but repeated successes when we should be shooting for "one large win."  It’s easy to see why:  Big payoffs come only rarely, and perhaps late in life; in the meantime, who wants to keep on feeling like a loser?

Here is another bit:

Oddly, Taleb’s argument is weakest in the area he knows best, namely finance. Only on Wall Street do people seem to give proper credence–not too much, not too little–to very unlikely events. It is easy enough to use hindsight to identify the black swans Wall Street has missed, such as stock-price crashes. But it is harder to argue that the market undervalues surprise more generally. Stock and bond markets offer simple ways to bet on black swans. In financial terminology, you can purchase an option that is "deeply out of the money"; for instance, you can bet that Google shares will rise or fall in value an enormous amount over the next three months. These investments pay off precisely when the rest of the market does not anticipate the scope for surprise. Yet "long-shot" strategies are well-studied, and they do not yield extra profit. In other words, organized securities markets track rare and unpredictable events as well as the current state of knowledge will allow. If you don’t believe me, it is easy enough to bet on the Los Angeles Clippers to win the 2008 NBA title, or to bet on the longest odds at the racetrack. Such actions are hardly the path to either happiness or riches.

Here is my previous post on the book.  Here is Taleb’s podcast on EconTalk.  If you’ve read the book, do tell us what you thought of it…

Bruce G Charlton June 14, 2007 at 6:35 am

Neither Tyler’s review nor the detailed interview on EconTalk make me want to read this book – or at least, I don’t want to buy it.

It is trivially obvious that human inferences are based on incomplete and biased data. But in life inferences _must_ be made, action must be taken – and often urgently.

So a general policy of ‘greater skepticism about stuff’ (because it’s probably just random noise) usually leads to decision-paralysis camouflagued by a smug sort of quietism – which I seem to detect in the style of Taleb.

Buzzcut June 14, 2007 at 9:52 am

I’m about halfway to 3/4 through with the book.

I find his style of writing annoying. Taleb is a name dropper. Taleb is full of himself, and while he decries overconfidence, is quite overconfident himself. In fact, “NNT” as he likes to refer to himself, is quite an ass.

On the positive side, the book is quite easy to read. So he’s got that going for him.

But at the end of the day, I don’t know if his whole book couldn’t have been summed up in one chapter!

Yan Li June 14, 2007 at 10:53 am

I thought the book stimulating, but wish Taleb has elaborated more on the difference between independent and dependent events, which, I think, is the key to the long-tail phenomenon. Here are some of my random thoughts: in China, where people are less individualistic and more subjected to peer pressure, events tend to be more connected, at least locally. When the pressure of a system increases, local connectivity can spread globally, and lead to a phase transition of some sort. WWI is a good example.

Kyle June 14, 2007 at 11:42 am

I was pretty unimpressed. I had trouble getting over his self-indulgent and arrogant writing style, and the book often felt like a long essay entitled, “Why I’m smarter than 99.9% of the population and how the other .1% and I have deep intellectual conversations.” Overall, I don’t think I learned much more from the book than could already be gleaned from the EconTalk podcast.

conor June 14, 2007 at 1:50 pm

I read both FBR and the Black Swan. I agree about the overlap of the topics and also that the Black Swan could have been considerably condensed. Those with ideas dont always convey them best.

In response to two of your points from the review:

1) I think NNT makes a meaningful distinction between betting on longshots, such as betting on the Clippers, and unknowable extreme events with no determinable outcome. Betting on the Clippers is a defined improbability. It will pay a certain amount, say 500 to 1, for a defined event. They win or they don’t. The other events he mentions, publishing a hit book or a drop in the stock market, have an almost unlimited floor or ceiling. We may know they will happen but we don’t know when they will happen or how big it will be.

2) I dont think he totally discounts hard work. He does talk about people that work hard, hoping that the Black Swan will be their novel, their scientific research, their cumulative effort recognized in some way.

joeo June 14, 2007 at 4:46 pm

Fooled by Randomness was a bad book. He never gives any proof that his investing methods are better than the normal investing methods. The book was a content-less sales pitch for his hedge fund.

jb June 14, 2007 at 8:11 pm

So, ah, utility functions tend to have diminishing returns? And this is somehow new?

Play often for low stakes. That minimizes your objective expected losses and maximizes your subjective expected utility.

michael webster June 15, 2007 at 2:45 pm

I read both of Taleb’s books and expected more than I got from the Black Swan.

Fooled by Randomness has a number of impressive observations, and is pitched as such. There is no overall theory about rare events, only a number of interesting examples about how wrong some pretty bright people are about the rareness of events. Normal distributions come in for a reasonable bashing.

What I was hoping for in the Black Swan was some stabs at a theory, series of models, different explanations of how it is possible that
a) we all agree that X has low chances of happening;
b) we are wrong, and;
c) this happens on a consistent enough basis, but we don’t get much better in estimating X’s.

The Black Swan does not contain any of this, hence my own disappointment.

As to the Taleb’s writing style, I find it engaging, amusing, but maddening at times.

I found Fooled by Randomness to repay careful re-reading – like Steve Gilbert’s book on Happiness, Taleb can be too fun to read. But since I expected more simple models or examples of the Black Swan, I am not sure that I will return to Taleb’s new book that often.

odograph June 16, 2007 at 6:15 pm

There’s another good old podcast at it converseations. I think I’d have rated him pompous if I hadn’t got the nature of his humor from that. My feeling was the same, that Fooled by Randomness was a great read, and that Black Swan is a little uneven (does he dismiss editors in the book, even as he needs an editor?).

I am reading Stumbling on Happiness now, and I see many parallels.

eli June 19, 2007 at 2:51 pm

Alex,

In the least, Taleb is preparing a full point by point response to the Slate review. He does this with some reviews if he thinks them unfavorable.

Mainly, he holds the ideas he writes about very seriously and spends a lot of time clarifying his points if he feels a reviewer didn’t get it.

The only other “negative” reviewer to get a full retort is Gregg Easterbrook of the New York Times.

Anyhoo, it does help to listen to him speak to get a sense of his humor.

There is also a section on his site for “Black Swan Debates/Critiques”

Dave B November 5, 2008 at 3:56 pm

Taleb’s response seems characteristically opaque. I’d have thought he would just reply that the studies showing ‘no extra profit’ are themselves contaminated by assuming the distribution is Gaussian and/or by eliminating outliers from any tests of the hypothesis that the distribution is Gaussian. I’ve forgotten nearly all the statistics I ever learned (not much to begin with), so maybe I am off the mark.

Further, I don’t think he ever claimed a variant on his strategy would ever make ‘extra profit’, he would probably reject the whole framework. His concern is that conventional strategies are vulnerable to black swans (he might say that they are likely to be wiped out before they are closed out). To him, it is all about whether or not you’re vulnerable to negative black swans and positioned to benefit from some positive black swans. Offered a choice between a portfolio that makes ordinary (or even extra) profits but is vulnerable to crashes and one that makes pathetic profits or even small losses during ‘ordinary times’ but never loses your shirt and could make enormous profits from a black swan, Taleb prefers the latter (not entirely clear when black swan lottery tickets cost too much). Of course, he vaguely mentions a strategy with most of the portfolio in treasury securities and a small portion in highly speculative options, but obviously there are black swans that could make treasury securities worthless, so it’s not clear why he is not recommending that everyone stock up on canned goods and become survivalists.

One of Taleb’s claims makes me wish I had enough stats to give me a better feel for whether he was right or wrong – (I should probably go look it up specifically, but I am too lazy, so forgive me if this is only approximately what he wrote). He claimed that the crash of 1987 was a six sigma outlier (six standard deviations from the mean), an event that could be expected (if the data are actually Guassian) to occur once in 10,000 or 100,000 years. So the fact that both that and the various other crashes, especially the crash of 1929, occurred within a few hundred years of the beginning of the stock market should be sufficient to reject the hypothesis that the data are actually Gaussian. My perhaps lame exposition of the Taleb position would be that the standard models exclude outliers such as these crashes to get a better fit for ordinary activity (Taleb’s ‘Mediocristan’). If we instead include the outliers in the regression, we end up with something that stinks even at predicting ordinary stuff, and still cannot predict crashes. Hence, Taleb rejects the models and ordinary measures of risk and extra profit.

I found the book very interesting but frustrating, the arguments often difficult to follow. He accuses his critics of ad hominem, but he (especially on his web page) uses ad hominem himself to an unfortunate degree. I found his ideas interesting, but not entirely convincing. I didn’t find Tyler’s review helpful. I’m not sure whether this means Tyler wasn’t taking Taleb seriously, or whether (as I think is sometimes the case with Tyler) he left out a lot of stuff he considers obvious, but which by no means is obvious to me.

Alexandra Hopkins April 11, 2010 at 1:47 am

“The Black Swan” is one of the most thought-provoking books I’ve read (or almost read — I’m 3/4 through). Yes, we all know that large unexpected events can change everything. But do we act like we know it? Do we insure ourselves in some way against 1929-type crashes? Few do. Why? Because we don’t fully understand and take seriously the nature of the Black Swan — the large, unexpected event.

Thus, the market crash of 2008. We are swayed by a number of fallacies that human thinking is subject to, like buying the “story” (the latest fad in thinking like “Nobel Prize winners have developed the math model that shows sub-prime mortgages are safe.”)

Yes, the book could have been written more clearly and in a more neutral style. True, one must take action when faced with uncertainty, something Taleb doesn’t focus a great deal on. And I’ve not read “Fooled by Randomness,” so can’t judge the amount of overlap. But developing deep understanding of common fallacious lines of thought is priceless. That’s what “The Black Swan” helps one with. Yey, Taleb! Your excessive ego is totally justified!

Comments on this entry are closed.

Previous post:

Next post: