Fooled by Randomness

I found a lot of common ground when I read Nassim Nicholas Taleb’s Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets, 2004, Second Edition. There are lots of good quotes that fit well with my own thinking. We also have some sharp differences. But they do not involve the book’s major points.

I especially liked his presentation of the problem with false positives. Assume that a disease affects 0.1% of the population and a test has a 5% chance of returning a false positive. If the test is positive (indicating that a patient has the disease), what are the odds that he really does?

The answer is that it is not very likely in spite of the test. In fact, it is highly unlikely. There are too many false positives.

Out of 1000 people, you would expect 1 to be sick. The test would indicate this. You would expect the test to indicate that 50 people (5% of 1000) who are well have the disease. This totals 51 people. Only one is sick. If the test indicates that a patient has the disease, the odds are 1 out of 51 that he actually does. The actual chance that the patient has the disease is under 2%.

The clarity of his presentation comes from the very low (0.1%) percentage of the population that the disease affects. It is easy for us to keep the factor of 1 in a 1000 separated in our minds. If Taleb had started with a disease that affects 10% of the population, his argument would have been much more difficult to understand.

Nassim Taleb’s profession is trading options. He purchases a large number of options that are likely to expire worthless. But he makes sure that those that do pay off, pay off handsomely. He makes sure that they pay enough to make up for his losses and much more.

Taleb focuses on the kind of rare events that happen all of the time. We see this when thousand year floods happen two or three times in a decade. Nassim Taleb looks for failure mechanisms. He looks deeply for flaws in our logic. How do we know that something really is a rare event? How do we know that those thousand year floods really are thousand year floods?

Nassim Taleb mentions his dependence on Monte Carlo models. Rest assured that he builds much more sophisticated models than we do.

Here are some areas of similarities and consistencies between what I do and what I see in the book:
1) I caution readers to expect to be blindsided at least once in every generation.
2) The book focuses on the logic behind statistics. I consider the logic to be much more important than the calculations.
3) I refuse to claim any statistical precision greater than 90% (two sided). I point to the underlying uncertainties in our knowledge about the probability distribution of the stock market.
4) We make extensive use of our models and Nassim Taleb makes extensive use of his. He points out that modeling gives many answers that elegant theorems cannot.
5) I emphasize cause-and-effect. I urge readers not to depend upon numbers by themselves. We use numbers to identify cause-and-effect relationships and to determine sensitivities. But we rely on the cause-and-effect relationships that we uncover.

Nassim Taleb spends a lot of time examining why we perceive information incorrectly. This is in common with many of today’s investment books. He looks at this from many vantagepoints.

Understanding our misperceptions helps us explain such things as the value premium and why it persists in spite of the fact that it is well known. Rest assured that Nassim Taleb would never accept any explanation as definitive. Instead, he would search for a flaw that he could exploit for profit.

My impression is that people take out of the book whatever they like or don’t like. Some people become horribly defensive. Some people miss the point entirely, seeing nothing of significance. Some readers have twisted the meaning. Some people are blinded by their personality differences, which is easy to do. They fail to see the merit behind Nassim Taleb’s assertions and arguments.

Accept the fact that Nassim Taleb comes from a different culture. Accept the fact that he has a strong personality. Expect to reject a handful of his opinions. What remains is worth knowing.

Have fun.

John Walter Russell
June 28, 2005.