The two books of Taleb together have been able to drag out these words many a time from my lips. I am infinitely grateful to him for writing a book which has obviously been a product of several years of intense introspection.
One of the ideas from “I never thought of it that way!” school was the problem of silent evidence. This problem is best illuminated by a story which Taleb tells in "The Black Swan:"
Diagoras, a nonbeliever in the gods, was shown painted tablets bearing the portraits of some worshippers who prayed, then survived a subsequent shipwreck. The implication was that praying protects you from drowning.[/i][i]Diagoras asked, “Where are the pictures of those who prayed, then drowned?”What Diagoras was pointing to is called "the problem of silent evidence." The worshiper who drowned did not have any pictures to prove that they prayed and drowned. They are evidence of the fact that “praying does not protect from drowning” but are silent because they are now at the bottom of the ocean.
The problem of silent evidence is prevalent in every conceivable area one can think of. Malcolm Gladwell in his essay, “The Talent Myth,” describes a research done by McKinsey & Company with the aim of documenting how the top companies in the U.S. differed from the way they hired and promoted their talent compared to other, less successful ones. The test concluded that “[The top companies] singled out and segregated their stars, rewarding them disproportionately, and pushing them into ever more senior positions.”
Gladwell goes on to describe the management of Enron and how closely they followed the mantra from McKinsey & Co. Enron went bust at the end of 2001. Enron, a poster child for disproportionately rewarding talent, played by the conclusions of McKinsey’s report. McKinsey conducted 20 different projects at Enron. Its directors regularly attended Enron's board meetings and the CEO Ken Lay was a previous McKinsey partner, and still the company did not survive. It now is the “silent evidence” for the failure of the theory.
The experiment itself was flawed in this sense from the beginning. Granted that all the top companies in the U.S. handled their talent in a pampered way, what about the companies who failed even after doing the same? What about the Enrons of the past? They were not part of the sample because they did not appear on McKinsey’s radar.
Let us take the New York Times bestselling book, “The Millionaire Next Door.” In this book Dr. Thomas J. Stanley aims to “reveal the secrets of building wealth in America.” The methodology of doing so is as follows. Dr. Stanley collects data from several millionaires on how they live, eat, shop and save. He comes up with the conclusion that most of the time the wealth is a result of “hard work, diligent savings and living below your means.”
Again, this methodology runs into the problem of silent evidence. Looking at a set of people and trying to explain their success does not give the complete picture. What about all the people who did everything correctly but their assets declined in value ?
* * *
Yesterday I was reading the essay “Blowing Up” by Malcolm Gladwell on Nassim Taleb. And I was quite surprised by the following passage:
Warren Buffett was known as the "sage of Omaha" because it seemed incontrovertible that if you started with nothing and ended up with billions then you had to be smarter than everyone else: Buffett was successful for a reason. Yet how could you know, Taleb wondered, whether that reason was responsible for someone's success, or simply a rationalization invented after the fact?It hit me that Gladwell seems to be under the impression that Taleb is implying that Buffett might have been just plain lucky. And sure enough after a few lines on George Soros, Gladwell explains himself:
For Taleb, then, the question why someone was a success in the financial marketplace was vexing. Taleb could do the arithmetic in his head. Suppose that there were ten thousand investment managers out there, which is not an outlandish number, and that every year half of them, entirely by chance, made money and half of them, entirely by chance, lost money. And suppose that every year the losers were tossed out, and the game replayed with those who remained. At the end of five years, there would be three hundred and thirteen people who had made money in every one of those years, and after ten years there would be nine people who had made money every single year in a row, all out of pure luck.I disagree with that conclusion entirely. It is not supported by any math that Buffett might have been plain lucky.
Let us humor Gladwell and take his argument to its inevitable conclusion. The most recent performance tally of Berkshire is given in Buffett’s 2011 letter to shareholders. Let us forget the amount Buffett has beaten the S&P over the last 46 years (which is 10.6% cumulative each year). Let us assume that each coin toss represents a year and winning the coin toss means that you beat the S&P that year. The following table gives the odds of winning as the number of years progresses.
|No. of coin tosses (no. of years)||Odds|
|10||1 in 1024|
|18||1 in 262,144|
|22||1 in 4,194,304|
|26||1 in 67,108,864|
|30||1 in 1,073,741,824|
|34||1 in 17,179,869,184|
The population of the world in 1975 (10 years after Buffett started in 1965) touched 4 billion. The odds for winning 34 years in a row is around 1 in 17 billion.
Following is the table of the Buffett partnership performance from this article.
|Year||Overall Results From Dow||Partnership Results||Limited Partners' Results|
Following is the table of the years in which Buffett was beaten by S&P.
|Year||Relative performance with S&P|
Let us take Buffett’s performance during 1957 to 2011. And because the partnership intersects with Berkshire Hathaway (NYSE:BRK.A)(NYSE:BRK.B), we will take the worst of the situations, i.e. in 1967 we assume that Buffett lost the toss.
Over the 54 years Buffett then has had 8 down years and 46 up years. In coin-tossing terms that means 46 heads and 8 tails. If a coin is tossed 54 times (assuming that it is equally likely to result in a head or a tail), getting 46 heads and 8 tails has odds of 1 in 17,313,782. The population of the entire U.S. in 1957 was 172 million. To get a track record of Buffett by sheer luck, one needs to start with 17 million managers in 1957! I don’t think that 1 out of 10 people in the U.S. were managers in 1957.
Furthermore, to actually match Buffett one needs to beat the S&P by a significant margin (10.6% cumulatively between 1965 and 2011). The probability of doing this will be significantly smaller but sadly, I do not have the necessary data to run that calculation.
In either case, Warren Buffett cannot be generated randomly.