This addendum to my previous post What is luck? is a response to the assertion that Warren Buffett could possibly have achieved extraordinary investment returns even if the probability was extraordinarily low. He could have been “incomprehensibly lucky,” using the phrase I used. This is similar to the theory that if you had enough monkeys typing for a long enough period of time, eventually one monkey will author William Shakespeare’s Hamlet.
This is clearly a possibility. As you increase the number of trials in the 1,000 coin flip experiment towards infinity, the probability of getting 99.5% heads at least once approaches 1. Like I said, you can’t completely rule out “luck” to explain Buffett’s performance.
However, rather than accept the conclusion that Buffett was lucky because it was possible that he was lucky, we need to use reasoning to conclude whether this is a likely explanation of the results. One method used by me and others is to determine whether people who consistently outperform the market have similar investment approaches.
The superinvestors of the “Superinvestors of Graham-and-Doddsville” essay clearly share a similar approach to the investment process. This approach appears to differ markedly from that used by the “market.” Some investors just dollar-cost average into an index. Some people use technical analysis. Some people try to use fundamental analysis but are really clueless about what the fundamentals mean. Some people are just plain clueless. Many investors are rather undisciplined and buy or sell at a whim motivated by greed or fear.
Many investors make their investment decisions based on the recommendations of conflicted sell-side analysts. Others don’t take time to fully research a company. Many investors have either positive or negative prejudices about a company that prevent them from being objective. Some investors restrict their investment opportunities based on moral grounds. And few investors seem to be able to insulate themselves from what the market has done recently or is currently doing, which, in my opinion, is a huge reason why most people are incapable of outperforming the market consistently.
As Albert Einstein said, “Few people are capable of expressing with equanimity opinions which differ from the prejudices of their social environment. Most people are even incapable of forming such opinions.” Translation: few people can really think for themselves. Albert knows.
This potpourri of investor strategies and philosophies comes together to determine the market averages. Is it reasonable to think that each investor’s strategy is as good as another’s? I think this is a laughable idea, but it is the conclusion that we must come to if we are to think that no strategy can beat the market consistently outside of luck. If some strategies are superior to others, then some investors would consistently outperform the others.
In other intellectual activities such as chess, business planning, or problem solving, few people would suggest that one strategy is just as good as the next or that the ability to think rationally and creatively is not a huge advantage. So why have so many people come to believe that an intellectual activity like investing is any different?
I think it has to do with what the academics have told us is true. They said, “See? We have evidence. We tested it. You can’t beat the market consistently.” And many of us bought it. We didn’t understand quite how they tested it, but they were “experts” with PhD’s and Nobel Prizes, and that was good enough for us. But I think there is a major problem with the way the hypothesis was tested by academics. Because they couldn’t possibly test whether a comprehensive investment philosophy (including the impact of an investor’s emotions and other extremely important factors) is superior to others, researchers had to simplify the tests dramatically.
To really understand whether the academics’ conclusions make sense, it would be helpful to understand how they tested for it. First, what is the Efficient Market Theory (EMT)? It is simply an hypothesis that suggests that security prices adjust rapidly to new information, making it impossible to outperform the market consistently.
There are actually three forms of the EMT:
1) weak form – all information contained in past price movements is fully reflected in current market prices
2) semi-strong form – current market prices reflect all publicly available information; and
3) strong form – current market prices reflect all pertinent information whether publicly available or privately held.
They tested these ideas in a number of ways. For the weak form, they tested to see if returns over time were independent of each other. One way was to see if there was any correlation of returns over a series of days (autocorrelation). Another test was to apply a “runs test.” They would compare the runs in the price changes (i.e. consecutive price increases or consecutive price decreases) to the number of runs that should occur in a random series. The goal of these tests was to test if investors were able to make abnormal returns from examining price movements. Another type of test was to compare the returns from some technical trading rule to the returns that would be achieved from a buy-and-hold strategy. Many of these technical tests seem uselessly simplified, like testing a trading rule that says buy after a stock rises 5% and sell a stock that declines 5%.
The tests for the semi-strong form test to see if any public information will provide superior estimates of returns for either a short-run or long-run horizon. They tested certain “fundamental” information like dividend yields, default spreads (difference in yields between lower-grade and AAA-rated bonds), term structure (the difference between long-term AAA yields and 1-month Treasury bills), and quarterly earnings (e.g. “earnings surprises”). Many calendar studies have been done to test if buying or selling at certain days or times can lead to superior returns (e.g. the “January Effect”). The number of things studied is tremendous, including price-earnings ratios, price-earnings/growth rate ratios, size effects (i.e. market capitalization), “neglected firms” (based on number of analysts covering the security or its trading volume), book value-market value ratios, stock splits, IPO’s, exchange listings (i.e. liquidity effects), unexpected world events, accounting change announcements, and corporate events (e.g. mergers, security offerings, etc.).
The strong form of the theory is tested to see if abnormal returns can be achieved using non-public information in addition to public information. Tests have been done to see if corporate insiders’ trades achieve abnormal returns. Other tests have tested whether stock exchange specialists achieve abnormal returns by virtue of their monopolistic access to information about unfilled limit orders. Some tests have also been done to test if security analysts have access to private information, or possibly are just superior at identifying undervalued securities.
The conclusions from all these tests generally support the weak and semi-strong forms of the theory, but the support for the strong form is generally much weaker. The question I have to ask is “So what?” These tests don’t even come close to testing the way securities are selected in real life. It shouldn’t be particularly surprising that one little piece of information generally can’t be used to outperform the market. If it were that easy, then everyone would do it, rendering the strategy ineffective. It’s kind of like expecting to find that football teams that blitz heavily on third downs are superior football teams.
That’s not how things work in the real world. Investors take massive amounts of both relevant and irrelevant information, sift through it, think it over, and come to a conclusion. Some investors will use irrelevant information, and some will come to irrational conclusions even if they use relevant information. Certain investors are able to take large amounts of diverse information that create complicated patterns and create a simple mental model out of it. The fact is that some people are just superior to others at analysis and rational thinking.
Warren Buffett had it right when he said:
“Amazingly, EMT was embraced not only by academics, but by many investment professionals and corporate managers as well. Observing correctly that the market was frequently efficient, they went on to conclude incorrectly that it was always efficient. The difference between these propositions is night and day.”
– 1988 Letter to Berkshire Hathaway shareholders
Of course, there is no way to test the theory that superior analysis leads to superior returns, and therein lies the problem. Any such test would be so complicated that the results would be impossible to interpret. In any case, it would be incredibly difficult to model the thought process of a “lucky” investor like Buffett. However, the simpler tests that have been conducted are very unconvincing, and reasoning otherwise suggests that superior strategy should lead to superior results, just like most other areas of life.
I think Eeyore from Winnie the Pooh (A.A. Milne) said it best:
“We can’t all, and some of us don’t. That’s all there is to it.”
Comments on this entry are closed.