Dark Bilious Vapors

But forget for a moment all of the factors that influence stock prices. Imagine, instead, that the market is a coin. Assemble 1,000 people and tell them that, to assess their powers of prediction, you are going to flip the coin. When they protest—coin-flipping is unpredictable—lie to them. Say that your coin flips will be predictable. Say that some of them—those who whose powers of perception allow them to spot patterns and behavior others can't see—will be able to predict the flips with greater-than-average frequency. If there is no deliberation before the first flip (i.e., if no one makes a compelling argument about why the coin will land on one side or the other), approximately 500 people should call "heads" and 500 "tails." Flip the coin.

When the coin lands, post the names of the "winners" on a gigantic scoreboard. Call for a new round of predictions and flip the coin again. And so on. After five flips, the scoreboard should display the names of about 30 people who will have been right five times in a row, along with hundreds of others who have been right two or three times. If you were to conduct interviews, you would probably find people starting to believe that those on the scoreboard were better-than-average flip predictors. Flip some more.

After eight flips, you should have about four people with perfect records left. After nine, two. After 10, one. One person whose powers of prediction are apparently so acute that he or she has correctly called the coin an astonishing 10 times in a row. Assuming this person's success has been noticed after flips four, five, or six, you can imagine the accolades he or she might now receive, especially if, prior to every flip, he or she has explained the prediction ("The prevailing air currents, combined with the average height of the previous tosses and the average rate of the previous spins, will cause the coin to develop such momentum on the downside that, even if it bounces on tails, it will flip up and land on heads. … ").

Everyone will want to know the genius' secrets—especially the poor sod who, via the same probabilities, will have been "wrong" every time. In truth, of course, this savant isn't good: He or she is just lucky. From luck alone, one person in 1,000 should correctly predict a 50-50 outcome 10 times in a row—and, more important, 500 of the 1,000 should make the "correct" prediction half the time. The difference between the stock market and a coin is that, unlike coin flipping, the market seems inherently predictable (and, for some participants, it is—at least more predictable than a coin flip). Recall, however, that like a coin flip, the potential market "outcome" is binary. Recall that only a tiny fraction of the investors are right almost all the time. Recall that, to be a successful investor, you don't have to be right every time, or even most of the time—you just have to be right some of the time. And next time you hear a forecaster predict that the market is going to go up, remind yourself that if a monkey were making this prediction, it would usually be right.
--Henry Blodget [slate.msn.com]

Len on 09.17.04 @ 06:54 AM CST