G is for Gambler's Fallacy

The Gambler's Fallacy occurs when someone believes that a run of a certain type in a random process makes a reversion more likely. A run of reds on a roulette wheel doesn't make black any more likely next time, but people persist in believing that it will. Equally a run of daily losses on a stock doesn't mean a profit is guaranteed tomorrow.

Example

In 1913, at the famed Monte Carlo Casino, black came up on the roulette wheel an astonishing fifteen times in a row. Gamblers rushed to bet on red ... and black came up again. They redoubled their bets on red ... and black came up again ... and again. By the time red finally did come up, on the twenty-seventh spin of the wheel, the casino was millions better off. If stock prices are essentially random, and most evidence suggests they are, then any given run is just as likely to be because of luck than it is because of some underlying cause.

Causes

The main reason for the gambler's fallacy is that people don't understand statistics - or at least don't understand when they're dealing with a random process and when they're not. Investing using charts requires a belief that the processes behind stock movements aren't random: but the thinking is flawed, even if the results are correct (after all, red could have turned up on the sixteenth spin). And arguing that the charts reflect human psychology is also flawed - after all, we're part of the system: if it's random, we're to blame.

The idea that stock prices follow a random walk isn't universally accepted, but if there are patterns within the overall chaos it seems that making people aware of them leads to them disappearing. The accrual anomaly, for instance, has all but vanished, arbitraged away by hedge funds.

Mitigation

In the short-term stocks and stockmarkets are roulette wheels but in the long-term they're not. Careful research and attention to valuation metrics and competitive advantage can reveal corporations that should outperform in the long-run and good levels of diversification will ensure those companies that don't succeed don't badly damage the overall returns. And otherwise a buy and hold portfolio of low cost index trackers across a diversified range of asset classes will mitigate the risk nicely, as long as you can afford to hold for a long time: 20 years should do the trick.

And also: Recency: Hot Hands and the Gambler's Fallacy
Read more: A Liberal Arts Investor's Reading List

The A to Z of Behavioral Bias