Want to be a better investor? Start learning from your mistakes

Remember when a special teacher in your life told you it was good to make mistakes as long as you could learn from them? It is probably one of the most important maxims for investors—and probably one of the most ignored. The evidence that investors don’t learn from mistakes is widespread. Every few years a new mania takes hold and investors bid up an asset class to unrealistic heights, such as in the dot.com stock market of the late 1990s or the recent residential real estate boom. Or, investors go through periods when all notion of risk is thrown out the window, and they are willing to pay high prices for securities previously seen as very risky.

It seems our overly protective mental structures serve to shield us from the pain of admitting mistakes, some cognitive psychological research has found. One protective attribute is called “self-attribution bias,” or, to put it in plain English, “Flipping heads is skill, but flipping tails is just bad luck.” How many times have you purchased or sold a stock with a good result and told yourself it was a result of your own superior investment knowledge, while a loss on a stock you attributed to bad luck or the market?

Another protective device is hindsight bias, the tendency to feel after the fact that we knew the outcome all along. Consider again the dot.com boom, when certain Internet stocks were priced at hundreds of times their projected earnings, vs. a typical average for the market of 12 to 14 times earnings. Years after the bubble burst, many investors say they saw it coming, even though the majority participated in the bubble at the time.

We also seem to have a hard time accepting that much of what happens in the world is probabilistic, i.e., the world is uncertain and there is not always clear cause and effect. We tend to associate probabilistic events in the world with our own actions, even though our actions likely had no effect on the real world outcomes.

Tests by psychologists have found that subjects engage in ritualistic and superstitious behavior to influence the outcome of a random action. This happened after a subject had a chance positive occurrence early on in the test. From then on the subject associated his action with the random result. Rather than look at the percentage of times a particular strategy paid off, we are prone to fall into the illusion that an early positive correlation will continue into the future.

There are several techniques an investor can use to better learn from past actions. You should keep a record of decisions as they are made, along with your reasoning. When the outcome is known, you can go back and see whether your justifications had anything to do with the outcome, or whether you experienced a chance coincidence or were just plain wrong. You can separate the outcomes four ways. If you were right, were you right for the right reason (in other words, did you display actual skill), or were you right but your reason was wrong, indicating chance? If you were wrong, was it for the wrong reason (learn from it) or did you lose even though your reasoning was correct?