Those portfolio managers who always have their "fingers on the trigger" ready to override their investment strategies contrary to their experienced judgment face the risk of cognitive bias.
People tend to override at the wrong times. In most cases, they would have been better off sticking to their investment disciplines. Judgmental overrides of value and fundamental forces driving markets can be very dangerous for performance and are a major cause of investment disappointment.
The reason is cognitive bias, a human phenomenon that is an acknowledged branch of psychology. Cognitive bias represents the tendency of intelligent, well-informed people to consistently do the wrong thing.
Cognitive bias is a subtle but useful tool in decision making in portfolio management. So how can we understand cognitive bias?
For example, the best way to convince someone to buy your favorite stock is to tell them a compelling story.
Investors can identify with stories. The path the story outlines from the company's current activities to its all-but-certain future profits will seem far more probable when presented in an engaging story form. Captivated listeners will naturally reject the myriad of other possible outcomes for the stock, since they cannot visualize a path to these alternatives. For this reason, "story stocks" are so important.
Hersh Shefrin and Meir Statman of Santa Clara University pointed out the implications of cognitive bias for investment managers last year at the Q Group, the Institute for Quantitative Research in Finance, by contrasting small- vs. large-capitalizations stocks.
As studies have shown, stocks of small companies with low price-to-book-value ratios provide excess returns over large-cap, high price-to-book stocks. Therefore, if given a choice between small, cheap stocks and a large, high-priced popular stocks, we should be able to predict that prominent investors will prefer the small, cheap stocks.
The facts suggest otherwise.
Investment professionals, such as financial analysts, senior company executives and company directors, consistently prefer large, high-priced stocks. Such prominent investors typically are not contrarians. These investors make a cognitive error reinforced by an aversion to regret.
Messrs. Shefrin and Statman use in their database a popular survey polling 8,000 senior executives, directors and financial analysts to rate companies in their industry on eight attributes, using a scale of 0 (poor) to 10 (excellent). The attributes are quality of management, quality of products or services, innovativeness, long-term investment value, financial soundness, ability to attract, develop and keep talented people, responsibility to the community and the environment, and wise use of corporate assets.
In one of the most recent surveys, the most admired company was Merck & Co. with an average score of 8.74. The least admired was Wang Laboratories Inc. with an average score of 1.99.
The attribute long-term investment value should be negatively correlated with size since small stocks provide superior returns. However, in the survey of prominent investors, long-term investment value had a positive correlation with size. Similarly, long-term investment value should have a negative correlation with the price/book ratio since low price-book stocks provide superior returns. But in the survey of prominent investors, long-term investment value had a positive correlation with price-book.
This result, certainly not a single-year phenomenon, has been stable over time.
Thus, prominent investors overwhelmingly believe that good companies make good stocks. This belief is representativeness, a cognitive bias heuristic. The representativeness heuristic evaluates the probability of an uncertain event by the degree which it is (1) similar in essential properties to its parent population and (2) reflects the salient features by which it is generated.
The study shows prominent investors overestimate the probability that a good company is a good stock because prominent investors rely on the representativeness heuristic. Good stocks are similar to good companies. Or, good stocks are now good companies. Therefore, they conclude incorrectly that good companies are good stocks.
The preference for good companies is also based on another cognitive bias, aversion to regret. What is aversion to regret? Peter drives to work following his usual route and has an accident; another driver ran a stop sign and dented the fender of Peter's car. The same day, John drives to work, deviating from his usual route. John also has an accident; another driver ran a stop sign and dented the fender of John's car.
John will have greater regret about the accident. John will feel partially responsible since he took a new route. John can more easily imagine taking the usual route (and not having the accident) than Peter. Regret is acute when the an individual must take responsibility for the final outcome.
Aversion to regret is different from aversion to risk. Consider our commuters, Peter and John. The alternative routes to work have equal probability of safe arrival. Risk is not the issue.
Aversion to regret leads to a preference for stocks of good companies. The choice of the stocks of bad companies involves more personal responsibility and higher potential for regret. Consider Merck and Wang Labs, the best and worst stocks in the survey. Merck is safe and respectable. If you own Merck and it goes down, it is due to unforeseeable events. You feel no regret. Wang's problems are highly publicized. Everyone knows it is a bad stock - bad company. Owning Wang seems more likely to lead to regret. Stocks whose selection involves little responsibility can be expected to yield lower returns than stocks involving a lot of responsibility. We have a double cognitive error: Good companies make good stocks (representativeness) and involve less responsibility (less aversion to regret).
The reason for cognitive bias, the tendency of intelligent, well-informed people to consistently do the wrong thing, is the fact that the human brain is not built for processing statistics. Our brains handle visual images much better than lists of numbers. It is the way we are and, in many cases, simplifies our daily lives.
If a child burned a finger by touching a hot stove, you would expect the child to learn from that single experience. The next time that the child is near the stove, the child will recall the pain and will not touch the stove. You would be quite surprised if the child said, "Well, that was painful. But, based on a single data point, I still do not believe the hot stove is dangerous. I must touch the hot stove a hundred times to get a statistically significant sample."
The child who learns quickly, often from individual experiences (single data points), is usually able to function quite well in daily life. In short, cognitive bias is part of all of us.
Because of cognitive bias, investment managers need disciplines, tools and quantitative techniques to allow us to be successful contrarian investors. It is too easy to do what everyone else is doing, to succumb to cognitive bias and own the popular investments with little responsibility. Quantitative investment techniques provide the discipline to enable investment managers to overcome cognitive bias, follow sound investment principles by investing in value and be successful contrarian investors.
Richard A. Crowell is vice chairman and chief investment officer, PanAgora Asset Management Inc., Boston.