Cognitive bias parallels in investing and gambling

A cognitive bias is defined as when “individuals draw inferences or adopt beliefs where the evidence for doing so in a logically sound manner is either insufficient or absent”. In other words, it is when human beings act in ways that are not entirely rational.

Cognitive biases exist in both investing and in gambling. In fact, one such cognitive bias is commonly referred to as the gambler’s fallacy. This occurs when a gambler or investor draws inferences about future outcomes based disproportionately on recent outcomes or events. The classic gambling iteration of this bias occurs in roulette. Casinos often will display a series of prior outcomes, including the outcome’s color (red, black or green for zero) adjacent to the roulette wheel. An unsophisticated gambler might assess that a consecutive series of “red” outcomes increases the probability that the next outcome will be “black.” This is a fallacy, since each roulette wheel spin is a statistically independent event with an identical set of probabilities for spin outcomes (48.65% black, 48.65% red and 2.70% green) (Figure 1).

In fact, there might even be a counterargument to be made in such cases. If a coin flipped 999 times always lands on heads, one arguably should consider the likelihood that the coin is weighted or double-headed and that the probability of the next coin flip being heads might not be 50%. The gambler’s fallacy might also appear in certain investment decisions. For instance, if the past nine investment decisions have resulted in losses, the investment manager might infer a higher probability that the next investment decision will generate a gain. (To be fair, in this instance, the investment decisions arguably are not statistically independent events and probably suggest reconsidering the choice of investment manager!)

Loss aversion is another example of a cognitive bias with parallels in both investing and gambling behaviour. Loss aversion is different from risk aversion, which is the preference investors have for lower risk or volatility associated with generating (either higher or lower) returns. Loss aversion bias suggests that the preference for avoiding risk is asymmetric, i.e., investors and gamblers both dislike losses more than they like comparable gains. This phenomenon is clearly evident in the overreaction anomaly, where market sell-offs tend to be more volatile than market rallies.

Disposition effect is another cognitive bias in which investors and gamblers alike tend to avoid realizing losses but seek to realize gains. In gambling, the gambler’s tendency to avoid leaving the casino when he or she is down just 50% of his or her allocated capital is rare (most will play until they have lost 100% of their allotted capital or until – in rare circumstances when their winnings are significant — they have a moment of clarity and cash out).

Investing suffers from a similar disposition effect: Investors have a tendency to sell winning positions but hold onto losing positions. Some might argue that this disposition effect is completely rational and should occur as part of a portfolio’s rebalancing out of outperforming companies or sectors. However, the counterargument is that the losing investment might in fact be a losing investment because the original investment thesis was wrong. Addressing this disposition requires reassessment of the investment portfolio against the original investment thesis to ensure congruency, rather than solely based on avoiding taking losses.

Confirmation bias is the tendency to search for, interpret or recall information in a way that confirms one’s beliefs or hypotheses. In blackjack, confirmation bias appears when the gambler recalls all of the instances when the third base player (the player who is dealt cards last) has a perceived adverse impact on the outcome but never recalls the instances when such impact is favorable. Confirmation bias also appears in bond investing: The investor seeks out research that perhaps supports his or her view on interest rates and places less emphasis on research that challenges that view. Confirmation bias is particularly dangerous in fixed income investments, where the risk profile is asymmetric given the low probability but high impact of default risk. For fixed income investments, the investment manager needs to be open to new information that might alter the analyst’s assessment of credit risks, even if that is not the analyst’s own view.

Finally, overconfidence itself can be a cognitive bias. Gambling unfortunately brings out the worst in human beings’ bias for overconfidence. (There is a reason some casinos offer free drinks to their gambling benefactors.) Investment managers are also not immune to overconfidence. In his 2006 study Behaving Badly, researcher James Montier found that 74% of the 300 professional fund managers surveyed believed they had delivered above-average job performance. Of the remaining 26% surveyed, the majority viewed themselves as average (Figure 2). Incredibly, almost 100% of the survey group believed that their job performance was average or better. Clearly, only 50% of the sample can be above average, suggesting the irrationally high level of overconfidence these fund managers exhibited.

Leave a Reply

Your email address will not be published. Required fields are marked *