Investors' memories of past performance are positively biased, according to new research by Daniel Walters, INSEAD Professor of Marketing and Philip Fernbach, Professor of Marketing at the University of Colorado Boulder.
In the paper, titled Investor Memory of Past Performance is Positively Biased and Predicts with Overconfidence, published in Proceedings of the National Academy of Sciences (PNAS), the professors found that investors tend to recall returns as better than achieved and are more likely to recall winners than losers. No published paper has previously shown these effects with investors.
By surveying investors about their interpretation of their past performance over a given time and then comparing those answers with their financial statements, the researchers witnessed over and over again that memory differed from true figures.
The professors concluded that this is the result of two memory biases: “Distortion,” which relates to the inclination to positively amplify reported returns; and “selective forgetting,” a tendency for investors to fail to remember their losses.
They also found that the participants who displayed larger memory biases were more overconfident in their investments. Overconfident investors trade more frequently, and the more they trade, the more they underperformed in the market. While research has examined overconfidence in other areas, such as top management teams and CEOs, this is the first-time investor overconfidence has been detected.
Crucially Walters and Fernbach found that investor overconfidence could be dispelled by exposing investors to their true past returns. When they encouraged half the investors surveyed to review the actual results of previous investments, they found that these participants were less overconfident and planned to trade less frequently than those who relied on just their memory.
“By validating a new methodology for reducing overconfidence we could help investors make better decisions,” explained Professor Walters.
While the study focused on investing, their findings do suggest that such memory distortion and selective forgetting could actually also be associated with overconfidence in many other fields, ranging from CEOs repeatedly overestimating the success of corporate acquisitions to chess players overestimating their future performance. Overconfidence in these areas is surprising because feedback ought to make CEOs and chess players become better calibrated. Biased memory might help to explain why overconfidence is persistent in these domains.
The conclusion: Don’t trust your memory. If you do so, you are likely to have an inflated sense of your own abilities, and this bias can lead to expensive mistakes.