Ask most investors to explain their trades and they’ll give you a list of reasonable-sounding justifications, but the strong disagreements between experienced, successful investors tell you that it doesn’t all come down to logic. The disposition effect is a cognitive bias that has been found to affect all kinds of investors, but recent research from Empiritrage found that it drops off significantly for traders with more than five years of experience.

Retail Investors

Investors exhibiting disposition effect

“Both experienced investors and inexperienced students exhibit the disposition effect,” writes researcher Wesley Gray, adding that “investors with more than five years of experience had a lower propensity toward [it].”

The disposition effect is when investors more likely to realize a profit than a loss, and will hold on to stocks that are losing value while selling stocks that are gaining value. Ideally, investment decisions should be based on the present financial state of the company. Taking your own personal starting price into account biases trade decisions.

To test whether experience could change the disposition effect, Gray divided participants into three groups: students who had classroom knowledge of finance but no actual experience, traders with 2 – 5 years’ experience, and traders with more than 5 years’ experience. He then had each of them manage a stock portfolio using a simulator based on the Sao Paulo exchange. His team found that a majority of participants in all groups showed bias in their trades, but that the effect was slightly less pronounced among the most experienced traders, with 69.2% demonstrating a bias in line with the disposition effect compared to 76.3% of students.

Experienced traders’ outperformance

Experienced traders also outperformed the students with cumulative returns of 230.8% against 201.6% over twenty years. Gray didn’t try to prove that there was a correlation between the disposition effect and cumulative returns (there is certainly a lot more that goes into portfolio management than this one cognitive bias), but it seems like it could be a factor.

He also points out that the participants weren’t spending their own money, so they might not have behaved as they would in real life. This is especially interesting because cognitive biases tend to play on our emotions, so if anything they would be suppressed during a simulation. The Empiritrage study should probably be seen as a floor for how much this bias affects people instead of an average measure.