By LARRY SWEDROE
In their 1985 study, The Disposition to Sell Winners Too Early and Ride Losers Too Long: Theory and Evidence, Hersh Shefrin and Meir Statman introduced into the behavioural finance literature the disposition effect — the tendency of individual investors to sell assets that have increased in value while keeping assets that have dropped in value. This trading behaviour has been documented using evidence from both individual and institutional investors across different asset markets and around the world.
In their 2018 study, Are Investors Really Reluctant to Realize Their Losses? Trading Responses to Past Returns and the Disposition Effect, Itzhak Ben-David and David Hirshleifer investigated this behavioural pattern further and found that the probability of selling as a function of profit is actually V-shaped: At short holding periods, investors are more likely to sell big losers than small ones — individual investors trade in response to the as well as the of profits (gains or losses). They also found that the V is asymmetric: The right branch is steeper than the left branch — the main source of the disposition effect. They also found that the Vs for both selling and buying are steeper for frequent traders and for male investors (who tend to be more overconfident)—consistent with a speculative motive for trading as a contributor to the V. And they found little evidence that tax-loss selling steepens the left branch in the last quarter or last month of the year.
New research
Li An and Bronson Argyle contribute to the behavioural finance literature with their study Overselling Winners and Losers: How Mutual Fund Managers' Trading Behavior Affects Asset Prices, published in the September 2021 issue of the Journal of Financial Markets, in which they examined the selling pattern of mutual fund managers to determine if they exhibited the same trading behaviours found in individual investors. They began by noting: “While studying the trading behaviour of retail investors is interesting and significant, mutual fund managers command much more capital, are arguably more sophisticated than retail investors, and play a larger role in deciding equilibrium prices.” Their database is from Thomson Reuters and covers the period 1980-2018. Following is a summary of their findings:
U.S. mutual fund managers, like presumably less sophisticated retail investors, are more likely to sell holdings with large unrealised gains and losses rather than those with small unrealised gains and losses (they exhibit a V-shaped pattern).
The V-shaped pattern creates selling pressure, pushing down current prices (for non-fundamental reasons), leading to higher future returns — their behaviour depresses the price of affected securities. As future prices revert to fundamental values, affected stocks will outperform, creating return predictability in the cross-section.
Aggregating across funds, securities for which investors have large unrealised gains and losses outperform in the subsequent month. The price effect is both economically and statistically significant — a 10 percentage point increase in the aggregate unrealised gains (losses) for a stock predicts a 9 (5) basis point increase in next month’s returns. (The price effect of gain overhang is 1.8 times as large as that of loss overhang.) A long-short portfolio strategy based on this effect can generate a monthly alpha of approximately 0.5 percent, with a Sharpe ratio equal to 1.2.
The results were strongest in smaller cap stocks.
Fund managers’ selling responses to unrealised profit weakens as the holding time becomes longer — consistent with findings on retail investors.
Funds with larger turnover, shorter holding period and higher expense ratios are significantly more likely to manifest this trading pattern, and unrealised profits from such funds have stronger return predictability.
Decomposing security-level unrealised gains and losses into those from “more-disposition-prone” funds and those from “less-disposition-prone” funds, they found that unrealised profits from the more-disposition-prone funds are stronger in predicting future returns.
The price impact of the V-shaped trading behaviour is unrelated to flow pressure — the trading tendency of mutual fund managers is the source of the price deviation from fundamentals.
Investors’ selling response to unrealised profit is not likely be the source of the momentum effect, as loss overhang and the loser leg in momentum have opposite return predictions.
Interestingly, An and Argyle found that stocks with large gain and loss overhangs tend to be those with high idiosyncratic volatility (IVOL). Yet, the historical evidence demonstrates that stocks with high IVOL tend to have low future returns. However, when An and Argyle controlled for IVOL, the results of their predictive overhang variables were strengthened.
Investor takeaways
An and Argyle demonstrated that retail investors are not alone in exhibiting V-shaped trading behaviour that affects asset prices in predictable ways — actively managed mutual fund managers are just as guilty of being subject to behavioural biases, as they trade in response to the magnitude as well as the direction of their gains or losses. And that behaviour negatively impacts their performance, just as it does the performance of individual investors. Thus, their evidence provides another contributing factor to the persistently poor performance of actively managed funds. For individual investors who still trade individual stocks, their findings should at least provide a wake-up call. Being aware of a bias is the first and necessary step in avoiding actions based on that bias. Forewarned is forearmed.
Have you discovered The TEBI Podcast? You’ll find it on SoundCloud, Spotify, Apple Podcasts, Podbean or your preferred podcast provider.
© The Evidence-Based Investor MMXXIV. All rights reserved. Unauthorised use and/ or duplication of this material without express and written permission is strictly prohibited.