The disposition effect, i.e. the propensity to sell winners and hang on to losers, is an all too human bias that every investor has to manage. But surely, professional fund managers must be able to deal with this bias better than everyday investors. Well, it depends…
During the early stages of my career, I was a value investor and worked in an environment full of people heeding the value investment philosophy. When stocks were cheap we would be buying and we would tell anyone that we would sell these stocks once they become overvalued. It’s the time-honoured way of making money depicted in the chart below.
Value investors try to buy undervalued assets…
Source: Liberum
The problem in practice, though, is less the buying when stocks are undervalued part, it is the selling when they are overvalued. I cannot count the times when we decided to sell existing holdings the moment, they became slightly more expensive than fair value. And I am guilty as charged. When I managed money, I was good at selling portfolio holdings at the first sign of trouble. That may have felt good because I locked in gains before a market downturn, but especially during an equity bull market, it was also counterproductive because, after a short setback, prices would continue to rise, and valuations would become even more expensive.
And according to a study by Can Yilanci that is a very common problem for value managers everywhere. He looked at the quarterly holdings of some 400 US value fund managers and another 400 US growth managers. Based on the fund holdings, he could then calculate how likely fund managers were to sell a stock held at a profit vs. stocks held at a loss. The chart below from his note shows in green bars the propensity of fund managers to sell a stock held at a gain (PGR) or at a loss in red bars (PLR). The difference between these two probabilities measures the disposition effect of the fund manager.
The disposition effect is visible for value but not for growth fund managers
Source: Yilanci (2022)
Note, how on average, value fund managers are more likely to sell stocks held at a profit than stocks held at a low while growth fund managers tend to do the opposite. It is the very nature of their investment style and investment philosophy that growth fund managers tend to hang on to investments for as long as possible in order to benefit from the expected strong growth. Meanwhile, value managers are always concerned about holding overvalued assets and tend to sell them as soon as these holdings made money for them.
Realising this bias exists should help value investors to think about techniques to manage it. I did this by introducing mechanical rules to sell losers sooner than I would normally have done, thus shifting the odds of selling losers vs. winners. Other fund managers have implemented rules where they have a hard and fast rule to sell a stock only after it has gone up 2x or 3x or once according to their metrics the stock is overvalued by x%, where x% is a pretty large number like 50%. This way, they force themselves to hand on to winners for longer than feels necessarily comfortable to them, but they remove the disposition effect that costs them performance.
Very good article, any book recommendation on when to sell an asset
Thanks for this post. Could you elaborate on the mechanical rules that you have created to sell losers sooner?