Earnings and interest rates may be the two most important ingredients for stock returns, but there is one more underappreciated component: human emotions. Take the disposition effect, for example, which suggests investors have a greater tendency to sell stocks at a gain compared to stocks at a loss. This violates the common market wisdom of “cut losses and hold onto winners,” of course, but the propensity of doing the opposite stems from prospect theory, whereby people fear a loss more so than they are encouraged by a gain. That impacts selling behavior, but what about buying behavior? Researchers from USC’s Marshall School of Business and Caltech published a novel study last year collecting neural data of participants trading stocks to find if regret influences investment decisions. They found that neural regret signals correlate to the size of the repurchase effect of their subjects. The regret investors feel after seeing the price of a stock they recently sold increase discourages them from repurchasing the stock because it would extend the regret they feel by having to buy it again at a higher price. Consequently, the disposition effect and repurchase effect cause investors to be more likely to buy losing rather than winning stocks, and sell stocks with positive price momentum over losing companies. Both, according to the study, are costly to returns.
Note from Nick: “You can always buy it back” is one of the “10 Trading Commandments”. It is meant to encourage good risk management – cutting your losses before they kill you. But as Jessica outlines today, it is easier said then done and thanks to some useful academic work now we know why.
“Always sell what shows you a loss and keep what shows you a profit. That was so obviously the wise thing to do and was so well known to me that even now I marvel at myself for doing the reverse.” Many of our readers may recognize this quote from the classic investment book, Reminiscences of a Stock Operator by Edwin Lefèvre. It chronicles the life of Jesse Livermore, a fictional speculative trader who provides fundamental principles of investing that still influence some of the best traders today despite being published all the way back in 1923.
Here are a couple of other quotes related to the old market mantra of holding onto winners and selling losers:
- “A loss never bothers me after I take it. I forget it overnight. But being wrong, not taking the loss, that is what does the damage to the pocketbook and to the soul.”
- "The only thing to do when a person is wrong is to be right, by ceasing to be wrong. Cut your losses quickly, without hesitation. Don’t waste time. When a stock moves below a mental-stop, sell it immediately. "
Easier said than done, that. Another market motif, “Buy low, sell high,” may also be a simple concept, but behavioral biases often get in our way of making logical choices when investing. Daniel Kahneman and Amos Tversky did much of the groundbreaking in this field with their work on prospect theory, the idea that people are loss averse which ultimately drives their (often poor) decisions. Given that we feel potential losses to a greater degree than potential gains, regret after the fact certainly plays with our psyche when investing as well.
Therefore, researchers Cary Frydman and Colin Camerer from USC’s Marshall School of Business and Caltech published a study last year that sheds unique light on this topic. They created an experimental stock market and collected neural data from participants in order to ascertain whether regret influences investor behavior. Here’s a breakdown of how they conducted the study:
- The researchers used functional magnetic resonance imaging (fMRI) to retrieve neural data in order to record the “precise moment when expost suboptimal trading outcomes are revealed, but before trading decisions are made.” They then tested “whether the regret generated upon news of a stock return is correlated with subsequent trading behavior.”
- Twenty-eight “Caltech subjects” traded three stocks and were given $350 at the beginning of two 16-minute sessions with 108 trials. They had to buy one share of each stock, leaving them with a $50 balance. Each trial was separated into two parts. The first section gave a price update on one of the three stocks, and the second randomly chose one of the three stocks and asked whether the participant wanted to trade it. Subjects only received price updates during the first 9 trials in order to “accumulate some information about the three stocks before having to make any trading decisions.” Participants could either choose to sell a stock or repurchase it, and were allowed to “hold a maximum of one share and a minimum of zero share of each stock at any point in time.”
- “The price path of each stock [was] governed by a hidden-state Markov process with a good state and a bad state.” In the good state, a stock’s price in a trial rose with “probability 0.55 and decreases with probability 0.45.” In a bad state, the opposite occurred: “its price increases with probability 0.45 and decreases with probability 0.55.” The researchers assigned each stock one of these states before the first trial. If a stock’s price was updated, then its state stayed the same as in the previous trial with a probability 0.8 and changed with probability 0.2.
- In other words, “if a stock increased on the last price update, it was probably in a good state for that” and its “next price change is likely to also be positive.” The notion here is that under an optimal strategy, an investor would “sell (or not buy) a stock when he believes that it is more likely to be in the bad state than in the good state; and to buy (or hold) the stock when he believes that it is more likely to be in the good state.” As Jesse Livermore knew, it followed the notion of holding onto or buying the stocks that are working and selling or not buying the ones that aren’t.
- The researchers analyzed the trades by focusing on the “regret-devaluation mechanism.” This theory suggests that investors feel regret after seeing the price of a stock they sold increase. The investor is then less likely to repurchase the stock because it would “prolong the regret, creating an undeniable fact that the stock was sold at a low price and then bought again at a higher price.” This is called the “repurchase effect.” The researchers also assumed the reverse is true: participants would be happy after the price of a stock they sold dropped, and would increase “the expected utility of repurchasing the stock.”
Now that we have laid out their methodology and assumptions, here were their findings:
- The researchers found that “investors do experience regret when receiving information that indicates a trading decision is ex-post suboptimal.” They observed a fall in brain activity in the location where studies have shown a response to “predication errors and regret-related signals” when participants saw that the price of a stock they had sold rose. Additionally, “the strength of this neural regret signal correlates with the size of the repurchase effect across [their] subject pool.” These results showed the size of the repurchase effect varied across participants, but was “greater than the level displayed by an optimal trader for 26 of the 28 subjects.”
- They also found that “neural activity in the vmPFC (our note: a specific part of the human brain) negatively correlates with the forgone capital gain at the moment when a subject is presented with the opportunity to repurchase a stock.” Since “decision values are encoded in the vmPFC,” this showed that respondents’ expected utility of repurchasing a stock is lower if it increased in price after they sold it. Moreover, they discovered that “the specific subarea of the vmPFC that we find to encode the foregone capital gain at a buying opportunity is nearly the exact subarea that we find to encode the capital gain at the time of a selling decision.”
- This led the researchers to analyze the relationship between the repurchase effect and the disposition effect (“a greater tendency to sell stocks at a gain compared to stocks at a loss”). They found the disposition effect was similarly costly as the repurchase effect given that their “experimental design induces positive short-term autocorrelation in price changes” on the basis that strong recent performance will likely continue. Selling a stock with positive price momentum sacrifices potential and likely future gains. Thus, they likened selling winning stocks to holding onto losers – both hurt returns.
- The study’s results also showed a high correlation between the repurchase effect and disposition effect (r=0.71). The researchers concluded that “subjects who exhibit purchasing mistakes are also more likely to exhibit selling mistakes, suggesting that there may be a common psychological mechanism that governs both types of trading mistakes.” For further evidence, they found “traders who are slow to realize losses are also slow to repurchase stocks that have gone up since the last sale” and that “traders who are quick to realize gains are also quick to repurchase stocks that have gone down since last sale.” With that said, they also recognized “traders who are quick to repurchase stocks that have gone down, are not necessarily slow to repurchase stocks that have gone up.”
- Lastly, they did a follow up experiment to find if investors were trading with certain beliefs, such as the idea that stocks revert to the mean. In this case, participants would buy losers and sell recent winners. Even still, they found “a significant number of suboptimal trading decisions” that an “irrational belief” – such as mean reversion – “cannot explain.”
In sum, investors fall privy to both the repurchase effect and the disposition effect as humans are driven by regret and the fear of loss. This causes investors to more likely buy losing rather than winning stocks, and sell winning rather than losing companies. Instead, the optimal strategy is to buy or hold onto stocks with positive price momentum and cut losses by selling stocks with negative price momentum. Even though most traders like Jesse Livermore understand these concepts, emotions often circumvent logic. That’s why keeping behavioral biases in check can be just as important to returns as studying fundamentals and the macro environment.
To end, we’ll leave you with this last Jesse Livermore quote: “It takes a man a long time to learn all the lessons of his mistakes. They say there are two sides to everything. But there is only one side to the stock market; and it is not the bull side or the bear side, but the right side.”