I have a relative—let’s call her Jane. Last year, in the early days of the pandemic, Jane had the foresight to buy shares in vaccine maker Moderna. With the benefit of hindsight, it was a smart decision. But in Jane’s view, it was not a difficult decision. It was no secret that the company was working on a Covid vaccine. And it was clear that vaccines would be in high demand. That made the investment case clear. And Jane was right. Since then, Moderna stock has risen tenfold. This year, it’s been the single best performing stock in the S&P 500.

So Jane is a happy investor—but she also has a problem: Deciding when to sell has turned out to be a much tougher decision than when to buy. She could, of course, sell now and declare victory. Who wouldn’t be happy with a tenfold gain? But then again, that might be premature. Moderna has lots of other drugs in its pipeline and might be just getting started.

To be sure, this is a good problem to have. But still, it isn’t an easy decision. It turns out, though, that Jane isn’t alone. According to a recent paper titled “Selling Fast and Buying Slow,” professional investors struggle with this same issue. That is, they struggle more with selling than with buying. In fact, they struggle more than you might guess.

Researchers for this paper selected a group of more than 700 large, professionally-managed portfolios, each averaging nearly $600 million. These included pensions, university endowments and the like. Then they examined the trades in these portfolios over a 16-year period. What they found was surprising in a few respects.

First, despite the mantra—and the data—that “active management underperforms,” this study revealed that active managers actually display significant skill when buying. The average stock they purchased went on to outperform by more than one percentage point per year. While that might not sound like a lot, consistent outperformance by a margin like this would be notable. That was the first surprising finding.

The second finding was that these same investors—who were so skilled at buying—completely fell apart when it came to selling. While their purchases substantially outperformed, their sales substantially underperformed.

The researchers wondered how this could be the case—that the same investors who were so good at buying had such a hard time with selling. They were seemingly two sides of the same coin. The authors’ conclusion: “an asymmetric allocation of limited cognitive resources towards buying and away from selling.” In other words, selling isn’t fundamentally harder than buying. Instead, investors simply think harder and put more effort into buying decisions, and that’s why those results are better. How were the researchers able to prove this? In studying millions of trades, they were able to identify four key patterns, each of which supports their conclusion:

1. When portfolio managers used more up-to-date information, they made better decisions. The study found that investors weren’t uniformly bad at selling decisions. One situation in which they did measurably better was when companies issued quarterly earnings reports. On those specific days—when there was more current data available—portfolio managers made much better decisions. Selling decisions on earnings release dates actually outperformed—proof that good selling decisions can be made; they just need to be made using more data.
2. When portfolio managers saw a stock as being important, they made better decisions. How did they know if a stock was important to a manager? The proxy they used was the stock’s percentage weight in the portfolio, the logic being that stocks with bigger weightings have more impact on a portfolio’s overall results and are thus more important. Sure enough, managers made better decisions with those more heavily weighted stocks. In other words, when they felt it was important enough, portfolio managers put in the time to make a better decision, and those efforts paid off.
3. When portfolio managers took their eye off the fundamentals, they made worse decisions. The data revealed that portfolio managers were much more apt to sell a stock that had either outperformed or underperformed by a wide margin in the recent past. In other words, instead of logically evaluating a stock’s future prospects—as investors are taught to do—portfolio managers spent time looking in the rearview mirror. While it might seem like the right thing to do to sell a stock that’s recently outperformed (because it might now be overpriced) or that’s recently underperformed (because it might indicate a problem at the company), it turns out that, on average, these judgments end up hurting rather than helping.
4. When portfolio managers were feeling stressed, they made worse decisions. How could the researchers have known when managers were under stress? The proxy they used was recent performance. If a fund had been doing poorly, they presumed that the portfolio manager would be feeling stress. And during those periods, selling decisions measurably underperformed.

What does all this mean for individual investors? I draw a few conclusions. First, and as you might guess, I see this as another reason to steer clear of trading individual stocks and to instead choose index funds. That can insulate you from agonizing decisions, like Jane’s challenge with her Moderna shares.

For better or worse, index funds that own Moderna stock right now aren’t giving it any thought; they’re simply holding it. Over the long term, that might help or it might hurt. But according to multiple studies, on average it seems to help. That is, it helps to remove the human element. This research helps illustrate why. Investing is a psychological minefield. And if even the most sophisticated fund managers—those running half-billion dollar accounts for pensions and endowments—are susceptible to the challenges outlined above, then no one really is immune. When you own an index fund, on the other hand, you can sidestep a large part of that minefield.

But what if you already own some individual stocks? Then what? I see a number of useful lessons:

First, treat selling with as much deliberation as buying. This seems like it ought to be obvious. But as the authors point out, investors put more effort into buying than selling because buying is inherently more enjoyable. The hunt for new ideas is a discovery process, and it’s interesting. Selling, on the other hand, tends to be viewed as more of a housekeeping chore—to raise cash for a withdrawal, to reallocate to a new idea, or to rebalance a portfolio.

Second, map out a plan with decision rules for selling shares. That can help you avoid reacting to anecdotes, snippets of news or recent performance. You could, for example, employ a dollar cost averaging approach to sales. Instead of trying to make judgments about where a stock is going, simply make a schedule to sell incrementally and mechanistically.

Third, consider how each stock fits into your overall portfolio. If it’s a large position that ends up distorting the overall balance of your portfolio, that’s worth more of your attention than a small holding that won’t hurt much if it underperforms but might help if it outperforms, as explained below.

Finally, and maybe most counterintuitively, be as careful with the small positions in your portfolio as with the large ones. That’s because one of the key ways you can go wrong when you sell a stock is to sell a winner prematurely. Imagine, for example, if you’d sold a small position in Apple or Amazon five years ago. In both of those cases, recent performance had been strong, and they might have looked overpriced. And yet, they went on to deliver such strong results that even a small holding might have made a big difference.

If all of this sounds like a lot of work, I agree. And I think that helps explain the study’s core finding, that investors tend to give short shrift to selling decisions. It’s not that people can’t sell well. It’s just exhausting. But if you venture down that road, I hope these guidelines are of some help.