When your stocks leave you feeling disappointed, should you hold on to them or get rid of them? Most investors plump for the selling solution. But three researchers, including NEOMA’s Eser Arisoy, have shown that the more disappointing a stock is, the better its rate of return down the line! The scientists ratified this so-called “regret premium” by analysing stock market prices for over a period of almost 60 years and by further confirming their results based on real-life trading transactions of 78,000 US households over the period from 1991 to 1996.

In psychology, regret is defined as the negative emotion – a mixture of pain and anger – that people experience when they recognise that a different decision would have led to a better outcome. And it’s a fate suffered by anyone who invests in the stock market: share prices fluctuate up and down, and it’s a game you can’t win all the time.

A regret premium hidden in plain sight

A number of studies have examined the impact of regret on the behaviour of investors. They have shown, for instance, that investors experience more regret about holding on to a losing asset for too long than about selling a winning asset too soon; or that their regret is also rooted in comparisons between the stock they hold and the stock they could have acquired.

By contrast, no previous research had looked into the relationship between investor regret and portfolio performance. And it is exactly this that makes the new study so fascinating, especially since its conclusion is counter-intuitive: the more tears you shed over an asset, the better it subsequently performs! In other words, there really is such a thing as a regret premium.

What is the gap between my stocks and the top performers in the same category?

But how can you measure an investor’s emotions objectively? The authors made the decision to base their research on comparisons. They quantified regret by calculating the difference in gains or losses over one month between stock held by an investor and stocks in the same industry that had the highest score over the same period.

For instance, let’s say an investor owns stock in a PC company that goes up 2% in one month. If the top performing manufacturer rises by 3%, regret has a value of 1 (3% – 2%), but if it leaps by 15%, regret is valued at 13. It is low in the first instance, and high in the second.

The more regret a stock generates, the more it is underpriced.

Why focus on a single industry? Because investors – especially private investors – don’t follow a financial market across the board: they hone in on a small pool of industrial sectors and companies. It is within this subset that they make comparisons and identify their low-regret and high-regret stocks.

This calibration is critical: investors are averse to disappointment, which means they turn their backs on high-regret stocks, investing instead in their low-regret equivalent. It follows that the former are bought cheaply, which will then increase future capital gains – and this is the much-vaunted premium. The latter, which are in high demand, become more costly, which reduces their upside potential.

Regret premium and risk premium: don’t muddle them up

Naturally enough, the authors checked that the high-regret stock premium was not linked to the risk premium, which is widely encountered among investors: the more risky a stock is thought to be, the more likely it is to have a high rate of return (… or make heavy losses). Mathematical checks performed on benchmark portfolios showed that they are separate phenomena: the regret premium is persistent irrespective of the risk level of the stock.

Nor is it the consequence of an everyday catch-up effect. We might have thought that a stock that is trailing behind other stocks in the same sector might eventually make up the lost ground. And yet, the regret premium persists when mathematically deducing the impact of this catch-up mechanism.

The level of regret predicts future returns

There is an alternative explanation: the regret premium might reflect an investor’s fear of losing. It would then be higher when the investor sees their stocks fall more than others than when they go up (but less than others). This isn’t the case: the two situations (loss or gain) elicit a regret premium that is more or less the same.

The authors analysed the stock market prices from 1963 to 2020 as well as the trading activity of 78,000 US households between 1991 and 1996to test whether or not their theory was connected to other factors: the type of industry, the geographical location of a company, stock market conditions, economic cycles, and so forth. And what verdict did they reach? The level of investor regret is always a robust predictor of future returns over a maximum period of five months irrespective of how these factors change.

Higher premium for younger, smaller firms

Anyone keen to cash in on this discovery by targeting high-regret premiums will be interested to learn that they should prioritise shares in small, young companies that are largely ignored by analysts and seldom held by institutional investors.

Conversely, shares in major, well-established firms that are closely followed by analysts and well represented in institutional portfolios generate smaller regret premiums. And this makes sense: information about these companies is easily accessible, which means it is immediately incorporated into stock prices. You can’t expect miracles.

But watch out: the regret premium isn’t a novel winning formula for making it big on the stock market. The authors remind us that high-regret stocks start by underperforming before beginning to hit the big time. This means that their overall track record isn’t so fantastic – unless, that is, you buy them after the disappointing period. What’s more, shares in small, young firms don’t attract many investors. Anyone who holds them can’t know for sure that they will be able to sell them at the best price when the time comes…

Find out more

2024. E. Arisoy, T. G. Bali and Y. Tang, Investor Regret and Stock Returns, Management Science, January 2024. https://doi.org/10.1287/mnsc.2022.03389