Jegadeesh–Titman (1993): The Research That Changed How We Understand Stock Momentum

For a long time, one of the basic ideas in investing was that markets are efficient. According to the Efficient Market Hypothesis (EMH), stock prices should quickly reflect available information. If a stock has already performed very well, investors should not be able to systematically earn extra returns simply by buying it based on its past performance. In other words, yesterday's winners should not predict tomorrow's winners. However, the research paper “Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency” by Narasimhan Jegadeesh and Sheridan Titman (1993) challenged this belief in an important way. Their study showed evidence of momentum in stock prices—stocks that had performed well in the recent past tended to continue performing well for some time, while stocks that had performed poorly tended to continue performing poorly.

The central idea of the paper is surprisingly simple. Imagine that we divide stocks into two groups based on their performance during the previous 3 to 12 months. The stocks that performed best are called winners, while the stocks that performed worst are called losers. Jegadeesh and Titman found that if an investor bought the recent winners and simultaneously sold or avoided the recent losers, the investor could earn significant abnormal returns over the following several months. This strategy became known as a momentum strategy.

The important point is that the researchers were not simply saying that stocks which are good companies perform well. Their finding was much more specific: recent price performance itself contained information about future relative performance. A stock that had recently gone up strongly had a tendency to continue outperforming a stock that had recently gone down strongly.

The researchers tested several strategies using historical U.S. stock-market data. They formed portfolios according to stocks' past returns and then examined how those portfolios performed in the future. They particularly studied formation periods and holding periods ranging from approximately 3 to 12 months. The results showed a clear pattern: recent winners continued to outperform recent losers during the subsequent holding period.

This was important because it appeared to contradict the simplest interpretation of market efficiency. If all relevant information was immediately incorporated into stock prices, why should past returns help predict future returns? If a stock had already risen substantially, investors should have no systematic reason to expect it to keep rising simply because it had risen before.

The momentum effect suggested that something else could be happening in financial markets.

One possible explanation comes from investor psychology. Investors do not always process information perfectly or immediately. When new information about a company appears, investors may initially react too slowly. Suppose a company announces unexpectedly strong earnings. The market may recognize that the news is positive, but investors may underestimate how important the information is for the company's future profits. The stock price may therefore rise gradually rather than adjusting completely in one moment. As more investors recognize the significance of the information, demand for the stock increases and its price continues to rise.

This creates a kind of underreaction. The market does respond to information, but it may not respond completely or immediately.

Another psychological explanation is herding or feedback trading. Investors often observe what other investors are doing. When a stock begins rising, its success attracts attention. More investors may become interested in it, which creates additional buying pressure. That buying can push the price higher, attracting even more investors. In this way, a trend can continue for a period of time.

This does not necessarily mean that investors are irrational in a simple sense. Rather, it suggests that human decision-making can interact with markets in complicated ways. Investors may pay attention to recent performance, become more confident in successful stocks, or react gradually to new information. These behaviors can contribute to price trends.

The Jegadeesh–Titman study therefore became important not only for finance but also for our understanding of behavioral finance. It helped demonstrate that financial markets could display patterns that were difficult to explain using a perfectly efficient-market framework.

However, the story becomes even more interesting when we look at what happened after the initial momentum period.

The researchers found that momentum did not continue forever. A portfolio of recent winners could outperform recent losers for several months, but the effect eventually weakened. Over longer horizons, another phenomenon could appear: long-term return reversal. Stocks that had performed extremely well over much longer periods could eventually underperform, while previous losers could recover.

This creates an interesting pattern in stock-price behavior. In the short to medium term, prices can show momentum: winners continue to win and losers continue to lose. But over a much longer horizon, prices can show reversal: previous winners may eventually become losers and previous losers may eventually become winners.

This distinction is extremely important. It tells us that stock-price behavior cannot simply be described as “prices always continue in the same direction.” Instead, the relationship between past and future returns depends heavily on the time horizon being studied.

For example, imagine three stocks. Stock A rises 30%, Stock B rises 5%, and Stock C falls 20% over the previous six months. A momentum investor would rank Stock A as a winner and Stock C as a loser. Based on the Jegadeesh–Titman evidence, the investor would expect Stock A to have a better chance of continuing to outperform Stock C over the following several months. But this does not mean Stock A will continue rising forever. At sufficiently long horizons, the relationship may weaken or even reverse.

This is one reason the paper became so influential. It changed the question investors asked. Instead of simply asking, “Is this stock fundamentally good?”, investors could also ask, “How has this stock behaved recently, and does that behavior contain information about its future performance?”

The research also had an important practical implication. It suggested that an investor could construct a systematic portfolio rather than trying to predict the exact future price of every individual stock. The investor could rank stocks according to their recent returns, buy the strongest performers, and sell or underweight the weakest performers. The strategy was based on relative performance, not on predicting a precise target price.

This idea eventually became one of the foundations of quantitative investing. Modern quantitative strategies frequently rank securities according to momentum measures and construct portfolios based on those rankings. The basic principle can be summarized as:

Buy relative winners and avoid or sell relative losers.

But there is an important warning. A momentum strategy is not a guaranteed money-making machine. Momentum can experience periods of severe losses, particularly when market conditions change suddenly. A portfolio that is heavily invested in recent winners can suffer when those trends reverse quickly. Therefore, momentum involves risk as well as potential return.

Another important issue is transaction costs. If an investor constantly changes positions because stocks move in and out of the winner and loser groups, brokerage costs, bid-ask spreads, taxes, and market impact can reduce actual returns. Therefore, a strategy that looks attractive in historical data may produce lower returns when implemented in the real world.

The paper also raises a deeper question about the meaning of market efficiency. The existence of momentum does not automatically prove that financial markets are completely inefficient. Markets can be highly competitive and still contain predictable patterns. Once investors discover a profitable pattern, they may attempt to exploit it. Their trading can then reduce the opportunity.

This creates an interesting cycle:

A market anomaly is discovered → investors trade on it → the trading changes prices → the anomaly may become weaker.

This helps explain why part of the momentum effect could dissipate over time. Once a strategy becomes widely known, investors may begin using it themselves. The increased competition can make the original opportunity less powerful.

The Jegadeesh–Titman research therefore has two sides. On one side, it provides evidence that past stock returns can contain predictive information, at least over certain horizons. On the other side, it demonstrates how difficult it is to maintain an abnormal return once a market pattern becomes known.

The study also became an important bridge between traditional finance and behavioral finance. Traditional models often assume that investors respond rationally and immediately to information. Behavioral explanations suggest that investors can make predictable psychological errors. Underreaction, overreaction, attention effects, herding, and other behavioral forces may influence how prices evolve.

The momentum phenomenon is particularly interesting because it appears to occupy a middle ground between underreaction and overreaction. In the short and medium term, investors may underreact to information, allowing trends to continue. But over very long periods, excessive optimism or pessimism may contribute to overreaction, eventually producing reversals.

Therefore, one possible story is:

Information arrives → investors initially underreact → prices gradually move in the same direction → momentum develops → sentiment becomes excessive → eventually prices reverse.

This framework provides a powerful way of thinking about financial markets.

The broader lesson from Jegadeesh and Titman's research is that financial markets are not simply machines that instantly convert information into perfectly accurate prices. Markets are collections of human decisions. Investors have different expectations, different information-processing abilities, different levels of confidence, and different reactions to uncertainty. These differences can create temporary patterns in prices.

The paper is also historically important because it helped establish momentum as one of the most influential empirical findings in modern asset pricing. Momentum later became one of the major factors studied alongside concepts such as value, size, and market risk.

In simple words, the research tells us something that sounds almost strange:

A stock that has recently been winning may continue winning for a while—not because its past performance magically causes its future performance, but because information, investor behavior, expectations, and trading decisions can cause price movements to persist.

At the same time, the research warns us not to extrapolate trends forever. What works over six months may not work over six years. A strategy can be profitable under one market environment and vulnerable under another.

The lasting importance of the Jegadeesh–Titman (1993) paper is therefore not merely that it discovered a trading strategy. Its deeper contribution was showing that the relationship between past and future stock returns is more complicated than traditional theories suggested. Short- and medium-term momentum can coexist with longer-term reversal, creating a fascinating pattern in financial markets.

Ultimately, the paper changed the way researchers and investors think about price trends. It showed that yesterday's winners can remain tomorrow's winners for a while, but no trend lasts forever. Understanding why that happens takes us beyond mathematics and into the psychology of investors, the speed at which information spreads, the interaction between expectations and prices, and the constantly changing competitive environment of financial markets.

The core idea in one line

Jegadeesh and Titman (1993) showed that recent winners tend to outperform recent losers over the next several months, providing powerful evidence for stock-price momentum and challenging the idea that past returns contain no useful information about future relative performance.

---