Momentum investing: a complete, evidence-based guide

If you have ever heard someone say "the trend is your friend" and wondered whether that is just a saying or something with real evidence behind it, here is the short answer: momentum investing is the observed pattern that assets which have been going up tend to keep going up for a while, and assets which have been going down tend to keep going down for a while, before eventually the pattern breaks. It is one of the most studied ideas in finance. It is also, as you will see, a lot less reliable and a lot more dangerous at the edges than the popular version of the idea suggests. For the deep dive into the data and its limits, the evidence for momentum, and what its critics get right is the place to go after this guide.

#Where the idea comes from

The modern story starts with a 1993 paper by two researchers, Narasimhan Jegadeesh and Sheridan Titman. They sorted thousands of U.S. stocks into groups based on how well they had done over the previous 3 to 12 months. Then they checked what happened next. The stocks that had been winners kept outperforming the losers, on average, for the following several months. That was surprising at the time, because the leading theory of markets said prices should already reflect everything knowable, and past price movement is about as public as information gets. If momentum were real and easy to see, in theory it should have been arbitraged away already.

It was not a one-off. Later studies extended the pattern across other countries, other decades, and other asset classes, including currencies and commodities. That breadth is a big part of why momentum is taken seriously by academics who are otherwise skeptical of most "systems." It shows up in Japan and Germany, not just the United States. It shows up in the 1930s and the 2010s. Consistency across time and geography is one of the strongest forms of evidence you can get in finance, where most patterns are one country, one decade, and gone.

#Why would this even happen

Markets are supposed to be full of people trying to spot free money, so a persistent pattern needs an explanation for why it is not immediately traded away. Researchers generally point to a mix of two kinds of causes.

The behavioral explanations are about people, not machines. One is underreaction: when good news comes out about a company, investors do not fully update their view all at once. The price drifts upward over weeks as the news slowly sinks into more and more decisions, rather than jumping instantly to a new fair value. Another is herding: once a trend is visible, more traders pile in simply because others are piling in, which can push a move further than the original news justified. A third is the disposition effect, a well-documented human habit of selling winners too early and holding losers too long, which can slow down how fast information about winners gets fully priced in.

The structural explanations are about the plumbing of markets rather than psychology. Index funds that rebalance on a schedule, mutual fund managers who chase recent winners to avoid looking foolish to their clients, and stop-loss orders that trigger more selling as a price falls, can all create momentum-like patterns as a side effect of how money is mechanically required to move, independent of anyone's opinion about value.

Nobody has proven definitively which explanation matters most, and that honesty matters. A pattern can be real and repeatedly observed while still lacking a fully agreed-upon cause.

#The three ways people actually measure it

"Momentum" is not one number. In practice it is usually captured with a few standard tools, and it helps to know all three because they answer slightly different questions.

• Trailing returns over 3, 6, or 12 months. This is the original academic version: simply, how much has the price moved over that window, compared to its own past. It answers "has this been a winner or a loser lately."
• Moving averages. A moving average is just the average price over a set number of recent days, recalculated every day so it "moves" with the market. The most common one watched by longer-term investors is the 200-day version, and the 200-day moving average, explained in 3 minutes walks through why that particular number became the informal line between an uptrend and a downtrend in so many people's heads.
• Relative strength. This compares one asset's momentum to another's, or to the market as a whole, rather than looking at it in isolation. A stock can be going up in absolute terms and still be a laggard if everything else in its group is going up faster. Relative strength: telling a leader from a laggard covers how to make that comparison properly.

A related but distinct tool is the 52-week range, which shows where today's price sits between its highest and lowest points over the past year. It is not momentum exactly, but it is often read alongside it, and how to read a stock's 52-week range explains what that position can and cannot tell you.

#What this looks like on a real day

Numbers make this less abstract. On a recent day, Momentu's data showed TSMC (ticker TSM) with a momentum score of 61 out of 100, trading around $422.06, with a positive 3-month trend and trading volume running at 0.65 times its own typical 20-day level, meaning quieter than usual. Eli Lilly (LLY) showed a score of 60, price near $1,215.02, also on light volume at 0.56 times normal, but with 25 news articles about the company in the prior 48 hours, well above its usual pace of coverage. Palantir (PLTR) sat at a score of 60 with volume close to normal at 0.89 times its 20-day average and 35 recent articles. JD.com (JD) had a lower score of 57 but the highest activity reading of this group at 72 out of 100, with volume running 1.72 times its own normal level and zero news articles in 48 hours, an unusual combination worth sitting with for a moment: elevated trading with no obvious headline explanation. Rivian (RIVN) rounded out the group with a score of 57 on quiet volume, 0.59 times normal.

Notice what these numbers are and are not doing. Every comparison is against that specific company's own recent history, not against some universal threshold, which is why a giant like TSMC and a much smaller name like Rivian can sit side by side on the same 0-to-100 scale without the bigger company automatically scoring higher just for being large. A score of 60 for Eli Lilly does not mean the stock is about to do anything in particular. It means Eli Lilly's own combination of trend, volume, and attention is running somewhat elevated relative to its own typical range. That is a description of the present, not a forecast of the future.

#Where momentum breaks, and why that matters more than where it works

The uncomfortable part of momentum investing is what happens at turning points. Academic work on "momentum crashes" has found that the strategy's worst losses tend to cluster right after sharp market downturns, exactly when a rebound begins. Stocks that had been falling hardest can snap back violently, and a strategy built around riding the trend gets caught leaning the wrong way at the worst possible moment. This is not a minor footnote. Some of the largest drawdowns in momentum strategies happened within a matter of weeks, wiping out gains that had taken years to build.

For an ordinary investor watching a chart, this often shows up as what looks like a fresh breakout inside what is actually still a longer downtrend, a brief rally that has all the visual excitement of a new uptrend starting but sits inside a bigger picture that has not actually turned. What a "pop inside a downtrend" looks like walks through how to tell the difference between a real shift and a bounce that fades.

Trading volume, the number of shares changing hands, often spikes around these turning points too, which is part of why volume is worth understanding as its own signal rather than folding it entirely into momentum. Unusual volume in stocks: the complete guide goes into how to read that separately.

#What Momentu's own testing actually found, plainly stated

Here is where we want to be very direct, because it is easy for a service like ours to overstate what a signal can do. In a one-year test across 50 stocks, roughly 10,500 individual observations, we found that periods of unusual trading volume preceded price moves that were, on average, about 18 percent larger than that stock's normal moves. That is a real, measurable effect on the size of what comes next. What it did not do is predict direction. Unusual volume did not reliably tell us whether the next move would be up or down, and we did not find that it beat simply following the broader market on that question. We think this distinction, between predicting how much something might move versus which way it will move, is the single most important thing to understand about any momentum-adjacent signal, including our own.

This is also why we would encourage skepticism, including toward us. Before trusting any market signal, including a score you see on Momentu, it is worth knowing what to check first, and how to evaluate any market signal before you trust it lays out the questions that separate a genuinely tested pattern from a plausible-sounding story.

#The honest summary

Momentum investing describes something real: prices that have been trending tend to keep trending, on average, for a while, across markets and decades, for reasons that are probably a mix of human psychology and market mechanics. It is not a promise about any individual stock, and it fails hardest exactly at the moments when following it feels most obvious, right around market turns. Treat every score, chart, and trend line as a description of what has already happened, not a guarantee of what happens next.

If you want to see how these patterns look in real names on an ordinary day, our free daily radar tracks momentum, volume, and attention across dozens of stocks, plainly labeled and updated each morning.

Momentu is an informational tool, not financial advice or a buy/sell recommendation, and it does not consider your personal circumstances. Any decision you make is your own.