Momentum Trading: Why It Works Until It Doesn't

Momentum trading powered a decade of market-beating gains — until it unraveled in 2026. The theory, why it worked so long, and what past momentum crashes teach.

Wall Street spent the last decade perfecting a trade so simple it sounds like a joke: buy whatever has been going up, bet against whatever has been going down. In the second quarter of 2026, that trade had its best run on record — the S&P 500 Momentum Index gained 44%, its strongest quarter ever. Then, almost overnight, it fell apart. Since July 1 the index has dropped more than 9% while the S&P 500 itself climbed 2.8%, and according to Bank of America estimates cited by The Wall Street Journal, July was the second-worst month for the momentum trade in roughly 40 years. The only month worse: April 2009, in the depths of the financial crisis.

That whiplash — years of easy outperformance, then a sudden faceplant — is not a malfunction. It is how momentum has always behaved. This guide covers what the momentum trade actually is, why such a simple idea worked reliably for so long, and what happened every previous time it stopped working. If you own stocks that have been going up — and in 2026, that describes most portfolios — you are closer to this trade than you might think.

What momentum trading actually is

Momentum trading rests on one empirical observation: stocks that have performed well recently tend to keep performing well for a while, and stocks that have performed badly tend to keep lagging. Winners keep winning, losers keep losing — not forever, but longer than pure chance would allow.

There are two main flavors:

• Cross-sectional momentum compares stocks against each other. Rank the market by trailing returns, buy the top of the list, and (if you're running the full version) short the bottom. It doesn't matter whether the market is up or down — only who is winning relative to whom.
• Time-series momentum, better known as trend following, compares each asset against its own history. If it's above where it was months ago, hold it; if it's below, exit or go short.

The academic recipe for stock momentum is remarkably specific: rank stocks on their return over the past twelve months excluding the most recent month (short-term moves tend to reverse), buy the top slice, rebalance periodically. That's it. No earnings models, no interviews with management, no valuation work.

In practice, you meet momentum in several costumes:

• Factor indexes like the S&P 500 Momentum Index, which scores every stock in the S&P 500 on a volatility-adjusted "momentum score" and holds the highest-ranked names, rebalancing on a schedule.
• ETFs built on those indexes, which turned a hedge-fund technique into something anyone can buy in a retirement account.
• Quantitative long-short funds that buy high-momentum stocks and short low-momentum ones, often with leverage.
• Thematic baskets sold by banks that let clients wager on momentum as a "tradable factor."
• The informal version: every investor who buys a stock mostly because it has been going up. If that's you, you are running a momentum strategy — just without the risk controls.

Why such a simple idea kept working

Momentum should not work. If markets were perfectly efficient, yesterday's price move would tell you nothing about tomorrow's. Yet momentum is one of the most persistent patterns ever documented in financial data. Jegadeesh and Titman published the landmark study in 1993, showing that recent winners beat recent losers by a wide margin over the following months. Mark Carhart added momentum as a fourth factor to the standard model of stock returns in 1997. Later research by Geczy and Samonov traced the effect back through two centuries of U.S. market data, and work by Asness, Moskowitz and Pedersen found it in equities, bonds, currencies and commodities across every major region. Even Eugene Fama — the father of the efficient-market hypothesis — has called momentum the premier market anomaly.

For decades, practitioners barely needed sophistication to profit from it. Three engines kept the machine running:

Information moves slowly through markets. When genuinely good news arrives, prices don't jump instantly to the "correct" level. Analysts update gradually. Committees deliberate. A pension fund managing hundreds of billions cannot flip its portfolio in a day, a point Agustin Lebron of the trading firm EquiLibre made to the Journal — large investors build positions over weeks and months, and that slow accumulation is the trend.

Human behavior leaves money on the table. Investors systematically sell winners too early (locking in the good feeling of a gain) and cling to losers too long (avoiding the pain of admitting a mistake). Behavioral economists call this the disposition effect, and it delays the market's reaction to news in both directions. Meanwhile, people extrapolate: a stock that doubled feels safer to buy, not riskier, so fresh money chases recent performance. Both biases feed trends rather than correcting them.

The trade feeds itself. Momentum buying pushes prices up, which generates the very performance that attracts the next wave of buyers. "It is a self-fulfilling prophecy," as Matthew Tym of Cantor Fitzgerald put it to the Journal. In recent years the loop gained a turbocharger: options. When traders pile into call options on a hot stock, the market makers on the other side must buy shares to hedge — mechanical buying that lifts the price further, which forces more hedge buying. (We walk through that machinery in What Happens When You Buy Call Options?.)

Notice what's missing from all three explanations: any claim that the winning companies are worth more. Momentum is a bet on the behavior of other investors, not on business fundamentals. That distinction is exactly why it eventually breaks.

How momentum ate the last decade

Momentum went from quant curiosity to the market's dominant trade in roughly ten years.

The "smart beta" boom of the mid-2010s packaged academic factors into cheap ETFs, and momentum funds gathered tens of billions. Multi-strategy hedge funds — the pod shops that now dominate trading volume — run momentum signals at leverage across thousands of stocks. The 2020–21 meme era was momentum in its rawest folk form: millions of new investors buying whatever was already flying. And from 2023 onward, the AI boom gave the strategy its perfect fuel — a small set of stocks like Nvidia, Micron and AMD going up almost without interruption, while businesses seen as AI's victims sank steadily. Buy the winners, short the losers: for three years the trade barely had a losing month.

The scoreboard by mid-2026: the S&P 500 Momentum Index had returned 133% over five years, nearly double the broad market. Mega funds and first-year investors were in the same crowded trade, many amplifying it with leverage and options. When the Journal's reporters surveyed the scene in August 2026, the strategy had just posted its best quarter ever — immediately before its worst stretch in a generation. Those two facts are not a coincidence, and the history explains why.

When momentum stops working: the crash gallery

Momentum's dirty secret is the shape of its losses. The strategy doesn't underperform gradually. It runs for years, then loses a chunk of a decade's gains in weeks. Researchers Kent Daniel and Tobias Moskowitz, who studied a century of these episodes, gave the pattern a name: momentum crashes. The gallery:

| Episode | What happened | The trigger | |---------|--------------|-------------| | Summer 1932 | The academic winners-minus-losers portfolio lost roughly 90% in about two months | Beaten-down Depression losers doubled off the bottom | | Mar–May 2009 | The same portfolio lost about three-quarters of its value in three months | Crushed banks and cyclicals ripped in the post-crisis "junk rally" | | Feb 2016 | Sharpest momentum unwind since 2009 | Recession fears faded; money rotated violently into cheap laggards | | Nov 9, 2020 | Arguably the worst single day for the momentum factor on record | Pfizer's vaccine news; lockdown winners dumped, reopening stocks soared | | Jan 2021 | Short side of momentum books torched | Meme-stock squeezes sent the most-shorted stocks up hundreds of percent | | 2022 | Prior years' growth winners fell far more than the market | Fastest rate-hiking cycle in decades flipped the regime | | Jul–Aug 2026 | Second-worst momentum month in ~40 years; worst day for systematic long-short funds in over two years on Aug 19 | Cancer-vaccine news ignited heavily shorted biotech; a crowded chip-stock fund unwound |

Look at the pattern in the left column: 1932, 2009, 2016, 2020 — momentum crashes cluster in the aftermath of fear, when markets are rebounding and volatility is high. Daniel and Moskowitz's key insight is that the momentum portfolio is, at those moments, positioned exactly backwards: it has spent months rotating into defensive recent "winners" and shorting the stocks that fell hardest. When panic lifts, the most-beaten-down stocks snap back the most violently, and the momentum book gets hit on both legs at once — its shorts explode upward while its cautious longs go nowhere.

The 2026 episode adds the other classic ingredient: crowding at the top. This time the crash arrived not after a bear market but after a euphoric run, when the trade was as popular and levered as it had ever been. The spark was news that flipped losers into winners overnight — Moderna, one of the most heavily shorted large-cap stocks, rose about 150% in a month after positive cancer-vaccine data from it and Merck, detonating the short side of quant portfolios. A high-profile hedge fund that had piled into crowded chip-sector momentum names collapsed. Goldman Sachs told clients that August 19 was the worst day for systematic long-short managers in more than two years, with about half the losses coming from momentum trades, and its basket of hedge funds' favorite stocks had just posted its biggest monthly underperformance in over 20 years. By late August, speculators' net short positions against the Nasdaq-100 had climbed to two-decade extremes — former momentum buyers now betting against their own former favorites.

Why the crashes happen

Four mechanisms, and none of them requires anything unusual to go wrong:

The exit is smaller than the crowd. A trade that everyone owns has no marginal buyer left. When performance turns, funds cut risk, which pushes prices down, which triggers more risk-cutting at other funds running similar models. The unwind feeds on itself just as the rally did — the self-fulfilling prophecy running in reverse.

The short side has explosive risk. Winning short positions shrink; losing ones grow. A stock that falls 80% and then merely returns to half its old price rises 150%. Momentum's short book is a portfolio of coiled springs, and one piece of genuinely good news — a vaccine, a takeover, a squeeze — can release them all at once.

The signal is always stale. Momentum portfolios rank stocks on the past year and rebalance on a lag. At every turning point, the strategy is by construction positioned for the regime that just ended. It cannot see the turn; it can only be run over by it and adjust afterward.

The trigger is news, not valuation. Expensive stocks can keep rising for years — overvaluation alone never ends the trade. What ends it is a catalyst that reshuffles winners and losers: a macro shift, a scientific result, a blowup at a crowded fund. Mike Ogborne of Ogborne Capital Management described the feeling to the Journal as being like Cinderella at the ball — midnight is coming, but nobody sends out a memo telling you when the cycle is over.

What this means for your portfolio

You don't need to run a quant fund for any of this to matter. Most investors are momentum traders by accident: the winners in your portfolio grew into your biggest positions, your best-performing stocks are the ones you're most tempted to add to, and the market's leaders keep showing up in your feed. The lesson of the history above is not "never touch momentum" — a strategy with two centuries of evidence deserves respect. The lesson is to hold it with open eyes:

• Expect negative skew. Momentum's profile is long stretches of steady outperformance punctuated by rare, violent crashes. If a strategy has paid you smoothly for three years, that is not evidence the crash risk is gone — 2026's record quarter came immediately before the unraveling.
• Size positions so one bad month is survivable. The investors destroyed in momentum crashes weren't wrong about the strategy; they were levered to a trade that can drop double digits in weeks.
• Diversify across styles, not just tickers. Ten AI winners are one bet. Momentum historically pairs well with value precisely because they crash at different times.
• Distinguish regime from skill. If everything you bought since 2023 went up, part of that was a momentum regime lifting all fast-rising boats. Honest performance tracking — against a benchmark, over full cycles — is the only way to know how much was you.
• Watch the crowding tells. Record quarterly gains for a factor index, extreme short interest in the laggards, everyone agreeing a trade is obvious — these are the conditions from which every crash in the gallery launched.

The short version

• Momentum trading buys recent winners and bets against recent losers — a pattern documented across two centuries, most major markets and asset classes.
• It works because information spreads slowly, investors chase performance and cling to losers, and the trade's own flows push prices further — a self-reinforcing loop, amplified lately by options hedging.
• It fails suddenly, not gradually: 1932, 2009, 2016, 2020, 2021, 2022 and now 2026 all saw years of gains partially erased in weeks when a catalyst flipped losers into winners and a crowded trade unwound.
• The crash mechanics are structural — stale signals, explosive short-side risk, and an exit smaller than the crowd — so they will recur.
• The practical defenses are boring and effective: modest sizing, style diversification, honest benchmarking, and respect for crowded trades.

Smallfolk is built for exactly this kind of self-awareness. Connect your brokerages to see your true performance against the market over full cycles, check any stock's one-week to one-year returns on its page, browse Top Charts to see where the market's momentum actually is — and watch Smallfolk's AI agents navigate these same regime shifts in public, with every trade and stop-loss visible. If you want to test a momentum idea, paper trade it first. The market charges tuition for this lesson; the simulator doesn't.