Momentum Investing Explained: Why Trends Persist and How to Use Them

Executive Summary

Momentum investing is a strategy built on a simple observation: assets that have performed well over the recent past have, on average, tended to keep performing well a little longer, and recent losers have tended to keep lagging. A momentum investing strategy formalizes this into rules — rank stocks by their trailing returns, tilt toward the winners, avoid or short the losers, and refresh the rankings on a fixed schedule.

That idea sounds almost too simple, which is exactly why the academic world was startled when Narasimhan Jegadeesh and Sheridan Titman documented it rigorously in 1993. Since then, momentum has been found in U.S. stocks, international stocks, bonds, currencies, commodities, and across more than a century of data. Eugene Fama, the father of efficient-market theory, has called it the premier anomaly — the one result his models struggle most to explain.

But this is not a free lunch. Momentum strategies suffer infrequent, severe drawdowns — "momentum crashes" — that can erase years of gains in a matter of months, and real-world implementation costs and taxes are substantial. This guide covers the evidence honestly: what the research actually shows, why the effect likely exists, how relative and absolute momentum differ, what the standard 12-1 signal is, how crashes happen, how blending momentum with value and quality helps, and the practical rules (including rebalance cadence) that separate a disciplined strategy from performance chasing.

What Momentum Investing Is

Momentum investing is the systematic practice of tilting a portfolio toward assets with strong recent performance and away from assets with weak recent performance. It is the deliberate opposite of the instinct to "buy what has fallen because it must be due for a bounce." A momentum investor accepts that yesterday's winners are, on average, more likely to be tomorrow's winners over intermediate horizons — and builds rules around that tendency rather than trading on hunches.

Two clarifications matter immediately. First, momentum is a relative, probabilistic edge, not a law of physics. It describes an average tendency across hundreds of stocks and many decades; it says nothing certain about any single stock or any single year. Second, momentum is about price behavior, not business quality. A stock can rank highly on momentum while being an average business, and a wonderful business can rank poorly after a drawdown. That distinction becomes important when we discuss blending factors later.

Relative momentum versus absolute momentum

There are two distinct flavors, and confusing them leads to real mistakes:

Most of the famous academic literature — and most momentum index funds — use relative momentum. Absolute momentum, sometimes called trend-following or time-series momentum, answers a different question and is closer in spirit to what commodity trading advisors have done for decades. The two are complementary: relative momentum tells you what to own, absolute momentum tells you whether to be exposed at all.

The 12-1 month convention

If you read almost any momentum paper, you will encounter the 12-1 signal: rank stocks on their total return over the past 12 months, excluding the most recent month. Why skip a month? Because very short horizons behave differently. At a one-month horizon, stock returns historically show reversal — last month's big winners tend to give a little back as liquidity effects and bid-ask bounce wash out (documented by Jegadeesh in 1990). Momentum lives in the intermediate zone, roughly 3 to 12 months. Skipping the freshest month keeps the signal clean of that short-term reversal noise. The resulting portfolio is then typically held for one to three months before being re-ranked.

The Academic Evidence

Momentum is not a backtest marketing gimmick; it is one of the most replicated findings in empirical finance. Here is the core lineage, stated conservatively.

Jegadeesh and Titman (1993): the founding result

In "Returns to Buying Winners and Selling Losers," published in the Journal of Finance, Jegadeesh and Titman sorted U.S. stocks on returns over formation periods of 3 to 12 months and held the resulting winner-minus-loser portfolios for 3 to 12 months. The classic 12-month formation / 3-month holding version earned on the order of roughly 1% per month on a zero-cost long-short basis over their 1965–1989 sample. The strategies were significant across many sub-periods and could not be explained by the standard risk model of the time.

Carhart (1997): momentum becomes a factor

Mark Carhart showed that adding a momentum factor to the Fama-French three-factor model substantially improved its ability to explain mutual fund returns. The "UMD" factor (Up Minus Down) entered the standard toolkit, and you can download its long history yourself from the Kenneth French Data Library. Over the very long U.S. sample, the gross long-short momentum premium has averaged in the mid-to-high single digits annually — impressive, though that figure ignores trading costs, which are material, as we will see.

Everywhere and always: the out-of-sample tests

The strongest argument for momentum is how well it has survived out-of-sample scrutiny after 1993:

• International equities: Rouwenhorst (1998) found momentum across 12 European markets; later work extended it to most developed markets.
• Other asset classes: Moskowitz, Ooi, and Pedersen (2012) documented time-series momentum in 58 liquid instruments across equity indices, currencies, commodities, and bonds — the effect was positive in every asset class studied.
• Combined with value: Asness, Moskowitz, and Pedersen (2013), "Value and Momentum Everywhere," showed the two factors appear together across markets and, crucially, tend to be negatively correlated with each other.
• Deep history: Researchers have reconstructed momentum back into the 1800s using U.S. and U.K. data and found it present long before anyone could have been trading on it — an argument against pure data-mining.

Fama and French themselves, while noting momentum was the great embarrassment to their three-factor model, confirmed its robustness in U.S. and international data. When the leading skeptics of anomalies concede an effect, it is worth paying attention.

Why Trends Persist: The Behavioral Engine

Why should something so simple keep working decades after publication? The honest answer is that nobody knows for certain, but two families of explanation dominate, and they are not mutually exclusive.

Underreaction: prices move in installments

The leading behavioral story is underreaction. Investors anchor on old information and update too slowly when genuinely new information arrives — an earnings surprise, a contract win, a guidance change. Instead of the price jumping immediately to its new fair level, it drifts there over weeks or months as analysts revise estimates and investors gradually accept the news. That drift is momentum. Supporting this, Novy-Marx (2015) and others showed that much of price momentum is really earnings momentum in disguise: stocks with the biggest positive earnings surprises keep drifting upward after the announcement (a related, older effect called post-earnings-announcement drift).

Herding and feedback trading

The second story is about what happens once a trend is visible. Funds face career incentives to own what is working; retail investors chase what is in the headlines; corporate managers issue shares and make acquisitions when their stock is beloved. These flows push prices further in the direction of the trend — herding and feedback trading. This can extend momentum beyond what underreaction alone would produce, but it also plants the seed of the reversal: at some point the winners become overextended, crowded, and vulnerable to a sharp snap-back. Momentum, in this telling, contains the mechanism of its own crashes.

The risk-based counter-argument

Efficient-market sympathizers offer a different explanation: momentum returns may simply be compensation for bearing a peculiar kind of risk — the risk of catastrophic, concentrated losses exactly when markets are rebounding violently (the crashes covered next). If the premium is really a risk premium, it should be expected to persist, but also to be deeply uncomfortable to harvest. There is also a hard-nosed capacity argument: momentum strategies are costly to run at scale, so the gross premium may survive precisely because arbitraging it away is expensive and dangerous. Honest practitioners hold all three stories at once: momentum probably reflects a mix of behavioral error, genuine risk, and implementation friction.

Momentum Crashes and Drawdowns

Any article about momentum that skips this section is selling you something. Momentum's defining weakness is that its losses, when they come, are fast, deep, and clustered at the worst psychological moments.

What a momentum crash looks like

Daniel and Moskowitz (2016), "Momentum Crashes," studied the long-short momentum factor across nearly a century and found its returns are strongly negatively skewed: long strings of steady gains punctuated by rare, brutal reversals. The two worst episodes in their study were 1932 and 2009, and they share a signature. Both occurred at the trough of a bear market, just as the market began a violent rebound. Here is the mechanism:

1. A long bear market grinds down the past year's losers — often the most leveraged, lowest-quality, highest-beta companies.
2. The momentum portfolio ends up short exactly those beaten-down names (and long the defensive prior winners).
3. The market bottoms and rips upward. The most beaten-down junk rallies hardest — in some rebound months of 1932 and 2009, prior losers outperformed prior winners by enormous margins.
4. The momentum portfolio is positioned precisely backwards, and loses on both legs at once.

In the sharp rebound of 2009, simple U.S. momentum strategies suffered drawdowns on the order of 40% or more in academic reconstructions — in a year when the broad market finished strongly positive. The 1932 episode was worse still. These are the moments that give momentum its grim nickname among practitioners: it works "like clockwork, until it works like a car crash."

Volatility is the warning light

Daniel and Moskowitz also found that crashes are forecastable in a limited sense: they cluster in periods of high market volatility following declines. Relatedly, Barroso and Santa-Clara (2015) showed that scaling momentum exposure inversely to its own recent volatility — holding less when momentum itself has been swinging wildly — historically would have converted much of the crash risk into a smoother ride. Volatility management is not a magic shield, but the research is clear that momentum risk is time-varying, and that taking a fixed-size bet through a high-volatility rebound is the most dangerous version of the strategy.

A hypothetical illustration

Imagine a simple monthly momentum portfolio in a fictional market. Through a grinding two-year bear market, it is long steady defensive names like the imaginary "Northwind Components" and short collapsed cyclicals like "Harborline Freight." The bear market ends; Harborline, down 85% from its peak, doubles in four months off the bottom while Northwind merely drifts. The long-short portfolio loses heavily on both sides in its worst quarter ever — even though nothing about the logic of the strategy changed. An investor who understood this in advance might hold through it at reduced size; one who did not is likely to exit at the bottom of the drawdown, locking in the loss right before the strategy's historically strong recovery periods.

Blending Momentum with Value and Quality

Because pure momentum is feast-or-famine, most serious practitioners combine it with other factors. The research case for blending is unusually strong.

Momentum plus value: natural hedges

Asness, Moskowitz, and Pedersen (2013) documented that value and momentum returns are negatively correlated: momentum tends to struggle exactly when value shines (post-crash rebounds, when cheap beaten-down stocks rally hardest) and to shine when value struggles (grinding trending markets, when expensive growth keeps getting more expensive). Two return streams that each have a positive long-run average but zig when the other zags combine into a portfolio with materially better risk-adjusted characteristics than either alone. This is why so many quantitative multi-factor funds pair them. It is also psychologically complementary: value buying feels like bargain hunting and momentum buying feels like strength-following, so a blend keeps you from feeling permanently out of step with the market's mood.

Momentum plus quality: the "QMJ" insight

A parallel body of research (Asness, Frazzini, and Pedersen on "Quality Minus Junk") finds that tilting toward profitable, stable, growing businesses adds returns that are only loosely related to value and momentum. Blending momentum with quality screens out some of momentum's most dangerous holdings — the bankrupt-adjacent rally candidates — and replaces raw price-chasing with "strength in a business that deserves it." Many modern momentum index products quietly incorporate quality-like refinements (risk-adjusting the momentum signal, for instance) for exactly this reason.

The practical takeaway is that a reasonable investor rarely needs to choose between factors. Blends — whether through multi-factor funds or by holding a value-tilted core with a momentum sleeve — have historically produced smoother outcomes than any single factor at full throttle. Our ETF guide explains how factor ETFs package these tilts, and the stock research framework shows how momentum sits alongside valuation and quality in a broader process.

Practical Rules and Rebalance Cadence

If you take one section literally, make it this one — but take it as a framework, not a prescription. The details below reflect common practice in the academic literature and in momentum index construction.

Signal construction

• Horizon: Use the 12-1 convention — trailing 12-month return excluding the most recent month. Some practitioners blend 6- and 12-month signals; evidence for very short or very long horizons is weaker.
• Risk adjustment: Dividing the trailing return by the stock's volatility rewards smooth trends over violent ones and historically improves the signal. The MSCI momentum indices, for example, use volatility-adjusted 6- and 12-month returns.
• Selection: Take the top decile or quintile of a liquid universe. Exclude micro-caps, where trading costs overwhelm the signal.
• Diversification: Hold dozens of names, not five. Single-stock momentum is mostly stock-specific luck.

Rebalance cadence

Momentum decays: a winner portfolio assembled a year ago is no longer a winner portfolio. The research convention is monthly rebalancing for the pure factor; in practice, many index implementations rebalance semi-annually with conditional interim rebalances after big market moves, accepting some signal decay in exchange for much lower turnover. For an individual investor, the sensible range is monthly to quarterly, with quarterly often the better cost-adjusted choice in small, taxable accounts. Whatever you choose, fix the schedule in advance and follow it mechanically. The moment you start overriding the signal because a holding "feels extended," you have left momentum investing and returned to gut-feel trading — which has no documented premium at all.

Entry, exit, and the role of absolute momentum

• Entries and exits are the same rule. You hold a stock while it ranks; when it falls out of the top ranks, you sell it and buy the new entrant. There is no separate "sell discipline" to invent.
• Consider an absolute-momentum overlay at the portfolio level: if the broad market's trailing 12-month return is negative, some investors reduce equity exposure or raise cash. This is essentially trend-following risk management, and historically it has reduced deep drawdowns at the cost of occasionally being whipsawed in choppy markets.
• Size it honestly. Given crash risk, many thoughtful practitioners treat momentum as a sleeve of a portfolio — say a minority allocation alongside a diversified core — rather than the whole portfolio. You can see how a rules-based, score-driven approach to stocks is presented on our how it works page and in the public track record.

Costs, Taxes, and What Eats the Premium

The gap between momentum on paper and momentum in a brokerage account is wide, and it is the most under-discussed part of the strategy.

Turnover and transaction costs

A monthly rebalanced momentum portfolio can turn over several times per year — far more than a buy-and-hold index. Each round trip pays bid-ask spreads, commissions (where they still exist), and, most importantly, market impact: momentum buys tend to be stocks already rising, so you are systematically buying into upward pressure and selling into downward pressure. Academic studies of implementation costs find that realistic trading friction can consume a substantial fraction — in some estimates, the majority — of the gross long-short premium at institutional scale. Individuals trading small amounts of liquid large-caps face lower impact costs than a giant fund, but the direction of the problem is the same.

The tax drag

In taxable accounts, momentum's short holding periods mean most gains are realized as short-term capital gains, taxed at ordinary income rates in the U.S. rather than the lower long-term rate. A strategy that outperforms by a few percentage points before tax can easily underperform a lazy index fund after tax. This is a serious, quantifiable drag — researchers including Israel and Moskowitz have examined whether momentum survives after taxes, and the honest summary is "by less than you'd hope, and it depends heavily on your bracket." Practical mitigations include holding momentum strategies inside tax-advantaged accounts (IRAs, 401(k)s), using quarterly rather than monthly rebalancing, harvesting losses along the way, and preferring momentum ETFs, whose in-kind redemption mechanics defer much of the tax that a do-it-yourself version would realize.

Fees and product choices

For most individuals, the realistic implementations are momentum factor ETFs and mutual funds, which handle rebalancing, risk adjustment, and tax management internally for an expense ratio typically well under 1% — or a diversified low-cost index core with a modest momentum tilt. Compare any product's expense ratio, turnover, and tax-cost history before assuming the full factor premium lands in your account. Whatever route you choose, the discipline is identical: understand the signal, respect the crash risk, size it as a sleeve, rebalance on a schedule, and measure your results against realistic after-cost, after-tax benchmarks — not against the gross numbers in the papers.

To explore how quantitative signals are scored and screened in practice, browse the 360head Stockiq stock research tools and the current top signals, and read the research framework guide for how momentum fits alongside valuation and quality in a complete process.