Momentum Anomaly

The momentum anomaly is the empirical pattern that stocks with the strongest recent returns tend to continue outperforming over the next 3–12 months, defying the efficient market hypothesis’s prediction that returns should be unpredictable. Momentum is one of the oldest and most robust market anomalies, and it remains unexplained by traditional finance theory.

The empirical fact

In 1993, finance researcher Narasimhan Jegadeesh and Sheridan Titman published a landmark paper documenting that US stocks ranked in the top decile (best performers) over the prior 3–12 months delivered annualized returns nearly 1% per month higher than the bottom decile (worst performers) over the subsequent 6–12 months. This outperformance persisted after accounting for market risk (beta), industry exposure, and firm size. The pattern was not a fluke—it held across multiple time periods and geographies.

The implication is stark: past returns predict future returns, at least over medium-term horizons. An investor who bought the best-performing stocks and shorted the worst-performing stocks would earn abnormal profits. The stock market, under efficient markets doctrine, should not permit this.

Momentum persists today. Academics and practitioners have documented the effect across asset classes: equities, bonds, commodities, foreign currencies, and cryptocurrencies. It is one of the few anomalies robust enough to underpin a live investment strategy—momentum factor investing is a multi-billion-dollar industry.

Why efficient markets predicted something else

The efficient market hypothesis asserts that asset prices fully reflect all available information. If so, price changes should be unpredictable. An investor learns that Tesla stocks rose 50% over the past year. Under efficiency, that information is already in the price; knowing it gives no edge in predicting the next year’s return.

Traditional finance also relies on mean reversion logic. If a stock has outperformed, it is either overvalued or has benefited from good luck. Both conditions are temporary. Overvaluation will correct. Luck will revert. Future returns should be lower, not higher.

Momentum violates both intuitions. It says: past winners stay overvalued for months, not immediately correct. Past luck continues, rather than reverts. The efficient markets framework was wrong, or at least incomplete.

Behavioral explanations

Finance researchers have offered several behavioural accounts of momentum:

Underreaction and gradual information diffusion: When a company announces good news (a successful product launch, a better-than-expected earnings report), the stock price does not immediately jump to its new equilibrium. Investors underreact—either because they miss the news, misinterpret it, or doubt it. Over weeks or months, more investors learn and act on the information, pushing the price higher. A momentum strategy captures this delayed adjustment.

Herding and trend chasing: Some investors follow price trends mechanically. When a stock rises, they perceive it as a signal of quality or momentum and buy. Their buying pushes the price higher, validating the trend and attracting more followers. This self-reinforcing dynamic can push prices above fundamental value, but the feedback continues for months before exhaustion.

Overconfidence and representativeness: Investors often overweight recent performance when assessing a company’s quality. A stock that has risen sharply feels like a winner and appears representative of a good investment. Investors overestimate the durability of past success and overcommit to it, driving further price appreciation.

Loss aversion and regret: An investor who did not buy a surging stock regrets missing the move. As it continues upward, regret aversion pushes them to chase the trend, arriving late but still contributing to momentum.

Reversals and the limits of momentum

Momentum is not eternal. Over very short horizons (days to weeks), prices exhibit reversals—bounces after sharp declines. Over very long horizons (2–5 years), mean reversion dominates; past winners underperform past losers. The sweet spot for momentum is 3–12 months.

This inversion—from reversal to momentum to reversal—suggests different market regimes. Over short horizons, price bounces reflect liquidity and overreaction. Over medium horizons, underreaction and gradual information diffusion drive momentum. Over long horizons, valuation and fundamentals reassert themselves.

Momentum also crashes periodically. In August 2020, during the COVID-19 selloff, momentum strategies suffered historic losses. High-momentum (recent winner) stocks plummeted while the worst performers rebounded—a brief but violent reversal. This tail risk is real and humbles momentum investors regularly.

Profiting from momentum

A typical momentum strategy ranks all stocks by past return over a lookback period (say, the prior 12 months, excluding the most recent month to avoid short-term reversals). It then buys the top quintile (best performers) and shorts the bottom quintile (worst performers), holding for 3–6 months before rebalancing.

Transaction costs and slippage erode returns. Momentum portfolios turn over frequently, incurring brokerage and market-impact costs. In simpler markets or backtests, momentum looks spectacular. In live trading with realistic costs, it is less impressive but still attractive to professional investors.

Momentum also exhibits factor correlation with other anomalies. It tends to work better in value-tilted portfolios and worse in growth-tilted ones. It has been observed to crash when volatility spikes, particularly in systemic crises. Sophisticated momentum practitioners adjust their strategy for market regime and diversify away from pure price-trend systems.

The anomaly that anomalies cannot explain

Despite decades of research, no single explanation fully accounts for momentum. It is too large to be merely transaction costs or data mining artifacts. It is too persistent across markets and time periods to be a temporary mispricing. Yet it also does not fit neatly into a single behavioral or rational framework.

Some researchers have argued that momentum is not an anomaly at all—that it reflects a genuine risk premium that the traditional capital asset pricing model fails to capture. Others contend that it is purely behavioral—a product of investor psychology that will eventually fade as markets become more efficient and participants internalize the pattern.

The truth likely involves both. Momentum captures part of a risk premium (trending stocks may be genuinely riskier than static ones) and part of a behavioral mispricing (investors underestimate the durability of trends). Understanding which element dominates in any given market regime remains an open question.

See also