Momentum Trading Strategies: How to Ride Market Trends with Data

What Is Momentum in Trading?

Momentum is the tendency of assets that have outperformed recently to continue outperforming in the near future, and vice versa.

This sounds like chasing performance — and conventional wisdom says not to do that. Yet momentum is one of the most extensively documented anomalies in academic finance. The Nobel laureate Eugene Fama called it the premier challenge to market efficiency. Jegadeesh and Titman (1993) demonstrated that buying past winners and shorting past losers produced significant positive returns across US equities. Hundreds of follow-up studies replicated it across time periods and markets.

Why does momentum persist?

• Investor underreaction: Markets initially underreact to information. Prices adjust slowly, creating a directional trend.
• Herding behavior: Institutional investors chase performance, amplifying moves.
• Trend-following: As more participants observe a trend, they join it — self-fulfilling to a degree.
• Risk premium: Momentum stocks may carry genuine risk (they are often expensive on valuation metrics), and the return is compensation for that risk.

---

Two Types of Momentum

Time-Series Momentum (Absolute Momentum)

Time-series momentum asks: Has this asset gone up or down over the past N months?

If an asset's return over the past 12 months is positive, the signal is positive (trend is up). If negative, the signal is negative (trend is down).

Signal = Return(t-12, t-1)   [exclude the most recent month to avoid short-term reversal]

Buy  if Signal > 0
Sell or hold cash if Signal < 0

Why exclude the last month? Research shows there is a short-term reversal effect over 1–4 weeks. Including the most recent month in momentum lookback reduces performance.

Time-series momentum is the foundation of trend-following strategies used by CTA (commodity trading advisor) hedge funds. It works across asset classes: equities, bonds, commodities, currencies.

Cross-Sectional Momentum (Relative Momentum)

Cross-sectional momentum asks: Which assets in this universe have outperformed each other?

Instead of trading each asset based on its own history, you rank all assets in a universe by their 12-1 month return and:

• Buy the top quintile (top 20% performers)
• Sell or short the bottom quintile (bottom 20% performers)
• Rebalance monthly

Rank all stocks by 12-1 month return
Long:  Top 20% (winners)
Short: Bottom 20% (losers)
Rebalance: Monthly

This is the original Jegadeesh-Titman momentum strategy. It exploits relative performance differences within a group of assets, independent of whether the market as a whole is rising or falling.

---

Measuring Momentum

Simple Return-Based Momentum

The cleanest measure:

Momentum(12,1) = (Price(t-1) / Price(t-13)) - 1

This calculates the total return from 13 months ago to 1 month ago (skipping the most recent month).

Risk-Adjusted Momentum

Instead of raw return, use return per unit of risk:

Risk-adjusted momentum = Return(12,1) / Volatility(12,1)

This ranks assets by their Sharpe-like momentum score, not just return. Low-volatility winners rank higher than high-volatility winners with the same absolute return.

Research shows risk-adjusted momentum produces better Sharpe ratios and lower crash risk than simple momentum.

Dual Momentum (Gary Antonacci)

Combines time-series and cross-sectional momentum:

1. Cross-sectional test: Does the asset outperform alternatives in its universe?
2. Time-series test: Does the asset have positive absolute return?

Only hold the asset if both conditions are met. Otherwise, hold cash or bonds.

This significantly reduces the crash risk of momentum strategies.

---

Momentum Strategy Performance

Based on decades of academic research and practitioner replication:

Metric	Cross-Sectional Momentum	Time-Series Momentum
Avg annual return	8–15% above benchmark	10–20% (asset class dependent)
Sharpe ratio	0.4–0.8	0.6–1.2
Max drawdown	40–60% during momentum crashes	20–35%
Best environment	Trending markets, bull or bear	Any persistent trend
Worst environment	Momentum crashes (sharp reversals)	Choppy, whipsawing markets

---

The Momentum Crash Problem

The biggest risk in momentum is the momentum crash — a sharp, rapid reversal that devastates momentum portfolios.

The most severe occurred in:

• March 2009: As the market bottomed during the Global Financial Crisis, momentum portfolios (short beaten-down financials, long defensives) suffered catastrophic losses as the pattern reversed violently.
• 2020 COVID crash and recovery: Momentum portfolios were positioned defensively. The rapid market recovery in April–May 2020 triggered a major drawdown.

Why crashes happen: Momentum is essentially long volatility. When volatility spikes suddenly (a crash), momentum unwinds quickly as participants exit simultaneously.

Mitigating Momentum Crash Risk

Several approaches reduce crash vulnerability:

Approach	Mechanism
Time-series filter	Go to cash when asset shows negative absolute momentum
Volatility scaling	Reduce position size when volatility is elevated
Skip short-term reversal	Exclude last 1–4 weeks from momentum signal
Sector-neutral	Neutralize sector exposures to avoid crowded sector bets
Blend with value	Momentum and value are negatively correlated — combining smooths drawdowns

---

Practical Implementation

Universe Selection

Not all stocks are equally suited to momentum. Research shows momentum works best in:

• Mid and large-cap liquid stocks (small-caps have higher transaction costs that eat the premium)
• Stocks with high institutional coverage (more herding behavior)
• Developed market equities (well-documented across US, Europe, Japan)

Avoid momentum in illiquid names where slippage will destroy the edge.

Rebalancing Frequency

Frequency	Trade count	Transaction costs	Performance
Weekly	Very high	Very high	Typically net negative
Monthly	Moderate	Manageable	Optimal for cross-sectional
Quarterly	Low	Low	Captures longer-horizon momentum

Monthly rebalancing is the industry standard for cross-sectional momentum strategies.

Position Sizing in Momentum

Most implementations use equal weighting within the winner/loser portfolios. More sophisticated approaches use:

• Inverse volatility weighting: Lower allocation to high-volatility winners
• Signal strength weighting: Higher allocation to stocks with stronger momentum scores

---

Momentum in Quantitative Systems

Individual stock momentum is one layer. Professional quantitative systems stack multiple momentum signals:

1. Price momentum (12-1 month return)
2. Earnings revision momentum (upward EPS estimate revisions)
3. Technical momentum (RSI trend, MACD direction)
4. Short-term reversal (1-month reversal as separate factor)

When multiple momentum signals align, the signal confidence is higher.

Gilito evaluates momentum configurations at multiple time horizons — from 1-week to 52-week lookback — per asset daily, ranking them by statistical robustness. This surfaces the specific momentum window that is currently working for each asset, rather than applying a one-size-fits-all lookback.

---

Frequently Asked Questions

Is momentum the same as trend following? Related but distinct. Time-series momentum is essentially trend following applied to a single asset. Cross-sectional momentum ranks assets relative to each other. Professional trend-following CTAs primarily use time-series momentum across multiple asset classes.

Does momentum work in bear markets? Time-series momentum works in bear markets — it simply goes short or to cash when the trend is negative. Cross-sectional momentum works too, going long defensive and short cyclical sectors. However, momentum crashes (rapid reversals) are most severe around market bottoms.

What period should I use for momentum? The 12-1 month window (12-month lookback, skip the last month) is the most academically validated. Some practitioners use 6-month or 3-month lookback. Test both in your specific universe.

Does momentum work for ETFs and indices, not just individual stocks? Yes. Dual Momentum and other strategies applied to broad asset class ETFs (US stocks, international stocks, bonds, REITs) have shown strong historical performance with lower transaction costs than individual stocks.

How much of a portfolio should be in momentum strategies? Due to crash risk, momentum strategies benefit from diversification. Many quantitative allocators combine momentum with value and quality factors, with momentum typically representing 20–40% of total factor exposure.

---

The Bottom Line

Momentum is not a market inefficiency that will disappear once discovered — it has persisted for decades across markets worldwide. It endures because the behavioral and institutional forces that create it (underreaction, herding, trend-following) are structural features of financial markets, not anomalies that arbitrage away.

The discipline is in execution: measuring momentum correctly, managing rebalancing costs, surviving momentum crashes without abandoning the strategy at the worst moment, and combining momentum with complementary strategies like value and mean reversion.

Quantitative systems that systematically evaluate momentum signals across thousands of combinations per asset — adjusting for volatility, time horizon, and current market regime — offer a more rigorous approach than manual momentum screening.