Momentum Trading

What Is Momentum?

Momentum is the empirical observation that recent winners tend to keep winning and recent losers tend to keep losing — over intermediate horizons of roughly 3 to 12 months. This "momentum effect" was first formally documented by Jegadeesh and Titman (1993) and has since been confirmed across virtually every asset class, geography, and time period studied.

Unlike trend following (which looks at whether a single asset is going up or down), momentum is a relative strategy. It asks: compared to all the other stocks in the universe, which ones have the strongest recent performance? You buy the top decile and sell (or avoid) the bottom decile. The outperformance of winners over losers — the "momentum spread" — has averaged 8–12% annually in US equities over multi-decade periods.

How Momentum Trading Works

Step 1: Rank the Universe

Start with a defined universe — S&P 500, Russell 1000, all US-listed stocks above $5, or a crypto watchlist. Calculate the total return over a lookback period (typically 12 months, excluding the most recent month) for every asset. Rank them from strongest to weakest.

Step 2: Buy the Winners

Buy the top 10–20% of the ranked list. In a universe of 500 stocks, that's the top 50–100 names. Equal-weight or volatility-weight the portfolio. Hold for a rebalancing period — typically one month or one quarter.

Step 3: Rebalance Periodically

At each rebalancing date, re-rank the entire universe. Sell holdings that have fallen out of the top tier; buy new names that have risen into it. This rotation keeps the portfolio concentrated in the strongest performers.

Step 4: Risk Management

• Sector constraints: Cap any single sector at 25–30% of the portfolio to avoid concentration in one industry (e.g., all tech stocks during a tech rally).
• Stop losses: Some implementations add individual position stops at 10–20% below purchase price to limit damage from sudden reversals.
• Market filter: Some systems shut off momentum entirely when the broad market is in a downtrend (below its 200-day MA), moving to cash or bonds.

Why Momentum Works

The academic debate over why momentum exists centres on two explanations:

• Behavioural: Investors underreact to new information initially (anchoring to old views), then overreact as the trend becomes obvious (herding). This creates a slow-to-start, fast-to-finish price pattern that momentum captures.
• Risk-based: Momentum stocks may carry hidden risks (concentration in certain factors, vulnerability to sudden reversals) that justify a return premium — similar to how value stocks outperform partly because they are genuinely riskier.

In practice, it doesn't matter which explanation is correct — the pattern is persistent enough across enough markets and time periods to be tradeable.

Momentum Crashes

Momentum's Achilles heel is the crash: sudden, violent reversals where past losers snap back and past winners collapse. The most famous momentum crash was March 2009, when deeply beaten-down financial stocks rocketed as the market bottomed, while momentum portfolios that were short those stocks suffered devastating losses in days.

Defences against momentum crashes include:

• Adding a market-regime filter (move to cash when broad market is below its 200-day MA)
• Combining momentum with value (the "value-momentum" barbell has historically produced smoother returns)
• Using stop losses on individual positions
• Reducing position sizes during high-volatility regimes (VIX > 30)

How AI Tools Enhance Momentum Trading

• Danelfin — scores every US stock daily using AI across technical, fundamental, and sentiment factors. The technical score component is effectively a momentum ranking. Screening for stocks with Danelfin scores of 8+ pre-filters the universe to high-momentum names. Free tier available. AIClarity Score: 8.2.
• Trade Ideas — Holly AI identifies the strongest-performing stocks each session and ranks them by projected continuation probability. This is real-time momentum scanning automated. AIClarity Score: 8.5.
• TrendSpider — multi-timeframe analysis can confirm that a stock showing momentum on a weekly chart also has supporting structure on daily and intraday charts — reducing the risk of entering a momentum name at the point of exhaustion. Review →

Key Takeaways

• Momentum — buying recent winners and avoiding losers — is one of the most robust and well-researched anomalies in finance.
• The standard lookback is 12 months (minus the most recent month) with monthly or quarterly rebalancing.
• Win rate is moderate (~50–55%) with consistent positive expectancy over multi-year periods.
• Momentum crashes are the main risk — sudden reversals where losers outperform winners violently.
• AI tools automate the ranking, screening, and signal generation that momentum strategies require.