Cross-Sectional Momentum: What Jegadeesh and Titman Found and What Single-Asset Backtests Show

What Jegadeesh and Titman Actually Proved

In 1993, Narasimhan Jegadeesh and Sheridan Titman published Returns to Buying Winners and Selling Losers in the Journal of Finance. Their core finding: U.S. stocks that outperformed their peers over the prior 3–12 months tended to keep outperforming over the next 3–12 months. Buy the top decile by past return, short the bottom decile, hold a few months, repeat.

The word cross-sectional is doing heavy lifting here. The signal is purely relative — how does this stock rank against everything else in the universe right now? A stock up 30% over twelve months is not a momentum bet if every other stock is up 40%. That relative-ranking structure is what separates the Jegadeesh-Titman strategy from simply buying an asset that has been rising.

The 52-Week High Connection

George and Hwang (2004) offered a behavioural explanation that sharpened the practical version of the trade: proximity to the 52-week high is itself a strong momentum signal. Traders anchor psychologically to prior highs and resist pushing price through them even when fundamentals justify it. When the breakout eventually happens, it tends to continue — the delayed reaction becomes the momentum.

This maps directly to breakout indicators. In our tested indicator set, Donchian channel breakout variants capture exactly this logic mechanically. The 52-week high is simply the upper Donchian band on a weekly chart. The academic result and the chart indicator are measuring the same thing from different angles.

What Our Backtests Found for Stocks

IndicatorEdge ran 660,005 out-of-sample tests across 903 assets and 382 indicators on 1-Hour, 4-Hour, Daily, and Weekly timeframes. For the Stock asset class, the five indicators that topped the most individual assets were: Fibonacci Pivots (22 assets), Projection Bands (16), Intraday Momentum Index (16), Camarilla Pivots (16), and Markov Regime (14).

The Intraday Momentum Index is explicitly a momentum measure — it evaluates open-to-close price action relative to prior range. Markov Regime detection identifies whether the current market state is trending or mean-reverting before any signal fires. Both embed the same intuition as Jegadeesh-Titman: momentum is real but regime-dependent. For the Index asset class, Rate of Change — ROC (30) — appeared among the top performers, a direct 12-month-lookback analogue.

Across all 903 assets, 63% had at least one indicator that beat buy-and-hold out-of-sample. The median best Sharpe ratio was 0.62. Only 26% of all indicator-asset-timeframe combinations beat buy-and-hold — most combinations fail. Finding the ones that work requires the empirical filtering, not assumption.

Where Cross-Sectional Momentum Has Limits

Momentum crashes. The short side of the JT strategy — selling prior losers — is the source of most of the risk. During sharp market recoveries, prior losers often rebound violently and simultaneously. The long-short portfolio can suffer severe drawdowns at precisely the moments when capital is most scarce.

International decay. Follow-up research tested the strategy outside the U.S. Results were mixed. Some developed markets showed similar patterns; results in several Asian markets were weaker or inconsistent. More capital explicitly targeting momentum over the decades since 1993 has also compressed the effect.

The single-ticker problem. You cannot implement Jegadeesh-Titman momentum on one chart. The strategy requires ranking a live universe of hundreds of stocks against each other every rebalance. What you can do is use momentum-aware indicators — breakout signals, ROC, regime detectors — that capture the underlying price-continuation logic on a single asset. Whether they outperform buy-and-hold depends on the specific asset, timeframe, and parameter set, which is exactly what the backtests measure.

These Are Hypothetical Backtests — Not Advice

All results on IndicatorEdge are hypothetical backtest results produced by applying rule-based indicator strategies to historical price data with realistic transaction cost assumptions. They do not represent actual trading, actual account performance, or a guarantee of future results. Momentum strategies, including those related to Jegadeesh-Titman or the 52-week high effect, can and do go through extended losing periods. Nothing on this site is investment advice. Use this research to inform your own due diligence, not as a direct trading signal.