How Trend and Mean Reversion Vary Across Trading Horizons
Summary
The article compares trend following and mean reversion across holding periods, drawing on the author's earlier tests and a cited study spanning minutes to decades. Its broad synthesis is that mean reversion appears at horizons beyond roughly two years and from two to 30 minutes, while trend following dominates between about an hour and two years. Trend strength is described as strongest over weeks to months; the cited study finds short intraday reversals, especially over several minutes. The author distinguishes predictive fit from Sharpe ratio, since faster turnover can raise risk-adjusted returns even when predictability weakens.
The discussion translates those patterns into practical choices by horizon, factoring in transaction costs, execution, and strategy risk. Medium speed trend following is presented as accessible to many futures traders, while very fast mean reversion would require passive fills and careful risk control. The author also retracts an earlier absolute mean reversion backtest after finding an error, and says the replacement strategy performs poorly relative to tested relative mean reversion. The horizon findings are not universal prescriptions: costs, data frequency, market, and the cited study's methodology limit what can be inferred for any trader.
Key ideas
- The article presents trend following as most useful from roughly an hour to two years, with strength varying by horizon.
- It describes mean reversion at very long horizons and at intraday horizons of two to 30 minutes.
- Predictive strength and Sharpe ratio can favor different holding periods because Sharpe ratio also depends on turnover.
- Transaction costs and fill quality constrain strategies as trading speed increases.
- The author retracts an earlier absolute mean reversion result after finding a backtest error.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.