Mean Reversion vs Trend Following

Mean reversion assumes price returns to an average. Trend following assumes price continues in its current direction. These cannot both be true at once, and yet both have long records of working.

The two return profiles

The difference is not just direction but the shape of the results.

• Mean reversion produces many small wins and occasional large losses. It feels good almost all the time. The rare loss arrives when price does not revert, and it can erase a long run of gains.
• Trend following produces many small losses and occasional large wins. Win rates around 35–40% are normal. It feels bad most of the time, and the discipline problem is psychological rather than analytical.

Mean reversion is comfortable and occasionally catastrophic. Trend following is uncomfortable and occasionally spectacular.

Why each works

Mean reversion exploits over-reaction. Markets overshoot on emotion and liquidity effects, and prices often retrace part of that overshoot. The edge is real but modest and short-lived.

Trend following exploits under-reaction. Information diffuses gradually, institutions accumulate positions over days or weeks rather than instantly, and behavioural anchoring slows the adjustment. Moves therefore persist longer than efficiency would imply.

The conditions that suit each

Mean reversion needs range-bound, mature, liquid markets where no strong new information is arriving. Trend following needs directional conviction, which usually accompanies a genuine change in fundamentals or policy.

The critical implication: the environment that maximises one is the environment that punishes the other. A trend follower's worst year is often a mean reverter's best.

Risk management differs fundamentally

For mean reversion, the stop loss is existential. The strategy's premise is that price returns; if it does not, losses grow without limit. A mean-reversion system without hard stops is not a strategy, it is a hope.

For trend following, the exit rule is what matters. The strategy depends on a small number of large winners, so cutting them early removes the entire edge. Letting losers run, by contrast, is survivable because they are individually small.

Choosing

The honest criterion is temperament, not backtest. A trader who cannot tolerate a 60% loss rate will abandon a trend system during a normal losing streak, which converts a viable approach into a realised loss. A trader who cannot enforce stops will eventually be destroyed by mean reversion.

Running both is coherent if you accept that one will usually be underperforming, and that this is the design working rather than failing.

Educational content only. Not financial advice. Trading involves substantial risk of loss.