A mean reversion strategy assumes price that has stretched far from its average or range will snap back - so you fade the move, selling overextended rallies and buying oversold drops, targeting the return to the mean. It is the mirror of trend following: high win rate, small winners, and a hidden tail risk when a range turns into a trend.
Why mean reversion works - in a range
In a ranging market, price oscillates between boundaries. Moves to the edge of the range are, on average, more likely to reverse than to continue, because there is no dominant force pushing one direction. Fading those extremes gives you a series of small, frequent winners. The catch is in that phrase "in a range" - the same behavior in a trend is how accounts die.
Step 1: define "overextended" with a rule
You need an objective trigger for "too far, too fast." Common ones:
- Range boundaries: price reaching the top or bottom of an established range.
- Distance from a moving average: price stretched a set number of pips or ATRs beyond a mean.
- Oscillator extremes: a momentum indicator reaching an overbought or oversold reading - used as a filter, not a lone signal.
Step 2: enter, stop, and target
Enter as price shows it is stalling at the extreme - a rejection candle, a failure to make a new high. Place the stop just beyond the extreme, so if the level breaks you are out for a small, defined loss. Target the mean - the middle of the range or the moving average. Because the target is near, your reward-to-risk is usually below 1, which is why the win rate has to be high for the math to work.
The danger is not the win rate - it is the occasional trade that never reverts. One trend can undo weeks of small wins without a hard stop.
The trend is the enemy
Every mean reversion blow-up has the same story: the trader kept fading a move that kept going, averaging in, sure it "had to" turn. It did not. A confirmed trend has to be a hard filter - if higher-timeframe structure is trending, you do not fade it, full stop. This is the single rule that separates a durable mean reversion strategy from a slow-motion disaster.
Important: a high win rate is the most seductive and most misleading number in trading. A mean reversion strategy can show 70% winners for months and still be net negative because of a few uncapped losses. Judge it on expectancy and worst drawdown, never win rate alone.
Backtest the losers, not the winners
When you backtest a fade, the winners are boring and reassuring - it is the losers that tell you whether the strategy survives. Force a hard stop into every test, and specifically sample strong-trend periods to see how badly the strategy does when it is out of its element. In a simulator you can replay a clean range and then a brutal trend back to back, and the equity curve will show you exactly where the risk lives.
Mean reversion FAQ
What is a mean reversion strategy?
It fades stretched price - selling overextended rallies, buying oversold drops - targeting a return to the average. High win rate, small winners, best in ranges.
Why is it dangerous in a trend?
A trend keeps making new extremes, so fading it stacks losses until one runaway move erases many small wins. A hard stop and a trend filter are essential.
How do I backtest it?
Define overextension, entry, a stop beyond the extreme, and a target at the mean, then study the losing trades and worst drawdown, not just the win rate.