Mean reversion is the most psychologically comfortable strategy family and the most dangerous one for exactly that reason. Buying something that just dropped feels clever. It feels like a bargain. That feeling is doing none of the analytical work.
What is mean reversion trading?
Mean reversion trades the tendency of price to return toward an average after moving unusually far from it. You buy near the low end of an established range and sell near the high end, with tight invalidation beyond the extreme. It works in ranging conditions and fails badly in trends.
The inverted payoff profile
Mean reversion is the mirror image of breakout trading. Breakouts win rarely and pay large. Mean reversion wins often — sometimes 60–70% of the time — and pays modestly, because you're taking profit at the middle or opposite edge of a range rather than riding a trend.
That creates a specific vulnerability. When most trades win small, a single oversized loss erases a long stretch of good ones. Eight wins at 1 unit and one loss at 8 units nets you nothing. The entire strategy depends on the losses staying small, which means the stop is non-negotiable in a way it isn't elsewhere.
And the stop is precisely what people abandon here, because the whole premise trains you to think of falling prices as opportunities. "It's even cheaper now" is the thought that ends accounts.
Where it works and where it doesn't
| Works when | Fails when |
|---|---|
| Clear range with tested boundaries | Strong directional trend |
| Higher timeframe is flat or sideways | Higher timeframe just broke structure |
| Extremes are being rejected sharply | Price closing beyond the range repeatedly |
| No major catalyst pending | Earnings, policy decisions, fresh news |
That last row deserves emphasis. Mean reversion assumes price moved for no durable reason. When something fundamental has genuinely changed, there's no average to revert to — the average itself is relocating. Fading a move driven by real news is betting against information.
Building the trade
- Confirm the regime is ranging before anything else. Check the higher-timeframe structure; if it's trending, the setup is off, full stop.
- Define the range edges as zones, using tested levels rather than exact lines.
- Wait for rejection at the edge — evidence that buyers or sellers actually showed up — instead of catching the move mid-fall.
- Stop goes beyond the range boundary, where the range is objectively broken. Use volatility to set the distance sensibly; see ATR-based stops.
- Target the middle or opposite edge. Modest, defined, taken without negotiation.
- Size from the stop distance so a single break costs one unit of risk, no more.
The uncomfortable truth about high win rates
A strategy winning 70% of the time feels like mastery. It isn't — it's a payoff profile. High accuracy with small wins can be identical in expectancy to low accuracy with big wins, and it's far more seductive because it's more pleasant to experience day to day.
Judge it the same way you'd judge anything else: expectancy per trade after costs, and worst drawdown. A pleasant strategy that loses money slowly is still losing money. Work through expectancy before trusting a win rate.
Frequently asked questions
Is mean reversion safer than trend following? No. It has a higher win rate and a fatter tail risk. Trend following loses often and small; mean reversion wins often and can lose big if stops are ignored.
Which indicators help? Oscillators like RSI, or bands around a moving average, can flag stretched conditions. They flag context, not entries — an oversold reading in a strong downtrend just means it's falling efficiently.
Can I average down instead of using a stop? That converts a defined risk into an undefined one. Adding to losers is the most reliably documented way retail accounts blow up — see why most day traders lose money.
Every setup family carries its own failure mode, and knowing them in advance is most of the job. The Asymmetric Trader's Playbook covers regimes, setups, sizing, and a failure-mode audit as one connected framework, $29 (reg. $39). Educational material, not financial advice.
More in the Make Money guides.