Mean Reversion Trading Strategy: Setups, Rules and Risks

Mean reversion trading is built on one idea: when price stretches unusually far from its recent average, it often drifts back toward that average. The trader's job is to measure the stretch, act only when it is extreme, and get out quickly when the market is trending instead. This guide covers the setups, where the strategy breaks and how to size it in a prop evaluation.

Quick answer: A mean reversion trading strategy buys when price is unusually far below its recent average and sells when it is unusually far above, aiming to profit as price returns toward the mean. Traders measure the stretch with Bollinger Bands, RSI or a z-score. It works best in ranges and fails in strong trends.

---

Mean reversion trading: key facts

What is mean reversion in trading?

Mean reversion is the tendency of a value to move back toward its average after moving away from it. In trading, the "mean" is usually a moving average of price over a fixed window, such as 20 candles, but it can also be a fair value spread between two related assets.

The idea has academic roots. Studies published in 1990 found that stocks with the worst returns over the previous week or month tended to do better than recent winners in the following period, a pattern known as short-term reversal (Jegadeesh, 1990; Lehmann, 1990). Over much longer horizons, De Bondt and Thaler (1985) found that portfolios of past three to five year losers went on to outperform past winners. These are findings across large groups of stocks, not signals for one chart.

Mean reversion sits on the opposite side of the trading strategies map from momentum. A momentum trader buys strength and expects more strength. A mean reversion trader sells strength and buys weakness, expecting the move to fade.

How do you measure how far price has stretched?

You cannot trade "too far from the mean" without defining both the mean and the distance. Three methods cover most setups.

Bollinger Bands

Bollinger Bands, created by John Bollinger, place a band two standard deviations above and below a 20-period simple moving average. Because the bands widen and narrow with volatility, a close outside a band means price has moved further than usual for current conditions. A basic mean reversion rule buys a close below the lower band and targets the middle band.

The chart below shows that pattern: price swings around its 20-period average, a sharp sell-off closes below the lower band with RSI in oversold territory, and price then returns to the middle band.

RSI

The Relative Strength Index measures the speed of recent gains against recent losses on a 0 to 100 scale. Readings below 30 are traditionally called oversold and above 70 overbought. Mean reversion traders often shorten the period, for example to 2 to 5 candles, so the indicator reaches extremes more often, and then require a second condition such as a close outside a Bollinger Band.

A low RSI alone is not a buy signal: in a strong downtrend, RSI can stay below 30 while price keeps falling.

Z-score

The z-score expresses the stretch in standard deviations:

z = (price minus moving average) / standard deviation of price over the same window

A z-score of minus 2 means price is two standard deviations below its average, which is the same line as the lower Bollinger Band. The z-score is one number you can compare across instruments. A rule might be: enter when z falls below minus 2, exit when z returns to 0, and stop out if z reaches minus 3.5.

What does a mean reversion trading strategy look like?

Here is a simple, rules-based version you can backtest before trading it. The numbers are a starting point, not a recommendation.

1. Filter the regime. Only trade when the 50-period moving average is roughly flat. If it slopes steeply, the market is trending and the setup is off.
2. Wait for the stretch. A candle closes below the lower Bollinger Band (20, 2) and RSI (14) is below 30. For shorts, reverse every condition.
3. Enter on confirmation. Enter when the next candle closes back inside the band, not while price is still falling.
4. Place the stop beyond the extreme. Put the stop-loss below the lowest low of the stretch, or a fixed multiple of the ATR below entry, such as 1 ATR.
5. Target the mean. Take profit at the 20-period moving average. Some traders take half at the mean and leave the rest for the opposite band.
6. Add a time stop. If price has not reached the mean within a set number of candles, such as 10, exit. A stretch that does not revert is information.

The time stop and the regime filter protect the edge; the entry is the easy part.

When does mean reversion fail?

Mean reversion fails when the average itself starts moving. In a trending market, a close below the lower band is often a sign of strength in the downtrend rather than an oversold extreme, and price can "walk the band" for many candles.

Common failure modes:

• Trends and breakouts. A range that turns into a trend leaves the mean reversion trader on the wrong side of a move that keeps going. This is the territory of breakout trading and trend following.
• News and regime shifts. Earnings, central bank decisions and macro data can reprice an asset permanently. The old mean stops being relevant.
• Averaging down. Adding to a losing position because "it has to come back" turns a small, planned loss into an account-level problem.
• Ignoring costs. Small targets mean commission and spread take a larger share of each win.

Because the strategy tends to win often and lose rarely, the losses that do arrive can be large if stops are moved or skipped. Look at the risk-reward ratio and the win rate together, not separately: a 70% win rate with an average win of 0.5R and an average loss of 1.5R loses money.

What is pairs trading and statistical mean reversion?

Pairs trading applies mean reversion to the relationship between two related assets rather than to one price. The trader finds two instruments that have historically moved together, watches the spread between them, and when the spread stretches unusually wide, sells the stronger one and buys the weaker one, expecting the gap to close.

Gatev, Goetzmann and Rouwenhorst tested a version of this on US stocks from 1962 to 1997, matching pairs by how closely their normalised prices tracked each other. They reported average annualised excess returns of up to 12% for portfolios of top pairs, while noting that part of the profit may come from market microstructure effects. Historical results do not promise future returns.

The statistical intuition: correlation is not enough, because two correlated assets can still drift apart for good. Professionals look for a spread that stays stable over time (cointegration), trade the spread's z-score rather than either price, and accept that relationships can break after a merger or regulatory change.

On a multi-asset account, pairs such as gold versus silver or SPY versus QQQ need testing, and both legs count toward your loss limits.

How do you size a mean reversion trade on a prop account?

Here is a worked example on a 50K CLASSIC 2-Step account. Product rules come from the Velotrade challenge page and FAQ; the trade prices are illustrative.

Account rules

• Account size: 50,000 USD
• Daily loss limit: 5%, so 2,500 USD, reset at 00:30 UTC from the higher of balance or equity
• Maximum drawdown: 10% static, so a fixed floor at 45,000 USD for the life of the account

Trade plan

• Instrument: EUR/USD on a 1-hour chart in a flat range
• Setup: close below the lower Bollinger Band at 1.0800, RSI (14) at 26, next candle closes back inside the band
• Entry: 1.0805
• Stop: 1.0780, below the low of the stretch, so 25 pips of risk
• Target: the 20-period average at 1.0845, so 40 pips
• Risk per trade: 0.5% of the account, so 250 USD

Position size

Position size = risk in dollars / (stop distance in pips x pip value per unit)

On EUR/USD, one pip on 100,000 units is worth 10 USD. A 25-pip stop on 100,000 units therefore risks 250 USD, which matches the budget. The position is 100,000 EUR, about 108,050 USD of notional. Velotrade lists forex majors at up to 50x leverage on challenges, so this uses only a small part of the available margin.

What the numbers mean

A sensible personal rule sits well inside the firm's limits: stop for the day after three losses, or 750 USD. That keeps the 2,500 USD daily limit as a backstop and blocks the most dangerous mean reversion habit, doubling down after each loss.

Use the position size calculator to repeat this for other instruments and stop distances.

How does mean reversion fit prop firm rules?

Mean reversion can suit an evaluation because trades are short and targets are defined. A few rules deserve attention.

• The static drawdown floor does not move. On Velotrade, the maximum drawdown is a fixed dollar floor set at activation. Gains do not drag it up behind you, so a winning streak builds a real buffer. Losses still count in full, so an unprotected trade that runs against you can use up that buffer quickly.
• The daily loss limit is checked against live equity. A position that is down on paper counts. Averaging into a falling market can breach the limit before any trade is closed.
• Floating profit at the reset raises the floor. The daily limit is set at 00:30 UTC from the higher of balance or equity. If a reversion trade is in profit at that moment, the day's breach floor is set from the higher equity figure.
• There is no maximum risk per trade. Velotrade does not cap position size per trade, so the discipline has to come from your trading plan, not from the platform.
• Choose the plan that fits the profile. CLASSIC 2-Step (5% daily, 10% static) gives the most room for a strategy with occasional large losses. PRO 1-Step (3% daily, 3% static) is the tightest and suits only very small, strictly stopped risk.

Funded accounts start at an 80% profit split, with a paid add-on for 90%. Compare plans and sizes on the challenges page, and see how different styles fit evaluation rules in best prop firm for swing traders.

Is mean reversion better than trend following?

Neither is better in general. Mean reversion tends to win in ranges and suffer in trends; trend following and momentum trading tend to do the opposite. Many traders learn both and use a regime filter, such as the slope of a long moving average, to decide which playbook applies.

Trading involves risk, and most retail traders lose money. Treat any strategy here as educational, test it on historical data, and size it so a run of losses cannot end your account.

Sources

• Jegadeesh, N. (1990), "Evidence of Predictable Behavior of Security Returns", The Journal of Finance: short-term (monthly) return reversal in US stocks.
• Lehmann, B. (1990), "Fads, Martingales, and Market Efficiency", The Quarterly Journal of Economics: weekly winners and losers tend to reverse.
• De Bondt, W. and Thaler, R. (1985), "Does the Stock Market Overreact?", The Journal of Finance: long-horizon losers outperform past winners.
• Gatev, E., Goetzmann, W. and Rouwenhorst, K. G., "Pairs Trading: Performance of a Relative Value Arbitrage Rule", NBER Working Paper 7032: pairs trading results on US stocks, 1962 to 1997.
• Bollinger Bands, "A complete explanation of Bollinger Bands": construction and default settings from the creator's site.
• Investopedia, "Mean Reversion": definition and common uses of mean reversion.
• Investopedia, "Z-Score: Meaning and Formula": how a z-score measures distance from the mean in standard deviations.