Mean Reversion

Mean reversion is based on the statistical observation that prices tend to oscillate around a central value. When a stock deviates significantly from its average — whether due to news overreaction, algorithmic momentum, or emotional selling — it creates a snapback opportunity.

The "mean" can be defined multiple ways:

• Moving average (20-day SMA or EMA) — the most common reference point for swing traders
• VWAP — the volume-weighted mean price for the day, used by intraday traders
• Bollinger Band midline — the 20-period SMA with standard deviation bands showing how far price has deviated
• Historical price range — the stock's typical trading range over weeks or months

The statistical foundation is rooted in the concept of stationarity. A stationary time series fluctuates around a constant mean and tends to return to it. Individual stock prices are generally non-stationary (they trend), but certain transformations — like RSI, distance from moving average, or the spread between two cointegrated stocks — are stationary and therefore mean-reverting.

Mean reversion works best in range-bound or choppy market regimes. During strong trends, mean reversion signals become "catching a falling knife" — the stock is not reverting, it is repricing to a new level.

RSI Reversal: Buy when the 2-period RSI drops below 10 (extreme oversold) and the stock is above its 200-day SMA (confirming the uptrend is intact). Exit when RSI crosses above 70. Larry Connors popularized this approach, and backtests show 70-80% win rates on large-cap stocks.

Bollinger Band Bounce: Buy when price touches or closes below the lower Bollinger Band (2 standard deviations below the 20-day SMA) and a reversal candle forms. Target the midline (20-day SMA). This works because moves beyond 2 standard deviations are statistically uncommon and tend to revert.

VWAP Reversion (intraday): For day traders, buy when price drops 1-2 ATR below VWAP and shows a reversal candle. VWAP acts as an intraday mean, and institutional algorithms anchor to it. Price tends to oscillate around VWAP throughout the session.

Pairs Trading: The purest form of mean reversion. Find two cointegrated stocks (their spread is stationary). When the spread widens beyond 2 standard deviations, buy the underperformer and short the outperformer. Exit when the spread reverts to the mean. This is market-neutral — you profit regardless of market direction.

Moving Average Envelope: Buy when price drops to the lower envelope (e.g., 5% below the 20-day EMA) and sell when it returns to the moving average. Simple but effective in range-bound stocks.

Mean reversion and momentum are opposing strategies — one profits from price extremes reverting, the other from price extremes continuing. Both work, but in different market environments:

Mean reversion works best when: the market is range-bound (no clear trend), VIX is elevated but not in crisis mode, breadth is mixed (some sectors up, some down), and the stock has a history of range-bound behavior. Mean reversion has higher win rates (60-80%) but smaller average wins.

Momentum works best when: the market is trending clearly (above 20-day MA, moving in one direction), VIX is low or declining, breadth is strong (most sectors moving together), and the stock is breaking out to new highs/lows. Momentum has lower win rates (40-55%) but larger average wins.

The regime detection question: The key challenge is knowing which regime you are in. A stock dropping 5% could be a mean reversion opportunity (buy the dip) or the start of a momentum breakdown (stay away). Tools like the HMM regime detector, ADX (trending vs ranging), and market breadth indicators help classify the current environment.

Practical approach: Many professional traders run both strategies simultaneously, allocating more capital to whichever strategy's conditions are currently present. Mean reversion during consolidation, momentum during breakouts. The total portfolio benefits from this diversification across strategy types.