Mean Reversion Strategies¶

Strategies that profit from prices returning to equilibrium.

Overview¶

Mean reversion strategies bet that extreme price moves will reverse: - Oversold stocks will rise - Overbought stocks will fall - Prices gravitate toward "fair value"

Why Mean Reversion Works¶

Overreaction: Markets overreact to news
Liquidity: Forced sellers create opportunities
Behavioral: Fear and greed cause extremes
Fundamental anchor: Prices tied to value long-term

Strategies in This Section¶

Strategy	Description	Complexity
Statistical Reversion	Z-score based reversion	Basic
RSI Reversion	Relative Strength Index	Basic
Bollinger Bands	Band-based signals	Intermediate
Short-Term Reversal	1-week reversals	Basic

Quick Example¶

Text Only

data:
  source = "prices.parquet"
  format = parquet

signal mean_reversion:
  // Price vs 20-day moving average
  ma_20 = rolling_mean(prices, 20)
  deviation = (prices - ma_20) / rolling_std(prices, 20)

  // Buy oversold (negative z), short overbought (positive z)
  emit -zscore(deviation)

portfolio main:
  weights = rank(mean_reversion).long_short(top=0.2, bottom=0.2)
  backtest from 2015-01-01 to 2024-12-31

Key Considerations¶

Works Best In¶

Range-bound markets
High volatility (more mispricings)
Liquid markets

Fails In¶

Strong trends
Regime changes
Momentum markets

Risk Management¶

Text Only

// Add trend filter to avoid trending markets
signal safe_reversion:
  reversion = mean_reversion_signal
  trend_strength = abs(ma_50 - ma_200) / ma_200

  // Only revert in range-bound markets
  low_trend = trend_strength < 0.05

  emit where(low_trend, reversion, reversion * 0.3)

Expected Performance¶

Metric	Typical Range
Sharpe	0.3 - 0.7
Annual Return	3% - 8%
Max Drawdown	10% - 25%
Turnover	400% - 800%

Research References¶

Jegadeesh (1990): "Evidence of Predictable Behavior of Security Returns"
Lo & MacKinlay (1990): "When Are Contrarian Profits Due to Stock Market Overreaction?"
DeBondt & Thaler (1985): "Does the Stock Market Overreact?"