Technique | Algo
Mean Reversion
"A strategy based on the hypothesis that prices tend to return to their historical average after significant deviations."
In-Depth Definition
Mean Reversion is one of the oldest and most studied strategies in quantitative finance. It rests on the statistical principle that prices, returns, or spreads that deviate significantly from their mean tend to revert to it. In practice, this means buying assets that have fallen sharply (considered oversold) and selling those that have risen sharply (overbought). Mean reversion strategies include pairs trading (spread arbitrage between two correlated assets), Bollinger Band trading, and RSI-based strategies. The key is distinguishing a temporary deviation (reversion opportunity) from a permanent regime change (trend). The main risk is that the mean itself evolves over time.
StarQuant Insight
StarQuant uses cointegration models and stationarity tests to identify asset pairs most suitable for mean reversion strategies, and dynamically calibrates entry/exit thresholds based on current spread volatility.
Pro Tip
Always test the stationarity of your time series before applying a mean reversion strategy. A strongly trending asset violates the mean reversion hypothesis and will make your strategy catastrophically losing.