Pairs trading is a strategy that involves taking long and short positions on two highly correlated securities to profit from temporary divergences between their price movements.
By neutralizing systematic market influences, pairs trading allows profiting from the natural tendency of correlated assets to revert to their historical relationship over time.
Identifying Suitable Pairs
Traders evaluate years of historical data to identify pairs with correlation coefficients above 0.8 that often deviate briefly before converging again.
Sectors like airlines, banks, oil producers provide candidates. Statistical mean-reversion tests establish average periods for convergence to occur.
Entry Signals
Positions are initiated when the spread between the two securities surpasses common historical limits, implying short-term relative over/undervaluation.
For example, entering long the lagging stock and short the leading when their 50-day moving average difference exceeds 2 standard deviations.
Managing Positions
Partial covers on a portion of the spread closing lock in profits. Stops are wide to tolerate usual weekly relative moves of 3-5%.
Leverage is low as volatility of the relative difference changes little versus the absolute price.
Statistical Performance Metrics
Backtesting quantifies optimal holding periods, average wins/losses and spread characteristics. Drawdown, Sharpe ratio and profit factor analyses establish robustness.
Studies show annual returns of 15-30% are achievable with minimal correlation to overall markets.
In summary, pairs trading exploits naturally resolving short-term divergences between inherently tied variables like market sector constituents.
Identification of historically predictive pairs through statistical testing combined with opportunistic entries based on robust patterns improves risk-adjusted returns.