Statistical Arbitrage and Pairs Trading with Spread Stationarity Tests
Summary
This overview explains statistical arbitrage as a family of strategies that trade relative mispricing across related instruments. It distinguishes cross-market, cross-asset, ETF, and market-neutral approaches, and gives pairs trading as a central example: when historically related stocks diverge, the trader buys the relative underperformer and shorts the outperformer, expecting the spread to narrow. The proposed workflow selects candidate pairs, examines prices and spread behavior, calculates a spread z-score, and tests stationarity with the Augmented Dickey-Fuller test before forming signals.
The article includes examples involving related stocks and reports a sample hedge ratio and stationarity-test output, but it does not provide a complete, independently assessable backtest. It notes that trading costs and slippage can erode results, and that external shocks or a breakdown in the relationship can invalidate mean-reversion assumptions. Statistical arbitrage therefore carries execution, liquidity, and model risk despite the apparent convergence logic.
Key ideas
- Statistical arbitrage seeks to trade relative price deviations among instruments with historically related behavior.
- Pairs trading typically buys the relative laggard and shorts the outperformer, assuming their spread will mean-revert.
- A candidate pair can be assessed using a spread, z-score, and stationarity test before signals are generated.
- Cross-market, cross-asset, ETF, and market-neutral strategies are presented as forms of statistical arbitrage.
- Transaction costs, slippage, liquidity, and changing relationships can undermine the strategy.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.