Manual traders chase direction — up or down, long or short, hope or despair. The problem is that markets spend enormous amounts of time going sideways, chopping through stops and rewarding nobody. Pairs trading sidesteps the directional question entirely by trading the relationship between two instruments rather than betting on either one moving a particular way.
The core mechanics are deceptively simple. Find two instruments whose prices move together over time — think two major bank stocks, or two commodity ETFs — then monitor the spread between them. When the spread widens beyond its historical norm, the strategy sells the outperformer and buys the underperformer, betting mean reversion closes the gap.
The word traders throw around here is cointegration. Correlation tells you two series move in the same direction. Cointegration tells you they share a long-run equilibrium they keep returning to — a much stronger claim. An algo screens for this statistically, using tests like the Augmented Dickey-Fuller to filter candidates before a single trade is placed.
Entry signals are generated when the spread crosses a threshold — typically one to two standard deviations from the rolling mean. The algo calculates a z-score on the spread continuously, and a signal fires when that z-score breaches the chosen level. Exit logic mirrors entry: the position closes when the spread reverts toward the mean, or a stop fires if it keeps moving the wrong way. The classic trap is overfitting the z-score threshold and lookback window to historical data until the equity curve looks magnificent — then watching live performance quietly destroy that illusion. Solid pairs strategies use walk-forward validation and out-of-sample testing to stress-test signal quality before any capital is committed. Resources worth studying include the Investopedia pairs trade overview, the Wikipedia article on pairs trading for the academic lineage, and cointegration on Wikipedia for the statistical underpinning that separates a real edge from a pretty chart.
The backtest always looks better than the live account — that's not pessimism, it's physics. Build the signal right, validate it ruthlessly, and the spread will do the rest.
This content is for educational purposes only and does not constitute financial product advice. Past performance is not indicative of future results. Profit Logic Ltd (ACN 688 669 936) accepts no responsibility for errors or omissions in this content or anywhere on this website. Always seek advice from a licensed financial adviser before making investment decisions.