Pairs trading profits from temporary dislocations in the price relationship between two correlated stocks. When the spread between them stretches beyond its historical norm, you trade the convergence — not the direction.
Two companies in the same industry often move together — they share the same macro drivers, the same interest rate sensitivity, the same sector sentiment. MSFT and GOOGL both benefit from enterprise spending. JPM and BAC both move with rate expectations. XOM and CVX both track oil prices.
Because they share these drivers, their prices tend to maintain a relatively stable spread (ratio or difference) over time. When one stock temporarily outperforms the other — perhaps due to a stock-specific event, a liquidity imbalance, or momentum — the spread stretches. Pairs trading is the bet that the spread will revert to its mean.
A pairs trade is constructed by going long the underperformer and short the outperformer simultaneously. If both stocks rise together (market goes up), you make money on your long and lose money on your short — they cancel out. If both fall together, same thing. You're only exposed to the relative performance of the two stocks, not the market's direction.
This makes pairs trading genuinely market-neutral — one of the few strategies that can generate returns in both bull and bear markets, because the bet is on convergence, not direction.
The spread is the price difference (or ratio) between the two stocks, adjusted for their relationship. The z-score normalizes this spread by expressing it in units of standard deviations from the historical mean:
Z = (Current Spread − Mean Spread) / Standard Deviation of Spread
A z-score of +2.0 means the spread is 2 standard deviations above its mean — the outperformer has moved unusually far ahead. A z-score of −2.0 means the underperformer has fallen unusually far behind. These extremes are the pairs trade entry signals.
Rule of thumb: |z| ≥ 2.0 is a potential entry signal. |z| ≥ 3.0 is a higher-conviction signal with less frequency. The trade closes when z returns to 0 (convergence) or exits at a stop loss of |z| ≥ 4.0 (spread has gone too far — re-evaluate the pair's relationship).
Pairs trading assumes the relationship between two stocks is stable. Sometimes it isn't. A regime break occurs when the fundamental relationship between two stocks changes — one gets acquired, one pivots its business model, one loses a key contract. When this happens, the spread doesn't converge — it keeps widening, and the pairs trade turns into a directional bet you didn't want.
This is why cointegration testing is critical before entering a pairs trade. Cointegration is the statistical test that verifies the relationship is stable enough to trade (see the Cointegration guide). A pair that isn't cointegrated shouldn't be traded as a pairs strategy.