Institutional Trading

Two Stocks That Move Together Until They Do Not

Pairs trading bets that two historically linked securities that have diverged will converge again. It profits from the relationship between two prices rather than the direction of either.

↩ Looking BackPart of the 2020 to 2026 retrospective, written in July 2026. The date below marks the 2022 events this piece revisits, not when it was published, so it draws on everything known through mid 2026.
Nathan Xiang·May 18, 2022

Trading the Relationship, Not the Direction

Most trading bets on direction: a stock will rise or fall. Pairs trading bets on something narrower, the relationship between two securities that historically move together. It does not care whether the market goes up or down, only whether two prices that usually track each other converge after diverging.

The classic example is two companies in the same industry. They face similar forces and their stocks usually move together. When one rises sharply and the other lags, the pairs trader shorts the outperformer and buys the underperformer, betting the gap will close.

The pairs trader has no view on the market or even on the two companies. The only bet is that a relationship that has held will reassert itself after a temporary divergence.

Why It Is Market Neutral

Because the trade is long one stock and short another of similar size, it is roughly market neutral: if the whole market rises, the long gains and the short loses, offsetting. If the market falls, the reverse. The profit comes only from the two stocks moving relative to each other, not from the market direction.

ScenarioLong legShort legNet
Gap closes, laggard catches upGainsFlat or small lossProfit
Market rises, gap unchangedGainsLosesRoughly flat
Gap widens furtherLagsRisesLoss

This independence from market direction is the appeal. In a falling market, a pairs trade can profit if the relationship converges, which makes it attractive as a source of returns uncorrelated with everything else.

Measuring the Normal Gap

The discipline of pairs trading is in defining what counts as an unusual divergence. A trader measures the normal historical relationship between the two prices and flags when the gap has grown large relative to its typical range, which signals a trade that bets on the pair reverting to its usual relationship.

The hard part is choosing the right pair. The two securities must be genuinely linked by shared economics, not merely correlated by coincidence over the sample studied, because a relationship that was never causal has no reason to reassert itself when it breaks.

The Risk That Kills It

The fatal risk in pairs trading is that the relationship breaks permanently rather than temporarily. The trade assumes the two securities will converge because they have in the past. But sometimes they diverge because something fundamental has changed: one company won a contract the other lost, one faces a scandal, one is being acquired.

When the divergence reflects a real change rather than a temporary dislocation, the gap does not close, it widens, and the trade loses on both legs. The short keeps rising and the long keeps falling. Distinguishing a temporary divergence from a permanent break is the entire skill, and it is genuinely hard, because the two look identical at the moment the trade is put on.

The Crowding Problem

Pairs and statistical arbitrage strategies became widely used, and widespread use created a specific danger. When many funds hold similar statistical arbitrage positions and one is forced to sell, perhaps to meet redemptions or margin calls, the selling moves the prices against everyone holding the same positions.

This can trigger a cascade, as losses force more selling, which deepens the losses. A notable episode saw many quantitative funds suffer sudden sharp losses simultaneously, as crowded statistical arbitrage positions unwound together in a matter of days, even though the underlying relationships were sound. The strategy was correct and the crowding was fatal, because everyone was in the same trades and the exit was narrow.

The Bottom Line

Pairs trading bets that two historically linked securities that have diverged will converge, profiting from their relationship rather than the market direction, which makes it roughly market neutral. It rests on the statistical history of the pair and its fatal risk is that a divergence reflects a permanent change rather than a temporary dislocation, in which case the gap widens and both legs lose. Its other danger is crowding, since widely held statistical arbitrage positions can unwind together violently even when the underlying logic is sound.

Explore Teen Biz News →