Definition
A pairs trade is a relative-value structure with a long leg and a short leg. The thesis is the spread between two exposures, not necessarily the absolute direction of the market.
Mechanics
- Identify two related exposures.
- Define why one should outperform the other.
- Size the legs to reflect volatility, beta, and risk budget.
- Monitor the spread and the evidence that should drive convergence or divergence.
- Exit when the spread thesis plays out or invalidates.
Worked example
An AI-disruption pairs trade could short a basket of seat-priced workflow software and go long a basket of AI infrastructure beneficiaries. If the whole market rises, the long side may offset some short-side losses. If the disruption thesis is right, the exposed software basket should underperform the beneficiary basket over time.
Pair design
| Design choice | Question |
|---|---|
| Theme | What economic shift connects the long and short legs? |
| Beta | How much broad market exposure remains? |
| Liquidity | Can both legs be adjusted without excessive cost? |
| Invalidation | What evidence means the spread thesis is wrong? |
Common mistakes
- Assuming two stocks are a pair just because they share a sector.
- Ignoring beta, volatility, and position sizing.
- Letting the long leg become a separate unrelated thesis.
- Ignoring borrow and dividend costs on the short leg.
Stock Shorter framing
Pairs trades fit the Stock Shorter thesis because AI disruption is a transfer of value, not just a destruction of value. The thesis argues that AI compresses certain SaaS economics while redirecting value toward infrastructure, platforms, and AI-native workflows.
Research only. Pairs trades can lose money on one or both legs.