Pairs trading is a market-neutral trading strategy that simultaneously takes a long position in one security and a short position in a correlated security. When the price relationship between the pair diverges from its historical norm, pairs traders bet on convergence—going long the underperformer and short the outperformer. Profits come from the spread normalizing, regardless of overall market direction.
- Long one stock, short another correlated stock
- Profit from relative movement, not market direction
- Works when correlated stocks temporarily diverge
How Pairs Trading Works
Pairs trading exploits temporary divergences in correlated securities:
Pairs Trading Logic:
1. Identify two highly correlated stocks (e.g., Visa & Mastercard)
2. Calculate their historical price ratio (spread)
3. When spread deviates 2+ standard deviations:
- Short the outperformer
- Long the underperformer
4. Exit when spread reverts to mean
5. Profit regardless of market direction
Quick Reference: Pairs Trading Setup
| Element | Description | Target |
|---|---|---|
| Correlation | How closely pairs move together | > 0.80 |
| Cointegration | Statistical relationship over time | Must pass test |
| Spread entry | Standard deviations from mean | 2+ SDs |
| Spread exit | When spread normalizes | 0-0.5 SDs |
| Holding period | Time for spread to converge | Days to weeks |
Example: HDFC Bank vs. ICICI Bank Trade
Setup: Both banks trade together historically (correlation 0.85)
Observation:
- HDFC Bank: Up 8% this month
- ICICI Bank: Up 2% this month
- Spread: 2.3 standard deviations from mean
Trade:
- Short ₹5 lakh of HDFC Bank
- Long ₹5 lakh of ICICI Bank
- Net market exposure: Zero
Two weeks later:
- HDFC Bank: Flat (from your short entry)
- ICICI Bank: Up 4% (from your long entry)
- Spread has normalized
Result:
- Short leg: Break-even
- Long leg: +4% = ₹20,000 profit
- Total: ₹20,000 on ₹10 lakh deployed
Pairs trading goes long on one stock and short on a correlated stock when their relationship diverges. It’s market-neutral because you profit from the spread converging, not from market direction. Find pairs with high correlation in the same sector.
Finding Tradable Pairs
Criteria for Good Pairs:
- Same industry – Affected by similar forces
- High correlation – 0.80+ over long periods
- Cointegrated – Statistical test for long-term relationship
- Sufficient liquidity – Can enter/exit both legs easily
- Borrowable – Shares available to short
Popular Pairs in Indian Markets:
- HDFC Bank / ICICI Bank
- Reliance / ONGC
- TCS / Infosys
- Maruti / Tata Motors
- SBI / Bank of Baroda
The Spread Calculation
Price Ratio Method:
Spread = Price of Stock A / Price of Stock B
Normal Ratio: 1.25 (historical mean)
Current Ratio: 1.40 (Stock A outperforming)
Deviation: 12% from mean (trade signal)
Z-Score Method:
Z-Score = (Current Spread - Mean) / Standard Deviation
Z > 2: Short A, Long B
Z < -2: Long A, Short B
Exit when Z returns to 0
Risk Management in Pairs Trading
1. Position Sizing
Equal dollar amounts on each leg. If one stock is more volatile, adjust for equal beta exposure.
2. Stop Loss
Set max spread deviation. If spread hits 3-4 standard deviations, the relationship may be broken.
3. Time Stop
If spread doesn’t converge within expected period (e.g., 30 days), exit and reassess.
4. Correlation Monitoring
Track correlation in real-time. If it’s breaking down, exit the position.
Why Pairs Trade?
-
Market-neutral – Don’t need to predict market direction
-
Reduced volatility – Hedged position has lower variance
-
Works in any market – Bull, bear, or sideways
-
Statistical edge – Mean reversion is a documented phenomenon
Common Mistakes
-
Ignoring cointegration – High correlation isn’t enough. Pairs must be statistically cointegrated to mean-revert.
-
Unequal position sizes – Both legs must be equal in value (or risk) for proper hedging.
-
Fighting permanent divergence – Sometimes the relationship breaks. Companies change. Cut losses when the thesis is broken.
-
Ignoring borrow costs – Short positions have costs. Factor them into profitability.
How JournalPlus Tracks Pairs Trades
JournalPlus tracks both legs of your pairs trades as a combined position, calculating net P&L, spread entry/exit levels, and whether your pairs are converging as expected. You can monitor correlation between your paired positions over time.