How Does Pairs Trading Work When Both Stocks Fall?

How Does Pairs Trading Work When Both Stocks Fall?

By Rami Alame (Akylles) | Trade Feeld | Level: Intermediate | Instruments: Stocks

Pairs trading can produce a gross profit when both stocks fall if the gain on the short position exceeds the loss on the long position. With equal dollar positions, that means the shorted stock falls more, in percentage terms, than the stock you own. Different position sizes change that calculation. Borrowing costs, dividends, execution costs, and financing can turn a small gross gain into a net loss. The trade is about relative performance, not simply whether the market rises or falls. This article is educational only, not financial advice.

1. What a pairs trade actually trades

A pairs trading strategy combines a long position in one stock with a short position in another. Usually, the trader expects a historical or economically justified relationship between the stocks to move back toward an estimated equilibrium.

The long leg is the stock considered relatively cheap under that framework. The short leg is the stock considered relatively expensive. Neither description means the stock is cheap or expensive in absolute terms.

Suppose two businesses face similar industry conditions. A trader might expect one to hold up better during a slowdown. If both stocks decline, the long can lose money while the short gains more.

But relative outperformance alone is not always enough. Position sizing determines how much each percentage move contributes to the result. The relevant calculation is:

Net trading result = long-leg profit or loss + short-leg profit or loss − trading and holding costs.

A short position gains when shares can be repurchased for less than their sale price. It loses when repurchasing them costs more. Unlike a long stock position, a short position has theoretically unlimited loss potential.

2. Worked example: both stocks fall

All figures in this example are hypothetical round numbers, not live prices, forecasts, or suggested trades.

Assume a trader establishes an equal-dollar pair:

• Buy 100 shares of Stock A at $100: $10,000 long exposure.
• Short 200 shares of Stock B at $50: $10,000 short exposure.
• Gross exposure is $20,000; initial net dollar exposure is zero.

Later, both stocks are lower:

• Stock A falls to $90, a 10% decline.
• Stock B falls to $40, a 20% decline.

The long position loses $1,000: 100 shares multiplied by a $10 decline. The short position gains $2,000: 200 shares multiplied by a $10 decline.

Combined gross profit: $1,000. If total trading and holding costs are hypothetically $200, the net trading profit is $800.

Now reverse the relative moves. If A falls 20% and B falls 10%, the long loses $2,000 while the short gains $1,000. The pair loses $1,000 before costs, even though both stocks fell.

If both decline 10%, equal initial dollar positions offset before costs. Costs then make the net result negative.

The $20,000 gross exposure is not necessarily the cash or margin required. Short-sale proceeds are subject to broker restrictions, and margin requirements can change. A return-on-capital figure needs a clearly defined capital base; dividing profit by net exposure, which starts at zero here, is meaningless.

3. Equal dollars versus a beta neutral stock spread

Equal-dollar sizing removes initial net dollar exposure. It does not necessarily remove broad-market sensitivity.

Beta estimates how strongly a stock has historically moved with a chosen market benchmark. If the long stock has a higher beta than the short stock, an equal-dollar pair can retain meaningful market exposure.

For a simplified beta neutral stock spread, the sizing relationship is:

Long dollar exposure × long beta = short dollar exposure × short beta.

For example, using hypothetical estimates, a $10,000 long with beta 1.2 would require a $15,000 short with beta 0.8 to balance estimated market sensitivity. Both sides then contribute $12,000 of beta-weighted exposure in opposite directions.

That pair is not dollar neutral. Nor is it guaranteed to remain beta neutral: estimates depend on the benchmark, observation window, and market conditions.

Dollar neutrality, beta neutrality, and a statistically estimated hedge ratio answer different questions. A hedge ratio used to model a stable price relationship is not automatically a market-beta hedge.

Whatever the method, calculate each leg's actual dollar contribution. “The short fell more” is insufficient when exposures differ.

4. Cointegration versus correlation

Understanding cointegration versus correlation helps separate a plausible relative-value setup from two stocks that merely move together.

Correlation measures how closely changes, commonly returns, move together over a selected period. Two stocks can have highly correlated returns while their relative price relationship drifts persistently. Correlation alone does not establish mean reversion.

Cointegration concerns whether a particular combination of otherwise nonstationary price series is stationary under the model and sample used. In plain English, the modeled spread may fluctuate around a stable relationship rather than wander indefinitely.

That distinction matters because a convergence trade needs a reason to expect the spread to normalize. Shared industry membership is a starting point, not proof.

A practical research process checks:

• Whether the relationship makes economic sense.
• Whether results survive different estimation periods and out-of-sample testing.
• Whether adjusted price data properly handle splits and distributions.
• Whether the spread has changed after shifts in leverage, products, or business strategy.

A standardized spread measure, such as a z-score, describes distance from a historical estimate. It is not a probability of profit or a countdown to convergence. Statistical evidence can weaken precisely when market stress increases.

5. Costs, borrow, and event risk

Pairs trading short borrow risk deserves its own review. A short requires available shares to borrow, and availability is not permanent. Fees can rise, shares can be recalled, and a broker may require a position to be closed under its policies.

Before treating a backtest as executable, check the broker's current stock-loan availability, annualized borrow fee, charging convention, margin requirements, and recall procedures. Those live terms belong in the broker's locate or stock-loan system—not an assumed universal cost.

Other practical frictions include:

• Dividend-related payments: A short seller generally owes a payment corresponding to a dividend on borrowed shares.
• Financing: Debit interest and any interest treatment of short proceeds depend on account terms.
• Execution: Bid-ask spreads, slippage, and partial fills can change the intended hedge.
• Corporate events: Earnings, mergers, spin-offs, and restructurings can invalidate the relationship.

Use SEC EDGAR to review company filings and business changes. Check the Nasdaq earnings calendar for scheduled reports, then confirm timing with each company's investor-relations announcements. Calendar estimates can change.

For general investor education on short selling and margin, consult FINRA. Your broker's account agreement governs its specific requirements.

6. Common mistakes when both stocks decline

Mistaking a hedge for protection against every loss. A pair may reduce one exposure while retaining sector, liquidity, earnings, and company-specific risks. Both legs can also lose if the long falls and the short rises.

Comparing dollar price drops instead of percentage moves and position sizes. A $5 decline means different things for stocks at different starting prices. Share counts then determine actual profit and loss.

Treating the historical average as fair value. A spread may widen because one company's economics have permanently deteriorated. A past relationship is evidence to investigate, not an obligation for prices to return.

Adding exposure simply because the spread is wider. A larger deviation can reflect a broken thesis rather than a better opportunity. Increasing size also increases funding and liquidation pressure.

Ignoring the path to the final result. An eventual convergence does not prevent an earlier margin call, borrow recall, or forced exit. Mark-to-market losses matter throughout the holding period.

7. A step-by-step checklist

1. Define the relationship. State why the businesses should have linked valuations and what would invalidate that explanation.
2. Choose the spread model. Specify price or log-price inputs, hedge ratio, estimation window, and whether the objective is convergence or another relative-performance thesis.
3. Validate the evidence. Test stability outside the fitting sample and account for repeated strategy searches. Use consistent, adjusted data.
4. Set the exposure framework. Distinguish dollar neutrality from beta neutrality. Estimate what happens if both stocks fall, both rise, or move against their respective legs.
5. Verify live constraints. Check borrow availability, fees, margin, liquidity, and upcoming company announcements before execution.
6. Define operating rules. Document entry, exit, maximum holding period, loss limits, and responses to broken relationships or lost borrow. Plan for partial fills.
7. Monitor attribution. Separate spread performance from market exposure, execution, financing, and borrow costs. Reassess the thesis rather than relying only on total profit and loss.

For macro context, FRED provides economic and financial series. Check each series' observation date, frequency, and release notes; historical data can be revised. Macro context does not validate an individual pair.

The bottom line

When both stocks fall, a pairs trade works only if the short's dollar gain exceeds the long's dollar loss and all applicable costs. Equal-dollar positions make the comparison simple; other hedge methods require explicit weighting.

The harder task is establishing a defensible relationship and keeping the trade executable through changing prices, borrow terms, and margin demands. Neutrality is a modeling objective, not a promise of safety.

Keep learning free on Trade Feeld, and follow @tradefeeld on X for more trading education. Use these concepts to understand mechanics and risk, not as a recommendation to enter a trade.