Pair Trading: The Market-Neutral Strategy That Works in Any Market

In August 2015, when the S&P 500 crashed 11% in four days on China fears, most traders were wiped out. But a certain class of hedge fund barely noticed. Their portfolios — built on pairs of long and short positions in related stocks — moved almost not at all. While the market collapsed, the relationship between their positions stayed stable.

That's the power of pair trading: a strategy that profits from the relative performance of two securities, not from whether the overall market goes up or down.

Pair trading is one of the oldest market-neutral strategies on Wall Street, pioneered in the 1980s by a quantitative team at Morgan Stanley. Today it's a cornerstone of hedge fund operations worldwide — and with modern tools and fractional shares, it's accessible to individual investors too.

This guide covers:

• The mechanics of how pair trades work
• How to identify and evaluate good pairs
• Classic real-world examples
• How professional quant funds run statistical arbitrage at scale
• A practical step-by-step approach for individual investors
• The risks that can break a pair — and how to manage them

The Core Concept: Betting on a Relationship, Not a Direction

Traditional trading is simple: You think a stock goes up, you buy it. You think it goes down, you short it. Your profit depends on being right about direction.

Pair trading is different. You don't need to predict direction at all. Instead, you exploit the relationship between two securities.

Here's the fundamental insight: Many stocks in the same industry move together over time, because they're subject to the same economic forces. When Coke and Pepsi both sell soda, their stocks tend to rise and fall together — same commodity input costs, same consumer spending trends, same regulatory environment.

But occasionally, their prices diverge temporarily. Maybe one company beats earnings, causing a short-term premium. Maybe investors rotate into one over the other. Maybe a short-term news event pumps one but not the other.

The pair trader's bet: This divergence is temporary. The relationship will mean-revert — prices will converge back to their historical ratio.

The trade:

• Long the underperforming stock (it should catch up)
• Short the outperforming stock (it should give back its premium)

Your profit comes from the spread between the two positions converging — regardless of whether the market goes up or down, because you're simultaneously long and short roughly equal amounts.

How Pair Trading Works: The Mechanics

Step 1: Find a Valid Pair

Not any two stocks make a pair. You need cointegrated securities — stocks with a stable, long-term price relationship.

Cointegration vs Correlation:

This distinction matters enormously:

Correlation measures whether two stocks move in the same direction at the same time. Many stocks are highly correlated during market trends, but that correlation breaks down quickly.

Cointegration measures whether two stocks share a long-term equilibrium — whether their spread reverts to a mean over time. This is the property that makes pair trading work.

Example:

• Coke and Pepsi are correlated — they both rise in risk-on markets
• Coke and Pepsi are cointegrated — their price ratio stays within a stable range over years

Two stocks can be correlated but not cointegrated (like any two random stocks in a bull market) and cointegrated but sometimes not correlated (they can diverge temporarily but always come back).

Pair trading requires cointegration, not just correlation.

Step 2: Calculate the Spread

The spread is the quantified price difference between the two securities.

There are two common ways to measure it:

Ratio method:

code-highlight
Spread = Price of Stock A / Price of Stock B

Example (Coke vs Pepsi):

• Coke (KO): $65
• Pepsi (PEP): $160
• Ratio: 65 / 160 = 0.406
• Historical average ratio: 0.410

Residual method (more sophisticated): Run a regression to find the hedge ratio (β) that minimizes the spread variance:

code-highlight
Spread = Price A - (β × Price B)

The β tells you exactly how many shares of B to short against each share of A to be dollar-neutral.

Step 3: Calculate the Z-Score

The z-score tells you how far the spread is from its historical average in units of standard deviations.

code-highlight
Z-Score = (Current Spread - Mean Spread) / Standard Deviation of Spread

Interpretation:

• Z-score near 0 → Spread is at historical average → no trade
• Z-score > +2 → Spread is 2 standard deviations above average → Stock A is overpriced relative to Stock B → Short A, Long B
• Z-score < -2 → Spread is 2 standard deviations below average → Stock A is underpriced → Long A, Short B
• Z-score returns to 0 → Spread mean-reverted → Close the trade, take profit

Why ±2 standard deviations? Statistically, a spread that's 2 standard deviations from its mean should revert about 95% of the time (assuming normal distribution). Traders enter at ±2 to ensure meaningful deviation from normal.

Step 4: Execute the Trade

When Z-score is +2.0 (Stock A is expensive relative to B):

• Sell short Stock A (the expensive one)
• Buy Stock B (the cheap one)
• Position size: Adjusted to be dollar-neutral (equal dollar amounts long and short)

When Z-score is -2.0 (Stock A is cheap relative to B):

• Buy Stock A (the cheap one)
• Sell short Stock B (the expensive one)

Step 5: Exit the Trade

Profit exit: When Z-score returns to 0 (spread mean-reverts), close both positions. Profit = the spread that closed.

Stop-loss exit: If Z-score reaches ±3 or beyond (spread keeps widening), cut the loss. The pair may be breaking down permanently.

Classic Pair Trading Examples

Example 1: Coca-Cola (KO) vs PepsiCo (PEP)

Why they're a pair:

• Both sell beverages and snacks globally
• Subject to same input costs (corn syrup, aluminum, plastic)
• Same consumer spending cycles
• Nearly identical distribution networks
• Compete in the same grocery store shelf space

The setup:

• Historical KO/PEP price ratio: 0.40 (average over 5 years)
• Normal range: 0.36 to 0.44 (one standard deviation = 0.04)

A pair trade:

• PepsiCo reports better-than-expected earnings, stock jumps 8%
• Coke's stock moves up only 2% (sympathetic, not company-specific)
• New ratio: 0.33 (Coke now cheap relative to Pepsi)
• Z-score: -1.75 (approaching the -2.0 entry signal)

Trade:

• Long 300 shares KO at $65 = $19,500
• Short 122 shares PEP at $160 = -$19,520 (dollar neutral)

Outcome (4 weeks later):

• Pepsi-specific earnings euphoria fades, both stocks normalize
• KO: $67 (+$2/share × 300 = +$600)
• PEP: $158 (-$2/share × 122 = +$244)
• Total profit: $844 on ~$19,500 capital (4.3% return in 4 weeks)
• Market direction didn't matter — both could have dropped 5% and the trade still profits from convergence

Example 2: Goldman Sachs (GS) vs JPMorgan Chase (JPM)

Why they're a pair:

• Both major US banks subject to Fed policy, interest rates, credit cycles
• Both have investment banking, trading, and consumer banking divisions
• Move with the same economic tailwinds and headwinds

The setup:

• Historical GS/JPM ratio: ~0.72 over 3 years
• Typical range: 0.65 to 0.79

A pair trade scenario:

• GS announces a major regulatory fine of $500M
• GS stock drops 6%, JPM unchanged
• New GS/JPM ratio: 0.65 (GS unusually cheap)
• Z-score: -2.3

Trade:

• Long 50 shares GS at $420 = $21,000
• Short 65 shares JPM at $320 = -$20,800 (dollar neutral)

Thesis: The regulatory fine was smaller than feared, GS's underlying business hasn't changed, the ratio will revert.

Outcome (2 weeks later):

• GS recovers 4%, JPM flat
• GS: $437 (+$17/share × 50 = +$850)
• JPM: $320 (+$0 × 65 = $0)
• Profit: $850 on $21,000 capital (4.0%)

Example 3: Ford (F) vs General Motors (GM)

Why they're a pair:

• Both US-based automakers with domestic manufacturing
• Same labor costs (UAW contracts), steel/aluminum inputs
• Same regulatory exposure (EPA, NHTSA, fuel economy standards)
• Both transitioning to electric vehicles on similar timelines

When this pair can trade:

• One announces a favorable UAW contract settlement, rallies
• One announces an EV partnership, temporary premium
• One beats quarterly production targets, stock jumps

The key: Both companies face the same long-term forces. Short-term divergences driven by company-specific news tend to fade.

Example 4: SPY vs IVV (ETF Pairs)

The cleanest pair that ever existed:

SPY (SPDR S&P 500 ETF) and IVV (iShares S&P 500 ETF) track the exact same index with virtually identical holdings. They should, in theory, always trade at almost identical prices relative to their NAVs.

Why they diverge:

• Slight differences in dividend treatment and timing
• Different liquidity profiles (SPY is more liquid)
• Institutional flow imbalances (pension funds may prefer IVV, retail traders prefer SPY)
• Arbitrage keeps spreads tight, but small opportunities exist

Who trades this:

• High-frequency trading firms running sub-millisecond arbitrage
• Not practical for most retail investors (spreads are fractions of a cent)

Why it matters: This shows that even nearly identical instruments diverge — and the mechanism that brings them back (arbitrage) is exactly the mechanism that makes pair trading work across all markets.

Example 5: ExxonMobil (XOM) vs Chevron (CVX)

Why they're a pair:

• Both integrated oil majors (upstream production + downstream refining)
• Nearly identical oil price sensitivity
• Same geopolitical exposure (Middle East risk)
• Similar dividend yields and capital allocation philosophies
• Both in Dow Jones Industrial Average

What makes a great pair trade setup:

• One announces a large acquisition (temporary premium from deal optimism)
• One has an operational incident (refinery fire, deepwater spill)
• Both face same oil price environment, but company-specific news creates divergence

The trade: Long the underperformer, short the outperformer. Wait for the company-specific noise to fade and fundamental similarity to reassert.

How the Pros Use It: Wall Street's Stat-Arb Operations

Professional pair trading at hedge funds is an entirely different beast from what individual investors do. Understanding how the pros operate gives you insight into the strategy's edges — and limits.

Statistical Arbitrage (Stat-Arb) at Scale

The major quant funds — Renaissance Technologies, Two Sigma, D.E. Shaw, Citadel, AQR — don't trade single pairs. They run systematic programs monitoring thousands of pairs simultaneously across all asset classes.

How a stat-arb desk operates:

Universe construction:

• Start with 500-5,000 liquid stocks
• Generate all possible pair combinations (500 stocks = 124,750 possible pairs)
• Filter by sector, cointegration tests, liquidity requirements
• Result: 500-2,000 actively monitored pairs

Signal generation (running 24/7):

• Calculate z-scores across all pairs in real-time
• Flag pairs where z-score exceeds threshold (±1.5 to ±2.0)
• Risk model checks: Is this pair's recent correlation stable? Any upcoming earnings or news?
• Position sizing: Allocate capital based on expected convergence time and confidence level

Execution:

• Automated order routing minimizes market impact
• Spread trades executed simultaneously (or within milliseconds)
• Both legs hedged: Long leg and short leg open within the same second

Portfolio management:

• At any given time: 100-500 active pair trades
• Overall portfolio: Roughly dollar-neutral and beta-neutral
• Daily P&L fluctuates in narrow band — high Sharpe ratio, low drawdowns

Risk management:

• Automatic stop if z-score reaches ±3.5
• "Pair divorce" detection: Algorithms detect when cointegration breaks down
• Correlation monitoring: Flag if a pair's correlation drops below historical minimum
• Event risk: Auto-reduce positions before earnings, FDA decisions, regulatory announcements

The "Quant Quake" of August 2007

One of the most instructive events in stat-arb history is the August 2007 Quant Quake, sometimes called the "quant meltdown."

What happened: Several large quant hedge funds were running similar stat-arb strategies. When one fund (believed to be Goldman Sachs Global Alpha) was forced to reduce risk (due to unrelated losses), they sold their stat-arb positions en masse. This caused the spread on hundreds of pairs to move against every other fund running similar strategies simultaneously.

The cascade:

1. Fund A forced to sell long positions and cover short positions
2. Long positions drop, short positions rise — exactly opposite of what pairs traders want
3. Other funds' risk models trigger automatic stop-losses
4. Those stops create further selling pressure
5. Self-reinforcing loop for 3-4 days

The lesson: When too many funds run similar strategies, crowding creates "factor crash" risk. Individual pair trades were fine in isolation — it was the systemic similarity of positions across hundreds of funds that caused the crash.

This matters for individual investors: The most popular pairs (Coke/Pepsi, GS/JPM) are heavily monitored by institutional players. Less obvious pairs — same-industry regional banks, international consumer staples — may offer cleaner opportunities with less crowding risk.

Long/Short Equity vs Pure Pair Trading

Many professional hedge funds use a long/short equity approach that's related to but different from pure pair trading:

Long/short equity:

• Long portfolio of stocks expected to outperform
• Short portfolio of stocks expected to underperform
• Net exposure: 40-60% long (not fully market neutral)
• Pairs may or may not be explicitly matched

Pure pair trading (stat-arb):

• Each long has a matched short in the same sector
• Zero net market exposure (truly neutral)
• Profit only from relative value, not market beta

Why pros prefer long/short: They want some market exposure (beta) for upside participation. A long/short fund with 50% net long captures market upside while reducing drawdowns.

Why pure pair trading: Capital preservation-focused funds, or risk parity allocations within a larger portfolio. Truly uncorrelated to market returns.

How Individual Investors Can Do It

You don't need a $1 billion fund or a PhD in statistics to trade pairs. With a margin account, basic tools, and discipline, individual investors can implement pair trades effectively.

What You'll Need

Account requirements:

• Margin account — Required for short selling (the short leg of every pair)
• Minimum equity: $2,000 (FINRA minimum), but $10,000-$25,000 for meaningful position sizes
• Broker: Interactive Brokers, TD Ameritrade, or any broker allowing margin short selling

Why Interactive Brokers is preferred:

• Low borrow rates for short selling (often 0.25-2% annually on liquid stocks)
• Portfolio margin available (more efficient capital use)
• Direct access routing for simultaneous execution
• Low commissions ($0.005/share, $1 minimum)

Tools you'll need:

• Spreadsheet (Google Sheets/Excel) — Track pair ratios, calculate z-scores
• Charting platform — View ratio charts (TradingView allows custom ratio charts)
• Stock scanner — Screen for unusual relative strength/weakness within sectors

Step-by-Step: Your First Pair Trade

Step 1: Pick your sector and candidates

Start with a sector you understand. Consumer staples is ideal for beginners (stable, slow-moving pairs).

Consumer staples pairs to consider:

• Coke (KO) vs Pepsi (PEP)
• Procter & Gamble (PG) vs Colgate (CL)
• Walmart (WMT) vs Target (TGT)

Banking pairs:

• Goldman Sachs (GS) vs Morgan Stanley (MS)
• Wells Fargo (WFC) vs Bank of America (BAC)
• JPMorgan (JPM) vs Citigroup (C)

Energy pairs:

• ExxonMobil (XOM) vs Chevron (CVX)
• ConocoPhillips (COP) vs Pioneer Natural Resources (PXD)
• Valero (VLO) vs Phillips 66 (PSX)

Tech pairs:

• Microsoft (MSFT) vs Alphabet (GOOGL)
• Visa (V) vs Mastercard (MA) — one of the most popular pairs
• AMD (AMD) vs Nvidia (NVDA) — highly volatile, not for beginners

Step 2: Build your ratio chart

On TradingView, type in the pair ratio directly:

• Search: KO/PEP → Shows the ratio chart
• Add a 252-day (1 year) simple moving average → That's your "mean"
• Add ±1 and ±2 standard deviation bands → Those are your entry/exit signals

Google Sheets approach:

• Column A: Date
• Column B: KO price
• Column C: PEP price
• Column D: =B/C (ratio)
• Column E: =AVERAGE(D2:D252) (rolling 1-year mean)
• Column F: =STDEV(D2:D252) (rolling 1-year standard deviation)
• Column G: =(D-E)/F (z-score)

Step 3: Validate the pair

Before trading, confirm the relationship is real:

Qualitative checks:

• Same sector (not just same industry, but truly similar business models)
• Both large-cap, liquid stocks (easy to short, low borrow cost)
• No pending merger/acquisition for either company
• Similar leverage ratios (high-debt company vs no-debt company can diverge structurally)

Quantitative checks:

• Correlation coefficient above 0.70 over past 3 years
• Z-score has crossed ±1.5 and returned to 0 multiple times (mean-reversion history)
• Z-score rarely stays above ±3 for more than 20-30 trading days
• No obvious structural breaks in the ratio chart (look for sudden permanent shifts)

Step 4: Size your position

Dollar-neutral sizing:

code-highlight
Dollar amount long = Dollar amount short

Example:

• KO/PEP ratio is at -2.0 z-score (KO cheap, PEP expensive)
• You want to risk $10,000 per leg
• KO at $65: Buy 153 shares ($9,945)
• PEP at $160: Short 62 shares ($9,920)
• Approximate dollar neutral ✓

Beta-neutral sizing (more advanced):

Dollar neutral isn't always beta neutral. If KO has beta of 0.55 and PEP has beta of 0.75:

• Your $10,000 long KO = $5,500 market exposure
• Your $10,000 short PEP = $7,500 market exposure
• Net: -$2,000 market exposure (you're slightly short the market)

Beta-neutral adjustment:

• Long KO: $13,636 ($10,000 / 0.55 × 0.75) to match PEP's market exposure
• Short PEP: $10,000

For beginners, dollar-neutral is fine. Beta-neutral is worth learning once you're comfortable.

Step 5: Execute both legs together

The key to pair trading is simultaneous execution — open both legs at the same time.

Why this matters:

• If you buy the long leg first, the stock might move before you execute the short
• You're now exposed to market direction until both legs are on
• In volatile markets, 30 seconds between legs can cost meaningful money

Execution approach:

• Use bracket orders if your platform supports them
• Otherwise: Have both order tickets pre-filled and submit within seconds
• Some platforms (Interactive Brokers, TradeStation) have "pair order" functionality

Step 6: Monitor and manage

Once the trade is on:

• Check z-score daily (or weekly if you're less active)
• Close the trade when z-score returns to 0 (or your target, like ±0.5)
• Cut the trade if z-score reaches ±3.5 (the pair may be breaking down)

Trade management scenarios:

Scenario A: Successful convergence

• Entered at z-score = -2.0
• After 2 weeks: z-score = -0.5
• Close both legs, take profit (pair almost fully converged)
• Result: ~75% of expected profit captured (good enough — don't wait for perfect z-score = 0)

Scenario B: Continuing divergence (danger zone)

• Entered at z-score = -2.0
• After 2 weeks: z-score = -2.8 (still widening)
• Reassess: Is there news explaining this? Is the pair breaking down?
• Wait until z-score = -3.0, then cut the trade

Scenario C: Permanent pair breakdown (worst case)

• Entered at z-score = -2.0
• Company A announces merger → stock gaps up 30%
• Z-score explodes to -6.0 overnight
• Cut immediately at market open (this is why stops exist)

ETF Pairs: Simpler, But Tighter Margins

For investors who want pair trading without short-selling individual stocks, ETF pairs offer a cleaner approach.

Why ETFs work for pairs:

• Lower borrow costs: Liquid ETFs are easy to short, borrow rates often 0-0.25%
• No earnings risk: ETFs don't have quarterly surprises
• No company-specific risk: Individual company crises won't blow up your trade
• Transparent holdings: You know exactly what you own

Popular ETF pairs:

Technology:

• QQQ (Nasdaq 100) vs XLK (Technology Select SPDR)
• Same assets, different weightings (QQQ is cap-weighted tech-heavy, XLK is slightly different)

Sector rotation pairs:

• XLE (Energy) vs XOM (Exxon, the largest energy stock)
• If XOM diverges dramatically from XLE, it may be specific to Exxon, not the sector

International pairs:

• EEM (Emerging Markets) vs EFA (Developed International)
• Bet on relative performance of emerging vs developed economies

Fixed income pairs:

• TLT (20+ Year Treasury) vs IEF (7-10 Year Treasury)
• Trade the yield curve shape (flattener or steepener trade)
• TLT/IEF ratio rises in bull market for bonds (curve steepening)

Commodity sector pairs:

• GDX (Gold Miners ETF) vs GLD (Gold ETF)
• When gold miners diverge from gold price itself, trade the reversion
• Miners should trade at a premium to gold (operating leverage), if that premium collapses, buy miners vs gold

Crypto-adjacent:

• MSTR (MicroStrategy, ~50% Bitcoin proxy) vs GBTC (Bitcoin Trust)
• Highly volatile, not for beginners, but a liquid pair

Practical Position Sizing for Individual Investors

Account size guidelines:

Capital allocation rules:

• Never put more than 20-25% of account in a single pair
• Keep 40-50% in cash/reserve (margin requirement buffer + new opportunities)
• Diversify across uncorrelated sectors (don't run 5 bank pairs + 0 others)

What Can Go Wrong: Pair Breakdown Risk

Pair trading sounds almost too good — market neutral, mean-reverting, quantifiably risky. But it has genuine failure modes.

The Fundamental Business Change

What it looks like: A company fundamentally changes its business model, competitive position, or capital structure in a way that permanently alters its relationship with its pair.

Real example: Sears and JC Penney were a long-running retail pair. Both were mid-market department stores facing similar competitive pressures. But when Amazon's impact accelerated differently for each (Sears went bankrupt faster), the pair diverged permanently and never converged.

Detection: Watch for:

• Major acquisitions (one company buys into a new business)
• Activist investors driving strategic changes
• Rapid shifts in revenue mix (Pepsi becoming more snacks-heavy vs Coke remaining beverage-focused)
• Leverage divergence (one company takes on massive debt)

The Acquisition Premium

What it looks like: A company in your pair gets acquired (or rumors begin). Its stock immediately prices in a takeover premium (usually 25-40%), causing a massive z-score divergence that won't converge.

Why it's dangerous: You're short the acquiree (you think it's expensive relative to its pair). It gets taken out at $50/share — 30% above your entry. The ratio doesn't mean-revert; the acquisition closes.

Detection: Monitor unusual options activity (high call volume, unusual put/call ratio), unusual stock volume, and M&A news in your sectors.

Prevention: Close pairs in sectors with high M&A activity (biotech, small-cap tech) before earnings or any rumor period.

The Short Squeeze

What it looks like: Your short leg (the overperforming stock) gets caught in a short squeeze — heavy short interest, positive news catalyst, shares start to cover rapidly. The stock you're short surges further, widening your pair spread.

Why it's dangerous: Pair trading is leveraged. If both legs are $10,000 each and the short leg rises 30%, you lose $3,000 on the short while the long (same-sector, sympathetic move) might only rise 10%, gaining $1,000. Net: -$2,000 on a $20,000 position.

Prevention: Before taking a pair short, check short interest on the intended short leg. Avoid heavily shorted stocks (above 20% of float) as pair trade shorts.

Crowding Risk (The Quant Quake Problem)

What it looks like: Many institutions run the same pairs. When they all need to exit simultaneously (for any reason), pairs diverge together, not independently.

Detection: If you're trading the most obvious pairs (Coke/Pepsi, GS/JPM, XOM/CVX), you're likely crowded with many other players. A broader market dislocation can cause simultaneous stop-outs.

Mitigation: Run less-popular pairs within sectors. Regional bank pairs, mid-cap consumer pairs, less-followed sectors. Less institutional competition means cleaner mean-reversion.

Borrow Costs Eroding Profit

What it looks like: The short leg of your pair has high borrow costs. You might be right about the pair converging, but the 15-25% annual borrow fee on your short eats all your profits.

Example:

• You enter a pair trade targeting a 4% spread convergence profit
• The short leg has an 8% annual borrow cost
• Your trade takes 3 months to converge: 8% annual ÷ 4 quarters = 2% borrow cost
• Net profit: 4% - 2% = 2% (acceptable)
• But if trade takes 8 months: 8% ÷ 12 × 8 = 5.3% borrow → Loss

Prevention: Check borrow cost before entering (your broker shows this). Avoid any pair trade where the borrow rate exceeds your expected profit.

Pair Trading Approaches: A Quick Reference

Pair Trading vs Other Market-Neutral Strategies

It's worth understanding how pair trading fits into the broader universe of market-neutral approaches:

Pure pair trading:

• Matched long/short within same sub-industry
• Truly sector and market neutral
• Mean-reversion based

Long/short equity:

• Long basket of strong stocks, short basket of weak stocks
• Usually some net long exposure (40-60%)
• Return = market beta + alpha (stock picking)

Risk arbitrage (merger arb):

• Buy acquisition target at discount to deal price
• Short acquirer when deal may be dilutive
• Return based on deal closing probability

Convertible bond arbitrage:

• Buy convertible bonds, short underlying stock
• Profit from mispricing of the embedded option

Statistical arbitrage:

• Pair trading at scale, using many pairs simultaneously
• Systematic, quantitative, often using machine learning signals
• Requires significant infrastructure

Pairs trading advantages vs alternatives:

• No fundamental research required (quantitative, not qualitative)
• Lower market impact (you're one of many small traders, not moving prices)
• Scalable for individuals (start with $10k, scale as you learn)
• Concrete entry/exit rules (no ambiguity about when to act)

Pair Trading Summary: Key Principles

The core rules:

1. Only trade cointegrated pairs — correlation alone is not enough
2. Enter at ±2 standard deviations — statistically meaningful divergence
3. Exit at zero — don't hold waiting for more; take the convergence
4. Cut at ±3.0 to ±3.5 — the pair may be breaking down permanently
5. Dollar-neutral sizing — equal dollar amounts long and short
6. Check borrow costs first — high borrow rates can turn profits into losses
7. Avoid M&A-prone pairs — acquisition premium permanently breaks pairs
8. Diversify across sectors — don't run 5 bank pairs simultaneously

What pair trading is:

• Market-neutral income generation with defined entry and exit signals
• A strategy for relative value, not directional market calls
• A tool for reducing correlation to broad market movements

What pair trading is not:

• Risk-free arbitrage (pairs can and do break down)
• A replacement for directional trading (you miss pure momentum)
• Suitable for illiquid stocks (execution becomes the problem)

The honest edge: Pair trading has a real edge in mean-reversion tendencies within industries. Companies in the same industry, facing the same macro environment, should trade at consistent relative valuations. When they don't — when one becomes temporarily too cheap or too expensive relative to its peer — that divergence is tradeable. Not always, not every time, but often enough to generate consistent returns with lower risk than directional trading.

Getting Started: Your Pair Trading Action Plan

Week 1 — Learn the mechanics:

• Build a ratio chart for KO/PEP on TradingView (type KO/PEP in the search bar)
• Add a 52-week moving average and ±2 standard deviation bands
• Watch how the ratio oscillates around the mean without trading anything

Week 2 — Build your screening process:

• Pick 3-5 sectors you know well
• Identify 2-3 candidate pairs per sector
• Build a simple Google Sheet to track daily ratios and z-scores

Week 3 — Paper trade:

• When a z-score hits ±2.0 on one of your monitored pairs, record it as a paper trade
• Track the outcome: Did it converge? How long did it take?
• Paper trade 5-10 pairs before risking capital

Week 4+ — First real trade:

• Use a small position (5-10% of your account) on the pair you're most confident in
• Execute both legs simultaneously, set your stop in your head (don't forget it)
• Close when z-score returns to ±0.5, whether that's 2 weeks or 8 weeks

Resources to go deeper:

• "Pairs Trading: Quantitative Methods and Analysis" by Vidyamurthy — The academic foundation
• "Statistical Arbitrage" by Andrew Pole — How hedge funds implement at scale
• TradingView: Free ratio charts for any two symbols
• Interactive Brokers: Best broker for pair traders (low borrow costs, good execution)
• Stock Alarm Pro: Set relative alerts to notify you when one stock in a pair significantly outperforms the other