What is Pair Trading? Pair trading is a market-neutral strategy where you trade two correlated currency pairs: Buy (Long) the underperforming currency Sell (Short) the outperforming currency

This takes advantage of the mean reversion principle, assuming the two currencies will eventually move back to their historical correlation.

🔹 How to Implement Forex Pair Trading?

1️⃣ Choose Highly Correlated Currency Pairs

• EUR/USD & GBP/USD (strong positive correlation)
• USD/JPY & EUR/JPY
• AUD/USD & NZD/USD
• USD/CAD & Crude Oil Prices (inverse correlation)

2️⃣ Calculate Correlation (Using Pearson's Coefficient)

• A correlation above +0.8 means strong positive relation.
• A correlation below -0.8 means strong negative relation.

3️⃣ Find Trading Signals

• Use Z-Score (Mean Reversion Strategy): Z=(Pricepair1−Pricepair2)−MeanStandard DeviationZ = frac{(Price_{pair1} - Price_{pair2}) - Mean}{Standard Deviation}
• If Z-Score > +2, short the stronger currency and long the weaker.
• If Z-Score < -2, do the opposite.

4️⃣ Execute Trades Simultaneously

• Go Long on the weaker currency and Short the stronger.
• Close the trade when the spread reverts to mean.

🔹 Python Code for Forex Pair Trading

This Python script fetches forex data, calculates correlation, and generates trading signals using Z-score.

import pandas as pd
import numpy as np
import yfinance as yf
import matplotlib.pyplot as plt

# Fetch historical forex data
pair1 = "EURUSD=X"
pair2 = "GBPUSD=X"

data1 = yf.download(pair1, period="6mo", interval="1d")["Close"]
data2 = yf.download(pair2, period="6mo", interval="1d")["Close"]

# Combine into a DataFrame
df = pd.DataFrame({pair1: data1, pair2: data2})

# Calculate spread (price difference)
df["Spread"] = df[pair1] - df[pair2]

# Calculate moving average and standard deviation
df["Mean"] = df["Spread"].rolling(window=20).mean()
df["Std"] = df["Spread"].rolling(window=20).std()

# Calculate Z-score
df["Z-Score"] = (df["Spread"] - df["Mean"]) / df["Std"]

# Trading signals
df["Long"] = df["Z-Score"] < -2 # Buy pair1, Sell pair2
df["Short"] = df["Z-Score"] > 2 # Sell pair1, Buy pair2

# Plot Z-Score and trading signals
plt.figure(figsize=(10, 5))
plt.plot(df.index, df["Z-Score"], label="Z-Score", color="blue")
plt.axhline(2, color='red', linestyle='--', label="Sell Signal")
plt.axhline(-2, color='green', linestyle='--', label="Buy Signal")
plt.legend()
plt.title("Pair Trading Z-Score Signals")
plt.show()

print(df.tail(10)) # Display last 10 rows with signals

🔹 How the Code Works

✅ Fetches EUR/USD & GBP/USD data
✅ Computes the spread between the two pairs
✅ Calculates Z-score (deviation from the mean)
✅ Generates Buy/Sell signals based on mean reversion
✅ Plots the Z-score chart with trading signals

🔹 Key Takeaways

✔ Hedge Risk: Trading correlated forex pairs reduces risk.
✔ Market-Neutral: Works in both bullish and bearish markets.
✔ Statistical Approach: Based on math, not just technicals.
✔ Automate with AI: Use machine learning to predict optimal entry & exit points.