📈 Interactive Pairs Trading Backtester

Pairs Trading Explained

Pairs trading is a market-neutral strategy that aims to profit from temporary mispricings between two related assets (a "pair").

Core Idea:

• Identify two assets (e.g., stocks in the same sector) whose prices historically move together.
• Calculate the "spread" between their prices (often using price ratios or differences).
• Assume this spread is mean-reverting – meaning it tends to return to its historical average over time.
• When the spread widens significantly beyond its average, bet on it reverting:
  • If Spread is unusually HIGH: Short the outperforming asset and Long the underperforming asset.
  • If Spread is unusually LOW: Long the underperforming asset and Short the outperforming asset.
• Close the position when the spread returns closer to its average.

How This Backtester Works:

• Data: Fetches daily closing prices from the TradingMRI API (limited to 120 points).
• Spread Calculation: Uses the natural logarithm of the price ratio: log(Price Asset 1 / Price Asset 2).
• Mean Reversion Signal (Z-Score):
  • Calculates a rolling average (mean) and standard deviation of the spread over the specified "Lookback" period.
  • Determines the Z-Score for the current spread: (Current Spread - Rolling Mean) / Rolling Standard Deviation.
  • The Z-Score measures how many standard deviations the current spread is away from its recent average.
• Trading Signals:
  • Entry: Enters a trade when the absolute Z-Score (|Z-Score|) exceeds the "Entry Z-Score Threshold".
    • If Z-Score > Entry Threshold: Short the spread (Short Asset 1, Long Asset 2).
    • If Z-Score < -Entry Threshold: Long the spread (Long Asset 1, Short Asset 2).
  • Exit: Exits the trade when the absolute Z-Score falls back below or equals the "Exit Z-Score Threshold".
• Position Sizing: Attempts to make the trade approximately dollar-neutral at entry by allocating roughly equal capital to the long and short legs based on current prices and portfolio value.
• Simulation: Steps through the historical data day by day, tracking cash, positions, and calculating portfolio value.

Important Limitations:

• No Costs/Slippage: Ignores commissions, fees, and the potential difference between the desired trade price and the actual execution price (slippage). These can significantly impact real-world returns.
• Simplified Cointegration: Uses Z-Score deviation as a proxy for mean reversion. A formal statistical test (like Augmented Dickey-Fuller) is typically used to confirm cointegration before trading.
• Data Limitation: TradingMRI free tier provides only 120 data points, limiting the backtest period.
• Look-Ahead Bias Avoided (Mostly): Uses rolling calculations, but assumes trades execute at the closing price of the signal day.
• No Risk Management: Doesn't include stop-losses or sophisticated risk controls.
• Educational Use Only: This tool is for learning and exploration, NOT financial advice.