pair-trading

Mean-reversion trading signals are generated by computing the price ratio between two correlated instruments and applying a Z-score to the spread. Configurable lookback, entry, and exit thresholds trigger long or short signals to capture reversion back to the mean.

Pair trading exploits mean reversion between two highly correlated instruments by trading the price ratio around its mean. The Z-score measures the spread's deviation, triggering long or short pairs when the entry threshold is crossed and exiting upon reversion to the exit threshold.

Yes, this mean-reversion workflow applies to both equities and crypto trading. It requires exactly two instruments with aligned OHLCV data to compute the price ratio and generate deterministic, backtestable trade signals for your execution engine.

NaN values during the initial lookback window are handled automatically by the deterministic workflow. The engine computes the price ratio and Z-score while managing the incomplete data period, ensuring per-instrument signals are only generated once the lookback window is fully populated.

The approach is limited to exactly two correlated instruments with aligned OHLCV data. It assumes the price ratio mean-reverts, so it may not work for instruments whose correlation breaks down, and it requires configuring appropriate lookback and threshold parameters to avoid false signals.

Yes, aligned OHLCV data for exactly two instruments is required to calculate the price ratio and Z-score. The deterministic workflow depends on this aligned historical data to output accurate per-instrument mean-reversion trade signals for backtesting.