Pearson Correlation Pairs Trading with Comover Portfolios

Summary

This article presents a pairs trading method that selects stocks using correlations between their returns. In a formation period, it calculates monthly returns, finds each stock’s most correlated peers, and forms an equal-weighted peer portfolio. Regression estimates the relationship between each stock and that portfolio. During trading, the prior month’s divergence from the fitted relationship ranks stocks: the strategy goes long the highest-ranked group and short the lowest-ranked group, holding positions for a month.

Compared with the basic distance method, this approach focuses on return correlations rather than minimizing squared price differences. Its diversified peer portfolio may help distinguish stock-specific moves from broader movements, and cited research reports stronger, more robust excess returns for quasi-multivariate methods across a range of thresholds. However, return correlation does not establish an equilibrium relationship, and the article notes that reversal is not theoretically guaranteed. It suggests testing divergence reversion and potentially adding cointegration screening; the discussion offers no new empirical test of its own.

Key ideas

• The method selects stock peers by Pearson correlation of returns during a formation period.
• A peer portfolio provides a benchmark for estimating each stock’s return relationship.
• Prior-month divergence ranks stocks for long and short portfolio selection.
• Correlation-based selection is less restrictive than minimizing squared price differences.
• Return correlation alone does not establish cointegration or guarantee divergence reversal.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.