What problem does it solve?
This Skill helps analysts identify strong co-movement and long-run relationships between assets so they can build and validate pairs-trading candidates instead of relying on correlation alone.
Core Features & Use Cases
- Co-Movement Discovery: Scan a universe by Pearson/Spearman correlations and shortlist Top-K candidates for follow-up testing.
- Deep Return-Correlation Analysis: Compute multiple correlation measures (Pearson, Spearman, Kendall), OLS beta/R², rolling correlation, and spread Z-scores.
- Sector Clustering: Use hierarchical clustering on the correlation matrix to uncover sector-like structure and diversify exposures.
- Realized (Regime-Conditional) Correlation: Measure rolling correlation and analyze how correlation changes under bull/bear/sideways/high-vol regimes.
- Cointegration Testing Framework: Run Engle-Granger and Johansen cointegration tests to confirm long-run equilibrium relationships.
- Pairs Trading Readiness Metrics: Estimate spread half-life and (optionally) use a Kalman filter for a dynamic hedge ratio.
- Signal Generation Workflow: Convert correlation/cointegration results into actionable Z-score-based long/short/exit signals.
Quick Start
Use correlation-analysis to scan a candidate universe, test cointegration, estimate half-life and (optionally) dynamic hedge ratios, and output Z-score-based pair-trading signals from two asset price series.
