Using Correlation Measures to Identify Trends and Build Trading Signals
Summary
This article explains how correlation between price and time can quantify the direction and strength of a market move. It introduces Pearson correlation on raw values, Spearman correlation on ranks, Kendall’s rank measure, and Fechner’s sign-based measure. A worked downtrend example calculates several coefficients and finds strong negative relationships, illustrating that different methods can describe the same broad movement with differing magnitudes. The article then applies correlation measures in indicators and a correlation-based trading system, and describes testing that system.
The central use is to track a coefficient against a threshold: its sign indicates direction, its magnitude reflects association, and persistence above a threshold can help characterize trend duration. The text also discusses linear versus nonlinear relationships and limitations of rank methods. Correlation is a statistical description, not proof of predictive causality or a guarantee of entry and exit timing. Although the article includes an example and system testing, the provided excerpt omits much of the later implementation and test results, limiting assessment of performance and robustness.
Key ideas
- Correlation between price and bar number can summarize trend direction and strength.
- Pearson uses raw values, while Spearman and Kendall use ranks and Fechner compares deviation signs.
- Tracking coefficient magnitude and persistence against a threshold can help describe trend strength and duration.
- Different correlation measures can produce different coefficient magnitudes on the same price sequence.
- Correlation describes association and cannot guarantee predictive timing or profitable trades.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.