Covariance and Correlation for Financial Market Analysis
Summary
The document explains covariance and correlation as measures of how two variables move together. Covariance indicates the direction of their linear co-movement, while correlation also expresses its strength on a standardized scale from negative one to positive one. It compares their ranges, units, and sensitivity to scaling, then presents population and sample covariance formulas and the relationship between covariance and correlation. Examples use stock prices to illustrate positive and negative associations.
The article also describes applications in trading analysis, including examining relationships between financial variables and calculating rolling estimates as those relationships change over time. It outlines rolling calculations using a 20-day window for Microsoft and Tesla, with plots showing that their co-movement can shift between positive and negative. These examples are educational rather than evidence of a trading edge: the article gives no systematic performance results, and linear association alone does not establish causation or guarantee future behavior.
Key ideas
- Covariance indicates whether two variables tend to move in the same or opposite direction.
- Correlation expresses direction and strength on a standardized scale from negative one to positive one.
- Covariance depends on measurement units and scale, while correlation is unitless and scale invariant.
- Sample covariance uses a different denominator from population covariance to account for sample estimation.
- Rolling covariance and correlation can track changing relationships between assets over time.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.