Meaning and portfolio use
Covariance shows the joint deviation of two return series from their respective means. Unlike correlation, it depends on the scale and representation of the returns. In the portfolio-variance formula, the covariance term captures how joint movement interacts with the portfolio weights.[104](#source-cv-covariance)
Illustrative calculation
With assumed volatilities of 15% and 25% and a correlation of 0.2, covariance in decimal form is 0.0075. This is not a portfolio return or a promised percentage. If returns are recorded in percentage points rather than fractions, the numerical scale of covariance changes.
Data and interpretation limits
The calculation requires aligned observations, a consistent frequency and stated estimation rules. Pairwise measures based on different date sets may not form a coherent overall picture without additional checks. Covariance does not explain the cause of a relationship and does not automatically measure default or maximum drawdown. Use it as a component of a statistical model while separately describing price quality, the observation window and the possibility that relationships may change.