Betta Signals

Learn · Risk and performance

Covariance

Covariance measures whether two variables, such as two return series, tend to move above and below their averages together.

By Betta Signals Research Desk · Reviewed October 24, 2026

Plain-English definition: Covariance measures whether two variables, such as two return series, tend to move above and below their averages together.[1]

How the term is used

Positive covariance indicates same-direction co-movement in the sample; negative covariance indicates opposite movement. Unlike correlation, its magnitude depends on the units and variability of the inputs.[1]

What to check before relying on it

Covariance is historical and sample-dependent. Dates, return frequency, missing observations, currency, and outliers need to be aligned before the result is meaningful.

Example

Beta divides the covariance between an asset and its benchmark by the benchmark's variance, turning co-movement into a benchmark-sensitivity estimate.

BetaCorrelationVolatility

Research last reviewed: October 24, 2026. This definition is educational and does not evaluate a particular security.

Sources

  1. TU Delft MUDE Textbook: Covariance and Correlation Coefficient — covariance definition, sign, units, and its relationship to correlation. Accessed October 24, 2026.