covariance

Covariance

A measure of how two variables change together — positive covariance means they tend to increase together; negative means when one increases the other tends to decrease.

When you'd see it: Portfolio management (covariance of asset returns drives diversification math), in the underlying math of PCA and regression, and in multivariate statistics generally. Covariance is typically converted to correlation to make it interpretable.

Why it matters: Correlation is the normalized version of covariance and is easier to interpret (bounded -1 to +1). Covariance is the raw building block — it's what the math needs. Understanding the relationship between the two helps when reading technical documentation about multivariate models.

Common mistakes: Trying to interpret covariance magnitudes directly. A covariance of 500 vs 5 doesn't mean one relationship is stronger — the scale depends on the units of the original variables. Always normalize to correlation when you need to compare relationship strengths.

Related correlation, Pearson correlation, variance, principal component analysis