What is Vector AutoRegression (VAR) and how does it model multivariate time series?

VAR models multiple time series where each variable depends on past values of itself and other variables.

Vector AutoRegression (VAR) extends AR models to multivariate settings. Instead of modeling a single series, VAR treats every variable as a function of lagged values of all variables in the system. This allows capturing complex interdependencies such as feedback loops. It is widely used in econometrics, finance, and macroeconomic forecasting. The key assumption is that all variables are jointly stationary.