<b>BVAR</b>: Bayesian Vector Autoregressions with Hierarchical Prior Selection in <i>R</i>

Vector autoregression (VAR) models are widely used for multivariate time series analysis in macroeconomics, finance, and related fields. Bayesian methods often employed to deal with their dense parameterization, imposing structure on model coefficients via prior information. The optimal choice of the degree informativeness implied by these priors is subject much debate can be approached hierarchical modeling. This paper introduces BVAR, an R package dedicated estimation VAR selection. It implements functionalities options that permit addressing a wide range research problems, while retaining easy-to-use transparent interface. Features include structural impulse responses, forecasts, most commonly conjugate priors, as well framework defining custom dummy-observation priors. BVAR makes user-friendly provides accessible reference implementation.