A new algorithm for structural restrictions in Bayesian vector autoregressions

A comprehensive methodology for inference in vector autoregressions (VARs) using sign and other structural restrictions is developed. The reduced-form VAR disturbances are driven by a few common factors identification can be incorporated their loadings the form of parametric restrictions. Gibbs sampler derived that allows parameters to sampled efficiently one step. key benefit proposed approach it treating parameter estimation as joint problem. An additional scale large VARs with multiple shocks, extended accommodate non-linearities, asymmetries, numerous interesting empirical features. excellent properties new algorithm explored synthetic data experiments, revisiting role financial economic fluctuations based on