Inference in Bayesian Proxy-SVARs

Motivated by the increasing use of external instruments to identify structural vector autoregressions (SVARs), we develop an algorithm for exact finite sample inference in this class time series models, commonly known as Proxy-SVARs. Our makes independent draws from any posterior distribution over parameterization a Proxy-SVAR. approach allows researchers simultaneously proxies and traditional zero sign restrictions shocks. We illustrate our methods with two applications. In particular, show how generalize counterfactual analysis Mertens Montiel-Olea (2018) identified