Factor-Augmented Vector Autoregression with narrative identification. An application to monetary policy in the US

Abstract

We extend the Bayesian Factor-Augmented Vector Autoregressive model (FAVAR) to incorporate an identification scheme based on an external instrument approach. Using this novel modelling framework, we show that a monetary policy tightening in the United States has contractionary effects on the economy. Moreover, accounting for a large information set seems to help mitigate price and real economic puzzles in the estimated impulse responses.