Measuring the impact of monetary policy: a factor-augmented vector autoregressive (favar) approach under bayesian framework

In this paper we provide evidence of the impact of monetary policy on a broad range of macro-economic variables for U.S, Canada, U.K., and Japan using factor-augmented vector auto regressive (FAVAR) model developed by Bernanke, Boivin and Eliasz (2003). Traditional approaches, such as vector auto regressive (VAR) models have not yielded satisfactory results because of the sparse information sets employed in these models. The recently developed FAVAR approach resolves this issue by augmenting VAR model with factors summarizing the information of a vast data set that is used by central banks in monetary policy decision making process. By using monthly data of 55 to 70 macroeconomic variables from the period starting as early as 1990 ending in 2010, we first show that the factors have additional information in summarizing the behavior of major economic variables and second that how contractionary monetary policy impacts a broad range of macroeconomic variables.