Monetary Policy and Factor-Augmented VAR Model

Measuring the transmission of monetary policy is the main subject in a large empirical literature, the conclusions about the role that monetary policy plays in the economy and the way it should be conducted by the central banks depending on the way the monetary policy affects each economy. Although the VAR approach to measuring the effects of monetary policy shocks appears to deliver a great deal of useful structural information, there are some problems that might arise during implementation. Factor Augmented Vector Autoregression models have been introduced into the economic literature in order to solve the issues that have been raised during a wide number of studies on the effect of monetary policy innovations on macroeconomic variables. For estimation purposes, 92 variables representing the evolution of production index, producer price index, consumer prices, unemployment rate, over the period 2001:M1 and 2013:M6 are considered. Principal component analysis is used in order to estimate the factors included in the model and Bayesian inference is used afterwards, more precisely Gibbs sampling algorithm, with the purpose of computing the posterior distribution of the unobservable states and the hyper-parameters.