Assessing monetary policy in the euro area: a factor-augmented VAR approach

Abstract

In order to overcome the omitted information problem of small-scale Vector Autoregression (VAR) models, this study combines the VAR methodology with dynamic factor analysis and assesses the effects of monetary policy shocks in the euro area in the period during which there is a single monetary policy. Using the Factor-Augmented Vector Autoregressive (FAVAR) approach of Bernanke et al. (2005), we summarize the information contained in a large set of macroeconomic time series with a small number of estimated factors and use them as regressors in recursive VARs to evaluate the impact of the nonsystematic component of the European Central Bank's (ECB's) actions. Overall, our results suggest that the inclusion of factors in the VAR allows us to obtain a more coherent picture of the effects of monetary policy innovations, both by achieving responses easier to understand from the theoretical point of view and by increasing the precision of such responses. Moreover, this framework allows us to compute impulse-response functions for all the variables included in the panel, thereby providing a more complete depiction of the effects of policy disturbances. However, the extra information generated by the FAVAR also delivers some puzzling responses, in particular those relating to exchange rates.