Monetary Policy Transmission in Canada: A FAVAR Analysis

Abstract

This paper employs the Factor Augmented Vector Autoregression (FAVAR) method used in Boivin, Giannoni and Stevanovic (2010) to analyze traditional channels of monetary policy transmission in Canada and to estimate the industrial and provincial effects of monetary policy. The benchmark model includes 409 quarterly macroeconomic time series from 1981Q2 to 2016Q2. The impulse response results are broadly consistent with the theoretical prediction of a small open economy model, except for CPI and total exports that are relatively nonresponsive to the monetary policy shock. The identified monetary policy shocks have heterogeneous sectorial effects, which are the largest for industrial GDP in manufacturing and wholesale trade, and for employment in manufacturing and construction. Furthermore, this paper finds that there are provincial differences in the responses to monetary policy shocks, with employment in Quebec and Ontario and housing price in Ontario and British Columbia being the most affected.