Exchange Rate Dynamics and Global Monetary Policy Spillovers: Time-Varying Dynamic Causal Effects

Abstract

We show how to study time-varying dynamic causal effects of structural shocks using external instruments in a generalized Factor-Augmented-VAR(FAVAR) model with time-varying parameters and stochastic volatility. Specifically, we employ the Bayesian MCMC estimation methodology and focus on global factors of outputs, inflation rates, interest rates, and exchange rates in five representative advanced economies, namely the United States, Canada, Germany, Japan, and the United Kingdom, to study time-varying impacts of an exogenous U.S. monetary policy shock on these open economies. We find uniformly strong evidence over time in support of Dornbusch's theoretical prediction of the exchange rate overshooting in response to an exogenous monetary policy shock. The delayed overshooting puzzle commonly documented in the literature disappears, likely thanks to the better identification of the exogenous U.S. monetary policy shock via external instruments. We also find that the U.S. monetary policy shock has significant contributions to the dynamics of the exchange rates overall, and the contributions were particularly large during the period of global financial crisis in 2007-2008. We document a great deal of time variations of impacts of the U.S. monetary policy shock on global factors of all other economic variables including interest rates, inflation rates, and outputs. These empirical results lend strong support to the extension of the baseline model to the more general model with time-varying parameters and stochastic volatility.