Wenting Liao, Jun Ma, Chengsi Zhang

Read Full Paper

This paper presents a novel econometric approach to estimating time-varying policy effects in the presence of time-varying instrument relevance, using a factor-augmented VAR model with data from the United States, Canada, Germany, Japan, and the United Kingdom. The study finds that US monetary policy shocks are a key driver of exchange rate movements in other advanced economies, without delayed overshooting. It also highlights that without considering the time variation in instrument relevance, estimates of the spillover effects of US monetary policy shocks on inflation and real economic activity could be distorted. The time variation in policy effects primarily reflects changes in the shock size rather than their transmission.

The core of the approach lies in developing an SVAR-IV model with time-varying coefficients and stochastic volatility, allowing for better capture of the time variations in monetary policy effects and addressing the issue of constant instrument relevance in traditional models. By estimating the spillover effects of US monetary policy shocks on other advanced economies, the paper shows that these shocks have significant effects on exchange rates, inflation, and real economic activity. These effects tend to be particularly pronounced during economic recessions, especially during the 2008–2009 global financial crisis.

The paper compares its findings with existing literature, pointing out that most studies assume a constant relationship between instruments and policy shocks, which may not hold, especially during the zero lower bound period. The proposed method allows for a more robust estimation of time-varying policy effects. The study also finds that the exchange rate response to US monetary policy shocks does not exhibit delayed overshooting, corroborating similar findings from other research. Furthermore, the paper provides new insights into the global spillover effects of US monetary policy by jointly estimating the time-varying FAVAR model, a more advanced method than two-step approaches used in previous studies.

Overall, this paper provides a new method for accurately estimating the time-varying effects of monetary policy shocks, especially in multi-country settings, and offers a fresh perspective and methodological support for research on global monetary policy spillovers.