Real exchange rate fluctuations and monetary shocks: A revisit

In this paper, I first estimate a structural VAR model by following Clarida and Gali (1994) and obtain results indicating that the variance of real exchange rates can be attributed more to monetary shocks when the sample span is extended. In order to further investigate this aspect, I then employ a VAR model with long-run US-UK annual data from 1889 to 1995. According to the variance decompositions, I find that monetary shocks can explain nearly 50% of real exchange rate variance in the long-run sample periods. All the evidence suggests that monetary shocks are indeed more important in a larger sample set. Copyright (C) 2003 John Wiley Sons, Ltd.