ON PREDICTING THE RELATIVE EFFECTS OF SUPPLY, DEMAND, AND MONEY SHOCKS ON REAL EXCHANGE RATE FLUCTUATIONS

Abstract

This study builds off the work of Clarida and Gali (1994), who use structural VAR modeling and forecast-­‐error variance decompositions to find the percentages of real exchange rate (RER) fluctuations that can be attributed to supply, demand, and nominal shocks. Applying a similar methodology, this study analyzes the effects of economic development level, geographical region, and economic openness on the percentages generated from the SVAR model. This work contributes to the understanding of international monetary finance in several ways: first, through the inclusion of an additional twenty years worth of data; secondly, by including many economies that have not yet been studied in this way, including the Euro zone; and thirdly, by going further in the analysis than most studies of sources of RER fluctuations have, analyzing the effects of economic characteristics on the variance decomposition results. Though the work of Hau (2002) and the Balassa-­Samuelson model suggest that geography, economic development, and economic openness will all be significant predictors of the generated regressors, the empirical evidence fails to support these expectations. One reason for this observed lack of significance is that the geography variable could be simultaneously capturing opposing effects caused by geographical region and economic development; countries in similar regions tend to have roughly the same level of economic development as well. Thus, separating out these dual effects is among the suggestions for further research presented in this paper.