Sources of real exchange-rate fluctuations: empirical evidence from four Pacific Basin countries.

• Author: Chen, Show-Lin

1. Introduction

   In international finance, the competing hypothesis that the logarithms of real exchange rates appear to be well-described as random walks during a period of floating rate, has found some support. Studies by Adler and Lehmann [2], Hakkio [12] and Mark [18] are examples of work supporting this view. If a real exchange rate process has been approximated to be non-stationary it implies there is little tendency for the real exchange rate to be mean-reverting, and also that deviations from the purchasing power parity (PPP) are permanent.

   Modern models of exchange rate determination suggest that real shocks can induce permanent changes in the real exchange rate. Stockman [21] pointed out that according to the equilibrium approach, the behavior of real exchange rates since the collapse of Bretton Woods could reflect, not the importance of sluggish price-level adjustment, but rather the influence of real shocks with significant permanent components. Moreover, the evidence in Campbell and Clarida [5] and Huizinga [13] suggests that real exchange rates contain both permanent and transitory components, with most of their variation explained by the permanent components.

   However recent studies employing long-run data, or more powerful tests, tend to confirm the PPP hypothesis. Studies by Abuaf and Jorion [1], Kim [16], Whitt [22] and Diebold, Husted, and Rush [7] are four such examples. Their evidence shows that real exchange rates are stationary, which implies that the behavior of real exchange rates are influenced by transitory nominal disturbances. Hence, the validity of PPP remains at issue.

   Identifying the sources of exchange rate fluctuations is important not only for establishing the validity of PPP but also for achieving successful exchange rate stabilization. Attempts to stabilize exchange rate changes that are due to economic fundamentals could be futile and even harmful to the economy. In addition, measuring the relative importance of permanent and transitory shocks on exchange rates is essential for exchange rate modeling. If exchange rates are dominated by real shocks then the "equilibrium approach" offered by Stockman [21] is appropriate in analyzing the behavior of exchange rates. Conversely, if the evidence suggests the contrary, then the "disequilibrium approach" of Dornbusch [8] should be considered as an alternative. As a consequence, it is important to develop an empirical analysis to measure the relative importance of permanent and transitory shocks on exchange rates.

   Recently, the approach of long-run structural VAR proposed by Blanchard and Quah [4] has been used to examine the sources of the macroeconomic fluctuations.(1) Lastrapes [17] and Evans and Lothian [10] extend this approach in examining the sources of real exchange rate fluctuations. In general, their findings support the importance of real shocks. To date, the majority of studies concerning the sources of real exchange rate fluctuations have been conducted for developed countries; most of these studies indicate that real shocks dominate in the variability of real exchange rates.

   However, it is interesting to consider the behavior of the real exchange rate for the Pacific Basin countries. Over the past decade, countries in the Pacific Basin have instituted a continuing policy of financial market liberalization and as a result have experienced rapid growth, which has lead to increasingly strong trading ties to the United States.

   The purpose of this paper is to apply the method of long-run structural VAR to analyze the influence of real shocks on the fluctuations of real exchange rates for four Pacific Basin countries. Our paper attempts to distinguish between empirically real, as opposed to nominal sources of exchange rate fluctuations. We find that real shocks dominate the variability of real exchange rates. This is consistent with the finding of Lastrapes [17].

   Our paper is organized as follows: Section II provides the empirical model, discusses the econometric method, and examines the model's identification. Section III presents empirical results, and our conclusions are summarized in section IV.

2. The Structural VAR Model

   In this section, we apply the long-run structural VAR approach to examine the influence of real shocks on the fluctuations of real exchange rates. We utilize Blanchard and Quah's [4] method to show how our assumptions help in characterizing the exchange rate process, and how this process can be recovered from the data.

   Suppose that the economic system is driven by two structural shocks: real disturbances, and monetary disturbances. Furthermore, let us assume the monetary shock has no long-term effect on the real exchange rate; at the same time these two disturbances are uncorrelated and have unit variance.

   Firstly, we will consider a vector of stationary variables [X.sub.t] and a vector of structural shocks [U.sub.t]. In this paper, [X.sub.t] includes the change in real and nominal exchange rate and [U.sub.t] consists of the real and monetary innovations. Then [X.sub.t] can be represented by:

   [X.sub.t] = [summation of] [A.sub.k][U.sub.t-k] where k=0 to [infinity] = A(L)[U.sub.t], Var([U.sub.t]) = I (1)

   where [X.sub.t] = {[Delta][s.sub.t], [Delta][q.sub.t]}[prime], [Delta][s.sub.t] and [Delta][q.sub.t] are the change in nominal and real exchange rates respectively, and...