This paper reexamines whether the real exchange rate follows a random walk. We test the null hypothesis of a unit root against the alternative of stationarity and also the null hypothesis of stationarity against the alternative of a unit root. The test proposed by Kwiatkowski, Phillips, Schmidt and Shin (KPSS, 1992) is modified and applied to the monthly and annual data. While our monthly series suggest somewh at mixe d results, the results of annual data favor the stat i o n a rity hypothesis

more »

... n our both tests. We conclude that the real exchange rate may have a long mean-reversion component that the conventional unit root tests are not powerful enough to detect in a short span sample. [F 31] of the real exchange rate. Abuaf and Jorion (1990), Diebold, Husted and Rush (1991), Liu and He (1991) , Grilli and Kaminsky (1991), and Whitt (1992) all find evidence * We are ve ry grateful to Robin Cart e r, Bruce Hansen, Alfred Haug, John Knight, Alan S t o ckman and the participants in the Econometrics Wo rkshop at the Unive rsity of We s t e rn Ontario for their valuable comments and suggestions. We also thank an anonymous referee and the editor of the Journal for their useful comments. However, they are not responsible for any errors. Baizhu Chen thanks the University of Saskatchewan for providing a starting research grant.