This paper reconsiders several recently published but controversial results about the behaviour of exchange rates. In particular, it explores finite-sample problems in the application of cointegration tests and shows how these may have affected the conclusions of recent research. It also demonstrates how simple simulation methods may be used to check the robustness of cointegration tests in particular applied settings, and provides information on the potential sources of size distortion in these tests.

Three case studies are presented. The first is the literature on cointegration and prediction of nominal spot exchange rates spawned by Baillie and Bollerslev (1989). The second is work on the long-run validity of the monetary model of exchange rate determination, particularly the contributions of MacDonald and Taylor (1993; 1994a). The final case study looks at the evidence presented by Kasa (1992) on common stochastic trends in the international stock market. Our results suggest that Baillie and Bollerslev's results are unaffected by finite-sample problems, but that the opposite is true for the other two case studies.