A Bootstrap Granger Causality Test from Exchange Rates to Fundamentals

This study is in response to claims of a Granger causality relationship from exchange rates to fundamentals. The alleged relationship is based on an asymptotic test performed in a previous study, which has been taken as evidence to support the present-value model for exchange rates. We adopted a bootstrap method to reassess the evidence supporting Granger causality, the results of which contradict the findings of the previous work. A further Monte Carlo experiment suggests that the causality test implemented in the previous study tended to spuriously reject null hypotheses, implying that the existing evidence for the present value model of exchange rates is very weak.