An empirical investigation of the long-run behavior of real exchange rates comment on the huizinga paper

Abstract

Like the curate's egg, Huizinga's paper is good in parts. Sections II and III provide interesting univariate characterizations of real exchangerate behavior. The conclusions about the presence of mean-reverting behavior for the real exchange rate and about important departures from random-walk behavior are, however, directly contradicted by the evidence presented by the author. Section IV which relates the univariate time series evidence on the real exchange rate to the overshooting hypothesis is misconceived. Sections V and VI approach the real exchange rate through interesting multivariate data descriptions. The absence of any explicit theoretical framework for the joint determination of this vector of variables means, of course, that no interpretation can be attempted of any empirical regularities that may be established.