Mean reversion in equilibrium real exchange rates

This paper reports evidence that an equilibrium model of real exchange rate determination with time non-separable preferences can generate mean reversion in six real exchange rates. Using the generalized method of moments. I obtain plausible parameter estimates and cannot reject the model by the over-identifying restrictions test. Monte-Carlo experiments cannot reject the null hypothesis that the variance rations estimated with the data are drawn from the empirical distribution generated by the model.[F31].