• preprint
    • Published in RePEc
Abstract
A Markov-switching model is fit for eighteen exchange rates at quarterly and monthly frequencies. This model fits well in-sample at the quarterly frequency for many exchange rates. By the mean-squared-error or mean-absolute-error criterion. the Markov model does not generate superior forecasts at a random walk or at the forward rate. There appears to be some evidence that the forecast of the Markov model are superior at predicting the direction of change of the exchange rate.
All Related Versions

This publication has 0 references indexed in Scilit: