Forecasting euro exchange rates: How much does model averaging help?

摘要: We analyze the performance of Bayesian model averaged exchange rate forecasts for euro/US dollar, euro/Japanese yen, euro/Swiss franc and euro/ British pound rates using weights based on out-of-sample predictive likelihood. The paper also presents a simple stratified sampling procedure in spirit Sala i Martin et alia (2004) to obtain accuracy. Our results indicate that accounting explicitly uncertainty when constructing predictions euro leads improvements accuracy as measured by mean square forecast error. While forecasting error combined tends be systematically smaller than individual would have been chosen test sample, random walk cannot beaten significantly terms squared errors. Direction change statistics, other hand, are improved averaging.