Evaluating the Forecast Accuracy of Policymakers, Private Banks and Exchange Rate Forecasting Models

Abstract

This paper compares the out-of-sample forecast accuracy of policymakers, private banks and three classes of exchange rate models in predicting the yearly Norwegian kroner/Euro, I-44 and KKI exchange rate. The three classes are time series models, fundamental models, and general models (simple models that combine various variables that in the literature have found to hold predictive power on exchange rates). My findings support the evidence of Meese and Rogoff (1983) that the naïve random walk model is difficult to outperform out-of-sample. Further, I find that Policymakers and Nordic Banks are reliable forecasters producing stable and precise forecasts. Finally, I find evidence for the stable and accurate forecasting power of the Taylor Rule and the output gap differential between Norway and the Euro-zone.