Exchange Rate Forecasting: Evidence from the Emerging Central and Eastern European Economies

There is a vast literature on exchange rate forecasting focusing on developed economies. Since the early 1990s, many developing economies have liberalized their financial accounts, and become an integral part of the international financial system. A series of financial crises experienced by these emerging market economies ed them to switch to some form of a flexible exchange rate regime, coupled with inflation targeting. These developments, in turn, accentuate the need for exchange rate forecasting in such economies. This paper is a first attempt to compile data from the emerging Central and Eastern European (CEE) economies, to evaluate the performance of versions of the monetary model of exchange rate determination, and time series models for forecasting exchange rates. Forecast performance of these models at various horizons are evaluated against that of a random walk, which, overwhelmingly, was found to be the best exchange rate predictor for developed economies in the previous literature. Following Clark and West (2006, 2007) for forecast performance analysis, we report that in short horizons, structural models and time series models outperform the random walk for the six CEE countries in the data set.