Forecasting Spot Exchange Rates Using Different Forecasting Techniques

Abstract

Forecasting of exchange rates has always been a popular research topic among scholars. The thesis tried to forecast the future change in spot rate using the difference between forward rate and the current spot rate. Forward rates of three-time horizons were used as the time horizon’s included 1 month, 3 month and 6 months. The study employed two exchange rates and both the exchange rates were forecasted using three forecasting methodologies to determine the best forecasting tool. Several prior studies have produced different results and the thesis explores possible to improve the forecast of structural exchange rate models, by explicitly accounting for the time variability of the parameters when estimating these models. The study was not able to show the impact of time varying parameter characteristic on forecasting as the results indicated that both the exchange rate series did not possess any time-varying characteristics.