Exchange rate prediction using fuzzy system neural network approach

Forecasting exchange rate is very important for many international agents e.g. investors, money managers, investment banks, funds makers and others. We forecasted the daily Bangladeshi and Canadian exchange rate series for the period of October 1996 to January 2013. Paying attention with recently developed econometric noises, we considered widely-used non-linear forecasting model namely the fuzzy extension of artificial neural network model and compared results with the Markov switching autoregressive model. Our target is to investigate whether selected model can serve as a useful forecasting model to find volatile and non-linear behaviors of the considered exchange rate series. By most commonly used statistical measures: Root mean square error and correlation coefficient we found that fuzzy extension of the artificial neural network model is a superior predictor than the other selected predictor for the Bangladeshi series and the reverse observed for the Canadian series. The findings will have implications for many kinds of businessmen and multinational organizations. We believe findings of this paper will be helpful to make a wide range of policies for multinational companies who are involved with various international business activities.