Predicting exchange rates using a novel "cointegration based neuro-fuzzy system"

The present study focuses upon the applications of currently available intelligence techniques to forecast exchange rates in short and long horizons. The predictability of exchange rate returns is investigated through the use of a novel cointegration-based neuro-fuzzy system, which is a combination of a cointegration technique; a Fuzzy Inference System; and Artificial Neural Networks. The Relative Price Monetary Model for exchange rate determination is used to determine the inputs, consisting of macroeconomic variables and the type of interactions amongst the variables, in order to develop the system. Considering exchange rate returns of three ASEAN countries (Malaysia, the Philippines and Singapore), our results reveal that the cointegration-based neuro-fuzzy system model consistently outperforms the Vector Error Correction Model by successfully forecasting exchange rate monthly returns with a high level of accuracy.