An Enhanced Neural-based Methodology for Combining Linear and Nonlinear Models for Exchange Rate Forecasting

difficult task in international finance. Various linear and nonlinear theoretical models have been developed in the literature; however, few are more successful in out-of-sample forecasting than a simple random walk model. Artificial neural networks (ANNs) are flexible computing frameworks and universal approximators that can be applied to a wide range of forecasting problems with a high degree of accuracy. Recent applications of neural networks in the exchange rate forecasting have yielded mixed results, and hence; it is not wise to apply them blindly to any type of data. This is the reason that hybrid methodologies incorporating linear and nonlinear models have been proposed in the literature. In this paper, a new methodology is proposed in order to combine the linear auto-regressive integrated moving average (ARIMA) models with nonlinear artificial neural networks in order to overcome the limitations of traditional hybrid methodologies and yield more general and more accurate hybrid models. Empirical results of exchange rate forecasting indicate that the proposed methodology can be a more effective way in order to combine linear and nonlinear models together than traditional hybrid methodologies.

Keywords:

Time Series Forecasting , Hybrid Methodologies , Artificial Neural Networks (ANNs) , Auto-Regressive Integrated Moving Average (ARIMA) , Exchange Rate