Forecasting exchange rate of Solomon Islands dollar against Euro using artificial neural network

With continual changes made and reviews of the exchange rate regime of Solomon Islands it is imperative that a proper forecasting modelling tool is established. The use of neural network models in exchange rate forecasting has received much attention in recent research. In this paper we propose an artificial neural network (ANN) model for forecasting exchange rates of the Solomon Islands dollar (SBD) against Euro (EUR). We use daily exchange rate data during the period of January 5, 1998 to June 30, 2014. The proposed model is compared with a naive method as a benchmarked method. Further, it is compared with single exponential smoothing; double exponential smoothing with trend; and Holt-Winter multiplicative and additive seasonal and multiple linear regression models. The performance of the models was measured by using various error functions such as root mean square error, mean absolute error, and mean absolute percentage error. The validation tests of the models were also carried out using different goodness of fit measures such as R-square, bias and tracking signal. The empirical result reveals that the proposed model is an efficient tool for forecasting SBD against Euro more accurately than are regression and time series models.