Forecasting Malaysian Foreign Exchange Rate Using Artificial Neural Network and ARIMA Time Series

In view of today's global economy, countries around the world are getting more connected and linked to each other via trade, technically known as import and export activities. To get ahead in the world economy, a country must actively participate in these activities. Otherwise, a country which adopt a closed door policy will be left behind due to this intra connectivity. To allow a monetary transaction to take place during a trading process in a systematic and coherent manner, exchange rate plays a prominent role. However, due to the uncertainty and volatility in the world's economy, the participants of such trading activities are vulnerable and exposed to risk. Therefore, the ability to accurately forecast the exchange rate offers a solution to mitigate this risk. The main purpose of this paper is to forecast the exchange rate for US Dollar expressed in terms of Malaysian Ringgit. The exchange rate forecasting is conducted by using two methods; Artificial Neural Network (ANN) and Autoregressive Integrated Moving Average (ARIMA) time series. Feed forward neural network has been chosen as neural network's method to forecast the exchange rate because this method has been proven to be inherently stable. On the other hand, ARIMA (0, 1, 1) is chosen as the best model for the time series based on Box Jenkins approach. After comparing the forecasting method using ANN and ARIMA (0, 1, 1) time series, we find that feed forward neural network exhibit a smaller mean square error and root mean square error as compared to ARIMA (0, 1, 1). This result suggest that in this research, ANN approach using the feed forward neural network is a more suitable forecasting method to predict the exchange rate for US Dollar expressed in terms of Malaysian Ringgit compared to ARIMA (0,1,1) time series model.