A Suitable Model for the Forecast of Exchange Rate in Nigeria (Nigerian Naira versus US Dollar)

This project presents an empirical study of modeling and forecasting time series data of the official Exchange rate of Nigeria Naira to the US Dollar. The Box-Jenkins ARIMA and ARMA methodology were used for forecasting the monthly data collected from January 2000 to December 2012. Result analysis revealed that the series became stationary at first difference. The diagnostic checking has showed that ARIMA (1, 1, 2) and ARMA (1, 1) are appropriate or optimal model based on the Akaike-s information criterion (AIC), Shcwarzt information criterion (SIC), and Hannan Quinn criterion (HQC). The performance of the models (ARIMA and ARMA model) for both in-sample and out-of-sample also show that ARIMA (1, 1, 2) has Minimum Mean Error (ME), Mean Squared Error (MSE), Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), which indicates that ARIMA (1, 1, 2) model is the best or optimal model for the period forecasted. These forecasts would be helpful for policy makers in Nigeria to foresee ahead of time the Exchange rate, and the possible fluctuation intervals of Nigerian Naira to the US Dollar for future forecasted.