Evaluation of Models for Forecasting Daily Foreign Exchange Rates Between Nigerian Naira and US Dollar Amidst Volatility

In her quest to put the nations’ foreign exchange policy in line with global practice, the Central Bank of Nigeria (CBN) in the mid of 2016 made a paradigm shift by unveiling flexible foreign exchange policy driven purely by market forces. Consequently, forecasting models reported during pegging policy era may be deeming obsolete. This paper proposed a prediction model for future daily selling exchange rate between Nigerian Naira and United States Dollar in the interbank market amidst volatility using daily rates made available by the Central Bank of Nigeria over the periods July, 2016 to October, 2017. We applied the Box-Jenkins methodology to the Naira – US Dollar exchange rate series in order to build an adequate and well specified prediction model. On the first phase of the study, we checked the stationarity of the series and based on Augmented Dickey-Fuller test and correlogram, the series is stationary at level. The findings revealed that the future daily exchange rate of Naira per US Dollar depends on its past rates and associated innovations. This is evident from the fact that turned out parsimonious model for forecasting the parity between the two currencies on the basis of Bayesian Information Criterion. The said model yielded error of forecast (MAPE) within ten percent.