Modeling of Nigerian Foreign Exchange (Naira/1.0$) Using Disbursement Data (1981-2010)

Abstract

The study aimed at developing a model that best describe foreign exchange disbursement data. The data was obtained from CBN statistically Bulletin 2012. Time series technique of ARIMA was used to model the data. The Time series plot of the data shows that the data has a trend which it is not stationary. ADF and KPSS were employed to test for stationarity of the data which revealed that data is non-stationary at levels but became stationary after the first difference. The residual ACF and PACF suggest mixed process (ARMA) is appropriate. Several models were developed and tested and of all the models developed ARIMA (2,1,2) was the best fit in terms of coefficients significant and information criterion (Akaike, Hann-Quin and Schwartz). The model was then recommended for the forecast disbursement of Nigerians foreign exchange which will help policy makers in the economic sector to foresee the likely disbursement of the Foreign Exchange