Econometrics modelling and forecasting Nigerian crude oil prices

Moving Average (MA) method adopted by the Nigerian government in projecting crude oil price benchmark has proved to be inefficient due to observed large discrepancy between domestic oil price benchmarks and the actual international crude oil prices. Therefore, the main objective of this research is to identify the best econometrics model for forecasting Nigerian crude oil prices. The research employs time series models using monthly bonny light crude oil prices data for the period of April 1986 to December, 2015. The study further used annual Bonny light crude oil prices from 1979 to 2015. The unit root tests results using ADF, PP, KPSS and correlogram indicated that both DMBLP and DABLP possess unit root, but become stationary series at first difference transformation. The best model identified for DMBLP is ARIMA (6, 1, 6). ARIMA (2, 1, 6) is found to be the suitable and parsimonious model for DABLP. Symmetric GARCH (1, 1)-GED is also found to be the parsimonious model and perform better forecast than other GARCH family model.On the other hand GARCH model performs better than the ARIMA model in term of forecasting ability of the monthly bonny light oil prices.