ARIMA Model for Gross Domestic Product (GDP): Evidence from Nigeria
Atanu, Enebi Yahaya *
Department of Statistics, Federal Polytechnic of Oil and Gas, Bonny Island, Rivers State, Nigeria.
Ette, Harrison Etuk
Department of Mathematics/Computer Science, Rivers State University of Science and Technology, Port Harcourt, Nigeria.
Nwuju, Kingdom
Department of Mathematics/Computer Science, Rivers State University of Science and Technology, Port Harcourt, Nigeria.
Nwaoha, William Chimee
School of General Studies, Federal Polytechnic of Oil and Gas, Bonny Island, Rivers State, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
A nation’s GDP is an important index reflecting development in economy and incomes. This paper uses the annual data of Nigeria’s GDP from 1981 to 2019 as the research data. An Augmented Dick Fuller test was used to test for stationarity of the data and was seen to be stationary at the second differencing. ARIMA (1, 2, 1) was identified as an appropriate model using Eviews 11 software after comparing the AIC values. The Ljung-Box test of the Residual satisfied that the model was adequate and was used to forecast the out of sample data. And with a Theil inequality of 0.022008, the model forecasting ability is deemed be a good.
Keywords: GDP, ARIMA modelling, residual analysis, forecasting.