Time Series Analysis of Petroleum Products Sales Using Autoregressive Model
O. Ihekuna, Stephen *
Department of Statistics, Imo State University, PMB 2000, Owerri, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
Nigeria, Africa’s largest oil producer and a key member of the Organization of Petroleum Exporting Countries (OPEC), has an economy deeply rooted in petroleum production and exports. Since the discovery of oil at Oloibiri in 1956, the sector has driven national revenue, foreign exchange earnings, and infrastructural growth, while also contributing to economic volatility, environmental degradation, and regional unrest. This study examines the historical trajectory of oil exploration and production in Nigeria and applies a statistical time series approach to forecast petroleum product sales by the Nigerian National Petroleum Corporation (NNPC) as published in NNPC Annual Statistical Bulletin. Quarterly sales data from 2015- 2021 were analyzed using the Autocorrelation Function (ACF) and Autoregressive (AR) models to assess the dependency structure within the data and predict future sales trends. The results show significant short-term correlation in the series, with the partial autocorrelation indicating an AR(1) process as the best-fit model. The estimated autoregressive coefficient (ϕ₁ = 0.29) was used to forecast sales for the 2022 quarters, revealing a gradual decline compared to previous years—reflecting market instability and supply constraints. The findings highlight the persistent sensitivity of Nigeria’s petroleum sales to internal inefficiencies and external price shocks. This research underscores the importance of robust forecasting and diversification strategies in managing Nigeria’s oil-dependent economy amid global energy transitions.
Keywords: Autocovariance, autocorrelation, forecasted value, fluctuations, auto regressive process, discrete-time process