ARIMA-based forecasting of Nigerian crude oil prices (2006–2023): Long-term dynamics, optimal model selection, and policy implications

Abstract

The purpose of this study is to investigate Nigerian monthly crude oil prices from 2006 to 2023 using the Autoregressive Integrated Moving Average (ARIMA) model in order to provide reliable forecasts for fiscal and economic planning. Crude oil remains central to Nigeria’s economy, and its unstable price patterns have significant implications for government budgeting, revenue generation, and long-term policy design. The study employs monthly price data sourced from the National Bureau of Statistics. Preliminary inspection of the series revealed sharp fluctuations without a clear long-term direction. The Augmented Dickey-Fuller test confirmed that the series was non-stationary at the level but became stationary after first differencing. Autocorrelation and partial autocorrelation plots suggested three possible models: ARIMA (0,1,1), ARIMA (1,1,0), and ARIMA (1,1,1). Model performance was compared using AIC, AICc, and BIC values, and further validated with residual diagnostics. The findings indicate that ARIMA (1,1,0) is the best-fitting model, showing that present changes in oil prices are strongly linked to immediate past changes, which reflects the short-term memory property of the oil market. Forecasts from the model point to moderate price stability around US$90–95 per barrel in the near term, though widening confidence intervals highlight rising uncertainty over longer horizons. The practical implication is that accurate short-term forecasts can guide budgetary and fiscal policies in Nigeria and other oil-dependent economies, while underscoring the importance of diversifying revenue sources to reduce vulnerability to oil price shocks.