Abstract

Author:
Akunna, O. A., Aronu, C. O., Ugwu, N. D.

Doi: 10.26480/egnes.02.2025.39.46

This is an open access article distributed under the Creative Commons Attribution License CC BY 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited

This study develops a robust predictive framework to model and forecast Premium Motor Spirit (PMS) pump prices in Nigeria using monthly macroeconomic data from January 2006 to January 2024. Recognizing the limitations of traditional econometric models in capturing complex and nonlinear economic relationships, the research integrates a Grey Box (GB) model with a Random Forest (RF) machine learning model, culminating in a GreyForest Hybrid Regression Model (GFHRM). The study leverages five key macroeconomic indicators crude oil price, crude oil production, crude oil export, exchange rate (NGN/USD), and inflation rate—alongside their lagged and lead transformations to capture temporal dependencies in PMS pricing. The RF model demonstrated superior predictive accuracy (R² = 0.982, RMSE = 6.061), outperforming the GB model (R² = 0.7285, RMSE = 78.030). However, the hybrid GFHRM achieved the highest explanatory power (R² = 0.989) and the lowest mean squared error (MSE = 86.428), validating the synergy of combining theory-based and data-driven approaches. Sensitivity analyses revealed that exchange rate fluctuations and inflation particularly their lagged and lead components are the dominant drivers of PMS prices, while global crude oil prices showed limited direct influence. These findings highlight the need for proactive, forward-looking economic policies. The study recommends exchange rate stabilization, inflation targeting, and investment in local refining infrastructure to reduce import reliance. The GFHRM provides policymakers with a robust and interpretable tool for managing fuel price volatility and planning subsidy reforms. This study contributes significantly to the literature on hybrid modelling and macroeconomic forecasting in petroleum-dependent economies.