TITLE:
Forecasting Retail Diesel Prices in Malaysia Using an ARIMA Approach
AUTHORS:
Rahaini Mohd Said, Norhafizah Hussin, Nur Azura Noor Azhuan, Nurul Hajar Mohd Yusoff, Mohamad Faiz Dzulkalnine
KEYWORDS:
Forecasting, Autoregressive Integrated Moving Average (ARIMA), Diesel Price, Autocorrelation Function (ACF), Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF)
JOURNAL NAME:
Journal of Power and Energy Engineering,
Vol.13 No.10,
October
27,
2025
ABSTRACT: Accurate forecasting of diesel prices is essential for Malaysia, where transportation, agriculture, and industry are highly dependent on this fuel. This study investigates the effectiveness of the Autoregressive Integrated Moving Average (ARIMA) model in predicting Malaysian retail diesel prices. Using the Box-Jenkins methodology, the research process involves model identification, estimation, and validation. Data stationarity is assessed through visual inspection and the Augmented Dickey-Fuller (ADF) test, with first-order differencing applied to achieve non-stationarity. Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF) analysis guide the parameter selection, leading to the estimation of several ARIMA (p, 1, q) models. Model adequacy is determined using the Akaike Information Criterion (AIC) with ARIMA (1, 1, 0) identified as the optimal specification. Performance evaluation based on Mean Squared Error (MSE), Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) indicates high predictive accuracy, demonstrating the model’s robustness in capturing diesel price dynamics. The results practically provide value for policymakers and industry stakeholders by supporting evidence-based decision-making, facilitating effective economic planning, and optimizing resource management.