This paper aims to reveal how the refining industry's inputs in Saudi Arabia affect its output and to forecast refining industry dynamics. The variables used in this paper are the refined petroleum products representing the dependent variable, with natural gas liquids, crude oil, labor, and capital acting as explanatory variables covering the period 1990-2020. The long run cointegration of the variables was observed. An error correction model utilizing the Cobb-Douglas production function framework was performed. Furthermore, this study applied the vector autoregressive model (VAR) and its diagnosis tests, including forecast-error variance decomposition (FEVD) and impulse response functions (IRFs). The results indicate that natural gas liquids and crude oil have a significant influence on the refining industry's output. Although capital and labor are significant determinants of output, they do not contribute significantly to output creation in the refining industry. This might be related to some parts of the capital and human resources being directed toward supporting activities, such as administration, technical support, maintenance, transportation, logistics and assigning third-party contractors to perform the main duties related to the production process. Additionally, the petroleum refining industry requires substantial capital resources for construction and maintenance. Thus, the actual measurement of capital input's influence on output was observed in the long run. The results reveal that the refining industry's variation is influenced by both its own characteristics and that natural gas liquid, crude oil, capital, and labor factors have a significant impact on the accuracy of industry forecasts. This study concludes that Saudi Arabia's petroleum refining industry operates under decreasing returns to scale, while the shocks in the refining industry are influenced and caused by external factors.
Keywords: Cobb–Douglas production function; Economies of scale; Refinery; VAR model.
© 2024 The Authors.