Author(s) :

Rohan Jha, Rishabh Jha.

Volume/Issue :

Abstract :

Energy demand will increase due to the increasing population, improved standard of living, and rising artificial intelligence applications. The existing energy sources may not be sufficient, and we need multiple sources of energy in the near future to meet the increased demand. The oil & gas energy sector will face several challenges such as stricter regulation, the rise of renewable energy, and aging assets. The industry, therefore, needs sophisticated tools to forecast the price and cost more accurately than ever for reliable and sustainable profits with reduced downside risks. In this paper, we develop statistical models for a profitable refinery to forecast the key cost driver as crude oil price, and the key revenue driver as gasoline/diesel prices, 6 months in advance. We first identify statistically significant variables and then refine further using collinearity to identify key variables. Our analysis indicates that both oil and gasoline prices depend on the Dow Jones industrial average, WTI oil price, average miles used per gallon, average well production per day, number of refineries, refinery capacity, and rig count. These variables are statistically significant with p-values less than 0.05. We tested the 6-month future price and found that the model can predict the oil price within 5% variability and the gasoline price within 7%. For the oil price variation, the Dow Jones industrial average accounts for ~24%, the number of refineries for ~21%, average well production per day for ~20%, and the remaining variation is attributed to other variables. For the gasoline price variation, oil price contributes ~40%, the Dow Jones industrial average covers ~25%, and the remaining variables account for the remaining ~35%. The Dow Jones industrial average impacts both the cost and revenue aspects of a refinery business. Thus, we recommend contracting crude oil and gasoline using an index tied to the broader economy to reduce the downside risk and improve profitability. These results could enable the refinery management team to seek long-term contracts to lock in high-margin crude oil supply and favorable gasoline price contracts with customers when a decline in price is forecasted to improve the profit margin. This methodology and strategy are applicable beyond the refinery business to any cyclical business to maintain margin, particularly in a down cycle.