DEA/AR efficiency and profitability of 14 major oil companies in U.S. exploration and production

With a U.S. focus, the efficiency and profitability of 14 integrated oil companies (Majors) were analyzed for the years 1980-1991. The data base represented primarily Arthur Andersen’s oil and gas disclosures. Data Envelopment Analysis (DEA) and Assurance Region (AR) methods were applied. Unique optimal solution pairs of primal slacks and dual multipliers were found for almost all of the inefficient firms; these solutions provided unique projections to the DEA frontier. Separable input and output AR bounds were placed on the modelled prices (multipliers) to proceed from technical toward overall efficiency. Significantly fewer extreme-efficient DMUs and lower levels of efficiency were found in the presence of these bounds than in their absence. there was a 21% level of average inefficiency present in the findings across the 12 years analyzed. Linked input and output bounds provided measures of maximum and minimum profit ratios. Profit potential was found for a high percentage of the Majors; only modest downside risk indications were found. In this linked case, efficiency measurement requires a practical algorithm to compute optimal solutions to the nonlinear problem.