Modeling and forecasting the supply of oil and gas

Abstract

This paper surveys the literature in empirical oil and gas supply modeling. The models fall into two broad categories: geologic/engineering and econometric. Two types of geologic/engineering models are surveyed — play analysis, or simulation, models and discovery process models. A third category of supply models, ‘hybrids’, which contain features of both econometric and discovery process models are also discussed. Particular attention is paid to whether or not the models have linkages between a dynamic model of producer optimizing behavior and the factors governing supply of the resource; whether or not expectations of future prices, costs, and other stochastic variables are incorporated; whether the physical characteristics of non-renewable resources are captured; and how well the models perform. The paper concludes that the best path for future research efforts is a hybrid approach where the econometric component is derived from a stochastic dynamic optimization model of exploration behavior.