Price and inventory dynamics in petroleum product markets

Abstract

Unlike many studies of commodity inventory behavior, this paper estimates a model with endogenous spot and forward prices, inventories, production, and net imports. Our application involves markets for refined petroleum products in the United States. Our model is built around the supply and demand for storage. We estimate the model using Generalized Method of Moments and perform dynamic, simultaneous simulations to estimate the impacts of supply and demand shocks. Supply curves for the industry are inelastic and upward sloping. High inventory levels depress prices. Inventories fall in response to higher sales, consistent with production smoothing. Under higher input prices, refiners reduce their stocks of crude oil but increase their product inventories, consistent with cost smoothing. In some cases, imports of products are more variable than production or inventories.