A Stylized Model of the Oil Squeeze

Abstract

The paper proposes a practical alternative to a traditional theory of storage, which is difficult to apply to oil markets due to structural differences, such as low-price elasticities of demand and supply, and the dominant role of oil as an investment asset. We model futures time spreads as a derivative of the observable state variable, oil inventories. Inventories, normalized by the storage capacity, are assumed to follow a mean-reverting stochastic process. The boundary condition at the futures expiry is specified by a pair of Dirac delta functions which correspond to the events of default by the futures trader when either oil inventory or the storage capacity are no longer available. The model is calibrated to data for Cushing, Oklahoma, the delivery location for WTI futures, and is used to analyze the market dynamics that led to negative oil prices. We argue that while the limitation of storage capacity was the initial catalyst leading to the event, the actual episode of negative prices was the result of the financial squeeze in the futures market.