An Energy Market Modeling Approach for Valuing Real Options

Abstract

Demand and supply in the market for natural gas are both affected by specific features that other commodities do not share. Demand has a strong seasonal component and mean-reverting short-run variation that is typically modeled simply as noise. Supply is provided by storage facilities that hold the gas extracted from the ground until it is consumed. A storage facility operator earns an expected, but risky, return by injecting gas into storage when it is cheap (in summer) and selling it when prices rise (in winter). Standard models of the optimal storage strategy have focused on the dynamics of the spot price, but they suffer from the same type of problems that arise in modeling the interest rate term structure: The model is generally not consistent with observed prices of traded futures contracts; market participants cannot be assumed to know the model parameters exactly; and, in this case, the models typically fail to consider the possibility of enhancing return by short-term speculative trading to offset short-run demand imbalances. One further problem is that future gas price risk does not fall on a specific date but rather applies to gas consumption over a period of time, so traded futures are based on average, not point in time, prices.

In this article, Uhrig-Homburg and Unger develop a model to resolve these difficulties by first constructing an implied continuous term structure of forward prices in the framework of a market model. Applying the model to U.S. natural gas spot and futures prices from 1997–2015, they show that the model fits the data well and that the ability to increase returns by short-run trading should be considered an important component in the overall profitability of a storage facility.