# 15-104/VI/DSF95 (2015-08-28)

Author(s)
Lin Zhao, University of Amsterdam, the Netherlands; Sweder van Wijnbergen, University of Amsterdam, the Netherlands
Keywords:
stochastic volatility, Generalised Autoregressive Score modeling, incomplete markets, real options, utility indifference pricing, gas storage, capacity constraints
JEL codes:
C61, C63, G12, G13, Q41

We investigate the relationship between the gas spot market and the price of gas storage capacity. Contrary to the common belief, the auction prices for gas storage are mostly affected by the volatility of current market prices rather than by the winter-summer price differences. This paper provides a numerical solution for pricing storage capacity, by taking investor's activities through the spot market and storage service into account. A bivariate Generalized Autoregressive Score (GAS) model is employed for modeling the dynamics of the day-ahead and month-ahead spot market prices, as well as the time-varying volatilities and correlations. Under an incomplete market setting, our model is able to approximate the realized auction prices. Moreover, one interesting implication is that the implied average risk aversion of investor for a storage contract increases with the volatility of the spot market. This is an intuitive result because storage capacity can serve as an effective hedging product for the spot market, and the demand for this product is high when the market becomes risky: more risk averse investors are participating in the auctions. Moreover, a sensitivity analysis on different injection/withdrawal rates is also included, and particularly, contracts with higher capacity rates are priced at a higher level.