An improved principal components methodology for the valuation of natural gas storage contracts.

摘要

Given the increasing demand for natural gas, over the past fifteen years, a number of industry participants have attempted to value their storage contracts. A critical component of such valuation is the forecast of the term structure of natural gas forward prices. Leading industry firms have utilized Principal Component Analysis in conjunction with simulation to develop future values for natural gas forward contracts. In their approaches, they accept an evolution of the covariance matrix of forward returns generated by historical data on an ad-hoc monthly basis.; In this work, we developed an improved procedure for applying the evolution of the historical covariance matrix. Our methodology involves the determination of a set of weights and the ensuing application of those weights on the results of Monte Carlo simulations under a variable-period switching of the VCV matrices used. We call our methodology Seasonal Principal Components-based Post-Weighted Monte Carlo Simulation or, SPCA-PWMC. When applied to the simulation of natural gas forwards and spreads for out-of-sample time horizons, our methodology results in a significantly improved approximation of the actual values of such forwards and spreads.; Additionally, we developed procedures in Mathematica to determine the operational feasibility of any combination of forward contracts, as used in storage valuation. Finally, we developed a procedure, also in Mathematica, to calculate the present values of all operationally feasible combinations of spreads and European call options.