Date of Original Version



Working Paper

Rights Management

All Rights Reserved

Abstract or Description

A global liquefied natural gas (LNG) market is quickly emerging, with several significant development projects very recently completed or underway; these projects consist of extraction, liquefaction, shipping, regasification, and storage facilities. Exact valuation of the real option to store LNG at the downstream terminal of an LNG value chain is computationally intractable. Thus, we develop a novel and tractable model for the heuristic valuation of this real option. This model uses a shipping model to represent upstream LNG production and shipping to the downstream regasification facility; a reduced form model of the evolution of the spot price in the wholesale natural gas market where regasified LNG is sold; a stochastic dynamic programming model to determine a policy for inventory control at the storage facility and sale into this market; and a final Monte Carlo simulation step to estimate the value of this policy. The basestock type structure that we prove for our model's LNG inventory release policy is central to make the final simulation step computationally efficient; this makes our model practical. We incorporate real and estimated data to quantify the value of the real option to store LNG at a regasification terminal, the dependence of this value on the level of stochastic variability in the shipping model and the type of natural gas price model used, and the relative value of this option for different parties involved in an LNG value chain. We also develop an upper bound, based on sample path optimization, to assess the effectiveness of our heuristic and find that our method is highly accurate. Our model has the potential to be used to value the real option to store other commodities in storage facilities located downstream of a commodity production or transportation stage, or the real option to store the input used in the production of a commodity.


Tepper Working Paper 2006-E99