Valuations vendor Pricing Partners has upgraded its existing commodity module with the release of the new Gabillon and Gibson Schwartz model. The vendor claims it now able to better calibrate and price a wide range of commodity related derivatives, in spite of market volatility.
The Gabillon model and the Gibson Schwartz model are market standard two-factor models specific to the commodity market, able to capture future curve movement. These two models consider that the convenience yield itself follows a stochastic process correlated to the spot process. This stochastic convenience yield provides a much better forward dynamic compared to Black Scholes, local volatility or more generally any other deterministic convenience yield model as they take into account the risk of futures curve reshaping changing from backwardation to contango situation and vice versa.
For the two models, Pricing Partners implemented not only analytical formulas for futures and vanilla options allowing fast analytical calibration but also numerical compatibility with its generic Monte Carlo and PDE engine. Key differences in the two models are that the Gibson Schwartz model, widely used in the commodity market, assumes stochastic convenience yield per se, while the Gabillon model introduces indirectly stochastic convenience yield by means of a ‘long term price’ concept that may be very natural in certain commodity market with strong mean reversion property.
Zaizhi Wang, financial engineer at Pricing Partners, comments: “The improved module taking into consideration both Gibson Schwartz and Gabillon models enables us a better modelling of the dynamic of the convenience yield. We have seen unprecedented level of curve reshaping like for instance on the crude oil market where the spot price went up to more than 140 dollars a barrel up to a current 70 dollar, while the long term futures did not move as fast reshaping considerably the futures curve. It made sense for us to have proper models to account for fast curve reshaping to price commodity accurate. This new development is obviously much more accurate on products sensitive to the stochasticity of the convenience yield, and already concerns simple products like options on futures.”
