Preprint B36/2010
Calibration of the Schwartz-Smith Model for Commodity Prices
Ana Luiza AbrĂ£o
Keywords: Calibration | Commodities | Schwartz and Smith
This work develops a two-stage calibration methodology for commodity prices, using nonlinear least squares and maximum likelihood. The theoretical model is based on Schwartz and Smith (2000), which decomposes the spot price of energy commodities in two components: a short term and a long term one. The short term component is modeled by a mean reverting process while the long term one is modeled by a geometric brownian motion. The calibration procedure is validated by numerical tests. We create a database with futures prices obtained from known parameters and show that it is possible to recover them by the methodology proposed. The robustness of the method is proven by disturbing the initial parameters and showing that the true parameter values are recovered precisely. The methodology is also applied to real data from the natural gas market, obtaining equally satisfactory results.