A1741
Title: Multi-Factor Polynomial DIffusion Models and Inter-Temporal Futures Dynamics in Energy Markets
Authors: Peilun He - Macquarie University (Australia) [presenting]
Pavel Shevchenko - Maquarie University (Australia)
Nino Kordzakhia - Macquarie University (Australia)
Gareth Peters - University of California Santa Barbara (United States)
Abstract: In stochastic multi-factor commodity models, it is often the case that futures prices are explained by two latent state variables which are used to represent the short and long term stochastic dynamic factors for modelling of commodity future term structure dynamics. In this work we focus on development of the family of stochastic models based on polynomial diffusion structures to represent the underlying spot price stochastic dynamics. The polynomial family of diffusion models allows one to incorporate a variety of non-linear higher order effects into a multi-factor stochastic model as well as interaction terms between stochastic variables, volatility dynamics and non-linear trend dynamics. We will compare the predictive performances of the polynomial diffusion models and the Schwartz and Smith two-factor model. The risk neutral prices of the futures contracts in the Schwartz and Smith model are given in the exponential affine closed form. In the polynomial diffusion setting we must extend these estimation methods to be able to estimate the unknown model parameters jointly with hidden state variables under a non-linear model dynamics with measurement equations given in non-closed form. Finally, we illustrate the performance of these models on TOCOM Platts Dubai crude oil futures. The results demonstrate that the polynomial diffusion model extensions can be advantegeous for explaining the inter-temporal futures dynamics more accurately.