Describing the Phelix forward electric-power market using Bayesian estimators for stochastic volatility models

This paper applies the EMM methodology to build a stochastic volatility (SV) model of the mean and latent volatility for the Phelix future electric power market. The main objective is to find appropriate descriptions emphasising schemes for derivative pricing purposes. A Bayesian estimator is used to estimate a general scientific SV model and is computed adapting MCMC simulation proposed by Chernozhukov and Hong (2003). The approach helps circumvent the computational curse of dimensionality and is substantially superior to conventional derivative based hill climbing optimizers. The paper finds that MCMC estimation of stochastic volatility models are successful describing the energy market¿s two financial contracts. The success suggests that the dynamics of the market contains features known to general SV-models; that is - a preference for simulation based derivative pricing schemes mainly caused by volatility clustering. High market volatility caused by the German market¿s rather low transparency and credibility, the shut-down options of producers, and coal plant threshold price production decisions, induces derivative contracts both important and expensive risk management instruments. Higher market transparency and credibility may therefore suggest a potential for lower hedging costs and increased market liquidity.