Stochastic multifactor modeling of spot electricity prices

摘要

In this paper, a stochastic multifactor model is proposed for modeling of the daily spot market electricity prices. Stochastic part of the model is composed of three jump processes and a Brownian motion where two of the jump processes are assumed to be mean reverting with different mean reversion rates. The multistep algorithm proposed for model estimation utilizes an iterative threshold function (constructed on GARCH(1,1) volatility estimates) in separation of the jumps. The factor model is applied to Turkish day ahead electricity market. In order to evaluate the performance of the proposed multifactor model, estimation results are also compared to the results acquired by application of mean reverting jump diffusion model of Cartea and Figueroa (2005) and Markov regime switching model of Janczura and Weron (2010) to the same data set.