Abstract
This paper presents a single-factor model to describe the fluctuation of the electricity futures price for its trading risk management. An autoregressive moving-average model (ARMA(2, 1) process) was used to express the stochastic process of the price, instead of a conventionally used Malkov process such as the AR(1) process, where the ARMA(2, 1) process becomes a hybrid of short- and long-term mean-reversion processes in the continuous time model. This model was applied to the analysis of the price of the electricity futures (the PJM Monthly) traded at the New York Mercantile Exchange (NYMEX). The result showed that the model well explained the term structure of the volatility of futures price with respect to the time to maturity, which is important for estimating its trading risk. The expected long-term fixed electricity price and its confidence interval were also estimated by using the obtained model function of the forward curve and its parameters.
