抄録
For considerations of risk managements of electricity market trading, this paper presents a time series analysis of the day-ahead spot price traded in the PJM power market, taking correlations among hourly price motions into account. The time series was decomposed into the deterministic and stochastic components, the latter of which was analyzed by using the AR-GARCH model both for hourly data (single-factor model) and daily data of hourly prices (multi-factor model), where hourly data were given as a series of hourly prices. The Monte-Carlo price simulation with resulting model parameters well reconstructed the original hourly time series. It was demonstrated to estimate optimum bidding prices in the day-ahead electricity spot market by using this price simulation for an example of its application.
