Comparison of extended mean-reversion and time series models for electricity spot price simulation considering negative prices

摘要: Abstract This paper evaluates different financial price and time series models, such as mean reversion, autoregressive moving average (ARMA), integrated ARMA (ARIMA) general conditional heteroscedasticity (GARCH) process, usually applied for electricity simulations. However, these models are developed to describe the stochastic behaviour of prices, they extended by a separate data treatment deterministic components (trend, daily, weekly annual cycles) spot prices. Furthermore jumps considered implemented within regime-switching model. Since 2008 market design allows negative prices at European Energy Exchange, which also occurred several hours in last years. Up now, only few approaches exist, able capture presents new approach incorporating The evaluation presented points out that reversion deliver lowest root square error between simulated historical gained from Exchange. These posses lower errors than GARCH models. Hence, more suitable simulate well-fitting paths. it is shown daily structure curves better captured applying or ARIMA processes instead mean-reversion Another important outcome consideration via proposed lead significant improvement simulation.