Factor models in the German electricity market: Stylized facts, seasonality, and calibration
系统梳理了算术因子模型在电力市场中的应用，比较了不同季节性函数对德国/奥地利日前现货价格去季节化的效果，并提出了基于ARMA估计均值回复速度的新方法。
The class of arithmetic factor models is flexible enough to model all stylized facts occurring in electricity markets, including negative prices, while still yielding tractable derivative prices. In this paper we conduct a thorough review of the requirements and possibilities of factor models. We compare different seasonality functions and study their power to deseasonalise day-ahead spot prices from the EPEX Germany/Austria market. Furthermore, we introduce an alternative method to estimate mean reversion speed based on ARMA time series and a method to evaluate the distributional fit of the model to realised market prices, which we apply to two non-Gaussian estimated models.