Pricing electricity day-ahead cap futures with multifactor skew-t densities
针对日前电力市场缺乏对冲工具的问题，提出用多因子GAMLSS模型结合偏斜t分布预测日前电价分布，为上限期货定价，并证明其优于正态分布等低阶模型。
Short-term risk management is becoming increasingly significant in power trading as the intermittent renewable generators introduce more weather risk into the price formation dynamics. There is a vacuum in hedging instruments at the day-ahead stage to protect retailers in particular from such volatility and price spikes. Motivated by this requirement, this paper analyses a flexible hedging product, day-ahead cap futures. For pricing this product, we parametrically predict the probability distribution of day-ahead prices using the multifactor Generalized Additive Model for Location, Scale and Shape (GAMLSS) based upon the skew-t distribution with weather forecasts and calendar information as explanatory variables. In particular, we reveal that this higher-order moment model is superior to several lower-order models such as the normal distribution in all the following three aspects: fairness as pricing method, underwriting risk of the risk taker, and the variance reduction effect of the risk hedger.