论文部分内容阅读
有效地评估电力市场价格波动风险是电力市场风险管理的基础。在对电价的基本特征及其影响因素综合分析的基础上,建立了考虑电价多季节性、异方差性、波动集聚性、尖峰厚尾特征及其与负荷相关性的GARCH-VaR计算模型,分析了残差的分布形式设定及分布参数的时变性对VaR估计精度的影响。对PJM电力市场历史数据的分析表明:计及偏度和峰度时变特征的基于正态分布密度函数的Gram-Charlier展开的GARCH模型的VaR估计结果准确有效,而正态分布GARCH模型在高置信水平下低估了电力市场风险,t分布GARCH模型在低置信水平下高估了电力市场风险。该结果对于电力市场参与者准确地测度价格波动风险和制定有效的风险规避策略具有重要的指导意义。
Effectively assessing the risk of price fluctuation in the electricity market is the basis of power market risk management. Based on the comprehensive analysis of the basic characteristics of electricity price and its influencing factors, a GARCH-VaR calculation model considering the multi-seasonal, heteroscedastic, fluctuating concentration, peak-thick tail characteristics and their correlation with load is established and analyzed. The influence of the distribution of residuals and the time variability of distribution parameters on the accuracy of VaR estimation. The analysis of the historical data of PJM power market shows that the VaR estimation of GARCH model based on Gram-Charlier based on the normal distribution density function which accounts for the skewness and kurtosis is accurate and effective. However, Under the confidence level, the electricity market risk is underestimated. The t-distribution GARCH model overestimates the electricity market risk under the low confidence level. This result is of great guiding significance for the power market participants to accurately measure the risk of price fluctuation and to formulate an effective risk aversion strategy.