摘要
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、异方差、波动集聚、尖峰厚尾及其与负荷相关性的GARCH-VaR计算模型,以正态分布、t分布、有偏t分布和广义误差分布(GED)为例,分析了残差的概率分布设定对VaR估计精度的影响.对PJM电力市场历史数据的分析表明:低置信水平下GED分布的估计结果较为精确,高置信水平下t分布的估计结果更加有效,考虑残差的偏度可在一定程度上改善VaR的估计精度.
： How to effectively evaluate price of volatility risk is the basis of risk management in electricity market. With comprehensive analysis of the basic features and influencing factors of electricity prices, a GARCH model of computing value-at-risk （VaR） is proposed, in which the seasonalities, heteroscedasticities, kurtosises and heavy-tails, volatili- ty-clustering and its relationship with system loads are jointly considered. The impacts of probability distribution as- sumption for four innovation＇ s distributions, normal, student-t, skewed student-t and general error distribution （GED）, on VaR estimation are analyzed. The numerical example based on the historical data of the PJM market shows that the model with GED distribution performs better in predicting one-period-ahead VaR under lower confidence level, while the one with student-t distribution performs better under higher confidence level. Moreover, when taking the innovation＇ s skewness into account, the estimated accuracy can be improved to some extent. These results present several potential implications for electricity markets risk quantifications and hedging strategies.
出处
《海南师范大学学报（自然科学版）》
CAS
2012年第1期36-40,共5页
Journal of Hainan Normal University（Natural Science）
基金
海南省高等院校科研基金项目(Hjkj2011-48)
关键词
风险价值
GARCH模型
概率分布设定
电力市场
value-at-risk
GARCH model
probability distribution assumption
electricity market
