摘要
气温预测一直是气温衍生品精确定价的难点及关键点.为有效捕捉气温变化的缓慢衰减过程及其波动率的非对称动态特征,本文在ARMA-EGARCH模型的基础上构建一个PCA-ARMA-EGARCH气温预测模型.该模型融入众多与每日平均气温有联动作用的其他气象因子信息,能进一步提高模型拟合及预测精度.我们以中国杭州市为例进行实证分析,比较气温残差服从不同分布(正态分布、t分布与GED分布)时对预测模型精度的影响.结果表明PCA-ARMA-EGARCH模型与已有模型相比有更好的拟合及预测效果.杭州市每日平均气温缓慢衰减特征明显,气温波动率具有非对称性,即高温热浪对气温冲击的影响程度远大于强冷空气.气温残差服从GED分布,具有尖峰、厚尾特征,已有研究中正态分布假设可能低估风险的发生.论文结果可为杭州市高温风险管理提供依据.
Forecasting of temperature is known to be difficult and key to temperature derivatives’pricing.In this paper,based on the ARMA-EGARCH model,a PCA-ARMA-EGARCH temperature forecasting model is established to catch the slowly decay process and the asymmetry of volatility of the temperature.The model involes much information contained in other meteorological factors related to daily average temperatures so that the fitting and forecasting accuracy is improved.We take Hangzhou city as an example for empirical analysis.The fitting effects of the models under different distributions of residuals(the normal distribution,student t-distribution and GED distribution)are compared with each other.The results show that the PCA-ARMA-EGARCH model has a better fitting and forecasting effect than the existing models.The daily average temperature of Hangzhou decays slowly remarkably and appears asymmetry of volatility,which means that the influence of heat waves on temperature’s shock is much bigger than that of the strong cold air.It is found that residuals of the temperature of Hangzhou obeys the GED distribution and has the peak and thick tail characteristics,which suggests that the normal distribution assumption of residuals may underestimate the occurrence of risk.Our results can certainly provide supports for the high temperature risk management in Hangzhou.
作者
崔海蓉
周颖
鲁训法
Cui Hairong;Zhou Ying;Lu Xunfa(Nanjing University of Information Science&Technology,School of Management Science and Engineering,Nanjing 210044,China)
出处
《数学理论与应用》
2019年第3期90-103,共14页
Mathematical Theory and Applications
基金
国家自然科学基金资助项目(71701104)
教育部人文社会科学资助项目(17YJC790102)
江苏高校品牌专业建设工程资助项目(PPZY2015A072)
江苏省高校哲学社会科学基金项目(2019SJA0153)
关键词
气温衍生品
波动率
非对称性
GED分布
主成分分析
Temperature Derivatives
Volatility
Asymmetry
GED Distribution
Principal Component Analysis
