天气衍生品气温预测模型探索研究  

Research on weather derivative temperature prediction model

作　　者：郑涛涛 韩笑笑 陶祥兴[1] 季彦颋 ZHENG Taotao;HAN Xiaoxiao;TAO Xiangxing;JI Yanting(School of Science,Zhejiang University of Science and Technology,Hangzhou 310023,China)

机构地区：[1]浙江科技学院理学院,浙江杭州310023

出　　处：《浙江工业大学学报》2023年第5期553-558,共6页Journal of Zhejiang University of Technology

基　　金：国家自然科学基金资助项目(11626213);浙江省自然科学基金资助项目(LQ17A010002)。

摘　　要：针对随机误差项不满足正态性假设且低精度模型会导致期权定价错误被放大的问题,首先在OU过程描述气温变化的基础上采用时变的均值回复速度修正气温波动项的变化;然后采用蒙特卡洛方法,通过与时间序列和随机过程方法构建的多个模型作对比,评估模型的有效性;最后采用风险中性定价法对期权合约定价。结果显示:修正后的模型不仅在杭州市气温数据集上的表现优于其他模型,而且随机误差项通过了正态性检验,同时也证实了精度越高的气温预测模型越能有效改善期权定价错误被放大的问题,从而减小经济损失。In view of the problem that the random error term does not satisfy the normality hypothesis and the low precision model will lead to the magnification of the option pricing error,based on the OU process describing the temperature change,the continuous-time mean reversion parameter is used to correct the change of temperature fluctuation.Furthermore,Monte Carlo simulation method is used to evaluate the validity of the model by comparing with mutiple models based on times-series approaches and stochastic process.Finally,the risk-neutral pricing method is adopted to price the option contract.Result shows that the revised model performs better on the Hangzhou temperature data set than the other models,and the random error meet the normality assumption.At the same time,it is proved that the more accurate the temperature prediction model is,the more effective it is to solve the problem of magnification of option pricing error and thus the economic losses can be reduced.

关 键 词：时变Ornstein-Uhlenbeck过程 正态性假设 期权定价 

分 类 号：F830.91[经济管理—金融学]