管理评论

›› 2020, Vol. 32 ›› Issue (4): 3-11.

• 经济与金融管理 •    下一篇

时变O-U模型在农业气温指数保险定价中的适用性研究

李永, 侍欢, 李海英   

  1. 同济大学经济与管理学院, 上海 200092
  • 收稿日期:2017-06-16 出版日期:2020-04-28 发布日期:2020-05-07
  • 通讯作者: 李永(通讯作者),同济大学经济与管理学院副教授,硕士生导师,博士
  • 作者简介:侍欢,同济大学经济与管理学院硕士研究生;李海英,同济大学经济与管理学院讲师,博士。
  • 基金资助:

    国家社会科学基金重大项目（16ADA052）；国家社会科学基金一般项目（15BJY127）。

Study on the Applicability of Time-varying O-U Model in the Pricing of Agricultural Temperature Index Insurance

Li Yong, Shi Huan, Li Haiying   

  1. School of Economics & Management, Tongji University, Shanghai 200092
  • Received:2017-06-16 Online:2020-04-28 Published:2020-05-07

PDF (PC)

609

摘要：

农业气温指数保险定价的首要环节是提升气温预测值的精确度，本文通过在均值回复速率设定中引入时间序列模型，构建了时变O-U模型，分别拟合武汉、大连、郑州1951—2015年的日均气温变动特点，并检验了模型预测精准度。在此基础上，以武汉为例测算了一份气温指数保险合约中保险双方的收益。研究发现，时变O-U模型较好地拟合了气温数据变动趋势，提升了预测精确度；模型改进之后，使保险合约价格上升，同时也使农民收益为正的概率上升。这一方面能够使保险公司获得较高的保费收入，有利于冲减经营成本；另一方面虽然使农民支付了稍高的保费，但是最终获得收益为正的概率却提高了。

关键词: 天气衍生品, 气温指数保险, 时变O-U模型, 均值回复速率

Abstract:

The first part of the agricultural temperature index insurance pricing is to improve the accuracy of the temperature prediction. In this paper, the time series model is introduced in the speed of mean reversion to construct a time-varying O-U model. On this basis, we fit the characteristics of daily temperature variation in Wuhan, Dalian and Zhengzhou from 1951 to 2015 respectively, and examine the accuracy of model prediction. On this basis, Wuhan is taken as an example to calculate the income of both the insurer and the insured in a temperature index insurance contract. It is found that the time-varying O-U model can better fit the trend of temperature data, improve the prediction accuracy, boost the price of insurance contracts and increase the probability of farmers' income ending up positive. As a result, on the one hand, it can make the insurance company get higher premium income, thus reducing the cost of operation. On the other hand, although the farmers pay slightly higher premium, their incomes are more likely to end up positive.

Key words: weather derivatives, temperature index insurance, time-varying O-U model, speed of mean reversion