Research on Crop Area Yield Insurance Pricing in China——Case of Yunnan Rice with the ENN and ERF Methods
Rice is one of the most fundamental crops in China and the area yield insurance helps to protect people's quality of life,maintain stable farmers'incomes,and promote sustainable economic development.The key issues to the area yield insurance lie in accurate yield forecasting,risk zoning and ratemaking.This paper gathers rice unit yield data and six meteorological indicators from 16 prefecture-level cities in Yunnan Province and attempts to introduce several expectile ma-chine learning methods to train the data and make more accurate forecasts of future unit yields by utilizing the excellent characteristics of expectile,which focuses more on the extreme losses in the tail regions.A non-parametric kernel density is used to estimate the density function of the distribution of the rice unit yield.Pure premium rates are calculated and three risk areas are obtained according to those rates:low,medium,and high.We design a novel rice unit yield insurance and ex-plore the difference between the subsidy policies derived from our models and the policies implemented in the real world.Finally,a mean-semivariance model is applied to test the effectiveness of the new unit yield insurance in protecting farmers'incomes.The results show that the novel unit yield insurance can effectively reduce the fluctuation of farmers'income and potential losses,and ensure the stability of farmers'incomes.Based on research findings,we propose the adoption of a province-wide uniform pricing insurance design scheme,the establishment of a pre-disaster prevention and post-disaster loss reduction mechanism,and the optimization of the government's financial subsidy measures and other measures,as well as advice on the business strategy of agricultural insurance companies,to improve the ability of the production area to cope with disasters,while reducing the risk of production of farmers and to promote the sustained increase in farmers'incomes.
NETL
NSTL
万方数据
