| Since2007the central finance to carry out a pilot premium subsidy that Agriculturalinsurance in China has achieved leapfrog development and the function of it has beenprominent day by day under China Insurance Regulatory Commission’s active promotion.However, the transaction cost of traditional agriculture insurance in China is generallyrelatively high due to adverse selection and moral hazard problems, thus impeding thedevelopment of agricultural insurance. Area-based Crop Yield Insurance developed in1980s as the product innovation of agricultural insurance can effectively avoid the adverseselection and moral hazard problem in insurance, so that it favored by all countries inpractice. The pilot area-based yield insurance is an effective way to solve the shortage ofagricultural insurance, but also the development trend of China’s agricultural insurance inthe future. But, scientific premium rate determination is a realistic problem urgently to besolved in domestic production of the pilot area insurance. Therefore, this paper carries onthe study of premium rate determination of area-based crop yield insurance is of greatsignificance for the development of agricultural insurance and its pricing, but also toprovide the necessary theoretical support for China’s agricultural insurance pilot.This paper firstly analyzes the characteristics of area yield crop insurance and status ofagricultural insurance pilot in China and designs the area yield insurance drawing lessonsfrom foreign mature crop area yield insurance model. Secondly, this paper chosesproduction distribution statistical model method as the pricing method, through analyze andcompare the advantages and disadvantages of crop area yield insurance ratemaking theoryand method. Under the framework of the production distribution statistical model method,wavelet analysis and nonparametric statistical method is introduced into the pricing modelto make the corresponding improvement in view of the previous defects of premiumratemaking method:On the one hand, considering that the wavelet analysis reflects the superiority innon-stationary time series analysis, it is applied to the trend analysis and forecast of crop yield. Through the wavelet multi resolution analysis better separate period periodic term,trend item and random item, then obtain the trend component reacting technology progressand the weather fluctuation components. Namely, separate the trend yield(in response tooutput growth trend caused by technological progress, natural geographical environmentchange, improvement of infrastructure and laborers’ quality) and fluctuation yield (reactingthe yield fluctuation caused by meteorological conditions).Then in combination withsupport vector machine (SVM) for a more accurate fitting and forecast to get the predictionof trend yield. On the other hand, in order to solve the heteroscedasticity of the data in thefluctuating yield distribution fitting, this paper uses the relative yield loss rate as thealternative values of fluctuation yield for probability density estimation. Finally, this paperuses the nonparametric estimation method to simulate distribution of relative yield loss rateand get the probability density function of yield, because the nonparametric methodwithout prior assumptions distribution type and suitable for any distribution.Eventually obtain the improved area-based crop yield insurance pricing model basedon wavelet-nonparametric method. Choose Shandong province a big agriculture provincein China’s corn yield data from1949to as samples for empirical research, get the cornyield insurance premium rate under different protection levels and a comparative analysis.Obtain the following conclusions:It helps to improve the accuracy of the ratemaking to de-trend crop yield time seriesthrough wavelet analysis method study its trend and fluctuation term; On the same securitylevel, parameter estimation under normal distribution assumption that the rate level is lowerthan using nonparametric kernel density estimation was carried out on the yield loss ratedistribution fitting to get pure rates. Mainly due to crop yield data distribution tend to havepositive skewness and rush fat-tailed features, so that the premium ratemaking technologyof parameters estimation under the normal distribution assumption underestimated the risksand nonparametric method of loss distribution fitting more accord with the actualdistribution of data is more accurate; With the protection level’s decreasing, corn area yieldinsurance rate showed a trend of decline, and reasonably determining the security level hasthe vital significance for successful pricing and pilot development on crop insurance; Inaddition, corn insurance premium rate concluded in this paper is significantly lower than the actual current rate in Shandong. Mainly, due to the higher transaction cost in survey toassess the damage and claims and other aspects of agricultural insurance’s actual operation,and thus raised the total premium level. |
PDF Full Text Request