Abstract: In order to effectively reduce the adverse selection and moral hazard in the traditional forest insurance market, reduce the cost of claims, and improve the efficiency of claims, authors designed drought weather index insurance based on the development of the walnut industry in Yunnan.First of all, the drought index was calculated based on anomaly percentage of annual precipitation.Secondly, the HP filter method was used to calculate the walnut yield reduction rate. Finally, by constructing a regression model to determine the correlation between the drought index and the rate of yield reduction, the pure rate and risk zoning were determined.The time span of the panel data is from 2007 to 2017. Meteorological data and walnut output data come from the China Forestry Statistical Yearbook compiled by the State Forestry Administration, Yunnan Statistical Yearbook compiled by the Yunnan Provincial Statistics Bureau, Yunnan Yearbook compiled by the Yunnan Provincial People's Government, Yunnan Provincial Government Report on National Economic and Social Development and statistical yearbook sharing platform. Others are from The official website of the Yunnan Forestry and Grassland Bureau (http://lcj.yn.gov.cn/) Government Report on National Economic and Social Development of Yunnan Province. At the same time, refer to the relevant data of the 2019 Yunnan Province Climate Bulletin. There are 8 steps in the innovative design of insurance products. (1) Determine the homogeneity of the region, and apply weather index insurance when the homogeneity is high. (2) Determine the drought weather index, based on the impact of precipitation on the growth of walnuts, and refer to the existing drought grade standards, to determine the different drought indexes in the region. (3) Calculate the yield reduction of walnuts due to lack of water, and isolate the impact of precipitation on crop growth potential, that is, isolate the drought weather yield. (4) Quantify the regression relationship between walnut yield reduction rate and drought index, and establish a regression model. (5) Determine the critical value of production reduction due to water shortage and determine the trigger compensation value of insurance. (6) The price of insurance contracts is determined, and the pure rate is determined based on the principle of "equal expected value", and the contract is designed. (7) Feedback on product pilots, and revise and optimize products based on market demand. (8) Market promotion and popularization of products, and do a good job in product publicity. The results showed that the pure premium rate of walnut drought index insurance in different states (cities) in Yunnan is between 1.26% and 34.39%. Cluster analysis found that there are significant differences in the risk loss probability of various states (cities). Among them, Qujing, Pu'er and other areas are in mild ecologically fragile areas with fewer drought disasters, while Zhaotong and Honghe are in strong ecologically fragile areas with more drought disasters. Which provide a theoretical basis for insurance companies to further calculate the rate. Based on the risk zoning cluster analysis of pure premium rate, different states (cities) can set different premium subsidy ratios, so as to improve the premium subsidy policy system. According to the distribution of walnut planting in Yunnan and the regional drought situation, Yiliang county, Zhaotong city, which has a higher pure premium rate in northern Yunnan, Changning county, Baoshan city, which is ecologically fragile in western Yunnan, and Dayao county, Chuxiong prefecture, the "hometown of walnuts in China" in eastern Yunnan and Yangbi county, Dali prefecture, can conduct pilot projects and optimize insurance products based on feedback from actual conditions.

Key words: Drought weather index, Walnut yield reduction rate, Rate determination, Insurance product design, Cluster analysis