广东水稻生产风险评估与保险费率厘定研究

发布时间：2018-04-29 07:02

本文选题：水稻 + 风险评估　；参考：《中国农业科学院》2013年硕士论文

【摘要】：广东作为我国自然灾害高发地区，每年因各种灾害造成的经济损失近120亿元。与此同时，同样作为经济强省和产粮大省，广东农业保险开展远远落后于浙江、江苏等省，农业保险对恢复农业生产、促进农村经济和保障农民生活的缓冲作用远没有得到发挥。为促进农业保险在广东快速发展，本研究以当地农业主导产业之一水稻，主要从生产风险评估、风险区划和保险费率厘定三方面对广东省水稻保险进行了相关研究。首先，在风险评估部分，，遵循“单产—趋势—波动—评估”范式，运用直线滑动平均计算广东各地市早晚稻趋势单产和波动率，利用非参数信息扩散模型评估水稻生产风险。结果表明，广东水稻生产风险区域性差别较大，如江门、湛江早稻生产损失率超过10%的概率分别高达0.103917和0.084829，而广州仅0.030074；整体来讲水稻单产减产10%以上的概率较小；早稻生产风险小于晚稻生产风险。其次，在风险区划部分，利用风险分散原则、对价交换理论和农业气象学理论，选取“有效灌溉率”、“旱涝保收率”、“年平均气温”、“年日照时数”、“单产变异系数”、“单产损失超过5%的概率”和“单产损失超过10%的概率”等11个指标作为因子分析的初始分析指标，利用因子分析和聚类分析，在对广东水稻生产风险评估的基础上开展风险区划。根据综合得分排名，将广东20个水稻种植地市分为低风险区、较低风险区、一般风险区、较高风险区和高风险区5个等级。研究结果表明：整体而言，粤东和粤西沿海地区水稻生产风险最高，粤北山区次之，珠三角地区风险较低。最后，在保险费率厘定部分，先拟合各地区早晚稻单产损失分布，再利用Matlab计算各地区纯费率。在水稻单产损失模型分布拟合和纯费率计算中，遵循“模型构建—模型检验—实证研究”，以@Risk中的40多个参数分布，经统计检验得出各地市早晚稻损失单产的最佳分布，再利用Matlab编程计算各地市早晚稻保险纯费率。计算结果表明：与风险区划研究结果类似，湛江、阳江、茂名、江门等沿海地区水稻保险纯费率较大；韶关、肇庆、云浮等北部区域次之；广州、佛山、东莞等珠三角地区水稻保险纯费率最小。本研究结果表明：（1）广东各市早晚稻保险纯费率在1.5%—2.5%之间，保障单产在300—400公斤/亩之间；且不同地区费率差异较大，但当前广东水稻保险实行的是全省统一费率。（2）对水稻保险实际业务，应根据各地市早晚稻生产风险的不同，制定分地区分品种的差异化保险费率。
[Abstract]:Guangdong, as a region with a high incidence of natural disasters, suffers an economic loss of nearly 12 billion yuan per year due to various disasters. At the same time, as a strong economic province and a big grain producing province, Guangdong's agricultural insurance development lags far behind that of Zhejiang and Jiangsu provinces. Agricultural insurance has been used to restore agricultural production. The buffering effect of promoting rural economy and protecting farmers' life is far from being brought into play. In order to promote the rapid development of agricultural insurance in Guangdong, rice insurance in Guangdong Province was studied from three aspects: production risk assessment, risk regionalization and premium rate determination. Firstly, in the part of risk assessment, following the paradigm of "unit yield-trend-fluctuation-assessment", the trend yield and volatility rate of early and late rice were calculated by linear moving average, and non-parametric information diffusion model was used to evaluate rice production risk. The results showed that there were significant regional differences in rice production risk in Guangdong Province, such as Jiangmen and Zhanjiang, where the probability of loss rate of early rice production was as high as 0.103917 and 0.084829, respectively, while that of Guangzhou was only 0.030074, and the probability of reducing rice yield per unit yield by more than 10% was smaller. The production risk of early rice is lower than that of late rice. Secondly, in the risk regionalization part, using the principle of risk dispersion, the theory of price exchange and the theory of agrometeorology, we select "effective irrigation rate", "waterlogging rate", "annual mean temperature", "annual sunshine time". "coefficient of variation per unit yield", "probability of loss per unit yield over 5%" and "probability of loss per unit yield over 10%" were used as the initial analysis indexes of factor analysis, using factor analysis and cluster analysis. On the basis of risk assessment of rice production in Guangdong Province, risk regionalization was carried out. According to the rank of comprehensive score, 20 rice planting cities in Guangdong were divided into five grades: low risk area, low risk area, general risk area, high risk area and high risk area. The results showed that the risk of rice production was the highest in the coastal areas of eastern Guangdong and western Guangdong, followed by the mountainous areas of northern Guangdong, and the risk was lower in the Pearl River Delta. Finally, in the part of premium rate determination, the distribution of loss per unit yield of early and late rice was fitted first, and then Matlab was used to calculate the net rate of each region. In the distribution fitting of rice yield loss model and the calculation of pure rate, the best distribution of loss per unit yield of early late rice was obtained by statistical test according to the distribution of more than 40 parameters in "model-examination-empirical study". The net rate of early and late rice insurance was calculated by Matlab programming. The calculated results show that, similar to the risk regionalization study, the pure rates of rice insurance in coastal areas such as Zhanjiang, Yangjiang, Maoming and Jiangmen are higher; Shaoguan, Zhaoqing, Yunfu and other northern regions are second; Guangzhou, Foshan, Dongguan and other Pearl River Delta area rice insurance net rate is the smallest. The results show that the net insurance rate of early and late rice in Guangdong Province is between 1.5% and 2.5%, and the guaranteed yield is between 300 kg and 400 kg / mu, and the rates vary greatly in different regions. However, the current rice insurance in Guangdong Province is based on the unified rate of the whole province. (2) the actual business of rice insurance should be based on the different risks of early and late rice production in different cities and cities, and the differential insurance rates for different regions and varieties should be established.
【学位授予单位】：中国农业科学院
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F326.11;F842.66

【参考文献】