根据茶叶基地历年茶叶逐日经济产出,确定茶叶进入开采期或开采期前遭受霜冻的经济损失率与最低气温的关系。利用支持向量机将位于各乡镇、街道的中尺度自动气象站的最低气温资料延长到30 a;最低气温和茶叶经济损失率相对应,因此通过根据最低气温资料计算不同等级霜冻出现风险确定茶叶经济损失风险,解决了茶叶品种种植时间短、中尺度自动气象站资料积累时间短,不能满足风险分析要求的问题。利用多种风险分析模型拟合分析各乡镇、街道茶叶处于不同开采期时的最低气温分布,从中选择最优的理论概率分布函数进行序列的风险概率估算,得到较为稳定并符合实际的风险评估结果。在风险定量分析基础上,从浙江省政策性农业保险经营的实际需要出发,综合区域产量保险和气象指数保险的优点,设计了精细化到乡镇一级的茶叶霜冻气象指数保险,降低了农业保险中存在的基差风险、逆选择和道德风险,在灾后理赔时不需要大量的人力、物力勘查定损,理赔时效高、理赔成本低,为开展茶叶政策性农业保险提供技术支撑。
This paper takes tea production in Xinchang County of Zhejiang as an example, under the principle of disaster risk analysis and requirement for agricultural insurance, combined with characteristics such as huge topographic relief and significant difference in microclimate in southern mountain area of China, carries out risk evaluation and agricultural insurance product design through elaborating frost disaster which affected tea production in March at village & township level. Firstly, the paper confirms tea's economic output changes from time to time in productive life based on tea's daily economic output in normal years of tea planting base; combined with economic loss rate due to frost during/before productive life in previous years to confirm frost weather index—minimum temperature in a period before productive life or after productive life, different minimum temperature is in accordance with tea's economic loss rate. Support Vector Machine (SVM) is used to extend mesoscale automatic weather station's minimum temperature data to 30 years in specific to non-linear relation existing among meteorological data of mesoscale automatic weather station in villages, towns and county; minimum temperature corresponds to tea's economic loss rate, therefore it's possible to calculate different levels of frost risk based on minimum temperature data and confirm tea's economic loss risk, and to solve problems such as tea's short planting period, short accumulation period for mesoscale automatic weather stations and unable to meet risk analysis requirement. There're certain differences between risk results calculated on different risk analysis models. The article applies probability density function fitting distributions such as Beta, Exponential, Gumbel, Gamma, Generalized Extreme Value, Inverse Gaussian, Logistic, Log-Logistic, Lognormal, Lognormal2, Normal, Pareto, Pareto2, Pearson Type V, Pearson Type VI, Student, Weibull to minimum temperature data sequence in every township and street in the study area. Parameter estimation in distribution model applies maximum likelihood method, both of Anderson-Darling and Kolmogorov-Smirnov examination selected from different production periods in different townships and villages passed significance level of 0.05. Almost coincidence distribution of P-P drawing's tailer and diagonal line confirms that Generalized Extreme Value distribution's risk probability estimation as the best theoretical probability distribution function, thus being relatively stable with realistic risk evaluation result. On the basis of quantitative risk analysis and considering actual needs of policy-based agricultural insurance operation, combined with advantages of regional output insurance and weather index, the paper designs tea's frost weather index insurance. It is to confirm insurance rate and compensation agreement in a pre-designated area on the basis of frost caused economic loss rate and occurrence risk in the case that frost happens during tea's production life. Tea's frost weather index is weather index corresponds to pre-determined frost meteorological event, every index value corresponds with certain tea economic loss rate and claim ratio. This paper suggests to carry out compensation according to frost weather index confirmed by meteorological data observed by mesoscale automatic weather stations in every township, village and street, compensation index for rural household and insurance company is measured by representative weather stations closest to rural households, therefore basis risks, adverse selection and moral risk are reduced. And there is no need to use huge human and non-human sources to inspect and judge losses in claim settlement after disaster, and the claim settlement efficiency is high but the claim settlement cost is low, hence this is a positive and effective technological measure to carry out tea insurance and promote sustainable development of agricultural insurance.
[1] Motha R P. Development of an agricultural weather policy [J]. Agricultural and Forest Meteorology, 2007, 142: 303-313.
[2] Coble K H, Hanson T, Miller J C, et al. Agricultural insurance as an environmental policy tool [J]. Journal of Agricultural and Applied Economics, 2003, 35(2): 391-405.
[3] Lou W P, Qiu X F, Wu L H, et al. Scheme of weather-based indemnity indices for insuring against freeze damage to citrus orchards in Zhejiang, China [J]. Agricultural Sciences in China, 2009, 8(11): 1321-1331.
[4] GlobalAgRisk. Designing Agricultural Index Insurance in Developing Countries: A GlobalAgRisk Market Development Model Handbook for Policy and Decision Makers [M]. Lexington, KY: GlobalAgRisk, 2009: 12-13.
[5] Makki S S. Crop insurance: inherent problems and innovative solutions//Tweeten L, Thompson S R. Agricultural Policy for the 21 Century. Ames: Iowa State Press, 2002.
[6] Smith V H, Chouinard H H, Baquet A E. Almost ideal area yield crop insurance contracts [J]. Agricultural and Resource Economics Review, 1994, 23: 75-83.
[7] Wenner M, Arias D. Agricultural insurance in Latin American: Where are we? Paving the Way Forward for Rural Finance an International Conference on Best Practices. 2003.
[8] Ozaki V A, Ghosh S K, Goodwin B K, et al. Spatio-temporal modeling of agricultural yield data with an application to pricing crop insurance contracts [J]. American Journal of Agricultural Economics, 2008, 90(4): 951-961.
[9] Zeng L. Weather derivatives and weather insurance: Concept, application, and analysis [J]. Bulletin of the American Meteorological Society, 2000, 81(9): 2075-2082.
[10] Yang C C, Brockett P L, Wen M M. Basis risk and hedging efficiency of weather derivatives [J]. Journal of Risk Finance, 2009, 10(5): 517-536
[11] Botts R R, Boles J N. Use of normal-curve theory in crop insurance rate making [J]. Journal of Farm Economics, 1958, 40: 733-740.
[12] Just R E, Weninger Q. Are crop yields normally distributed? [J] American Journal of Agricultural Economics, 1999, 81(2): 287-304.
[13] Ramirez OA, Misra S K, Nelson J. Efficient estimation of agricultural time series models with non-normal dependent variables [J]. American Journal of Agricultural Economics, 2003, 85(4): 1029-1040.
[14] Chen S, Miranda M. Modeling multivariate crop yield densities with frequent extreme values . Paper Presented at the American Agricultural Economics Association Annual Meeting, Denver, Colorado, 2004.
[15] Goodwin B K, Roberts M C, Coble K H. Measurement of price risk in revenue insurance: Implications of distributional assumptions [J]. Journal of Agricultural and Resource Economics, 2000, 25(1): 195-214.
[16] Lu Y, Ramirez O A, Rejesus R M, et al. Empirically evaluating the flexibility of the Johnson family of distributions: A crop insurance application [J]. Agricultural and Resource Economics Review, 2008, 37(1): 79-91.
[17] Ker A P, Goodwin B K. Nonparametric estimation of crop insurance rates revisited [J]. American Journal of Agricultural Economics, 2000, 83(5): 463-478.
[18] Ozaki V, Silva R. Bayesian ratemaking procedure of crop insurance contracts with skewed distribution [J]. Journal of Applied Statistics, 2009, 36(4): 443-452.
[19] Featherstone A M, Kastens T L. Non-parametric and semi-parametric techniques for modeling and simulating correlated, non-normal price and yield distributions: applications to risk analysis in Kansas agriculture [J]. Journal of Agricultural and Applied Economics, 2000, 32(2): 267-281.
[20] 王丽红, 杨讷华, 田志宏, 等. 非参数核密度法厘定玉米区域产量保险费率研究——以河北安国市为例[J]. 中国农业大学学报, 2007, 12(1): 90-94.
[21] 娄伟平, 吴利红, 邱新法, 等. 柑桔农业气象灾害风险评估及农业保险产品设计[J]. 自然资源学报, 2009, 24(6): 1330-1340.
[22] Sherrick B J, Zanini F C, Schnitkey D G, et al. Crop insurance valuation under alternative yield distributions [J]. American Journal of Agricultural Economics, 2004, 86(2): 406-419.
[23] Glauber J W. Crop insurance reconsidered [J]. American Journal of Agricultural Economics, 2004, 86(5): 1179-1195.
[24] Gunnar B, Raushan B, Olaf H. Evaluating the potential of index insurance schemes to reduce crop yield risk in an arid region [J]. Journal of Agricultural Economics, 2008, 59(2): 312-328.
[25] Barnett B J, Barrett C B, Skees J R. Poverty traps and index-based risk transfer products [J]. World Development, 2008, 36(10): 1766-1785.
[26] Mapfumo S. Weather index insurance—The case for South Africa . Micro Insurance Agency Holdings LLC, 2007.
[27] United States Agency for International Development. Index insurance for weather risk in lower income countries . Prepared by GlobalAGRisk, Inc. Lexington, 2006: 23.
[28] World Bank. China: Innovations in agricultural insurance, promoting access to agricultural insurance for small farmers . East Asia Agricultural and Rural Development Department and Finance and Private Sector Unit, Washington, D C, 2007.
