摘要
对农作物巨灾风险进行科学度量,并在此基础上进行风险分层管理,对于理论深化与管理实践有重要意义。文章阐述了农作物巨灾风险度量的步骤和方法,利用农作物灾情数据,计算得到农作物灾害损失率数据;对数据进行厚尾性检验,选择最优的分布模型,进行蒙特卡洛模拟,获得新样本空间。综合采用超额均值图、Hill图、峰度法确定阈值,运用超阈值模型(POT),采用广义帕累托分布(GDP)拟合新损失率数据,根据灾害重现期,测算VaR、TVaR,最后选择福建省农作物洪涝灾害作为案例进行实证研究。
Scientific measurement of crop catastrophe risk and layered risk management on this basis are of great significance for theory deepening and management practice.This paper describes the steps and methods of crop catastrophe risk measurement,and uses the data of crop disaster to calculate the loss rate of crop disaster.And then,the paper conducts a thick tail test on the data,selecting the optimal distribution model for Monte-Carlo simulation so as to obtain the new sample space.The paper also comprehensively uses the excess mean graph,Hill graph and kurtosis method to determine the threshold value,employing the super threshold model(POT),and the generalized Pareto distribution(GDP)to fit the new loss rate data,and calculating VaR and TVaR according to the disaster recurrence period.Finally,the paper selects the case of crop flood disaster in Fujian Province to conduct an empirical study.
作者
占纪文
郑思宁
徐学荣
Zhan Jiwen;Zheng Sining;Xu Xuerong(College of Economics,Fujian Agriculture and Forestry University,Fuzhou 350002,China;School of Public Management,Fujian Agriculture and Forestry University,Fuzhou 350002,China)
出处
《统计与决策》
CSSCI
北大核心
2020年第3期37-41,共5页
Statistics & Decision
基金
国家自然科学基金资助项目(71703023)。
