中国农业风险分析及管理

AnalysisandmanagementofagriculturalriskinChina中国农业风险分析和管理Prof.ZhangQiaoAgriculturalInformationInstitute(AII),ChineseAcademyofagriculturalsciences(CAAS)April22.2008Contents（目录）1.Agriculturalriskanalysis农业风险分析2.Agriculturalriskmanagement农业风险管理3.Effectivenessofcropinsurance农业保险效果1.Agriculturalriskanalysis农业风险分析1.1Agriculturalriskclassification农业风险分类1.2Agriculturalriskfeature农业风险特征1.3Agriculturalriskanalysis农业风险分析1.1Agriculturalriskclassification农业风险分类•Agriculturalrisk（概念）Thedegreethatindividual’sactualrevenuelowerthanexpectedrevenue,foruncertaintyandboundedrationality,duringagriculturalproductionandoperationactivities风险是指在一定时期和一定客观环境下，由于不确定性的存在和人的有限理性，致使经济行为主体的实际收益低于预期收益的程度。•Agriculturalriskclassification（农业风险分类）—naturalrisk—IndependentRisk—NormalRisk（自然风险）（独立风险）（一般风险）—marketrisk—SystemicRisk—CatastrophicRisk（市场风险）（系统风险）（巨灾风险）•Agriculturalyieldrisk（农业产量风险）•Priceriskofagriculturalcommodity（农产品市场风险）•Agriculturalcatastrophicrisk（农业巨灾风险）1.2Agriculturalriskfeature农业风险特征•YieldRisk（生产风险）Relativeindependent,stochastic相对独立性、随机性.•PriceRisk（市场风险）Systemic,conjunctive系统性、关联性.•CatastrophicRisk（巨灾风险）Lowprobability,highLoss,strongimpactive发生概率低、损失程度大、破坏性强AnExample:CornyieldriskanalysisandevaluationinChina（中国玉米生产风险分析和评估）¾Background（背景）¾Analysisobjective（分析目标）¾Analysisapproach（分析方法）¾Analysisresults（分析结果）1.3Agriculturalriskanalysis农业风险分析Background（背景）CornisoneofthemaincropinChina,andtheCornBeltinChinaisfromNortheasttoSouthwest.玉米是中国主要粮食作物之一，我国玉米生产分布大致呈现一个从东北到西南的斜长带。CornBeltWholecountryPercentAcreage(104Ha)2213.432464.077989.8%Output(104t)10826.4412114.72589.4%Averageof2000-2005TheacreageandproductioninCornBeltandwholecountry主产区玉米播种面积和产量占全国的比重Background（背景）WehadselectedeightprovinceasthesampleofChineseCropBelt，thereareHeilongjiang,Jilin,Liaoning,Shandong,Henan,Sichuan,GuizhouandGuangxi)我们选取了八个省份作为我国玉米主产区的样本代表，分别为：黑龙江、吉林、辽宁、山东、河南、四川、贵州和广西Averageof2000-2005ThepercentofSampleacreage&productiontoCornBelt八省份玉米播种面积和产量占主产区的比重HLJJLLNHNSDSCGXGZSumAcreage9.5%11.8%7.0%10.6%11.3%5.4%2.5%3.2%61.4%Output8.5%14.0%8.2%10.1%13.8%4.9%1.6%3.1%64.1%Analysisobjective(分析目标)¾Analysisandcomparetheyieldrisk(CV,Mean,Variation)betweendifferentregions分析和比较不同地区的产量风险（均值、标准差和变异系数等）¾Establishtheriskdistributionmodelsfordifferentregions建立各个地区生产风险分布模型（单产分布模型）¾Estimateandrankyieldriskindifferentregions估算和比较不同地区的生产风险大小Analysisapproach(分析方法)•Step1:Collectyielddataandgottherelativestochasticvariation(RSV)第一步：收集单产数据，计算相对随机波动（RSV）a:Detrend（去除趋势）And:acturaltrendacturaltrendyyuuyy=+⇒=−trenduRSVy=Analysisapproach(分析方法)b:Weusethelinearmovementaverage(LMA)tosimulatetheyieldtrend我们用线性移动平均法（LMA）计算单产趋势(Why?Causethismethodshavetheadvancedofobjectively&didn’tlosesampledata)为什么？因为该方法具有既不必主观假定产量历史演变的曲线类型，同时又不损失样本量的优点Analysisapproach•Ademo0100200300400500505560657075808590950005yield-120-80-4004080120505560657075808590950005afterdetrend-.5-.4-.3-.2-.1.0.1.2.3505560657075808590950005RSVAnalysisapproach(分析方法)•Step2:Selectcandidateparameterizationofyielddistributionfordifferentregions第二步：采用参数方法，为不同地区选定合适的单产分布形式a:ADtestAD检验b:Comparetheskewnessandkurtosisofsampleswiththemomentratiodiagrams将各样本的偏度、峰度值与距比率图进行比较Analysisapproach(分析方法)¾Step3:UsingMLEtoestimatetheparametersofyielddistributions,thengotthePDF&CDFfordifferentregions第三步：利用极大似然估计法估计各单产分布的参数，得到各样本的概率密度函数和累积概率密度函数Analysisresults(分析结果)•RESULT(1)Theyieldrisk(CV,Mean,Variation)betweendifferentregions不同地区生产风险（均值、方差和变异系数）图示-0.0100-0.00500.00000.00500.01000.01500.02000.02500.0300HeiJILIAOYULVCHUANQIANGUIMean&Variation-15-10-50510CVMeanVariationCVAnalysisresults(分析结果)•RESULTS(2)Thehistogram,PDFandyielddistributionofthesesampleis:各样本省的单产分布直方图、概率密度函数和单产分布模拟图：Analysisresults(分析结果)¾Heilongjiang(Weibulldistribution)黑龙江（服从Weibull分布）6.5350()4.2873()exp[(1.75751.7575xxfx7.5350=××−)]02468101214-0.2-0.00.20.40.60.8Analysisresults(研究结果)¾Jilin(Normaldistribution)吉林（服从正态分布）21(0.002014)()exp[]0.29520.0277xfx+=−0123456789-0.2-0.1-0.00.10.2Analysisresults(研究结果)¾Liaoning(Weibulldistribution)辽宁（服从Weibull分布）2.1316()1.0836()exp[(2.892.89xxfx3.1316=××−)]02468101214-0.4-0.2-0.00.2Analysisresults(研究结果)¾Henan(Weibulldistribution)河南（服从Weibull分布）1.5212()1.0455()exp[(2.41142.4114xxfx2.5212=××−)]0246810-0.3-0.2-0.1-0.00.1Analysisresults(研究结果)¾Shandong(Weibulldistribution)山东（服从Weibull分布）2.0047()1.0428()exp[(2.88142.8814xxfx3.0047=××−)]024681012-0.2-0.1-0.00.1Analysisresults(研究结果)¾Sichuan(Logisticdistribution)四川（服从Logistic分布）0.0020680.00206820.0483350.048335()0.048335(1)xxfxee++−−−=××+024681012-0.3-0.2-0.1-0.00.10.2Analysisresults(研究结果)¾Guizhou(Logisticdistribution)贵州（服从Logistic分布）0.0002980.00029820.0492860.0492861()(1)0.049286xxfxee−−−−=××+024681012-0.250-0.125-0.0000.1250.250Analysisresults(研究结果)¾Guangxi(Logisticdistribution)广西（服从Logistic分布）0.0006220.00062220.0517010.0517011()(1)0.051701xxfxee++−−=××+048121620-0.2-0.00.20.42.Agriculturalriskmanagement农业风险管理2.1Agriculturalriskmanagement农业风险管理2.2Agricultureriskmanagementtools农业风险管理工具2.3Frameworksystemofintegratedriskmanagement农业风险管理系统框架设计2.1Agriculturalriskmanagement农业风险管理•Noteliminateoravoidagriculturalrisk并不是消灭或消除农业风