【作者】 张岳；

【导师】 田玲；

【作者基本信息】 武汉大学 ， 保险学， 2012， 博士

【摘要】 随着巨灾发生频率和损失金额的扩大，利用保险市场分散巨灾风险得到学者、公司及政府的关注。在目前的实践及理论研究中，多关注与巨灾保险制度的设计，即供给端的问题，对需求端研究不多。但在巨灾保险实践中发现的诸如参保率不高的问题等说明需求端也会导致市场失灵，使供给端的制度设计效果大大折扣。从理论研究看，目前巨灾保险需求的研究亦出现总体文献少、实证文献缺乏、研究对象集中于美国等现象。在我国呼吁建立巨灾保险制度的声音越来越高的情况下，有必要对需求端进行基础性的研究。本文集中于巨灾保险需求的特征及影响因素这个焦点，发现巨灾保险需求曲线呈现先后弯曲的特征，这导致了巨灾保险市场出现多处均衡，也导致市场失灵。文章将影响巨灾保险需求的因素分为客观因素和主观因素，分别建立了相应的理论模型分析了各因素的影响路径及影响程度。理论研究是实证研究的基础，实证研究是理论研究的证明与延伸。在理论研究的基础上，使用九个省的数据分别验证了我国洪水、台风、地震保险需求的影响因素及影响程度，分别使用了面板分析技术与背景分析方法。文章还设计了实验的方法验证风险态度对洪水保险的需求影响，具体设计了实验方案。最后，结合理论研究与实证研究的结论，从需求方提出了对巨灾保险制度设计的三点建议。第一章解决了为什么研究巨灾保险需求以及研究的起点问题。利用近几年尤其是2011年及2012年初的国外文献，梳理了巨灾保险需求的研究现状及问题，并重点对三类实证方法进行了归类分析，为全文的方法论奠定了基础。第二章构建了巨灾保险需求的研究框架，将巨灾保险需求研究置于保险需求研究体系下，回顾了个人及企业保险需求的主要特征及研究方法。提出了巨灾保险需求研究的主要方法，并分析了巨灾与一般自然灾害的区别，这将巨灾保险需求与一般保险需求的分析区别开来。第三章分析了巨灾保险的特征及客观影响因素，使用了静态分析方法，利用一般的期望效用模型及等级依赖效用模型发现，巨灾保险需求曲线具有向后弯曲的特征，而这种特征导致供需均为零也是一个市场均衡点，这会导致市场失灵。出现这种情况的影响因素是投保人预期保险人面对巨灾会出现偿付能力不足导致的拒赔现象，这种现象在实践并不鲜见。博弈论的分析方法也在本章应用，将此方法的使用由保险主体扩展到了需求方的研究。在第三章单周期静态分析的基础上，第四章使用了多周期动态分析，利用贝尔曼方程的性质，讨论了影响投保人的两个重要客观因素的影响路径，一是流动性约束，即收入问题，二是对投保人违约的担心。对投保人违约的担心也是巨灾损失的系统性决定的。通过对贝尔曼方程横截面条件和充分性的分析，发现投保人在流动性约束和对投保人违约的担心下，购买不足额保险是最优的，而且随着约束的增强，保额会下降，即巨灾保险需求会下降。在此基础上，本文提出了一种新的缴费方法，即事后缴费，这可以放松投保人的流动性约束和对保险人违约的担心约束，研究证明，这种方法可以提高巨灾保险需求。但带来的新问题如投保人违约及保险的利益受损等需要得到关注。文中还对收入的绝对值与巨灾保险的价格这两个在文献中被认为最重要的影响因素进行了定性分析。第五章主要分析了投保人对低概率事件往往高估及忽视这种主观因素对巨灾保险需求的影响，使用的方法是前景理论。根据投保人不同参考点的设置，利用前景理论中的价值函数研究了投保人的购买行为变化，研究发现，投保人对低概率事件的误判会严重影响巨灾保险的购买。但当投保人购买过或强制购买巨灾保险后，其自主需求会上升。第六章讨论了巨灾保险制度中影响最大的主体之一政府对巨灾保险需求的影响。利用一个图形分析，文章分析了政府救济对巨灾保险需求的挤出效用。在此基础上，建立了两地区的理论模型讨论了政府的介入方式对巨灾保险需求的影响方式及程度。研究发现，政府通过干预市场提高保险人的偿付能力或降低保险人的破产概率，以及提高参保率可以提高投保人的自主的巨灾保险需求。在前六章的基础上，第七章聚焦我国，分别分析了台风保险、洪水保险、地震保险等三类典型巨灾保险的影响因素。台风风险及洪水风险我们使用的面板数据，被解释变量分为经济变量（包括GDP、农村居民房屋价值）、灾害损失、居民风险态度、平均家户规模、政府救灾支出等。解释变量采用《中国保险年鉴》中企财险、农业险及其他保险和保费收入之和。对台风和洪水两大风险的分析结果显示，灾害损失与保费收入正相关，而政府救灾支出与保费收入负相关，这进一步解释了政府救济对巨灾保险需求的挤出效用。但也说明，我国居民对政府的依赖较强。其他解释变量的解释程度不高，可能与数据来源有关。对地震的分析采用了背景风险分析法，结果显示，汶川地震的发生并未带来保费收入的异常增长，结合航空事故发生后航意险脱销的情况，说明我国地震保险的供给还不足，也可能是我国居民对特大自然灾害对政府的依赖很强，没有使用市场机制的意识和习惯。第八章我们设计了一个对风险态度的实验，利用实验方法测量人们对洪水保险的购买行为，实验对象根据国外研究的基础选择了社会人士。实验的方案采用了台湾学者研究台湾洪水保险需求的实验方案，但对购买机制进行了改进。但不足的是受成本制约，实验没有使用有偿的方式，可能对结果造成影响。研究结果发现了投保人高估低概率事件的现象。第九章结合全文理论和实证研究结果，提出了对巨灾保险制度设计的三点建议：一是根据中国国情，重视灾害救助制度的基础性作用，但要与巨灾保险形成有效的互补，降低挤出效用。二是充分利用现有保险条款中已经对洪水、台风等承担保险责任，地震也通过附加条款的形式予以承保。在巨灾保险制度设计中应考虑这些因素，避免引起保险业的震动，但也要设计基础性的巨灾保险产品，满足不同层次需要。三是提高巨灾保险的参保率，主要通过发挥财政补贴的杠杆作用、采取强制或半强制保险的形式以及加强对巨灾保险的宣传等。本研究目标较为明确，即从理论和实证两个方面对巨灾保险需求做一个系统性梳理和研究。在理论研究上，本文在国内较系统的使用博弈论、等级依赖效用函数及贝尔曼方程等工具对巨灾保险需求的影响进行了模型分析。而实证研究则是尝试性的进行了面板分析，由于我国尚无专门的洪水、台风及地震保险方面的数据，实证研究结果的普遍性受到了限制。而使用实验方法测量巨灾保险的购买行为在国内为首先应用。但由于国内巨灾保险需求的研究文献较少，且专门针对我国巨灾保险需求的研究几乎为空白，因此本文在实证研究的数据、理论模型在我国的适用性等方面还存在很大的不足。更多还原

【Abstract】 With expansion of the frequency and the amount of loss as the catastropheoccurred, from the academic, corporate and Government concern catastrophe riskdisperse using insurance market. In the study of current practice and, moreconcerned with design of catastrophe insurance system, that is, problems on thesupply side, have not much demand-side research. Found in the catastropheinsurance practices, such as problems with low insurance rates the demand-side canlead to market failure, system design on the supply side effects greatly discounted.From the perspective of theoretical research, current catastrophe insurance needsalso appears less general references, lacks empirical literature, and participants focuson United States and so on ’s voice rising cases of, There is a need to study on thebasis of the demand-side. calling for catastrophe insurance system in our country.This article focuses on the catastrophe insurance demand characteristics andinfluence factors, finding catastrophic insurance demand curve bendingcharacteristics, which led to the catastrophe insurance market appear more balanced,also lead to market failure. Factors that will affect the demand for catastropheinsurance is divided into objective and subjective factors, respectively theestablishment of corresponding theoretical model analysis of the influence factors ofthe path and how much. On the basis of theory study, we use data validation innine provinces in China China’s flood, Typhoon, earthquake insurance requirementinfluence factors and impact of, respectively, use the a panel with backgroundanalysis methods. Article also designed the experimental methods to verify that therisk of effects on flood insurance requirements, and specific design of theexperimental programme. Finally, the conclusions of the study on the combination oftheory and empirical research, from the demand-side of catastrophe insurance systemdesign was made three suggestions.Chapter I addresses why study on catastrophe insurance needs as well as of thebeginning of the study. In recent years, especially the foreign literature in early2011 and2012, carding status and problems of study on catastrophe insurance needs, andfocus on three types of empirical methods and analysis, lays the Foundation for thefull text of the methodology.Build catastrophe insurance requirements in chapter II of the researchframework, study on catastrophe insurance requirements on the insurancerequirements under the research system, reviewed the main features of personal andbusiness insurance needs and research methods. Made the primary method of studyon catastrophe insurance requirements, and analysis of the differences betweencatastrophe and natural disasters in General, the demand for catastrophe insuranceand general insurance needs analysis distinguishes.Chapter III analyses the characteristics and objective factors of catastropheinsurance, using static analysis methods, using the expected utility model and therank-dependent utility model found, catastrophe insurance demand curve has a bendback feature, this feature causes are zero is also a market equilibrium between supplyand demand, this can lead to market failure. This impact factor is the expected policyholders of the insurer solvency shortfall of repudiation of catastrophe occurs, thisphenomenon is not uncommon in practice. Analysis of game theory methods in thischapter also apply, use of this method by extending insurance subject to thedemand-side research.During static analysis in chapter III on the basis of, during the fourth chapterusing the dynamic analysis, using the Bellman equation, discussed the effects of thetwo important factors that affect the insured path, a liquidity constraint, that is,income, and is worried about the breach of policy holders. Worried about the breachof policy holders are also systemic decisions of catastrophe losses. By Bellmanequation analysis of cross section and sufficiency of the condition and found thatpolicyholders under liquidity constraints and concerns about policy holders defaulted,purchase is not a qualified insurance is optimal, but with the constraint of enhanced,sum insured is reduced, that decline in demand for catastrophe insurance. On thisbasis, this article proposes a new payment method, after the payment, this can relaxliquidity constraints and fears of a default of the insurer of the insured, and studies have shown that this method can improve the catastrophe insurance needs. But newproblems such as the default policy holders and insurance interests, need to beconcerned about. Article also discusses revenue in absolute terms and prices ofcatastrophe insurance, two in the literature is considered the most important factorsfor the qualitative analysis.The fifth chapter analyses the policy holders tend to overestimate lowprobability events and ignored the influence of subjective factors on the demand forcatastrophe insurance, method is to prospect theory. According to different referencepoints of policy holders set, using value function in prospect theory to study theinsured purchase behavior changes, found that policyholders misjudge severe effectson the low-probability events of catastrophe insurance purchase. But after buying orforcing out catastrophe insurance policy holders, the independence needs areincreased.In the sixth chapter discuss catastrophe insurance system affect the Governmentof one of the largest main impact on catastrophe insurance needs. Using a graphicalanalysis, article analysis of the Government relief on extrusion effect of catastropheinsurance needs. On this basis, building theoretical models discussed in the two areasof government intervention affecting the manner and to the extent of the demand forcatastrophe insurance. Studies have found that government intervention in themarket to improve the solvency of the insurer or reduce the probability ofbankruptcy of the insurer, and increased insurance rates can increase the insuredcatastrophe insurance of independent demand.In the first six chapters on the basis of the seventh, focusing, respectivelyanalysis of typhoons, floods, earthquakes, and other effects of three kinds of typicaldemand for catastrophe. Typhoon and flood our panel data, are explanatory variablesinto economic variables (including GDP, housing values of rural residents), disasterlosses, residents ’ attitude towards risk, the average size of households, governmentspending for disaster relief. Yearbook of the explanatory variables in the Chinainsurance enterprise in casualty, agriculture and insurance and other insurancepremiums. Risk analysis revealed two typhoons and floods, disaster losses and premiums are related, and Government disaster relief payments and premiumreceipts, which further explains the Government relief on extrusion effect ofcatastrophe insurance needs. But also note, our residents ’ dependence ongovernment strong. Other explanatory variables explain high level, may beassociated with the data source. Background risk analysis method for seismicanalysis, results showed that earthquake occurred in Wenchuan has not causedabnormal growth in premium income, combined with the airlines after an aviationaccident insurance stock, supply enough description of earthquake insurance inChina, may also be residents for catastrophic natural disasters in our country’sdependence on Government is so strong that no consciousness and habit of usingmarket mechanisms.The eighth chapters we have designed an experiment on risk attitudes, usingexperimental method to measure people’s purchase of flood insurance, selectstudents as study abroad Foundation of experimental objects. Experimentalprogramme was conducted in a Taiwan scholar research on Taiwan experimentscheme of flood insurance requirement, but the purchase mechanism has beenimproved. But is the cost constraints of insufficient, experiments do not use paid,may have an impact on the results.Nineth combining full-text theory and empirical findings of catastropheinsurance system design was made three suggestions: first, according to China’snational conditions, importance to the fundamental role of disaster relief system, andform an effective complementarity of catastrophe insurance, reducing the extrusioneffect. Second is to have full use of the existing insurance terms insuranceresponsible for floods, typhoons, earthquakes through the additional terms shall be inthe form of underwriting. Catastrophe insurance system should be taken into accountin the design of these factors, avoidance of vibrations in the insurance industry, butalso to design a basic catastrophe insurance products to meet needs at different levels.Third, increase the rate of insurance of catastrophe insurance, mainly throughfinancial subsidies for leverage, coercive or half-form of compulsory insurance aswell as strengthen the publicity on catastrophe insurance. The research goals more specific, that is, from the theoretical and empiricalaspects of catastrophe insurance needs to be a systematic in-depth study and research.On the theoretical research, this article at home, a system using game theory, rely ontools such as Berman and utility function equation to model the effect of catastropheinsurance needs analysis. And the empirical research is trying to make the Panelanalysis, because there is no dedicated data flood, Typhoon and earthquake insurance,limited empirical findings of universality. The use of experimental methods formeasurement of catastrophe insurance purchases at home for the first application.But because of the domestic demand for catastrophe insurance study less, andspecifically for study on catastrophe insurance demand in China is almost blank,because this data in empirical research, theoretical model’s applicability in China andalso there is a big shortage.更多还原