基于非参数方法的火灾公众责任保险定价研究
发布时间：2018-10-08 10:12
【摘要】：近年来,火灾事故频发,给社会带来了巨大的损失。一场场公众场所火灾事故催生出人们对火灾公众责任保险的迫切需求。但目前,我国火灾公众责任保险承保范围窄、定价技术不够发达等问题的存在,使得火灾公众责任保险的发展困难重重。本文以火灾公众责任保险为视角,探讨了我国火灾公众责任保险定价问题,为火灾公众责任保险科学合理定价提供了思路。 本文在研究火灾公众责任保险定价之前,首先明确了火灾公众责任的承保范围,在总结国外发展经验的基础上,分析了我国火灾公众责任保险的发展现状及存在的问题,为我国火灾公众责任保险的发展提供了思路。为保证火灾公众责任保险定价理论的科学性和先进性,本文系统梳理了目前发展比较成熟且相对先进的责任保险定价理论,通过比较分析,选择发展最为成熟且假设条件少的精算定价理论作为火灾公众责任保险定价方法,并呈献出火灾公众责任保险的整个定价过程。 在利用精算方法对火灾公众责任保险进行定价时,正确的损失分布拟合方法是损失分布拟合精度的重要保证。在对火灾损失数据进行拟合时,本文引进了非参数损失分布拟合方法,与参数损失分布拟合方法相比,该方法并不事先假定损失数据服从特定分布,而是从样本数据自身特点出发,使拟合出的损失分布更加符合现实分布,保证了损失分布的精确性及保费的合理性。在利用非参数核密度估计方法拟合了损失分布后,得到了火灾损失的期望损失均值,进而得到纯保费。 在此基础上,考虑到火灾公众责任保险的特殊性,结合不同行业、营业面积、建筑物结构及防火设施等因素对总保费的影响,对火灾公众责任保险实行差别费率,这使得测算结果更加科学合理。 为验证该定价方法的可行性,本文以我国批发零售场所火灾事故损失数据为对象,进行了实证研究。 本文在以下几个方面进行了创新性探讨： (1)在对火灾公众责任保险损失分布进行拟合时,创新性的引入了非参数核密度估计方法。该方法并不事先假设样本数据服从特定的分布,而是直接从样本数据的特点出发拟合损失分布,使得拟合结果与实际分布更加吻合,更具实际利用价值,克服了精算定价方法中利用参数方法拟合损失分布的缺陷。 (2)完善了火灾损失分布赔付标准。在处理实证数据时,对伤亡人员赔付标准进行细化,保证了损失赔付的准确性。 (3)检验了非参数方法在火灾公众责任保险中的适用性,拓展了火灾公众责任保险的定价思路。以批发零售业火灾为研究对象,给出了火灾公众责任保险的定价过程。
[Abstract]:In recent years, frequent fire accidents have brought huge losses to the society. A fire accident in public places has given rise to the urgent need for public liability insurance in fire. However, at present, the insurance coverage of fire public liability insurance in China is narrow, pricing technology is not developed, and so on, which makes the development of fire public liability insurance difficult. From the perspective of fire public liability insurance, this paper discusses the pricing problem of fire public liability insurance in China, which provides a way of thinking for the scientific and reasonable pricing of fire public liability insurance. Before studying the pricing of fire public liability insurance, this paper firstly clarifies the coverage scope of fire public liability insurance. On the basis of summarizing the development experience of foreign countries, this paper analyzes the present situation and existing problems of fire public liability insurance in China. It provides ideas for the development of fire public liability insurance in China. In order to ensure the scientific and advanced nature of the pricing theory of fire public liability insurance, this paper systematically combs the relatively mature and relatively advanced liability insurance pricing theory, and through comparative analysis, The most mature and hypothetical actuarial pricing theory is chosen as the pricing method of fire public liability insurance, and the whole pricing process of fire public liability insurance is presented. When the fire public liability insurance is priced by actuarial method, the correct fitting method of loss distribution is an important guarantee for the accuracy of loss distribution fitting. When fitting the fire loss data, this paper introduces the non-parametric loss distribution fitting method. Compared with the parametric loss distribution fitting method, the method does not presuppose the loss data from a specific distribution. On the basis of the characteristics of the sample data, the distribution of the loss is more in line with the actual distribution, which ensures the accuracy of the loss distribution and the rationality of the premium. After the loss distribution is fitted by the nonparametric kernel density estimation method, the expected loss mean of fire loss is obtained, and then the pure premium is obtained. On this basis, considering the particularity of fire public liability insurance, combined with different industries, business area, building structure and fire prevention facilities and other factors on the total premium, the fire public liability insurance rate is different. This makes the results more scientific and reasonable. In order to verify the feasibility of the pricing method, this paper makes an empirical study on the fire loss data of wholesale and retail establishments in China. In this paper, the following aspects are discussed: (1) when fitting the loss distribution of fire public liability insurance, the nonparametric kernel density estimation method is innovatively introduced. This method does not presuppose that the sample data is distributed according to the specific distribution, but directly matches the loss distribution according to the characteristics of the sample data, which makes the fitting result more consistent with the actual distribution and has more practical use value. It overcomes the defect of using parameter method to fit the loss distribution in actuarial pricing method. (2) the standard of fire loss distribution and compensation has been improved. In dealing with empirical data, the standard of compensation for casualties is refined to ensure the accuracy of compensation for loss. (3) the applicability of non-parametric method in fire public liability insurance is tested, and the pricing thinking of fire public liability insurance is extended. Taking wholesale and retail fire as the research object, the pricing process of fire public liability insurance is given.
【学位授予单位】：中国海洋大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F842.69
本文编号：2256387
[Abstract]:In recent years, frequent fire accidents have brought huge losses to the society. A fire accident in public places has given rise to the urgent need for public liability insurance in fire. However, at present, the insurance coverage of fire public liability insurance in China is narrow, pricing technology is not developed, and so on, which makes the development of fire public liability insurance difficult. From the perspective of fire public liability insurance, this paper discusses the pricing problem of fire public liability insurance in China, which provides a way of thinking for the scientific and reasonable pricing of fire public liability insurance. Before studying the pricing of fire public liability insurance, this paper firstly clarifies the coverage scope of fire public liability insurance. On the basis of summarizing the development experience of foreign countries, this paper analyzes the present situation and existing problems of fire public liability insurance in China. It provides ideas for the development of fire public liability insurance in China. In order to ensure the scientific and advanced nature of the pricing theory of fire public liability insurance, this paper systematically combs the relatively mature and relatively advanced liability insurance pricing theory, and through comparative analysis, The most mature and hypothetical actuarial pricing theory is chosen as the pricing method of fire public liability insurance, and the whole pricing process of fire public liability insurance is presented. When the fire public liability insurance is priced by actuarial method, the correct fitting method of loss distribution is an important guarantee for the accuracy of loss distribution fitting. When fitting the fire loss data, this paper introduces the non-parametric loss distribution fitting method. Compared with the parametric loss distribution fitting method, the method does not presuppose the loss data from a specific distribution. On the basis of the characteristics of the sample data, the distribution of the loss is more in line with the actual distribution, which ensures the accuracy of the loss distribution and the rationality of the premium. After the loss distribution is fitted by the nonparametric kernel density estimation method, the expected loss mean of fire loss is obtained, and then the pure premium is obtained. On this basis, considering the particularity of fire public liability insurance, combined with different industries, business area, building structure and fire prevention facilities and other factors on the total premium, the fire public liability insurance rate is different. This makes the results more scientific and reasonable. In order to verify the feasibility of the pricing method, this paper makes an empirical study on the fire loss data of wholesale and retail establishments in China. In this paper, the following aspects are discussed: (1) when fitting the loss distribution of fire public liability insurance, the nonparametric kernel density estimation method is innovatively introduced. This method does not presuppose that the sample data is distributed according to the specific distribution, but directly matches the loss distribution according to the characteristics of the sample data, which makes the fitting result more consistent with the actual distribution and has more practical use value. It overcomes the defect of using parameter method to fit the loss distribution in actuarial pricing method. (2) the standard of fire loss distribution and compensation has been improved. In dealing with empirical data, the standard of compensation for casualties is refined to ensure the accuracy of compensation for loss. (3) the applicability of non-parametric method in fire public liability insurance is tested, and the pricing thinking of fire public liability insurance is extended. Taking wholesale and retail fire as the research object, the pricing process of fire public liability insurance is given.
【学位授予单位】：中国海洋大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F842.69
【参考文献】
相关期刊论文 前10条
1 杨新华;试论我国公众责任保险的发展[J];琼州大学学报;2005年01期
2 吕迎;周文兴;;CAPM模型的非参数估计理论[J];商场现代化;2009年10期
3 姜利琴;关于大力发展上海市公众责任保险之我见[J];上海保险;2005年01期
4 王晓丽;席金平;吴润衡;;基于Copula函数和非参数估计的尾部相关性[J];数学的实践与认识;2009年07期
5 屈文建;熊国经;;非参数密度估计法比较分析及应用[J];沈阳农业大学学报;2008年04期
6 毕晓辉,张维;APT 理论的非参数解释[J];天津大学学报;1998年04期
7 区诗德;邢国东;张江红;;股票收益率密度的非参数估计及投资策略[J];统计与决策;2006年05期
8 赵丽琴;籍艳丽;;Copula函数的非参数核密度估计[J];统计与决策;2009年09期
9 陈希镇;胡兆红;;Copula函数的非参数核密度估计方法[J];统计与决策;2010年14期
10 镇志勇;李军;;非参数核密度估计在恒生指数收益率分布中的应用[J];统计与决策;2011年09期
，本文编号：2256387
本文链接：https://www.wllwen.com/jingjilunwen/bxjjlw/2256387.html
