摘要
传统的准备金估计方法主要基于流量三角形,通过稳定的发展因子预测未决赔款准备金。流量三角形中近期事故的准备金是未决赔款准备金中最大的部分,其受发展因子的影响也是最显著的,因此所有发展因子的稳定性对近期未决赔款准备金的估计尤为重要。但由于大额赔付的影响,发展因子的稳定性受到很大挑战;由于保单生效日期的不同,近期事故的暴露单位一般不同,相应赔付额度也将不同,但传统的准备金估计方法基本不考虑暴露单位的影响。为解决上述问题,本文提出了一种近期未决赔款准备金估计方法,充分利用个体保单的赔付额和暴露单位信息,将保单按风险强度分类,避免由于赔付额差异大造成的准备金预测不准确等问题,并进一步给出保单组的未决赔款准备金汇总方法;最后,基于一组车险的索赔数据,比较了经典方法和本文提出的新方法在近期未决赔款准备金估计上的表现,说明本文构建的方法的可行性和有效性。
The traditional method for reserve estimation is mainly based on the liquidity triangle and predicts the outstanding reserves with stable development factors.The reserves for latest accidents in the triangle are the largest part of the outstanding reserves,and they are also most significantly affected by development factors.Therefore,the stability of all development factors is particularly important for the estimation of the latest reserves.However,the stability of development factors is greatly challenged by the impact of large losses.Furthermore,the number of risk exposures of recent accidents is generally different due to the different effective dates of the insurance policies,and the corresponding loss amounts will also be different.The traditional method does not consider the impact of the number of exposures.In order to solve these problems,this paper proposes a new method,which makes full use of the loss amount and exposure information of individual insurance policies.It classifies the insurance policies according to the risk intensity,avoiding the inaccurate reserve prediction caused by the large difference of loss amount,and further offers the method of summarizing the outstanding reserves of insurance policies.Finally,based on a set of vehicle insurance claim data,the performances of the classical method and the new method proposed in this paper in the latest reserve estimation are compared to illustrate the feasibility and effectiveness of the new method.
出处
《保险研究》
CSSCI
北大核心
2023年第1期43-50,共8页
Insurance Studies
基金
西南财经大学科技创新项目“基于机器学习的保单分组与准备金评估”(JBK210301)
西南财经大学社科项目“机器学习框架下基于个人驾驶里程行为的新能源车险的定价研究”(JBK2203004)。
关键词
近期事故
未决赔款准备金
泊松过程
风险分类
latest accidents
outstanding reserve
Poisson Process
policy classification