基于因子随机选择的非寿险准备金进展法

发布时间：2018-06-17 22:00

本文选题：未决赔款准备金 + 进展法　；参考：《天津财经大学》2014年硕士论文

【摘要】：非寿险业务责任准备金的评估十分重要,一直是保险公司面临的重要课题。进展法是评估未决赔款准备金的重要方法。但进展法评估准备金的过程中,未来准备金进展率和支付率的选择是确定的数值,使得这两个因子的选择带有主观性,最终得到的未决赔款准备金仅仅是一个点估计值,不能提供准备金的波动信息。为了度量进展法计提准备金的不确定性,本文提出进展率和支付率这两个因子随机选择的进展法。基于因子随机选择的进展法是将未来的进展率和支付率视为随机变量,分别从相应进展年的历史进展率和支付率的分布中随机选择一个值,分别作为该进展年的未来进展率和支付率,采用蒙特卡洛模拟得到无数个准备金进展率和支付率流量三角形,然后采用进展法可以得到无数个未决赔款准备金总额,最终得到准备金的预测标准误差和预测分布。使得进展法具有随机性的特征,从而完善了进展法的功能和实用性。在基于因子随机选择进展法的过程中,当历史进展率和支付率的分布为离散时,称为基于离散因子随机选择的进展法；当历史进展率和支付率的分布为连续时,称为基于连续因子随机选择的进展法。基于连续因子随机选择的进展法采用了核密度估计方法,核密度估计是用来估计给定样本点的密度函数,该方法不利用数据的先验信息,对数据的分布也不做任何假定,是一种从数据本身出发研究分布特征的方法,是一种典型的非参数估计方法。
[Abstract]:The evaluation of non-life insurance liability reserve is very important and has always been an important subject faced by insurance companies. Progress method is an important method to evaluate the reserve for outstanding claims. However, in the process of evaluating the reserve by the progress method, the choice of the rate of future reserve progress and the rate of payment is a certain value, which makes the choice of these two factors subjective, and the resulting reserve for outstanding claims is only a point estimate. Unable to provide information on fluctuations in reserves. In order to measure the uncertainty of the progress method in the calculation of reserve, this paper proposes the progress method of random selection of two factors, progress rate and payment rate. The progress method based on factor random selection considers the future progress rate and the payment rate as random variables, and selects a value randomly from the distribution of historical progress rate and payment rate of the corresponding progress year. As future progress rates and payment rates for that year, Monte Carlo simulations are used to obtain numerous reserve rate of progress and payment rate flow triangles, and then, using the progress method, the total number of outstanding claims reserves is obtained, Finally, the prediction standard error and forecast distribution of reserve are obtained. The progressive method has the randomness characteristic, thus consummates the function and the practicability of the progressive method. In the process of factor-based stochastic selection progress method, when the distribution of historical progression rate and payment rate is discrete, it is called the progress method based on discrete factor random selection, and when the distribution of historical progress rate and payment rate is continuous, It is called the progressive method based on continuous factor random selection. The progressive method based on continuous factor random selection uses kernel density estimation, which is used to estimate the density function of a given sample point. The method does not utilize the prior information of the data, nor does it make any assumptions about the distribution of the data. It is a typical nonparametric estimation method.
【学位授予单位】：天津财经大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F840.4

【相似文献】