ESG模拟平台的构建与分红保单负债评估

发布时间：2018-04-28 19:37

本文选题：蒙特卡洛模拟 + 经济情景发生器　；参考：《湖南大学》2014年硕士论文

【摘要】：保险公司是负债经营的金融机构,传统的寿险保单借助生命表和定价假设就能够较好的进行负债评估,但分红保单由于涉及到投资收益率和金融市场波动,需要考虑其特有的市场风险。而分红险在我国保险市场占有不小的份额,因此如何准确地对分红险进行评估成为我国乃至全球寿险业健康发展的关键因素。传统的偿付能力评估模型苦于没有较好的经济指标预测方法,仅仅以保费收入在未来年份的保障收益率贴现作为市场风险的测度,忽略了不同投资组合对负债变动的影响,对分红保单的负债评估并不准确,这为我国保险业风险监管埋下了隐患。而经济情景生成器(ESG)作为蒙特卡洛模拟方法在精算评估与风险管理领域的实现手段,凭借其强大的仿真与预测能力,能够较好的模拟金融资产在未来年份的波动状况,弥补传统评估方法的缺点。本文论述了课题的研究背景和意义,在新一代偿付能力监管框架下,根据风险模块细分的原则,首先分析了传统的保险风险模块下的保单负债评估方法,然后详细阐述了利率期限结构的计算与参数估计、相关系数和不同的风险测度等要素,并在此基础上探讨市场风险模块下内含最低保障收益的特殊保单的负债最优估计方法,随后引入了蒙特卡洛模拟法的理论知识和基于该方法搭建的ESG,并详细梳理了构建ESG平台的步骤与算法。本文以股票、债券和不动产三类资产作为资产组合,采用Hull-White模型进行拟合即参数估计,同时按照三类资产的相关系数构建ESG模拟平台,最后对一份基于标准假设的分红保单进行完整的负债评估,计算保单的最优负债估计。研究结果表明,基于ESG模拟的评估方法能够测度投资组合差异对负债的影响,克服了传统评估方法单纯以保费收入为评估标准的缺陷,提高了分红保单负债评估的准确性。
[Abstract]:Insurance companies are financial institutions operating in debt. Traditional life insurance policies can better evaluate liabilities by means of life tables and pricing assumptions, but dividend policies involve investment returns and financial market fluctuations. Need to consider its unique market risk. The dividend insurance has a large share in the insurance market in China, so how to accurately evaluate the dividend insurance has become the key factor for the healthy development of the life insurance industry in China and the whole world. The traditional solvency assessment model suffers from the lack of a better economic index forecasting method. It only takes the discount of premium income in the future year as the measure of market risk, neglecting the influence of different investment portfolio on the change of debt. The liability assessment of dividend policy is not accurate, which lays hidden trouble for insurance risk supervision in our country. As a method of Monte Carlo simulation in the field of actuarial evaluation and risk management, ESGG can better simulate the volatility of financial assets in the future year by virtue of its powerful simulation and forecasting ability. Make up for the shortcomings of traditional evaluation methods. This paper discusses the research background and significance of the subject. Under the framework of the new generation solvency supervision, according to the principle of risk module segmentation, this paper first analyzes the traditional insurance risk module under the policy liability assessment method. Then, the calculation and parameter estimation of term structure of interest rate, correlation coefficient and different risk measures are discussed in detail. On this basis, the optimal method for estimating the liabilities of special insurance policies with minimum guaranteed income under the market risk module is discussed. Then the theoretical knowledge of Monte Carlo simulation method and the ESG based on this method are introduced, and the steps and algorithms of constructing ESG platform are discussed in detail. In this paper, stock, bond and real estate assets are used as portfolio, Hull-White model is used to fit and estimate parameters, and ESG simulation platform is constructed according to the correlation coefficient of three kinds of assets. Finally, a dividend policy based on standard assumption is evaluated and the optimal liability estimate is calculated. The results show that the evaluation method based on ESG simulation can measure the impact of portfolio differences on liabilities, overcome the shortcomings of traditional evaluation methods based on premium income, and improve the accuracy of dividend policy liability assessment.
【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F842

【相似文献】