PROST模型精算假设的一致性检验与改进
摘要
PROST(养老金改革可选方案模拟工具包)是由世界银行社会保障部开发的一个通用的、基于微机的预测模型。它以人口状况、年龄结构、宏观经济形势等统计资料作为依据,可以模拟未来75年的人口结构、养老保险金收支和隐性债务发展趋势等情况,为养老保险政策设计提供参考。PROST目前在多个国家运用,但长期缺乏必要的升级。本文以其采用的精算假设为研究对象,探究其合理性与一致性,以期提高模型的适用性与精确性。
本文首先采用了全面数据质量管理的观点,运用TDQM方法检验了PROST模型的精算假设的一致性,对模型假设中的多处不一致之处予以指出。然后分章节针对各处进行了修改。修改之处有死亡率假设、退休人口假设、缴费工资假设等等。为改进死亡率假设,本文分析并比较了各种修匀方法,最终采用了分段参数修匀法。与此同时,对各个修改处运用实际数据作了改进的检验。最后,本文运用质量控制的观点,将PROST作为软件进行了整体评价。
本文验证了PROST中数据管理的重要性,同时可为PROST日后升级提供参考。
PROST (Pension Reform Options Simulation Tool-kit) is a generic, computer-based forecasting model, which was developed by Department of Social Security, The World Bank. Based on demographic situations, age structures, macroeconomic situations and other statistical information, it could simulate future structure of population, pension payments and implicit debt of the next 75 years, and also provide reference for pension policy design. Now PROST is used in many countries, but it has been in Lack of necessary upgrades for a long time. This paper considers the rationality and consistency of actuarial assumptions of PROST, intended to improve the accuracy and applicability of this model.
This paper first adopts the view of Total Data Quality Management, and tests the consistency of actuarial assumptions of PROST using TDQM, then points out and improves the inconsistent assumptions. The thesis improves the mortality table assumption, amends the assumption of retired population, wage and contribution. Before improving the mortality table, the paper analyzes and compares 5 different graduation methods, and finally chooses spline graduation because of its accuracy and applicability. At the end of this article, PROST is evaluated as a software using the view of quality control.
The thesis stresses the importance of data quality management. I hope this thesis could dedicate to the upgrade of PROST in the future.
本文首先采用了全面数据质量管理的观点,运用TDQM方法检验了PROST模型的精算假设的一致性,对模型假设中的多处不一致之处予以指出。然后分章节针对各处进行了修改。修改之处有死亡率假设、退休人口假设、缴费工资假设等等。为改进死亡率假设,本文分析并比较了各种修匀方法,最终采用了分段参数修匀法。与此同时,对各个修改处运用实际数据作了改进的检验。最后,本文运用质量控制的观点,将PROST作为软件进行了整体评价。
本文验证了PROST中数据管理的重要性,同时可为PROST日后升级提供参考。
PROST (Pension Reform Options Simulation Tool-kit) is a generic, computer-based forecasting model, which was developed by Department of Social Security, The World Bank. Based on demographic situations, age structures, macroeconomic situations and other statistical information, it could simulate future structure of population, pension payments and implicit debt of the next 75 years, and also provide reference for pension policy design. Now PROST is used in many countries, but it has been in Lack of necessary upgrades for a long time. This paper considers the rationality and consistency of actuarial assumptions of PROST, intended to improve the accuracy and applicability of this model.
This paper first adopts the view of Total Data Quality Management, and tests the consistency of actuarial assumptions of PROST using TDQM, then points out and improves the inconsistent assumptions. The thesis improves the mortality table assumption, amends the assumption of retired population, wage and contribution. Before improving the mortality table, the paper analyzes and compares 5 different graduation methods, and finally chooses spline graduation because of its accuracy and applicability. At the end of this article, PROST is evaluated as a software using the view of quality control.
The thesis stresses the importance of data quality management. I hope this thesis could dedicate to the upgrade of PROST in the future.
引文
[1] CAS Data Management Educational Materials Working Party, “Actuarial I.Q. (Information Quality) ”, CAS E-Forum,Winter 2008
[2] Mortality Sub-Committee of the CMI, “Mortality Improvements and the Cohort Effect”, the Staple Inn Actuarial Society, 11 March 2003
[3] Richard Y. Wang, “A Product Perspective on Total Data Quality Management”, Communications of the ACM, February 1998/Vol. 41, No. 2
[4] “Calculus for Prost”, The Social Protection Team, The World Bank
[5] Macdonald AS, Cairns AG, Gwilt PL, “International comparison of recent trends in population mortality [J]”, British Actuarial Journal, 1998, 4 (16):1210
[6] London D. Graduation, “The revision of estimates [M]”, NY: ACTEX Publications, 2000
[7] John Turner, “Mortality Improvements and Pension Liabilities”, the Society of Actuaries, 2005
[8] Society of Actuaries, “Report of the Society of Actuaries: Mortality Improvement Survey Subcommittee”, http://www.soa.org, March 2003
[9] Emily S. Andrews, “Pension Reform and the Development of Pension Systems: An Evaluation of World Bank Assistance”, the World Bank, 2006
[10] “Final IAA Guidelines of Actuarial Practice for Social Security Programs”, IAA, January 1, 2003
[11] 刘惠,《软件开发过程中的质量控制问题》,海洋技术,1996年12月
[12] 郭小琼,《软件质量的控制与质量保证》,软件开发,2001年第5期
[13] 兰雨晴,高静,《软件质量度量与软件过程度量》,专论综述,2003年第九期
[14] 刘明亮,卢仿先,《修匀法的案例分析》,经济数学,1995年6月
[15] 张镱锂,张玮,摆万奇,李双成,郑度,《行政单元与自然地理单元之间的数据耦合方法初探——以青藏高原人口统计数据为例》,土地变化科学与生态建设,2008年
[16] 余芳东,《国外统计数据质量评价和管理方法及经验》,北京统计,2003年7月
[17] 姜作勤,《数据质量的研究与实践的现状及空间数据质量标准》,信息化论坛,2004年3月
[18] 李晓林,孙佳美,《生命表基础(中国精算师资格考试用书)》,中国财政经济出版社,2006
[19] 周江雄, 刘建华, 黎颖芳,《生命表的构造理论》,南开大学出版社,2001
[20] 宋立新,王洪曾,冯敬海,《死亡率的贝叶斯修匀与分段参数修匀》,中国现场统计研究会第十三届学术年会论文,2007年8月
[21] 王晓军,《社会保障精算原理》,中国人民大学出版社.2000
[22] 原俊青,杨兵,李泽慧,《养老保险的连续精算模型》.兰州大学学报(自然科学版),2003年6月:17～20
[23] 李荣敏,张立,侯文才,张慈恩,《上海市在职职工死亡率估计与改进的整体参数修匀法》,复旦学报(自然科学版),2002年10月
[24] 王军 等,《巧学巧用Excel 2003 VBA与宏》,电子工业出版社,2005
[25] 李南 等,《基于时间-区域序列的随机死亡率预测及对中国数据的应用》,人口研究,1995年04期:58～63
[26] 侯长荣、余松林、陈心广,《一种新的预测人群死亡率方法的应用》,中国卫生统计,2000年05期:304～306
[2] Mortality Sub-Committee of the CMI, “Mortality Improvements and the Cohort Effect”, the Staple Inn Actuarial Society, 11 March 2003
[3] Richard Y. Wang, “A Product Perspective on Total Data Quality Management”, Communications of the ACM, February 1998/Vol. 41, No. 2
[4] “Calculus for Prost”, The Social Protection Team, The World Bank
[5] Macdonald AS, Cairns AG, Gwilt PL, “International comparison of recent trends in population mortality [J]”, British Actuarial Journal, 1998, 4 (16):1210
[6] London D. Graduation, “The revision of estimates [M]”, NY: ACTEX Publications, 2000
[7] John Turner, “Mortality Improvements and Pension Liabilities”, the Society of Actuaries, 2005
[8] Society of Actuaries, “Report of the Society of Actuaries: Mortality Improvement Survey Subcommittee”, http://www.soa.org, March 2003
[9] Emily S. Andrews, “Pension Reform and the Development of Pension Systems: An Evaluation of World Bank Assistance”, the World Bank, 2006
[10] “Final IAA Guidelines of Actuarial Practice for Social Security Programs”, IAA, January 1, 2003
[11] 刘惠,《软件开发过程中的质量控制问题》,海洋技术,1996年12月
[12] 郭小琼,《软件质量的控制与质量保证》,软件开发,2001年第5期
[13] 兰雨晴,高静,《软件质量度量与软件过程度量》,专论综述,2003年第九期
[14] 刘明亮,卢仿先,《修匀法的案例分析》,经济数学,1995年6月
[15] 张镱锂,张玮,摆万奇,李双成,郑度,《行政单元与自然地理单元之间的数据耦合方法初探——以青藏高原人口统计数据为例》,土地变化科学与生态建设,2008年
[16] 余芳东,《国外统计数据质量评价和管理方法及经验》,北京统计,2003年7月
[17] 姜作勤,《数据质量的研究与实践的现状及空间数据质量标准》,信息化论坛,2004年3月
[18] 李晓林,孙佳美,《生命表基础(中国精算师资格考试用书)》,中国财政经济出版社,2006
[19] 周江雄, 刘建华, 黎颖芳,《生命表的构造理论》,南开大学出版社,2001
[20] 宋立新,王洪曾,冯敬海,《死亡率的贝叶斯修匀与分段参数修匀》,中国现场统计研究会第十三届学术年会论文,2007年8月
[21] 王晓军,《社会保障精算原理》,中国人民大学出版社.2000
[22] 原俊青,杨兵,李泽慧,《养老保险的连续精算模型》.兰州大学学报(自然科学版),2003年6月:17～20
[23] 李荣敏,张立,侯文才,张慈恩,《上海市在职职工死亡率估计与改进的整体参数修匀法》,复旦学报(自然科学版),2002年10月
[24] 王军 等,《巧学巧用Excel 2003 VBA与宏》,电子工业出版社,2005
[25] 李南 等,《基于时间-区域序列的随机死亡率预测及对中国数据的应用》,人口研究,1995年04期:58～63
[26] 侯长荣、余松林、陈心广,《一种新的预测人群死亡率方法的应用》,中国卫生统计,2000年05期:304～306
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |