基于非参数方法的社保基金股票收益率研究

发布时间：2018-02-14 23:08

本文关键词： 社保基金 VaR模型 ES模型非参数核密度估计方法　出处：《南京大学》2014年硕士论文　论文类型：学位论文

【摘要】：2008年,次贷危机在美国华尔街爆发,并最终发展成席卷全球的金融危机。我国经济发展受到了较大的冲击,长久以来备受质疑的养老金支付能力和隐形债务问题也再次被推到了风口浪尖。与此同时,全国社会保障基金(以下简称社保基金)作为养老金给付的后备基金,其投资收益问题也随之受到众多社会学者、民众和政策制定者的关注。作为备用金,社保基金暂时不需要承担和面临支付的压力,但是随着社保基金投资资本市场渠道和规模不断扩大,对投资回报的稳定高效性和风险测度的精准性也提出了更高的要求。本文所要探索的问题即是如何选用合适的方法和模型对我国社保基金股票投资的风险进行度量。在理论层面,本文首先简要介绍社保基金、投资风险及其度量模型、非参数核密度估计理论及方法,通过对风险度量模型的比较,本文认为基于非参数核密度估计的ES模型是适合社保基金投资风险度量模型,并利用K-Means聚类分析方法从另一角度对社保基金投资风险进行分析；在实证层面,本文的样本数据选自社保基金板块股票2013年第四个季度投资的648家涉及公司、18个行业的收益率,并对其进行分析,结果证明基于核密度估计的非参数方法计算出来的风险大小能够反映我国社保基金的股市投资风险点及风险大小,适用于比较分析社保基金不同行业的股票投资决策。同时,利用K-Means聚类分析方法得到的结论与核密度估计的非参数方法ES模型所得到的结论是一致的。
[Abstract]:In 2008, the subprime mortgage crisis broke out on Wall Street in the United States, and finally developed into a global financial crisis. The question of pension ability to pay and hidden debt, which has long been questioned, is again being pushed to the fore. Meanwhile, the National Social Security Fund (NSSF) acts as a reserve fund for pension payments. The issue of its investment returns has also attracted the attention of many social scholars, people and policy makers. As a reserve fund, the Social Security Fund does not need to bear and face the pressure to pay for the time being. But with the expansion of the social security fund's investment capital market channels and scale, Higher requirements are also put forward for the stability and efficiency of investment return and the accuracy of risk measurement. The problem this paper will explore is how to select appropriate methods and models to carry out the risk of social security fund stock investment in China. Quantity. At the theoretical level, Firstly, this paper briefly introduces social security fund, investment risk and its measurement model, nonparametric kernel density estimation theory and method, and compares the risk measurement model. This paper holds that the es model based on nonparametric kernel density estimation is suitable for the investment risk measurement model of social security fund, and uses K-Means cluster analysis method to analyze the investment risk of social security fund from another angle. The sample data of this paper are selected from the 648 companies that invested in the fourth quarter of 2013 in the Social Security Fund sector, which involved the returns of 18 industries, and analyzed them. The results show that the non-parametric method based on nuclear density estimation can reflect the investment risk point and risk of social security fund in China, which is suitable for comparing and analyzing the stock investment decision of different industries of social security fund. The results obtained by K-Means clustering method are consistent with those obtained by the non-parametric method es model for kernel density estimation.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F842.67;F832.51

【相似文献】