【作者】 许星剑；

【导师】 安实；

【作者基本信息】 哈尔滨工业大学 ， 管理科学与工程， 2008， 博士

【摘要】 随着全球老龄化危机的到来,社会保障以及社会保障基金的研究成为现代金融学研究的热点和难点之一。其核心问题是如何达到有效保值增值的目标来应对老龄化危机。全国社会保障基金是中央政府集中的社会保障资金,是国家重要的战略储备,主要用于弥补今后人口老龄化高峰时期的社会保障需要。全国社会保障基金股票投资组合是实现保值增值的有效途径之一,是具有代表性意义的投资。投资组合绩效评估理论则是针对投资组合的时机运作成果进行评估,这是投资理论的重要组成部分。投资绩效评估(performance measurement)是衡量投资管理的重要环节,将投资组合绩效评价和全国社保基金股结合起来进行全面、系统地分析全国社保基金在最初阶段的业绩表现具有前瞻性和实际指导意义。投资组合绩效评价是投资组合管理理论的一个重要分支,作为一种反馈机制,尽管不能完全预测未来,但是通过对历史业绩的评价和未来投资决策的制定,它的确是开拓了一条在未来实现更好业绩的途径。全国社会保障基金发展迅猛,但是国内对社保基金绩效评价方法和评价机制发展方面则非常滞后,严重阻碍了社保基金的进一步发展。建立有效适用的全国社会保障基金绩效评价体系具有重要的现实意义。因此,本文将全国社会保障基金股票投资组合实际情况与投资组合理论、投资组合绩效评价理论紧密结合,按照理论的发展脉络,分别选取了不同指标,分析了风险调整收益方面的全国社会保障基金股票投资组合的整体性收益情况、绩效归属来源以及持续性表现。首先,利用单因素指标对全国社保基金进行整体收益评价。本文从介绍基准组合的选择出发,分别对Sharpe、Treynor、Jensen、M2等指标进行实证检验。在此基础上,针对资产收益率序列实际具有的序列具有尖峰、厚尾、有偏的特性,利用Copula拟合实际分布得到了VaR值,提出了修正的Sharpe指数,并进行了实证研究。其次,采用了对基准组合有显著影响的持股比例作为先决信息变量对全国社保基金进行绩效归属分析,对绩效时变特征的条件性参数和非条件性参数进行检验,着重对社保基金管理人的择时和选股能力进行实证研究,并且利用广义矩估计法可以减轻利用离散数据估计连续时间模型给残差项分布带来的不利影响的特性对实证结果进行估计。结果显示全国社会保障基金股票投资组合在整体上没有显示出良好的预测市场走势的能力,但显示出来一定的正的选股能力。最后,探讨了股票投资组合投资业绩的持续性,利用改进的回归系数法实证发现样本组合在总体上业绩不存在持续性,但是在特定期间内存在着一定的业绩持续性,但呈现了反转性。这在理论上检验和支持了市场有效性理论及其经济内涵。本文将投资组合理论、绩效衡量方法和评价理论三者紧密结合,在绩效评价核心领域内对全国社会保障基金股票投资组合进行探索性研究,旨在为全国社保理事会如何正确衡量与评价全国社保基金股票投资组合的绩效表现提供理论与应用的支持,其成果有望对于遴选合格的优秀投资管理人、实现分散风险和收益最大化方面做出贡献,最终对全国社保理事会股票投资实现保值增值目标提供理论和方法的借鉴。更多还原

【Abstract】 With the global population aging, social security and social security fund now become one of the difficult and hot issues in scientific research. Its ultimate target is how to solve the population aging crisis with keeping value and increment of the social security funds. NSSF aims to be a solution to the problem of aging and serves as a strategic reserve fund accumulated by the central government to support future social security expenditures.Performance evaluation theory estimates actual operation result of portfolio and as one the most important application part. The fund performance measurement has both theoretical and practical concerns,it is the key issue of performance evaluation theory. China social security fund has just been established less than 5years, although there are some difficulties we have to face,such as few sample portfolios financial data available and short evaluation period. It is still necessary and meaningful to comprehensively and systematically analyze the China social security funds in its initial stage. This empirical study mainly employs the econometric approaches and quantitative analyze. The 3 main parts of the paper are as follows:Firstly, the paper measure the total performance of the NSSF stock portfolios,applying 5 various single factor indexes,the results illustrate that the portfolios performanve as a whole is better when compared to the benchmark index. And, the results indicate that the returns are more affected by the benchmark chosen.all in all,the Sharpe ratio and M2 measure are rather meaningful and practical apporoaches than the Treynor index and Jensen index. The return of portfolio is disobedient normally distribution and has obvious peak and fat tail, In empirical study Copula theory is applied to research the heavy-tail characteristic of the joint distribution of the Shanghai and Shenzhen stock market .and considering NSSF more concern downside risk, so put forward a modified sharpe index on the basis of VaR, Subsequently, its efficiency poved.Secondly,this section is about performance attribution analysis.focusing on the security selectivity and market timing of the portfolios with different models,especially using the predetermined variable of ownership percentage to measure the conditional model. The empirical study shows that the SSF stock portfolios as a whole do not exhibit excellent skills for forecasting the market trend,as the most results obtained both home and abroad, the evidence shows that the SSF stock portfolios have positive security selectivity coefficient and negative market timing coefficient, althouth both not significant by the large.and ,the results showsthat the conditional model is preferable than the unconditional model, but the divergence between the conditional model and the total-conditional model is not that distinctly.Finally,the study investigates the performance persistence .there is no legible evidence support the obvious persistence of the SSF stock portfolios,but the study shows that the sample has certain persistence in the medium term(6、9、12months),in the short term and long term ,the study presents reversion.更多还原