【作者】 郭健；

【导师】 马超群；

【作者基本信息】 湖南大学 ， 管理科学与工程， 2011， 硕士

【摘要】 随着中国资本市场的快速发展，中国的证券投资基金规模不断扩大，品种日益丰富，成为证券市场中举足轻重的机构投资者。并且，证券投资基金绩效研究已经成为金融工程热点研究领域，这也对科学、客观的基金绩效评价方法提出了更为迫切的需求。以夏普指数、特内诺指数和詹森指数为代表的传统绩效评价方法都是以CAPM为理论基础，而CAPM理论本身存在诸多局限性。非常重要的一点在于资产收益的高阶矩特征被忽略，而大量理论研究和实证研究不仅证实了资产组合收益的非正态性，并且指出了高阶矩特征在投资效用中有着重要作用，致使传统评价方法的有效性受到质疑，由此本文试图提出一个更有效、更符合投资者效用偏好的基金绩效评价方法。首先，本文结合中国证券市场实际情况，总结回顾了国内外基金绩效的研究成果及各种观点，介绍了传统绩效评价方法的理论基础并分析了其在应用中存在的局限性。然后，针对CAPM理论的缺陷，本文详细阐述了资产收益的高阶矩特征在投资效用中的作用及经济意义，研究表明基金收益的高阶矩特征是一个非常重要的基金绩效评价影响因素。在此基础上，将数据包络分析(DEA)思想引入基金绩效评价，建立了一个基于高阶矩特征的基金绩效DEA评价模型。最后，由于开放式基金在数量与规模上已经远远超过了封闭式基金，日益成为基金业的主流，因此本文选择了27只开放式基金作为样本，在高阶矩框架下对基金绩效评价进行了实证分析，研究表明考虑高阶矩特征后，绩效得分会根据投资者效用偏好而发生变化；同时运用投影分析为非DEA有效的基金提升其绩效提供了未来的调整方向和程度；另一方面，采用三大指数作为对比研究，结果表明在总体绩效评价方面，高阶矩框架下的DEA方法与夏普指数、特内诺指数关系密切，但与詹森指数不相关；本文也通过考察不同市场状态下基金绩效，分析了基金抵御风险和把握市场机会能力。总的来说，基于高阶矩特征的基金绩效评价结果更切合实际情况，更符合投资者偏好。文章最后在总结实证研究结论的同时，结合当前绩效评价研究的发展趋势，提出了下一步工作的重心和方向。更多还原

【Abstract】 With the fast development of the securities market, the China securities fundshave become larger and more various, which have become important institutionalinvestors in the stock market. And, the research of funds performance evaluation arebecoming an important topic in the field of financial engineering. It is essential tofound a scientific and objective method of funds performance evaluation. Thetraditional methods of funds performance evaluation, such as Sharpe, Treynor andJensen, which are based on CAPM models, are limited in a numble of ways. The keypoint is that the higher moment of the assets return was typically neglected. However,a great amount of research, both theoretical and empirical, has supported the existenceof nonnormality of portfolio return and that the higher moment of the return hasplayed an important role in the investment utility. This has led to widespreadsuspicion of the valid for the traditional evaluation methodology. So, this paperexpects to propose a more effective performance evaluation method.Firstly, the paper reviews and summarizes the foreign and domestic research onfunds performance. Meanwhile, the traditional methods of performance evaluationand their theoretical basis are introduced in the paper and the limitations in theapplication are also analyzed. Then, to solve these problems, the higher momentcharacteristics of the assets return are described, including the role in the investors’utility. The research indicates that the higher moments are important factors on fundsperformance evaluation. Based on higher moments characteristics, I represent a newperformance evaluation models with data envelopment analysis (DEA).Finally, this paper choose27open-end funds, which have become the mainstream in China funds market, as samples of empirical research. The performance ofthe27funds are evaluated in the consideration of higher moments. The results showthat the evaluation score is related to the utility preference of the investors; and theprojection analysis of funds provide the direction and extent of adjustment fornon-efficient funds to improve their performance; on the other hand, comparing theDEA method in higher moments framework to three traditional indexes, thecorrelations with Sharpe index and Treynor index are significant, but not with Jensenindex; this paper also anlysis the abilities of funds to avoid the risk and seizeopportunities, via evaluate the performance of funds in different market trends. In general, the results of funds performance evaluation in higher momentsframework are more realistic and consistent with the investors’ preference.Summarising the results of empirical research and also cosidering the current trend offunds performance evaluation, the paper proposed the focus and directions of futureresearch.更多还原