贝叶斯方法在私募基金业绩评价中的应用

发布时间：2018-01-12 03:01

本文关键词：贝叶斯方法在私募基金业绩评价中的应用　出处：《浙江财经学院》2012年硕士论文　论文类型：学位论文

【摘要】：近年来,随着我国证券投资基金业的迅速发展,如何正确分析和评价基金业绩,建立适用于中国证券市场的基金业绩评价方法和体系对投资者正确决策以及基金业的健康稳定发展具有重要的理论和现实意义。但是目前的基金业绩评价方法主要是针对公募基金,一般是根据历史数据采用传统的风险调整方法,而新方法却提出较少,传统方法尽管有利于评价的客观性,但是却没有重视投资者在实际决策时将自己关于基金业绩状况的的直觉和经验融入到其业绩评价当中。另一方面,作为证券市场新兴力量的私募基金在业绩评价方面尚处于起步阶段,发展相对滞后,也存在很多不足。基于以上原因,本文以传统基金业绩评价方法和体系为基础,将私募证券投资基金业绩评价作为研究核心,结合中国私募基金及业绩评价现状,研究了贝叶斯统计推断方法在私募基金业绩评价中的应用。与传统方法不同,该方法从投资者角度出发,强调了投资者先验信息对基金业绩评价的影响。本文按照基金业绩评价理论方法回顾——贝叶斯方法研究——建立贝叶斯业绩评价模型——实证分析的逻辑思路评价私募基金业绩,首先通过直觉问题引出先验信息并由此建立了一个关于基金管理技能的先验信息集合;然后将这些先验信息与Jensen测度模型相结合,进而导出模型中的截距项的后验期望代数解;最后运用贝叶斯业绩评价指标对我国2008年7月1日前成立的87只阳光私募基金样本进行实证分析(样本期为2008年7月1日到2011年6月30日)并全面总结分析结果。实证研究表明,贝叶斯方法不仅使我们看到先验信息如何影响投资者的最终决策,同时也可以较好的解决经典评价方法的结果很大程度受数据精度和显著性水平因素影响的问题。全文由以下五部分组成:第一章:深入分析国内外相关研究背景及现状,提出选题背景和意义。第二章:首先简要论述私募基金发展历史及现状,然后对基金业绩评价原理及现有业绩评价方法进行了回顾与小结。第三章:详细介绍贝叶斯业绩评价方法,主要包括介贝叶斯统计推断的思想及其本质,举例说明它和经典业绩评价方法的不同之处,建立相关模型并进行数学推导。第四章:对我国证券市场的阳光私募证券投资基金进行贝叶斯方法的实证分析并得出实证结果和结论。第五章,对全文进行总结并指出本文的不足。在研究方法上,本文采用了从理论到实证,定性分析和定量分析相结合的方法以及比较分析法等,定性分析的方法如对先验信息与业绩评价的经济环境的探讨等,定量分析的方法如似然率与先验分布模型的建立,先验分布的引出与后验分布模型的建立,以及以基准超额收益率为自变量,基金的超额收益率为因变量进行一元回归分析与显著性检验等。比较分析法,如将贝叶斯业绩评价方法与传统的频率方法的评价结果进行对比分析,说明其相对传统频率分析方法的优越性所在。本文的创新之处主要有以下几点:(1)近年来国内少有私募基金业绩评价方法的研究,本文可以算是这方面的一个尝试;(2)首次将贝叶斯模型用于私募基金业绩评价,充分利用了先验信息的作用;(3)对原有的贝叶斯方法的基金评价模型中的先验分布模型进行了改进,以使其更真实的反映现实情况。面对我国私募基金市场的快速发展和市场参与方对正确客观评价私募基金业绩水平的现实需要,本文的研究不仅展示了贝叶斯方法同经典业绩评价方法的不同之处,即投资者的先验信息会如何影响其投资决策,而且还为私募证券投资基金业绩评价提供了一个崭新的视角和方法,具有很强的理论价值,这可能会丰富目前的私募基金业绩评价理论,使我国私募基金的参与者受到启发,有助于引导社会资源更多的流向拥有成熟、理性投资理念的基金管理人,实证结论对于指导投资者进行投资决策也具有积极的现实意义。
[Abstract]:In recent years, with the rapid development of China's securities investment fund industry, how to correctly analyze and evaluate the performance of the fund, has important theoretical and practical significance to establish funds performance evaluation methods and system for China stock market steady development of investors to make decisions as well as the fund industry's health. But the present methods of fund performance evaluation is mainly aimed at the public offering of the fund, is generally based on historical data of risk adjusted using the traditional method, the new method has less traditional methods proposed, although conducive to the objectivity of the evaluation, but no attention in the actual investment decision when he condition on the performance of the fund's intuition and experience to the performance evaluation. On the other hand, as a rising power in the securities market, private equity funds in terms of performance evaluation is still in its infancy, the development is lagging behind, there are also many problems based on. Based on the above reasons, the traditional fund performance evaluation method and system as the basis, the private securities investment fund performance evaluation as the core of research, combined with the status quo of Chinese private fund and performance evaluation, study the Bayesian statistical inference method in evaluation of private equity fund performance should be used. Unlike the traditional method, this method from the perspective of investors, stressed the influence of investors prior information on fund performance evaluation.
In this paper, according to the evaluation of fund performance -- a review of Bias theory research on the methods of establishing Bias performance evaluation model -- An Empirical Analysis of the logic of private equity fund performance evaluation, the first problem by intuition a priori information and thus establish a fund management skill set of prior information; and then combine these prior information and Jensen model to measure the phase. And then export the model in the intercept of the posterior expectation of algebraic solution; finally using the Bias performance evaluation index to analyze the 87 sunshine private equity fund samples established in China before July 1, 2008 (sample period from July 1, 2008 to June 30, 2011) and a comprehensive summary of the analysis results. The empirical study shows that the Bias method not only enables us to see how the effect of prior information the final decision of the investors, but also can solve the classic evaluation well The result of the method is largely influenced by the accuracy of the data and the level of the level of significance.
This paper consists of five parts: the first chapter: in-depth analysis of relevant research background and current situation at home and abroad, put forward the background and significance. The second chapter: first briefly discusses the history and current situation of the development of private equity funds, then the principle of fund performance evaluation of existing performance evaluation methods for review and summary. The third chapter introduces the method of Bias: performance evaluation, including the idea and essence of Bias statistical inference, illustrate the difference between it and the classical performance evaluation methods, models and mathematical derivation. The fourth chapter is the empirical analysis of Bias of China's securities market sunshine private securities investment fund and draw empirical results and conclusion. In the fifth chapter. A summary of the thesis and points out the shortcomings.
In research methods, this paper from theory to empirical methods, qualitative analysis and quantitative analysis and comparative analysis, qualitative analysis methods such as the prior information and performance evaluation of the economic environment and so on, the quantitative analysis methods such as the likelihood ratio and the establishment of the prior distribution of the model, the prior distribution of lead with the establishment of the posterior distribution model, and the benchmark excess return rate as independent variables, the fund excess return rate as the dependent variable for regression analysis and significance test. The comparison analysis method, such as the evaluation method of Bias frequency performance evaluation method and the results were analyzed to illustrate the relative superiority the traditional frequency analysis method. The main innovations of this paper are as follows: (1) research in recent years, few domestic private equity fund performance evaluation method, this paper can be considered in this regard One attempt; (2) first, the Bias model was applied to the performance evaluation of private equity funds, making full use of the role of a priori information; (3) the prior distribution model of the original Bias's fund evaluation model was improved to make it more realistic.
We need to face the correct evaluation of private equity fund performance level of reality in the rapid development of the market of China's private equity fund and the market, this research not only shows the difference with the classical performance evaluation method of Bayesian method, in which investors prior information will be how to influence their investment decisions, but also provides a brand-new method the performance evaluation of private securities investment fund, has a strong theoretical value, which may enrich the current private equity fund performance evaluation theory, so that China's private equity fund participants were inspired to help guide social resources to more mature, rational investment fund management, the empirical conclusion also has the reality positive significance in guiding the investor to make investment decisions.

【学位授予单位】：浙江财经学院
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.51;F224

【参考文献】