基于VaR和CVaR约束的投资组合模型

发布时间：2018-01-02 07:43

  本文关键词：基于VaR和CVaR约束的投资组合模型　出处：《山东大学》2013年硕士论文　论文类型：学位论文

  更多相关文章： VaR CVaR 风险约束 投资组合

【摘要】：投资组合选择(Portfolio Selection)是指把财富分配到不同的资产中,以达到分散风险、确保收益的目的。1952年,Markowitz用方差来量化股票收益的风险,提出了投资组合选择的均值-方差分析方法,揭开了现代金融学研究的序幕。 VaR与CVaR风险度量方法有多方面的应用,如信用风险的测量、内部风险资本金的确定、资本配置、金融监管等。VaR与CVaR作为一种测量投资组合风险的新方法近几年来迅速发展。目前许多学者从定义、性质、计算等方面对VaR和CVaR之间进行了比较分析,还有的比较了均值VaR、均值一CVaR模型的不同之处,但没有对基于VaR、CVaR约束的投资组合模型进行实证研究。 本文运用理论研究与实证研究相结合的方法。全文共分五章,第一章是研究背景及意义,并综述了国内外的研究概况。第二章和第三章分别从总体上介绍了VaR与CVaR风险度量方法的定义、性质及其相应的均值-VaR和均值-CVaR模型,并对均值-CVaR模型的边界与有效前沿进行了探讨,给出了均值-CVaR模型的数学表述与图形形状。第四章从理论上对VaR和CVaR约束下的投资组合进行比较,并结合几何知识给出了一种较为简单的计算方法。第五章根据上一章建立的模型进行实证分析,利用几何方法通过MATLAB编程进行求解,得出了不同置信区间下的投资组合权重,然后根据VaR和CVaR作为约束条件所得出的不同权重在我国证券市场进行实际模拟投资,分析比较所得结果,得出CVaR比VaR更能体现投资组合的潜在风险,更具有安全性的结论。最后,讨论了我国CVaR应用研究的问题,并提出了投资组合研究的进一步发展方向。
[Abstract]:Portfolio selection refers to the distribution of wealth among different assets for the purpose of diversifying risk and securing returns. 1952. Markowitz quantifies the risk of stock returns with variance, and puts forward the mean-variance analysis method of portfolio selection, which opens the prelude of modern finance research. VaR and CVaR risk measurement methods have many applications, such as the measurement of credit risk, the determination of internal risk capital, and the allocation of capital. As a new method to measure portfolio risk, financial supervision and CVaR have developed rapidly in recent years. Calculation and other aspects of the comparative analysis between VaR and CVaR, there is a comparison of the mean VaR, the difference between the mean-#en2# model, but not on the basis of VaR. The CVaR-constrained portfolio model is an empirical study. The thesis is divided into five chapters by combining theoretical research with empirical research. The first chapter is the background and significance of the research. Chapter two and chapter three introduce the definition of VaR and CVaR respectively. The properties and the corresponding mean-VaR and mean-CVaR models are discussed, and the boundary and efficient frontier of the mean-CVaR model are discussed. The mathematical representation and figure shape of the mean value CVaR model are given. Chapter 4th compares the portfolio under the constraints of VaR and CVaR theoretically. Combined with geometric knowledge, a relatively simple calculation method is given. Chapter 5th carries on empirical analysis according to the model established in the previous chapter, and uses geometric method to solve it by MATLAB programming. The portfolio weights under different confidence intervals are obtained, and then according to the different weights obtained by VaR and CVaR as constraint conditions, the actual simulated investment in China's securities market is carried out, and the results are analyzed and compared. It is concluded that CVaR can reflect the potential risk of portfolio and is more secure than VaR. Finally, the application research of CVaR in China is discussed. The further development direction of portfolio research is also put forward.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F830.59;F224

【参考文献】

相关期刊论文 前4条

1 王春峰,李刚;基于分布拟合法的VaR估计[J];管理工程学报;2002年04期

2 赵睿,赵陵;VaR方法与资产组合分析[J];数量经济技术经济研究;2002年11期

3 田新民,黄海平;基于条件VaR(CVaR)的投资组合优化模型及比较研究[J];数学的实践与认识;2004年07期

4 曲圣宁,田新时;投资组合风险管理中VaR模型的缺陷以及CVaR模型研究[J];统计与决策;2005年10期

，

本文编号：1368330