基于均值-CVaR投资组合优化模型实证分析

发布时间：2018-06-16 18:37

本文选题：投资组合 + VaR　；参考：《重庆大学》2014年硕士论文

【摘要】：投资组合理论研究的是如何进行投资决策，将资金按一定比例分配到不同的风险资产上，在获取一定收益的同时又达到分散风险的目的。理性的投资者以追求期望效用最大化为目标，，在可承受风险水平下获取最大收益或预期收益目标。风险与收益是息息相关、密不可分的。均值-方差模型(简称MV模型)是投资组合理论的基础，其他形式的投资组合模型大多以它为基础拓展而得。均值-方差模型以资产收益率的均值来反映投资收益，以收益率的方差来描述风险。 VaR和CVaR方法是近些年被广泛使用的风险测量方法。本文将CVaR方法与均值-方差模型相结合，构建了均值-CVaR模型，以CVaR代替方差刻画了收益率时间序列的尾部风险的大小，并通过线性规划求解方法，计算出最优的投资组合权重。本文还将均值-CVaR模型运用到国内的证券市场，实证分析结果表明均值-CVaR模型不仅对国内的证券市场适用，而且可有效地描述和分散投资组合的潜在风险。本文的选题具有一定的理论意义和实用价值。均值-CVaR模型既可以帮助投资者实现资产的最优合理配置，还可以对投资风险进行测量，并通过选择不同的置信水平来控制风险，得到不同的投资组合有效前沿。本文采用历史模拟法，将均值-CVaR模型运用到我国证券市场，进行投资组合优化的实证分析，并通过一定时期对模型效果的观察，验证了该模型在中国证券市场的适用性。
[Abstract]:Portfolio theory studies how to make investment decisions and allocate funds to different risk assets according to a certain proportion to achieve the goal of dispersing risks while obtaining certain income. Rational investors aim to maximize the expected utility and obtain the maximum return or expected return under the risk tolerance level. Risk and income are closely related and inseparable. Mean-Variance Model (MV Model) is the basis of portfolio theory, and most other forms of portfolio models are based on it. The mean-variance model reflects the return on investment by the average of the return on assets, and describes the risk by the variance of the return. VaR and Cvar are widely used risk measurement methods in recent years. In this paper, we combine the Cvar method with the mean-variance model, construct the mean-CVaR model, and use CVaR instead of variance to characterize the tail risk of the yield time series, and calculate the optimal portfolio weight by linear programming solution method. This paper also applies the mean-CVaR model to the domestic securities market. The empirical results show that the mean-CVaR model is not only applicable to the domestic securities market, but also can effectively describe and disperse the potential risks of the portfolio. The topic of this paper has certain theoretical significance and practical value. Mean-CVaR model can not only help investors achieve the optimal and reasonable allocation of assets, but also measure the investment risk, control the risk by choosing different confidence levels, and obtain the effective frontier of different portfolio. In this paper, the mean value CVaR model is applied to the stock market in China, and the empirical analysis of portfolio optimization is carried out, and the applicability of the model is verified by observing the effect of the model in a certain period of time.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F830.59;F224

【参考文献】