Markowitz模型与Black-Litterman模型比较研究——投资人情绪对资产组合的影响

Comparition Study of Markowitz Model and Black-Litterman Model:Investor Sentiment on Portfolio

由于马尔科维茨资产组合模型存在放大方差和对输入变量异常敏感以及无卖空约束下的无意义的权重问题,使均值方差模型很难在实际中应用。针对该模型的缺陷,利用BMA模型优化Black-Litterman模型观点加入方式,详细设定了模型观点矩阵,利用非贝叶斯法则构建了投资人对观念的置信度,明确计算了隐含均衡收益和上证指数数据,再次证实了投资人情绪对传统资产组合模型结果的影响,并从数理模型上阐释了两个模型所依据的理论基础。

The mean variance model is hardly used in real life because of the problems of variance amplification,unusual sensitiveness to the input variable and the meaningless weight issue based on the no short sales constraints existed in the Markowitz Portfolio Model. According to these defects of this model, Fisher Black and Robert Litterman（1992）set up a new model connecting with the investor＇s subjective expectation. This new model could improve the predictive ability of portifilio model. And the predictive ability is the inverstor＇s mood. In the theory of Behavioral Finance, the investor＇s mood could influence the portfolio selection. Thus in this paper, the author optimizes the B--L Model on the basis of BMA Model, then designs the model view matrix in detail. Moreover, this paper constructs the investor＇s view confidence level that counts the implied balanced return precisely. Meanwhile, with the reference of the Shanghai Composite Index data, the influence of investor＇s mood on the traditional Portfolio Model is verified further. At last, the theoretical basis of these two models are explained by the mathematical model.