摘要
运用收集到的道琼斯中国88指数股票数据,基于随机矩阵理论对相关系数进行"去噪",构建金融网络,以关键社团为研究对象,分别应用"去噪"前后的相关系数矩阵构建投资组合的均值-方差模型,并且对两种投资组合模型的风险进行比较。结果表明:在相同的收益下,"去噪"后构建的投资组合模型的风险要小于"去噪"前构建的投资组合模型的风险,因此,"去噪"后构建的投资组合模型更有利于在投资决策中分散风险,从而能更有效地对我国股票投资组合风险进行控制。
Dow Jones 88 China Index stock data were investigated.The correlation coefficient matrix was improved by the denoising method based on random matrix theory.The financial network model was built.With key community as the research subjects,the mean-variance portfolio selection model was constructed through the original matrix and the denoised matrix;and their risks were then compared.The results show that in the same income,the risk of portfolio model constructed by the denoised matrix is smaller than that of the original matrix.Therefore,the portfolio model constructed by the denoised matrix is more conducive to diffuse the risk in the investment decision-making.It is more effective to control the risk of stock portfolio in our country.
出处
《武汉理工大学学报（信息与管理工程版）》
CAS
2014年第4期585-588,共4页
Journal of Wuhan University of Technology：Information & Management Engineering
基金
国家自然科学基金资助项目(71140015)
中央高校基本科研业务费专项基金资助项目(2011-1a-034)
关键词
金融网络
关键社团
投资组合
financial network
key community
investment portfolio
