基于熵度量风险的投资组合优化模型——来自深证100的数据分析  

Portfolio Optimization Model Based on Entropy-measuring Risk——An Analysis of Shenzhen 100 Data

作　　者：郑勇[1] 刘超[2] 

机构地区：[1]山东财经大学财务处,山东济南250014 [2]山东财经大学金融学院,山东济南250014

出　　处：《山东财政学院学报》2014年第4期29-35,F0003,共8页Journal of Shandong Finance Institute

基　　金：国家自然科学基金项目"货币政策多目标交互行为协调控制研究"(61273230);2012年度教育部"新世纪优秀人才支持计划"(NCET-12-1027);山东省"金融产业优化与区域发展管理协同创新中心"首席科学家平台"金融产业系统优化机制创新平台"

摘　　要：在马科维茨的均值—方差模型的基础上,将熵理论引入投资组合模型中,建立基于熵度量风险的均值—熵模型,并选取深证100中的10只股票进行实证研究,比较了所选股票的方差与初始熵值,验证了新模型建立的可行性以及必要性。通过matlab计算两种模型的投资方案得出:均值—熵模型较均值—方差模型在实际应用中更具有实用性,在同样收益率水平下能够为投资者提供具有较少证券数量的投资方案。这不仅节约了过度分散化给投资者带来的交易费用和管理费用,而且节省了信息资源,增强了投资者的信息处理能力。Based on Markowitz＇ s Mean-Variance Model and by introducing the entropy theory into the invest- ment portfolio model, the Mean-Entropy model based on entropy-measuring risk is established before an empirical study is conducted with 10 stocks selected from Shenzhenl00 to compare the variance of the selected stocks with the initial entropy value and verify the feasibility and necessity of establishing the new model. The calculation via matlab of the investment plans of the two models shows that the Mean-Entropy Model is more practical than Mean-Variance Model, and the former can offer investment programs with a smaller number of securities yet with the same yield, which can save not only the transaction costs and management fees resulting from excessive decentralization, but also the information resources, thus enhancing investors＇ capability of information processing

关 键 词：熵 风险 投资组合 均值—方差模型 

分 类 号：F830.59[经济管理—金融学]