考虑大宗交易的均值-方差投资组合优化模型及其分支定界算法  

Portfolio optimization model and branch-bound algorithm under block trading

作　　者：薛宏刚[1] 张川[2] 胡春萍[3] 徐成贤[2] 

机构地区：[1]西安交通大学经济与金融学院,西安710061 [2]西安交通大学理学院,西安710049 [3]西安交通大学公共政策与管理学院,西安710049

出　　处：《系统工程理论与实践》2011年第9期1617-1627,共11页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(71171158);教育部新世纪优秀人才支持计划项目(NCET-10-0646);教育部人文社会科学研究项目基金(09XJAZH005;10YJCZH043)

摘　　要：由于大宗交易下边际交易费用递减,因此用线性加凹的函数拟合实际交易费用函数,建立了均值-方差框架下的组合优化模型并给出了相应的求解算法.通过对恒生指数样本股的实证分析发现:考虑大宗交易的组合有效边缘介于线性交易费用和无交易费用的组合有效边缘之间;大宗交易稀释了"分散化降低风险"的效应;大宗交易下交易费用越大,相对于线性交易费用而言组合集中度越高.In this paper,we consider the portfolio optimal problem under block trading.As the decrease of marginal transaction cost under block trading,we establish the mean-variance portfolio optimization model and the corresponding algorithm by means of fit transaction cost function with linear and concave form.Computational results are presented by considering the stocks involved in Hang Seng Index to show that：The efficient frontier under block trading is between that with linear transaction cost and without transaction cost;The investment risk can be eliminated by portfolio diversification,but the effect of that is dilute under block trading;In terms of linear transaction cost,the greater the transaction cost under block trading,the higher the portfolio concentration.

关 键 词：大宗交易 线性加凹交易费用 组合规模 分散化效应 分枝定界算法 

分 类 号：F830.9[经济管理—金融学]