考虑现实因素的模糊多目标投资组合策略  

A Fuzzy Multi-objective Portfolio Strategy with Realistic Factors

作　　者：杨兴雨[1] 陈思豆 刘伟龙 张永[1] YANG Xing-yu;CHEN Si-dou;LIU Wei-long;ZHANG Yong(School of Management,Guangdong University of Technology,Guangzhou 510520,China)

机构地区：[1]广东工业大学管理学院,广东广州510520

出　　处：《模糊系统与数学》2021年第2期76-84,共9页Fuzzy Systems and Mathematics

基　　金：国家自然科学基金资助项目(71501049);教育部人文社会科学研究基金资助项目(18YJA630132)。

摘　　要：真实金融市场上的红利和税收等现实因素对投资者的决策活动有着直接的影响。考虑风险资产的红利、税收及交易费用等因素及最小交易单位与投资比例边界等约束,将每个风险资产的收益率和红利率视为梯形模糊数,以最大化投资组合净收益的期望与偏度、最小化其下半方差与模糊性为目标函数,建立了带红利和税收的模糊多目标投资组合优化模型。然后,采用模糊多目标规划方法将该模型转换成一个单目标规划模型,并设计了一个改进的多种群遗传算法对模型进行了求解。最后,利用真实股票数据进行数值算例分析,说明了所提出的优化模型及求解算法的可行性和适用性。In the real financial market, many realistic influence factors such as dividend and tax have a direct effect on investors’ decision-making. We take the factors of dividend, tax and transaction cost and the constraints of minimum transaction lots and threshold constraints into account and regard the return rate and dividend rate of each risky asset as trapezoid fuzzy numbers, and build a fuzzy multi-objective portfolio model with dividend and tax, which aims to maximize the possibilistic mean and skewness and minimize the possibilistic lower semi-variance and fuzziness of the portfolio. Then, transform the proposed model into a single objective programming model by employing a fuzzy multi-goal programing method, and design a modified multi-population genetic algorithm to solve it. Finally, give a numerical example using real stock data to illustrate the feasibility and applicability of the proposed model and algorithm.

关 键 词：模糊投资组合 红利 税收 多种群遗传算法 

分 类 号：F830[经济管理—金融学]