求解带有交易费的CVaR投资组合模型的L-S算法

张茂军; 南江霞; 高爱华

求解带有交易费的CVaR投资组合模型的L-S算法

引用本文：张茂军, 南江霞，高爱华.求解带有交易费的CVaR投资组合模型的L-S算法[J].经济数学,2012,(2):73-78

摘要点击次数: 1350

全文下载次数: 170

作者	单位
张茂军, 南江霞，高爱华	(1. 桂林电子科技大学数学与计算科学学院，广西桂林541004 2. 大连理工大学数学科学学院，辽宁大连116024)

中文摘要:本文假设投资者是风险厌恶型，用CVaR作为测量投资组合风险的方法.在预算约束的条件下，以最小化 CVaR为目标函数，建立了带有交易费用的投资组合模型.将模型转化为两阶段补偿随机优化模型，构造了求解模型的随机L-S算法.为了验证算法的有效性，用中国证券市场中的股票进行数值试验，得到了最优投资组合、VaR和CVaR的值.而且对比分析了有交易费和没有交易费的最优投资组合的不同，给出了相应的有效前沿.

中文关键词:CVaR 投资组合交易费 L-S算法二阶段补偿优化

L-S Algorithm for the CVaR Portfolio Models with Transaction Costs

Abstract:Assuming that investors are risk averse, CVaR was used as a method to measure a portfolio risk. A portfolio model with transaction costs was constructed under the budget constraint condition and minimizing the CVaR as an objective function. The model was transformed into a two-stage recourse problem of stochastic programming, and a stochastic L-S algorithm was designed to solve the model. The efficiency of our algorithm was illustrated by a portfolio selection experiment on the China market, and the optimal portfolio, VaR and CVaR were obtained. Moreover, the differences between the optimal portfolio with transaction costs and the one with no transaction costs were analyzed, and the corresponding efficient frontiers were given.

keywords:CVaR portfolio transaction cost L-S algorithm two-stage recourse optimization

查看全文 查看/发表评论 下载pdf阅读器