主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2008, Vol. 16 ›› Issue (3): 102-108.

• 论文 • 上一篇    下一篇

最优交易策略问题的动态求解方法

李志生   

  1. 中南财经政法大学新华金融保险学院, 湖北武汉43007
  • 收稿日期:2007-01-19 修回日期:2008-05-25 出版日期:2008-06-30 发布日期:2008-06-30
  • 作者简介::李志生(1978- ),男(汉族),湖北武汉人,中国财经政法大学新华金融学院,副教授,研究方向:金融、风险管理、决策优化.
  • 基金资助:

    中南财经政法大学振兴基金(90407006105)

A Dynamic Programming Approach for Constructing Optimal Trading Strategy

LI Zhi-sheng   

  1. Zhongnan University of Economics and Law, Wuhan 430073, China
  • Received:2007-01-19 Revised:2008-05-25 Online:2008-06-30 Published:2008-06-30

摘要: 资产配置包括资产在空间和时间上的配置。现代投资组合理论为资产在空间上的配置提供了比较完备的模型和应用框架,但是资产在时间上的配置问题,学者们的研究甚少。资产在时间上的配置的核心问题是在不同时间对不同资产做出合理的买进、持有和卖出决策,即交易策略设计。本文应用动态规划的原理,分别讨论了存在和不存在最大交易次数限制的情况下,基于总收益率最大的交易策略的求解算法,并利用香港股票市场的数据进行实例分析。本文提出的算法是关于交易的时间跨度和资产数量的多项式算法,计算量和存储空间不因二者的增大而过度增大,在解决大规模问题时也是非常有效的。

关键词: 交易策略, 交易费用, 收益率, 动态规划, 最优解

Abstract: Asset allocation means dividing investors' investments among different assets both in space and time.Although modern portfolio theory has developed to a highly sophisticated level and provided us with valuable theatrical models and application frameworks in steering assets in space,little is known about asset allocation in time.The core question of asset allocation in time is how to optimally select buying and selling time for different assets at different time,that is,optimal trading strategy design.Based on dynamis programming principles,this paper proposes an efficient algorithm for return-optimal trading strategy both for the case of trading with and without the constraint of maximum trading times,and implements the algorithm by using the data from Hong Kong stock market.The computation algorithm proposed in this paper is a linear time algorithm with respect to the number of trading periods and number of assets,and can be used in large-scale problem efficiently.

Key words: trading strategy, transaction cost, rate of return, dynamic programming, optimal solution

中图分类号: