基于遗传编程的中国股票市场有效性检验.pdfVIP

Vo 1. 43 No. 6A 第 43 卷第6A 期 计算机科学 2016 年 6 月 Computer Science June 2016 基于遗传编程的中国股票市场有效性检验 王红霞1 曹波2 (北京青年政治学院计算机系 北京 100102)1 (北京工业大学计算机学院 北京 100124)2 摘 要 现代资本市场理论与金融投资实践之间存在着有效市场假说与技术分析之间的矛盾，使用流行的技术交易 规则检验股票市场有效性可能导致两种结论偏差。遗传编程使用树形结构表示问题的候选解，可以很好地描述技术 交易规则。利用遗传编程算法生成一种技术交易策略，并用其检验上证综合指数和 5 个沪深股市个股。回测结采表 明，提出的方法相对于买入-持有策略能够获得超额收益，并且优于常用的流行技术指标，也说明我国股票市场并未 达到弱式有效。 关键词 遗传编程，市场有效性，中国股票市场 中固法分类号 TP391 文献标识码 A Chinese Stock Market Efficiency Testing ßased on Genetic Programming WANG Hong-xia 1 CAO Bd (Computer Department ,Beijing Youth Politics College ,Beijing 100102 ,China)1 (School of Computer Science ,Beijing University of Technology ,Beijing 100124 ,China)2 Abstract There is a contradiction between the modem capital marl王et theory and the financial investment practice. And the contradiction is about the effective marl优t hypothesis and the technical analysis. The use of the popular technology trading rules to examine the effectiveness of the stock market may lead to two types of conclusion deviation. The tree structure is used to represent the candidate solutions in genetic programming which can well describe the technical trading rules. The genetic programming algorithm is used to generate technical trading strategy in this paper. The strat egy is used to test Shanghai indexes and five stocks in the Shanghai and Shenzhen stock markets. The back test results show that genetic programming generates the best technical trading strategy with significant excess profit com