摘要
尽管图形技术分析在金融市场上被投资者广泛应用,但信息含量的学术研究却很不充分。本文参考Lo等(2000)和Brock等(1992)的研究框架,开辟了检验图形技术分析信息含量的新视角。通过将技术分析所使用的图形进行分解,设计出了一种衡量图形技术分析信息含量的检验途径,并提出了一个基于bootstrap原理的可用于检验各种交易策略的统计检验方法。此外,本文对原有文献模型的参数选择、形态定义等也提出了较大改进。根据本文对于图形技术分析的改进研究方法,我们首次对中国股市1233支A股进行图形技术分析,研究其包含的信息含量,证明恰当的图形技术分析确实可以提供十分显著的信息含量,从而有力地反驳了技术分析无效的观点。
Based on the framework of Lo et al (2000) and Brock et al (1992), this paper provided a new perspective of the research on the information of ‘charting’ By decomposing the patters that technical analyst used, this paper designed a testing method to gauge the information of ‘charting’ . This paper also proposed a bootstrap method that could be used to test the profitability of all kinds of trading strategies. In addition, this paper improved the original models in parametric selection and pattern definitions. Using the method improved in this paper, we tested the information of‘charting’, given the trading data of 1233 stocks from Chinese stock market. We proved that there do exist some technical patterns that can provide significant information, thus strongly refuted the viewpoint that technical analysis is useless.
出处
《数量经济技术经济研究》
CSSCI
北大核心
2005年第9期73-82,共10页
Journal of Quantitative & Technological Economics