论文部分内容阅读
针对常用的参数分析方法在证券风格投资分析中容易导致结果的差异性问题,提出将亚超度量空间方法引入到风格投资研究中,因为具有准确定义拓扑序列的亚超度量空间与指数分层结构、风格能够一一对应。首先利用股票价格计算出股间距离,并采用关联该证券组合的最小生成树(Kruskal)优化方法,从而得到亚超度量空间;其次将此亚超度量空间转换成指数分层结构树图,最后得到风格的分布特征。2004~2007年的上证50实证结果表明:横向来看以行业分类的风格较为显著,其中以银行和钢铁也最为明显,此外还有与地区差异、交叉持股相关的风格效应显现;纵向则在股改后行业风格效应更显著,其它的风格也显现得更多。结论为此方法是有效的,同时结果也可为证券投资组合配置提供依据和参考。
Aiming at the common problem that the method of parameter analysis easily lead to the difference of the result in the analysis of securities style investment, the sub-metric space method is proposed to be introduced into the study of style investment because of the sub-metric space and exponential hierarchy with exactly defined topological sequence , The style can correspond one by one. Firstly, the stock price is used to calculate the distance between stocks and the Kruskal optimization method is used to get the sub-metric space. Second, the sub-metric space is transformed into an exponential hierarchical tree map. Finally, Get the style of distribution. The empirical results of Shanghai stock market from 2004 to 2007 show that: the horizontal classification of the industry is more obvious, of which the banking and steel are also the most obvious, in addition there are regional differences, cross-shareholdings related style effects appear; After the share reform industry-style effect is more pronounced, other styles also appear more. The conclusion is that this method is effective, and the result also provides the basis and reference for the securities investment portfolio allocation.