基于复杂网络方法的上证与创业板比较研究

Comparisons of Shanghai Stock Exchange Index and GEM Index Based on Complex Network Method

利用可见图算法,构建由证券市场价格时间序列生成的复杂网络,并对其特征进行分析。研究结果表明:上证和创业板网络的度分布都服从幂律分布而不是泊松分布,与许多自然界网络一样,呈现出典型的无标度特征。通过计算网络的全局效率、模块化指数、同配系数等网络拓扑参数可以发现,两个市场在不同时间区间上的值变拐点、增减趋势、波动特征均存在差异。

This paper constructed complex networks generated by the time series of stock market price.The empirical study on the characteristics of the network showed that the degree distribution of Shanghai Stock Exchange Index and Growth Enterprise Market Index both obey the power law distribution rather than Poisson distribution.Hence,they both showed typical scale-free characteristics like many natural networks.In particular,we also found that there were differences in the inflection point of value change,trend of increase and decrease,and fluctuation characteristics between the two markets at different time intervals by calculating network topology parameters such as global efficiency,modularization index and assortativity coefficient.