Nobi Ashadun, Lee Jae Woo
Department of Physics, Inha University, 100 Inha-ro, Nam-gu, Incheon 402-751, South Korea.
Chaos. 2017 Jun;27(6):063107. doi: 10.1063/1.4978925.
In this paper, the change in topological hierarchy, which is measured by the minimum spanning tree constructed from the cross-correlations between the stock indices from the S & P 500 for 1998-2012 in a one year moving time window, was used to analyze a financial crisis. The hierarchy increased in all minor crises in the observation time window except for the sharp crisis of 2007-2008 when the global financial crisis occurred. The sudden increase in hierarchy just before the global financial crisis can be used for the early detection of an upcoming crisis. Clearly, the higher the hierarchy, the higher the threats to financial stability. The scaling relations were developed to observe the changes in hierarchy with the network topology. These scaling relations can also identify and quantify the financial crisis periods, and appear to contain the predictive power of an upcoming crisis.
在本文中，通过在一个一年移动时间窗口内由1998 - 2012年标准普尔500指数之间的互相关性构建的最小生成树来衡量的拓扑层次结构的变化，被用于分析一场金融危机。在观察时间窗口内的所有小危机中，层次结构都有所增加，但在2007 - 2008年全球金融危机发生时的急剧危机期间除外。全球金融危机之前层次结构的突然增加可用于提前检测即将到来的危机。显然，层次结构越高，对金融稳定的威胁就越大。开发了标度关系以观察层次结构随网络拓扑的变化。这些标度关系还可以识别和量化金融危机时期，并且似乎包含对即将到来的危机的预测能力。
