Gao Xiangyun, Huang Shupei, Sun Xiaoqi, Hao Xiaoqing, An Feng
School of Humanities and Economic Management, China University of Geosciences, Beijing 100083, People's Republic of China.
Key Laboratory of Carrying Capacity Assessment for Resource and Environment, Ministry of Land and Resources, Beijing 100083, People's Republic of China.
R Soc Open Sci. 2018 Mar 28;5(3):172092. doi: 10.1098/rsos.172092. eCollection 2018 Mar.
Microscopic factors are the basis of macroscopic phenomena. We proposed a network analysis paradigm to study the macroscopic financial system from a microstructure perspective. We built the cointegration network model and the Granger causality network model based on econometrics and complex network theory and chose stock price time series of the real estate industry and its upstream and downstream industries as empirical sample data. Then, we analysed the cointegration network for understanding the steady long-term equilibrium relationships and analysed the Granger causality network for identifying the diffusion paths of the potential risks in the system. The results showed that the influence from a few key stocks can spread conveniently in the system. The cointegration network and Granger causality network are helpful to detect the diffusion path between the industries. We can also identify and intervene in the transmission medium to curb risk diffusion.
微观因素是宏观现象的基础。我们提出了一种网络分析范式，从微观结构视角研究宏观金融系统。我们基于计量经济学和复杂网络理论构建了协整网络模型和格兰杰因果关系网络模型，并选择房地产行业及其上下游行业的股票价格时间序列作为实证样本数据。然后，我们分析协整网络以理解稳定的长期均衡关系，并分析格兰杰因果关系网络以识别系统中潜在风险的扩散路径。结果表明，少数几只关键股票的影响能够在系统中便利地传播。协整网络和格兰杰因果关系网络有助于检测行业间的扩散路径。我们还可以识别并干预传导媒介以遏制风险扩散。
