基于时频波动溢出网络的中国金融机构关联性

Interconnectedness of China's financial institutions based on time-frequency volatility spillover networks

金融机构关联性是系统性风险的主要来源,厘清时频域下金融机构关联性特征有助于防范与化解系统性风险.本文基于30家A股上市金融机构在2011年~2020年期间的高频交易数据,构建时频波动溢出网络研究中国金融机构关联性.通过引入网络密度与全局效率两个系统层面指标和入度、出度与相对影响值三个机构层面指标,以测度时频域下静态与动态关联性.研究发现:中国金融机构间的关联性主要为20天~60天的中长期关联性;在极端事件期间,往往表现为长期关联性增强和短期关联性减弱;银行部门倾向于接收偏短期的波动关联性并发出偏短期波动关联性,而证券和保险部门倾向于接收偏长期的波动关联性并发出偏短期的波动关联性.

The interconnectedness of financial institutions is the main source of systemic risk,and understanding the interconnectedness characteristics of financial institutions in the time-frequency domain can help prevent and resolve systemic risk.Based on the high-frequency trading data of 30 A-share listed financial institutions from 2011 to 2020,time-frequency volatility spillover networks are constructed to study the inter-connectedness of China's financial institutions.By introducing two system-level indicators(network density and global efficiency)and three institutional-level indicators(in-degree,out-degree and relative influence),the static and dynamic interconnectedness in the time-frequency domain are measured.The research results show that the connectedness of China's financial institutions is mainly the medium-long-term interconnectedness with 20 days~60 days.During extreme events,the interconnectedness often manifests as an enhancement of long-term interconnectedness and a weakening of short-term interconnectedness.The banking sector tends to receive short-term volatility interconnectedness and send short-term volatility interconnectedness,while the security and insurance sectors tend to receive long-term volatility interconnectedness and send short-term volatility interconnectedness.