By introducing a new generalized forecast error variance decomposition (GFEVD) approach that splits the same into its contemporaneous and lagged components, we investigate the risk spillover effects of different order moments, derived from intraday data, for the top 10 banks and top 10 oil and gas companies in the U.S., covering the period from December 29, 2017 to December 30, 2022. The study finds that, first, the dynamic total connectedness of all order moments is heterogeneous over time driven by economic events. Second, except realized volatility spillovers, the vast majority of overall spillovers are attributable to contemporaneous spillovers, while only a tiny fraction is associated with lagged spillovers. Finally, realized skewness (crash risk) and realized kurtosis (extreme events) in banks and oil and gas companies originate mainly from intra-industry rather than inter-industry transmission.

中文翻译：

---

石油银行网络的连接程度如何？公司层面的高频证据


通过引入一种新的广义预测误差方差分解 (GFEVD) 方法，将其分为同期和滞后部分，我们研究了从日内数据得出的不同订单时刻的风险溢出效应，针对排名前 10 的银行和排名前 10 的石油和天然气公司。美国天然气公司的数据，涵盖2017年12月29日至2022年12月30日期间。研究发现，首先，在经济事件驱动下，随着时间的推移，所有订单时刻的动态总连通性是异质的。其次，除了已实现的波动性溢出之外，总体溢出的绝大多数可归因于同期溢出，而只有一小部分与滞后溢出相关。最后，银行和油气公司的已实现偏度（崩溃风险）和已实现峰度（极端事件）主要源于行业内而不是行业间传播。