Dynamic network analysis of North American financial institutions

Abstract We propose a state-space model to estimate the dynamic network among financial institutions selected from STOXX600 North America in the period from January 2005 to May 2020. We measure the network strength and find that the spillover effect increases significantly during the 2008 financial crisis and the coronavirus pandemic. Using weekly updates of the weight matrix, we detect four time-varying communities. Three communities mostly include companies of the financial supersectors, while the remaining includes Canadian companies. Furthermore, the communities centralities peak during the 2008 financial crisis, while during the COVID-19 period lower values are estimated.