Using Granger Causality to Determine Interconnectedness in Unlisted Banking Markets

Abstract

The proposed model allows to identify the interconnectedness of the banking sector where banks are nonlisted by using causal relationships (direct and indirect) between bank performance indicators. Model was applied to three separate banking markets in Baltics for time period 2014-2018. The results show that least interconnected banking sector is in Lithuania, followed by averagely interconnected Estonian market and highly interconnected Latvian banking market. Model was also able to distinguish bilateral bank pairs with no interconnectedness. This study shows that interconnectedness tests should be based on various performance indicators as banks are able to herd through various channels.