A compositional analysis of systemic risk in European financial institutions

Abstract

Systemic risk is a complex and multifaceted phenomenon that needs to be addressed from different perspectives. In this work we propose a Compositional Data (CoDa) approach to analyze the distribution of relative contributions to systemic risk associated with major European countries during the period 2008–2021. We represent systemic risk measures corresponding to those countries as percentage shares, or parts, of a compositional dataset and we perform a multivariate statistical analysis using specific CoDa procedures. The proposed approach sheds new light on some variability patterns and cross-country relationships that appear to be linked to the composition of systemic risk parts in the system.