Modeling Risk Patterns of Russian Systemically Important Financial Institutions

The world financial crisis of 2008-2009 has shown that the existence of systemically important financial institutions (SIFIs) poses serious policy challenges to both developed and developing economies’ authorities. As for now there are different approaches to identifying SIFIs focused on contagion, concentration, correlation and conditions effects. The paper aims at testing a new approach to SIFIs’ identification based on the Russian banking data panel. It is hypothesized that SIFIs are characterized by unique behaviour in terms of risks undertaken. Automatic clustering procedure is being run to find homogeneous groups of banks in terms of their risk patterns. Risk patterns include proxies for credit, market, operational risk values for each bank in a sample. In order to reconstruct aggregate risk patterns for the banking clusters, copula models are used. Time variances in risk profile are accounted by identifying copula structural shift moment. The paper also tests a hypothesis about the key role of the institution’s size in determining systemic importance. Finally the effectiveness of SIFIs’ identification based on their risk profile is evaluated. When concluding, recommendations on SIFIs’ regulation in Russia are provided.