Measuring Systemic Risk: Robust Ranking Techniques Approach

Amirhossein Sadoghi

DOI: https://doi.org/10.48550/arXiv.1503.06317

2015-03-21

Risk Management

Abstract:In this research, we introduce a robust metric to identify Systemically Important Financial Institution (SIFI) in a financial network by taking into account both common idiosyncratic shocks and contagion through counterparty exposures. We develop an efficient algorithm to rank financial institutions by formulating a fixed point problem and reducing it to a non-smooth convex optimization problem. We then study the underlying distribution of the proposed metric and analyze the performance of the algorithm by using different financial network structures. Overall, our findings suggest that the level of interconnection and position of institutions in the financial network are important elements to measure systemic risk and identify SIFIs. Results show that increasing the levels of out- and in-degree connections of an institution can have a diverse impact on its systemic ranking. Additionally, on the empirical side, we investigate the factors which lead to the identification of Global Systemic Important Banks (G-SIB) by using a panel dataset of the largest banks in each country. Our empirical results supports the main findings of the theoretical model.