Modelling cross-border systemic risk in the European banking sector: a copula approach

Raffaella Calabrese,Silvia Osmetti

DOI: https://doi.org/10.48550/arXiv.1411.1348

2014-11-06

Abstract:We propose a new methodology based on the Marshall-Olkin (MO) copula to model cross-border systemic risk. The proposed framework estimates the impact of the systematic and idiosyncratic components on systemic risk. Initially, we propose a maximum-likelihood method to estimate the parameter of the MO copula. In order to use the data on non-distressed banks for these estimates, we consider times to bank failures as censored samples. Hence, we propose an estimation procedure for the MO copula on censored data. The empirical evidence from European banks shows that the proposed censored model avoid possible underestimation of the contagion risk.

Risk Management,Statistical Finance