Global systemic risk measures and their forecasting power for systemic events, Working Paper, this version September 2017, by Peter Grundke and Michael Tuchscherer
Since the financial crisis 2007 – 2009, many market-based systemic risk measures have been proposed. Prominent examples are MES, SRISK or ΔCoVaR. Based on a simulation study in an extended banking network model that incorporates several sources of systemic risk, we analyse how good these systemic risk measures indeed perform in indicating the risk of a systemic event. For this, the systemic risk measures of those banks that default and for which after their default a systemic event occurs are compared with the systemic risk measures of those defaulting banks for which no subsequent systemic event can be observed. Within the simulation study, we find that many systemic risk measures are statistically significant in explaining the likelihood of a systemic event after a bank’s default. However, the economic significance of the systemic risk measures is located on a low level.
Keywords: banking network model, contagion, systemic risk measures
JEL classification: G01, G21, G28