Systemic Risk Indicators in the Eurozone: An Empirical Evaluation

In this brief paper, we use combination of Markov-switching models and dynamic conditional correlation models to ex-post evaluate the performance of three widely used systemic risk measures (SRISK, CISS and term-spread) based on their ability to predict financial turmoil. We first compare systemic risk measures based on their dynamic correlations. Second, we identify three regimes for each indicator and evaluate them based on their ability to capture crisis information. We found that in practice, the correlations between studied systemic risk measures are indeed high and indicators are successful in capturing regimes of high financial stress. We have however, identified a few periods when the indicators are not overlapping, especially in pre- and post-crisis period.