Comparing the systemic risk of Italian insurers and banks

In this paper we assess the systemic risk of listed Italian insurers and banks by estimating four different measures based on conditional value-at-risk and marginal expected shortfall. Daily estimates in the period from 2007 to 2023 are obtained by assuming a parametric model able to capture volatility clustering phenomena. We keep the framework as simple as possible to get closed formulas or straightforward simulations for the estimation of the risk measures empirically studied in this paper. This allows us to compare the systemic risk of the entities in our sample without resorting to complex model calibration and risk measure evaluation, and to explore the dynamics of systemic risk on more than 4,000 daily observations for the 14 banks and 4 insurers in our sample. Our findings, partly justified by the composition of our sample (i.e. the few insurance companies considered have significantly higher market capitalizations compared with some of the banks in the sample), reveal that, across three out of four risk metrics, insurance entities exhibit slightly higher systemic risk levels than banks, on average. The least systemic banks consistently maintain lower risk profiles compared with the least systemic insurance companies. Conversely, the most systemic bank is slightly riskier than the most systemic insurance company. However, there is more variability among banks compared with insurance companies.