Systemic risk in the banking sector is usually associated with long periods of economic downturns and very large social costs. On one hand, shocks coming from correlated exposures towards the real economy may induce correlation in banks’ default probabilities thereby increasing the likelihood for systemic-tail events like the 2008 Great Financial Crisis. On the other hand, financial contagion also play an important role in generating large-scale market failures, amplifying the initial shocks coming from the real economy. To study the sources of these rare phenomena, we propose a new definition of systemic risk (i.e. the probability to have a large number of banks going into distress simultaneously) and thus we develop a multilayer microstructural model to study empirically the determinants of systemic risk. The model is then calibrated on the most comprehensive granular dataset for the euro area banking sector, capturing roughly 96% or EUR 23.2 trillion of euro area banks’ total assets over the period 2014-2018. The outputs of the model decompose and quantify the sources of systemic risk showing that correlated economic shocks, financial contagion mechanisms, and their interaction are the main sources of systemic events. The results obtained with the simulation engine nicely resemble common market-based systemic risk indicators and empirically corroborate findings from the existing literature. This framework represents to regulators and central bankers a tool to study systemic risk and its developments, pointing out that systemic events and banks’ idiosyncratic defaults have different drivers, hence implying different policy responses.