Financial Networks and Systemic Fragility

We study cascade of failures in multilayer financial networks with heterogeneous agents. We develop a flexible framework that allows for the evaluation of systemic risk and demonstrate that the model converges to a unique fixed point. We model a financial accelerator engine at the microeconomic level that incorporates feedback between the real and the financial sectors using contagion transmission channels such as loan defaults, bank credit crunches and deposit withdrawals. Using this model, we provide theoretical analysis both on the cross-section and time components of systemic risk. In the first, we demonstrate how idiosyncratic shocks to microeconomic agents can develop into systemic events. In the second, we investigate the structural causes of systemic risk variations in evolving financial networks. We find closed form expressions in both approaches with clear economic interpretation. We illustrate and confirm our theoretical predictions using a comprehensive and unique dataset on micro-level firm and bank data for Brazil.