Systemic Risk in Interbanking Networks

We develop a mean field model of interbanking borrowing and lending activities. Each bank borrows from or lends to other counterparties at an idiosyncratic rate and is exposed to sudden shocks affecting the level of its monetary reserves. Using weak convergence analysis, we provide an explicit characterization of the measure-valued process associated with a large interbanking system. We use the limit process to construct law of large number (LLN) approximations for systemic indicators assessing average distance to default and measuring the total volume of interbanking activities. We illustrate the predictive power and accuracy of our framework via a detailed numerical analysis, showing that indicators are sensitive to lending preferences, volatility, and occurrences of negative events.