A Network Model for Bank Lending Capacity

Abstract

We provide a structural framework to study the impact of a macroeconomic shock on the banking sector’s lending capacity. Our model consists of three building blocks: first, the evolution of non-bank assets that provide the entrance point for shocks from the real economy into the banking sector. Second, a network model that describes the propagation of the shock. Third, regulatory capital constraints that restrict banks’ lending behaviour. Therefore, the framework combines the literature on systemic risk and the literature on the bank lending channel. The building blocks ensure that our model captures six important aspects: firstly, by measuring systemic risk as the reduction in lending capacity, it aligns the model definition of systemic risk to the definition of the European Systemic Risk Board. Secondly, it models banks’ lending behavior as a function of counterparty risk addressing the usual critique of missing microfoundations of bank behavior. Thirdly, it emphasizes that not only defaulted banks, but also distressed banks are a thread for the real economy due to reduced lending capacity. Fourthly, our model shows the procyclicality of current (static) capital requirements. Fifthly, it allows us to study the effects of micro-prudential capital requirements in a network setting. And finally, the model describes the feedback loop of a shock that stems from the real economy through the banking sector back into the real economy.