Working Paper: CEPR ID: DP8332
Authors: Ethan Cohen-Cole; Eleonora Patacchini; Yves Zenou
Abstract: We propose a novel mechanism to facilitate understanding of systemic risk in financial markets. The literature on systemic risk has focused on two mechanisms, common shocks and domino-like sequential default. Our approach is a formal model that provides an intellectual combination of the two by looking at how shocks propagate through a network of interconnected banks. Transmission in our model is not based on default. Instead, we provide a simple microfoundation of banks' profitability based on classic competition incentives. As competitors lending quantities change, both for closely connected ones and the whole market, banks adjust their own lending decisions as a result, generating a `transmission' of shocks through the system. We provide a unique equilibrium characterization of a static model, and embed this model into a full dynamic model of network formation with n agents. Because we have an explicit characterization of equilibrium behavior, we have a tractable way to bring the model to the data. Indeed, our measures of systemic risk capture the propagation of shocks in a wide variety of contexts; that is, it can explain the pattern of behavior both in good times as well as in crisis.
Keywords: financial networks; interbank lending; interconnections; network centrality; spatial autoregressive models
JEL Codes: C21; G10
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| interconnectedness among banks (F65) | systemic risk (E44) |
| one-dollar increase in lending by a bank (G21) | systemic risk multiplier effect (E44) |
| structure of the network (D85) | lending behavior (G21) |
| centrality (D80) | liquidity and risk exposure (G33) |
| network centrality shifts (D85) | nature of systemic risk during financial crises (E44) |