Optimal Stress Tests in Financial Networks

Thumbnail Image



Journal Title

Journal ISSN

Volume Title

Repository Usage Stats



Bank stress test has become a centerpiece of post crisis bank supervision. Current studies have thus far examined the optimal policies on stand-alone single banks, but financial systems are interconnected in practice, and disclosure about banks influences the counterparty risks of other banks. This dissertation studies the optimal stress test design in a financial network, where banks' endogenous default outcomes are determined by a fixed point payment problem that accounts for both project qualities and interbank contagion.

The first part examines the joint stress test design on all banks in the financial network. In addition to the cross-state risk sharing in models of single banks, this model highlights the novel cross-bank risk sharing that arises from the spillover effects of disclosures via interbank payments. When expected bank profitability is high or counterparty exposures are large, disclosure is non-discriminatory, and either all banks pass or all banks fail the stress tests; otherwise only less impaired banks may pass. For network structures, I find: (i) in a ring network, banks at least a specific distance away from the nearest bank with asset impairment may pass; (ii) a more connected network is not necessarily more stable under the optimal disclosure; (iii) typically more connected banks receive preferred treatment.

The second part studies a selective stress test in a financial network, where the regulator selects an optimal subset of banks for stress tests and accordingly design the optimal disclosures only on these banks. Compared with the first part, systemic risk becomes more important as the regulator is less able to fine tune beliefs about contagion and needs to contain the risks from unselected banks. For network structures, I find: (i) in a ring network, stress test is either "balanced'' on banks positioned evenly and disclosure is non-discriminatory, or on "connected'' banks and disclosure is truth-telling on potential shocks; (ii) in a star network, stress test is conducted on the center bank when counterparty exposure is either sufficiently small or sufficiently large.





Huang, Jing (2020). Optimal Stress Tests in Financial Networks. Dissertation, Duke University. Retrieved from https://hdl.handle.net/10161/21455.


Dukes student scholarship is made available to the public using a Creative Commons Attribution / Non-commercial / No derivative (CC-BY-NC-ND) license.