Optimal Systemic Risk Mitigation in Financial Networks

We propose a novel multi-period bilateral clearing framework, where the level of systemic risk is mitigated through an optimal bailout allocation strategy. The interbank liability network evolves stochastically over time, with default events having a persistent impact on the balance sheet structure of the network. The optimal bailout policy and associated clearing payments are recovered as the solution of a constrained stochastic dynamic programming problem. We develop a numerical analysis showing that optimal bailout allocations are able to reduce significantly the systemic risk level when liability exposures are heterogeneous and volatile. Our analysis provides a tool to support regulator decisions of when and where to intervene so to prevent the onset of potential threats to financial stability. ‡School of Industrial Engineering, Purdue University, West Lafayette, IN, 47906. Email:[email protected] School of Industrial Engineering, Purdue University, West Lafayette, IN, 47906. Email:[email protected]