Reducing Systemic Risk in Financial Networks using Reinforcement Learning

Richard Le
Thursday, July 28 at 4 pm
Ross N638

An essential property of a financial system is the interconnectedness between each of its participants. Commonly, these connections present themselves as lending and borrowing relationships. Unfortunately, these relationships also serve as a channel for amplifying economic shocks leading to contagion. In this talk, we present how reinforcement learning can be used to discover network configurations with a lower level of systemic risk.