Agent-Based Models for Financial Crises

Abstract

This article describes the agent-based approach to modeling financial crises. It focuses on the interactions of agents and on how these interactions feed back to change the financial environment. It explains how these models embody the contagion and cascades that occur owing to the financial leverage and market concentration of the agents and the liquidity of the markets. This article also compares agent-based models to the standard economic approach to crises and shows the ways in which agent-based models overcome limitations of economic models when dealing with financial crises. In particular, this article demonstrates how agent-based models replace homogeneous, representative agents with heterogeneous agents and optimization with heuristics, and how such models move away from a focus on equilibrium, allowing non-ergodic dynamics that are manifest during financial crises to emerge.