Minority Games: A Physics-Based Analysis of Financial Market Behavior

Abstract

This thesis investigates the Minority Game as a simplified yet powerful agent-based model to study the collective behavior of financial markets through the lens of statistical physics. The model captures how individual decisions, based on limited historical information, can lead to emergent market dynamics such as herding, volatility, and phase transitions. By extending the original Minority Game, we analyze the effects of introducing randomness, imitation and individualbased learning on system efficiency and stability. Furthermore, entropy analysis provides insights into coordination efficiency and the degree of disorder within the system. The results highlight the Minority Game’s relevance as a simple framework but important for understanding the relation between individual behavior and macroscopic market phenomena