Heterogeneous agent models in financial markets: A nonlinear dynamics approach

Abstract

Studies on financial markets have accumulated consistent evidences of stylized facts and anomalies, which can be characterized by stochastic switching among different co-existing market states but yet difficult to reconcile with traditionally rational expectation theory. When agents are heterogeneous and boundedly rational, recent developments on the role of the adaptive behavior of interacting heterogeneous agents in financial markets have provided a nonlinear dynamics channel to such co-existence of different market states, shedding light into these stylized facts and anomalies. This survey focuses on the nonlinear dynamics approach to model the feedback of evolutionary dynamics of heterogeneous agents and to characterize the underlying mechanisms of the stylized facts and anomalies in financial markets, of which the authors and several coauthors have contributed in several papers.