Behavioral Finance and Agent Based Model: the new evolving discipline of quantitative behavioral finance ?

The financial crisis of recent years has deeply questioned the ability of the traditional economic models to help to govern the complexity of the modern financial world. A growing number of scholars, practitioners, and regulators agree that the recurring financial crisis as well as the overwhelming evidence of market anomalies could be explained only resorting to behavioral finance. Behavioral finance has been able to identify the individual investor irrationality but unable to quantify its total effect on the market in terms of price deviation from fundamental. Quantitative Behavioral Finance (QBF) is an emerging discipline that attempts to model the impact of human cognitive biases over asset prices. The aim of this paper is to provide an overview of its theoretical foundations and its challenges. The paper is divided in two parts. In the first one, we present a much selected literature review of the key theoretical foundations. Why does this new field of study emerge ? What topics does it study ? Which disciplines have contributed the most and why ? In the second part, the paper sketches an outline and provides a preliminary, set of references about the agent-based model approach as one of the most promising line of research in quantitative modeling the behavioral investorsÕ impact on the market. The literature surveyed supports the use of this class of models because of their capability in copying with heterogeneous agentsÕ behaviours either rational or bounded rational without losing the ability to identify and examine how each of them operates separately or in interaction. Taken as a whole, the articles reviewed here indicate that many open issues remains both on the theoretical design of agent based models, due to the large degree of freedom of modelers, and on the empirical use of this class of models for real political economic implications, due to the arduous methods for the model validation, calibration and estimation.