Heterogeneous Beliefs, Risk and Learning in a Simple Asset Pricing Model精编.pdfVIP

Computational Economics 19: 95–132, 2002. 95 © 2002 Kluwer Academic Publishers. Printed in the Netherlands. Heterogeneous Beliefs, Risk and Learning in a Simple Asset Pricing Model CARL CHIARELLA and XUE-ZHONG HE School of Finance and Economics, University of Technology, Sydney, P.O. Box 123, Broadway, NSW 2007, Australia, E-mails: carl.chiarella@uts.edu.au and tony.he1@uts.edu.au Abstract. Trade among individuals occurs either because tastes (risk aversion) differ, endowments differ, or beliefs differ. Utilising the concept of ‘adaptively rational equilibrium’ and a recent frame- work of Brock and Hommes [6, 7] this paper incorporates risk and learning schemes into a simple discounted present value asset price model with heterogeneous beliefs. Agents have different risk aversion coefficients and adapt their beliefs (about future returns) over time by choosing from differ- ent predictors or expectations functions, based upon their past performance as measured by realized profits. By using both bifurcation theory and numerical analysis, it is found that the dynamics of asset pricing is affected by the relative risk attitudes of different types of investors. It is also found that the external noise and learning schemes can significantly affect the dynamics. Compared with the findings of Brock and Hommes [7] on the dynamics caused by change of the intensity of choice to switch predictors, it is found that many of their insights are robust to the generalizations considered: however, the resulting dynamical behavior is considerably enriched and exhibits some significant differences. Key words: asset pricing, risk, heterogeneous beliefs, bifurcation 1. Introduction In recent years a body of evidence on the role of heterogeneous beliefs in financial markets has presented a sharp challenge to the traditional view that assets in finan- cial markets are rationally priced to