Macroeconomic Dynamics, 11 (Supplement 1), 2007, 8–33. Printed in the United States of America. DOI: 10.1017/S1365100512060208 ARTICLES LEARNING IN COBWEB EXPERIMENTS CARS HOMMES,JOEP SONNEMANS,JAN TUINSTRA, AND HENK VAN DE VELDEN University of Amsterdam Different theories of expectation formation and learning usually yield different outcomes for realized market prices in dynamic models. The purpose of this paper is to investigate expectation formation and learning in a controlled experimental environment. Subjects are asked to predict the next period’s aggregate price in a dynamic commodity market model with feedback from individual expectations. Subjects have no information about underlying market equilibrium equations, but can learn by observing past price realizations and predictions. We conduct a stable, an unstable, and a strongly unstable treatment. In the stable treatment, rational expectations (RE) yield a good description of observed aggregate price fluctuations: prices remain close to the RE steady state. In the unstable treatments, prices exhibit large fluctuations around the RE steady state. Although the sample mean of realized prices is close to the RE steady state, the amplitude of the price fluctuations as measured by the variance is significantly larger than the amplitude under RE, implying persistent excess volatility. However, agents’ forecasts are boundedly rational in the sense that fluctuations in aggregate prices are unpredictable and exhibit no forecastable structure that could easily be exploited. Keywords: Expectations, Learning, Cobweb Dynamics, Excess Volatility 1. INTRODUCTION The question whether “expectations matter” and may cause excess price volatility, above and beyond volatility driven by news about underlying economic funda- mentals, has been a matter of heavy debate among economists for many decades already. In a pioneering paper, Shiller (1981), for example, argued that stock prices are excessively volatile. The present paper may be viewed as an experimental test- ing of expectation formation and learning in a dynamic market setting. We employ We would like to thank Vernon Smith and Shyam Sunder for stimulating discussions. This research has been supported by the Netherlands Organization for Scientific Research (NWO) under a NWO-MaG Pionier grant. Earlier versions of this paper were presented at the Annual Meeting of the Economic Science Association, Lake Tahoe, May 27–30, 1999, the CeNDEF workshop on Economic Dynamics, Amsterdam, January 13–15, 2000 and the 7th Viennese Workshop on Optimal Control, Dynamic Games and Nonlinear Dynamics: Theory and Applications in Economics and OR/MS, Vienna, May 24–26, 2000 and the conference on “Belief Formation and Fluctuations in Economic and Financial Markets,” Heidelberg, December 12–13, 2002 sponsored by the Volkswagenstiftung. Stimulating discussions with participants are gratefully acknowledged. Address correspondence to: Cars Hommes, Center for Nonlinear Dynamics in Economics and Finance (CeNDEF), University of Amsterdam, Roetersstraat 11, NL-1018 WB Amsterdam, The Netherlands; e-mail: C.H.Hommes@uva.nl. c  2007 Cambridge University Press 1365-1005/07 $18.00 8