To explore why bubbles frequently emerge in the experimental asset market model of Smith, Suchanek & Williams (1988), we vary the fundamental value process (constant or declining) and the cash-to-asset value-ratio (constant or increasing). We observe high mispricing in treatments with a declining fundamental value, while overvaluation emerges when coupled with an increasing C/A-ratio. A questionnaire reveals that the declining fundamental value process confuses subjects, as they expect the fundamental value to stay constant. Running the experiment with a different context ("stocks of a depletable gold mine" instead of "stocks") significantly reduces mispricing and overvaluation as it reduces confusion. JEL: C92, D84, G10Keywords: Experimental economics, asset market, bubble, market efficiency, confusion * Kirchler: University of Innsbruck, Department of Banking and Finance, Universitätsstrasse 15, 6020 Innsbruck, and University of Gothenburg, Centre for Finance, 40530 Gothenburg, Sweden, michael.kirchler@uibk.ac.at.Huber: University of Innsbruck, Department of Banking and Finance, Universitätsstrasse 15, 6020 Innsbruck, juergen.huber@uibk.ac.at. Stöckl: University of Innsbruck, Department of Banking and Finance, Universitätsstrasse 15, 6020 Innsbruck, thomas.stoeckl@uibk.ac.at. We thank Peter Bossaerts, Michael Hanke, Cars Hommes, Charles Noussair, Jörg Oechssler, Charlie Plott, Matthias Sutter, and the participants at ESA 2010 in Copenhagen, Experimental Finance 2010 in Gothenburg, 5th Nordic Conference on Behavioral and Experimental Economics 2010 in Helsinki, Conference on Quantifying and Understanding Dysfunctions of Financial Markets 2010 in Leuven, and participants at seminars at the University of Amsterdam and the University of Innsbruck for helpful comments. Financial support by the Austrian National Bank (OeNBgrant 12789), the Austrian Science Foundation (FWF-grant 20609), and the University of Innsbruck (Nachwuchsfoerderung Kirchler) is gratefully acknowledged. 1We explore possible causes for the emergence of "bubbles" in experimental asset markets replicating the seminal design introduced by Vernon L. Smith, Gerry L. Suchanek & Arlington W. Williams (1988, henceforth SSW). In particular, we separate the effect of a declining/constant fundamental value (FV) from the effect of an increasing/constant cash-to-asset-value-ratio (C/A-ratio). 1We observe that confusion in treatments with declining fundamental value is the main driver for mispricing and leads to overvaluation when coupled with a high C/A-ratio.2 This is supported by findings from a questionnaire and from two control treatments where the notion "stocks" is replaced by "stocks of a depletable gold mine". This change in context reduces subjects' confusion about the FV (elicited by a questionnaire) and leads to significantly smaller mispricing and overvaluation. The design is robust to treatment changes like a higher variance of the dividend process and a higher initial C/A-ratio that have rekindled bubbles in earlier experiments.In 1988,...

Experimental economics, Asset market, Bubble, Bubble measure, C90, D84, G14,

Hot hand belief, Gambler’s fallacy, Experimental economics, Decision making under risk, C91, D81, G10,

No abstract

a b s t r a c tIn laboratory experiments we explore the effects of communication and group decision making on investment behavior and on subjects' proneness to behavioral biases. Most importantly, we show that communication and group decision making do not impact subjects' overall proneness to the hot hand fallacy and to the gambler's fallacy. However, groups decide differently than individuals, as they rely significantly less on useless outside advice from "experts" and choose the risk-free option less frequently. Furthermore we document gender differences in investment behavior: groups of two female subjects choose the riskfree investment more often and are marginally more prone to the hot hand fallacy than groups of two male subjects.