Although a number of theoretical studies explain empirical puzzles in nance with ambiguity aversion, it is not a given that individual ambiguity attitudes survive in markets. In fact, despite ample evidence of ambiguity aversion in individual decision making, most studies nd no or only limited ambiguity aversion in experimental nancial markets, even when they exclude arbitrage. We argue that ambiguity eects in markets depend on market feedback and on a suciently strong bias toward ambiguity among the participants. Accordingly, we nd signicant ambiguity eects in lowfeedback call markets for assets that provoke high ambiguity aversion, but no ambiguity eects in high-feedback double auctions. * Radboud University Nijmegen, IMR, Department of Economics, Thomas van Aquinostraat 5, 6525 GD Nijmegen, The Netherlands † Universität Erlangen-Nürnberg, Lehrstuhl für Volkswirtschaftslehre, insbesondere Wirtschaftstheorie, Lange Gasse 20, 90403 Nürnberg, Germany ‡ Radboud University Nijmegen, IMR, Department of Economics, Thomas van Aquinostraat 5, 6525 GD Nijmegen, The Netherlands (u.weitzel@fm.ru.nl) § We thank seminar participants at the ESA World Meeting in New York (2012), the Experimental Finance Conference in Luxembourg (2012), and the ESA European Meeting in Cologne (2012) for helpful comments. All remaining errors are ours.
1Electronic copy available at: http://ssrn.com/abstract=2227335
IntroductionMany real-life decisions are characterized by ambiguity, in which we lack important information such as the objective probabilities of the relevant states.Keynes (1921) proposed a simple thought experiment to illustrate the eects of ambiguity.Imagine [...] the two cases following of balls drawn from an urn. In each case we require the probability of drawing a white ball;in the rst case we know that the urn contains black and white in equal proportions; in the second case the proportions of each color is unknown, and each ball is as likely to be black as white. It is evident that in either case the probability of drawing a white ball is 1 /2, but that the weight of the argument in favor of this conclusion is greater in the rst case. (Keynes, 1921, Chapter VI.6) 1 Ellsberg (1961) used this experimental design, commonly referred to as the '2-color Ellsberg urn' to show that a preference for the risky urn (with measurable probabilities) over the ambiguous urn (with immeasurable probabilities) violates the Subjective Expected Utility Theory and the Sure-thing Principle of the Savage axioms (Savage, 1954). Since then a large body of individual choice experiments have conrmed that, on average, decision makers are 'ambiguity averse' when confronted with the above-quoted choice.
2In a recent survey of the experimental literature, Trautmann and Van De Kuilen (forthcoming) conclude that there is clear evidence that on the average, and across various elicitation methods, ambiguity aversion is the typical qualitative nding.Ambiguity aversion is a possible cause for a number of empirical puzzles in nancial economics, wh...