Framing effects on asset markets: An experimental analysis

In this paper we investigate four hypotheses which are inconsistent with expected utility theory, but may well be explained by prospect theory. It deals with framing, the non-linearity of subjective probabilities, the disposition effect, and the correspondence of different experimental risk elicitation methods. Overall, 64 participants traded two assets on eight markets in a computerized continuous double auction. The results (i) indicate that the framing of information influenced individual trading behavior and asset holdings. However (ii), the variation of the probability of the framed information had no influence on trading volume. In addition, the results (iii) confirm the disposition effect. Participants who experienced a gain sold their assets more rapidly than participants who experienced a loss. In line with previous empirical results, we (iv) found little correspondence between different experimental risk elicitation methods.