摘要：correct information about the expected dividends and their probabilities is also available. METHOD: In two experiments, totaling34 Smith-Suchanek-Williams type double-auction continuous experimental markets (238 subjects), participants were exposed to misinformation regarding dividend payouts in a previous game, with the correct dividend matrix also provided. The misinformation stated the dividends in the previous game to have been much lower or much higher than according to the expected value function. The misinformation was either homogenous for all participants or provided to only half of the investors in a market (heterogeneously). RESULTS: Homogenous misinformation stating that the last game’s dividend payouts were high, led to larger overpricing throughout the game, as compared to baseline (no misinformation) and homogenous misinformation stating that the last game’s dividends were low. In informationally heterogeneous markets, where half of the participants received “high dividends” misinformation and half remained non-misinformed, transaction prices were the lowest compared to the aforementioned treatments. It was also discovered that agents receiving the ‘high dividend’ misinformation had lower returns than non-misinformed participants in the same heterogeneous market.
