This project consists of six experiments exploring the effects of informational heterogeneity on asset markets in the laboratory. It is the first empirical application of a growing theoretical literature showing that asset bubbles arise because people have different prior beliefs about dividends. Bayesian learning explains important aspects of data from previous experiments, and this project explores whether subjects can be induced to have common beliefs agreeing with the actual distribution of an asset's fundamentals. The treatments control subjects' experience, use a tangible randomization device, and attempt to induce common priors by showing how dividends are drawn. This research will have significant implications for a general theory of asset prices.
