Strategic reporting in exponential family prediction markets

Prediction markets are a platform for aggregating information from a population. We perform market simulations on the exponential family models of prediction markets. We verify the market dynamics and provide some extensions to the previous models. We also consider the incentive compatibility problem in such markets with risk neutral Bayesian traders. We show that while the market is guaranteed to achieve information aggregation, whether traders express their beliefs promptly depends on their beliefs and initial market state.