Do Crowds Learn Wisdom? Theory and Evidence from Trading During the World Cup

Abstract:     
The paper presents a theoretical and empirical analysis of learning in a prediction market. The paper presents necessary and sufficient conditions for prices to converge to the probability of an event as information is incorporated into the market. The theoretical results are augmented with numerical analysis of convergence when assumptions are relaxed. The predictions of the theoretical model are tested against price dynamics from trading on Intrade futures during the 2002 World Cup. The results suggest that while the price may not equal the probability of an event, prices tend to converge to the probability.

Keywords: Prediction Markets, World Cup, probabilities

JEL Classifications: D83

Working Paper Series