Trading Simulations and Real Money Outcomes

Abstract

Using a novel dataset of trades made on a trading simulator in Brazil, we examine the link between simulation experience and (1) the determinants of opening a real money account, and (2) the performance in real money accounts. We find that more active simulator users, simulator users who take on more risk, and simulator users with the best stock level performance (but not portfolio level performance) are more likely to open real money accounts. These users are also likely to be more active real money traders and significantly underperform in their real money trades. These results suggest that without proper education regarding the hazards of active trading and accurate performance evaluation, stock simulator users may draw incorrect inferences regarding their trading skill and experience real-money underperformance.