Wishful thinking - a large bonus for accurate predictions did not result in a smaller bias

From Wishful Thinking by Guy Mayraz. The abstract:

This paper presents a model and an experiment, both suggesting that wishful thinking is a pervasive phenomenon that affect decisions large and small. Agents in the model start out with state-dependent payoffs, and behave as if high-payoff states are more likely. Subsequent choices maximize subjective-expected utility given these beliefs. Subjects in the experiment were paid in accordance with the future value of a financial asset. Despite incentives for hedging, subjects gaining from high prices made higher predictions than subjects gaining from low prices. Comparative statics agreed with predictions. In particular, a large bonus for accurate predictions did not result in a smaller bias.