Loss aversion and learning to bid

Bidding challenges learning theories since experiences with the same bid vary stochastically:the same choice can result in either a gain or a loss. In such an environment thequestion arises how the nearly universally documented phenomenon of loss aversion affectsthe adaptive dynamics. We analyze the impact of loss aversion in a simple auctionfor different learning theories. Our experimental results suggest that a version of reinforcementlearning which accounts for loss aversion fares as well as more sophisticatedalternatives.