Quantal response and nonequilibrium beliefs explain overbidding in maximum-value auctions

We analyze data from the second-price maximum-value auction experiment reported in Ivanov, Levin and Niederle (2010) in order to investigate the extent to which the winner’s curse can be explained by quantal response, in combination with diffferent assumptions about equilibrium (QRE) or nonequilibrium beliefs (QCH or QCE). We find a close correspondence between the theoretical predictions of those models and experimental behavior, even in the presence of frequent (and extreme) overbidding. The basic pattern in the data consists of a combination of flat overbidding for low signals, and monotonically increasing bidding for higher signals. The logit QRE model fits this pattern reasonably well. Incorporating nonequilibrium beliefs into the QRE model, in the form of either levels of strategic sophistication or cursed beliefs, leads to an even better match with the data, as these models imply slightly higher bids. We also show that these imperfect best response models predict essentially no treatment effects across different versions of the game, consistent with the experimental findings. Overall, our study indicates that the winner curse phenomenon in this auction is plausibly attributable to limits on strategic thinking combined with imperfect best response. ∗We thank Asen Ivanov for promptly supplying the original data from the experiments in Ivanov, Levin and Niederle (2010), and to audience members of ESA Meeting Fall 2010, especially Muriel Niederle, for comments. Please do not quote or circulate without permission.