Reserve Prices in Repeated Auctions

I consider a model of repeated auctions in which the distribution of bidders’ values is only known to the bidders and the seller attempts to learn this distribution to inform her choice of reserve prices in the future. I find that in any equilibrium bidders will shade their bids to act as if their values are drawn from a lower distribution than they actually are. The bid shading may be so severe that the seller would prefer to simply commit to setting the reserve price that would be optimal if bidders’ values were drawn from the lowest possible distribution to eliminate the incentive for bidders to shade their bids.

• Algorithms and Theory

• Economics and Electronic Commerce