Learning to Bid: The Design of Auctions Under Uncertainty and Adaptation

Abstract

We examine auction design in a context where symmetrically informed buyers and sellers of a good with a common but uncertain value learn through experience. Buyer strategies, even in the very long run, do not converge to the Bertrand-Nash strategy of bidding the expected value of the good. Moreover, first- and second-price auctions are not revenue equivalent. The outcomes of the auctions are sensitive to both the number of participating bidders and the reservation price. When only a small number of bidders participate, the sellers tend to employ a first-price auction even though it generates a lower average revenue than a second-price auction.

Keywords: Auction design, adaptive learning, genetic algorithm

JEL Classification: D44, D83