Robust Bidding in First-Price Auctions: How to Bid without Knowing What Others Are Doing

Abstract

Bidding optimally in first-price auctions is complicated. In the classical equilibrium framework, optimal bidding relies on detailed beliefs about other bidders' value distributions and bidding functions. This article shows how to and a robust bidding rule that does well with minimal information and thus achieves good performance in many situations. Robust bidding means to minimize the maximal difference between the payoff and the payo that could be achieved if one knew the other bidders' value distributions and bidding functions. We derive robust bidding rules under dierent scenarios, including complete uncertainty. Our bid recommendations are evaluated with experimental data.