Reserve Pricing in Repeated Second-Price Auctions with Strategic Bidders
This addresses revenue optimization for sellers in auction settings with strategic bidders, representing a strong specific gain in algorithmic performance.
The paper tackles the problem of revenue optimization in repeated second-price auctions with strategic bidders by proposing a novel algorithm that achieves a strategic regret upper bound of O(log log T) for worst-case valuations.
We study revenue optimization learning algorithms for repeated second-price auctions with reserve where a seller interacts with multiple strategic bidders each of which holds a fixed private valuation for a good and seeks to maximize his expected future cumulative discounted surplus. We propose a novel algorithm that has strategic regret upper bound of $O(\log\log T)$ for worst-case valuations. This pricing is based on our novel transformation that upgrades an algorithm designed for the setup with a single buyer to the multi-buyer case. We provide theoretical guarantees on the ability of a transformed algorithm to learn the valuation of a strategic buyer, which has uncertainty about the future due to the presence of rivals.