Budget-Constrained Auctions with Unassured Priors

In today's online advertising markets, it is common for an advertiser to set a long-period budget. Correspondingly, advertising platforms adopt budget control methods to ensure that any advertiser's payment is within her budget. Most budget control methods rely on value distributions of advertisers. However, due to the complex environment advertisers stand in and privacy issues, the platform hardly learns their true priors. Therefore, it is essential to understand how budget control auction mechanisms perform under unassured priors. This paper gives a two-fold answer. First, we propose a bid-discount method barely studied in the literature. We show that such a method exhibits desirable properties in revenue-maximizing and computation when fitting into first-price auction. Second, we compare this mechanism with another four in the prior manipulation model, where an advertiser can arbitrarily report a value distribution to the platform. These four mechanisms include the optimal mechanism satisfying budget-constrained IC, first-price/second-price mechanisms with the widely-studied pacing method, and an application of bid-discount in second-price mechanism. We consider three settings under the model, depending on whether the reported priors are fixed and advertisers are symmetric or not. We show that under all three cases, the bid-discount first-price auction we introduce dominates the other four mechanisms concerning the platform's revenue. For the advertisers' side, we show a surprising strategic-equivalence result between this mechanism and the optimal auction. Extensive revenue dominance and strategic relationships among these mechanisms are also revealed. Based on these findings, we provide a thorough understanding of prior dependency in repeated auctions with budgets. The bid-discount first-price auction itself may also be of further independent research interest.