Title:Tight Bounds for the Price of Anarchy of Simultaneous First Price Auctions

Abstract:We study the Price of Anarchy of simultaneous First-Price auctions for buyers with submodular and subadditive valuations. The current best upper bounds for the Bayesian Price of Anarchy of these auctions are $e/(e-1)$ by Syrgkanis and Tardos and $2$ by Feldman et al., respectively. We provide matching lower bounds for both cases {\em even} for the case of the full information and for mixed Nash equilibria. An immediate consequence of our results, is that for both cases, the Price of Anarchy of these auctions stays the same, for mixed, correlated, coarse-correlated, and Bayesian Nash equilibria.
We bring some novel ideas to the theoretical discussion of upper bounding the Price of Anarchy in Bayesian Auctions settings. We suggest an alternative way to bid against price distributions. Using our approach we were able to re-provide the upper bounds of $e/(e-1)$ for XOS bidders. An advantage of our approach, is that it reveals a worst-case price distribution, that is used as a building block for the matching lower bound construction.
Finally, we apply our techniques on Discriminatory Price multi-unit auctions. We complement the results by Keijzer et al. for the case of subadditive valuations, by providing a matching lower bound of $2$. For the case of submodular valuations, we provide a lower bound of 1.109. For the same class of valuations, we were able to reproduce the upper bound of $e/(e-1)$ using our non-smooth approach.