Ascending auctions with bidder asymmetries

Abstract

We present a partial identification approach for ascending auctions with bidder asymmetries, where bidders' asymmetric types may be unobservable to the econometrician. Our approach yields sharp bounds and builds on and generalizes other recent bounds approaches for correlated private values ascending auctions. When bidder identities are observable, our approach yields tighter bounds than previous approaches that ignore asymmetry, demonstrating that bidder asymmetries can function as an aid rather than a hindrance to identification. We present a nonparametric estimation and inference approach relying on our identification argument and apply it to data from U.S. timber auctions, finding that bounds on optimal reserve prices and other objects of interest are noticeably tighter when exploiting bidder asymmetries.

Estimation in English auctions with unobserved heterogeneity

“…Although the model of valuations is extremely general (except for assuming ex ante symmetry), this method requires accurate observation of the number of bidders in each auction, and the resulting bounds can be fairly wide unless the number of bidders varies a lot. Coey, Larsen, Sweeney, and Waisman (2017) demonstrate similar bounds for a model with asymmetric bidders. Another limitation of this method is that it makes no attempt to fully recover underlying model primitives, making it useful for certain counterfactuals but not for others.…”

The bidder exclusion effect

“…In the symmetric correlated private values setting of Aradillas‐López, Gandhi, and Quint () (or in the related asymmetric correlated private values setting of Coey et al., ), the authors demonstrate that their two‐sided bound on seller profits is only available when valuations are independent of the number of bidders. If valuations and the bidder count are positively dependent, a one‐sided bound on seller revenue still holds, but no meaningful bound on the optimal reserve price is available.…”

“…These results include Aradillas‐López, Gandhi, and Quint () and Coey et al. (), which apply to ascending auctions with correlated private values, and Hernandez, Quint, and Turansick (), Freyberger and Larsen (), and Chesher and Rosen (), which apply to ascending auctions with separable unobserved heterogeneity. Our approach contributes to this literature by providing identification arguments for ascending auctions in a particular type of correlated private values environments, that of conditionally independent private values.…”

The Simple Empirics of Optimal Online Auctions

“…Our paper also contributes to a literature in economics that provides direct inference about reserve prices or other objects of interest to auction design rather than attempting to estimate the full distribution of valuations, such as Li et al (2003), Haile and Tamer (2003), Tang (2011), Aradillas-López et al (2013), Coey et al (2014, Chawla et al (2014), and Coey et al (2017). We also relate to other work empirical focusing on online e-commerce auctions, such as Song (2004) or Platt (2017), and online ad auctions.…”