Nonparametric Identification and Estimation of Multi-Unit, Sequential, Oral, Ascending-Price Auctions with Asymmetric Bidders

Within the independent private-values paradigm, we derive the data-generating process of the winning bid for the last unit sold at sequential English auctions when bidder valuations are draws from one of several different classes of distributions; i.e., in the presence of asymmetries. When the identify of the winner as well as the number of units won by each bidder in previous stages of the auction are observed, we demonstrate nonparametric identification, and then propose two estimation strategies, one based on the empirical distribution function of winning bids for the last unit sold and the other based on Laguerre polynomials. Subsequently, we apply our method to data from fish auctions held in Grenaa, Denmark. For single-unit supply, we use our estimates to compare the revenues a seller could expect to earn were a Dutch auction employed instead