Asymmetric Auctions

Abstract

The revenue‐equivalence theorm for auctions predicts that expected seller revenue is independent of the bidding rules, as long as equilibrium has the properties that the buyer with the highest reservation price wins and any buyer with the lowest possible reservation price has zero expected surplus. Thus, in particular, the two most common auction institutions—the open ‘English’ and the sealed high‐bid auction—are equivalent despite their rather different strategic properties.

Simultaneous Allocation of Bundled Goods through Auctions: Assessing the Case for Joint Bidding

“…As in our experiment, joint bidding reduces the number of bidders to two. He leverages previous work by Lebrun (1999) and Maskin and Riley (2000) to show that (under general assumptions) the strong bidder (producing the joint bid) always bids a smaller fraction of her valuation than the weak (individual)…”

Anglo-Dutch Premium Auctions in Eighteenth-Century Amsterdam

“…This has been shown in a variety of different models. In a private-values model, Maskin and Riley (2000) show that if one bidder has ex-ante stochastically lower valuations than his opponent, he wins less often, and earns lower payoffs, in an ascending auction than in a first-price or Dutch auction. Klemperer (1998) shows that if values are "mostly common" but one bidder has a slight private-value disadvantage, he cannot earn positive profits in an ascending auction, and drops out immediately in equilibrium, leading to low revenue, while he could profitably compete in a first-price auction.…”

“…In a descending auction, only a single bidder's bid determines the price; a bidder has no information about his opponents when he bids, and might, upon winning the auction and then learning what the other bidders had been planning to bid, realize he is worse off than if he had not participated. 33 In addition, in a private-values setting, ascending auctions allocate an object efficiently even in the presence of asymmetries among bidders, while descending auctions may not (Maskin and Riley 2000). Finally, ascending auctions are strategically simple; 34 auction complexity is often ignored by theorists, but could potentially deter less sophisticated bidders.…”

“…Le Moine de l'Espine (1801) stressed, for instance, that the size of the premium was in proportion to the value of the auctioned asset ("near de waerdije der geveilde Goederen"). 45 Evidence shows that in Amsterdam 44 The order of the sellers was determined by the order in which they had petitioned the government to participate, or via lottery (Le Moine de l'Espine 1801, pp 310-311). ACA, Willige verkopingen, inv.nr.…”

“…Securities sold together often had consecutive issuing numbers, suggesting they were purchased together at original issuance. 45 We have no reason to believe the premium varied systematically across sales other than with the size of the security; our empirical approach implicitly assumes that any variation in the premium (as a fraction of the security's face value) is independent of our explanatory variables.…”