“Asymmetric Sequential Auctions”
Ro’i Zultan, Ella Segev, Ben-Gurion University of the Negev
Auctions are all around us, from government procurement procedures to on-line platforms such as eBay. An auction designer—be it the government or an individual—seeks to achieve the best outcome in the auction. If the bidders in the auction are not evenly matched, competitive pressure is low and the auction may not yield the desired outcome. In our experiment the theoretical analysis suggests that it may be beneficial to give the weak bidder(s) an advantage in the form of allowing them to bid after observing the bid(s) of stronger player(s). This may not only increase expected revenue but also increase the weak bidder’s expected payoff and can thus serve as a tool to encourage participation of weak bidders. We study sequential auctions, in which two or three bidders place bids one after another, with all previous bids observable to the current bidder. Our experiments will test whether such sequential bidding is better suited to auctions with asymmetric bidders than simultaneous auctions, in which all bidders place their bids without knowing the others’ bids. We also test how sequential auctions fare when the auction designer only has partial information about who is the stronger and who is the weaker bidder. Specifically, we predict that identifying the strongest bidder—and placing this bidder first—is sufficient to obtain a satisfactory outcome in terms of revenue.