Even though information cascades with and without direct effects have dominated our classroom discussion lately, this entry is going to go back to our discussion of auctions. The article discusses some of the accomplishments of William Vickery (the William Vickery of VCG prices) and how he integrated the concepts of game theory and auctions. In multiple contexts, we have affirmed that in a second price auction with 2 bidders, it is a dominant strategy to bid your true value. When I first heard that, I took the fact as a given; however, this article opened my mind to the fact that an auction is just one big game, each person reacting to the actions of other buyers. What I failed to consider is that the auction is a bilateral game. Bidders must choose which strategy is best to maximize their utility. However, sellers also face a decision as to which type of auction to use to maximize payoff. As is obvious from our dissection of different auctions, clearly the choice of auction to employ is directly related to bidder behavior. The seller should choose the auction mechanism that gets buyers to reveal their true value. The question that arises then is what auction to employ?

As noted in our text and by the article, there are four general types of auctions: the English ascending bid auction, the Dutch descending auction, the first price auction in which the buyer pays what he or she bid and the second price auction in which the buyer does not pay his or her bid but the bid of the second highest bidder. We assumed that all of these auctions were in the “private” context in which each bidder’s value is independent of the other bidders’ values (with common values, we get into the context of the “winner’s curse”). What I learned in this article is that the Second Price bid auction was actually designed by Vickery himself. In my reading of our text, I thought that Vickery played an instrumental role in studying the effects of such an auction, but I failed to recognize that he actually generated the auction that has served as the basis for much of our studies. The article gives an example (much like many that we have gone over in class) why a bidder’s best strategy is to bid his or her true value no matter what the bidding behavior is of other bidders. If you bid below your true value, there is a chance that the other bidder will obtain the object when you actually valued it more and could have gained a greater surplus. If you bid above your true value, there’s a likely chance that you may overpay for the auctioned item. So Vickery’s created a system in which buyer behavior was consistent and predictable: always bid your true value. But why would a seller want to use this method? Sure a buyer will reveal his or her true value, but the seller doesn’t get this price, the seller only receives the second highest price. It would seem more logical for the profit-maximizing seller to use a first-price auction framework whereby the winner pays whatever his or her bid was.

However, such a conclusion would ignore the fact that the buyer’s bid would then not only be a factor as to whether the buyer obtains the good but also determine the price he or she pays. In the second price framework proposed by Vickery, the price was an exogenous factor not determined by the bidder, so the buyer only had to worry about obtaining the good. Therefore, the rational bidder would bid his or her true value. If that bidder valued it the highest, he or she would obtain the auctioned item and wouldn’t have to worry about overpaying. If the bidder did not get the item, then he or she would intuit that another bidder just valued it more. Under a first price auction scheme, the dominant strategy to bid your true value does not exist. If it were, there would be no chance of consumer surplus as seen in second price auctions. As discussed in our text and reiterated in the article, those competing in a first price auction will “shade” their bid to something lower than their true value. The level of shading depends on the number of other buyers in the auction with bids approaching true values as the number of bidders increases. Though the level of shading varies in different contexts, Vickery found that (contrary to initial beliefs) sellers generate the same revenue in a first-price sealed bid auction as well as a second price auction. At first, such a conclusion may seem surprising, but once considering bidder behavior, it becomes clear. Under the second price auction, buyers have no incentive to hide their true values, so even though the seller won’t receive the highest bidder’s value, he or she will receive that of the second highest bidder. Under the first-price scheme, buyers have an incentive to hide their true values, so the seller will receive a discounted amount of the highest value which is exactly equal to the revenue obtained under the second-price scheme.

The most interesting result is that under certain circumstances that all auctions generate the same seller revenue as long as the item goes to the bidder who values it the most and there is no fee or reward for entering in the auction. These two caveats to Vickery’s auction theory are what keep research into auctions alive. In non-theoretical auctions, we observe these caveats in action. For example, in Ebay or in auctions for airwave frequencies, the buyers have to pay a fee to participate. The initial fee may extort behavior and may eliminate some of the high-value buyers thus altering seller revenue. Furthermore, the Vickery auction scheme does not consider redistribution in auctions. Perhaps the seller provides a bid matching scheme for some bidders rather than others which would greatly affect seller revenue. All in all, Vickery introduced an auction framework that was molded to form the basis of auctions we see used by many companies such as Ebay and Google. The 1996 Nobel prize winner has definitely contributed to a profitable form of e-commerce as well has added an interesting component to network analysis and game theoretic motives of profit maximization.

http://www.beyonddiscovery.org/content/view.page.asp?I=3685