Ascending Bid Auctions with Sealed Bids
https://site.stanford.edu/sites/default/files/cho_s_empirical_analysis.pdf
This paper is about a type of auction that a rental car company uses to sell cars. In this auction, people don’t know the other bidders’ bids and they can keep submitting bids for a specific amount of time. At any point in time, if a bidder has the highest bid, (s)he will know. The person who has the highest bid at the end of the auction has to pay however much (s)he bid for the car. The fact that people can bid early and see whether or not they are the highest bidder appears to help people find out more information about the other bidders. However, often, people to whom the paper refers as “bid snipers” bid high at the end so any information gathered in the beginning about how people are bidding might not ending being that useful. Interestingly, the paper found that there was a higher probability that the winning bid would come later in the game. The paper also found that the addition of the ability to gain some information about other bidders in the early part of the auction actually ends up making people bid more than they would in an auction in which people only bid once. This is another example in which providing an additional option to people make them less happy, even though they don’t have to use that option. We saw this with some of the other games we discusses in class.
This article relates to concepts from this class because it discusses a type of auction and analyzes the different strategies that players could have during this type of auction. We talked about ascending bid auctions and first price sealed-bid auctions. The auction described in this paper is like a first price auction, except people can submit multiple bids. According to the article, this modification makes the dominant strategy close to the dominant strategy for ascending bid auctions. Since people can bid repeatedly, if someone were to keep incrementally raising his bid until either he reached his value of the car or he had the highest bid, then he would pay the value of the second highest bidder. Supposing person A does that, close to the end of the auction, someone could bid an amount that is less than A’s value of the car, but more than A’s current bid, which would mean A’s strategy was not optimal. In the paper, this auction was analyzed using similar methods to those we used in class to analyzes the auction about which we learnt.