Optimal Bidding in Sequential Online Auctions

Auctions are widely used online to conduct commercial transactions. An important feature of online auctions is that even bidders who intend to buy a single object frequently have the opportunity to bid in sequential auctions selling identical objects. This paper studies key features of the optimal bidding strategy, assuming rational, risk-neutral agents with independent private valuations and sealed-bid second-price sequential auctions. In contrast to previous work on this topic, we develop our theory using the concept of the “option value” of an upcoming auction – a measure of the expected payoff from being able to participate in a future auction. This option value depends, among other things, upon the mean and variance of the future number of bidders. We derive an optimal bidding strategy in sequential auctions that incorporates option value assessment. Furthermore, we establish that our optimal bidding strategy is tractable since it is independent of the bidding strategies of other bidders in the current auction and is only dependent on the option value assessment. We test and find support for our theory using data collected on 327 eBay auctions on digital cameras in first two months of 2001.