Multiple overlapping online auction market: Bidder strategic mixture analysis using entropy

摘要

Online auction bidders are sensitive to the information revealed from other overlapping auctions as well as the auction they are bidding on. As the Internet Technology enables signaling of auction information across multiple overlapping auctions, it affects overall bidder behavior, more specifically, their entry and bidding strategies. In this paper, we take an approach of entropy to capture essential information regarding the mixture of heterogeneous bidder strategies. While prior research demonstrated the influence of surrounding market forces on various aspects of bidder behavior, we develop a concise yet more integrated proxy that contains comprehensive information about these aspects. We find that entropy state of an auction is endogenous to the overlap of multiple auctions. Given that an auctioneer has limited control over entry and bidding activities of bidders, identification of unique strategic pattern based on entropy state provides an effective estimate of the market revenue.