Essays in Modeling the First Bid in Retail Secondary Market Online Auctions: A Bayesian Approach Öffentlichkeit
Herunterladbarer InhaltPDF Herunterladen
The online commerce has greatly changed the trading markets for both businesses and consumers. A big part of this environment is the way online auctions are conducted. In this dissertation, I develop a series of models to study some of the elements relevant to the formation of first bids in the secondary market online auctions.In the first essay, I propose models to study bidders' participation in auctions and their first bidding behavior in their participated auctions. Through these models I predict two major elements of given auctions: who will bid and who will place the first bid. I apply these models to different sets of auctions and compare their fit measures and predictive performances. In developing the models, in addition to some auction-specific characteristics, I consider some time-varying bidder's characteristics as well which show how bidders change their strategies by learning from their past activities. In addition, I show how the information from other overlapping auctions may influence bidders' decisions in participating in the focal auction and the time they place their first bid. In the second essay, I focus on the time that auctions receive their first bid and study the effect of auctions features on arrival of the first bid. I propose a model of finite mixture of beta distributions to model the time to the first bid. The developed model helps in studying the auctions that receive their first bids very early or very late and how this may impact their dynamics. I also discuss its predictive aspects and its managerial implication.