Reserve prices in repeated auctions (Q1742150): Difference between revisions

This paper presents a theoretical analysis of equilibria in repeated auctions in which the seller attempts to learn the distribution of bidders' values so as to set better reserve prices in the future and strategic bidders anticipate this and use this to refine their bidding strategies. The author's research follows this model: In each time period \(t\in \{1,\ldots, T\}\), an infinite number of auctions are held in which \(n\) bidders participate. In each auction in each time period, each bidder's value \(v_i\) is an independent and identically distributed draw from the c.d.f. \(F(\cdot |\theta)\) with corresponding continuously probability function \(f(\cdot |\theta)\) and supports \([0,\infty]\). Here \(\theta\) is a random variable drawn from the c.d.f. \(G(\cdot)\) with corresponding continuous probability function \(g(\cdot)\) and supports that \([\underline {\theta},\overline {\theta}]\) is common in all periods. The realization of \(\theta\) is known for the bidders but not to the seller, and that for any \(\theta \in [\underline {\theta},\overline {\theta}]\), \(F(v|\theta) = F((\underline {\theta}/\theta)v|\underline{\theta})\). A precise analysis of the important ``supersaturation'' characteristic of the cumulative sigmoid is given. In general, the topic is interesting because research has a wide application in the area of financial mathematics and insurance mathematics.