Identifying Demand from Online Auction Data

38 Pages Posted: 4 Dec 2009

Date Written: December 2, 2009

Abstract

The paper presents a framework for estimating a demand system from price data generated by a large scale auction platform such as eBay. Auction prices are used to identify characteristics of the joint distribution using an order statistics like approach where the bidder's "revealed preference" in equilibrium enables inference of valuations. Two approaches are presented for identifying the full joint distribution. Sklar's Theorem is used to show the complete joint distribution can be inferred from combining information from the first step with information on the marginal distributions inferred from prices of auctions with no competition. A multiple market model is used to show variation in the number of auctions for the same product can be used to provide bounds on the joint distribution. The paper presents additional results for when the number of participants is unobserved, when the number of competing auctions is unobserved and when there is observed bidder heterogeneity.

Keywords: demand, differentiated, products, copula, ebay, auctions

JEL Classification: C14, D44

Suggested Citation

Adams, Christopher, Identifying Demand from Online Auction Data (December 2, 2009). Available at SSRN: https://ssrn.com/abstract=1517442 or http://dx.doi.org/10.2139/ssrn.1517442

Christopher Adams (Contact Author)

CBO ( email )

Ford House Office Building
2nd & D Streets, SW
Washington, DC 20515-6925
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
105
Abstract Views
777
Rank
463,225
PlumX Metrics