Correlated Signals Against Monotone Equilibria

27 Pages Posted: 29 May 2000

See all articles by Michael Landsberger

Michael Landsberger

University of Haifa - Department of Economics

Boris Tsirelson

Tel Aviv University - School of Mathematical Sciences

Date Written: April 2000

Abstract

We consider symmetric auctions that may be multi-unit, with multi-dimensional bids and correlated multi-dimensional signals. Payment and allocation mechanisms are quite arbitrary. There are n potential bidders and the process of submitting a bid involves cost, which may be random and may vary across bidders. The number of effective bidders is part of an equilibrium. In fact, our model addresses more general games of which auction games are a special case. However, given the importance of auctions in economic theory and their prevalence in reality, they can be considered as a perfect archetype of our model.

We prove that if n is sufficiently large and signals behave in a burst mode (which appears to be rather typical under interdependence), there is no equilibrium supported by monotone strategies. Since results obtained in the auction literature rely on monotonicity of strategies, even very basic results, such as existence and uniqueness of a symmetric equilibrium, become open questions, even for single unit first price auctions.

We establish some properties of equilibria. They seem very innocuous and yet, they have power in the sense that they are incompatible with monotone strategies.

JEL Classification: D44

Suggested Citation

Landsberger, Michael and Tsirelson, Boris, Correlated Signals Against Monotone Equilibria (April 2000). Available at SSRN: https://ssrn.com/abstract=222308 or http://dx.doi.org/10.2139/ssrn.222308

Michael Landsberger (Contact Author)

University of Haifa - Department of Economics ( email )

Haifa 31905
Israel
972-4-8240023 (Phone)
972-4-8240059 (Fax)

Boris Tsirelson

Tel Aviv University - School of Mathematical Sciences ( email )

Tel Aviv 69978
Israel
972-3-6409638 (Phone)

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

Paper statistics

Downloads
206
Abstract Views
3,225
Rank
269,741
PlumX Metrics