Pricing With Local Interactions On Agent-Based Electronic Marketplaces

Benoit Leloup*

^a Institute for Advanced Applications of the Internet (IAAI), 12, avenue du Général Leclerc, 13003, Marseilles, France
^b Ecole Nationale Supérieure des Telecommunications de Bretagne (ENST de Bretagne), Brest, France

Received 30 October 2002; revised 30 December 2002; accepted 28 February 2003. ; Available online 14 June 2003.

Abstract

This paper investigates the optimal pricing strategies of a selling agent that is randomly matched with several heterogeneous buying agents whose reservation prices are initially unknown. The seller perceives the behaviors of the buying agents through a logistic distribution with unknown parameters. We study the optimal learning by experimentation model of the logistic distribution. We extend this framework to a dynamic pricing model in which the selling agent is randomly matched with buying agents that are able to communicate their purchase experience to other buying agents. We carry out multi-agent system simulations of this dynamic pricing decision problem and we discuss some properties of the price dynamics one can observe on such marketplaces.

Author Keywords: Electronic commerce; Dynamic pricing; Discrete choice model; Autonomous agent; Neighborhood; Multi-agent system simulations

*Corresponding author. Present address: Institute for Advanced Applications of the Internet (IAAI), 12, avenue du Général Leclerc, 13003 Marseilles, France. Tel.: +33-4-9110-8576; fax: +33-4-9110-8561.

www.ABA-RD.org
"Agents for Business Automation" Ó Research & Development