Foundations for automated trading - It's the information that matters

Abstract

Business relationships, and the sense of trust that they embody, provides an environment in which trade may be conducted with confidence. Trading involves the maintenance of effective relationships, and refers to the process of: need identification, partner selection, offer-exchange, contract negotiation, and contract execution. So we use the term in a sense that includes the business of e-procurement. Of particular interest are: the selection of trading partners, the development of trading relationships, and the negotiation and execution of contracts in the context of a relationship. The potential value of the e-business market - including e-procurement - is enormous. Given the current state of technical development it is surprising that a comparatively small amount of automated trading is presently deployed.1 The technologies we refer to are first multiagent systems, and second the theory of competitive interaction. As a discipline, multiagent systems dates back to around 1975 when it was referred to as distributed artificial intelligence. The problems of building and deploying autonomous, intelligent agents are far from resolved but the technology has considerable maturity. The theory of competitive interaction, game theory, that principally dates back to the 1920s, has been developed mainly by the microeconomic community, and more recently also by the multiagent systems community in the area of agent-mediated electronic commerce. Why is it then that virtually no automated trading takes place? © Springer-Verlag Berlin Heidelberg 2005.