A Methodology for Building Trading Agents in Electronic Markets

In this paper we present a methodology that guides the development of a multiagent system for trading in electronic markets. Our approach uses a strategy of decomposing a complex trading problem into sub-problems, and defines agent roles that tackle these sub-problems. We also present an incremental development process for building asynchronous and distributed agents, which enable effortlessly the combination of intelligent agent strategies. This creates a highly adaptable system for unknown situations that is easy to scale up. We use the Trading Agent Competition (TAC) environment as a case study to illustrate the suitability of our approach.