We present a framework for dening trading scenarios based on sh market auctions. In these scenarios, agents of arbitrary complexity can participate in electronic auctions under a collection of standardized market conditions and be evaluated against their actual market performance. We argue that such competitive situations constitute convenient problem domains in which to study issues related with agent architectures in general and agent-based trading strategies in particular. Our proposal involves a set of conventions for the typication of goods, bidding protocols, availability of goods, buyer endowments and performance evaluation criteria. The proposed framework is implemented as a multi-agent testbed which is an extension of FM96.5 {a Java-based version of the Fishmarket auction house. A simple tournament is used to illustrate these elements.
Juan A. Rodríguez-Aguilar, Francisco J. Mar