Speculation agents for dynamic multi-period continuous double auctions in B2B exchanges

Business to business (B2B) exchanges provide opportunities for companies to streamline their supply chains in dynamic business situations, but they also create new management challenges. Managers are faced with more frequent buying and selling decisions, more available information, and even new problems such as speculation. Due to these new challenges, trading support systems play an important role in helping companies to achieve maximum profits in B2B exchanges. In this paper, we focus on providing support for two core capabilities, bidding and speculating, in the context of B2B exchanges with continuous double auctions (CDA). Our work extends previous research on bidding agents to more dynamic and realistic B2B exchange situations where both demands and supplies change from period to period. Changes in demand and supply cause price fluctuations and motivate companies to speculate on inventory. The action of speculation is one fundamental aspect of exchanges in general, and is the main way in which B2B exchanges can hedge against costly shortages. We develop a multiagent system in which a speculation agent makes speculation decisions for a given buyer or seller and interacts with a set of bidding agents. The related algorithms are discussed in detail. We also propose a theoretical model for the economic equilibrium in this dynamic situation. Under various market conditions, the experiments show that our system significantly increases the overall profit of the entire market.