Based on the timing of the bidding protocol, double auctions can be classified as synchronous (or synchronized ) or asynchronous double auctions. In a synchronized double auction, all agents submit their bids in lockstep. Bids are ``batched'' during the trading period, and then cleared at the end of the period. This type of auction can be seen in a clearing house. In an asynchronous double auction, also called a continuous double auction, agents offer to buy or sell and accept other agents' offers at any moment. Continuous double auctions have been widely used in stock exchange markets [Friedman1993] and Internet auctions. The auctions we have designed agents for are synchronized double auctions.
The book edited by Friedman and Rust [Friedman & Rust1993] collects several studies of double auctions, including both simulations and game-theoretic analyses. Game-theoretic studies [Satterthwaite & Williams1993] [Friedman1993] on double auctions generally adopt the framework of static (one-shot) games with incomplete information, for which the equilibrium solution is Bayesian Nash equilibrium. Double auctions are essentially dynamic games. Since agent interaction takes more than one round, the static game framework fails to address the basic dynamics of the system. Other theoretical studies [Easley & Ledyard1993] try to explain the experimental data generated from human subjects. They assume that each buyer or seller has a reservation price and has a way to recalculate its reservation price after trading. While the study of human behavior is interesting, we are more interested in designing artificial agents who can bid as intelligently as possible to get maximum payoffs.
Gode and Sunder [Gode & Sunder1993] designed zero-intelligence agents who submit random bids within a range such that their utilities never decrease. To improve upon zero-intelligence agents, Cliff [Cliff1998] designed zero-intelligence-plus agents who submit bids within the utility increasing range, but the bids are chosen so that their utilities will increase by some proportion which is adjusted over time. The effectiveness of the learning depends on several parameters including the learning rate. Cliff implemented a genetic algorithm (GA) to let the agent learn about these parameters. The training for the GA requires the agent to know the final convergence price of the whole auction. It is not clear how such GA training can be applied to online settings.
Other types of intelligent agents have also been designed for double auctions. Park et al [Park, Durfee, & Birmingham1998] designed an adaptive p-strategy agent to participate in continuous double auctions. Her experiment showed that the p-strategy out-performed other strategies. In the Santa Fe Tournament [Rust, Miller, & Palmer1993], 30 different intelligent programs competed in a synchronized double auction. A simple non-adaptive agent won that tournament by always waiting in the background and letting the others do the negotiation. When the bid and ask prices were sufficiently close the agent would jump in and steal the deal. Such an agent is not applicable when there is no negotiation process in the auction.