多双边拍卖下交易策略和竞争市场机制的研究

As the development of Internet economy, trading agents(buyers and sellers) can trade in multiple double auction marketplaces simultaneously. In this context, they need trading strategies to decide which marketplaces to participate in and how to bid in the selected marketplaces. Furthermore, in this competing environment, the market agents need to design effective market mechanisms to attract trading agents and make profits by charging fees to them.In this project, we intend to investigate the market selection and bidding strategies of trading agents, and the competing market mechanisms of market agents by considering two classical settings with budget constraint and value constraint. Specifically, in the context of multiple static double auctions, we use game theory to analyze the Nash equilibrium market selection and bidding strategies of trading agents with continuous and private types. Then in the context of multiple market agents competing with each other, we analyze how the market agents set their competing market mechanisms (in order to maximize the revenue) and how the trading agents select marketplaces and bid in the selected marketplaces in Nash equilibrium. Specifically, this is a multi-stage game where the strategies of market agents and trading agents are affected by each other. Finally, we design a practical trading algorithm and a competing double auction mechanism, and design a platform to evaluate them. This project will provide useful insights for designing practical trading algorithms and market mechanisms in the real-world Internet economy, and also can provide insights for the strategic interaction of self-interested Internet entities and the interaction protocol design of Internet authorities under the economic constraints.

随着网络经济的发展，交易智能体可同时在多个双边拍卖中交易，它们必须决定选择进入哪些市场，以及如何在市场中确定交易数量并报价；在多市场竞争环境下，市场智能体必须设置高效市场机制来吸引交易者并获得收益。本项目拟在预算约束和价值约束这两种典型的设定下对交易智能体的市场选择和报价策略，以及市场智能体的双边拍卖竞争机制进行分析。首先在多双边拍卖静态环境下，使用博弈论方法分析具有连续且私有类型的交易智能体的纳什均衡策略。然后在多市场竞争环境下，考虑市场机制和交易策略互相影响，在多阶段博弈下分析市场智能体的竞争市场机制和交易者的交易策略。在理论工作基础上设计实际的交易算法和双边拍卖竞争机制，开发交易仿真平台对交易算法和竞争机制进行评估。项目的研究成果对网络经济中交易算法和竞争市场机制设计有重要意义，并为Internet环境中遵循经济学约束下的自利实体交互策略设计和网络管理者的交互机制设计提供思路。

随着网络经济的发展，交易智能体可同时在多个双边拍卖中交易，它们必须决定选择进入哪些市场，以及如何在市场中确定交易数量并报价；在多市场竞争环境下，市场智能体必须设置高效市场机制来吸引交易者并获得收益。本项目拟在预算约束和价值约束这两种典型的设定下对交易智能体的市场选择和报价策略，以及市场智能体的双边拍卖竞争机制进行分析。首先在多双边拍卖静态环境下，使用博弈论方法分析具有连续且私有类型的交易智能体的纳什均衡策略，具体分析在预算约束下，考虑商品价值是累加、可替代、互补情况下的交易策略。然后在多市场竞争环境下，考虑市场机制和交易策略互相影响，在多阶段博弈下分析市场智能体的竞争市场机制和交易者的交易策略。对于双边拍卖市场机制的定价策略，我们发现差异定价策略和均衡定价策略都有助于双边拍卖市场高效分配资源。另外，我们发现差异定价有助于吸引更多的交易者。在收费策略研究中，我们分析不同类型的费用如何影响交易者的纳什均衡报价策略，并分析哪种类型费用有助于市场获得高利润并保留交易者。项目的研究成果对网络经济中交易算法和竞争市场机制设计有重要意义，并为Internet环境中遵循经济学约束下的自利实体交互策略设计和网络管理者的交互机制设计提供思路。