【作者】 潘超杰；

【导师】 张勤生；

【作者基本信息】 中国海洋大学 ， 管理科学与工程， 2010， 硕士

【摘要】 网上拍卖作为传统拍卖的延伸和发展,在电子商务领域发挥着越来越重要的作用。网络拍卖从拍卖物品的数量上可以分为单物品拍卖和多物品拍卖。近年来多物品网络拍卖发展极其迅速,网上交易额数以百亿计,网上拍卖的方式也不断推陈出新,但多物品拍卖尚未形成一套成熟的理论体系,拍卖效率高低不一。如何设计拍卖规则使得买卖双方利益最大化的研究显得尤为重要。本文以网上拍卖多物品拍卖为研究背景,以传统拍卖理论,博弈论等为研究工具,探讨C2C多物品最优拍卖问题。首先,对国内外的拍卖理论和发展现状进行对比、分析和综合研究,采用定性判断与定量计算相结合、运用收入等价理论、博弈论、最优化理论等方法,对实际拍卖过程进行分析研究。然后,在各种机制模型的基础上,引入机制设计与显示原理,并对多物品的各种拍卖机制进行分析,得出各种机制的最优拍卖形式。接着,探讨了多物品网络拍卖机制——逢低买入机制,针对该机制建立了相应的模型。最后,在以上基础上对此机制进行分析,得到了逢低买入机制的最优策略。通过研究,获得了以下研究成果与研究结论：本文在同价拍卖分析中通过数据验证了VCG拍卖机制激励相容,对每个投标者来说真实地宣布自己VCG的估价是最优策略,而且其他投标者的投标对他自己的效用没有影响。在命题3.1的基础上本文进一步分情况讨论得出：多物品序列拍卖无论是否公开保留价都不影响卖方的最优分配策略以及获得的相应期望总利润。且无论是否设置保留价,卖方应最优选择比较小的投标最小增量。本文基于多物品收益等价定理对序列拍卖采用英式拍卖分析得出：在第一场拍卖中,对于卖家设定的保留价格,会存在一个大于该保留价格的阈值—(?)。且存在最优策略SS,P即如果竞价者估价大于该阈值,那么竞价者真实竞价；否则,竞价者不参加第一场拍卖,第二场拍卖中,只要竞价者估价不小于卖家设定的保留价格,就会参加拍卖,并且真实竞价；在连续两场的英式拍卖中,存在唯一的对称的纯战略均衡----精炼的贝叶斯纳什均衡,就是竞价者采取半参与策略Sb,P。在竞价者对单个物品估价服从二项分布且两种物品只有两种可能估价的假设下,本文基于得出的多物品最优拍卖的三个重要性质对独立拍卖,绑定拍卖及混合拍卖机制进行分析得出：根据(?)是否成立,及各种类型投标者期望收益Rij的不同有着不同的最优策略组合。在定理5.2.1的基础上本文结合具体数据分析得出逢低买入存在最优策略S,即估价v低于qN不投标,否则投标θ(v)。本文共分为六章进行讨论：本文第一章阐述了论文选题的背景与意义,探讨了国内外的研究现状,并对论文的研究方法、研究思路作了简要的说明。第二章是对网络拍卖进行了概述讲解,包括拍卖的定义和分类,公有估价模型和私有估价模型,拍卖的基本形式,网上拍卖的起源和发展以及与传统拍卖的区别。第三章是对重点分析了同质多物品、异质多物品以及多物品网络拍卖机制。第四章在前面各种机制模型的基础上,引入机制设计与显示原理,并对多物品的各种拍卖机制进行分析,得出各种机制的最优拍卖形式。第五章探讨了一种多物品网络拍卖机制——逢低买入机制,并针对该机制建立了相应的模型,最后在前文的基础上对此机制进行分析从而得到了逢低买入机制的最优策略。第六章是对全文的总结,总结了全文的结论并对未来的研究提出了一些方向性建议。更多还原

【Abstract】 Online auction, as a extension and development of traditional auction, plays a more and more important role in e-commerce field. Online auction can be divided into single-item auction and multi-item auction based on the number of items. In recent years, the development of multi-item online auctions is extremely fast, online transactions are tens of billions of dollars, online auction patterns have also continued to emerge. But multi-items auction theory has not yet formed a mature system, with different efficiency. How to design the auction rules allow maximum benefit of buyers and sellers is very important. In this thesis, based on multi-item auction online auction research, with the traditional auction theory, game theory as the research tool to explore the issue of optimal mechanism auction of C2C multi-items auction.First, compared with the development of domestic and international status of the auction theory, analysis and synthesis of research, I adopt qualitative judgments and quantitative calculation, use the revenue equivalence theory, game theory, optimization theory methods, and analyze the actual auction process. Next, the thesis brings into mechanism design and display principle based on the models in a variety of mechanisms, analyzes multi-Item various auction mechanisms, obtains the optimal auction mechanisms. Following, it discusses multi-item auction mechanism-Buying a mechanism to establish the mechanism for the corresponding model. Finally, this mechanism based on the above analysis, it obtains the optimal strategy of Group-buying auction mechanism.Sequential auctions, in the first auction, for the seller’s reservation price, there is a threshold- bigger than the reservation price.Whether to set the reserve price or not, the seller should Optimally choice relatively smaller Minimum bid increment.Sequential auction exists optimal strategy Ss,p, that if the bidder’s valuation is bigger than the threshold, then the bidder will real bid; otherwise, the bidders will not participate in the first auction; while in the second auction when bidders’valuations are not less than the seller’s reservation price, the bidders will participate in the auction with the real bidding. In two successive English auction, there exists a unique symmetric pure strategy equilibrium----Refined Bayesian Nash equilibrium, that is bidder takes semi-participation strategy Sb,P. Sequential auctions for multiple items, whether or not open the reserve price, will not affect the seller’s optimal allocation strategy and access to the corresponding expected total profit.Two different item auctions in the independent, binding auction and mixed auction will have its own mixed optimal strategy. According to sets up or not, and the different expect profits with all kinds of bidders, it exists the different optimal strategy combination.Under the assumption of the baseline model, that the same as in the independent private model, if the bidder’s risk is neutral and symmetric, the optimal strategy of Group-buying mechanism is not bidding with the valuation v less than qN otherwise biddingθ(v).This thesis is divided into six chapters:The first chapter of this thesis mainly introduce the background and significance of the topic, research status at home and abroad, a brief description of the research methods and research ideas.The second chapter is a brief overview of the online auction, including the definition and classification of the auction, public valuation model、private valuation model, the basic form of the auction, and the origin and development of online auctions as well as the difference between the traditional auction and online auction.The third chapter is focused on the homogeneous multi-items, heterogeneous multi-items and multi-items auction optimal mechanism.The fourth, the thesis brings into mechanism design and display principle based on the models in a variety of mechanisms, analyzes multi-Item various auction mechanisms, obtains the optimal auction mechanisms.Chapter V introduces a multi-items auction mechanism-Group-buying mechanisms and making the corresponding model. The thesis analyzes the corresponding mechanism based on foregoing chapters, conclude the optimal strategy about the Group-buying mechanism.Chapter VI is a summary of the full text. We summarize the full text and put forward some suggestions about the future research.更多还原