Dissertation Research: Bidding Behavior in Internet Auction Markets

互联网拍卖市场是当今最具活力和令人兴奋的交易机构之一。典型的在线拍卖是一种复杂的机制，为卖家提供了许多设计选择。此外，拍卖很少是单独存在的。例如，在互联网上最大的市场eBay上，每天都有数百个相同或相关物品的拍卖结束。因此，买方面临着众多的选择，她对如何出价的决定远非简单。这项博士论文改进奖将资助对两种理论的研究，这些理论是关于市场力量如何影响人们可能在网上找到的拍卖中的出价激励。这些理论将通过实验室实验进行验证。从eBay和其他拍卖行收集的数据建立了两个强有力的经验规律。例如，一组竞标者倾向于在拍卖后期出价，而另一组则喜欢逐步修改他们的出价。我们认为，这些竞价模式可能是买家对拍卖市场产生的激励做出理性反应的产物。虽然我们的理论在很大程度上与经验证据一致，但本项目的主要目标是使用实验方法进一步检验其稳健性。我们的实验将为竞标者如何应对互联网上普遍实施的复杂拍卖机制带来新的见解。他们将为正在进行的关于观察到的竞标模式的解释及其对拍卖市场表现的总体影响的辩论做出贡献。这项研究的发现将使我们能够在众多关于互联网拍卖中竞价行为的解释中进行选择。基于我们的结果，我们将能够评估替代模型的经验有效性，并为设计更好、更有效的拍卖市场提供启示。

Internet auction markets are one of the most vibrant and exciting trading institutions of today. A typical online auction is a complicated mechanism offering many design options for the seller to choose from. Furthermore, auctions rarely stand alone. For example, on eBay, the largest marketplace on the Internet, hundreds of auctions for the same or related object close each day. Thus, the buyer is confronted with multitude of options and her decision on how to bid is far from being simple. This doctoral dissertation improvement award will fund an examination of two theories about how market forces influence bidding incentives in the kind of auctions one might find online. The theories will be tested using lab experiments. The data collected from eBay and other auction houses establish two strong empiricalregularities. For example, one group of bidders tends to bid late in the auction while other group likes to progressively revise their bids. We argue that these bidding patterns could be a product of rational response of buyers to the incentives that auction markets generate. Although our theories are largely consistent with the empirical evidence, the main goal of this project is to use experimental methods to further test their robustness. Our experiments will bring new insights into how bidders respond to complicated auctionmechanisms that are commonly implemented on the Internet. They will contribute to the ongoing debate about the interpretation of observed bidding patterns and their overall impact on the performance of auction markets. Findings from this research will allow us to select between numerous explanations that have been proposed about bidding behavior in Internet-type auctions. Based on our results we will be able to evaluate the empirical validity of alternative models and provide implications for designing better and more efficient auction markets.