【摘要】 互联网的发展和数字经济的兴起催生了一系列新型网络拍卖机制和销售模式。本文关注付费竞价式拍卖这一新型机制,将预算约束引入拍卖理论模型,并进行严格的均衡刻画。通过考虑竞价是否免费和拍卖品价格是否提升两个维度,本文在二人拍卖的设定下将三类拍卖机制——英式拍卖、升价便士拍卖、序贯全支付拍卖纳入统一性分析框架,并利用实验设计与分析对理论预测进行检验。实验结果在商品售出率、拍卖持续期数、拍卖成交价格上均与理论预测排序相符。但收益等价性理论结果未得到实验结果支持,这是由于参与者在升价便士拍卖中表现出的沉没成本谬误为卖方创造了超额收益。但随着参与者经验的积累,三种拍卖形式下的卖方利润差异逐渐弱化。本文对于付费竞价式拍卖的研究,在数字经济成为国家重大发展战略的背景下具有重要的理论与实践意义。更多还原

【Abstract】 The rapid development of Internet technology and digit economy has given birth to many nascent online auction mechanisms, one of which is the pay-to-bid auction(also known as penny auction). The feature of penny auctions can be described as follows. The auction usually begins with a price of zero. Any bid placed within a pre-specified time limit extends the auction by resetting the time limit. Each bid costs the bidder a certain amount of bidding fee and increases the price by a fixed amount. The auction ends when there are no more bids before time expires, and the last bidder wins the object at the current price. It is noteworthy that although the transaction prices are typically much lower than the retail prices, the penny auction is remarkably profitable for the sellers by extracting bidding fees from a large number of bidders.On one hand, since the penny auction shares the feature of pay-to-bid with the all-pay auction, and the feature of dynamically increasing price with the English auction, it will be nice to have a unified framework that can incorporate all three auction mechanisms. On the other hand, given that there are many scenarios in the real life in which bidders face budget constraints, it will be useful to develop a penny auction model with budget constraint, which has not been done in the previous literature.In this paper, we propose a general setting on pay-to-bid auction, which allows bidders to have their own budget constraints and also includes free bidding and fixed price as special cases. Firstly, we characterize the equilibrium solutions in all three situations: no budget constraint, loose budget constraint and tight budget constraint, respectively. By doing so, we not only broaden the application scope of the auction model by making the model more realistic, but also enhance the existing theory for the situation of no budget constraint by finding a new set of symmetric Markov perfect equilibrium. Secondly, since the three auction mechanisms unified in our setting, namely, the ascending-price free-to-bid auction, the ascending-price pay-to-bid auction, and the fixed-price pay-to-bid auction, will become the English auction, the ascending-price penny auction, and the sequential all-pay auction when there are only two bidders, we conduct theoretical comparison among these three auction mechanisms in a 2-bidder setting. Lastly, to further test the validity of our theory, we study the bidding behavior using college students as subjects under different auction settings in a laboratory environment with monetary incentives.Our findings are as follows.(1) In terms of the transaction rate and auction duration, English auction>ascending-price penny auction>sequential all-pay auction;(2) in terms of the transaction price, English auction>ascending-price penny auction>sequential all-pay auction; which are both consistent with the ranking order predicted by the theory. However,(3) the theoretical result of revenue equivalence is not supported by the experimental data. The ascending-price penny auction can bring excessive profits to the auctioneer, while the sequential all-pay auction with a tight budget constraint makes negative profits. In comparison, only the results of English auction are aligned with the theoretical prediction. Yet, the learning effect would bring the outcome closer to the equilibrium, mitigating the revenue difference among the three auction mechanisms over time.Currently, many industries in China, including the traditional auction industry, are going through a digital transformation. While under a set of standard assumptions, no auction mechanism is able to yield positive profits from the theoretical perspective, our experimental data along with the existing empirical data do suggest that bidders are likely to suffer from sunk cost fallacy, aggressiveness, and some other irrational behavioral biases, resulting in the transaction outcomes greatly deviating from the theoretical prediction. For this reason, the government should evaluate the underlying operation principle and the potential behavioral bias which may be induced by the auction mechanism before it is implemented. Some important factors such as efficiency, fairness, sustainability and stability should also be taken in account. Furthermore, in order to meet people’s continuously increasing demand on material consumption and legal rights, it is necessary and important for the regulators to establish the relevant legal procedures regulating Internet auctions, as well as to cultivate the consumers’ risk perception of the online activities.更多还原