多物品网上拍卖的最优设计

摘要

Compared with traditional auctions, online auctions have some unique characteristics, which include bidders' stochastic arrival, snipping effects, the existence of auction platform and the fee charged to sellers' as well as multi-unit and multiple stages. Taking into account these unique characteristics, this paper explores some important questions of multi-unit online auctions systematically. The result shows that goods with larger valuation disparity are more suitable to be on sale by auction but under the precondition that the network user basement should achieve a certain scale; under a unified analysis framework, we solve the seller' s optimal public reserve price and private reserve price respectively, and also verify that the seller with private reserve price can gain more profits than the seller with public reserve price; considering the inconsistency between the auction platform and the seller' s objectives, we indicate that, to decrease the listing fee, increasing commission ratio and shortening the auction duration are both helpful for us to design the incentive compatible optimal auction mechanism; to auction more goods is not always optimal for the seller, we solve for the optimal number of goods to be sold in one auction under one stage and multiple stage auctions respectively.%考虑网上拍卖与传统拍卖相比的特性,包括竞标者随机到达,末尾抢标效应,拍卖平台的存在和收费,以及多物品多批次,系统地研究多物品网上拍卖中的一些重要问题.结果表明,竞标者估值离差越大的商品越适合通过拍卖来进行销售,但前提是网络用户基础必须要达到一定规模；在统一分析框架下,分别求解了卖者最优的公开保留价格和隐藏保留价格,并证明了隐藏保留价格能给卖者带来更多的期望利润；在考虑到拍卖平台和卖者目标不一致的条件下,研究表明降低陈列费增加佣金比例的同时缩短拍卖的持续时间有利于设计出激励相容的最优拍卖机制；由于在一次拍卖中拍卖的物品数量并非越多越好,分别在单批次和多批次拍卖下求解了卖者在单批次中应该拍卖的最优批量.