Are Incremental Bidders Really Naive Theory and Evidence from Competing Online Auctions.ppt

上传人:时光煮雨 文档编号:20347398 上传时间:2023-05-14 格式:PPT 页数:19 大小:1.04MB
返回 下载 相关 举报
Are Incremental Bidders Really Naive Theory and Evidence from Competing Online Auctions.ppt_第1页
第1页 / 共19页
Are Incremental Bidders Really Naive Theory and Evidence from Competing Online Auctions.ppt_第2页
第2页 / 共19页
Are Incremental Bidders Really Naive Theory and Evidence from Competing Online Auctions.ppt_第3页
第3页 / 共19页
亲,该文档总共19页,到这儿已超出免费预览范围,如果喜欢就下载吧!
资源描述

《Are Incremental Bidders Really Naive Theory and Evidence from Competing Online Auctions.ppt》由会员分享,可在线阅读,更多相关《Are Incremental Bidders Really Naive Theory and Evidence from Competing Online Auctions.ppt(19页珍藏版)》请在三一文库上搜索。

1、Are Incremental Bidders Really Naive Theory and Evidence from Competing Online Auctions Incremental-bidding Bid more than once in one auction Proxy-bidding Submit the maximum amount he would like to bid and let eBay bid for himMotivation Literature Most papers treat online auctions as independent of

2、 one another.Proxy-bidding system is commonly used.In real world Simultaneous or overlapping online auctions are not rare.The portion of incremental bidders is large.Related Literature Analysis of incremental-bidding Considered as non-equilibrium strategies(Ockenfels and Roth,2006)or strategies of n

3、aive bidders who know little about their valuation(Hossain,2008)Peters&Severinov(2006)Empirical studies on online auctions Anwar,McMillan&Zheng(2006)Questions Can we find a reason for incremental bidding behavior?Do the bidders benefit from this strategy?How does late bidding relate to incremental b

4、idding and concurrent auctions?Data Items:Wii console and Wii bundle Every auction ended between 06/04/2008 0:00 am and 06/12/2008 0:00 am Bidding behaviors from the time when the first auction started 4256 observationsFrequency of AuctionsAuctions with No Incremental Bidders2.72%Average Interval b/

5、w Auctions162.7 sAuctions with Concurrent Auctions(within 2 mins)86.2%Auctions with Concurrent Auctions(within 5 mins)97.5%Average Number of Concurrent Auctions(2 mins)2.23Average Number of Concurrent Auctions(5 mins)5.69Wii Console Sold on eBaySummary Statistics of Auction CharacteristicsCharacteri

6、sticsPercentageFree Shipping93.56%Starting Price S).Each auction has a minimum initial bid mi.Bidders arrive sequentially and bidder is private value vi is drawn i.i.d from some distribution F.Each buyer can observe the bidding history and the standing bid of all the auctions.It takes a little time

7、for a bidder to react another bidders bid.The standing bid of each auction is the second highest bid received on it.There is a smallest increment d by which bids must be raised.Simultaneous Auctions with Flexible Ending Time Lemma 1:When there are more than one auction,proxy-bidding strategy may not

8、 construct an equilibrium.Theorem 1:There can exist perfect Bayesian equilibrium in which every bidder plays the completely incremental bidding strategy.Simultaneous Auctions with Fixed Ending TimeTheorem 2:It is a perfect Bayesian equilibrium for all bidders to use the revised incremental bidding s

9、trategy,in which everyone bids his true value at the last minute.Theorem 3:If some incremental bidders fail to bid their true value at the last minute,late bidding could be more profitable.A buyer is more likely to snipe if S is small.Predictions from the Theory Winning probability:Incremental Proxy

10、 Incremental Sniper Winning price:Sniper Incremental Proxy Switching among auctions Lower winning priceDistribution of Bidders 1:snipers 2:incremental 3:proxy-incremental 4:incremental-proxy 5:proxyProbit Regression of Winners on Bidder TypesVariablesAll BiddersSerious BiddersConstant-7.089(0.096)-7

11、.062(0.098)Highest bid0.019(0.000)0.019(0.000)Dummy for Last-Minute0.386(0.026)0.386(0.026)Dummy for Incremental0.452(0.025)0.453(0.025)Pseudo R-sqr0.360.26Effects of Cross-Bidding:Dependent Variable:realprice=price+shipcost w/o Fixed Effects w/Fixed Effects VariableEstimatet ValueEstimatet ValueInt

12、ercept304.4629.25316.7627.03cross-bidding-3.25-3.21-3.27-3.27buyer_score0.00-1.510.00-1.36Seller_rate60.315.8460.055.85duration0.130.370.200.59start_price0.1825.530.1825.46peak hours-3.27-3.12-#of bids0.548.810.559.10#of bidders0.221.370.080.49dummy for bold title0.620.580.310.30dummy for featured i

13、tems35.3526.3136.7027.28dummy for wii conslole-80.15-90.2-79.84-90.30adjusted R-square0.7332 0.7392 Effects of Cross-Bidding by Groups(w Fixed Effects)Dependent Variable:realprice=price+shipcostEstimates on Auctions Characteristics Omittedgroup 1:sniper group2:proxy group3:incrementalVariableEstimat

14、et ValueEstimatet Valuegroup1-8.81317-8.94-9.14383-8.35group20.724240.710.109540.09group1*crossbid -4.45877-2.25group2*crossbid -2.18-1.13group3*crossbid -4.62213-3.25Effects of Number of Concurrent Auctions on Sniping Regression of the number of snipersVariableEstimateEstimateIntercept1.049(0.199)1

15、.081(0.199)standing bid-0.002(-0.000)-0.003(-0.000)#of concurrent(5 mins)-0.011(0.006)-#of concurrent(2 mins)-0.031(0.012)Conclusion Incremental bidders will pay less than proxy bidders when they really understand the advantage and of this strategy and make use of it.When more auctions are held together,bidders are less likely to snipe.

展开阅读全文
相关资源
猜你喜欢
相关搜索

当前位置:首页 > 社会民生


经营许可证编号:宁ICP备18001539号-1