Estimation Method Study Of The Value Distribution Of Online English Auction Objects

The auction is a special form of market mechanism,which do simliely reflect the price generation process.In essence,the auction is a game of incomplete information.Since the joint value distribution of participants effectively reveals information of both the buyers and the seller it is the basis of the price generation process.Research which combines auction value distribution with demand analysis is a breakthrough of applying auction theory into practice.In private value information environment,the joint distribution function of the order statistic of the value F,(·)completely contains the information about needs of participants.through which(the joint value distribution of the mining order)we can recover demand curve,and then evaluate whether the different market mechanism or the surplus allocation is efficient.However,it needs a large number of empirical data to do emperical research of joint value distribution.Online auction could provide abundant examples for the empirical research because of its wide range of goods,high trading frequency,In addition,various kinds of goods auctioned online makes the competition of both buyers and seller intenser,closely to the perfect competition market transactions.The joint value distribution of the Internet auction is of more practical significance.In order to obtain the joint value distribution of online auction bidders',we refer to classical auction theory of Haile and Tamer based on the assumption that bidders with private values bid independently.It is a empirical research about establishing the incomplete bid auction model which combines bid and bidders' evaluation closely.We deduce the upper and lower bounds of the joint value distribution function and estimate the bounds by using functional data analysis technique.Current researches about auction value distribution are to establish one-to-one mapping relationship between observable bid and bidders' value to identify the joint value distribution,and then to use parametric or semiparametric method to do estimation.Our method is more simple and practical compared with other research and make the identification and estimation process brief.In addition,we carry empirical analysis of Taobao second-hand iPhone auctions.The estimated upper and lower bounds well reflect the real value of the auction objects at tighter bounds of the estimated distribution area.In the range of 15%below the market price,the upper and lower bounds on the distribution fits well the value of the second-hand iPhone6 mobile phone distribution.