| dc.description.abstract | Transactions through Internet auction have become more popular today. The considerable and available data on auction websites (either eBay or Yahoo) attract the economists’ attentions to review the general auction theory, which include theoretical and empirical studies. There are three main topics in the essay. The first two topics (chapter 2 and 3) empirically use mobile phone auctions from Yahoo auctions website. Further, the data from eBay motors’ auctions is applied in the third topic (chapter 4).
The focus in chapter 2 might be the research foundation in Internet auctions. The available information on auction sites are classified into four parts, which are seller’s reputation, price information, item’s information, and other possible information, and these four information categorizations are the independent variables in the estimations. Otherwise, the number of received bids, the probability of auction success, and the auction transaction price may be the three important outcomes in online auctions. Consequently, due to different types of dependent variables, three different econometric models are applied. And different information types indeed have different effect on the variables of auction outcomes.
In chapter 3, the study re-tests the effect of seller’s reputation on the transaction price. In the regression, the other control variables are still contained in the empirical model. Two-step estimation method is applied to conclude that the sellers’ reputation may have no effect on the transaction prices. However, it affects the dependent variable in the first stage, which is the probability of auction success. Hence, a different conclusion compared to the previous literatures is taken by the result of this chapter.
Finally, the study uses the data from eBay Toyota motors’ auctions to test the famous auction theorem: revenue equivalence theorem (Vickrey, 1961), and this research utilize 74 daily observations (from 2008/2/8 to 2008/4/21) in each format from eBay Toyota motors’ auctions to estimate the time series econometric model developed by Ravallion (1986). The transaction price from different auction formats would converge, while the expected sellers’ revenue of different formats would be equal. | en_US |