eBay Auctions Research Articles

Digital trading networks and competitive advantage in a buyer-seller network

PurposeFirms embedded in networks of relations are theorized through Gnyawali and Madhavan’s (2001) (G&M) structural embeddedness model to gain competitive advantage from topological characteristics. Empirical studies to support their theory have never been executed in full. Our study provided a full empirical test of their model in a digital trading network to achieve a higher degree of certainty that those network structural characteristics can have a major impact on the degree to which certain firms lead to competitiveness in a digital trading network environment.Design/methodology/approachTo examine how firms respond in competitive situations, we chose the hyper-active digital trading network, eBay as our empirical context. We used eBay auction data to analyze how the network characteristics of eBay resellers impact their competitive behaviors.FindingsOur study found strong support for the G&M model of competitiveness. We offer explanations for where support was not as strong as the Gynawali and Madavan theory proposes.Research limitations/implicationsOur research is limited by our chosen context and findings in support of part of G&M model. Future studies in other digital contexts are needed to enhance the modeling of network topologies and further study the impacts of network density and structural autonomy on competitive action.Practical implicationsOur study suggests that managers proceed cautiously in forming partnerships, weighing circumstances where the firm can find itself with increased information power and avoiding, to the greatest extent possible, situations where the playing field is roughly equal.Social implicationsTheory-making in this domain has begun as well as initial empirical testing. Much more needs to be accomplished, though, before embeddedness modeling can be thought of as being well established.Originality/valueThe G& M Model of competitiveness is an SNA explanation of why some competitive units succeed and others do not. Our study is the first, full blown empirical analysis of the theory.

A Comprehensive Exploration of Machine Learning Models for Predicting Online Auction Prices

The transformative impact of traditional commerce by online marketplaces is exemplified through eBay, a global platform that facilitates diverse transactions via auctions. In this research, the dynamics of eBay auctions, crucial for buyers, sellers, and researchers, are delved into. The central inquiry revolves around the key factors shaping auction outcomes, examining bid behaviors and types. The study leverages a robust dataset from eBay, meticulously curated to encompass auction identifiers, bid details, pricing information, auction types, and temporal aspects. A comprehensive approach involves data preprocessing, ensuring reliability by addressing missing values and outliers. Rigorous exploration and validation validate the dataset's integrity. Machine Learning Techniques, including MLP, SVR, Linear Regression, Extra Trees, and Gradient Boosting, form the analytical backbone. Model evaluation reveals top-performing candidates, such as MLP Regressor (0.8084), SVR (0.8210), and Linear Regression (0.8173), exhibiting superior accuracy and reliability. These models are identified for adoption in future work, emphasizing nuanced predictions in eBay auctions. This research contributes to understanding online auction dynamics, offering practical insights for eBay users and the broader e-commerce community. The models identified pave the way for enhanced predictive capabilities and continuous refinement in deciphering factors influencing auction outcomes.

A Comprehensive Exploration of Machine Learning Models for Predicting Online Auction Prices

The transformative impact of traditional commerce by online marketplaces is exemplified through eBay, a global platform that facilitates diverse transactions via auctions. In this research, the dynamics of eBay auctions, crucial for buyers, sellers, and researchers, are delved into. The central inquiry revolves around the key factors shaping auction outcomes, examining bid behaviors and types. The study leverages a robust dataset from eBay, meticulously curated to encompass auction identifiers, bid details, pricing information, auction types, and temporal aspects. A comprehensive approach involves data preprocessing, ensuring reliability by addressing missing values and outliers. Rigorous exploration and validation validate the dataset's integrity. Machine Learning Techniques, including MLP, SVR, Linear Regression, Extra Trees, and Gradient Boosting, form the analytical backbone. Model evaluation reveals top-performing candidates, such as MLP Regressor (0.8084), SVR (0.8210), and Linear Regression (0.8173), exhibiting superior accuracy and reliability. These models are identified for adoption in future work, emphasizing nuanced predictions in eBay auctions. This research contributes to understanding online auction dynamics, offering practical insights for eBay users and the broader e-commerce community. The models identified pave the way for enhanced predictive capabilities and continuous refinement in deciphering factors influencing auction outcomes.

Shill bidding and information in eBay auctions: A Laboratory study

In online auction platforms, like eBay, sellers have frequently been observed to bid on their item to artificially increase its price, and this is known as shill bidding. We represent the eBay auction in a sequential auctions environment using lab experiments and study the behavioral consequences of sellers being able to participate as shill bidders and of being informed about buyers’ past bidding histories. We find that the possibility of shill bidding in ongoing and future auctions benefits sellers and mostly affects high private-value buyers. At the same time, buyers seem to overreact to the threat of shill bidding in the future auctions by biding too high in the current auction. However, providing sellers with buyers’ bidding histories between auctions has little impact on auction outcomes and players’ bidding behavior. Moreover, there are significant differences between buyers’ and sellers’ dynamic bidding behavior during auctions, which can be used to identify shill bidding sellers from buyers.

(Un)Obtainable Luxury and Performance: The Semiotics of Michael Jordan’s 1996 V12 Mercedes S600 Lorinser

On August 23rd, 2020 a model of 1996 V12 Mercedes S600 Lorinser was sold at an eBay auction. While the asking price for this particular model, in 2021, is approximately between $4,500 and $11,000, this particular car was sold for a hefty sum of $202,200. What makes this Mercedes car unique and pricey is the fact that it used to belong to a basketball’s great, Michael Jordan. The fact that the auction in question was finalized in the midst of global, Coronavirus pandemic, and following the release of The Last Dance documentary on Jordan’s career, factors into the interpretation of the semiotics of the automobile itself. The aim of this paper is to analyze: the semiotics meanings attached to Michael Jordan as a persona and a brand/sign; the semiotics meanings attached to Mercedes brand and cars; and, ultimately, shifting semiotic meanings of Jordan’s 1996 V12 Mercedes S600 Lorinser in the context of time and the selling price, with nostalgia as a factor.

Last minute only bidding is implausible in eBay sealed bid type-of-auctions

In recent years internet auctions have attracted much attention. This paper discusses the "last minute bidding" (sniping) phenomenon, first investigated by Roth and Ockenfels (American Economic Review 92, 1093-1103) in eBay, fixed-end, auctions. Unlike standard auctions where each offer is processed by the auctioneer due to system traffic, and connection time, in eBay auctions very late bids may be left unprocessed. In the paper we consider a simple two-players, two-periods sealed-bid auction model, inspired by eBay auctions, with private values and complete information. The main difference with respect to e-Bay auctions is given by the first period. Indeed, players can submit at most one bid per period. In the first stage bids are processed with certainty by the system while in the second stage, with positive probability, bids may not be processed due to system congestion. Unlike eBay we show how in this model last minute, that is second stage only, bidding may have low plausibility, as it can be a Nash equilibrium only under very specific circumstances and is never a unique best reply. Intuitively this is because the first stage is sealed-bid, with at most one offer per bidder, and players have no guarantee that they could outbid the opponent in the second period, once after the first round they realise that their bid is not the best.

How do sellers benefit from Buy-It-Now prices in eBay auctions?

In eBay “Buy-It-Now” auctions, sellers make a take-it-or-leave-it price (BIN-price) offer prior to the auction. A buyer can either buy the good at the BIN-price or start the auction. This article studies how eBay sellers set BIN-prices and whether they benefit from using them. We apply a new experimental approach that combines the advantages of the lab and the field: The experiment is conducted in the lab while using the eBay auction platform and eBay traders. We find that the eBay auction format leads to prices substantially below those expected in second-price auctions. Moreover, the publicly available information on eBay about buyer experience and bidding behavior correlates with the observed price deviation and sellers use this information strategically when deciding on their BIN-price. More experienced sellers ask for higher BIN-prices, whereas seller risk aversion does not affect their BIN-price. Compared to an eBay auction without a BIN-price, an auction with a BIN-price is more efficient and generates significantly higher revenue.

When bargaining meets auctions: assessing the effect of best offer on eBay auctions

ABSTRACT In this paper, we apply a recently popularized methodology called ‘matched listings’ to eBay API data to study the effect of enabling eBay’s bargaining feature on auction outcomes. Our results indicate that enabling the bargaining feature increases sale probability by 3.17%. In addition, hypothesis testing indicates this estimate is consistent with eBay’s claim where enabling bargaining increases sale probability by 5%. Finally, we find that enabling bargaining has no impact on transaction price. We provide an innovative evaluation of whether the matched listings methodology works well. We find that this methodology is prone to endogeneity with longer time windows, causing a downward bias when estimating the treatment effect of enabling bargaining.

Online Auction Markets Before and During the Lockdown

We use a novel dataset on more than 35,000 eBay auctions for electronics to analyse Covid19-induced lockdown effects. The dataset covers pre- and lockdown periods and, for each auction, contains both auction settings fixed by the seller and the whole bidding record. The interplay of agents reactions has led to an overall equilibrium price decrease, with some heterogeneity across brands and products. We identify a mechanism that partially explains this decrease. Sellers in eBay choose the ending day of their auctions. The most selected day of the week in the pre-lockdown was less selected that most selected one during the lockdown. That increased the competition among auctions. Our regression analysis delivers statistically significant evidence of a negative impact of that competition on the stop-out prices precisely for products and brands which had a larger price decrease, which presumably coincide with those with a broader set of close substitutes. The effect of other variables on prices, either related to bidding behavior or to eBay specificities is discounted.

Matching algorithm of the auctions on Ebay

This paper explores eBay auction properties that match buyers and sellers and generates millions of sales every month. eBay’s auction is now a well known mechanism designed to make buyers and sellers feel comfortable doing business without meeting each other. In a theoretical point of view, the current matching algorithm has not solved the online auction problems because the main conditions of agents’ preferences do not satisfy when bidders are unobservable and a set of bidders is not identified. Therefore, we construct a new simplified model of matching with a given object for sale to form a seller-bidder pair to overcome the online auction issues. Specially, our model may extend for the matching mechanism of the job market.

A Transaction Utility Approach for Bidding in Second-Price Auctions

In both the Vickrey and eBay auctions, bidding the reservation price is the optimal strategy within the conventional utility framework. However, in practice, buyers tend to bid less than their reservation prices, and bid multiple times, thus increase their bids, in the course of an auction. In this paper, we show that both underbidding and multiple bidding behaviors can be consistent with utility maximization, if buyer's utility incorporates a transaction utility (reference price dependent) component. Transaction utility is based on the difference between the buyer's reference price and actual price paid; it captures the perceived value of the deal. More specifically, we show that the optimal bid is lower than the reservation price, but higher than the reference price. Furthermore, buyer may re-bid (above the prior optimal level) if the reference price is revised upon observing a higher current price.

Competing trade mechanisms and monotone mechanism choice

We investigate the choice between posted prices and auctions of competing sellers with private valuations. Assuming that buyers face higher hassle costs in auctions, we show the existence of monotone pure strategy equilibria where sellers offer posted prices rather than auctions if and only if they have a sufficiently high reservation value. Posted prices sell with lower probability but yield a larger revenue in case of trade. Using an empirical strategy to compare revenues of posted prices and auctions that takes selling probabilities explicitly into account, we find our theoretical predictions supported by data from eBay auctions on ticket sales for the EURO 2008 European Football Championship.

Revenue in first-price auctions with a buy-out price and risk-averse bidders

Abstract We model the “Buy-It-Now or Best Offer” eBay listing option as a first-price auction with a buy-out price and consider seller revenue when bidders are risk-averse. In this model we consider a discrete setting and a continuous setting. We show that a buy-out price in a first-price auction can increase seller revenue in the discrete setting but not in the continuous setting. Furthermore, we prove that in either setting a second-price auction with an optimally chosen buy-out price will never generate more expected revenue than a first-price auction with an optimally chosen buy-out price. These theoretical results support the thesis that a first-price mechanism is superior in expected revenue generation to a second-price mechanism in an online marketplaces where sellers are risk-neutral and buyers are risk-averse and explain the prevalence of “Buy-It-Now or Best Offer” listings on eBay over the traditional eBay auction.

A Transaction Utility Approach for Bidding in Second-Price Auctions

In both the Vickrey and eBay auctions, bidding the reservation price is the optimal strategy within the conventional utility framework. However, in practice, buyers tend to bid less than their reservation prices, and bid multiple times, thus increase their bids, in the course of an auction. In this paper, we show that both underbidding and multiple bidding behaviors can be consistent with utility maximization, if buyer’s utility incorporates a transaction utility (reference price dependent) component. Transaction utility is based on the difference between the buyer’s reference price and actual price paid; it captures the perceived value of the deal. More specifically, we show that the optimal bid is lower than the reservation price, but higher than the reference price. Furthermore, buyer may re-bid (above the prior optimal level) if the reference price is revised upon observing a higher current price.

Detecting Shill Bidding Using eBay’s 30-Day Bid Summary

Shill bidding is a fraudulent act whereby a seller places bids on his/her own auction to drive up the final price for the winning bidder. eBay currently masks bidder usernames in auctions and has adopted a 30-Day Bid Summary that restricts the availability of data on the previous auctions a bidder has participated in. This has made shill bidding detection more difficult, as tracking users across multiple auctions is essential in identifying shills. This paper presents two pieces of software to aid in the detection of shill bidding in light of masked bidder usernames and limited bid history data. First, we propose the Auction Data Collector to automatically acquire vital eBay auction data required for shill bidding detection. Second, we propose the BidderLinker Algorithm, which links bidders across multiple auctions they have bid on with the same seller by utilising the 30-Day Bid Summary. The 30-Day Bid Summary provides information about bidders that can be used to uniquely identify them, even if their usernames are masked. This allows for shill detection algorithms to be applied to the data gathered from multiple auctions with the same seller, where previously this was not possible.

Trust and experience in online auctions

ABSTRACTThis paper aims to shed light on the complexities and difficulties in predicting the effects of trust and the experience of online auction participants on bid levels in online auctions. To provide some insights into learning by bidders, a field study was conducted first, to examine auction and bidder characteristics from eBay auctions of rare coins. We proposed that such learning is partly because of institutional-based trust. Data were then gathered from 453 participants in an online experiment and survey, and a structural equation model was used to analyze the results. This paper reveals that experience has a nonmonotonic effect on the levels of online auction bids. Contrary to previous research on traditional auctions, as online auction bidders gain more experience, their level of institutional-based trust increases and leads to higher bid levels. Data also show that both a bidder’s selling and bidding experiences increase bid levels, with the selling experience having a somewhat stronger effect. This paper offers an in-depth study that examines the effects of experience and learning and bid levels in online auctions. We postulate this learning is because of institutional-based trust. Although personal trust in sellers has received a significant amount of research attention, this paper addresses an important gap in the literature by focusing on institutional-based trust.

EBay Memorabilia Auctions vis-à-vis Predictions for the 2016 U.S. Presidential Election

The present paper investigates whether there is a relationship between eBay auction outcomes for memorabilia and various predictions from the FiveThirtyEight site for the 2016 U.S. presidential election. We find that changes in the value of various FiveThirtyEight prediction variables were almost always correlated with the number of bids. For the national prediction, an increase in the probability of a Clinton victory was associated with a reduction in the number of bids and the willingness to pay for an item. Similar results were found for the subsample of Trump items, suggesting that the auctions for his items dominated the results for the full sample. JEL Classifications: A20, D44, L81

Made by Mistake: When Mistakes Increase Product Preference

Abstract Mistakes are often undesirable and frequently result in negative inferences about the person or company that made the mistake. Consequently, research suggests that information about mistakes is rarely shared with consumers. However, we find that consumers actually prefer products that were made by mistake to otherwise identical products that were made intentionally. This preference arises because consumers perceive that a product made by mistake is more improbable relative to a product made intentionally, and thus, view the product as more unique. We find converging evidence for this preference in a field study, six experiments, and eBay auction sales. Importantly, this preference holds regardless of whether the mistake enhances or detracts from the product. However, in domains where consumers do not value uniqueness (e.g., utilitarian goods), the preference is eliminated.

What Did Consumers Pay for 2016's Hot Toy? A Descriptive Analysis of Hatchimal Prices in Ebay Auctions during the 2016 Holiday Shopping Season

While it is undisputed that 2016’s “hot toy” was the Hatchimal, it is unclear how much shoppers paid for the toy in the secondary market or how resellers fared. Using data on 34,657 completed successful Hatchimal eBay auctions by 16,726 sellers between October 13 and December 24, 2016, I find that shoppers paid an average of $143.67 for the $60 toy. Hatchimal prices varied considerably over this time period. They were lowest ($65.24) on October 16 and highest ($185.22) on November 24, Thanksgiving Day. 62.1% of eBay shoppers paid between $125 and $175, and just over 3% paid more than $200 for a Hatchimal. The Toys R Us exclusive Bearakat was the most sought after Hatchimal version, earning a price premium of 23%. In contrast, the Walmart and Target exclusive Hatchimal versions earned rather modest price premiums of 4-5%. Most sellers were amateurs, listing one or two Hatchimal auctions and earning an estimated $150 of total profit from their endeavor. Shoppers paid higher prices in Hatchimal auctions with high shipping fees ($10 ) and those listed by experienced eBay sellers. Based on the findings from this descriptive analysis, I conclude this report with specific recommendation to shoppers about how to shop smart for a hot toy during the holiday shopping season.

Reserve Prices in Online Auctions:Theory and Evidence

The paper proposes a simple model of auctions with an impatient seller who chooses the reserve price and the buy-it-now (BIN) price to maximize revenue. The three main sales channels in the online auction (the pure auction, the BIN auction, and the fixed-price sale) are each shown to be an optimal solution for the sellers with certain degrees of time-preference. The theory also predicts that the optimal posted price for the fixed-price listing is greater than the optimal reserve price for the BIN auction, which in turn is greater than that for the pure auction. This prediction is confirmed by data from eBay's auction of iPods. Using the seller's inventory as a proxy for his degree of time-preference, the data also confirm the theoretical prediction that there is an inverse relationship between the optimal reserve price and the seller's time-discount factor.