Auction Theory and The Online Auction

James Counts

Humboldt State University

Econ 310: Intermediate Microtheory & Strategy

Professor Steven Hackett

Wednesday, November 15, 2000

 

Introduction

Information networks, such as the Internet, are slowly transforming the properties of markets across the world. One of the more dominant changes is the popular renewal of variable-price markets. Haggling, which seemed on deaths door a few years ago, is back„in the form of online auctions. I began to wonder if the new markets were substantially different from the old. Assuming the new markets were different, I wanted to know if the changes benefited buyers, sellers or both.

To take a position on these matters required knowledge of some basic competition theory, auction theory, and an examination of an online market. For this paper, I chose to examine eBay, the largest online auction host. I found evidence that it is possible to purchase items at below-retail prices on eBay, and that both buyers and sellers can benefit from using eBay.

Auction Theory

William Vickrey laid the foundation of modern auction theory in his 1961 paper, ñCounterspeculation, auctions, and competitive sealed tenders.î The model auction described in the paper serves as a useful measuring stick when evaluating auction structures. When allocating goods using VickreyÍs auction model the results are Praeto-optimal. Any allocation other than the normal result would decrease the overall welfare of those involved.

Auctions with the following characteristics achieve the normal result:

‡         [T]he resource to be allocated comes in a small number of discrete indivisible units rather than consisting of a fungible commodityƒ

‡         [T]o be sold to one of a number of potential purchasers.

‡         Price can vary continuously andƒ there is no minimum increment between bids.

‡         Rational bidders.[1]

‡         [Bidding stops] at a level approximately equal to the second highest value among the values that the purchasers place on the item, since at this point there will be only one interested bidder leftƒ[2] (14)

Irrationality and information asymmetries are likely to spoil the results of the model auction. To insure optimal results Vickrey proposed two mechanisms: sealed bids and second-bid pricing (20). Sealed bids force bidders to employ optimal bidding strategies. To maximize the probability of winning, while eliminating the possibility of overpaying, each bidder must place a bid equal to their private valuation of the item at auction. Second-bid pricing is a system where the highest bidder wins the auction, but pays the bid submitted by the second highest bidder. This systemÍs advantage is that the highest bidder does not lose the entire consumer surplus to the seller, and so is motivated to participate in the auction.

eBay & the Theory

EBay is the largest online auction site and claims to be the worldÍs largest market:

ñToday, we continue to take great pride in our business benchmarks.

‡         The eBay community includes more than 10 million registered users.

‡         Each day, eBay hosts nearly 4 million auctions, with over 450,000 new items joining the ïfor saleÍ list every 24 hours.

‡         In January 2000, eBayÍsƒ number of unique visitors on an average daily basis set a new record of 1.782 million.î

(eBay Company Overview: Benchmarks, 2000)

* * *

eBay created a new market: efficient one-to-one trading in an auction format on the Webƒ Individuals„not big business„use eBay to buy and sell itemsƒ

(eBay Company Overview, 2000)

While eBay is indubitably large, its assertions about its own efficiency should be tested against auction theory and competition theory. Most economists are intimately familiar with the criteria describing perfect competition, and accept their use as a ñmeasuring stick.î For reference the criteria are reprinted here:

‡         There are many buyers and sellers, each of whom is small in size relative to the overall market

‡         There are no entry/exit costs or other difficulties associated with entry/exit

‡         Products made by different sellers are effectively identical

‡         Economies of scale in production are exhausted at low output levels, which explains in part why there are many relatively small sellers

‡         Costs are identical across firms

‡         There are no transaction, travel or information costs

(Hackett, 2000)

With ten million registered users and 1.8 million unique daily visitors, eBay meets the first criteria: many small buyers and sellers. Buyers face no entry/exit costs, while sellers pay a small listing fee no greater than $2.[3] Product differentiation varies across auctions. All the auctions surveyed for this paper were for new ñnever openedî items, and were identical to each other and their retail counterparts. But many items auctioned on eBay are for used goods (differentiated by wear) or antique/collectible goods (rare or unique and highly differentiated.) Since no production occurs in these auctions, the only way to be large on eBay is to have multiple listings. The eBay user agreement limits each user to ten simultaneous listings for identical goods„leveling the playing field for ñeconomies of scale.î The eBay fee schedule sets identical selling costs for all sellers, buyers face no browsing costs. Information costs are insignificant, because it is easy to compare prices by visiting various Internet sites. Shipping and handling charges represent the travel costs and vary widely depending on the good. In some cases, however, travel costs may be less costly than local sales taxes. So, eBay measures up to competition theory well, after all.

A little more work is required to compare eBay practices with VickreyÍs auction model. To compare eBay with the model, I sampled 34 auctions to determine the average number of bidders and I surveyed 25 auctions to compare the winning bids to retail prices. In most eBay auctions only one item is up for bid, so called ñDutchî auctions involve multiple discrete items, but usually in small amounts. The auction sample shows that the average unit-to-bidder ratio is 1:33 (Appendix A). Thirty-three bidders interested in a single item are enough to meet VickreyÍs second criterion of ñmany purchasers.î Although minimum bid increments are the sellerÍs prerogative on eBay, I only observed one auction with a minimum bid increment. By default then, the bid increment is $0.01 and is unlikely to prevent a bidder from placing bids at exactly their private valuation.

It is easy to determine eBay meets VickreyÍs first three criteria. It is more difficult to ascertain that eBay meets the final two criteria. Since we cannot know the private valuations held by eBay users, we cannot say if the bidding on eBay stops at approximately the second bidderÍs valuation. However, since this condition requires bidder rationality, testing eBay bidders for rationality should give us information regarding private valuations. If the bidders seem rational, we can assume that the auctions meet the second price criterion.

To test bidder rationality on eBay I compared 25 high bid prices to retail prices for the same items (Appendix B). I obtained the retail prices from online, fixed-price, non-discount retailers and this allowed me to ignore shipping and handling costs (which buyers pay either way.) If bidders pay less than or equal to the online retail prices then they are probably rational, if they pay more then they are irrational or poorly informed.[4]

I found that high bids on eBay were generally lower than prices found at fixed price online retailers. A bidder who paid 30% of retail for an Intel Pentium III CPU enjoyed the greatest individual saving. A bidder who paid 94% of retail for a Palm Pilot m100 enjoyed the smallest savings. Only one bidder in the sample paid more than retail. This unfortunate person paid 247% of retail for a pair of Nike athletic shoes.[5] On average bidders on eBay paid 79% of retail, lending some support to the conclusion that they are„for the most part„rational.

Conclusion

The examination of eBayÍs structure as defined by its user agreement and fee schedule, as well as a survey of eBay usersÍ behaviors provide evidence that eBay is a close approximation to auction and competition theory. The price sample shows that it is possible to find below-retail prices at auction on eBay. While eBay auctions obviously benefit consumers in this way, they are also beneficial to retailers. Online auctions allow small retailers to penetrate geographically distant markets and achieve higher prices than their home markets would bear. Neither of these benefits is free. The participants in this market must give up the time and effort it takes to participate. Does eBayÍs efficiency and generally lower prices mean we are headed for the ballyhooed ñauction economy?î

Personally, I hope this is not the case. Though difficult to measure, time and effort are real costs of doing business in the auction economy. While it may be worthwhile to participate in an auction and save $250 on a computer system, it would not be worthwhile to participate in an auction to save $0.025 on a pencil. Fixed price retailers will offer convenience and instant gratification that auctions do not. In general eBay and other e-commerce concerns will probably promote price variability, and erode some advantages held by spatial monopolists, but I doubt they will eliminate either institution.

Works Cited

EBay (2000) About eBay: Company Overview: Benchmarks. Available 14 November 2000 HTTP: pages.eBay.com/community/aboutebay/overview/benchmarks.html

EBay (2000) About eBay: Company Overview. Available 14 November 2000 HTTP: pages.eBay.com/community/aboutebay/overview/index.html

Hackett, S. (2000) ñIntroduction to Market Structure Analysis: Competition and Monopolyî Lecture notes. Internet. Available: 14 November 2000 HTTP: www.humboldt.edu/~microeco/note/mktstr.htm

Vickrey, W. (1961) Counterspeculation, Auctions, and Competitive Sealed Tenders. Journal of Finance, 16, 8-37.

Bibliography

Dvorak, J. (1999) Inside Track: The Next Big Trend Dept. PC Magazine, July, 89

EBay (2000) Help: Rules & Safety: User Agreement. Available 14 November 2000 HTTP: pages.eBay.com/community/png-user.html

EBay (2000) Help: Seller Guide: Fees. Available 14 November 2000 HTTP: pages.eBay.com/help/sellerguide/selling-fees.html

Hamel, G. & Sampler, J. (1998) The e-Corporation. Fortune, December 7, 80-92.

Hof, R. (2000) Will Auction Frenzy Cool? Business Week, September 18, EB140

Klemperer, P. (1999) Auction Theory: A Guide to the Literature. Journal of Economic Surveys, Vol. 13(3) July, 227-286.

Klemperer, P. (2000) What Really Matters in Auction Design. Available: 15 November 2000 HTTP: www.nuff.ox.ac.uk/economics/people/klemperer.htm

Klemperer, P. (2000) Why Every Economist Should Learn Some Auction Theory. Available: 15 November 2000 HTTP: www.nuff.ox.ac.uk/economics/people/klemperer.htm

 

Klemperer, P & Bulow J. (1999) Prices and the WinnerÍs Curse. Available: 15 November 2000 HTTP: www.nuff.ox.ac.uk/economics/people/klemperer.htm

Lucking-Reiley, D. (1999) Using Field Experiments to Test Equivalence Between Auction Formats: Magic on the Internet. The American Economic Review, December, Vol. 85 No 5, 1063-1080

Razzi, E. (2000) Auction Fever. KiplingerÍs Personal Finance, May, 115-118.

Smith, M. Bailey, J & Brynjolfsson, E. (2000) Understanding Digital Markets: Review and Assessment. Available 15 November 2000 HTTP: ecommerce.mit.edu/papers/ude

Tebbe, M. (1998) Confessions of an online auction addict: Not even the CFO can stop me. InfoWorld, May 18, 140

Thaler R. (1992) The WinnerÍs Curse: Paradoxes and Anomalies of Economic Life. New York: Free Press

 

Appendix A: Bidder Survey

Located in this Appendix:

Sample of 34 auctions and results

EBay listing of sampled auctions

Appendix B: Price Survey

Located in this Appendix:

Sample of 25 auctions compared to retail prices

Example auction listing

Example bid listing

Example retail listing