“The Bidder’s Curse”

I was looking at the latest (April 2011) issue of The American Economic Review and came across an article titled “The Bidder’s Curse,” by Ulrike Malmendier of Cal Berkeley and Young Han Lee of Virtu Financial, a computer-driven market-making firm in New York City.  The article describes bidding patterns for online auctions on e-Bay.

The study has two parts:

1.  The first is an analysis of bidding in auctions of a board game, aptly named (for the sellers, anyway) Cashflow 101, over an eight month period in 2004.  Two different sellers with excellent feedback from customers conducted the auctions of brand-new games.  The important characteristic of the auctions for the study is that each seller continuously displayed a fixed “Buy It Now” price of $129.95 (later raised to $139.95) for the game on the same webpage as the auction.

Their result: despite the availability of Buy-It-Now, 42% of the auctions ended with the “winning” bidder paying more than the Buy It Now price.  If you factor in the higher shipping costs charged to auction winners, the percentage of auction winners overpaying rises to 73%.  In 27% of these cases, winners paid an extra $10+; in 16% of the cases, winners paid an additional $20+.

2.  The second aspect of the study was a snapshot view of 1,929 different auction results involving a wide variety of product categories during three different months in 2007.  This was to address the possibility that bidders’ (crazy) behavior might be limited only to Cashflow 101.  It wasn’t.  Overbidding happened in these auctions as well–between 44% and 52% of the time, depending on the sample taken.  The average overpayment was 10%.

What should we make of this behavior?

According to the authors, it isn’t that the winners don’t understand what e-Bay is about.  Both rookie bidders and veterans overpay.  Nor is there a relationship between the number of bids a person makes and overbidding, nor one with the total time span in which bids are made (both are ways of calculating the effort a bidder is spending on the auction).  Yes, sellers may create a favorable atmosphere for the auction by, say, setting the initial price low and the Buy-It-Now price higher than what you could buy the article for elsewhere (the latter is almost always true).  But none of this window-dressing explains why anyone bids more than the Buy-It-Now price.

The study finds a few factors that do seem to influence the bidding process:

–the closer the Buy-It-Now price is physically located to the auction price on the webpage, the less likely it is that overbidding will occur

–anyone who has already bid and gets a “You have been outbid!” notice by e-mail tends to respond by overbidding

–overbidding is sparked by a small number of auction participants.  Overbidding is slow at first but feeds on itself if these participants make up more than 10%-15% of total auction bidders.  From a seller’s perspective, then, it becomes very important to create conditions that will attract this segment to the auction–or at least to let in as many people as possible.

my thoughts

People talk about the “winner’s curse” in an auction.  In its simplest form, this is the idea that an auction winner is the individual who has the highest estimate of the value of the item being sold.  No one else thinks it’s worth that much (otherwise, they’d continue to bid), so chances are good that the winner has overpaid.

The bidder’s curse is a lot worse than that.  The Bidder’s Curse study cites earlier research whose conclusion is that in 98.8% of instances, an item being auctioned on e-Bay can easily be located elsewhere on the internet for less than the Buy-It-Now price.   In a world where bidders collected even the most basic, readily available information, everyone would know how much the item is worth and auctions would never reach the BIN price.  IN addition, of course, the BIN price should be an upper bound to bidding.  In contrast, the article documents the prevalence in the e-Bay world of an ill-informed and economically indifferent cadre of overbidders who end up determining many auction results.

As far as the stock market goes, I think there are two questions for any investor to ask:

–The more critical is whether we’re members of the class of chronic overbidders.  In this case, active investing will only lose us money; passive investment is the only way to go.

–Marketers distinguish between impulse purchases–like anything within reach in the checkout line–for which consumers have little concern about price, and research purchases–like a house or a car–where consumers prepare themselves through careful study and price comparison before acting.  It may be easy to dismiss the e-Bay results as being in the impulse category and to conclude, therefore, that they have no relevance for buying stocks.  I’m not sure that’s right.  The Cashflow 101 behavior seems to me to have a lot to say about how small-cap tech stocks work in the US.  And the overbidding phenomenon may have wider applications, as well.