Linkedin’s IPO followed the familiar pattern. Priced to initial investors at $45/share, the stock soared to a peak of $120/share on the first day of trading and closed at around $90. “Money was left on the table.” Felix Salmon has a good summary of different opinions about this and he briefly discusses auctions as an alternative to the traditional door-to-door IPO. You will recall that Google used an auction when it went public.
Like Mr. Salmon you might think that the winner’s curse would prevent an auction from discovering the right price and ensuring that money is not left on the table. The price-discovery properties of an auction rely on partially informed investors bidding according to their private information and this information thereby being reflected in the price. But a (potentially) winning bidder foresees that his win would mean that most other bidders bid less than he and that therefore “the market” is more pessmistic about the value of the firm than he is. Anticipating this, he underbids. Since all informed investors do this, the auction price will understate the value of the firm. Money will be left on the table again.
But an IPO is a multi-unit auction. Many shares are up for sale. A new strategic feature arises in these auctions, the loser’s curse. If you are outbid, then you know that many informed investors were more optimistic than you and this will make you regret having bid too low. This strategic force pushes you to raise your bid.
And in fact, as shown by Pesendorfer and Swinkels, for large auctions, under some quite general conditions, a uniform-price sealed bid auction produces a sales price which comes very close to the market’s best estimate of the value of the firm. No money is left on the table.
Cheap Talk 
3 comments
Comments feed for this article
June 8, 2011 at 3:31 am
Anonymous
Costly information acquisition is important here. Talk to your colleague Ravi Jagannathan.
June 8, 2011 at 5:53 pm
Anonymous
As the earlier commenter said, you should look at the paper that Felix Salmon refers to, by Ravi Jagannathan and Ann Sherman (there are more recent versions available), IPO auctions have been tried in at least 25 countries and have been rejected pretty much everywhere, because the method led to a lot of noise and uncertainty and problems.
The Pesendorfer and Swinkels auction theory doesn’t include the features that have led to problems with IPOs in practice: a cost to information (i.e. people don’t magically wake up knowing how much IPO shares are worth), and a huge number of potential bidders, each deciding independently whether or not to enter. Here’s an auction model that includes both of those features:
http://ssrn.com/abstract=276124
The paper was published in Journal of Financial Economics
Volume 78, Issue 3, December 2005, Pages 615-649
In an environment that is appropriate for IPOs, theory predicts that the method will lead to large variations in outcomes, even if everyone becomes informed and is bidding optimally. The reason is simple: even if everyone correctly shaves their bids for the winner’s curse, they can only do this based on their ex ante expectation of the number of bidders. Ex post, oops! If more people than expected happened to bid in that auction, then the bidders didn’t shave their bids enough and the auction could lead to overpricing. If fewer people than expected happened to bid, bidders shaved too much and the auction could lead to underpricing.
Tossing the auction open to literally millions of inexperienced potential bidders is not likely to result in the exact same, perfect number showing up each and every time. The paper Salmon referred to gave plenty of real world examples of big fluctuations in the number of bidders (for example, look at the IPOs of Japan Railway East vs. Japan Railway West; Japan forced all issuers to use auctions from 1989 to 1997, and had all kinds of problems that finally led them to offer issuers a choice of methods; once issuers had a choice, the auction method promptly vanished).
If we add the possibility that even a few potential bidders might not be well informed and highly sophisticated in their bidding strategy, there’s even more noise in the process.
Thus, theory predicts that IPO auctions will be risky for both buyers and sellers, and the experience of many countries indicates that IPO sellers, once they become familiar with the auction method, refuse to use it unless forced.
October 2, 2013 at 2:11 am
iwc レディース
bvlgari ネックレス iwc レディース http://www.cfta4.com/