Most firms try to sell their initial public offerings at predetermined prices, rather than just holding an auction. After the shares are sold, there appear to be immediate excess returns, which suggests some money may have been left on the table. Why do things this way?
Ravi Jagannathan, Andrei Jirnyi, and Ann Sherman have a new and comprehensive study (NBER) of this long-standing question. Here is the abstract:
At least 25 countries have used IPO auctions, but most have since abandoned them. We argue that this is because auctions, being indirect mechanisms, require a level of sophistication above that of many investors. Through suitably calibrated examples, we show that even sophisticated investors can make mistakes while bidding in auctions, especially when facing uncertainty about the number and type of bidders, and such mistakes impose costs on other participants. We provide empirical support for our arguments. IPO auctions have been plagued by unexpectedly large fluctuations in the number of participants, return chasing investors, and high-bidding free riders. Our analysis suggests that a direct mechanism that resembles a transparent version of book building would be preferable to auctions.
Here is one summary of the piece. Here is a different, 2007 paper on the question. Here is another related paper, and here (by Jagannathan and Sherman, much older draft, minus Jirnyi). Here is a one-year-old ungated version of the main paper, I am not sure how much it differs. Do any of you know of an ungated version of the current draft?
Here is an earlier Alex post on the Google IPO, which was held as an auction.