Information Production by Institutions and Information Extraction by Underwriters in Hybrid IPO Auctions
WFA 2018 Coronado Meetings Paper
68 Pages Posted: 27 Jul 2017 Last revised: 2 Apr 2018
Date Written: July 26, 2017
We analyze the informational properties of hybrid IPO auctions using a large and unique database of 115,707 bids from institutional investors in 715 auctioned IPOs between 2009 and 2012 in China, where an IPO auction is first conducted among institutions, after which a fixed price offering is conducted for retail investors. We find convincing evidence of information production of significant precision by institutions about the intrinsic values of firms going public and find that this information is effectively extracted by underwriters and used in setting the IPO offer price. In particular, we find that the offer price is more sensitive to bidding by institutions able to produce more precise information: institutions geographically close to the IPO firm; large institutions; and mutual funds and security companies. We also find that institutional bidding information from the IPO auction has predictive power for IPO initial returns and for long-run post-IPO stock returns, with the bidding information from institutions able to produce more precise information having stronger predictive power for these variables. Finally, we find that institutions are compensated for their information production through greater IPO share allocations in better performing IPOs, with institutions able to produce more precise information receiving larger share allocations in better performing IPOs, yielding them greater profits. Our results suggest that “dirty” uniform price hybrid IPO auctions are an effective alternative to the traditional bookbuilding mechanism, worthy of adoption in countries where regulators may not wish to give underwriters complete discretion in IPO share allocation, a necessary requirement for implementing the bookbuilding mechanism.
Keywords: Initial Public Offerings, Hybrid IPO Auctions, Information Production
JEL Classification: G24, G23
Suggested Citation: Suggested Citation