Abstract

We analyze the informational properties of hybrid IPO auctions using a large and unique database of institutional bids from Chinese IPO auctions. We find strong evidence of information production by institutions about the intrinsic values of IPO firms and of underwriters extracting and using this information in IPO pricing. The IPO offer price is more sensitive to bids from institutions able to produce more precise information. In particular, the offer price is more sensitive to bids from domestic institutions, compared to bids from foreign institutions who likely have less knowledge or experience about Chinese firms and the Chinese financial market due to geographic, legal, and cultural distances. Institutional bidding information also has predictive power for IPO initial returns and long-run post-IPO stock returns. Finally, institutions are compensated for their information production: institutions able to produce more precise information get significantly larger allocations in better performing IPOs; this difference in IPO allocations between better and worse performing IPOs is greater for institutions able to produce more precise information. Overall, “dirty” uniform price hybrid IPO auctions appear to be a reasonably effective alternative to the traditional book-building IPO mechanism, so that our analysis has important implications for IPO mechanism reform in developing countries.