380 A. Ljungqvist
indications of interest, which are recorded in a ‘book’, and the state of the market, the
investment bank proposes an offer price to the company. Once priced, investors are
asked to confirm their indications of interest, shares are allocated, and a few hours later,
trading begins. This process is known as bookbuilding.
The precise details of the institutional framework potentially have a bearing on the
efficiency of the capital-raising process. For instance, regulatory constraints imposed
on the bank conducting the deal concerning the pricing or allocation of IPO shares
can influence the extent of underpricing, as can the way pricing-relevant information is
gathered, aggregated, and paid for. This recognition has recently sparked another trend:
interest in the IPO experience of countries other than the U.S. Despite the fact that IPO
practices appear to become more homogeneous around the world (seeLjungqvist, Jenk-
inson, and Wilhelm, 2003), institutional frameworks differ in ways that allow sharper
tests of theoretical predictions. The United Kingdom, for example, is interesting for the
fact that integrated (one-stop-shop) securities houses familiar from Wall Street com-
pete with financial intermediaries that specialize in either corporate finance advice
or stockbroking, but do not perform both functions. What services the intermediary
offers very likely affects the internal conflicts of interest it is subject to. Or take Tai-
wan. The Taiwan Stock Exchange does not permit bookbuilding and instead operates
a discriminatory-price auction system that prices IPOs based on investors’ bids, and
investors pay what they bid. This would seem a suitable way to price IPOs from a
revenue-maximization point of view, except that the market regulator in Taiwan also
imposes various constraints on the auction process which typically lead to widespread
underpricing.
The empirical IPO literature is now fairly mature—the main stylized facts have been
established, and most theories have been subjected to rigorous empirical testing. We
know that IPOs are underpriced and that the extent of underpricing, and the number of
companies going public, fluctuates over time. Broadly speaking, there is a large body
of evidence supporting the view that information frictions (including agency conflicts
between the issuing company and its investment bank) have a first-order effect on un-
derpricing. Still, there is continued interest in at least four areas: behavioral approaches
to explain why the extent of underpricing varies over time, peaking during the recent
‘dot-com bubble’; tests exploiting cross-country differences in institutional frameworks;
work shedding light on the allegedly conflicted behavior of investment banks during the
stock market boom of the late 1990s; and the potential for using auction mechanisms to
price and allocate IPOs.^3
(^3) There is surprisingly little literature on IPO auctions, especially regarding the potential costs and ben-
efits of moving from bookbuilding to auctions for pricing IPOs.Jagannathan and Sherman (2006)surveys
the international experience of using IPO auctions in a large number of countries, concluding that auctions
have fallen out of favor in the last ten or 15 years.Derrien and Womack (2002)show that in France, where
issuers can choose between bookbuilding and auctions, auctions are associated with lower and less variable
underpricing than are bookbuilding IPOs.