412 A. Ljungqvist
cated pro rata. Stoughton and Zechner, on the other hand, model a bookbuilding regime
with discretionary allocations. In a pro-rata regime Stoughton and Zechner would have
difficulty allocating enough stock to the large shareholder to ensure effective monitor-
ing. In a bookbuilding regime, Brennan and Franks would not need to underprice as
much to discriminate against large investors: absent pro rata allocation rules, the issuer
(and underwriter) could simply select which investors to exclude from allocations. This
illustrates the importance of the institutional assumptions made in IPO modeling.
Second, Stoughton and Zechner assume that managers internalize the agency costs
they impose on outside investors, via the lower price that investors are willing to pay
for the stock. This internalization is absent from the Brennan–Franks model.
The ownership and control dimension is a promising, albeit nascent, field in the study
of IPO underpricing. Much more empirical evidence is needed before we can assess the
validity of the theoretical contributions and before we can say whether control consid-
erations are of first or second-order importance when offer prices are set.
- Behavioral explanations
In the late 1990s initial returns increased substantially. As pointed out in Section2,U.S.
issuers left an aggregate of $62 billion on the table in 1999 and 2000 alone. Many re-
searchers are doubtful whether informational frictions, the risk of lawsuits, or control
considerations could possibly be severe enough to warrant underpricing on this scale.
As a consequence, some argue we should turn to behavioral explanations for IPO un-
derpricing. Behavioral theories assume either the presence of ‘irrational’ investors who
bid up the price of IPO shares beyond true value, or that issuers are subject to behavioral
biases and therefore fail to put pressure on the underwriting banks to have underpricing
reduced. This literature is still in its infancy.^18
The IPO market is a good setting in which to study the effect of ‘irrational’ investors
on stock prices. IPO firms by definition have no prior share price history and tend to
be young, immature, and relatively informationally opaque. Not surprisingly, therefore,
they are hard to value, and it seems reasonable to assume that investors will have a wide
range of priors about their market values. In Section6.2, we will review one recent
theory of IPO underpricing that builds on this assumption. In Section6.3, we will turn
to a model of behaviorally challenged managers. We begin, however, with a discussion
of a model of rational ‘informational cascades’.
6.1. Cascades
Welch (1992)shows that ‘informational cascades’ can develop in some forms of IPOs
if investors make their investment decisions sequentially: later investors can condition
(^18) For a survey of behavioral corporate finance more generally, seeChapter 4by Baker, Ruback, and Wurgler
in this volume.