400 A. Ljungqvist
time period. Interestingly, however, in 1999 the point estimate falls to only 19 cents
in trading commissions per dollar of underpricing gain. Thus at the height of the IPO
bubble, the ‘price’ of underpriced IPO allocations seems to have dropped substantially.
In fact, in aggregate dollar terms, almost the entire increase in money left on the table
in 1999 appears to have accrued to mutual funds, with banks’ revenue from trading com-
missions largely unchanged in 1999 compared to earlier years. This is hard to reconcile
with the view that banks deliberately increased underpricing during the IPO bubble: if
they did, they were curiously inept at profiting from it.
3.4. Underpricing as a signal of firm quality
The final group of asymmetric information models reverses Rock’s assumption regard-
ing the informational asymmetry between issuing firms and investors. If companies have
better information about the present value or risk of their future cash flows than do in-
vestors, underpricing may be used to signal the company’s ‘true’ high value. This is
clearly costly, but if successful, signaling may allow the issuer to return to the market
to sell equity on better terms at a later date. In the words ofIbbotson (1975),whois
credited with the original intuition for the IPO signaling literature, issuers underprice in
order to ‘leave a good taste in investors’ mouths’.Allen and Faulhaber (1989), Grinblatt
and Hwang (1989), andWelch (1989)have contributed theories with this feature.
Suppose there are two types of firms, denoted high-quality and low-quality, which
look indistinguishable to investors. Firms raise equity in two stages, via an IPO and at
a later date. High-quality firms have incentive to credibly signal their higher quality, in
order to raise capital on more advantageous terms. Low-quality firms have incentive to
mimic whatever high-quality firms do. The proposed signal in the IPO signaling models
is the issue price.
With some positive probability, a firm’s true type is revealed to investors before the
post-IPO financing stage. This exposes low-quality issuers to the risk that any cheating
on their part will be detected before they can reap the benefit from imitating the high-
quality issuers’ signal. This makes separation between the two types possible. Provided
the risk of detection and the implied reduction in IPO proceeds are sufficiently great to
deter the low-quality firms from imitating the high-quality ones, a high-quality firm can
influence investors’ after-market beliefs about its value by deliberately leaving money
on the table at the IPO. This money is ‘recouped’ when the firm returns to the market at a
later date. Low-quality firms refrain from mimicking the signal (i.e., from underpricing)
because the risk of detection means they may not be able to recoup the cost of the signal
later.
Signaling models are open to the challenge that the proposed signaling device may
be dominated by other signals. Would firms really choose the underpricing signal
if they had a wider range of signals to choose from? Such a range could include
the choice of particularly reputable underwriters (Booth and Smith, 1986), auditors
(Titman and Trueman, 1986), or venture capitalists (Megginson and Weiss, 1991;
Lee and Wahal, 2004), each of whom could perform a certification-of-quality role; the