Ch. 7: IPO Underpricing 385
allocation is below the simple average underpricing return shown in Section2.Inthe
extreme case, the uninformed are rationed completely in underpriced IPOs and receive
100 percent allocations in overpriced IPOs, resulting in average returns that are negative.
When conditional expected returns are negative, uninformed investors will be unwill-
ing to bid for IPO allocations, so the IPO market will be populated only with (equally)
informed investors. Rock assumes that the primary market is dependent on the continued
participation of uninformed investors, in the sense that informed demand is insufficient
to take up all shares on offer even in attractive offerings.^4 This requires that conditional
expected returns are non-negative so that the uninformed at least break even.^5 In other
words, all IPOs must be underpriced in expectation. This does not remove the alloca-
tion bias against the uninformed—they will still be crowded out by informed investors
in the most underpriced offerings—but they will no longer (expect to) make losses on
average, even adjusted for rationing. Note that it is not rationingper sethat necessitates
underpricing; it is instead the bias in rationing, with uninformed investors expecting
more rationing in good than in bad offerings.
Rock’s model requires one more assumption. Collectively, firms seeking to go public
benefit from underpricing, because it is the key to ensuring the continued participation in
the IPO market of the uninformed, whose capital is needed by assumption. Individually,
on the other hand, underpricing is clearly costly to a firm going public. This creates an
incentive for an individual firm to free-ride by underpricing too little.Beatty and Ritter
(1986)argue that as repeat players, investment banks have an incentive to ensure that
new issues are underpriced by enough lest they lose underwriting commissions in the
future. Investment banks thus coerce issuers into underpricing. Of course, they cannot
underprice too much for fear of losing underwriting market share.
3.1.1. Testable implications and evidence
Adjusted for rationing, uninformed investors earn zero initial returns. Informed in-
vestors’ conditional returns just cover their costs of becoming informed.
(^4) This ad hoc assumption is actually unnecessary, because a situation where everyone is informed is not
in fact an equilibrium. Imagine that all remaining investors are informed. Only attractively priced IPOs will
succeed and all others will fail for lack of buyers. But then, assuming that becoming informed is costly, this
creates an incentive to stay uninformed and to free-ride on the information of the other investors instead. The
investor would simply bid for IPO shares indiscriminately, receiving shares in the attractive IPOs but not in
the unattractive ones (which will still fail)—clearly a profitable strategy. Since every investor faces the same
incentive, no one would choose to become informed, so unattractive offerings would no longer fail. But if
no one is informed, there is an incentive to become informed, in order to avoid the unattractive IPOs. So a
situation in which no one is informed is not an equilibrium either, unless becoming informed is prohibitively
expensive.
(^5) How realistic is the assumption that issuers must pay for the uninformed investors’ participation in an
offering? If, as Rock asserts, the resources of the informed are limited, the uninformed could simply invest
through the informed investors, in exchange for a fee, to avoid the mistake of buying into overpriced issues.
(Renaissance Capital Corporation, for instance, manages a mutual fund called ‘IPO Plus Aftermarket Fund’.)
This is one of the reasons why investment funds exist in the first place: there are economies of scale in
becoming informed.