efficiency. The fact that we find six such cases indicates that the highly pub-
licized Palm example was not unique.^5
Second, all the cases we study show a similar time pattern of returns
whereby the stub becomes less negative over time and eventually becomes
positive. This suggests that market forces act to mitigate the mispricing, but
slowly. We return to this slow adjustment, which reflects the difficulty of
shorting due to the sluggish functioning of the market for lending stocks, in
section 4.
3.Risk and Return on StubsIn this section, we investigate the returns to an investment strategy of buying
the parent and shorting the subsidiary. We find that this strategy produces
142 LAMONT AND THALER
(^5) It is hard to say whether the ratio of one-third overestimates or underestimates the preva-
lence of mispricing. On the one hand, perhaps firms tend to do carve-outs when they think
that their subsidiaries are overpriced, in which case the eighteen firms are not a representative
sample (firms should issue equity when that equity is overpriced, as argued by Stein [1996]).
Further, it could be that 1998–2000 was a time in which mispricing was prevalent, but in most
years mispricing is rare. Ritter and Welch (2002) show that this period was one with extraor-
dinary IPO first-day returns, and Ofek and Richardson (2001) show that Internet-related IPOs
had especially high first-day returns in this period. On the other hand, mispricing could occur
more than one-third of the time. We show that only six of the eighteen have negative stubs.
Perhaps the other twelve have stubs that are too low or too high. So in that sense, perhaps
one-third is a lower bound for relative mispricing.
! "
#$ ##
%
#&! "#$ ##
'
'
'
'
!"
#$$
(#
$$#(#
Figure 4.4. Methode/Stratos stub, June 27, 2000–May 7, 2001.