Ch. 6: Security Offerings 355
Value-weighted issuer portfolios. The results reported above are based on equal-
weighted issuer portfolios. With value-weights, relatively successful firms gradually in-
crease their portfolio weights. If the relatively low return of issuers is driven by “losers”,
then value-weighting increases the average portfolio return and possibly the abnormal
performance parameter alpha. The literature is fairly unanimous on this issue: alphas
with value-weighted issuer portfolios appears less negative than for equal-weighted
portfolios, and they sometimes provide evidence of issueroverperformance relative to
matched firms.
5.4.3. Issue-purged factors
Loughran and Ritter (2000) argue that it is counterproductive to generate factor-
mimicking portfolios without excluding security issuers from the stock universe. In-
clusion of security issuers in the factor portfolios results in the factor regressions having
the same firm on both sides of the regression (albeit with a small weight in the factor
portfolio). They argue that this substantially reduces power to detect abnormal return
via the estimated alpha.
Note that, under the null hypothesis of zero abnormal performance, purging the
factor-mimicking portfolios for ex post issuing firms biases the tests in favor of find-
ing a significant alpha. This, of course, means that failing to reject the null hypothesis
even with purged factor portfoliosa fortiorisupports the market efficiency hypothesis
over the market over/underreaction proposition.
Eckbo, Masulis, and Norli (2000)report that, on average, 11.1% of the firms in the
factor-mimicking portfolios also make SEOs during the subsequent five years. They
purge their factors by eliminating a firm from the factor-mimicking portfolios if the
firm issued equity (primary offerings) over the previous five years.Lyandres, Sun, and
Zhang (2005)also report results based on purged factors. The main conclusion of both
studies is zero abnormal returns when using issuer-purged factor regressions.
- Conclusions and issues for future research
The economics of security offerings has generated considerable empirical research in-
terest over the past two decades. This survey alone identifies more than 280 studies
largely restricted to public seasoned security offerings for cash—and we have surely
missed some. In addition, there are a large number of related studies discussed in other
surveys in this Handbook, including those on IPO underpricing (Ljungqvist, 2007), se-
curity swaps associated with corporate takeovers and restructurings (Betton, Eckbo, and
Thorburn, 2007; Eckbo and Thorburn, 2007; Hotchkiss et al., 2007), stock compensa-
tion to employees (Aggarwal, 2007), private equity (Gompers, 2007), and credit markets
(Drucker and Puri, 1989). In all of these settings, the issuer faces both direct and indi-
rect flotation costs that depend on (1) constraints imposed by security regulations, (2)