Handbook of Corporate Finance Empirical Corporate Finance Volume 1

Ch. 6: Security Offerings 351

Ta b l e 2 1
Average monthly abnormal equal-weighted portfolio return (α) for three-to-five year holding periods follow-
ing securities offerings by U.S. firms

Study Issuer
type

Sample

size

Sample

period

Holding

period

α

A. IPOs

Brav and Gompers (1997) All 3 , 407 1972–1992 5 yrs − 0 .49%∗a
Brav and Gompers (1997) All 934 1972–1992 5 yrs 0 .09%b
Brav, Geczy, and Gompers (2000) All 3 , 501 1975–1992 5 yrs − 0 .19%
Ritter and Welch (2002) All 6 , 249 1973–2001 3 yrs − 0 .21%
Eckbo and Norli (2005) All 5 , 365 1972–1998 5 yrs 0 .40%c
Eckbo and Norli (2005) All 5 , 365 1972–1998 5 yrs 0 .18%d

B. SEOs

Jegadeesh (2000) All 2 , 992 1970–1993 5 yrs − 0 .31%∗
Brav, Geczy, and Gompers (2000) All 3 , 775 1975–1992 5 yrs − 0 .19%
Eckbo, Masulis, and Norli (2000) Ind 3 , 315 1964–1995 5 yrs − 0 .05%d
Eckbo, Masulis, and Norli (2000) Ind 3 , 315 1964–1995 5 yrs − 0 .14%e
Eckbo, Masulis, and Norli (2000) Util 880 1964–1995 5 yrs − 0 .13%d
Bayless and Jay (2003) Ind 1 , 239 1971–1995 5 yrs − 0 .54%∗
Krishnamurthy et al. (2005) All 1 , 477 1983–1992 3 yrs − 0 .36%∗
Eckbo and Norli (2005) Ind 1 , 704 1964–1995 5 yrs − 0 .03%c
Lyandres, Sun, and Zhang (2005) All 6 , 122 1970–2003 3 yrs 0 .02%f
D’Mello, Schlingemann, and Subramaniam (2005) All 1 , 621 1982–1995 3 yrs − 0 .31%∗

C. Private placements of equity

Hertzel et al. (2002) All 619 1980–1996 3 yrs − 1 .18%∗
Krishnamurthy et al. (2005) All 276 1983–1992 3 yrs − 0 .77%∗

D. Straight debt offerings

Spiess and Affleck-Graves (1999) All 392 1975–1989 5 yrs − 0 .29%∗
Eckbo, Masulis, and Norli (2000) Ind 981 1964–1995 5 yrs − 0 .10%
Eckbo, Masulis, and Norli (2000) Util 348 1964–1995 5 yrs − 0 .22%
Butler and Wan (2005) Ind 799 1975–1999 5 yrs − 0 .18%g

(Continued on next page)

fiedportfolios as proxies for pervasive risks. Either approach suffers from potential “bad
model” problems in terms of representing the true asset pricing model. Since tests for
abnormal returns are always a joint test of the risk factors assumed to generate expected
return, it is therefore useful to provide information on the sensitivity of abnormal re-
turn estimates to alternative model specifications. Moreover, factor regressions may
suffer from non-stationarity in the estimated parameters that may be predictable us-
ing publicly available information. Also,Loughran and Ritter (2000)point out that the
factor mimicking portfolios used in the regressions for estimating alphas contain issu-
ing firms, and they argue that this “contamination” may reduce the power of the tests.