346 B.E. Eckbo et al.
Ta b l e 1 8
(Continued)Study Issuer
type
Sample
sizeSample
periodHolding
periodBHRi
−BHRmE. Convertible debt offerings
Spiess and Affleck-Graves (1999) All 400 1975–1989 5 yrs − 37 .0%∗
Lee and Loughran (1998) All 986 1975–1990 5 yrs − 30 .4%∗
Eckbo, Masulis, and Norli (2000) Ind 459 1964–1995 5 yrs − 16 .1%
Kahle (2000) Ind 527 1981–1992 3 yrs − 18 .1%∗
Lewis, Rogalski, and Seward (2001) All 566 1979–1990 5 yrs − 26 .5%∗
Butler and Wan (2005) Ind 303 1975–1999 5 yrs − 24 .0%∗h
Buy-and-hold percent returns are defined as:
BHR≡^1
N∑N
i= 1[∏Tit=τi( 1 +Rit)− 1]
× 100.Superscript∗indicates significantly different from zero at the 1% level.
aSample of non-venture-backed IPOs.
bSample of venture-backed IPOs.
cSample of primary issues. Matching firms are matched on size, book-to-market and prior annual return.
dSample of completed SEOs.
eSample of cancelled SEOs.
fReturns to non-participating investors (who do not buy shares in the private placement).
gReturns to participating investors (those who also capture the discount in the offering).
hAbnormal returns are insignificant when also matching on liquidity.
often thought to be associated with managerial beliefs that the firm’s future earnings
prospects are good (e.g.,Ross, 1977). So why would debt issuers underperform non-
issuing firms matched on size and B/M? Moreover, why would this underperformance
be close to the magnitude for SEOs? The answer may reflect a combination of statis-
tical problems with buy-and-hold returnBHR, as well as the matched firm technique
producing the wrong benchmark for measuring the true systematic risk of issuing firms.
In the subsequent section, we address this issue by measuring abnormal performance to
issuing firms using both a monthly return horizon and a risk adjustment emanating from
factor regressions.
Eckbo and Norli (2005)also examine the frequency of company delistings from the
stock exchange due to bankruptcy/liquidation over the five-year period following IPOs.
The idea is that low post-issue returns may be driven by a greater exit due to bank-
ruptcy/liquidation compared to the rate for the matched firms. However, they find no
evidence that the rate of bankruptcy/liquidations (or delisting due to takeover) differs
across issuer and their matches.