Seo Cycles
SEO Cycles
John S. Howe
University of Missouri-Columbia
Shaorong Zhang*
Marshall University
Forthcoming, The Financial Review. Please monitor www.thefinancialreview.org for
the publication schedule. Please consult the published version for exact wording
and pagination before finalizing any verbatim quotes.
Public equity offerings by seasoned firms (SEOs) exhibit similar but less
volatile cycles than initial public offerings of newly public firms. Our
paper provides a comprehensive examination of the factors that cause
variation in the number of firms issuing SEOs. Specifically, we use four
factors from studies of IPOs as potential determinants of SEO cycles. We
find that whether tested separately or collectively, only the demand for
capital and market timing hypotheses receive strong empirical support in
explaining SEO volume. Investor sentiment is not an important factor in
explaining SEO volume, nor is information asymmetry.
JEL Classifications: G10, G30
Keywords: SEO, IPO, Cycles, Information asymmetry, Investment opportunities, Investor
sentiment, Market timing
* Corresponding author: Division of Finance and Economics, Marshall University, One John Marshall Drive,
Huntington, WV 25755; Tel: (304) 696-2605; Fax: (304) 696-3662; Email: zhangs@marshall.edu.
We thank Re-Jin Guo for helpful comments and discussions, as well as seminar participants at the University
of Missouri and Seattle University. The suggestions of the referee and Arnie Cowan, the Editor, were very
useful. We thank IBES for providing analyst earnings forecast data. We are responsible for all remaining
errors.
1. Introduction
Many studies have reported cycles in initial public offerings (IPOs) and have
examined the factors associated with the wide variation of those equity offering activities
over time (e.g., Ibbotson and Jaffe, 1975; Ibbotson, Sindelar, and Ritter, 1988; Lowry and
Schwert, 2002; among others). However, cycles in seasoned equity offerings (SEOs) have
received less attention. Choe, Masulis, and Nanda (1993) and Bayless and Chaplinsky
(1996) both report the variation of SEO volume over time, but they each focus on a single
explanation for this variation. Ours is the first comprehensive study of SEO cycles, that is,
the first to consider a wide range of explanations for the time-varying nature of SEOs.
Lowry (2003) tests three hypotheses that seek to explain the fluctuations in IPO
volume: information asymmetry, demand for capital, and investor sentiment. She finds
that all three are related to changes in IPO volume over time. However, the demand for
capital and investor sentiment hypotheses receive stronger support than the information
asymmetry hypothesis. Pástor and Veronesi (2005) develop a model of IPO market timing
and predict that IPO waves should be preceded by high market returns and followed by
low market returns. Although studies have reported the cyclical phenomenon of SEOs, no
study has examined factors other than information asymmetry as the cause for the variation
of SEOs.
We believe that the hypotheses tested in Lowry (2003) and Pástor and Veronesi
(2005) apply to SEOs as well. Our primary objective is thus to investigate the
determinants of SEO cycles in a comprehensive manner, that is, by considering the
hypotheses of Lowry and Pástor and Veronesi. Because SEOs and IPOs are both equity-
raising events, they can be influenced by common factors. Therefore, our second objective
1
is to compare SEO and IPO cycles to better understand the differences and similarities
between the two.
The hypotheses include the following predictions about SEOs. First, the
information asymmetry hypothesis predicts that SEO volume is negatively related to
information asymmetry. Second, the demand for capital hypothesis predicts that the
volume of SEOs is positively related to investment opportunities. Third, the investor
sentiment hypothesis predicts that SEO volume is positively related to sentiment. Fourth,
the market timing hypothesis predicts that the volume of SEOs is positively related to the
market valuation of equity.
Descriptively, we find that there is significant variation in the monthly (quarterly)
SEO volume, but this variation is smaller than that of monthly (quarterly) IPO volume.
For example, there are months with no IPOs. However, seasoned firms issue equity every
month in the 33 years from 1970 to 2002. Our main results concern the factors that
influence SEO variation over time. Whether tested separately or collectively, only the
demand for capital and market timing hypotheses receive strong empirical support in
explaining SEO volume. Investor sentiment is not an important factor in explaining SEO
volume, nor is information asymmetry.
Our results are mostly consistent with Pástor and Veronesi (2005) that market
timing is an important consideration when firms raise equity. Similar to Lowry (2003), we
also find that demand for capital is a significant factor for SEO cycles, and information
asymmetry is not an important factor. However, seasoned firms appear to be less
susceptible to investor sentiment and information asymmetry than newly public firms.
2
2. Literature review and hypotheses
2.1. Information asymmetry
Early studies on SEOs (e.g., Mikkelson and Partch, 1986; Asquith and Mullins
1986; Masulis and Korwar, 1986) find that the market reacts negatively to the
announcement of an SEO, an average decrease of two to three percent on or around the
announcement date. The negative announcement effect is generally interpreted as
consistent with the Myers and Majluf (1984) adverse selection model. Korajczyk, Lucas,
and McDonald (1991) develop a time-varying information asymmetry model of equity
issuance. An implication of their model is that more firms issue when the information
asymmetry problem is less pronounced.
Choe, Masulis, and Nanda (1993) argue that firms sell seasoned equity when they
face lower adverse selection costs, which occurs in periods with more promising
investment opportunities and with less uncertainty about assets in place. They find that
adverse selection is significantly lower in expansionary periods and in periods with a
relatively larger volume of equity financing.
Bayless and Chaplinsky (1996) find that the price reaction to equity issue
announcements in high volume periods is smaller than in low equity volume periods. They
interpret this evidence as supportive of the existence of “windows of opportunity” for
equity issues that can be partially explained by reduced levels of information asymmetry.
They use seasoned stock issue volume as a measure of market conditions but do not
investigate why the SEO volume changes over time.
Based on these models, we expect SEO volume to be negatively related to proxies
for information asymmetry.
3
2.2. Demand for capital
The demand for capital hypothesis says that firms issue more equity when they
have better investment opportunities. When overall economic conditions improve and all
firms have better expected growth, financing activity increases. Although firms can
finance by issuing securities other than equity, firms may prefer equity to debt in such
periods (Baker and Wurgler, 2000). Lowry (2003) finds evidence that IPO volume is
positively related to demand-for-capital proxies. Harjoto and Garen (2003) find that IPO
firms with higher unanticipated positive growth are more likely to conduct an SEO during
four years after their IPOs. Brau, Ryan, and DeGraw (2006) provide survey evidence from
CFOs that funding for future growth and market timing, among others, are the most
important decision when firms make IPO decisions.
We thus expect SEO volume to be positively related to future investment
opportunities in the aggregate economy.
2.3. Investor sentiment
The investor sentiment hypothesis predicts that changes in investor sentiment cause
the costs of issuing equity, and thus SEO volume, to fluctuate over time. Baker and
Wurgler (2006) study the effect of investor sentiment on a cross section of stock returns.
Cornelli, Goldreich, and Ljungqvist (2006) find that overoptimistic sentiment among small
retail investors is related to high initial returns and subsequent poor long-run performance.
To our knowledge, no other study to date examines how investor sentiment affects
seasoned equity offerings.
4
During periods when investors are especially optimistic and are willing to pay
higher prices for equity, firms should try to issue more equity. Thus, we expect SEO
volume to be positively related to investor sentiment.
2.4. Market timing
The market timing hypothesis predicts that firms time market conditions to issue
securities and choose to issue equity when market valuation of equity is high. This
hypothesis appears in studies of both IPOs and SEOs. For example, Lucas and McDonald
(1990) present a model showing that stock issues should be preceded by positive market
returns, and there is also a positive correlation between volume of issues and general
market returns. Benninga, Helmantel, and Sarig (2005) develop a model explaining
entrepreneurs’ going public and reprivatizing decisions. The model predicts that when
firm values are higher because of a better economic environment, more firms go public.
Because of strong cross-correlation between cash flows generated by all firms, IPOs come
in waves.
Baker and Wurgler (2000) find that the equity share in new issues has predictive
power for aggregate stock market returns. High equity shares predict low market returns.
They interpret this evidence as market timing by firms and as inconsistent with market
efficiency. The variable of interest in their paper is the equity share, which is the total
equity issue as a fraction of total equity and debt issues. They do not distinguish between
issues by IPO firms and SEO firms.
Lowry and Schwert (2002) attribute the positive relation between IPO initial
returns and future number of IPOs as evidence of firms timing market conditions. They
find that more firms file for IPOs and fewer firms withdraw offerings after observing high
5
initial returns of recent IPOs. Graham and Harvey (2001) conduct a survey asking CFOs
directly what they think to be the most important factors affecting their firms’ stock
issuance decisions. They report that firms are reluctant to issue common stock when they
perceive that the stock is undervalued. Recent stock price performance is the third most
popular factor affecting equity-issuance decisions, also supportive of the “windows of
opportunity” argument.
Pástor and Veronesi (2005) develop a model of IPO timing and predict that IPO
waves are caused by three broad market conditions: declines in expected returns, increases
in expected profitability, and increases in a priori uncertainty. Thus, IPO waves should be
preceded by high market returns and followed by low market returns.
In this paper, we test the market timing hypothesis as applied to SEOs: if firms time
market valuations to issue more equity when the entire market valuation of equity is high,
then SEO volume should be positively related to the market valuation of equity.
3. Data and sample description
3.1. Data
We use Thomson Financial Global New Issues data for our SEO and IPO samples.
We obtain a total of 26,172 common stock offerings in the United States from 1970 to
2002. Among these, 14,575 are seasoned equity offerings and 11,597 are initial public
offerings. We discard closed end funds, unit offerings, real estate investment trusts, and
ADRs. We delete issues with an offer price less than three dollars or total proceeds of less
than $0.5 million. We also require that the issues have at least 25% primary shares. Using
these criteria, we identify 8,250 SEOs and 8,136 IPOs. We measure monthly and quarterly
6
SEO volume and IPO volume by the number of issues (N
).1 Following Choe,
SEO, NIPO
Masulis, and Nanda (1993) and Lowry (2003), we also scale both monthly and quarterly
series by the total number of public companies at the end of last year (N
). The
SEO,% , NIPO,%
scaled monthly and quarterly number of offerings control for changes in economic activity
over time.
To investigate how SEO cycles are affected by various economic and noneconomic
factors, we also create proxies for the factors. As discussed earlier, there are four
hypotheses that have been proposed to explain the variation in IPO market activity. We
construct proxies for the factors in the four hypotheses, following Lowry (2003) and Pástor
and Veronesi (2005), to examine their explanatory power for SEO cycles. We measure the
variables at quarterly intervals for two reasons. The first is to be consistent with prior
studies and to facilitate the comparisons of our results with other studies, and the second is
that many variables are not available at monthly intervals.
To measure the effect of asymmetric information on equity issuance, we construct
three proxies. The first is the standard deviation of abnormal returns around earnings
announcement days (ABRETSTD). For ABRETSTD, we first calculate the three-day
cumulative abnormal returns for all firms making earnings announcement in a quarter, and
then compute the standard deviation of all abnormal returns observed around the earnings
announcement day in that quarter. The abnormal return is the difference between the stock
return and the return of the market. We use equally weighted market returns.
1 Another way to measure monthly IPO or SEO volume is to use total proceeds raised each month deflated by
the market capitalization. Lowry (2003) shows that the number of issues series and proceeds series are
highly correlated with a correlation of 0.90 between 1970 and 1996. Thus, we only use the monthly number
of issues as a measure of IPO and SEO volume.
7
Another measure of information asymmetry is the standard deviation of analysts’
forecasts of long-term earnings growth for each quarter (IBESSTD). This is the average
standard deviation of analyst forecasts of long-term earnings growth for all companies
during a quarter. The construction of these two variables follows Lowry (2003). Our third
measure of information asymmetry is the standard deviation of daily market returns within
a quarter (MRETSTD). We expect these proxies to be negatively related to SEO volume.
We use five variables to test the demand for capital hypothesis: the one-year GDP
growth rate following a quarter (GDP+1, +4), the one-year sales growth rate of all public
companies (Sales+1, +4), the aggregate market-to-book ratio (MTB), the discount rate (T-
bond rate), and expected future long-term earnings growth (ΔIBESMEAN). We obtain
quarterly GDP data from NBER. The sales growth rate is measured by aggregating the
revenues of all public companies on Compustat for two consecutive quarters, computing
the quarterly sales growth rate, then using the quarterly growth rate to find the one-year
growth rate after each quarter. Firms should issue equity before rapid economic growth if
the issues are driven by the demand for capital. Thus, we expect SEO issue volume to be
positively related to future GDP and sales growth.
We also construct the quarterly aggregate market-to-book (MTB) ratio across all
public firms available on CRSP and Compustat. We calculate the market value (number of
shares outstanding multiplied by quarter-end closing price) for each firm each quarter for
which we have market value at the end of the quarter and book value at the end of previous
fiscal year. We aggregate market value and book value of all available firms and divide
the aggregate market value by the aggregate book value. The market-to-book ratio is
8
widely used as a measure of investment opportunities. We expect SEO volume to be
positively related to prior and concurrent MTB.
A lower discount rate could result in more attractive corporate investment projects.
We use the ten-year T-bond rate as a proxy for the discount rate. We obtain monthly T-
bond rates from the Federal Reserve. We also use financial analysts’ forecast of long-term
earnings growth (ISEBMEAN) as another proxy for future investment opportunities. We
obtain analysts’ long-term earnings growth forecasts from IBES. For each quarter, we
calculate the average forecasted long-term growth rate. SEO volume is expected to be
negatively related to the discount rate and positively related to the long-term earnings
growth rate.
To test the investor sentiment hypothesis, we use future market returns, as in Lowry
(2003). We also use the consumer sentiment index (SINDEX) from the website of The
Conference Board, Inc.
To test the market timing hypothesis, we construct an aggregate market return
series. The market returns are equally weighted quarterly returns. For each quarter, we
calculate issuing quarter market return (MRET), previous quarter market return (MRET-1),
one year prior market returns (MRET-4, -1), one quarter after market returns (MRET+1), and
one year after market returns (MRET+1, +4). We use these market returns to test whether
firms successfully time the market to sell shares when the overall market valuation is high.
We expect past market returns to be positively related to SEO volume, and future market
returns to be negatively related to SEO volume.
To control for seasonality in the equity issuance market, we use an indicator
variable for the first quarter of the year (QTR1). Lowry (2003) reports that there are
9
significantly fewer IPOs and SEOs in the first quarter of the year. We also use a business
cycle indicator EXPANSION, which has a value of one if two consecutive months in a
quarter are in the expansionary phase of a cycle.
3.2. SEO and IPO cycles
Table 1 presents descriptive statistics of both monthly and quarterly SEO volume
and IPO volume for comparison from 1970-2002. The average monthly numbers of IPOs
and SEOs are similar, 20.83 for SEOs and 20.55 for IPOs. However, the standard
deviation of monthly number of SEOs (12.27) is much lower than that of IPOs (18.44),
which suggests that the IPO market is more volatile than the SEO market. The smallest
monthly number of SEOs for the entire sample period is two. However, there are 32
months in the 33-year sample period without any IPOs. For the quarterly SEO and IPO
series, the mean number is 62.5 for SEOs and 61.64 for IPOs. Again the standard
deviation of the SEO quarterly series (32.98) is significantly lower than that of the IPO
series (52.00).2 The IPO market is more volatile than the SEO market and clustering
during certain periods is more pronounced for unseasoned firms than for seasoned firms.
[Insert Table 1 Here]
The scaled monthly (quarterly) number of SEOs ranges between 0.03% (0.18%)
and 1.34% (3.26%) with a mean of 0.33% (1.00%). The scaled number of monthly
(quarterly) IPOs ranges between zero (zero) and 1.43% (3.40%) with a mean of 0.30%
(0.91%). Again, the standard deviations of the monthly and quarterly scaled number of
2 Although not reported in Table 1, we conduct an F-test to determine whether the variances of monthly and
quarterly SEO series are equal to those of monthly and quarterly IPO series. For the monthly series, the F
value is 2.26, significant at 1% level. The F value is 2.48 for the quarterly test, also significant at the 1%
level.
10
SEOs (0.21% and 0.59% respectively) are smaller than those of monthly and quarterly
scaled number of IPOs (0.27% and 0.75% respectively).
3.3. Descriptive statistics of explanatory variables
[Insert Table 2 Here]
Table 2 shows descriptive statistics for the explanatory variables used in our
multiple regressions. The moving one-year GDP growth after each issuing quarter is 3.1%
and the one-year sales growth of all public firms averages 9.4%. The aggregate market-to-
book ratio of equity has a mean of 2.10 and varies between 1.02 and 4.45. The T-bond rate
averages 7.95% over the sample period. We are only able to get 85 quarterly observations
for the mean and standard deviation of analysts’ long-term earnings growth forecasts
because this variable is available only after December 1981. The average long-term
earnings growth forecast made in a quarter across all firms is 17.2% with a standard
deviation of 4.2%. The quarterly consumer confidence index has a mean of 85.7 and
varies between 54.4 and 110.1. The average quarterly standard deviation of abnormal
returns around earnings announcement day is 7.8%. The average quarterly market return is
3.7%, and the average standard deviation of daily market returns is 0.6%. Eighty-three
percent of the quarters are designated as in an expansionary phase of business cycle.
4. Multiple regression analysis of SEO cycles
We use multiple regressions to investigate which theories best explain SEO
volume. The dependent variable is the quarterly number of SEOs (NSEO, %) scaled by the
11
number of public firms at the prior year-end. In all regressions, we control for seasonality
using a first quarter indicator, QTR1.
When we use OLS to estimate the relation between SEO volume and explanatory
variables, we find strong evidence of first-order serial correlation of the error terms; the
Durbin-Watson D statistics are very close to zero. Thus, we use AR (1) feasible GLS or
the Cochrane-Orcutt procedure (Greene, 2003) to correct for serial correlation.
[Insert Table 3 Here]
4.1. Information asymmetry
Model 1 of Table 3 shows the effect of information asymmetry on SEO volume.
The average standard deviation of financial analysts’ forecast (IBESSTD) shows a positive
sign while the standard deviation of abnormal return around earning announcement day
during the quarter (ABRETSTD) and the standard deviation of analyst long-term earnings
forecast (IBESSTD) both have their expected negative signs, but none is significant. The
EXPANSION indicator is positive but only marginally significant at the 10% level, and the
first quarter indicator is significant at the 5% level. Thus, information asymmetry
variables are not related to SEO volume after controlling for business cycles. This
evidence is in contrast to Choe, Masulis, and Nanda (1993) that information asymmetry
affects SEO cycles.
4.2. Demand for capital
Model 2 of Table 3 shows tests of the effect of future investment opportunities on
SEO volume. GDP growth, change in analysts’ long-term earnings forecast (ΔIBESMEAN),
and prior one-year change of long-term T-bonds are all significant with the expected signs.
The coefficient for GDP growth is 0.18, significant at the 5% level; the coefficient for
12
change in analysts’ long-term earnings forecast (ΔIBESMEAN) is 0.40, significant at the 5%
level; and the coefficient for prior one-year change of long-term T-bond is -0.16, also
significant at the 5% level. The EXPANSION dummy is not significant in Model 2. The
change of aggregate market-to-book ratio (ΔMTB) has the expected positive sign but is not
significant, while sales growth is marginally significant at the 10% level with the wrong
sign.
Thus, not all proxies for the demand for capital are individually significant with the
expected signs. However, this result is likely caused by multicollinearity among the
explanatory variables. We conclude that SEO volume is related to demand for capital.
4.3. Investor sentiment
Model 3 of Table 3 tests the investor sentiment hypothesis. SINDEX has the
expected positive sign but is not significant, and future market return has the expected
negative sign (-0.007) and is significant at the 5% level. Because the future one-year
market return can also be considered a market timing proxy, we conclude that we do not
find reliable evidence that investor sentiment affects SEO cycles.
4.4. Market timing
Model 4 of Table 3 shows tests of the timing hypothesis. Both the past and future
one-year market returns are significant with the expected signs. In addition, the past
quarterly return has a coefficient of 0.006 and is significant at 5%, and the future quarterly
return has the expected sign but is not significant. Thus, seasoned firms issue equity after
market run-ups and prior to market declines, consistent with the market timing hypothesis.
4.5. The horse race
13
So far, when we test the four hypotheses individually with their respective proxy
variables, we find the demand for capital and market timing to be significantly related to
SEO volume. Whether the demand for capital and market timing remain significant after
controlling for competing hypotheses is uncertain. Thus, we estimate a multiple regression
with all explanatory variables.
Six variables in Model 5 of Table 3 have the expected sign and are significant at
the 5% level. The standard deviation of daily market return, the future one-year GDP
growth, the change in analyst long-term earnings growth, the prior one-year T-bond rate
change, and the prior one-year and one-quarter market returns are all significant with the
expected signs. The results are consistent with all hypotheses except investor sentiment.
However, many variables have unexpected signs. Joint F-tests for groups of variables, not
reported in detail, support the demand for capital and timing hypotheses at the 1% level.
The information asymmetry variables as a group are not significant, and the investor
sentiment variables are significant but of the wrong sign.
5. Summary
We empirically examine the cycles of seasoned equity offerings. We find hot and
cold markets over the last three decades and report several new findings. Using the
monthly number of SEOs as a measure of volume, we find that there is significant
variation in monthly SEO volume but that this variation is smaller than that of monthly
IPO volume. It appears that in certain periods, private firms simply find the cost of going
public to be prohibitively high.
14
We further test how economic and noneconomic factors affect SEO cycles. These
factors are suggested by four hypotheses: information asymmetry, demand for capital,
investor sentiment, and market timing. When tested separately or collectively, only the
demand for capital and market timing hypotheses receive strong empirical support in
explaining SEO volume. Investor sentiment is not an important factor in explaining SEO
volume, nor is information asymmetry.
Our results are consistent with Pástor and Veronesi (2005) that market timing is an
important consideration when firms raise equity although they do not control for investor
sentiment in their models. Similar to Lowry (2003), we also find that demand for capital is
a significant factor for SEO cycles and that information asymmetry is not an important
factor. However, contrary to Lowry (2003), we do not find evidence that investor
sentiment affects SEO cycles. After controlling for competing hypotheses, we do not find
evidence that information asymmetry affects SEO cycles in our sample period, which is in
contrast to Choe, Masulis, and Nanda (1993). Seasoned firms are less susceptible to
investor sentiment and information asymmetry than newly public firms.
15
References
Asquith, Paul and David W. Mullins, Jr., 1986. Equity issues and offering dilution, Journal
of Financial Economics 15, 61-89.
Baker, Malcolm and Jeffrey Wurgler, 2000. The equity share in new issues and aggregate
stock returns, Journal of Finance 55, 2219-2257.
Baker, Malcolm and Jeffrey Wurgler, 2006. Investor sentiment and the cross-section of
stock returns, Journal of Finance 61, 1645-1680.
Bayless, Mark and Susan Chaplinsky, 1996. Is there a window of opportunity for seasoned
equity issuance? Journal of Finance 51, 253-278.
Benninga, Simon, Mark Helmantel, and Oded Sarig, 2005. The timing of initial public
offerings, Journal of Financial Economics 75, 115-132.
Brau, James C., Patricia A. Ryan, and Irv DeGraw, 2006. Initial public offerings: CFO
perceptions, The Financial Review 41, 483-511.
Choe, Hyuk, Ronald W. Masulis, and Vikram Nanda, 1993. Common stock offerings
across the business cycle, Journal of Empirical Finance 1, 3-31.
Cornelli, Francesca, David Goldreich, and Alexander P. Ljungqvist, 2006. Investor
sentiment and Pre-IPO markets, Journal of Finance 61, 1187-1216.
Graham, John R. and Campbell R. Harvey, 2001. The theory and practice of corporate
finance: Evidence from the field, Journal of Financial Economics 60, 187-243.
Greene, William H., 2003. Econometric Analysis, 5th edition. (Pearson Education, Inc.,
Upper Saddle River, New Jersey).
Harjoto, Maretno and John Garen, 2003. Why do IPO firms conduct primary seasoned
equity offerings? The Financial Review 38, 103-125.
Ibbotson, Roger G. and Jeffrey F. Jaffe, 1975. “Hot issue” markets, Journal of Finance 30,
1027-1042.
Ibbotson, Roger G., Jody L. Sindelar, and Jay R. Ritter, 1988. Initial public offerings,
Journal of Applied Corporate Finance 1 (Summer), 37-45.
Korajczyk, Robert A., Deborah J. Lucas, and Robert L. McDonald, 1991. The effect of
information releases on the pricing and timing of equity issues, Review of Financial
Studies 4, 685-708.
16
Lowry, Michelle, 2003. Why does IPO volume fluctuate so much? Journal of Financial
Economics 67, 3-40.
Lowry, Michelle and G. William Schwert, 2002. IPO market cycle: Bubbles or sequential
learning? Journal of Finance 57, 1171-1200.
Lucas, Deborah J. and Robert L. McDonald, 1990. Equity issues and stock price dynamics,
Journal of Finance 45, 1019-1043.
Masulis, Ronald W. and Ashok N. Korwar, 1986. Seasoned equity offerings, Journal of
Financial Economics 15, 91-118.
Mikkelson Wayne H. and M. Megan Partch, 1986. Valuation effects of security offerings
and the issuance process, Journal of Financial Economics 15, 31-60.
Myers, Stewart C. and Nicholas Majluf, 1984. Corporate financing and investment
decisions when firms have information the investors do not have, Journal of
Financial Economics 13, 187-221.
Pástor, Ĺuboš and Pietro Veronesi, 2005. Rational IPO waves, Journal of Finance 60,
1713-1757.
17
Table 1
Descriptive statistics of SEOs and IPOs from 1970-2002
This table shows descriptive statistics of SEOs and IPOs from 1970-2002. We report the
mean, median, standard deviation (STD), minimum (Min), maximum (Max), and number
of observations (N) for each variable. NSEO, M (NIPO, M) denotes the number of SEOs (IPOs)
in a month from January 1970 to December 2002. N%SEO, M (N%IPO, M) denotes the number
of SEOs (IPOs) as a percentage of the number of total public firms at the end of last year in
a month from January 1970 to December 2002. NSEO, Q (NIPO, Q) denotes the number of
SEOs (IPOs) in a quarter from 1970 to 2002. N%SEO, Q (N%IPO, Q) denotes number of SEOs
(IPOs) as a percentage of the number of total public firms at the end of last year in a
quarter from 1970 to 2002.
Variable Mean
Median
STD
Min
Max
N
NSEO, M
20.83
18.50
12.27
2
74
396
NSEO, Q
62.50
55.00
32.98
13
180
132
NIPO, M
20.55
17.00
18.44
0
83
396
NIPO, Q
61.64
50.00
52.00
0
202
132
N%SEO, M
0.334 0.283 0.213 0.028 1.340 396
N%SEO, Q
1.002 0.884 0.586 0.182 3.260 132
N%IPO, M
0.303 0.253 0.266 0
1.431 396
N%IPO, Q
0.909 0.779 0.749 0
3.404 132
18
Table 2
Descriptive statistics of explanatory variables
GDP+1, +4 is the one-year GDP growth rate after an issuing quarter. Sales+1, +4 is the one-
year sales growth rate for all public firms after an issuing quarter. MTB is the average
market-to-book ratio of all firms available on CRSP and Compustat in a quarter.
IBESMEAN is the average analyst forecast of long-term earnings growth for all firms made
in a quarter. IBESSTD is the average of standard deviations of analysts’ forecast of long-
term earnings growth for all firms made in a quarter. T-bond is the one-year T-bond rate at
the end of a quarter. SINDEX is the consumer confidence index published by The
Conference Board Inc. ABRETSTD is the standard deviation of abnormal returns around
earnings announcements across all firms which made earnings announcements in a quarter.
MRET is the market return for an issuing quarter, and MRETSTD is the standard deviation
of daily market returns during that quarter. MRET-4, -1 (MRET+1, +4) is the one-year market
return prior to (after) an issuing quarter. EXPANSION is a business cycle indicator
variable with a value of 1 if NBER defines a quarter is in an expansionary phase of a
business cycle.
Variable Mean
Median
STD
Min
Max
N
GDP+1, +4
3.096
3.444
2.323
-2.861
8.646
129
Sales+1, +4
9.408
10.158
6.401
-9.117
19.858
132
MTB
2.100
1.939
0.828
1.024
4.449
132
IBESMEAN
17.198
16.339
2.519
14.291
23.631
85
IBESSTD
4.214
3.850
0.809
3.371
6.263
85
T-bond
7.953 7.460 2.373 3.870
15.320 132
SINDEX
85.748 88.800 12.539 54.400 110.100 132
ABRETSTD 7.815 7.324 2.169 4.570
13.516 126
MRET
3.697
3.253
12.772
-29.415
48.085
132
MRETSTD
0.641
0.549
0.334
0.250
2.484
132
MRET-4,-1
14.560 16.206 25.114 -38.469 107.089 132
EXPANSION
0.826
1.000
0.381
0.000
1.000
132
19
Table 3
Multiple regression analysis of SEO volume (NSEO, %)
We use the AR (1) feasible GLS (Cochrane-Orcutt procedure) to correct for
autocorrelation in error terms. The dependent variable is the quarterly number of SEOs
(NSEO, %) scaled by the number of public firms at the prior year-end. IBESSTD is the
average of standard deviations of analysts’ forecast of long-term earnings growth for all
firms made in a quarter. ABRETSTD is the standard deviation of abnormal returns around
earnings announcements across all firms who made earnings announcements in a quarter.
MRETSTD is the standard deviation of daily market returns during a quarter. GDP+1, +4 is
the one-year GDP growth rate after an issuing quarter. Sales+1, +4 is the one-year sales
growth rate for all public firms after an issuing quarter. IBESMEAN is the average analysts’
forecast of long-term earnings growth for all firms made in a quarter. MTB is the average
market-to-book ratio of all firms available on CRSP and Compustat in a quarter. T-Bond-4,
-1 is the one-year T-bond rate change at the prior quarter-end. T-Bond+1, +4 is the one-year
T-bond rate change after an issuing quarter. SINDEX is the consumer confidence index
published by The Conference Board Inc. MRET-1 (MRET+1) is the quarterly market return
prior to (after) an issuing quarter. MRET-4, -1 (MRET+1, +4) is the one-year market return
prior to (after) an issuing quarter. EXPANSION is a business cycle indicator variable with
a value of 1 if NBER defines a quarter is in an expansionary phase of a business cycle. We
report the number of observations (OBS), adjusted R-squared (Adj. R2), and F-value for
each model.
20
Table 3 continued
1 2 3 4 5
Variables
Coeff.
t-stat Coeff. t-stat Coeff. t-stat Coeff. t-stat Coeff. t-stat
IBESSTD 0.211
1.66
0.042
0.70
ABRETSTD -0.020
-0.54
0.023
1.07
MRETSTD -0.201
-1.56
-0.250
-2.02**
GDP+1, +4
0.177
4.77**
0.141
4.83**
Sales+1, +4
-0.020
-1.99
-0.015
-2.01**
ΔIBESMEAN 0.401
2.91**
0.264
2.38**
ΔMTB
0.190
1.32
0.210
1.30
T-Bond-4, -1
-0.164
-4.68**
-0.107
-3.99**
T-Bond+1, +4
-0.012
-0.30
-0.009
-0.30
SINDEX
0.001
0.14
-0.013
-2.81**
MRET-4,-1 0.005
2.69**
0.006
2.91**
MRET-1
0.006
2.70**
0.014
5.11**
MRET+1
-0.001
-0.60
-0.003
-0.99
MRET+1,+4
-0.007
-3.91**
-0.006
-3.22**
0.001
0.46
EXPANSION
0.310
1.68
0.084
0.53
-0.002
-0.01
QTR1
-0.254 -3.73** -0.182 -2.39** -0.264 -5.11** -0.239 -4.71** -0.083 -1.30
N
84 80 130 130 80
Adj. R2
0.43
0.85
0.45
0.58
0.95
F
11.51 52.12 27.93 30.47 82.96
** denotes significance at the 5% level.
21
