Video 3

International Journal of Managerial Finance
Dynamic relationships and technological innovation in hot and cold issue markets
Thomas J. Walker Michael Y. Lin
Article information:
To cite this document:
Thomas J. Walker Michael Y. Lin, (2007),"Dynamic relationships and technological innovation in hot and
cold issue markets", International Journal of Managerial Finance, Vol. 3 Iss 3 pp. 200 - 228
Permanent link to this document:
http://dx.doi.org/10.1108/17439130710756899
Downloaded by FREIE UNIVERSITAT BERLIN At 02:44 04 March 2017 (PT)
Downloaded on: 04 March 2017, At: 02:44 (PT)

References: this document contains references to 44 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 639 times since 2007*
Users who downloaded this article also downloaded:
(2003),"Strategic management of technological innovations in the small to medium enterprise", European
Journal of Innovation Management, Vol. 6 Iss 1 pp. 18-31 http://dx.doi.org/10.1108/14601060310456300
(2004),"The use of strategies in managing technological innovation", European Journal of Innovation
Management, Vol. 7 Iss 3 pp. 229-250 http://dx.doi.org/10.1108/14601060410549919
Access to this document was granted through an Emerald subscription provided by emerald-srm:514603 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.

The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1743-9132.htm
IJMF
3,3 Dynamic relationships and
technological innovation in hot
and cold issue markets
200
Thomas J. Walker and Michael Y. Lin
Department of Finance, Concordia University, Montreal, Canada
Abstract
Purpose – The puzzle of hot and cold issue markets has attracted substantial interest in the academic
community. The behavior of IPO volume and initial returns over time is well documented. Few studies,
however, investigate the dynamic interrelationship between these two variables. This paper aims to
fill this gap. In addition, the technological innovations hypothesis of hot issue markets is tested. Welch
and Hoffmann-Burchardi suggest that the clustering of new issues is caused by IPO volume spikes in
industries that have recently experienced technological innovations or favorable productivity shocks.
Design/methodology/approach – This paper employs a sample of 8,160 initial public offerings
filed in the USA between January 1972 and December 2001. A simultaneous equation approach is used
to examine the endogenous relationship between IPO volume and initial returns. In addition, the paper
analyzes the industry correlation matrix of new issue activity and estimates a fixed-effects model
based on industry-level data to examine the impact of technological innovations on new issue activity.
Findings – It is found that higher IPO volume causes higher initial returns, but not vice versa. In
addition, evidence is found against the technological innovations hypothesis. The findings suggest
that economy-wide rather than industry-specific factors are responsible for the observed variations in
IPO volume.
Research limitations/implications – As with any empirical study, the results may be
sample-specific.
Originality/value – The paper extends the prior literature on the relationship between IPO volume
and initial returns by applying two-stage and three-stage least squares models that go beyond prior
methodological approaches used in the extant literature. In addition, the paper provides some of the
first empirical evidence on the effect of technological innovations and productivity shocks on IPO
activity.
Keywords Issues, Returns, Innovation, Mathematical modelling
Paper type Research paper
1. Introduction
Both the volume and average initial returns of initial public offerings (IPOs) fluctuate
over time. Cycles in the IPO market have attracted the interest of numerous academics.
Extremes in the cycles are often referred to as hot and cold issue markets. A hot issue
market is generally associated with a period in which the average initial return of new
equity issues is abnormally high, while cold issue markets are associated with low
under-pricing. What factors cause these cycles in the IPO market is still largely
International Journal of Managerial
unknown.
Finance While the extant finance literature provides numerous theories that partially
Vol. 3 No. 3, 2007
pp. 200-228 explain the fluctuations of IPO volume and under-pricing over time, few of them have
q Emerald Group Publishing Limited
1743-9132
explored the dynamic interrelationship between the two variables. We fill this gap by
DOI 10.1108/17439130710756899 examining the endogeneity of IPO volume and initial returns in a simultaneous
equations framework. By using two-stage and three-stage least squares estimation, we Hot and cold
uncover dynamic interrelationships between IPO volume and initial returns that the issue markets
existing literature has not captured. In addition, we add to the extant literature by
examining whether hot issue markets are driven by technological innovations in
certain industries. Chemmanur and Fulghieri (1999) and Hoffmann-Burchardi (2001)
suggest that hot issue markets are the result of positive productivity shocks and
positive surprises about industry prospects in a small number of related industries. 201
Our paper is the first to use a fixed effects model to examine the cross-sectional effects
in IPO issuing activity. Our findings indicate that hot issue markets are not driven by
increased issuing activity in a small number of industries, but that they occur across a
broad range of industries at the same time.
The remainder of this paper is organized as follows. In Section 2, we review the
literature relating to IPO under-pricing, IPO volume and the technological innovations
hypothesis. In Section 3, we describe our data and provide summary statistics for our
sample. In Section 4, we model IPO volume and initial returns in a simultaneous
equations framework. Section 5 tests the technological innovations hypothesis. Section
6 provides concluding remarks.
2. Literature review
2.1. Cycles in IPO volume and initial returns
There are both demand-side and supply-side explanations for the hot and cold issue
market puzzle. Ibbotson and Jaffe (1975) advanced the argument that if issue markets
are demand-driven, entrepreneurs may be better off by going public in a cold issue
period to raise more money for the firm and get a higher price for their stakes in the
firm. In contrast, Loughran and Ritter (1995) found that the long-term performance of
hot issue market issues is lower than the post-issue performance of cold market issues.
They argued that hot markets provide a window of opportunity for even those firms to
go public that would otherwise spark little investor interest.
Ritter (1984) examines the hot issue market of 1980 by studying the risk
characteristics of issuing firms, and finds that the changing risk composition cannot
explain the unusually high average initial returns during this hot issue period. Rock
(1986) argued that riskier firms should have higher average initial returns compared to
firms that are easier to evaluate. He argued that a hot issue market takes place when a
large proportion of new issues have high risk, and that a cold issue market takes place
when a large proportion of new issues have low risk. Similarly, Choe et al. (1993) found
that capital can be raised at a lower cost when certain periods offer a window of
opportunity.
Consistent with Choe et al.’s (1993) findings, Bayless and Chaplinsky (1996)
examined volume trends in seasoned equity issuance. They found that the price
reaction to seasoned equity issue announcements, measured as the cumulative
announcement date prediction errors from day 2 1 to day 0, in periods with high issue
volume is approximately 200 basis points lower than during periods with low equity
issue volume. In a nutshell, a typical hot market issuer foregoes about 2 percent in
additional equity value compared to a cold market issuer. Similarly, Ljungqvist (1997)
documented that a positive macroeconomic climate raises the average amount of
under-pricing.
IJMF In addition to the variations in IPO under-pricing over time, we can also observe
3,3 significant variations in IPO volume. Lee et al. (1991) argued that the large
fluctuations in IPO volume are the result of market irrationality, i.e. firms time their
IPOs by taking advantage of investor over-optimism. Similarly, Rajan and Servaes
(1997) argued that IPO volume is influenced by the level of over-optimism in
securities markets. Consistent with the over-optimism hypothesis, Pagano et al.
202 (1998) suggest that fluctuations in IPO volume are primarily driven by owners’
attempts to exploit sectoral mispricing. Lerner (1994), Loughran and Ritter (1995)
and Pagano et al. (1998) provide empirical results that are consistent with this
hypothesis. Lerner et al. (2003) interpret that during low IPO volume periods private
firms cannot access the stock markets under favorable conditions, thus they have to
seek less favorable financing methods.
Empirical findings suggest that the under-pricing and IPO volume cycles are not
perfectly synchronized. Ibbotson et al. (1994) find that initial returns lead IPO
volume by six to 12 months in the US market. Lee et al. (1994) and Ljungqvist
(1995) document a similar lead-lag relationship in Australia as well as in Germany
and Sweden, respectively. Hoffmann-Burchardi (2001) claims that the IPO price of
one firm serves as a feedback mechanism to other issuing firms since it can reveal
information about the common value factor and therefore change the value of other
firms. Similarly, Lowry and Schwert (2002) suggest that high initial returns lead to
higher IPO volume in subsequent periods and that both the cycles in initial returns
and the lead-lag relationship between initial returns and IPO volume are mostly
driven by information learned during the registration period. Their findings reveal
that the relationship is caused by a positive information feedback, which causes
more companies to file their IPOs after periods of high initial returns. Ibbotson and
Jaffe (1975) employ simple OLS regressions to try to reveal the relation between
initial returns and IPO volume but observe no significant relationship between the
two variables. Using firm-specific data, Cliff and Denis (2004) indicate a negative
relationship between under-pricing and IPO volume, i.e. under-pricing is lower when
the IPO volume is high.
2.1.1 Business conditions/capital demand. There are various theories that try to
explain the variation of IPO volume over time. The business conditions hypothesis
argues that issue activities mirror the economic climate and will therefore be positively
related to various macroeconomic measures. When the economy is expanding, the cost
of equity capital is lower and firms tend to have higher demand for capital. The ability
to issue new equity at a lower cost motivates private firms to go public. Fama and
French (1989) distinguish between hot and cold issue markets by monthly industrial
production. Choe et al. (1993) suggest that more firms have seasoned equity offerings
when they face better economic conditions. Similarly, Ljungqvist (1997) finds that a
positive macroeconomic environment increases average initial returns. Ritter and
Welch (2002) argue that market conditions are the most important factor in the decision
to go public.
Lowry (2003) provides empirical evidence that companies’ demand for capital
explains a significant amount of the fluctuations in IPO volume. Lowry uses
percentage growth in the real gross domestic product, percentage growth in real
private, fixed non-residential investment, the change in the number of new
corporations and average real sales growth of public firms as proxies for capital Hot and cold
demand. Consistent with the capital demand hypothesis, she finds that sales growth issue markets
is positively related to IPO volume both on an aggregate level and on an industrial
level.
2.1.2 Asymmetric information/adverse-selection. Rock (1986) argued that riskier
firms should have higher initial returns than firms that are easier to evaluate. He
suggested that a hot issue market takes place when a large proportion of the new 203
issues have high risk, and that a cold issue market takes place when a large
proportion of new issues have low risk. Lucas and McDonald (1990) argued that
IPO volume is driven by time variations in the amount of asymmetric information
over time. Choe et al. (1993) argued that periods of economic growth are associated
with both a greater volume of equity issues and lower adverse selection costs.
During periods of high information asymmetry, it is less attractive for issuers to

pursue new issues since the benefits of issuing may not surpass the direct issue cost
plus adverse selection cost. This causes a cyclical pattern in new issues markets as
issuing companies will postpone IPOs or seek alternative financing. Lucas and
McDonald (1990) and Bayless and Chaplinsky (1996) found that hot market issuers
tend to be better quality firms. Similarly, Stoughton et al. (2001) suggested that
being the first firm to go public may convey a strategic advantage relative to rival
firms in the same market segment.
Lowry (2003) used the dispersion of abnormal returns around public firms’ earnings
announcements and the dispersion of analyst forecasts of public firms’ earnings,
obtained from IBES, as proxies for the degree of information asymmetry. Pastor and
Veronesi (2003) used a measure that proxies for new firm excess volatility by
calculating the difference between the median return volatility of new firms and the
volatility of the market. In addition, they used the log difference between the median
market-to-book ratio of new firms and the median market-to-book ratio across all firms
as a proxy for prior uncertainty. Lowry (2003) did not find a significant relationship
between asymmetric information and IPO volume. The findings of Pastor and Veronesi
(2003) were mixed. Consistent with the asymmetric information hypothesis, they found
that new firm excess volatility is positively related to IPO volume, but that the excess
market-to-book ratio of new firms is not.
2.1.3 Investor sentiment. Some researchers argue that changes in investor sentiment
are a driving force behind the observed variations in IPO volume. The investor
sentiment hypothesis is a synonym of the irrational investor hypothesis. Lee et al.
(1991) argued that the large fluctuations in IPO volume are the result of market
mispricing, i.e. firms time their IPOs by taking advantage of investor irrationality.
Pagano et al. (1998) suggested that fluctuations in IPO volume are primarily caused by
issuers’ attempts to exploit industrial overvaluation, which is consistent with the
over-optimism hypothesis. Lerner (1994) and Loughran and Ritter (1995) provided
empirical results that are consistent with this hypothesis. Lerner et al. (2003) argued
that during low IPO volume periods private firms cannot access the stock markets
under favorable conditions, and thus they have to seek less favorable financing
methods. Similarly, Jindra (2001) found that firms are willing to have seasoned equity
offerings if they are overvalued.
IJMF Lee et al. (1991) argued that the discount rate of closed-end funds reflects sentiment
3,3 changes of individual investors that invest in these funds relative to a clientele that
invests in the underlying assets. Following Lee et al. (1991), Lowry (2003) used the
discount rate of closed-end funds and future market returns as proxies for the level of
investor optimism. Consistent with the investor sentiment hypothesis, Lowry (2003)
found a significant negative relationship between the discount rate of closed-end funds
204 and IPO volume.
Similarly, Charoenrook (2002) used the index of consumer sentiment, constructed by
the Survey Research Center at the University of Michigan, as a proxy for the level of
optimism of a broad number of investors. Charoenrook (2002) found that changes in
consumer sentiment predict one-month and one-year stock market returns.
2.2 Technological innovations

The technological innovations hypothesis suggests that hot issue markets can occur
when there is a wave of IPOs by companies in the same industry, which has
experienced a technological innovation. Chemmanur and Fulghieri (1999) suggested
that hot issue markets occur when there are correlated productivity shocks across
firms. Similarly, Maksimovic and Pichler (2001) suggested herding of IPOs in
industries where new-entry risk is significant. Hoffmann-Burchardi (2001) argued that
both the clustering and under-pricing phenomena result from positive surprises about
industry prospects. Benveniste et al. (2002) contended that bundling IPOs in an
industry by investment banks induces hot issue markets. Ivanov and Lewis (2003)
suggested that technological innovations are unlikely to be highly correlated across
industries and that hot issue activities are concentrated within particular industries. In
other words, hot issue markets are dominated by issues from a relatively small number
of industries.
Empirical findings related to the technological innovations hypothesis are mixed.
Ivanov and Lewis (2003) examined the correlation structure of new issue activity using
industry-level data. They found that the correlations across industries are significantly
different, which is supportive of the technological innovation hypothesis.
In contrast, Helwege and Liang (2004) found that both hot and cold IPO markets are
largely concentrated in the same narrow set of industries and that hot markets for
many industries occur at the same time. Thus, they argued that technological
innovations are not the major determinant of hot and cold issue markets. Rather, they
argued that hot markets are driven by investor optimism.
3. Data
Data on IPOs were obtained from the Securities Data Company (SDC) New Issues
database (see Appendix). The database includes 10,184 offerings on US exchanges over
the period 1972-2001. While the SDC database also provides information on non-US
IPOs, we decided to concentrate our analysis on the US market for several reasons.
First, the US accounts for the largest proportion of worldwide IPO activity during our
sample period and has been affected by few regulatory or political changes in the past
30 years. Second, data for non-US IPOs is generally more difficult to obtain and the
inclusion of more than one country introduces potential cross-country biases that a
single-country study is exempt from. We exclude offerings from closed-end mutual
funds, American Depository Receipts (ADRs), Real Estate Investment Trusts (REITs), Hot and cold
unit offerings and mutual-to-stock conversions. We also require price data for the firms issue markets
to be available in the Center for Research in Security Prices (CRSP) database during the
first 45 trading days after the IPO. This yields a sample of 8,160 observations. Taking
into account that new issues of penny stocks are part of the IPO market, we do not
exclude new issues in which the offer price is below $5.00 as Loughran and Ritter
(2002) and Lowry (2003) did. Loughran and Ritter (2002) suggested eliminating penny 205
stocks in IPO studies, since stocks with an offer price below $5.00 per share are subject
to the provisions of the US Securities Enforcement Remedies and Penny Stock Reform
Act of 1990, aimed at reducing fraud and abuse in the penny stock market. We include
penny stocks in our study in order to observe dynamic changes in the IPO market from
a whole market perspective. In a robustness test, we will discuss interrelationships in
the IPO market without the influence of penny stocks.

Table I provides summary statistics for our IPO sample and reports IPO volume,
average initial returns and aggregate gross proceeds of new issues per year. Table I
clearly shows the cyclical patterns of both IPO volume and initial returns. During our
30-year sample period, 8,160 companies went public. Their average initial return was
approximately 17.19 percent, with total gross proceeds of $349.24 billion. Yearly IPO
volume during our sample period ranges from a low of four IPOs in 1974/1975 to a
high of 715 in 1996. Although we can observe a general upward trend in IPO volume
over time, there have been intermittent troughs, particularly in 1974/1975, 1988-1990
and in 2001. On the other hand, we can observe a particularly strong IPO market in
the years 1983, 1993-1996 and 1999. A similar pattern can be observed in initial
returns. IPOs in 1974 had the lowest first day returns (0.68 percent), while the average
first day returns during the 1999/2000 internet bubble were 67.3 percent and 53.9
percent, respectively.
Table II provides a univariate comparison of firm-specific data for high-tech firms
versus non-high-tech firms, and of venture capital backed firms versus non-venture
capital backed firms[1]. Venture-backed IPOs are identified through venture capitalist
information contained in the SDC database. We observe that, on average, high-tech
IPOs are significantly more under-priced and have a higher post-IPO standard
deviation of returns than low-tech IPOs. Assuming that high-tech firms have a higher
level of ex ante uncertainty than firms in traditional industries, our results are
consistent with Rock (1986). Rock (1986) attributed the large under-pricing returns of
high-tech firms to the increased level of information asymmetry among informed and
uninformed investors. Megginson and Weiss (1991) argued that new issues backed by
venture capitalist have higher proceeds and lower initial returns after controlling for
risk. While our proceeds variable is somewhat inconclusive, our univariate
under-pricing results suggest that in contrast to Megginson and Weiss (1991),
venture capital backed issues were significantly more under-priced than non-venture
capital backed IPOs during our sample period[2]. Yet, this is not surprising, as we also
find that venture capital backed IPOs have a higher standard deviation in post-IPO
returns, suggesting that they are riskier issues[3].
Table III provides summary statistics for key variables used in our research. For
each variable we report the number of observations as well as the mean, median,
standard deviation, minimum, and maximum. We define IPO volume as the number of
IJMF
Year Number of IPOs Average first day return (percent) Aggregate gross proceeds ($m)
3,3
1972 27 4.18 1,013.7
1973 45 6.52 789.3
1974 4 0.68 85.8
1975 4 8.68 155.4
206 1976 28 3.74 622.1
1977 17 9.31 368.0
1978 22 20.32 521.6
1979 48 12.90 798.1
1980 98 26.16 2,208.2
1981 250 8.68 4,814.6
1982 93 11.87 1,969.2
1983 587 12.18 19,201.8

1984 278 3.88 4,975.7
1985 287 16.25 12,300.6
1986 583 8.99 23,701.2
1987 438 9.00 17,672.5
1988 184 10.99 6,478.7
1989 174 11.85 8,584.4
1990 149 11.79 5,418.3
1991 330 12.64 15,192.7
1992 447 10.08 22,964.1
1993 697 10.15 51,178.4
1994 528 8.81 28,784.2
1995 488 21.01 26,400.1
1996 715 17.77 36,121.4
1997 486 13.19 30,975.8
1998 299 19.36 21,831.1
1999 467 67.30 37,727.6
2000 308 53.90 29,832.0
2001 79 13.23 8,973.2
Total 8,160 17.19 421,660.1
Notes: This Table provides descriptive statistics for all IPOs between 1972 and 2001. The sample
consists of all IPOs available in the Securities Data Company (SDC) New Issues database. The
database includes 10,184 offerings over the period 1972-2001. We exclude offerings from closed-end
mutual funds, ADRs, REITs, unit offerings and mutual-to-stock conversions. We also require that the
firms have price data available in the Center for Research and Security Prices (CRSP) database during
Table I. the first 45 trading days after the IPO. This yields a sample of 8,160 observations. For each year, we
Descriptive statistics for report the number of IPOs, the equally weighted first day return during the year and aggregate
IPOs, 1972-2001 proceeds in US$ million, converted to year 2000 dollars
IPOs per quarter. Following Lowry (2003), we deflate the number of IPOs per quarter
by the number of public firms at the end of the previous quarter in order to avoid
problems related to non-stationarity and series correlation. We calculate the initial
return during a quarter as the equally weighted mean IPO underpricing of firms that
went public during that quarter[4]. Proceeds, obtained from SDC, are converted to year
2000 dollars based on CPI data from the Bureau of Labor Statistics, and are averaged
across all firms that went public in a given quarter. Following Alti (2006), we expect
IPOs issued during periods of high IPO volume to have higher proceeds as issuers take
Hi-tech mean Non-hi-tech mean t-test Wilcoxon test VC-backed mean Non-VC-backed mean t-test Wilcoxon test
median median ( p-value) median median ( p-value)
n 2,683 5,477 2,661 5,499

Under-pricing
(percent) 25.00 13.36 ,0.001 25.54 13.13 , 0.001
9.10 3.95 ,0.001 8.92 3.92 , 0.001
Post-IPO standard
deviation 0.0456 0.0334 ,0.001 0.0468 0.0333 , 0.001
0.0388 0.0299 ,0.001 0.0397 0.0291 , 0.001
Proceeds 48.43 53.26 0.011 45.00 54.91 , 0.001
29.04 24.72 ,0.001 34.12 20.61 , 0.001
Notes: This Table provides a univariate comparison of hi-tech firms versus non-hi-tech firms, and of venture capital backed firms versus non-venture
capital backed firms using firm-specific data. Under-pricing is the percentage return from the offer price, obtained from SDC, to the first-day closing price,
obtained from CRSP. Post-IPO standard deviation is the standard deviation of IPO returns during the first 30 trading days. Proceeds are the offer proceeds
converted to year 2000 dollars based on CPI data from the Bureau of Labor Statistics (BLS). We follow the classification by Loughran and Ritter (2002) and
Cliff and Denis (2004) into high-tech and low-tech firms. High-tech firms are those firms whose SIC codes match one of the following: 2833, 2834, 2835,
2836, 3571, 3572, 3575, 3577, 3578, 3661, 3663, 3669, 3674, 3812, 3823, 3825, 3826, 3827, 3829, 3841, 3845, 4812, 4813, 4899, 7370, 7371, 7372, 7373, 7374,
7375, 7377, 7378, 7379. Venture-backed IPOs are identified through venture capitalist information contained in the SDC database. In columns four and
seven we report p-values of a t-test for differences in means and of a Wilcoxon test for differences in medians for each comparison of subsamples
issue markets
firm-specific data
Hot and cold
Univariate statistics for

207
Table II.
3,3
208
IJMF
Table III.
Descriptive statistics
Variables n Mean Median Standard deviation Minimum Maximum
Number of IPOs 120 68.00 51.50 61.18 0 238

Initial return 114 15.53a 11.67a 16.26a 2 0.79a 95.09a
Prior market volatility 120 0.0061 0.0052 0.0032 0.0025 0.0248
Proceeds ($m) 120 3,513.83 1,879.41 3,846.00 0 16,201.02
Post-IPO standard deviation 114 0.0364 0.0338 0.0140 0.0166 0.0946
Log(GDP) 120 3.800 3.802 0.1157 3.602 3.9951
Prior market return 120 1.15a 1.14a 4.06a 2 10.84a 13.98a
Risk-free rate 120 6.60a 5.74a 2.65a 1.91a 15.05a
ICS 120 86.12 90.33 12.94 54.43 110.13
HT ratio 114 0.3056 0.2996 0.1554 0 0.7115
VB ratio 114 0.3124 0.2843 0.1625 0 1
Notes: This Table provides summary statistics for variables used in our two-stage regression and three-stage least-squares regression. For each variable
we report the number of observations as well as the mean, median, standard deviation, minimum, and maximum. aFigures given are percentages
advantage of windows of opportunity; that is, they issue more equity to benefit from Hot and cold
the temporarily low cost of capital during those periods. On the other hand, Habib and issue markets
Ljungqvist (1998) observe a negative relationship between proceeds and initial returns.
We use two proxies for the degree of asymmetric information. One is the post-IPO
standard deviation of a firm’s return, and the other is prior market volatility. Following
Rock (1986) and McGuinness (1992), we use the standard deviation of IPO returns
during the first 30 trading days, averaged quarterly. The post-IPO standard deviation 209
captures the degree of asymmetric information from the viewpoint of investors. If
investors think that the degree of asymmetric information is high, then issuers have to
underprice their IPOs sufficiently enough to attract investors. Thus, we hypothesize
that initial returns are positively related to post-IPO standard deviation. On the other
hand, prior market volatility is used to capture the degree of asymmetric information
from the viewpoint of issuers. Prior market volatility is the annualized standard
deviation of returns on the CRSP equally weighted market index during the three
months prior to the firm’s IPO.
We also investigate the influence of prior stock market returns on the IPO market,
as in Benveniste et al. (2003). Prior market returns are the annualized returns during
the three months prior to the firm’s IPO using the equally weighted portfolio of all
NYSE, AMEX, and NASDAQ stocks as provided by CRSP. Following Pastor and
Veronesi (2003), we use prior market returns as a proxy for the equity premium. To
proxy for capital demand, we use the quarterly US Gross Domestic Product (GDP),
obtained from the US Department of Commerce, Bureau of Economic Analysis (see
www.bea.doc.gov). To adjust for inflation, we convert all GDP data to year 2000 dollars
based on CPI data from the Bureau of Labor Statistics. The risk-free rate is the
annualized nominal yield on a three-month T-Bill during a quarter, obtained from the
Federal Reserve website (see www.federalreserve.gov). The consumer sentiment index
is constructed by the Survey Research Center at the University of Michigan. The index
serves as a proxy for consumer sentiment during a given quarter. To examine whether
the IPO market is influenced by productivity boosts that affect the overall economy we
also include a variable that captures changes in productivity. Productivity shocks are
measured as the percentage change in non-farm business output per hour during the
prior four quarters, based on productivity data available through the Bureau of Labor
Statistics (see www.bls.gov). Finally, we construct a high-tech ratio, defined as the
number of high-tech IPOs divided by the total number of IPOs during a quarter, and a
venture capital backed ratio, defined as the number of venture capital backed IPOs to
the total number of IPOs during a quarter. Both variables allow us to control for any
effects that a concentration of high-tech IPOs or venture capital backed IPOs may have
on IPO volume and initial returns in a given quarter.
4. Simultaneous equations approach

4.1. Methodology
We take into account that IPO volume and initial returns during a given time period
may influence each other. In order to reveal the endogeneity between IPO volume and
initial returns, we follow the methodology in Lowry and Shu (2002) and apply a
simultaneous equations approach. In the simultaneous system of equations,
endogenous variables are determined jointly rather than sequentially. Thus a
IJMF two-stage least squares (2SLS) regression of IPO volume and initial returns controls for
3,3 endogeneity. We consider the following model:
Initial return ¼ g1 IPO volume þ a1 X þ b1 X 1 þ 11 ; ð1Þ
IPO volume ¼ g2 Initial return þ a2 X þ b2 X 2 þ 12 ; ð2Þ

210
where initial return is the quarterly average amount of under-pricing and IPO volume
is the number of IPO firms that went public during a given quarter, deflated by the
total number of public companies in the last quarter obtained from CRSP (Lowry,
2003). Both IPO volume and initial return are endogenous variables, i.e. they are
determined within the system of equations. X, X1 and X2 are vectors of instrument
variables. In detail, X is a vector of IPO characteristics that are common to both

equations. X contains lagged initial returns, lagged volume, proceeds, prior market
returns, the consumer sentiment index, the productivity shock, and a season dummy.
The lagged endogenous variables are predetermined variables, in that they are
determined within the system by past values of the variables. Prior market returns, the
consumer sentiment index and our season dummy are exogenous variables. X1 is a
vector of exogenous IPO characteristics that are idiosyncratic to initial returns, but not
IPO volume. X1 includes prior market volatility, the high tech ratio, the venture capital
backed ratio, and the post-IPO standard deviation of returns. X2 is a vector of
exogenous IPO variables that are directly related to IPO volume but not to initial
returns. X2 includes the logged GDP and the risk free rate. The first stage of the
two-stage least squares (2SLS) estimation involves the creation of instrumental
variables. The second stage substitutes the first-stage predicted values for endogenous
variables and generates unbiased parameters in a standard OLS regression.
When there are high correlations between the structural form equations of the
two-stage least squares model, a two-stage least squares estimation is inefficient. To
correct for this problem, we also present results for a three-stage least squares (3SLS)
regression which takes the correlations between error terms of equations into account.
If the disturbances in the different structural form equations are uncorrelated, 3SLS
reduces to 2SLS. There are three steps in the process:
(1) first-stage regressions to get predicted values for the endogenous regressors;
(2) a second-stage to get residuals of each equation to estimate the cross-equation
correlation matrix; and
(3) the final stage to secure generalized least-squares parameters.
One potential shortcoming of a 2SLS or 3SLS approach is the delay between the time a
firm decides to go public and the time its shares are first traded, i.e. the time its
under-pricing return is first observed. Specifically, we observe that the median time
that passes between the date a firm first files its first S1 filing with the Securities and
Exchange Commission (SEC) and its IPO date is 58.3 days in our sample. While we
only report results for models that are based on aggregated quarterly volume and
initial returns, we also perform a series of unreported robustness tests based on
six-month and one-year periods. In addition, we perform a robustness test in which we
measure IPO volume in terms of the number of IPO filings (irrespective of whether they
were ultimately followed through or withdrawn). Finally, we perform a series of Hot and cold
Granger causality tests, as in Lowry and Schwert (2002), to ensure that our results are issue markets
unaffected by methodological choices. Our results are qualitatively and quantitatively
little affected by these variations in sample selection, variable definition or
methodology.
4.2 Empirical results 211

Before we run the 2SLS regression, we analyze the correlation matrix of control
variables grouped by variables that are hypothesized to have an impact on initial
returns or IPO volume, respectively. The results are reported in Table IV. Judging by
the magnitude and significance of the correlation coefficients, there appears to be no
severe multicollinearity in our model.
The results of our 2SLS and 3SLS regressions are reported in Table V. All the
results are robust to serial correlation and heteroscedasticity. Because there are lagged
endogenous variables in our model, a Durbin-Watson test is not appropriate for testing
for autocorrelation in the model. We choose a Durbin h test instead. The p-value of the
Durbin h test is 0.12, indicating that there is no significant autocorrelation in our
model. We also perform a Lagrange multiplier test and exclude the possibility of
heteroscedasticity.
Using a 2SLS model, we find no significant relationship between IPO volume and
initial returns. Both IPO volume and initial returns show seasonal characteristics, i.e.
we observe lower IPO volume and higher initial returns in the first season of the year.
As expected, the high-tech ratio, prior market returns, the consumer sentiment index,
and post-IPO standard deviation have a significant positive relationship with initial
returns.
On the other hand, we observe that the prior market volatility is significantly
negatively related to initial returns. This is somewhat surprising, as we would expect a
higher level of uncertainty to be positively associated with initial returns (see, for
example, Rock, 1986). We can only speculate on the possible reasons for the observed
relationship. Most likely, since prior findings apply at the firm level, our results are
methodology-driven; that is, they are caused by the quarterly aggregation of firms.
Why the relationship reverses (instead of just becoming insignificant) remains
puzzling, however. This could be an interesting area for further research.
In contrast to Alti (2006), we do not find a significant relationship between IPO
volume and proceeds. In addition, unlike Lowry (2003), we do not find support for the
capital demand hypothesis, i.e. GDP does not have a significant relationship with IPO
volume in our model[5]. We find that prior market returns have a significant positive
relationship with IPO volume, which suggests that the higher the prior market returns,
the higher the IPO volume during the following quarter. Similarly, positive
productivity shocks result in higher IPO volume. On the other hand, we do not find
a significant relationship between initial returns and productivity shocks or the
venture capital backed ratio.
In Panel B of Table V, we provide a cross-model correlation matrix computed from
the two-stage least square residuals. We calculate cross-model correlations,
covariances, the inverse of the correlation matrix, and the inverse of the covariance
matrix. We find that the disturbances in the different structural form equations are
3,3
212
IJMF
Table IV.
Correlation matrix of
independent variables
Prior market Post-IPO standard Prior market Productivity Season VB
volatility Proceeds deviation return ICS shock dummy HT ratio ratio
Panel A. Variables having impact on initial IPO returns
Prior market volatility
Correlation 1
p-value
n 120
Proceeds
Correlation 20.1695 1
p-value (0.056)
n 120 120
Post-IPO standard deviation
Correlation 0.5288 0.3476 1
p-value (,0.001) (,0.001)
n 114 114 114
Prior market return
Correlation 0.4933 20.1824 0.0912 1
p-value (0.002) (0.048) (0.028)
n 120 120 114 120
ICS
Correlation 20.2356 20.0871 20.1019 20.0771 1
p-value (0.152) (0.244) (0.491) (0.387)
n 120 120 114 120 120
Productivity shock
Correlation 0.0916 0.1773 0.0401 0.0267 0.3220 1
p-value (0.535) (0.126) (0.792) (0.323) (0.115)
n 120 120 114 120 120 120
Season dummy
Correlation 0.0022 20.0575 20.0063 0.3302 20.0077 0.0266 1
p-value (0.980) (0.519) (0.946) (,0.001) (0.931) (0.743)
n 120 120 114 120 120 120 120
HT ratio
Correlation 0.1557 0.1461 0.2329 0.2138 0.1556 0.4183 0.2940 1
p-value (0.095) (0.119) (0.012) (0.510) (0.604) (0.004) (0.001)
n 114 114 114 114 114 114 114 114
VB ratio
Correlation 0.1046 0.0680 0.1647 0.1022 0.0221 0.3790 0.2698 0.5212 1
p-value (0.264) (0.470) (0.077) (0.719) (0.677) (0.009) (0.003) (,0.001)
n 114 114 114 114 114 114 114 114 114
(continued)
Proceeds Log(GDP) Prior market return Risk-free rate ICS Productivity shock Season dummy
Panel B. Variables having impact on IPO volume
Proceeds
Correlation 1
p-value
n 120
Log (GDP)
Correlation 0.6776 1
p-value (,0.001)
n 120 120
Prior market return
Correlation 20.1824 0.0026 1
p-value (0.048) (0.977)
n 120 120 120
Risk-free rate
Correlation 20.4549 20.3196 20.1109 1
p-value (,0.001) (,0.001) (0.213)
n 120 120 120 120
ICS
Correlation 20.0871 0.6164 20.0771 20.4226 1
p-value (0.244) (,0.001) (0.387) (,0.001)
n 120 120 120 120 120
Productivity shock
Correlation 0.1773 0.2958 0.0267 0.1415 0.3220 1
p-value (0.126) (0.038) (0.323) (0.481) (0.115)
n 120 120 120 120 120 120
Season dummy
Correlation 20.0575 20.0239 0.3302 0.0147 20.0077 0.0266 1
p-value (0.519) (0.788) (,0.001) (0.869) (0.931) (0.743)
n 120 120 120 120 120 120 120
Notes: We present a correlation matrix for all independent variables used in our two-stage and three-stage least squares estimation. For each pair of
variables, we report the correlation coefficient, p-value and the sample size in quarters. Panel A presents the correlations matrix for variables that are
hypothesized to have an impact on initial returns. Panel B presents correlation coefficients for variables hypothesized to have an impact on IPO volume.
Note that there were no IPOs in six quarters. Therefore, some of the variables during these quarters are undefined and are omitted from our calculations
issue markets
Hot and cold
213
Table IV.
IJMF
Model
3,3 2SLS 3SLS
Initial return IPO volume Initial return IPO volume
Panel A. Simultaneous regression results

IPO volume 11.3611 16.6709
214 (0.145) (0.029)
Initial return 20.0012 2 0.0010
(0.901) (0.747)
Intercept 20.1856 20.0144 20.2106 2 0.0318
(0.008) (0.597) (0.039) (0.233)
Prior market volatility 214.6571 213.1037
(0.004) (0.016)
Proceeds 20.4259 0.0190 20.4054 0.0250
(0.402) (0.126) (0.490) (0.103)
Post-IPO standard deviation 10.1478 10.2037
(,0.001) (,0.001)
Log (GDP) 0.0040 0.0068
(0.175) (0.157)
Prior market return 0.1600 0.0516 0.1729 0.0511
(0.031) (0.004) (0.037) (, 0.001)
Risk-free rate 0.0144 0.0201
(0.561) (0.289)
ICS 0.0081 0.0039 0.0062 0.0075
(0.049) (0.068) (0.072) (0.099)
Productivity shock 0.4517 0.0192 0.5528 0.0304
(0.291) (0.032) (0.240) (0.044)
Season dummy 0.0952 20.0043 0.1260 2 0.0051
(0.010) (,0.001) (,0.001) (, 0.001)
HT ratio 0.2517 0.2733
(0.040) (0.034)
VB ratio 20.0831 20.1267
(0.534) (0.260)
Volume lag 1 28.1145 0.7585 212.3370 0.7843
(0.091) (,0.001) (0.009) (, 0.001)
Volume lag 2 0.2558 0.1989 21.2142 0.2042
(0.812) (0.062) (0.502) (0.073)
Volume lag 3 20.3822 20.2423 0.8805 2 0.2170
(0.793) (0.063) (0.725) (0.049)
Volume lag 4 20.4457 20.0020 20.4820 2 0.0279
(0.755) (0.931) (0.730) (0.701)
IR lag 1 0.12031 0.0045 0.0739 0.0033
(0.227) (0.360) (0.461) (0.471)
IR lag 2 20.2305 0.0021 20.2384 0.0020
(0.031) (0.589) (0.022) (0.702)
IR lag 3 0.1552 20.0030 0.1537 2 0.0057
(0.228) (0.445) (0.114) (0.4006)
IR lag 4 20.0515 20.0076 0.0131 2 0.0070
Table V. (0.681) (0.061) (0.823) (0.065)
Simultaneous equations Adjusted R 2/system-weighted R 2 0.64 0.70 0.79
approach (continued)
Hot and cold
Model
2SLS 3SLS issue markets
Initial return IPO volume Initial return IPO volume
Initial return IPO volume

Panel B. Cross-equation relationships
Cross-equation correlation 215
Initial return 1 20.3809
IPO volume 20.3809 1
Cross-equation covariance
Initial return 0.0107 20.0002
IPO volume 20.0002 0.00002
Cross-equation inverse correlation

IPO volume 0.4455 1.1697
Cross-equation inverse covariance
IPO volume 962.3 58,301.6
Notes: In Panel A, we provide results of a two-stage least squares regression and a three-stage least
squares regression to test the relationship between aggregate quarterly IPO volume and initial returns.
Our sample consists of 8,160 IPOs from 1972 to 2001. Coefficients are reported with p-values below. In
a simultaneous system of equations, endogenous variables are determined jointly rather than
sequentially. Thus a two-stage least squares regression of IPO volume and initial returns controls for
endogeneity between the variables. For details of the model, see section 4.1. In Panel B, we provide a
cross-model correlation matrix computed from the two-stage least square residuals. We calculate the
cross-model correlations, covariances, the inverse of the correlation matrix, and the inverse of the
covariance matrix. For each regression model we report parameter estimates with p-values in
parentheses below Table V.
correlated. Therefore, a 3SLS estimation approach is more appropriate than a 2SLS

estimation.
The system weighted R 2 of the 3SLS model is 0.76, which is higher than either of
the adjusted R 2s of the 2SLS model. This supports our belief that a 3SLS model
outperforms a 2SLS model when examining the relationship between initial returns
and IPO volume. The results of the 3SLS estimation show that IPO volume has a
significant positive influence on initial returns, but not vice versa. Consistent with
Booth and Chua’s (1996) spillover hypothesis, we also find that past IPO volume during
the preceding three months has a significant negative influence on initial returns. On
the other hand, IPO volume is positively influenced by IPO volume during the
preceding two quarters and negatively influenced by IPO volume during the preceding
third quarter.
Our findings reveal a dynamic relationship between IPO volume and initial returns.
We do not find support for the conventional wisdom that high IPO activity may follow
high under-pricing. Our results give empirical support to the model of
Hoffmann-Burchardi (2001), who hypothesized that hot issue markets often coincide
with more pronounced under-pricing than cold issue markets. Consistent with
McGuinness (1992), our proxy for ex ante uncertainty has a significant positive
relationship with initial returns.
IJMF As expected, the high-tech ratio is significantly positively related to under-pricing.
3,3 We do not find a significant relationship between initial returns and average quarterly
proceeds. Beatty and Ritter (1986) and Habib and Ljungqvist (1998) observed a
negative relationship between these variables at the firm level. Similar to the findings
of Aggarwal et al. (2002), under-pricing is not significantly related to our venture
capital backed indicator. The puzzling negative relationship between prior market
216 volatility and initial returns remains in our 3SLS estimation, albeit at a somewhat
lower significance level. As in our 2SLS estimation, proceeds are not significantly
related to IPO volume, which is inconsistent with the significant positive relationship
documented by Alti (2006). In the same vein, GDP is not significantly related to IPO
volume, suggesting that firms’ demand for capital and investor sentiment are not
important determinants of IPO volume once we control for productivity shocks, which
is inconsistent with Choe et al. (1993) and Lowry (2003). We find that prior market
returns are significantly positively related to IPO volume, which is consistent with the
conventional belief that hot issue markets tend to follow periods of high market return,
i.e. more firms choose to go public during bull market periods and keep waiting during
bear markets. This empirical result is supportive of Pastor and Veronesi’s (2003)
model. Moreover, consistent with Loughran and Ritter (1995), prior market returns
exhibit a significant positive relationship with initial returns. Finally, the index of
consumer sentiment is significantly positively related to IPO volume, although the
significance level is considerably lower than in our 2SLS estimation.
4.3 Robustness tests

We perform robustness tests using three different data sources and time periods. The
first data set was collected from Jay Ritter’s website (see http://bear.cba.ufl.edu/ritter/
ipodata.html). It contains information on 12,105 IPOs from 1970 to 2001, excluding
closed-end funds, ADRs, REITs, units, mutual-to-stock conversions, and issues in
which the offer price was less than $5. The second data set is the same as our original
data set in Table I, but excludes IPOs with an offer price below $5. The third data set is
also based on our original data set, but is limited to the period from 1980 to 2001. We
cannot calculate the high-tech ratio and the venture capital backed ratio as well as
proceeds and the standard deviation of post-IPO returns for Ritter’s data. Therefore we
drop these variables from our model (see Table VI).
For each sample, the 3SLS regression results are robust with respect to significance
and economic magnitude to those in Table V, suggesting that the inclusion or
exclusion of penny stocks does not have any significant impact on the relationship
between IPO volume and initial returns and that the interaction of IPO volume and
initial returns is persistent across different time periods[6]. We argue that a 3SLS
model is suitable to capture the dynamic endogeneity between IPO volume and initial
returns, and that it overcomes some of the problems in 2SLS estimation. The dynamic
relationship between IPO volume and initial returns cannot be observed in 2SLS
estimation, but it is clearly evident in a 3SLS model.
5. Technological innovations
One hypothesis related to the IPO volume puzzle is that technological innovation leads
to a hot issue market. With data from 1977 to 1982, Ritter (1984) found that the hot
Data source
IPOs excluded with offer price
Ritter’s data below $5 IPOs issued after 1980
Initial return IPO volume Initial return IPO volume Initial return IPO volume
IPO volume 27.6875 15.8741 12.8203

(0.033) (0.106) (0.023)
Initial return 20.1479 20.0019 2 0.0014
(0.269) (0.757) (0.854)
Intercept 2 0.0793 20.0746 20.1778 20.0237 20.1552 0.0781
(0.423) (0.419) (0.005) (0.355) (0.032) (0.150)
Prior market volatility 12.1016 216.4502 216.9022
(0.226) (,0.001) (,0.001)
Proceeds 20.0252 0.0098 20.0279 0.0175
(0.662) (0.313) (0.459) (0.224)
Post-IPO standard deviation 9.6728 14.3451
(,0.001) (,0.001)
Log (GDP) 0.0135 0.0063 2 0.0202
(0.199) (0.195) (0.112)
Prior market return 0.1348 0.2164 0.2141 0.0501 0.2086 0.0461
(0.040) (0.182) (0.002) (0.001) (0.017) (0.014)
Risk-free rate 0.1150 0.0083 2 0.0297
(0.362) (0.652) (0.438)
ICS 0.0061 0.0004 0.0130 0.0089 0.0131 0.0178
(0.084) (0.480) (0.028) (0.245) (0.032) (0.021)
Productivity shock 0.2441 0.0301 0.1968 0.0199 0.2318 0.0341
(0.346) (0.027) (0.481) (0.043) (0.409) (0.015)
Season dummy 0.1489 20.0013 0.1006 20.0057 0.1210 2 0.0042
(0.013) (0.715) (0.020) (,0.001) (,0.001) (, 0.001)
(continued)
issue markets
three-stage least squares

Hot and cold
regression
Robustness test of the
217
Table VI.
3,3
218
IJMF
Table VI.
Data source
IPOs excluded with offer price
Ritter’s data below $5 IPOs issued after 1980
Initial return IPO volume Initial return IPO volume Initial return IPO volume
HT ratio 0.2048 0.0493

(0.009) (0.725)
VB ratio 20.0928 20.2790
(0.307) (0.045)
Volume lag 1 2 24.3712 0.7298 213.1051 0.8361 211.4003 0.7274
(0.032) (0.017) (0.019) (,0.001) (0.002) (, 0.001)
Volume lag 2 2 5.9810 0.0921 20.1015 0.2011 0.0769 0.1997
(0.218) (0.049) (0.881) (0.075) (0.871) (0.077)
Volume lag 3 5.3026 20.4431 2.7481 20.2424 20.7571 2 0.2304
(0.272) (0.260) (0.269) (0.029) (0.785) (0.054)
Volume lag 4 2 0.3451 0.0960 22.9803 0.0133 21.1671 2 0.0685
(0.841) (0.701) (0.148) (0.850) (0.562) (0.424)
IR lag 1 0.3768 0.0716 0.1942 0.0180 20.0140 0.0090
(0.012) (0.199) (0.094) (0.336) (0.762) (0.423)
IR lag 2 2 0.1581 20.0130 20.2834 0.0021 20.2656 0.0015
(0.202) (0.564) (0.016) (0.790) (0.028) (0.673)
IR lag 3 0.3403 0.0398 0.0889 20.0094 0.1200 2 0.0049
(0.014) (0.305) (0.501) (0.532) (0.292) (0.552)
IR lag 4 0.0843 20.0156 0.0869 20.0131 20.0415 2 0.0116
(0.493) (0.267) (0.416) (0.142) (0.629) (0.102)
System-weighted R 2 0.63 0.84 0.79

Notes: We perform robustness tests using three different data sources and time periods (see section 4.3). For each dataset, we perform a three-stage least
squares regression. We report coefficient estimates with p-values in parentheses below. Note that for Ritter’s dataset, we could not construct variables
related to proceeds, post-IPO standard deviation, the high-tech ratio and the venture capital backed ratio. Thus, we perform the regression for Ritter’s
dataset without these variables
issue market during that period is driven exclusively by natural resource issues and Hot and cold
does not carry over to other industries. Hoffmann-Burchardi (2001) argued that the IPO issue markets
clustering and high under-pricing phenomena in a hot issue market result from the
same underlying fact, i.e. a positive surprise about industry prospects. Helwege and
Liang (2004) find that both hot and cold market IPOs are largely concentrated in the
same narrow set of industries and that hot markets for many industries occur at the
same time. 219
We partition our aggregate data set by industry in order to capture patterns in new
issue activity on an industrial level. Following Helwege and Liang (2004), we categorize
new issues into 11 major industries based on their truncated two-digit SIC code. These
are the top 11 industries in terms of IPO activity, and account for about 65 percent of
total IPO volume. Table VII provides summary statistics for our IPO sample,
categorized by industry. The business services sector (SIC 73), which includes software
and internet companies, has both the highest average initial return and the most IPOs
during our 30-year sample period. On average, there are about 11.53 new issues per
quarter in the business services industry, which is more than twice the volume of any
other industry. In total, the business services industry alone accounts for more than 16
percent of all offerings during the last 30 years.
5.1 Cross-industry correlation of issue activity

Following Ivanov and Lewis (2003), we first test the technological innovations
hypothesis by examining the correlation structure of issue activity by industry. If the
hypothesis holds, the correlation coefficients for issue activity between different
industries will be different. Only industries that share similar underlying technologies
should have positive and relatively high cross-industry correlations. On the other hand,
IPO volume in less related industries is not expected to be as highly correlated.
Table VIII reports the correlation matrix of the number of new issues across
industries. The correlation coefficients are all positive but of different magnitude.
While some of the correlation coefficients are close to one, others are close to zero. We
observe particularly high correlations between “chemicals and allied products” (SIC 28)
and the “engineering and scientific instruments” sector (SIC 38), between “chemicals
and allied products” (SIC 28) and “health services” (SIC 80), between “communications”
(SIC 48) and the “business services” sector (SIC 73), and between “retailing” (SIC 52-59)
and “health services” (SIC 80). For each of these pairs, the correlation coefficient is
above 0.7. Under the technological innovations hypothesis the high correlation
coefficients between those industries would suggest that they benefit from the same
technological innovations. Although some of the correlated industry pairs make
intuitive sense, others are somewhat dubious. In particular, there is no apparent link
between retailing and health services, yet these industries are highly correlated.
On the other hand, we observe low correlations between what are arguably highly
related industries, such as between the “industrial and commercial machinery and
computer equipment” sector (SIC 35) and the “communications” sector (SIC 48),
between “industrial and commercial machinery and computer equipment” (SIC 35) and
“business services” (SIC 73), and between “wholesale trade” (SIC 50-51) and “business
services” (SIC 73).
3,3
220
IJMF
Table VII.
Descriptive statistics of
issue activity by industry
Number of IPOs
Initial return
Industry Two-digit SIC code Total Mean Median SD (percent)
Oil and gas 13, 29 153 1.32 1 1.97 7.44

Chemicals and allied products 28 440 3.79 2 4.46 12.3
Industrial and commercial machinery and computer
equipment 35 492 4.24 3 3.78 17.22
Electronic, electrical equipment and components 36 602 5.19 4 5.04 26.44
Engineering and scientific instruments 38 503 4.34 3 4.78 17.18
Communications 48 295 2.54 1 3.56 19.34
Wholesale trade 50-51 338 2.91 2 3.11 15.03
Retailing 52-59 634 5.47 3 5.77 14.14
Business services 73 1,337 11.53 6 14.90 35.19
Health services 80 293 2.52 1 3.19 10.35
Engineering, accounting, research, management and
related services 87 280 2.41 2 2.67 15.19
Other 2,793 24.09 16 24.49 10.32
Notes: We provide summary statistics for our IPO sample, categorized by industry. We follow the industry classification by Helwege and Liang (2004)
and group firms into 11 major industries based on their two-digit SIC code. In the third column, we report the total number of IPOs in each industry. In
columns four to six we calculate the mean, median and standard deviation for the number of IPOs per quarter. In the last column, we calculate the equally
weighted mean under-pricing for all firms in each sector
SIC 13, 29 SIC 28 SIC 35 SIC 36 SIC 38 SIC 48 SIC 50-51 SIC 52-59 SIC 73 SIC 80 SIC 87
SIC 13, 29 Correlation 1

p-value
SIC 28 Correlation 0.0967 1
p-value (0.302)
SIC 35 Correlation 0.3834 0.5738 1
p value (, 0.001) (,0.001)
SIC 36 Correlation 0.1790 0.6415 0.6397 1
p-value (0.055) (,0.001) (,0.001)
SIC 38 Correlation 0.2807 0.7800 0.5915 0.6047 1
p-value (0.002) (,0.001) (,0.001) (, 0.001)
SIC 48 Correlation 0.0890 0.4173 0.3609 0.6147 0.3834 1
p-value (0.342) (,0.001) (,0.001) (, 0.001) (, 0.001)
SIC 50-51 Correlation 0.1902 0.5269 0.6827 0.5253 0.5395 0.4207 1
p-value (0.041) (,0.001) (,0.001) (, 0.001) (, 0.001) (,0.001)
SIC 52-59 Correlation 0.1220 0.6368 0.6839 0.6408 0.6437 0.4769 0.7167 1
p-value (0.192) (,0.001) (,0.001) (, 0.001) (, 0.001) (,0.001) (,0.001)
SIC 73 Correlation 0.0532 0.4495 0.3607 0.5715 0.4279 0.8229 0.3858 0.4298 1
p-value (0.571) (,0.001) (,0.001) (, 0.001) (, 0.001) (,0.001) (,0.001) (,0.001)
SIC 80 Correlation 0.0727 0.7423 0.5751 0.5241 0.6474 0.2465 0.5767 0.7193 0.2870 1
p-value (0.438) (,0.001) (,0.001) (, 0.001) (, 0.001) (0.008) (,0.001) (,0.001) (0.002)
SIC 87 Correlation 0.2433 0.6605 0.5411 0.6362 0.5723 0.5459 0.4731 0.5724 0.4919 0.4790
p-value (0.009) (,0.001) (,0.001) (, 0.001) (, 0.001) (,0.001) (,0.001) (,0.001) (,0.001) (, 0.001)
Notes: We present a correlation matrix for quarterly IPO volume by industry. For each pair of industries, we report the correlation coefficient and
p-value. Our sample consists of 120 quarters of IPO volume aggregated by industry
industry
issue markets
volume classified by
Hot and cold
Correlation of IPO
Table VIII.
221
IJMF Overall, an analysis of the cross-industry correlation matrix appears to provide at best
3,3 mixed results. While related industries tend to have high correlation coefficients, we
also observe high correlation coefficients for seemingly unrelated industries. On the
other hand, industries that are intuitively more related frequently display
comparatively low correlation coefficients. An objective interpretation is further
complicated by the lack of predefined benchmarks. To avoid these shortcomings, we
222 employ an alternative method for testing the technological innovations hypothesis,
which we discuss in the next section.
5.2 Fixed-effects model

To evaluate the technological innovations hypothesis in more detail, we apply a
fixed-effects model to examine cross-sectional influences between the 11 industries.
The fixed-effects model can be written as follows:
X
N 21 X
T 21
Y it ¼ a þ bX it þ gi W it þ dt Z it þ 1it ; ð3Þ
i¼1 t¼1
where Yit is the aggregate IPO volume during quarter t excluding the IPO volume in
industry i; Xit is a matrix of independent variables; Wit is an industry dummy with
W it ¼ 1, i ¼ 1 to N 2 1 and W it ¼ 0 otherwise; Zit is a time period dummy with
Z it ¼ 1, t ¼ 1 to T 2 1; and Z it ¼ 0 otherwise; N is the number of industries; and T is
the number of quarters.
Specifically, Xit includes industrial IPO volume, that is, the number of IPOs in
industry i during a given quarter, as well as industrial initial return, calculated as the
average return of those IPOs. In addition, it comprises the same factors we
hypothesized to have an impact on IPO volume in our simultaneous equations
model[7]. Following Lowry (2003), both the aggregate IPO volume as measured by Yit
and industrial IPO volume are deflated by the total number of public companies in the
previous quarter.
Our goal is to examine whether the number of new issues in each individual
industry, industrial IPO volume, has a significant relationship with the number of new
issues in the rest of the market. If the coefficient on industrial IPO volume is positive
and significant, the technological innovations hypothesis will be rejected. If the
coefficient is insignificant, there will be support for the technological innovations
hypothesis.
Results for the fixed-effects model are provided in Table IX[8]. We find that
industrial IPO volume has a significant positive impact on aggregate IPO volume.
Consistent with Helwege and Liang (2004), this result is unsupportive of the
technological innovations hypothesis. Specifically, our findings provide evidence
against the notion that aggregate hot markets are characterized by clusters of firms
from particular industries for which a technological innovation or positive productivity
shock has occurred and that IPO volume in the remainder of the market is unaffected.
Instead, the significant positive relationship between Yit and industrial IPO volume
suggests that potential IPO volume spikes in one industry are, at least to some extent,
reflected by concurrent volume spikes in other industries.
Variable Parameter estimate p-value
Hot and cold
issue markets
Intercept 2 0.1206 , 0.001
Industrial IPO volume 2.0390 , 0.001
Industrial initial return 2 0.0017 0.375
Productivity shock 0.0421 0.094
Proceeds 0.0108 0.356 223
Prior market return 0.0137 0.018
Log (GDP) 0.0332 0.084
Risk-free rate 0.0126 0.009
ICS 0.0057 0.004
Season dummy 0.0012 , 0.001
F testfixed effects ( p-value) 0.0001
R2 0.46
Notes: We apply a fixed-effects model to examine the cross-sectional influences between 11
industries. Our industry classification follows Helwege and Liang (2004). For details on the
fixed-effects model, see section 5.2. We regress aggregate IPO volume excluding industry i against the
industrial IPO volume and the initial return in industry i, as well as the economy-wide productivity
shock. Other regressors include proceeds, prior market returns, logged GDP, the risk-free rate, the
index of consumer sentiment and a season dummy. For each variable, we report the estimated
parameter and the corresponding p-value. In the last two rows we provide the p-value for an F-test of Table IX.
the null hypothesis that all industry dummies are zero (i.e. that there are no fixed effects) and the R 2 Fixed-effects analysis of
for our regression IPO volume
At the same time, we observe that economy-wide productivity shocks tend to increase
aggregate IPO volume, although the relationship is only marginally significant[9]. All
other coefficients are consistent with our 3SLS analysis. Both prior market returns and
the risk-free rate have a significant positive influence on aggregate IPO volume, which
is supportive of the hypothesis that more firms go public when they face better market
conditions. Consistent with the capital demand hypothesis (Lowry, 2003), GDP is
significantly positively related to IPO volume. A significant and positive relationship is
also found between the index of consumer sentiment and aggregate IPO volume. This
result lends support to the hypothesis that investor sentiment is an important
determinant of IPO volume (Lowry, 2003). Finally, unlike Alti (2006) we find no
support for the notion that higher IPO volume is associated with higher proceeds, i.e.
we find no evidence that firms time their IPOs and issue more equity during periods in
which the cost of equity capital is low.
6. Conclusions
We study the dynamic relationship between IPO volume and initial returns over the
period from 1972 to 2001. By taking the endogenous interrelationship between IPO
volume and initial returns into account, and applying a three-stage least squares
regression model, we find that higher IPO volume in the concurrent quarter leads to
higher initial returns. At the same time, we observe a negative relationship between
initial returns and lagged IPO volume. In other words, the under-pricing of a new issue
is affected not only by the concurrent number of new issues but also by the number of
new issues in prior periods. On the other hand, we find no support for the hypothesis
IJMF that higher initial returns cause increased IPO volume. Our findings shed light on the
3,3 dynamic relationships in the IPO market and help explain the occurrence of hot and
cold issue markets. We find that measures of ex ante uncertainty are positively related
to initial returns. This is supportive of the hypothesis that under-pricing is positively
related to the degree of asymmetric information. We also find a significant positive
relationship between prior market returns and IPO volume, which is consistent with
224 the hypothesis that IPO waves are preceded by favorable market conditions (Pastor
and Veronesi, 2003).
Lowry (2003) argues that firms’ capital demands and investor sentiment are
important determinants of IPO volume. Our findings for the two hypotheses are mixed.
In our 3SLS model, we do not find support for the theories, i.e. there is no support for
Lowry’s hypotheses when considering aggregate level data, particularly after
controlling for changes in economy-wide productivity levels. However, in our

fixed-effects model, which is based on industry level data, we find that GDP and the
index of consumer sentiment are significantly positively related to the aggregate
number of new issues. Hence, we argue that business conditions and investor
sentiment have more importance on issue activity on an industrial level.
In addition to exploring the hot and cold issue market puzzle and the
interrelationship between IPO volume and initial returns, we examine the
technological innovations hypothesis that suggests that hot issue markets are
driven by a small number of industries that have benefited from technological
innovations or positive productivity shocks. We analyze the correlation matrix of the
11 most active IPO industries, and find mixed evidence for the hypothesis that hot and
cold IPO markets are concentrated in specific industries. Furthermore, we employ a
fixed-effects model in our analysis and find that the number of new issues in the 11
most active IPO industries is significantly positively related to aggregate IPO volume
in the rest of the market. This provides empirical evidence against the technological
innovations hypothesis and suggests that aggregate hot issue markets occur across a
broad range of industries at the same time.
Notes
1. We follow the classification by Loughran and Ritter (2002) and Cliff and Denis (2004) into
high-tech and low-tech firms. Hi-tech firms are those firms whose SIC codes match one of the
following: 2833, 2834, 2835, 2836, 3571, 3572, 3575, 3577, 3578, 3661, 3663, 3669, 3674, 3812,
3823, 3825, 3826, 3827, 3829, 3841, 3845, 4812, 4813, 4899, 7370, 7371, 7372, 7373, 7374, 7375,
7377, 7378, 7379.
2. Note that IPO proceeds in our sample are highly skewed and that our mean and median tests
for the proceeds of hi-tech versus non-hi-tech and VC-backed versus non-VC-backed firms
provide conflicting results.
3. Note that our univariate results should be viewed with caution as they do not control for
potential other influences.
4. For robustness, we also perform our analysis by using value-weighted mean under-pricing
during a quarter. Our results are quantitatively and qualitatively robust to either variable
definition.
5. Note that the GDP variable is significant if the productivity shock variable is excluded but
becomes insignificant after its inclusion.
6. As an additional robustness test, we performed all estimations in Table V by including two Hot and cold
time-period dummies, one for the period from 1982 to 1991, and one for the time period from
1992 to 2001, thus dividing our sample into three periods of ten years each. Our results are issue markets
qualitatively and quantitatively little affected if we add these variables.
7. Note that we do not include the high-tech ratio in our model because industry differences are
captured in more detail by the industry dummies Wit.
8. Note that Table IX also provides results for an F-test of the null hypothesis that all industry 225
dummies Wit are zero, i.e. that there are no fixed effects (see Wooldridge, 2002). Given that
the p-value is 0.0001, the null hypothesis is clearly rejected, making a fixed-effects model
indeed appropriate.
9. To examine whether industry-specific productivity shocks affect IPO volume in the
respective industry sectors, we also performed a robustness test in which we regressed
industry-specific IPO volume on prior-year productivity shocks in that industry and a series
of control variables. The robustness test was performed for a subsample of IPOs after 1987
because industry-specific productivity data is not available through the Bureau of Labor
Statistics (BLS) prior to 1987. Our results (not reported here for brevity) suggest no
significant relationship between industry-specific IPO volume and industry-specific
productivity shocks, after controlling for other factors, which may also be seen as
evidence against the technological innovations hypothesis.
References
Aggarwal, R., Krigman, L. and Womack, K. (2002), “Strategic IPO underpricing, information
momentum, and lockup expiration selling”, Journal of Financial Economics, Vol. 66 No. 1,
pp. 105-37.
Alti, A. (2006), “How persistent is the impact of market timing on capital structure?”, Journal of
Finance, Vol. 61 No. 4, pp. 1681-710.
Bayless, M. and Chaplinsky, S. (1996), “Is there a window of opportunity for seasoned equity
issuance?”, Journal of Finance, Vol. 51 No. 1, pp. 253-78.
Beatty, R. and Ritter, J. (1986), “Investment banking, reputation and the underpricing of initial
public offerings”, Journal of Financial Economics, Vol. 15 No. 1, pp. 213-32.
Benveniste, L., Busaba, W. and Wilhelm, W. (2002), “Information externalities and the role of
underwriters in primary equity markets”, Journal of Financial Intermediation, Vol. 11
No. 1, pp. 61-86.
Benveniste, L., Ljungqvist, A., Wilhelm, W. and Yu, X. (2003), “Evidence of information
spillovers in the production of investment banking services”, Journal of Finance, Vol. 58
No. 2, pp. 577-608.
Booth, J. and Chua, L. (1996), “Ownership dispersion, costly information, and IPO underpricing”,
Journal of Financial Economics, Vol. 41 No. 2, pp. 291-310.
Charoenrook, A. (2002), “Change in the consumer sentiment index and aggregate stock returns”,
working paper, Vanderbilt University, Nashville, TN.
Chemmanur, T. and Fulghieri, P. (1999), “A theory of the going-public decision”, Review of
Financial Studies, Vol. 12 No. 2, pp. 249-79.
Choe, H., Masulis, R. and Nanda, V. (1993), “Common stock offerings across the business cycle”,
Journal of Empirical Finance, Vol. 1 No. 1, pp. 3-33.
Cliff, M. and Denis, D. (2004), “Do IPO firms purchase analyst coverage with underpricing?”,
Journal of Finance, Vol. 59 No. 6, pp. 2871-901.
IJMF Fama, E. and French, K. (1989), “Business conditions and expected returns on stocks and bonds”,
Journal of Financial Economics, Vol. 25 No. 1, pp. 1089-108.
3,3
Habib, M. and Ljungqvist, A. (1998), “Underpricing and IPO proceeds: a note”, Economic Letters,
Vol. 61 No. 3, pp. 381-3.
Helwege, J. and Liang, N. (2004), “Initial public offerings in hot and cold markets”, Journal of
Financial and Quantitative Analysis, Vol. 39 No. 3, pp. 541-69.
226 Hoffmann-Burchardi, U. (2001), “Clustering of initial public offerings, information revelation and
underpricing”, European Economic Review, Vol. 45 No. 2, pp. 353-83.
Ibbotson, R. and Jaffe, J. (1975), “‘Hot issue’ markets”, Journal of Finance, Vol. 30 No. 4,
pp. 1027-42.
Ibbotson, R., Sindelar, J. and Ritter, J. (1994), “The market’s problems with the pricing of initial
public offerings”, Journal of Applied Corporate Finance, Vol. 7 No. 1, pp. 66-74.
Ivanov, V. and Lewis, C. (2003), “The determinants of issue cycles for initial public offerings”,
working paper, Vanderbilt University, Nashville, TN.
Jindra, J. (2001), “Seasoned equity offerings, overvaluation, and timing”, working paper,
Cornerstone Research, Boston, MA.
Lee, C., Shleifer, A. and Thaler, R. (1991), “Investor sentiment and the closed-end fund puzzle”,
Lee, P., Taylor, S. and Walter, T. (1994), “The voluntary disclosure of forecast data by Australian
IPOs”, working paper, University of Sydney, Sydney.
Lerner, J. (1994), “Venture capitalists and the decision to go public”, Journal of Financial
Economics, Vol. 35 No. 3, pp. 293-316.
Lerner, J., Shane, H. and Tsai, A. (2003), “Do equity financing cycles matter? Evidence from
biotechnology alliances”, Journal of Financial Economics, Vol. 67 No. 1, pp. 411-46.
Ljungqvist, A. (1995), “When do firms go public? Poisson evidence from Germany”, working
paper, Oxford University, Oxford.
Ljungqvist, A. (1997), “The pricing of initial public offerings: further evidence from Germany”,
European Economic Review, Vol. 41 No. 7, pp. 1309-20.
Loughran, T. and Ritter, J. (1995), “The new issues puzzle”, Journal of Finance, Vol. 50 No. 1,
pp. 23-51.
Loughran, T. and Ritter, J. (2002), “Why don’t issuers get upset about leaving money on the table
in IPOs?”, Review of Financial Studies, Vol. 15 No. 2, pp. 413-43.
Lowry, M. (2003), “Why does IPO volume fluctuate so much?”, Journal of Financial Economics,
Vol. 67 No. 1, pp. 3-40.
Lowry, M. and Schwert, G. (2002), “IPO market cycles: bubbles or sequential learning?”, Journal
of Finance, Vol. 57 No. 3, pp. 1171-201.
Lowry, M. and Shu, S. (2002), “Litigation risk and IPO underpricing”, Journal of Financial
Economics, Vol. 65 No. 3, pp. 309-36.
Lucas, D. and McDonald, R. (1990), “Equity issues and stock price dynamics”, Journal of Finance,
Vol. 45 No. 4, pp. 1019-43.
McGuinness, P. (1992), “An examination of the underpricing of initial public offerings in Hong
Kong: 1980-90”, Journal of Business Finance & Accounting, Vol. 19 No. 2, pp. 165-85.
Maksimovic, V. and Pichler, P. (2001), “Technological innovation and initial public offerings”,
Review of Financial Studies, Vol. 14 No. 2, pp. 459-94.
Megginson, W. and Weiss, K. (1991), “Venture capitalist certification in initial public offerings”, Hot and cold
issue markets
Pagano, M., Panetta, F. and Zingales, L. (1998), “Why do companies go public? An empirical
analysis”, Journal of Finance, Vol. 53 No. 1, pp. 27-64.
Pastor, L. and Veronesi, P. (2003), “Stock prices and IPO waves”, working paper, University of
Chicago, Chicago, IL.
Rajan, R. and Servaes, H. (1997), “Analyst following of initial public offerings”, Journal of 227
Finance, Vol. 52 No. 2, pp. 507-29.
Ritter, J. (1984), “The ‘hot issue’ market of 1980”, Journal of Business, Vol. 57 No. 2, pp. 215-40.
Ritter, J. and Welch, I. (2002), “A review of IPO activity, pricing, and allocations”, Journal of
Finance, Vol. 47 No. 4, pp. 1795-828.
Rock, K. (1986), “Why new issues are underpriced”, Journal of Financial Economics, Vol. 15 No. 1,
pp. 187-212.
Stoughton, N., Wang, K. and Zechner, J. (2001), “IPOs and product quality”, Journal of Business,
Vol. 74 No. 3, pp. 375-408.
Wooldridge, J. (2002), Econometric Analysis of Cross Section and Panel Data, MIT Press,
Cambridge, MA.
Further reading
Lee, P., Taylor, S. and Walter, T. (1999), “IPO underpricing explanations: implications from
investor application and allocation schedules”, Journal of Financial and Quantitative
Analysis, Vol. 34 No. 4, pp. 425-44.
Welch, I. (1989), “Seasoned offerings, imitation costs, and the underpricing of initial public
offerings”, Journal of Finance, Vol. 44 No. 2, pp. 421-49.
IJMF Appendix
3,3
Variable Data sources Description
Under-pricing SDC, CRSP Percentage return from offer price (SDC) to first-day
closing price (CRSP)
Initial return SDC Average IPO under-pricing of firms that went public
228 during a quarter
Number of IPOs SDC Number of firms that went public during a quarter
IPO volume SDC, CRSP Number of firms that went public during a quarter,
deflated by the total number of public companies in
the previous quarter
Proceeds SDC, BLS Offer proceeds converted to year 2000 dollars based
on CPI data from the Bureau of Labor Statistics
(BLS), averaged across all firms that went public in a
given quarter
Tech dummy SDC Tech dummy ( ¼ 1 if the firm is a technology firm, 0
otherwise). We follow Loughran and Ritter (2002)
and Cliff and Denis (2004) and categorize firms with
the following SIC codes as tech firms: 2833, 2834,
2835, 2836, 3571, 3572, 3575, 3577, 3578, 3661, 3663,
3669, 3674, 3812, 3823, 3825, 3826, 3827, 3829, 3841,
3845, 4812, 4813, 4899, 7370, 7371, 7372, 7373, 7374,
7375, 7377, 7378, 7379
High-tech ratio (HT ratio) SDC Ratio of the number of high-tech IPOs divided by the
total number of IPOs during a quarter
Prior market return CRSP Annualized return on the CRSP equally weighted
market index during the three months prior to the
firm’s IPO
Prior market volatility CRSP Annualized standard deviation of returns on the
CRSP equally weighted market index during the
three months prior to the firm’s IPO
Risk-free rate Federal Reserve web site Annualized average nominal yield on a three-month
(www.federalreserve.gov) T-Bill during a quarter
Log (GDP) US Department of Log of the quarterly US gross domestic product
Commerce, Bureau of converted to year 2000 dollars based on CPI data
Economic Analysis from the Bureau of Labor Statistics (BLS)
(www.bea.doc.gov), BLS
Venture capital backed SDC Ratio of the number of venture capital backed IPOs
ratio (VB ratio) to the total number of IPOs during a quarter
Post-IPO standard CRSP Standard deviation of returns during the first 30
deviation trading days after an IPO, averaged quarterly
Index of consumer University of Michigan Consumer sentiment index as constructed by the
sentiment (ICS) Survey Research Center of the University of
Michigan. The index serves as a proxy for consumer
sentiment during a given quarter
Productivity shock BLS Percentage change in non-farm business output per
hour during the prior four quarters based on
productivity data from the Bureau of Labor
Statistics (BLS)
Table AI. Season dummy Dummy variable ( ¼ 1 if the IPO was issued in the
Variable definitions first quarter, 0 otherwise)
Corresponding author
Thomas J. Walker can be contacted at: twalker@ jmsb.concordia.ca
To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints
This article has been cited by:
1. Arjan Premti, Jeff Madura. 2013. Motives and Consequences of IPOs in Cold Periods. The Quarterly
Review of Economics and Finance . [CrossRef]
2. James AngDepartment of Finance, College of Business, Florida State University, Tallahassee, Florida,
USA Carol BoyerDepartment of Finance, College of Management, Long Island University CW Post,
Brookville, New York, USA. 2009. Performance differences between IPOs in new industries and IPOs in
established industries. Managerial Finance 35:7, 606-623. [Abstract] [Full Text] [PDF]

Video 3

Uploaded by

Copyright:

Available Formats

Video 3

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Video 3

Uploaded by

Copyright:

Available Formats

International Journal of Managerial Finance

Downloaded on: 04 March 2017, At: 02:44 (PT)

*Related content and download information correct at time of download.

During periods of high information asymmetry, it is less attractive for issuers to

2.2 Technological innovations

the IPO market without the influence of penny stocks.

1983 587 12.18 19,201.8

n 2,683 5,477 2,661 5,499

Univariate statistics for

Number of IPOs 120 68.00 51.50 61.18 0 238

4. Simultaneous equations approach

IPO volume ¼ g2 Initial return þ a2 X þ b2 X 2 þ 12 ; ð2Þ

variables. In detail, X is a vector of IPO characteristics that are common to both

4.2 Empirical results 211

Panel A. Simultaneous regression results

Initial return IPO volume

Initial return 1.1697 0.4455

correlated. Therefore, a 3SLS estimation approach is more appropriate than a 2SLS

4.3 Robustness tests

IPO volume 27.6875 15.8741 12.8203

three-stage least squares

HT ratio 0.2048 0.0493

System-weighted R 2 0.63 0.84 0.79

5.1 Cross-industry correlation of issue activity

Oil and gas 13, 29 153 1.32 1 1.97 7.44

SIC 13, 29 Correlation 1

5.2 Fixed-effects model

The fixed-effects model can be written as follows:

controlling for changes in economy-wide productivity levels. However, in our

To purchase reprints of this article please e-mail: [email protected]

You might also like