Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Page 1
Earnings Quality and International IPO Underpricing
Thomas J. Boultona,*
, Scott B. Smartb, Chad J. Zutter
c
aFarmer School of Business, Miami University, Oxford, OH 45056, USA
bKelley School of Business, Indiana University, Bloomington, IN 47405, USA
cKatz Graduate School of Business, University of Pittsburgh, Pittsburgh, PA 15260, USA
Draft: July 2009
Abstract
Examining 7,306 IPOs from 34 countries, we find that IPOs are underpriced less in countries
where existing public firms produce higher quality earnings information. This finding persists
after controlling for other deal- and country-specific factors that affect IPO underpricing, and it is
driven neither by the large and relatively transparent markets in the U.S. and U.K. nor by the
relatively opaque Japanese market. The impact of going public in a country with relatively low
earnings quality is partially offset by the use of a high-quality underwriter.
JEL classification: G15; G24; G30; G32; G34
Keywords: international finance; earnings quality; initial public offerings; underpricing.
Data Availability: Data used in this study are available from public sources.
For valuable comments, the authors express their thanks to Oya Altinkiliç, Utpal Bhattacharya,
Robin Grieves, Robert Jennings, Kenneth Lehn, Gershon Mandelker, Jay Ritter, Shawn Thomas, Greg
Udell, Xiaoyun Yu, and seminar participants at Central Michigan University, Duquesne University,
Financial Management Association meetings (Orlando), Louisiana Tech University, Marquette University,
Miami University, Oklahoma State University, Seton Hall University, Texas A&M University, University
of Nebraska, University of Oklahoma, University of Otago, University of Pittsburgh, West Virginia
University, and Wright State University. Any remaining errors or omissions remain the responsibility of the
authors.
* Corresponding author. Tel.: + 1-513-529-1563; fax: + 1-513-529-8598.
E-mail address: [email protected] (T. Boulton).
2
I. Introduction
Few corporate events garner more attention from researchers, practitioners, the media, and
the public than initial public offerings (IPOs). Generally, the focus is on the large, sometimes
spectacular, first-day gains to new issues observed not only in the U.S., but in virtually all of the
s stock markets. Summarizing the findings of dozens of papers, the majority of which
focus on underpricing in a single country, Loughran et al. (1994) confirm that IPOs earn positive
first-day returns everywhere, and that underpricing varies dramatically across countries. What
drives underpricing differences across markets is still a largely unexplored question.
Theoretical explanations of the underpricing phenomenon often propose that underpricing
arises from information asymmetries between participants in the IPO process. To cite just one
prominent example, Rock (1986) shows that when investors have different information sets,
underpricing is necessary to induce less informed investors to bid for IPO shares in equilibrium.1
In a recent survey of the empirical evidence, Ljungqvist (2007) concludes that information
asymmetries -
the cross-country underpricing variation documented in the literature may be driven by
differences in the quality of information available to investors in different markets. Specifically,
we are interested in whether international differences in the quality of accounting information
help explain the international cross-section of IPO initial returns.
1 Other information-based underpricing theories endow different participants in the IPO process with
superior information. For example, in Baron (1982) investment bankers know more about an IPO firm s
true value than the issuer does, and underpricing becomes a solution to the resulting principal-agent
problem. In Allen and Faulhaber (1989), Grinblatt and Hwang (1989), and Welch (1989), issuers use their
information advantage to signal firm quality, whereas in Benveniste and Spindt (1989), underpricing
induces well-informed investors to reveal what they know before the offer price is set.
3
The accounting and finance literatures confirm that earnings quality influences a range of
capital market outcomes. For example, Bhattacharya et al. (2003) find that in countries with
greater earnings opacity, firms face a higher cost of capital. Biddle and Hilary (2006) establish
that better accounting information leads to more efficient firm-level investment decisions,
particularly in countries where financing comes through arms-length transactions rather than
through relationships with creditors. Gelos and Wei (2005) find that emerging markets mutual
fund managers invest more heavily in countries with greater transparency. Jin and Myers (2006)
argue that greater opacity leads to lower firm-specific risk borne by investors and higher R-square
values.
These findings, combined with the central role that information asymmetries play in IPO
underpricing, give rise to our earnings quality hypothesis do firms endure higher underpricing
when they go public in countries where the quality of earnings information is generally low? We
approach this question in an international setting for several reasons. First, as noted above,
underpricing varies dramatically across countries, and relatively little work has been done to
understand why that is the case. Second, within a single country, firms operate in the same legal
and regulatory environment, and that environment likely imposes some limit on the variability of
earnings quality across firms. Thus, it is plausible that the quality of accounting information
varies more across markets than within a single market. Third, in most cases very little
accounting information is available about a particular firm prior to its IPO. An IPO prospectus
typically offers little more than two years of pre-IPO financial information. This
ability to capture cross-sectional earnings quality differences as of the IPO date. In contrast, the
country-level earnings quality measures that we use draw upon several years of data generated by
many different firms. While these measures do not provide any information on the quality of
earnings from one IPO to another in a given country, they do capture differences in the
information environment in which firms from different countries go public. Fourth, we are
interested in whether reputable financial intermediaries, such as top-tier investment banks, play a
4
role in mitigating the information asymmetries related to earnings quality. Given the global scope
of the investment banking industry, and given the difficulties in identifying cross-sectional
earnings quality variance in a single country, the question of whether investment banks mitigate
earnings-related information asymmetries seems ideally suited to an international sample.
Despite the advantages of studying underpricing using an international sample, we must
mention one caveat associated with our research design. Underpricing differences across
countries may be influenced by omitted variables, such as differences in offering methods, which
could be correlated with our earnings quality measures. Naturally we have taken many steps to
minimize this possibility, for example by controlling for numerous country and deal
characteristics; including offering method. In addition, we appeal to evidence that IPO pricing
methods around the world are converging over time, with the U.S. bookbuilding approach
dominating in most countries. For instance, Ljungqvist et al. (2003) find that in a sample of 2,143
IPOs from 65 countries, 46.2 percent were priced using the bookbuilding method in 1994, but by
the end of their sample period (the first seven months of 1999), the fraction of new issues priced
via bookbuilding rose to 80 percent. Similarly, Jagannathan and Sherman (2006) examine the use
of auctions and bookbuilding in 46 countries and find that in all but four countries, auctions have
been abandoned entirely, and auctions are rare in the few countries that still use them. In contrast,
they find that the bookbuilding method has been gaining market share over time and has become
the dominant pricing method in most countries. So while we cannot rule out the possibility of an
omitted variables problem in our analysis, the timeliness of our sample (2000-2006) combined
with our regression control variables and robustness tests reduce this problem to the extent
possible.
Examining 7,306 IPO events, we find evidence of a statistically and economically significant
association between country-level earnings quality and IPO-firm-level underpricing. Comparing
two countries that differ in earnings quality by one standard deviation, firms going public in the
country with better earnings information experience 4.8 percentage points lower underpricing
5
(relative to a sample mean initial return of 27.5 percent). Our regression models control for many
deal-specific (e.g., offer size, industry, underwriter, and underwriting method) and country-
specific (e.g., market returns, liquidity, and IPO activity) variables that influence underpricing,
and our results are robust to the adoption of various minimum offer price screens. Furthermore,
the link between earnings quality and underpricing is driven neither by the large and relatively
transparent U.S. and U.K. markets nor by the relatively opaque market in Japan.
We also explore the influence of underwriters on underpricing and its interaction with
earnings quality. We develop a measure of underwriter quality and interact it with our earnings
quality measures. We corroborate findings from other recent studies, which suggest that since the
early 1990s IPOs underwritten by higher quality investment banks experience higher
underpricing.2 However, better underwriters mitigate the effect of poor earnings quality on
underpricing, such that firms listing in countries with relatively low earnings quality experience a
net reduction in underpricing if they are backed by higher quality underwriters.
The remainder of this paper is structured as follows. Section 2 highlights previous research on
earnings quality and IPO underpricing related to this study. Section 3 describes our sample
construction and descriptive statistics. Section 4 contains our primary results on the relations
between IPO underpricing and the quality of earnings and illustrates that many of the
conventional variables used in single-country studies of IPO underpricing also explain the
international cross-section. Section 5 summarizes and concludes.
II. Earnings quality and IPOs
A number of researchers have studied the link between accounting information disclosed in
the IPO prospectus and the market value of going-public firms. Essentially this strain of the
literature asks whether IPO firms manipulate their financial statements to obtain a higher share
price. Early papers offer modest affirmation for this hypothesis. In a study of Canadian IPOs,
2 See for example Aggarwal et al. (2002), Smart and Zutter (2003), Loughran and Ritter (2004).
6
Clarkson et al. (1992) find that some firms voluntarily disclose earnings projections in their
prospectuses, while other firms do not. Firms that choose to disclose generally report positive
earnings projections, and these projections do affect the IPO offer price. However, they also find
that the stock market is able to adjust for bias in these forecasts. Aharony et al. (1993) find weak
evidence that U.S. firms attempt to inflate earnings prior to going public. Only among very small
firms and firms with very high leverage do they find evidence of earnings manipulation around
the IPO. They also find weak evidence that high-quality underwriters and auditors limit earnings
management behavior. Friedlan (1994) reports that U.S. firms make income-increasing
discretionary accruals in interim financial statements published in the IPO prospectus, but not in
the final statements released after the IPO.
Even stronger evidence suggesting that IPO issuers manipulate the accounting numbers in
their prospectuses appears in subsequent papers. Teoh et al. (1998) provide evidence that IPO
firms have high issue-year earnings and abnormal accruals, followed by poor long-run earnings
and negative abnormal accruals. They indicate that abnormal accruals at the IPO help explain
subsequent poor stock returns. Similarly, Teoh et al. (1998) find that IPO firms that are the most
aggressive in using accruals to report cash flows in excess of earnings earn 20 percent lower stock
returns in the three years after the IPO compared to the firms reporting the most conservative
earnings figures. Firms using accruals aggressively also issue about 20 percent fewer seasoned
equity offerings. This finding leads the authors to conclude that investors naively focus on
artificially high earnings reported by some firms when they go public.3 Teoh and Wong (2002)
suggest that analysts fare no better than investors at disentangling earnings manipulation at the
3 One caveat with this study is that it uses accounting data produced after the IPO rather than before to
measure earnings manipulation. The reason provided by the authors is that pre-IPO data are not widely
available. Our study takes advantage of data from existing public companies to characterize earnings
quality in each country prior to any given IPO in that country.
7
IPO. They find that accounting accruals predi
non-issuing firms, and the predictive power is greater for discretionary rather than
nondiscretionary accruals. For IPO firms, the relation between accruals and analyst forecast errors
is independent of underwriter affiliation, and predicted forecast errors are strongly tied to long-
run underperformance of IPO shares. Because accruals help predict analyst forecast errors for
firms that do not issue new equity, the authors conclude that analysts contribute to market
inefficiency, broadly speaking, because analysts are not sufficiently skeptical of earnings figures
influenced by discretionary accruals.4 Fan (2007) offers mixed support for Teoh et al. and Teoh
and Wong by reporting that discretionary accruals are indeed highest in the IPO year, consistent
with the notion that IPO issuers manage earnings. However, Fan finds no evidence of a negative
relation between IPO earnings management and subsequent stock returns. She concludes that
market participants are not fooled by the earnings management activities of IPO firms and
challenges the notion that IPO earnings management is opportunistic behavior on the part of
issuing firms.
Venkataraman et al. (2008) overturn this emerging conventional wisdom with their finding
that pre-IPO audited accruals are not opportunistically high. They examine pre-IPO financial
statements and find that pre-IPO accruals tend to be negative and less than post-IPO accruals.
They conclude that higher litigation exposure related to IPOs results in higher audit fees and more
conservative audits. In the same vein, Ball and Shivakumar (2008) find
practices become more conservative when they transition from private to public status through an
IPO. They examine a sample of U.K. IPOs that initially prepared financial statements as private
companies, but then later restated these statements as part of the process of going public. Because
Ball and Shivakumar rely on statements that reflect different reporting choices for the same firms
4 Two other studies that find evidence of pre-IPO earnings manipulation are Aharony et al. (2000) and
DuCharme et al. (2001).
8
in the same years, their research design controls for firm-specific factors such as growth and
profitability that may be correlated with the decision to go public. At the same time, their research
design exploits a change in the regulatory environment in which firms operate.5
Morsfield and Tan (2006) study the relation between venture capital funding and earnings
management. Building on prior studies that suggest that venture capitalists play a monitoring role
in IPO firms (e.g., Gompers and Lerner 2002; Hellman and Puri 2002), Morsfield and Tan predict
a negative relation between venture capital financing and earnings management. Controlling for
the endogeneity of venture capital funding, they find that abnormal accruals are lower for venture
capital-backed IPO firms.
All of these papers focus on the extent to which earnings manipulation leading up to the IPO
between the IPO offer price and the market price established once trading begins), which is our
focus. However, the accounting literature does offer several insights into the influence of
accounting disclosures on IPO underpricing.
In a study of micro-cap Nasdaq IPOs, Willenborg and McKeown (2001) find that firms going
public with going-concern audit opinions are more likely to delist within two years of the IPO,
but these firms also endure less underpricing, consistent with the notion that the audit opinion
reduces information asymmetries between issuers and investors. Jog and McConomy (2003)
examine the impact of voluntary disclosure of management earnings forecasts in the IPO
prospectus on IPO valuation and performance. They find higher underpricing for IPOs that do not
include earnings forecasts, though this difference is concentrated among small firms. Schrand and
Verrecchia (2005) study the relation between underpricing and the frequency of pre-IPO
5 Lewis (2006) finds that earnings and income-increasing accruals at the IPO are more persistent when
the investment bank assisting the firm is more reputable. She also finds that investors capitalize earnings at
higher multiples when high-reputation banks are present.
9
disclosure and find lower underpricing for firms with more frequent disclosures prior to the IPO.
An exception is Internet firms, where this relation is reversed. They also find that more frequent
disclosure ameliorates adverse selection in the aftermarket, with lower bid-ask spreads and
greater depths for firms that disclosure more frequently pre-IPO. Finally, Leone et al. (2007) find
Collectively these papers, all of which focus on IPOs in a
single country, suggest that accounting disclosures influence underpricing.
Our study extends this analysis to a multi-country setting to determine if earnings quality at
the country level influences underpricing costs borne by firms going public in different markets.
In so doing we add to the very limited evidence on the determinants of cross-country
underpricing variation as well as contribute to another strain of the literature on the quality and
value of accounting information in different countries. For example, Defond et al. (2007) find that
earnings announcements are more informative in countries with better overall earnings quality.
Bhattacharya et al. (2003) study the extent to which cross-country variation in the quality of
accounting information influences the cost of capital and trading volume in international equity
markets. Using data from 34 countries covering 1985-1998, they measure three dimensions of
earnings opacity for each country earnings aggressiveness, loss avoidance, and earnings
smoothing. They find robust evidence that an increase in overall earnings opacity leads to an
increase in the required return demanded by shareholders and a decrease in trading activity. In a
similar vein, Leuz et al. (2003) examine earnings management practices in 31 countries and find
that firms engage in greater earnings management in countries with weaker investor protections.6
6 Bushman and Piotroski (2006) construct earnings quality measures based upon accounting
conservatism and link those measures to cross-country variation in legal and political institutions. See also
Hung (2000) and Ball et al. (2003).
10
These studies and others suggest that the quality of earnings information available to outside
investors influences information asymmetries in the market. Given the central role of asymmetric
information in theories of IPO underpricing, we anticipate that differences in earnings quality
across countries will influence the underpricing costs that firms going public in different
countries face. Recognizing that earnings quality is a multi-faceted concept and that no single
measure can capture all aspects of earnings quality, we employ a wide range of earnings quality
measures drawn from the extant literature and study their association with firm-level
underpricing. The measures we use are designed to capture elements of earnings quality such as
earnings smoothing, loss avoidance, earnings aggressiveness, and accounting conservatism,
among others. Below, we construct and use the earnings quality measures in Leuz et al. (2003),
Bhattacharya et al. (2003), Bushman and Piotroski (2006), and La Porta et al. (1998) to test our
earnings quality hypothesis.
III. Sample construction and descriptive statistics
Sample construction
We begin our sample construction by retrieving IPOs for all countries reported in the
Thomson Financial SDC Platinum New Issues database from 2000 through 2006. We exclude
countries with fewer than five IPOs during our sample period. Next we attempt to match each
IPO firm with Datastream to collect secondary market prices and calculate initial returns. We
identify 7,455 IPOs with valid prices and positive trading volume.7 We calculate the initial return
as the percentage price change from the offer price to the secondary market closing price (in local
7 A valid price is the first price observation in Datastream with positive trading volume that occurs
within -3 to +60 days of the SDC IPO issue date. In other words, if Datastream is reporting prices prior to -
3 or does not begin reporting prices until after +60 then the event is discarded. We require positive trading
volume to ensure that the Datastream price that we extract is a market price and not simply the IPO offer
price recorded by Datastream. Moreover, for France and Taiwan we use the tenth valid price, rather than
the first, because in these countries initial IPO prices are constrained by daily volatility limits.
11
currency). To eliminate the impact of outliers, which likely result from difficulties in matching
SDC information with valid Datastream prices, we trim our sample by removing the top and
bottom one percent based on initial return. These steps result in a final sample of 7,306 IPOs
listed in 34 countries.
Although common in the IPO literature, we do not impose a minimum offer price restriction.
For example, Ritter (1991) evaluates U.S. IPOs with a minimum offer price of $1 to mitigate the
bid-ask bounce effect. However, imposing this filter would not only greatly reduce the number of
IPOs in many countries, but it would also eliminate some countries entirely. Applying a $1
minimum offer price (converting local currency to U.S. dollars based on the exchange rate as of
the IPO date) eliminates over one-third of the sample events. Thus, the main analysis presented
here imposes no minimum offer price, but we do verify that our results are not driven by the
inclusion of IPOs with very low offer prices.8
Descriptive statistics
Figure 1 displays the average IPO underpricing and number of IPOs by year. The figure
clearly illustrates the slowdown in equity offerings following the decline in equity markets in
2000. IPO volume falls more than 50 percent from 2000 to 2001, then remains relatively flat for
2002 and 2003. The number of new issues rebounds sharply to 1,278 deals in 2004, and then falls
1984), average underpricing is positively correlated with yearly IPO volume (correlation = 0.58).
[Place Figure 1 about here]
8 Table 7 reports robustness tests using minimum offer price restrictions. In particular, the bottom 2, 5,
10, and 20 percent of offer prices are excluded from respective country-level samples. In terms of the U.S.
distribution of offer prices, these cutoffs correspond to minimums of $5, $6, $7, and $9. The main
conclusions of our analysis are the same when employing any of these restrictions.
12
Coffee (1999, 2002) suggests that firms list abroad to bond themselves to foreign listing
standards. For example, firms listing in the U.S. subject themselves to SEC oversight, agree to
meet generally accepted accounting principles (GAAP), and face the scrutiny of financial
intermediaries involved in the security markets. While most of our IPOs originate and list in the
same country, some companies choose to list outside their home country. Most of these firms list
standards, the country where the firm lists is the relevant location for this study.
Table 1 shows the IPO volume and average underpricing for each country in our sample.
Average underpricing varies widely, ranging from 57.3 percent in Japan to less than 3 percent in
Mexico. However, this variation appears to be largely unrelated to IPO volume. Not surprisingly,
the U.S. has the most IPOs in the sample, followed closely by Japan and the U.K. The aggregate
gross proceeds for the entire sample is about $573 billion of which the U.S. represents about 41
percent.
[Place Table 1 about here]
For each sample country, we calculate several earnings quality measures, described in more
constructed such that a higher score implies lower quality earnings. Figure 2, groups countries
into
average underpricing and average aggregate earnings management score.
[Place Figure 2 about here]
Figure 2 offers preliminary evidence that earnings quality influences underpricing. Average
underpricing is almost 75 percent higher (or over 15 percentage points higher) in the highest
AggEM quartile compared to the lowest quartile. The simple correlation between underpricing
and aggregate earnings management scores in Figure 2 is 95 percent. Of course, many differences
in the IPO process and in the types of firms going public exist across countries, so in the next
13
section we test for a link between earnings quality and underpricing, holding constant deal-level
and other country-level characteristics.
Table 2 provides descriptive statistics for the variables used in our regression analysis,
starting with various earnings quality measures drawn from Leuz et al. (2003), Bhattacharya et al.
(2003), Bushman and Piotroski (2006), and La Porta et al. (1998). For most of these measures, we
provide up-to-date calculations (described below) for each of our sample countries and scale
these measures such that a higher value indicates lower earnings quality.9
The unit of observation in our regressions is the IPO firm, and the dependent variable is the
first-day return. The sample mean initial return equals 27.5 percent, which is roughly double the
long-run mean for U.S. underpricing. The cross-sectional standard deviation of initial returns is
52.3 percent, indicating large variations in IPO returns. Unquestionably some of this variation can
be explained by deal-specific and market-specific factors not related to earnings quality, so we
use the remaining variables listed in Table 2 to control for these effects.
Because the extent to which an IPO is underpriced may be influenced by the quality of the
underwriter, we create an underwriter market share variable similar to that proposed by
Megginson and Weiss (1991). In our analysis, the underwriter reputation measure equals the
fraction of IPO inflation-adjusted proceeds underwritten by a particular underwriter in a particular
country for our sample period. Table 2 indicates that the mean underwriter has a market share of
7.2 percent.10
In some countries, underwriters can engage in price support once trading begins.
9 We do not recalculate the accounting conservatism measure or the accounting standards index,
relying instead upon the values published in Bushman and Piotroski (2006) and La Porta et al. (1998)
respectively.
10 Our results are broadly similar with a global market share measure.
14
market price begins to fall below the IPO offer price. If underwriters engage in price support to
limit the occurrence of negative initial returns, then the first-day returns distribution will have a
higher mean than it would have in the absence or price support. We do not have detailed data on
the regulations and practices with respect to price stabilization in all countries, so we attempt to
control for price stabilization activity in two ways. First, we construct a country-level proxy for
price stabilization activity. If price stabilization is widespread, then we expect to see an unusually
large probability mass in the distribution of first-day returns just to the right of zero and an
unusually small probability mass to the left of zero. Therefore, for each country we calculate the
difference in the number of IPOs with initial returns between zero and one percent and the
number of IPOs with initial returns between zero and negative one percent, and then divide the
difference by the total number of IPOs. The more prevalent is price stabilization in a given
market the higher should be this ratio, and the higher should be average underpricing. Table 2
reports a mean price stabilization ratio of 0.007, indicating a slightly greater incidence of small
positive initial returns than small negative returns. Our second approach to control for price
stabilization exploits the fact that underwriters typically provide price support for a very limited
time, so the effects of stabilization activities on the IPO returns distribution should diminish over
time. We calculate the IPO initial return based on the market price roughly one month (22 trading
days) after the IPO rather than on the first trading day, and we test to see if our regression results
are robust to this change.
As suggested in numerous papers, including Ritter (1984), underpricing tends to be higher
when IPO volume is high and when overall stock market returns are high. We include two
variables to control for hot market effects. First, our IPO activity measure equals the
number of IPOs in a given country in each year divided by the total number of listed equities in
Datastream for that country in 2006. Therefore, this measure takes the same value for all IPOs
from a single country in a particular year, but within a country it varies across time, and within a
single year it varies across countries. Second, for each IPO, we calculate the return on the
15
Datastream market index in the three months leading up to the offer. Consequently, two firms
will share the same market return value only if they go public at the same time in the same
country.
Ellul and Pagano (2006) suggest that IPOs in less liquid markets will exhibit larger initial
returns. Higher underpricing compensates IPO investors for the illiquidity risk that they bear. To
control for differences in liquidity across national markets, we include a country-level stock
market turnover ratio, which is defined by Beck et al. (2000) as the ratio of total value of shares
traded to market capitalization.11 liquidity relative
to its size, and we expect lower initial returns in more liquid markets.
Because the state of economic development varies dramatically across our sample nations, we
include a measure of economic development in our regressions. The variable,
Underdevelopment is the sum of the rankings of infant mortality, Internet users per capita,
literacy, unemployment rate, and paved airport runways as reported by Butler and Fauver (2006).
The underdevelopment index is standardized on a 0 to 1 scale, with higher values indicating less
development. To the extent that IPO investors bear more risk when investing in underdeveloped
markets, we expect higher underpricing in countries with higher development index values.
Another country-level control variable is the Index of Economic Freedom, which is produced
by the Heritage Foundation and The Wall Street Journal. The index is an aggregate measure of
several different measures of economic freedom including: (1) the ease of starting a new
business, (2) the level of trade barriers, (3) the top marginal income tax rate, (4) government
spending as a percentage of GDP, (5) the inflation rate, (6) the ease with which foreigners can
11 Beck et al. (2000) report the ratio of total value of shares traded to market capitalization for the
period 1999-2004. Since our sample period includes 2005 and 2006, we take the average of all years
reported for a country by Beck et al. as our measure of country-level stock market turnover.
16
invest in a given country, (7) the extent of government control of the banking system, (8) legal
protection of property rights, (9) corruption, and (10) the ability of a business to hire and fire
workers without restrictions. We include the freedom index because of the literature spawned by
the work of La Porta et al. (2000) illustrating the importance of various investor protections on a
range of capital market outcomes. Unlike the La Porta et al. measures, the Index of Economic
Freedom is updated annually, so we can match it to our sample years rather than using somewhat
dated measures from older papers.12 We also include as a control variable the updated anti-
director rights index from Djankov et al. (2008) which is based on laws in place as of May 2003.
The anti-director rights index is based on a collection of legal and regulatory variables related to a
Most IPO underpricing studies include measures designed to capture information
asymmetries, including the natural logarithm of the offer size, which we report in Table 2.
Because offering sizes vary widely across countries, we also construct an offer size measure that
captures the size of an IPO relative to other deals in the same country. This offer size ratio equals
the offer size divided by the mean offer size for a given country.13 We obtain broadly similar
results with both absolute and relative deal-size measures, so in Table 2 and in subsequent
regressions we simply report natural log of the CPI-adjusted total offer size for each IPO.
[Place Table 2 about here]
Bradley et al. (2004) report that IPOs with integer prices experience higher underpricing. To
bargain over finer offer price increments as the uncertainty surrounding firm value declines.
12 Moeller and Schlingemann (2005) use this index to proxy for the restrictiveness of the institutional
environment related to corporate acquisitions, and they find that in cross-border acquisitions, bidder returns
are higher when targets come from countries with greater freedom.
13 All offer sizes are measured in millions of CPI-adjusted 2006 U.S. dollars. CPI data is from the U.S.
Bureau of Labor Statistics website. SDC reports offer size in U.S. dollars.
17
Thus, for IPOs that are particularly difficult to value, and hence should have higher underpricing,
offer prices tend to fall on integers. Slightly more than 50 percent of our IPOs have an integer
price which compares to 76 percent of U.S. IPOs as reported in Bradley et al. We include in our
regression a dummy variable equal to one when the offer price is an integer.14
Sherman (2005) notes that methods for taking firms public worldwide are converging
towards the traditional U.S. bookbuilding approach, and she argues that bookbuilt offers are
associated with lower underpricing than offers conducted via auctions. Sherman argues that book
building reduces the risk faced by IPO issuers and investors, and therefore should lead to less
underpricing. However, she also notes that book building affords issuers discretion over
information expenditures, which she suggests can lead to either more or less underpricing
tter (1987) provides evidence that firm commitment
IPOs are underpriced less than best efforts IPOs. Sixty-one percent of the firms in our sample are
taken public through a firm commitment offering, while 63 percent of the deals are bookbuilt
offerings. Deals that are both firm commitment and bookbuilt represent 37.4 percent of the
sample. We include dummy variables for bookbuilt deals and for firm-commitment offerings.
Schipper and Smith (1986) and Prezas et al. (2000) provide evidence that equity carve-outs are
underpriced less than original IPOs. To capture this effect, we include a dummy variable for
carve-outs.
IV. Earnings quality
Country-level earnings quality and IPO underpricing
In this section, we ask whether earnings quality affects underpricing. We recalculate two
different sets of earnings quality measures from prior studies, and we also take two other
measures directly from published studies. For all measures, the underlying principle is that
14 In Japan all IPOs are priced on an integer. For Japanese IPOs, our integer dummy equals one if the
offer price (in yen) is perfectly divisible by 100.
18
managers can take various actions that cause the reported earnings distribution to obscure the true
underlying performance distribution. For the measures that we replicate, we use firm-level data
from Compustat Global over the years 1995 to 2005. All non-financial firms with a minimum of
three consecutive years of accounting data are included in the calculation of the earnings
measures. Each country included in the sample has a minimum of 200 firm-year observations
from Compustat Global and five IPOs with valid price and volume data. Each earnings
management measure is constructed annually over the 2000-2006 sample period using accounting
data from the prior five years.
Our first set of earnings measures come from Leuz et al. (2003). They construct four country-
level earnings management measures. The first of these, denoted EM1, is an earnings smoothing
measure equal to the median value in each country of the ratio of the firm-level standard
deviation of operating earnings divided by the standard deviation of cash flow from operations.
Their second measure, EM2, also attempts to capture earnings smoothing behavior. EM2 equals
the cross-sectional correlation in country i in year t between the change in accruals and the
change in cash flows (both scaled by assets). In order to standardize the interpretation of our
earnings quality measures, we transform our EM1 and EM2 measures by multiplying their values
by -1. Hence, higher values of EM1 and EM2 correspond to more aggressive earnings smoothing
by managers and are more likely to obscure the true variability in performance.
accruals scaled by the absolute value of cash flow from operations. A higher value of EM3
indicates greater earnings management. The final metric, EM4, measures loss avoidance
behavior. EM4 equals the ratio of the number of firms reporting small profits over the number of
or minus one percent. The intuition for this measure is that if managers manipulate earnings to
avoid showing losses, then there will be a missing probability mass in the earnings distribution
just to the left of breakeven, and a higher-than-expected frequency of firms reporting earnings
19
just above zero (see Degeorge et al. 1999). Therefore, the higher is the loss avoidance ratio, the
greater is the incidence of loss avoidance behavior in a given country and the more opaque are the
Finally, Leuz et al. (2003) calculate an overall earnings management
figure, aggregate earnings management (AggEM), for each country by ranking each country on
each of the four earnings management measures and then taking the average ranking, where
higher rankings signify more earnings management.
Bhattacharya et al. (2003), construct their own earnings quality measures that focus on three
aspects of earnings opacity. First, they define earnings aggressiveness as the tendency to
accelerate the recognition of gains and delay the recognition of losses. Their earnings
aggressiveness measure is the median for country i in year t of the ratio of total accruals divided
by lagged assets. A higher value of this ratio implies more aggressive (and more opaque)
earnings.
Bhattacharya et al. (2003) construct two other earnings opacity measures, which essentially
replicate two measures from Leuz et al. (2003). Bhattacharya et al. calculate a loss avoidance
measure for each country by taking the ratio of the number of firms with small positive earnings
divided by the number of firms with small negative earnings, and they construct an earnings
smoothing measure using the cross-sectional correlation between changes in accruals and cash
flows. Finally, Bhattacharya et al. construct a single aggregate earnings opacity measure for each
country based on a ranking methodology. Specifically, they construct deciles for each earnings
measure and then average the decile ranks across the three measures to arrive at an overall
earnings opacity ranking for each country. Countries that earn a higher average ranking have
more opaque earnings. In the tables that follow, we report regression results for all four Leuz et
al. measures as well as their aggregate earnings management score, and we report the two
(earnings aggressiveness and aggregate earnings opacity) from Bhattacharya
et al.
20
From extant research we borrow two additional variables that capture different elements of
earnings quality. The first variable measures the extent to which firms recognize bad news
quickly and good news slowly, i.e., accounting conservatism. Bushman and Piotroski (2006)
calculate accounting conservatism measures using a sample of countries and years that
substantially overlaps our sample, so we use their estimates here rather than replicating them. The
accounting conservatism measures come from a regression of the following form:
RETNEGRETNEGNI *3210 (1)
where the dependent variable is net income before extraordinary items, NEG is a dummy variable
equal to 1 when RET is negative, and RET is the fiscal-
stock. 3 captures the incremental speed of bad news
recognition in earnings relative to good news. 3 = 0, then there is no difference between the
speed of recognition of bad and good news 3 > 0, the firms recognize bad news faster
than good news. Bushman and Piotroski estimate cross-sectional regressions for each country in
3 coefficient for each country. We borrow these estimates from
Table 2 in their paper. We multiply their estimates by -1 so that a higher value indicates less
conservative accounting.
All of the earnings quality variables discussed thus far are based on quantitative analysis
of firm-level data. Our final earnings quality measure is the more qualitative Accounting
Standards Index from La Porta et al. (1998). La Porta et al. created this index by examining
annual reports from companies and rating them based on the inclusion or omission of 90
characteristics related to the quality of corporate disclosures. We multiply this index by -1 so that
a higher value indicates weaker accounting standards. A significant disadvantage of this measure
is that it is based on data from 1990, so we include it here only as an additional robustness check.
Table 3
from Table 2 as well as the earnings quality measures from Leuz et al. We estimate regressions
21
with each earnings measure separately (Models 1-4) as well as a regression that includes the
aggregate earnings management measure (Model 5). The dependent variable in the Table 3
regressions is the IPO initial return, which is calculated as the secondary market closing price
divided by the final offer price, minus one. To the extent that earnings quality is related to
information asymmetry, we expect poorer earnings quality to be associated with higher
underpricing on average.
[Place Table 3 about here]
Recall that for EM1, EM2, EM3, EM4, and the overall AggEM variable, a higher value
means more earnings management. Therefore, our earnings quality hypothesis predicts positive
coefficients for the Leuz et al. earnings quality measures. The results presented in Models 1-4
provide strong support for the earnings quality hypothesis. Each of the Leuz et al. measures enters
the regression with a positive and significant coefficient as expected. When it is common practice
in a country for managers to engage in activities to manipulate earnings, IPO underpricing is
higher. Given that all four of the individual earnings management variables link to underpricing
as expected, it is not surprising that in Model 5 the aggregate earnings management score also
indicates that poor earnings quality is associated with higher IPO underpricing.
Other results in Table 3 are broadly consistent with the extant underpricing literature. For
example, smaller IPOs and IPOs taken public after a period of high market returns have higher
underpricing. IPOs priced on integers endure higher underpricing as do firms going public in less
liquid markets. As expected, our proxy for price stabilization indicates that underpricing is higher
when price stabilization is more prevalent. We also find evidence of higher underpricing for
bookbuilt and firm commitment deals, and deals that occur following a period of low IPO
volume. The adjusted R-square values indicate that our models explain between twelve and
thirteen percent of the variation in the international underpricing cross-section.
Economically, the earnings quality effects presented in Table 3 are quite dramatic. The results
in Model 5 suggest that for a one standard deviation increase in the Leuz et al. (2003) AggEM
22
measure, underpricing increases by roughly 4.8 percentage points, or greater than one-sixth of the
sample mean initial return. As a point of reference, consider that the aggregate earnings
management measure for the Philippines is slightly more than one standard deviation above that
of the United Kingdom. Clearly, country-level earnings quality has a considerable influence on
underpricing around the world.
Table 4 repeats the cross-section analysis of IPO returns from Table 3 except that in Models 1
and 2 we replace the Leuz et al. earnings management variables with the Bhattacharya et al.
(2003) earnings aggressiveness and earnings opacity scores. Model 3 focuses on the Bushman
and Piotroski (2006) earnings conservatism variable, and Model 4 includes the La Porta et al.
accounting standards index. All of these measures are normalized such that a higher value implies
more opaque earnings, less conservative reporting practices, or weaker accounting standards.
Therefore, we expect positive signs in the regressions. Indeed, all coefficients are positive and
significant, consistent with the hypothesis that underpricing is higher in countries where investors
receive lower quality accounting information.
The Table 4 results confirm the economically significant relation between earnings quality
and IPO underpricing discussed above. Consider the accounting conservatism results reported in
Model 3. The Netherlands represents the median country in terms of accounting conservatism
(value = -0.265). Switzerland (value = -0.380) is an approximately one standard deviation
improvement from the Netherlands. The coefficient on accounting conservatism (0.652) suggests
that, all things equal, IPOs in Switzerland will be underpriced by 7.5 percentage points less than
IPOs in the Netherlands.
[Place Table 4 about here]
In Table 5 we provide a robustness check with an alternative control for price stabilization
activity by underwriters. In these regressions, we measure the IPO initial return as the percentage
difference between the offer price and the secondary market closing price 22 trading days after
the IPO. The intuition for doing this is that the effects of price stabilization dissipate over time,
23
generally within a month of the IPO, as reported in Ruud (1993) and elsewhere. Thus, if price
stabilization temporarily obscures the left tail of the IPO returns distribution, then our finding that
lower quality earnings leads to higher underpricing might be the result of greater price support in
countries with lower earnings quality. Table 5 indicates that measuring the initial return over a
longer horizon does not fundamentally change our results. All earnings quality measures have the
expected signs and are significant, and most other results from Tables 3 and 4 carry through as
well.15
[Place Table 5 about here]
The post price stabilization results reported in Table 5 continue to point to an economically
opacity (value = 5.743) is slightly above the sample mean (value = 5.477). Based on the earnings
opacity score, earnings in the United States (value = 3.295) are more transparent than those in
South Korea. The earnings opacity coefficient reported in Table 5, Model 2 suggests that, all
things equal, South Korean IPOs would reduce their underpricing by approximately 563 basis
points, on average, by improving their earnings opacity measure to that of the United States.
Together, the results in Tables 3 through 5 suggest an important link between earnings quality
and IPO underpricing.
Earnings quality and the certification effect of financial intermediaries
The idea that uncertainty, or the lack thereof, can influence underpricing is well established in
the IPO literature. One mechanism that could mitigate this uncertainty is the presence of a
reputable intermediary who, in effect, certifies the earnings of IPO firms. Carter and Manaster
15 The regression models in Tables 3 and 4 do not contain country dummies. With several right-hand-
side variables measured at the country level, some country dummies display multicollinearity with other
controls. If we introduce country dummies, the results in Tables 3 and 4 are qualitatively similar, but
slightly weaker.
24
(1990), Megginson and Weiss (1991), Barry et al. (1991) all report results consistent with a
certification effect (i.e., lower underpricing) when firms go public with the assistance of reputable
underwriters or venture capital investors.16 However, most recent underpricing studies find a
positive relation between underwriter quality and initial returns. Why IPOs underwritten by more
prestigious investment banks are underpriced more remains an unanswered question, but our
results indicate that this pattern continues for recent years and across the globe.
In spite of the positive coefficient for underwriter reputation, underwriters can still play a role
in reducing underpricing in countries with high levels of information asymmetries. In Table 6 we
include an interaction between our underwriter reputation measure and each of the earnings
quality measures. For example, Table 6 Model 1 includes an interaction term between the
underwriter quality and aggregate earnings management measure. If higher quality underwriters
certify the earnings of IPO firms and thereby reduce the uncertainty faced by investors, then we
expect the interaction terms to have the opposite sign of the relevant earnings quality measure. In
all four regression specifications in Table 6, the coefficient on the interaction between
underwriter reputation and earnings quality is significant and takes the opposite sign of the
earnings quality coefficient.
[Place Table 6 about here]
The coefficients on the interaction terms indicate that IPO issuers benefit more from
significance of this result, consider an issuer in India with an average earnings opacity score of
8.0. Suppose one firm in India goes public with the assistance of an underwriter with the sample
median reputation (i.e., market share) of 0.021, and another firm uses an underwriter with a
reputation at the 75th percentile, 0.081. The coefficient on the interaction term in Model 2 of
16 Due to missing data, reported results do not control for venture capital backing. Unreported tests
indicate that all results are robust to controlling for venture capital backing where available.
25
Table 6 is -0.259, and the coefficient on the underwriter reputation variable is 1.763. Moving
from the median to the 75th percentile in terms of underwriter quality has a significant net effect
on underpricing for Indian IPOs. Underpricing increases through the underwriter reputation
coefficient by 0.06 times 1.763, but underpricing falls through the interaction term by 0.06 times
the product of -0.259 and 8.0, thus the net effect is an underpricing reduction of 185 basis points.
We obtain similar results when examining the interaction between the other proxies for earnings
quality and underwriter reputation. These results are consistent with underwriter reputation acting
as certification for new issues and are particularly important in countries with lower earnings
quality.
Table 7 provides additional robustness checks. In Models 1-4 we impose increasingly
stringent restrictions on our country-level sample by imposing a minimum offer price cutoff. In
particular, the bottom 2, 5, 10, and 20 percent of offer prices are excluded from respective
country-level samples. In terms of the U.S. distribution of offer prices, these cutoffs correspond to
minimums of $5, $6, $7, and $9. Increasing the minimum offer price reduces the odds of finding
spurious results driven by market microstructure effects. In this table, we report only one earnings
quality measure, the aggregate earnings management score, but we obtain similar results with the
other earnings quality measures. In all cases, the results indicate that an economically and
statistically significant increase in underpricing occurs when earnings quality deteriorates. As
before, high-quality underwriters mitigate this effect. In models 5-7, we exclude each of the three
largest IPO markets in our sample, the U.S., the U.K, and Japan. We do this simply to illustrate
that the earnings quality results are not driven exclusively by these markets. Excluding any of
these three countries, lower earnings quality is still associated with higher underpricing.
Similarly, the underwriter certification effect persists when we impose stringent minimum offer
price restrictions on the sample, or when we exclude the U.S., the U.K., or Japan.
[Place Table 7 about here]
26
V. Conclusion
In this paper we examine IPO underpricing across many different countries. Many of the
relations documented previously in single-country studies hold up in an international setting.
However, our primary contribution is not simply to verify the robustness of established effects
within the international cross-section, but to study how country-level differences in earnings
informativeness influence underpricing.
Using a wide range of earnings quality measures, we find higher underpricing in countries
with lower earnings quality even after controlling for many country- and deal-specific
characteristics. This evidence is consistent with asymmetric information explanations for
underpricing. Just as other researchers have found that poor accounting information can lead to a
higher cost of capital, our evidence suggests that the cost of going public rises when investors
have greater difficulties interpreting financial information. However, financial intermediaries
such as investment banks play a role in mitigating the effects of low earnings quality on
underpricing. These results are robust to controlling for deal- and macroeconomic-specific
characteristics.
27
References
Aggarwal, R. K., L. Krigman, and K. L. Womack. 2002. Strategic IPO underpricing, information
momentum, and lockup expiration selling. Journal of Financial Economics 66 (October):
105 137.
Aharony, J., C. Lin, and M. P. Loeb. 1993. Initial public offerings, accounting choices, and
earnings management. Contemporary Accounting Research 10 (Fall): 61-81.
Aharony, J., C. J. Lee, and T. J. Wong. Financial packaging of IPO firms in China. 2000. Journal
of Accounting Research 38 (Spring): 103-126.
Allen, F., and G. R. Faulhaber. Signaling by underpricing in the IPO market. 1989. Journal of
Financial Economics 23 (August): 303-323.
Ball, R., A. Robin, and J. S. Wu. 2003. Incentives versus standards: properties of accounting
income in four East Asian countries. Journal of Accounting and Economics 36 (December):
235-270.
Ball, R., and L. Shivakumar. 2008. Earnings quality at initial public offerings. Journal of
Accounting and Economics 45 (August): 324-349.
Baron, D. P. 1982. A model of the demand for investment banking advising and distribution
services for new issues. Journal of Finance 37 (September): 955-976.
Barry, C. B., C. J. Muscarella, J. Peavy, III, and M. R. Vetsuypens. 1990. The role of venture
capital in the creation of public companies: evidence from the going-public process. Journal
of Financial Economics 27 (October): 447 472.
Beck, T., A. Demirgüç-Kunt, and R. Levine. 2000. A new database on financial development and
structure. World Bank Economic Review 14 (September): 597 605.
Benveniste, L. M., and P. A. Spindt. 1989. How investment bankers determine the offer price and
allocation of new issues. Journal of Financial Economics 24 (October): 343-361.
Bhattacharya, U., H. Daouk, and M. Welker. 2003. The world price of earnings opacity. The
Accounting Review 78 (July): 641 678.
28
Biddle, G., and G. Hilary. 2006. Accounting quality and firm-level capital investments. The
Accounting Review 81 (October): 963 982.
Bradley, D. J., J. W. Cooney, Jr., B. J. Jordan, and A. K. Singh. 2004. Negotiation and the IPO
offer price: a comparison of integer vs. non-integer IPOs. Journal of Financial and
Quantitative Analysis 39 (September): 517 539.
Bushman, R. M., and J. D. Piotroski. 2006. Financial reporting incentives for conservative
accounting: The influence of legal and political institutions. Journal of Accounting and
Economics 42 (October): 107-148.
Butler, A. W., and L. Fauver. 2006. Institutional environment and sovereign credit rating.
Financial Management 35 (Autumn): 53 79.
Carter, R. B., and S. Manaster. 1990. Initial public offerings and underwriter reputation. Journal
of Finance 45 (September): 1045 1067.
Clarkson, P. M., A. Dontoh, G. Richardson, and S. E. Sefcik. 1992. The voluntary inclusion of
earnings forecasts in IPO prospectuses. Contemporary Accounting Research 8 (Fall): 601-
626.
Coffee, Jr., J. C. 1999. The future as history: the prospects for global convergence in corporate
governance and its implications. Northwestern University Law Review 93 (Issue 3): 641 708.
Coffee, Jr., J. C. 2002. Racing towards the top? The impact of cross-listings and stock market
competition on international corporate governance. Columbia Law Review 102 (November):
1757 1831.
Defond, M., M. Hung, and R. Trezevant. 2007. Investor protection and the information content of
annual earnings announcements: international evidence. Journal of Accounting and
Economics 43 (March): 37-67.
Degeorge, F., J. Patel, and R. J. Zeckhauser. 1999. Earnings management to exceed thresholds.
Journal of Business 72 (January): 1 33.
29
Djankov, S., R. La Porta, F. Lopez-De-Silanes, and A. Shleifer. 2008. The law and economics of
self-dealing. Journal of Financial Economics 88 (June): 430-465.
Ducharme, L. L., P. H. Malatesta, and S. E. Sefcik. 2001. Earnings management: IPO valuation
and subsequent performance. Journal of Accounting, Auditing, and Finance 16 (Fall): 369-
396.
Dyck, A., and L. Zingales. 2004. Private benefits of control: An international comparison.
Journal of Finance 59 (April): 537-600.
Ellul, A., and M. Pagano. 2006. IPO underpricing and after-market liquidity. Review of Financial
Studies 19 (Issue 2): 381 412.
Fan, Q. 2007. Earnings management and ownership retention for initial public offering firms:
Theory and evidence. The Accounting Review 82 (January): 27-64.
Friedlan, J. M. 1994. Accounting choices of issuers of initial public offerings. Contemporary
Accounting Research 11 (Summer): 1-31.
Gelos, R. G., and S. Wei. 2005. Transparency and international portfolio holdings. Journal of
Finance 60 (December): 2987 3020.
Gompers, P., and J. Lerner. 2002. The Venture Capital Cycle. Cambridge, MA: The MIT Press.
Grinblatt, M., and C. Y. Hwang. 1989. Signaling and the pricing of new issues. Journal of
Finance 44 (June): 393-420.
Hellmann, T., and M. Puri. 2002. Venture capital and the professionalization of start-up firms:
empirical evidence. Journal of Finance, 57 (February): 169 197.
Hung, M. 2000. Accounting standards and value relevance of financial statements: An
international analysis. Journal of Accounting and Economics 30 (December): 401-420.
Jaganathan, R., and A. Sherman. 2006. Why do IPO auctions fail? NBER working paper no.
12151.
Jin, L., and S. C. Myers. 2006. R2 around the world: new theory and new tests. Journal of
Financial Economics 79 (February): 257 292.
30
Jog, V., and B. McConomy. 2003. Voluntary disclosure of management earnings forecasts in IPO
prospectuses. Journal of Business Finance and Accounting 30 (January): 125-167.
La Porta, R., F. Lopez-De-Silanes, A. Shleifer, and R. Vishny. 1998. Law and finance. Journal of
Political Economy 106 (December): 1113-1155.
La Porta, R., F. Lopez-De-Silanes, A. Shleifer, and R. Vishny. 2000. Investor protection and
corporate governance. Journal of Financial Economics 58 (October/November): 3-27.
La Porta, R., F. Lopez-De-Silanes, and A. Shleifer. 2006. What works in securities laws? Journal
of Finance 61 (February): 1-32.
Leone, A. J., S. Rock, and M. Willenborg. 2007. Disclosure of intended use of proceeds and
underpricing. Journal of Accounting Research 45 (March): 111-153.
Leuz, C., D. Nanda, and P. D. Wysocki. 2003. Earnings management and investor protection: an
international comparison. Journal of Financial Economics 69 (September): 505 527.
Lewis, M. 2006. Assessing earnings quality at the IPO: The role of reputable investment banks.
University of Utah working paper,
Ljungqvist, A. P. 2007. IPO Underpricing. In Handbook of Corporate Finance, Volume 1:
Empirical Corporate Finance, edited by Espen Eckbo, 375-422. New York, NY:
Elsevier/North Holland.
Ljungqvist, A. P., T. Jenkinson, and W. J. Wilhelm, Jr. 2003. Global integration in primary equity
markets: The role of U.S. banks and U.S. investors. Review of Financial Studies 16 (Issue 1):
63-99.
Loughran, T., J. R. Ritter, and K. Rydqvist. 1994. Initial public offerings: international insights.
Pacific Basin Journal 2 (May): 165 199.
Loughran, T., and J. R. Ritter. 2004. Why has IPO underpricing changed over time? Financial
Management 33 (Autumn): 5 37.
Megginson, W. L., and K. A. Weiss. 1991. Venture capitalist certification in initial public
offerings. Journal of Finance 46 (July): 879 903.
31
Moeller, S. B., and F. P. Schlingemann. 2005. Global diversification and bidder gains: A
comparison between cross-border and domestic acquisitions. Journal of Banking and Finance
29 (March): 533-564.
Morsfield, S. G., and C. E. L. Tan. 2006. Do venture capitalists influence the decision to manage
earnings in initial public offerings? The Accounting Review 81 (October): 1119-1150.
Prezas, A. P., M. Tarimcilar, and G. K. Vasudevan. 2000. The pricing of equity carve-outs.
Financial Review 35 (November): 123 137.
Ritter, J. R. 1984. The hot issue market of 1980. Journal of Business 57 (April): 215 240.
Ritter, J. R. 1987. The costs of going public. Journal of Financial Economics 19 (December):
269 281.
Ritter, J. R. 1991. The long-run performance of initial public offerings. Journal of Finance 46
(March): 3 27.
Rock, K. 1986. Why new issues are underpriced. Journal of Financial Economics 15 (January):
187 212.
Ruud, J. S. 1993. Underwriter price support and the IPO underpricing puzzle. Journal of
Financial Economics 34 (October): 135-151.
Schipper, K., and A. Smith. 1986. A comparison of equity carve-outs and seasoned equity
offerings: share price effects and corporate restructuring. Journal of Financial Economics 15
(January/February): 153 186.
Schrand, C., and R. E. Verrecchia. 2005. Information disclosure and adverse selection
explanations for IPO underpricing. Wharton working paper.
Sherman, A. E. 2005. Global trends in IPO methods: book building versus auctions with
endogenous entry. Journal of Financial Economics 78 (December): 615 649.
Smart, S. B., and C. J. Zutter. 2003. Control as a motivation for underpricing: a comparison of
dual and single-class IPOs. Journal of Financial Economics 69 (July): 85 110.
32
Teoh, S. H., I. Welch, and T. J. Wong. 1998. Earnings management and the long-run market
performance of initial public offerings. Journal of Finance 53 (December): 1935-1974.
Teoh, S. H., and T. J. Wong. 2002. Why new issues and high-accrual firms underperform: The
. Review of Financial Studies 15 (Issue 3): 869-900.
Teoh, S. H., T. J. Wong, and G. R. Rao. 1998. Are accruals during initial public offerings
opportunistic? Review of Accounting Studies 3 (March): 175-208
Venkataraman, R., J. P. Weber, and M. Willenborg. 2008. Litigation risk, audit quality, and audit
fees: Evidence from initial public offerings. The Accounting Review 83 (September): 1315-
1345.
Welch, I. 1989. Seasoned offerings, imitation costs, and the underpricing of initial public
offerings. Journal of Finance 44 (June): 421-449.
Willenborg, M., and J. C. Mckeown. 2001. Going-concern initial public offerings. Journal of
Accounting and Economics 30 (December): 279-313.
33
0
200
400
600
800
1000
1200
1400
1600
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
2000 2001 2002 2003 2004 2005 2006
New
Issu
es
Ave
rage
Und
erpr
icin
g
Average Underpricing New Issues
Fig. 1. Average underpricing and number of IPOs by year. Bars show average underpricing by year and correspond to the left axis. Line points show the number of new issues by year and correspond to the right axis. Underpricing is the first-day secondary market closing price divided by the final offer price, minus one.
34
0.00
5.00
10.00
15.00
20.00
25.00
30.00
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
Ave
rage
Agg
rega
te E
arni
ngs
Man
agem
ent
Ave
rage
Und
erpr
icin
g
Aggregate Earnings Quality Quartile
Average Underpricing Ave. Aggregate Earnings Management Score
Lowest earningsquality quartile
Highest earningsquality quartile
Fig. 2. Underpricing and aggregate earnings management (AggEM) of IPOs by AggEM quartiles. Sample countries are separated into quartiles based on their average AggEM score over the 2000-2006 sample period. Bars show average underpricing by quartile and correspond to the left axis. Underpricing is the first-day secondary market closing price divided by the final offer price, minus one. Line points show the average AggEM score by quartile and correspond to the right axis. AggEM is the average country i ranking across the following four earnings management measures: EM1, EM2, EM3, and EM4. EM1 is the median ratio in country i of the firm-level standard deviations of operating earnings over the cash flow from operations (both scaled by lagged total assets), multiplied by -1. EM2 is the cross-sectional correlation in country i between the change in accruals and change in cash flows from operations (both scaled by lagged total assets), multiplied by -1. EM3 is the median ratio in country i of the absolute value of accruals over the absolute value of cash flow from operations. EM4 is the ratio in country i of the number of firms reporting small profits over the number of firms reporting small losses. A small profit (loss) is defined as a value of net earnings scaled by lagged total assets in the range [0, 0.01] ([-0.01, 0)).
35
Table 1 Country-level Descriptive Statistics Country N Average Underpricing Aggregate Gross Proceeds (US$ MM)Australia 711 19.96% 15,575.33Austria 25 14.31% 4,194.38Belgium 24 8.11% 3,669.78Brazil 28 8.96% 5,735.15Canada 21 37.03% 2,462.26Denmark 15 23.18% 1,409.35Finland 21 18.94% 1,904.69France 282 12.63% 21,473.64Germany 223 29.96% 25,350.67Greece 49 28.02% 1,370.63Hong Kong 521 16.10% 43,397.99India 97 38.73% 7,558.13Indonesia 53 38.89% 1,549.59Italy 99 9.62% 12,319.03Japan 1,092 57.29% 51,969.59Malaysia 331 35.04% 2,816.09Mexico 5 2.53% 419.23Netherlands 12 13.49% 3,691.84New Zealand 35 15.03% 967.92Norway 60 4.18% 7,628.94Philippines 23 14.08% 674.42Poland 23 50.97% 1,805.38Portugal 8 10.65% 1,676.34Singapore 360 20.37% 7,295.99South Africa 5 16.85% 221.70South Korea 327 49.08% 13,619.99Spain 20 7.95% 6,209.96Sweden 42 6.22% 7,882.64Switzerland 43 14.86% 8,544.51Taiwan 431 17.90% 6,626.98Thailand 153 18.35% 5,016.80Turkey 5 19.51% 746.84United Kingdom 1,034 17.70% 63,976.36United States 1,128 24.60% 233,222.79
Full Sample 7,306 27.53% 572,984.96 This table presents country-level descriptive statistics for the entire sample of 7,306 IPOs. N is the number of sample IPOs listing in the country over the 2000-2006 sample period. Underpricing is the first-day secondary market closing price divided by the final offer price, minus one. Aggregate gross proceeds is the sum of CPI-adjusted offer value in 2006 U.S. dollars for the country of listing.
36
Table 2 IPO Sample Descriptive Statistics Variable N Mean Std Dev Minimum MaximumEM1 7,306 -0.486 0.135 -0.894 -0.254EM2 7,306 0.829 0.235 0.074 1.000EM3 7,306 0.612 0.135 0.348 0.998EM4 7,306 2.771 1.230 0.000 11.000Aggregate EM 7,306 17.052 8.022 4.500 32.250Earnings aggressiveness 7,306 -0.039 0.012 -0.086 0.023Earnings opacity 7,306 5.477 1.392 2.533 8.800Accounting conservatism 7,283 -0.257 0.118 -0.618 0.145Accounting standards index 7,283 -69.535 8.538 -83.000 0.000Initial return 7,306 0.275 0.523 -0.527 3.885Underwriter reputation 6,834 0.072 0.111 0.000 0.766Price stabilization 7,306 0.007 0.023 -0.125 0.200IPO activity 7,306 0.045 0.027 0.000 0.134Market return 7,306 0.021 0.093 -0.412 0.656Stock market turnover ratio 7,306 1.082 0.648 0.173 2.946Underdevelopment 7,306 0.285 0.092 0.200 0.680Index of economic freedom 7,306 73.875 8.961 45.748 91.367Antidirector rights index 7,306 3.859 0.888 2.000 5.000Offer size 7,294 78.556 301.247 0.100 10568.200Integer offer price 7,306 0.503 0.500 0.000 1.000Bookbuilt 7,169 0.629 0.483 0.000 1.000Firm commitment 7,256 0.608 0.488 0.000 1.000Equity carve-out 7,241 0.038 0.190 0.000 1.000 This table presents descriptive statistics for the entire sample of 7,306 IPOs. EM1 is the median ratio in country i of the firm-level standard deviations of operating earnings over the cash flow from operations (both scaled by lagged total assets), multiplied by -1. EM2 is the cross-sectional correlation in country i between the change in accruals and change in cash flows from operations (both scaled by lagged total assets), multiplied by -1. EM3 is the median ratio in country i of the absolute value of accruals over the absolute value of cash flow from operations. EM4 is the ratio in country i of the number of firms reporting small profits over the number of firms reporting small losses. A small profit (loss) is defined as a value of net earnings scaled by lagged total assets in the range [0, 0.01] ([-0.01, 0)). AggEM is the average country i ranking across the following four earnings management measures: EM1, EM2, EM3, and EM4. Earnings aggressiveness is the median ratio in country i of total accruals over the lagged total assets. Earnings opacity is the average country i decile ranking across the following three earnings management measures: EM2, EM4, and earnings aggressiveness.
(2006). Accounting conservatism reported by Bushman and Piotroski is multiplied by -
et al. (1998). Accounting standards index reported by La Porta, et al. is multiplied by -1. Initial return is the secondary market closing price divided by the final offer price, minus one. Underwriter reputation is the country of listing Megginson-Weiss underwriter market share measure, which is the fraction of total CPI-adjusted offer value underwritten by a given underwriter for the sample. Price stabilization is the difference in the number of IPOs with small positive first day returns (greater than zero and less than or equal to one percent) and the number of IPOs with small negative first day returns (less than zero and greater than or equal to negative one percent) divided by the total number of IPOs issued in the country of listing. IPO activity is the ratio of the total number of IPOs in the issue year divided by the number of Datastream listed equities for the country of listing as of 2006. Market return is the return on the Datastream index for the country of listing over the three months preceding the offering. Stock market turnover ratio equals the ratio of the total value of shares traded to aggregate market capitalization and reported in Beck et al. (2000). Underdevelopment is the sum of the rankings of infant mortality, internet users per population, literacy, unemployment rate, and paved airport runways as reported by Butler and Fauver (2006). Index of economic freedom is a product of The Heritage Foundation and The Wall Street Journal and is an aggregate measure covering the following ten freedoms: business, trade, monetary, freedom from government, fiscal, property
37
rights, investment, financial, freedom from corruption, and labor. Antidirector rights index measures shareholder rights by considering the following issues: (1) vote by mail, (2) shares not blocked or deposited prior to shareholder meetings, (3) cumulative voting in director elections, (4) oppressed minority mechanisms, (5) pre-emptive rights to new issues, and (6) minimum capital requirements and is reported in Djankov et al. (2008). Offer size is the CPI-adjusted offer value in millions of U.S. dollars. Indicator variables are set equal to one for integer offer price, bookbuilt, firm commitment, and equity carve-out deals.
Page 38
Table 3 Underpricing Regressions on Earnings Quality
Model 1 Model 2 Model 3 Model 4 Model 5Intercept 0.905*** 0.425*** 0.495*** 0.563*** 0.501***EM1 0.526***EM2 0.166***EM3 0.272***EM4 0.022***Aggregate EM 0.006***Underwriter reputation 0.200*** 0.243*** 0.240*** 0.254*** 0.224***Price stabilization 1.464*** 1.422*** 1.577*** 1.291*** 1.560***IPO activity -1.259*** -0.500 -0.503 -0.221 -0.859**Market return 1.161*** 1.132*** 1.147*** 1.135*** 1.152***Stock market turnover ratio -0.028** -0.001 -0.013 -0.021 -0.006Underdevelopment -0.754*** -0.602*** -0.711*** -0.605*** -0.672***Index of economic freedom -0.004*** -0.004*** -0.005*** -0.003*** -0.004***Antidirector rights index 0.073*** 0.072*** 0.072*** 0.048*** 0.070***Offer size -0.050*** -0.056*** -0.053*** -0.054*** -0.052***Integer offer price 0.149*** 0.154*** 0.161*** 0.154*** 0.154***Bookbuilt 0.029* 0.070*** 0.060*** 0.031* 0.050***Firm commitment 0.113*** 0.128*** 0.132*** 0.135*** 0.122***Equity carve-out -0.019 -0.013 -0.014 -0.015 -0.017
Adjusted R-square 13.02% 12.52% 12.36% 12.35% 12.56%Number of observations 6,650 6,650 6,650 6,650 6,650 This table presents OLS regressions of IPO underpricing on country-level earnings quality measures. The dependent variable is the IPO initial return, which is the secondary market closing price divided by the final offer price, minus one. EM1 is the median ratio in country i of the firm-level standard deviations of operating earnings over the cash flow from operations (both scaled by lagged total assets), multiplied by -1. EM2 is the cross-sectional correlation in country i between the change in accruals and change in cash flows from operations (both scaled by lagged total assets), multiplied by -1. EM3 is the median ratio in country i of the absolute value of accruals over the absolute value of cash flow from operations. EM4 is the ratio in country i of the number of firms reporting small profits over the number of firms reporting small losses. A small profit (loss) is defined as a value of net earnings scaled by lagged total assets in the range [0, 0.01] ([-0.01, 0)). AggEM is the average country i ranking across the following four earnings management measures: EM1, EM2, EM3, and EM4. Underwriter reputation is the country of listing Megginson-Weiss underwriter market share measure, which is the fraction of total CPI-adjusted offer value underwritten by a given underwriter for the sample. Price stabilization is the difference in the number of IPOs with small positive first day returns (greater than zero and less than or equal to one percent) and the number of IPOs with small negative first day returns (less than zero and greater than or equal to negative one percent) divided by the total number of IPOs issued in the country of listing. IPO activity is the ratio of the total number of IPOs in the issue year divided by the number of Datastream listed equities for the country of listing as of 2006. Market return is the return on the Datastream index for the country of listing over the three months preceding the offering. Stock market turnover ratio equals the ratio of the total value of shares traded to aggregate market capitalization and reported in Beck et al. (2000). Underdevelopment is the sum of the rankings of infant mortality, internet users per population, literacy, unemployment rate, and paved airport runways as reported by Butler and Fauver (2006). Index of economic freedom is a product of The Heritage Foundation and The Wall Street Journal and is an aggregate measure covering the following ten freedoms: business, trade, monetary, freedom from government, fiscal, property rights, investment, financial, freedom from corruption, and labor. Antidirector rights index measures shareholder rights by considering the following issues: (1) vote by mail, (2) shares not blocked or deposited prior to shareholder meetings, (3) cumulative voting in director elections, (4) oppressed minority mechanisms, (5) pre-emptive rights to new issues, and (6) minimum capital requirements and is reported in Djankov et al. (2008). Offer size is the natural log of the CPI-adjusted offer value in millions of U.S. dollars. Indicator variables are set equal to one for integer offer price, bookbuilt, firm commitment, and equity carve-out deals. Regressions
Page 39
include IPO year and industry dummies. Industry classifications reflect those reported by Dyck and Zingales (2004). Respectively, ***, **, and * denote significance of the coefficient at the 1, 5, and 10 percent level.
Page 40
Table 4 Underpricing Regressions on Alternative Earnings Quality Measures
Model 1 Model 2 Model 3 Model 4Intercept 0.844*** 0.474*** 0.614*** 1.277***Earnings aggressiveness 4.018***Earnings opacity 0.029***Accounting conservatism 0.652***Accounting standards index 0.012***Underwriter reputation 0.251*** 0.234*** 0.180*** 0.221***Price stabilization 1.673*** 1.449*** 0.382 1.289***IPO activity 0.062 -0.682** -0.797** -0.398Market return 1.142*** 1.138*** 1.162*** 1.165***Stock market turnover ratio -0.021 -0.006 -0.043*** -0.030**Underdevelopment -0.681*** -0.608*** -0.397*** -0.883***Index of economic freedom -0.005*** -0.004*** 0.001 -0.002**Antidirector rights index 0.054*** 0.054*** 0.019* 0.094***Offer size -0.052*** -0.053*** -0.053*** -0.055***Integer offer price 0.162*** 0.153*** 0.150*** 0.140***Bookbuilt 0.057*** 0.046*** 0.044*** 0.032*Firm commitment 0.095*** 0.115*** 0.105*** 0.139***Equity carve-out -0.013 -0.018 -0.013 -0.011
Adjusted R-square 12.63% 12.47% 13.22% 13.28%Number of observations 6,650 6,650 6,632 6,585 This table presents OLS regressions of IPO underpricing on country-level earnings quality measures. The dependent variable is the IPO initial return, which is the secondary market closing price divided by the final offer price, minus one. Earnings aggressiveness is the median ratio in country i of total accruals over the lagged total assets. Earnings opacity is the average country i decile ranking across the following three earnings management measures: EM2, EM4, and earnings aggressiveness. EM2 is the cross-sectional correlation in country i between the change in accruals and change in cash flows from operations (both scaled by lagged total assets), multiplied by -1. EM4 is the ratio in country i of the number of firms reporting small profits over the number of firms reporting small losses. A small profit (loss) is defined as a value of net earnings scaled by lagged total assets in the range [0, 0.01] ([-0.01, 0)). Accounting
(2006). Accounting conservatism reported by Bushman and Piotroski is multiplied by -1. Accounting standards ind
et al. (1998). Accounting standards index reported by La Porta, et al. is multiplied by -1. Underwriter reputation is the country of listing Megginson-Weiss underwriter market share measure, which is the fraction of total CPI-adjusted offer value underwritten by a given underwriter for the sample. Price stabilization is the difference in the number of IPOs with small positive first day returns (greater than zero and less than or equal to one percent) and the number of IPOs with small negative first day returns (less than zero and greater than or equal to negative one percent) divided by the total number of IPOs issued in the country of listing. IPO activity is the ratio of the total number of IPOs in the issue year divided by the number of Datastream listed equities for the country of listing as of 2006. Market return is the return on the Datastream index for the country of listing over the three months preceding the offering. Stock market turnover ratio equals the ratio of the total value of shares traded to aggregate market capitalization and reported in Beck et al. (2000). Underdevelopment is the sum of the rankings of infant mortality, internet users per population, literacy, unemployment rate, and paved airport runways as reported by Butler and Fauver (2006). Index of economic freedom is a product of The Heritage Foundation and The Wall Street Journal and is an aggregate measure covering the following ten freedoms: business, trade, monetary, freedom from government, fiscal, property rights, investment,
Page 41
financial, freedom from corruption, and labor. Antidirector rights index measures shareholder rights by considering the following issues: (1) vote by mail, (2) shares not blocked or deposited prior to shareholder meetings, (3) cumulative voting in director elections, (4) oppressed minority mechanisms, (5) pre-emptive rights to new issues, and (6) minimum capital requirements and is reported in Djankov et al. (2008). Offer size is the natural log of the CPI-adjusted offer value in millions of U.S. dollars. Indicator variables are set equal to one for integer offer price, bookbuilt, firm commitment, and equity carve-out deals. Regressions include IPO year and industry dummies. Industry classifications reflect those reported by Dyck and Zingales (2004). Respectively, ***, **, and * denote significance of the coefficient at the 1, 5, and 10 percent level.
Page 42
Table 5 Underpricing Regressions on Earnings Quality Following Price Stabilization
Model 1 Model 2 Model 3 Model 4Intercept 0.484*** 0.455*** 0.556*** 1.260***Aggregate EM 0.004***Earnings opacity 0.023**Accounting conservatism 0.624***Accounting standards index 0.013***Underwriter reputation 0.328*** 0.334*** 0.287*** 0.313***IPO activity -1.060** -0.945* -0.925** -0.777*Market return 1.297*** 1.286*** 1.298*** 1.303***Stock market turnover ratio -0.017 -0.016 -0.045** -0.036*Underdevelopment -0.497*** -0.457*** -0.319*** -0.792***Index of economic freedom -0.003** -0.002* 0.002 0.000Antidirector rights index 0.045*** 0.034** 0.013 0.078***Offer size -0.055*** -0.055*** -0.057*** -0.058***Integer offer price 0.189*** 0.188*** 0.183*** 0.171***Bookbuilt 0.104*** 0.101*** 0.092*** 0.086***Firm commitment 0.106*** 0.099*** 0.082*** 0.118***Equity carve-out -0.055 -0.057 -0.051 -0.050
Adjusted R-square 7.43% 7.42% 7.87% 7.97%Number of observations 6,617 6,617 6,600 6,553 This table presents OLS regressions of IPO underpricing on country-level earnings quality measures. The dependent variable is the post-price stabilization IPO initial return, which is the 22nd secondary market closing price divided by the final offer price, minus one. AggEM is the average country i ranking across the following four earnings management measures: EM1, EM2, EM3, and EM4. EM1 is the median ratio in country i of the firm-level standard deviations of operating earnings over the cash flow from operations (both scaled by lagged total assets), multiplied by -1. EM2 is the cross-sectional correlation in country i between the change in accruals and change in cash flows from operations (both scaled by lagged total assets), multiplied by -1. EM3 is the median ratio in country i of the absolute value of accruals over the absolute value of cash flow from operations. EM4 is the ratio in country i of the number of firms reporting small profits over the number of firms reporting small losses. A small profit (loss) is defined as a value of net earnings scaled by lagged total assets in the range [0, 0.01] ([-0.01, 0)). Earnings opacity is the average country i decile ranking across the following three earnings management measures: EM2, EM4, and earnings aggressiveness. Earnings aggressiveness is the median ratio in country i of total accruals over the lagged total assets. ed by Bushman and Piotroski (2006). Accounting conservatism reported by Bushman and Piotroski is multiplied by -1.
reported in La Porta et al. (1998). Accounting standards index reported by La Porta, et al. is multiplied by -1. Underwriter reputation is the country of listing Megginson-Weiss underwriter market share measure, which is the fraction of total CPI-adjusted offer value underwritten by a given underwriter for the sample. Price stabilization is the difference in the number of IPOs with small positive first day returns (greater than zero and less than or equal to one percent) and the number of IPOs with small negative first day returns (less than zero and greater than or equal to negative one percent) divided by the total number of IPOs issued in the country of listing. IPO activity is the ratio of the total number of IPOs in the issue year divided by the number of Datastream listed equities for the country of listing as of 2006. Market return is the return on the Datastream index for the country of listing over the three months preceding the offering. Stock market turnover ratio equals the ratio of the total value of shares traded to aggregate market capitalization and reported in Beck et al. (2000).
Page 43
Underdevelopment is the sum of the rankings of infant mortality, internet users per population, literacy, unemployment rate, and paved airport runways as reported by Butler and Fauver (2006). Index of economic freedom is a product of The Heritage Foundation and The Wall Street Journal and is an aggregate measure covering the following ten freedoms: business, trade, monetary, freedom from government, fiscal, property rights, investment, financial, freedom from corruption, and labor. Antidirector rights index measures shareholder rights by considering the following issues: (1) vote by mail, (2) shares not blocked or deposited prior to shareholder meetings, (3) cumulative voting in director elections, (4) oppressed minority mechanisms, (5) pre-emptive rights to new issues, and (6) minimum capital requirements and is reported in Djankov et al. (2008). Offer size is the natural log of the CPI-adjusted offer value in millions of U.S. dollars. Indicator variables are set equal to one for integer offer price, bookbuilt, firm commitment, and equity carve-out deals. Regressions include IPO year and industry dummies. Industry classifications reflect those reported by Dyck and Zingales (2004). Respectively, ***, **, and * denote significance of the coefficient at the 1, 5, and 10 percent level.
Page 44
Table 6 Underpricing Regressions on Earnings Quality with Underwriter Certification
Model 1 Model 2 Model 3 Model 4Intercept 0.465*** 0.398*** 0.646*** 1.507***Aggregate EM 0.010***Earnings opacity 0.049***Accounting conservatism 0.821***Accounting standards index 0.016***Underwriter reputation 1.278*** 1.763*** -0.161 -2.117***Underwriter reputation * earnings measure -0.054*** -0.259*** -1.580*** -0.034***Price stabilization 1.744*** 1.541*** 0.439 1.229***IPO activity -0.939*** -0.754** -0.794** -0.466Market return 1.146*** 1.136*** 1.163*** 1.163***Stock market turnover ratio -0.002 -0.003 -0.047*** -0.028**Underdevelopment -0.858*** -0.715*** -0.411*** -0.941***Index of economic freedom -0.004*** -0.004*** 0.001 -0.002*Antidirector rights index 0.083*** 0.065*** 0.020* 0.105***Offer size -0.055*** -0.055*** -0.054*** -0.055***Integer offer price 0.151*** 0.153*** 0.147*** 0.136***Bookbuilt 0.053*** 0.051*** 0.045*** 0.036**Firm commitment 0.127*** 0.121*** 0.109*** 0.138***Equity carve-out -0.023 -0.024 -0.015 -0.012
Adjusted R-square 13.16% 12.94% 13.41% 13.46%Number of observations 6,650 6,650 6,632 6,585 This table presents OLS regressions of IPO underpricing on country-level earnings quality measures and the interaction of underwriter reputation and earnings quality. The dependent variable is the IPO initial return, which is the secondary market closing price divided by the final offer price, minus one. AggEM is the average country i ranking across the following four earnings management measures: EM1, EM2, EM3, and EM4. EM1 is the median ratio in country i of the firm-level standard deviations of operating earnings over the cash flow from operations (both scaled by lagged total assets), multiplied by -1. EM2 is the cross-sectional correlation in country i between the change in accruals and change in cash flows from operations (both scaled by lagged total assets), multiplied by -1. EM3 is the median ratio in country i of the absolute value of accruals over the absolute value of cash flow from operations. EM4 is the ratio in country i of the number of firms reporting small profits over the number of firms reporting small losses. A small profit (loss) is defined as a value of net earnings scaled by lagged total assets in the range [0, 0.01] ([-0.01, 0)). Earnings opacity is the average country i decile ranking across the following three earnings management measures: EM2, EM4, and earnings aggressiveness. Earnings aggressiveness is the median ratio in country i of total accruals over the lagged total assets.
(2006). Accounting conservatism reported by Bushman and Piotroski is multiplied by -1.
et al. (1998). Accounting standards index reported by La Porta, et al. is multiplied by -1. Underwriter reputation is the country of listing Megginson-Weiss underwriter market share measure, which is the fraction of total CPI-adjusted offer value underwritten by a given underwriter for the sample. Price stabilization is the difference in the number of IPOs with small positive first day returns (greater than zero and less than or equal to one percent) and the number of IPOs with small negative first day returns (less than zero and greater than or equal to negative one percent) divided by the total number of IPOs issued in the country of listing. IPO activity is the ratio of the total number of IPOs in the issue year divided by the number of Datastream listed equities for the country of listing as of 2006. Market return is the return on the Datastream index for the country of listing over the three months preceding the offering. Stock market turnover ratio equals the ratio of the total value of shares traded to aggregate market capitalization and
Page 45
reported in Beck et al. (2000). Underdevelopment is the sum of the rankings of infant mortality, internet users per population, literacy, unemployment rate, and paved airport runways as reported by Butler and Fauver (2006). Index of economic freedom is a product of The Heritage Foundation and The Wall Street Journal and is an aggregate measure covering the following ten freedoms: business, trade, monetary, freedom from government, fiscal, property rights, investment, financial, freedom from corruption, and labor. Antidirector rights index measures shareholder rights by considering the following issues: (1) vote by mail, (2) shares not blocked or deposited prior to shareholder meetings, (3) cumulative voting in director elections, (4) oppressed minority mechanisms, (5) pre-emptive rights to new issues, and (6) minimum capital requirements and is reported in Djankov et al. (2008). Offer size is the natural log of the CPI-adjusted offer value in millions of U.S. dollars. Indicator variables are set equal to one for integer offer price, bookbuilt, firm commitment, and equity carve-out deals. Regressions include IPO year and industry dummies. Industry classifications reflect those reported by Dyck and Zingales (2004). Respectively, ***, **, and * denote significance of the coefficient at the 1, 5, and 10 percent level.
Page
46
Tabl
e 7
Und
erpr
icin
g R
egre
ssio
ns o
n Ea
rnin
gs Q
ualit
y R
obus
tnes
s
Elim
inat
e bo
ttom
2%
Elim
inat
e bo
ttom
5%
Elim
inat
e bo
ttom
10%
Elim
inat
e bo
ttom
20%
Excl
udin
g U
.S.
Excl
udin
g U
.K.
Excl
udin
g Ja
pan
Inte
rcep
t 0
.462
***
0.4
83**
* 0
.514
***
0.5
80**
* 0
.301
***
0.4
48**
* 0
.329
***
Agg
rega
te E
M 0
.010
***
0.0
11**
* 0
.011
***
0.0
10**
* 0
.011
***
0.0
09**
* 0
.002
*U
nder
writ
er re
puta
tion
1.1
75**
* 1
.126
***
0.9
72**
* 0
.924
***
0.7
50**
* 1
.154
***
0.7
81**
*U
nder
writ
er re
puta
tion
* A
ggre
gate
EM
-0.0
50**
*-0
.049
***
-0.0
43**
*-0
.039
***
-0.0
30**
*-0
.051
***
-0.0
32**
*Pr
ice
stab
iliza
tion
0.2
99 0
.295
0.2
55-0
.034
0.2
83-0
.011
0.6
18**
*IP
O a
ctiv
ity-1
.390
***
-1.5
56**
*-1
.796
***
-2.0
30**
*-1
.267
***
-1.2
92**
*-0
.830
***
Mar
ket r
etur
n 1
.183
***
1.1
76**
* 1
.200
***
1.2
60**
* 1
.120
***
1.1
80**
* 1
.045
***
Stoc
k m
arke
t tur
nove
r rat
io-0
.005
-0.0
08-0
.016
-0.0
27-0
.005
0.0
12 0
.096
***
Und
erde
velo
pmen
t-0
.836
***
-0.8
44**
*-0
.856
***
-0.8
45**
*-0
.778
***
-0.9
65**
*-0
.059
Inde
x of
eco
nom
ic fr
eedo
m-0
.003
***
-0.0
03**
*-0
.003
***
-0.0
02*
-0.0
03**
*-0
.004
***
-0.0
03**
*A
ntid
irect
or ri
ghts
inde
x 0
.064
***
0.0
59**
* 0
.052
***
0.0
32**
0.0
67**
* 0
.100
***
0.0
69**
*O
ffer s
ize
-0.0
55**
*-0
.056
***
-0.0
56**
*-0
.060
***
-0.0
62**
*-0
.050
***
-0.0
30**
*In
tege
r offe
r pric
e 0
.148
***
0.1
51**
* 0
.150
***
0.1
49**
* 0
.171
***
0.1
52**
*-0
.008
Boo
kbui
lt 0
.058
***
0.0
61**
* 0
.063
***
0.0
77**
* 0
.077
***
0.0
81**
*-0
.017
Firm
com
mitm
ent
0.1
40**
* 0
.143
***
0.1
56**
* 0
.174
***
0.0
90**
* 0
.114
***
0.0
56**
*Eq
uity
car
ve-o
ut-0
.030
-0.0
32-0
.029
-0.0
35-0
.050
-0.0
34-0
.006
Adj
uste
d R
-squ
are
13.1
9%13
.51%
14.3
3%16
.04%
13.5
3%13
.06%
11.1
0%N
umbe
r of o
bser
vatio
ns6,
273
6,09
05,
801
5,18
35,
524
5,66
05,
562
Th
is ta
ble
pres
ents
OLS
regr
essi
ons
of IP
O u
nder
pric
ing
on c
ount
ry-le
vel e
arni
ngs
qual
ity m
easu
res
and
the
inte
ract
ion
of u
nder
writ
er re
puta
tion
and
earn
ings
qu
ality
for v
ario
us s
ubsa
mpl
es. T
he fi
rst f
our m
odel
s re
stric
t the
sam
ple
to IP
Os
with
an
offe
r pric
e ab
ove
the
2nd, 5
th, 1
0th a
nd 2
0th p
erce
ntile
rela
tive
to a
ll de
als
with
in a
giv
en c
ount
ry, r
espe
ctiv
ely.
Mod
els
5, 6
, and
7 e
xclu
de I
POs
liste
d in
the
Uni
ted
Stat
es, U
nite
d K
ingd
om, a
nd J
apan
, res
pect
ivel
y. T
he d
epen
dent
va
riabl
e is
the
IPO
initi
al re
turn
, whi
ch is
the
seco
ndar
y m
arke
t clo
sing
pric
e di
vide
d by
the
final
off
er p
rice,
min
us o
ne. A
ggEM
is th
e av
erag
e co
untry
i ra
nkin
g ac
ross
the
follo
win
g fo
ur e
arni
ngs
man
agem
ent m
easu
res:
EM
1, E
M2,
EM
3, a
nd E
M4.
EM
1 is
the
med
ian
ratio
in c
ount
ry i
of th
e fir
m-le
vel s
tand
ard
devi
atio
ns
of o
pera
ting
earn
ings
ove
r the
cas
h flo
w fr
om o
pera
tions
(bot
h sc
aled
by
lagg
ed to
tal a
sset
s), m
ultip
lied
by -1
. EM
2 is
the
cros
s-se
ctio
nal c
orre
latio
n in
cou
ntry
i be
twee
n th
e ch
ange
in
accr
uals
and
cha
nge
in c
ash
flow
s fr
om o
pera
tions
(bo
th s
cale
d by
lagg
ed to
tal a
sset
s), m
ultip
lied
by -
1. E
M3
is th
e m
edia
n ra
tio in
co
untry
i of
the
abso
lute
val
ue o
f acc
rual
s ov
er th
e ab
solu
te v
alue
of c
ash
flow
from
ope
ratio
ns. E
M4
is th
e ra
tio in
cou
ntry
i of
the
num
ber o
f firm
s re
porti
ng
smal
l pro
fits
over
the
num
ber o
f firm
s re
porti
ng s
mal
l los
ses.
A s
mal
l pro
fit (l
oss)
is d
efin
ed a
s a
valu
e of
net
ear
ning
s sc
aled
by
lagg
ed to
tal a
sset
s in
the
rang
e [0
, 0.0
1] (
[-0.
01, 0
)). U
nder
writ
er r
eput
atio
n is
the
coun
try o
f lis
ting
Meg
gins
on-W
eiss
und
erw
riter
mar
ket s
hare
mea
sure
, whi
ch is
the
frac
tion
of to
tal C
PI-
adju
sted
offe
r val
ue u
nder
writ
ten
by a
giv
en u
nder
writ
er fo
r the
sam
ple.
Pric
e st
abili
zatio
n is
the
diff
eren
ce in
the
num
ber o
f IPO
s w
ith s
mal
l pos
itive
firs
t day
re
turn
s (g
reat
er th
an z
ero
and
less
than
or e
qual
to o
ne p
erce
nt) a
nd th
e nu
mbe
r of I
POs
with
sm
all n
egat
ive
first
day
retu
rns
(less
than
zer
o an
d gr
eate
r tha
n or
eq
ual t
o ne
gativ
e on
e pe
rcen
t) di
vide
d by
the
tota
l num
ber
of I
POs
issu
ed in
the
coun
try o
f lis
ting.
IPO
act
ivity
is th
e ra
tio o
f the
tota
l num
ber
of I
POs
in th
e is
sue
year
div
ided
by
the
num
ber o
f Dat
astre
am li
sted
equ
ities
for t
he c
ount
ry o
f lis
ting
as o
f 200
6. M
arke
t ret
urn
is th
e re
turn
on
the
Dat
astre
am in
dex
for t
he
coun
try o
f lis
ting
over
the
thre
e m
onth
s pre
cedi
ng th
e of
ferin
g. S
tock
mar
ket t
urno
ver r
atio
equ
als t
he ra
tio o
f the
tota
l val
ue o
f sha
res t
rade
d to
agg
rega
te m
arke
t ca
pita
lizat
ion
and
repo
rted
in B
eck
et a
l. (2
000)
. U
nder
deve
lopm
ent
is t
he s
um o
f th
e ra
nkin
gs o
f in
fant
mor
talit
y, i
nter
net
user
s pe
r po
pula
tion,
lite
racy
,
Page
47
unem
ploy
men
t rat
e, a
nd p
aved
airp
ort r
unw
ays a
s rep
orte
d by
But
ler a
nd F
auve
r (20
06).
Inde
x of
eco
nom
ic fr
eedo
m is
a p
rodu
ct o
f The
Her
itage
Fou
ndat
ion
and
The
Wal
l Stre
et J
ourn
al a
nd is
an
aggr
egat
e m
easu
re c
over
ing
the
follo
win
g te
n fr
eedo
ms:
bus
ines
s, tra
de, m
onet
ary,
free
dom
fro
m g
over
nmen
t, fis
cal,
prop
erty
rig
hts,
inve
stm
ent,
finan
cial
, fre
edom
from
cor
rupt
ion,
and
labo
r. A
ntid
irect
or ri
ghts
inde
x m
easu
res
shar
ehol
der r
ight
s by
con
side
ring
the
follo
win
g is
sues
: (1)
vo
te b
y m
ail,
(2) s
hare
s no
t blo
cked
or d
epos
ited
prio
r to
shar
ehol
der m
eetin
gs, (
3) c
umul
ativ
e vo
ting
in d
irect
or e
lect
ions
, (4)
opp
ress
ed m
inor
ity m
echa
nism
s, (5
) pr
e-em
ptiv
e rig
hts
to n
ew is
sues
, and
(6)
min
imum
cap
ital r
equi
rem
ents
and
is r
epor
ted
in D
jank
ov e
t al.
(200
8). O
ffer
siz
e is
the
natu
ral l
og o
f th
e C
PI-
adju
sted
offe
r val
ue in
mill
ions
of U
.S. d
olla
rs. I
ndic
ator
var
iabl
es a
re s
et e
qual
to o
ne fo
r int
eger
offe
r pr
ice,
boo
kbui
lt, fi
rm c
omm
itmen
t, an
d eq
uity
car
ve-o
ut
deal
s. R
egre
ssio
ns in
clud
e IP
O y
ear
and
indu
stry
dum
mie
s. In
dust
ry c
lass
ifica
tions
ref
lect
thos
e re
porte
d by
Dyc
k an
d Zi
ngal
es (
2004
). R
espe
ctiv
ely,
***
, **,
an
d *
deno
te si
gnifi
canc
e of
the
coef
ficie
nt a
t the
1, 5
, and
10
perc
ent l
evel
.