accelerated equity offers: the role of internal and...
TRANSCRIPT
Accelerated Equity Offers: The Role of Internal and External Certification
Mechanisms
Hardjo Koerniadi
Auckland University of Technology
Private Bag 92006, Auckland, New Zealand
Chandrasekhar Krishnamurti
University of Southern Queensland
Toowoomba, QLD 4350, Australia
Sie Ting Lau
Nanyang Technological University
Singapore
Alireza Tourani-Rad
Auckland University of Technology
Private Bag 92006, Auckland, New Zealand
Ting Yang
Auckland University of Technology
Private Bag 92006, Auckland, New Zealand
2
Accelerated Equity Offers: The Role of Internal and External Certification
Mechanisms
Abstract
We examine the role of internal and external certification mechanisms in the issuance choice of
SEOs between accelerated offers and fully marketed offers. Our empirical work supports the
view that a firm’s internal corporate governance structure and audit fee are associated with the
issuance choice between accelerated and fully marketed offers. Furthermore, we find that after
controlling for the self-selection problem, firms paying higher audit fee incur lower flotation
costs as measured by gross spreads. Finally, we find a significant positive association between
audit fee and the net proceeds raised in SEOs.
JEL Classifications: G14, G24, G32
Keywords: Accelerated offers, seasoned equity, flotation costs, information asymmetry, due
diligence
3
1. Introduction
In their seminal work, Myers and Majluf (1984) modelled the implications of information
asymmetry on issuing new stock. In their words:
If managers know more than the market does, firms should avoid situations in which
valuable investment projects have to be financed by stock issues. Having slack solves the
problem, and one way to get the slack is to issue stock when there is no asymmetric
information.
However, asymmetric information is always an issue and investment bankers conduct due
diligence before underwriting an issue to mitigate the problem. During the last 15 years,
accelerated bookbuilt offers have become a popular method for issuing seasoned equity offerings
(SEOs).1 Accelerated offers are typically completed in one day compared to 31 days between
filing to offer for fully marketed offers. This raises the question of how issuing firms deal with
asymmetric information while using accelerated offers method for issuing SEOs. We address this
issue by examining the role of certification in reducing the information asymmetry in the context
of issuance choice between accelerated offerings and fully marketed SEOs.
Gao and Ritter (2010) examine the choice between accelerated and fully marketed SEOs.
They also report that fully marketed offers pay higher average gross spreads than accelerated
offers. Furthermore, fully marketed offers are more underpriced compared to accelerated offers.2
Overall, fully marketed offers take longer to issue and incur higher transaction costs. This raises
the question as to why firms use the slower and more expensive method of issuing fully
marketed SEOs. Gao and Ritter (2010) posit that that firms use fully marketed issues when faced
with relatively more inelastic demand curve for their stocks. Autore et al. (2011) suggest that
lower quality issuers prefer accelerated offerings to avoid rigorous pre-issue scrutiny by
1 Gao and Ritter (2010)
2 However, accelerated offers have slightly larger discount and experience slightly more negative stock price
reaction than fully marketed offers.
4
investment bankers. On closer analysis, it appears that only one of the seven information
asymmetry variables considered by them is significantly associated with the choice of using
accelerated offers as opposed to fully marketed offers3. Therefore, we consider the evidence,
regarding the influence of the quality of issuing firm on issuance choice between fully marketed
and accelerated offers, presented in Autore et al. (2011) to be inconclusive. Hence, in this paper
we examine the role of firm quality proxies, such as their internal and external certification
mechanisms on the issuance choice between accelerated offers and fully marketed offers.
Although accelerated offers are advantageous in terms of speed of issuance and
transaction costs from the perspective of issuers, they pose additional challenges to investment
bankers. First, it increases the competition faced by investment bankers thereby reducing the
underwriting fee charged by them in accelerated SEO offerings.4 Second, accelerated offerings
place additional pressures on investment bankers in performing their due diligence
investigations. The reduced time available for conducting due diligence activities could
potentially increase the cost of such investigations. Alternately, the time constraint imposed by
the accelerated method could necessitate an upper limit on the extent of due diligence
evaluations that can be accomplished by investment bankers5. Although documented due
diligence cases are relatively rare, when they do occur, investment bankers tend to take a
substantial hit. For example, in the case of WorldCom, which conducted two speedy shelf offers,
investors sued underwriters for negligence and Citigroup, one of the firms that sold WorldCom
bonds paid $2.65 billion in settlement6. There is no evidence to indicate that investors or judges
3 Further, as these results are untabulated, we are unable to assess the strength of this relationship.
4 Please see Bortolotti et al. (2008) for empirical evidence (Table 3).
5 These arguments were advanced by Sherman (1999) in the context of shelf offerings. They apply with equal if not
greater force in the case of accelerated offerings. 6 See Morgenson (2004) for further details.
5
take a more lenient view of due diligence standards for accelerated offerings compared to fully
marketed offers (Morgenson, 2004).
How do investment bankers cope with the challenge of time pressures in conducting due
diligence of accelerated offerings? What can issuing firm managers do to mitigate the difficulties
faced by investment bankers? There is as yet no work on this issue. We contribute to the
literature by examining the importance of potential alternate certification mechanisms that firms
can use in the context of accelerated SEOs.
We suggest that firms can potentially use an internal certification mechanism such as
their corporate governance structure to diminish the role of due diligence activities performed by
investment bankers. We construct a firm’s internal corporate governance score using raw data
on a firm’s reported corporate governance practices and use this as our proxy of internal
certification quality. We posit that firms with better quality internal governance are more likely
to choose accelerated offerings while making SEOs.
We also propose that firms can use external certification mechanisms such as the
purchase of high quality audit services to mitigate the importance of due diligence investigations
by underwriters. We postulate that firms may use audit fee as an external certification
mechanism to reduce the information asymmetry between managers of issuing firms and
potential investors with investment bankers mediating between them. We therefore suggest that
firms which pay higher audit fees are more likely to choose accelerated offerings while making
SEOs.
Our empirical work suggests that these internal and external certification mechanisms
perform three significant roles in the issuance of SEOs. First, by reducing information
asymmetry, certification mechanisms influence the choice between accelerated offerings and
6
fully marketed offers. Second, we find that certification mechanisms help to reduce the issuance
costs as measured by gross spread. Finally, we find that there is an association between the
efficacy of certification and the net proceeds of SEO issuance.
The remainder of the paper is organized as follows. In the next section, we provide the
theoretical underpinnings for our empirical tests and develop the hypotheses. In section 3, we
describe the data sources, our sample selection procedures, and the characteristics of our sample.
In section 4, we report our empirical results, discuss the implications, and report results from
robustness checks. Our concluding comments are offered in section 5.
2. Theoretical Underpinnings
2.1 Shelf Registration and Underwriter Certification
The Securities and Exchange Commission (SEC) adopted Rule 415, also known as shelf
registration, which allowed large firms to register the total amount of securities it wishes to sell
over the subsequent two year period and sell portions of them whenever it chooses. This rule
became effective from November 1983. The SEC stipulated that companies that wish to register
their offerings under Rule 415 must satisfy the following conditions: (i) the aggregate market
value of the firm’s outstanding shares and unaffiliated voting stock should be greater than $150
million; (ii) the firm should not have defaulted on its preferred stock, debt, or rental obligations
in the last three years; (iii) the firm should have met all the disclosure requirements of the
Securities Exchange Act of 1934 during the last three years; and (iv) the firm’s bonds should
have investment grade rating.
Some researchers (see Bhagat et al., 1985; Kadapakkam and Kon, 1989 among others)
argue that shelf offerings hold several benefits for issuing firms. First, Bhagat et al. (1985)
contend that shelf offerings enable the issuer to time the offering dates to take advantage of
7
favourable market conditions. By aligning the issuance with demand for its shares, issuers would
be able to sell their equity at better prices. Second, the shelf offering process may intensify
competition between underwriters thereby lowering the issuance costs. Finally, shelf offerings
may lower the fixed costs associated with the SEC registration. Under traditional registration
process, there are large fixed cost components due to the SEC registration requirements. With a
shelf registration, the fixed costs of an offering are substantially reduced since it does not require
the preparation or distribution of detailed prospectuses. Empirical evidence supporting the view
that issuance costs are lower for shelf offerings is provided by Bhagat et al. (1985) and more
recently by Autore et al. (2008).
Investors interested in purchasing newly issued equity securities face an adverse selection
problem since managers have privileged access to information that outsiders do not have.
Therefore managers have incentives to issue new shares when the market overvalues them.
Typically, the market price of shares falls when a firm announces a new seasoned equity issue.
In this context, underwriters provide a valuable service by certifying the validity of the current
market price. Underwriters have incentives to correctly price each issue in order to preserve their
reputation.
Shelf registration exacerbates the information asymmetry problem faced by investors.
When a firm utilizes Rule 415, it is not obligated to appoint an underwriter until the date of
offering. In reality, once a shelf registration has been filed, investment bankers make competitive
bids for the issue. Investment bankers have incentives to charge lower fees in order to win the bid
while facing two impediments in performing their due diligence investigations. First, since there
is no guarantee that a given investment banking firm will be chosen to underwrite the offering, it
has no incentives to incur the costs of investigating the firm. Second, the investment banker will
8
have little time to conduct the due diligence investigation. This is because oftentimes an
underwriter is chosen on the same day that the offering is completed. Denis (1991) attributes
impediments to underwriter certification as the major factor behind the steep decline in shelf
offerings of industrial firms during the years 1984-1988.
However, more recently, Autore et al. (2008) document a resurgence in shelf offerings
during 1990-2003. They acknowledge the difficulties faced by investment bankers in conducting
due diligence of firms that use shelf registration. They suggest that firms that use shelf
registration mitigate the under-certification problem by using shelf offerings during periods
when there is less need for underwriter certification. These include periods following low
abnormal stock price runups, and after prior certification in previous seasoned offerings.
2.2 Accelerated Offers
Shelf offerings may be further classified into accelerated and traditional bookbuilt offers
based on the speed of issuance. Accelerated offerings may be completed quickly. There are two
types of accelerated offers – bought deals and accelerated bookbuilt offers. In a bought deal, the
issuing firm announces the amount of stock it wishes to sell and underwriters bid for these
shares. The underwriter that offers the highest net price wins the deal. The winning underwriter
typically resells the shares to institutional investors within the following 24 hours. Essentially,
bought deals are auctions to underwriters followed by open market sales. In accelerated
bookbuilt offers, investment banks submit proposals specifying a gross spread, but not
necessarily the offer price, for the right to underwrite the shares. The winning bank then forms a
syndicate and markets the shares to institutional investors.
Gao and Ritter (2010) examine the factors that determine issuing firms’ choice of using
accelerated offerings versus traditional fully marketed offerings. They find that accelerated
9
bookbuilt offerings are typically completed in a single day while fully marketed SEOs take, on
average, 31 days to complete. Also, accelerated offers have lower gross spreads than fully
marketed offers. Accelerated offers further have lower underpricing compared to fully marketed
offers.
From the underwriter’s point of view, in an accelerated offer, there is no time to conduct
an accurate due diligence analysis. Typically, there is far less information gathering and
marketing effort required from underwriters for accelerated offers. In order to win the deal in a
timely manner, the underwriter must quickly assess the market demand before committing an
offer price. Thus, only the largest and most capital intensive underwriters conduct accelerated
offers. The underwriting syndicates are smaller, take much more risk and generate comparable
revenues over shorter periods and effectively “buy” market share and league table rankings. For
the underwriters, insufficient due diligence raises the possibility of increased legal liabilities.
For example, in the case of WorldCom, which conducted two speedy shelf offers, investors sued
underwriters for negligence and were paid over $6 billion in settlement. (Schor, 2006)
Taken together, it appears that issuing firms will benefit significantly from using
accelerated offers rather than fully marketed offers. This raises an important issue. If accelerated
offers are highly beneficial compared to fully marketed offers, then why do issuing firms
continue to use the slower and more expensive method of issuing SEOs? Gao and Ritter (2010)
find evidence consistent with the view that issuing firms that have a more inelastic demand curve
for its shares tend to use fully marketed offers. Also firms making larger size issues use a fully
marketed offer.
2.3 Certification Mechanisms
10
In this paper, we utilize two certification mechanisms. First, we score a firm’s internal
corporate governance features and consider this as an internal certification device. Second, we
use audit fee paid by the firm as the second certification mechanism. The audit fee paid by a frim
can be considered as an external certification mechanism.
2.3.1 Internal Corporate Governance
Prior literature on the impact of a firm’s internal corporate governance on its valuation,
informativeness of stock prices, and liquidity motivates our selection of internal corporate
governance as a suitable certification mechanism for firms issuing SEOs. Durnev and Kim
(2005) hypothesize that firms with high quality governance are valued higher due to better
investor protection. Their empirical evidence is based on cross-country firm-level governance
data from 27 countries.
Ferreira and Laux (2007) argue that the absence of antitakeover provisions creates
incentives for private information collection. Their rationale follows Grossman and Stiglitz
(1980), who predict that enhancing the cost-benefit trade-off on information collection facilitates
more informed trading and more informative pricing. Furthermore, strong investor protection,
articulated by openness to takeovers is associated with a diminished likelihood of investors
expropriating outside investors. Therefore openness to takeovers has the potential to foster
uninformed ownership and trading, thereby offering more cover for privately informed trading.
Ferreira and Laux (2007) find strong supportive evidence linking a firm’s antitakeover
provisions to its informativeness as measured by idiosyncratic volatility. The more restrictive are
a firm’s antitakeover provisions the less informative is its stock price.
Further evidence on the link between a firm’s internal corporate governance and stock
market liquidity is provided by Chung et al. (2010). They conjecture that corporate governance
11
affects stock market liquidity because effective governance increases financial and operational
transparency which reduces informational asymmetries between insiders and outsiders. Further,
they posit that governance provisions could potentially improve financial transparency by
mitigating management’s ability and incentives to distort information disclosures. Specific
governance features could preclude managers from acting in their self-interest. Features such as
the use of independent audit committees serve to improve the quality of financial statements.7
Corporate governance features may also improve operational transparency by increasing the
ability of shareholders to discern the quality of management and the actual value of the firm.
Governance provisions strengthen the disciplinary threat of removing the management and
therefore limit the extent to which management can expropriate firm value through shirking,
empire building, overconsumption of perquisites, and risk aversion.8 Summing up, prior
literature provides abundant evidence that a firm’s internal corporate governance improves the
information environment of a firm and thereby reduces the information asymmetry between
insiders and outsiders.
2.3.2 Audit Fee
Both policy makers and academicians believe that the audit fee paid by a firm is
positively associated with the quality of financial reporting. In the words of Lynn Turner, former
chief accountant at the Securities and Exchange Commission:
7 See also Leuz et al. (2003), Ajinkya et al. (2005) and Karamanou and Vafeas (2005).
8 See also Babchuk and Cohen (2005) and Bebchuk et al. (2009).
12
Certainly throughout the 1980s and 1990s, corporations, sometimes with the
assistance of their audit committees, “twisted” the arms of independent auditors to
reduce their audit fees. Our experience includes corporations who competitively
bid their independent audit work solely to reduce their fees below levels that could
generate a reasonable return for their auditors. In turn, the audit firms reduced the
level of work they needed to perform in their role as gatekeepers for investors.
Inevitably inferior audits resulted.9
Empirical evidence in support of the view that lower audit fees are associated with lower audit
quality is provided by Blankley et al. (2012) who found that auditors of restatement firms charge
lower fees in the years prior to the announcement of restatements.
Ball et al. (2012) posit that managers seeking to disseminate private information need a
mechanism for credibly committing to be truthful. They propose that committing to the provision
of high quality audited financial statements is a potentially effective mechanism. Therefore, they
use the amount of excess audit fees paid by a firm as the proxy for the level of its financial
statement verification. They consider that audit service is a differentiated product that permits
clients to choose their auditor and several aspects of audit quality and effort.
Furthermore, audit fees are affected by the choice of audit firm (Big 4 versus others), the
level of seniority of the audit engagement partner, the number and hourly rate of audit personnel
deployed for the job, the degree of verification sought by the firm, and the frequency of
communication with the audit committee and other key executives. Audit fees are directly linked
to the level and price of audit activity and therefore to the degree of independent verification of
financial reporting.10
2.4 Hypotheses
Issuance Choice
We argue that a firm’s internal corporate governance arrangement could serve as a
potential certification device. Thus firms with high quality internal governance mechanisms may
9 See Turner (2005).
10 See also Simunic (1980) and Watts and Zimmerman (1983).
13
reduce the extent of due diligence performed by underwriters. Given the benefits of shelf
offerings as compared to non-shelf offerings, good governance firms will choose shelf offerings
over non-shelf offerings. Given our arguments in the previous subsection, we postulate that the
audit fee paid by a firm will serve as an external certification mechanism. We thus posit the
following hypothesis:
H1: Firms with better certification quality are more likely to choose shelf offerings while making
SEOs, other things being equal.
Given the risky nature of the accelerated offer method, it makes sense for underwriters to
use a certification device before accepting a deal. We suggest that firms may use their internal
corporate governance structure and/or audit fee as a potential certification device. In order to
protect their reputation and lower their risk, investment bankers will prefer to underwrite offers
made by high quality firms that certify their quality through internal corporate governance
structure and / or audit fee. Thus we posit:
H2: Firms with better certification quality are more likely to choose accelerated offerings while
making SEO, other things being equal.
Flotation Costs
A key variable of interest is flotation cost. When underwriters bid for an SEO issue, they
are compensated for their risk-bearing and marketing services. Underwriters typically buy an
SEO from the issuers at an offer price discount which is their compensation. In investment
banking parlance, this compensation is called the gross spread. The main components of gross
spread are a) management fee, b) underwriting fee and c) selling concession (see, Lee and
Masulis 2009). Typically, gross spread is scaled by offer price and expressed as a percentage.
Since gross spread includes a compensation for risk, it stands to reason that less risky firms
14
would have lower gross spreads. Also, a firm’s internal corporate governance setup is an
indicator of agency risks faced by the underwriter. Thus well-governed firms are expected to pay
lower gross spreads as compared to poorly governed firms. Audit fee is an external certification
device and firms paying higher audit fee are purchasing higher quality verification and
authentication services to ensure their financial reporting quality. We therefore expect that firms
paying higher audit fee will enjoy lower flotation costs as investment bankers are likely to be
cognizant of this aspect of a firm while conducting its SEO. We thus posit:
H3: Firms with better certification quality will have lower gross spreads while making SEOs,
other things being equal.
Net Proceeds
The net proceeds of an offer (offer size) could potentially be affected by asymmetric
information. Following the arguments of Myers and Majluf (1984), greater information
asymmetry increases the incentives of mangers to time equity offerings when their stock is
overvalued. This could lead to a negative relation between corporate governance quality and
issue proceeds. Alternately, underwriters could be reluctant to manage large issues of firms with
poor quality. This could lead to a positive relation between the quality of certification (internal
corporate governance or audit fee) and issue size. We prefer the second position as investment
bankers would recognise the risk to their reputation and would be cautious when they suspect
that the firm is overvalued or is of lower quality. We therefore posit:
H4: Firms with better certification quality will raise larger net proceeds while making SEOs,
other things being equal.
15
3. Data, Sample, and Measures of Governance Quality
3.1 Data and Sample Selection
In this paper, we rely on four sources of data. Equity issuance data are from Thomson Reuter’s
SDC database, raw corporate governance data are obtained from RiskMetrics11
, and accounting
data are extracted from Osiris database maintained by Bureau van Dijk. Finally, audit fee data is
sourced from Audit Analytics database. The sample selection process is detailed as follows. We
first merge all the equity issuances in the US market during 2001 through 2007 with annual
corporate governance data from RiskMetrics. Excluding all the equity issues made by firms
without corporate governance data, resulting in 5,751 issues. After issuances by firms in the
financial industry are further deleted, there are 5,225 issues.12
We then match these issues with
accounting data from Osiris. There are 3,933 issues by firms who have a record in Osiris.
Because we run three separate groups of regressions using three different dependent variables (a
dummy variable indicating shelf offers, a dummy variable for accelerated offers, and the gross
spread as a measure of flotation costs), to maximize the number of observations, we use three
different samples. All these three samples are obtained from the 3,933 issues and formed by
further requiring that data be available for all variables used in our regressions and that all issues
be secondary issues (IPOs and rights offers are excluded). The first sample, shelf offers sample,
has 1,250 issues, and is used to run regressions with the dummy, shelf (one for shelf-registration
and zero otherwise), as the dependent variable. The second sample, accelerated offers sample,
has 1,201 issues, and is used to run regressions using the dummy, accelerated (it equals one if the
11
We compile composite corporate governance measures used in our regressions (G1, G2, G3, G4, and CGI4) from
these raw data. 12
Based on SDC definition of industry, the following industries are excluded: commercial bank, credit institution,
insurance, investment bank, investment fund, other finance, and S&L/thrift.
16
issue is an accelerated issue and zero otherwise), as the dependent variable. The third sample,
gross spread sample, has 1,040 issues, and is the sample used to run regressions with gross
spread as the dependent variable.13
3.2 Firm Level Corporate Governance Quality
Our primary source of data on corporate governance aspects is from the RiskMetrics Corporate
Governance database. We first describe the sample of firms covered by the database provided by
Institutional Shareholder Services (ISS) of RiskMetrics. For each firm, for each governance
feature, RiskMetrics provides information regarding compliance or its absence. Next, we use
this information and based on best practices, we score each firm on each governance attribute
and construct our own governance ratings, aggregating these attributes. Our choice of this data
source is driven by its widespread use in academic work (Brown and Caylor, 2006; Chung et al.,
2010; Aggarwal et al., 2011).
RiskMetrics provides firm level Corporate Governance Quotient (CGQ) for companies from
2001. The CGQ rankings are designed to measure a firm’s internal governance as represented by
its adoption of governance attributes that increase the power of its minority shareholders. ISS
reports CGQ_Industry (hereafter IndustryCGQ) which gives a firm’s percentile ranking within
its GICS industry group In order to compute these indices, ISS collects information on a set of
governance attributes for a large number of companies. In this paper, we use corporate
governance data for the 2001-2007 period.
13
We use these three samples to run regressions that examine the determinants of issue methods and flotation costs.
For the sake of brevity and exposition, we rely on the largest sample, the shelf offers sample, to describe sample
summary statistics (Table 1), correlations between key variables of interest (Table 2), and univariate analysis (Table
3).
17
ISS compiles fifty-five governance attributes for each firm. A firm’s performance on each
attribute is determined by examining its regulatory filings, website and annual reports. Firms do
not pay to get rated but are allowed to access their ratings and verify the accuracy of ratings.
Firms can only change their ratings by altering the governance structure and publicly disclosing
it. ISS scores each firm on each attribute depending on whether it meets a threshold level of
acceptability. The fifty five attributes cover four broad categories: board composition and
effectiveness, anti-takeover provisions, compensation and ownership, and audit practices. The
board component of governance encapsulates the aspects of functioning of the board of directors
pertaining to board independence, size, composition of committees, transparency and the conduct
of work. Anti-takeover provisions include the firm’s charter and bylaws, dual-class structure,
role of shareholders, poison pill, and blank check preferred. The compensation and ownership
component deals with executive and director compensation issues, options, stock ownership and
loans. The audit component captures the independence of audit committee and the role of
auditors. Appendix A contains a list of the variables used and the acceptable standards used in
the scoring.
The rating provided by ISS (that is, IndustryCGQ) evaluates the strength, deficiencies, and
overall quality of a company’s corporate governance practices and is designed on the premise
that good corporate governance ultimately results in increased shareholder value. The exact
weighting of the different features of governance in computing the index is not available to us.
Also, ISS claims that CGQ is a “reliable tool for identifying portfolio risk related to governance
and leverages governance to drive increased shareholder value”. However, in addition to the
index scores provided by ISS, we also construct our own index using the raw data provided to us
18
by RiskMetrics. A clear benefit of constructing our own governance indicator is that we are able
to capture a wide variety of governance features employed by firms.
We create our own governance rating based on the raw data provided by ISS. The
overall score (CGI4) is the sum of the ratings for the four categories of governance practices
mentioned above. The first corporate governance rating, G1, represents board composition and
effectiveness. G2 is the rating that measures anti-takeover defences. G3 is the governance score
that measures managerial compensation and incentives. G4 is the rating that measures the
strength of audit practices. We score each firm based on whether or not it meets the threshold of
good governance for each attribute. Based on this binary coding of each attribute, we first
compute the G1, G2, G3, and G4 scores and then aggregate them to arrive at the overall score for
each firm, CGI4. In our study, we use G1, G2, G3, and G4 to pinpoint the specific governance
mechanisms that are at work. We also use CGI4, and IndustryCGQ, as proxies for the overall
governance quality.
3.3 External Certification
We argue that the certification of an issuer’s quality should be made by players in the equity
issue process, that is, by the issue firm (self-certification), or its underwriters, auditors, or
potential investors interested in the shares issued. Based on this reasoning, we use natural
logarithm of a firm’s total audit fee as a proxy for the certification from its auditors based on the
premise that higher audit fee indicates more effort and work done by the auditors and hence
19
stronger certification.14
Following Coulton et al. (2012), we use three-year average audit fees to
proxy for external certification.
3.4 Sample Characteristics
Table 1 provides descriptive statistics of the key variables used in this study.
Approximately 64% of the sample equity issues are shelf-registered offers, which is consistent
with the finding in Autore et.al (2008) that there has been a significant revival in the use of shelf
offering since the 1990s. A substantial portion, about 40%, of the sample issues are accelerated
offers.15
This confirms what Gao and Ritter (2010) point out that accelerated offers have gained
increasing popularity during the last decade. The average issue cost for our sample is around
4.3% of the total proceeds. Our four governance components ratings (G1 to G4) are all scaled to
have a maximum possible rating of 0.25, and hence the highest possible value for the overall
rating CGI4 is 1. IndustryCGQ, as discussed before, is the percentile ranking assigned by
RiskMetrics, and therefore have the maximum value of 100. Table 1 indicates that both our four
governance components ratings (G1 to G4) and the two overall governance ratings (CGI4,
IndustryCGQ) show a fair amount of variation across the sample firms.
In addition to the frequency of issue methods, issue costs, and governance quality, we
also provide other firm and issue characteristics that we think may affect firms’ choice of issue
methods and their respective costs. In terms of firm characteristics, an average sample firm has
net sales (used as a proxy for firm size) of US $1.35 billion. A median sample firm generates
about 5% earnings before interest and taxes (EBIT) from its total assets, experiences around 18%
14
Another proxy is the identity of the auditor, e.g., whether the auditor is one of the Big 5 accounting firms.
However, only less than 8% of our observations do not have a Big-5 auditor. We therefore do not include it in the
regressions. 15
The portion of accelerated offers in our sample is comparable to that in Gao and Ritter (2010) (42%) and Autore
et al. (2011) (43%).
20
growth in sales, and has a long-term debt to assets ratio of 0.28.16
Regarding issue
characteristics, an average sample equity issue raises US $ 171 million. Approximately 75% of
the shares issued are primary shares. In about 45% of the issues, stocks are listed on Nasdaq.
4. Empirical Results
4.1 Preliminary Results
In order to obtain a preliminary understanding between the relation between the quality of
the firm corporate governance and the decision to issue SEO using shelf/ accelerated methods,
we calculate the correlation coefficients between dummy variables for shelf and accelerated,
gross spread and several proxies for governance quality ratings. We present the coefficients and
the p-values for the test of significance in Table 2. The table shows that the choice of shelf offer
and the choice of accelerated offer are both positively correlated with all the governance quality
ratings except for G2 (anti-takeover). Further, the issue costs, as measured by gross spread, are
negatively correlated with all the governance quality ratings except for G2. All the reported
correlations are highly statistically significant. These findings are generally consistent with our
hypotheses that firms with better governance quality are more likely to use shelf-registered or
accelerated offers, and they enjoy lower financing costs. The only exception being the
correlation between G2 (the anti-takeover measure) and issue methods and issue costs that do not
show the expected sign. Another key finding reported in Table 2 is that issue costs are
significantly negatively correlated with shelf and accelerated offers. This is consistent with
Bhagat et al. (1985), who find that stocks sold through shelf offerings incur lower issue costs
than those sold through regular (non-shelf) offerings. Ours results are also consistent with Gao
16
We focus on the medians for accounting measures, because they are less affected by outliers.
21
and Ritter (2010) who report untabulated results indicating that gross spreads are lower for firms
making accelerated offers.
We then take a step further by utilizing our panel data and sorting all sample issues into
quintiles by our overall governance rating, CGI4. We examine the choice of issue methods, issue
costs, and other important firm and issue characteristics for each of the 5 portfolios (quintiles)
and report our results in Table 3. Panel A shows that as we move from Quintile 1, which contains
firms with the lowest governance quality, to Quintile 5, which contains firms with the highest
governance quality, there is a strict monotonicity in the frequency of shelf offers and accelerated
offers, as well as in the magnitude of issue costs. For instance, only 25.6% of firms in the lowest
governance quality quintile use accelerated offers increasing to 51.7% of firms in the highest
governance quality quintile. These findings provide initial support for hypothesis H2. Except for
firm size, which increases with governance quality, all the other firm and issue characteristics do
not show any clearly discernible patterns across governance quintiles.
We then formally test the significance of the monotonicity found in Panel A of Table 3.
In Panel B of Table 3, we calculate the difference in the frequency of shelf and accelerate offers
and in the issue costs between every 2 consecutive governance quintiles, and test the significance
of the differences. Panel B shows that all the differences in the frequency of shelf offers across
consecutive governance quintiles are statistically and economically significant. For example,
approximately 13% more equity issues are offered through shelf registration when we compare
Quintile 4 with Quintile 3. The differences in the frequency of accelerated offers demonstrate
very similar pattern, though the increase in the use of accelerated offering from Quintile 4 to
Quintile 5 firms is not statistically significant. As for the issue costs, the differences have the
expected sign, two of which are statistically significant. Overall, our univariate analyses as
22
reported in Table 3 support our hypotheses: firms with better governance quality are more likely
to conduct shelf or accelerated offers, and they also enjoy lower issue costs.
4.2 Regression Results
In this subsection, we investigate the relation between equity offer methods, issue costs, and
certification mechanisms controlling for firm and issue characteristics that may affect issue
methods. First, we examine the determinants of equity offer method by running the following
panel logit regressions:
titiproxiesioncertificat
tiNasdaq
tiprimary
tisizeissue
tileverage
tigrowth
tiROA
tiMarketCap
tidaccelerateor
tishelf
,,8,7,6,5
,4,3,2,1)
,(
,
(1)
The dependent variables are shelf and accelerated which are dummy variables that take the value
of one if an issue is shelf registered or accelerated offer, respectively, and zero otherwise. Firm
size is measured as the natural logarithm of net sales in thousands of US dollars at the last
financial year end before an offering17
. ROA is earnings before interest and taxes scaled by total
assets at the fiscal year end before an equity offering. Growth is the growth rate of net sales
during the year of the equity offering. Leverage is total long-term debt divided by total assets at
the financial year end before an equity offering. Issue size is the natural logarithm of the offer
proceeds in thousands of US dollars. Primary is the proportion of primary shares in the total
shares offered. Nasdaq is a dummy variable that equals one if a firm is listed on Nasdaq and zero
otherwise. The internal certification proxies included are each of the four governance component
ratings and the overall governance quality ranking. We use audit fee as an external certification
proxy. In order to ensure that our results are not distorted by outliers, we winsorize all variables
17
We also used alternate proxies for firm size such as total assets and market capitalization.
23
except for the dummies at the 1st and the 99
th percentiles. We also estimate the standard errors
through bootstrapping using 200 replications.
Table 4 presents the panel logit regression results for the determinants of shelf-registered
offerings. Our focus is on the relation between shelf offerings and certification mechanisms such
as internal governance quality. We find that all the governance quality ratings except for G2
(anti-takeover) are significantly positively related to the likelihood of a shelf offering. This is
consistent with the univariate results in Table 3 and supports our first hypothesis (H1), that is,
firms with better internal governance quality are more likely to use shelf-registered offerings
when conducting SEOs. Audit fee, our proxy for external certification, is highly positively
associated with the choice of using shelf registration. We also find that firms with higher
financial leverage and offerings which include a higher proportion of primary shares are more
likely to use shelf registered offerings. This indicates that firms utilizing shelf offerings are those
facing tighter financial conditions and having a greater need for external financing (Heron and
Lie, 2004). In addition, Nasdaq firms tend to use shelf offerings less often. To the extent that
Nasdaq firms are typically of higher risk and potentially more difficult to value, this finding
suggests that firms for whom certification by underwriters is more important rely less on shelf
offerings. We also find that larger firms are more likely to use shelf offerings. There is a negative
relationship between issue size and the choice of shelf registration. Firms raising higher proceeds
tend not to use shelf registration.
Results from panel logit regression analyses of the determinants of accelerated offerings
are reported in Table 5. The probability of conducting an accelerated offer increases with a
firm’s governance quality, which is evidenced by the positive and significant coefficient
estimates for all governance ratings except G2. Audit fee, the external certification proxy, is
24
significantly positively associated with the choice of accelerated offerings in four of the six
models used. This supports our second hypothesis (H2). We also find that larger firms are more
likely to use accelerated offering, which is consistent with Gao and Ritter (2010). Firms with
higher return on assets (ROA) are less likely to use accelerated offerings. The negative
relationship between ROA and the decision to use accelerated offering method is consistent with
the recommendation of Myers and Majluf (1984) who suggest that firms should build up a slack
when conditions are favourable. The positive coefficient estimate for leverage suggests that a
firm’s preference for an accelerated offering may be driven by its tight financial situation. We
also find that firms raising a larger amount of capital or listed on Nasdaq are less likely to use
accelerated offering. This is in line with the finding of Gao and Ritter (2010), as the marketing
service provided by underwriters during a traditional bookbuilt offer is more valuable for such
firms, they tend to use accelerated offerings less often.
In order to perform rigorous tests for the potential impact of certification quality on gross
spreads we need to control for the possibility of self-selection bias. Firms which self-select the
accelerated offerings methods for SEOs may have intrinsic qualities that may also affect the
gross spread. Li and Prabhala (2005) describe in detail the self-selection models commonly used
in corporate finance and provide the rationale for using them. Following Lee and Masulis (2009)
and Bethel and Krigman (2008), we examine the determinants of gross spreads paid by
estimating a Heckman two-stage model to control for the potential selection bias implicit in firms
choosing the accelerated method. The first stage logit estimation uses Model 5 of Table 5. Our
second stage estimation uses gross spread as the dependent variable in the following model:
titiproxiesioncertificat
tiimary
tiLeverage
tiGrowth
tiROA
tiMarketcap
tidGrosssprea
,,6,Pr
5
,4,3,2,1,
25
We control for firm size, ROA, leverage, growth and primary offerings. Finally, we include
lambda, the inverse mills ratio calculated by the first stage logit model, which is the predicted
probability that a firm uses accelerated method.
Our results are shown in Table 6. We find that after controlling for self-selection, that a
firm’s internal corporate governance has a weak effect on gross spread. However, audit fee, our
external certification variable, has a significant negative effect on gross spread. The panel
regression results show that firms which are more profitable enjoy significantly lower issue
costs. We also find that firms that sell a higher proportion of primary shares bear higher issue
costs, which may be because they require more marketing efforts from the underwriters. Our
empirical results support our third hypothesis (H3). We find that lambda has a reliable positive
impact on gross spread. This finding implies that the decision to select accelerate offerings
method has a positive impact on gross spread after controlling for the self-selection problem.
We investigate the determinants of net proceeds by estimating a Heckman two-stage
model to control for the potential selection bias implicit in firms choosing the accelerated
method. The first stage logit estimation uses Model 5 of Table 5. Our second stage estimation
uses net proceeds (issue size) as the dependent variable in the following model:
titiproxiesioncertificat
tiimary
tiLeverage
tiGrowth
tiROA
tiMarketcap
tisNetproceed
,,6,Pr
5
,4,3,2,1,
We control for firm size, ROA, leverage, growth and primary offerings. Finally, we include
lambda, the inverse mills ratio calculated by the first stage logit model, which is the predicted
probability that a firm uses accelerated method.
Our empirical results, which are reported in Table 7, indicate strong empirical support for
a positive relation between external certification and issue size. Audit fee is strongly positively
26
associated with net proceeds. Thus firms which face a smaller adverse selection risk are able to
make larger size SEO issues. Internal certification mechanism as proxied by good quality
corporate governance has a weak effect on net proceeds. Firms which have higher ROA are able
to raise more issue proceeds consistent with the view that the market and investment bankers are
less concerned about the misuse of free cash flows when firms have a good track record of
profitability. Market capitalization has a significant positive impact on net proceeds implying
that large firms suffer less from information asymmetry and are therefore able to raise more
money from SEOs. The coefficient of lambda is positive and significant at the one per cent level
indicating that even after controlling for firm characteristics that drive the choice of accelerated
offerings, firms using accelerated offerings raise higher proceeds than firms using non-
accelerated methods. Our empirical evidence supports our fourth hypothesis (H4).
Overall, our results show that both internal and certification devices play a role in
affecting key aspects of a firm’s SEO issuance. Both certification methods (internal and external)
influence the choice of registration method (shelf versus non-shelf) and issuance method
(accelerated versus non-accelerated). However, only the external certification device (audit fee)
is associated significantly with flotation costs and issue size. This evidence suggests that
investment bankers lend greater credibility to external certification provided by audit fee in
setting their gross spread and issue size.
4.3 Robustness Checks
We check the robustness of our empirical findings in two ways. First, we use alternate
proxies for certification. For internal certification, we use Industry CGQ as an alternate proxy.
Our results reported in Tables 4 through 7 show that we get qualitatively similar results in
issuance choice, issuance cost and issuance size models. As an alternate proxy for external
27
certification (the certification by potential investors), we compare the file offer price with the
initial offer price range and use two dummy variables, below range and above range, to capture
the situations where the final offer price is below or above the initial range, respectively.
Because the difference between the final offer price and the initial price range reflects the
deviation of the potential investors’ assessment of the equity offering from the issuer’s own
assessment, we argue that the larger the difference is, the weaker is the certification from
potential investors. Our empirical evidence supports the validity of this alternate certification
proxy in issuance choice models reported in Tables 4 and 5.
Second, we use alternate proxies for firm size. We use net sales, total assets, and market
capitalization to proxy for size in our tests. Our empirical results are qualitatively similar. We do
not tabulate these results in order to conserve space.
5. Conclusions
Recent work has documented the increased use of accelerated offers in SEOs in the US.
For issuers, the reduction in time taken to issue and the lower issuance costs in the accelerated
offers method compared to fully marketed offers are attractive features. However, the issuance
of new capital via SEOs is fraught with the information asymmetry problem between issuers and
investors. Underwriters typically use due diligence to reduce information asymmetry in order to
sell the issue to potential investors. The reduction in issuance time in accelerated offers affects
the conduct of due diligence. How do high quality issuers differentiate themselves from low
quality issuers in order to issue equity via accelerated offers?
We contribute to the emerging literature by examining this crucial issue. Given the fact
that firms have been paying increasing attention to corporate governance over the last two
decades, we hypothesize that internal corporate governance may serve as a certification device
28
and allow issuers with strong governance to overcome the under certification problem and take
advantage of shelf and accelerated offerings. We also posit that an external certification
mechanism such as the audit fee paid by the firm may provide a credible means of overcoming
the under certification problem in the context of SEO issuance.We postulate that firms may use
internal and external certification mechanisms to signal lower asymmetry.
Our empirical results confirm our conjecture and indicate that firms with higher
certification quality are more likely to use the accelerated offers method while issuing SEOs.
Using a panel dataset of US SEOs, we find evidence suggesting that these internal and external
certification mechanisms perform three significant roles in the issuance of SEOs. First, by
reducing information asymmetry, certification mechanisms influence the choice between
accelerated offerings and fully marketed offers. Second, we find that certification mechanisms
help to reduce the issuance costs as measured by gross spread. Finally, we find that there exists
an association between the efficacy of certification and the net proceeds of SEO issuance. Our
paper contributes to the securities issuance and corporate governance literature by proposing a
linkage between governance quality and the choice of securities issue techniques and providing
empirical evidence that supports the existence of the linkage.
29
References
Aggarwal, R., Erel, I., Ferreira, M., Matos, P., (2011). Does governance travel around the world?
Evidence from institutional investors. Journal of Financial Economics 100, 154–181.
Ajinkya, B.; S. Bhojraj; and P. Sengupta (2005). “The Association between Outside Directors,
Institutional Investors and the Properties of Management Earnings Forecasts.” Journal of
Accounting Research, 43, 343–376.
Autore, D. M., Hutton, I., & Kovacs, T. (2011). Accelerated equity offers and firm quality.
European Financial Management 17(5), 835–859.
Autore, D. M., Kumar, R., & Shome, D. K. (2008). The revival of shelf-registered corporate
equity offerings. Journal of Corporate Finance 14, 32–50.
Ball, R., S. Jayaraman and L. Shivakumar. (2012). Audited financial reporting and voluntary
disclosure as complements: A test of the confirmation hypothesis. Journal of Accounting and
Economics 53(1): 136-166.
Bebchuk, L. A., and A. Cohen (2005). “The Costs of Entrenched Boards.” Journal of Financial
Economics, 78, 409–433.
Bebchuk, L. A.; A. Cohen, and A. Ferrell (2009). “What Matters in Corporate Governance?”
Review of Financial Studies, 22, 783–827.
Bethel, J.E., & Krigman, L., (2008). Managing the costs of issuing common equity: The role of
registration choice. Quarterly Journal of Finance and Accounting 47(4), 57-85.
Bhagat, S., Marr, M., & Thompson, G. (1985). The rule 415 experiment: equity markets. Journal
of Finance 40, 1385–1401.
Blackwell, D. W., Marr, M. W., & Spivey, M. F. (1990). Shelf registration and the reduced due
diligence argument: Implications of the underwriter certification and the implicit insurance
hypotheses. Journal of Financial and Quantitative Analysis 25, 245-259.
Blankley, A., D. Hurtt and J. McGregor. (2012) Abnormal audit fees and restatements. Auditing: A
Journal of Practice and Theory 31 (1): 79-96.
Bortolotti,B., Megginson,W., Smart,S., (2008). The rise of accelerated seasoned equity
underwritings. Journal of Applied Corporate Finance 20 (3), 35–57.
Brown, L., & Caylor, M. (2006). Corporate Governance and Firm Performance. Journal of
Accounting and Public Policy 25, 409-434..
Chung, K. H., Elder, J., & Kim, J-C. (2010). Corporate governance and liquidity. Journal of
Financial and Quantitative Analysis 45, 265-291.
30
Coulton, J.,Livne, G., Pettinicchio, A., and S. Taylor. (2012). Audit fees and accounting quality:
Inferences from single-period versus multi-period perspectives. Working paper, The University of
Technology, Sydney, Australia.
Denis, D. (1991). Shelf registrations and the market for seasoned equity offerings. Journal of
Business 64, 189–212.
Doidge, C., Karolyi, G. A., & Stulz, R.M. (2007). Why Do Countries Matter So Much For
Corporate Governance. Journal of Financial Economics 86, 1-39.
Durnev, A. & Kim, E.H., (2005) To Steal or Not to Steal: Firm Attributes, Legal Environment,
and Valuation. The Journal of Finance 60(3), 1461-1493.
Ferreira, M., Laux, P., 2007. Corporate governance, idiosyncratic risk, and information flow.
Journal of Finance 62, 951–989.
Gao, X., & Ritter, J.R. (2010). The marketing of seasoned equity offerings. Journal of Financial
Economics 97, 33–52.
Grossman, Sanford, and Joseph Stiglitz, (1980). On the impossibility of informationally efficient
markets, American Economic Review 25, 393–408.
Heron, R. A., & Lie, E. (2004). A comparison of the motivations for and the information content
of different types of equity offerings. Journal of Business 77, 605-632.
Jensen, M. C. & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs
and ownership structure. Journal of Financial Economics 3, 305–360.
Kadapakkam, P. & Kon, S. J. (1989). The value of shelf registration for new debt issues. Journal
of Business 62, 271-292.
Karamanou, I., and N. Vafeas. “The Association between Corporate Boards, Audit Committees,
and Management Earnings Forecasts: An Empirical Analysis.” Journal of Accounting Research,
43(2005), 453–486.
Lee, G., & R. W. Masulis (2009). Seasoned equity offerings: Quality of accounting information
and expected flotation costs. Journal of Financial Economics 92(3), 443-469.
Leuz, C.; D. Nanda; and P.Wysocki. “Earnings Management and Investor Protection: An
International Comparison.” Journal of Financial Economics, 69 (2003), 505–527.
Li K., & Prabhala, N.R. (2005). Self-selection models in corporate finance. Working paper,
University of Maryland.
Mande, V., Park, YK., & Son, M. (2012). Equity or Debt Financing: Does Good Corporate
Governance Matter?. Corporate Governance: An International Review 20(2), 195-211.
31
Morgenson, G., (2004). Judge in WorldCom action sides with plaintiffs on issue of due diligence
by banks. The New York Times, December 16, 2004.
Myers, S., Majluf, N. (1984). Corporate financing and investment decisions when firms have
information that investors do not have. Journal of Financial Economics 13, 187–221.
Schor, G.A., (2006). The Due Diligence and Reliance Defenses in WorldCom: Retrospect and
Prospect. Working paper
Sherman, Ann, (1999). Underwriter certification and the effect of shelf registration on due
diligence, Financial Management 28, 5–19.
Simunic, D. A. (1980). The pricing of audit services: Theory and evidence. Journal of
Accounting Research 18 (1): 161–190.
Turner, L. (2005). Comment letter to the Securities and Exchange Commission. April 12.
Available at: http://www.sec.gov/spotlight/soxcomp/soxcomp-turner.pdf
Watts, .L., Zimmerman, J.L., (1983). Agency problems, auditing, and the theory of the firm:
some evidence. Journal of Law and Economics 26, 613–633.
32
Appendix A
CGQ Rating Variables Summary
Acceptable Governance Standards
Board Composition and Effectiveness (G1) 1. All directors attended 75% of board meetings or had a valid excuse
2. CEO serves on the boards of two or fewer public companies
3. Board is controlled by more than 50% independent outside directors
4. Board size is greater than 5 but less than 16
5. CEO is not listed as having a related-party transaction
6. No former CEO on the board
7. Compensation committee composed solely of independent outsiders
8. Chairman and CEO are separated or there is a lead director
9. Nominating committee composed solely of independent outsiders
10. Governance committee exists and met in the past year
11. Shareholders vote on directors selected to fill vacancies
12. Governance guidelines are publicly disclosed
13. Annually elected board (no staggered board)
14. Policy exists on outside directorships (four or fewer boards is the limit)
15. Shareholders have cumulative voting rights
16. Shareholder approval is required to increase/decrease board size
17. Majority vote requirement to amend charter/bylaws (not supermajority)
18. Board has the express authority to hire its own advisors
19. Performance of the board is reviewed regularly
20. Board-approved succession plan in place for the CEO
21. Directors are required to submit resignation upon a change in job
22. Board cannot amend bylaws without shareholder approval or can do so only under limited circumstances
23. Does not ignore shareholder proposal.
24. Company has policy on mandatory retirement age or term limits for directors
25. All board members participate in accredited director education programs. Anti-takeover (G2) 1. Single class, common
2. Majority vote requirement to approve mergers (not supermajority)
3. Shareholders may call special meetings
4. Shareholder may act by written consent
5. Company either has no poison pill or a pill that was shareholder approved
6. Company is not authorized to issue blank check preferred Compensation and Ownership (G3) 1. Directors are subject to stock ownership requirements
2. Executives are subject to stock ownership guidelines
3. No interlocks among compensation committee members
4. Directors receive all or a portion of their fees in stock
5. All stock-incentive plans adopted with shareholder approval
6. Options grants align with company performance and reasonable burn rate
7. Company expenses stock options
8. All directors with more than one year of service own stock
9. Officers’ and directors’ stock ownership is at least 1% but not over 30% of total shares outstanding
10. Repricing is prohibited
11. An option pricing model is used to measure the cost of all stock-based incentive plans.
12. Non-employee directors should not participate in pension plans
13. Corporate loans should not be given to participants of stock option plans. Audit Practices (G4) 1. Consulting fees should be less than audit fees.
2. Shareholders should be permitted to ratify management’s selection of auditors each year.
3. The entire audit committee is composed of independent directors.
4. The entire audit committee should be composed of financial experts.
33
Table 1. Summary statistics
Variable Mean Std. Dev. Minimum 1st Quartile Median 3
rd Quartile Maximum
shelf 0.635 0.482 0.000 0.000 1.000 1.000 1.000
accelerated 0.399 0.490 0.000 0.000 0.000 1.000 1.000
spread 4.264 1.659 0.106 3.500 4.850 5.476 10.693
G1 0.147 0.049 0.000 0.113 0.150 0.188 0.250
G2 0.142 0.036 0.026 0.120 0.145 0.169 0.226
G3 0.144 0.037 0.023 0.114 0.136 0.159 0.250
G4 0.142 0.056 0.028 0.083 0.139 0.194 0.250
CGI4 0.576 0.121 0.261 0.482 0.579 0.667 0.853
IndustryCGQ 54.478 26.311 0.500 33.600 55.000 75.750 100.000
firm size 1,351,474 4,499,174 22 81,484 286,915 861,916 113,000,000
roa -0.017 0.245 -1.891 -0.006 0.054 0.089 0.637
growth 0.441 2.474 -1.000 0.052 0.181 0.403 74.581
leverage 0.291 0.246 0.000 0.040 0.280 0.457 1.680
issue size 171,000 273,000 730 57,800 100,000 180,000 4,180,000
primary 0.751 0.397 0.000 0.543 1.000 1.000 1.000
Nasdaq 0.451 0.498 0.000 0.000 0.000 1.000 1.000
This table presents the summary statistics for our sample firms over 2001 to 2007. Shelf is a dummy variable that equals one if an
equity offer is a shelf-registered offering and zero otherwise. Accelerated is a dummy variable that equals one if an equity offer is an
accelerated offer and zero otherwise. Spread is the gross spread as a percentage of the principal amount offered. G1, G2, G3, and G4
are ratings of governance quality concerning board of directors, anti-takeover provisions, executive and director ownership and
compensation, and audit practices and other progressive practices, respectively. CGI4 is the sum of G1, G2, G3, and G4. These ratings
are compiled by ourselves using raw data provided by RiskMetrics. IndustryCGQ which is compiled by RiskMetrics, is the percentile
rankings of the governance quality for a firm vis-à-vis its industry group. Higher governance rankings indicate better governance
quality. Firm size is measured by net sales at the last financial year end before an equity offering in thousands of US dollars. ROA is
earnings before interest and taxes scaled by total assets at the last financial year end before an equity offering. Growth is the growth
rate of net sales during the year of the equity offering. Leverage is total long-term debt divided by total assets at the last financial year
end before an equity offering. Issue size is the offering proceeds in thousands of US dollars. Primary is the proportion of primary
shares in the total shares offered. Nasdaq is a dummy variable that equals one if a firm is listed on Nasdaq and zero otherwise.
34
Table 2. Correlations between equity offering methods and governance quality ratings
shelf accelerated spread G1 G2 G3 G4 CGI4 IndustryCGQ
shelf 1.000
accelerated 0.482 1.000
(0.000)
spread -0.311 -0.491 1.000
(0.000) (0.000)
G1 0.238 0.162 -0.117 1.000
(0.000) (0.000) (0.000)
G2 -0.071 -0.067 0.069 0.023 1.000
(0.009) (0.014) (0.015) (0.408)
G3 0.277 0.204 -0.237 0.398 -0.099 1.000
(0.000) (0.000) (0.000) (0.000) (0.000)
G4 0.309 0.220 -0.240 0.453 0.108 0.544 1.000
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
CGI4 0.303 0.210 -0.211 0.742 0.328 0.690 0.843 1.000
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
IndustryCGQ 0.155 0.141 -0.143 0.533 0.099 0.550 0.344 0.574 1.000
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
This table shows the pairwise correlation coefficients between equity offering methods and governance quality ratings. Numbers in the
brackets indicate the p-values for the test that a correlation coefficient is zero. All variables are as defined in Table 1.
35
Table 3. Choice of equity issuance methods and governance quality: Univariate analysis
Panel A
Quintile CGI4 Shelf Accelerated Spread Firm Size ROA Growth Leverage Issue Size Primary Nasdaq
1 0.405 0.433 0.256 4.684 785,239 0.067 0.144 0.263 136,600 0.744 0.511
2 0.502 0.522 0.311 4.492 1,054,828 0.057 0.191 0.260 135,300 0.754 0.544
3 0.580 0.641 0.404 4.391 1,225,436 0.049 0.195 0.256 141,700 0.758 0.511
4 0.652 0.767 0.507 3.937 1,289,779 0.053 0.215 0.307 209,300 0.734 0.367
5 0.743 0.814 0.517 3.750 2,405,994 0.047 0.160 0.307 230,400 0.766 0.323
Panel B
2-1 0.089**
(0.019)
0.056*
(0.076)
-0.192*
(0.060)
-269,589
(0.130)
3-2 0.119***
(0.003)
0.093**
(0.012)
-0.101
(0.240)
-170,608
(0.290)
4-3 0.126***
(0.001)
0.104***
(0.008)
-0.454***
(0.002)
-64,343
(0.407)
5-4 0.047*
(0.088)
0.009
(0.414)
-0.187
(0.125)
-1,116,215**
(0.015)
Firm-years are assigned to quintiles based on their CGI4 ratings. Panel A presents the mean value for issuance variables (shelf,
accelerated, spread, issue size, and primary) and firm characteristics (firm size and Nasdaq) for each quintile, and the median of ROA,
Growth, and Leverage for each quintile. Panel B tests the difference in mean shelf, accelerated, spread, and firm size between
consecutive quintiles. For shelf and accelerated, one-sided proportion tests are used. For spread and firm size, one-sided t-tests are
used. All variables are as defined in Table 1. Numbers in the brackets are the p-values. One, two, and three asterisks indicate
significance at the 10%, 5%, and 1% level, respectively.
36
Table 4. Determinants of shelf-registered offerings: Internal and external certification
1 2 3 4 5 6
Constant -4.532*** (0.000)
-4.664*** (0.000)
-4.947*** (0.000)
-4.375***
(0.000)
-4.949***
(0.000)
-4.615***
(0.000)
Market cap 0.378*** (0.000)
0.372*** (0.000)
0.375*** (0.000)
0.387*** (0.000)
0.401*** (0.000)
-0.371*** (0.000)
ROA -0.927*** (0.000)
-0.990*** (0.000)
-0.968*** (0.000)
-1.05*** (0.000)
-0.967*** (0.000)
-0.993*** (0.000)
Growth 0.079 (0.277)
0.081 (0.267)
0.070 (0.327)
0.039 (0.585)
0.047 (0.488)
0.075 (0.330)
Leverage 0.907*** (0.000)
0.836*** (0.000)
0.894*** (0.000)
0.965*** (0.000)
1.004*** (0.000)
0.855*** (0.000)
Issue Size -0.232*** (0.006)
0.232*** (0.007)
-0.223*** (0.007)
-0.215** (0.016)
-0.229*** (0.008)
-0.240*** (0.003)
Primary 0.643*** (0.000)
0.709*** (0.000)
0.653*** (0.000)
0.690*** (0.000)
0.678*** (0.000)
0.702*** (0.000)
Nasdaq -0.319*** (0.003)
-0.313*** (0.004)
-0.257** (0.019)
-0.254** (0.020)
-0.300*** (0.006)
-0.311*** (0.004)
Below Range -1.535*** (0.000)
-1.554*** (0.000)
-1.474*** (0.000)
-1.495*** (0.000)
-1.470*** (0.000)
-1.530*** (0.000)
Above Range -1.385*** (0.000)
-1.447*** (0.000)
-1.360*** (0.000)
-1.366*** (0.000)
-1.316*** (0.000)
-1.425*** (0.000)
Audit Fee 0.137** (0.017)
0.200*** (0.001)
0.132** (0.029)
0.058 (0.360)
0.048 (0.435)
0.192*** (0.001)
G1 3.750*** (0.000)
G2 -0.326 (0.815)
G3 6.059*** (0.000)
G4 6.357*** (0.000)
CGI4 2.682*** (0.000)
IndustryCGQ 0.003 (0.10)
Log likelihood -487 -494 -485 -473 -477 -493
No. of Obs. 1040 1040 1040 1040 1040 1040
The sample period is 2001 to 2007. The dependent variable is shelf. Market Cap is the
Market Capitalization in US dollars at the last financial year end before an offering. The
coefficients of market cap are multiplied by 10-7
for convenience. Issue size is the natural
logarithm of total proceeds in US dollars. Below range and above range are dummy variables
that equal one if the issue price is below or above the filing price range, respectively, and
zero otherwise. Audit fee is the natural logarithm of a firm’s three-year average total audit fee
in US dollars as of the financial year end before an offering. All the other variables are as
defined in Table 1. All except for the dummy variables are winsorized at the 1st and the 99
th
percentiles. Numbers in the brackets are the p-values. Standard errors are based on the
bootstrapping method using 500 replications. One, two, and three asterisks indicate
significance at the 10%, 5%, and 1% level, respectively.
37
Table 5. Determinants of accelerated offerings: Internal and external certification methods
1 2 3 4 5 6
Constant 2.435*
(0.068)
2.328*
(0.062)
2.262*
(0.069)
2.884**
(0.024)
2.461*
(0.056)
2.274*
(0.073)
Market Cap 0.547** (0.024)
0.508** (0.036)
0.533** (0.015)
0.591** (0.011)
0.583** (0.015)
0.503** (0.038)
ROA -0.602** (0.011)
-0.642*** (0.003)
-0.644*** (0.005)
-0.668*** (0.004)
-0.635*** (0.005)
-0.648*** (0.003)
Growth 0.026 (0.703)
0.031 (0.643)
0.017 (0.790)
-0.003 (0.966)
0.003 (0.964)
0.024 (0.730)
Leverage 0.915*** (0.000)
0.887*** (0.000)
0.907*** (0.000)
0.949*** (0.000)
0.950*** (0.000)
0.896*** (0.000)
Issue Size -0.208*** (0.006)
-0.207*** (0.003)
-0.209*** (0.003)
-0.213** (0.003)
-0.210*** (0.003)
-0.212*** (0.003)
Primary 0.136 (0.324)
0.175 (0.211)
0.145 (0.283)
0.149 (0.253)
0.147 (0.298)
0.178 (0.186)
Nasdaq -0.451*** (0.000)
-0.436*** (0.000)
-0.419*** (0.000)
-0.405*** (0.000)
-0.442*** (0.000)
-0.444*** (0.000)
Below Range -1.719*** (0.000)
-1.741*** (0.000)
-1.700*** (0.000)
-1.691*** (0.000)
-1.683*** (0.000)
-1.719*** (0.000)
Above Range -1.438*** (0.000)
-1.490*** (0.000)
-1.437*** (0.000)
-1.400*** (0.000)
-1.391*** (0.000)
-1.466*** (0.000)
Audit Fee 0.139*** (0.008)
0.182*** (0.001)
0.147*** (0.008)
0.091 (0.117)
0.096* (0.089)
0.172*** (0.001)
G1 2.456** (0.014)
G2 -0.871 (0.494)
G3 3.030** (0.021)
G4 3.992*** (0.000)
CGI4 1.541*** (0.000)
IndustryCGQ 0.003 (0.119)
Log likelihood -531 -534 -531 -524 -527 -622
No. of Obs. 1040 1040 1040 1040 1040 1040
The sample period is 2001 to 2007. The dependent variable is accelerated. Market Cap is the
Market Capitalization in US dollars at the last financial year end before an offering. The
coefficients of market cap are multiplied by 10-7
for convenience. Issue size is the natural
logarithm of total proceeds in US dollars. Below range and above range are dummy variables
that equal one if the issue price is below or above the filing price range, respectively, and
zero otherwise. Audit fee is the natural logarithm of a firm’s three-year average total audit fee
in US dollars as of the financial year end before an offering. All the other variables are as
defined in Table 1. All variables except for the dummy variables are winsorized at the 1st and
the 99th
percentiles. Numbers in the brackets are the p-values. Standard errors are based on
the bootstrapping method using 500 replications. One, two, and three asterisks indicate
significance at the 10%, 5%, and 1% level, respectively.
38
Table 6. Determinants of gross spread: Heckman two-stage model estimations
1 2 3 4 5 6
Constant 3.889*** 3.844*** 3.970*** 4.131*** 3.76*** 3.830***
(0.001) (0.000) (0.001) (0.001) (0.002) (0.002)
Market Cap -1.11*** -1.15*** -1.16*** -1.15*** -1.10*** -1.17***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
ROA -1.727*** -1.753*** -1.789*** -1.884*** -1.847*** -1.814***
(0.001) (0.000) (0.000) (0.000) (0.000) (0.001)
Growth 0.181* 0.171 0.182 0.160 0.151 0.186*
(0.094) (0.131) (0.112) (0.165) (0.179) (0.094)
Leverage -0.403 -0.417 -0.458 -0.392 -0.354 -0.445
(0.352) (0.312) (0.288) (0.395) (0.409) (0.297)
Primary 0.844*** 0.997*** 0.919*** 0.888*** 0.895*** 0.911***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Audit Fee -0.216** -0.190** -0.182** -0.225** -0.243** -0.185**
(0.015) (0.042) (0.049) (0.019) (0.013) (0.042)
G1 4.01**
(0.030)
G2
1.966
(0.328)
G3
-0.703
(0.770)
G4
2.955*
(0.083)
CGI4
1.706*
(0.056)
IndustryCGQ
0.005
(0.138)
Lambda 1.693*** 1.622*** 1.624*** 1.706*** 1.737*** 1.666***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) R-squared 0.445 0.446 0.450 0.447 0.446 0.444
No. of Obs. 1040 1040 1040 1040 1040 1040
The sample period is 2001 to 2007. The dependent variable is spread. Market cap is in
thousands of US dollars and is measured at the financial year end before an offering. The
coefficients of market cap are multiplied by 10-7
for convenience. Audit fee is the natural
logarithm of a firm’s three-year average total audit fee in US dollars as of the financial year
end before an offering. All the other variables are as defined in Table 1. All variables except
for the dummy variables are winsorized at the 1st and the 99
th percentiles. Numbers in the
brackets are the p-values. Standard errors are based on the bootstrapping method using 500
replications. One, two, and three asterisks indicate significance at the 10%, 5%, and 1% level,
respectively.
39
Table 7. Determinants of net proceeds: Heckman two-stage model estimations
1 2 3 4 5 6
Constant 15.087*** 15.084*** 14.978*** 15.153*** 15.005*** 15.054***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Market Cap 1.79*** 1.79*** 1.81*** 1.80*** 1.82*** 1.79***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
ROA 0.363** 0.363* 0.319* 0.300* 0.337* 0.355**
(0.048) (0.052) (0.079) (0.099) (0.069) (0.050)
Growth 0.093** 0.092* 0.087** 0.079 0.081* 0.092**
(0.046) (0.064) (0.035) (0.102) (0.097) (0.043)
Leverage 0.538*** 0.539*** 0.543*** 0.576*** 0.574*** 0.541***
(0.001) (0.000) (0.000) (0.000) (0.000) (0.000)
Primary -0.052 -0.049 -0.081 -0.075 -0.063 -0.055
(0.627) (0.669) (0.476) (0.481) (0.543) (0.589)
Audit Fee 0.188*** 0.188*** 0.177*** 0.162*** 0.166*** 0.187***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
G1 0.034
(0.962)
G2
0.057
(0.945)
G3
1.683
(0.113)
G4
1.658**
(0.027)
CGI4
0.579*
(0.098)
IndustryCGQ
0.001
(0.557)
Lambda 0.456*** 0.456*** 0.484** 0.508*** 0.498*** 0.465***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
R-squared 0.465 0.496 0.457 0.437 0.486 0.464
No. of Obs. 1040 1040 1040 1040 1040 1040
The sample period is 2001 to 2007. The dependent variable is net proceeds. Market cap is in
thousands of US dollars and is measured at the financial year end before an offering. The
coefficients of market cap are multiplied by 10-7
for convenience. Audit fee is the natural
logarithm of a firm’s three-year average total audit fee in US dollars as of the financial year
end before an offering. All the other variables are as defined in Table 1. All variables except
for the dummy variables are winsorized at the 1st and the 99
th percentiles. Numbers in the
brackets are the p-values. . Standard errors are based on the bootstrapping method using 500
replications. One, two, and three asterisks indicate significance at the 10%, 5%, and 1% level,
respectively.