-
The Effect of Regulator Oversight on Firms’ Information
Environment:
Securities and Exchange Commission Comment Letters
Reining Chen
MIT Sloan School of Management Cambridge, MA 02142
[email protected]
Rick Johnston Krannert School of Management
Purdue University West Lafayette, IN 47907
[email protected]
July 9, 2010
Abstract:
We explore the content and determinants of Securities and
Exchange Commission (SEC) comment letters and then examine whether
letter resolution affects both the firm’s information environment
and that of its peers. Our content analysis confirms that the
comment emphasis is disclosure. If the resolution of an enquiry
improves disclosure thereby enhancing the preannouncement
information environment, we would expect dampened market reactions
around ensuing earnings announcements. Our results support the
hypothesis showing reduced return volatility and trading volume
after the comment letter for targeted firms, and where SEC scrutiny
is intense, reduced return reactions for industry peers. This
spillover effect is consistent with our determinants model, which
shows that the probability of receiving a letter is higher for
industry leaders suggesting that peers mimic industry leaders. We
conclude the SEC’s oversight has both direct and indirect
effects.
JEL Classification: G12, G14, G18, M48
Key Words: Securities and Exchange Commission (SEC), Comment
Letter, Disclosure, Enforcement, Regulation
This paper is a revision of a previous manuscript titled
“Securities and Exchange Commission Comment Letters: Enforcing
Accounting Quality and Disclosure”. We appreciate the helpful
comments and suggestions from Andrew Karolyi, Sundaresh Ramnath,
Cathy Schrand, Ro Verrecchia, and workshop participants at the 2009
AAA Financial Accounting and Reporting Section meeting and the
annual meeting, the 2010 LBS conference and the following
universities: City University London, MIT, National University of
Singapore, The Ohio State University, University of Pennsylvania,
and University of Technology Sydney. We thank Anthony Meder and
Yunyan Zhang for data assistance. Johnston thanks The Wharton
School, University of Pennsylvania and The Ohio State University
for financial support.
-
1
1. Introduction
In this paper we investigate whether regulatory oversight of
financial reporting affects
companies’ information environment. We address the question by
examining the content,
determinants, and ensuing market consequences of the Securities
and Exchange Commission’s
(SEC) comment letters. The SEC reviews a significant portion of
the filings (10Qs, 10Ks, S1-4
etc.) submitted to them. If the SEC staff identify potential
deficiencies, they send a comment
letter to the company seeking clarification, more information,
or revision of the filing or future
filings. Unless the company voluntarily discloses it, the SEC
enquiry is unknown to the public
during its existence. In 2005, the SEC began to publicly release
comment letters of resolved
cases. 1 These letters provide a unique opportunity to
investigate the monitoring role of the SEC
in the U.S. capital market.
The stated goal of the SEC review process is to improve the
quality of material disclosure
to investors in a timely manner. However, whether the review
process has any effect on the
information environment is unclear. On the one hand, disclosure
may improve as a result of the
review, thereby providing useful information to investors and
enhancing a firm’s information
environment. For example, an expanded or clarified revenue
recognition policy disclosure could
improve user forecasts of earnings, resulting in less surprise
at future earnings announcements.
Alternatively, any additional disclosure may be an oversupply of
information and thus have little
economic consequence. This alternative seems highly plausible,
given that under current
accounting and disclosure standards, the U.S. information
environment is already rich.
1 Section 2 describes additional institutional details on
comment letters.
-
2
Moreover, we might expect cross-sectional variation in any
potential effects. Some letters might
address more issues or issues of greater severity and thus
create a larger improvement.
Alternatively, effects may vary across firm size. Smaller firms
may not have the same quality of
information environment, so the impact of the SEC’s comments may
be greater. Finally, the
changes in disclosure that arise from comment letters could be
mimicked by industry peers,
hence we explore industry spillover effects. We expect any such
effects to be most pronounced
where SEC scrutiny is intense within an industry.
We collect the comment letters from the SEC’s website for
2003-2006 and retain those
related to 10Ks and 10Qs. We conduct content analysis on a
subsample and document that the
vast majority of letter comments address disclosure issues. We
then explore the attributes of
letter recipients. Whether or not a firm receives a comment
letter depends on two factors: the
type of firms that the SEC targets for review in that year, and
the firm’s reporting quality. The
SEC's review selection criteria are unknown. However, the
Sarbanes-Oxley Act of 2002 (SOX)
outlines several factors for the SEC to use when it selects
firms for review. We use those criteria
and other factors related to financial reporting quality and
find that firms more likely to receive a
SEC comment letter are industry leaders, those that have been
public longer and those that have
previously restated their financial results. Greater cash flow
volatility and higher Earnings/Price
ratios also increase the likelihood of receiving a letter.
To evaluate the impact of the comment letters on the firms’
information environment we
study the changes in market behavior around earnings
announcements. Prior research shows how
information asymmetry and differential information processing by
investors affect price and
volume reactions to public announcements (Diamond and
Verrecchia, 1991; Kim and
-
3
Verrecchia, 1991a, b, 1994, 1997; Harris and Raviv, 1993; Kandel
and Pearson, 1995). In
particular, a decline in price reactions around earnings
announcements suggests that the quality
of the information environment prior to the announcement has
improved. Lower volume
reactions indicate less investor disagreement regarding the
information content of the earnings
announcement and higher consensus of firm value.2
We find that abnormal return volatility and trading volume
around earnings
announcements decline subsequent to the SEC comment letters, and
that the magnitude of the
change is economically significant. The return volatility change
appears to be concentrated in
small firms and the volume results apply to letter cases which
are severe. Our findings are robust
to various approaches that address potential letter selection
biases.
In our test of industry spillover effect, we do not find any
change in return volatility for
industry peers, on average. However, in industries that receive
greater attention from the SEC,
peer firms do show a reduction in price reactions. The magnitude
of this reduction is as large as
the comment letter firms. We find no change in volume reactions
for peer firms.
Our study is related to four streams of research. The first
investigates the economic
consequences of companies that voluntarily commit to higher
levels of disclosure (e.g., Welker,
1995; Leuz and Verrecchia, 2000; Brown et al., 2004). These
studies provide evidence that the
quality or quantity of disclosure has positive effects.
Increases in disclosure levels tend to
generate greater stock liquidity and reduce a firm’s cost of
capital and information asymmetry.
2 The above argument is similar to Bailey et al. (2003) and
Bailey et al. (2006). Some might conjecture that the reverse of the
story is the case, that a higher quality disclosure is associated
with a stronger market reaction. This argument is true if earnings
quality improves, leading to a stronger ERC. In Section 2 we
examine a subsample of 157 comment letters and find that the
majority of the comments are related to disclosure, not earnings
quality. Therefore, on average, we expect dampened market reactions
after the letters.
-
4
Our paper complements these studies by examining the market
consequences of disclosure
changes that arise from the regulator’s direct monitoring of
corporate reporting.
The second research path studies the market consequences when
the SEC issues a new
regulation, for example, Regulation Fair Disclosure (FD) (Bailey
et al., 2003; Heflin et al.,
2003). This research finds that the quantity of voluntary
disclosure increases after the adoption of
the regulation, but provides mixed evidence on market related
aspects, such as volatility around
earnings announcements and the degree of information asymmetry.
Our study explores the
impact of monitoring and enforcement of regulations rather than
of issuing regulations.
A third research area which explores private and public
enforcement of securities laws
concludes that public enforcement of securities laws has limited
value (La Porta et al., 2006;
Djankov et al., 2008a). Jackson and Roe (2009) however, find
that these papers underestimate
the extent to which public enforcement is associated with
capital market development. Leuz and
Hail (2006) find that firms from countries with more extensive
disclosure requirements, stronger
securities regulation, and stricter enforcement mechanisms have
a significantly lower cost of
capital. Leuz and Hail base their enforcement construct on a
survey of lawyers. In contrast, we
examine a sample of actual enforcement activities undertaken in
the U.S. by the SEC. Unlike
LaPorta et al. (2006) and Djankov et al. (2008a) but similar to
Jackson and Roe (2009), our
results suggest that there are positive benefits of public
enforcement.
The fourth research stream explores the SEC Accounting and
Auditing Enforcement
Releases (AAER) (see for example, Feroz et al., 1991; Dechow et
al., 1996; Beatty et al., 1998;
Beneish, 1999; Farber, 2005). This research examines the impact
of these enforcement actions on
corporate governance, managers, auditors, underwriters, and
market participants. Comment
-
5
letters can lead to AAERs but AAERs are rare. By studying
comment letters, we add another
dimension to the research that assesses the impact of the SEC on
the U.S. markets.
A working paper by Ertimur and Nondorf (2006) examines a sample
of comment letters
for 95 firms that undergo the Initial Public Offering (IPO)
process.3 They find no association
between the SEC comment letters and IPO underpricing, bid-ask
spreads, or market depth. Our
study differs from theirs along several important dimensions.
First, they focus on IPO firms
whereas we study public companies which allows us to examine the
change in information
environment after the comment letters. Second, they categorize
the content of the letters into
several groups and that grouping may not be representative of
the relative importance of various
comments. We use a more objective measure, time to resolution,
to proxy for the severity of the
letter content. Finally, while they fail to find any association
between comment letter attributes
and the firm’s information environment, we find evidence that
targeted firms and, in some
circumstances, their industry peers experience an improvement in
their information environment
subsequent to the SEC comment letters.
Our study offers early evidence on the content, determinants and
consequences of SEC
comment letters and adds to the literature on the positive
effect of quality accounting and
disclosure on the information environment of U.S. firms. We
document the beneficial effect of
the oversight role played by the SEC in enhancing and
maintaining the quality of the information
environment of firms listed in the U.S. markets. This oversight
evidence is important, because
practitioners and academics often focus on the SEC’s role in
terms of creating regulations. Our
3 Their paper was undertaken independently and at the same time
as ours.
-
6
findings suggest that the SEC’s review process is also an
important factor that contributes to the
quality of the U.S. markets. The results could be of interest to
policymakers and the Commission
itself, particularly in terms of demonstrating the value of
their review efforts. Moreover, the
results could have implications for other countries who wish to
replicate the success of U.S.
markets. Finally, this paper is of relevance to the financial
statement analysis literature, in that it
demonstrates the potential value of reviewing financial
reports.
The paper is organized as follows. Section 2 provides
institutional details about the
SEC’s comment letter process and the results of our letter
content analysis. We develop our
hypotheses in Section 3 and in Section 4 we outline our research
design. Section 5 details the
data and explores the determinants of receiving a letter.
Section 6 presents the empirical analyses
and Section 7 concludes.
2. SEC comment letters: institutional background and content
analysis
The Securities Exchange Act of 1934 requires public companies to
file quarterly (10Q)
and annual reports (10K) with the SEC. The Sarbanes-Oxley Act of
2002 (SOX) requires the
SEC review a company’s filings at least once every three years.
Prior to SOX, the SEC reviewed
approximately 20% of the filings each year. The SEC motivates
the review program as follows:4
“The full disclosure system for public companies is the
foundation of the federal securities laws. Currently, the Division
of Corporation Finance achieves the goal of improving the quality
and timeliness of material disclosure to investors by selectively
reviewing the periodic financial and other disclosures made by
public companies.” (emphasis added)
4 Taken from sec.gov, 2008.
-
7
Various stated review objectives include identifying potential
or actual material accounting,
auditing, financial reporting or disclosure deficiencies;
influencing accounting standards and
practices; proposing new and amended disclosure rules; and
offering guidance and counseling,
either informally or through no action letters.5 Feroz et al.
(1991) cite an SEC official who
claimed that half of all SEC enforcement leads came from reviews
of financial statements and
securities filings.
The SEC does not disclose when a firm will be subject to review,
so only if a firm
receives a comment letter does it become aware of the review.
Many reviews are completed
without issuing any comments. Section 408 (b) of SOX requires
the Commission to consider the
following factors in scheduling reviews:
(1) issuers that issued a material restatement of financial
results; (2) issuers that
experience significant volatility in their share price as
compared to other issuers;
(3) issuers with the largest market capitalization; (4) emerging
companies with
disparities in price to earnings ratios; (5) issuers whose
operations significantly
affect any material sector of the economy; and (6) any other
factors that the
Commission may consider relevant.
When comments are issued, the company receives a letter from the
SEC and has ten
business days to respond. The company can either submit a
response letter or amend the filing
under review. Follow-up comment letters and responses can be
made until the issues are resolved
to the Commission’s satisfaction, at which point the SEC staff
advises the filer that the review is
complete.
5 In a speech on July 19, 2000, the SEC Chief Accountant, Robert
A Bayless made the following remark, “the review and comment
process in the Division of Corporation Finance unearths a
surprising number of accounting errors, disclosure deficiencies,
and tortured interpretations of GAAP in filings with the
Commission.”
-
8
Prior to 2005, public access to comment letters and the related
responses were only
available through a Freedom of Information Act request. Due to
an increasing number of such
requests, in 2005 the SEC began to publicly release, on their
website, comment letters relating to
filings made after August 1, 2004 but no earlier than 45 days
after the review is completed.
Therefore, in the absence of a company’s voluntary disclosure
regarding the letter, the public can
learn about the existence and content of the letter only when
the SEC releases it. Of course by
then, all the relevant issues have been resolved and interim
filings have been enhanced.6
We provide two complete letters as examples, one for the Landec
Corporation (Appendix
1) and a second one for Charles Schwab Corp. (Appendix 2). In
addition, to provide some
context for the nature and frequency of comments in the letters,
we read and manually code the
first batch of the letters posted to the SEC website. The first
batch contains 157 letters from
2004-2005. We apply extracts of the Ertimur and Nondorf (2006)
taxonomy, since we find that
they accurately represent the nature of comments found in the
letters. We present the 80 possible
comment types and a description of each in Appendix 3.
The comments fall into four groups. The first group, Accounting
Issues, address big-
picture problems. Comments relate to issues such as adherence to
GAAP, materiality, and
auditor issues. The second group, Accounting/Financial
Reporting/Disclosure Topics, comprise
comments specific to accounting balances or transactions such as
revenue recognition, inventory,
6 Proprietary costs are a major counter force to the incentive
to disclose. For sensitive information, companies can request
confidential treatment under Rule 83 (17 CFR 200.83). If the
request is granted, companies can exclude the confidential
information from publicly available filings, and only provide the
information to the SEC.
-
9
related party transactions, and capital expenditures. The third
group, Business Issues, address
more generic business issues, such as liquidity, competitive
environment, and risk factors. The
fourth group, Tone and Level of Disclosure Issues, is editorial
in nature, the comments address
presentation and requests are to emphasize or de-emphasize,
clarify or disaggregate, certain
items.
The 157 letters contain 1,504 comments, slightly less than ten
per letter, on average.
Forty-five percent of the comments fall into the second group.
Within the second group,
questions about revenue recognition are the most frequent,
followed by claims, commitments and
contingencies, and then expenses. The other three groups are
approximately equal in terms of the
percentage of comments. In the first group, the most common
comment is a request for a cite
from authoritative literature to support an accounting
treatment. Other frequent comments
include a request to clarify an accounting policy, reasons to
explain why the company is not
following GAAP, and a request to disclose certain material
information. In the third group, both
MD&A disclosure and liquidity issues receive substantial
attention. In the editorial category, the
most common comment is a request for something to be clarified.
The second most common
request is to quantify an amount related to a disclosure.
The descriptions of the 80 comment items in Appendix 3 indicate
that most of the issues
deal with disclosure. It seems unlikely that earnings or
accounting balances would be altered
frequently as a result of the review process. However, there is
one exception, item 11 in the first
group, "Not Following GAAP" (see Appendix 3). If this SEC claim
is supported, then earnings
may change. To provide further evidence on the nature of the
comments, we code the second
group comments into disclosure-related or change in accounting
numbers. If the comment
-
10
suggests that the filer provide more information in the 10K or
10Q, we code it as a disclosure
issue. If the comment is more likely to require a change in a
financial statement figure, then we
treat it as an accounting issue. We find ninety-six percent of
the comments are disclosure related.
This preliminary analysis suggests that the SEC review process
is more likely to impact
disclosure, rather than to revise accounting figures.
3. Hypothesis Development
A large body of research has established a positive relation
between enhanced disclosure
and firms’ information environment. For example, Welker (1995)
shows that a well-regarded
disclosure policy reduces information asymmetry and increases
liquidity in equity markets.
Healy, Hutton, and Palepu (1999) find that when firms expand
their voluntary disclosure, they
can attract more analyst following and improve stock liquidity.
Leuz and Verrecchia (2000)
examine German firms that commit to higher disclosure levels by
adopting International
Accounting Standards or U.S. GAAP and find that such firms
experience a decline in the
information asymmetry component of the cost of capital,
subsequent to adoption. Brown et al.
(2004) show that conference calls lead to long-term reduction in
information asymmetry among
equity investors. Brown and Hillegeist (2007) find that
disclosure quality reduces the likelihood
that investors discover and trade on private information.
If the disclosure revisions that result from the comment letter
process are substantive,
then we would expect an enhanced information environment
thereafter. Further, if an effect
exists, then ex ante it seems reasonable to expect
cross-sectional differences along at least two
dimensions. First, the resolution of severe letters is more
likely to improve the information
-
11
environment. For example, some letters may address a greater
number of issues, thus creating a
cumulative impact. Or some letters could address an issue that
is uniquely important. Neither
scenario is mutually exclusive, but it suggests that some
measure of severity of the letter would
be related to the market effect arising from the resolution of
the issue(s). Second, we conjecture
that the effect of the comment letters on the information
environment of small firms is larger. On
average, small firms have poorer information environments. They
generally have lower analyst
following and institutional ownership and are likely to have
higher disclosure costs (Lang and
Lundholm 1993). This poor (pre-letter) information environment
suggests that if there is any
impact created by the comment letters, then it should be
relatively greater for small firms.
The effect of comment letters, if any, may extend beyond
targeted firms. Peer companies
may mimic comment letter firm disclosure to adhere to industry
norms or avoid review. Further
Big 4 audit firms may also encourage conformity across clients.
If peers do mimic the
disclosures arising from comment letters, then the information
environments of firms that do not
receive letters could be enhanced. We hypothesize that this
indirect or spillover potential is most
likely where the SEC performs an intense review of a particular
industry.
However, a long-standing economic question is the justification
of regulating corporate
disclosures (Healy and Palepu, 2001). Schulte (1988) summarizes
the paradox of regulation by
arguing that information is similar in nature to public goods,
because it is difficult to exclude
others from using it and its consumption by one user does not
diminish its availability to others.
Without regulation, public goods tend to be underproduced
because of the free-rider problem.
The tendency in regulation, however, is to oversupply the public
good, because users of
information always overstate their demand.
-
12
Based on Schulte’s (1988) argument, if the SEC is merely
creating excess disclosure or
an oversupply, then comment letters should contain little
substance and result in minimal
improvement in the firm's information environment. Given the
high quality accounting
standards, disclosure requirements, and the requirement in the
U.S. that public companies be
audited, we question whether there would be any benefit from an
SEC review, which in
substance is merely the SEC staff reading the corporate
filings.
Another rationale for why we might expect comment letters to
have no economic effect is
regulatory capture theory, which suggests that regulated firms
manipulate the agency responsible
for regulating them (see Dal Bo, 2006 for a review). If the SEC
is subject to filer influence, then
comment letters may avoid substantive issues, thus creating no
economic benefits. Ertimur and
Nondorf’s (2006) lack of results would support either of these
rationales.
Some anecdotal evidence of the concern about the effectiveness
of the comment letters
appears in the Commission’s own annual reports (2005 and 2006),
in which they report various
metrics to track and report on SEC effectiveness. For comment
letters, they state that they are
currently unable to quantify significant improvements or actions
related to the letters (see
footnote for exact statement). 7
7 “For corporate filings, comments are issued to elicit better
compliance with applicable disclosure requirements and improve the
information available to investors. Many instances, amendments
involve financial restatements. Determination of “significance”
stems from the nature of the change (e.g., restating positive
income as a loss) or the size of the company. Analysis of divisions
of corporation finance and management continued to work toward
establishing a means for accurately tracking data on comments that
result in significant enhancements in financial and other
disclosures or other significant actions to protect shareholders.
Divisions will provide data for this indicator once such tracking
methods are in place.” See Exhibit 2.23 in the 2006 SEC annual
report available at www.sec.gov.
-
13
In summary, whether the SEC’s review of registrant filings
improves firms’ information
environment is an empirical question. On one hand, if the review
process is successful in
enhancing disclosure to investors, we would expect a better
information environment. This
improvement is likely to vary with the severity of the letter
and the firm’s original information
environment. On the other hand, if any additional information
required by the review is an
unnecessary oversupply or lacks substance, we would expect no
change in the firms’ information
environment.
4. Research Design
To assess whether SEC oversight enhances firms’ information
environments, we compare
market behavior around earnings releases for the two quarters
before and after the receipt of an
SEC comment letter. Our focus on earnings announcement reactions
as a proxy for firms’
information environment is motivated by theoretical work of
Diamond and Verrecchia (1991)
and Kim and Verrecchia (1991a, b; 1994; 1997). These papers show
that stock price reaction to a
public announcement decreases with the precision of
pre-announcement information. Therefore,
a decrease in return volatility around earnings announcements
indicates an improvement in the
pre-announcement information environment. These papers further
show that stock trading
volume arises from differences in the quality (precision) of
investor’s private information.
Consequently, a reduction in trading volume around earnings
announcements reflects less
information asymmetry across investors and greater consensus
regarding firm value. Several
-
14
papers in accounting and finance have used a similar approach to
study information environment
changes.8
4.1 Baseline model
We measure stock price reaction as the four-day absolute
cumulative abnormal return
(ACAR) around quarterly earnings announcements. We define the
earnings announcement date
as day zero (� = 0), and compute ACAR as ���� = | ∏ (1 + ��) −
1|
2
��1 where we compute
AR as the abnormal return based on one-factor market model
residuals estimated over the period
t-11 to t-200 trading days. CAV is the four-day cumulative
abnormal trading volume around the
quarterly earnings announcement.9 We define abnormal trading
volume as the difference
between announcement-window (-1, +2) trading volume and the mean
of pre-announcement
window (-200, -11) trading volume, normalized by the mean
volume.
To test the effect of the SEC oversight on the information
environment, we employ the
following model:
������ �������� = �0 + �1������� + �2�������_���� + ������!� + "
(1)
where Market Reaction is either absolute return or volume
reaction to earnings announcements;
ComtLtr equals one if the firm receives a SEC comment letter,
and zero otherwise; and
8 For example, Heflin, Subramanyam and Zhang (2003) use return
volatility around earnings announcements to study the change in the
flow of financial information to the capital markets before and
after the implementation of Regulation FD. Another FD study, Bailey
et al. (2003) use return and volume reactions to investigate the
information environment change around the standard. Bailey et al.
(2006) use return and volume reactions around earnings releases to
study the change in a firm’s information environment when it
cross-lists in the US market. 9 Some papers in the trading volume
literature use a “volume market model” in the preannouncement
window to calculate expected trading volume (e.g., Tkac (1999);
Bailey et al. (2006)). We opt not to follow this approach because
given the highly skewed volume data, a linear model tends to poorly
specify the underlying data structure. Moreover, such model
requires more computational cost but provides little improvement in
power (Bamber, Barron and Stevens 2009).
-
15
ComtLtr_Post equals one if the firm receives a letter and the
observation is from the post-letter
period, and zero otherwise. We define the pre-letter period as
the two quarters before the date of
the first letter and the post-letter period as the second
quarter after the date of the last letter. We
omit the first quarter after the letter to ensure that the
equity market has been exposed to any
change in disclosure. We require the comment letter firms to
have both a pre-letter and a post-
letter observation.
In Equation (1), β0 measures the average market reactions of the
benchmark firms, β1
captures the difference in market reactions between the comment
letter firms and the benchmark
firms, and β2 captures the change in market reactions for
comment letter firms. If the SEC’s
review process substantively improves corporate disclosures,
then we would expect more muted
market responses to earnings announcements following the
resolution of the letters, which would
be reflected by a negative β2. In contrast, if the additional
information resulting from the review
process is just an oversupply, then we would expect no change in
market responses and thus β2 to
be indistinguishable from zero.
Consistent with earlier studies, we include a set of control
variables related to market
behavior around earnings announcements, all of which we discuss
in the results section. We
estimate Equation (1) with industry fixed effects, where we
define industry as in Fama and
French (1997). We cluster standard errors by firm and year to
correct for possible correlations
across observations (Rogers, 1993; Petersen, 2009).
To investigate whether the change in the information environment
varies with the
severity of the letters, we alter Equation (1) by replacing the
comment-letter dummy with a
variable based on the severity of letter content. We use the
duration of the letter period as our
-
16
proxy for the seriousness of the letter content. We conjecture
that if it takes longer to resolve the
comment letter issues, then the issues are more likely to be
substantial, or there are more issues
to resolve, or both.
The change in a firm's information environment is also likely to
vary with the firm’s pre-
letter information environment. We use firm size as our proxy
for a firm’s original information
environment. To capture the differential effects, we partition
the comment letter dummy in
Equation (1) into three size dummies.
4.2 Comment letter selection issues
If a firm receives an SEC comment letter, it is unlikely to be a
random event. For
example, Section 408 of the SOX identifies various firm
characteristics for consideration by the
SEC staff in choosing which companies to review. Therefore,
certain types of firms are more
likely to attract the SEC’s attention. In addition, firms with
certain characteristics may be more
likely to have reporting deficiencies and hence, receive a
letter. This non-random treatment
assignment implies that there may be systematic differences
between firms that receive a SEC
letter and firms that do not. The systematic differences could
lead to biased estimates of the
treatment effect, i.e., changes in market reactions to earnings
announcements.
One way to address the selection concern would be to construct
the benchmark control
sample based on firms subject to SEC review, but for which no
letter is issued. However, we are
unable to identity all the firms the Commission chooses to
review.10 Hence, we construct the
10 Despite making several phone calls to the Division of
Corporate Finance and also requesting the data under the Freedom Of
Information Act.
-
17
benchmark control sample based on firms not receiving a letter
in the year, i.e., these firms can
either have been reviewed by the SEC but did not receive a
letter or not have been reviewed at
all. However, we further screen the benchmark firms, which we
describe next.
We start by building the determinant model which relates firm
characteristics to the
probability of receiving a letter.11 The firm characteristics we
consider are those listed on Section
408 of the SOX and also the firm’s operating uncertainty and
audit quality. We conjecture that
firms are more likely to get a comment letter if they fall under
the SOX Section 408 criteria and
are more subject to reporting errors. For every year, we run a
logit model and assign each firm a
predicted probability of receiving a comment letter in that
year. Then we require both the
comment letter firms and the benchmark firms to have a common
predicted probability range.
We remove firms that fall out of the range and exclude any
unmatched comment-letter firm.
In addition, we conduct several robustness tests. First, we
apply a standard two-stage
Heckman model. The first stage is a probit regression that
models the probability of receiving a
letter. The second stage is Equation (1) augmented with the
inverse Mills ratio from the probit
model to control for any self-selection bias.
Second, we use propensity score matching to select a matched
control sample. We then
conduct a difference-in-difference test, examining the change in
the information environment
between the comment-letter sample and the matched control
sample. The propensity score for a
firm is the probability of receiving a SEC comment letter
conditional on the firm’s observable
characteristics. Propensity score matching provides two
advantages. First, it allows us to control
11 (See Section 5.2)
-
18
for many company covariates simultaneously by matching on a
single scalar variable (i.e., the
propensity score), and second, we can create a quasi-randomized
experiment (D’Agostino,
1998). If we find two firms, one in the comment letter group and
one in the control group, with
the same propensity score, then it is as if these two firms were
randomly assigned to each group
in the sense that they are equally likely to be treated or
control (conditional on the observed firm
characteristics). Each matched control firm has a hypothetical
comment letter period based on
the corresponding comment-letter firm.
The difference-in-difference model is:
������ �������� = �0 + �1���� + �2������� + �3������� ∗ ����
+
+ ������!� + "
(2)
�1 captures how both treatment and control firms are influenced
by time. The time-invariant
difference, if any, in information environment between the
treatment and the control is captured
by �2. β3 represents the differential change in information
environment between the comment
letter firms and the matched control firms. If the SEC’s review
process is effective in enhancing
a firm’s information environment, then we would expect �3 <
0.
The difference-in-difference design ensures that our results are
not driven by some
unspecified macro time trend. However, propensity score matching
is also subject to the
limitation that it can only remove selection biases based on
observable firm characteristics. It is
less robust to (unobservable) omitted conditioning variables
than the Heckman approach
(Heckman and Navarro-Lozano, 2004). Since the Heckman and the
matching approach have
their own strengths and weaknesses, we use both to test the
robustness of our findings.
-
19
5. Data and comment letter determinants
5.1 Sample description
We search EDGAR, the SEC database of public company filings, for
comment letters
relating to 10Qs and 10Ks. For the period 2003 to 2006 we obtain
9,212 letters relating to 4,138
cases for 3,818 firms.
Table 1, Panel A summarizes the sample selection process. Of the
3,818 firms in our
sample, there are 307 firms that have 627 cases. For our pre-
and post-letter research design, we
want to ensure that issues are fully resolved in the post-letter
period. Therefore, for these 627
cases, we use two criteria to decide whether the later cases are
actually a continuing investigation
of an earlier case and hence the two should be combined. First,
if the two cases have overlapping
letter periods, we consider them as one case. We find 90 cases
meet this criterion and merge
them into 45 cases. Second, if the period between the end date
of the earlier case and the
beginning date of the later case is less than six months, 232
cases meet this criteria. Merging
these cases reduces the number of cases by 118. (Two firms have
three cases and one has four
cases.) In addition, we require the 3,818 sample firms have a
Compustat GVKEY. This
requirement eliminates 903 firms and 921 cases. Our selection
process results in a final sample
of 3,054 cases that represent 2,915 firms.
Table 1, Panel B shows that most of the cases in the sample
arise in 2005 and 2006. This
time clustering exists because the SEC began publicly disclosing
comment letters in 2005 and
only intended to post letters relating to filings made after
August 1, 2004. Panel C shows that the
majority of the sample firms are the focus of only one SEC
investigation. Out of 2,915 firms,
only 135 firms are the subject of two cases and two firms are
the subject of 3 cases. In Panel D,
-
20
we see that on average, a case lasts for 69 days and has two
comment letters. Panel E reports the
industry distribution of both comment letter firms and the
Compustat universe in the sample
period. The industry classification follows Fama and French
(1997). Insurance is slightly over-
represented among comment letter firms (4.32% compared to
2.66%), and utilities are slightly
underrepresented (2.95% compared to 4.05%). Otherwise, letter
representation appears to be
proportional.12
5.2 Determinants of receiving an SEC comment letter
Whether a firm receives an SEC comment letter in a particular
year depends on whether
the firm is selected for review in that year and its reporting
quality. We develop our determinant
model based on these two factors.
Our proxies for the SOX review criteria follow. Restate is a
firm’s restatement history,
which we define as the number of restatements the firm has filed
based on the Government
Office of Accountability database. We measure a firm’s price
volatility as its idiosyncratic
volatility in the stock market so %&���'�������(�! = ln
(1�)2
)2), and �2 is the R-square from the
market model estimated one year prior to receiving the comment
letter (Durnev et al., 2004;
Ferreira and Laux, 2007). This proxy is relative to market-wide
variation which we believe best
captures the SOX criteria. MarketCap is the firm’s market
capitalization at the fiscal year-end
12 In untabulated analysis we explore the market reaction when a
firm receives the first comment letter from the SEC. Of 2,915
firms, we are able to find 2,324 firms (2,376 cases) with
non-missing return data on the first letter date. We find no
statistically significant stock market reaction in terms of either
daily abnormal return or cumulative abnormal return over (-1, +2).
This result is not surprising, since the existence of an SEC
enquiry is unknown to the public until the case is resolved and the
SEC releases the letters. Obviously, we are interested in the
market reaction on the date that the SEC releases the letters, but
we do not have those dates of release.
-
21
prior to receiving a comment letter. To measure whether a firm
is an emerging company with a
disparate PE ratio, we include a firm’s age and its
earnings-per-share (EPS) to share price ratio
(E/P). Age is the number of years the firm appears on CRSP. E/P
is calculated at the fiscal year-
end prior to the comment letter. We use the E/P ratio rather
than a P/E ratio because some of the
sample firms have zero earnings. To measure the impact of a
firm’s operation on any material
sector of the economy, we calculate each company’s proportion of
their respective industry
revenue at the fiscal year end prior to the comment letter and
denote the variable as
RevProportion. We define industries based on their two-digit SIC
codes.
In addition to the SOX review criteria we include factors that
relate to a firm’s reporting
quality. First, firms with high uncertainty in their operating
environment are likely to use greater
estimation and more approximations in their financial reports.
Accordingly, we expect such firms
to be more subject to reporting errors. We use the volatility of
a firm’s operating cash flow as our
proxy for operating uncertainty. CFOVol is the standard
deviation of cash flows from operation
(CFO) over the five years prior to receiving the comment letter.
We scale CFO by total assets.
Dominant audit suppliers are likely to provide higher quality
audits because they have more
resources and also face a greater risk to their reputation if
they conduct poor-quality audits. We
expect companies audited by these large audit firms to have
higher reporting quality, and hence
to be less likely to receive a comment letter. The dominant
audit suppliers in our sample period
are the so-called "Big 4" public accounting firms: Deloitte and
Touche, Ernst and Young,
KPMG, and PricewaterhouseCoopers. Big4 equals one if the firm is
audited by one of these audit
firms and zero otherwise.
-
22
Of the 3,054 comment letter cases, 2,190 have complete data for
the determinant
variables. There are 9,098 non-comment-letter firm-years in
Compustat that have similarly
complete data. Table 2, Panel A provides descriptive statistics
and the results of the univariate
tests we use to compare firms that receive an SEC comment letter
with those that do not. For
comment-letter firms, we measure all variables prior to the date
of the first letter. For non-
comment-letter firms, we measure all variables at the prior
year’s fiscal year-end date.
On average, we find that firms that receive an SEC comment
letter are more likely to
have a restatement history and lower idiosyncratic volatility,
be larger in market capitalization
and represent a larger proportion of their industry revenue,
have been listed longer, and have a
higher E/P ratio (median only). The univariate results on
operating cash flow volatility that we
use as a proxy for operating uncertainty are mixed; the average
CFOVol of the comment letter
sample is significantly larger than the non-comment-letter
sample, while the median CFOVol of
the comment-letter firms is significantly smaller than the
non-comment-letter firms. We find no
significant difference in the proportion of comment letter firms
and non-comment-letter firms
that are audited by a Big 4 audit firm.
Panel B of Table 2 presents Pearson correlations. The largest
significant correlations are a
negative correlation of -0.377 between Big4 and
IdiosyncraticVol, followed by a positive
correlation of 0.259 between RevProportion and MarketCap. The
majority of other correlations
fall between ±0.15, which suggests that the variables included
in our determinant model capture
distinct firm attributes.
We model the probability of a firm receiving a comment letter as
a function of the above-
mentioned firm characteristics by using the following logistic
regression.
-
23
���,(�������) = -(�0 + �1������� + �2%&���'�������(�!
+�3��������. + �4��/���.������ + �5�0�
+�61� + �7�23(�! + �84�04)
(3)
Panel C of Table 2 presents the results. In addition to the
coefficient estimates, we calculate the
marginal effect of each variable. Doing so can provide insight
into which firm attributes are most
important in determining the likelihood of receiving a letter
from the Commission. We find that
share of industry revenue has the largest marginal effect on the
probability of receiving an SEC
comment letter. This result suggests that the SEC pays more
attention to these firms because they
play an important role in the economy. A previous restatement
and large P/E disparity also
increase the probability of receiving a letter. However,
inconsistent with the SOX guideline, we
find that older firms are more likely to receive a letter. And
we find that firms with a more
uncertain operating environment, as reflected in the operating
cash flow volatility, face a higher
probability of receiving a letter. Our determinant model yields
a Wald 52 of 124.81, which is
significant at the 1% level or better.
Overall, the univarite analyses and the logit model results in
Table 2 suggest that the SEC
tends to pay more attention to industry leaders and more
established firms and firms that have a
restatement history and higher operating uncertainty.
6. Market reactions to earnings announcements
Following Heflin, Subramanyam, and Zhang (2003), in Figure 1 we
plot the mean
absolute cumulative abnormal returns (ACARs) from 64 trading
days (the approximate number
of trading days in a quarter) before to two days after earnings
announcements for both the pre-
-
24
and post-comment-letter periods. The post-comment-letter ACARs
are consistently smaller than
are their pre-comment-letter counterparts, as reflected in the
line below in Figure 1. Since
ACARs represent the information gap between the pre-announcement
price and the full-
information post-announcement price, the figure suggests a
reduction in the information gap
after the comment letters, and therefore an enhanced
pre-announcement information
environment.
6.1. Changes in market reactions
Our requirement that comment-letter and benchmark firms have a
common range of
predicted probability of receiving a letter in the year leaves
2,070 comment-letter firms and
3,832 benchmark firms. We further require non-missing data from
Compustat, CRSP, and IBES,
so for the share price volatility test, our sample includes
1,286 comment-letter firms and 2,067
benchmark firms. For the trading volume test, we have 1,897
comment-letter firms and 3,540
benchmark firms. The price volatility test sample is smaller
because it requires IBES data.
Panel A of Table 3 provides descriptive statistics for the
market reaction variables. The
univariate comparisons show a decline in both ACAR and CAV for
the comment-letter firms
following resolution of the letter, although only the ACAR
change is statistically significant. On
average, we see no difference in ACARs between benchmark firms
and comment-letter firms
prior to the letter, but we do observe larger CAVs for the
comment-letter firms. Comparing the
post-comment-letter reactions for comment-letter firms to the
benchmark firms, we see lower
ACAR reactions for comment-letter firms and no statistically
significant difference in CAV
reaction.
-
25
Panels B1 and B2 of Table 3 present the regression results when
we use ACAR as the
dependent variable. We present nine columns of analysis. In
Panel B1, Columns (1) and (2), we
present the baseline model, with and without industry fixed
effects respectively; in Columns (3)
and (4) we present some robustness tests. Panel B2 presents the
additional analyses that we use
to address the potential selection bias problem, and also the
cross-sectional analyses based on
letter severity and firm size. The results of the baseline model
show that, after controlling for
various firm characteristics, the price reactions of the
comment-letter firms are significantly
different from those of the benchmark firms as the coefficients
on ComtLtr are positive and
significant. More importantly, for comment-letter firms, the
price reactions become significantly
lower after the receipt of an SEC comment letter as the
coefficients on ComtLtr_Post are
negative and statistically significant at the 1% level.
To give a sense of the magnitude of the changes in ACARs, we
compare the coefficients
on ComtLtr_Post with the sum of the coefficients on the constant
and on ComtLtr. The
comparison using the coefficients in Column (1) indicates that
comment-letter firms experience a
decrease of about 155% in ACARs. Column (2) shows that including
the industry fixed effects
reduces the magnitude of the decline to 32%.
The control variable coefficients generally have the expected
sign. Firms with inherently
higher price volatility tend to have higher price reactions
around earnings releases, as indicated
by the positive coefficients on RetVol and NegCar (Heflin et al.
2003; Black, 1976; Christie,
1982; Nelson, 1991). The positive coefficients on AbsCar suggest
that larger information flow
yields greater market reactions (Heflin et al., 2003). The
coefficients on Loss are negative and
significant, consistent with the theory that the market reacts
less when the earnings numbers are
-
26
less informative (Hayn, 1995). The coefficients on Size are
negative, suggesting larger reactions
for smaller firms (Atiase, 1985). For smaller firms, investors
may have less incentive to gather
pre-disclosure information, and therefore the market reacts more
to an earnings shock. We do not
find that bond yield (Collins and Kothari, 1989) or analyst
forecast error are significant in our
specification. The results of Columns (3) and (4) in Table 3
confirm that these results are robust
to using only the one quarter before the comment letter in
contrast to two in the earlier
specification (balanced panel). Our results are also robust to
calculating the abnormal return
market model only once, just prior to the comment letters (no
overlap) for both the pre- and post-
comment-letter reactions.
Panel B2 begins with a standard Heckman model. Column (1)
presents the probit model
results. The results are consistent with those presented in
Section 5. Large industry leaders and
firms with a restatement history are more likely to receive an
SEC comment letter. Column (2)
presents the second-stage OLS results. The coefficient on the
inverse Mills ratio, λ, is significant
and positive, indicating the presence of an upward selection
bias toward the coefficient on
ComtLtr. After including the inverse Mills ratio, the
coefficient on ComtLtr switches from
positive (Panel B1 column (2)) to negative. More importantly,
the coefficient on the
ComtLtr_Post term remains negative and significant. After
correcting for the potential selection
bias, we find an 87.5% reduction in return volatility for
post-comment-letter earnings events.
In addition, we select a propensity-score-matched control sample
and conduct a
difference-in-difference test. Among the 1,286 comment-letter
firms in the return volatility test,
we find matches for 1,227 firms. After the match, the means of
all the determinant variables
(Restate, IdiosyncraticVol, MarketCap, RevProportion, Age, E/P,
CFOVol, and Big4) are not
-
27
significantly different between the comment letter firms and the
control sample firms (not
tabulated). This lack of differences confirms that the matching
successfully removes observable
systematic differences between the two samples.
Column (3) of Panel B2 presents the matched sample results. The
coefficient on ComtLtr
is no longer significantly different from zero, which further
confirms that the matched control
sample has firm features similar to those of the treatment
sample. The coefficient on Post is also
not significant, indicating that the control firms do not
experience any change in their abnormal
return reactions to earnings announcement. The sum of the
coefficients on Post and ComtLtr x
Post is negative and significantly different from zero (p-value
= 0.0043), implying a decline in
ACARs for comment-letter firms. The coefficient on the
interaction term ComtLtr x Post
captures the difference-in-difference between the comment-letter
firms and the control firms and
it is negative and significant at the 1% level. The results of
all four analyses consistently show
that companies generally experience a reduction in return
reactions around earnings
announcements after the resolution of SEC comment letters. The
reduced reactions imply an
enhanced pre-announcement information environment and a smaller
information gap between
firm managers and investors.
In Table 3, Columns (4) and (5) of Panel B2 we present the
cross-sectional analysis using
the matched sample. As our proxy for letter severity we use the
time to resolve the comment
letter case. We introduce two dummy variables, Severe for cases
that are the top decile of the
distribution, and Non-Severe for the others. Although the Column
(4) results are consistent for
both groups, the reduction magnitude is much greater for the
severe cases as the interactive post
coefficient is -0.018 compared to -0.006 for the non-severe
cases.
-
28
We also split the sample into three groups based on market
capitalization and present the
results in Table 3, Column (5). The results show that the
comment-letter effect is concentrated in
the small firm group. These firms likely have a lower-quality
initial information environment.
However, regulatory capture theory is an alternative explanation
for our size results. If larger
firms have influence over the SEC that small firms do not, then
they may be able to avoid any
substantive changes required from a review.
In Panels C1 and C2 of Table 3, we present the results where
abnormal trading volume is
the dependent variable. In these models, we control for the
contemporaneous price reaction
because prior research finds a positive association between
volume reaction and the same
window price reaction (e.g., Kim and Verrecchia 1991a, 1997;
Atiase and Bamber 1994). ACAR
is the absolute value of the cumulative abnormal return in the
announcement window (-1, 2).
Given the importance of the relation between ACAR and trading
volume, we also allow the
coefficient on ACAR to differ in the pre- and
post-comment-letter period. Consistent with prior
research, we find that ACAR is positively associated with volume
reactions. Similar to the price
reaction specification, we control for the amount of surprise to
investors. However, rather than
using analyst forecasts as our proxy for investor expectations,
we use last year’s quarterly
earnings since Bamber (1986) shows that price reactions to
earnings announcements is best
related to analyst-based unexpected earnings, while trading
volume reactions is more related to
random-walk based unexpected earnings. EarnSurprise is the
absolute value of seasonal changes
in earnings per share deflated by price at the end of the
quarter. Surprisingly, we find no relation
to earnings surprise.
-
29
The structure and presentation of Panel C is similar to Panel B.
The baseline model
(Columns (1) and (2)) show that the coefficients on ComtLtr_
Post are significant and negative,
indicating a reduction in abnormal trading volume around
earnings releases for comment-letter
firms. The results imply that the SEC comment letters are
effective in enhancing disclosure
levels, which leads to a lower divergence of opinion among
investors. The results shown in
Column (3) confirm that our results are robust to using only the
quarter before the comment
letter (balanced panel). However, the results do not hold when
calculating the abnormal volume
model strictly prior to the comment letters. We believe this
result arises due to the time series
trend of increasing trading volume.
Panel C2 presents the test results which control for selection
bias as well our cross-
sectional analysis. The baseline model result is robust to a
Heckman self-selection bias
correction shown in Column (2). Of the 1,897 comment-letter
firms in the volume test, we are
able to find 1,873 matched control firms. Column (3) shows the
difference-in-difference results
with the matched control sample. The coefficient on ComtLtr x
Post becomes insignificant in this
specification. The cross-sectional analysis shows that the lack
of results in Column (3) seems to
be in part atrributable to the non-severe cases, which show no
change in Column (4) in contrast
to the severe cases. We do not find any cross-sectional
differences by firm size (Column 5).
We find some evidence of a trading volume decline following
comment letters, but the
results are sensitive to the specification and the effect
appears to be strongest for the severe cases.
-
30
6.2 Industry externality
The previous results suggest that once the SEC’s comments are
resolved, comment-letter
firms experience a reduction in price and volume reactions
around earnings announcements. The
lower reactions imply an enhanced information environment and an
effective SEC review
process. In this section, we explore whether the SEC review
process creates an industry spillover.
We define industry peers as all companies in the same four-digit
SIC industry level as
comment-letter firms that do not receive a comment letter in the
same year. We further require
any industry year to have at least five firms. For every year,
we denote an industry comment-
letter period as starting from the earliest date of all the
cases issued to the industry in that
industry year and ending on the last date of those cases. We
define the pre-comment-letter period
for the industry peers as the the two fiscal quarters before the
starting date of the industry
comment-letter period and the post-comment-letter period is the
second fiscal quarter after the
end date of the industry comment-letter period. After we apply
the data requirements, we are left
with 861 comment-letter firms and 1,331 industry peers for the
return volatility test; for the
trading volume test we have 1,424 comment letter firms and 2,499
industry peers.
Panel A of Table 4 reports the regression results of changes in
return reactions. In
Columns (1) and (2), we compare the change in return reactions
among peer firms. In Column (1)
the coefficient on Post is not statistically significant,
indicating that on average, peer firms
experience no change in their information environment,
suggesting no positive industry
externality. However, Column (1) does not differentiate between
whether or not the industry is
receiving intense SEC attention. We investigate this possibility
in Column (2). We define an
industry as receiving intense SEC attention if the industry has
received more than eight
-
31
comment-letter cases in the year. Eight represents the upper
quartile. We partition the Post
dummy into Post_ Intense, and Post_ Nonintense, according to the
industry group to which each
firm belongs. The results indicate that peer firms in nonintense
industries do not experience any
change in their return reactions around earnings announcements,
however, peer firms in intense
industries experience a decrease in theirs. The negative and
significant coefficient on
Post_Intense provides some evidence that issuing comment letters
creates positive industry
spillover, and that firms in industries to which the SEC reviews
intensely improve their
information environment even without receiving a letter.
In Table 4, Columns (3), (4), and (5), we compare the change in
return volatility between
peer and comment letter firms. If the externality is strong,
then the reduction in return volatility
for the industry peers may be the same as comment-letter firms.
Column (3) presents the overall
results. The coefficient on Post x Treat is negative and
significant, suggesting that the decrease
in return reactions is larger for firms that actually receive a
letter. The coefficient on Post is not
significant. Peer firms do not experience any change in their
price reactions, on average. From
the results of Columns (1) and (2), however, we know that the
nonsignificance of Post comes
from peer firms in the nonintense industries.
Therefore we examine intense industries and nonintense
industries seperately. Column (4)
presents the results for intense industries. Consistent with the
result in column (2), the coefficient
on Post is negative and significant: firms in intense industries
experience a reduction in return
reactions even without receiving a comment letter. The
nonsignificant coefficient on Post x Treat
indicates that comment-letter firms and peer firms experience a
similar decrease in return
reactions.
-
32
Column (5) shows that peer firms in nonintense industries do not
experience any change
in return reactions, and that comment-letter firms experience a
greater decrease in return
reactions. The evidence of an industry spillover suggests that
the Table 3 results may actually
understate the comment-letter effect, since the benchmark or
control firms are also affected by
SEC oversight, thereby reducing the contrast between
comment-letter and peer firms.
We repeat the externality analysis with trading volume as the
dependent variable and
present the results in Panel B of Table 4. From Columns (1) and
(2), we find that peer firms do
not experience a decrease in their trading volume reactions.
When we compare the decrease in
volume reactions between peer firms and comment-letter firms in
Columns (3) – (5), we find that
whether or not a firm receives a comment letter makes no
difference. These results differ from
Table 3 due to the different sample applied for each.
In summary, Table 4 documents some evidence of a positive
industry externality from
comment letters. Peer firms who do not receive a letter but are
in the same industry as comment
letter firms receiving intense SEC scrutiny also experience a
reduction in return reactions around
earnings releases and the magnitude of the reduction for
industry peers is equivalent to that for
comment letter firms.
7. Conclusion
We study both the direct and indirect impact of SEC comment
letters on the information
environment of firms by examining stock market reactions to
earnings announcements before
and after a firm resolves a letter. Our sample includes 2,915
firms that receive comment letters in
the 2003-2006 period. Our results provide evidence consistent
with a change in the firm’s
-
33
information environment following a comment letter. This change
is evidenced by lower
abnormal return volatility and trading volume around ensuing
earnings announcements.
Generally, our results are robust to correction for potential
selection bias associated with which
firms receive a letter. However, the trading volume results
appear to be concentrated in
comment-letter cases that are severe. Moreover, we find some
evidence that there is an industry
spillover effect. We find that industry peers that do not
receive letters also experience benefits,
suggesting that they adopt disclosure changes of comment-letter
firms. Hence, we conclude that
the SEC comment letters enhance the quality of firms’ accounting
and disclosure, which results
in improved information environments.
Our paper provides what we believe is the first large-sample
evidence on the financial
reporting oversight role of the SEC. We find that regulators can
improve firms’ information
environment by monitoring corporate reporting. These results
contrast with recent papers that
suggest no benefit to public enforcment (La Porta et al. (2006);
Djankov et al. (2008a)). Our
findings could be of interest to policy makers, both domestic
and foreign, as well as the SEC
who wish to evaluate the effectiveness of this regulatory
effort.
We note that our paper is not without its limitations. Our
sample of SEC comment letters
is clustered in a short time frame and hence the
generalizability of our results may be a concern.
In addition, the SEC comment-letter process has many different
objectives (see Section 2) and
thus may create other effects that we do not explore.
-
References
Alexander, C.R., and K.W. Hanley, 2007. Regulatory monitoring
under the Sarbanes-Oxley Act. SEC Working Paper.
Atiase, Rowland Kwame, 1985. Predisclosure information, firm
capitalization, and security price
behavior around earnings announcements. Journal of Accounting
Research 23: 21-36. Atiase, Rowland Kwame, 1987. Market
implications of predisclosure information: Size and
exchange effects. Journal of Accounting Research 25 168-176.
Atiase, R.K., and L.S. Bamber, 1994. Trading volume reactions to
annual accounting earnings announcements: The incremental role of
predisclosure information asymmetry. Journal of Accounting and
Economics 17, 309-329.
Bailey, W., Li, H., Mao, C.X., and R. Zhong, 2003. Regulation
Fair Disclosure and earnings
information: market, analyst, and corporate responses. Journal
of Finance, 2487-2514. Bailey, W., Karolyi, A., and C. Salva, 2006.
The economic consequences of increased
disclosure: Evidence from international cross-listings. Journal
of Financial Economics, 175-213.
Bamber, Linda Smith, 1986. The information content of annual
earnings releases: A trading
volume approach. Journal of Accounting Research 24 40-56.
Bamber, Linda Smith, Orie E. Barron, and Douglas E. Stevens,
2009. Trading volume around
earnings announcements and other financial reports: Theory,
research Design, empirical
Evidence, and directions for future research. Working Paper,
Penn State University.
Bayless, R. A., 2000. “Speech by SEC Staff: How the Division of
Corporation Finance Influences Accounting Practices and Financial
Reporting Standards”
Beatty, R. P., 1989. Auditor reputation and the pricing of
initial public offerings. The
Accounting Review 64, 693.709.
Beatty, R.P., Bunsis, H., and J.R. Hand, 1998. The indirect
economic penalties in SEC investigations of underwriters. Journal
of Financial Economics 50, 151-186.
Beneish, M., 1999. Incentives and Penalties Related to Earnings
Overstatement That Violate
GAAP. The Accounting Review 74 (4): 425-457 Black, Fischer,
1976. Studies in stock price volatility changes. Proceedings of the
1976 Meetings
of the American Statistical Association, Business and Economics
Statistics Section, 177-181.
-
35
Brown, Stephen, and Stephen A. Hillegeist, 2007. How disclosure
quality affects the level of
information asymmetry. Review of Accounting Studies 12
443–477.
Brown, Stephen, Stephen A Hillegeist, and Kin Lo, 2004.
Conference calls and information asymmetry. Journal of Accounting
and Economics 37 343–366.
Chordia, Tarun, Richard Roll, and Avanidhar Subrahmanyam, 2001.
Market liquidity and trading activity. Journal of Finance 56
501-530.
Christie, Andrew A, 1982. The stochastic behavior of common
stock variances: Value, leverage and interest rate effects. Journal
of Financial Economics 10 407-432.
Collins, Daniel W., and S.P. Kothari, 1989. "An analysis of
intertemporal and cross-sectional determinants of earnings response
coefficients." Journal of Accounting and Economics 11 143-181.
D’Agostino, R.B., 1998. Tutorial in biostatistics propensity
score methods for bias reduction in the comparison of a treatment
to a non-randomized control group. Statistics in Medicine 17,
2265-2281.
Dal Bo, E., 2006. Regulatory Capture: A review. Oxford Review of
Economic Policy, Vol 22 Dechow, P. M., and I.D. Dichev, 2002. The
quality of accruals and earnings: The role of accrual
estimation errors. The Accounting Review 77, Supplement. Dechow,
P., R. Sloan and A. Sweeney, 1996. Causes and Consequences of
Earnings
Manipulation: An Analysis of Firms Subject to Enforcement
Actions by the SEC. Contemporary Accounting Research 13 (1):
1-36
DeFond, M. L., and J. Jiambalvo, 1991. Incidence and
circumstances of accounting errors. The
Accounting Review 66, 643-655. Diamond, D., and R. Verrecchia,
1991. Disclosure, Liquidity and Cost of Capital. Journal of
Finance, September. Djankov, S., La Porta, R, Lopez-de-Silanes,
F., and A. Shleifer, 2008a. The law and economics
of self-dealing. Journal of Financial Economics 88. 430-465
Durnev, A., Morck, R., and B. Yeung, 2004. Value-enhancing capital
budgeting and firm-
specific stock return variation. Journal of Finance 59 Issue
1.
-
36
El-Gazzar, S.M., 1998. Predisclosure information and
institutional ownership: A cross-sectional examination of market
revaluations during earnings announcement periods. The Accounting
Review 73 No. 1, 119-129.
Ertimur, Y., and M.E. Nondorf, 2006. The SEC comment letter
process and its effects on the
quality of disclosure for IPO firms. Working Paper, Duke
University. Fama, Eugene F., and Kenneth R. French. "Industry costs
of equity." Journal of Financial
Economics 43 (1997): 153-193.
Farber 2005. Restoring Trust after Fraud: Does Corporate
Governance Matter? The Accounting Review 80 (2): 539–561
Feroz, E., Park, K., and V. Pastena, 1991. The financial and
market effects of the SEC’s
Accounting and Auditing Enforcement Releases. Journal of
Accounting Research. Ferreira, M. A., and P.A. Laux, 2007.
Corporate governance, idiosyncratic risk, and information
Flow. Journal of Finance 62 Issue 2, 951-989. Harris, M., and A.
Raviv, 1993. Differences of opinion make a horse race. Review of
Financial
Studies 6, 473-506. Hayn, Carla, 1995. The information content
of losses. Journal of Accounting and Economics 20
125-153.
Healy, Paul M., Amy P. Hutton, and K. G. Palepu, 1999. Stock
performance and intermediation changes surrounding sustained
increases in disclosure. Contemporary Accounting Research 16
485-520.
Healy, P., and K. Palepu, 2001. Information asymmetry, corporate
disclosure and the capital markets: A review of the empirical
disclosure literature. Journal of Accounting and Economics.
Heckman, J., Ichimura, H., and P. Todd, 1997. Matching as an
econometric evaluation estimator:
Evidence from evaluating a job training program. Review of
Economic Studies 64, Issue 222, 605-654.
Heckman, J., Ichimura, H., and P. Todd, 1998. Matching as an
econometric evaluation estimator.
Review of Economic Studies 65, Issue 223, 261-294. Heckman,
James, and Salvador Navarro-Lozano, 2004. Using matching,
instrumental variables,
and control functions to estimate economic choice models. Review
of Economics and Statistics 86 30-57.
-
37
Heflin, Frank, K. R. Subramanyam, and Yuan Zhang, 2003.
"Regulation FD and the financial information environment: Early
evidence." The Accounting Review 78 1-37.
Jackson, H.E., and M.J. Roe, 2009. Public and private
enforcement of securities laws: Resource-based evidence, Journal of
Financial Economics 93, 207-238
Kandel, E., and N. Pearson, 1995. Differential interpretation of
public signals and trade in
speculative markets. Journal of Political Economy 103, 831-872.
Kim, O., and R.E. Verrecchia, 1991a. Trading volume and price
reactions to public
announcements. Journal of Accounting Research 29, 302-321. Kim,
O., and R.E. Verrecchia, 1991b. Market reaction to anticipated
announcements. Journal of
Financial Economics. Kim, O., and R.E. Verrecchia, 1994. Market
liquidity and volume around earnings
announcements. Journal of Accounting and Economics 17, 41-67.
Kim, O., and R.E. Verrecchia, 1997. Pre-announcement and event
period private information.
Journal of Accounting and Economics 24, 395-419. Lang, M. H.,
and R. J. Lundholm, 1996. Corporate disclosure policy and analyst
behavior. The
Accounting Review 71 467-492.
Lang, M., and R. Lundholm, 1993. Cross-sectional determinants of
analyst ratings of corporate disclosures. Journal of Accounting
Research 31 246-271.
La Porta, R. Lopez-de-Silanes, F., and A. Shleifer, 2006 What
works in securities laws? Journal
of Finance 61, 1-32 Leuz, C., and L. Hail, 2006. International
differences in the cost of equity capital: Do legal
institutions and securities regulation matter. Journal of
Accounting Research 44, Issue 3. Leuz, C., and R.E. Verrecchia,
2000. The economic consequences of increased disclosure.
Journal of Accounting Research 38, Issue 3, 91-124. Nelson,
Daniel B, 1991. Conditional heteroskedasticity in asset returns: A
new approach.
Econometrica 59 347-370.
Petersen, Mitchell A, 2009. Estimating standard errors in
finance panel data sets: Comparing approaches. Review of Financial
Studies 22 435-480.
Rogers, William, 1993. Regression standard errors in clustered
samples. Stata Technical Bulletin 13 19-22.
-
38
Rosenbaum, P. R., and D.B. Rubin, 1983. The central role of the
propensity score in observational studies for causal effects.
Biometrika 70, 41-55.
Rosenbaum, P.R., and D.B. Rubin, 1985. Constructing a control
group using multivariate
matched sampling models that incorporate the propensity score.
American Statistician 39 Issue 1, 33-38.
Schulte, D., 1988. The debatable case for securities disclosure
regulation. Journal of
Corporation Law 88, Winter. Tkac, P. A., 1999. A trading volume
benchmark: Theory and evidence. Journal of Financial and
Quantitative Analysis 34, 89-114. Utama, S., and W.M. Cready,
1997. Institutional ownership, differential predisclosure
precision
and trading volume at announcement dates. Journal of Accounting
and Economics 24, 129-150.
Welker, Michael, 1995. Disclosure policy, information asymmetry,
and liquidity in equity
markets. Contemporary Accounting Research 11 801-827.
-
39
Appendix 1
Mail Stop 0510 February 10, 2005 Via U.S. mail and facsimile
Gary T. Steele, President and Chief Executive Officer Landec
Corporation 3603 Haven Avenue Menlo Park, CA 94025 RE: Form 10-KSB
for the fiscal year ended May 30, 2004 Form 10-QSB for the period
ended August 29, 2004 File No. 0-27446 Dear Mr. Steele: We have
reviewed these filings and have the following comments. If you
disagree with a comment, we will consider your explanation as to
why our comment is inapplicable or a revision is unnecessary.
Please be as detailed as necessary in your explanation. In some of
our comments, we may ask you to provide us with supplemental
information so we may better understand your disclosure. After
reviewing this information, we may or may not raise additional
comments. Please understand that the purpose of our review process
is to assist you in your compliance with the applicable disclosure
requirements and to enhance the overall disclosure in your filing.
We look forward to working with you in these respects. We welcome
any questions you may have about our comments or on any other
aspect of our review. Feel free to call us at the telephone numbers
listed at the end of this letter. FORM 10-K FOR THE YEAR ENDED MAY
30, 2004 Comments applicable to your overall filing 1. Where a
comment below requests additional disclosures or other revisions to
be made, please show us in your supplemental response what the
revisions will look like. These revisions should be included in
your future filings. Item 7. Management`s Discussion and Analysis
of Financial Condition and Results of Operation Critical Accounting
Policies and Use of Estimates Revenue Recognition, page 23 2.
Please expand your disclosure to define what you refer to as
"recycled" revenue. Results of Operations Revenues Apio Trading,
page 25 3. Please expand your disclosure here and in footnote 12 to
include further information regarding the concentration of your
International sales in Asia and any other material geographies.
Corporate, page 26 4. You have disclosed the reason for the
decrease in revenue is due to a decrease in licensing revenue with
UCB and a decrease in research and development revenue associated
with a medical device company. Please expand your disclosure to
include further details regarding the closing of these agreements.
Please include in your disclosure whether the product licensed to
UCB can and will be licensed to other potential customers; whether
any additional revenue from royalties or licensing is expected as a
result of the research and development work performed for the
-
40
medical device company; and what your expectations are for the
coming year relating to licensing and research and development
revenue. Gross Profit Apio Trading, page 27 5. You have disclosed
on page 26 a change in certain export contracts. Please expand your
disclosure to include any impact these contract changes had or will
have on gross profit, if any. Liquidity and Capital Resources, page
32 6. You have disclosed on page 12 you are currently shipping
products to L`Oreal of Paris. You have also disclosed you will
receive royalty payments from Alcon on sales of the PORT(tm) device
through 2012. You have further disclosed on page 39, that you may
not receive royalties on future sales of QuickCast(tm) and PORT(tm)
because you no longer have control over the sales of these
products. Please expand your disclosure to include your
expectations regarding revenue from these products and any other
new products, product lines, or licensing and research and
development agreements. Also, please include in your disclosure how
not having control of these products may affect your ability to
receive royalties on these products. 7. You have disclosed on page
13 information regarding potential milestone payments relating to
an exclusive licensing and one year research and development
collaboration with a medical device company. Please expand your
disclosure to discuss the terms and status of this agreement and
whether or not you expect to meet any of these milestones. Please
also disclose the timing on if and when you anticipate revenue will
be earned through royalties. Contractual Obligations, page 34 8.
Please revise your table of contractual cash obligations to include
estimated interest payments on your debt. Because the table is
aimed at increasing transparency of cash flow, we believe these
payments should be included in the table. Please also disclose any
assumptions you made to derive these amounts. Additional Factors
That May Affect Future Results Our Indebtedness Could Limit Our
Financial and Operating Flexibility, page 35 9. You have disclosed
you may be obligated to make future payments to the former
shareholders of Apio of up to $1.2 million for the future supply of
produce. Please expand your disclosure to include the terms and
conditions that would cause you to incur this additional liability.
Please include in your disclosure any amounts that were accrued for
the periods presented and where these amounts were recorded in the
balance sheet and statement of operations. Please also indicate
when payments on these amounts are expected to be paid, if
applicable. Financial Statements Statements of Operations, page 49
10. Please revise your statements of operations to breakout
separately the cost of service revenue, related party. Statements
of Cash Flows, page 51 11. Please tell us which of the cash
outflows and inflows related to your notes and advances receivable
are included in operating activities and which are included in
investing activities. Please explain to us how you determined which
amounts belonged in each classification. In providing us a
response, please also tell us where the cash flows related to each
of the loans shown in Note 4 are included and explain why each loan
was classified where it was. Naturally, we understand that interest
earned on these notes and advances receivable would be included in
operating activities, regardless of where the principal amounts are
classified. In the event the repayments you receive exceed the
original principal amounts, for reasons other than stated interest
payments, please tell us how these amounts are treated in your cash
flow statement as well. If a portion of the repayments on these
receivables occurs with consideration other than cash, please
disclose how this works and how you take into account these
non-cash payments in preparing your statement of cash flows. If all
of the cash flows related to your investments in farming activities
are not included in the notes and advances receivable cash flows,
please separately address your classification for these cash flows
as well. Refer to paragraphs16, 17, 22 and 23 of SFAS 95. 12.
Please present the cash inflows and outflows related to your notes
and advances receivable on a gross basis. Otherwise, please explain
to us how they meet the criteria in SFAS 95 for netting.
-
41
Only cash flows stemming from investments, loans and debt with
original maturities of three months or less may be reported on a
net basis. 13. Please present cash flows related to the change in
other assets separately from those related to the change in other
liabilities. Please also present these cash flows on a gross basis,
rather than a net one. Please supplementally tell us how you
determined that these cash flows represented investing cash flows.
Refer to paragraphs 16 and 17 of SFAS 95. 14. Please present sales
of common stock and repurchases of common stock on a gross basis.
Please also present your stock repurchases separately in your
statement of changes in shareholders` equity. Please disclose in a
footnote the timing, nature and terms of your stock repurchases. If
these stock repurchases occurred under a stock repurchase program,
please discuss it as well. Notes to Financial Statements 15. Please
disclose the types of expenses that you include in the cost of
sales line item and the types of expenses that you include in the
selling, general and administrative expenses line item. Please also
disclose whether you include inbound freight charges, purchasing
and receiving costs, inspection costs, warehousing costs, internal
transfer costs, and the other costs of your distribution network in
the cost of sales line item. With the exception of warehousing
costs, if you currently exclude a portion of these costs from cost
of sales, please disclose: * in a footnote the line items that
these excluded costs are included in and the amounts included in
each line item for each period presented, and * in MD&A that
your gross margins may not be comparable to those of other
entities, since some entities include all of the costs related to
their distribution network in cost of sales and others like you
exclude a portion of them from gross margin, including them instead
in a line item, such as selling, general and administrative
expenses. 1. Organization, Basis of Presentation, and Summary of
Significant Accounting Policies Related Party Transactions, page 56
16. Your disclosure states that you have loss exposure on the
subleases fr