Does Overlapping Membership on Audit and Compensation Committees Improve a Firm’s Financial Reporting Quality?
Nandini Chandar LeBow College of Business
Drexel University [email protected]
Hsihui Chang* LeBow College of Business
Drexel University [email protected]
Xiaochuan Zheng Accounting Department
Bryant University [email protected]
September 2008 *Please address correspondence to: Hsihui Chang KPMG Professor of Accounting LeBow College of Business Drexel University Matheson Hall, Suite 400 3141 Chestnut Street Philadelphia, PA 19104 Phone: 215.895.6979 Email: [email protected]
Acknowledgement: We appreciate the helpful comments and suggestions of participants at the 2008 annual meeting of the American Accounting Association at Anaheim, CA, and, in particular, those of Phil Stocken, our discussant.
Does Overlapping Membership on Audit and Compensation Committees Improve a Firm’s Financial Reporting Quality?
ABSTRACT
We investigate whether audit committee members of the board prove to be better monitors if they are also on the compensation committee, as they would be more attuned to compensation related earnings management incentives. Analyzing data on a sample of S&P 500 firms over the period 2003-2005, we find that ceteris paribus, firms with overlapping audit and compensation committees have higher financial reporting quality, proxied for by discretionary accruals than those without such overlap. Further, we find that there is an inverted U shaped relationship between overlapping magnitude and financial reporting quality, with an optimum overlap of about 47%. Our findings are robust to controlling for the percentage of CEOs’ incentive compensation and using accounting restatements as another proxy for financial reporting quality. Overall, our results suggest that there is knowledge spillover from the compensation committee to the audit committee, as reflected in higher financial reporting quality. We interpret this to suggest that some overlapping of the committee memberships may be beneficial to audit committee monitoring effectiveness.
Keywords: overlapping memberships, audit committee, compensation committee, financial reporting quality.
Data Availability: Data are publicly available from sources identified in the paper.
1. Introduction
Recent accounting irregularities in high profile public companies have been
attributed to earnings management practices by corporate managers. These practices
have raised concerns about the quality of public companies’ reported earnings1 and
garnered increasing attention by the Securities and Exchange Commission (SEC). At the
same time, the role of audit committees in ensuring the quality of financial reporting has
come under considerable scrutiny. The Sarbanes Oxley Act of 2002 (SOX) in particular
represents the most sweeping reform of US securities laws relating to corporate
governance, disclosures, and reporting requirements since the 1930s. For instance, SOX
requires that the audit committee be fully independent and financially literate. The role
of audit committees in monitoring a firm’s financial reporting quality is recognized as a
central concern in the smooth functioning of capital markets, and is a primary focus of
corporate governance reform in the post-SOX era.
Audit committee members need information about their company’s business
activities and economic contexts in order to perform their monitoring functions
effectively. An important source of this information to audit committees is senior
management, who possess information that is crucial to the valuation of their firm.
However, management incentives in making representations, disclosures and accounting
choices are likely driven by the structure of their compensation contracts (Watts and
Zimmerman, 1986, 1990). If audit committee members are more attuned to the
incentives provided by management compensation, they are more likely to incorporate
1 In an influential speech in September 1998, SEC Chairman Arthur Levitt raised concerns about firms’ earnings management practices. Following this speech, the SEC staff issued three staff accounting bulletins (SABs) intended to improve financial reporting transparency. These SABs describe several earnings management techniques used by managers and point to the undesirable consequences of earnings decreasing as well as earnings increasing manipulations.
1
them into their monitoring of the firm’s financial reporting process. Membership on the
board’s compensation committee is likely to provide directors with intimate knowledge
of senior management’s compensation-related incentives. We therefore investigate
whether membership on both compensation and audit committees is beneficial to audit
committee members in performing their financial reporting oversight function. We
predict that such overlapping membership is likely to result in higher financial reporting
quality due to this knowledge spillover effect.
The corporate governance consequence of overlapping membership on audit and
compensation committees has been a contentious issue, particularly in the post-SOX
period. In a special comment entitled “Best Practices in Audit Committee Oversight of
Internal Audit,” Moody’s Investors Service (2006) argues in favor of audit committee
members serving on the compensation committee, the rationale being that “…the audit
committee should have a thorough understanding of executive incentives and goals so
that it is aware of management’s motivations.” (p. 8, Moody’s Investors Service 2006).
In contrast, the Higgs Report (2003) considers it undesirable for any one individual to be
placed simultaneously on the audit, remuneration and nomination committees “in order
not to concentrate too much influence on one individual” (p. 59, Higgs 2003). Given the
significant role of audit committees in monitoring financial reporting quality and the
increasing concerns about the effect of compensation arrangements on earnings
management, there are potentially positive monitoring outcomes that can arise with
overlapping membership on the two committees. This study provides evidence on this
issue as it considers the corporate governance consequence of the micro structure of the
board that is yet to be studied in accounting research.
2
The corporate governance implications of board characteristics as well as auditee
characteristics have been investigated in prior accounting research (e.g., Klein 2002a).
However, little empirical research has been conducted to date on board structure and
monitoring effectiveness in general and the consequences of delegating board functions
to committees in particular. Understanding the costs and benefits of the board committee
structure is particularly important as boards typically operate through the use of
committees. We contribute to this area by considering the effect of overlapping
memberships on two of the most active and important board committees – the
compensation and audit committees – on the monitoring effectiveness of the audit
committee.
Additionally, prior literature on audit committees has been conducted almost
exclusively in the pre-SOX era. Due to the widespread regulatory changes as a result of
SOX as well as the likely adjustments made by capital market participants in response to
the changed environment, we choose to study audit committee effectiveness in the post-
SOX period. Thus, our study contributes to recent, post-SOX debates on board structure
and composition in practice as well.
Following prior research on audit committee effectiveness like Klein (2002a), we
choose non-financial S&P 500 companies as the starting point for our sample. As SOX
possibly reflects a major inflection point with respect to corporate governance and
disclosure, we examine the post-SOX sample period 2003-2005. Our analysis proceeds
in the following stages. First, we investigate whether firms that have overlapping audit
and compensation committee members have higher financial reporting quality compared
to firms in which there is no overlap of membership between these two committees.
3
Second, we examine whether the magnitude of this overlap is related to financial
reporting quality; in other words, if there is a benefit to overlapping membership, is this a
monotonically increasing effect?
We document several interesting findings. First, we find that after controlling for
other commonly used determinants of financial reporting quality, firms with overlapping
audit and compensation committees have higher financial reporting quality than those
without such overlap. However, we do not find that this association is monotonic; rather,
we find that there is an inverted U shaped relationship between the magnitude2 of the
overlap and financial reporting quality. This suggests that overlapping membership is
beneficial to an extent; beyond a certain point, however, the benefits of committee
overlap decline. Our findings have implications for policy deliberations on the
composition of board committees in general and audit committees in particular, as it
clarifies the benefits of overlapping committee members.
The remainder of the paper proceeds as follows. Section 2 reviews the relevant
literature and formulates the research hypotheses. Section 3 describes the research design
including sample selection and estimation models. Section 4 presents empirical results
and robustness checks. Section 5 concludes.
2. Related Literature and Hypotheses Development
Corporate boards assume an oversight role in mitigating agency problems
resulting from the separation of corporate ownership and managerial control (Jensen and
Meckling 1976; Fama and Jensen 1983). This oversight role involves appointing the
2 The magnitude of overlap is defined as the proportion of audit committee members that is also on the compensation committee.
4
CEO, setting incentive compensation for top management, approving and reviewing
business strategies, monitoring control systems, liaising with external auditors, etc.
Given its diverse responsibilities, the board of directors typically delegates its oversight
activities to different sub-committees. The audit committee is one of the most important
sub-committees and its main responsibility is to oversee financial reporting. Another key
board committee is the compensation committee, which has responsibility for
determining appropriate compensation packages for top management and assessing
executives’ performance.
The finance literature has documented the link between compensation structure
and financial performance, and the accounting literature has demonstrated that corporate
governance and compensation structures have a substantial impact on earnings
management. However, the link between the compensation committee’s oversight on
compensation contracts and performance and the audit committee’s oversight of financial
reporting quality is as yet unexplored. Below we review the literature on these separate
literature strands that we integrate in this paper.
2.1 Literature review
Accounting research has long examined the effectiveness of audit committees in
monitoring the quality of firms’ financial reporting. Early research investigates whether
the existence of an audit committee improves a firms’ financial reporting quality. The
general finding is that there is a positive association between the presence of an audit
committee and the firm’s financial reporting quality. For example, DeFond and
Jiambalvo (1991) document that overstatement errors in annual earnings are less likely
among firms that have an audit committee. Wild (1996) finds that the earnings response
5
coefficients for companies with audit committees are higher following the appointment of
the audit committee.
Significant recent changes in regulations relating to audit committees started in
February 1999, when the Blue Ribbon Committee (BRC) made ten recommendations for
improving the effectiveness of audit committees, as well as five guiding principles for
audit committees to follow as they devised company-specific policies. The BRC
recommendations resulted in changes in the audit committee requirements for firms listed
on all major exchanges. The Sarbanes Oxley Act of 2002 has significantly shaped and
restricted corporate board structure with the objective of improving corporate
governance. It has expanded the authority and responsibilities of audit committees, and
increased membership requirements and committee composition to include more
independent and well qualified directors, i.e., those who are designated as “financial
experts.” The SEC and the exchanges also responded to strengthen audit committee
requirements.
Recent research has therefore focused on the characteristics of audit committees
and their association with proxies of financial reporting quality of the firm. Two major
audit committee attributes examined thus far in the literature are independence and
financial expertise (e.g. Klein 2002a, 2002b; DeFond et al. 2005). A common argument
by proponents of independence is that audit committees consisting of independent
directors would lead to higher financial reporting quality, because they would have
reduced conflicts of interest and less incentive to sacrifice objectivity (SEC 1999).
6
Consistent with this argument, Klein (2002a) observes a negative relationship between
board or audit committee independence and abnormal accruals.3
While prior accounting research has focused on attributes like audit committee
independence and found a positive association with a firm’s earnings quality, the post-
SOX environment calls for an investigation of other attributes of audit committees that
potentially determine their effectiveness. This is because, in the post-SOX era, as audit
committees are required to be fully independent (Section 301, SOX), audit committee
independence is not likely to be associated with a firm’s financial reporting quality,
because of lack of variability in the attribute.
Given the complexities of financial reporting particularly for public companies, at
least a subset of audit committees should have financial reporting expertise (BRC 1999).
Consistent with this notion, Xie et al. (2003) document that an audit committee member
with financial expertise enhances committee performance. In addition, DeFond et al.
(2005) provide evidence that it is specific accounting expertise rather than the broadly
defined “financial expertise” that improves the audit committee’s effectiveness. Other
dimensions of audit committees that have been found to be associated with financial
reporting outcomes include audit committee size (DeZoort and Salterio 2001; Anderson
et al. 2004) and number of meetings (Menon and Williams 1994; Beasley et al 2000).
This stream of prior research has largely focused on incentives (e.g.,
independence) and information processing abilities (e.g., financial expertise and number
of meetings) of the audit committee in performing monitoring activities. Audit
committee independence and financial expertise are mainly related to how given
3 However, she does not find a significant difference between an audit committee with a majority of independent members and an audit committee made up of the more stringent 100% independent members as required by the NYSE and the NASDAQ.
7
information sets are processed; independence is a quality that results in unbiased
information processing, while financial expertise is associated with more effective
information processing. Since effective information processing depends on the
availability of sufficient information and knowledge, the validity of the emphasis on audit
committee independence and financial expertise depends on the accessibility of
information by audit committee members (Nowak and McCabe 2003). We argue that
audit committee members who are also members of the compensation committee of the
board have more direct knowledge of the managements’ compensation related incentives
in making accounting choices. They can therefore prove to be better monitors by
neutralizing managements’ propensity to make accounting choices that are driven by
their incentives to maximize their compensation.
There is a long stream of literature that has examined the relation between
executive compensation and earnings management. Theoretically, providing executives
with incentive compensation, such as stock options should align managers’ and
shareholders’ interests (Jensen and Meckling 1976). However, both conventional
wisdom and prior empirical results suggest that incentive compensation might provide
managers incentives to manage earnings.4 Accounting research has established that
compensation related incentives are associated with opportunistic earnings management
behavior.
Early research focused on accruals management incentives based on bonus plans
(Healy 1985; Healy et al. 1987; Gaver et al. 1995; Holthausen et al. 1995). More recent
research examines stock price based incentives, and the use of earnings management to
4 See, for example, the 9 January, 2004 New York Times article by Gretchen Morgenson: ‘‘Options packages encourage executives to fiddle books.”
8
affect managements’ wealth. Sloan (1996) for instance finds that managers use
aggressive earnings management to increase stock price at least in the short run. Beneish
and Vargus (2002) find that significantly positive abnormal accruals are associated with
insider sales of shares. There is also considerable evidence that stock options and
restricted stock compensation provide incentives for earnings management. Studies have
found that that stock option exercises provide incentives to manage earnings upward. For
example, Bartov and Mohanram (2004) find evidence of earnings management leading
up to option exercises. Cheng and Warfield (2005) find that stock-option exercises and
holdings provide incentives for firms to meet or beat earnings targets. Bergstresser et al.
(2006) find that firms make more aggressive assumptions on their defined benefit pension
plans in the period in which they exercise their options.
Research has also argued that option grants provide managers with an incentive to
manage earnings downward to lower the exercise price at the grant date. The general
argument is that earnings can be used to signal weaker performance before the grant,
indirectly lowering the exercise price of the options (Baker et al. 2003, Balsam et al.
2003). McAnally et al. (2008) extend this research by examining the relation between
option grants and missed earnings targets, an event that could have a large influence on
exercise prices. McAnally et al.’s (2008) results suggest that accruals are used to
intentionally miss earnings targets in advance of option grants.
Other recent studies have found a nondirectional relationship between stock based
incentive compensation and earnings management. Burns and Kedia (2006) find that
firms whose CEOs have significant options exposure are more likely to file earnings
restatements. Similarly, Bergstresser and Philippon (2006) and Cheng and Warfield
9
(2005) find that the level of earnings management is positively related to managements’
stock based incentives, particularly those provided by stock options.
We bring together these two separate streams of prior research by examining whether
audit committee members’ intimate knowledge of compensation related incentives to manage
earnings will improve their ability to monitor the financial reporting process, thereby
reducing earnings management.
2.2 Research hypotheses
Compensation plans are established to generate appropriate incentives to align the
interests of management and shareholders. These incentives are frequently tied to accounting
performance measures (Lambert 1993). Since the compensation committee is responsible for
establishing management’s compensation plan, it is very likely that members of this
committee would have a deep and nuanced understanding of the incentives that the plan
generates. This would also imply that compensation committee members have a more
thorough knowledge of management’s incentives to manipulate earnings. Thus, if an audit
committee member also serves on the firm’s compensation committee, this member could
use the knowledge about management’s compensation-driven incentives to unravel
opportunistic accounting choices that the management might make. Therefore, the overlap
between compensation committee and audit committee might be beneficial in reducing the
information asymmetry between audit committee and management, resulting in higher
financial reporting quality induced by improved monitoring by the audit committee.
Furthermore, directors’ liabilities have been increased substantially after SOX due to
increased accountability of directors (Linck et al. 2008). As a result, audit committee
10
members who also sit on compensation committees likely have greater incentives to use
knowledge and information obtained from serving on the compensation committee. This
may accentuate the positive association between overlapping committees and financial
reporting quality after SOX, which is the time period of our study.
As we have previously argued, overlapping membership with the compensation
committee potentially results in knowledge spillover that is useful to the audit committee’s
financial reporting oversight function. However, there are costs related to the degree of
overlap as well. As Laux and Laux (2007) argue, a potential advantage of delegating board
functions to committees is that using smaller subgroups can reduce the free-rider problem
that plagues larger groups. If there is a complete overlap committee, the subgroup structure
and its advantages break down. Since setting CEO pay and overseeing the financial
reporting system are two major, but different functions of a board (Laux and Laux 2007),
membership in both the audit and compensation committees is very time consuming,
exacerbating the free-rider problem with increased committee overlap. Further, high degree
of committee overlap likely leads to the dilution of effort and diffusion of responsibility
because of multi-accountability, which could mitigate the oversight function of the audit
committee.
If such overlapping membership is expected to improve financial reporting quality,
we further examine whether the magnitude of this overlap matters. Specifically, if
overlapping membership is positively associated with financial reporting quality, is it true
that the magnitude of overlap is monotonically associated with financial reporting quality?
In the limit, is the best outcome a 100% overlapping of audit and compensation committees?
We do not expect that there is such a monotonic relationship. Instead, we anticipate an
11
inverted U-Shaped relationship between the magnitude of the overlap and financial reporting
quality as there may be costs related to overlap that needs to trade-off against the benefits of
having overlapping committee members. This trade-off is expected to result in an
“optimum” degree of overlap.
Based on the above discussion, we specify the following hypotheses in alternate
form:
H1: Ceteris paribus, a firm with an overlap of audit committee and compensation committee members of the board has higher financial reporting quality than a firm that does not have such overlapping membership H2: Ceteris paribus, there is an inverted U-Shaped relationship between the magnitude of overlap and financial reporting quality.
3. Research Design
3.1 Sample selection
Consistent with prior research on audit committee effectiveness, our sample
consists of all S&P 500 firms that appear in the RiskMetrics database5 from 2003 to
2005. The RiskMetrics database contains corporate governance information on company
directors such as committee membership, independence, qualifications, work experience,
tenure and other such data. We use the Executive Compensation (EXECUCOMP)
database to gather data on CEO incentive pay, CEO tenure and CEO ownership. In
addition, we obtain accounting data from COMPUSTAT. Our sample selection and
distribution is reported in Table 1. Following Klein (2002a), we exclude all financial
institutions and foreign firms. There are 104 firm-year observations missing in the
RISKMETRICS database. We also eliminate 76 firm-year observations because of
missing data in EXECUCOMP, and 25 firm-year observations with missing 5 This database was formerly known as the Investor Responsibility Research Center (IRRC) database.
12
COMPUSTAT data. In addition, to minimize the impact of extreme observations, we
trim our data by removing the extreme one percent of observations based on the
distribution of discretionary accruals. Thus, our final sample as shown in Panel A of
Table 1 contains 1032 firm-year observations relating to 399 public companies and 1782
directors.
Table 1 Panel B provides the distribution of our sample by year and industry. The
number of firm-year observations across our sample years is somewhat uniform. More
than half of our sample firms are manufacturing firms.
[Insert Table 1 about here]
3.2 Empirical models
In our empirical models we use performance matched discretionary accruals (see
Kothari et al. 2005; Francis et al. 2005) as our primary measure of earnings quality since
Kothari et al. (2005) provide evidence that performance-matched abnormal accrual
measures have smaller error rates than abnormal accruals determined by the Jones model
and the modified Jones model. We examine whether our findings are robust to using
accounting restatements as an alternate proxy of financial reporting quality.
Our primary independent variable of interest is OVERLAP which we measure in
two ways – as a dummy specification and as a continuous variable. We provide below
more precise definitions of these and other variables related to audit committee and firm
characteristics.
To test our first hypothesis (H1), we specify the following estimation model:
)1()ln(__/_
___)_ln(
10
98765
43210
itit
itititititit
ititititit
ASSETSLEVNIABSNINEGBVMVHOLDCEO
TENURECEOSIZEBDINDBDOVERLAPPMDAABS
εββββββ
βββββ
+++Δ+++
+++++=
13
where ln(ABS_PMDA)6, the dependent variable, is the natural log of the absolute value
of abnormal accrual measured by the performance matched modified Jones model
(Kothari et al. 2005). In keeping with prior literature, firms with large abnormal accruals
(in absolute value) are assumed to have lower financial reporting quality compared to
firms with smaller absolute accruals. OVERLAP is the main independent variable, which
takes on a value of one if at least one audit committee member sits on the compensation
committee and zero otherwise. BD_IND measures board independence and is defined as
the proportion of independent board directors. We include BD_IND as an additional
independent variable as a strong board environment might influence audit committee
monitoring effectiveness. Further, prior literature documents that board independence is
positively related to financial reporting quality (e.g. Klein 2002a). BD_SIZE is the
number of board directors. We include BD_SIZE as a control variable because prior
studies document that audit committee size is negatively associated with earnings
management and the cost of debt (e.g. Andersen et al. 2004) and smaller boards are
associated with higher governance effectiveness (Yermack 1996). CEO_TENURE is
measured as the number of years the current CEO has been in his or her position.
Shivdasani and Yermack (1999) observe that there are entrenchment effects associated
with CEO tenure. CEO_HOLD is the proportion of common equity owned by the CEO.
MV/BV, measuring firm growth potential, is the ratio of market value to book value
measured at the beginning of the fiscal year. NEG_NI is an indicator for firms having
two or more consecutive years of negative income, ending on the fiscal year prior to the
shareholders’ meeting. ABS_ΔNI is the absolute value of the change in net income
between previous year and current year deflated by last year’s assets. Klein (2002a) 6 For simplicity, we suppress subscripts i and t.
14
documents that both NEG_NI and ABS_ ΔNI are positively related to earnings
management (e.g. Klein 2000a). LEV is financial leverage defined as the long-term debt
divided by last year’s assets. ln(ASSETS) represents firm size measured as the natural
log of total assets. LEV and ln(ASSETS) are included as control variables because
previous studies found that financial leverage and political costs (proxied by firm size)
affect earnings management (e.g. Warfield et al. 1995; Dechow and Sloan 1995).
Our second hypothesis examines whether the extent of overlapping membership
between audit and compensation committees matters. As we argue in Section 2, because
there are costs and benefits to such overlap, we expect an inverted U-Shaped relationship
between the percentage of overlap and financial reporting quality. Thus, we specify the
following estimation model to test our second hypothesis (H2):
)2()ln(__/____
__)_ln(
1110987
6543
2210
ititititititit
itititit
ititit
ASSETSLEVNIABSNINEGBVMVHOLDCEOTENURECEOSIZEBDINDBD
POVERLAPPOVERLAPPMDAABS
εβββββββββ
βββ
+++Δ++++++
+++=
where OVERLAP_P is measured as the proportion of audit committee members who also
sit on the compensation committee. OVERLAP_P2 is the square of OVERLAP_P. All
other variables are as defined in model (1).
Table 2 summarizes the variable definitions used in the above estimation models.
[Insert Table 2 about here]
4. Empirical Results
4.1 Descriptive statistics
Table 3 presents descriptive statistics for the key variables used in our study. A
typical board has approximately 10 directors, 4 of whom, on average, sit on the audit
15
committee.7 Practically all of the boards and audit committees in our sample are
independent, which is not surprising given the fact that we examine S&P 500 firms,
which are required to have independent directors. The mean OVERLAP is 0.64, which
means that almost two-thirds of the companies in our sample have at least one director
who is on both the audit and compensation committees. The average OVERLAP_P is
about 31%, indicating that about a third of the members of audit committees are also
members of the compensation committee. The average CEO tenure is around 6 years and
most of the CEOs have a very small percentage ownership (the upper quartile is 0.004) in
the company. Regarding accounting variables, the average absolute abnormal accrual is
4% of previous years’ total assets; the average market to book ratio is 3.71; around 15%
of our observations have two or more years of consecutive losses; and the average long
term debt is about 21% of the previous years’ total assets.
[Insert Table 3 about here]
Table 4 presents spearman correlations between all sets of variables. The
correlation between OVERLAP and the natural log of absolute value of abnormal
accruals (ln(ABS_PMDA)) is negative but not statistically significant. Similarly, the
magnitude of the overlap (OVERLAP_P) is insignificantly and negatively correlated with
ln(ABS_PMDA) before controlling for other contextual variables. In addition,
OVERLAP is negatively (p-value<0.01) correlated with BD_SIZE. Also, OVERLAP is
negatively (p-value<0.01) correlated with BD_IND, which is probably due to the fact that
both audit committee and compensation committee are now required to be 100%
independent .
[Insert Table 4 about here]
7 The NYSE and the NASDAQ require that each company at least has three audit committee members.
16
4.2 Regression results
Since we have unbalanced panel data, we run fixed-effects regressions for all of
our estimation models.
4.2.1 Overlapping committees and financial reporting quality
Table 5 reports the fixed-effects regression results of relating financial reporting
quality to the presence of overlapping committee members and other control variables.
The coefficient on OVERLAP is highly significant at conventional levels, providing
strong support to our hypothesis H1. This suggests that firms with at least one audit
committee member who also sits on the compensation committee have, on average, lower
levels of discretionary accruals than firms without such overlap in the post-SOX period.
The coefficients on the control variables are generally consistent with prior literature.
For example, as in Klein (2002a), we also find that firm size is negatively associated with
the discretionary accruals.
[Insert Table 5 about here]
4.2.2 Overlapping magnitude and financial reporting quality
Given that our hypothesis H1 is confirmed, we evaluate H2 to examine whether
the extent of overlap of membership between the audit and compensation committees
matters. The results, presented in Table 6, indicates that the coefficients for both
variables of interest, the percentage of overlap, OVERLAP_P and the square of the
overlap percentage, OVERLAP_P2 are statistically significant (p-value <0.05); however,
they have opposite signs (-0.7871 and 0.8385 respectively), indicating that there is an
inverted U-shaped relation between overlapping magnitude and financial reporting
quality. This suggests that overlapping membership is beneficial only to certain extent;
17
beyond that the costs of overlapping membership outweigh the benefits, resulting in
deterioration in financial reporting quality.8
To determine the kink point in the relationship between financial reporting quality
and overlapping percentage (the optimal overlapping level), we use the first order
condition (FOC) of the function [-0.7871*OVERLAP_P+0.8385*OVERLAP_P2 ]. Since
the FOC is: -0.7871 +2*0.8385 OVERLAP_P=0, we solve to obtain OVERLAP_P
=0.4694. That is, on average, holding everything else constant, when the overlapping
percentage is 47% in the post-SOX period, the abnormal accruals are the lowest,
implying that the financial reporting quality at that point is highest. Thus, our second
hypothesis (H2) is supported.
[Insert Table 6 about here]
4.3 Robustness checks
If knowledge of compensation related incentives in general makes audit
committee members better monitors of the financial reporting outcomes of the firm, a
concern might arise regarding whether the magnitude of CEOs’ incentive compensation
affects the association between overlapping committee membership and financial
reporting quality. Prior research has shown that a greater level of incentive compensation
provides management with higher incentives to manage earnings, resulting in a larger
monitoring benefit from the transfer of incentive-related information to audit committee
members. To address this concern, we introduce a new control variable, INCE, the
proportion of CEOs’ non salary compensation and its interaction term, OVERLAP*INCE
into our estimation models. The results are reported in Panel A of Table 7.
8 Our results remain qualitatively unchanged when we use a reduced sample comprised of only observations with overlapping committee members to evaluate the relationship between overlapping magnitude and financial reporting quality.
18
After controlling for the proportion of CEOs’ compensation-related incentives, we
still find a significantly (p-value <0.05) negative relation between overlapping
committees and the financial reporting quality. Also, we find that there is a significant
(p-value <0.1) mediating effect of OVERLAP on the association between INCE and
financial reporting quality. A joint test of the coefficients of OVERLAP and the
interaction term, OVELAP*INCE, shows negative and significant (p-value <0.05)
results, which indicates that, overall, there is a significantly positive relation between
OVERLAP and financial reporting quality, even after controlling for the percentage of
incentive compensation to CEOs.
To further evaluate the robustness of our empirical results, we repeat our tests of
H1, and H2 using another proxy for financial reporting quality, namely, the incidence of
accounting restatements (REST).9 REST is a dummy variable that takes on a value of 1
if the current year’s final statements need to be restated and zero otherwise. The results
are presented in Panel B of Table 7. Consistent with our prior results, we find a
significantly (p-value <0.1) negative relation between OVERLAP and the incidence of
accounting restatement. Also, as with our discretionary accrual measure of financial
reporting quality, we find a U shape relationship between OVERLAP_P and the
incidence of accounting restatement (p-value<0.05). These results indicate that our
previous findings regarding the association between overlapping memberships,
overlapping magnitude and financial reporting quality are robust to an alternate
specification of financial reporting quality.
[Insert Table 7 about here]
9 Only the restatements caused by the violation of GAAP are considered. In other words, those restatements due to routine accounting changes are excluded. We get the data from Audit Analytic database.
19
There may be reasons to believe that overlapping membership on audit and
compensation committees is affected by financial reporting quality. For example, it is
possible that a firm with better financial reporting quality is more likely to structure
boards with overlapping membership between the two committees. Thus, our analyses
suffer from a potential endogeneity issue. We address this self-selection or reverse
causality problem using the Heckman (1979) two step treatment effects model.10 In the
treatment effects model, we model OVERLAP as an endogenous choice. Note that the
first-stage probit model includes the proxy for financial reporting quality (the absolute
value of performance matched discretionary accruals) and all other control variables that
also enter the second-stage (outcome) regression. Results (untabulated) of the second-
stage regression show that the OVERLAP (treatment) coefficient is negative and
significant at the 5% level using absolute value of performance matched discretionary
accruals as dependent variable. In brief, the Heckman model estimations support our
main analysis, thereby providing robust evidence to suggest that on average, firms with
overlapping memberships between the audit and compensation committees have higher
financial reporting quality than firms without such overlapping membership.
Furthermore, our empirical results remain qualitatively unchanged even when we control
for number of audit committee meetings.11
5. Conclusion
This study investigates whether overlapping membership on the audit and
compensation committees is positively associated with financial reporting quality. In
contrast to prior literature on audit committee effectiveness, our study focuses on the 10 We choose Heckman model since the endogenous repressor (overlapping) is dichotomous. 11 Results after controlling for the frequency of audit committee meetings are available upon request.
20
post-SOX era as a way to enhance our understanding of the monitoring role of audit
committees in the new regulatory regime. Based on a sample of S&P 500 firms, we find
that firms with such overlapping members are associated, on average, with higher
financial reporting quality. In addition, we find that the magnitude of the overlap
matters. The relationship between overlapping magnitude and financial reporting quality
is non-linear and inverted U-shaped. This suggests that the benefits of overlapping
membership decline after a certain point. Additional overlapping beyond that point
appears to provide detrimental effects, as the costs of overlapping exceed the benefits.
Our results help clarify some of the on-going debate surrounding board structure
and have important practical implications. Given the limited number of qualified board
members available, the optimal allocation of individual board members to committees is
important. The findings of this study provide some guidance in implementing this task.
For instance, for a firm with four (the average audit committee size in our sample is 3.99)
audit committee members, assigning two12 audit committee members to the
compensation committee seems to be the optimal choice.
Audit committees, and their role in monitoring financial reporting practices of
firms have occupied center stage in the post-SOX era. In such a vastly changed
environment, extrapolating pre-SOX evidence into the future must be done with care. By
providing evidence that relies exclusively on data subsequent to SOX, our study takes an
initial step in understanding the effectiveness of audit committee governance in this new
stage of the evolution of our capital markets.
12 This is derived as follows: 4*0.4694 (the optimum overlap proportion) ≈2.
21
References Anderson, R.C.; S.A. Mansi; and D.M. Reeb. “Board characteristics, accounting report integrity, and the cost of debt.” Journal of Accounting and Economics 37 (2004): 315-342.
Baker, T.; D. Collins; and A. Reitenga. “Stock option compensation and earnings management incentives.” Journal of Accounting, Auditing and Finance 18 (2003): 557-582. Balsam, S.; H. Chen; and S. Sankaraguruswamy. “Earnings management prior to stock option grants.” Unpublished paper, Temple University and Georgetown University, 2003. Available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=378440. Bartov, E. , and P. Mohanram. “Private information, earnings manipulations, and executive stock-option exercises.” The Accounting Review 79 (2004): 889-920. Beasley, M.S.; J.V. Carcello; D.R. Hermanson; and P. D. Lapides. “Fraudulent financial reporting: consideration of industry traits and corporate governance mechanisms.” Accounting Horizons 14 (2000): 441-454. Beneish, M.D., and M.E. Vargus. “Insider trading, earnings quality, and accruals mispricing.” The Accounting Review 77 (2002): 755-791. Bergstresser, D.; D. M. Desai; and J. Rauch. “Earnings manipulation, pension assumptions and managerial investment decisions.” Quarterly Journal of Economics 121 (2006): 157-195. Bergstresser, D., and T. Philippon. “CEO incentives and earnings management.” Journal of Financial Economics 35 (2006): 371-390. Blue Ribbon Committee on Improving the Effectiveness of Corporate Audit Committees (BRC). 1999. Report and Recommendations of the Blue Ribbon Committee on Improving the Effectiveness of Corporate Audit Committees. New York, NY: New York Stock Exchange and National Association of Securities Dealers. Burns, N., and S. Kedia. “The impact of CEO incentives on misreporting.” Journal of Financial Economics 35 (2006): 371-390. Cheng, Q., and T.D. Warfield. “Equity incentives and earnings Management.” The Accounting Review 80 (2005): 441-476.
Dechow, P. , and R. Sloan. “Detecting earnings management.” The Accounting Review 70 (1995): 193-225.
DeFond, M.L., and J. Jiambalvo. “Incidence and circumstances of accounting errors.” The Accounting Review 66 (1991): 643-655.
22
DeFond, M.L.; R.N. Hann; and X. Hu. “Does the market value financial expertise on audit committees of boards of directors?” Journal of Accounting Research 43 (2005): 153-193. DeZoort, F.T., and S. Salterio. “The Effects of corporate governance experience and financial and audit knowledge on audit committee members judgments.” Auditing: A Journal of Practice & Theory 20 (2001): 31-47. Fama, E., and M.C. Jensen. “Separation of ownership and control.” Journal of Law & Economics. 26 (1983): 301-326. Francis, J.; R. LaFond; P. Olsson; and K. Schipper. “The market pricing of accruals quality.” Journal of Accounting and Economics 39 (2005): 295-327. Gaver, J.J.; K.M. Gaver; and J.R. Austin. “Additional evidence on bonus plans and income management.” Journal of Accounting and Economics 19 (1995): 3-28. Healy, P. M. “The impact of bonus schemes on the selection of accounting principles.” Journal of Accounting and Economics 7 (1985): 85–107. Healy, P.M.; S.K. Kang; and K.G. Palepu. “The effect of accounting procedure changes on CEO’s cash salary and bonus compensation.” Journal of Accounting and Economics, 9 (1987): 7-34. Heckman, J.J. “Sample selection bias as a specification error.” Econometrica 47 (1979): 153–161. Higgs, D. “Review of the role and effectiveness of non-executive directors.” 2003. Available at www.dti.gov.uk/cld/non_exec_review. Holthausen, R.; D. Larcker; and R. Sloan. “Annual bonus schemes and the manipulation of earnings. ” Journal of Accounting and Economics 19 (1995): 29-74. Jensen, M. C., and W. H. Meckling. “Theory of the firm: Managerial behavior, agency costs, and ownership structure.” Journal of Financial Economics 3 (1976): 305–360.
Klein, A. “Audit committee, board of director characteristics, and earnings management.” Journal of Accounting & Economics 33 (2002a): 375-400.
_______. “Economic determinants of audit committee independence.” The Accounting Review 77 (2002b):435-452.
Kothari, S. P.; A.Leone; and C. Wasley. “Performance matched discretionary accrual measures.” Journal of Accounting & Economics 39 (2005): 163-197.
23
Lambert, R. A. “The use of accounting and security price measures of performance in managerial compensation contracts.” Journal of Accounting and. Economics 16 (1993): 55-100.
Laux, C., and V. Laux. “Board committees, CEO compensation, and earnings management.” Unpublished paper, University of Texas, 2007. Available at SSRN: http://ssrn.com/abstract=887492.
Levitt , A. “The Numbers Game.” CPA Journal 68 (1998): 12-14. Linck, J.; J. Netter; and T. Yang. “The effects and unintended consequences of the Sarbanes_Oxley Act on the supply and demand for director.” AFA Meeting (2008), Boston, MA. McAnally, M. L.; A. Srivastava; and C. Weaver. “Executive stock options, missed earnings targets and earnings management.” The Accounting Review 83 (2008): 185-216. Menon, K., and J. D. Williams. “The use of audit committees for monitoring.” Journal of Accounting and Public Policy 13 (1994): 121-139. Moody’s Investors Service. “Best practices in audit committee oversight of internal audit.” 2006. Report Number: 99909
Nowak, M.J., and M. McCabe. “Information costs and the role of the independent corporate director.” Corporate Governance: An International Review 11 (2003): 300-307.
Securities and Exchange Commission (SEC). “Audit Committee Disclosure.” 1999 Release No. 34-42266. File No. S7-22-99. Government Printing Office, Washington, DC,.
Shivdasani, A., and D. Yermack. “CEO involvement in the selection of new board members: An empirical analysis.” Journal of Finance 54 (1999): 1829-1853. Sloan, R. G. “Do stock prices fully reflect information in accruals and cash flows about future earnings?” The Accounting Review 71 (1996): 289–315. Watts, R., and J. Zimmerman. Positive Accounting Theory. Prentice-Hall, New York, 1986. Watts, R., and J. Zimmerman. “Positive Accounting Theory: A Ten Year Perspective.” The Accounting Review, 65 (1990), 131-156. Warfield, T.D.; J.J. Wild; and K.L. Wild. “Managerial ownership, accounting choices, and informativeness of earnings.” Journal of Accounting and Economics 20 (1995): 61–92.
24
Wild, J. “The audit committee and earnings quality.” Journal of Accounting, Auditing & Finance 11 (1996): 247-276.
Xie, B.; W. Davidson III; and P. DaDalt. “Earnings management and corporate governance: The role of the board and the audit committee.” Journal of Corporate Finance 9 (2003): 295-316. Yermack, D. “Higher market valuation of companies with a small board of directors.” Journal of Financial Economics 40 (1996): 185–211.
25
TABLE 1 Sample Selection
Panel A: Sample selection process
Firm-YearsInitial S&P 500 Sample for 2003–2005 1500
Financial Institutions (four-digit SIC code: 6000–6999) (230)Foreign firms (11)Missing data in RISKMETRICS Directors Database (104)Missing data in ExecuComp Database (76)Missing data in Compustat (25)Trimming top and bottom 1% on discretionary accruals (22)
Final sample 1032 Panel B: Sample distribution by year and industry
Year Construction Industries Manufacturing Mineral
Industries Retail Trade Service Industries
Transportation and Communication Wholesale Trade Total
2003 3 199 12 28 47 52 9 350
2004 5 195 11 31 45 51 9 347
2005 5 191 15 29 41 47 7 335
Total 13 585 38 88 133 150 25 1032
26
TABLE 2 Variable Definitions
Variables Operationalization Acronym Sources 1.Dependent variable Earnings quality Absolute value of performance matched abnormal
accruals measured as abnormal accrual minus the median abnormal accrual for a portfolio of firms (with the same 2- digit SIC code Year) in the same ROA decile
ABS_PMDA
COMPUSTAT
2. Test variables Overlap between audit and compensation committees Extent of committee overlap
Dummy variable which takes value 1 if at least one audit committee member sits on compensation committee, and zero otherwise Total number of audit committee members who also sit on the compensation committee divided by the audit committee size
OVERLAP OVERLAP_P
RISKMETRICS RISKMETRICS
3. Board characteristics Board independence Proportion of independent board director BD_IND RISKMETRICS Size of board Number of board directors BD_SIZE RISKMETRICS 4.Other control variables
CEO ownership Proportion of common equity owned by the CEO CEO_HOLD ExecuComp
CEO power
Number of years the current CEO has been in the position
CEO_TENURE
ExecuComp
Firm growth Market value over book value of assets, measured at the beginning of the fiscal year
MV/BV COMPUSTAT
Firm leverage Long-term debt divided by previous year’s assets LEV COMPUSTAT Firm financial distress
An indicator if the firm had two or more consecutive years of negative income, ending on the fiscal year prior to the shareholders’ meeting.
NEG_NI COMPUSTAT
Volatility The absolute value of the change in net income between previous year and current year (deflated by last year’s assets)
ABS_∆NI COMPUSTAT
Firm size The natural log of the book value of assets.
Ln(ASSETS) COMPUSTAT
27
28
Variable Mean Std DevLower
Quartile MedianUpper
Quartile
This See Table 2 for all variable definitions.
TABLE 3
Descriptive Statistics of Variables (N=1032)
ABS_PMDA
OVERLA P
OVERLA P_P
BD_SIZE
BD_IND
CEO_TENURE
CEO_HOLD
0.041
0.641
0.306
10.199
0.715
6.091
0.008
0.039
0.480
0.295
2.103
0.183
6.550
0.027
0.013
0.000
0.000
9.000
0.625
2.000
0.000
0.028
1.000
0.250
10.000
0.750
4.000
0.001
0.055
1.000
0.500
11.000
0.833
8.000
0.004
7
0
4
4
5
MV/BV
NEG_NI
ABS_∆NI
LEV
ln(ASSETS)
3.708
0.152
0.045
0.207
9.040
12.626
0.359
0.115
0.154
1.158
1.812
0.000
0.008
0.095
8.146
2.949
0.000
0.021
0.191
8.987
4.40
0.00
0.04
0.30
9.81
29
Variabl
TABLE 4 Spearman Correlations of Major Variables
e ln(ABS_PMDA)
OVERLAP OVERLAP_P
BD_IND BD_SIZE CEO_TENURE
CEO_HOLD
MV/BV Neg_NI ABS_∆NI LEV ln(ASSETS)
ln(ABS_PMDA) 1.00
OVERL
OVERL
BD_IN
BD_SIZ
CEO_T
CEO
AP -0.05 1.00
AP_P -0.01 0.85*** 1.00
D 0.00 -0.13*** -0.18*** 1.00
E -0.12*** -0.14*** -0.23*** 0.04 1.00
ENURE 0.00 0.05* 0.10*** -0.05 -0.13*** 1.00
_HOL
MV/BV
Neg_NI
D 0.00 0.10*** 0.11*** -0.14*** -0.16*** 0.50*** 1.00
0.05 -0.02 0.00 -0.04 -0.03 0.03 -0.06** 1.00
0.08*** 0.04 0.03 0.01 -0.09*** -0.13*** -0.08*** -0.25*** 1.00
NI 0.22*** -0.02 0.00 -0.01 -0.16*** 0.00 -0.06* 0.05 0.39*** 1.00
-0.15*** 0.09*** 0.05* 0.08** 0.21*** -0.11*** -0.02 -0.23*** 0.05* -0.18*** 1.00
ET S) -0.18*** -0.09*** -0.11*** 0.09*** 0.44*** -0.1*** -0.25*** -0.22*** -0.02 -0.19*** 0.31**
*
1.00
ABS_∆
LEV
ln(ASS
This table reports the Spearman correlations of major variables. See table two for all variable definitions. *, **, ***, indicate significance at the 10, 5 and 1 percent level respectively.
TABLE 5 Association between Overlapping Committees and Financial Reporting Quality
Variables Pred. Signs
Estimated Coefficients p-value
INTERCEPT -2.3951 0.0049
OVERLAP - -0.2111 0.0073
BD_IND - 0.0548 0.4020
BD_SIZE ? -0.0059 0.3950
CEO_TENURE + 0.0002 0.4892
CEO_HOLD + 2.3405 0.0768
MV/BV + -0.0009 0.3894
NEG_NI + 0.1897 0.0552
ABS_NICHANGE + 0.7565 0.0192
LEV ? 0.1196 0.7100
ln(ASSETS) - -0.1442 0.0006
Adj R-Sq
N
0.1015
1032
.
This table presents the fixed effects (year & industry ( SIC two digits)) regression results for following cross-sectional model:
itit
itititititit
ititititit
ASSETSLEVNIABSNINEGBVMVHOLDCEO
TENURECEOSIZEBDINDBDOVERLAPPMDAABS
εββββββ
βββββ
+++Δ+++
+++++=
)ln(__/_
___)_ln(
10
98765
43210
See table 2 for variable definitions. P-values are based on one tailed (two tailed) tests in instances where we have directional (nondirectional ) predictions.
TABLE 6 Association between Overlapping Magnitude and Financial Reporting Quality
Pred. Signs
Estimated Coefficients p-value
INTERCEPT -2.5321 0.0033
OVERLAP_P - -0.7871 0.0194
OVERLAP_P2 + 0.8385 0.0275
BD_IND - 0.0821 0.3563
BD_SIZE ? 0.0028 0.4519
CEO_TENURE + -0.0012 0.4338
CEO_HOLD + 2.5621 0.0601
MV/BV + -0.0007 0.4125
NEG_NI + 0.1862 0.0589
ABS_NICHANGE + 0.7622 0.0186
LEV ? 0.1026 0.7499
ln(ASSETS) - -0.1461 0.0006
Adj R-Sq 0.0990
N
1032
.
Optimal OVERLAP_P 0.4694
This table presents the fixed effects (year & industry ( SIC two digits)) regression results for following cross-sectional model:
ititititititit
itititit
ititit
ASSETSLEVNIABSNINEGBVMVHOLDCEOTENURECEOSIZEBDINDBD
POVERLAPPOVERLAPPMDAABS
εβββββββββ
βββ
+++Δ++++++
+++=
)ln(__/____
__)_ln(
1110987
6543
2210
See table 2 for variable definitions. P-values are based on one tailed (two tailed) tests in instances where we have directional (nondirectional) predictions. By using the first order condition (FOC) of the function [-0.7871*OVERLAP_P+0.8385*OVERLAP_P2 ], we find that the optimal OVERLAP_P is 0.4694. That is, on average, holding everything else constant, when the overlapping percentage is 47% in the post-SOX period, the abnormal accruals are the lowest.
31
TABLE 7
Robustness Checks Panel A: Controlling for the level of CEO’s incentive compensation (INCE )
Without Interaction With Interaction
Variables
Pred.
Signs
Estimated
Coefficients p-value
Estimated
Coefficients p-value
INTERCEPT -2.4241 0.0047 -2.7750 0.0022
OVERLAP - -0.2107 0.0074 0.3476 0.2220 0.0527 OVERLAP*INCE - -0.6807
INCE + 0.0745 0.3934
0.4059
0.3964
0.4986
0.0734
0.3887
0.0549
0.0200
0.7070
0.0006
0.5026 0.1261
0.3859
0.3846
0.4558
0.0523
0.3844
0.0555
0.0211
0.6811 0.0005
BD_IND - 0.0527 0.0642 BD_SIZE ? -0.0058 -0.0065 CEO_TENURE + 0.0000 -0.0008 CEO_HOLD + 2.4123 2.7289 MV/BV + -0.0009 -0.0009 NEG_NI + 0.1903 0.1895 ABS_∆NI + 0.7518 0.7433 LEV ? 0.1210 0.1323 ln(ASSETS) - -0.1459 -0.1483 Adj R-Sq 0.1007 0.1010
N 1032 1032
Joint test: OVERLAP+OVERLAP*INCE=0
-0.3330
0.0053
32
Panel B: Accounting restatements (REST) as a proxy for financial reporting quality
Equation (a) Equation (b)
Variables
Pred.
Signs
Estimated
Coefficients p-value
Estimated
Coefficients p-value
INTERCEPT -5.4990 0.4574 0.0816
-5.3800 0.4583
OVERLAP - -0.2715
OVERLAP_P - -1.6307 0.0270
0.0572
0.2020
0.1330
0.1274
0.3855
0.2059
0.4823
0.0015
0.8148 0.0671
OVERLAP_P2 + 1.4947 BD_IND - -0.4230 0.2004
0.1126
0.1137
0.3686
0.1944
0.4835
0.0015
0.8095 0.0548
-0.4254 BD_SIZE ? -0.0619 -0.0594 CEO_TENURE + 0.0179 0.0172 CEO_HOLD + -1.2253 -1.0701 MV/BV + -0.0051 -0.0049 ABS_∆NI + -0.0269 -0.0293 NEG_NI + 0.7121 0.7143 LEV ? 0.1738 0.1690 ln(ASSETS) - 0.1572 0.1480 N 1032 1032
Likelihood-Ratio-Pr>Chi-Sq <0.0001 <0.0001
Pseudo R-Sq 0.1364 0.1381
Panel A of this table presents the results of robustness checks by controlling for the magnitude of CEO’s incentive compensation. The results are from the fixed effects (year & industry ( SIC two digits)) regression for following cross-sectional models: (i) Without interaction of OVERLAP and INCE
ititit
itititititit
ititititit
ASSETSLEVNIABSNINEGBVMVHOLDCEOTENURECEO
SIZEBDINDBDINCEOVERLAPPMDAABS
εβββββββ
βββββ
+++Δ++++
++++=
)ln(__/__
__)_ln(
1110
98765
43210
(ii) Including interaction of OVERLAP and INCE
ititititit
itititititit
ititititit
ASSETSLEVNIABSNINEGBVMVHOLDCEOTENURECEOSIZEBDINDBD
INCEINCEOVERLAPOVERLAPPMDAABS
εβββββββββ
ββββ
+++Δ++++++
++++=
)ln(__/____
*)_ln(
1211109
87654
3210
INCE is the proportion of CEOs’ non salary compensation. See table 2 for other variable definitions. P-values are based on one tailed (two tailed) tests for variables whose relation to the dependent variables is (is not) predicted.
33
Panel B shows the results when we use the incidence of accounting restatement as an alternative proxy for earning quality. The results are from following fixed effects (year & industry ( SIC two digits)) logistic regression models:
)()ln(__/_
___
10
98765
43210
aASSETSLEVNIABSNINEGBVMVHOLDCEO
TENURECEOSIZEBDINDBDOVERLAPREST
itit
itititititit
ititititit
εββββββ
βββββ
+++Δ+++
+++++=
)()ln(__/__
____
1110
98765
432
210
bASSETSLEVNIABSNINEGBVMVHOLDCEOTENURECEO
SIZEBDINDBDPOVERLAPPOVERLAPREST
ititit
itititititit
ititititit
εβββββββ
βββββ
+++Δ++++
+++++=
REST is a dummy variable that takes on a value of 1 if current year’s final statements need to be restated and zero otherwise. See table 2 for other variable definitions. P-values are based on one tailed (two tailed) tests for variables whose relation to the dependent variables is (is not) predicted.
34
Top Related