Audit committee characteristics and repeatedly meeting-beating analyst forecasts.
Rickling, Maria
I. INTRODUCTION
In recent years, the Securities and Exchange Commission (SEC 1999a,
2003), legislators (U.S. Senate 2002a, 2002b; SOX, 2002), and others
have stressed the importance of the audit committee in providing
effective oversight of the financial reporting process. Spurred by such
interest from legislators and regulators, many prior studies have
examined the association between audit committee characteristics and a
variety of financial reporting outcomes. As noted by Beasley et al.
(2009) and Carcello et al. (2011), prior research related to audit
committee characteristics has focused on audit committee composition and
diligence. Further, studies examining audit committee composition have
concentrated on two characteristics of audit committee directors:
independence and financial expertise. While earlier research tended to
focus on independence, in the aftermath of actions by the SEC (1999a,
2000) and the enactment of SOX (requiring all audit committee directors
to be independent), later research has focused on issues related to the
presence of financial experts on the audit committee (e.g., Yang and
Krishnan, 2005; Krishnan and Visvanathan, 2008; Krishnan and Lee, 2009;
Dhaliwal et al., 2010).
In an effort to identify what constitutes a "good" audit
committee, DeFond and Francis (2005) and Carcello et al. (2011) call for
additional research addressing specific characteristics of the audit
committee beyond independence and expertise. Audit committee member
tenure and the number of other directorships are two issues that have
received increasing attention from good governance advocates and others
in recent years.
DeZoort et al. (2002) call for an enhancement of the richness of
audit committee composition measures, specifically suggesting that the
"research could address whether ... current audit committee tenure
affects overall ACE [audit committee effectiveness]." A survey by
Heidrick and Struggles (2007) finds that 21 percent of companies had a
term-limit policy for directors and that this proportion had doubled
since 2000. Shareholder activists have recently put forth proposals at
many companies that there should be a limit on the number of years a
director can serve. (1) Reflecting this trend, publications issued by
the Big 4 accounting firms now discuss audit committee director tenure
issues (PricewaterhouseCoopers, 2000; Deloitte, 2010). For example,
Deloitte (2010) notes in its Audit Committee Brief that "[t]o be
most effective, audit committees should periodically reassess the
optimal mix of committee members, taking into account ... the skills,
experiences, diversity, time commitments, tenure, and rotation of its
members." Such views echo those espoused by some academics earlier
(Lapides et al., 2007).
During the Senate hearings related to the Enron failure, senators
and others raised the possibility of limiting the number of boards an
individual could serve on at any time (U.S. Senate, 2002b).
Increasingly, companies are having formal policies that restrict the
number of other directorships that can be held by their outside
directors. For example, General Motors noted in its 2007 proxy statement
that under its Corporate Governance Guideline No. 13,
"non-management directors are encouraged to limit the number of
other boards of U.S. public companies on which they serve, to no more
than four. ... Moreover, the Directors and Corporate Governance
Committee and the Board annually review whether members of GM's
Audit Committee serve on audit committees of other companies, and
whether that service compromises their ability to fulfill their duties
on GM's Audit Committee."
Thus, there is now an increased focus on audit committee member
characteristics beyond independence and financial expertise. Along these
lines, Carcello et al. (2011) note that "for good audit committees,
most of the focus is on audit committee financial expertise and
independence" and suggest that there is a "need to develop
better measures of board and audit committee characteristics." Yet,
there is limited prior research on audit committee member tenure and
"busyness."
Yang and Krishnan (2005) find that earnings management is lower
when audit committee members have longer tenure and have multiple other
board memberships; in contrast, Dhaliwal et al. (2010) show that
accruals quality is positively related to lower tenure and fewer board
memberships of the audit committee financial expert. In addition, Barua
et al. (2010) find that the extent of investment in internal auditing is
higher when audit committee members have shorter tenure. With respect to
busy-boarding, Sharma and Iselin (2006) find that restatements are less
likely when audit committee members have multiple board memberships, but
Sharma et al. (2009) find that multiple directorships are negatively
associated with audit committee meeting frequency. Thus, the results
from prior studies are not consistent; in addition, data from four of
the above five studies are from the pre-SOX period when there was also
greater variation in audit committee member characteristics, such as
independence and the presence of financial experts. Given the focus on
audit committee director tenure and busyness, in this paper I examine
the association between these two audit committee characteristics and
the likelihood that a company will repeatedly meet or just beat earnings
forecasts versus repeatedly missing earnings forecasts.
As noted by Carcello et al. (2011), many prior studies have
examined the notion of earnings management or earnings quality by
focusing on a variety of dependent variables, such as fraudulent
financial reporting, restatements, and accruals quality. Fraud and
restatement are relatively infrequent occurrences, while the accruals
quality metrics are prone to problems with model specification (e.g.,
Krishnan and Yang, 2007; Ball, 2009). In this paper, I use another
metric that has been used by prior researchers as a measure of earnings
management, namely meeting or just beating analysts' forecasts.
Importantly, the SEC's (1999d, 2009) statements and actions
indicate that meeting-just beating analysts' forecasts is
considered by the SEC to be a factor in evaluating the quality of
financial reporting. (2)
While some studies suggest that the likelihood of meeting or just
beating analyst earnings expectations has declined in the post-SOX
period (Koh et al., 2008; Bartov and Cohen, 2008), firms still appear to
engage in such activity and regulators are interested in such behavior.
For instance, in August 2009, General Electric agreed to pay a $50
million penalty to settle charges stemming from an investigation by the
SEC alleging that GE accounting executives approved improper application
of accounting standards, with one specific occasion allowing GE to
directly avoid missing analysts' expectations. The SEC (2009)
specifically noted in its complaint that GE consistently met or exceeded
financial analysts' consensus EPS expectations each quarter from
1995 through 2004.
In my analyses, I compare firms that repeatedly meet or just beat
analyst's forecasts against firms that just failed to meet
analyst's forecasts on multiple occasions during the period from
2005 to 2007. I find that audit committee director tenure and busyness
are positively associated with the likelihood of a firm repeatedly
meeting or just beating analyst forecasts. These results are consistent
with suggestions from governance advocates about the benefits related to
restricting audit committee member tenure (e.g., Lapides et al., 2007;
Deloitte, 2010) and service on multiple boards (NACD, 1996; CII, 1998;
U.S. Senate, 2002b).
The next section discusses the background. This is followed by a
development of hypotheses and then a description of method and data.
After a discussion of results, the paper ends with a summary and
conclusions.
II. BACKGROUND
A. Meet/Just Beat Strategy of Earnings Management
Analysts' forecasts serve as a proxy for the market's
expectations and are one of three performance measures that managers
seek to meet (Degeorge et al., 1999). There are at least two reasons as
to why managers wish to meet this benchmark: an expected valuation
premium and a reduced cost of capital.
The importance of the meet/just beat (MB) strategy of earnings
management could be attributed, at least in part, to empirical evidence
showing that investors reward firms with earnings that meet or beat
analysts' estimates by assigning a valuation premium and penalize
those that fall short of such estimates (Barth et al., 1999; Bartov et
al., 2002; Kasznik and McNichols, 2002). Graham et al. (2005) provide
additional survey-based evidence that is consistent with this notion of
a MB premium. They report that more than eighty percent of survey
respondents (CFOs) agree that meeting earnings benchmarks helps
"maintain or increase the stock price" and "build
credibility with capital markets." Furthermore, even in situations
where investors are capable of discerning the effect of earnings
management in order to achieve this expectation threshold, investors
discount the MB premium assigned but the discount is economically minor
and statistically insignificant (Bartov et al., 2002).
In addition, managers may want to consistently engage in the MB
strategy due to a lower cost of capital associated with a decrease in
information asymmetry. Brown et al. (2009) find that the decrease
(increase) in information asymmetry is larger for MB (Miss) firms who
have regularly met or beaten (missed) expectations over the prior eight
quarters and that this average reduction in asymmetry is significant
only when it is part of a repeated pattern. Chevis et al. (2007) find
that the valuation premium awarded by the market to MB firms increases
as the length of the period of continuous meet/beat activity increases.
They also present evidence suggesting that repeat MB firms enjoy higher
valuations of income and book value of equity than firms that only
periodically meet/just beat or do not meet/just beat at all. Graham et
al. (2005) report that seventy-eight percent of surveyed CFOs believe
that missing a benchmark "creates uncertainty about the firm's
future prospects."
III. DEVELOPMENT OF HYPOTHESES
A. Audit Committees Characteristics and Financial Reporting
The SEC (1999a) noted that "audit committees play a critical
role in the financial reporting system by overseeing and monitoring
management's and the independent auditors' participation in
the financial reporting process." Thus, the audit committee serves
as an important monitoring mechanism in ensuring high quality financial
reporting and can influence whether a firm will have the ability to
engage in the MB strategy of earnings management.
In their detailed summaries of research related to audit
committees, DeZoort et al. (2002), DeFond and Francis (2005), and
Carcello et al. (2011) note that almost all research related to audit
committee composition has focused on independence and financial
expertise. While earlier research on audit committees focused on
director independence, after the listing related changes mandated by the
NYSE and NASDAQ in 1999 (SEC 1999b, 1999c) and the requirement in SOX
that all audit committee members be independent, recent research has
tended to focus on audit committee financial expertise. Generally, prior
research finds audit committees that have financial experts, and are
diligent, are associated with higher quality financial reporting and
auditing using a variety of measures, such as fraudulent financial
reporting (Beasley et al., 2010), accruals (Yang and Krishnan, 2005;
Carcello et al., 2008), internal controls (Krishnan, 2005; Naiker and
Sharma, 2009), and going-concern reporting (Carcello and Neal, 2000).
Audit committees can also influence management's ability to
engage in earnings management through the extent of their support for
the positions of the external auditor. Carcello and Neal (2003) and
DeZoort et al. (2003) find that audit committees with financial experts
are more likely to support the external auditor.
In this study, I focus on two issues related to audit committee
composition that have started receiving attention from corporate
governance advocates and others yet have received sparse attention in
prior research: tenure on the audit committee and the number of other
corporate boards simultaneously served on by audit committee directors.
B. Audit Committee Member Tenure
During the Enron hearings, Charles Elson, Director of the Center
for Corporate Governance noted that:
"[T]here is concern about the length of directors' terms.
Directors who are on a board for too long, are viewed as becoming
effectively tired, not as sharp as they once were in reviewing the
company and much more willing to accept management representations than
not. That is why a number of folks have called for term limits for
directors."
(U.S. Senate, 2002b)
Directors with longer tenure are likely to be more closely
affiliated with management and less likely to challenge management
decisions (Boeker and Goodstein, 1993). Vafeas (2003) asserts that term
limits for board members should be considered by regulators in light of
evidence that long-tenured board members, as opposed to short-tenured
members, are more closely affiliated with management as exhibited by
long-tenured members' proclivity to approve higher compensation to
the CEO.
While the above arguments relate to directors in general, it is
likely that they are particularly applicable in the case of audit
committee directors. The 21st Century Governance and Financial Reporting
Principles (Lapides et al., 2007), endorsed by the Institute of Internal
Auditors, recommends that "the board should consider limiting the
number of years an individual can serve on the audit committee to ensure
adequate rotation of its members." DeZoort et al. (2002) note the
paucity of research related to director tenure and suggest that future
research address whether audit committee member tenure affects overall
audit committee effectiveness. While Yang and Krishnan (2005) find that
earnings management is less likely when audit committee members had
longer tenure, Dhaliwal et al. (2010) show that shorter tenure of the
financial expert on the audit committee is associated with better
quality of accruals in the post-SOX era.
The above discussion suggests the first hypothesis (stated in the
null form):
H1: The likelihood of a company repeatedly engaging in the MB
strategy of earnings management is not related to audit committee
director tenure.
C. Number of Other Directorships Held by Audit Committee Members
Fama (1980) and Fama and Jensen (1983) suggest that the number of
other board memberships can be viewed as a signal of the market's
assessment about a particular director. Under this reputation argument,
directors establish reputations for being effective monitors and are
rewarded with additional directorships. Thus, the higher the number of
other board memberships, the greater the expertise of a director and the
better the quality of monitoring provided by a director.
The counterpoint is that too many board memberships spread a
director thin, and thus reduces the quantity and/or quality of the
oversight provided by the director. For example, during the Enron
hearings held by the Permanent Subcommittee on Investigations of the
U.S. Senate's Committee on Governmental Affairs, the Senators posed
the following written question to some witnesses:
"Some directors of the Enron Board have been criticized for
their membership on numerous boards, calling into question their ability
to dedicate time and focus to issues at Enron. Would you be in favor of
limiting the number of corporate boards an individual may serve
simultaneously?"
(U.S. Senate, 2002b)
This view is espoused, among others, by the National Association of
Corporate Directors (NACD, 1996), the Council of Institutional Investors
(CII, 1998), and the New York Stock Exchange (SEC, 2003).
In the context of audit committees, Yang and Krishnan (2005) find
that earnings management is lower at firms where the audit committee
directors serve on multiple boards, but Dhaliwal et al. (2010) show that
accruals quality is positively related to accounting experts who hold
low levels of multiple directorships. Sharma and Iselin (2006) find that
restatements are less likely at firms where audit committee members have
multiple board memberships, while Fich and Shivdasani (2007) show that
"busy boards" significantly increase a firm's probability
of facing financial litigation.
Given such inconsistent evidence, I do not make a directional
prediction for the next hypothesis. Thus, hypothesis two, in the null
form, is:
[H.sub.2]: The likelihood of a company repeatedly engaging in the
MB strategy of earnings management is not related to the number of
additional board memberships held by audit committee directors. (3)
IV. METHOD AND DATA
A. Model
I base the empirical model on prior research seeking to explain the
likelihood of firms engaging in the MB strategy of earnings management.
The model is as follows:
MBMiss = [[beta].sub.0] + [[beta].sub.0]*Horizon +
[[beta].sub.2]*ForStd + [[beta].sub.3]*NumAnaly + [[beta].sub.4]*Lev +
[[beta].sub.5]*LitRisk + [[beta].sub.6]*MtoB + [[beta].sub.7]*LogMktVal
+ [[beta].sub.8]*Loss + [[beta].sub.9]*Big4 + [[beta].sub.10]*ACSize +
[[beta].sub.11]*ACXprt + [[beta].sub.12]*ACMeet +
[[beta].sub.13]*ACTenure + [[beta].sub.14]*ACBusy + [epsilon]
where MBMiss = 1 if actual earnings reported repeatedly exceeds the
analyst's forecast by one cent per share or less (i.e., $0.00 [less
than or equal to] Forecast Error [less than or equal to] $0.01), 0
otherwise (see below, following description of the sample, for a more
detailed explanation of "repeatedly"); Horizon = Forecast
horizon, measured as number of days between earnings announcement and
the day the most recent earnings forecast was made; ForStd = Forecast
dispersion, calculated as the standard deviation of earnings forecasts
during the 4th quarter of 2007; NumAnaly = Number of analysts making an
earnings forecast during the 4th quarter of 2007; Lev = Leverage, equal
to total liabilities divided by total assets; LitRisk = 1 if firm's
primary SIC code is 2833-2836, 3570-3577, 3600-3674, 5200-5961, or
7370-7370, 0 otherwise; MtoB = Market to book ratio, calculated as stock
price at fiscal year-end divided by book value per share; LogMktVal =
Natural log of market value of equity; Loss = 1 if the firm had a net
loss for fiscal year 2007, 0 otherwise; Big4 = 1 if auditor is a Big 4
auditor, 0 otherwise; ACSize = Square root of the number of audit
committee members during 2007; ACXprt = Square root of the number of
audit committee financial experts during 2007; ACMeet = Square root of
the number of audit committee meetings during 2007; ACTenure = Ratio of
the number of members serving on the audit committee for more than 7
consecutive years as of 2007; and ACBusy = Ratio of the number of
members on the audit committee who hold more than three other outside
directorships during 2007.
Before I discuss the rationale for including these specific control
variables in the model, I note that as part of my sensitivity tests I
use a variety of different measures for the audit committee related
variables.
Prior research has shown that the closer the earnings forecast is
made to the earnings announcement, the smaller the forecast error
(Crichfield et al., 1978; O'Brien, 1988; Brown, 1991; Sinha et al.,
1997). Since earnings forecasts and forecast revisions occur at
different times for different firms, I include the variable Horizon to
capture the number of days between the earnings announcement day and the
most recent earnings forecast available. Given that more recent
forecasts, or forecasts with a smaller horizon, tend to be more
accurate, I expect the coefficient on Horizon to be negative in the
above regression.
I attempt to control for cross-sectional differences in the
information environment that may affect forecast accuracy by including
the following variables: forecast dispersion (ForStd) as measured by the
standard deviation of all forecasts issued for the firm during the
fourth quarter, the number of analysts following the firm (NumAnaly), as
well as the logged market value of the firm (LogMktVal) as a measure of
firm size. Forecast dispersion captures the degree of uncertainty that
analysts have about the performance of the target firm, so I anticipate
the coefficient of ForStd to be negative. Conversely, since the number
of analysts following a firm represents the degree to which a firm is
followed, I anticipate the coefficient of NumAnaly will be positive. It
can be argued that larger firms have more resources to engage in the MB
strategy of earnings management. Therefore, I expect the coefficient on
LogMktVal to be positive. These directional predictions are based on
results from prior research, which show that forecast horizon and
dispersion are negatively related to forecast accuracy but positively
related to firm size and the number of analysts following the firm
(Atiase, 1985; Lys and Soo, 1995; Brown, 1997; Cheng and Warfield, 2005;
Chevis et al., 2007; Davis et al., 2009).
Chevis et al. (2007) find that the likelihood of meeting or just
beating forecasts is positively associated with highly leveraged firms.
I therefore include leverage (Lev) as a control variable and expect the
coefficient to be positive. Following Cheng and Warfield (2005) and
Frankel et al. (2002), I include variables to capture the growth of the
firm as represented by the firm's market-to-book ratio (MtoB),
whether the firm operates within a litigious industry (LitRisk), and
whether the firm experienced a loss for the fiscal year (Loss). In their
analysis of equity incentives and the probability of meeting or just
beating analysts' forecasts by $0.01, Cheng and Warfield (2005)
find that growth is significantly and negatively associated with the
likelihood to meet or just beat earnings forecasts, while firms in
litigious industries are more likely to meet or just beat forecasts.
Frankel et al. (2002) find that firms reporting small earnings surprises
are less likely to meet or just beat earnings forecasts. I predict the
coefficients of MtoB and Loss to be negative, and the coefficient of
LitRisk to be positive.
I include three other audit committee related variables in the
model. I include ACSize because as the number of audit committee members
increases it is likely that the extent of audit committee oversight will
increase (Raghunandan and Rama, 2007). Based on prior research
indicating audit committees that have members with financial expertise
are associated with generally better quality financial reporting
(Beasley et al., 2009), I include ACXprt in the model and expect this
variable to have a negative coefficient. Audit committees that hold more
frequent meetings are said to be more diligent and more effective in
their monitoring (DeZoort et al., 2002; Carcello et al., 2011). Based on
the above, I expect that more frequent audit committee meetings lead to
better monitoring and less earnings management; hence, I expect the
coefficient of ACMeet to be negative.
B. Data
Table 1 describes the sample selection process. I begin by
obtaining the universe of December 31 year-end firms having earnings and
forecast data in the Institutional Brokers Estimate System (I/B/E/S)
during the twelve quarters beginning in January 2005 and ending in
December 20074, yielding 3,521 firm-quarter observations.
Analysts make their earnings forecasts throughout the year, making
revisions as they receive new earnings-relevant information concerning
their target firms. Consequently, forecasts issued closer to the
earnings announcement date are based on a more rich information set and
thus tend to be more accurate than the preceding forecasts (Sinha et
al., 1997). Prior studies have documented this positive association
between forecast recency and forecast accuracy (Crichfield et al., 1978;
O'Brien, 1988; Brown, 1991; Sinha et al., 1997). Hence, I use the
most recent forecast issued prior to the earnings announcement date as
the analyst forecast measure.
I then calculate the forecast error, as actual earnings per share
less forecasted earnings per share. Meeting or just beating (just
missing) analyst earnings expectations are firm-quarter observations for
which actual earnings reported either meets or exceeds (misses) the
analyst's forecast by a cent per share or less, i.e., $0.00 [less
than or equal to] forecast error [less than or equal to] $0.01 (-$0.01
[less than or equal to] forecast error < $0.00). After removing
irregular observations or observations with missing data and foreign
firms, the initial sample is reduced to 3,205 firm-quarter observations.
I deleted 45 firms with either missing proxy data or missing
control variable data in Compustat. In addition, I deleted two firms
that switched to a non-December 31 fiscal year end month during the
analysis period, and one outlying firm. (5) This process yielded a final
overall sample of 3,157 firm-quarter observations.
Table 2 provides empirical evidence about the number of quarters in
which firms either meet or just beat analyst forecasts, or just missed
analyst forecasts. The table shows that 727 of the 3,157 firms (23
percent) did not meet, just beat, or just miss even once during the 12
quarters examined in this study.
The data in Table 2 show a very interesting pattern. While 2,204 of
the 3,157 firms (70 percent) either meet or just beat analyst forecasts
at least once during the 12 quarters, only 1,277 firms (40 percent) just
missed analyst forecasts at least once during the same time period.
Similarly, 563 of the 3,157 firms (18 percent) meet or just beat analyst
forecasts at least four times during the study period; in contrast, only
34 of the 3,157 firms (1 percent) just missed analyst forecasts at least
four times during the 12 quarters. Overall, fewer firms are likely to
just miss than to meet or just beat analyst forecasts.
I then partition the overall sample of firms into repeated
meet/just beat and repeated just miss groups. A firm is classified as a
repeated "Meet/Just Beat" firm if it had a net MB for at least
7 out of the 12 quarters. Thus, a firm that met or just beat the
earnings forecasts by $0.01 for 9 of the 12 quarters but also just
missed forecasts by $0.01 for 2 of the 12 quarters is still considered a
repeated "Meet/Just Beat" firm. A firm is classified as a
repeated "Just Miss" firm as long as the firm misses the
earnings forecast by $0.01 or less, at minimum, a net of 2 out of the 12
quarters. Thus, a firm that just missed the earnings expectation for 5
out of 12 quarters, but also met or just beat expectations for 3 of the
12 quarters is considered a repeated "Just Miss" firm. The
above process yields 77 repeated Meet/Just Beat firms and 79 repeated
Just Miss firms. (6)
All audit committee data were hand-collected from the firms'
proxy statements obtained from the SEC's website. After deleting
observations with missing data for variables in the regression model,
the final sample for the regression analysis includes 75 firms that
repeatedly meet or just beat analyst forecasts and 64 firms that
repeatedly just missed analyst forecasts.
V. RESULTS
A. Descriptive Statistics
Table 3 presents univariate results of differences between the two
groups. The repeated Meet/Just Beat firms have a shorter forecast
horizon (p < .10), less forecast dispersion (p < .01), lower
leverage (p < .05), and higher market value (p < 0.01); the
repeated Meet/Just Beat firms are also more likely to be in risky
industries (p < .05), and less likely to have losses (p < 0.05).
Turning to audit committee related variables, there are no
significant differences between the two groups in terms of the number of
audit committee directors, the number of experts or the number of
meetings. Further, the proportion of audit committee directors with more
than three other board memberships is not significantly different
between the two groups. However, the repeated Meet/Just Beat group of
firms has a higher proportion of long-tenured audit committee members (p
< .01).
B. Regression Results
Table 4 reports the results obtained from the regression model. The
explanatory power of the model, as measured by the Pseudo [R.sup.2], is
0.50; this is in line with those reported in prior research relating to
meeting-beating analysts' forecasts (Cheng and Warfield, 2005;
Chevis et al., 2007; Davis et al., 2009).
Consistent with expectations, the likelihood of a firm repeatedly
meeting or just beating analyst's forecasts is negatively
associated with number of days between the earnings forecast and the
earnings announcement (Horizon), forecast dispersion (ForStd),
market-to-book ratio (MtoB), loss (Loss) and a Big 4 auditor (Big4).
Also consistent with expectations and prior research, the likelihood of
engaging in repeated meet-just beat behavior is positively associated
with firm size (LogMktVal) and industry type (LitRisk).
With respect to the audit committee variables, consistent with
expectations, the coefficients of ACSize, ACXprt and ACMeet are negative
and significant indicating that meeting or just beating analyst's
forecasts is less likely in firms that have audit committees that have
more (a) members, (b) experts, and (c) meetings. The coefficient of
ACTenure is negative and significant indicating that when the audit
committee has a higher proportion of long-tenured members there is a
higher likelihood of a firm repeatedly meeting or just beating
analyst's forecasts. This result is consistent with the
management-friendliness hypothesis; that is, long tenured outside
directors become friendly with organizational management thereby
creating a less stringent oversight environment. The coefficient of
ACBusy is also negative and significant, indicating that as the
proportion of directors who serve on more than three boards increases,
the likelihood of the firm repeatedly meeting or just beating analyst
forecasts increases. This evidence supports the argument espoused by
good governance advocates that companies should consider limiting the
number of boards on which audit committee members serve concurrently.
C. Sensitivity Analyses
I perform the following additional analyses as part of sensitivity
tests. First, I recognize that my use of seven years (for audit
committee director tenure) and more than three boards (for concurrent
directorships) is necessarily arbitrary. Hence, I use a number of other
cutoff measures. For the tenure variable, I use the following
alternative cutoffs: five years and ten years. With each of these
alternative measures, the ACTenure variable remains negative and
significant (as in Table 4). However, for the ACBusy variable, when I
use more than two as the cutoff (instead of more than three), ACBusy is
not significant at conventional levels. Thus, it appears that the
difference arises once a director sits on more than three boards. In
such alternative regressions, the sign and significance of the other
variables in the model are generally similar to those presented in Table
4.
Next, instead of using the proportion of audit committee members
with tenure or board memberships above a specific threshold, I use the
average tenure and board membership measures for the audit committee.
With this alternative specification, the ACTenure variable is once again
negative and significant but the ACBusy variable is not significant.
Some prior studies use dummy variables for audit committee related
variables (e.g., if the committee met more than a specified number of
times a year). Further, some other studies have sought to distinguish
between accounting experts and other types of experts (e.g., Dhaliwal et
al., 2010). Hence, I use dummy variables for meetings and experts, in
lieu of ACMeet and ACXprt, as follows: ACMeetD = 1 if the committee met
more than the median number of meetings (8) of the sample, 0 otherwise;
ACAccXprt = 1 if an accounting expert is present on the audit committee,
0 otherwise. (7) With such alternative specification of the regression
model, I find that ACMeetD is negative and significant but ACAccXprt is
not significant. More importantly, the significance of the variables of
interest, namely ACTenure and ACBusy remains substantively similar to
those presented in Table 4.
I include an additional control variable measuring internal control
quality in the regression model. I define this dummy variable as
follows: ICW =1 if there is a material weakness in internal controls for
fiscal year 2007, 0 otherwise. ICW is not significant in the model, and
the sign and significance of the other variables remains substantively
similar to those presented in Table 4.
VI. SUMMARY AND CONCLUSIONS
The audit committee's role in ensuring high quality financial
reporting has long been recognized by the SEC and others. In this paper,
I examine the association between audit committee characteristics and
the likelihood of firms repeatedly meeting or just beating or just
missing analyst's earnings forecasts. Almost all prior research on
audit committee composition has focused on independence and financial
expertise of the audit committee directors. I extend the literature on
audit committee composition by focusing on two factors that have
received little attention in prior research, yet have become the focus
of legislators, good governance advocates, and others: audit committee
members' tenure and the number of other directorships.
I examine the propensity of 3,157 firms to meet/just beat or just
miss analyst forecasts by $0.01 or less during 2005 to 2007. I find that
firms are much more likely to repeatedly meet or just beat analyst
forecasts than to repeatedly just miss analyst forecasts. While this
finding may not be surprising, the frequency of such repeated behavior
may be informative: 70 percent of the sample firms either meet or just
beat analyst forecasts at least once during the 12 quarters, but only 40
percent of the sample firms just missed analyst forecasts at least once
during the same time period. Further, 18 percent of the sample firms
meet or just beat analyst forecasts at least four times during the study
period but only 1 percent of the sample firms just missed analyst
forecasts at least four times during the 12 quarters.
In my analyses, I focus on 75 firms that had a net meet or just
beat count of 7 out of the 12 quarters and 64 firms that had a net just
miss count of 2 out of the 12 quarters. I find that audit committee
director tenure is positively associated with the likelihood of a firm
repeatedly meeting or just beating analyst forecasts. This finding holds
whether I use the average number of years of audit committee member
tenure, or the proportion of directors with more than five, seven, or
ten years of audit committee tenure. These results provide strong
support for the argument that too long a service may lead to audit
committee members becoming less vigilant or more permissive of earnings
management, and support calls (e.g., Lapides et al., 2007) to restrict
the tenure of directors on the audit committee.
My results for the number of other directorships held by audit
committee members is mixed: when I use the proportion of directors
holding more than three other board memberships, the ACBusy variable is
positive and significant in the regression, indicating that audit
committees that have a higher proportion of members with four or more
other board memberships are less likely to prevent earnings management.
However, when I use average number of board memberships, or when I use
the proportion of audit committee directors holding more than two other
directorships, the ACBusy variable is not significant in the regression.
Since the NYSE and others have typically sought to limit the number of
board memberships to four, my results may be viewed as providing partial
support for efforts seeking to limit busy-boarding by audit committee
directors.
My results suggest many interesting avenues for future research.
One avenue is to examine the association between audit committee member
tenure and busy-boarding with other measures of audit quality and
financial reporting quality, particularly in the post-SOX period.
Another interesting area is to examine the reaction of external and
internal auditors when audit committee members have long tenure and/or
multiple board memberships. Finally, it is also interesting to examine
how audit committee member interactions and processes vary with the
tenure and busy-boarding of audit committee members.
REFERENCES
Atiase, R, 1985, "Predisclosure Information, Firm
Capitalization and Security Price Behavior around Earnings
Announcements," Journal of Accounting Research, 23(1), 21-36.
Ball, R, 2009, "Market and Political/Regulatory Perspectives
on the Recent Accounting Scandals," Journal of Accounting Research,
47(2), 277-323.
Barth, M., J. Elliott, and M. Finn, 1999, "Market Rewards
Associated with Patterns of Increasing Earnings," Journal of
Accounting Research, 37(2), 387-414.
Bartov, E., D. Givoly, and C. Hayn, 2002, "The Rewards to
Meeting or Beating Earnings Expectations," Journal of Accounting
and Economics, 33(2), 173-204.
Bartov, E., and D. Cohen, 2008, "Mechanisms to Meet/Beat
Analyst Earnings Expectations in the Pre-and Post-Sarbanes-Oxley
Eras," Working Paper, New York University.
Barua, A., D. Rama, and V. Sharma, 2010, "Audit Committee
Characteristics and Investment in Internal Auditing," Journal of
Accounting and Public Policy, 29, 501-513.
Beasley, M.S., J.V. Carcello, D.R. Hermanson, and T.L. Neal, 2009,
"The Audit Committee Oversight Process," Contemporary
Accounting Research, 26(1), 65-122.
Beasley, M.S., J.V. Carcello, D.R. Hermanson, and T.L. Neal, 2010,
Fraudulent Financial Reporting 1998-2007: An Analysis of U.S. Public
Companies, Durham, NC, Committee of Sponsoring Organizations of the
Treadway Commission (COSO).
Boeker, W., and J. Goodstein, 1993, "Performance and Successor
Choice: The Moderating Effects of Governance and Ownership,"
Academy of Management Journal, 36(1), 172-186.
Brown, L, 1991, "Forecast Selection When All Forecasts Are Not
Equally Recent," International Journal of Forecasting, 7(3),
349-356.
Brown, L, 1997, "Analyst Forecasting Errors. Additional
Evidence," Financial Analysts' Journal, 53(6), 81-88.
Brown, S., S.A. Hillegeist, and K. Lo, 2009, "The Effect of
Earnings Surprises on Information Asymmetry," Journal of Accounting
& Economics, 47(3), 208-225.
Carcello, J.V., and T.L. Neal, 2000, "Audit Committee
Composition and Auditor Reporting," The Accounting Review, 75(4),
453-467.
Carcello, J.V., and T.L. Neal, 2003, "Audit Committee
Characteristics and Auditor Dismissals Following "New"
Going-Concern Reports," The Accounting Review, 78(1), 95-117.
Carcello, J.V., C. Hollingsworth, A. Klein, and T. Neal, 2008,
"Audit Committee Financial Expertise, Competing Corporate
Governance Mechanisms, and Earnings Management," Working Paper,
University of Tennessee.
Carcello, J.V., D. Hermanson, and Z. Ye, 2011, "Corporate
Governance Research in Accounting and Auditing: Insights to Date and
Directions for Future Research," Working Paper, University of
Tennessee.
Cheng, Q., and T. Warfield, 2005, "Equity Incentives and
Earnings Management," The Accounting Review, 80(2), 441-476.
Chevis, G., S. Das, and K. Sivaramakrishnan, 2007, "Does it
Pay to Consistently Meet Analysts' Earnings Expectations?"
Working Paper, Baylor University.
Council of Institutional Investors (CII), 1998, Core Policies,
Positions and Explanatory Notes. Washington, DC.
Crichfield, T., T. Dyckman, and J. Lakonishok, 1978, "An
Evaluation of Security Analysts' Forecasts," The Accounting
Review, 53(3), 651-668.
Dhaliwal, D., V. Naiker, and F. Navissi, 2010, "The
Association between Accruals Quality and the Characteristics of
Accounting Experts and Mix of Expertise on Audit Committees,"
Contemporary Accounting Research, 27(3), 787-827.
Davis, L., B. Soo, and G. Trompeter, 2009, "Auditor Tenure and
the Ability to Meet or Beat Earnings Forecasts," Contemporary
Accounting Research, 26(2), 517-548.
Deloitte, 2010, Audit Committee Brief (June), Deloitte Development
LLC.
DeFond, M.L., and J.R. Francis, 2005, "Audit Research after
Sarbanes-Oxley," Auditing: A Journal of Practice & Theory,
Supplement, 5-30.
Degeorge, F., J. Patel, and R. Zeckhauser, 1999, "Earnings
Management to Exceed Thresholds," Journal of Business, 72(1), 1-33.
DeZoort, F.T., D.R. Hermanson, D.S. Archambeault, and S.A. Reed,
2002, "Audit Committee Effectiveness: A Synthesis of the Empirical
Audit Committee Literature," Journal of Accounting Literature, 21,
38-75.
DeZoort, F.T., D.R. Hermanson, and R.W. Houston, 2003, "Audit
Committee Support for Auditors: The Effects of Materiality Justification
and Accounting Precision," Journal of Accounting and Public Policy,
22(3), 175-199.
Fama, E. F, 1980, "Agency Problems and the Theory of the
Firm," Journal of Political Economy, 88, 288-303.
Fama, E.F, and M.C. Jensen, 1983, "Separation of Ownership and
Control," Journal of Law and Economics, 26(2), 301-325.
Fich, E.M., and A. Shivdasani, 2007, "Financial Fraud,
Director Reputation, and Shareholder Wealth," Journal of Financial
Economics, 86, 306-336.
Frankel, R., M. Johnson, and K. Nelson, 2002, "The Relation
between Auditors' Fees for Nonaudit Services and Earnings
Management," The Accounting Review, 77, 71-105.
Graham, J., C. Harvey, and S. Rajgopal, 2005, "The Economic
Implications of Corporate Financial Reporting," Journal of
Accounting and Economics, 40(1), 3-73.
Heidrick and Struggles, 2007, 10th Annual Corporate Board
Effectiveness Study 2006-2007, Chicago, IL.
Kasznik, R., and M.F. McNichols, 2002, "Does Meeting Earnings
Expectations Matter? Evidence from Analyst Forecast Revisions and Share
Prices," Journal of Accounting Research, 40(3), 727-759.
Koh, K., D. Matsumoto, and S. Rajgopal, 2008, "Meeting or
Beating Analyst Expectations in the Post-Scandals World: Changes in
Stock Market Rewards and Managerial Actions," Contemporary
Accounting Research, 25(4), 1067-1098.
Krishnan, G., and G. Visvanathan, 2008, "Does the SOX
Definition of an Accounting Expert Matter? The Association between Audit
Committee Directors' Accounting Expertise and Accounting
Conservatism," Contemporary Accounting Research, 25(3), 827-858.
Krishnan, J, 2005, "Audit Committee Quality and Internal
Control: An Empirical Analysis," The Accounting Review, 80(April),
649-675.
Krishnan, J, and J. Yang, 2007, "Were Former Andersen Clients
Treated More Leniently than Other Clients? Evidence from Going-Concern
Modified Audit Opinions," Accounting Horizons, December, 423-435.
Krishnan, J, and J. Lee, 2009, "Audit Committee Financial
Expertise, Litigation Risk, and Corporate Governance," Auditing; A
Journal of Practice & Theory, 28(1), 241-261.
Lapides, P.D., M.S. Beasley, J.V. Carcello, F.T. DeZoort, D.R.
Hermanson, T.L. Neal, and J.G. Tompkins, 2007, 21st Century Governance
and Audit Committee Principles, Corporate Governance Center, Kennesaw
State University.
Lys, T., and L. Soo, 1995, "Analysts' Forecast Precision
as a Response to Competition," Journal of Accounting, Auditing and
Finance, 10(4), 751-765.
Naiker, V., and D.S. Sharma, 2009, "Former Audit Partners on
the Audit Committee and Internal Control Deficiencies," The
Accounting Review, 84(2), 559-587.
National Association of Corporate Directors (NACD), 1996, Report of
the NACD Blue Ribbon Commission on Director Professionalism, NACD,
Washington, DC.
O'Brien, P, 1988, "Analysts' Forecasts as Earnings
Expectations," Journal of Accounting & Economics, 10(1), 53-83.
PricewaterhouseCoopers, 2000, Audit Committee Effectiveness: What
Works Best, 2nd edition, The Institute of Internal Auditors Research
Foundation, Altamonte Springs, FL.
Raghunandan, K., and D. Rama, 2007, "Determinants of Audit
Committee Diligence," Accounting Horizons, 21(3), 265-279.
Sharma, V.D., and E.R. Iselin, 2006, "Reputation, Tenure and
Compensation of Independent Audit Committee Members and Financial
Restatements," Paper presented at the American Accounting
Association Annual Conference, Washington D.C.
Sharma, V.D., V. Naiker, and B. Lee, 2009, "Determinants of
Audit Committee Meeting Frequency: Evidence from a Voluntary Governance
System," Accounting Horizons, 23(3), 245-263.
Sinha, P., L. Brown, and S. Das, 1997, "A Re-Examination of
Financial Analysts' Differential Earnings Forecast Accuracy,"
Contemporary Accounting Research, 14(1), 1-42.
Securities and Exchange Commission (SEC), 1999a, Letter from Lynn
E. Turner, Chief Accountant: Call for Academic Research on Key
Accounting Issues, Available at:
http://www.sec.gov/news/extra/aaacall.htm.
SEC, 1999b, NYSE Rulemaking, Release No. 34-42233; File No.
SR-NYSE-99-39, Washington, D.C., Government Printing Office.
SEC, 1999c, NASD Rulemaking, Release No. 34-42231; File No.
SR-NASD-99-48, Washington, D.C., Government Printing Office.
SEC, 1999d, SEC Staff Accounting Bulletin No. 99: Materiality.
Washington, D.C., Government Printing Office.
SEC, 2000, Final Rule: Audit Committee Disclosure, Release No.
34-42266; File No. S7-22-99, Washington, D.C., Government Printing
Office.
SEC, 2003, NASD and NYSE Rulemaking, Relating to Corporate
Governance, Release No. 34-48745, Washington, D.C., Government Printing
Office.
SEC, 2009, SEC Charges General Electric with Accounting Fraud,
Available at: http://www.sec.gov/news/press/2009/2009-178.htm.
Sarbanes-Oxley Act (SOX), 2002, Public Law No: 107-204, Washington,
D.C., Government Printing Office.
U.S. Senate, 2002a, Oversight Hearing on Accounting and Investor
Protection Issues Raised by Enron and Other Public Companies, Committee
on Banking, Housing and Urban Affairs, Washington, D.C., Government
Printing Office.
U.S. Senate, 2002b, The Role of the Board of Directors in
Enron's Collapse, Hearing before the Permanent Subcommittee of
Investigations of the Committee on Governmental Affairs, S. Hrg.
107-511, Washington, D.C., Government Printing Office.
Vafeas, N, 2003, "Length of Board Tenure and Outside Director
Independence," Journal of Business Finance and Accounting, 30(7/8),
1043-1064.
Yang, J., and J. Krishnan, 2005, "Audit Committees and
Quarterly Earnings Management," International Journal of Auditing,
9(3), 201-219.
Maria Rickling
M.E. Rinker, Sr. Institute of Tax and Accountancy
Stetson University, Deland, FL 32723
mricklin@stetson.edu
ENDNOTES
(1.) AT&T, American Express, Home Depot, Pfizer and United
Technologies are some examples of companies that have had shareholder
proposals related to director term-limits in recent years.
(2.) The SEC (1999d) noted that "among the considerations that
may well render material a quantitatively small misstatement of a
financial statement item" is "whether the misstatement hides a
failure to meet analysts' consensus expectations."
(3.) I note that for the first hypothesis I focus on the number of
years of service on the audit committee but for the second hypothesis I
focus on the number of additional board memberships. I do this for two
reasons. The first reason is to be consistent with the approach taken in
prior research (e.g, Yang and Krishnan, 2005; Barua et al., 2010;
Dhaliwal et al., 2010). The second reason is the fact that, among
others, the suggestions of Lapides et al. (2007) and Deloitte (2010)
focus on audit committee tenure presumably because a new member on the
committee may be more likely to view things from a different perspective
and/or may be more skeptical. However, for the second hypothesis I focus
on the total number of additional board memberships because, in the case
of busy-boarding related discussions, the focus is on the time it takes
to serve on boards; further, very few directors serve on more than three
audit committees.
(4.) I stop with 2007 because I also wanted to examine if the
propensity to repeatedly meet / just beat or miss analyst forecasts is
associated with subsequent negative restatements. Since, on average, it
takes more than two years before most subsequent restatements are
disclosed, I stopped with fiscal year 2007. However, the subsequent
analysis indicates that only 8 of the 156 sample firms (selected as
described below) had a negative restatement until June 30, 2010,
reinforcing the earlier suggestion that restatements are relatively
infrequent.
(5.) This firm was deemed an outlier at the sub-group regression
analysis stage. It had a market-to-book ratio value that was
approximately 40 times the size of the average ratio for the Just Miss
group.
(6.) I recognize that the above process is somewhat arbitrary.
However, because I had to hand-collect audit committee related data, I
wanted the sample to be of manageable size. Further, the very definition
of "repeated" implies more than two, so I used the "net
of 2" for the repeated Just Miss firms; using a net of 3 or more
for the Just Miss firms drastically reduces the sample size of this
group. With respect to the repeated Meet/Just Beat firms, I used
alternative cutoffs (net of 8 or 9 quarters); the results with such
alternative cutoffs are similar to those reported in the paper.
(7.) I define an accounting expert as someone who has experience as
an auditor or as a senior corporate executive in accounting or finance
(e.g., CFO, CAO, VP-Finance, etc.).
Table 1
Sample selection
Number of Firms
Total number of firm-quarter observations in I/B/E/S 3,521
Less: Foreign firms (316)
Less: Firms missing financial, proxy, or analyst data (45)
Less: Firms that switched fiscal year ends (2)
Less: Outlier firm (1)
Sample Size 3,157
Table 2
Number of quarters of meet/just beat or just miss analyst forecasts
This table presents the frequency of meeting or just beating
analyst forecasts, or just missing analyst forecasts, during the 12
quarters ending December 31, 2007. The shaded areas represent the
"repeated Meet/Just Beat" and "repeated Just Miss" groups of firm
observations. These two groups combined represent the sample of
firm observations used for the regression analysis.
Meet or Just Beat
Just Miss
# of
Quarters 0 1 2 3 4 5 6 7 8
0 727 492 282 162 95 54 23 15 20
1 176 208 159 122 73 52 18 16 9
2 34 64 54 36 37 25 22 11 2
3 12 18 10 17 14 16 5 8 1
4 3 1 3 7 4 1 3 -- 1
5 1 2 1 2 2 1 -- -- --
6 -- -- 1 -- 1 -- -- -- --
Total 953 785 510 346 226 149 71 50 33
# of
Quarters 9 10 11 Total
0 4 4 2 1,880
1 9 5 3 850
2 6 -- -- 291
3 1 -- -- 102
4 -- -- -- 23
5 -- -- -- 9
6 -- -- -- 2
Total 20 9 5 3,157
Table 3
Univariate tests of differences
This table presents univariate tests of differences between 75
firms that repeatedly meet or just beat analyst forecasts and 64
firms that repeatedly just missed meeting analyst forecasts. The
sample sizes differ from those depicted in the top-right and
bottom-left corners of Table 2 due to missing data for variables in
the regression model. The variables are defined as follows: Horizon
= Forecast horizon, equal to the number of days between earnings
announcement and the day the most recent earnings forecast was
made; ForStd = Forecast dispersion, calculated as the standard
deviation of earnings forecasts during the 4th quarter of 2007;
NumAnaly = Number of analysts making an earnings forecast; Lev =
Leverage, equal to total liabilities divided by total assets;
LitRisk = Indicator variable if firm's SIC code is 2833-2836,
3570-3577, 3600-3674, 5200- 5961, or 7370-7370, 0 otherwise; MtoB =
Market to book ratio, calculated as stock price at fiscal year-end
divided by book value per share; LogMktVal = Logged market value of
equity; Loss = Indicator variable equal to 1 if the firm had a net
loss for fiscal year 2007, 0 otherwise; Big4 = 1 if auditor is a
Big 4 firm, 0 otherwise; ACSize = Square root of the number of
audit committee members during 2007; ACXprt = Square root of the
number of audit committee financial experts during 2007; ACMeet =
Square root of the number of audit committee meetings during 2007;
ACTenure = Ratio of audit committee members having consecutive
tenure on the committee greater than 7 years; ACBusy = Ratio of
members on the committee who hold more than three other outside
directorships.
Repeated Meet/Just Beat Firms
(n = 75)
Variable Mean Median Std.
Dev.
Horizon 31.730 20.000 30.458
ForStd 0.020 0.011 0.028
NumAnaly 7.910 6.000 4.902
Lev 0.520 0.480 0.277
LitRisk 0.390 0.000 0.490
MtoB 3.249 2.797 2.645
LogMktVal 3.181 2.968 0.745
Loss 0.070 0.000 0.251
Big4 0.850 1.000 0.356
ACSize 1.901 1.732 0.200
ACXprt 1.268 1.000 0.375
ACMeet 2.775 2.828 0.492
ACTenure 0.298 0.250 0.287
ACBusy 0.078 0.000 0.145
Repeated Just Miss Firms p-value
(n = 64) from
tests of
differences
Variable Mean Median Std.
Dev.
Horizon 45.730 26.000 41.066 0.051
ForStd 0.083 0.030 0.250 0.000
NumAnaly 6.940 5.000 5.188 0.146
Lev 0.608 0.627 0.282 0.044
LitRisk 0.190 0.000 0.393 0.015
MtoB 4.340 2.041 6.716 0.174
LogMktVal 2.889 2.723 0.767 0.015
Loss 0.220 0.000 0.417 0.013
Big4 0.830 1.000 0.380 0.816
ACSize 1.923 1.732 0.234 0.756
ACXprt 1.284 1.000 0.408 0.784
ACMeet 2.801 2.828 0.601 0.774
ACTenure 0.174 0.000 0.240 0.006
ACBusy 0.056 0.000 0.137 0.243
Table 4
Regression results
This table presents the results from a logistic regression with
MBMiss as the dependent variable. MBMiss = 1 if a firm observation
repeatedly met or just beat the analyst's forecast by one cent per
share or less (i.e., $0.00 [less than or equal to] Forecast Error
[less than or equal to] $0.01), 0 otherwise (i.e. -$0.01 [less than
or equal to] Forecast Error < $0.00). The sample includes 75 firms
that repeatedly met or just beat analyst forecasts and 64 firms
that repeatedly just missed meeting analyst forecasts. Other
variables are defined as in Table 3.
Model: MBMiss = [[beta].sub.0] + [[beta].sub.1]*Horizon +
[[beta].sub.2]*ForStd + [[beta].sub.3]*NumAnaly +
[[beta].sub.4]*Lev + [[beta].sub.5]*LitRisk+ [[beta].sub.6]*MtoB +
[[beta].sub.7]*LogMktVal + [[beta].sub.8]*Loss +
[[beta].sub.9]*Big4 + [[beta].sub.10]*ACSize +
[[beta].sub.11]*ACXprt + [[beta].sub.12]*ACMeet +
[[beta].sub.13]*ACTenure + [[beta].sub.14]*ACBusy + [epsilon]
Model Chi-square = 65.511, p < .001; Nagelkerke [R.sup.2] = .502.
Variable Predicted sign Coefficient p-value
Intercept 5.618 0.030
Horizon - -0.016 0.027
ForStd - -25.108 0.002
NumAnaly + 0.000 0.498
Lev + 0.522 0.288
LitRisk + 2.539 0.001
MtoB - -0.212 0.002
LogMktVal + 1.451 0.005
Loss - -1.198 0.070
Big4 - -0.947 0.085
ACSize - -2.083 0.047
ACXprt - -0.973 0.093
ACMeet - -1.210 0.009
ACTenure ? 2.476 0.009
ACBusy ? 4.108 0.018
