Accounting Research Journal

Audit quality and overvalued equity

Robert Houmes, Maggie Foley, Richard J. Cebula,
Article information:
To cite this document:
Robert Houmes, Maggie Foley, Richard J. Cebula, (2013) "Audit quality and overvalued equity", Accounting
Research Journal, Vol. 26 Issue: 1, pp.56-74, https://doi.org/10.1108/ARJ-08-2011-0024
Permanent link to this document:
https://doi.org/10.1108/ARJ-08-2011-0024
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

Downloaded on: 12 January 2019, At: 03:49 (PT)

References: this document contains references to 44 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 2513 times since 2013*
Users who downloaded this article also downloaded:
(2016),"Audit quality indicators: perceptions of junior-level auditors", Managerial Auditing Journal, Vol.
31 Iss 8/9 pp. 949-980 <a href="https://doi.org/10.1108/MAJ-01-2016-1300">https://doi.org/10.1108/
MAJ-01-2016-1300</a>
(2013),"Audit firm rotation and audit quality: evidence from academic research", Accounting Research
Journal, Vol. 26 Iss 1 pp. 75-84 <a href="https://doi.org/10.1108/ARJ-11-2012-0087">https://
doi.org/10.1108/ARJ-11-2012-0087</a>

Access to this document was granted through an Emerald subscription provided by emerald-srm:216788 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.

*Related content and download information correct at time of download.

 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1030-9616.htm

ARJ
26,1 Audit quality and
overvalued equity
Robert Houmes, Maggie Foley and Richard J. Cebula
56 Department of Accounting and Finance, Jacksonville University,
Jacksonville, Florida, USA

Abstract
Purpose – Audit quality studies document that accruals decrease when the audit firm is large, or the
audit firm is an industry specialist, or the audit-client tenure is long. The purpose of this paper is to
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

posit that incentives related to highly-valued equity mitigate these results, as managers use income
increasing accruals to augment earnings.
Design/methodology/approach – To test this assertion, the authors regress discretionary accruals
on: controls, a highly valued equity indicator variable equal to 1 if the client’s lagged price-to-earnings
ratio is in the highest P/E quintile, indicator variables equal to 1 for alternative measures of audit
quality, and interaction terms between the highly valued equity indicator variable and audit quality
indicator variables.
Findings – Results of tests show positive and statistically significant coefficients for each of the
highly-valued equity-audit quality interaction terms, suggesting that when a firm is highly valued the
accruals’ decreasing effect of high quality auditors is reduced.
Originality/value – Beginning with Jensen’s article regarding the agency costs of overvalued
equity, a stream of research examining factors associated with highly priced firms has developed. The
paper extends these findings, as well as the considerable body of audit quality studies, by examining
the ability of a high quality auditor to attenuate this result.
Keywords Auditing, Equity capital, Price earning ratio, Audit quality, Highly valued equity,
Discretionary accruals
Paper type Research paper

Accounting Research Journal
Vol. 26 No. 1, 2013
pp. 56-74
q Emerald Group Publishing Limited
1030-9616
DOI 10.1108/ARJ-08-2011-0024
Introduction
Classical agency theory asserts that alignment of management-shareholder interests
increases incentives for value creation (Jensen and Meckling, 1976). Jensen (2005) argues,
however, that when a firm becomes overvalued, i.e. the price of the firm becomes greater
than its underlying economic value, managers are motivated to perpetuate
overvaluation, which is consistent in principle with arguments in Renas and Cebula
(2005). Since an overvalued firm, by definition, lacks the operational capability to achieve
performance levels reflected in its price, managers are motivated to use aggressive
accounting policies to maximize earnings. Although numerous reporting alternatives
are available to achieve earnings management goals, accruals are an especially attractive
choice since they are a normal part of the financial reporting process and their amounts
require forward looking estimates over which managers have considerable discretion.
In a recent study Houmes and Skantz (2010) assert that incentives associated with over
valued equity induce managers to support extreme valuations by using discretionary
accruals to manage earnings higher. In particular, they regress discretionary accruals
on highly valued equity proxy variables measured as companies in the highest quintile
of P/E firms and controls. Results show that in the year following valuation and relative
to other firms, highly valued firms report higher discretionary accruals. Among other
 things they conclude that directors and audit committees should be especially sensitive to Audit quality
earnings management when the firm is highly valued. In the spirit of these remarks, we
extend the findings in Houmes and Skantz (2010) by investigating the effect that
overvaluation has on audit quality. This study provides evidence that the previously
documented inverse relation between, large audit firms, industry specialist audit
firms, long tenure audit firms, and discretionary accruals is attenuated when the firm is
highly valued. 57
Using high price-to-earnings firms to proxy for highly valued equity and the level of
income increasing discretionary accruals to proxy for earnings quality, our results
show that, in accordance with prior studies, high quality auditors generally constrain
accruals. By contrast, however, accruals for highly valued clients of high quality
auditors are statistically significantly higher. In particular, accruals for clients of
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

large Big N (Big 6, Big 5, or Big 4, depending on the period), industry specialist and
long tenure auditors increase for clients in the previous year’s highest quintile of
price-to-earnings ratios. Hence, the widely documented tendency for high quality audit
firms to reduce accruals is mitigated when the firm is highly valued. We rationalize
these findings within the context of traditional audit quality literature and Jensen’s
(2005) overvalued equity hypothesis by asserting that the accruals decreasing effect of
high quality auditors on earnings is reduced as informed managers of highly valued
firms with greater incentives to perpetuate values prevail upon auditors to report
higher levels of income increasing discretionary accruals.
The value of a firm is a function of its prospective performance, and the greater the
performance expectations the higher the value. Since expectations are greater when a
firm is richly priced, investors’ reactions to unexpectedly negative performance
outcomes should also be greater. For example, Skinner and Sloan (2002) document that
the market’s reaction to negative earnings surprises is greater than that for positive
earnings surprises and that this asymmetric reaction increases for high market-to-book
firms. Given that shareholder disappointments will be greater when the market’s
expectations are the highest, the potential findings of this study have implications for
audit practitioners as they conduct audits for highly valued firms. Prior studies report
that shareholder suits are a common source of auditor litigation (Palmrose, 1988;
Goldwasser and Eickemeyer, 2004). Extant research also documents that the likelihood
of a securities fraud class action increases with the firm’s market-to-book ratio (Grimm,
2009). If high valuations impair audit quality, then auditors are performing lower
quality audits on the very firms that historically have tended to disappoint investors
the most. Results of this study should also be relevant for boards of highly valued
firms. If overvaluation induces lower earnings quality and by extension lower audit
quality, boards should be especially vigilant in their internal control function. This
should be particularly true for the audit committee.
From a research perspective, a substantial body of work currently exists on audit
quality and more recent studies have examined factors related to highly valued equity.
To date, however, published studies have not considered audit quality within a
valuation context. We synthesize and provide additional insight into both streams of
research by investigating the accruals constraining effect of high quality auditors
vis-à-vis other auditors when the client is highly valued.
The remainder of this study is organized as follows. The section thereafter includes
relevant literature, along with a discussion of Jensen’s (2005) overvalued equity
 ARJ hypothesis as it relates to earnings and audit quality. The next section includes testable
26,1 hypotheses and their motivations, followed by a description of the sample and
methodology used to conduct tests. Results of these tests are reported in the subsequent
section. The study concludes with a discussion and summary of the findings.

Literature and theory

58 Highly valued equity
Beginning with Jensen (2005), a significant stream of research has developed regarding
the overvalued equity hypothesis. Efendi et al. (2007) use a sample of 190 firms from a
2003 General Accounting Office report to investigate factors related to overvalued
equity and document among other things that earnings restatements increase for firms
that are constrained by debt covenant violations or that raise new debt. They also
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

show that the incidence of restatements increases when CEOs hold large amounts of
in-the-money stock options.
Chi and Gupta (2009) report that overvaluation-induced earnings management as
proxied for by discretionary accruals is negatively related to following year abnormal
returns. They document that, in the following year, abnormal returns of firms with
high discretionary accruals are 11.88 percent lower than the returns of firms with lower
discretionary accruals.
A long-standing anomaly of efficient equity markets is the tendency for highly
valued, high price-to-earnings firms to earn lower returns going forward. While
explanations for the persistence of this anomaly are beyond the scope of this study,
they include the following three:
(1) economic theory predicts that competition mitigates the ability of firms to
achieve and sustain abnormal performance levels over time;
(2) since highly valued firms are less risky, lower returns reflect lower risk; and
(3) investors (irrationally) tend to over-price high-growth-glamour firms.

Notwithstanding these conjectures, however, the tendency for high price-to-earnings

firms to under perform is well documented (Basu, 1977; Chopra et al., 1992;
Lakonishok et al., 1994; Campbell and Shiller, 2001). In a more recent study, Anderson
and Brooks (2006) document that the difference in returns between value and glamour
firms almost doubles when P/Es are calculated using average earnings over the last eight
years. The following graphical depiction of companies from the 1991 to 2008 period
shows that firms in the highest quintile of price-to-earnings ratios earned lower following
year mean and median abnormal returns in every year but one (Figures 1 and 2).
Hence, from a market value perspective and relative to other companies, the very
firms that are expected to perform the best, on average, tend to perform worse. Since
expectations are particularly high for highly valued firms, when managers foresee the
operational inability of their firms to meet expected performance targets, incentives to
manage earnings increase. An important deterrent against these incentives is the audit.

Audit quality
The value of accounting information is a function of its credibility, and a central
objective of audits is to enhance the quality of financial reporting. Prior studies have
provided several empirical surrogates to measure audit quality. These include audit
firm size, audit industry specialization, and the length of the auditor-client relationship.
 0.5
Mean Returns of Firms in the Highest Quintile of Price-to-Earnings Ratios Mean Returns Other
Audit quality

0.4

0.3 59
Mean Returns

0.2
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

0.1

0
Year 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Figure 1.
–0.1
Annual mean CAPM
returns highly valued vs
other 1991-2008
–0.2

Median Returns of Firms in the Highest Quintile of Price-to-Earnings Ratios Median Return Other
0.35

0.3

0.25

0.2

0.15
Median Returns

0.1

0.05

0
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
–0.05

–0.1 Figure 2.
Annual median CAPM
–0.15 returns highly valued vs
other 1991-2008
–0.2

Beginning with DeAngelo (1981), decades of research have shown that large audit firms
with greater resources and more reputation at stake perform higher quality audits
(Palmrose, 1986; Beatty, 1989; Craswell and Taylor, 1995; Lennox, 1999; Teoh and
Wong, 1993; Menon and Williams, 2004; Houmes et al., 2012, etc.). Using accruals to
 ARJ proxy for earnings quality, Becker et al. (1998) show that clients of large (Big 6) audit
26,1 firms report lower discretionary accruals. In a similar study, Francis et al. (1999)
examine a large sample of NASDAQ firms and find that although Big 6 audit
clients have higher levels of total accruals, they have lower estimated discretionary
accruals. Krishnan (2003) provides evidence that investors ascribe higher values to
the discretionary accruals of Big 6 clients than non-Big 6 clients. Heninger (2001)
60 documents that the likelihood of litigation increases with levels of total and
discretionary accruals, but decreases if a Big 5 auditor. Houmes and Skantz (2010)
provide evidence that the magnitude of the inverse relation between operating cash flow
and accruals increases if the firm is highly valued and assert that incentives to manage
earnings increase when operating cash flow is weak. Although they do not offer any
direct tests relating to the effect of high valuation on audit quality after excluding loss
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

firms, they find (not tabulated) evidence large Big N auditors reduce this tendency.
Balsam et al. (2003) shows that clients of within Big 6 and Big 5 auditor industry
specialists have lower accruals and higher earnings response coefficients. Myers et al.
(2003) document that discretionary accruals decrease with the length of audit
firm-client relationship. Mansi et al. (2004) show that the cost of debt decreases with
tenure and Carcello and Nagy (2004) document an increase in the incidence of financial
reporting fraud when audit firm tenure is less than three years.
Although audit opinions enhance the credibility and reliability of financial reports,
they also reflect a negotiation dimension and, within the ethical and technical confines
of accounting standards, a firm’s published financial report may be perceived as a
joint statement from the manager and auditor (Antle and Nalebuff, 1991). Gibbons et al.
(2001) use a sample of 93 experienced audit partners to report that auditor-client
negotiation occurs on a regular basis. In particular, results show that negotiation is
common, with 67 percent of audit partners experiencing negotiation with 50 percent or
more of their clients. Hence, the final reporting product is often the result of a
compromise between management and auditors (Ellingsen et al., 1989).
Since managers of highly valued firms are under increased pressure to meet
optimistic earnings forecasts, they have incentives to assume a more aggressive
negotiating stance, prevailing on auditors to, within the confines of existing accounting
standards, report higher earnings. Hence, income increasing negotiations that bias
earnings upward should be more prevalent for highly valued firms. If the auditor relents,
earnings quality will be impaired. Consequently, audit quality for highly valued firms
could be negatively affected. Antle and Nalebuff (1991) demonstrate that when joint
auditor-client welfare is maximized, ex post income reporting is biased upward.
Although executives have numerous opportunities to manage earnings, accruals are
particularly appealing since they are a normal, frequent, and expected component of
the financial reporting process. Numerous prior studies provide evidence that earnings
are managed with discretionary accruals (Subramanyam, 1996; Bergstresser and
Philippon, 2006, etc.).
Since highly valued companies are under greater pressure to meet earnings
expectations, we expect that managers will assume a more intransigent negotiating
stance with the auditor and utilize accruals to increase earnings. Consequently, the
tendency for high quality audit firms to constrain accruals will be diminished for highly
valued firms as auditors acquiesce towards the upper bounds of accounting standards
constraints in the face of increased client pressure.
 Finally, auditors may suffer, at least in some measure, from the same information Audit quality
asymmetry that characterizes the relationship between managers and owners. While the
asymmetry may not be as great, managers nevertheless are closer to the firms they lead
than external auditors. At the very least, since managers of highly valued firms have
better knowledge of their company’s economic performance, they will be inclined to
accept undetected statement errors in their favor and protest only those that are adverse.
We test these assertions with the following hypotheses stated in alternative form: 61
H1. The magnitude of the inverse relation between the discretionary accruals of
the clients of large audit firms and the discretionary accruals of the clients of
other audit firms decreases if the clients are highly valued.
H2. The magnitude of the inverse relation between the discretionary accruals
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

of the clients of industry specialist audit firms and the discretionary accruals
of the clients of other audit firms decreases if the clients are highly valued.
H3. The magnitude of the inverse relation between the discretionary accruals of
audit firms’ clients with long tenure and the discretionary accruals of clients
with shorter tenure decreases if the clients are highly valued.
H4. The magnitude of the inverse relation between the discretionary accruals of the
clients of audit industry specialist audit firms with long tenure and the
discretionary accruals of other clients decreases if the clients are highly valued.

Sample and methodology

Sample
For years 1998-2009, we acquire panel data for financial statement variables from the
Compustat files of companies. Similar to prior accruals studies, we remove all financial
services (SIC codes 6000-6999) and regulated (4900-4999) firms. As explained below,
to control for the Big N audit quality effect on our specialists and long tenure auditor
variables, we separate the overall sample of firms into two groups. One group, our
Big N sample, contains Big N and non-Big N auditors. The other group, our
specialist-tenure sample contains Big N audit clients only. After deleting firms with
missing data for model variables, the total number of firm year observations in our Big
N sample is 29,405. Eliminating non-Big N audit firms reduces our specialist-tenure
sample to 23,937 observations. Although we also provide regression results for all
discretionary accruals, consistent with our assertion that managers of highly valued
firms manage earnings higher, our primary tests involve income increasing
discretionary accruals only. Eliminating negative discretionary accruals reduces our
Big N (specialist-tenure) sample to 16,346 (13,172) firm year observations.

Discretionary accruals
Our dependent variable is discretionary accruals (DACit). For all firms, discretionary
accruals are estimated using the cross sectional version of the modified Jones model (Jones,
1991). The modified Jones model has been used in a variety of research settings (Becker et al.,
1998; Francis et al., 1999; Reynolds and Francis, 2000, etc.). The model is specified as follows:
 
1
TAC it ¼ b0 þ b1 þ b2 ðDREV it 2 DARit Þ þ b3 PPE it þ 1it ð1Þ
AT it21
 ARJ where TACit is the difference between firm i’s year t earnings before extraordinary items
and net cash flow from operations scaled by beginning of year (t 2 1) assets; ATit2 1 is firm
26,1 i’s beginning of the year t total assets; DREVit is the difference in year t and year t 2 1 sales;
DARit is the difference between year t and year t 2 1 trade account receivables; and PPEit is
net property plant and equipment both scaled by ATit2 1. For each year and two-digit SIC
industry group, discretionary accruals (DACit) are estimated cross-sectionally as the
62 residuals from the above regression. Stated alternatively, coefficients from equation (1) are
used to estimate predicted accruals for each year in each SIC group. Discretionary accruals
are the difference between each firm’s actual and predicted accruals, i.e.:
 
1
DAC it ¼ TAC it 2 ½b0 þ b1 þ b2 ðDREV it 2 DARit Þ þ b3 PPE it
: ð2Þ
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

1=AT it21

To mitigate the potential effects of extreme observations on results, we winsorize accruals at

the top and bottom 1 percent levels across the entire pooled sample. In our models we also
include controls that may affect the cross sectional variability in the dependent variable,
DACit.

Control variables
Relative to other companies, certain industries or firms may tend to generate higher
accruals. In addition, it is natural that growth companies with increasing earnings and
investments in working capital are more likely to produce greater accruals, and prior
studies show that growth firms report higher accruals (McNichols, 2000). To control for the
possibility that companies with greater total accruals may also have larger discretionary
accruals that our accruals model does not capture, we include total accruals (ACRLit) in our
multivariate tests measured as the difference between firm i’s year t earnings before
extraordinary items and net cash flow from operations scaled by beginning of year assets.
Accruals studies typically control for size effects. Dechow and Dichev (2002) show
that larger firms record larger accruals. Also, larger firms with larger investor
following and more developed and sophisticated financial reporting systems may
affect accrual levels (Becker et al., 1998; Reynolds and Francis, 2000). For each firm i we
include the end of fiscal year t natural log of total assets (LnASSETit).
Reynolds and Francis (2000) provide evidence that the tendency to manage earnings
increases with leverage. DeFond and Jiambalvo (1994) show that accruals are related to
debt covenant breeches. In addition, debt may serve as a monitoring mechanism that
constrains earnings management. To control for the effect that high debt levels may
have on accruals, we include the variable LEVit measured as firm i’s end of year t long
term debt scaled by t 2 1 total assets.
Operating cash flows are a component of earnings and their levels correspond
inversely with accruals. Further, the level of cash flow may affect the ability and/or need
to use accruals, causing firms with higher (lower) operating cash flows to report lower
(higher) discretionary accruals (Becker et al., 1998). We control for these effects by
including operating cash flow deflated by the beginning of the year total assets (OCFit).
Kothari et al. (2005) show that discretionary accruals are impacted by financial
performance. Accordingly, weinclude ROAit, income before extraordinary items
divided by the beginning of the year total assets. Finally, for 47 of the 48 two-digit SIC
industry codes included in our study we assign dummies equal to 1 to control for
 potential industry effects[1]. Similar to our dependent variable, DACit, we winsorize all Audit quality
continuously measured independent variables at the 1 percent levels.

Variables of interest: audit quality and highly valued equity

Using our initial all accruals sample, we identify highly valued clients (HVit2 1) as firms in
the highest quintile of PEs (i.e. P/E . 27.204) and assign an indicator variable equal to 1 if
the client is in this quintile of prior fiscal year end price-to-earnings ratios and 0 otherwise. 63
Our audit quality variables are as follows: BIGNit2 1 equal to 1 and 0 otherwise if the
audit firm is one of the largest four, five, or six audit firms (depending on the time period);
SPECit2 1, equal to 1 and 0 otherwise if the audit firm is an industry specialist; and
TENit2 1 equal to 1 and 0 otherwise if the length of the auditor-client relationship is equal
to or over five years. We also include an additional audit quality indicator variable if the
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

auditor is both an industry specialist and has a long tenure auditor-client relationship
(SPECTENit2 1). In accordance with prior studies (Palmrose, 1986), we define an auditor
as an industry specialist according to the auditor’s two-digit SIC code industry share and
designate the firm as a specialist based on the number of clients and total client sales.
In particular, an indicator variable equal to 1 and 0 otherwise is assigned to auditors with
both the largest number of clients and the highest total client sales within each two-digit
SIC code. To eliminate the brand name effect of BN auditors, we restrict the sample of
specialist and tenure firms to Big N clients only. We eliminate SIC code industries with
fewer than 15 observations. The income increasing discretionary accruals, Big N, and
non-Big N sample consists of 16,346 firm year observations. The income increasing
discretionary accruals, specialist-tenure sample contains 13,172 observations.
To investigate the relation between high valuations and the tendency of high audit
quality auditors to mitigate accruals, we interact the audit quality variables with our
highly valued equity dummy. Statistically significantly positive estimates for the audit
quality, highly valued equity interaction terms: BIGN *HVit2 1, SPEC *HVit2 1,
TEN *HVit2 1 and SPECTEN *HVit2 1, provide support for hypotheses that incentives
associated with high valuations reduce the tendency of high quality audit firms to
constrain accruals. Our models are as follows:

DAC it ¼ a0 þ a1 ACRLit þ a2 LnASSET it þ a3 LEV it þ a4 OCF it þ a5 ROAit

þ a6 BIGN it21 þ a7 HV it21 þ a8 BIGN * HV it21 þ a9 INDUS þ 1it ð3Þ

DAC it ¼ a0 þ a1 ACRLit þ a2 LnASSET it þ a3 LEV it þ a4 OCF it þ a5 ROAit

þ a6 SPEC it21 þ a7 HV it21 þ a8 SPEC * HV it21 þ a9 INDUS þ 1it ð4Þ

DAC it ¼ a0 þ a1 ACRLit þ a2 LnASSET it þ a3 LEV it þ a4 OCF it þ a5 ROAit

þ a6 TEN it21 þ a7 HV it21 þ a8 TEN * HV it21 þ a9 INDUS þ 1it ð5Þ

DAC it ¼ a0 þ a1 ACRLit þ a2 LnASSET it þ a3 LEV it þ a4 OCF it þ a5 ROAit

ð6Þ
þ a6 SPECTEN it21 þ a7 HV it21 þ a8 SPECTEN * HV it21
þ a9 INDUS þ 1it
 ARJ Results
26,1 Descriptive statistics
Table I defines variables used in models. Tables II and III provide descriptive
statistics for variables. Table II Panels A and B shows the maximum, minimum, mean,
and median values of variables in our income increasing Big N (Panel A) and
specialist-tenure (Panel B) samples. Table III provides univariate correlations. Mean
64 discretionary accruals (scaled) for both samples are 0.064 and 0.059, respectively[2].
In our BN auditor sample 82 percent of the observations involve large audit firms. In our
specialist-tenure sample, 20 percent of the observations involve specialists and
54 percent have audit-client tenures over five years. Hence most Big N audit-client
relationships are greater than five years. 10 percent of the observations are identified as
both a specialist and long tenure audit client.
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

Univariate correlations in Table III reveal that size, leverage and operating cash flow,
are inversely related to discretionary and total accruals. ROA is positively (negatively)
related to total (discretionary) accruals. ROA and operating cash flow are also positively
correlated with our HVit2 1 indicator variable. While coefficients are generally positive
between each of our audit quality variables and HVit2 1, only Big N is statistically
significant. Overall, the signs and significance of our other model variables are
consistent across all proxies for audit quality. BIGN, SPEC, TEN, and SPECTEN are
positively (negatively) correlated with size, operating cash flow and ROA (discretionary
accruals).
Table IV provides the differences in mean scaled discretionary accruals between each
of the high quality auditor measures and their non-high quality auditor counterparts.

DACit are Firm i’s fiscal year t income increasing discretionary accruals
ACRLit are Firm i’s fiscal year t total accruals
LnASSETit are Firm i’s fiscal year t natural log of total assets
LEVit are Firm i’s fiscal year t long term debt divided by fiscal year t 2 1 total assets
OCFit are Firm i’s fiscal year t cash flow from operating activates divided by fiscal year
t 2 1 total assets
ROAit are fFrm i’s fiscal year t income before extraordinary items divided by fiscal year
t 2 1 total assets
HVit2 1 are An indicator variable equal to 1 for firms in the highest quintile of fiscal year
t 2 1 price-to-earnings ratios
AQ it2 1 are Four measures of audit quality defined as follows
BIGNit2 1 are An indicator variable equal to 1 if at the end of fiscal year t 2 1 the auditor is a
Big N auditor
SPECit2 1 are An indicator variable equal to 1 if at the end of fiscal year t 2 1 the auditor is an
industry specialist
TENit2 1 are An indicator variable equal to 1 if at the end of fiscal year t 2 1 the auditor
tenure is greater than or equal to five years
SPECTENit2 1 are An indicator variable equal to 1 if at the end of fiscal year t 2 1 the auditor is an
industry specialist and tenure is greater than or equal to five years
BIGN *HVit2 1 are An interaction term between BIGNit2 1 and HVit2 1;
SPEC *HVit2 1 are An interaction term between SPECit2 1 and HVit2 1
TEN *HVit2 1 are An interaction term between TENit2 1 and HVit2 1
SPECTEN * are An interaction term between SPECTENit2 1 and HVit2 1
Table I. HVit2 1
Variable definitions INDUS are An indicator variable equal to1for firms in each two-digit SIC code
 Audit quality
Variables Maximum Minimum Mean Median SD

Panel A. Big N auditor-income increasing discretionary accruals (n ¼ 16,346)

DACit 0.331 0.000 0.064 0.044 0.066
ACRLit 0.256 2 0.544 20.014 2 0.026 0.077
LnASSETit 32.140 22.120 26.603 26.520 2.157
LEVit 1.196 0.000 0.183 0.123 0.215 65
OCFit 0.428 2 0.759 0.040 0.068 0.159
ROAit 0.247 2 1.318 0.010 0.043 0.177
BIGNit 1.000 0.000 0.820 1.000 0.388
HVit2 1 1.000 0.000 0.213 1.000 0.401
Panel B. Specialists and long tenure auditors – income increasing discretionary accruals (n ¼ 13,172)
DACit 0.331 0.000 0.059 0.041 0.062
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

ACRLit 0.269 2 0.544 20.017 2 0.028 0.073

LnASSETit 32.360 21.990 27.021 26.978 0.890
LEVit 1.498 0.000 0.197 0.144 0.223
OCFit 0.498 2 1.708 0.048 0.073 0.168
ROAit 0.848 2 2.576 0.015 0.045 0.187
SPECit 1.000 0.000 0.195 0.000 0.396
TENit 1.000 0.000 0.540 1.000 0.498
SPETENCit 1.000 0.000 0.099 0.000 0.298
HVit2 1 1.000 0.000 0.231 0.000 0.398
Table II.
Note: Variables are defined in Table I Descriptive statistics

Results show that relative to other audit firms, discretionary accruals are lower for each
of the four audit quality proxy measures ( p , 0.001) with the greatest difference
occurring between Big N and non-Big N auditors (2 0.025).
This study predicts that incentives related to high valuations reduce the accruals
decreasing effect of high quality auditors. To initially investigate this assertion, within
each price-to-earnings group (top 20 PE and bottom 80 PE) we compare the mean
discretionary accruals between clients of high quality auditors and other audit clients.
Results show that while high quality audit clients report lower accruals vis-à-vis other
clients in both price-to-earnings groups, for three of our audit quality measures
differences generally decrease when the client firm is in the highest PE quintile.
In particular, for three of the four audit quality variables, the difference in scaled
discretionary accruals between the clients of high quality auditors and the clients of
other auditors is not significant when the firm is highly valued. For example, the
difference in mean accruals between specialist and non-specialist auditors in the top
quintile of price-to-earnings ratios is an insignificant 2 0.003 (0.349), while the difference
in mean accruals for firms in the bottom 80 percent is 2 0.006 (0.000). In addition, the
differences in accruals between the top 20 PE, TEN and SPECTEN audit clients and the
accruals of the corresponding top 20 PE, NTEN, and NSPECTEN audit quality clients is
an insignificant 2 0.002 and 2 0.001 ( p ¼ 0.296 and 0.788, respectively). Differences in
accruals between the bottom 80 PE, TEN, and SPECTEN are, however, significantly
lower for high quality auditors. Differences in accruals between the discretionary
accruals of high quality specialists, long tenure, and specialist – long tenure audit
clients and other lower quality audit clients decrease if the client is highly valued.
The difference, however, in mean accruals for the top quintile of Big N versus NBN
auditors is significant at the p , 0.001 level. Hence, the reduction in accruals between
 Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

66
26,1
ARJ

Table III.
Panel A. Variable correlations for Big N auditors – income increasing discretionary accruals (n ¼ 16,346)
Variable DACit ACRLit LnASSETit LEVit OCFit ROAit BIGN HVit2 1
DACit 1 0.626 (0.000) 2 0.289 (0.000) 2 0.046 (0.050) 2 0.353 (0.000) 2 0.070 (0.000) 2 0.146 (0.000) 2 0.029 (0.000)
ACRLit 1 2 0.167 (0.000) 2 0.072 (0.000) 2 0.336 (0.000) 0.119 (0.000) 2 0.073 (0.000) 2 0.003 (0.746)
LnASSETit 1 0.275 (0.000) 0.379 (0.000) 0.330 (0.000) 0.407 (0.000) 0.101 (0.000)
LEVit 1 0.031 (0.000) 0.022 (0.004) 0.117 (0.000) 2 0.043 (0.000)
OCFit 1 0.801 (0.000) 0.118 (0.000) 0.122 (0.000)
ROAit 1 0.085 (0.000) 0.116 (0.000)
BIGNit 1 0.033 (0.000)
HVit2 1 1
Panel B. Variable correlations for specialists and long tenure auditors – income increasing discretionary accruals (n ¼ 13,172)
Variable DACit ACRLit LnASSETit LEVit OCFit ROAit SPEC TEN SPEC TEN HVit2 1
DACit 1 0.650 (0.000) 2 0.262 (0.000) 2 0.025 (0.004) 2 0..331 (0.000) 2 0.062 (0.00) 2 0.036 (0.000) 2 0.047 (0.000) 2 0.041 (0.000) 2 0.024 (0.006)
ACRLit 1 2 0.178 (0.000) 2 0.071 (0.000) 2 0.319 (0.000) 0.088 (0.000) 2 0.009 (0.319) 2 0.011 (0.224) 2 0.021 (0.016) 2 0.008 (0.382)
LnASSETit 1 0.246 (0.000) 0.369 (0.000) 0.323 (0.000) 0.055 (0.000) 0.154 (0.000) 0.090 (0.000) 0.086 (0.000)
LEVit 1 0.013 (0.142) 0.007 (0.448) 0.035 (0.000) 2 0.002 (0.780) 0.025 (0.004) 2 0.049 (0.000)
OCFit 1 0.828 (0.000) 0.037 (0.000) 0.072 (0.000) 0.051 (0.000) 0.106 (0.000)
ROAit 1 0.028 (0.001) 0.075 (0.000) 0.039 (0.000) 0.102 (0.000)
SPECit 1 2 0.032 (0.000) 0.673 (0.000) 0.005 (0.591)
TENit 1 0.307 (0.000) 0.005 (0.559)
SPECTENit 1 2 0.006 (0.471)
HVit2 1 1
Notes: Significance ( p: two-tailed) is shown in parentheses under coefficients; variables are defined in Table I
 Audit quality
Audit quality Mean disc. Audit quality Mean disc.
variables accruals n variables accruals n

BN 0.060 13,320 TEN 0.057 7,125

NBN 0.085 3,026 NTEN 0.063 6,047
Difference 20.025 (0.000) Difference 20.006 (0.000)
SPEC 0.055 2,568 SPECTEN 0.053 1,300 67
NSPEC 0.061 10,604 NSPECTEN 0.061 11,872
Difference 20.006 (0.000) Difference 20.008 (0.000) Table IV.
Difference in mean
Notes: Variable definitions: BN – firm is audited by a Big N auditor; NBN – firm is not audited by a income increasing
Big N auditor; SPEC – auditor is an industry specialist; NSPEC – auditor is not an industry specialist; discretionary accruals
TEN – audit-client firm tenure is greater than or equal to five years; NTEN – audit-client firm tenure is
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

between high quality:

less than five years; SPECTEN – audit-client firm tenure is greater than or equal to five years and the Big N, specialist and
auditor is an industry specialist; NSPECTEN – firm’s auditor is not a SPECTEN auditor; n – number long tenure auditors and
of firm year observations other auditors

BN and NBN auditors is consistent for both PE groups. An important limitation of this
analysis is that it omits control variables that may affect accruals levels. The following
section includes results of our main tests using multivariate analysis that includes
previously described control variables (Table V).

Discretionary accruals models

Table VI shows results of main tests for our income increasing discretionary accruals
sample for models depicted in equations (3) through (6). Adjusted R 2 exceed 50 percent.
Total accruals, ACRLit, is significant and positive. As expected and in accordance with
prior accruals studies, estimates for assets and operating cash flow are negative,
suggesting that greater firm size, and operating cash flow enhance accruals quality. The
coefficients for ROAit are positive. Discretionary accruals increase with (scaled) income.

BN n NBN n Difference n
Top 20 PE 0.057 2,939 0.078 543 20.021 (0.000) 3,482
Bottom 80 PE 0.060 10,381 0.086 2,483 20.026 (0.000) 12,864
SPEC n NSPEC n Difference n
Top 20 PE 0.054 603 0.057 2,450 20.003 (0.349) 3,053
Bottom 80 PE 0.056 1,965 0.062 8,154 20.006 (0.000) 10,119
TEN n NTEN n Difference n
Top 20 PE 0.056 1,695 0.058 1,358 20.002 (0.296) 3,053
Bottom 80 PE 0.057 5,430 0.064 4,689 20.007 (0.000) 10,119
SPECTEN n NSPECTEN n Difference n
Top 20 PE 0.056 297 0.057 2,756 20.001 (0.788) 3,053
Bottom 80 PE 0.052 1,003 0.062 9,116 20.010 (0.000) 10,119 Table V.
Difference in mean
Notes: Variable definitions: top 20 PE – firm is in the highest 20 percent of price-to-earnings ratios; discretionary income
bottom 80 PE – firm is in the lowest 80 percent of price-to-earnings ratios; BN – firm is audited by a increasing accruals
Big N auditor; NBN – firm is not audited by a Big N auditor; SPEC – auditor is an industry specialist; between audit quality
NSPEC – auditor is not an industry specialist; TEN – audit-client firm tenure is greater than or equal variables for firms in
to five years; NTEN – audit-client firm tenure is less than five years; SPECTEN – audit-client firm the highest 20 and
tenure is greater than or equal to five years and the auditor is an industry specialist; NSPECTEN – lowest 80 percent of
firm’s auditor is not a SPECTEN auditor; n – number of firm year observations price-to-earnings ratios
 ARJ
BN Specialist Tenure Specialist and tenure
26,1 coefficient coefficient coefficient coefficient
( p-value) ( p-value) ( p-value) ( p-value)

ACRLit 0.546 (0.000) 0.563 (0.000) 0.561 (0.000) 0.561 (0.000)

LnASSETit 20.005 (0.000) 2 0.004 (0.000) 2 0.004 (0.000) 20.004 (0.000)
68 LEVit 0.008 (0.000) 0.009 (0.000) 0.008 (0.000) 0.008 (0.000)
OCFit 20.042 (0.091) 2 0.046 (0.000) 2 0.046 (0.000) 20.046 (0.000)
ROAit 0.001 (0.000) 0.011 (0.024) 0.010 (0.028) 0.010 (0.030)
HVit2 1 20.006 (0.007) 2 0.003 (0.006) 2 0.004 (0.004) 20.003 (0.004)
BIGNit2 1 20.006 (0.000)
BIGN *HVit2 1 0.004 (0.044)
SPECit2 1 2 0.002 (0.061)
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

SPEC *HVit2 1 0.003 (0.087)

TENit2 1 2 0.001 (0.364)
TEN *HVit2 1 0.003 (0.055)
SPECTEN 20.002 (0.065)
SPECTEN *HVit2 1 0.006 (0.003)
Table VI. INDUS Not reported Not reported Not reported Not reported
Results for effects Adjusted R 2 0.507 0.533 0.532 0.532
of overvaluation on the F( p) 0.000 0.000 0.000 0.000
income increasing
n 16,346 13,172 13,172 13,172
discretionary accruals
of Big N and audit Notes: DAC it ¼ a0 þ a1 ACRLit þ a2 LnASSET it þ a3 LEV it þ a4 OCF it þ a5 ROAit þ a6 HV it21 þ
specialist clients a7 AQit þ a8 AQ*it HV it21 þ a9 INDUS þ 1it; variables are defined in Table I

HVit2 1 is negative. Since our hypotheses are directional, tests for all our audit quality
variables (both main effects and interaction terms) are one-tailed. Similar to univariate
tests in the previous sections, Table VI shows that the coefficients for the Big N, specialist,
and long tenure-specialist audit quality measures are negative with estimates and ( p)
values of 20.006 (0.000), 20.002 (0.061), and 20.002 (0.065), respectively. However, the
estimate for our audit quality variable, TENit, is an insignificant 20.001 (0.364).
We test hypotheses with interaction terms between our highly valued equity and audit
quality variables: BIGN *HVit2 1, SPEC *HVit2 1, TEN *HVit2 1, and SPECTEN *HVit2 1.
Respective estimates and p-values for the income increasing accruals sample are 0.004
(0.044), 0.003 (0.087), 0.003 (0.055), and 0.006 (0.003). For each of the four alternative audit
quality measures, discretionary accruals increase when the client firm is highly valued.
As described above, since incentives related to managing earnings higher should be
associated with higher discretionary accruals, we restrict tests in Table VI to clients
reporting income increasing discretionary accruals. Nevertheless, alternative factors may
induce negative accruals. For example, auditors with a conservative bias may provoke
income decreasing accounting choices (Becker et al., 1998) and this may be especially likely
if the firm is highly valued. To consider this possibility and other factors related to lower
accruals, we re-estimate equations (3)-(6) using signed (positive and negative)
discretionary accruals as the dependent variable. Table VII provides findings from
tests of models using this all accruals sample. While results for controls are similar across
both samples, in contrast to Table VI, leverage is inversely related to discretionary
accruals. The coefficients for the Big N, specialist, and long tenure-specialist audit quality
measures are negative with estimates and ( p) values of 20.005 (0.000), 20.002 (0.026),
 Audit quality
BN Specialist Tenure Specialist and tenure
coefficient coefficient coefficient coefficient
( p-value) ( p-value) ( p-value) ( p-value)

ACRLit 0.798 (0.000) 0.809 (0.000) 0.809 (0.000) 0.809 (0.000)

LnASSETit 20.003 (0.000) 20.002 (0.000) 2 0.002 (0.000) 2 0.002 (0.000)
LEVit 20.004 (0.004) 20.003 (0.024) 2 0.003 (0.057) 2 0.003 (0.048) 69
OCFit 20.023 (0.000) 20.018 (0.000) 2 0.019 (0.000) 2 0.018 (0.000)
ROAit 0.011 (0.000) 0.010 (0.218) 0.010 (0.000) 0.010 (0.000)
HVit2 1 20.006 (0.002) 20.003 (0.000) 2 0.004 (0.000) 2 0.003 (0.000)
BIGNit2 1 20.005 (0.000)
BIGN *HVit2 1 0.003 (0.061)
SPECit2 1 20.002 (0.026)
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

SPEC *HVit2 1 0.004 (0.004)

TENit2 1 2 0.001 (0.320)
TEN *HVit2 1 0.002 (0.073)
SPECTEN 2 0.002 (0.026)
SPECTEN *HVit2 1 0.004 (0.005)
INDUS Not reported Not reported Not reported Not reported
Adjusted R 2 0.747 0.760 0.760 0.760 Table VII.
F( p) 0.000 0.000 0.000 0.000 Results for effects
N 29,405 23,937 23,937 23,937 of overvaluation on the
discretionary accruals
Notes: DAC it ¼ a0 þ a1 ACRLit þ a2 LnASSET it þ a3 LEV it þ a4 OCF it þ a5 ROAit þ a6 HV it21 þ of Big N and audit
a7 AQit þa8 AQ*it HV it21 þ a9 INDUS þ 1it ; variables are defined in Table I specialist clients

and 2 0.002 (0.026), respectively, and once again the estimate for audit quality variable,
TENit2 , is not significant 20.001 (0.320). Big N and specialist estimates are similar to
other audit quality – accruals studies. For example, Becker et al. (1998) use a sample of
10,881 firm year observations from the 1989 to 1992 period to report that non-Big
6 audit accruals are 1.5 percent higher than Big 6 accruals. In a similar study Balsam et al.
(2003) reports 20.006 estimate on an audit quality dummy from a regression of
discretionary accruals on controls and an industry leadership indicator variable.
Regarding TENit, prior audit tenure studies conflict on how differing lengths of
auditor-client tenure reflect audit quality. For example, Johnson et al. (2002) report that
relative to medium tenures of four to six years financial reporting quality as measured
by accruals decreases when the tenure is short, but find no reduction in audit quality
for long tenure auditors greater than nine years. Myers et al. (2003) generally confirm
these findings. However, Davis et al. (2003) find that discretionary accruals increase
over years when audit tenure is continuously measured. Although we believe our
choice of tenure equal to or greater than five years reflects the most efficacious measure
of audit quality from the existing literature, a possible explanation for the lack of
significance may simply reflect that documented in prior studies.
As depicted in Table VII, results of our main tests are unchanged from Table VI as
all audit quality-highly valued equity interaction terms, BIGN *HVit2 1, SPEC *HVit2 1,
TEN *HVit2 1, and SPECTEN *HVit2 1 remain statistically significant and positive
0.003 (0.061), 0.004 (0.004), 0.002 (0.073), and 0.004 (0.005).
These findings provide evidence that relative to other firms, the accruals decreasing
effect of high quality auditors is reduced when the client is highly valued. Furthermore,
these results are robust across several measures of audit quality with interaction term,
 ARJ SPECTEN *HVit2 1, showing the highest and most statistically significant estimate
26,1 (0.006) from our income increasing – Big N only sample. Hence, high valuation
attenuates the accruals decreasing effect of a long tenure – industry specialist auditor
in that accruals are 0.6 percent of assets higher for clients in the highest quintile of PEs.
Unreported mean assets and common shares outstanding for SPECTEN clients are
$4,989,790,883 and 222,265,446 shares, respectively. Thus, on average net income is
70 increased by 0.006 times $4,989,790,883 or $29,938,745. Average earnings per share
increases by $0.135 per share ($29,938,745 divided by 222,265,446 shares). While not as
pronounced, the statistically significant estimates for all other audit quality – high
valued client interaction terms similarly illustrate the potential economic impact of
increasing accruals on net income. Indeed, the unreported average diluted earnings
before extraordinary items per share for highly valued SPECTEN firms is $0.644,
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

suggesting that relatively small increases in our interaction coefficients can have a
significant economic impact on earnings per share.

Robustness
We conduct several sensitivity tests. Regarding valuation variable, HVit2 1, we
also alternatively measure price-to-earnings valuation by stratifying firms into
deciles (quintiles) and assign a 1 to companies in the lowest PE decile and quintile and
10 (5) to clients in the highest. Results are similar to our main findings. In particular,
for both income increasing and all discretionary accruals samples, results show
significantly negative (positive) coefficients for audit quality variables (interaction
terms), BIGN, SPEC, and SPECTEN (BIGN * HV it2 1 , SPEC *HVit2 1 , and
SPECTEN *HVit2 1). TEN *HVit2 1 is not significant for either price-to-earnings
measure.
To control for factors occurring over years that our models may not capture, using
1998 as the base year, we include year dummies in each of our equations. With the
exception of TEN *HVit2 1 ( p ¼ 0.125) in our all accruals model, audit quality-highly
valued equity interaction terms are significant and positive. Since financing and/or
acquisition activities could impact accruals, we also assign indicator variables equal to
1 if common stock or debt increased by more than 10 percent and/or the firm reported
an acquisition. Results do not alter our findings.
Negative prior year earnings may impact managers’ incentives to report current
year higher net income. We also run tests with a dummy equal to 1 if the firm reported
a loss in the prior year. In addition, to control for the impact of events not captured
in our models on our dependent variable, discretionary accruals, we further include
an abnormal returns variable measured according to the capital asset pricing model
(CAPM).
That is:

ARit ¼ Rit 2 ERit ð7Þ

CAPM:

ERit ¼ Rft þ bit ðRmt 2 Rft Þ ð8Þ

where, ERit is firm i’s expected end of fiscal year t return; Rft is the one month treasury
bill rate; and bit is the end of fiscal year t market model b calculated over the prior year
 with daily closing prices. Rmt is the value-weighted return from the NYSE, AMEX, and Audit quality
NASDAQ (from CRSP). Inclusion of these variables does not alter findings as audit
quality – highly valued equity interaction terms remain generally significant and
positive.

Discussion of results and conclusion

Using audit quality proxies from prior literature, this study examines the accruals 71
decreasing effect of high quality auditors on highly valued audit clients. We posit that
management incentives associated with highly valued equity reduce the tendency of
high quality auditors to reduce accruals. Although results generally support
hypotheses, reported increases in accruals may be affected by the particular measure of
audit quality used. While not as pronounced, the findings of this study seem to provide
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

some evidence that high quality, Big N auditors continue to maintain their audit
quality edge, albeit marginally, over other audit firms. To reiterate, as per Table V, and
in contrast to our other audit quality characteristics, the difference in mean accruals
between Big N and non-Big N auditors is significant even when the client is highly
valued. Hence, perhaps large audit firms with more clients and greater reputation at
stake are better able to mitigate the accruals increasing tendency of highly valued
firms. Nevertheless, multivariate results with controls do reveal higher discretionary
accruals for highly valued-Big N audit clients. In addition to our main multivariate
tests, we investigate this conjecture further by eliminating long tenure and
specialist auditors from the Big N sample. For both our increasing and signed
accruals model unreported findings show that while estimates for the BIGN *HVit2 1
interaction terms are positive (0.004 and 0.005, respectively), neither are significant
(0.170 and 0.206). While we do not further address this issue here, additional research
that attempts to disentangle the characteristics of large auditors, specialist auditors,
and long tenured auditors as they relate to highly valued audit clients might provide
additional insight.
Regarding other areas for further research, consideration might also be given to the
impact of corporate governance on the audit quality-highly valued equity relation.
In addition, future studies could investigate temporal aspects of audit quality and
highly valued equity both separately and in conjunction with each other. For example,
does the audit quality effect change before, during, and after periods of market tops
when prices and P/Es are at historic highs?

Notes
1. We include SIC dummies in regression models to provide an additional control for industry
effects that may not be captured by the within industry Jones model estimates.
2. As discussed above, discretionary accruals (the residuals from equation (1)) are estimated
using all (positive and negative) accruals. For the all accruals sample, the unreported mean
discretionary accruals are 0.002 for the Big N sample and 0.000 for the specialist and tenure
sample. As reported in Table II, average discretionary accruals are higher for our income
increasing accruals samples.

References
Anderson, K. and Brooks, C. (2006), “The long-term price-earnings ratio”, Journal of Business
Finance & Accounting, Vol. 7, pp. 1063-1086.
 ARJ Antle, R. and Nalebuff, B. (1991), “Conservatism and auditor-client negotiations”, Journal of
Accounting Research, Vol. 29, pp. 31-54.
26,1
Balsam, S., Krishnan, J. and Yang, J.S. (2003), “Auditor industry specialization and earnings
quality”, Auditing: A Journal of Practice & Theory, Vol. 22, pp. 71-97.
Basu, S. (1977), “Investment performance of common stocks in relation to their price
earnings ratios: a test of the efficient market hypothesis”, Journal of Finance, Vol. 32,
72 pp. 663-682.
Beatty, R. (1989), “Auditor reputation and the pricing of initial public offerings”, The Accounting
Review, Vol. 4, pp. 693-709.
Becker, C.L., Defond, M.L., Jiambalvo, J. and Subramanyam, K.R. (1998), “The effect of
audit quality on earnings management”, Contemporary Accounting Research, Vol. 15,
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

pp. 1-24.
Bergstresser, D. and Philippon, T. (2006), “CEO incentives and earnings management”, Journal of
Financial Economics, Vol. 80, pp. 511-529.
Campbell, J. and Shiller, R. (2001), “Valuation ratios and the long-run stock market outlook:
an update”, Cowles Foundation Discussion Paper No. 1295, Cowles Foundation for
Research in Economics, Yale University, New Haven, CT, March.
Carcello, J.V. and Nagy, A.L. (2004), “Audit firm tenure and fraudulent financial reporting”,
Auditing: A Journal of Practice & Theory, Vol. 23, pp. 55-69.
Chi, J.D. and Gupta, M. (2009), “Overvaluation and earnings management”, Journal of Banking
& Finance, Vol. 33, pp. 1652-1663.
Chopra, N.J., Lakonishok, L. and Ritter, J.R. (1992), “Measuring abnormal performance: do stocks
overreact?”, Journal of Financial Economics, Vol. 31, pp. 235-268.
Craswell, A.J. and Taylor, S. (1995), “Auditor brand name reputations and industry
specialization and earnings quality”, Auditing: A Journal of Practice & Theory, Vol. 23,
pp. 297-322.
Davis, L.R., Soo, B.S. and Trompeter, G.M. (2003), “Auditor tenure and the ability to meet or beat
earnings forecasts”, Contemporary Accounting Research, Vol. 26, pp. 517-548.
DeAngelo, L.E. (1981), “Audit size and audit quality”, Journal of Accounting and Economics,
Vol. 3, pp. 183-199.
Dechow, P.M. and Dichev, D. (2002), “The quality of accruals and earnings: the role of accruals
estimation errors”, The Accounting Review, Vol. 77, pp. 35-59.
DeFond, M.L. and Jiambalvo, J. (1994), “Debt covenant violation and manipulation of accruals”,
Journal of Accounting and Economics, Vol. 17, pp. 145-176.
Efendi, J., Srivasta, A. and Swanson, E. (2007), “Why do corporate managers misstate financial
statements? The role of option compensation and other factors”, Journal of Financial
Economics, Vol. 85, pp. 667-708.
Ellingsen, J.E., Pany, K. and Fagan, P. (1989), “SAS No. 59: how to evaluate going concern
opinions”, Journal of Accountancy, January, pp. 24-31.
Francis, J., Mayhew, E. and Sparks, H. (1999), “The role of big six auditors in
the credible reporting of accruals”, Auditing: A Journal of Practice & Theory, Vol. 18,
pp. 17-34.
Gibbons, M., Salterio, S. and Webb, A. (2001), “Evidence about auditor-client management
negotiation concerning clients’ financial reporting”, Journal of Accounting Research,
Vol. 39, pp. 535-564.
 Goldwasser, A. and Eickemeyer, J.H. (2004), “Introduction”, Accountant’s Liability, Chapter 1, Audit quality
Practicing Law Institute, New York, NY.
Grimm, S.D. (2009), “The role of accounting quality in securities class action lawsuits”, working
paper, University of Minnesota, St Paul, MN, June.
Heninger, W.G. (2001), “The association between auditor litigation and abnormal accruals”,
The Accounting Review, Vol. 76, pp. 111-126.
Houmes, R. and Skantz, T. (2010), “Discretionary accruals and highly valued equity”, Journal of
73
Business Finance & Accounting, Vol. 37, pp. 60-92.
Houmes, R., Dickins, D. and O’Keefe, R. (2012), “New evidence on the incremental information
content of earnings reported using the LIFO inventory method”, Advances in Accounting,
Vol. 28, pp. 235-242.
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

Jensen, M. (2005), “Agency costs of highly valued equity”, Financial Management, Vol. 34,
pp. 5-20.
Jensen, M.C. and Meckling, W.H. (1976), “Theory of the firm: managerial behavior, agency costs
and ownership structure”, Journal of Financial Economics, Vol. 3, pp. 305-360.
Johnson, V.E., Khurana, I.K. and Reynolds, J.K. (2002), “Audit-firm tenure and the quality of
financial reports”, Contemporary Accounting Review, Vol. 19, pp. 637-660.
Jones, J. (1991), “Earnings management during import relief investigations”, Journal of
Accounting Research, Vol. 29, pp. 193-228.
Kothari, S.P., Leone, A. and Wasley, C. (2005), “Performance matched discretionary accrual
measures”, Journal of Accounting and Economics, Vol. 39, pp. 163-197.
Krishnan, G.V. (2003), “Audit quality and the pricing of discretionary accruals”, Auditing:
A Journal of Practice & Theory, Vol. 22, pp. 243-263.
Lakonishok, J., Shleifer, A. and Vishny, R. (1994), “Contrarian investment extrapolation and
risk”, Journal of Finance, Vol. 49, pp. 1541-1578.
Lennox, C. (1999), “Audit quality and auditor size: an evaluation of reputation and deep
pockets hypothesis”, Journal of Business Finance & Accounting, Vol. 26, pp. 779-805.
McNichols, M. (2000), “Research design issues in earnings management studies”, Journal of
Accounting & Public Policy, Vol. 19, pp. 313-345.
Mansi, S.A., Maxwell, W.F. and Miler, D.P. (2004), “Does audit quality and tenure matter to
investors? Evidence from the bond market”, Journal of Accounting Research, Vol. 42,
pp. 755-793.
Menon, K. and Williams, D. (2004), “Former audit partners and abnormal accruals”,
The Accounting Review, Vol. 79, pp. 1095-1118.
Myers, J.N., Myers, L.A. and Omer, T.C. (2003), “Exploring the term of the auditor-client
relationship and the quality of earnings: a case for mandatory auditor rotation?”,
The Accounting Review, Vol. 78, pp. 779-799.
Palmrose, Z. (1986), “Audit fees and auditor size: further evidence”, Journal of Accounting
Research, Vol. 24, pp. 97-110.
Palmrose, Z. (1988), “An analysis of auditor litigation and audit service quality”, The Accounting
Review, Vol. 63, pp. 55-73.
Renas, S.M. and Cebula, R.J. (2005), “Enron, herding, and the deterrent effect of
disclosure of impropriety”, American Journal of Economics and Sociology, Vol. 64,
pp. 743-756.
 ARJ Reynolds, J.K. and Francis, J.R. (2000), “Does size matter? The influence of large clients on
office-level auditor-reporting decisions”, Journal of Accounting and Economics, Vol. 30,
26,1 pp. 375-400.
Skinner, D. and Sloan, R. (2002), “Earnings surprises, growth expectations and stock returns or
don’t let an earnings torpedo sink your portfolio”, Review of Accounting Studies, Vol. 7,
pp. 289-312.
74 Subramanyam, K.R. (1996), “The pricing of discretionary accruals”, Journal of Accounting and
Economics, Vol. 22, pp. 249-281.
Teoh, S.H. and Wong, T.J. (1993), “Perceived auditor quality and the earnings response
coefficient”, The Accounting Review, Vol. 68, pp. 346-367.

Corresponding author
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)

Robert Houmes can be contacted at: rhoumes@ju.edu

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com

Or visit our web site for further details: www.emeraldinsight.com/reprints
 This article has been cited by:

1. Ahsan HabibDepartment of Accounting, Auckland University of Technology, Auckland, New Zealand

Rong GongJames and Monroe, Auckland, New Zealand Mahmud HossainDepartment of Accounting,
Curtin University of Technology, Perth, Australia. 2013. Overvalued equities and audit fees: a research
note. Managerial Auditing Journal 28:8, 755-776. [Abstract] [Full Text] [PDF]
Downloaded by University of Sri Jayewardenepura At 03:49 12 January 2019 (PT)