Professional Documents
Culture Documents
Lee and Park 2013 - Subjectivity in Fair Value Estimates
Lee and Park 2013 - Subjectivity in Fair Value Estimates
a r t i c l e i n f o a b s t r a c t
JEL classification: This study empirically examines whether difference in audit quality is reflected in the pricing of other com-
M41 prehensive income (OCI). Specifically, we first investigate whether OCI measures of Big 4 clients are more
M42 value-relevant than those of non-Big 4 clients. Considering different degrees of subjective management judg-
ment involved in the OCI reporting process, we then explore whether the differential valuation effect of OCI
Keywords:
between Big 4 and non-Big 4 clients is more pronounced for more subjective OCI components (e.g., minimum
Other comprehensive income
pension liability and foreign currency-translation adjustment) than a less subjective component (e.g., mar-
Audit quality
Subjectivity in fair-value estimate ketable securities adjustment). We predict that the aggregate OCI of a Big 4 client is more value-relevant
than that of a non-Big 4 client. We also hypothesize that the differential valuation effect between Big 4
and non-Big 4 clients can be attributed to the amount of subjective assumption and judgment required in es-
timating OCI. Consistent with our predictions, we find that aggregate OCI audited by a Big 4 auditor has
incremental information content over earnings, compared to OCI audited by a non-Big 4 auditor. More inter-
estingly, our results also show that the differential valuation effect between Big 4 and non-Big 4 clients is
stronger for OCI components of a more subjective nature. Our results are robust even after controlling for
self-selection bias, the potential effect of the financial crisis, and other related effects.
© 2013 Elsevier Ltd. All rights reserved.
0882-6110/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.adiac.2013.05.003
C. Lee, M.S. Park / Advances in Accounting, incorporating Advances in International Accounting 29 (2013) 218–231 219
some components of OCI may involve more subjective estimates, there- is available. Thus, we argue that among OCI components, marketable-
by adding noise to financial reporting. securities adjustments are less subject to management judgment and
There is a widespread belief that because of the subjectivity assumption than other components of OCI.4
of fair-value estimates, managers have incentives for substantial Given that perceived audit quality is valued by the capital market,
reporting judgment and discretion, and thus it is difficult for auditors we predict that Big 4 auditors in general provide higher-quality audits
to challenge. Recently, the PCAOB (Fall of 2011) issues a detailed re- with respect to the amount of subjectivity and management judg-
port outlining certain high risk-audit areas and challenges that it ment in fair-value estimates compared to non-Big 4 auditors, for sev-
identified during its regular inspections of registered accounting eral reasons. First, Big 4 auditors are more prone to litigation risk than
firms. Fair-value measurement is identified as one of the more signif- smaller auditors (see Becker, DeFond, Jiambalvo, & Subramanyam,
icant high-risk audit areas in the PCAOB's report.3 Primary deficien- 1998; Dye, 1993; Francis & Wang, 2008; Kim, Chung, & Firth, 2003).
cies identified by the PCAOB relating to the fair-value measurement Dye (1993) argues that auditors with more wealth at risk from litiga-
of financial instruments include (1) whether fair-value measure- tion have greater incentive to issue accurate reports. It is also known
ments are determined using appropriate valuation methods; and that Big 4 auditors are more likely to be sued in case of misstatements
(2) the reasonableness of management's significant assumptions because of their “deeper pockets” (Kim et al., 2003) and that audit
used to measure fair value, such as discount rates, and credit loss ex- failures produce greater reputation losses, especially for Big 4 audi-
pectations (see PCAOB, 2011). Especially, the PCAOB's report states tors, because they possess greater reputational capital (Becker et al.,
that the attention and additional scrutiny directed towards these 1998). Second, Big 4 firms mitigate information asymmetry more
high-risk areas will be considerable. In addition, Mark Olson, the for- than non-Big 4 auditors. Prior studies (Francis, Maydew, & Sparks,
mer chairman of the PCAOB, expresses concern about the challenges 1999; Willenborg, 1999) document that, by providing higher quality
for assurance services related to fair-value estimates, saying that auditing services, Big 4 auditors help reduce information asymmetry
“The increased use of fair value accounting poses a challenge for audi- between shareholders and managers. Khurana and Raman (2004)
tors and the PCAOB” (PCAOB, 2007a). Considering the nature of OCI, provide empirical evidence that the cost of equity capital is lower
which is mainly derived from fair-value application and subject to for Big 4 clients than for non-Big 4 clients. Third, large accounting
managerial judgment and discretion, subjective fair-value estimates firms are subject to a regular annual inspection by the PCAOB (see
of OCI should be of particular concern and scrutinized closely by PCAOB inspection rule 4003). Although four second-tier audit firms
auditors. Specifically, more extensive audit work on subjective OCI es- (see 2010 PCAOB inspection report) are also inspected annually by
timates will be necessary for auditors, because a high degree of man- the PCAOB, Big 4 firms are more likely to be subject to the inspection
agement judgment and subjectivity may present risks of material than non-Big 4 auditors, including the second-tier firms, because Big 4
misstatement, thereby increasing the audit risk (SAS No.99, 2002). In- audit firms have far more issuer clients than second-tier firms and a
tensive scrutiny by auditors, in turn, reduces the effect of subjective higher number of issuers inspected.5 Closer monitoring by the PCAOB
management judgment, thereby enhancing the quality of financial induces large auditors to increase their audit effort and make more con-
reporting. Therefore, it is an interesting research question to deter- servative decisions in their audit engagements. That is, large auditors
mine whether the valuation effect of OCI between Big 4 and non-Big are likely to be more diligent and watchful (e.g., DeFond, 2010). Fourth,
4 clients is contingent upon the amount of subjectivity in professional Big 4 auditors have a greater ability to constrain questionable account-
judgment over estimates of OCI measures. ing decisions, because they in general have better audit technology,
In this study, we empirically examine whether differences in audit superior knowledge, and a strong negotiation stance with clients in
quality are reflected in the pricing of OCI. Specifically, we first investi- terms of financial-statement requirements (DeFond & Jiambalvo,
gate whether OCI items for Big 4 clients have incremental information 1993; Francis et al., 1999).6 Finally, the market values a perceived differ-
content in explaining stock returns compared to those of non-Big 4 cli- ence in audit quality between Big 4 and non-Big 4 (Knechel, Naiker, &
ents. The basic premise of this study is that audit quality reflects a form Pacheco, 2007; Menon & Williams, 1991; Palmrose, 1988; Teoh &
of service distinction that the capital market values differently. Given Wong, 1993; Watts & Zimmerman, 1986).
that Big 4 auditors provide higher-quality assurance regarding the cred- We partition OCI components into two groups based on the degree
ibility of reported earnings (e.g., Khurana & Raman, 2004), we predict of subjective judgment: (1) a less subjective component—the change
that the OCI audited by a Big 4 auditor is more value-relevant than in unrealized gains and losses on available-for-sale marketable securi-
that audited by a non-Big 4 auditor. ties (ΔSEC), and (2) more subjective components—the sum of other
Based on the level of subjectivity and degree of management dis- OCI components, such as change in pension or other postretirement
cretion involved in the OCI reporting process, we then explore wheth- benefits (ΔPEN), change in adjustments for foreign currency translation
er the differential valuation effect of OCI between Big 4 and non-Big 4 (ΔFOR), and change in adjustments for derivative instruments that are
clients is stronger for more subjective OCI components compared to a designated as cash-flow hedges (ΔDER).7 Following prior research
less subjective component (e.g., marketable-securities adjustment). (e.g., Becker et al., 1998; Craswell, Francis, & Taylor, 1995; DeAngelo,
Some OCI components, such as foreign-currency translation and min- 1981; Khurana & Raman, 2004), we rely on auditor type (i.e., Big N) as
imum pension-liability adjustments may utilize more subjective
estimates than the marketable-securities adjustment (e.g., Dhaliwal
et al., 1999) and therefore require more assurance from auditors.
When an active market for a security exists, quoted market prices 4
Not all available-for-sale securities are classified as level 1 estimates. Some of such
are the best evidence of fair value and should be used as the basis financial instruments can be grouped as level 2 or level 3. This is even more severe for
for measurement, because investors and creditors regularly rely on financial institutions, because available-for-sale securities are often classified as level 1,
those prices to make their decisions (PCAOB, 2008). The level of man- level 2, or level 3 estimates. In this study, we exclude financial institutions from our
sample to mitigate such a potential impact. We thank an anonymous reviewer for this
agement judgment required in establishing the fair value of financial
insight.
instruments may be minimal when a quoted price in an active market 5
Church and Shefchik (2012) report that, in 2009 client portfolios, Big 4 firms have
many more issuer clients (6648) than second-tier firms (1432), and 267 issuers of Big
4 audit firms (compared to 104 issuers of the second-tier audit firms) are inspected in
3
Specifically, some of the more significant high-risk audit areas identified by the 2009.
6
PCAOB’s report are: (1) fair-value measurements for financial instruments; (2) fair- DeFond and Jiambalvo (1993) find that Big 8 (now Big 4) auditors are more likely
value measurement for non-financial assets; (3) impairment of goodwill, indefinite- to oppose questionable accounting methods that increase reported earnings.
7
lived intangible assets, and other long-lived assets; and (4) revenue recognition, and We also examine each component of OCI separately and show the results. Details
others (e.g., valuation of inventory, income taxes). are discussed in a later section.
220 C. Lee, M.S. Park / Advances in Accounting, incorporating Advances in International Accounting 29 (2013) 218–231
a surrogate for audit quality.8 We also use auditor industry specializa- auditor change is positive (negative) when a client switches to (from)
tion as an alternative proxy for audit quality. a high-quality auditor because prestigious auditors offer better moni-
Consistent with our predictions, we find that aggregate OCI audited toring capabilities. Recently, Knechel et al. (2007) find that firms
by Big 4 auditors has incremental information content in explaining switching between Big 4 auditors experience significant positive (nega-
stock returns relative to OCI audited by non-Big 4 auditors. This implies tive) abnormal returns when the successor auditor is an industry spe-
that investors account for audit quality in their valuation processes for cialist (not a specialist). They also find that such market reactions are
reported OCI items. Interestingly, our results also show that the differ- more likely to be caused by changes in perceived audit quality rather
ential valuation effect of OCI between Big 4 and non-Big 4 clients is than differential costs of using specialist auditors.
more pronounced for more subjective OCI components, such as mini- Numerous studies (e.g., Balsam, Krishnan, & Yang, 2003; Becker et
mum pension liability and foreign-currency translation adjustments, al., 1998; Francis et al., 1999; Reynolds & Francis, 2000; Teoh & Wong,
compared to the less subjective OCI item, marketable-securities adjust- 1993) find that audit quality is associated with earnings quality. For
ments. These results are robust after controlling for self-selection bias, instance, Becker et al. (1998) and Reynolds and Francis (2000) assert
potential effect of financial crisis, and other related effects. that high-quality auditors (e.g., Big 6 auditors in their cases) are able
This study contributes to the literature in several ways. Our study to detect earnings management better because of their superior
is the first to provide evidence that audit quality plays an important knowledge and act to constrain clients' opportunistic accounting de-
role in determining the reporting quality of OCI. Extant empirical cisions. Specifically, these studies find that clients of high-quality au-
research (e.g., Bamber, Jiang, Petroni, & Wang, 2010; Biddle & Choi, ditors (e.g., Big 4) have lower discretionary accruals than clients of
2006; Chambers et al., 2007; Choi, Das, & Zang, 2007; Kanagaretnam, lower-quality auditors.
Mathieu, & Shehata, 2009; Louis, 2003) reports that market prices infor- In addition, high-quality auditors can reduce the perceived uncer-
mation embedded in OCI measures, which is consistent with the com- tainty and noise in reported earnings, thereby resulting in a higher
prehensive income argument. Although the aforementioned studies earnings-response coefficient (ERC). Holthausen and Verrecchia (1988)
highlight the value-relevance of OCI, the effect of audit quality on the contend that investors' responses to an earnings surprise depend on
usefulness of OCI remains unexplored. To the best of our knowledge, the perceived credibility of the earnings report. In their theoretical
there is little evidence on the valuation implication of OCI conditioned framework, Teoh and Wong (1993) predict a negative association be-
on the degree of management subjectivity and judgment. Therefore, tween the noise in earnings and the ERC (see Eq. (5)). They also show
by extending prior work, this study fills a gap in the prior literature. Fur- empirical evidence of higher ERCs for clients who are audited by Big 8
thermore, by showing that our findings still hold even after controlling audit firms. Balsam et al. (2003) report that high-quality auditors,
for potential effects of the financial crisis, we add value to the literature proxied by industry specialization, result in a higher ERC than
on fair-value audits. Evidence from this study can also help standard lower-quality auditors. Hackenbrack and Hogan (2002) show that dis-
setters, auditors, and regulators better understand the reporting impli- closures about reasons for auditor changes signal information about
cations for OCI items in light of the amount of subjectivity, management earnings precision, leading to changes in ERC following auditor changes.
judgment, and audit quality. In sum, these studies suggest that auditors are an integral part of the
The remainder of this paper proceeds as follows. We discuss back- financial-reporting process and that perceived differences in audit qual-
ground and develop hypotheses in Section 2. In Section 3, we present ity explain variation in investors' responses to financial reports.
the research design. Section 4 discusses the empirical results. A sum-
mary and conclusion follow in Section 5. 2.2. Value-relevance of OCI reporting
2. Background and hypotheses development Prior research provides mixed evidence of the value-relevance of
OCI. With the exception of available-for-sale securities adjustments,
2.1. Audit quality, market perception, and earnings quality Dhaliwal et al. (1999) find little evidence that comprehensive income
is associated with stock returns. Their findings suggest that among
Audit quality is defined as the probability that the auditor will both the components of OCI, only the marketable securities adjustment
discover and report a breach in the client's accounting process improves the association between income and returns. 10 Using a
(DeAngelo, 1981). Investors are unable to directly observe audit quality sample of U.K. firms, O'Hanlon and Pope (1999) report that investors
and determine whether the reported information is an unbiased indica- do not price OCI and its components consistently. In contrast, several
tor of firms' financial performance. It is well-known that reputable audi- prior studies document that investors price OCI differently (e.g.,
tors perform higher-quality audits and better certify the reliability of Biddle & Choi, 2006; Carroll, Linsmeier, & Petroni, 2003; Chambers
financial statements (e.g., Balvers, McDonald, & Miller, 1988; Beatty, et al., 2007; Choi et al., 2007; Kanagaretnam et al., 2009; Soo & Soo,
1989; Datar, Feltham, & Hughes, 1991; DeAngelo, 1981; Francis & 1994). For instance, Soo and Soo (1994) find that the market incorpo-
Wilson, 1988; Krishnan, 2003; Palmrose, 1988; Teoh & Wong, 1993).9 rates information about foreign-translation adjustments reported in
Prior studies (Knechel et al., 2007; Menon & Williams, 1991; stockholders' equity. Biddle and Choi (2006) find that comprehensive
Palmrose, 1988) document that the market reacts differently to per- income dominates both traditional net income and fully comprehen-
ceived audit-quality differences. Menon and Williams (1991) find that sive income, measured by the change in retained earnings plus the
Big 8 clients get better pricing of stock issues compared to those hiring common stock dividend, in explaining stock returns. Chambers et al.
non-Big 8 auditors. A number of prior studies (Eichenseher, Hagigi, & (2007) provide empirical evidence that as-reported OCI components,
Shields, 1989; Johnson & Lys, 1990; Knechel et al., 2007; Nichols & such as unrealized gains and losses on marketable securities and
Smith, 1983) also present evidence that the market response to an foreign-currency translation adjustments, are priced positively.11 In
addition, Louis (2003) finds that a positive foreign-currency transla-
8 tion adjustment is associated with a loss of value and that such a
Lawrence et al. (2011) measure audit quality using discretionary accruals, the ex
ante cost-of-equity capital, and analyst forecast accuracy. They find that there are no
10
differences in audit quality between Big 4 and non-Big 4 auditors. They conclude that Dhaliwal et al. (1999) use the return/earnings association approach to examine
observed differences in quality are likely driven by client attributes, such as client size. whether adding the foreign-translation adjustment to net income increases the associ-
We conduct additional analyses after controlling for clients’ size and discuss the results ation of earnings with returns. They find no evidence that adding the translation ad-
in the sensitivity analysis section. justment to net income affects the return/earnings association.
9 11
Prior studies use auditor brand name to proxy for audit quality and examine the as- Chambers et al. (2007) claim that previous studies fail to find the value-relevance
sociation between brand name and earnings quality (e.g., Balsam et al., 2003; Becker et of OCI items because they use an estimated as-if measure of OCI that is subject to mea-
al., 1998; Reynolds & Francis, 2000). surement error.
C. Lee, M.S. Park / Advances in Accounting, incorporating Advances in International Accounting 29 (2013) 218–231 221
negative association between the translation adjustment and change method will be used for translation. Thus, the foreign-currency trans-
in value is largely attributable to those firms that are the most lation adjustment can be a subjective measure, thereby requiring
labor-intensive. more assurance from auditors.
Based on those observations, if OCI captures all sources of value
creation and investors incorporate audit quality in their valuation 2.3.3. Pensions/other postretirement benefits
processes, we expect OCI of a Big 4 client to show an incrementally The provisions of ASC 820 (“Fair value measurements and disclo-
positive association with stock returns. Thus, given that OCI is priced sures”) apply to the measurement and valuation of pensions/other
by the market (e.g., Biddle & Choi, 2006; Carroll et al., 2003; postretirement benefits plan assets in financial statements. It is well
Chambers et al., 2007; Louis, 2003), we posit the following hypothesis known that option valuation models require highly subjective inputs.
in the alternative form: Several assumptions are considered in determining a firm's pension
obligation, such as the discount rate, expected return on plan assets,
H1. The aggregate OCI of a Big 4 client is more value-relevant than
and rate of compensation increase. The expected rate of return on
that of a non-Big 4 client.
plan assets is a particularly important element in the determination
of pension expense (PCAOB, 2008). Actuarial assumptions used to de-
2.3. Subjectivity in OCI components
termine future values of plan assets and liabilities are also subject to
management judgment. Accordingly, pension-related estimates are
Under ASC 820 (“Fair value measurements and disclosures”), a cli-
likely to be classified as either level 2 or 3 inputs.
ent must determine the appropriate level in the fair-value hierarchy
for each fair-value measurement from level 1 to level 3. The level 1 in-
puts are quoted prices in active markets for identical assets or liabili- 2.3.4. Derivative instruments
ties, and the level 3 inputs indicate unobservable inputs for the asset Derivatives depend on the value of other instruments or events.
or liability. Since different consequences associated with each of the Derivatives may be valued through the use of complex models and as-
three levels of inputs may appear, managers have an incentive to in- sumptions (e.g., credit default probabilities) or the use of pricing
appropriately classify fair-value measurements within the hierarchy. specialists. The assumptions used in these models can be highly sub-
For example, an asset or liability with level-2 estimates is measured jective, and even a slight difference in assumptions could result in sig-
using observable market inputs other than quoted prices, which nificant valuation discrepancies (PCAOB, 2008).12 In particular, the
would be included in level 1. Thus, level-2 measurements might re- measurement of the fair value of interest-rate-related derivative in-
quire management to make more subjective assumptions or judg- struments is based on estimates of future interest rates, which may
ments than level-1 measurements (see PCAOB, 2007b). As a result, change based on economic or other factors. When market data are
the estimate of fair value tends to be highly subjective and requires not available, key inputs to the fair-value measurement include
significant judgment regarding the estimate of the magnitude and quotes from counterparties and other market data. Hence, a firm's de-
timing of future reporting of unit cash flows. The reporting of OCI is rivative fair-value measurements consider assumptions about coun-
subject to fair-value measurement because OCI items are in general terparty and its own non-performance risk, such as credit-default
a “mark-to-market” type of adjustments, which are derived from probabilities. Changes in the fair value of these derivative instruments
changes in interest rates, exchange rates, and other random walk pro- are recorded either through current earnings or as other comprehen-
cesses (Smithson, Smith, & Wilford, 1995). sive income, depending on the type of hedge designation. In particu-
lar, gains and losses on derivative instruments designated as cash-
flow hedges are reported in OCI and reclassified into earnings in
2.3.1. Unrealized gains or losses on available-for-sale v securities
the periods during which earnings are impacted by the hedged item
The fair value of equity securities, such as available-for-sale secu-
(ASC 815). The assurance about derivatives may require the auditor
rities, is determined based on quoted prices in active marketplaces.
to use a considerable amount of effort and time, because the valuation
That is, the determination of fair value of marketable securities is
of derivatives is based on highly subjective assumptions or is par-
based on direct observations of transactions, rather than judgments
ticularly sensitive to changes in underlying circumstances (see AU
or assumptions, thereby providing superior reliability. Not all
Section 332.41).
available-for-sale securities are classified as level-1 estimates; some
Based on our discussions above, we argue that the fair-value mea-
can be grouped as either level 2 or level 3. Nonetheless, there is still
surement of marketable securities are less subjective but the
less room for management subjectivity and judgment in determining
fair-value estimates of other OCI components are more subjective
the fair value of marketable securities compared to other components
(i.e., level 2 or 3 inputs).
of OCI, because an active market for a security exists, and quoted mar-
ket prices are usually available as evidence of fair value. Therefore, we
consider unrealized holding gains or losses on available-for-sale secu- 2.4. Auditing of fair-value measurements for OCI
rities as a less subjective OCI item.
The PCAOB's auditing-practice standards provide guidance to
2.3.2. Foreign-currency translation adjustment auditors on auditing fair-value and other accounting estimates.13
Foreign-currency translation adjustments require management's For instance, AU section 328, Auditing Fair Value Measurements and
judgment to determine the functional currency in which the foreign Disclosures, applies to auditing fair-value measurements and disclo-
entity primarily conducts its operations and generates/expends its sures in financial statements. It states that the auditor should evaluate
cash (ASC 830-10-45-6). Managers determine the functional currency whether the fair-value measurements and disclosures in the financial
of the firm and its subsidiaries in foreign countries (IAS No. 21). De- statements are in conformity with GAAP.
termination of functional currency depends on judgments about the During the audit, the auditor obtains an understanding of the
firm's current and future cash flows and could change in the future client's process for determining fair-value measurements and dis-
if there is a significant change in the firm's economic environment. closures. The auditor should also evaluate whether management's
Specifically, managers make the choice between the two translation
12
methods, temporal and current-rate methods, based on GAAP criteria. In September 2008, the FASB issued a staff position aimed at improving disclosures
about credit derivatives.
Holt (2011) contends that managers generally manipulate the func- 13
See AU section 328, Auditing Fair Value Measurements and Disclosures; AU section
tional currency choice. Pinto (2002) finds that managers opportunis- 332, Auditing Derivative Instruments; and AU section 342, Auditing Accounting
tically determine whether the temporal method or the current-rate Estimates.
222 C. Lee, M.S. Park / Advances in Accounting, incorporating Advances in International Accounting 29 (2013) 218–231
assumptions are reasonable, and whether the client's method for statements (DeFond & Jiambalvo, 1993; Francis et al., 1999). DeFond
determining fair-value measurements is applied consistently (see and Jiambalvo (1993) find that Big 8 (now Big 4) auditors tend to op-
PCAOB, 2007b). pose questionable accounting methods to increase reported earnings.
Audit firms are subject to a regular annual inspection by the PCAOB Francis et al. (1999) argue that hiring a Big 4 auditor by a client sig-
(see PCAOB inspection rule 4003), which is intended to identify the nals that the client's current and future earnings are likely to be less
quality of financial statements, including fair-value audit deficiencies. subject to managers' discretion than earnings of non-Big 4 clients.
Palmrose (2006) argues that closer monitoring by the PCAOB may in- Finally, the market values the perceived difference in audit quality
duce auditors to increase their audit effort and make more conservative between Big 4 and non-Big 4 auditors (Knechel et al., 2007; Menon &
judgments and decisions in their audit engagements. Thus, considering Williams, 1991; Palmrose, 1988; Teoh & Wong, 1993; Watts &
the subjective nature of the fair-value estimates of OCI, the role of audi- Zimmerman, 1986). For instance, Menon and Williams (1991) find
tors in evaluating and assuring management's estimates and judgment that clients employing Big 8 auditors get better pricing of stock issues
on OCI should also be a main concern to investors, because they reflect compared to those hiring non-Big 8 auditors. In a similar vein, Teoh
the perceived audit-quality difference in valuing the firm (see Knechel and Wong (1993) find that Big 4 auditors are more likely to reduce
et al., 2007; Menon & Williams, 1991; Palmrose, 1988; Teoh & Wong, the perceived uncertainty and noise in reported earnings, thereby
1993). resulting in a higher earnings response coefficient (ERC).15
We predict that Big 4 auditors in general provide higher-quality Dhaliwal et al. (1999) assert that some components of OCI, such
audits regarding the audit of fair-value estimates compared to as the foreign-currency translation and minimum pension liability
non-Big 4 auditors, for several reasons. First, Big 4 auditors are likely adjustments may bear more on subjective estimates than the
to be more concerned about litigation risk than smaller auditors. Big 4 marketable securities adjustment and thus may be noisier than the
auditors have a higher probability of being sued in case of misstate- marketable-securities adjustment. Accordingly, these components re-
ments because of their “deeper pockets” and higher insurance cover- quire more assurance from auditors compared to marketable securi-
age (Kim et al., 2003). In addition, auditors lose their reputation ties adjustments. That is, more extensive audit work on such
capital when audit-market participants perceive that they have components is required for auditors, because a high degree of subjec-
allowed misreporting (Abbott, Gunny, & Zhang, 2008; Skinner & tive assumptions, judgments, and estimates may present increased
Srinivasan, 2012; Weber, Willenborg, & Zhang, 2008). Reputation risk of material misstatement, thereby increasing the audit risk (SAS
capital is a primary factor that motivates auditors to provide No.99, 2002). The auditor's greater scrutiny, in turn, reduces the effect
high-quality auditing (DeAngelo, 1981). Thus, audit failures are likely of management's subjective judgment, thereby enhancing the quality
to produce greater reputation losses, especially for Big 4 auditors, be- of financial reporting. Therefore, given that the capital market values
cause they possess greater reputational capital (Becker et al., 1998, audit quality, we would expect the differential valuation effect be-
p.8). Francis and Wang (2008) find that Big 4 auditors impose ac- tween Big 4 and non-Big 4 clients to be more pronounced for those
counting conservatism on their clients in order to mitigate their expo- more subjective OCI components than for less subjective marketable
sure to litigation and reputation risks. securities adjustments. We posit the following hypothesis in an alter-
Second, Big 4 auditors tend to mitigate information asymmetry more native form:
than non-Big 4 auditors. Prior studies (Francis et al., 1999; Willenborg,
1999) document that Big 4 auditors provide higher-quality auditing ser- H2. The differential valuation effect of OCI between Big 4 and non-Big
vices than non-Big 4 auditors and that such high-quality service enables 4 clients is stronger for more subjective OCI components than for
firms to decrease information asymmetry between shareholders marketable securities adjustments.
and managers. In a similar vein, Khurana and Raman (2004) report a
lower cost of equity capital for Big 4 clients compared to the clients of 3. Research design
non-Big 4 firms.
Third, large accounting firms are subject to a regular annual 3.1. Data and sample
inspection by the PCAOB (see PCAOB inspection rule 4003). The
PCAOB conducts inspections of registered public accounting firms as We obtain our sample from three sources: (1) Compustat Funda-
required by the Sarbanes–Oxley Act of 2002. Specifically, the PCAOB mentals Annual for financial data, (2) CRSP Monthly Stock File for
performs annual inspections of audit firms that regularly provide stock price and computing returns, and (3) AuditAnalytics for auditor
their audit reports with respect to more than 100 issuers. Although information. Because Compustat only provides financial data for OCI
four second-tier audit firms (see 2010 PCAOB inspection report) are and components of OCI since 2001, our sample period covers eight
also inspected annually by the PCAOB, Big 4 firms are more likely to years, 2002–2009.16 First, we eliminate observations for which audi-
be subject to the inspection than non-Big 4 auditors, including the tor or audit-fee data are not available. We also exclude firms in the fi-
second-tier firms, because Big 4 audit firms have far more issuer cli- nancial services industry.17 To gather financial data, we require that
ents than second-tier firms and a higher number of issuers inspected. all firms have a non-zero value of OCI, non-negative book value of eq-
Closer monitoring by the PCAOB induces large auditors to increase uity, and non-negative value of sales. Further, we require firms to
their audit effort and make more conservative decisions in their have return data over twelve consecutive months, eight months
audit engagements. That is, large auditors are likely to be more dili-
gent and watchful (e.g., DeFond, 2010). In contrast, for smaller 15
Specifically, Teoh and Wong (1993) show empirical evidence of higher ERCs for cli-
auditing firms that audit public registrants, the PCAOB conducts in- ents who are audited by Big 8 audit firms. In addition, Balsam et al. (2003) report that
spections at least once every three years. high-quality auditors, proxied by industry specialization, result in a higher ERC than
Fourth, Big 4 auditors have a greater ability to constrain question- lower-quality auditors. Hackenbrack and Hogan (2002) show that disclosures about
reasons for auditor changes signal information about earnings precision, leading to
able accounting decisions, thus reducing uncertainty about reported
changes in ERC following auditor changes.
earnings.14 This is because Big 4 auditors in general have better 16
Since 2001, Compustat has reported financial data for OCI and components of OCI.
audit technology and superior knowledge, interpretations of GAAP 17
We delete observations that belong to the financial services industry for the fol-
that reduce the scope for management subjectivity and a strong ne- lowing two reasons: First, operations of firms in the financial industry are not compa-
gotiation stance with clients in terms of requirements for financial rable to those of firms in other industries. Second, the available-for-sale securities can
be primary holding assets of financial firms. Thus, to avoid the potential effects by any
particular industry or particular assets composition, we exclude financial firms. We,
14
The term Big 4 to denote the four largest audit firms is referred to previously as the however, conduct additional analysis including financial industry in our sample. We
Big 8, Big 6, or Big 5. find that the results are unchanged.
C. Lee, M.S. Park / Advances in Accounting, incorporating Advances in International Accounting 29 (2013) 218–231 223
we truncate earnings and return variables at the top and bottom 0.5% Panel A: Sample selection
to mitigate the effects of extreme values, yielding 17,878 firm-year
Procedures Number
observations as the final sample.18 Panel A of Table 1 presents the of obs.
sample-selection process.
Firms with non-zero other comprehensive income on the Compustat 47,868
Panel B of Table 1 presents the sample distribution by the Annual Database for 2002–2009
two-digit SIC code industry. The most heavily represented industry Less firms with missing auditor information on the AuditAnalytics 6200
is manufacturing (52.76%, SIC codes 20–39), followed by Services Less firms in financial sectors (two-digit SIC code with 60–67) 9799
(19.75%, SIC codes 70–89), and Transportation, Communication, Elec- Less firms with missing 12-month buy and hold returns on the CRSP 10,245
Less firms with negative value of equity, non-positive value of sale, and 3040
tric, and Gas (10.75%, SIC codes 40–49). Panel C of Table 1 reports the missing variables in regression model
annual distribution of our sample observations. As shown, the sample Less firms with truncated at top and bottom 0.5% for variables of change 706
observations are equally distributed over the sample period. in net income, change in other comprehensive income and 12-month
holding returns.
Final sample 17,878
3.2. Components of OCI
Panel B: Sample distribution by industry
OCI includes several components. Following the FASB ASC Glossa-
Two-digit SIC Industry # of % of
ry, firms are required to report unrealized gains and losses on code Obs. sample
available-for-sale marketable securities (SEC), gains or losses associ-
01–09 Agriculture, Forestry, and Fishing 80 0.45%
ated with pension or other postretirement benefits (PEN), adjust- 10–14 Mining 978 5.48%
ments for foreign-currency translation (FOR), and gains and losses 15–17 Construction 182 1.02%
on derivative instruments that are designated as cash-flow hedges 20–39 Manufacturing 9416 52.76%
(DER).19 We compute OCI as the change in accumulated OCI 40–49 Transportation, Communication, 1919 10.75%
Electric, and Gas
(Compustat variable name—ACOMINC). Similarly, the components 50–51 Wholesale 626 3.62%
of OCI are also computed by the change in accumulated unrealized 52–59 Retail 1024 5.74%
gains or losses on available-for-sale securities (Compustat variable 70–89 Services 3525 19.75%
name—MSA), the change in the accumulated balance of minimum 91–99 Public Administration 78 0.44%
Total 17,878 100.00%
pension liabilities (Compustat variable name—AOCIPEN), the change
in the accumulated adjustment of foreign-currency translation Panel C: Sample distribution by year
(Compustat variable name—RECTA), and the change in accumulated
Year # of Obs. % of sample
derivatives' unrealized gains or losses (Compustat variable name—
2002 1931 10.80%
AOCIDERGL). Relying on these components, we examine both the ag-
2003 2104 11.77%
gregate measure of OCI and its individual components by replacing 2004 2238 12.52%
total OCI with its components. 2005 2376 13.29%
2006 2441 13.65%
3.3. Empirical models 2007 2433 13.61%
2008 2349 13.14%
2009 2006 11.22%
With respect to the information content of OCI, prior studies (e.g., Total 17,878 100.00%
Chambers et al., 2007; Choi et al., 2007; Dhaliwal et al., 1999) express
stock returns as a function of OCI and its components. To test our H1
and H2, we estimate the following OLS regression models:
ΔNSEC change in adjustments of OCI components other than
RET it ¼ a0 þ a1 ΔNIit þ a2 ΔOCI it þ a3 BIG4it þ a4 ΔNI BIG4it available-for-sale securities, measured as the sum of change
ð1Þ
þa5 ΔOCI BIG4it þ eit in the accumulated balance of minimum pension liability
(ΔPEN), plus change in accumulated adjustment of foreign
RET it ¼ a0 þ a1 ΔNI it þ a2 ΔSEC it þ a3 ΔNSEC it þ a4 BIG4it currency translation (ΔFOR), plus change in accumulated
þa5 ΔNI BIG4it þ a6 ΔSEC BIG4it þ a7 ΔNSEC BIG4it þ eit ; derivatives unrealized gains or losses that are designated
ð2Þ as cash-flow hedges (ΔDER).
Table 3
Correlation matrix for selected variables.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
(1) RET
(2) ΔΝΙ 0.216
(3) ΔΟCI 0.139 0.046
(4) ΔSEC 0.039 0.005 0.303
(5) ΔNSEC 0.134 0.046 0.953 0.015
(6) ΔPEN 0.083 0.031 0.567 0.007 0.596
(7) ΔFOR 0.113 0.039 0.749 0.018 0.783 0.070
(8) ΔDER 0.017 −0.001 0.291 0.069 0.313 0.019 −0.002
(9) BIG4 0.022 0.006 −0.000 0.061 −0.004 −0.035 0.021 −0.004
(10) NI 0.091 0.301 0.009 0.141 0.007 −0.027 0.034 −0.011 0.073
(11) SIZE 0.058 0.024 0.008 0.115 0.001 −0.052 0.043 −0.008 0.408 0.368
(12) MB 0.007 0.002 −0.004 0.101 −0.004 0.001 0.001 −0.013 −0.015 −0.001 0.010
(13) LEV −0.005 −0.052 −0.038 -0.089 −0.042 −0.043 −0.007 −0.044 0.104 0.018 0.145 0.028
(14) LOSS −0.066 −0.208 −0.002 -0.003 −0.014 0.018 −0.037 0.010 −0.107 −0.671 −0.388 −0.001 −0.029
Pearson correlation coefficients are presented. The coefficients in bold are statistically significant at the 5% level (two-tailed test). All variables are defined in Appendix A.
and adjustment for derivative instruments that are designated as into its individual components: ΔPEN, ΔFOR, and ΔDER. We repeat our
cash-flow hedges (DER). As discussed earlier, we assume that some OCI analysis by replacing ΔNSEC with these three components as indepen-
components, such as foreign-currency translation, minimum pension- dent variables. The last column in Table 4 reports the results using
liability adjustment, and adjustment of derivatives, contain more sub- these individual components. As shown, the coefficient on ΔFOR ∗ BIG4
jective management judgments and estimates (e.g., Dhaliwal et al., (ΔPEN ∗ BIG4) is positive and significant at the 1 (10)-percent level, in-
1999), thereby requiring more audit work and increased auditor scrutiny. dicating that a Big 4 premium plays an important role in pricing, espe-
We include two variables, ΔSEC and ΔNSEC, as well as their interaction cially for adjustments of both pension liability and foreign-currency
terms with the Big 4 variable in our regression models. Table 5 reports translation.21
the results. Furthermore, to delve into the differential valuation effect of OCI
As shown in the first column, the coefficients on ΔSEC and ΔNSEC are components, we partition our sample into high- and low-ΔNSEC
positive and significant at the 1-percent level. This indicates that the groups based on median value of ΔNSEC. Panel B of Table 5 presents
market values OCI components. Interestingly, as presented in the sec- the results by the level of ΔNSEC. In the last column, the coefficient
ond column, the coefficient on ΔNSEC ∗ BIG4 is positive and significant on ΔNSEC ∗ BIG4 is positive and significant at the 1-percent level
at the 1-percent level (t-value = 3.63, p b 0.001), indicating that (t-value = 3.19, p b 0.001), indicating that the audit-quality differ-
other OCI components of Big 4 clients have stronger valuation implica- ence appear in the client group with a high level of non-securities
tions, compared to those of non-Big4 clients. In contrast, we find little OCI components. This is not the case for the Big 4 clients with a low
evidence on differential-valuation effect of the marketable securities level of non-securities OCI components. Overall, our findings reported
adjustment (ΔSEC ∗ BIG4) between Big 4 and non-Big 4 auditors. in Table 5 suggest that the differential valuation effect of OCI compo-
The coefficient difference between ΔSEC ∗ BIG4 and ΔNSEC ∗ BIG4 is sta- nents between Big 4 and non-Big 4 clients can be attributed to the
tistically significant at the 5-percent level. Next, we decompose ΔNSEC amount of subjective judgment. This is consistent with our H2.
Table 5 explanatory variable. Including this ratio helps control for the likeli-
Association between returns and components of other comprehensive income condi- hood of self-selecting into the auditor. Specifically, we estimate the
tioned on audit quality.
following probit regression in the first stage:
Panel A: OLS regression results of Big 4 vs. Non-Big 4
Table 6 Table 7
Two-stage regressions controlling for self-selection of auditors. Regression results based on propensity score matching approach.
Table 8 (continued) quartile. This may cause the lower coefficient of ΔSEC due to the pre-
Dependent variable = RET ponderance of zero values. To investigate this possibility, we utilize
Variable Predicted Model 1 Model 2 Model 3 non-zero ΔSEC observations only and repeat our analyses. Panel C
sign Coefficient Coefficient Coefficient of Table 8 displays the results using non-zero values of ΔSEC. We
(t-statistics) (t-statistics) (t-statistics) find that the results using non-zero ΔSEC observations only are qual-
Panel D: Regression results using auditor industry specialization as an alternative itatively similar to those previously reported, except for the result of
proxy for audit quality minimum pension adjustment. While the coefficients on ΔOCI ∗ BIG4
ΔDER + 0.220 (in model 1), ΔNSEC ∗ BIG4 (in model 2), and ΔFOR ∗ BIG4 (in model
(0.11)
3) are all positive and significant at the conventional level, the coeffi-
ISA ? 0.020** 0.020** 0.020**
(1.84) (1.84) (1.77) cient on ΔPEN ∗ BIG4 is positive but insignificant. Such an insignifi-
ΔNI ∗ ISA + 0.141 0.142 0.145 cant result of minimum pension adjustment may be caused by the
(1.13) (1.14) (1.17) small number of observations.
ΔOCI ∗ ISA + 1.682***
(2.77)
ΔSEC ∗ ISA + −0.848 −0.853 4.7. Analysis using auditor industry specialization as an alternative proxy
(−0.23) (−0.24) for audit quality
ΔNSEC ∗ ISA + 1.878***
(3.03)
Prior literature (e.g., Craswell et al., 1995; DeFond, 1992; Deis &
ΔPEN ∗ ISA + 1.757*
(1.52) Giroux, 1992; Hogan & Jeter, 1999; Owhoso, Messier, & Lynch, 2002;
ΔFOR ∗ ISA + 1.857*** Solomon, Shields, & Whittington, 1999) suggests that auditor industry
(2.27) specialization is a good proxy for auditor quality. To examine whether
ΔDER ∗ ISA + 2.831 our results are robust to an alternative measure for audit quality, we
(1.20)
Adjusted R2 6.70% 6.83% 6.85%
conduct an additional analysis by replacing BIG4 with ISA (i.e., auditor
N 17,878 17,878 17,878 industry specialization). Panel D of Table 8 reports the results.
Test of Coefficients: Consistent with the results reported earlier, the coefficients on
2
ΔSEC ∗ ISA = ΔNSEC ∗ ISA χ = 0.66 (p = 0.4158) ΔOCI ∗ ISA (in model 1), ΔNSEC ∗ ISA (in model 2), and ΔPEN ∗ ISA
Panel A reports the results of OLS regressions after including control variables, where and ΔFOR ∗ ISA (in model 3) are all positive and significant at conven-
NI = the level of net income before extraordinary item deflated by average total assets; tional levels. Therefore, our results are robust even with an alterna-
ΔOTHER = component of ΔOCI other than ΔSEC, ΔPEN, ΔFOR, and ΔDER; SIZE = firm's tive proxy for audit quality.
size, measured as natural logarithm of market value of equity; MB = the market to
book ratio; LEV = debt-to-asset ratio, measured as long-term liabilities divided by
total assets; LOSS = an indicator variable that takes the value of 1 if current period 4.7.1. Analysis using Big 6 Auditors as an alternative proxy for audit
net income is less than 0, and 0 otherwise; and YEAR = a firm's fiscal year dummy var- quality
iable. The t-statistics are calculated using heteroscedasticity-consistent standard error.
Recent studies (e.g., Boone, Khurana, & Raman, 2010; Lawrence
The variables of RET, ΔNI, and ΔOCI are trimmed at the top and bottom 0.5 percentile
levels. All other variables are defined in Appendix A. *, **, *** indicate statistical signif- et al., 2011) show significant similarities in audit quality between
icance at the 0.10, 0.05, and 0.01 levels, respectively, based on one-tailed tests if the co- Big 4 and non-Big 4, especially second-tier auditors. For instance,
efficient has a predicted sign, and two-tailed tests otherwise. Panel B reports the results Boone et al. (2010) report that although there is a more pronounced
of OLS regression after excluding financial crisis periods of 2008–2009. Panel C presents difference in perceived audit quality between Big 4 and second-tier
the results of OLS regression based on sample with positive SEC. In Panel D, ISA is an
indicator of industry specialized auditor (ISA) as an alternative proxy for audit quality,
auditors, little difference in actual audit quality is observed. Hence,
measured by market share threshold of 15% (based on audit fees). All other variables by adding two second-tier audit firms, Grant Thornton and BDO
are defined in Appendix A. t-statistics are calculated using heteroscedasticity-consistent Seidman, to Big 4 audit firm group, we utilize Big 6 auditors as an al-
standard error. The variables of RET, ΔNI, and ΔOCI are trimmed at the top and bottom ternative proxy for audit quality and replicate our analyses.27
0.5 percentile levels. *, **, *** indicate statistical significance at the 0.10, 0.05, and 0.01
Untabulated results confirm that our findings are robust to Big 6 audi-
levels, respectively, based on one-tailed tests if the coefficient has a predicted sign, and
two-tailed tests otherwise. tors. This evidence supports the view of similarities in audit quality
between Big 4 and second-tier auditors.28
for auditors and users of financial information (see Christensen et al., American Institute of Certified Public Accountants (AICPA) (2002). Consideration of
fraud in a financial statement audit. Statement on Auditing Standards No. 99. New
2012; PCAOB, 2007b, 2008; SEC, 2008). In particular, financial reporting York, NY: AICPA.
of OCI may be subject to management's assumptions and judgments. Balsam, S., Krishnan, J., & Yang, J. S. (2003). Auditor industry specialization and the
Extending prior literature, this study empirically explores whether earnings response coefficient. Auditing: A Journal of Practice & Theory, 22(2), 71–97.
Balvers, R., McDonald, B., & Miller, R. (1988). Underpricing of new issues and the choice
OCI items of Big 4 clients have incremental information content in of auditors as a signal of investment banker reputation. The Accounting Review,
explaining stock returns relative to non-Big 4 clients. Assuming that 63(4), 605–622.
the higher the level of subjectivity and professional judgment involved Bamber, L., Jiang, J., Petroni, K., & Wang, I. (2010). Comprehensive income: Who's afraid
of performance reporting? The Accounting Review, 85(1), 97–126.
in the financial reporting process, the more auditor scrutiny increases, Beatty, R. (1989). Auditor reputation and the pricing of initial public offerings. The
we then investigate whether the differential valuation effect of OCI be- Accounting Review, 64(4), 693–709.
tween Big 4 and non-Big 4 clients is attributable to the degree of subjec- Becker, C. L., DeFond, M. L., Jiambalvo, J., & Subramanyam, K. R. (1998). The effect of
audit quality on earnings management. Contemporary Accounting Research, 15(1),
tive judgment and estimates required in estimating OCI components.
1–24.
We find that OCI of Big4 clients is more value-relevant than that of Biddle, G., & Choi, J. (2006). Is comprehensive income useful? Journal of Contemporary
non-Big4 clients. Our results also show that the differential valuation Accounting and Economics, 2(1), 1–32.
effect of OCI between Big 4 and non-Big 4 clients is more pronounced Boone, J. P., Khurana, I., & Raman, K. (2010). Do the Big 4 and the Second-tier firms
provide audits of similar quality? Journal of Accounting and Public Policy, 29(4),
for the more subjective OCI components, such as minimum pension 330–352.
liability and foreign-currency translation adjustment, compared to Carroll, T., Linsmeier, T., & Petroni, K. (2003). Fair value vs. historical cost accounting:
the less subjective marketable-securities adjustments. These results Evidence from closed-end mutual funds. Journal of Accounting, Auditing and Finance,
18, 1–23.
are robust after controlling self-selection bias, potential effect of fi- Chambers, D., Linsmeier, T., Shakespeare, C., & Sougiannis, T. (2007). An evaluation of
nancial crisis, and other related effects. SFAS No. 130 comprehensive income disclosures. Review of Accounting Studies,
Our findings provide implications for accounting professionals, reg- 12(4), 557–593.
Chaney, P., Jeter, D., & Shivakumar, L. (2004). Self-selection of auditors and audit pric-
ulators, and future research. Since the financial reporting paradigm is ing in private firms. The Accounting Review, 79(1), 51–72.
being moved toward fair-value accounting, accounting professionals Choi, J., Das, S., & Zang, Y. (2007). Comprehensive income, future earnings and market
and regulators have cast doubt on auditors' ability or willingness to ex- mispricing. Working Paper. Seoul National University.
Christensen, B., Glover, S., & Wood, D. A. (2012). Extreme estimation uncertainty in fair
ercise professional judgment for handling fair-value audits (e.g., Jamal value estimates: Implications for audit assurance. Auditing: A Journal of Practice &
et al., 2010). Given that OCI is largely measured from the fair-value ap- Theory, 31(1), 127–146.
plication, our results imply that Big 4 auditors provide greater assurance Church, B., & Shefchik, L. (2012). PCAOB inspections and large accounting firms.
Accounting Horizons, 26(1), 43–63.
in the context of fair-value audits, especially for OCI components with
Craswell, J., Francis, R., & Taylor, S. L. (1995). Auditor brand name reputations and in-
more subjective judgment. Since our findings portray only a limited dustry specializations. Journal of Accounting and Economics, 20(3), 297–322.
scope of fair-value audits, future research may provide additional in- Datar, S., Feltham, G., & Hughes, J. (1991). The role of audits and audit quality in valuing
sights of the effectiveness of Big 4 auditors on fair-value accounting by new issues. Journal of Accounting and Economics, 14(1), 3–49.
DeAngelo, L. (1981). Auditor independence, ‘low-balling’, and disclosure regulation.
extending the scope. Journal of Accounting and Economics, 3(2), 113–127.
DeFond, M. L. (1992). The association between changes in client firm agency costs and
auditor switching. Auditing: A Journal of Practice & Theory, 11(1), 16–31.
Appendix A. Variable definitions
DeFond, M. L. (2010). How should the auditors be audited? Comparing the PCAOB in-
spections with the AICPA peer reviews. Journal of Accounting and Economics,
RETit buy-and-hold return measured over a one-year window 49(1–2), 104–108.
(eight months before to four months after the fiscal year DeFond, M. L., & Jiambalvo, J. (1993). Factors related to auditor–client disagreements
over income-increasing accounting methods. Contemporary Accounting Research,
end); 9(2), 415–431.
ΔNI it change in net income after extraordinary of firm i in year t, Deis, D. R., & Giroux, G. (1992). Determinants of audit quality in the public sector. The
scaled by average total assets; Accounting Review, 67(3), 462–479.
Dhaliwal, D., Subramanyam, K., & Trezevant, R. (1999). Is comprehensive income supe-
ΔOCIit change in accumulated other comprehensive income of
rior to net income as a measure of firm performance? Journal of Accounting and
firm i in year t, scaled by average total assets; Economics, 26(1–3), 43–67.
ΔSECit change in accumulated unrealized gains or losses on Dye, R. A. (1993). Auditing standards, legal liability, and auditor wealth. Journal of
available-for-sale securities of firm i in year t, scaled by Political Economy, 101(5), 887–914.
Easton, P. D., & Harris, T. S. (1991). Earnings as an explanatory variable for returns.
average total assets; Journal of Accounting Research, 29(1), 19–36.
ΔNSECit change in adjustments of OCI components other than Eichenseher, J., Hagigi, M., & Shields, D. (1989). Market reaction to auditor changes by
available-for-sale securities, measured as the sum of change OTC companies. Auditing: A Journal of Practice & Theory, 9(1), 29–40.
Fama, E. F., & French, K. R. (1992). The cross-section of expected stock returns. Journal
in the accumulated balance of minimum pension liability of Finance, 47(2), 427–465.
(ΔPEN), plus change in accumulated adjustment of foreign Francis, J. R., Maydew, E. L., & Sparks, H. C. (1999). The role of Big 6 auditors in the
currency translation (ΔFOR), plus change in accumulated credible reporting of accruals. Auditing: A Journal of Practice & Theory, 18(2), 17–34.
Francis, J., & Wang, D. (2008). The joint effect of investor protection and Big 4 audits on
derivatives unrealized gains or losses (ΔDER), scaled by earnings quality around the world. Contemporary Accounting Research, 25(1),
average total assets; 157–191.
ΔPENit change in accumulated balance of minimum pension ad- Francis, J. R., & Wilson, E. R. (1988). Auditor changes: A joint test of theories relating to
agency costs and auditor differentiation. The Accounting Review, 63(4), 663–682.
justment of firm i in year t, scaled by average total assets; Gow, I. D., Ormazabal, G., & Taylor, D. (2010). Correcting for cross-sectional and time-
ΔFORit change in accumulated adjustment of foreign currency series dependence in accounting research. The Accounting Review, 85(2), 483–512.
translation of firm i in year t, scaled by average total assets; Hackenbrack, K. E., & Hogan, C. E. (2002). Market response to earnings surprises condi-
tional on reasons for an auditor change. Contemporary Accounting Research, 19(2),
ΔDERit change in accumulated derivatives unrealized gains or
195–223.
losses of firm i in year t, scaled by average total assets; Heckman, J. (1979). Sample selection bias as a specification error. Econometrica, 47(1),
BIGit an indicator variable that takes a value of 1 if firm i's auditor 153–161.
in year t is a Big 4, and 0 otherwise. Hogan, C. E., & Jeter, D. C. (1999). Industry specialization by auditors. Auditing: A Journal
of Practice & Theory, 18(1), 1–17.
Holt, P. (2011). Some effects of alternative foreign currency translation methodologies
on two short-term liquidity ratios. American Journal of Economics and Business
References Administrations, 3(4), 645–651.
Holthausen, R., & Verrecchia, R. (1988). The effects of sequential information releases
Abbott, L. J., Gunny, K., & Zhang, T. (2008). When the PCAOB talks, who listens? on the variance of price changes in an intertemporal multi-asset market. Journal
Evidence from client firm reaction to adverse, GAAP deficient PCAOB inspection re- of Accounting Research, 26(1), 82–106.
ports. Working paper. University of Memphis, University of Colorado at Boulder, Jamal, K., Bloomfield, R., Christensen, T., Colson, R., Moehrle, S., Ohlson, J., et al. (2010).
and Singapore Management University. A research-based perspective on the SEC's proposed rule—Roadmap for the
C. Lee, M.S. Park / Advances in Accounting, incorporating Advances in International Accounting 29 (2013) 218–231 231
potential use of financial statement prepared in accordance with International Public Company Accounting Oversight Board (PCAOB) (2007a). Speech by Chairman of
Financial Reporting System (IFRS) by U.S. issuers. Accounting Horizons, 24(1), Board Mark W. Olson at AICPA National Conference on Current SEC and PCAOB De-
139–147. velopments. Washington, DC. http://pcaobus.org/News/Speech/Pages/12102007_
Johnson, W., & Lys, T. (1990). The market for audit services: Evidence from voluntary OlsonAICPASECandPCAOBDevelopments.aspx
auditor changes. Journal of Accounting and Economics, 12(1–3), 281–308. Public Company Accounting Oversight Board (PCAOB) (2007b). Staff Audit Practice
Jones, D., & Smith, K. (2011). Comparing the value relevance, predictive value, and per- Alert No. 2: Matters related to auditing fair value measurements of financial instru-
sistence of other comprehensive income and special items. The Accounting Review, ments and the use of specialists Washington, DC.
86(6), 2047–2073. Public Company Accounting Oversight Board (PCAOB) (2008). Auditing fair value mea-
Kanagaretnam, K., Mathieu, R., & Shehata, M. (2009). Usefulness of comprehensive in- surements and disclosures. AU Section 328 (Washington, D.C.).
come reporting in Canada. Journal of Accounting and Public Policy, 28(4), 349–365. Public Company Accounting Oversight Board (PCAOB) (2011). Report on observations
Khurana, I., & Raman, K. (2004). Litigation risk and the financial reporting credibility of PCAOB inspectors related to audit risk areas affected by the economic crisis.
of Big 4 versus non-Big 4 audits: Evidence from anglo-american countries. The Washington, D.C.
Accounting Review, 79(2), 473–495. Reynolds, J. K., & Francis, J. R. (2000). Does size matter? The influence of large clients on
Kim, J. B., Chung, R., & Firth, M. (2003). Auditor conservatism, asymmetric monitoring office-level auditor reporting decisions. Journal of Accounting and Economics, 30(3),
and earnings management. Contemporary Accounting Research, 20(2), 323–359. 375–400.
Knechel, R., Naiker, V., & Pacheco, G. (2007). Does auditor industry specialization Securities and Exchange Commission (2008). SEC Office of the Chief Accountant and
matter? Evidence from market reaction to auditor switches. Auditing: A Journal of FASB Staff Clarifications on Fair Value Accounting-For immediate Release 2008-234
Practice & Theory, 26(1), 19–45. Washington, D.C.
Krishnan, G. V. (2003). Audit quality and the pricing of discretionary accruals. Auditing: Skinner, D., & Srinivasan, S. (2012). Audit quality and auditor reputation: Evidence
A Journal of Practice & Theory, 22(1), 109–126. from Japan. The Accounting Review, 87(5), 1737–1765.
Lawrence, A., Minutti-Meza, M., & Zhang, P. (2011). Can Big 4 versus Non-Big 4 differences Smithson, C., Smith, C., & Wilford, D. (1995). Managing financial risk: A guide to derivative
in audit-quality proxies be attributed to client characteristics? The Accounting Review, products, financial engineering and value maximization. New York: Irwin Professional
86(1), 259–286. Publishing.
Louis, H. (2003). The value relevance of the foreign translation adjustment. The Solomon, I., Shields, M. D., & Whittington, O. R. (1999). What do industry-specialist
Accounting Review, 78(4), 1027–1047. auditors know? Journal of Accounting Research, 37(1), 191–208.
Menon, K., & Williams, D. (1991). Auditor credibility and initial public offerings. The Soo, B. S., & Soo, L. G. (1994). Accounting for multinational firms: Is the translation pro-
Accounting Review, 66(2), 313–332. cess valued by the stock market? The Accounting Review, 69(4), 617–637.
Nichols, D., & Smith, D. (1983). Auditor credibility and auditor changes. Journal of Teoh, S., & Wong, T. (1993). Perceived auditor quality and the earnings response coef-
Accounting Research, 21(2), 534–544. ficient. The Accounting Review, 68(2), 346–366.
O'Hanlon, J., & Pope, P. (1999). The value-relevance of U.K. dirty surplus accounting Titman, S., & Trueman, B. (1986). Information quality and the valuation of new issues.
flows. The British Accounting Review, 31(4), 459–482. Journal of Accounting and Economics, 8(2), 159–172.
Owhoso, V. E., Messier, W. F., & Lynch, J. G. (2002). Error detection by industry- Watts, R. L., & Zimmerman, J. L. (1986). Positive accounting theory. Upper Saddle River,
specialized teams during sequential audit review. Journal of Accounting Research, NJ: Prentice Hall.
40(3), 883–900. Weber, J., Willenborg, M., & Zhang, J. (2008). Does auditor reputation matter? The case
Palmrose, Z. -V. (1988). An analysis of auditor litigation and audit service quality. The of KPMG Germany and ComROAD AG. Journal of Accounting Research, 46(4),
Accounting Review, 63(1), 55–73. 941–972.
Palmrose, Z. -V. (2006). Maintaining the value and viability of auaditors as gatekeepers White, H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a
under SOX: An auditing master proposal. In Y. Fuchita, & R. E. Litan (Eds.), Financial direct test for heteroskedasticity. Econometrica, 48(4), 817–838.
gatekeepers: Can they protect investors? (pp. 103–135). Washington, D.C.: Brookings Willenborg, M. (1999). Empirical analysis of the economic demand for auditing in the
Institution Press and the Nomura Institute of Capital Markets Research. initial public offering market. Journal of Accounting Research, 37(1), 225–238.
Pinto, J. (2002). Foreign currency translation method choice: insights from Game
Theory. Journal of Applied Business Research, 18(4), 25–34.