Are Key Audit Matter Disclosures Useful in Assessing Financial Distress?

Full text

Are Key Audit Matter Disclosures Useful in Assessing Financial Distress?
María-del-Mar Camacho-Miñano
Associate Professor of Accounting
Accounting and Finance Department
Faculty of Business Administration and Economics
Complutense University of Madrid, Madrid
marcamacho@ccee.ucm.es
Nora Muñoz-Izquierdo
Assistant Professor of Accounting
Accounting and Finance Department
CUNEF, Madrid
nmunoz@cunef.edu
Morton Pincus
Dean’s Professor of Accounting
The Paul Merage School of Business
University of California, Irvine
949-824-4062
mpincus@uci.edu
Patricia Wellmeyer *
Clinical Assistant Professor of Accounting
The Paul Merage School of Business
University of California, Irvine
949-824-9359
patricia.wellmeyer@uci.edu
December 4, 2020
* Corresponding author.
The authors appreciate the helpful comments provided by two anonymous reviewers from the
AAA Audit-Mid Year meeting. Pincus and Wellmeyer thank the Paul Merage School of Business
at the University of California-Irvine (UCI) for financial support. Camacho-Miñano thanks to the
“Beca del Amo” 2019-Complutense University of Madrid for the financial support for her visit to
UCI and the research project PR87/19-22586 entitled "Regulation and Financial Information to
mitigate business risks" Banco Santander-UCM call (2019).

2
Are Key Audit Matter Disclosures Useful in Assessing Financial Distress?
Abstract
This study examines the usefulness of new expanded audit report key audit matters (KAM)
disclosures in assessing firm financial distress. Using KAM disclosures from a manually-collected
sample of Premium listed firms in the U.K. for 2013 to 2018, we explore the association between
the number, level (entity versus account), financial category (profitability, liquidity, or solvency),
and nature (individual type) of auditor disclosed KAMs and firm financial distress. We find
evidence that the greater the number of KAMs disclosed on a firm’s audit opinion, the higher the
financial distress level of the firm. Results also show an association between the level and nature
of KAM(s) disclosed by an auditor and financial distress level. Specifically, we find evidence that
certain entity-level (e.g. going concern, restructuring and discontinued operations, and mergers
and acquisitions) and individual account-level (e.g. revenue and management estimates) KAMs,
as well as account-level KAMs with primary impact on a firm’s profitability and solvency, are
more likely to be disclosed than others when firms are in financial distress. Our findings also
suggest that KAMs have predictive ability in assessing the subsequent period financial distress of
a firm. In all, the results provide evidence of a significant relation between KAM disclosures and
firm financial distress levels, and in so doing, show that expanded auditor report regulation can
help financial statement users assess and predict one of the main risks associated with a firm - the
risk of failure.
Keywords: expanded audit report, critical/key audit matters (KAMs), risk of material
misstatement, ISA 700, firm risk, financial distress, auditor characteristics.

3
I. Introduction
The recent enactment of expanded auditor report regulation in the U.K., U.S., and E.U. has
spurred much debate on whether the benefits to financial statement users of enhanced transparency
into the audit process exceed the costs of requiring auditors to disclose potentially sensitive firm-
specific information in their audit opinions. A large part of this debate centers on a requirement in
the presently enacted expanded auditor reporting regulations that auditors disclose the financial
reporting area(s) they deem carry the most significant risk of material misstatement for the client
firm. In requiring these disclosures - termed Key Audit Matters (KAMs) under U.K./European
regulation and Critical Audit Matters (CAMs) under U.S. PCAOB regulation
1
- regulators aim to
make the auditor’s report more informative to stakeholders by providing visibility into those areas
having the greatest effect on the overall audit strategy and requiring more challenging audit
judgments (PCAOB 2017; FRC 2013).
Academic research has begun to examine the utility of KAMs, yielding mixed findings on
the usefulness of these disclosures to financial statement users. Recent studies such as Smith
(2019) and Loew and Mollenhaur (2019) find KAM disclosures enhance the usefulness of the
annual report with respect to firm characteristics and client-specific audit risks, as does Moroney
et al. (2020) who condition by audit firm size. However, studies such as Carvel and Trinkle (2017)
and Kohler et al. (2020) find no evidence that investors consider KAM disclosures as incrementally
informative. Lennox et al. (2019) report mixed results, indicating that while investors do not find
KAM disclosures as incrementally informative in the short-term, KAM disclosures seem to
1
While named and defined with slightly different wording in IAASB and PCAOB expanded auditor regulations, a KAM/CAM in
both standards is intended to represent a matter assessed by the auditor as a significant area of risk of material misstatement for the
client firm. For consistency and readability, as well as the fact that we focus on U.K. expanded audit reports, we denote these
disclosures as “KAMs” for the remainder of the paper.

4
reliably capture the uncertainty in accounting measurements in the longer-term. The relative
scarcity of research examining the impact of mandated KAM reporting and the mixed findings
they have yielded indicate that further research is needed to provide clarity on the extent to which
these new auditor disclosures may be useful in formulating judgments about audited firms.
One of the most important yet complex judgments to make about a firm is the extent to
which it remains a viable operating and financial entity. Audit standards require that auditors make
assessments of a client’s ability to remain a going concern (GC) and report in their audit opinions
when they believe there exists substantial doubt in this regard (IAASB 2015; PCAOB 1989).
Studies have shown, however, that auditors often misreport going concern classifications (Barnes
2004; Geiger et al. 2005; Feldmann and Read 2010; Geiger et al. 2014; Pincus et al. 2017; Choi et
al. 2018; Read and Yezegel 2018). In general, the evidence suggests that 40 to 50 percent of firms
going bankrupt did not receive a prior going concern opinion (GCO) (Feldmann and Reed 2010).
The proportion of firms that do not fail in the year subsequent to receiving a GCO has been found
to be even higher, at approximately 76 to 80 percent (Lennox 1999; Geiger and Rama 2006). The
2008 global financial crisis and its consequences have propelled renewed interest from regulators
and financial statement users in auditors’ ability to provide accurate GCO assessments and has
fueled calls for new approaches to measuring the extent to which a firm remains a viable entity
(Carson et al. 2012).
The aim of this study is to examine the extent to which auditor reported KAMs, which shed
light on the likelihood of material misstatements in client firms’ financial statements, are useful in
assessing firm financial distress. Specifically, we investigate whether the total number, level,
category, and nature of reported KAMs are associated with levels of financial distress. We focus
our examination on U.K. firms with Premium listings on the London Stock Exchange as the U.K.’s

5
Financial Reporting Council (FRC) was the first to implement expanded auditor reporting pursuant
to the passage of ISA 700 in 2013. To carry out our analysis, we hand-collect KAM disclosures
for firms receiving U.K. expanded auditor reports for fiscal years 2013 to 2018. To examine the
relation between KAMs and financial distress, we partition KAMs and regress them on several
established proxies of financial distress (Altman et al. 2017). Specifically, in addition to examining
the relation between the number of KAMs a firm receives and its risk of financial distress, we read
all individual KAM descriptions as well reviewed professional guidance (FRC 2015) and prior
literature and partition KAMs as follows: (1) by level, whether the KAMs represent an entity-wide
or account-level risk of material misstatement to a firm, (2) by category for account-level KAMs,
whether their primary impact is on a firm’s profitability, liquidity, or solvency, and (3)
individually, according to the nature (individual type) of risk of material misstatement each
presents to the company’s financial statements. The partition into entity versus account level risks
follows prior literature (IAS 701; Lennox et al. 2019; Sierra-García et al. 2019). Our study expands
upon this KAM categorization by classifying and analyzing KAMs by their primary effect on the
variables of company financial health ( profitability, liquidity, or solvency). These categorizations
form the basis of most financial analysis ratios included in well-known and popular bankruptcy
prediction models (Altman 1968; Altman et al. 2017; Lukason and Laitinen 2019). Second, we
also contribute to the archival studies on the usefulness of the expanded audit reporting
requirements (Lennox et al. 2019; Sierra-García et al. 2019) by considering both entity and account
level KAMs individually, as we aim to discover the different nature (types) of risks of material
misstatement reflected in KAMs most useful in assessing firm financial distress.
As noted above, the identification and disclosure of KAMs is intended to inform financial
statement users of areas within a client firm deemed by the auditor to carry significant risks of

6
material misstatement(Andreicovici et al 2020). Several recent studies have found client
characteristics are determinants of the type of KAMs disclosed by an auditor, and that differences
exist in the extent to which entity-wide versus account-level KAMs capture value relevant risks at
the firm level (Lennox et al. 2019; Sierra-García et al. 2019). Prior research into bankruptcy shows
the extent to which a client firm’s financial ratios indicate risks of failure varies depending on
whether the primary impact is on a firm’s profitability, liquidity, or solvency (Altman 2017;
Lukason and Laitinen 2019). We first hypothesize the greater the number of KAMs disclosed in a
firm’s audit report, the greater the level of financial distress the firm faces. Second, we expect the
level, category, and nature of KAMs an auditor discloses are useful in assessing the firm’s level
of financial distress. Hence, we hypothesize the extent to which individual KAMs reflect entity-
wide versus account-level risks of material misstatement and the extent to which account-level
risks impact a firm’s profitability, liquidity, or solvency are useful in assessing a firms’ financial
distress. Additionally, we explore the impact individual KAMs have on the assessment of financial
distress.
Findings from our study show auditor disclosed KAMs have useful explanatory and
predictive power in assessing the risk of firm financial distress. Specifically, results show the
greater the number of KAMs disclosed by an auditor on a client firm’s audit report, the higher the
level of financial distress for the firm. Findings also reveal that KAMs describing an entity-level
risk and a primary impact on a company’s profitability and solvency are more likely to be disclosed
than KAMS reflecting liquidity when firms are in financial distress. Additionally, results indicate
that when assessing financial distress, the usefulness of KAMs increases when considering the
type of risk they represent to a firm individually. Moreover, we find that individual KAMs
describing the risk of material misstatements related to some entity-level KAMs (such as going

7
concern and exceptional items presentation and disclosure) and to certain account-level individual
KAMs (such as revenue recognition, intangibles, pension and accruals and estimates) are most
significantly associated with higher levels of firm financial distress.
Findings from our study contribute to both the financial distress and expanded auditor
reporting streams of literature. While a growing number of studies have examined the extent to
which investors perceive expanded auditor reports as incrementally useful, we believe our study
is the first to examine the usefulness of expanded auditor disclosures in helping financial statement
users make judgments about a firm’s viability. By presenting evidence of the useful value of KAMs
in assessing and predicting the extent to which a firm faces financial distress, this study also
provides evidence of another way in which audit information, in this case KAMs, can be useful in
assessing a firm’s financial distress. These findings should be useful to academics in conducting
future research where the risk of financial distress is an important consideration, and to financial
statement users in making judgments about the extent of a firm’s financial distress. Lastly, findings
from this study should be of interest to regulators and practitioners as they evaluate the benefits
and costs of mandated engagement-specific audit disclosures.
The remainder of this paper is organized as follows. In the next section, we provide
additional background and develop our hypothesis. Section III describes our research design and
section IV our data and sample. In section V we present our findings and we conclude in the last
section.
II. Prior Literature and Hypothesis Development
The 2008 global financial crisis fueled renewed attention on the ability of auditors and the
audit process to detect and report on the true and fair presentation of firms’ financial statements
(Read and Yezegel 2018). This attention led to calls from financial statement users for enhanced

8
transparency from auditors into the audit process and the identification and resolution of firm-
specific audit matters. In response, expanded auditor reporting regulation enacted by both the
PCAOB and IAASB now requires auditors to disclose a description of the most significant
assessed firm risks of material misstatement and corresponding audit responses in the body of a
firm’s audit opinion. The major aim of these disclosures is to leverage auditors’ proprietary
insights and expertise on client firms to provide information useful to financial statement users in
understanding where the significant financial reporting risks exist for a firm and in assessing how
changes in the economy may affect the firm’s future performance (IAASB 2016; PCAOB 2013).
Recent research has emerged on KAMs, providing preliminary insights on their perceived
usefulness as well as the client and auditor factors that may influence the magnitude and type of
KAM disclosures auditors report for client firms.
2.1 Prior Literature and Hypotheses
Perceived utility value of KAMs
Studies examining the informative value of KAMs from a user perspective have yielded
mixed conclusions on the importance and interpretations financial statement users give these
disclosures in formulating judgments about firm risk. Focusing on lenders' reactions to KAM
disclosures, Porumb et al. (2019) find evidence suggesting KAMs improve lenders’ ability to
assess borrowing risks. Moreover, Trpeska et al. (2017) find lenders place high importance on the
number of risks disclosed, assessing borrowers with more risk of material misstatement disclosures
as riskier firms. In contrast, Boolaky and Quick (2016) examine the credit approval decisions of
bank loan officers (directors) and find no evidence that the inclusion of KAMs in audit reports
significantly affects directors’ credit-granting decisions or their perceptions about the quality of a
firm’s financial reporting quality.

9
In examining equity investor reactions to KAMs, market studies such as Carver and Trinkle
(2017), Bédard et al. (2019a, 2019b), Liao et al. (2019), Burke et al. (2020), and Gutierrez et al.
(2020) yield no evidence that markets react to these disclosures, suggesting investors do not find
them incrementally informative. Köhler et al (2020) suggest KAM disclosures are likely to have
lessened communicative value for investors given potential difficulties in processing the new
information. Lennox et al. (2019) report mixed results, finding that while investors do not seem to
find KAM disclosures to be incrementally informative using short window reactions, using long-
window association tests reveal KAM disclosures seem to reliably capture the risks inherent in a
firm’s accounting measurements. Supporting this latter finding, Smith (2019) reports evidence
suggesting audit reports post-ISA 700 are easier to read and better reflect the risk-related nature of
financial statement audits. Experimental studies to date also yield inconclusive findings with
respect to the usefulness of KAMs to shareholders. While Christensen et al. (2014) find that non-
professional investors who are presented with an audit report’s KAM disclosures are more likely
to change their investment decisions compared with investors who receive a standard audit report,
Kohler et al. (2020) show KAMs do not seem to affect the investment decisions of non-
professional investors. They also find, however, that professional investors perceive KAMs to
significantly reflect a company’s economic situation. Recently, Moroney et al (2020) provides
evidence that investors perceive the audit to be more valuable when KAMs are disclosed than
when KAMs are absent.
With preliminary studies on mandated KAM disclosures yielding mixed results on the
perceived usefulness of KAMs to financial statement users, the extent to which the number and
type of KAM disclosures documented in a firm’s audit opinion may convey useful information
about the complexity and financial condition of the firm is an open question.

10
Number of KAMs and financial distress risk
While the enactment of expanded auditor report regulation newly requires auditors to
publicly disclose their assessments and audit plan response to client-generated risks, the use of a
risk-based approach in planning and executing audit strategies has long been in place. Academic
literature provides evidence of the link between auditor client risk assessments and audit opinion
disclosures. Research suggests auditors are more likely to document the existence of client
business risks in audit planning when financial implications are significant and modify their audit
opinion when a client’s financial condition is poor (Basioudis 2008; Kochetova-Kozloski et al.
2013). Similarly, other studies find auditors modify client risk assessments in light of client
characteristics such as financial health, firm strategy, management integrity, industry, or relative
firm performance (Nelson 2009; Allen et al. 2006).
The implementation of the expanded auditor reporting regulation now enhances
transparency into auditor judgments regarding material client risk areas beyond what can be
gleaned from a modification to a standard audit opinion by requiring auditors to specifically report
these assessments via KAM disclosures. While regulators leave the determination of the number
of KAMs issued largely to auditor judgment, in identifying a KAM both IAASB and PCAOB
regulations require auditors to consider client areas assessed as significant risks in the audit,
including matters requiring complex and/or subjective estimation.
Recent studies examine the association between the number of KAMs and client
characteristics. Examining the audit reports issued in the U.K. and the E.U. under the expanded
audit report regulation, Pinto and Morais (2019) find a significant and positive association exists
between the number of issued KAMs and the business segments and specific accounting standards
reported by a firm. In their analysis of U.K. FTSE 100 firms during the period 2013-2016, Sierra-

11
Garcia et al. (2019) show client risks, such as leverage and structural complexity, and client
industry are significant determinants of the number of reported KAMs. Clients with higher
leverage and reported losses disclose a higher number of KAMs (Sierra-Garcia et al. 2019). Using
data from a sample of Brazilian listed firms, Ferreira and Morais (2019) obtain similar results.
They find that the most complex clients - measuring complexity by the number of segments
reported in the financial statements - tend to have more KAMs in their audit reports.
With regard to the number of KAMs and auditor characteristics, studies find mixed results.
Both Pinto and Morais (2019) and Sierra-Garcia et al. (2019) report a positive association between
audit fees and the number of reported KAMs, as audit fees are positively related with the
company’s specific financial, strategic, operational, and governance risks (Bortolon et al. 2013;
Yang 2018), and internal control weaknesses (Munsif et al. 2011). However, Ferreira and Morais
(2019) seem to find the opposite evidence, attributing their finding to the notion that KAM
disclosures may reflect a trade-off between maintaining the auditor’s reputation and maintaining
a certain level of audit fees. Additionally, Ferreira and Morais (2019) also find the number of
KAMs disclosed is associated with (1) auditor size, with Big 4 audit firms issuing, on average, a
greater number of KAM disclosures in their audit opinions,
2
and (2) type of audit opinion issued,
finding that clients with unmodified audit opinions have a greater number of auditor reported
KAMs. The authors attribute this finding as possibly an outcome of auditor-client negotiation as
clients seek to avoid the issuance and negative consequences linked to annual reports filed with
modified audit opinions (e.g. Securities and Exchange Commission (SEC) considers annual reports
filed with adverse opinions to be in violation of securities laws).
2
This is likely because large audit and professional service firms are more exposed to litigation risk (seen as having “deep pockets”)
or to regulatory sanctions (Dye 1993).

12
With the implementation of expanded audit report regulation, auditors now have a new
mechanism by which to alert users of material client risks without the need to qualify the audit
opinion. In communicating material risks to a company’s ability to continue as a going concern,
auditors might use KAMs as a way of signaling client financial distress risk to the market while
avoiding the negative consequences that could accrue to the client from issuing a going concern
opinion. Consequently, we propose there is a relation between the extent of financial distress and
the number of KAMs such that including the number of KAMs reported as an explanatory (or
predictive) variable to evaluate (or predict) firms’ financial distress can enhance financial distress
modeling. Our first hypothesis is:
H1: The greater the number of KAMs reported in a client firm’s audit report, the greater
the firm’s level of financial distress.
It is possible, however, that the number of KAMs may not be indicative of client financial
distress if auditors are more concerned with the litigation risk that might increase should they fail
to disclose a critical audit matter than with the risk of losing a client due to issuing too many
KAMs. If this is the case, the auditor may err on the side of issuing more KAMs as a way of
“covering their bases” and reduce the likelihood of litigation against the audit firm should a
financial reporting failure associated with a client subsequently occur. In a similar vein, it is not
obvious that an increase in the risk of a material misstatement in the financial statements is
necessarily related to the risk of financial distress. Hence, a link between number of KAMs and
financial distress is an empirical question.
Types of KAMs and Financial Distress Risk
As with the number of KAMs, expanded auditor reporting regulation allows the
determination of the type or nature of KAMs disclosed to be based largely on auditor judgment.
While scarce, academic research studying pre-expanded auditor report modifications and emphasis

13
of matter paragraphs provides some insights on the usefulness of auditor disclosures. For example,
in a natural experiment performed in Canada, Bédard et al. (2019a) find that when annual reports
include a going concern uncertainty disclosure, an emphasis of matter paragraph in the auditor’s
report may have incremental negative value to investors in contrast to when the uncertainty
disclosure is not accompanied by this paragraph. In addition, Casterella et al. (2020) suggest that
an impact on the share pricing of industry peer firms occurs when there is no warning from auditors
about an imminent bankruptcy when a rival goes bankrupt without having received a prior year
going concern audit opinion.
Moreover, a line of research has recently emerged providing insights on the usefulness of
types of audit report disclosures in assessing financial distress risk. Many of the well-established
models developed and used in the literature to assess financial distress risk rely on the analysis of
accounting-based ratios computed from financial statement data (Altman et al. 2017). Several
studies, however, question the predictive power of accounting-only variable models, documenting
evidence of enhancements to these models with the addition of market-based and non-financial
variables (Hillegeist et al. 2004; Bellovary et al. 2007; Laitinen and Laitinen 2009; Hernández-
Tinoco and Wilson 2013; Mai et al. 2019; Serrano-Cinca et al. 2019). One such category of non-
financial variables is information extracted from external auditing data, including the type of audit
opinion (Altman and McGough 1974; Altman et al. 2010) and content of audit report disclosures
(Muñoz-Izquierdo et al. 2019).
In general, prior literature using traditional pass/fail audit opinions has observed
associations between financial distress risk and modified opinions, finding the type of audit
opinion (e.g., going concern opinion), the accumulation of opinion modifications, and high auditor
rotation contribute significantly to the assessment of financial distress (Kim et al. 2008; Altman et

14
al. 2010; Piñeiro-Sánchez et al. 2012, 2013; Cenciarelli et al. 2018). Muñoz-Izquierdo et al. (2019)
extend prior findings on modified opinions and financial distress assessment by analyzing the
content of those opinions. Using pass/fail Spanish firms’ audit opinions, they suggest the quantity
and content of audit report disclosures (included in both emphases of matter and modification
paragraphs of these opinions) contribute to assessments of financial distress risk.
With the implementation of the new expanded audit report regulation, research has
emerged providing some insights on the content and determinants of KAM disclosures. No studies
that we are aware of, however, have examined how the type of KAMs reported by the auditor may
be useful in assessing a client’s financial distress risk. If new expanded audit report disclosures
provide a mechanism for auditors to inform financial statement users about client firm risk, then
factors reflecting the risk of material misstatements disclosed as KAMs might help inform users
of matters contributing to firm financial distress. Thus, in general, we posit there exists an
association between financial distress risk and the types of KAMs disclosed by auditors, so that
the inclusion of the type of KAMs as explanatory variables to shed light on client firms’ financial
distress could enhance the accuracy of predictions in financial distress modeling.
We read all of the individual KAM descriptions in our sample firms’ audit reports as well
as guidance from professional standards and prior literature to develop a classification of KAMs
by their nature (i.e., individual type). As discussed later in connection with Table 3, we identify
18 different types of KAMs. We first identify two groups: individual KAMs that represent risks to
the entity as a whole and account-specific KAMs, building on IAS 701 and Lennox et al. (2018).
Our next hypothesis is:
H2a: The risk level (entity-wide or account specific) of auditor reported KAMs (i.e., risks
of material misstatement) is associated with the firm’s level of financial distress.

15
In exploring the relation between type of KAM disclosed and firm financial distress risk,
we first examine the contemporaneous association (and later the predictability) between the level
of financial reporting risk a KAM represents for the firm (entity-wide vs account level) and the
level of a firm’s financial distress.
Next, we expand upon the entity and account KAM categorizations. Focusing on the
account-level KAMs, we divide them into profitability, liquidity, or solvency disclosures as these
are expected to be useful for assessing firm financial health. That is, these areas form the basis of
most financial ratios included in well-known and popular bankruptcy prediction models (Altman
1968; Altman et al. 2017; Lukason and Laitinen 2019), and several prior studies have documented
that profitability, liquidity, and solvency financial ratios are significant predictors of financial
distress risk (Altman 1968; Balcaen and Ooghe 2006; Cultrera and Brédart 2016; among others).
Lukason and Laitinen (2019), for instance, report a negative profitability ratio is the most
important contributor to financial distress risk predictions. Based on these findings, thus, we
believe a classification of auditor reported KAMs by their primary effect on the variables of a
firm’s financial health (profitability, liquidity, or solvency) may be useful in the assessment of
firm financial distress risk. Our next hypothesis is:
H2b: The primary impact auditor reported KAMs (i.e., risks of material misstatement)
have on the categories of a firm’s financial health (profitability, liquidity, or
solvency) is associated with the firm’s level of financial distress.
Finally, since our KAM classification system identifies all types of individual KAMs, we
explore the extent to which each individual KAM is useful in assessing financial distress. Our final
hypothesis is thus:
H2c: The nature (individual type) of auditor reported KAMs (i.e., risks of material
misstatement) is associated with the firm’s level of financial distress.

16
III. Research Design
3.1 Sample Selection
Our initial sample consists of all firms listed in London Stock Exchange in 2013 (2,149
firms). We then narrow our sample down to firms with a Premium listing classification as
expanded auditor report regulation in the UK applies only to audit reports of firms in this
classification (899 firms). Premium listed companies must present expanded auditor reports
beginning with fiscal year-ends on or after September 30, 2013.
3
Due to the costs and effort of
hand collecting data, we focus our sample on firms with a Premium listing in FY 2013 and collect
KAMS for these forms beginning with their first presentation of expanded auditor reporting (2013
or 2014 depending on their fiscal year end date) through 2018. In cases where a firm becomes
insolvent or is acquired by another entity during our sample period, we include data through the
last available fiscal year-end expanded auditor report. Financial data for sample firms are obtained
from ORBIS database. We drop firms without the necessary financial variables and also exclude
firms in the financial industry (financial firms and investment trusts) due to their distinctive
operating and regulatory nature. Table 1 summarizes the sample selection process and firm-year
observations for the final sample of 482 firms and 2,214 firm-years.
4
Place Table 1 About Here
We hand collect audit firm name and KAM disclosures from expanded audit reports for
each available firm-year in our sample. To ensure consistency in coding of KAMs by level,
category and individual type, two members of the research team independently categorized each
KAM. Any differences in coding were discussed and resolved through discussion with a
3
Firms with 2013 fiscal year ends prior to September 30 would not be required to provide expanded audit reports in their fiscal
year 2013 annual reports. Implementation of expanded auditor reporting would be required for these firms beginning with fiscal
year 2014.
4
Lack of data availability in a given firm-year can reduce the sample size of particular analyses.

17
professional auditor. Appendix A provides a sample firm’s auditor reported KAM disclosures
along with the categorization we gave to each KAM.
3.2 Model variables
3.2.1. Dependent variable
To test our hypotheses, we measure firm financial distress (FDi,t) by employing Altman’s
Z’’-Score model as the continuous dependent variable. This model is an updated version of the
original Z-Score model developed by Altman in 1983 (Altman 1983). The updated Altman model,
denoted the Z’’-Score model, is widely accepted in the academic literature as a leading bankruptcy
prediction and analysis tool for both manufacturing and non-manufacturing firms (Altman et al.
2017).
The Altman’s Z’’-Score is calculated using the following four-variable model:
Z’’-Score = 3.25 + 6.56×Z1 + 3.26×Z2 + 6.72×Z3 + 1.05×Z4
The four financial ratios of the Z’’-Score are as follows. Z1 is working capital to total assets, a
liquidity ratio, expressing the book value of net current assets of a firm over total assets; firms with
low liquidity are expected to be more financially distressed than firms with no liquidity issues. Z2
is retained earnings to total assets and reflects cumulative profitability as a proportion of total
assets. Profitability is negatively linked to bankruptcy, so a negative correlation between this long-
term profitability measure and bankruptcy is expected. Z3 is earnings before interest and taxes to
total assets. It shows how productive a firm is in generating earnings before deducting interest and
taxes, thus a low value of Z3 occurs when firms are under financial distress.
5
Lastly, Z4 is the book
value of equity to total liabilities. It captures financial leverage or capital structure by measuring
5
The Z3 ratio appears to be the most powerful predictor of bankruptcy (Altman et al. 2017), as it continually outperforms other
measures in assessing the risk of failure.

18
the relation between the firm’s shareholders’ equity and its obligations to external parties. A
decrease in this ratio indicates a warning signal for financial difficulties, as it is expected that
distressed firms are highly leveraged.
We use Z”-Score as our dependent variable in most of our analyses and label it FD. Altman
(1983) also established Z’’-Score thresholds to determine whether or not financial distress might
be present. A Z’’-Score above 2.6 indicates that a company is in a safe zone regarding financial
distress. A Z’’-Score between 1.1 and 2.6 situates a firm in a grey zone, suggesting financial
distress in the short term. Finally, a Z’’-Score below 1.1 positions a firm in the distress zone,
suggesting a high probability of financial difficulties in the current period.
6
Thus, the lower our
dependent variable (FD) is, the higher the financial distress of a firm in the sample.
Table 2 presents sample descriptive statistics. On average, firms in our sample report
$6,954 million of assets (SIZE) and $285 million in net income after tax (NI), indicating that
sample firms are large and profitable. The first five rows of Table 2 show statistics for FD, the
dependent variable, and for Z1-Z4, the four ratios that make up the Altman’s Z’’-Score model.
After winsorizing at the 1st and 99th percentiles, the average Z’’-Score (FD) is 7.41 across the
sample.
Place Table 2 About Here
3.2.2. Explanatory variables
In examining the relation between firm financial distress and auditor KAM disclosures, the
first variable we consider is the number of KAMs (No.KAM) reported in the annual audit opinions
of each firm. Consistent with Sierra-García et al. (2019) and Lennox (2019), we calculate our
No.KAM variable as the total number of individual KAMs for each firm in our sample.
6
In a robustness check, we use a dummy variable of the Z”-Score, that takes the value of 1 if the score is above 2.6 (distressed
firm) and 0 otherwise (non-distressed firm), and label it FDdum.

19
We next consider the nature of KAMs and firm financial distress. To explore this
association, we examine the impact of KAMs by level (entity-wide, ENTKAM, or account-specific,
ACCKAM), category (account-level KAMs categorized by profitability, liquidity, or solvency),
and nature (individual type). We identified seven entity-level (ENTKAM) and eleven account-level
(ACCKAM) individual KAM types. Entity-level KAMs relate to risks of material misstatement
that have a pervasive impact on the organization and its financial statements such as those related
to a client’s internal controls (ICFRAUD) and its ability to remain a going concern (GC). Account-
level KAMs represent misstatement risks affecting a specific account or disclosure in a client’s
financial statements, for example, the valuation of a firm’s intangible assets (INTANG) or the
recognition of revenues (REV). Findings in the financial distress literature show the most relevant
ratios for predicting firm financial distress levels include variables related to profitability, liquidity,
and solvency; hence, we further categorize account-level KAMs by whether their main impact is
on a company’s profitability (PROF), liquidity (LIQU), or solvency (SOLV). Finally, we examine
the relation of each individual type of KAM and firm financial distress.
Two members of the research team (one of whom had had extensive auditing experience
in public accounting) independently categorized each KAM the auditor disclosed based on the risk
the KAM potentially would have on the firm’s financial statements. The researchers’ independent
categorizations where merged and any differences in coding were discussed and reconciled with
an experienced auditor. Table 3 presents and describes our KAM categorizations.
Place Table 3 About Here
We include control variables related to auditor and client characteristics, which based on
prior literature could affect the disclosure of KAMs. The variable BIG4 identifies the audit firm
employed by each company in our sample; it equals 1 if a Big-4 audit firm is employed, and 0

20
otherwise. This is considered a proxy of audit quality (Francis and Wang, 2008). The variable
CHANGEAUDIT is a dummy variable that takes the value of 1 if there is a change in auditor from
the prior year, and 0 otherwise. We also control for firm size (SIZE) using the natural logarithm of
total assets (Sierra-García et al. 2019). Lastly, we include industry and year fixed effects to mitigate
concerns that firm industry characteristics and time-series trends may impact the frequency and
type of KAM disclosures (Lennox et al. 2019).
7
3.3. Empirical Models
The following OLS regression model is used to test H1. Recall that H1 hypothesizes a
relation between the number of disclosed KAMs and firm financial distress. Our Model 1 tests the
first hypothesis:
Model 1: FDit = β0 + β1No.KAMit + β2BIG4it + β3CHANGEAUDITit + β4SIZEit
+ CONTROLSit + μit
The coefficient β1 captures the relation between the level of financial difficulties and the number
of KAMs. Therefore, if H1 is supported, β1would be expected to be negative and significant.
H2a, H2b, and H2c, respectively, hypothesize an association between the level, category, and
individual type of KAMs disclosed and firm financial distress.
To test H2a, H2b, and H2c, we examine the contemporary association in three distinct OLS
regression models (Models 2, 3, and 4). Model 2 examines the relation between the level of risk a
KAM represents to a firm’s financial statements (i.e., entity-wide or account-specific) and firm
financial distress:
7
Financial ratios from client financial statements that have been used in other KAM archival studies are not included
in our analysis due to collinearity issues with our FD dependent variable (the Altman’s Z’’-Score model, which
consists of four financial ratios explained in Section 3.2.1).

21
Model 2: FDit = β0 + β1ENTKAMit + β2ACCKAMit + β3BIG4it
+ β4CHANGEAUDITit + β5SIZEit + CONTROLSit + μit,
where ENTKAM is the sum of the number of entity-level KAMs (KAMs 1 to 7 in Table 3)
and ACCKAM is the sum of the number of account-specific KAMs (KAMs 8 to 18 in Table 3)
from a client firm’s audit opinion for a given year. The coefficients β1 and β2 capture the relation
between the level of financial distress and the number of specific KAMs at the level of risk (entity
as wide and account as specific). If H2a is supported, β1 and β2 would be negative and significant.
Model 3 examines the relation between KAMs classified by their primary impact on a
firm’s profitability, liquidity, or solvency and firm financial distress:
Model 3: FDit = β0 + β1ENTKAMit + β2PROFit + β3LIQUit + β4SOLVit + β5BIG4it
+ β6CHANGEAUDITit + β7SIZEit + CONTROLSit + μit,
where PROF is the number of KAMs classified as having a primary effect on a firm’s profitability;
LIQU is the number of KAMs classified as having a primary effect on a firm’s liquidity; and SOLV
is the number of KAMs classified as having a primary effect on a firm’s solvency. If H2b is
supported, β1- β5coefficients should be negative and significant.
Lastly, Model 4 examines the relation between individual types of KAMs and firm
financial distress.
Model 4: FDit = β0 + β1GCit + β2ICFRAUDit + β3RDOit + β4MAit + β5TAXit + ….
+ β18PENSit + β19BIG4it + β20CHANGEAUDITit + β21SIZEit
+ CONTROLSit + μit ,

22
where GC, ICFRAUD, RDO, MA, TAX, …, PENS are KAM types as described in Table 3. If H2b
is supported, β1-β18 coefficients should be negative and significant.
The models include industry and year dummy variables. The p-values are calculated using
statistics estimated from robust standard errors clustered by firm. All continuous variables are
winsorized at the 1st and 99th percentiles to avoid the results being biased for the presence of
extreme values.
IV. Results
4.1. Descriptive statistics
Table 4 provides descriptive statistics of firm-year observations and KAMs disclosed in
the audit reports of our sample firms for the period 2013-2018. Panel A shows the industry
distribution of our sample firms by year. The most frequently represented industry in our final
sample is ‘manufacturing’ (33.1%), followed by ‘other services (10.2%) and ‘wholesale and retail
trade’ (9.7%). Panel B displays the distribution of audit firms by year. Over 90% of our sample
firms are audited by Big4 firms
8
because the expanded audit report was only required for premium
listed firms in the U.K. PWC audited the largest percent of firms in our sample (29.4%), followed
by KPMG (28.5%), Deloitte (27.1%), and lastly EY (15.0%). Hence, the majority of the KAMs
have been issued by the first three of these Big4 firms. Only 4.7% of firms from our sample have
been audited by a mid-tier company and 1.9% by a small audit firm.
Table 4 Panel C displays the distribution of the number of KAMs per auditor and year. The
mean number of KAMs disclosed across each auditor firm-year is 3.58. However, the mean of
8
This is not a surprising result given our sample consists of Premium listed firms. Firms with this classification in the U.K. are
expected to adhere to the highest standards of reporting and corporate governance, and thus, are most likely to obtain services
from a Big 4 auditor.

23
KAMs from small auditors (2.79) is lower than KAMs from the Big4 (3.59) and mid-tier auditors
(3.69). The number of KAMs disclosed range from 1 to 10 per year. The mean number of KAMs
disclosed decreased over our sample period from 3.92 in 2013 to 3.44 in 2018. As for number of
KAM disclosures by audit firm-year, the highest mean number of disclosed KAMs in 2013 and
2014 were issued by the mid-tier firms (4.36 and 4.48, respectively). Beginning in 2015, the
highest mean number of KAMs per firm-year were issued by the Big 4 (3.61). The smallest audit
firms consistently issued the fewest KAMs.
Table 4 Panel D shows the distribution of KAMs by classification and year. There is a total
of 7,918 KAMs across our 2,214 firm-years. When comparing the levels of KAMs (ENTKAM vs.
ACCKAM), there are generally more account-level (69.6%) than entity-level (30.4%) KAMs
reported in all the years. In the categories among the account-level KAMs, solvency related
(SOLV) KAMs are the most commonly reported (35.2%), followed by liquidity related (LIQU)
KAMs (19.2%) and profitability related (PROF) KAMs (15.2%). With respect to frequency of
individual KAMs, the most common type of KAM reported is related to revenue recognition (REV)
(16.5%). This is not surprising given revenue recognition can involve significant management
estimates and complex contract arrangements; also note revenue recognition is commonly an area
where earnings management is more likely to occur in profit-driven firms (Jansen et al. 2012).
Following revenue recognition, the next two most frequently reported types of KAMs are
intangibles (INTANG), tangibles (PPE) and tax (TAX) at 14.7%, 9.2% and 8.9%, respectively.
Place Table 4 About Here
We hypothesize that the greater the number of KAMs disclosed, the more financially
distressed a firm is (H1). Table 5, Panel A displays the interconnections between FD values and
the number of KAMs. Firms in our sample have from one to 10 KAMs disclosed in their audit

24
reports. Firms with one KAM in their audit report have a mean FD is 10.83, indicating a healthy
firm condition per Altman (1983). As the number of KAMs rise, the level of financial distress rises
as reflected in lower values of FD (i.e., Z’’-Scores).
Furthering our analysis, we divide our sample into two sub-samples: firms reporting the
fewest number of KAMs in our sample (one to two) and firms reporting the highest (6 to 10)
KAMs. Table 5, Panel B includes a t-test of differences in means. The results indicate there are
substantial differences between the two sub-samples, as the mean FDs for the two groups are
significantly different. Essentially, firms disclosing one to two KAMs tend to be less financially
distressed than firms disclosing 6 or more KAMs. Overall, these findings support H1 using
univariate tests.
Place Table 5 About Here
4.2. Correlations, regression models and results
Table 6 reports the Pearson correlations among the dependent variable (FD) and the
independent variables used to test our hypotheses. Generally, the correlations are statistically
significant but relatively low. The dependent variable is correlated with all of the independent
variables used in our models except with the number of KAMs classified as liquidity KAMs
(LIQU) and the change in auditors during the period considered (CHANGEAUDIT). The majority
of the signs are negative due to the inverse relation between financial distress and KAMs.
Place Table 6 About Here
The correlation between FD and No.KAM is -0.259 and significant, consistent with the
expectation that FD falls as No.KAM rises. Additionally, we find negative and significant
correlations for FD and both ENTKAM and ACCKAM, and for FD and both PROF and SOLV.

25
There are several significant correlations in Table 6 between the variables included in the
regression models, but no evidence that multicollinearity is a problem.
9
Table 7 presents the results of estimating four regression models to test our hypotheses.
The models estimate the interconnections between the presence of KAMs in the audit report and
the risk of financial distress. Model 1 represents the baseline model to test H1, with the number of
KAMs as the only independent variable. To test H2, we estimate Models 2 to 4. Model 2 replaces
the number of KAMs with the number of KAMs by level (i.e., entity-wide vs. account-specific
KAMs) as independent variables. Model 3 further divides account-level KAMs into profitability,
solvency, and liquidity KAMs. Finally, Model 4 includes the number of each of the 18 individual
KAMs, from going concern (GC) to pension (PENS). For all models, the goodness of fit, the Prob
> chi2, and the Wald chi-squared test are significant.
Model 1 tests H1. Results show a negative and significant relation between the number of
KAMs and financial distress, indicating that the higher the number of KAMs reported for a firm,
the lower the Z’’-score is, and thus, the greater the level of financial distress risk for a firm. This
finding is consistent with our expectations and is in line with recent studies which show a relation
between the number of KAMs and specific client characteristics, specifically client structural
complexity (Ferreira and Morais 2019; Pinto and Morais 2019) and client leverage and reported
losses (Sierra-Garcia et al. 2019).
Place Table 7 About Here
To test H2a-H2c, we examine the association between KAM type and firm financial
distress in three distinct OLS regression models (see Table 7). The results of Model 2 support H2a
9
A diagnostic test for multicollinearity through the estimation of the variance inflation factor (VIF) coefficients for
all regressions was carried out. The VIF coefficients are always below the threshold of 10 (Kennedy, 2008), suggesting
that multicollinearity does not affect the analyses.

26
as both the number of entity-wide KAMs and the number of account-specific KAMs have a
negative and statistically significant relation with firm Z’’-Score.
Model 3 further classifies account-level KAMs by whether their primary impact is on a
firm’s profitability, liquidity, or solvency. Consistent with H2b, the results show that profitability
and solvency related KAMs have a significant and negative association with firm financial distress
level. Liquidity related KAMs, however, are not significant. The profitability results are consistent
with findings in the financial distress literature as studies examining ratios-based financial distress
prediction models have found measures of profitability to outperform other measures in predicting
the risk of business failure (Altman 1968, 1983; Altman et al. 2017; Lukason and Laitinen 2019).
Model 4 in Table 7 examines whether individual types of KAMs differ in their contribution
to financial distress assessments. This model has the highest R-squared (21.9%), suggesting that
identifying KAMs by nature (individual types) enhances the explanatory power of KAMs over
simply the number of KAMs with respect to assessing financial distress. The results show that of
the 18 individual KAM types we examine, 12 have negative coefficients, eight of which are
negative and significantly related to firm financial distress values (i.e., the more financially
distressed the client firm is likely to be, the lower the Z’’-Score). The KAM variable with the
largest coefficient is GC. This finding supports Gutierrez et al. (2020), among others, who show
going concern opinions provide incremental information in the prediction of corporate defaults.
Results further reveal that KAMs related to exceptional items/presentation and disclosure
(EIPD), revenues (REV) and accruals, deferrals, and management estimates (ACCREST) are also
associated with higher levels of firm financial distress. These findings are intuitive as revenues
tend to be the account most associated with earnings management and accruals are an earnings
management proxy. Also significant are pension and defined benefit plan accounting (PENS) and

27
leases and log-term debt (LLTD). This finding is consistent with Muñoz-Izquierdo et al. (2019)
who show auditor disclosures related to liabilities and contingencies in emphasis of matter and
qualification paragraphs are higher in financially distressed than in non-financially distressed firms
and that a positive and significant relation between business failure and liability and contingency
disclosures exists. These findings also affirm that firms that are generally more leveraged exhibit
greater financial distress than non-financially distressed firms (Altman et al. 2017). The other two
significant individual types of KAMs, merger and acquisition accounting (MA) and intangibles
(INTANG) reflect harder to value elements.
Lastly, we highlight that in all the estimated models, the coefficient on SIZE is negative
and significant. This is consistent with findings documented in the financial distress literature
(Back 2005; Altman et al. 2010; Kim et al. 2008; Cultrera and Brédart 2016).
4.3 Robustness tests
To assess the robustness of our results, we use three alternative measures of financial
distress (FD). These are the Altman’s Z’’-Score dummy variable (FDdum, in Table 8), the
Charitou et al.’s (2004) score (FDCha in Table 9), and the Charitou’s score dummy variable
(FDChadum in Table 10).
10
FDdum is an indicator variable coded as 0 if firm financial distress is present and 1 in the
absence of it. Recall that the Altman’s Z’’-Scores categorizes firms with Z’’-Scores above 2.6 as
being is in a safe zone regarding financial distress (i.e. absent financial distress).We thus codify
our dependent variable FDdum as 1 if a firm’s Z’’-Score is above 2.6 and as 0 if below 2.6
10
Untabulated Kruskal-Wallis independent samples tests show the distribution of No.KAM differs across the two
dependent variable categories (1, 0) for FDdum and for FDChadum.

28
(financial distress is present).
11
Table 8 presents the results of estimating four logistic regression
models to test our hypotheses related to the interconnections between the presence of KAMs in
the audit report and the risk of financial distress measured with a dummy variable. Model 1 shows
a significant relation between the number of KAMs and financial distress, corroborating our
previous findings. Model 1’s area under the curve (untabulated) is 0.831,
12
and it is presumed that
the number of KAMs represents an efficient variable to explain financial distress.
The remaining results in Table 8 are also generally robust with use of the continuous FD
variable in Table 7. The results for model 2 indicate FDdum is significantly and negatively related
to liquidity and solvency, rather than to profitability and solvency; the latter is the case for model
2 in Table 7 where the dependent variable is the continuous variable FD. Consistent with Table 8,
model 3 has the best prediction value, with a pseudo R-squared of 32.40%. Table 8 results for
Model 4 differ from those in Table 7 in that MA, INTANG, and REV are not significant while
restructuring and discontinued operations (RDO) is significant.
Place Table 8 About Here
We also consider another measure of financial distress used specifically in the financial
distress literature for U.K. firms (Charitou et al., 2004). The components of Charitou et al.’s Score
winsorized are defined in Appendix B. Table 9 shows the inferences based on using Charitou et
al.’s continuous variable approach are similar to our Table 7 results for Models 1-3, while for
model 4 MA and LLTD are not significant while internal control fraud (ICFRAUD), RDO, and
investments and related impairment issues (INVEST) are significant.
11
Our dummy variable FDdum is coded as 0 when there is financial distress and 1 when there is no financial distress
to have the same signs in our β coefficients as in the linear regressions. Also, a three-category definition [Z”-Score <
1.1 (coded as 0); 1.1 < Z”-Score < 2.6 (coded as 1); and, Z”-Score > 2.6 (coded as 2)] yields qualitatively similar
results (untabulated).
12
In order to measure the ability to explain financial distress, AUCs (aka, ROCs) are used. The closer the value of an
AUC gets to 1, the more precise its discriminating ability.

29
Place Table 9 About Here
In the same way as we tested the dummy variable of Z’’-Score financial distress index
(FDdum), Table 10 shows the logistic regression models for Charitou et al.’s indicator variable
(FDChadum). Similar to Table 8’s Model 2 results, profitability is not significant while liquidity
is significant. Regarding Model 4, MA, REV, ACCREST, and LLTD are insignificant while
ICFRAUD, INVEST, and tax-related KAMs (TAX) are significant.
Place Table 10 About Here
In summary, across Tables 7-10, No.KAM is always significant in Model 1; ENTKAM
and ACCKAM are each always significant in Model 2; ENTKAM is always significant as is
SOLV, while profitability is significant when the dependent variable is continuous and liquidity
is significant when the dependent variable is an indicator for Model 3. Regarding Model 4,
individual KAMS for GC, EIPD, and PENS are always significant across the four tables, while
ACCREST and INTANG are significant in three of the tables.
4.4 Additional analysis
Having found generally consistent and significant evidence in support of H1 and H2a-H2c,
we examine their predictive power as a way to further highlight the utility of KAMs in the
expanded audit report for users of financial reports. For this reason, we have replicated the same
models but with lag variables. The idea is to show whether KAMs disclosed in period (t) help
predict financial distress (FD) in the subsequent period (t+1).
Table 11 presents our results. The results are generally robust with our previously reported
findings. Specifically, the coefficient on No.KAMs is negative and significant in Model 1. The
coefficients on both ENTKAM and ACCKAM are negative and significant for Model 2. In Model
3, ENTKAM, PROF and SOLV each have negative and significant coefficients. For Model 4, six
individual KAMs- GC, EIPD, PENS, INTANG, REV and MA.- have negative and significant

30
coefficients. These results accentuate the utility of auditor reported KAMs for investors interested
in predicting a firm’s financial distress risk.
Place Table 11 About Here
IV. Conclusions
This study contributes to the auditing literature by investigating the usefulness of new
expanded audit reports KAM disclosures in assessing firm financial distress; atopic relevant and
timely due to the recent enactment of expanded auditor report regulation in the U.K., U.S., and
E.U.. This study sheds light on whether enhanced transparency into the audit process from auditor
disclosures of key (critical) audit matters are useful in assessing the risk of financial distress
present at a firm.
To examine the explanatory power of KAMs in the assessment of financial distress risk,
we test the relation between the number, level, category, and individual type of KAMs disclosed
on a firm’s audit report and the firm’s financial distress level, measured using Altman’s Z’’-score
and three alternate financial distress proxies to test the robustness of our main results. In exploring
the usefulness of KAM disclosures, in addition to considering the total number of KAMs disclosed,
we separately examine the extent to which a KAM’s (1) risk level (entity-wide vs. account level),
(2) primary impact on a firm’s financial health (profitability, liquidity, and solvency), and (3)
individual type (e.g. revenue recognition, going concern, etc.) can be used to assess firm financial
distress.
To test our hypotheses, we identify Premium listed firms on the London Stock Exchange
starting in 2013 and hand collect KAM disclosures and obtain financial information for these firms
through 2018 (resulting in 2,214 firm-year observations). Results from our analyses show that
audit report KAM disclosures are useful in assessing firm financial distress risk. Specifically, we
find that the greater the number of KAMs disclosed by an auditor for a firm, the greater the

31
contemporaneous level of financial distress of the client firm; this also holds for separating KAMs
into entity and account categories and into profitability and solvency categories. Moreover, we
find that the usefulness of KAMs increases when considering the individual type of KAMs
disclosed by an auditor for a particular firm. Results show that going concern, exceptional
items/presentation and disclosure, and pensions are consistently important individual KAMs in
assessing financial distress. Accruals and estimates and intangibles are also frequently important.
Finally, we demonstrate that KAMs have predictive ability for next year’s level of financial
distress, with results being quite similar to results from our contemporaneous analyses.
Findings from this study have important implications for regulators, financial statement
users, and auditors alike. By showing that KAMs are informative in the assessment of firm
financial distress, our results aid regulators in the cost-benefit assessment of new expanded auditor
reporting regulation, and specifically, of the requirement that auditors disclose key (critical) audit
matter in their audit opinions. Results are also useful to financial statement users who may use
KAM disclosures in assessing the extent of financial distress at a firm as they make investment
decisions and credit. Lastly, this study should be useful to auditors in (1) their assessments of
engagement risk for prospective clients and (2) as they consider audit report disclosure alternatives
for signalling firm going concern risk to financial statement users.

32
References
Abdelfattah, T., Elmahgoub, M., & Elamer, A. A. (2020). Female audit partners and extended
audit reporting: UK evidence. Journal of Business Ethics, 1-21.
AICPA (1989) Reports on Audited Financial Statements.
Allen, R. D., Hermanson, D. R., Kozloski, T. M., & Ramsay, R. J. (2006). Auditor risk assessment:
Insights from the academic literature. Accounting Horizons, 20(2), 157-177.
Altman, E. I. (1968). Financial ratios, discriminant analysis and the prediction of corporate
bankruptcy. The Journal of Finance, 23(4), 589-609.
Altman, E. I. (1983). Corporate Financial Distress. A Complete Guide to Predicting, Avoiding,
and Dealing with Bankruptcy. Wiley Interscience, John Wiley and Sons, New York, USA.
Altman, E.I., & McGough, T.P. (1974). Evaluation of a company as a going-concern. Journal of
Accountancy, 138, 50–57.
Altman, E. I., Sabato, G. & Wilson, N. (2010). The value of non-financial information in small
and medium-sized enterprise risk management. Journal of Credit Risk, 2(6), 95-127.
Altman, E. I., Iwanicz‐Drozdowska, M., Laitinen, E. K., & Suvas, A. (2017). Financial distress
prediction in an international context: A review and empirical analysis of Altman's Z‐score
model. Journal of International Financial Management & Accounting, 28(2), 131-171.
Andreicovici, I., Jeny, A., & Lui, D. (2020). Do Firms Respond to Auditors’ Red Flags? Evidence
from the Expanded Audit Report. Evidence from the Expanded Audit Report (June 24,
2020). Available at SSRN; https://ssrn.com/abstract= 3634479.
Back, P. (2005). Explaining financial difficulties based on previous payment behavior,
management background variables and financial ratios. European Accounting Review,
14(4), 839-868.
Balcaen, S., & Ooghe, H. (2006). 35 years of studies on business failure: an overview of the classic
statistical methodologies and their related problems. The British Accounting Review, 38(1),
63-93.
Barnes, P. (2004). The auditor’s going concern decision and types I and II errors: The Coase
Theorem, transaction costs, bargaining power and attempts to mislead. Journal of
Accounting and Public Policy, 33(6), 415-440.
Basioudis, I. G. (2008). The introduction of ‘surprise tests’ in teaching accounting. Accounting
Education: an International Journal, 17(2), 205-208.
Bédard, J.; Carl Brousseau, and Ann Vanstraelen (2019a) Investor Reaction to Auditors' Going
Concern Emphasis of Matter: Evidence from a Natural Experiment. Auditing: A Journal of
Practice & Theory, 38, (2), 27-55.
Bédard, J., Gonthier-Besacier, N. & Schatt, A. (2019b) Consequences of Expanded Audit Reports:
Evidence from the Justifications of Assessments in France. Auditing: A Journal of Practice
& Theory, 38(3), 23-45.
Bellovary, J.L., Giacomino, D.E. & Akers, M.D. (2007). A review of bankruptcy prediction
studies: 1930 to present. Journal of Financial Education, 33, 1–42.
Boolaky, P. K., & Quick, R. (2016). Bank directors’ perceptions of expanded auditor's reports.
International Journal of Auditing, 20(2), 158-174.

33
Bortolon, P. M., Sarlo Neto, A., & Santos, T. B. (2013). Custos de auditoria e governança
corporativa. Revista Contabilidade & Finanças, 24(61), 27-36.
Burke, J., Hoitash, R., Hoitash, U., & Xiao, X. (2020). An Investigation of US Critical Audit
Matter Disclosures. Available at SSRN; https://ssrn.com/abstract=3635477.
Carson, E., Fargher, N. L., Geiger, M. A., Lennox, C. S., Raghunandan, K., & Willekens, M.
(2012). Audit reporting for going-concern uncertainty: A research synthesis. Auditing: A
Journal of Practice & Theory, 32(sp1), 353-384.
Carver, B. T., & Trinkle, B. S. (2017). Nonprofessional Investors’ Reactions to the PCAOB's
Proposed Changes to the Standard Audit Report. Available at SSRN:
https://ssrn.com/abstract=2930375
Casterella, J. R., Desir, R., Stallings, M. A., & Wainberg, J. S. (2020). Information Transfer of
Bankruptcy Announcements: Examining the Impact of Auditor Opinions. Accounting
Horizons, 34(1), 45-66.
Cenciarelli, V. G., Greco, G., & Allegrini, M. (2018). External audit and bankruptcy prediction.
Journal of Management and Governance, 22(4), 863-890.
Charitou, A., Neophytou, E., & Charalambous, C. (2004). Predicting corporate failure: empirical
evidence for the UK. European Accounting Review, 13(3), 465-497.
Choi, J. H., Ha, W., & Lee, H. (2018). Do Banks Care About Litigation Against Auditors of
Borrower Firms? Evidence from Bank Loan Pricing. Evidence from Bank Loan Pricing
(May 1, 2018).
https://pcaobus.org/EconomicAndRiskAnalysis/Conference/Documents/Do-Banks-Care-
About-Litigation-Lee-Ha-Choi.pdf
Christensen, B. E., Glover, S.M. and Wolfe, C.J. (2014) Do Critical Audit Matter Paragraphs in
the Audit Report Change Nonprofessional Investors' Decision to Invest?. Auditing: A
Journal of Practice & Theory, 33(4), 71-93.
Cipriano, M., Hamilton, E. L., & Vandervelde, S. D. (2017). Has the lack of use of the qualified
audit opinion turned it into the “Rotten Kid” threat? Critical Perspectives on Accounting,
47, 26-38.
Cultrera, L., & Brédart, X. (2016). Bankruptcy prediction: the case of Belgian SMEs. Review of
Accounting and Finance, 15(1), 101-119.
DeAngelo, L. E. (1981). Auditor size and audit quality. Journal of Accounting and Economics,
3(3), 183-199.
DeFond, M., & Zhang, J. (2014). A review of archival auditing research. Journal of Accounting
and Economics, 58(2-3), 275-326.
Dye, R. A. (1993). Auditing standards, legal liability, and auditor wealth. Journal of Political
Economy, 101(5), 887-914.
Feldmann, D. A., & Read, W. J. (2010). Auditor conservatism after Enron. Auditing: A Journal of
Practice & Theory, 29(1), 267-278.
Ferreira, C., & Morais, A. I. (2019). Analysis of the relationship between company characteristics
and key audit matters disclosed. Revista Contabilidade & Finanças, (AHEAD).
https://doi.org/10.1590/1808-057x201909040
Francis, J.R., & Wang, D.(2008). The joint effect of investor protection and Big 4 audits on
earnings quality around the world. Contemporary Accounting Research, 25 (1), 157–191.

34
FRC Financial Reporting Council (2013). International Standard on Auditing (UK and Ireland)
700. The independent auditor’s report on financial statements. London, England.
https://www.frc.org.uk/getattachment/501de004-b616-43c3-8d65-aeaebde19f8d/ISA-
700-(UK-and-Ireland)-700-(Revised)-Independent-auditors-report-June-2013.pdf
Geiger, M. A., Raghunandan, K., & Rama, D. V. (2005). Recent changes in the association
between bankruptcies and prior audit opinions. Auditing: A Journal of Practice & Theory,
24(1), 21-35.
Geiger, M. A., & Rama, D. V. (2006). Audit firm size and going-concern reporting accuracy.
Accounting Horizons, 20(1), 1-17.
Geiger, M.A.; Raghunandan, K & Riccardi, W. (2014). The global financial crisis: US
bankruptcies and going concern audit opinions. Accounting Horizons, 28(1), 59-75.
Gimbar, C., Hansen, B., & Ozlanski, M. E. (2016). Early evidence on the effects of critical audit
matters on auditor liability. Current Issues in Auditing, 10(1), 24-33.
Goh, B. W., Li, D., & Wang, M. (2019). Informativeness of the Expanded Audit Report: Evidence
from China. Available at SSRN; https://ssrn.com/abstract=3432107.
Gutierrez, E., Minutti-Meza, M., Tatum, K. W., & Vulcheva, M. (2018). Consequences of
adopting an expanded auditor’s report in the United Kingdom. Review of Accounting
Studies, 23(4), 1543-1587.
Gutierrez, E., Krupa, J., Minutti-Meza, M., & Vulcheva, M. (2020). Do going concern opinions
provide incremental information to predict corporate defaults? Review of Accounting
Studies, 1-38. https://doi.org/10.1007/s11142-020-09544-x
Hernández-Tinoco, M. & Wilson, N. (2013). Financial distress and bankruptcy prediction among
listed companies using accounting, market and macroeconomic variables. International
Review of Financial Analysis, 30, 394-419.
Hillegeist, S. A., Keating, E. K., Cram, D. P., & Lundstedt, K. G. (2004). Assessing the probability
of bankruptcy. Review of Accounting Studies, 9(1), 5-34.
IAS 701. Communicating key audit matters in the independent auditor's report
IAASB (2015). Reporting on audited financial statements – New and revised auditor reporting
standards and related conforming amendments.
IAASB (2016)?
Jansen, I. P., Ramnath, S., & Yohn, T. L. (2012). A diagnostic for earnings management using
changes in asset turnover and profit margin. Contemporary Accounting Research, 29(1),
221-251.
Jayasekera, R. (2018). Prediction of company failure: Past, present and promising directions for
the future. International Review of Financial Analysis, 55, 196-208.
Jermakowicz, E. K., Chen, C. D., & Donker, H. (2018). Financial statement effects of adopting
IFRS: the Canadian experience. International Journal of Accounting & Information
Management, 26(4), 466-491.
Kennedy, P. (2008). A Guide to Econometrics. Wiley-Blackwell, Hoboken, NJ.
Kim, M., Kim, M. & McNiel, R. D. (2008). Predicting survival prospect of corporate restructuring
in Korea. Applied Economics Letters, 15(15), 1187-1190.

35
Köhler, A., Ratzinger-Sakel, N. V., & Theis, J. (2020). The effects of key audit matters on the
auditor’s report’s communicative value: Experimental evidence from investment
professionals and non-professional investors. Accounting in Europe, 17(2), 105-128.
Kochetova-Kozloski, N., Kozloski, T. M., & Messier Jr, W. F. (2013). Auditor business process
analysis and linkages among auditor risk judgments. Auditing: A Journal of Practice &
Theory, 32(3), 123-139.
Laitinen, E.K. & Laitinen, T. (2009). Effect of accruals on financial, non-financial, and audit
information in payment default prediction. International Journal of Accounting, Auditing
and Performance Evaluation 5, 353-83.
Lennox, C. S. (1999). The accuracy and incremental information content of audit reports in
predicting bankruptcy. Journal of Business Finance & Accounting, 26(5‐6), 757-778.
Lennox, C. S., Schmidt, J. J., & Thompson, A. (2019). Is the expanded model of audit reporting
informative to investors? Evidence from the UK. Available at SSRN:
https://ssrn.com/abstract=2619785
Liao, L., Minutti-Meza, M., Zhang, Y., & Zou, Y. (2019). Consequences of the Adoption of the
Expanded Auditor’s Report: Evidence from Hong Kong. Available at SSRN 3392449.
Loew, E., & Mollenhauer, T. (2019). An Empirical Analysis of Key Audit Matters in the Financial
Industry. European Banking Institute Working Paper Series 2019 – no. 40. Available at
SSRN: https://ssrn.com/abstract=3379324
Lukason, O., & Laitinen, E. K. (2019). Firm failure processes and components of failure risk: An
analysis of European bankrupt firms. Journal of Business Research, 98, 380-390.
Mai, F., Tian, S., Lee, C., & Ma, L. (2019). Deep learning models for bankruptcy prediction using
textual disclosures. European Journal of Operational Research, 274(2), 743-758.
McKeown, J. C., Mutchler, J. F., & Hopwood, W. (1991). Towards an explanation of auditor
failure to modify the audit opinions of bankrupt companies. Auditing: A Journal of Practice
& Theory, 10, 1-13.
Menon, K., & Williams, D. D. (2010). Investor reaction to going concern audit reports. The
Accounting Review, 85(6), 2075-2105.
Moroney, R., Phang, S. Y., & Xiao, X. (2020). When Do Investors Value Key Audit Matters?
European Accounting Review, 1-20. DOI: 10.1080/09638180.2020.1733040
Munsif, V., Raghunandan, K., Rama, D. V., & Singhvi, M. (2011). Audit fees after remediation
of internal control weaknesses. Accounting Horizons, 25(1), 87-105.
Muñoz-Izquierdo, N., Camacho‐Miñano, M. M., & Pascual‐Ezama, D. (2017). The content of the
audit report in the year prior to bankruptcy filing. Empirical evidence from Spain. Spanish
Journal of Finance and Accounting, 46(1), 92-126.
Muñoz-Izquierdo, N., Camacho-Miñano, M. M., Segovia-Vargas, M. J., & Pascual-Ezama, D.
(2019). Is the external audit report useful for bankruptcy prediction? Evidence using
artificial intelligence. International Journal of Financial Studies, 7(2), 20.
Muñoz‐Izquierdo, N., Laitinen, E. K., Camacho-Miñano, M.M. & Pascual‐Ezama, D. (2020).
Does audit report information improve financial distress prediction over Altman's
traditional Z‐Score model? Journal of International Financial Management & Accounting,
31(1), 65–97.

36
Myers, L. A., Shipman, J. E., Swanquist, Q. T., & Whited, R. L. (2018). Measuring the market
response to going concern modifications: The importance of disclosure timing. Review of
Accounting Studies, 23(4), 1512-1542.
Nelson, M. W. (2009). A model and literature review of professional scepticism in auditing.
Auditing: A Journal of Practice & Theory, 28(2), 1-34.
Pincus, M., Tian, F., Wellmeyer, P., & Xu, S. X. (2017). Do clients’ enterprise systems affect audit
quality and efficiency? Contemporary Accounting Research, 34(4), 1975-2021.
Pinto, I., & Morais, A. I. (2019). What matters in disclosures of key audit matters: Evidence from
Europe. Journal of International Financial Management & Accounting, 30(2), 145-162.
Piñeiro-Sánchez, C., de Llano-Monelos, P. & Rodríguez-López, M. (2012). Evaluation of the
likelihood of financial failure. Empirical contrast of the informational content audit of
accounts. Spanish Journal of Finance and Accounting, 46(156), 565-587.
Piñeiro-Sánchez, C., De Llano-Monelos, P. & Rodríguez-López, M. (2013). A parsimonious
model to forecast financial distress, based on audit evidence. Contaduría y Administración,
58(4), 151-173.
Porumb, V. A., Karaibrahimoglu, Y. Z., Lobo, G. J., Hooghiemstra, R., & De Waard, D. (2019).
Is More Always Better? Disclosures in the Expanded Audit Report and their Impact on
Loan Contracting. Available at SSRN: https://ssrn.com/abstract=3216492.
Public Company Accounting Oversight Board (PCAOB) (2013). Proposed Auditing Standards—
The Auditor’s Report on an Audit of Financial Statements When the Auditor Expresses an
Unqualified Opinion. PCAOB Release No. 2013-005. Washington, DC: PCAOB.
Available at: https://pcaobus.org/Rulemaking/Docket034/Release_2013-005_ARM.pdf
Public Company Accounting Oversight Board (PCAOB) (2016). Proposed Auditing Standards—
The Auditor’s Report on an Audit of Financial Statements When the Auditor Expresses an
Unqualified Opinion. PCAOB Release No. 2016-003.
Public Company Accounting Oversight Board (PCAOB) (2017). Annual Report on the Interim
Inspection Program Related to Audits of Brokers and Dealers (PCAOB Release No. 2018-
003 – August 20, 2018) https://pcaobus.org/Inspections/Documents/Broker-Dealer-
Auditor-Inspection-Annual-Report-2018.pdf
Read, W. J., & Yezegel, A. (2018). Going-concern opinion decisions on bankrupt clients: Evidence
of long-lasting auditor conservatism? Advances in Accounting, 40, 20-26.
Reynolds, J. K., & Francis, J. R. (2000). Does size matter? The influence of large clients on office-
level auditor reporting decisions. Journal of Accounting and Economics, 30(3), 375-400.
Serrano-Cinca, C., Gutiérrez-Nieto, B., & Bernate-Valbuena, M. (2019). The use of accounting
anomalies indicators to predict business failure. European Management Journal, 37(3),
353-375.
Sierra-García, L., Gambetta, N., García-Benau, M. A., & Orta-Pérez, M. (2019). Understanding
the determinants of the magnitude of entity-level risk and account-level risk key audit
matters: The case of the United Kingdom. The British Accounting Review, 51, 227-240
Smith, K. (2019). Tell me more: A content analysis of expanded auditor reporting in the United
Kingdom. Available at SSRN 2821399. Retrieved from
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2821399

37
Trpeska, M., Atanasovski, A., & Lazarevska, Z. B. (2017). The relevance of financial information
and contents of the new audit report for lending decisions of commercial banks. Accounting
and Management Information Systems, 16(4), 455-471.
Yang, L., Brink, A. G., & Wier, B. (2018). The impact of emotional intelligence on auditor
judgment. International Journal of Auditing, 22(1), 83-97.

38
Appendix A
Extract from the expanded auditor’s report issued to John Wood Group PLC by KPMG,
LLP in the annual accounts of 2018
The key audit matters (KAMs) addressed in this expanded auditor’s report were the following:
1. The impact of uncertainties due to the UK exiting the European Union on our audit
2. Revenue recognition on fixed price contracts
3. Goodwill impairment
4. Litigation, investigations and contingent liabilities
5. Uncertain tax positions
6. U.S. asbestos related claims provision
7. Gross defined benefit pension liability
8. Amec Foster Wheeler Plc acquisition measurement period adjustments
9. Parent Company risk: Recoverability of parent company’s investment in subsidiaries
On our 18-item classification of KAMs, the previous KAMs are included in the following
categories (using a one-to-one match from above):
1. Litigation, macroeconomic, and system implementation
2. Revenue recognition
3. Intangibles and related impairment issues (included research and development)
4. Litigation, macroeconomic and system implementation
5. Tax
6. Accruals, deferrals and management estimates
7. Pension and defined benefit plan accounting
8. Mergers and acquisitions (M&A)
9. Investments and related impairment issues

39
Appendix B
Variable definitions
Dependent variables:
FD
Continuous dependent variable used for main results:
Financial distress (FD) =
= Altman’s Z’’- Score for non-manufacturers and listed companies = Z’’ =
= 3.25 + 6.56×Z1 + 3.26×Z2 + 6.72×Z3 + 1.05×Z4
• Z1
Z1 = (CA-CL) / TA
First ratio of Altman’s Z’’-Score: Z1 = (Current assets - Current
liabilities) / Total assets
• Z2
Z2 = RE / TA
Second ratio of Altman’s Z’’-Score: Z2 = Retained earnings / Total
assets
• Z3
Z3 = EBIT / TA
Third ratio of Altman Z''-Score: Earnings before interest and taxes /
Total assets
• Z4
Z4 = BV of equity / TL
Fourth ratio of Altman Z''-Score: Book value of equity / Total
liabilities
FDdum
Dummy dependent variable used for robustness:
Financial distress dummy (FDdum) =
= Altman’s Z’’-Score for non-manufacturers and listed companies categorized
into 2 categories:
Z’’ > 2.6 (Safe zone) = value of 1
Z’’< 2.6 (grey and distressed zone) = value of 0
FDCha
Continuous dependent variable used for additional analysis:
Financial distress Charitou (FDCha) =
= Charitou et al. (2004) Score for U.K. firms =
= 1 / 1 + EXP{-7.1786 + [12.3826×(TL/TA)] - [20.9691×(NI/TL)] –
[3.0174×(OPCF/TL)]}
FDChadum
Financial distress Charitou dummy (FDChadum) = Charitou’s index categorized
into 2 categories:
Score > 0.2 (Non-distressed) = value of 1
Score < 0.2 (Distressed) = value of 0

40
Independent variables and sample descriptives:
SIZE
Logarithm of total assets
TA
Total assets in thousands of dollars
NI
Profit or loss for the period (net income) in thousands of dollars
CA
Current assets in thousands of dollars
CL
Current liabilities in thousands of dollars
EBIT
Earnings before interest and taxes in thousands of dollars
SE
Shareholders’ funds in thousands of dollars
TL
Total liabilities in thousands of dollars
OPCF
Net cash from operating activities in thousands of dollars
No.KAM
Number of KAMs disclosed by an auditor
ENTKAM
Entity level KAMs
ACCKAM
Accounting level KAMs
PROF
Profitability KAMs (includes MGFEES, REV and EXP)
SOLV
Solvency KAMs (includes ACCREST, INV and CASHREC)
LIQU
Liquidity KAMs (includes INVEST, INTANG, PPE, LLTD and PENS)
BIG4
Audit firm dummy variable (1 if Big 4 auditors, 0 if Non-Big 4)
CHANGEAUDIT
Change in audit firm from the prior year (1 if audit firm has changed, 0
otherwise)
See Table 3 for explanation of KAMs included under PROF, SOLV and LIQU variables.

41
Table 1. Sample selection criteria
Panel A. Firms and firm-year observations of the sample
Initial sample: All firms listed in London Stock Exchange in 2013
2,149
(-) Firms not listed as premium
(1,250)
All premium firms listed in London Stock Exchange in 2013
899
(-) Financial firms
(-) Investment trusts
(360)
(51)
All premium firms listed in London Stock Exchange in 2013, excluding financial
firms and investment trusts
488
(-) Firms with no consolidated annual report data available in ORBIS database
(6)
Total firms of the sample: Firms with consolidated annual report data in ORBIS
database for all or some years of the sample (2013-2018)
482
Total firm-year observations
2,214
The table reports the sample selection criteria, starting with the initial sample, which consists of all listed companies
in the London Stock Exchange (LSE) in 2013, extracted from the LSE website. The first filter is all premium-listed
firms (a total of 899 companies), as expanded audit report regulation in the UK applies only to premium listed
companies beginning with fiscal year-ends on or after September 2013. We then search for the industry and
consolidated annual report data in ORBIS, and eliminate 411 financial companies and investment trusts,
13
and 6 firms
without any consolidated data in the period studied (2013-2018).
14
The final sample comprises 482 companies, which
are all non-financial premium listed companies on the LSE reporting extended audit report and with consolidated
annual report data in ORBIS database for all or some of the years of the sample (2013-2018). The total firm-year
observations is 2,214. A total of 678 firm-year observations are eliminated because companies publish the short format
of the audit report or ORBIS database does not provide sufficient data to implement our empirical analyses and the
annual report is not available on the internet. The unavailability of the annual report is due to the fact the company is
being liquidated or has being acquired.
13
In ORBIS database, financial companies and investment trusts are classified under the following industry codes: “K
– Financial” and “S – Other services: Banking”, respectively.
14
The consolidated annual report data is coded in ORBIS database as “C1” (consolidated accounts with no
unconsolidated companion) and “C2” (consolidated accounts with an unconsolidated companion). Companies without
any data included under these codes are excluded from our sample.

42
Table 2. Sample descriptive statistics
Variable
No.
Obs.
Mean
Median
Std. Dev.
Min.
Max.
FD
2,214
7.41
6.64
4.34
-0.46
33.97
Z1
2,214
0.11
0.07
0.19
-0.31
0.73
Z2
2,214
0.39
0.41
0.26
-0.59
0.88
Z3
2,214
0.07
0.07
0.10
-0.37
0.40
Z4
2,214
1.58
0.87
2.95
-0.24
23.55
TA
2,214
6,954,393
1,329,097
19,200,000
9,715.4
136,000,000
NI
2,214
285,894.5
52,432.5
912,653.7
-1,175,000
6,259,487
CA
2,214
1,819,755
346,902.3
4,501,999
2734
35,900,000
CL
2,214
1,726,354
239,087.9
5,319,040
164
38,966,414
EBIT
2,214
424,014.3
80,750.1
1,214,790
-786,000
8,274,000
SE
2,214
2,678,590
526,142.3
7,946,536
-158,191.8
60,700,000
TL
2,214
4,329,942
635,202.4
12,399,000
1543.5
83,300,000
OPCF
2,214
613,117.5
89,257.2
1,845,919
-76,641.5
13,200,000
The table presents summary statistics for the financials of our sample: number of observations, mean, standard
deviation, minimum and maximum. See Appendix B for variable definitions.

Table 3. Classification of key audit matters (KAMs)
Classification
Type
(Variable)
Description of categorical variables
A.
Entity-level KAMs (ENTKAM)
1.
Going concern
GC
Number of going concern KAMs disclosed
2.
Internal control and fraud
ICFRAUD
Number of internal control and fraud KAMs disclosed
3.
Restructuring and discontinued operations
RDO
Number of restructuring and discontinued operations KAMs disclosed
4.
Merger and acquisition (M&A) accounting
MA
Number of merger and acquisition (M&A) accounting KAMs
disclosed
5.
Tax-related
TAX
Number of tax-related KAMs disclosed
6.
Exceptional items and presentation and disclosure
EIPD
Number of exceptional items, presentation and disclosure KAMs
disclosed
7.
Litigation, macroeconomic and system implementation
LITMACRO
Number of litigation, macroeconomic and system implementation
KAMs disclosed
B.
Account-level KAMs (ACCKAM) Category
8.
Management and/or performance fees
Profitability
(PROF)
MGFEES
Number of management and/or performance fees KAMs disclosed
9.
Revenue recognition
REV
Number of revenue recognition KAMs disclosed
10.
Expenses recognition
EXP
Number of expense recognition KAMs disclosed
11.
Accruals, deferrals and management
estimates
Liquidity
(LIQU)
ACCREST
Number of accruals, deferrals and management estimates KAMs
disclosed
12.
Inventory
INV
Number of inventory KAMs disclosed
13.
Cash and receivables
CASHREC
Number of cash and receivables KAMs disclosed
14.
Investments and related impairment
issues
Solvency
(SOLV)
INVEST
Number of investments and related impairment issues KAMs
disclosed
15.
Intangibles and related impairment
issues
INTANG
Number of intangibles and related impairment issues KAMs disclosed
16.
Property, plant and equipment and
related impairment issues
PPE
Number of property, plant and equipment and related impairment
issues KAMs disclosed
17.
Leases and long-term debt
LLTD
Number of leases and long-term debt KAMs disclosed
18.
Pension and defined benefit plan
accounting
PENS
Number of pension and defined benefit plan accounting KAMs
disclosed
The table reports the variables that represent the 18-item codification of key audit matters (KAMs), segregated into two sections: entity level KAMs (ENTKAM)
and accounting level KAMs (ACCKAM). The table shows the section in the first column. The following columns present the item number and its name, abbreviated
name and the variable definition. In the accounting level KAMs section, KAMs are additionally classified depending on their impact on the profitability, liquidity
or solvency of the company.

Table 4. Descriptive statistics of the sample
Panel A. Industry distribution of firm-year observations
Industry
2013
2014
2015
2016
2017
2018
Totals
%
Agriculture, forestry, fishing and
mining
32
37
32
30
30
30
191
8.6%
Information and communication
24
39
38
35
33
31
200
9.0%
Manufacturing
85
143
137
128
125
114
732
33.1%
Other services
25
47
43
39
36
36
225
10.2%
Professional, scientific and technical
activities
18
28
26
22
21
20
135
6.1%
Real estate
15
34
34
32
32
32
179
8.1%
Transportation and accommodation
18
32
32
32
29
29
172
7.8%
Utilities and construction
18
32
31
31
26
26
164
7.4%
Wholesale and retail trade
16
44
42
41
37
36
216
9.7%
Total
251
436
415
390
369
354
2,214
100%
Table 4, Panel A reports the industry distribution of firm-year observations in absolute figures. Industry information
comes from ORBIS database and the nine industry categories are based on NACE codes, following Lennox et al.
(2019). Percentages over total sample firms are reported in the last column.

45
Panel B. Auditor distribution of sample firms by year
2013
2014
2015
2016
2017
2018
Totals
%
Deloitte
71
109
101
99
93
88
561
27.1%
EY
41
63
56
51
52
48
311
15.0%
KPMG
62
122
117
106
92
90
589
28.5%
PWC
62
110
115
111
107
102
607
29.4%
Big 4
236
404
389
367
344
328
2,068
100%
94.0%
92.9%
93.7%
94.1%
93.2%
92.7%
93.4%
BDO
7
14
10
10
10
10
61
59.2%
Crowe Clark Whitehill
0
0
0
0
1
0
1
1.0%
Grant Thornton
4
7
7
6
7
6
37
35.9%
RSM
0
0
1
1
1
1
4
3.9%
Mid-tier
11
21
18
17
19
17
103
100%
4.4%
4.8%
4.4%
4.4%
5.2%
4.8%
4.7%
Baker Tilly
2
2
0
0
0
0
4
9.3%
BSG Valentine
0
1
1
1
1
1
5
11.6%
Carter Backer
0
1
1
1
1
1
5
11.6%
Chantrey Vellacott
0
3
1
0
0
0
4
9.3%
French Duncan
0
1
1
1
1
1
5
11.6%
Larking Gowen Limited
1
1
1
0
0
0
3
7.0%
Mazars
1
1
1
1
1
2
7
16.3%
Moore Stephens
0
0
2
2
2
2
8
18.7%
Scrutton Bland
0
0
0
0
0
1
1
2.3%
UHY Hacker Young
0
0
0
0
0
1
1
2.3%
Small
4
10
8
6
6
9
43
100%
1.6%
2.3%
1.9%
1.5%
1.6%
2.5%
1.9%
Total
251
435
415
390
369
354
2,214
100%
Table 4, Panel B presents the auditor distribution of sample firms by year in absolute figures. Auditors are classified
by size into Big 4 audit firms, mid-tier auditors, and small audit firms. Percentages over the total sample (row) and
over auditor size (last column) are reported.

46
Panel C. Mean number of KAMs issued per auditor by year
2013
2014
2015
2016
2017
2018
Means
Deloitte
4.06
3.98
3.84
3.61
3.28
3.39
3.69
EY
4.07
4.17
3.82
3.45
3.56
3.69
3.80
KPMG
3.19
2.93
2.89
2.88
3.03
3.56
3.05
PWC
4.39
4.38
4.05
3.71
3.64
3.40
3.90
Mean No. of KAMs by Big 4
3.92
3.80
3.61
3.41
3.37
3.48
3.59
BDO
4.14
4.00
3.50
3.40
3.30
3.30
3.61
Crowe Clark Whitehill
3.00
3.00
Grant Thornton
4.75
5.43
3.71
3.67
3.57
3.50
4.08
RSM
3.00
1.00
1.00
1.00
1.50
Mean No. of KAMs by Mid-tier
4.36
4.48
3.56
3.35
3.26
3.24
3.69
Baker Tilly
2.50
3.00
2.75
BSG Valentine
3.00
3.00
3.00
3.00
2.00
2.80
Carter Backer
1.00
1.00
1.00
3.00
3.00
1.80
Chantrey Vellacott
3.67
3.00
3.50
French Duncan
4.00
4.00
3.00
3.00
3.00
3.40
Larking Gowen Limited
4.00
4.00
4.00
4.00
Mazars
3.00
4.00
3.00
3.00
2.00
1.50
2.57
Moore Stephens
3.00
3.00
2.00
2.00
2.50
Scrutton Bland
1.00
1.00
UHY Hacker Young
4.00
4.00
Mean No. of KAMs by small
auditor
3.00
3.30
3.00
2.67
2.50
2.22
2.79
Mean N. of KAMs by year
3.92
3.83
3.60
3.39
3.35
3.44
3.58
Median N. of KAMs by year
3.00
3.00
3.00
3.00
3.00
3.00
3.00
Std. Dev. N. of KAMs by year
1.40
1.50
1.44
1.43
1.55
1.53
1.49
Minimum N. of KAMs by year
1.00
1.00
1.00
1.00
1.00
1.00
1.00
Maximum N. of KAMs by year
10.00
10.00
8.00
9.00
9.00
9.00
10.00
Table 4, Panel C shows the mean number of KAMs issued per auditor by year. Auditors are partitioned by size into
Big 4 audit firms, mid-tier auditors, and small audit firms.

47
Panel D. Distribution of KAMs by level, category, and nature by year
Classification of KAMs
2013
2014
2015
2016
2017
2018
Totals
%
ENTKAM
1.
GC
28
46
45
30
13
27
189
2.4%
2.
ICFRAUD
78
82
28
20
22
18
248
3.1%
3.
RDO
11
28
21
19
10
7
96
1.2%
4.
MA
47
96
115
96
79
92
525
6.6%
5.
TAX
116
149
135
111
111
92
704
8.9%
6.
EIPD
38
85
67
65
69
63
387
4.9%
7.
LITMACRO
24
47
41
43
41
61
257
3.3%
Subtotal ENTKAM
332
533
452
384
345
360
2,406
30.4%
33.7%
32.0%
30.3%
29.0%
27.9%
29.6%
30.4%
ACC
KAM
PROF
8.
MGFEES
0
2
1
0
0
2
5
0.1%
9.
REV
167
275
247
223
203
191
1,306
16.5%
10.
EXP
17
38
40
40
38
34
207
2.6%
LIQU
11.
ACCREST
67
112
89
92
87
76
523
6.6%
12.
INV
49
101
97
84
70
68
469
5.9%
13.
CASHREC
26
53
42
33
34
27
215
2.7%
SOLV
14.
INVEST
37
60
48
35
64
93
337
4.3%
15.
INTANG
142
233
230
208
183
166
1,162
14.7%
16.
PPE
88
144
141
123
124
110
730
9.2%
17.
LLTD
14
19
19
18
11
18
99
1.3%
18.
PENS
46
94
88
83
76
72
459
5.8%
Subtotal ACCKAM
653
1,131
1,042
939
890
857
5,512
69.6%
66.3%
68.0%
69.7%
71.0%
72.1%
70.4%
69.6%
Total KAMs
985
1,664
1,494
1,323
1,235
1,217
7,918
12.4%
21.0%
18.9%
16.7%
15.6%
15.4%
100%
PROF
184
315
288
263
241
227
1,518
19.2%
28.2%
27.9%
27.6%
28.0%
27.1%
26.5%
27.5%
LIQU
142
266
228
209
191
171
1,207
15.2%
21.7%
23.5%
21.9%
22.3%
21.4%
20.0%
21.9%
SOLV
327
550
526
467
458
459
2,787
35.2%
50.1%
48.6%
50.5%
49.7%
51.5%
53.5%
50.6%
Table 4, Panel D reports the distribution of KAMs by level, category, and nature per year. KAMs are first divided by
level (entity- and account-level KAMs). Then, account-level KAMs are also segregated by category (KAMs disclosing
issues related to profitability, liquidity, or solvency). Also, both entity- and account-level KAMs are split into
individual KAMs according to the nature of the disclosure. See Table 3 for KAM definitions and Appendix B for
other variable definitions.

48
Table 5. Descriptive statistics and univariate analysis
of financial distress (FD) and number of KAMs
Panel A. Descriptive statistics
No. KAM
Firm-year
obs.
Mean
FD
Median
FD
Standard
Deviation
Percentile
5
Percentile
95
1
125
10.83
8.17
8.60
4.25
33.97
2
425
8.42
7.90
4.06
4.28
14.83
3
584
7.54
6.84
3.71
3.99
12.86
4
547
7.05
6.33
4.39
3.08
12.24
5
320
6.43
6.05
2.87
3.14
10.97
6
130
5.83
5.51
2.60
1.85
11.17
7
51
4.79
4.94
2.17
1.35
8.38
8
24
5.16
5.20
1.57
2.42
7.20
9
6
3.74
4.24
1.85
0.18
5.39
10
2
2.58
2.58
4.30
-4.63
5.62
Total
2,214
7.41
6.64
4.34
3.31
13.32
Panel B. Univariate analysis
No. KAM
Firm-year
obs.
Mean
FD
Standard
Deviation
p-value
t test
1-2
550
8.97
5.52
0.000
-9.03***
6-10
213
5.42
2.46
Table 5 provides descriptive statistics of financial distress (FD) by the number of KAMs. All audit reports present
from one KAM up to ten KAMs, and financial data to calculate the financial distress score are winsorized to the 1st
and 99th percentile. Panel A includes mean, median, standard deviation, percentile 5 and 95 of FD by KAMs disclosed
(from one to ten KAMs). Panel B reports the t test differences in means of two sub-samples: firms that disclose one
and two KAMs, and firms disclosing 6-10 KAMs.

49
Table 6. Pearson correlation matrix
Variable
FD
No.
KAM
ENT
KAM
ACC
KAM
PROF
LIQU
SOLV
BIG4
CHANGE
AUDIT
SIZE
Panel A. Pearson correlation matrix
FD
1.000
No.KAM
-.259
1.000
.000
ENTKAM
-.208
.660
1.000
.000
.000
ACC
KAM
-.158
.735
-.023
1.000
.000
.000
.276
PROF
-.132
.346
.022
.440
1.000
.000
.000
.0299
.000
LIQU
-.020
.348
-.055
.513
-.070
1.000
.344
.000
.009
.000
.000
SOLV
-.009
.446
-.004
.600
-.143
-.070
1.000
.001
.000
.862
.000
.000
.001
BIG4
-.108
.027
.040
-.000
-.089
.008
.065
1.000
.000
.204
.005
.969
.000
.743
.002
CHANGE
AUDIT
.010
.010
.001
.012
-.001
.037
-.005
-.036
1.000
.636
.636
.945
.567
.589
.007
.808
.092
SIZE
-.2600
.380
.317
.220
.026
.063
.230
.314
.022
1.000
.000
.000
.000
.000
.225
.003
.000
.000
.300
Table 6 presents the Pearson correlation matrix of the dependent and independent variables included in the regression
models, except for the classification of KAMs per type. All financial data included are winsorized to the first and 99th
percentile to avoid extreme values. We report p-values below each Pearson coefficient and show the probability of
observing this correlation under the null hypothesis that the correlation is zero. Additionally, we tested the correlations
of the classification of KAMs per type and there are no multicollinearity issues among them (untabulated). See Table
3 for KAM definitions and Appendix B for other variable definitions.

Table 7. Linear regression models (contemporaneous)
Model 1
Model 2
Model 3
Model 4
Dependent variable
FD
FD
FD
FD
INTERCEPT
15.994***
15.637***
15.676***
17.354***
(1.96)
(1.98)
(1.97)
(2.03)
No.KAM
-0.526***
(0.09)
ENTKAM
-0.716***
-0.701***
(0.12)
(0.12)
ACCKAM
-0.382***
(0.12)
PROF
-0.690***
(0.23)
LIQU
0.001
(0.20)
SOLV
-0.452**
(0.19)
GC
-3.119***
(0.61)
ICFRAUD
-0.480
(0.31)
RDO
-0.539
(0.34)
MA
-0.468**
(0.19)
TAX
0.004
(0.23)
EIPD
-0.728***
(0.23)
LITMACRO
0.153
(0.42)
MGFEES
-1.336
(1.76)
REV
-0.717***
(0.25)
EXP
-0.018
(0.42)
ACCREST
-0.333*
(0.20)
INV
0.510
(0.40)
CASHREC
0.494
(0.63)
INVEST
0.201
(0.40)
INTANG
-0.614**
(0.25)
PPE
0.650
(0.55)
LLTD
-0.830**
(0.38)
PENS
-1.085***
(0.28)
BIG4
-0.295
-0.297
-.391
-0.235
(1.05)
(1.05)
(1.04)
(0.99)
CHANGEAUDIT
0.466
0.468
0.430
0.345
(0.34)
(0.34)
(0.33)
(0.32)
SIZE
-0.476***
(0.14)
-0.460***
(0.14)
-0.453***
(0.14)
-0.623***
(0.15)

51
Model 1
Model 2
Model 3
Model 4
Dependent variable
FD
FD
FD
FD
Observations
2,214
2,214
2,214
2,214
R-squared
0.154
0.157
0.163
0.219
F-Stat
4.83***
5.07***
4.69***
4.92***
Year and Industry
dummies
Yes
Yes
Yes
Yes
VIF of the model
1.92
1.88
1.82
1.59
Table 7 shows the results of our regression models examining the relation between KAMs and firm financial distress.
KAM disclosures are included as independent variables in our models as follows: Model 1, number of KAMs only,
Model 2, KAMs by risk level (entity-wide and account), Model 3, account-level KAMs by primary impact on a firm’s
financial health (profitability, liquidity, and solvency), and Model 4, KAMs by individual type (based on the 18 item
classification of KAMs presented in Table 3). In the models, all financial data are winsorized to the first and 99th
percentiles to avoid extreme values. Standard errors are reported in parentheses. ***, ** and * indicate (two tailed)
statistical significance at the 1, 5 and 10 percent levels, respectively. Significant coefficients are shown in bold. All
specifications use robust standard errors. See Appendix B for other variable definitions.

52
Table 8. Logistic regression models
Model 1
Model 2
Model 3
Model 4
Dependent variable
FDdum
FDdum
FDdum
FDdum
INTERCEPT
-1.689
-1.769
-1.977
-0.625
(1.83)
(1.85)
(1.92)
(2.12)
No.KAM
-0.678***
(0.11)
ENTKAM
-0.745***
-0.773***
(0.13)
(0.13)
ACCKAM
-0.617***
(0.18)
PROF
-0.197
(0.27)
LIQU
-0.627**
(0.28)
SOLV
-0.825***
(0.24)
GC
-2.762***
(0.33)
ICFRAUD
-0.476
(0.38)
RDO
-0.187***
(0.52)
MA
0.280
(0.35)
TAX
0.178
(0.35)
EIPD
-1.161***
(0.28)
LITMACRO
-0.153
(0.46)
MGFEES
-
-
REV
-0.354
(0.33)
EXP
0.763
(0.70)
ACCREST
-0.657**
(0.32)
INV
-0.542
(0.53)
CASHREC
0.036
(0.48)
INVEST
-0.519
(0.40)
INTANG
-0.481
(0.32)
PPE
-0.011
(0.42)
LLTD
-1.308**
(0.64)
PENS
-1.532***
(0.42)
BIG4
0.530
0.539
0.685
0.539
(0.62)
(0.62)
(0.65)
(0.69)
CHANGEAUDIT
1.182*
1.164*
1.094
0.769
(0.72)
(0.71)
(0.70)
(0.66)
SIZE
0.523***
(0.09)
0.524***
(0.09)
0.539***
(0.09)
0.405***
(0.11)

53
Model 1
Model 2
Model 3
Model 4
Dependent variable
FDdum
FDdum
FDdum
FDdum
Observations
2,035
2,035
2,035
2,035
Pseudo R-squared
0.181
0.182
0.194
0.324
Wald chi2
173.35***
176.60***
208.56***
257.56***
Year and Industry
dummies
Yes
Yes
Yes
Yes
VIF of the model
6.00
5.70
5.21
3.73
Table 8 shows the results of our logistic regression models examining the relation between KAMs and firm financial
distress, measured by the dummy variable of the Altman’s Z’’-Score (FDdum). KAM disclosures are included as
independent variables in our models as follows: Model 1, number of KAMs only, Model 2, KAMs by risk level (entity
and account), Model 3, account-level KAMs by primary impact on a firm’s financial health (profitability, liquidity,
and solvency), and Model 4, KAMs by individual type (based on the 18 item classification of KAMs presented in
Table 3). In the models, all financial data are winsorized to the first and 99th percentiles to avoid extreme values.
Standard errors are reported in parentheses. ***, ** and * indicate (two tailed) statistical significance at the 1, 5 and
10 percent levels, respectively. Significant coefficients are shown in bold. All specifications use robust standard errors.
Statistically significant odds ratios are displayed between square brackets. Other variable definitions are in Appendix
B. Also, 179 observations were dropped since these were not used by STATA for the models.

54
Table 9. Linear regression models: Charitou et al.’s (2004) Approach
Model 1
Model 2
Model 3
Model 4
Dependent variable
FDCha
FDCha
FDCha
FDCha
INTERCEPT
0.665***
0.623***
0.617***
0.846***
(0.13)
(0.13)
(0.13)
(0.13)
No.KAM
-0.065***
(0.01)
ENTKAM
-0.087***
-0.087***
(0.01)
(0.01)
ACCKAM
-0.048***
(0.01)
PROF
-0.046**
(0.02)
LIQU
-0.021
(0.02)
SOLV
-0.067***
(0.02)
GC
-0.405***
(0.04)
ICFRAUD
-0.083**
(0.04)
RDO
-0.067*
(0.04)
MA
-0.011
(0.02)
TAX
-0.027
(0.02)
EIPD
-0.105***
(0.03)
LITMACRO
-0.018
(0.03)
MGFEES
-0.185
(0.22)
REV
-0.044**
(0.02)
EXP
-0.026
(0.03)
ACCREST
-0.040*
(0.02)
INV
0.043
(0.03)
CASHREC
-0.039
(0.04)
INVEST
-0.051*
(0.03)
INTANG
-0.065***
(0.02)
PPE
-0.021
(0.03)
LLTD
-0.060
(0.04)
PENS
-0.068**
(0.03)
BIG4
-0.015
-0.015
-0.014
-0.007
(0.06)
(0.06)
(0.06)
(0.05)
CHANGEAUDIT
0.035
0.035
0.033
0.010
(0.03)
(0.03)
(0.03)
(0.03)
SIZE
0.028***
(0.01)
0.030***
(0.01)
0.031***
(0.01)
0.011
(0.01)

55
Model 1
Model 2
Model 3
Model 4
Dependent variable
FDCha
FDCha
FDCha
FDCha
Observations
2,214
2,214
2,214
2,214
R-squared
0.114
0.120
0.124
0.205
F-Stat
8.93***
8.76***
8.36***
10.56***
Year and Industry
dummies
Yes
Yes
Yes
Yes
VIF of the model
1.92
1.88
1.82
1.59
Table 9 shows the results of our regression models based on Charitou et al.’s (2004) approach examining the relation
between KAMs and firm financial distress, measured by the Charitou’s Score (FDCha). KAM disclosures are included
as independent variables in our models as follows: Model 1, number of KAMs only, Model 2, KAMs by risk level
(entity-wide and account), Model 3, account-level KAMs by primary impact on a firm’s financial health (profitability,
liquidity, and solvency), and Model 4, KAMs by individual type (based on the 18 item classification of KAMs
presented in table 3). In the models, all financial data are winsorized to the first and 99th percentiles to avoid extreme
values. Standard errors are reported in parentheses. ***, ** and * indicate (two tailed) statistical significance at the 1,
5 and 10 percent levels, respectively. P-values calculated from standard errors clustered by firm. Significant
coefficients are shown in bold. All specifications use robust standard errors. Other variable definitions are in Appendix
B.

56
Table 10. Logistic regression models: Charitou et al.’s (2004) Approach
Model 1
Model 2
Model 3
Model 4
Dep. variable
FDChadum
FDChadum
FDChadum
FDChadum
INTERCEPT
-0.426
-0.655
-0.707
0.742
(0.87)
(0.89)
(0.91)
(1.03)
No.KAM
-0.483***
(0.07)
ENTKAM
-0.611***
-0.621***
(0.08)
(0.08)
ACCKAM
-0.381***
(0.09)
PROF
-0.220
(0.14)
LIQU
-0.332**
(0.14)
SOLV
-0.497***
(0.11)
GC
-2.119***
(0.25)
ICFRAUD
-0.736***
(0.28)
RDO
-0.293
(0.31)
MA
0.023
(0.21)
TAX
-0.412*
(0.21)
EIPD
-0.687***
(0.19)
LITMACRO
-0.149
(0.26)
MGFEES
-1.019
(1.12)
REV
-0.193
(0.17)
EXP
-0.172
(0.29)
ACCREST
-0.414**
(0.18)
INV
-0.073
(0.26)
CASHREC
-0.310
(0.35)
INVEST
-0.423
(0.25)*
INTANG
-0.545***
(0.16)
PPE
-0.117
(0.22)
LLTD
-0.570
(0.35)
PENS
-0.536**
(0.26)
BIG4
-0.205
-0.205
-0.173
-0.194
(0.39)
(0.39)
(0.40)
(0.43)
CHANGEAUDIT
0.233
0.231
0.227
0.053
(0.27)
(0.27)
(0.27)
(0.27)
SIZE
0.330***
(0.06)
0.341***
(0.06)
0.347***
(0.06)
0.235***
(0.07)

57
Model 1
Model 2
Model 3
Model 4
Dependent
variable
FDChadum
FDChadum
FDChadum
FDChadum
Observations
2,214
2,214
2,214
2,214
Pseudo R-
squared
0.113
0.117
0.120
0.174
Wald chi2
98.45***
106.15***
119.75***
189.83***
Year and
Industry
dummies
Yes
Yes
Yes
Yes
VIF of the
model
5.76
5.50
5.05
3.62
Table 10 shows the results of our logistic regression models based on Charitou et al.’s (2004) approach examining the
relation between KAMs and firm financial distress, measured by the dummy variable of the Charitou’s Score
(FDChadum). KAM disclosures are included as independent variables in our models as follows: Model 1, number of
KAMs only, Model 2, KAMs by risk level (entity and account), Model 3, account-level KAMs by primary impact on
a firm’s financial health (profitability, liquidity, and solvency), and Model 4, KAMs by individual type (based on the
18 item classification of KAMs presented in table 3). In the models, all financial data are winsorized to the first and
99th percentiles to avoid extreme values. Standard errors are reported in parentheses. ***, ** and * indicate (two tailed)
statistical significance at the 1, 5 and 10 percent levels, respectively. Significant coefficients are shown in bold. All
specifications use robust standard errors. Other variable definitions are in Appendix B.

58
Table 11. Linear regression models (prediction)
Model 1
Model 2
Model 3
Model 4
Dependent variable
FD_lead
FD_lead
FD_lead
FD_lead
INTERCEPT
15.41***
15.17***
15.16***
16.76***
(1.95)
(1.97)
(1.97)
(2.08)
No.KAM
-0.4966***
(0.10)
ENTKAM
-0.638***
-0.623***
(0.12)
(0.12)
ACCKAM
-0.387***
(0.12)
PROF
-0.681***
(0.22)
LIQU
0.015
(0.21)
SOLV
-0.476**
(0.21)
GC
-2.826***
(0.72)
ICFRAUD
-0.234
(0.31)
RDO
-0.438
(0.34)
MA
-0.502***
(0.18)
TAX
0.100
(0.26)
EIPD
-0.634***
(0.24)
LITMACRO
0.104
(0.48)
MGFEES
-0.335
(1.87)
REV
-0.799***
(0.24)
EXP
0.096
(0.46)
ACCREST
-0.276
(0.21)
INV
0.472
(0.42)
CASHREC
0.432
(0.66)
INVEST
0.155
(0.41)
INTANG
-0.744***
(0.27)
PPE
0.701
(0.63)
LLTD
-0.549
(0.38)
PENS
-1.045***
(0.28)
BIG4
0.106
0.098
0.009
0.172
(1.14)
(1.13)
(1.12)
(1.05)
CHANGEAUDIT
0.259
0.276
0.237
0.206
(0.31)
(0.31)
(0.32)
(0.29)
SIZE
-0.495***
(0.15)
-0.484***
(0.15)
-0.474***
(0.15)
-0.639***
(0.17)

59
Model 1
Model 2
Model 3
Model 4
Dependent variable
Z-score
winsorized t+1
Z-score
winsorized t+1
Z-score
winsorized t+1
Z-score
winsorized t+1
Observations
1,774
1,774
1,774
1,774
R-squared
0.1514
0.1533
0.1596
0.2165
F-Stat
4.16***
4.17***
3.97***
3.85***
Year and Industry
dummies
Yes
Yes
Yes
Yes
VIF of the model
1.88
1.84
1.78
1.57
Table 11 shows the results of our regression models examining the predictive power of different classifications of
KAMs and firm financial distress. KAM disclosures are included as independent variables in our models as follows:
Model 1, number of KAMs only, Model 2, KAMs by risk level (entity-wide and account), Model 3, account-level
KAMs by primary impact on a firm’s financial health (profitability, liquidity, and solvency), and Model 4, KAMs by
individual type (based on the 18 item classification of KAMs presented in Table 3). As the dependent variable is in
t+1, some observations are missing. In the models, all financial data are winsorized to the 1st and 99th percentiles to
avoid extreme values. Standard errors are reported in parentheses. ***, ** and * indicate (two tailed) statistical
significance at the 1, 5 and 10 percent levels, respectively. Significant coefficients are shown in bold. All specifications
use robust standard errors. See Appendix B for other variable definitions.