“The impact of key audit matter (KAM) disclosure in audit reports on stakeholders’ reactions: a literature review”

This article presents a literature review of 49 empirical studies on key audit matter (KAM) disclosure in audit reports. The study involves a structured literature review on KAM disclosure based on the reactions of stakeholders. The limitations of former stud- ies and useful recommendations for research are stressed. Five major streams of empirical research that analyze the impact of KAM disclosure on stakeholders’ reactions are focused: (1) shareholders (e.g. investors’ perceptions of auditors’ responsibility and litigation, value relevance and investors’ decisions); (2) debtholders (e.g. loan contract- ing terms); (3) external auditors (e.g. audit processes and audit fees); (4) boards of directors (e.g. earnings management); and (5) other stakeholders (e.g. informational value for suppliers and customers). The authors stress that most of the included studies use experimental or archival data and analyze the impact of KAM disclosure on investor reactions in a US-American setting. As the international standard setters assume a positive impact of KAM on stakeholder reactions, mixed empirical results are found. Although there are some indications of decreased earnings management behavior, most studies find no significant changes in auditor behavior. Furthermore, there are many insignificant results with regard to shareholders’ reaction in line with our stakeholder and behavioral agency framework. The literature review is especially use- ful for management decisions, because firm reputation may be positively or negatively influenced by KAM regulations.

The impact of key audit matter (KAM) disclosure in audit reports on stakeholders' reactions: a literature review

INTRODUCTION
After the 2008-2009 financial crisis, stakeholders widely criticized public interest entities' (PIEs') financial reporting and external auditors' reporting. Longer and more complex annual and audit reports are associated with a high risk of information overload and impaired usefulness for decision-making regarding the capital market (Bédard et al., 2016;Gimbar et al., 2016). In particular, private investors and other kinds of unsophisticated stakeholder groups find it difficult to extract relevant information for their financial analyses. Information asymmetries and conflicts of interest among the board of directors, auditors, share-and debtholders and other stakeholders lead to an expectation gap (Bédard et al., 2016;Gold et al., 2012). In reaction to the huge concern among stakeholders, many regulators introduced extended auditor reporting (see Table 1) for PIEs in recent years. Reduced information asymmetry, increased financial reporting and audit quality and increased value relevance of audit reports are the main goals of audit reporting regulations. In France, justifications of assessments (JOA) have been required since 2003 (Haut Conseil des Commissaires aux Comptes, 2006). Moreover, the UK Financial Reporting Council (FRC) implemented new disclosure rules regarding the significant risks of material misstatement (RMM) in the audit reports of companies with premium listings of equity shares on the main market of the London Stock Exchange (LSE) since the fiscal year beginning on 1 October 2012 (FRC, 2013). In 2015, the International Auditing and Assurance Standards Board (IAASB), as the international audit standard-setter, implemented key audit matters (KAM) for business years after December 14, 2016(IAASB, 2015. Finally in 2017, the US-American Public Company Accounting Oversight Board (PCAOB) introduced the disclosure of critical audit matters (CAM) (PCAOB, 2017). Table 1 presents the four main types of extended auditor reporting. Considering the notable interdependency between the four items, the low amount of JOA and RMM studies, the recent change from RMM and JOA to KAM in the UK and France and the many researchers who use the term KAM, we favor KAM in our literature review. If there are specific differences between CAM/KAM, RMM and JOA in our research results, we will explain them in detail. Otherwise we will only use the term KAM in our analysis.
In comparison to recent literature reviews on extended auditor reporting (e.g. Gimbar et al., 2016;Bédard et al., 2016), we chose a different review strategy and identified five major streams of empirical research that analyze the impact of KAM disclosure on stakeholders' reactions. Most of the 49 empirical studies concentrate on the influence of KAM on (1) shareholders (e.g. investors' perceptions of auditor responsibility and litigation, value relevance and investors' decisions) and found mixed results. The same applies for (2) debtholders as another relevant stakeholder group, but very few studies have been conducted on it so far. KAM disclosure should also have an impact on (3) external auditors' behavior (e.g. audit processes and audit fees). But most of the empirical research does not find any significant results. While KAM disclosures in the audit report mainly addressed external stakeholders, another stream of research analyses the impact of KAM disclosure on (4) the board of directors. Most of the

Content
The auditor shall determine, from the matters communicated with those charged with governance, those matters that required significant auditor attention in performing the audit. In making this determination, the auditor shall take into account the following: • areas of higher assessed risk of material misstatement or significant risks identified in accordance with ISA 315 (revised); • significant auditor judgments relating to areas in the financial statements that involved significant management judgment, including accounting estimates that have been identified as having high estimation uncertainty; • the effect on the audit of significant events or transactions that occurred during the period included studies find a negative impact on earnings management as assumed. Finally, some studies also include (5) other stakeholders, typically based on stakeholders' broad comments on regulations, and state an increased informational value of KAM disclosure.
Our structured literature review contains useful recommendations for future research and is also relevant for regulators and practice. Based on the five streams of research, we identify whether the goals of the regulators have been achieved. We also determine how different stakeholders are integrated into recent research activities. After providing a foundation on stakeholder and behavioral agency theory, we highlight the main results of empirical KAM research on different stakeholder groups, explain the studies' key limitations and give useful recommendations for future research. Finally, we provide a summary of the main results. The results of our literature review are especially relevant for management strategies (e.g. earnings management, management reporting) as extended auditor reporting can mainly influence stakeholder reaction and thus firm reputation. 1 According to DeAngelo (1981), audit quality represents the ability of an auditor to detect material misstatements and to report the breach, if not corrected. Thus, auditor reporting is part of audit quality. 2 According to Gaynor et al. (2016), higher quality financial reports "are more complete, neutral, and free from error and provide more useful predictive or confirmatory information about the company's underlying economic position and performance". Research on audit quality and financial reporting quality is classified by various interdependencies. Audit reporting refers to the audited financial statements and is clearly linked with financial reporting from a stakeholder perspective. Thus, the use of financial reporting variables (e.g. degree of earnings management) in order to analyze audit quality is common in empirical research. However, as the external auditor only evaluate the legality of financial reporting and not its firm profitability (e.g. usefulness of accounting policies), financial reporting quality does not automatically lead to better audit quality and vice versa (Gaynor et al., 2016).

AGENCY THEORETICAL FOUNDATION
According to stakeholder agency theory, an external auditor acts as an agent for shareholders and other stakeholders (Hill & Jones, 1992; Ross, 1973;Jensen & Meckling, 1976;Chow, 1982). External audits are intended to increase stakeholders' trust in financial statements. The auditor serves as a gatekeeper or public watchdog for the stakeholders (Kraakman, 1986). Publication of the audit report with KAM disclosures should ensure appropriate audit quality 1 and financial reporting quality 2 in line with stakeholders' interests (Ittonen, 2012). Empirical audit research has proved that audit reports can significantly influence capital market reactions (Bédard et al., 2016;Gimbar et al., 2016). As audit quality cannot be observed by stakeholders, the audit report is a key information tool. However, the informational value of the audit report and related market reactions are mainly affected by the expectation gap phenomenon due to agency conflicts (Liggio, 1974). The expectation gap represents the difference between stakeholders' assumptions concerning the purpose and range of the audit and the real audit. According to the famous approach proposed by Porter (1993), the expectation gap can be divided into a performance gap and a reasonableness gap (see Figure 1).  Prior research states that financial statement users often expect an absolute and not a reasonable level of external audits, and a guarantee of the absence of fraud or financial distress (Gold et al., 2012), resulting in the reasonableness gap. This gap occurs when stakeholders lack knowledge about the audit risk model of the external auditor. Audit risk can be separated into inherent risk, control risk and detection risk. Audit risk is the risk of submitting an unqualified audit opinion with substantial errors in the financial reporting. Inherent risks (e.g. business risk) and control risks (e.g. an insufficient internal control system) cannot be influenced by the external auditor himself; he can only influence detection risk, as he may choose between different kinds of audit proof. A key goal of external audits is to lower audit risk to an appropriate level (5% on average) with reference to materiality and efficiency. KAM disclosure can lower the reasonableness gap as the stakeholders can be better informed about the range and limits of external audits (Velte, 2018). The performance gap, the second part of expectation gap, can be linked to deficient performance and standards (Porter, 1993; see Figure 1). International regulations on KAM disclosure indicate that prior auditor reporting standards were not decision-useful for stakeholders as they did not get firm-specific information about the audit process and the outcome (Bédard et al., 2016). Furthermore, as external auditors are also economic agents, a lower degree of auditor reporting can lead to decreased performance and quality incentives. Thus, KAM disclosure may decrease all main sources of the expectation gap in line with regulators' assumptions (FRC, 2013; IAASB, 2015; PCAOB, 2017).
Empirical research on the expectation gap can be separated into three major streams (Gold et al., 2012). One stream focuses on the existence of the expectation gap in several countries. The second stream analyzes the influence of stakeholders' sophistication (i.e. level of experience and knowledge) regarding financial reporting and auditing on the expectation gap. In contrast to professional investors, private investors and other stakeholders with lower knowledge are more dependent on audit reports (Porter, 1993). The third research stream relates to the impact of differences in wording in the audit report on the expectation gap. Prior research stresses that a precise and transparent audit report can decrease the expectation gap and increase stakeholders' trust (Gold et al., 2012). KAM disclosures include firm-specific information about an audit of financial statement (Bédard et al., 2016;Gimbar et al., 2016). They portray an individual picture of the main critical accounting topics or items in a company (Velte, 2018). According to stakeholder agency theory, KAM disclosures contribute to lowering information asymmetry and conflicts of interest between management and stakeholders and decreasing the expectation gap (Fuller, 2015). Stakeholder agency theory assumes heterogeneous interests between stakeholders and expects homogeneous interests within a stakeholder group. Furthermore, it accepts the rational behavior of the principal(s) and the agent(s), and the risk-neutrality of the agents. As we will discuss in the next chapter, archival research is the dominant research method used to analyze the macro-economic impacts of KAM disclosure on stakeholders' reactions (e.g. Almulla & Bradbury, 2018 . According to these approaches, audi-tors feel more morally licensed to accept material misstatements by the client after the introduction of KAM disclosure. Thus, KAM disclosure is not an accountability mechanism, but a tool or license for unconsciously justifying an auditor's decision to waive an adjustment (Asbahr & Ruhnke, 2018).
In summary, KAM disclosure may be linked with increased, decreased or no impact on accounting quality, audit quality and stakeholder reactions based on stakeholder agency and behavioral agency theory.

Sample selection, research framework and methods
As the development of empirical KAM research is relatively new, there have been no explicit literature reviews to date. Bédard et al. (2016) conducted a literature review on US reform activities concerning auditor reporting (e.g. going concern opinions, audit partner signatures, CAM), and KAM disclosure is only one topic that is covered. Furthermore, Gimbar et al. (2016) presented the results of selected experimental studies concerning CAM's influence on auditor litigation. Our structured literature review contributes to the present body of literature, because it focuses on KAM disclosures without any limitations regarding research methods or region, stresses the limitations of former studies and gives useful recommendations for research and practice. We rely on a stakeholder agency and behavioral agency frame-work and present the impact of KAM disclosure on (1) shareholders, (2) debtholders, (3) external auditors, (4) the board of directors, and (5) other stakeholders (see Figure 2).
To select our studies, we used several international databases (e.g. Web of Science, Google Scholar, Social Science Research Network (SSRN), EBSCO, ScienceDirect) and specific terms (e.g. "key audit matters", "critical audit matters") in combination with "audit reporting", "audit quality", or comparable terms. Restriction to a specific timeframe was not useful in light of the discipline's newness. to the extended auditor reporting variables, we will emphasize them. Otherwise, in order to increase readability of our literature review, we only use the term KAM. Table 2 provides an overview of the number of included contributions per stakeholder and their year of publication, the regions examined and the journals in which the contributions were published or the working papers. All studies were published or prepared within the last years (2014-2018), with a clear increase in the number of studies published in 2018. Most empirical research on KAM focuses on investors' reactions in a US-American setting and adopt an experimental design. In addition, most research was published as working paper. In view of the latest regulation initiatives regarding KAM and the long review duration in top international journals, this is not surprising. We also identify more research activity in 2018 concerning the implications of KAM on the board of directors (e.g. management reporting, audit committee) and external auditors. Accounting and auditing journals are the primary publication strategies.

Results
As  Table 3). There is some evidence that KAM disclosures increase investors' perception of auditors' negligence (  In sum, we argue that recent experimental and archival research on the influence of KAM disclosure on investors' perceptions is rather heterogeneous. This result is in line with our stakeholder and behavioral agency framework. On the one hand, KAM disclosure may lead to an increased expectation gap, as risk-averse shareholders that perceive an increased audit risk may leave the company. They may blame external auditors for the increase in firm risks and don't trust them with regard to irrational behavior. On the other hand, KAM disclosure can lead to a lower expectation gap as shareholders are better informed about the scope of external audits and, at minimum, risk-neutral or risk-seeking shareholders honor information transparency with capital engagement.

Limitations and research recommendations
Empirical research on behavioral auditing can be still classified as a niche (

Results
In contrast to shareholders, we identify only three studies on debtholders ( Porump et al. (2018) found that an increased amount of RMM is connected with less favorable loan contracting terms.

Limitations and research recommendations
Recent empirical KAM research on debtholders reactions is not valid and of low quantity unlike research on investors. As debtholders are key corporate stakeholders, especially in small and medium-sized entities, we encourage researchers to include this stakeholder group in future designs. Interestingly, no experimental research has been conducted so far. This research gap should be decreased in light of the heterogeneous preferences of debtholders and their unclear reaction to KAM disclosures. In line with the behavior of shareholders, individual risk preferences may lead to increased or decreased loan contracting conditions for a company. It is most useful to analyze the link between KAM disclosure and debtholders' reactions using archival data. In line with our recommendations regarding investors' reactions, future research should conduct cross-country studies on the impact of KAM disclosure on debtholders' reactions.

Limitations and research recommendations
Interestingly, the surveys and experiments directly address external auditors. However, with the exception of Asbahr and Ruhnke's (2018) study, the number of participants in these studies is rather low. Therefore, the validity of the studies is limited. Pelzer (2016) mentioned the risks of misinterpretation and use of boilerplate information when discussing KAM. Thus, we encourage future researchers to survey external auditors to determine how they actively contribute to better transparency and readability of KAM disclosures, as regulators have criticized the lack of transparency of audit reports. As accounting topics, which lead to KAM disclosures, are often very complex for external stakeholders, the tone of explanations in the audit report may have a major impact on the expectation gap. Future archival research should analyze audit-related determinants (e.g. Big Four audits, industry specialists, (non) audit fees, audit report lag) on KAM readability. Most recent archival studies used audit fees as a proxy for audit quality ( . We are aware that the audit fee proxy has limited validity, because it can reflect both audit independence and audit efficiency. Non audit fees, auditor tenure and audit partner signature can be more useful for analyzing auditor independence. As many countries have recently implemented KAM disclosure rules, learning effects of external auditors will have a major impact on auditor reporting. To do so, time series analyses are recommended. As we will describe in the next section, cooperation between audit committees and external auditors should be included in future research. Also, research should analyze whether the audit committee has a complementary or substitutional relationship with the external auditor, which may lead to increased or decreased audit fees.

Results
Not only share-and debtholders and external auditors, but also a major impact of KAM disclosure on board of directors is expected in empirical research. Thus, KAM disclosure are linked with management decisions, e.g. earnings management or management reporting. According to our stakeholder agency-theoretical framework, extended auditor reporting should lead to decreased information asymmetries and conflicts of interests between management, external auditor and capital market. We identify five experiments ( Table 6).
Most studies explore the impact of KAM disclo- Three studies examined the role of the audit committee as a monitoring institution (Fuller, 2015;Kang, 2018;Velte, 2018). While Kang (2018) stated that the audit committee decreases their overall questioning behavior regarding management when KAM is disclosed, Fuller (2015) found that the strength of the audit committee's oversight moderates the positive association between KAM disclosure and management reporting. Velte (2018) is the only author in our sample that analyzed the impact of audit committee composition (gender diversity) on RMM readability and stated a positive link.  Note: * Study that analyzes possible determinants of KAM disclosure.

Limitations and research recommendations
With one exception (Klueber et al., 2018), the experiments were performed only in a US-American setting. We already stated that the empirical results are not transferable to other regimes. From an international perspective, one-and two-tier systems of corporate governance are very different in terms of composition and to KAM disclosure. In addition, the US-American one-tier system (e.g. by Bentley et al., 2018) is not comparable to the German two-tier system (Klueber et al., 2018). This is mainly important in relation to the role of audit committees in one-tier and two-tier systems. Audit committees in one-tier systems tend to have increased information access, but can have decreased independence in comparison to committees in two-tier systems. For this reason, research on two-tier systems is recommended.
Few earnings management studies based on experiments have been conducted to date (Klueber et al., 2018). We identify a research gap with regard to empirical qualitative research (surveys, interviews, case studies) on the impact of KAM disclosure on the reactions of boards of directors.
Both management and audit committees should be addressed in future research to gain more information about the cooperation process between the audit committee and external auditor and the correspondence between management reporting (e.g. in the notes or management report) and the KAM disclosure of the external auditor. It is also relevant to analyze the impact of audit committee composition (e.g. members with financial expertise or industry expertise) on KAM disclosure, as the audit committee and external auditor must discuss the audit focal points, which can be relevant for the external auditor's choice of KAM to disclose in the audit report (Velte, 2018 Table 7). Prasad

Limitations and research recommendations
Not only the amount of studies on the impact of KAM disclosures on other stakeholders, but also the focus on qualitative research (content analyses) lowers their relevance. Future researchers should conduct interviews, surveys and case studies on other stakeholders' perceptions of KAM (e.g. customers' awareness, suppliers' attraction, employee relationships). Furthermore, empirical quantitative research on other stakeholders should be both experimental and archival nature. The variety of regulations on KAM disclosure implemented in recent years are not limited to share-and debtholders, external auditors, and boards of directors. Thus, increasing trust of other stakeholders is an important goal for future research. PIEs are confronted with increased regulations concerning not only financial accounting, auditing and corporate governance, but also non-financial reporting (e.g. Corporate Social Responsibility (CSR) reporting or integrated reporting). As non-financial reporting and CSR assurance can be classified as key topics in research, regulation and practice, the interdependencies between KAM disclosure, integrated reporting and non-financial issues should be addressed from a broader stakeholder perspective in future research.

SUMMARY
Capital market trust in financial reporting and external audit quality has decreased after the financial crisis of 2008-2009. Information overload in annual reports and audit reports is linked with decreased decision usefulness for shareholders and other stakeholders. As a result, auditor reporting (e.g. KAM disclosures) have been extended by various institutions (e.g. FRC, 2013; IAASB, 2015; PCAOB, 2017). Regulators assume KAM disclosures are linked with lower information asymmetry and expectation gap and thus increase financial reporting and audit quality. In contrast to former literature reviews on extended auditor reporting (Gimbar et al., 2016;Bédard et al., 2016), we rely on 49 empirical studies on KAM disclosure and conduct a stakeholder-oriented structured literature review. We identified five major streams of empirical research that analyze the impact of KAM disclosure on stakeholders' reactions: those focusing on (1) shareholders (e.g. investors' perceptions of auditors' responsibility and litigation, value relevance and investors' decisions); (2) debtholders (e.g. loan contracting terms); (3) external auditors (e.g. audit processes and audit fees); (4) boards of directors (e.g. earnings management); and (5) other stakeholders (e.g. informational value for suppliers and customers). Most of the included studies use experimental or archival data and analyze the impact of KAM disclosure on investor reactions in a US-American setting. However, the intended goal of KAM regulations is still controversial, as evidenced by the mixed empirical results. Although there are some indications of decreased earnings management behavior, most studies find no significant changes in auditor behavior. Furthermore, there are many insignificant results with regard to shareholders' reaction in line with our stakeholder and behavioral agency framework. In order to increase the validity of empirical KAM research, we recommend a variety of research topics for the five streams of research. In particular, cross-country studies are needed due to the international regulation of KAM during recent years. Interestingly, behavioral research on auditing has been established within the field of empirical KAM research, but not related to other audit topics. We see a major need to include a mixture of research methods to explain the impact of KAM disclosure on stakeholders' reactions. As stakeholders' preferences are heterogeneous and many stakeholder groups have limited knowledge and experience with KAM disclosure, external auditors should provide a transparent and readable audit report without any boilerplate information. Otherwise, KAM regulations result in opposite stakeholder reactions (e.g. increased expectation gap and decreased stakeholder trust in financial accounting and external audits). Thus, our results are not only relevant for future research but also for regulators and practice. Future developments in KAM regulation (e.g. increased precision of audit reporting standards, possible extension of companies) have to be critically discussed in light of the limited evidence regarding increased accounting and audit quality. The results of our literature review are especially relevant for the board of directors and top management decisions. Recent empirical research on KAM disclosure indicates that KAM disclosure may have an impact on earnings management and management reporting behavior (e.g. risk reporting). Furthermore, the audit committee as a monitoring institution of the board of directors is linked to auditor cooperation and KAM disclosure. KAM disclosure and external audit in total should be classified as a major challenge to increase stakeholder trust after the financial crisis. Thus, management should be aware of the price competition and supplier concentration on the audit market. Perceived audit quality may be low if the audit committee does not grant enough resources and fees for the external auditors.