Readability of Financial Footnotes, Audit Fees, and Risk Management Committee

Abstract

We find that the readability of financial footnotes and risk management committees contributes to audit fees. We use observations from 758 companies listed in Indonesia for 2014–2018, and moderated regression analysis is used for statistical analysis. The results show that the readability of financial footnotes will affect audit fees paid, and RMC strengthens the relationship between the readability of financial footnotes and audit fees. In addition, we also used robustness assays to address endogeneity issues with consistent results as moderated regression analysis (hereafter MRA). These findings provide evidence for policymakers about the relationship between the readability of financial footnotes, RMC, and audit fees.

1. Introduction

The main goal of the company is to increase the value of the company. A high-value company can increase prosperity for shareholders so that shareholders will invest their capital in the company. The company’s annual report has always been one of the essential resources for decision making by capital market practitioners (as well as creditors and financial analysts). Easy to read and understand the information in the annual report by shareholders and prospective shareholders is undoubtedly an important point so that it does not become information asymmetry (Salehi et al. 2020). Readability is also an assessment of effective communication for information relevant to the company’s assessment and easy to analyze (Loughran and McDonald 2014; Lehavy et al. 2011). Therefore, easy-to-read management information can be of interest to analysts, reducing the time they spend analyzing information, thereby making them easier to access. Although the purpose of annual reports is to provide users with accounting information, several studies have highlighted that the language used in these reports is often complex, making it challenging to measure firm value accurately (Dalwai et al. 2021).

Deloitte (2015) argues that there is much information conveyed through corporate reporting frameworks, which, due to their volume and complexity, often does not really help people in deciding where to invest their money or how to understand the business better. On the other hand, long-winded and difficult-to-read financial reports will waste the time of shareholders and potential shareholders, and when there are many unnecessary words (verbosity), investors will tend to stop reading the report (SEC 2007). Public companies are required to provide comprehensive information to users of financial statements. Managers may find it advantageous to release an annual report that is difficult to understand. Managers can use the readability of annual reports as a strategic tool, especially when a company is in financial difficulty. In this case, they may intentionally make the information unreadable to send a biased message. The aim is to present the company as positively and optimistically as possible, for example, by presenting limited information about the company’s risk to influence investors’ perceptions of the company’s financial situation (Le Maux and Smaili 2021). Verbosity can also be used to hide poor performance (Li 2008) and income manipulation (Lo et al. 2017). Previous studies show that firms bear economic consequences for issuing hard-to-read annual reports. Specifically, firms with less-readable reports experience lower stock liquidity (Cho et al. 2022). This allows investors to stay away from financial markets. Because the purpose of the annual report fails to provide useful information for investment decisions, regulators, creditors, and researchers, the legibility of financial statements will impact market risk. An essential product of financial statement analysis is accounting adjustments, calculated using information disclosed in financial statement notes. For example, note information can be used to capitalize the assets and liabilities associated with operating leases. These adjustments correct missing or misclassified expense and revenue items and thus make the adjusted income statement numbers more informative as to a company’s economic performance. Like another example, equity analysts use note information to assess and calculate the necessary accounting adjustments. When the amount of the adjustments required differs from the analysts’ prior expectations, it will affect their assessments of the company’s future performance and valuation. Given the new information, analysts revise their target price estimates, thereby impounding financial statement note information into stock prices. For this reason, we expect accounting adjustments to be informative and, hence, valued by investors (De Franco et al. 2011a; Abernathy et al. 2019).

Auditors have an important public interest role in the financial markets by providing independent assurance of the credibility of accounting information leading to improved resource allocation and increased contracting efficiency. They also play an “informational role” in enhancing information from financial statements and providing useful information for investor decision making (Blanco et al. 2021; DeFond and Zhang 2014). Because in making decisions, companies must present information properly and correctly. Managers who operate will have more information and tend to take actions that the owners do not have. This is commonly known as agency conflict (Jensen and Meckling 1976). To overcome this problem, an external party who is considered a guarantor for the preparation of the annual report from risks is needed, which is a public accountant. Public accountants act as external auditors within the public accounting firm, their audit services require fees commonly known as audit fees, which include all fees paid and paid to the auditor by a company for audit services performed, and the amount of audit fees is influenced by the size of the client company, audit complexity, and audit risk (Prabhawa and Nasih 2021; Simunic 1980). From an audit perspective, the readability of the financial footnotes provides information about engagement risk because bad news can be more difficult to include in the annual report (Bloomfield 2008). Readability can also be affected by information being kept secret by managers by making it difficult for investors to obtain information that managers do not want to disclose, managers covering up bad news with more complex language and highlighting good news in clear language (Bloomfield 2002; Li 2008; Lo et al. 2017; Courtis 1998). This has implications for audit engagements because an illegible note to financial statements in the financial statements can indicate poor performance. Management uncertainty over such performance may indicate higher client business risk and possibly higher litigation risk. Auditors will generally respond to higher engagement risk by increasing effort or charging an additional risk premium (Simunic 1980; Hay et al. 2006; DeFond and Zhang 2014; Simunic and Stein 1996; Johnstone and Bedard 2004). The risk premium has several considerations, such as the client company, audit complexity, audit risk, and others. However, on the other hand, high audit fees can lead to dependence between auditors and clients by presenting more profitable reports to retain relationships with clients and higher costs (Kirkos et al. 2007). This statement is also supported by research by Salehi et al. (2020), which revealed that auditors in Iran charge higher fees to companies with more legible records on their finances; that will be our motivation to fill the gap in this study to reveal how the relationship between financial footnotes and audit fees is related.

Research by Walker et al. (2002) found that the risks that may occur within the company are of concern to auditors, and there has been an increase in risk management awareness caused by the many corporate scandals and unexpected business bankruptcies in previous years. Good corporate governance is one of the things that how companies control their business risks. Previous research by Abdullah et al. (2015) also shows that the risk management committee (hereafter, RMC) has a role in risk control, detection, and prevention, particularly financial risk inherent and control risks may be related to audit fees based on the auditors’ production costs (Badertscher et al. 2014). Thus, corporate governance can reduce agency costs generated by information asymmetry between managers and shareholders, and reducing these costs will increase firm value (Staszkiewicz and Karkowska 2021). Interestingly, there is positive evidence about the relationship between RMC and audit fees (Hines et al. 2015). High expectations of the performance of the risk committee make senior executives involved in risk monitoring. To address this problem, members of the company’s board of directors initiated a new organizational structure to assist the company’s risk-tracking process (Beasley 1996).

There are several previous studies that discuss the readability of several components of the company’s annual report. For example, Salehi et al. (2020) use a sample of all companies listed on the Tehran Stock Exchange (TSE) for 2012–2017. This article analyzes how the auditor’s response to the readability of the notes to the financial statements of a company. Initially, this study hypothesized that the lower the readability of financial footnotes, the auditor would tend to set a higher audit fee, assuming the more complex the information, the greater the risk borne by the auditor, where one of the determinants of the size of the audit fee is the risk factor. They use three auditors’ measurements: audit report lag, audit fees, and going concern opinion (GCO). The results of this study indicate that report lag and fees are positively related to readability, while GCO is negatively related. This is based somewhat on the fact that auditors will increase more effort in easy-to-read reports. They are less likely to charge clients a higher audit fee.

A similar study was written by Abernathy et al. (2019), which in this study examines the readability and results of the company’s audit, using samples from the U.S. from 2001 to 2014. In this study, readability was measured by the number of words in the report (LENGTH), words that were easily accessible (COMMONWORDS), and financial words that were affordable by the general public (FINANCIALWORDS). Audit results are measured using lag, fees, and going concern opinions. This study’s results align with the hypothesis that hard-to-read financial footnotes can be costly and delay the audit process. The results of this study are also supported by a robustness test with propensity score matching (PSM), which separates samples with good readability and less legibility from the median of readability measurement.

Xu et al. (2019) try to examine these two variables from the reverse side. This study uses the readability of management discussion and analysis (MD&A) as the dependent variable and audit costs as the independent. This article argues that audit is a proxy for audit quality; the higher the audit quality, the more errors will be detected and corrected. The results of a good audit will also provide better information to shareholders. Coupled with the signaling theory, which illustrates that management with good audit quality indicates that they do not want to report that they are difficult to read because they have paid much money, they will ask for the quality of financial reports to be better and legible. This study uses a U.S. sample from 2000 to 2016. The results of this study reveal that there is a positive relationship between audit fees and MD&A readability.

This study focuses on the relationship between the readability of the notes to financial statements and the audit fees in Indonesia and how to memorize the relationship between the two. This research contributes in several ways. First, this study provides empirical evidence about the level of quality of financial statements in terms of legibility with audit fees. Second, to provide insight to management, investors, and credit regarding the role of good corporate governance in dealing with risks, particularly from the legibility of the notes to financial statements. Third, add to the audit literature in Indonesia, especially audit fees with company risk.

The purpose of this study is “To examine the relationship between the readability of the financial footnotes and audit fees”. Then to “examine relationship between RMC and audit fees” and “ how the role of the RMC as a moderator on the relationship between the readability of financial footnotes and audit fees”. This study uses data on the financial statements of companies listed on the Indonesia Stock Exchange, that gives us the opportunity to assess audit fees and footnotes’ readability. We focus our analysis on Indonesia because by limiting the scope of the study to one homogeneous country and industry, we can control for a range of variables that cannot be controlled in cross-country and cross-industry data. Thus, the effect of readability is isolated from the standard factors and other country characteristics. The data used are from 758 companies for the 2014–2018 period. The independent variable for this study is the readability of financial footnotes, the dependent variable is the audit fee, and the moderating variable is the risk management committee.

The results of this study implicate both theoretically and practically. In general, this study uses agency theory where there are differences in the interests of agents and principals, then the possibility of information asymmetry in financial statements and other risks. The readability of financial footnotes will tend to add information to the auditor, and from the supply side, the auditor will match the information so that the risk of misstatement can be reduced, while the relationship between RMC and audit fees is on the demand side, where RMC will indirectly ask the auditor to conduct a more thorough examination. Both from the supply side and the demand side, both parties will increase audit fees because the effort and time of auditors will increase.

From a practical point of view, investors, potential investors, and others, besides paying attention to the company’s quantitative information, also have to pay attention to qualitative information. For auditors, this research encourages auditors to view financial footnotes as information or as a risk threat in audit assignment risk; in addition, the presence or absence of a risk management committee in the company will play an essential role in the workload of auditors. For management, to reduce the risk of misstatements that may occur and go undetected, the readability of financial footnotes can be paid more attention. At the same time, the presence of RMC in the company’s organization will help management monitor the risks in the company.

The remainder of this paper has the following structure: Introduction, this chapter describes the background of the research, which contains a brief explanation of the importance of corporate stakeholders in understanding the information provided by managers through annual reports and the audit function to ensure that information asymmetry does not occur and the background for conducting research, as well as how good corporate governance plays a role in monitoring company risk. Literature Review: This chapter explains previous research regarding the readability of financial footnotes, audit fees, risk management committee, hypothesis development, and conceptual framework regarding the relationship between variables. Materials and Methods: This chapter describes the research approach used, namely the quantitative approach. In addition, this study also explains the identification of variables, details of sample selection from the population, and an explanation of the analysis used in the study. This study uses the MRA (moderated regression analysis) analysis method. Results and Discussion: This chapter shows the empirical results of the data analysis process, including descriptive statistics, Pearson correlation, and moderated regression analysis, accompanied by a discussion of each hypothesis. To strengthen the statistical results, this study also tested the robustness of the model with coarsened exact matching (CEM) regression. Furthermore, this study also takes a model with different proxies or independent measures. Conclusions and Suggestions: The last part of this study contains the conclusions from the results of this study. Second, this study discusses the limitations contained in the study; then, there are implications from the research both from the management and shareholder side. Furthermore, the authors discuss suggestions for future researchers to make up for the shortcomings and make future research more perfect.

2. Literature Review

2.1. Agency Theory

This study uses agency theory as a basis that supports the difference in interests between management and shareholders. According to Jensen and Meckling (1976), agency theory is a theory that discusses the relationship between principals and agents. In this case, management is considered as an agent, while shareholders are the principal. The agent’s task is to give responsibility for his work and decisions taken within the company to the principal. In this case, the agent sometimes has a different goal or interest from the principal. This can lead to unintentional or intentional errors, commonly called fraud. Therefore, from the audit perspective difference in interest is called agency theory.

Furthermore, agency costs are further divided into three things, namely monitoring costs, bonding costs, and residual loss. Monitoring Costs: Costs are costs incurred to monitor agents’ behavior, namely to measure, observe, and control the behavior of agents. This benefits ensuring that the information provided by the manager or agent is correct. In ensuring this, monitoring costs are incurred to hire an external part of the company, one of which is to use an external audit. Audit fees are an essential part of monitoring costs because auditors must ensure that managers behave appropriately and check company accounts. Bonding costs are costs borne by the agent to establish and comply with the mechanism that ensures that the agent will act in the interests of the principal. Therefore, in paying the salaries of the audit committee and the internal audit division, it is paid through bonding costs. Residual loss is a cost incurred due to fraud or manipulation by managers or agents issued by shareholders or principals, which are considered detrimental.

2.2. Audit Fee

Audit fees (fee for services) are all fees paid to auditors by a company for audit services performed. The size of the audit fee is influenced by the size of the client company, audit complexity, and audit risk. (Simunic 1980). According to Ettredge et al. (2014) required, higher costs ensure higher effort to deal with higher risks. Better audit quality can detect more errors and result in fewer misstatements. So in its implementation, the auditor must make a greater effort.

The auditor’s function is to ensure that the information in the company’s annual report is at a fair value, both legally and ethically. An auditor appointed by the audit committee and in the position of principle will make the external auditor responsible and fulfill principle desires, one of which is the absence of information asymmetry, whether intentional or not, in the company’s annual report. According to Dey (2008), auditing is one of the measurements of corporate governance, which in this study sees that the agency profile that occurs by principals and agents can be reduced by implementing good corporate governance. Auditors consider the client’s overall business risk when determining engagement risk (DeFond and Zhang 2014). As a result, auditors usually increase audit effort or charge a risk premium to compensate for the high client risk (Simunic 1980). Therefore, in determining the cost of the auditor, it is necessary to consider the right time to assess the company’s risk and how the auditor will make much effort in the future in carrying out the audit process (Lobo and Zhao 2013).

2.3. Financial Footnotes Readability

The aim of annual reports is to convey accounting information to users. Numerous studies have emphasized that the language used in these reports is often complicated, making it difficult to accurately gauge firm value (Dalwai et al. 2021; Hassan et al. 2018; Fisher et al. 2019; Luo et al. 2018). Financial statements can only be presented in short words and numbers, so the notes to the financial statements are needed to provide additional important information that cannot be explained in the report (Worthington 1977). In addition, the existence of the notes to the financial statements is also needed in correcting some of the shortcomings of the financial statements, and investors also take into account the information contained in the notes to the financial statements after the financial statements and audit reports (De Franco et al. 2011b; Schwarzkopf 2007). This makes the information from the notes to the financial statements need to be understood and understood by stakeholders. Readability is one way to measure the understanding of narrative texts in a company’s annual report. Narrative text in management analysis and discussion is essential information about the company’s performance for investors to see. Readability is also one of the aspects that annual report users consider when evaluating a company. Readability can increase the company’s value in investors’ eyes. Management needs to consider legibility to maintain the quality of the annual report because a quality annual report greatly affects the company’s success (Pivac et al. 2017).

According to Chall (1958), readability refers to a different set of factors involving interest, ability, and ease of reading comprehension. Sun et al. (2014) also state that readability can be interpreted as a person’s level of difficulty in understanding something written. Robert Gunning formulated readability in 1952, which later became known as the Fog indicator. The notes to the financial statements contain additional information for the financial statements for the reader. The information presented is in the form of narrative text where the readability of the notes to financial statements is an aspect that users of the annual report consider when evaluating a company.

2.4. Risk Management Committee

The global financial crisis that occurred in the past has caused more and more companies to form risk management committees (RMC) at the board level (Malik et al. 2021). The risk management committee’s emergence is one of the tools for companies can control risk effectively and help investors better understand company risks (Linsley and Shrives 2006). In addition, the expertise of RMC will assist in communicating risk management information between divisional operations and strategy. For example, you can specify what important information should be obtained from the operations division (Financial Reporting Council 2011). Previous research has also found that risk management reduces asymmetric information and can increase firm value (Miihkinen 2013; Abdullah et al. 2015). However, information on risk management is still low and related to nonfinancial risk management; it is done voluntarily (Linsley and Shrives 2006; Larasati et al. 2019). In addition, the presence of RMC in the firm can increase audit demand (increase fees) or assess auditor risk (reduce costs) (Carcello et al. 2002).

Modern corporate governance systems are divided into internal and external mechanisms, which vary depending on the particular environment (Weir et al. 2001). Indicators of the internal governance mechanism consist of the number of directors, the proportion of independent commissioners, and external debt-to-equity governance in the form of institutional ownership (Beiner et al. 2004). When these two mechanisms work together, the corporate governance system will enable managers to maximize shareholder value (von Nendelstadh and Rosenberg 2003).

3. Hypothesis Development

In several previous studies by Salehi et al. (2020), Abernathy et al. (2019), Xu et al. (2019) related to readability and audit results, inconsistent results were found. Readability and audit costs have an agency theory background in explaining how the two variables are related. Agency theory, which argues that there is a difference of interest between agents (management) and principles (shareholders), has been developed and used by many related studies (Hassan et al. 2021; Prabhawa and Nasih 2021; Larasati et al. 2019; Datta et al. 2020; Salehi et al. 2022). The agent’s job is to assign managers responsible for the work and decisions made within the company. In this case, the agent sometimes has goals or interests that are different from the principal (Jensen and Meckling 1976). To overcome this problem, there are costs to overcome agency problems, namely costs incurred by shareholders to unify the actions of managers, namely monitoring costs (monitoring costs). To ensure this, costs are incurred to hire an external part of the company, namely the external auditor (Jensen and Meckling 1976). In addition, there is a supply side from the auditor. By looking at litigation, reputation, and regulatory risks, whether or not the notes to the financial statements are legible will increase or decrease audit fees. In terms of litigation risk, the legal consequences for the auditor can be quite significant. Evidence suggests that this threat affects audit fees (Habib et al. 2014). The audit reputation in question is the risk that when the auditor fails to gain the trust of stakeholders regarding his audit opinion, where there is something that supports lowering the reputation, the auditor will charge higher fees along with the increase in reputation.

Narrative disclosures are important and evaluated by auditors early in the audit process (Blanco et al. 2021). Readability is one of the measurements of how stakeholders understand the descriptive narrative in the company’s annual report. Previous research has shown that financial statements provide important information to capital market participants (Ball and Brown 1968; Beaver 1968; Abernathy et al. 2019). Therefore, companies need to maintain readability quality because it will increase investor value and good quality annual reports also contribute to the company’s success (Pivac et al. 2017).

Auditors have much greater information about client performance and risk than principles (Hay et al. 2006; DeFond and Zhang 2014). Therefore, auditors must be able to report how the report’s readability can affect the risk and usefulness for investors (Sharma et al. 2018). Auditor engagement risk consists of three components: (1) client’s business risk, namely risks related to the client’s life and profitability; (2) audit risk, namely the risk of the auditor failing to change the wrong opinion about material financial statements; and (3) auditor business risk, potential litigation risk and auditor reputation (Johnstone 2000; Stanley 2011; DeFond et al. 2016). From these risk components, the legibility of the notes to the financial statements can affect the risk of auditor involvement. Bloomfield (2008) suggests that communicating bad news may be more complicated. Thus, legible financial statement footnotes may indicate the client’s business risk from delivering bad news.

Previous research revealed that readability and audit results had a gap between positive and negative correlations. The difference in the results of this study is based on geographical differences in where the sample was carried out. From the research gap above, this study tries to reveal how the relationship between the notes on financial statements and audit fees in Indonesia. Based on this explanation, the hypothesis proposed by this study is non-directional, considering that the readability of the notes to financial statements has a different relationship in previous studies. The research hypothesis is stated as follows:

H1. 

Companies with the readability of financial footnotes have a relationship with company audit fees.

3.1. Risk Management Committee and Audit Fee

The role of the risk management committee will provide broader risk oversight within the company (Aebi et al. 2012). RMC that operates independently will carry out its role separately from the audit committee and can work more effectively to oversee risk management (Buckby et al. 2015). RMC is a governance mechanism that monitors corporate risks and communicates with stakeholders (Nahar et al. 2016). In its supervisory role, the RMC must oversee the enterprise risk management framework through a process to identify, assess, and respond to all current and future risks to the organization’s existence (Schlich and Prybylski 2009). Establishing a risk management committee demonstrates the board’s commitment and awareness of the importance of an internal control system and good corporate governance (Cummins et al. 2009). Good corporate governance and board composition show a positive relationship with corporate responsibility (Tumwebaze et al. 2018).

Knechel and Wilekens (2006) suggest that when a firm’s level of control is subject to internal demand forces gathered from multiple stakeholders, it enhances external security. Therefore, an independent risk management committee may require the auditor to conduct a more rigorous audit and require a higher level of control within the scope of the audit. In other words, external auditors need to pay more attention to financial statements, spend more time, and may incur higher audit fees (Malik et al. 2021). While RMC does not directly purchase audit services, it may recommend a broader range of risk response services for your risk control tasks. This can lead to a higher demand for external guarantees. On the demand side, from an audit point of view, the authors argue that the establishment of a risk management committee requires a larger audit scope. Based on the description above, the hypothesis can be formulated as follows:

H2. 

There is a positive relationship between the risk management committee and the audit fee.

3.2. Relationship between Readability of Notes on Financial Statements, Risk Management Committee, and Audit Fees

Based on the agency concept, RMC plays a role in standardizing activities and monitoring various company risks (Aebi et al. 2012). RMC also acts as a governance mechanism to oversee the company and inform stakeholders about these risks (Nahar et al. 2016). Having good corporate governance in a company can increase audit demand (increase fees) or reduce auditor risk assessment (reduce costs) (Carcello et al. 2002).

On the demand side of the audit, the risk management committee is responsible for monitoring risk across the company. Because risk can be detected by providing advice to management on how to deal with it, in this case, the auditor is indirectly obliged to carry out a more detailed investigation, which increases audit fees (Deloitte 2014). RMC coordinates with the entire board of directors regarding the risks faced by various committees, especially between the audit committee and the risk committee (Deloitte 2014). The nature of the risk management committee will increase its independence in dealing with risks. They can apply better audit methods because they are not exposed to the pressure of conflict of interest with management, thereby increasing their managerial ability (Carcello and Neal 2003). We argue that a risk management committee will provide more robust risk management based on previous research. The risk management committee can share their risk assessments and identify ways to address them as part of their risk control responsibilities. This indicates that the more visible a firm’s risk, in this case, the more financial footnotes readable, the more they will respond objectively and demand more audit coverage for higher audit quality and hence associated higher audit fees.

Based on the description above, the hypothesis can be formulated as follows:

H3. 

The risk management committee strengthens the relationship between the readability of notes to financial statements and audit fees.

4. Materials and Methods

4.1. Samples and Data Sources

This study uses a population of nonfinancial companies listed on the Indonesia Stock Exchange for the period 2014–2018, as many as 758 companies with several criteria, such as this study without using the SIC (standard industrial classification) code 6 in the financial industry because they have different accounting treatment characteristics. Second, we excluded all missing control variables. After applying these criteria, the final sample consisted of 758 observations. Finally, we combine the overall continuous variables at the 1st and 99th percentiles to reduce outliers.

4.2. Operational Definition and Variable Measurement

This study uses the type of data in the form of documentary data, and the source of the data used is secondary data. This study uses panel data consisting of various companies from 2014 to 2018. The data source comes from reports from 2014 to 2018 and is equipped with the OSIRIS database. The following research variables are described in

Table 1. Variable definition.

Variable	Operational Variable	Source
Financial Footnotes Readability (GF)	Gunning Fog readability index (t − 1)	Annual report
Audit Fees (LNFEE)	The amount of the natural logarithm of audit fees paid by the company to the auditor	Annual report
Risk Management Committee (RMC)	The dummy variable is coded 1 if the company has a risk management committee and 0 if otherwise	Annual report
Independent Commissioner (DIBOC)	Total independent commissioners included in the annual report	Annual report
Total Employee (EMP)	Total employees from the level of directors to staff within the company	Annual report
Board Size (BSIZE)	The total number of the board of directors and board of commissioners	Annual report
BIG 4	Dummy variable, coded 1 if the company is audited by BIG 4 public accounting firm (EY, KPMG, PwC, Delloitte)	OSIRIS
ARL	The period days between the end of the fiscal year to the date of the audit report	Annual report
Return on Assets (ROA)	Earning before tax/total assets	OSIRIS
Firm Size (FSIZE)	Natural logarithm of the company’s total assets	OSIRIS
Leverage (LEV)	Total long-term debt divided by total assets	OSIRIS
INVREC	Inventory and accounts receivable divided by total company assets	OSIRIS

, variants used in this article are summarized in

Table 2. Derivation of samples.

Sample Criteria	Observation
Firms-Year Observations	3573
SIC 6	(911)
Missing Audit Fees Variable	(1687)
Missing Control Variable	(217)
Final Sample	758

These data are missing due to the fact that companies are not required by law to disclose audit fees in their annual reports.

, sample distribution by year in

Table 3. Sample distribution by year.

Year	Frequency	Percentage	Cumulative
2014	151	19.92	19.92
2015	157	20.71	40.63
2016	195	25.73	66.36
2017	113	14.91	81.27
2018	142	18.73	100.000
Total	758	100.00

This table shows the distribution of years in the 2014–2018 sample.

, and

Table 4. Sample distribution by industry.

SIC	Frequency	Percentage	Cumulative
(SIC 0) Agriculture, Forestry, and Fisheries	14	1.85	1.85
(SIC 1) Mining	157	20.71	22.56
(SIC 2) Manufacturing	248	32.72	55.28
(SIC 3) Construction Industries	124	16.36	71.64
(SIC 4) Transportation, Communications and Utilities	105	13.85	85.49
(SIC 5) Wholesale and Retail Trade	63	8.31	93.80
(SIC 7) Service Industries	39	5.15	98.94
(SIC 8) Health, Legal, and Educational Services and Consulting	8	1.06	100.00
Total	758	100.00

This table shows the distribution of years in the 2014–2018 sample.

sample distribution by industry.

5. Methodology

This study uses STATA 14 software, and then the analytical techniques used are descriptive statistical analysis, Pearson correlation, and moderated regression analysis with fixed effects of industry years and the combined standard error (Petersen 2009), which aim to examine the effect of the independent variable on the dependent variable and the effect of the moderating variable. We use MRA following research by Park and Yi (2022). MRA can well interpret regression in which there is a multiplication between two or more variables. This is in line with our research to see the moderating effect of RMC multiplied by audit fees. However, before that, the Winzor technique is carried out on the data used to overcome the extreme data problem that comes from the outlier effect. We used the first search model (1) to test Hypotheses 1 and 2, while our third hypothesis was tested using the second research model (2). Based on our arguments in Hypotheses 1 and 2, we expect the readability of financial footnotes and RMC to be correlated with audit fees.

GF_i,t = β₀ + β₁ LNFEE_i,t + β₂ RMC_i,t + β_3–11 CONTROL_i,t + β₁₂IFE_i,t + β₁₃YFE_i,t ε

(1)

GF_i,t = β₀ + β₁ RMC_LNFEE_i,t + β₂ LNFEE_i,t + β₃ RMC_i,t + β_4–12 CONTROL_i,t + β₁₃IFE_i,t + β₁₄YFE_i,t + ε

(2)

where:

• β₀ = Constanta
• β₁–β₁₃ = Regression coefficient
• GF = Gunning Fog readability index
• IBOC = Proportion of independent board of commissioners
• EMP = Number of employees
• BSIZE = Number of total board
• BIG4 = Audited by BIG4
• ARL = Audit report lag
• ROA = Return on asset
• FSIZE = Firm size
• LEVERAGE = Total Debt/Total Asset
• INVREC = Inventory + Receivables/Total asset
• IFE = Industry fixed effect
• YFE = Year fixed effect
• i = Firm
• t = Year

6. Result and Discussion

Descriptive statistics are presented in

Table 5. Descriptive statistics.

	Mean	Median	Minimum	Maximum
LNFEE	20.533	20.500	17.910	23.519
GF	−19.798	−19.701	−25.360	−17.684
RMC	0.175	0.000	0.000	1.000
IBOC	0.877	1.000	0.000	1.000
EMP	7.065	7.137	0.693	11.391
BSIZE	9.572	9.000	4.000	21.000
BIG4	0.474	0.000	0.000	1.000
ARL	82.225	81.000	7.000	1024.000
ROA	5.692	4.330	−22.270	51.190
FSIZE	22.105	22.048	18.461	25.243
LEV	1.334	0.882	−2.084	9.384
INVREC	0.286	0.243	0.006	0.841

This table shows descriptive statistics for all the variables used in this study. The sample used in this study amounted to 758 companies listed on the IDX from 2014 to 2018. All variables are winsorized at 1% and 99% levels.

. The readability index in this study was measured using an index in English. The lower the value of this index indicates a lower level of readability or more difficult to read. The average value of notes on financial statements in Indonesia is −19.812 using the Gunning Fog index. The audit fee in this study uses the natural logarithm of the audit fee, which has an average value of 20,537. The control variables in this study were IBOC, EMP, BSIZE, BIG4, ROA, FSIZE, LEV, and INVREC. Independent commissioners have an average score of 1654, and the highest score is 4 people in the company. The companies audited by BIG4 in this study were 46% of the total sample, which indicates that BIG4 holds almost half of the market. The average value of ROA in this study is 5682. FSIZE, or the company’s total assets in this study, have passed the natural logarithm stage with a maximum value of 25,243. Furthermore, LEV has an average value of 1354, which means that the average company in this research sample has a liability that is 1.3 times greater than its total assets.

Pearson correlations for this study are shown in

Table 6. Pearson correlations.

Panel A: From variables LNFEE to BSIZE
	[1]	[2]	[3]	[4]	[5]	[6]
[1] LNFEE	1.000
[2] GF	0.104 ***	1.000
[3] RMC [4] DIBOC	0.270 *** −0.105 ***	0.045 −0.019	1.000 −0.045	1.000
[5] EMP	0.379 ***	0.078 **	0.213 ***	0.013	1.000
[6] BSIZE	0.353 ***	0.101 ***	0.223 ***	−0.105 ***	0.520 ***	1.000
[7] BIG4	0.403 ***	0.016	0.217 ***	−0.078 **	0.359 ***	0.389 ***
[8] ARL	−0.088 **	−0.049	−0.148 ***	−0.063 *	−0.212 ***	−0.208 ***
[9] ROA	0.116 ***	0.036	0.023	0.023	0.172 ***	0.149 ***
[10] FSIZE	0.499 ***	0.095 ***	0.379 ***	−0.120 ***	0.625 ***	0.556 ***
[11] LEV	0.088 **	0.055	0.072 **	0.048	0.056	−0.005
[12] INVREC	−0.156 ***	−0.010	−0.224 ***	0.096 ***	0.027	−0.122***
Panel B: From variables BIG4 to INVREC
	[7]	[8]	[9]	[10]	[11]	[12]
[7] BIG4 [8] ARL	1.000 −0.132 ***	1.000
[9] ROA	0.264 ***	−0.229 ***	1.000
[10] FSIZE	0.400 ***	−0.155 ***	0.051	1.000
[11] LEV	−0.048	0.034	−0.159 ***	0.128 ***	1.000
[12] INVREC	−0.057	0.020	0.214 ***	−0.352 ***	−0.030	1.000

This table displays the Pearson correlation of all variables used in this study. The sample uses 758 firms on the IDX listed for the years 2014–2018. Significance is at * p < 0.1, ** p < 0.05, *** p < 0.01.

; from the results of the Pearson correlations test, the independent variable of this study (GF) correlates with audit fees (LNAFEE). Data that significantly correlate with LNAFEE are RMC, DIBOC, EMP, BSIZE, BIG4, ROA, FSIZE, LEV, and INVREC. In addition, almost all controlled variables in this study have a very positive correlation with audit fees, except DIBOC and INREC, which have a negative correlation.

6.1. Readability Financial Footnotes, Audit Fees, and RMC

Based on the results of the regression

Table 7. Moderated regression analysis.

	(1)	(2)
	LNFEE	LNFEE
RMC_GF		0.251 **
		(2.28)
GF	0.083 ***	0.053 **
	(3.26)	(2.04)
RMC	0.230 **	5.154 **
	(2.20)	(2.36)
DIBOC	−0.155 *	−0.145
	(−1.69)	(−1.57)
EMP	0.108 ***	0.111 ***
	(2.91)	(3.03)
BSIZE	0.033 **	0.032 **
	(2.43)	(2.42)
BIG4	0.694 ***	0.720 ***
	(10.29)	(10.87)
ARL	0.003 *	0.003 *
	(1.66)	(1.72)
ROA	0.006 **	0.005 *
	(1.97)	(1.72)
FSIZE	0.277 ***	0.270 ***
	(7.13)	(6.92)
LEV	0.006	0.004
	(0.32)	(0.19)
INVREC	−0.181	−0.191
	(−1.18)	(−1.26)
_cons	15.049 ***	14.617 ***
	(16.20)	(15.76)
Year FE	Included	Included
Industry FE	Included	Included
r2_a	0.574	0.578
N	758	758

This table reports the regression result of the study’s main analysis. The first column is our first regression model, and the second column is to test the interaction effect between RMC and readability of financial footnotes (GF). This test was performed after winsorizing the data for 1% and 99%; t statistics in parentheses * p < 0.1, ** p < 0.05, *** p < 0.01.

, in the first column, this study documents a positive relationship between the readability of notes to financial statements (GF) and audit fees (LNAFEE) with a coefficient of 0.083 (t = 3.26) at a significance level of 1%. This coefficient means that each company with different readability values using the gunning fox index will increase audit fees by 0.083. Furthermore, the relationship between GF and RMC in this study has a significant positive relationship with a coefficient of 0.230 (t = 2.20) at a significant level of 5%. Therefore, based on model 1, it can be believed that the proposed Hypotheses 1 and 2 can are accepted. For model 2, documenting the regression with the moderating variable in the model with the variable name RMC_GF also showed significant positive results with audit fees with a coefficient of 0.251 (t = 2.28) with a significant value of 5%, which indicates that hypothesis 3 has been accepted. Several other control variables show a significant positive relationship with audit fees, such as DIBOC, EMP, BSIZE, BIG4, ARL, ROA, and FSIZE. Each of these relationships has a coefficient of −0.155 (t = −1.69); 0.108 (t = 2.91); 0.033 (t = 2.43); 0.694 (t = 10.29); 0.003 (t = 1.66); 0.006 (t = 1.97); and 0.277 (t = 7.13) with a significance of 1% to 10%. This indicates that these variables will affect the audit fees in this study.

Table 7 column 1 shows that the readability of financial footnotes (GF) has a significant positive relationship with audit fees with a significance value of 1%. Therefore, hypothesis 1 in this study can be accepted. However, the results have a different point of view from previous research. For example, research conducted by Abernathy et al. (2019) with a sample of companies in the U.S resulted in the conclusion that with the difficulty of legibility of a report, in this case, the notes to the financial statements, the risk accepted by the auditor is due to the possibility of things being deliberately covered by management and the risk of low earnings quality, in this case. as an auditor, it will be sensitive to the risks owned by the client company, which causes an increase in audit fees. This is done because the auditor is required to conduct a more thorough audit, which will increase the time and effort to prevent misstatement and also risk could be bad for the auditor’s reputation in the future (Lo et al. 2017; Loughran and McDonald 2014) In contrast to research conducted by Salehi et al. (2020) in Iran, in this study it was found that the easier to read the notes on the financial statements had a significant positive relationship with the company’s audit fees, this was explained by two things. First, the difference in the sample carried out, and secondly, with the readability of a report, the auditors tend to have references to compare and check with their findings, which will increase the auditor’s time and effort, which leads to higher costs.

This study tries to prove the relationship between the readability of financial footnotes and audit fees with a sample of companies in Indonesia, with regression results showing a significant positive relationship, in line with research conducted by (Salehi et al. 2020). This relationship is an illustration of how important the notes to the financial statements are in the annual report so that they can affect audit fees because, basically, the financial footnotes are not separate from the financial statements, and financial statements can only be presented with short words and numbers so that the notes to the report financial statements can add important information that cannot be explained in the report (Worthington 1977).

The results of model 1 of this study also see the relationship between RMC and LNFEE, which has a positive relationship with a significant level of 5%. RMC is a component of corporate governance that plays an independent role in overseeing corporate risk (Buckby et al. 2015). Similar results were also found in a recent study by Larasati et al. (2019) and Malik et al. (2021), each with samples in Indonesia and Malaysia. With the RMC in the company, it creates a demand for the audit demand side, where management asks the auditor to increase the quality of supervision in the risk management activities of a company. Carcello and Neal (2003) concluded that independent members could increase reporting transparency because they tend to understand the organization’s standards, policies, and regulations. Although RMC cannot directly determine audit fees, they suggest a more comprehensive service, increased risk management, and human resources to external auditors as a form of risk response in their risk management role. By improving services more thoroughly, the auditor will require greater time and effort.

In Table 7 column 2, this study also examines the relationship between the readability of financial footnotes, RMC, and audit fees. This study finds out how the presence of RMC in the company as an independent committee that handles risks will impact the relationship between financial footnotes and audit fees. The results of the relationship between these variables are positive and significant, which means that the presence of RMC will strengthen the relationship between GF and audit fees. This can be explained by the supply and demand side of the auditor. From the supply side, the auditor tries to ensure that by reading the financial footnotes, the auditor will obtain more information about the financial statements, so the auditor will spend more time analyzing and comparing their results with the notes to the financial statements that can be read, which may require effort and higher audit fees. Meanwhile, from the demand side, with the existence of an independent committee such as the RMC, the company will carry out supervisory functions such as identifying, evaluating, and responding to all risks that seem to threaten the existence of the organization (Schlich and Prybylski 2009). Auditors, as external parties who conduct audits of the fairness of financial statements, will also take into account the risks that may exist in the company. With the presence of RMC in the company, they may recommend more extensive services in response to risk control assignments, which leads to an increase in demand for external assurance, and broader audit requests will affect the audit fees paid (Khan and Abdul Subhan 2019).

In this study, we try to add additional analysis by using a different readability method, the purpose of the difference in measurement is because the author wants to ensure that with different readability proxies, the results are consistent with the main regression. For this additional test, the researcher used the SMOG index to measure the readability of financial footnotes. The SMOG index (SMOG) was invented by McLaughlin (1969) to measure readability, which is generally used. The SMOG index requires that the sentences used must be more or at least 30 sentences. The result of the SMOG calculation will use t − 1 and be multiplied by minus 1, which means that if it is positively related, it will be easy to read, and vice versa if it is negatively related.

Table 8. SMOG index.

	(1)	(2)
	LNFEE	LNFEE
RMC_SMOG2		0.204 *
		(1.87)
SMOG2	0.116 ***	0.081 **
	(3.31)	(2.18)
RMC	0.232 **	3.695 **
	(2.21)	(1.97)
DIBOC	−0.153 *	−0.152 *
	(−1.67)	(−1.65)
EMP	0.108 ***	0.110 ***
	(2.91)	(3.00)
BSIZE	0.032 **	0.032 **
	(2.39)	(2.37)
BIG4	0.696 ***	0.711 ***
	(10.32)	(10.71)
ARL	0.003 *	0.003 *
	(1.68)	(1.73)
ROA	0.006 *	0.005 *
	(1.93)	(1.73)
FSIZE	0.279 ***	0.274 ***
	(7.20)	(7.08)
LEV	0.006	0.004
	(0.33)	(0.23)
INVREC	−0.187	−0.183
	(−1.22)	(−1.20)
_cons	15.343 ***	14.849 ***
	(15.83)	(15.08)
Year FE	Included	Included
Industry FE	Included	Included
r2_a	0.575	0.577
N	758	758

This table reports the regression results from our additional study. The first column is our first regression model, and the second column is to examine the effect of the interaction between RMC and readability of financial footnotes (SMOG). This test was performed after winsorizing the data for 1% and 99%; t statistics in parentheses * p < 0.1, ** p < 0.05, *** p < 0.01.

presents the regression results of the readability of financial footnotes with audit fees, the readability of financial footnotes in this table uses the SMOG index measurement to distinguish it from the main regression. The first column shows that SMOG has a significant positive relationship with audit fees with a coefficient value of 0.116 (t = 3.31) with a significance level of 1%. Then, still in the first column, it can be seen that there is a significant positive relationship between RMC and LNFEE with a coefficient value of 0.232 (t = 2.21) with a significance of 5%. In column two, it can be seen how RMC moderates the relationship between SMOG and LNFEE written with the variable RMC_SMOG. The result is a significant positive with a coefficient value of 0.204 (t = 1.87). Some of the control variables in this study have a significant positive relationship with audit fees, namely DIBOC, EMP, BSIZE, BIG4, ARL, ROA, and FSIZE.

6.2. Robustness Test

This study uses the coarsened exact matching (CEM) regression method as a regression model robustness test. However, it is also possible that the results are influenced by the observed variables included in the regression model. Therefore, this study re-examines the results using the coarsened exact matching approach to address this potential problem. As an alternative approach to dealing with the problem of self-selection, CEM focuses on the possible observed variables that could influence the outcome of the main regression. The basic idea of CEM regression analysis is to test research samples that have the same characteristics. In this case, it means that they have the same value in several variables.

CEM testing uses 10 independent variables (including control). CEM in this study divided 2 sample groups, namely treated and control groups, which were divided into three strata based on the value of each control variable in this study. The treated group is a sample group with a value above the median (<−19.79835) on the GF, while the control group is the opposite. The sample used is minimal in stratum and has at least one treated and one control unit. It will automatically compare the readability above the median and below the median with similar control variables. Therefore, the number of samples in the CEM test will be less than in the main regression.

Panel A in

Table 9. Coarsened exact matching.

Panel A
All	378	380
Matched	332	344
Unmatched	46	36
Panel B
	(1)	(2)
	LNFEE	LNFEE
RMC_GF		0.230 **
		(2.04)
GF	0.078 ***	0.050 *
	(3.02)	(1.86)
RMC	0.250 **	4.761 **
	(2.24)	(2.13)
BIG4	−0.234 **	−0.224 **
	(−2.24)	(−2.13)
DIBOC	0.121 ***	0.123 ***
	(3.04)	(3.13)
EMP	0.030 **	0.030 **
	(2.17)	(2.18)
BSIZE	0.626 ***	0.655 ***
	(8.64)	(9.17)
ROA	0.003 *	0.003 *
	(1.77)	(1.76)
FSIZE	0.008 ***	0.007 **
	(2.69)	(2.49)
LEV	0.297 ***	0.290 ***
	(7.57)	(7.34)
INVREC	0.004	0.001
	(0.15)	(0.05)
_cons	14.613 ***	14.220 ***
	(15.48)	(15.05)
Year FE	Included	Included
Industry FE	Included	Included
r2_a	0.606	0.609
N	658	658

t statistics in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.

presents the matching CEM summary. A total of 332 out of 378 connected observations were matched with 344 out of 380 unconnected observations. In panel B, the sample studied in this test was reduced to 658 from the previous 758. The results of the CEM test on the GF and LNFEE variables had a significant positive relationship with a coefficient of 0.078 (t = 3.02) with a significance level of 1%. Furthermore, for testing the relationship between the variables RMC and LNFEE also has a significant positive relationship with a coefficient of 0.250 (t = 2.24) with a significance of 5%, and finally to see the resilience of the moderating relationship with the variable name RMC_GF has a significant positive result with a coefficient of 0.230 (t = 2.04) with a significance level of 5%. From the results of Table 9, it can be seen that all the results of observations of both independent, dependent, and moderating relationships have results that are in line with the main regression. For the relationship between GF and LNFEE, the significance remains 1%, and the RMC relationship with LNFEE remains at 5% in the CEM regression model. This shows that all hypotheses in this study still have the same results, namely GF is related to LNFEE, RMC is related to LNFEE, and RMC strengthens the relationship between GF and LNFEE the same as the main regression results, which indicates the results of this test are robust.

7. Conclusions

This study has a sample of 758 companies for 5 years (2014–2018) without using the financial industry (SIC 6) due to differences in recognizing assets. The sample from this study was significantly reduced because many companies did not disclose information related to audit fees. This study aimed to examine the relationship between the readability of notes to financial statements and company audit fees. This study proposes a directionless hypothesis due to differences of opinion regarding the readability of the notes on the financial statements with audit fees. The result of this research hypothesis is that the easier it is to read the notes on the financial statements, the greater the audit fees paid by the company. In line with the research by Salehi et al. (2020) with Iran setting in 2012–2017 that with the legibility of the notes on the financial statements, the auditor will obtain more information related to the financial statements; therefore, the auditor will examine more deeply related to the information contained in the financial footnotes so that audit risk can be minimized, but the time and effort expended will increase and result in an increase in audit fees.

Then the second hypothesis in this study wants to see how the risk management committee (RMC) relates to audit fees. RMC is chosen as a form of management responsibility to oversee company risk, one of which is asymmetric financial statement information. RMC in this study has a significant positive result with audit fees, Also in line with the research by Malik et al. (2021) by adopting 208 nonfinancial listed firms in Bursa Malaysia for the year-end 2014 and research by (Rahayu et al. 2021) in Indonesia setting, which means that the presence of RMC in a company will increase the demand for audits due to RMC’s demand to ensure that there are no misstatements, which causes an increase in the number of audit coverage and increases the audit time to be completed and directly correlated with audit fees. To complete this research, the researcher tries to moderate RMC with the relationship between finance footnotes readability and audit fees. In the theory of supply and demand audit, financial footnotes readability and audit fees act as supply, while the relationship between RMC and audit fees will act as demand. Therefore, the third hypothesis in this study is also accepted: RMC will strengthen the relationship between financial footnotes readability and audit fees.

The results of this study have implications or contribute both theoretically and practically. This study uses agency theory where there are differences in the interests of the agent and principal, then the possibility of information asymmetry in financial statements and other risks. The readability of financial footnotes will tend to add information to the auditor, and from the supply side, the auditor will match the information so that the risk of misstatement can be reduced, while the relationship between RMC and audit fees is on the demand side, where RMC will indirectly ask the auditor to conduct an audit regularly more broadly. Both the supply and demand sides will increase audit fees because the auditor’s effort and time will increase. From a practical point of view, investors, potential investors, and others pay attention to the company’s quantitative information and qualitative information. For auditors, this study encourages auditors to view financial footnotes as information or as a risk threat in audit assignment risk; the presence or absence of a risk management committee within the company will play an important role in the workload of auditors. For management, to reduce the risk of misstatements that may occur and go undetected, the legibility of financial records can be something that can be improved again. At the same time, the presence of RMC in the company’s organization will help management to monitor the risks that occur in the company.

This study also has several limitations. First, related to our total sample of 758 companies only, it is because not all public companies in Indonesia disclose their audit fees in annual reports. Second, the legibility of the notes on the company’s financial statements is sometimes in an image format, making it difficult for authors to use the data accurately. Furthermore, the notes on the financial statements are carried out only by measuring the legibility of the gunning fox and SMOG. Third, we use five years of data from 2014 to 2018; we cut off 2019 onwards to avoid any discrepancies in financial statement performance and other segments such as financial footnotes due to the COVID-19 pandemic, which will be handled differently from other years. Suggestions for further research to improve the quality of research data from readability data, audit fee data, and other measurements of good corporate governance other than RMC. Then, the next researcher can use other measurements of legibility outside the gunning fox and SMOG index. Finally, researchers can use other audit measures such as audit report lag, specialization, audit quality, etc.