The effect of audit committee characteristics on managements’ IFRS 7 risk disclosure quality in UK companies

(1)

The effect of audit committee characteristics on

managements’ IFRS 7 risk disclosure quality in

UK companies

By Mike Jonkman1

Master Thesis for the MSc Accountancy

Supervisor: dr. Y. Karaibrahimoglu Date: 21-06-2020

Abstract

This research examines the effect of audit committee characteristics on managements’ IFRS 7 disclosure quality. Furthermore, the moderating role of CEO power in this relationship is examined. By addressing the role of the audit committee and CEO power, an important gap in the current IFRS 7 risk disclosure literature will be filled. Using panel data from 2010-2016 for companies in the United Kingdom, this study provides evidence that certain audit committee characteristics affect the quality of management’s IFRS 7 risk disclosures. In addition, this study provides evidence that CEO power can moderate the relationship between audit committee characteristics and management’s IFRS 7 risk disclosure quality.

Keywords: IFRS 7, risk disclosures, audit committee, CEO power Wordcount: 7.857

1_Email:_{m.h.m.jonkman@student.rug.nl}_{student number: S3189635 course: Master’s Thesis Accountancy} phone: 0643113618

(2)

2

1. Introduction

Financial statements have always been the main source for creditors, market analysts and investors to accurately evaluate the financial health and future profitability of companies. In the past however, financial statements weren’t transparent enough to inform the users of financial statements about the entity’s exposure to risks on the market and how those risks were managed by the company (Muthupandian, 2008). The structure of an entity’s financial instruments can be very complex and therefore, it can be difficult to obscure their true

financial health. This led to a growing demand for quantitative and qualitative risk disclosures (Elshandidy, Shrives, Bamber, and Abraham, 2018).

There already were some regulations in place that covered the topic of financial instruments, namely IAS30 and IAS32. IAS30 covered the topic of disclosures in the financial statements of banks and similar financial institutions, while IAS32 covered the topic of the presentation of financial instruments. However, there was a demand for more and better disclosures and also for disclosures in non-financial firms that where not captured by IAS30. This growing demand eventually led to the introduction of IFRS 7 by the International Accounting Standards Board (IASB), since stakeholders needed more information regarding risks of entities with financial instruments. IFRS 7 is intended for all entities with financial

instruments (financial and non-financial entities) and it requires them to provide disclosures on: “(1) the significance of financial instruments for the entity’s financial position and

performance and (2) the nature and extent of risks arising from financial instruments to which the entity is exposed during the period and at the end of the reporting period, and how the entity manages those risks” (IFRS 7). This way, stakeholders get a better overview of the risks an entity is bearing regarding their financial instruments and how they manage those risks. This is in line with the agency theory, since previous research has shown that the quality of mandatory risk disclosures significantly mitigates information asymmetry between

managers and investors (Miihkinen, 2013). For this reason, the focus in this research will be on IFRS 7 risk disclosure quality.

Not much later, in 2007, the big financial crisis happened due to excessive risk taking, incorrect pricing of risk, easy credit conditions and subprime lending by banks and other organizations. The question that remained was: How could this happen and how do we make sure something like this doesn’t happen again? This question led to an even growing demand for quantitative and qualitative risk disclosures and a big revision of IFRS 7 in 2009. This

(4)

4

shows that the introduction and implementation of the first IFRS 7 draft in 2005 was a timely event given the big financial crisis that happened just a few years later.

Up until this day, a lot of research has been done on the quality of management’s risk disclosures. Research has been done on the determinants and incentives for management’s risk disclosure (Marshall & Weetman, 2007; Ntim et al., 2013), the informativeness of management’s risk disclosures for users of the financial statements (Bao & Datta. 2014; Elbannan and Elbannan, 2015), as well as on both these topics combined (Hope et al., 2016). An interesting research, is the research of Barakat and Hussainey (2013). They investigated the effect of governance, regulation and supervision on the quality of risk reporting of

European banks. However, in their conclusion they mention that still little is known about the effect of audit committee characteristics on management’s risk reporting quality. They investigated the effect of audit committee activity, but only on the quality of operational risk disclosures and only of European banks. So a lot of research has been done on the effect of corporate governance characteristics on risk disclosure quality, but little is known about the effect of audit committee characteristics. For this reason, there is still a big gap in the risk disclosure literature regarding the effect of audit committee characteristics on management’s risk reporting quality that should be researched. Therefore, this study aims to examine the effects of audit committee characteristics on management’s risk disclosure quality.

Moreover, this study also aims to examine the moderating role of CEO power on the relation between audit committee characteristics and management’s risk disclosure quality. In the past, research has shown that more powerful CEOs affect the effectiveness of audit committees and their oversight quality (Bruynseels & Cardinaels, 2014; Lisic et al., 2016). Audit committees can become less effective, which results in lower oversight quality. This way the audit committee might be unable to reduce the information asymmetry between management and the users of the risk disclosures. For this reason, the moderating effect of CEO power on the relationship between audit committee characteristics and managements’ risk disclosure quality will be researched as an additional sub-question.

This study makes various contributions to the extant literature. Firstly, when investigating the effect of audit committee characteristics on management’s risk disclosure quality, an existing gap in the literature will be filled. To the best of my knowledg e, this is the first research that will address the impact of audit committee characteristics on this topic. Recent work in accounting examines the effects of audit committee characteristics on

(5)

5

earnings management (Sun, Lan & Liu, 2014), firm performance (Aldamen et al., 2012) and accounting conservatism (Sultana, 2015). Closely related to the topic of this paper, research has been done on the effect of audit committee characteristics on voluntary risk management disclosure (Abdullah et al., 2017) and on operational risk disclosures (Barakat & Hussainey, 2013). This research paper will add to prior literature by considering the effect of audit committee characteristics on management’s IFRS 7 risk disclosures. Secondly, by considering the moderating effect of CEO power on the relationship between audit committee characteristics and management’s risk disclosures, this research will add an extra dimension to the current corporate governance and risk disclosure literature. Besides the contribution to the literature, it will also have a practical contribution. The results of this research can be used by market participants in practice. If a significant effect was to be found in this research, it will be possible for companies, regulators or supervisors to use those results to improve the quality of future management’s risk disclosures.

2 Theoretical explanation and Hypothesis development

2.1 Institutional theory

In order to explain the relationship between audit committee characteristics and

management’s risk disclosure quality, I ground my arguments on the agency theory and the stakeholder theory.

The agency theory states that within organizations, a principle-agent problem arises (Eisenhardt, 1989). The agency problem arises, because there is a separation between ownership and control. Because of this separation, the goals and desires of the principal and the agent conflict with each other and it is difficult and expensive for the principal to verify what the agent is doing. There is information asymmetry between the principal and the agent. In this metaphor, the agent is the management and the principal represents the shareholders. Because of this information asymmetry, the management can act in their best interest, instead of the interest of the shareholders. Research has shown that high quality risk reporting will reduce this information asymmetry between management and shareholders, since they disclose more to the public (Elzahar & Hussainey, 2012).

Where the agency theory solely looks at the shareholders, the stakeholder theory has a broader perspective. Freeman (1984), who was the first one to really conceptualize the stakeholder theory in his book, defines a stakeholder as ’any group or individual who can affect or is affected by the achievement of the organization’s objectives’. So where the agency theory

(6)

6

looks at the principal-agent relationship, the stakeholder theory expands this by considering the dynamic and complex relationship between the organization and its environment (Amran et al., 2008). This is why according to the stakeholder theory, entities might disclose

information to more effectively interact and better communicate with influential stakeholders (Barakat & Hussainey, 2013).

2.2 Audit committee

The audit committee is one of the major operating committees of a company’s board of directors and it is their role to assist the board of directors in overseeing the financial reporting and disclosure process (Sultana, 2015).

A competent, committed, independent, and tough-minded audit committee has been described as “one of the most reliable guardians of the public interest” (Levitt, 2000). In general, all listed entities around the world are required to have an audit committee within their

organization. Jurisdiction differs all around the world, but the core DNA of audit committees is essentially the same everywhere (KPMG, 2017). In general, most jurisdictions prescribe that the audit committee consists of at least three non- executive directors. Those three directors should be independent and most regulations and corporate governance codes prescribe that at least one member of the audit committee should possess a certain degree of competence in finance, accounting and/or auditing or overseeing the performance of a company.

Since the audit committee has the power to influence the financial reporting and disclosure practices, the agency theory and the stakeholder theory are good theories to explain the relationship between audit committee characteristics and risk disclosure quality. Audit committees are responsible for overseeing the financial reporting and disclosure process (Sultana, 2015) and are therefore likely to influence the amount of information asymmetry and management’s disclosure quality. The agency theory explains the monitoring part of the audit committee that might reduce information asymmetry and the stakeholder theory expands this by shifting the scope from shareholders to all stakeholders interested in the disclosures of companies.

The audit committee characteristics that will be used in this research are derived from previous studies on audit committees (Sultana, 2015; Sultana et al., 2015). These characteristics are audit committee financial expertise, audit committee size and audit committee tenure.

(7)

7 2.2.1 Audit committee financial expertise

The first audit committee characteristic that will be discussed is the audit committee financial expertise. A financial expert in an audit committee can be defined as someone with

experience in accounting, supervising financial professionals and overseeing the performance of a company (Mustafa & Youssef, 2010; Sultana, 2015). According to previous research, audit committees that possess a greater amount of financial expertise are considered more effective (Krishnan & Visvanathan, 2008). They are better able to effectively perform their monitoring function. This increased monitoring ability can lead to less information

asymmetry, since the audit committee will observe managerial opportunism quicker and are able to act on it adequately. Furthermore, they are better able to judge if the interests of all other important stakeholders are dealt with the right way. Such as the users of financial statements that want information about the disclosures on financial instruments. This leads to the first hypothesis:

H1: Financial expertise of audit committee members is positively associated with management’s risk disclosure quality.

2.2.2 Audit committee size

The second audit committee characteristic that will be discussed is the size of the audit committee. Audit committee size will be defined as the number of persons that make up the committee. As mentioned before, most jurisdictions prescribe that the audit committee consists of at least three non-executive members. Agency theory suggests that bigger audit committees are prone to less cohesion and are less dynamic than smaller audit committees (Hillman & Dalziel, 2003; Sultana et al., 2015). They argue that increasing the size of the audit committee will lead to less control and a decreasing ability to effectively monitor for example management or auditors. Because of the increased size, collective decision making can become harder and less effective. This way they will also lose track of the demands of certain stakeholders, such as the users of their financial statements for information about risk disclosures on financial instruments. The above mentioned problems can therefore lead to more information asymmetry between the company and stakeholders and also generate opportunistic behavior. This leads to the second hypothesis:

(8)

8 2.2.3 Audit committee tenure

The last audit committee characteristic that will be discussed is the tenure of audit committee members. Audit committee tenure will be defined as the number of years a director is active as a member of the audit committee. According to prior research, directors that serve on the audit committee for a longer time, acquire greater knowledge and experience about the firm that helps them in performing their role (Sharma & Iselin, 2012). This greater knowledge and experience helps them in better performing their role in the oversight process. In line with the agency theory, an audit committee member with a longer tenure is better able to reduce the information asymmetry since their monitoring ability is enhanced. Sharma & Iselin (2012) also mention that a longer audit committee tenure can lead to the fact that audit committee members become friends with the management. This can lead to a lower oversight quality and therefore higher information asymmetry. However, according to the extant literature, the increased knowledge and experience outweighs the downsides of longer audit committee tenure (Chan, Liu & Sun, 2013). This leads to the third hypothesis:

H3: Audit committee tenure is positively associated with management’s risk disclosure quality.

2.3 The moderating role of CEO power

In addition to the relationship between audit committee characteristics and risk disclosure quality, the agency theory and stakeholder theory can also be used to fundament the moderating effect of CEO power. Some CEOs might benefit from reducing the monitoring effectiveness of audit committees, since it can help them to extract for example more rents (Lisic et al., 2016). The increased information asymmetry that arises between providers of capital and the company benefits the CEO. This is in line with the agency problem, which stated that there are conflicting goals between the owners and directors of a company. The stakeholder theory expands this once again by looking at other stakeholders such as providers of loans. For this reason, the two theories seem suited to explain the moderating effect of a CEO’s power on the relationship between audit committee characteristics and

managements’ risk disclosure quality.

As mentioned above, more powerful CEOs might benefit from reducing the monitoring effectiveness of audit committees. Reducing the effectiveness of audit committees gives the CEO the possibility to extract more rents or behave more opportunistically (Lisic et al., 2016). A less effective audit committee can lead to more information asymmetry. Furthermore, when

(9)

9

the CEO has more power over the audit committee, the CEO might try to make sure not every risk about the financial instruments is disclosed if this benefits him or her. Or that only those things are disclosed that makes the company look healthy. We therefore expect to see a moderating effect of CEO power on the relationship between audit committee characteristics and management’s risk disclosure quality. It is expected that audit committee characteristics are less related to risk disclosure quality if a powerful CEO is involved. This leads to the last hypotheses:

H4: The positive relationship between financial expertise of audit committee members and management’s risk disclosure quality is reduced when CEO power is high.

H5: The negative relationship between audit committee size and management’s risk disclosure quality is strengthened when CEO power is high.

H6: The positive relationship between audit committee tenure and management’s risk disclosure quality is reduced when CEO power is high.

3 Method

3.1 Sample and Statistical Models

This research made use of archival data that was obtained from different databases. To test the effect of audit committee characteristics on managements’ risk disclosure quality, a sample of UK non-financial premium listed companies between the years 2010 to 2016 is used for which data on risk disclosure quality could be retrieved. This led to an initial sample of 2842 firm-year observations from 406 companies. As part of the project Karaibrahimoglu and Porumb2, data on risk disclosure quality was hand collected using the notes to the financial statements in the companies’ annual reports. Data on audit committee characteristics and CEO power were retrieved from BoardEx. Data on the control variables were retrieved from

CompuStat. Both BoardEx and CompuStat could be accessed using the Wharton Research Data Services. I eliminated observations with missing information on risk disclosure quality, audit committee characteristics or CEO Power. This led to a final sample of 850 firm-year observations.

2_{IASB project by Karaibrahimoglu and Porumb:} https://www.ifrs.org/news-and-events/2019/10/academic-research-input-to-the-iasb-work/

(10)

10

Research Model 1

To address the first research question in the paper, whether audit committee characteristics influences managements’ risk disclosure quality, the following equation is developed:

IFRS7_Disclosure_Qualityi,t = β0 + β1AC_Characteristicsi,t + β2ROAi,t + β3Leveragei,t +

β4Firm_Sizei,t + Year_Di,t + Industry_Di,t + ε

Where IFRS7_Disclosure_Qualityi,t is defined as a quality index measuring the extent to

which firm’s disclosures are in line with the best practices described by IFRS 7. This

definition in the same as used in the project Karaibrahimoglu and Porumb mentioned before. More details about the measurement of IFRS7_Disclosure_Qualityi,t will be given in section

3.2.

In Model 1, the variable of interest is AC_Characteristicsi,t. To test H1, H2 and H3, I used (1)

AC_FinancialExpertisei,t, (2) AC_Sizei,t, and (3) AC_Tenurei,t, respectively.

I expect AC_FinancialExpertisei,t to have a positive and significant coefficient. Similarly, the

coefficient of AC_Sizei,t is expected to be negative and significant. Lastly, I expect

AC_Tenurei,t to have a positive and significant coefficient.

Research Model 2

In the second research model, the moderating effect of CEO power on the relationship

between audit committee characteristics and risk disclosure quality is tested. This leads to the following equation:

IFRS7_Disclosure_Qualityi,t = β0 + β1AC_Characteristicsi,t + β2 CEO_Poweri,t + β3

AC_Characteristicsi,t * CEO_Poweri,t + β4ROAi,t + β5Leveragei,t + β6Firm_Sizei,t +

Year_Di,t + Industry_Di,t + ε

The variable of interest in Model 2 is the interaction term AC_Characteristicsi,t *

CEO_Poweri,t. Similar to Model 1, for H4, H5, and H6, I replace AC_Characteristicsi,t with

(1) AC_FinancialExpertisei,t,, (2) AC_Sizei,t, and (3) AC_Tenurei,t, respectively. In H4, I

expect the coefficient for AC_FinancialExpertisei,t * CEO_Poweri,t to be negative and

significant. Similarly, the coefficient of AC_Sizei,t *CEO_Poweri,t is expected to be negative

and significant. Lastly, I also expect the interaction term AC_Tenurei,t *CEO_Poweri,t to be

negative and significant.

In both research models, IFRS7_Disclosure_Qualityi,t has been used as the dependent

(11)

11

separate variables: (1) IFRS7_Disclosure_Quality_Assurance, (2)

IFRS7_Disclosure_Quality_Risk, and (3) IFRS7_Disclosure_Quality_Other. Both in research model 1 and in research model 2, alternatively to IFRS7_Disclosure_Quality,assurance (IFRS7_Disclosure_Quality_Assurance) and risk (IFRS7_Disclosure_Quality_Risk) quality will be used separately as additional tests. The “Other” component is left disregarded. This way, a distinction between the assurance component and the risk component of risk quality is made. Users of the disclosed information may react different to the different components of IFRS 7 disclosure quality.

As I already mentioned in the introduction section, IFRS 7 disclosures can comprise of (1) information about the amount and nature of risks associated with a firms’ financial

instrument, as well as (2) how the firms’ management addresses and manages those risks and uncertainties. This is where the distinction between the risk component and the assurance component becomes clear. The first part concerns disclosures about risk exposure

(IFRS7_Disclosure_Quality_Risk), where the second part concerns the assurance part of IFRS 7 disclosures (IFRS7_Disclosure_Quality_Assurance). For risk exposure, literature suggests that risk disclosures are associated with negative information an can increase the risk perceptions of users of the disclosed information (Kravet and Muslu, 2013). On the other hand, disclosures regarding assurance are found to be able to lower the users’ risk perception and are associated with positive information (Hodder et al. 2001). This is why users of IFRS 7 disclosures might react different to the separate components of IFRS 7 disclosure quality. Therefore, the additional tests will be performed to see what the different outcomes of the regressions will be.

All variables used in the research models, as well as their descriptions, can be found in Appendix A1.

3.2 Measurement of Dependent Variable

To measure IFRS7_Disclosure_Quality in annual reports of banks, content analysis was used. Content analysis is a research tool that involves classifying text units into categories (Beattie, McInnes & Fearnley, 2004). For this research, a disclosure index has been used to measure the quality of risk disclosures. This disclosure index consists of 13 items that are most likely used by the users of the annual report in order to consider if firms have high quality IFRS 7 risk disclosures. Those 13 items cover points like the disclosure of risk exposure, the amount of hedge activities and the presence of sensitivity analyses. As mentioned before, this data

(12)

12

was hand collected. The data that had to be collected was subdivided among several students of the Accountancy department of the Rijksuniversiteit Groningen. Every student had to collect a part of the data. In the end, this led to a complete dataset in a less time consuming way. The companies could score a maximum amount of points on each item based on

prescribed criteria and requirements. Each item comes with a description on how to assign the points. This way the subjectivity amongst the different students was limited to a minimal level which resulted in higher quality data. In the end, IFRS7_Disclosure_Quality is constructed as the normalized value of the equal weighted average of the 13 different items. The higher the value, the higher the quality of managements’ risk disclosures. The full disclosure index can be found in Appendix A2. Furthermore, the composition of the different IFRS 7 quality components used in the additional regressions can be found in Appendix A3.

3.3 Measurement of Independent Variables

As mentioned before, we distinguish between two kinds of independent variables. Audit committee characteristics and CEO power.

3.3.1 Audit Committee Characteristics

In order to measure the three different audit committee characteristics, I made use of the proxy measures that have also been used in the research of Sultana (2015) and the research of Sultana et. al (2015).

AC_Size is measured as the number of persons that make up the audit committee in each year. AC_Tenure is measured by dividing the sum of all individual member tenures by the amount of audit committee members. In other words, the average tenure of all audit committee members. AC_FinancialExpertise is measured in a different way than it has been done in the research of Sultana (2015). An audit committee member will be considered a financial expert if he or she either (1) is considered a financial expert in the BoardEx database or (2) has functional experience related to being a member of the audit committee, which can also be found in the BoardEx database. This functional experience can involve a former role as CEO, CFO or other related financial function. This is in line with the “supervising financial

professionals and overseeing the performance of a company“ part of the definition of a financial expert as has been mentioned in the theory section. After this has been done, for each firm-year I measure the percentage of financial experts in the audit committee.

(13)

13 3.3.2. CEO Power

In order to measure CEO_Power, we make use of four different variables that are commonly used in the extant literature to measure CEO power. Just like Finkelstein (1992), I use

CEO_Compensation and the number of qualifications (CEO_Qualifications) to measure the amount of structural power the CEO possesses. Next to this and in line with the research of Combs et al. (2007), we use the CEO’s board tenure (CEO_BoardTenure) as another measure for CEO power. The last variable that was added to measure CEO power, is

CEO_NetworkSize. From those four variables, an index was created. This index is created as the normalized value of the equal weighted average of the 4 different items. Details on the composition of this index can be found in Appendix A4.

3.4 Measurement of the Control Variables

Apart from the mentioned variables above, it is important to make use of control variables. This way we can better establish the relationship between audit committee characteristics and the quality of risk disclosures. These control variables are separated in firm characteristics and fixed effects.

3.4.1. Firm-Specific Controls

Research from Khlif and Hussainey showed (2016) that there is a positive relationship between firm size and the quality of risk disclosures. Next to firm size, they also found a positive relationship between leverage ratio and the quality of risk disclosures. Therefore, firm size and leverage ratio are two variables that can be used as control variables in this research. To measure Firm_Size, the natural logarithm of the total assets of the company will be used and Leverage will be measured as the ratio of total liabilities to common equity.

Furthermore, I control for firm performance by using ROA.

3.4.2. Fixed Effects

Apart from the control variables that can be found above, I also need to control for fixed effects. For this reason, I added two sorts of dummies. First, six different year-dummies (Year_D) have been added to control for the year-effects between 2010 – 2016. Next to this, I added eight different industry-dummies (Industry_D) to control for industry-effects. Details on the different industries present in this research can be found in appendix A5.

(14)

14

4 Results

In this chapter, the results will be discussed. First, the descriptive statistics will be presented, including the correlations. After that, the results of the main analysis will be presented. To test the three different research models mentioned in the method section, Ordinary Least Square Regression will be used for all the models. In addition, robust standard errors will be used to make sure the standard errors will be less biased.

4.1 Descriptive statistics and bivariate analyses

Table 1 presents the descriptive statistics of the final sample over the years 2010-2016. The table contains the number of observations, means, standard deviation, minimum and

maximum for all the variables used. The average IFRS7_Disclosure_Quality is 0.367. This indicates that among the 13 items that are important for users of IFRS7 disclosures, on average; firms comply by around 37% of the disclosures. The standard deviation for disclosure quality is 16.8%. The minimum and maximum are respectively 0% and 93.6%. This means that there is no firm that fully complies to the demand of the disclosures.

However, there are firms that do not even comply a little bit at all. These are companies that score 0%. When we consider the different components IFRS7_Disclosure_Quality_Risk and IFRS7_Disclosure_Quality_Assurance, we see the statistics are quite similar to the overall disclosure quality. Their means are 0.418 and 0.325 respectively.

IFRS7_Disclosure_Quality_Risk has a maximum of 1, which means that there is a firm that fully complies to the expected risk disclosures. In Appendix A6, a graph can be found that shows the different average disclosure scores over time. A slight increase over time can be seen. The average AC_Size is 3.480, with a minimum of one and a maximum of 8. The standard deviation for AC_Size is 0.957. The average AC_Tenure of an audit committee member is 3.454 years with a standard deviation of 1.753. When we consider

AC_FinancialExpertise, we see that on average 28.3% of the members are a financial expert with a standard deviation of 0.217. Furthermore, some audit committees consist of only financial experts and some audit committees exist without a financial expert. The CEOs have an average network size (CEO_NetworkSize) of 1080.204 with a large standard deviation of 1462.629. They have an average CEO_BoardTenure of 9.444 years with a minimum of 0.1 and a maximum of 37.9 years. Their average number of qualifications (CEO_Qualifications) is 1.827 with a minimum of 0 and a maximum of 6. For CEO_Compensation, the average is 1383.209 with a high standard deviation of 1046.246. The mean CEO_Power is 0.223, which indicates that on a scale from 0-1 the average CEO power in this sample is r low, with a

(15)

15

minimum of 0.021 and a maximum of 0.688. The average Firm_Size consists of a mean total assets of 3647.909 and the average Leverage is 1.463. Lastly, the average Return on Assets (ROA) is 0.050.

Table 1

Descriptive Statistics

Variable Obs Mean Std. Dev. Min Max

IFRS7_Disclosure_Quality 850 0.367 0.168 0 0.936 IFRS7_Disclosure_Quality_Risk 850 0.418 0.265 0 1 IFRS7_Disclosure_Quality_Assurance 850 0.325 0.243 0 0.917 AC_Size 850 3.48 0.957 1 8 AC_Tenure 850 3.454 1.753 0 12 AC_FinancialExpertise 850 0.283 0.217 0 1 CEO_Power 850 0.223 0.101 0.021 0.688 CEO_NetworkSize 850 1080.204 1462.629 6 5866 CEO_BoardTenure 850 9.444 6.939 0.1 37.9 CEO_Qualifications 850 1.827 1.174 0 6 CEO_Compensation 850 1383.209 1046.246 22 8685 Firm_Size 850 3647.909 10762.065 10.352 151220 Leverage 850 1.463 16.483 -77.546 60.152 ROA 850 0.05 0.103 -1.324 1.049

Table 2

Pearson Correlations

Panel A: Correlations Variables IFR7_Disclosure_Quality to AC_FinancialExpertise

Variables (1) (2) (3) (4) (5) (6) (1) IFRS7_Discl_Quality 1.000 (2) IFRS7_Discl_Q_Risk 0.624*** 1.000 (3) IFRS7_Discl_Q_Ass. 0.802*** 0.334*** 1.000 (4) AC_Size 0.143*** 0.019 0.122*** 1.000 (5) AC_Tenure -0.054 -0.004 0.027 -0.081** 1.000 (6) AC_Fin.Expertise 0.047 -0.092*** 0.026 -0.007 -0.150*** 1.000 (7) CEO_Power 0.128*** -0.025 0.141*** 0.218*** -0.053 0.010 (8) CEO_NetworkSize 0.146*** -0.004 0.135*** 0.214*** -0.080** 0.028 (9) CEO_BoardTenure -0.081** 0.001 -0.018 -0.117*** 0.037 -0.183*** (10) CEO_Qualifications 0.033 -0.097*** 0.049 0.098*** -0.006 0.094*** (11) CEO_Compensation 0.195*** 0.081** 0.142*** 0.309*** -0.062* 0.103*** (12) Firm_Size 0.186*** 0.115*** 0.148*** 0.266*** -0.017 0.092*** (13) Leverage 0.037 0.034 0.062* 0.034 0.021 -0.054 (14) ROA 0.037 0.048 -0.005 0.109*** -0.075** -0.032

(16)

16

Table 2 (Continued)

Panel B: Correlations Variables CEO_Power to ROA

Variables (7) (8) (9) (10) (11) (12) (13) (14) (7) CEO_Power 1.000 (8) CEO_NetworkSize 0.704*** 1.000 (9) CEO_BoardTenure 0.328*** -0.169*** 1.000 (10) CEO_Qualifications 0.617*** 0.208*** 0.004 1.000 (11) CEO_Compensation 0.403*** 0.216*** -0.077** 0.015 1.000 (12) Firm_Size 0.189*** 0.176*** -0.119*** 0.033 0.400*** 1.000 (13) Leverage 0.015 0.034 -0.007 -0.018 0.021 0.018 1.000 (14) ROA -0.015 -0.062* 0.033 -0.084** 0.163*** -0.005 0.030 1.000

*, **, *** Significant at the 0.1, 0.05 and 0.01 level, respectively

Table 2 presents the Pearson correlations between the different disclosure quality measures, the audit committee characteristics, the CEO power characteristics and the control variables. In line with my hypothesis, there is indeed a significant correlation between

IFRS7_Disclosure_Quality and AC_Size. However, in our hypothesis we expected a negative relationship and in the Pearson correlation matrix we find a positive correlation. Furthermore, the Pearson correlation matrix shows a positive significant correlation between AC_Size and IFRS7_Disclosure_Quality_Assurance. For AC_FinancialExpertise, we find a negative significant correlation with IFRS7_Disclosure_Quality_Risk. As expected, a significant correlation exists between Firm_Size and IFRS7_Disclosure_Quality. For the moderating variable CEO_Power, an interesting significant positive correlation is found with

IFRS7_Disclosure_Quality, IFRS7_Disclosure_Quality_Assurance and AC_Size.

The correlation matrix shows no signs of multicollinearity. A correlation between predictor variables of more than 0.700 or less than -0.700 is considered as multicollinearity. As can be seen, the correlation between Firm_Size and CEO_Compensation is the one that comes the closest with 0.400, but is still way below 0.700. When considering the correlation between IFRS7_Disclosure_Quality_Assurance and IFRS7_Disclosure_Quality, a correlation of 0.802 can be seen. However, this is not between two predictor variables, but between two dependent variables that will never be in the same regression model. Therefore, this is also no sign of multicollinearity.

Additionally, I performed a variance inflator factor (VIF) test to check whether

multicollinearity is present between the variables. When the VIF is 10 or higher, it means multicollinearity is present and that it is problematic. I performed a VIF test on all regressions

(17)

17

I ran and the highest VIF of all regression models was 1.33 . This, in combination with the correlation values, means that multicollinearity isn’t an issue in this research.

4.2 Main Analysis and findings

Table 3 presents the results of the regression analyses between AC_Characteristics and IFRS7_Disclosure_Quality. This reflects research model 1 in the method section. Table 3 consists of four different models. First, each different audit committee characteristic is tested individually and in model 4 all characteristics will be combined in one model. As table 3 shows, there can only be found one significant relationship between an audit committee characteristic and disclosure quality. This is for AC_Size. A small significant positive

relationship is present. This indicates that a larger audit committee leads to a higher quality of IFRS7 disclosures. This is in contrast with hypothesis 2. In our hypothesis, we expected a negative relationship between AC_Size and IFRS7_Disclosure_Quality.

Table 3

Regression results Research Model 1 Dependent Variable:

IFRS7_Disclosure_Quality

(1) (2) (3) (4)

Model 1 Model 2 Model 3 Model 4

Constant 0.489*** 0.496*** 0.491*** 0.501*** (0.067) (0.066) (0.066) (0.065) AC_FinancialExpertise 0.002 0.002 (0.006) (0.006) AC_Size 0.016*** 0.016*** (0.006) (0.006) AC_Tenure -0.008 -0.007 (0.006) (0.006) Firm_Size 0.035*** 0.031*** 0.035*** 0.030*** (0.005) (0.005) (0.005) (0.005) ROA 0.003 0.001 0.003 0.001 (0.005) (0.005) (0.005) (0.005) Leverage 0.004 0.004 0.005 0.004 (0.003) (0.003) (0.003) (0.003)

Industry-Dummies Included Included Included Included

Year-Dummies Included Included Included Included

Observations 850 850 850 850

R-squared 0.082 0.090 0.084 0.092

Highest VIF 1.21 1.28 1.20 1.31

Robust standard errors are in parentheses *** p<0.01, ** p<0.05, * p<0.1

(18)

18

The other two audit committee characteristics show no significant relationship. This means that hypothesis 1 and 3 can’t be accepted. As expected, a significant relationship between Firm_Size and IFRS7_Disclosure_Quality is present.

Furthermore, the moderating effect of CEO_Power on the relationship between

AC_Characteristics and IFRS7_Disclosure_Quality was tested. This reflects research model 2 in the method section. The results of this estimation can be found in table 4. For a better fitting interaction term, CEO_Power index was mean centered before running the regressions. As model 1 in table 4 shows, I found a negative and significant interaction term for

CEO_Power and AC_FinancialExpertise. This implies that a more powerful CEO reduces the positive relationship between AC_FinancialExpertise and IFRS7_Dislcosure_Quality or even makes it a negative relationship, since the coefficient of the interaction term is higher than the coefficient of AC_FinancialExpertise. However, despite the significant interaction term, in this model AC_FinancialExpertise individually is not significant. This was also the case in table 3 without the moderating variable. For this reason, hypothesis 4 cannot be fully accepted. As can be seen in model 2 in table 4 and in line with my expectations, the interaction term CEO_Power * AC_Size is significant and negative with a coefficient of β = -0.088. This indicates that a more powerful CEO reduces the significant positive relationship between AC_Size and IFRS7_Disclosure_Quality, or even makes it negative. This moderating effect has been visualized and can be seen in Appendix A7. A negative interaction term between CEO_Power and AC_Characteristics was expected in the hypothesis. However, since AC_Size is positively related to IFRS7_Disclosure_Quality instead of negatively, hypothesis 5 can’t be fully accepted. For AC_Tenure, there is no significant moderating effect of CEO_Power, so hypothesis 6 can’t be accepted.

(19)

19

Table 4

Regression results Research model 2 The moderating role of CEO_Power Dependent Variable:

IFRS7_Disclosure_Quality

(1) (2) (3)

Model 1 Model 2 Model 3

Constant 0.518*** 0.515*** 0.505*** (0.068) (0.066) (0.067) CEO_Power 0.183*** 0.176*** 0.172*** (0.059) (0.060) (0.060) AC_FinancialExpertise 0.004 (0.006) AC_FinancialExpertise * CEO_Power -0.133** (0.066) AC_Size 0.017*** (0.006) AC_Size * CEO_Power -0.088** (0.044) AC_Tenure -0.008 (0.006) AC_Tenure * CEO_Power -0.060 (0.058) Leverage 0.004 0.004 0.004 (0.003) (0.003) (0.003) ROA 0.004 0.002 0.004 (0.005) (0.005) (0.005) Firm_Size 0.030*** 0.029*** 0.031*** (0.005) (0.005) (0.005)

Industry-Dummies Included Included Included

Year-Dummies Included Included Included

Observations 850 850 850

R-squared 0.098 0.102 0.096

Highest VIF 1.27 1.33 1.27

As mentioned in the method section, I performed additional regressions to test for the different components of IFRS7_Disclosure_Quality separately, namely

IFRS7_Disclosure_Quality_Assurance and IFRS7_Disclosure_Quality_Risk. Results of the regressions without the moderation, but with the new dependent variables can be found in table 5 and table 6. As can be seen, AC_Size remains positive significant when tested against IFRS7_Disclosure_Quality_Assurance. The positive coefficient is even higher than in table 3.

(20)

20

However, table 6 shows that AC_Size is no longer significant when tested against

IFRS7_Disclosure_Quality_Risk. This implies that AC_Size is able to influence the quality of assurance related disclosures, but not the risk related disclosures. Furthermore, in contrast with table 3, table 6 shows a significant negative coefficient for AC_FinancialExpertise. This means that an audit committee with relatively more financial experts will lead to a lower IFRS7_Disclosure_Quality_Risk. In other words, firms with a higher percentage of financial experts on the audit committee, tend to provide lower quality disclosures regarding risk exposure. Just like in table 3, AC_Tenure shows a small insignificant coefficient in both table 5 and table 6. Furthermore, what is different from table 3, is that in table 5 Leverage shows a small positive significant relationship with IFRS7_Disclosure_Quality_Assurance. In table 6, ROA shows a small positive relationship with IFRS7_Disclosure_Quality_Risk. Both control variables ROA and Leverage weren’t significant before in table 3.

Table 5

Regression results IFRS7_Disclosure_Quality_Assurance Dependent Variable: IFRS7_Disclosure_Quality_Assurance (1) (2) (3) (4)

Constant 0.562*** 0.572*** 0.557*** 0.572*** (0.071) (0.070) (0.072) (0.071) AC_FinancialExpertise 0.001 0.004 (0.008) (0.009) AC_Size 0.022*** 0.023*** (0.008) (0.008) AC_Tenure 0.007 0.009 (0.009) (0.009) ROA -0.006 -0.009 -0.006 -0.008 (0.011) (0.011) (0.011) (0.011) Leverage 0.013** 0.013** 0.013** 0.013** (0.005) (0.005) (0.005) (0.005) Firm_Size 0.039*** 0.033*** 0.04*** 0.033*** (0.007) (0.008) (0.007) (0.008)

R-squared 0.057 0.065 0.058 0.066

Highest VIF 1.21 1.28 1.20 1.31

(21)

21

Table 6

Regression results IFRS_Disclosure_Quality_Risk Dependent Variable: IFRS7_Disclosure_Quality_Risk (1) (2) (3) (4)

Constant 0.468*** 0.499*** 0.504*** 0.466*** (0.116) (0.116) (0.115) (0.115) AC_FinancialExpertise -0.035*** -0.037*** (0.009) (0.009) AC_Size -0.007 -0.009 (0.009) (0.009) AC_Tenure -0.004 -0.009 (0.008) (0.008) ROA 0.013* 0.015** 0.014** 0.013* (0.007) (0.007) (0.007) (0.007) Leverage 0.006 0.008 0.008 0.006 (0.006) (0.006) (0.006) (0.006) Firm_Size 0.040*** 0.037*** 0.035*** 0.042*** (0.006) (0.006) (0.006) (0.006)

R-squared 0.118 0.103 0.102 0.120

Highest VIF 1.21 1.28 1.20 1.31

Table 7 and table 8 provide the results of the last two regressions that were done in this research. It represents the effect of the moderating role of CEO_Power, but with the different components of IFRS7_Disclosure_Quality split up. Both tables don’t provide interesting new insights, since no interaction term is found significant when tested against the different components of IFRS7_Disclosure_Quality separately.

(22)

22

Table 7

Regression results

The moderating role of CEO_Power. IFRS7_Disclosure_Quality_Assurance Dependent Variable:

IFRS7_Disclosure_Quality_Assurance

(1) (2) (3)

Constant 0.605*** 0.602*** 0.583*** (0.072) (0.070) (0.074) CEO_power 0.310*** 0.292*** 0.296*** (0.083) (0.088) (0.084) AC_FinancialExpertise 0.003 (0.008) AC_FinancialExpertise * CEO_Power -0.138 (0.100) AC_Size 0.019** (0.009) AC_Size * CEO_Power -0.062 (0.060) AC_Tenure 0.007 (0.009) AC_Tenure * CEO_Power -0.086 (0.088) Firm_Size 0.031*** 0.028*** 0.033*** (0.008) (0.008) (0.007) ROA -0.005 -0.007 -0.004 (0.011) (0.011) (0.011) Leverage 0.013** 0.012** 0.012** (0.005) (0.005) (0.005)

R-squared 0.075 0.078 0.075

Highest VIF 1.27 1.33 1.27

(23)

23

Table 8

Regression results

The moderating role of CEO_Power. IFRS7_Disclosure_Quality_Risk Dependent Variable:

IFRS7_Disclosure_Quality_Risk

(1) (2) (3)

Constant 0.472*** 0.499*** 0.498*** (0.117) (0.116) (0.116) CEO_power -0.038 -0.016 -0.041 (0.094) (0.094) (0.097) AC_FinancialExpertise -0.034*** (0.009) AC_FinancialExpertise * CEO_Power -0.110 (0.109) AC_Size -0.005 (0.009) AC_Size * CEO_Power -0.038 (0.073) AC_Tenure -0.004 (0.008) AC_Tenure * CEO_Power -0.042 (0.096) Firm_Size 0.040*** 0.038*** 0.036*** (0.006) (0.006) (0.006) ROA 0.013* 0.015** 0.014** (0.007) (0.007) (0.007) Leverage 0.006 0.008 0.008 (0.006) (0.006) (0.006)

R-squared 0.120 0.103 0.103

Highest VIF 1.27 1.33 1.27

(24)

24

5. Discussion and conclusion

In the final section of this paper, the findings concerning the results are explained.

Furthermore, the implications that come along with the results will be explained. Thereafter, the limitations of this paper and recommendations for future research will be addressed.

5.1 Findings

In this paper I performed, to the best of my knowledge, the first empirical analysis of the effects of audit committee characteristics on the quality of managements’ IFRS7 risk disclosures. Furthermore, the moderating effect of a CEO’s power on this relationship is tested. To perform this empirical analysis, a partially hand-collected dataset of 850 observations of UK companies between the years 2010 – 2016 is used.

The results of this research show that a bigger audit committee size leads to a higher quality of managements’ IFRS7 risk disclosures. This is in contrast with the predefined hypothesis. A possible explanation for this positive relationship lays in the research of Turley and Zaman (2007). They mention that a bigger audit committee can draw on a bigger and wider set of skills, such as more expertise, experience or knowledge. This viewpoint is supported by Chtourou and Courtea (2004), who argued that a larger audit committee has better control and oversight powers over the management. Therefore, it might be the case that the above

mentioned advantages of a bigger audit committee outweigh the downsides that were

mentioned in the theory section. This is a possible explanation for the contradicting result for audit committee size in this research. However, there must be a point where the audit

committee becomes too big and where the advantages of a big audit committee will be outweighed by the downsides mentioned earlier. So this finding has a caveat that must be considered. The other two audit committee characteristics showed no significant relationship with managements’ overall IFRS7 risk disclosure quality.`

Second, the results suggest that a more powerful CEO can indeed moderate the relationship between certain audit committee characteristics and IFRS7 risk disclosure quality. A more powerful CEO significantly reduces the effects of both audit committee size and audit committee financial expertise on IFRS7 risk disclosure quality. For audit committee tenure, no significant moderating effect was found.

Probably the most interesting finding of this research is the significant negative relationship found between the financial expertise of an audit committee and the risk related component of IFRS 7 disclosures. Where literature argues that an audit committee with more financial

(25)

25

expertise leads to improved monitoring and disclosure quality (Krishnan & Visvanathan, 2008), this seems not to be the case for risk related disclosures. Disclosures regarding risk exposure seem to be of less quality when there are relatively more financial experts involved. As mentioned before, risk related disclosures are associated with negative information and can increase the risk perception of users (Kravet and Muslu, 2013). This is why firms might be more reluctant in providing risk related disclosures compared to assurance related

disclosures. Therefore, one possible explanation for the negative relation lies in the fact that financial experts are the one good at presenting numbers and hiding bad or negative news to change the perception of the users. When audit committees possess relatively more financial expertise, they are better able to hide the negative information than when there are less

financial experts involved. This leads to a lower quality of IFRS 7 risk related disclosures, but might benefit the company by hiding negative information. Since the audit committee is supposed to improve the quality of the reporting process and be “a reliable guardian of the public interest” (Levitt, 2000), this is an interesting finding than contradicts literature.

5.2 Implications

This study contributes to the extant literature and provides new interesting insights in the current IFRS 7 risk disclosure literature. As mentioned before, to the best of my knowledge, this is the first time someone addresses the effect of audit committee characteristics on this topic. Furthermore, by adding the moderating role of CEO power, it provides even more valuable insights into this topic.

This study also has practical implications for investors and other stakeholders. Investors and stakeholders should be aware of the effect of audit committee characteristics and the role of CEO Power. Also, when management wants to improve their IFRS7 risk disclosure quality, they should keep in mind that certain audit committee characteristics can help them improve this.

5.3 Limitations and future research

Just like in every research, there are some limitations that should be considered. First of all, the data on IFRS7 risk disclosure quality was hand collected in different years by a group of people that have worked on the same project. Although the data collection process was guided by a manual, it is still prone to the presence of mistakes due to different interpretations. Second, for the sample, only large listed companies from the United Kingdom have been used. This limits the generalizability to smaller companies, private companies or companies

(26)

26

from outside the United Kingdom. Third, within my current time and resources, it wasn’t possible to find data on CEO power for all the observations of the initial sample of IFRS7 risk disclosure quality. This led to a data reduction, whereby the initial sample was eventually cut to 850 observations. This is still enough to work with, but with even more observations the results might be even better generalizable.

Those limitations are good starting points for future research on this topic. Executing this research with different companies from different countries will contribute to the

generalizability of this research. Furthermore, with even more time and resources on hand, more CEO power characteristics can be acquired. This will increase the sample and thus the generalizability of this research. As mentioned in the findings section, a bigger audit

committee enhances the quality of IFRS 7 disclosures to a certain point where adding another member will reduce the committee effectiveness. More research regarding the size of the audit committee might therefore be interesting and valuable. It might also be interesting to add even more audit committee characteristics to this research, such as audit committee meeting frequency or the gender or background of audit committee members.

(27)

27

References

Abdullah, M., Shukor, Z. A., & Rahmat, M. M. (2017). The influences of risk management committee and audit committee towards voluntary risk management disclosure. Jurnal Pengurusan (UKM Journal of Management), 50.

Aldamen, H., Duncan, K., Kelly, S., McNamara, R., & Nagel, S. (2012). Audit committee characteristics and firm performance during the global financial crisis. Accounting & Finance, 52(4), 971-1000.

Amran, A., Manaf Rosli Bin, A., & Che Haat Mohd Hassan, B. (2008). Risk reporting: An exploratory study on risk management disclosure in Malaysian annual reports. Managerial Auditing Journal, 24(1), 39-57.

Bao, Y., & Datta, A. (2014). Simultaneously discovering and quantifying risk types from textual risk disclosures. Management Science, 60(6), 1371-1391.

Barakat, A., & Hussainey, K. (2013). Bank governance, regulation, supervision, and risk reporting: Evidence from operational risk disclosures in European banks. International Review of Financial Analysis, 30, 254-273.

Bart, C., & McQueen, G. (2013). Why women make better directors. International Journal of Business Governance and Ethics, 8(1), 93-99.

Beattie, V., McInnes, B., & Fearnley, S. (2004). A methodology for analysing and evaluating narratives in annual reports: A comprehensive descriptive profile and metrics for disclosure quality attributes. Accounting Forum, 28(3), 205-236

Bruynseels, L., & Cardinaels, E. (2014). The audit committee: Management watchdog or personal friend of the ceo? The Accounting Review, 89(1), 113-145.

Chan, A. M. Y., Liu, G., & Sun, J. (2013). Independent audit committee members’ board tenure and audit fees. Accounting & Finance, 53(4), 1129-1147.

Chtourou, S. M., & Courteau, L. (2004). The effect of audit committee expertise, independence, and activity on aggressive earnings management. Auditing: a journal of practice & theory, 23(2), 13-35.

(28)

28

Combs, J., Ketchen, D., Perryman, A., & Donahue, M. (2007). The moderating effect of ceo power on the board composition?firm performance relationship. Journal of Management Studies, 0(0),

Eisenhardt, K. M. (1989). Agency theory: An assessment and review. Academy of management review, 14(1), 57-74.

Elbannan, M. A., & Elbannan, M. A. (2015). Economic consequences of bank disclosure in the financial statements before and during the financial crisis: Evidence from Egypt. Journal of Accounting, Auditing & Finance, 30(2), 181-217.

Elshandidy, T., Shrives, P. J., Bamber, M., & Abraham, S. (2018). Risk reporting: A review of the literature and implications for future research. Journal of Accounting Literature, 40, 54-82.

Elzahar, H., & Hussainey, K. (2012). Determinants of narrative risk disclosures in UK interim reports. The Journal of Risk Finance, 13(2), 133-147.

Finkelstein, S. (1992). Power in top management teams: Dimensions, measurement, and validation. The Academy of Management Journal, 35(3), 505-538.

Freeman, R.E. (1984) Strategic Management: A Stakeholder Approach. Pitman, Boston.

Hillman, A. J., & Dalziel, T. (2003). Boards of directors and firm performance: Integrating agency and resource dependence perspectives. Academy of Management review, 28(3), 383-396.

Hodder, Leslie, Lisa Koonce, and Mary Lea McAnally. (2001.) Sec Market Risk Disclosures: Implications for Judgment and Decision Making. Accounting Horizons 15(1):49–70.

Hope, O. K., Hu, D., & Lu, H. (2016). The benefits of specific risk-factor disclosures. Review of Accounting Studies, 21(4), 1005-1045.

IFRS 7. 2005. Financial Instruments: Disclosure. Available at: https://www.ifrs.org/issued-standards/list-of-standards/ifrs-7-financial-instruments-disclosures/#about

Khlif, H., & Hussainey, K. (2016). The association between risk disclosure and firm characteristics: A meta-analysis. Journal of Risk Research, 19(2), 181-211.

KPMG. (2017). Audit Committee Handbook. Available at:

(29)

29

Kravet, T. and Muslu, V., 2013. Textual risk disclosures and investors’ risk perceptions. Review of Accounting Studies, 18(4), 1088-1122

Krishnan, G. V., & Visvanathan, G. (2008). Does the SOX definition of an accounting expert matter? The association between audit committee directors' accounting expertise and

accounting conservatism. Contemporary Accounting Research, 25(3), 827-858.

Levitt, A. 2000. Remarks before the conference on the rise and effectiveness of new corporate governance standards. Federal Reserve Bank of New York, December 12

Lisic, L., Neal, T., Zhang, I., & Zhang, Y. (2016). Ceo power, internal control quality, and audit committee effectiveness in substance versus in form. Contemporary Accounting Research, 33(3), 1199-1237.

Marshall, A., & Weetman, P. (2007). Modelling transparency in disclosure: the case of foreign exchange risk management. Journal of Business Finance & Accounting, 34(5‐6), 705-739.

Miihkinen, A. (2013). The usefulness of firm risk disclosures under different firm riskiness, investor-interest, and market conditions: New evidence from finland. Advances in Accounting, Incorporating Advances in International Accounting, 29(2), 312-331.

Mustafa, S. T., & Youssef, N. B. (2010). Audit committee financial expertise and misappropriation of assets. Managerial Auditing Journal.

Thiruvadi, S., & Huang, H. (2011). Audit committee gender differences and earnings management. Gender in Management: An International Journal, 26(7), 483-498.

Muthupandian, K. S. (2008). IFRS 7 Financial Instruments: Disclosures-A Closer Look.

Ntim, C. G., Lindop, S., & Thomas, D. A. (2013). Corporate governance and risk reporting in South Africa: A study of corporate risk disclosures in the pre-and post-2007/2008 global financial crisis periods. International Review of Financial Analysis, 30, 363-383.

Sharma, V., & Iselin, E. (2012). The association between audit committee multiple-directorships, tenure, and financial misstatements. Auditing, 31(3), 149-175.

Sultana, N. (2015). Audit committee characteristics and accounting conservatism. International Journal of Auditing, 19(2), 88-102

(30)

30

Sultana, N., Singh, H., & Van der Zahn, J. (2015). Audit committee characteristics and audit report lag. International Journal of Auditing, 19(2), 72-87.

Sun, J., Lan, G., & Liu, G. (2014). Independent audit committee characteristics and real earnings management. Managerial Auditing Journal.

Turley, Stuart and Mahbub Zaman. 2007. “Audit Committee Effectiveness: Informal Processes and Behavioural Effects.” Accounting, Auditing & Accountability

(31)

31

Appendix

Appendix A1: All variables and their descriptions

Variable Description

IFRS7_Disclosure_Quality An index variable computed as the equal weighted average of the normalized value of 13 items presented in appendix A2.

IFRS7_Disclosure_Quality_Assurance An index variable computed as the equal weighted average of the normalized value of 6 items on hedging and collateral presented in the Appendix A2.

IFRS7_Disclosure_Quality_Risk An index variable computed as the equal weighted average of the normalized value of 3 items on risk exposure, risk concentration and impairment presented in the Table A2.

AC_Characteristics Audit committee characteristics. Can be split up into, AC_Size, AC_Tenure and

AC_FinancialExpertise

AC_Size Total number of directors that make up for the

audit committee

AC_Tenure The average number of years a director is active as

a member of the audit committee

AC_FinancialExpertise The average number of financial experts in the audit committee

CEO_Power An index variable computed as the equal weighted

average of the normalized values of CEO_NetworkSize, CEO_BoardTenure, CEO_Qualifications and CEO_Compensation. CEO_NetworkSize The total network size of the CEO, according to

data from BoardEx

CEO_BoardTenure The total amount of time a CEO has been part of the board of directors

CEO_Qualifications Total number of qualifications a CEO possesses

CEO_Compensation Total compensation for the CEO each year

Firm_Size The natural logarithm of the Total Assets

Leverage The ratio of total debts to total assets

ROA The ratio of earnings before interest and tax to

total assets at the end of the fiscal year.

Year_D A year dummy used to control for year-effects

Industry_D An industry dummy used to control for

(32)

32 Appendix A2: IFRS 7 Disclosure Index

Name of the item Explanation

Item #4

The existence of a separate Hedge Accounting Section explaining the impact

(1=Yes)

This is the only item we do not care whether it is in the “risk disclosure section or not. If you see that the firm has a separate section to explain its hedging activity in any part of the annual report, then you will code it as 1.

Hedge terms should be in the title of a section. For example. For the sky plc, it is in note 21, but risk disclosure is note 22. Still we will code it as 1.

Otherwise "0".

Item #7

Explicit and detailed disclosure of the monetary amount of hedge activity - Interest rate risk (1=Yes)

Search for the word hedge in the “risk disclosure

section”, not in the other parts of the annual report, you don’t need to read details.

You need to find that the company states hedging related to interest rate risk and if you see a monetary amount explanations as well, then it will be coded as 1. Otherwise "0". If it is cross referenced it will be considered as “0” as well.

If the company inform about the non-existence of hedging it is coded as 1 as well.

Item #10

Explicit and detailed disclosure of the monetary amount of hedge activity - Hedge activity - Currency Risk (1=Yes)

You need to find that the company states hedging related to currency risk and if you see a monetary amount explanations as well, then it will be coded as 1. Otherwise "0". If it is cross referenced it will be considered as “0” as well.

Item #13

Explicit and detailed disclosure of the monetary amount of hedge activity - Hedge activity - Other Price Risk (commodity and

equity price risks) (1=Yes)

You need to find that the company states hedging related to other price risk (risks for equity or

commodity prices) and if you see a monetary amount explanations as well, then it will be coded as 1.

Otherwise "0". If it is cross referenced it will be considered as “0” as well.

(33)

33 Item #16

Explicit and detailed disclosure of the monetary amount of hedge activity - Hedge activity - Liquidity Risk (1=Yes)

You need to find that the company states hedging related to liquidity risk and if you see a monetary amount explanations as well, then it will be coded as 1. Otherwise "0". If it is cross referenced it will be considered as “0” as well.

Item #22

Explicit and detailed disclosure of the monetary amount of hedge activity - Hedge activity - Credit Risk (1=Yes)

You need to find that the company states hedging related to credit risk and if you see a monetary amount explanations as well, then it will be coded as 1. Otherwise "0". If it is cross referenced it will be considered as “0” as well.

Item #18

Explicit disclosure of the amount of maximum exposure to credit risk

Search for “credit risk exposure” and/or “exposure to credit risk” in the “risk disclosure section”, not

in the other parts of the annual report, you

don’t need to read details.

If not disclosed, then the company will get the score “0”.

If the company gives a cross reference to any other notes or financial statement information other than the notes on risk disclosure, then the company will get the score “1”.

If the amount is disclosed in the risk disclosure, then the company will get the score “2”.

Item #19

Disclosure of the existence of collateral for credit risk exposure

Search for “collateral” in the “risk disclosure

If no disclosure on the level of collateral, then the company will get the score “0”.

If it is narrative explanation (just words) of the level of collateral in the risk disclosure, then the company will get the score “1”.

If there is tables containing quantitative data

regarding the level of collateral in the risk disclosure, then the company will get the score “2”. If the

The effect of audit committee characteristics on managements’ IFRS 7 risk disclosure quality in UK companies