The Influence of Board Independence on the Quality of Risk Disclosure and the Effects of Non-Executive
Board Members’ Industry Expertise
Master Thesis Accountancy Faculty of Economics and Business
Rijksuniversiteit Groningen
Author: K.T.M. Jansman S2960567
Word count: 10.348
Supervisor: dr. Y. Karaibrahimoglu
ABSTRACT
Starting from 2007, firms are required to apply IFRS 7 in their financial reporting. IFRS 7 requires firms to disclose the risk exposure and risk hedging of their operations in a more structured manner. This research will analyze factors that might influence this quality of risk disclosure by companies. First, this study analyzes the impact of board independence on the quality of risk disclosure by UK premium-listed firms from 2010-2016. This research finds that board independence has a positive effect on the quality of risk disclosure by firms.
Second, the moderating role of industry expertise of non-executive board members on the relationship between board independence and the quality of risk disclosure is analyzed.
Partially in accordance with what this study theorizes, a positive moderating effect of industry expertise is found on this relationship. However, the industry expertise as standalone variable had a negative significant effect on the quality of risk disclosure by companies. Thus, we can only partially agree with the previous literature regarding industry expertise and the
relationship with board independence and the quality of risk disclosure by companies. Finally, an additional analysis will be conducted regarding the existence of a separate Risk Committee and the relationship with the quality of risk disclosure by companies. The results show a negative significant relationship between the existence of a separate Risk Committee and the quality of risk disclosure by companies.
Keywords: IFRS 7, Risk Disclosure, Board Independence, Industry Expertise, UK Premium-
Listed Firms, Risk Committee
Table of Contents
1. Introduction 4
2. Theoretical Framework 8
2.1 Quality of Risk Disclosure (IFRS 7) 8
2.2 Board Independence 9
2.3 Industry Expertise 10
3. Methodology 12
3.1 Sample 12
3.2 Research Design 13
3.3 Variables 15
4. Results 20
4.1 Descriptive Statistics 20
4.2 Correlation Matrix 21
4.3 Multivariate analysis 23
4.4 Additional Test 27
5. Discussion and Conclusions 29
5.1 Discussion 29
5.2 Limitations 31
5.3 Future Research 32
6. References 33
7. Appendix: Disclosure Index Risk Disclosure Quality 38
1. Introduction
There is an ongoing debate regarding the risk information provided to the users of the annual reports of companies and the quality of risk management that is done by these companies. Linsley and Shrives mention (2005): “In identifying risk disclosures to be deficient, the institutional investors acknowledge that it is important to be able to assess the risk profile of a company and this is only possible if relevant risk information is provided.” Moreover, risk disclosure is important, because it will lead to a better risk management, improved accountability for stewardship, investor protection, and usefulness of financial reporting (Dobler, Lajili and Zéghal, 2011). This indicates that risk disclosure could help tackling managerial problems and evaluating the effectiveness of the
management in dealing with risk-issues. “Risk is an ever-present and ever-changing element of the business environment. Major risk events underscore the necessity for risk management to remain vigilant and continuingly evolve” (Moore, Brauneis, 2008). This illustrates the importance of the ongoing concern regarding the disclosure of risks by companies. Spira and Page (2003) mention that in the UK the risk disclosure requirements as stated in the Combined Code of Best Practice in Corporate Governance, produced by the Turnbull committee, enforced pressure on companies to better disclose their risks. In 2005, the International Accounting Standards Board (IASB) issued a regulatory initiative regarding disclosure requirements with the implementation of “IFRS 7 Financial
Instruments: Disclosures”. By this regulation, both financial and non-financial firms with financial instruments, which constitute of assets or packages of capital that the company can trade, need to disclose two aspects of risk. First, risk exposure, which refers to the significance, nature and extent of firms’ risk exposure. This information might be
perceived as negative, since it refers to the risks that the company might face. Second, the company should disclose information, which is perceived to be positive, with regards to assurance, which refers to hedging activities meant to address risks arising from the use of financial instruments. The IASB encouraged the early adoption of IFRS 7. However, the effective adoption date for firms was for annual reports for year-end 2007 (IASB, 2005).
This is right around the time that the financial crisis started to take shape in the world.
Since the financial crisis of 2008, the interest in the way companies disclose their risks
has further increased. Scheer (2011) suggests that: “The lack of disclosure is the root
cause of the financial crisis.” Furthermore, Scheer (2011) mentions that the financial
markets can be seen as “representative governments”. Just like a democracy, the financial
system requires transparency. This way, investors understand and place a value on the risks that are associated with assets or transactions (Scheer, 2011).
Governments and other regulators should also emphasize the need for risk disclosure.
“Regulations and initiatives on risk reporting are fundamental in distinguishing between voluntary and mandatory risk disclosure. This has already been initiated by the
implementation of IFRS 7 in 2005. However, neither the UK Accounting Standards Board nor the International Accounting Standards Board have issued a comprehensive mandatory risk reporting standard” (Elshandidy, Fraser, Hussainey, 2013). Thus, risk reporting
regarding operational risks, in the UK is voluntarily, as opposed to countries like the United States, that do have mandatory rules for risk disclosure. However, the UK Corporate Governance code of 2018 does state that boards should at least disclose information regarding the company’s emerging and principal risks. Finally, according to Elshandidy, Shrives, Bamber, and Abraham (2018), in the last decade, regulators aim at improving quantity and quality in firms’ risk reporting. The quality of risk disclosure will be examined by using the IFRS 7 regulation of the IASB, to which will be referred as the quality of risk disclosure in the remainder of this paper.
There is still limited research regarding the quality of risk disclosure by companies. This has been confirmed by Dobler, Lajili and Zéghal (2011): “Since risk disclosure is a relatively new field of research, there is only piecemeal evidence on its attributes and determinants to date.”
This research will try to extent the literature regarding the influence of independent board directors on the quality of risk disclosure. Previous research on independent board
directors has shown the impact of the presence of independent board directors on boards on the quality of risk disclosure by companies (Cheng and Courtenay, 2006, Cheng and Jaggi, 2002). Additionally, a moderator will be added namely the industry expertise of board directors, to study the influence of this variable on the relationship between independent board directors and the quality of risk disclosure.
The industry expertise of independent board directors was a relevant topic during the 2008- 2009 financial crisis. During this period, “the issue of independent directors’ industry expertise was brought to the forefront of the discussion on corporate governance and corporate boards in particular” (Wang, Xie, Zhu, 2015). There are many industry-specific accounting standards and practices, as described by the Financial Accounting Standards Board, and the application of such standards as for instance revenue recognition in the software industry, requires industry knowledge by the independent directors. (Beasley et al.
2000; Beasley et al. 2010). Furthermore, industry expertise has already been established as an
important trait for members of the audit committee of companies, since Olson (1999) argues that ‘‘The best qualified audit committee members will often be those who have practical management experience, and industry knowledge, as opposed to those with a financial or accounting background.’’ This is because industry expertise helps the audit committee members to understand the industry and business operations and they can better assess the potential risks that the company might be facing.
This study analyzes two separate hypotheses. First, the relationship between board independence and the quality of risk disclosure has been analyzed. In line with previous research of Cheng and Courtenay (2006), Cheng and Jaggi (2002) and other research, this study provides evidence that board independence has a positive relationship with the quality of risk disclosure by companies. Second, the moderating role of industry experts on the relationship between board independence and the quality of risk disclosure by companies is tested. The results are partially in accordance with previous studies. In line with Olson (1999), who already showed the importance of industry expertise, the results show that the interaction between industry expertise and board independence has a positive effect on the quality of risk disclosure by companies. However, industry expertise as a standalone variable has a
significant negative relationship with the quality of risk disclosure in this research. Finally an additional test is conducted to find the impact of the existence of a separate Risk Committee on the quality of risk disclosure by companies. Dobler (2008) already established the
importance of a separate Risk Committee for the quality of risk disclosure, on which the expectation is build that the relationship would be positive. However, the results show a negative impact on the quality of risk disclosure when there is a separate Risk Committee in place, which thus contradicts previous studies.
This research makes several contributions to the existing literature. First, this research is an extension to the research conducted by Porumb and Karaibrahimoglu (2019). In this research, the quality of risk disclosure by companies in the UK from 2010-2016 will be analyzed. This period has not been analyzed yet and thus contributes to the existing knowledge regarding the quality of risk disclosure by companies in the UK.
Second, it could be considered as an interesting extension to the relationship between
independent board directors and the quality of risk disclosure. The study will also
investigate the moderating effect of industry expertise on this relationship. The study
conducted by Elshandidy, Fraser and Hussainey (2013) already focused on the UK listed
companies. They investigated the role of board independence on the quality of risk
disclosure by companies in the UK. However, this research adds to this existing literature by re-examining this relationship in more recent years and to see whether there might be differences between this paper and the paper of Elshandidy, Fraser and Hussainey (2013).
Moreover, in the study of Elshandidy, Fraser and Hussainey (2013), they did not analyze the moderating effect of industry expertise, thus this research will also be an extension to the existing literature. Furthermore, this research analyzes the overall risk disclosure quality by companies and the two components, risk and assurance, individually. This way, it is possible to find differences in results regarding the hypotheses presented in this research. Finally, this research implements and additional test to analyze the impact of the existence of a separate Risk Committee on the quality of risk disclosure by companies.
There has been little research regarding Risk Committees and the relationship with the quality of risk disclosure. Thus, this research fills the gap in the existing literature by analyzing this relationship.
The remainder of this paper will be structured as followed. In the next chapter the
theoretical background and previous studies will be discussed. Based on these theories and
previous studies, the hypotheses will be developed that will be researched in this paper. In
the subsequent chapter the methodology of this paper will be described, consisting of the
data that is being used and the method of research. Then the results of the analysis will be
presented. Finally, the paper presents the conclusion and discussion, as well as limitations
of this study and future research possibilities.
2. Theoretical framework
Risk disclosure plays an important role in mitigating the information asymmetry between stakeholders of a company. Jorgensen and Kirschenheiter (2003) mention that: “In general terms, risk disclosure shall reduce the information asymmetry between managers and stakeholders by providing users of financial reports with information on the risks a company faces and on how these risks are managed.” several theories are relevant to the explanation of the importance of independent board members and industry expertise for the quality of risk disclosure.
2.1 Quality of Risk Disclosure (IFRS 7)
According to the legitimacy theory, as explained in the article of Al Hadi, Hasan and
Habib (2016): “Companies, including financial institutions, may disclose market risk
information in an attempt to satisfy external pressure for more transparent reporting.” This
indicates that there is a need for justification by companies to make sure that they handle
in the most transparent way towards their stakeholders. The legitimacy theory implies that
the companies have a social contract and if society views that the firm breached this
contract, the entity’s survival is at risk (Deegan, 2002). In 2005, the IASB introduced
IFRS 7 “Financial Instruments: Disclosure”. The main objective of this standard is to
improve the quality of the information provided by companies regarding their financial
instruments. The IASB (2005) mentions that this standard will help the users of annual
reports and other financial statements making better judgments on the risk profile of the
company. Even though, with the introduction of IFRS 7, the companies are mandated to
disclosure information regarding risks and risk management, the implementation of IFRS
7 requires a lot of judgment. This way, there is variability between the companies in the
way they adopt IFRS 7, which could emphasize their need for justification by adopting a
certain style of risk reporting. For example, some companies might adapt the minimal
requirements regarding risk disclosure as opposed to companies who report their risks
extensively. From an agency theory perspective, risk disclosure also decreases the
information asymmetry between external stakeholders and the management of the
company. This information asymmetry creates opportunities for managers or board
directors to not disclose all the information that they have available, for instance to avoid
sharing information to shareholders regarding risky investments or possible other risks
that the company could incur. Ang, Cole and Lin (2000) found that: “Investors make costs
to align the interests of the managers with their own”. These monitoring costs could be
reduced by the implementation of the IFRS 7 disclosure requirements for companies. By disclosing sufficient information to stakeholders regarding the risks that the company is facing, companies make sure that the investors receive enough information to be able to assess the risks themselves and potentially decide about their investment.
2.2 Board Independence
Regarding independent board directors, the stakeholder theory explains the positive relationship between this variable and the quality of risk disclosure. Rosenstein and Wyatt (1990) suggest that: “Independent directors can make impartial judgements in corporate decision making, increase the monitoring of managers behavior, and, thus, enhance stakeholders’ interest.” For example, Cheng and Courtenay (2006) found a positive relationship between independent board directors and financial disclosures. They also found that: “Independent directors play a more responsible role in enhancing both voluntary disclosure and investors’ interests in emerging economies.” Chen and Jaggi (2002) concluded that: “Non-executive directors influence the way the firm discloses its information, which is clearly visible in the disclosure of financial information”. Other previous studies further investigated the risk disclosure by companies and the role of independent board directors. Barakat and Hussainey (2013) conducted research on the role of independent board directors on the quality of risk disclosure. They found that
Operational Risk Disclosure (ORD) of European banks is of a higher quality if these banks have a higher proportion of outside independent board directors and concentrated outside non-governmental ownership. Furthermore, they advised that: “For the sake of enhancing risk reporting quality in banks, our findings recommend sustaining board independence.” Moreover, Neifar and Jarbouri (2018) found a positive relationship between the share of independent directors in the board of directors with the voluntary Operational Risk Disclosure by Islamic banks. Another interesting research was done by Ashfaq et al (2016) in which they found that: “Banks with a higher proportion of
independent board directors present their stakeholders a higher degree of risk disclosure in terms of volume of information as well as quality of the disclosure.” These studies mainly focused on the banking sector in different countries over the world. Elshandidy, Fraser and Hussainey (2013) investigated the role of board independence on the quality of risk disclosure by companies in the UK. They concluded that board independence affects the level of risk disclosure. Finally, Hillman and Dalziel (2003) found that effective
monitoring by non-executive directors have been found to reduce the agency costs for
investors. This suggest that the presence of independent board directors aids the investors by monitoring the disclosure of companies for them and thus ensuring higher quality of disclosure. Based upon the previous literature discussing the impact of independent board members on the quality of risk disclosure, this leads to the following hypothesis:
H1: There is a positive relationship between board independence and the quality of risk disclosure by companies.
2.3 Industry Expertise
The industry expertise of board directors could be an important moderating factor concerning the quality of risk disclosure. As mentioned in the guidelines of Deloitte (2014): “It is of vital importance that members of the risk committee have experience within the industry”. This is because it allows them to identify risk areas and to better be aware of methods of managing the company’s exposure to those risks. This guideline specifically focused on the role of the Risk Committee, but it also suggested that industry expertise is a vital trait of board directors, and the Risk Committee consists of board directors. Research regarding the audit committee at companies already established the importance of industry expertise for the quality of risk disclosure and reporting.
According to Abbott (2000), research supports the assertion that specialized industry knowledge is a component of audit quality and the quality of risk disclosure by companies. The importance of industry expertise can be linked to the agency theory, which states that information asymmetry is a problem in the relationship between
stakeholders and the company, and even within the company itself. According to Faleye et al (2014): “Industry expertise reduces the internal information asymmetry because it provides a deeper understanding of the risk and reward profiles of the firm’s industry.”
Regarding industry expertise, Wang, Xie and Zhu (2015) found evidence that: “Relevant
industry experience significantly enhances boards’ monitoring effectiveness in other key
areas of corporate policies and decision making.” They mention risk management as one
of these key areas. Cohen et al (2014) concluded that: “Audit Committees may lack
sufficient industry expertise to understand and thus properly monitor industry-specific
accounting issues.” This study focused on the audit committee, which is a committee to
the board of directors. In line with the research regarding audit committees, Olson (1999)
mentions that: “Audit committees should consider including industry specialists to shed a
light on any complexities that may be unique to the company or the industry.” Another
interesting study was conducted by Beasly et al (2000) in which they researched the impact of industry expertise on the accounting quality of companies. They concluded that:
“The application of industry-specific accounting standards requires industry knowledge.”
These recommendations strengthen the need for board directors with industry expertise.
Cohen et al (2014) conclude that: “Audit Committees and board directors may lack sufficient industry expertise to understand and thus properly monitor industry-specific accounting issues.” This implies that industry expertise is important for non-executive directors, and thus independent outside directors, to be able to assess and manage possible risks that the company might incur. Yatim (2009) further suggests that: “Because the audit committee members with financial backgrounds have the expertise and training to
understand the risk management activities, it is expected that firms with at least one financially knowledgeable director on their audit committees to engage more actively in risk management process.” This finding suggests that the industry background of the director can have an impact on the quality of risk management and disclosure. Although being related to the audit committee and financial expertise, it does have implications for this research, since the background and expertise of the director has been analyzed and proven to be beneficial to the quality of risk management. Finally, Wang, Xie and Zhu (2015) stated that: “Independent directors experienced in the industry in which a firm operates are potentially more effective monitors because their industry expertise enables them to better understand the firm’s unique challenges and opportunities, analyze any information pertinent to the firm’s operation and financial conditions, and ultimately evaluate the decision making by the firm’s managers.” These studies mainly focused on the standalone variable industry expertise. This research will implement industry expertise as a moderator, which has not been researched yet. However, these findings in previous literature suggest that industry expertise has a positive moderating role on the relationship between independent board directors and the quality of risk disclosure, which leads to the following hypothesis:
H2: Industry expertise has a positive moderating role on the relationship between board
independence and the quality of risk disclosure by companies.
Figure 1 Conceptual framework
3. Methodology 3.1 Sample
This research will analyze premium listed companies on the UK stock market, for the period 2010-2016. The UK has been chosen because this study is using data collected for the research paper of Karaibrahimoglu and Porumb (2019), which is aimed at the UK premium listed firms. The database used for this paper consists of hand collected data regarding the quality of risk disclosure for the period 2007-2016. This research did not include the years 2007-2009, since it was expected that there might be a confounding impact of the financial crisis on the results of this research. Thus, this research uses the database to collect the data for the years 2010-2016.
To collect all the data to analyze the two hypotheses, the BoardEx database and
Compustat database were used. The variables that have been collected are consisting of data from UK Premium listed firms. To combine the data, each set with data has been merged, by using a code that synchronized the variables into one dataset, namely the code
“-ISIN & year”. All UK premium listed firms have been considered in this research. This resulted in a total sample of 8,397 observations on IFRS 7 Disclosure quality. The
analysis conducted did not retrieve all these 8397 observations. By dropping the missing variables, a smaller sample was comprised to analyze the hypothesis. The IFRS Disclosure quality had 1771 missing variables, and board independence 3115. The control variable firm size finally had 517 missing variables. Subsequently, the total sample for the analysis of hypothesis 1 consisted of 2994 observations. For the second hypothesis, the variable industry expert was added to the analysis. This variable did not have any missing
Industry Expertise
Quality of Risk Disclosure Board Independence
H2: +
H1: +
variables, compared to the other variables, and thus the sample for hypothesis 2 also consisted of 2994 observations. Table 1 gives an overview of the sample that has been composed with the use of the database provided by Karaibrahimoglu and Porumb (2019), the BoardEx database, and Compustat Global.
Table 1
Sample Composition
Total observations IFRS 7 Disclosure Quality 8397
Missing data IFRS 7 Disclosure Quality -1771 Missing data Board Independence -3115 Missing data control variable Firm size -517
Sample size for Hypothesis 1 and Hypothesis 2 2994
3.2 Research Design
To test the first hypothesis; “There is a positive relationship between the board independence and the quality of risk disclosure”, this research will use the following model:
!"#$_&'
"#= )
$+ )
%+,'_-'
"#+ )
&._/+01
"#+ )
'._2&3.
"#+ )
(._$14
"#+ )
)51,'1&
"#+ )
*-3#&'_/+01
"#+ )
+61#&_'
"#+ )
,+,'"/7&6_'
"#+ 9
"#Model 1
!"#$_&'
"#represents the weighted average score of risk disclosure quality, which comprises of thirteen individual aspects. To enhance the test of the first and second hypothesis, as well as the additional analysis, the two aspects of risk disclosure quality, Risk and Assurance, will be measured individually for their relationships with the variables in the models, with their respective names !"#$_&'_&+/:
"#, and
!"#$_&'_#//"&#,;1
"#. In Model 1, the variable of interest is ,'_-'
"#, which represents the percentage of independent outside directors on the board of the companies.
This variable is, according to the first hypothesis, expected to be a positive coefficient.
Furthermore, this model uses the control variables firm size, firm profitability, firm leverage, firm growth, the percentage of female directors on the board of companies, and board size. Finally, this model comprises of two dummy variables, namely year and industry type. Each variable that has been used is explained in the variable paragraph of this research.
To test the second hypothesis: “Industry expertise has a positive moderating role on the relationship between independent board directors and the quality of risk disclosure”, the following model has been used:
!"#$_&'
"#= )
$+ )
%+,'_-'
"#+ )
&+,'"/7&6_1<21&7/
"#+ )
'+,'_-'
"#= +,'"/7&6_1<21&7/
"#+ )
(._/+01
"#+ )
)._2&3.
"#+ )
*._$14
"#+ )
+51,'1&
"#+ )
,-3#&'_/+01
"#+ )
-61#&_'
"#+ )
%$+,'"/7&6_'
"#+ 9
"#Model 2 +,'"/7&6_1<21&7/
"#describes the industry expertise of board members and calculated as the percentage of independent directors who have 5 years or more of experience in the industry in which the company operates. The variable of interest in Model 2
is +,'_-'
"#= +,'"/7&6_1<21&7/
"#. According to the second hypothesis, this
coefficient is expected to be positive. Furthermore, This model uses the control variables firm size, firm profitability, firm leverage, firm growth, the percentage of female directors on the board of companies, and board size. Finally, this model comprises of two dummy variables, namely year and industry type. Each variable that has been used is explained in the variable paragraph of this research.
This research uses an Ordinary Least Square (OLS) regression to test all hypotheses. Year
and industry indicators are implemented to control for year and industry fixed effects. To
execute the regressions, this analysis originally used clustering to make the regressions
more robust. The clustering was done by the use of the code unique. However, after
running the clustered regressions, the results were insignificant. Then, after running the
regressions without clustering, this research found significant results for the hypotheses
with the use of the two models. These results were used in the results section, which can
be found at the results section of this paper. The next section will briefly describe each variable that has been used in the models to test the two hypotheses.
3.3 Variables
Dependent Variable
The quality of risk disclosure
Risk disclosure consists of the reporting regarding risks that the company is facing or might be facing in the future. There are several approaches to assess the quality of risk disclosure by companies. Beretta and Bozzolan (2004) define risk disclosure as: “The communication of information concerning firms’ strategies, characteristics, operations, and other external factors that have the potential to affect expected results.” Because this definition was further explored and improved by Linsley and Shrives (2006), this research will use their definition of risk disclosure: “Informing the reader about any opportunity or prospect, or of any hazard, danger, harm, threat or exposure, that has already impacted upon the company or may impact upon the company in the future or of the management of any such opportunity, prospect, hazard, harm, threat or exposure.” To measure the quality of this risk disclosure, there are two aspects that need to be considered, namely the
exposure of the risks and how the company hedges these risks. This research will be conducted by using an existing index that assesses the quality of the risk disclosure by companies. The quality of risk disclosure will be measured by using 13 aspects of quality.
These 13 aspects are constructed by using the items provided by IFRS 7 which have the highest relevance for financial statement users. The classification of these items are described by the IASB’s 2017 discussion paper on Principles of Disclosure and the
document “User Perspective on Financial Instrument Risk Disclosures under International Financial Reporting Standards” issued by CFA Institute (2016). The data for 2010-2016 has been hand collected by the research project of Karaibrahimoglu and Porumb (2019).
This is beneficial for the validation of this research, because Fogarty (2006) mentions that
it is advised that at least one variable should be hand collected in order to increase the
reliability. The risk disclosure quality aspects can be found in the appendix. The score is
measured by the normalized value of the weighted average of the 13 items that determine
the quality of disclosure. Furthermore, the quality of risk disclosure will be broken down
into three aspects, namely the overall quality, risk quality and assurance quality. This way,
the hypotheses can be tested more thoroughly to see if they have significant associations
with each other.
Independent Variable Independent board directors
To measure the variable independent board directors, it is important to distinguish what is understood by this variable in this research. Independent board directors will be measured by the percentage of independent outside directors on the board of the company. Outside directors are, according to Lefort and Urzúa (2008): “Those board members elected without the controlling shareholder votes.” However, according to Peng (2004), there are outside directors who are not independent of the company. Thus, this research uses the definition of an independent outside director of Peng (2004): “Non-management directors who do not have family and/or professional relationships with the firm or firm
management.” This paper will use the BoardEx database to determine which director is independent. The database contains data regarding the function of the board directors and this way the independence can be calculated. This research will examine whether the percentage of independent outside directors on the board has an influence on the quality of risk disclosure by companies. Data will be retrieved from the BoardEx database.
Risk Committee
The independent variable Risk Committee will be used in the additional analysis of this paper. The variable RISK_COMMITTEE will be implemented in this analysis as a dummy variable, with a score of 1 if there is a separate Risk Committee in place, and a score of 0 if there is not a separate Risk Committee in place. Data will be retrieved from the BoardEx database.
Moderating Variable Industry Expertise
This research will focus on the industry expertise of members of the board of directors. To construct this variable, the study of Faleye et al (2014) will be used. Data will be collected from BoardEx to analyze the employment history of each independent director. Next, in line with Faleye et al (2014), the primary standard industrial classification is identified (SIC). A director is defined as an industry expert if he or she has had working experience of at least five years in the same industry in in which the company operates.
This study will examine whether the percentage of board directors with at least 5 years of
expertise in the industry in which the company operates has an influence on the quality of
risk disclosure. Data will be retrieved from the BoardEx database and the database of Porumb and Karaibrahimoglu.
Control variables
This research will control for several characteristics of the companies that are listed on the UK stock-market. These variables might infer with the quality of risk disclosure, board independence, and the moderating effect of industry expertise.
Firm Size
According to previous studies by Eng and Mak (2003): “Firm size is consistently
associated with increased disclosure levels.” This is why this research will implement this variable as a control variable. The firm size will be measured by the total assets at the end of the fiscal year. The data is collected by using the Compustat Global database.
Firm Profitability
The second control variable that will be implemented in this research is firm profitability.
Previous research by Singhvi and Desai (1971) suggests that: “Firms with good performance may have an incentive to disclose detailed information to support the continuation of their position and remuneration.” Firm profitability will be measured by using the return on equity (ROE) of the company at the end of the fiscal year. The data is collected by using the Compustat Global database.
Leverage
Thirdly, voluntary disclosure may reduce agency costs for companies with a high level of debt (Jensen & Meckling, 1976). Since the disclosure is voluntarily in the United
Kingdom, leverage is added as a control variable and it will be measured by comparing the total debt and total assets of the company at the end of the fiscal year. The data is collected by using the Compustat Global database.
Board size
Board size will be used as another control variable for this study. Boone et al (2007)
found that board size reflects a tradeoff between the firm-specific benefits of increased
monitoring and the cost of such monitoring. This monitoring role, according to Boone et
al (2007) is associated with the number of independent directors on the board. Thus, board
size could have a role in the percentage of independent board directors at companies and
therefore will be used in the analysis. Board size will be measured by assessing the total
number of directors in the board of directors at the company. The data is collected using the BoardEx database.
Board members gender
“A more gender diverse board may also improve a firm’s competitive advantage if it improves the image of the firm and if this has a positive effect on customers’ behaviour and thus on a firm’s performance” (Smith et al., 2006). The gender of board members thus has an effect on the reputation of the firm. Since IFRS 7 leaves room for judgment, there is much variability in the disclosure quality by companies. The reputation could play a role in the quality of risk disclosure. Board members gender will be measured by analyzing the percentage of female directors on the board of the company. The data is collected by using the BoardEx database.
Dummy variables
This research will control for two other variables. First, a dummy variable will be created to control for the various years, namely YEAR_D. The years will be analyzed individually, from 2010-2016. Furthermore, this study will control for the industry in which the
companies operate in, by creating another dummy variable, namely INDUSTRY_D. there are 9 types of industries that are analyzed with the use of the data, which are basic materials, consumer foods, consumer services, healthcare, industrials, oil and gas,
technology, telecommunications, and utilities. Table 2 gives an overview of the variables
that will be used in this paper.
Table 2 Variable definitions
Variable Description
Dependent variable
Quality of Risk Disclosure (QUAL_RD)
QUAL_RD_ASSURANCE QUAL_RD_RISK
Use of an index to give a score to assess the quality of risk disclosure (appendix 1). Score is calculated by the normalized value of the weighted average of the 13 aspects that determine the quality of risk disclosure.
Furthermore, the assurance and risk aspects will be measured individually. The index has separate scores for both the assurance and risk aspects of the quality of risk disclosure, and the overall score will be calculated in the same way as the overall risk disclosure quality by companies
Independent variable
Independent Board Directors (IND_BD) The percentage of independent outside board directors on the board of companies
Moderating variable
Independent directors with industry expertise (INDUSTRY_EXPERTS)
The percentage of independent directors who have 5 years or more of experience in the industry in which the company operates
Control variables
Firms size (F_SIZE) The natural logarithm of total assets at the end of the fiscal year
Firm profitability (F_PROF) Return on Equity (ROE) at the end of the fiscal year
Firm leverage (F_LEV) Total debt / total assets at the end of the fiscal year
Board size (BOARD_SIZE) The number of directors on the board of directors
Gender (GENDER) The percentage of female directors on the
board of the company
Year dummy (YEAR_D) Dummy variable to control for the year in which the company is analyzed
Industry dummy (INDUSTRY_D) Dummy variable to control for the industry in
which the company operates
4. Results 4.1 Descriptive Statistics
Table 3 presents the descriptive statistics that show that variables used in this paper to analyze the two hypotheses. The three dependent variables have been used in both models. Model one used the independent variable IND_BD with the control variables. Finally, to test the second hypothesis, the moderating variable INDUSTRY_EXPERTS was used, alongside with all the other variables.
TABLE 3 Descriptive Statistics
Sample used for model 1 and 2
Variables Obs. Mean Std. Dev. Min Max
Dependent Variables
QUAL_RD 2,994 .401 .164 0 .782
QUAL_RD_ASSURANCE 2,994 .372 .255 0 .917
QUAL_RD_RISK 2,994 .405 .246 0 1
Independent Variables
IND_BD (%) 2,994 50.035 21.630 5.580 91.570
Moderating Variables
INDUSTRY_EXPERTS 2,994 .111 .039 0 .205
Control Variables
F_SIZE 2,994 14.694 1.252 11.670 18.043
F_PROF 2,994 0.036 .096 -.318 .285
F_LEV 2,994 0.262 .185 0 .970
BOARD_SIZE 2,994 8.850 2.070 5 16
GENDER (%) 2,994 18.233 8.934 0 46.763
All the variables have been winsorized on a 1% and 99% level.
The variables have been winsorized at the 1% and 99% level. The average quality of risk disclosure in the samples is 0.401 (rounded), which is the weighted average of the scores given to the 13 aspects of risk disclosure quality, and the standard deviation is 0.164
(rounded). Breaking down this variable, the aspects which comprises of Assurance has a mean of 0.372 with a standard deviation of .255. The aspects that comprises of Risk have a mean of 0.405 and a standard deviation of 0.246 The independent variable Board Independence has averages 50,04% with a standard deviation of 21.63%. Finally, the moderating variable Industry Expertise averages 11,1% with a standard deviation of 4% The variables have been rounded to three decimals. Firm size was calculated as the natural logarithm of the total assets at the end of the fiscal year. Furthermore, it is worth mentioning that the amounts of
observations were equal for the first and second model.
4.2 Correlation matrix
In this section, the Pearson correlation matrix will be discussed, as described in table 4. The results of the first and second model are combined, since the number of observations with the respective variables did not differ between these models. In general, the explanatory power of the variables used is relatively low. However, there are a few correlations that are worth discussing. First, the quality of risk disclosure is highly correlated with risk assurance (0.84), which is rather logical since it is a component of the overall risk disclosure quality. However it is worth mentioning because the risk aspect of the quality of risk disclosure has a
significantly lower correlation with the quality of risk disclosure (0.51). Thus, there is a difference in the correlation between these aspects of risk disclosure quality. Furthermore, the correlation of the variable board independence with the quality of risk disclosure is
interesting. The overall risk disclosure quality is negatively correlated with board
independence (-0.0273), as well as the risk component (-0.0271), as opposed to the assurance component, which has a positive correlation (0.0375). This could partially be explained by the first hypotheses that will be tested and the support of the underlying theories; board
independence is positively associated with the quality of risk disclosure. Almost the same phenomena occurs if we look at the correlation between risk disclosure quality and industry experts. Here, a negative correlation can be seen at the overall (-0.09) and assurance quality (- 0.15) of risk disclosure, but a positive correlation with the risk component of the quality of risk disclosure (0.09). Furthermore, the negative correlation between industry experts and gender (-0.67) should be mentioned. This implies that the number of industry experts have a negative effect on the percentage of women on the board of companies. One possible
explanation could be that men have a longer tenure at the company than women. Finally, if
we look at the control variables, the correlation between gender and firm size is relatively
high (0.54). This implies that if the firms are larger, the percentage of women on the board of
the respective firm will increase.
Table 4 Correlation Matrix
The description of the variables in this correlation matrix can be found in table 2. The sample for this correlation matrix was the same for model 1 and model 2, since there was no difference in the number of observations between the models with the use of the variables needed.
Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
QUAL_RD (1) -
QUAL_RD_ASSURANCE (2) 0.8378 -
QUAL_RD_RISK (3) 0.5080 0.2233 -
IND_BD (4) -0.0273 0.0375 -0.0271 -
INDUSTRY_EXPERTS (5) -0.0866 -0.1505 0.0906 0.0840 -
F_SIZE (6) 0.0841 0.0652 -0.0363 0.0332 -0.4050 -
F_PROF (7) -0.0722 -0.0789 0.1223 -0.0313 0.0313 -0.0579 -
F_LEV (8) 0.1564 0.1282 0.1358 0.1110 0.0268 0.1231 0.0360 -
GENDER (9) 0.0520 0.1156 -0.1389 -0.0801 -0.6690 0.5380 -0.0831 0.0457 -
BOARD_SIZE (10) 0.0018 0.0001 -0.0437 0.1705 -0.0622 0.2870 0.1038 0.0789 0.0848 -
4.3 Multivariate Analysis
Table 5 represent the multivariate analysis for the two hypotheses of this research. The models have been displayed separately. Panel A represents model one (hypothesis 1) and Panel B represents model two (hypothesis 2). The observations will be discussed with regard to each of the two models.
In the first hypothesis, this research expected that board independence has a positive relationship with the quality of risk disclosure by companies. Model 1 shows the results of the first hypothesis. The overall risk disclosure quality has a positive significant (P<0.1) relationship with board independence. The assurance aspect of the quality of risk
disclosure also has a positive significant (P<0.01) relationship with board independence.
However, the risk aspect of the quality of risk disclosure has a negative significant (P<0.1) relationship with board independence. Because this research focuses on the overall quality of risk disclosure, this study mainly analyzes the relationship with this variable with board independence. The assurance aspect and risk aspect are taken into account in the tables, which makes the regression more robust and explainable. Since the results of the first model show a positive significant relationship between board
independence and the quality of risk disclosure by companies, the first hypothesis is accepted.
To test the second hypothesis, the positive moderating role of industry expertise on the relationship between board independence and the quality of risk disclosure, model 2 is used. The second part of table 5 shows the results of this model. The overall risk disclosure quality has a positive significant relationship (P<0.01) with the moderator, which constitutes of the role of industry experts and the board independence. This means that the interaction between industry experts and board independence has a positive effect on the quality of risk disclosure, in accordance with the hypothesis. This positive
significant relationship is further supported by the risk aspect of disclosure quality (P<0.05). The assurance aspect also illustrates a positive relationship, however this is not significantly supported. Given the results presented in model 2, there is enough evidence to accept the second hypothesis.
It is important to mention that in model 2, the independent variables board independence
and industry expertise are negatively significant (P<0.01) related to the quality of risk
disclosure. This is an interesting result, since the board independence in model 1 had a
positive significant relationship with the quality of risk disclosure. Industry expertise on
itself also had a negative relationship with the quality of risk disclosure, as opposed to the positive relationship between the interaction of board independence and industry expertise on the relationship with board independence and the quality of risk disclosure.
By analyzing the control variables, this study finds that in both models almost all control variables are significant, except board size in the first model. This research mainly analyzes the overall risk disclosure quality, and with this variable, the results are statistically significant.
Table 5
Multivariate Analysis Panel A: Model 1
*** p<0.01, ** p<0.05, * p<0.1. Standard errors are presented in parentheses. The variables all have been standardized, which means that they have a standard deviation of one and a mean of zero. The two models are separately displayed. Model one tests the first hypothesis, model two tests the second hypotheses.
(1) (2) (3)
VARIABLES QUAL_RD QUAL_RD_ASSURANCE QUAL_RD_RISK
IND_BD 0.0352* 0.103*** -0.0325*
(0.0187) (0.0190) (0.0190)
F_SIZE 0.103*** 0.0433** 0.0364**
(0.0174) (0.0177) (0.0178)
F_PROF -0.182*** -0.180*** 0.0761***
(0.0180) (0.0183) (0.0183)
F_LEV 0.113*** 0.0797*** 0.102***
(0.0174) (0.0177) (0.0177)
GENDER -0.125*** -0.102*** -0.0699***
(0.0202) (0.0206) (0.0206)
BOARD_SIZE 0.0334 0.137*** -0.121***
(0.0204) (0.0208) (0.0208)
Constant -1.492*** -0.649** -0.477*
(0.256) (0.261) (0.261)
YEAR_D YES YES YES
INDUSTRY_D YES YES YES
Observations 2,994 2,995 2,995
R-squared 0.201 0.217 0.195
Panel B: Model 2
*** p<0.01, ** p<0.05, * p<0.1. Standard errors are presented in parentheses. The variables all have been standardized, which means that they have a standard deviation of one and a mean of zero. The two models are separately displayed. Model one tests the first hypothesis, model two tests the second hypotheses.
(1) (2) (3)
VARIABLES QUAL_RD QUAL_RD_ASSURANCE QUAL_RD_RISK
IND_BD -0.164*** 0.0847* -0.133***
(0.0449) (0.0457) (0.0463)
INDUSTRY_EXPERTS -0.334*** -0.238*** -0.0237
(0.0395) (0.0402) (0.0407)
IND_BDxIN_EXPERTS 0.283*** 0.0280 0.141**
(0.0578) (0.0587) (0.0595)
F_SIZE 0.104*** 0.0364** 0.0416**
(0.0173) (0.0176) (0.0178)
F_PROF -0.177*** -0.187*** 0.0840***
(0.0179) (0.0182) (0.0185)
F_LEV 0.114*** 0.0845*** 0.0998***
(0.0172) (0.0174) (0.0177)
GENDER -0.133*** -0.108*** -0.0695***
(0.0200) (0.0203) (0.0206)
BOARD_SIZE -0.0783*** -0.00538 -0.0856***
(0.0251) (0.0255) (0.0258)
Constant -1.501*** -0.545** -0.554**
(0.254) (0.258) (0.262)
YEAR_D YES YES YES
INDUSTRY_D YES YES YES
Observations 2,994 2,995 2,995
R-squared 0.223 0.240 0.198
To visualize the interaction between board independence and industry expertise and the relationship with the quality of risk disclosure by companies, a graph was constructed, which can be found in figure 2. This graph shows that there is an interesting relationship between the variables that have been used. As can be seen in the graph, the more company boards consist of independent outside directors and industry experts, the higher the quality of risk disclosure will be of this respective company. This is visualized by the red, ascending line in this graph, where you can see this relationship between board independence, industry expertise and the quality of risk disclosure by the companies that have been analyzed in this research.
Figure 2
Visualization of the interaction and relationship
4.4 Additional Test
This research further comprises of an additional test. In this test, the existence of a Risk Committee at UK premium listed firms has been analyzed. According to the Corporate Governance code (2018), the establishment of a separate Risk Committee is not
mandatory in the United Kingdom. Field, Lowry and Mkrtchyan (2013) mentioned that:
“Given the complexity of the numerous committees in the financial reporting process, the audit committee might not have sufficient time, skills, and support to assess the firms’
overall risks.” Dobler mentions (2008): “A Risk Committee concentrates on, and
specializes in, risk monitoring and risk management.” Al-Hadi, Hasan and Habib (2016) found that: “Firms with a separate Risk Committee are associated with greater market risk-disclosure.” Building upon this previous research, it is expected that the existence of a separate Risk Committee has a positive influence on the quality of risk disclosure by companies. This relationship will be tested by adding the independent variable
RISK_COMMITTEE to model 2 of our analysis. By doing this, the impact of the existence of a Risk Committee on our principal test can be analyzed, namely the influence of board independence on the quality of risk disclosure, with the addition of the moderator industry expertise. The firms that did have a Risk Committee in place received a score of 1, and the firms that did not have a Risk Committee in place received a score of 0. The board committee names were collected from the BoardEx database. Every committee of which the name comprised of “Risk” or contained the word “Risk”, was considered as a separate Risk and thus received a score of 1 All other names of board committees were given a score of 0.
table 6 shows the results of the additional analysis.
The results show that the existence of a separate Risk Committee has a negative significant
(p<0.01) relationship with the quality of risk disclosure by companies. This implies that the
existence of a separate Risk Committee has a negative influence on the quality of risk
disclosure, which contradicts our prediction of a positive influence of the existence of a
separate Risk Committee. The significant negative relationship also exists at the assurance-
side of the risk disclosure quality, which means that the companies with a separate Risk
Committee negatively influence the assurance-aspect of risk disclosure quality.
Table 6
Addition Test – The existence of a separate Risk Committee
*** p<0.01, ** p<0.05, * p<0.1. Standard errors are presented in parentheses. The variables all have been standardized, which means that they have a standard deviation of one and a mean of zero.
