The influence of board characteristics and the occurrence of an
IC-failure on future risk reporting
Master thesis Accountancy, Faculty Economics and Business
June 13, 2017 C.M. Worst Student number: 2177692 + 31 (0) 6 19 69 52 90 Floresstraat 5b, Groningen c.m.worst@student.rug.nl Supervisor: G.C. Helminck Co-Supervisor: Prof. dr. J.A. Emanuels
Assessor: Dr. V.A. Porumb
Abstract
In the last few years, risk reporting has received considerable attention due to accounting scandals. As a result, stakeholders demand more non-financial information in annual reports, such as risk disclosure. This thesis examines the relationship between board characteristics, the occurrence of an IC-failure and the change over time in risk reporting. The sample of this study consists of 300 observations of firms that are listed on the U.S. stock exchange. In order to measure the change over time in risk reporting, the number of risk-related words in certain paragraphs of the annual report were counted and compared with the year after. By testing the hypotheses, the relationship between board size, gender diversity, occurrence of an IC-failure, and the change over time in risk reporting were measured. No significant associations were found in this study.
2 Table of Contents
1. Introduction ... 3
1.1 Motive ... 3
1.2 Relevance ... 4
1.3 Design of the study ... 5
2. Literature and background ... 6
2.1 Risk reporting ... 6
2.2 Agency theory and stakeholder theory ... 7
2.3 Upper Echelon theory ... 7
2.4 Change over time in risk reporting ... 8
2.5 Size of the board ... 8
2.6 Gender diversity ... 10
2.7 Conceptual framework ... 11
3. Research design ... 12
3.1 Data collection method and dataset ... 12
3.2 Dependent variable: change over time in risk reporting ... 12
3.3 Board size ... 13
3.4 Gender diversity ... 13
3.5 Occurrence of an IC-failure ... 14
3.6 Control variables ... 14
3.7 Model and statistical analysis ... 15
4. Results ... 17 4.1 Descriptive Statistics ... 17 4.2 Correlation analysis ... 18 4.3 Hypotheses ... 18 4.4 Robustness check ... 21 5. Conclusion ... 22
5.1 Conclusion main research question ... 22
5.2 Suggestions and limitations ... 23
3 1. Introduction
The emphasis of this study is on the change over time in risk reporting. In recent years, the subject risk disclosure has been gaining attention. However, little research has been done on the change over time in risk reporting: previous studies focus on risk reporting in one particular year. This thesis studies the influence of the Board of Directors (hereafter referred to as board) and the occurrence of an IC-failure on the change over time in risk reporting. This chapter introduces the topic, first discussing the motive of this study. Then the relevance of this study is explained. Subsequently, the design of this research is described.
1.1 Motive
In the last decade, there have been many accounting scandals due to problems with internal control (hereafter referred to as IC). These IC-failures arise from various sources, such as problems with financial reporting, strategic choices, business operations or compliance. This affected well-known organizations such as Ahold, Enron, and WorldCom. Recently, Wells Fargo, an American bank, is embroiled in a scandal over assertions that the employees of the bank secretly created millions of unauthorized bank- and credit card accounts, while the customers of the bank were not aware of this. The fake accounts earned the bank unwarranted fees and allowed employees to earn more money. This had serious consequences for Wells Fargo as a whole, in terms of reputational damage and financial loss, but also affected individual board members.1
In 2002, to prevent IC-failures, the U.S. government imposed the Sarbanes Oxley Act (hereafter referred to as SOx). This law contains legislation about the quality of IC reporting in order to improve corporate disclosure in terms of reliability and accuracy. The purpose of SOx is to protect investors and to restore the public confidence in firms. In addition, SOx discusses topics on risk control, risk disclosure, and provides a clear overview of the responsibilities of the board. The board makes important decisions about whether to publish additional non-financial information, such as risk reporting (Prado-Lorenzo & Garcia-Sanchez, 2010). In sum, this study aims to provide a more comprehensive understanding of the manner in which board characteristics and the occurrence of an IC-failure influence the change over time in risk reporting.
4 1.2 Relevance
The traditional financial section in annual reports is not sufficient for stakeholders to meet their information needs (Maines et al., 2002). For that reason, to provide additional disclosure, the non-financial section is receiving more attention. Examples of interest areas are matters relating to social-, environmental-, and risk reporting of a firm. The discussion on the importance of risk disclosure started in 1997, when the “Financial Reporting of Risk: Proposals for a Statement of Business Risk” was published. The report suggested that the board should provide risk information in the annual report to facilitate informed decision-making for the stakeholders.
The reporting style of a company is seen as an important factor on the relationship between the firm and their stakeholders. Companies can use this for their benefit and to divert attention from issues by highlighting positive points (Deegan, 2002). Furthermore, firms can affect their communication by choosing certain words and by the amount of information they disclose (De Waard, 2014). Therefore, a firm can change their risk reporting strategy to influence the opinions of the stakeholders. Prior research examined risk reporting at a certain moment (Beretta & Bozzolan, 2004; Linsley & Shrives, 2006; Abraham & Cox, 2007; Miihkinen, 2012; Ntim et al. 2013; Abraham & Shrives, 2014). This study however, investigates the change over time in risk reporting, which makes this research relevant and innovative.
Prior research highlighted the influence of board characteristics on corporate governance. A board needs members with diverse skills, experience, and knowledge to perform its tasks well. Therefore, board diversity is an important factor (Hillman & Dalziel, 2003). Some researchers classify a small board as effective, while other researchers argue for a large board. Besides board size, gender diversity also plays a role. Frias-Aceituno et al. (2013) show that the size and the gender diversity of the board affect the external non-financial disclosure such as risk disclosure. Therefore, board characteristics may affect the change over time in risk reporting. To enrich the literature, it is interesting to examine the influence of board characteristics; the occurrence of an IC-failure on the change over time in risk reporting.
5 Therefore this thesis will attempt to answer the following research questions:
- What is the effect of the size of the board on the change over time in risk reporting? - What is the effect of the occurrence of an IC-failure on the relationship between the size of
the board and the change over time in risk reporting?
- What is the effect of gender diversity on the change over time in risk reporting?
- What is the effect of the occurrence of an IC-failure on the relationship between gender diversity and the change over time in risk reporting?
1.3 Design of the study
The remainder of this thesis is organized as follows: Chapter 2 describes the theoretical framework used to investigate the relationships between board size, gender diversity, the occurrence of an IC-failure, and the change over time in risk reporting. Chapter 3 explains the research methodology. In Chapter 4, the results will be discussed. Finally, Chapter 5 provides the conclusion, the limitations of this study, and suggestions for future research.
6 2. Literature and background
This chapter starts with a definition of risk reporting, followed an explanation of the agency theory in combination with the stakeholder theory, and a description of the upper echelon theory (the following Sections 2.1, 2.2 & 2.3). The theories in this study are discussed to gain insight into the change over time in risk reporting (Section 2.4). This is used to draft the hypotheses for the influence of the size of the board, gender diversity, and the occurrence of an IC-failure on the change over time in risk reporting (Sections 2.5 & 2.6). Finally, Section 2.7 gives a conceptual model based on the established hypotheses.
2.1 Risk reporting
Previous research shows that the risk reporting of a firm is positively influenced by various factors, such as size (Linsley & Shrives, 2006; Abraham & Cox, 2007), profitability (Miihkinen, 2012), and leverage (Deumes & Knechel, 2008; Elshandidy et al., 2013). However, to perform a study on risk reporting, the concept ‘risk’ should first be investigated. There are many definitions of risk. COSO (2004) states that events can have both negative- and positive influences, but that something is only a risk if the event creates uncertainty for the firm to achieve their goals.
Emanuels et al. (2010b, p. 150) define risk management as “a system that enables management to identify, prioritize, analyse, and control the relevant risks which may be threatening the organization’ ability to meet its objectives”. There are also multiple definitions of risk disclosure (Beretta & Bozzolan, 2004; Miihkinen, 2012). However, in this thesis, risk reporting is defined by Linsley & Shrives (2006) as “informing the reader of any opportunity, prospect, hazard, 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”. This definition of risk disclosure contains both the positive- and the negative aspects of risks and will therefore be used when testing the hypotheses.
7 2.2 Agency theory and stakeholder theory
Based on the agency theory, the purpose of risk reporting could be placed into a broader perspective. Jensen & Meckling (1976) define the agency theory as “a contract under which one or more persons (the principals) engage another person (the agent) to perform some service on their behalf, which involves delegating some decision-making authority to the agent”. Information asymmetry occurs when agents of a firm have more information than the principals, this allows the agent to behave selfishly and make sub-optimal decisions, which is called the agency problem (Eisenhardt, 1989). A mechanism to reduce information asymmetry is issuing voluntary information, such as additional risk disclosure (Fama & Jensen, 1983).
In the stakeholder theory, not only shareholders, but also other stakeholders, are taken into account as principals: clients, investors, employees, suppliers, government, and environmental organizations (Hill & Jones, 1992). It is difficult for stakeholders to determine whether a company acts according to their interests. Therefore, the role of the board is to monitor the executive management to ensure that they act in the interest of the stakeholders (Erickson et al., 2005). Voluntary disclosure in the annual report can assist the stakeholders in verifying how well the board performed its monitoring function.
2.3 Upper Echelon theory
The board can be seen as a firm's governance mechanism, which monitors the actions of the executive directors and ensures that they act in the interest of the shareholders, which reduces the information asymmetry (Durnev et al., 2009). The Upper Echelon theory states that the strategy and decision-making of a firm is influenced by the characteristics of the board (Hambrick & Mason, 1984). Prior research shows that the board plays an important role on the decision to publish financial and non-financial information, since the board is responsible for the reporting policy of a firm, so whether to publish additional risk information (Chen & Jaggi, 2000; Bamber et al., 2010). A distinction can be made between the observable characteristics, such as age, ethnicity, and gender and unobservable characteristics such as personal values and beliefs (Milliken & Martins, 1996). In this study the focus is on observable board characteristics, namely board size and gender diversity.
8 2.4 Change over time in risk reporting
In this study, a change over time in risk disclosure is expected. In the last few years, risk reporting has received considerable attention (Power, 2004). Linsley & Shrives (2006) concluded that current annual reports disclose risk information in an incomprehensible way. Stakeholders use this information to get an understanding of the risk profile of the firm, which influence their future decision-making. The board monitors the executive management, to verify if they act in the interest of the stakeholders. The board discloses its findings to the stakeholders, so the annual report gives a representation that the board performed its tasks. In addition, the compensation of the board could be stock-based, publishing additional risk information could increase the share value of the firm. Finally, risk disclosure could be used to outperform other firms in the same industry, so the best-performing firm raises additional capital. Risk reporting is then used by the board as a tool to manage the stakeholders in their advantage (Deegan, 2002). Therefore, in this paper, the combined agency-, stakeholder- and Upper Echelon's theory could explain the change over time in risk reporting, which is expected to be strengthened by the occurrence of an IC-failure. In the next sections the independent- and moderating variables are discussed in more detail.
2.5 Size of the board
The size of the board has been extensively investigated in corporate governance studies (Larcker et al., 2007). The opinions about the board size are divided; some researchers argue that a small board is more effective than a large board. Previous studies show that smaller boards are related to good governance (DeFond et al., 2005; Hoitash et al., 2009; Hunton et al., 2011). Eng & Mak (2003) found that a larger board causes a decrease in the amount of firm specific disclosure. An explanation could be that the efficiency of decision-making and communication in large boards is lower, which increases the previous mentioned agency problems (Hermalin & Weisbach, 1991; Lipton & Lorch, 1992). Based on these results, a small board is preferable compared to a large board.
However, there is also evidence in favour of a large board. Larger boards will lead to an increase in knowledge, experience, and higher monitoring capacity, which increases the firms’ value (Bennett & Robson, 2004; Dalton & Dalton, 2005). In addition, a larger board has a stronger position to resist the executive board and to meet the expectations of the stakeholders with regard
9 to accountability (Abraham & Cox, 2007). Ntim et al. (2013) found that large boards are positively associated with the level of risk disclosure. This strengthens the assumption that the board plays a role in the risk reporting strategy of a company. In the study of Salmon (1993) the arguments mentioned above are combined. He argues that, a board must be large enough to have sufficient knowledge and expertise, but that it should not be too large that it prohibits efficient decision-making.
Based on the literature described above, this study expects to find that board size has an impact on the change over time in risk reporting. However, the direction of this relationship is unknown. There are no previous studies on the relationship between the size of the board and the change over time in risk reporting. Therefore, the following hypothesis is suggested to enrich the literature:
Hypothesis 1: There is a relationship between the size of the board and the change over time in risk reporting.
In this thesis, it is expected that the occurrence of an IC-failure strengthens the relationship between board size and the change over time in risk reporting. Because, after an IC-failure, the trust of the stakeholders in a firm can be decreased, the board publishes additional risk information to regain their confidence (Reynolds et al., 2006). Additionally, the board fulfils a monitoring function to oversee whether the executive management acts in the interest of the stakeholders. So, the board uses risk reporting to share their findings with the stakeholders. After an IC-failure, the urge of the board to show the stakeholders that they performed their tasks well increases, so the board uses risk disclosure as a means to show their accountability. The opinions about the size of a board are divided; some researchers argue that a large board is less effective than a small board, while other researchers claim the opposite. Therefore the following hypothesis is proposed:
Hypothesis 2: The occurrence of an IC-failure strengthens the relationship between the size of the board and the change over time in risk reporting after an IC-failure.
10 2.6 Gender diversity
Literature on gender diversity found considerable differences in behavioural characteristics between males and females. For this study, it is relevant to examine whether gender diversity in the board influences the change over time in risk reporting. Previous studies found a positive relationship between the relative number of women in the board and the effectiveness of the board monitoring, which resulted in lower information asymmetry (Bear et al., 2010; Boulouta, 2013). This positive relation is due to the fact that women, rather than men, take the needs of the stakeholders into account (Hillman et al., 2002). This is because previous studies show that female board members are more caring, empathetic, and tolerate opportunistic behaviour less in their decision-making (Krishnan & Parsons, 2008; Boulouta, 2013). Besides, women report more voluntary information, including risk disclosure (Adams & Ferreira, 2009).
Another advantage of women in the board is that they contribute to a higher level of innovation within a firm (Carter et al., 2003). A change in the risk reporting strategy can be seen as an example of an organizational innovation. Therefore, women in the board are a good attribution to the implementation of a more extensive risk disclosure. Empirical studies on gender diversity and risk reporting are scarce, but a positive relationship has been found between board diversity, based on ethnicity and gender, and risk disclosure (Ntim et al., 2013).
Some researchers question the positive effects of gender diversity. They argue that it has no effect, or may even have a negative impact on the boards functioning. The different characteristics between men and women are focused on the general population, however women in the board differ significantly from the population (Adams & Funk, 2012). They claim that gender diversity could lead to decision-making costs and conflicts between board members, which reduces the efficiency of the board as a monitoring body (Adams et al., 2015). Nonetheless, more support is found for the positive influence of gender diversity on risk reporting. Since there is no previous research on gender diversity that focuses on the change over time in risk reporting, the following hypothesis is suggested to enrich the literature:
Hypothesis 3: There is a positive relationship between gender diversity and the change over time in risk reporting.
11 In this study, it is expected that the occurrence of an IC-failure strengthens the urge for additional risk disclosure in the annual report. Since an IC-failure can decrease the trust the stakeholders have in a firm, a board could react by disclosing additional non-financial information (Reynolds et al., 2006). In addition, the board could use risk reporting as a means of showing that they have performed their duties well; that the IC-failure is not caused by the lack of supervision. Previous studies show that board characteristics, including gender diversity, play a role in the risk reporting strategy of a firm, since women monitor the actions of the executive management more stringently and disclose more information (Adams & Ferreira, 2009). Moreover, Ntim et al. (2013) found a positive relationship between gender diversity and the amount of risk disclosure. Therefore the following hypothesis is proposed:
Hypothesis 4: The occurrence of an IC-failure strengthens the positive relationship between gender diversity and the change over time in risk reporting after an IC-failure.
2.7 Conceptual framework
Based on the formulated hypotheses in the previous sections, a conceptual model can be drafted. The direct relationship between the moderator, the occurrence of an IC-failure and the change over time in risk reporting is tested as well. Figure 2.1 shows the conceptual framework of this study.
+
+
Figure 2.1: Conceptual model
Board size
Change over time in risk reporting Gender
diversity
IC-failure IC-failure
12 3. Research design
This chapter describes the research design and the research method used in this thesis. This study is executed quantitatively using regression analyses to test the hypotheses. Section 3.1 explains the data collection method. Section 3.2 describes how the dependent variable is measured. Sections 3.3, 3.4 & 3.5 explain how the independent- and moderating variables are measured. Section 3.6 discusses the control variables and finally in Section 3.7 the model for the regression analyses is described in more detail.
3.1 Data collection method and dataset
By means of random sampling, 300 observations between the years 2004 and 2012 of firms listed on the U.S. stock exchange have been selected. The selection has taken into account the equal representation of all countries. The size of the sample is determined with the statistical program G-power. The sample consists of 235 observations of firms who had an IC-failure during the fiscal year; the other observations are of companies that had no material weakness in that specific year. The data for the sample collection is retrieved through Audit Analytics.
3.2 Dependent variable: change over time in risk reporting
The change over time in risk reporting is quantitatively measured based on the number of risk-related words. Previous studies on risk disclosure also used this method (Linsley & Shrives, 2006; Abraham & Cox, 2007; Ntim et al., 2013). The data for this variable is retrieved from 10-K annual reports; the SEC requires this form for firms that are listed on the U.S. stock exchange. It gives a comprehensive summary of a firms’ financial performance. In this study, the number of words published under Part I item 1A ‘Risk factors’ and Part II item 7A ‘Quantitative and qualitative disclosures about market risk’ are counted and divided by the total number of words in the 10-K form. Subsequently, the ratio of risk-related words is compared with the year after.
In this study, the ratio of risk-related words is used instead of the absolute number of risk-related words under paragraphs 1A and 7A of the annual report, since a firm can increase the amount of risk disclosure. However, if the amount of non-risk information also increases the impact remains the same. Therefore, the hypotheses are tested again, using regression analyses to verify if the results are robust, based on the absolute number of risk-related words.
13 3.3 Board size
As noted earlier in the paper, the literature has established a relationship between the size of the board and risk reporting. Within the sample, companies with a single-tier as well as a two-tier board are included. The executive and supervisory functions are mixed in a one-tier board structure. A two-tier board structure consists of two bodies: the executive-board for the daily activities and the supervisory board for monitoring the executive directors (Belot et al., 2014). In this study, the size of the board is defined as the number of non-executive directors on the board (Cheng, 2008; Coles et al., 2008; Ntim et al., 2013). Therefore, the difference between one- or two-tier board structure has no influence on the results of this research. The data of the board is obtained by using BoardEx, this database contains board specific information, such as board structure, the remuneration and the education of board members. For companies that are not listed on BoardEx, the annual report is used to retrieve the number of board members.
3.4 Gender diversity
Diversity of gender in the board is measured by the normalized index Blau (Harrison & Klein, 2007; Van Ees et al., 2007; Solanas et al., 2012). This indicator is also known as the Herfindahl index or Hirschman's index. The data for this variable is retrieved through BoardEx; for the companies that are not listed in this database, the annual report is used.
The normalized Blau index is calculated based on the following formula:
𝐵𝑙𝑎𝑢 𝑖𝑗 = 1 − ∑ (𝑥𝑖𝑗 𝑛𝑗)
𝐾
𝑖=1
2
For example, there are seven board members in total: 5 males and 2 females. Then, the gender diversity is as follows:
14 3.5 Occurrence of an IC-failure
The occurrence of an IC-failure is measured by the number of material weakness(es) reported under SOx sections 302 and 404. A “material weakness” is defined as “a deficiency, or combination of deficiencies, in internal control over financial reporting, such that there is a reasonable possibility that a material misstatement of the registrant’s annual or interim financial statements will not be prevented or detected on a timely basis”2 by the U.S. Securities and Exchange Commission
(hereafter referred to as SEC). SOx section 302 concerns the control of information disclosure: the board of a company should report periodically on the effectiveness of the controls. Section 404 establishes rules for internal control and financial reporting: the board is obliged to publish explicit information on the reliability of the internal controls that are used in the firm. The data for this variable is obtained by Audit Analytics.
3.6 Control variables
In order to optimize the test results, three control variables are used in this study: firm size, auditor, and sector. First, the influence of the size of the firm is discussed. Larger companies are more complex and have many varied activities, which means that they have to deal with more risks than smaller firms (Deumes & Knechel, 2008). Besides, according to the stakeholder theory, larger organizations have to disclose more risk information because they have more stakeholders (Jensen & Meckling, 1976). In prior research, a positive relation is found between firm size and risk disclosure (Beretta & Bozzolan, 2004; Linsley & Shrives, 2006). In this study, firm size is measured as the logarithm of the total assets, to decrease the distance between the values, since large differences may give a distorted picture when they are used for regression analyses. This metric for determining the size of a firm is commonly used in accounting studies (Bhagat & Black, 2002; Francis et al., 2004; Andres et al., 2005). The data for this variable is retrieved through Audit Analytics.
Previous research shows that an auditor of a big-4 accounting firm delivers a higher quality of audit reporting than an auditor of a non-big-4 audit firms (Becker et al., 1998). Therefore, big-4 audit firms reduce monitoring costs by providing a higher level of audit quality. This provides certainty to the stakeholders that use the annual report. In addition, an increase in disclosure is found when
15 large audit firms perform the audit (Xiao et al., 2004; Wang et al., 2008). Nowadays, the big-4 audit firms are EY, PwC, KPMG, and Deloitte. Consistent with other studies, a value of 1 is assigned when the external auditor is a big-4 audit firm and 0 if it is not (Xiao et al., 2004; Wang et al., 2008). The data for this variable is obtained by Audit Analytics.
Sector is a particular segment in which a large group of companies can be categorized. In this study two sectors are distinguished: the financial and non-financial sector. This is in line with previous studies (Beck et al., 2000; Laurenceson, 2007; Iyare & Moore, 2011). Because more comprehensive legislation is applied to the financial sector than to the non-financial sector (Koelewijn, 2012). In this study, the financial sector includes banks, pension funds, investment companies, insurance companies, and intermediaries. Subsequently, a score of 1 is assigned if the firm is categorized in the financial sector. A score of 0 is assigned if the organization is categorized in the non-financial sector. The data for this variable is obtained by Audit Analytics.
3.7 Model and statistical analysis
This study tests the relationships between board size, gender diversity, and the change over time in risk reporting. Additionally, it analyses the moderating role of the occurrence of an IC-failure on the relationship between board size, gender diversity and the change over time in risk reporting, while controlling for other drivers of risk reporting. Table 3.1 provides a summary of the variables, including their definitions. The basic model is the following linear regression:
Change RR = β0 + β1Board_size + β2Gen_div + β3Log_assets + β4Auditor + β5Ind_fin + β6(Size_board x Mat_weakness) + β7(Gen_Div x Mat_weakness) + ε
16 Table 3.1
Definitions of the Variables
Definition Explanation
Dependent Variable
Change_RR The number of words under paragraphs 1A & 7A in the 10-K form divided by the total number of words in the 10-K form
Independent Variables
Board_size The number of non-executive board members Gen_div
Moderating Variable
The normalized Blau index
Mat_weakness Dichotomous variable that equals 1 if one or more material
weaknesses is reported and 0 if no material weaknesses is reported in the fiscal year
Control Variables
Log_assets Auditor
The logarithm of total assets at the end of the year
Dichotomous variable that equals 1 if one of the Big-4 audit firms (PwC, Deloitte, EY, KPMG) is employed; in all other cases 0
Ind_fin Dichotomous variable that equals 1 if the firm is in the financial sector and 0 for firms in the non-financial sector
17 4. Results
In this chapter, the results of the study are presented. The results were obtained using the statistical program SPSS. The first section discusses the changes that were made to the data and explains the descriptive statistics. Section 4.2 presents the correlation analyses. The results of the regression analyses are given in Section 4.3. Finally, in Section 4.4, a robustness check is executed.
4.1 Descriptive Statistics
Before the data is used to perform statistical analyses, some changes are made to the data. For fifteen of the 300 observations in the selection, the board data was not available or the company only had an executive board. One of the companies did not have any assets and one company had no independent auditor; therefore, these companies were removed from the selection. The final sample consists of 283 observations based upon 242 firms, between the years 2004 and 2012. The descriptive statistics of the variables in this study are given in table 4.1. As displayed, the average change in risk reporting is 0.338%, with a maximum of 32.0% and a minimum of -19.4%. The average board consists of five people, with a standard deviation of 2.594, ranging between 1 and 16 directors. Of the boards in the sample, the mean of gender diversity is 0.109, which means that most boards consist of only men. In the sample, 77% of the observations are of firms who had an IC-failure. The mean size of the firms in the sample is $12 billion (M = 8.255). Roughly 51.6% of the companies are engaged with one of the big-4 auditors. Finally, most firms in the selection are in the non-financial sector, average 86.9% (M = 0.131).
Table 4.1 Descriptive statistics
Variables Minimum Maximum Mean (M) SD
Change_RR -19.424 32.002 0.338 3.919 Board_size 1 16 5.367 2.594 Gen_div 0 0.500 0.109 0.161 Mat_weakness 0 1 0.770 0.421 Log_assets 3.573 12.197 8.255 1.153 Auditor 0 1 0.516 0.501 Ind_fin 0 1 0.131 0.338
18 4.2 Correlation analysis
The correlation matrix of Pearson is presented in table 4.2. This table shows that none of the variables in this study correlates with the dependent variable, the change over time in risk reporting. However, the independent variables, board size and gender diversity positively significantly correlate with each other (r = 0.302, p < 0.01), implying that larger boards are more diversified when it comes to gender. In addition, board size (r = 0.341, p < 0.01), and gender diversity (r = -0.286, p < 0.01) are significantly, negatively, associated with the occurrence of an IC-failure, suggesting that the occurrence of an IC-failure has an influence on size of the board and the presence of female directors in the board. Table 4.2 shows that the highest correlation that exists is r = 0.524, between the independent variable, size of the board and the control variable, firm size. This means that there is no multicollinearity in this study, since none of the variables has a higher correlation than 0.7 / 0.8 (Blumberg et al., 2005).
Table 4.2
Correlation of the research variables
Variables (1) (2) (3) (4) (5) (6) (7) Change_RR - Board_ size -.100 - Gen_div. .024 .302** - Mat_weakness , .045 -.341** -.286** - Log_assets -.050 .524** .144* -.262** - Auditor -.028 .304** .155** -.176** .336** - Ind_fin -.017 .196** .153* -.137* .057 -.107 -
**, Correlation is significant at the 0.01 level. * , Correlation is significant at 0.05 level.
4.3 Hypotheses
Table 4.3 shows the results of the linear regression analyses between the explaining variables and the change over time in risk reporting. In this research, multiple regression analyses and moderation analyses are preformed based on six models. For a better understanding of the models the adjusted R-squared and the F-value of the research variables are given in table 4.3.
19 Table 4.3
Linear regression analyses
Variables (1) (2) (3) (4) (5) (6) Intercept 1.669 1.082 -.062 1.690 1.582 -.004 Log_assets -.151 .009 .036 -.162 -.136 .037 Auditor -.112 .026 .043 -.154 -.110 .021 Ind_fin -.180 .046 -.044 -.250 -.219 -.164 Hypothesis 1 - -.156 - - - -.460 Hypothesis 2 - - -.252 - - -.291 Hypothesis 3 - - - .905 - .239 Hypothesis 4 - - - - .112 .167 Adjusted R-squared -.008 -.004 -.007 -.010 -.015 -.010 F-value .269 .709 .668 .292 .305 .660
Based on table 4.3, for the first model, which tested the relationship between the control variables and the change over time in risk reporting, the following conclusion can be drawn: the control variables are not significantly associated with the change over time in risk reporting in any model. The adjusted R-squared in the first model is -0.008 and the corresponding F-value 0.269.
In the second model, the first hypothesis is tested. The first hypothesis in this study is: “There is a relationship between the size of the board and the change over time in risk reporting”. From the second model can be deduced that there is no significant relationship between board size and the change over time in risk reporting (β = -0.156; p-value > 0.05). Therefore, the first hypothesis is rejected. The direction of this hypothesis is discussable; some researchers argue that a small board is more effective (DeFond et al., 2005; Hoitash et al., 2009; Hunton et al., 2011) and disclose more firm-specific information (Eng & Mak, 2003), while other studies show that a larger board increases the firms’ value (Bennett & Robson, 2004; Dalton & Dalton, 2005; Abraham & Cox, 2007) and is positively associated with the level of risk disclosure (Ntim et al., 2013). In this study both directions are not significant in the linear regression analysis.
Based on table 4.3, a conclusion can be drawn for the second hypothesis in the study. The second hypothesis in this study is: “The occurrence of an IC-failure strengthens the relationship between the size of the board and the change over time in risk reporting”. From the third model can be
20 deduced that the occurrence of an IC-failure does not significantly strengthen the relationship between board size and the change over time in risk reporting (β = -0.252; p-value > 0.05). So, since no significant evidence was found to support it, this hypothesis is rejected. This is contrary to the expectation, in which an increase in risk reporting was expected to keep the trust of the stakeholders and for the board as a means of showing their accountability to the stakeholders.
Based on the fourth model, the third hypothesis can be answered. The third hypothesis in this research is: “There is a positive relationship between gender diversity and the change over time in risk reporting”. No significant relationship was found between gender diversity in the board and the change over time in risk reporting (β = 0.905; p-value > 0.05). Therefore, this hypothesis is rejected. This is contrary to what is claimed in previous literature. Adams & Ferreira (2009) argue that women are better in monitoring the management and release additional information. In addition, Ntim et al. (2013) found a positive relationship between gender diversity and the disclosure of risk information.
In the fifth model, conclusions are drawn with regard to the fourth hypothesis. The fourth hypothesis of this thesis is: “The occurrence of an IC-failure strengthens the positive relationship between gender diversity and the change over time in risk reporting after an IC-failure”. No significant evidence was found that the occurrence of an IC-failure strengthens the relationship between gender diversity and the change over time in risk reporting (β = 0.112; p-value > 0.05). This hypothesis is therefore rejected. In this study, it was expected that, after an IC-failure, the confidence of the stakeholders in the firm would have declined, so the board discloses additional risk information to increase this. As a means, the board uses risk disclosure to show the stakeholder that the occurrence of the IC-failure is not caused by the lack of their supervision.
Finally, in the last model, the influence of all four hypotheses, in combination with the control variables on the change over time in risk reporting, is tested. The direct influence of the moderating variable (the occurrence of an IC-failure) is tested to see if this is significantly related to the change over time in risk reporting. However, it can be concluded that none of the variables are significantly associated with the change over time in risk reporting. The F-value of the fifth model is 0.660 and the corresponding adjusted R-squared is -0.010.
21 4.4 Robustness check
To see if the results are robust another measuring method for the dependent variable (the change over time in risk reporting) is used. The hypotheses are tested again, using the absolute number of risk-related words instead of dividing the number of words published under item 1A and 7A by the total number of words in the 10-K form. The number of absolute risk-related words is compared for two consecutive years. The change over time in risk reporting ranges from -7957 to 9144, and has a mean of 611 words. However, none of the hypotheses in the regression analyses are statistically significant related to the change over time in risk reporting (p-value < 0.05).
22 5. Conclusion
This chapter draws the conclusions of this study. The purpose of this study was to investigate the relationship between board characteristics, the occurrence of an IC-failure, and the change over time in risk reporting. In Section 5.1 the main research question is discussed. Subsequently, in Section 5.2, suggestions for future research are given and the limitations of this study are discussed.
5.1 Conclusion main research question
The main research question investigated in this study was: “Is there a relationship between board characteristics, the occurrence of an IC-failure, and the change over time in risk reporting?”. It has become apparent that the change over time in risk reporting when comparing two consecutive years was, on average, 611 words. Considering that a single-spaced page consists of approximately 500 words, it can be concluded that the average increase in risk-related words in annual reports is a little over one page long. However, the change over time in risk reporting when the number of risk-related words was divided by the total number words in the annual report was only 0.338%. This shows that the length of the average annual report is increased and that the growth in risk reporting is only a small portion of the total increase of words in the annual report.
In this study, the influence of board characteristics on the change over time in risk reporting was tested. No significant relationship was found between board size and the change over time in risk reporting. This might be caused by the fact that the direction of the relationship was unclear. While some studies show that a small board is more effective (DeFond et al., 2005; Hoitash et al., 2009; Hunton et al., 2011), other researchers argue that a large board is preferred (Bennett & Robson, 2004; Dalton & Dalton, 2005; Abraham & Cox, 2007). Another explanation could be that the change over time in risk reporting was too small. In addition, the influence of gender diversity on the change over time in risk reporting was tested. Contrary to expectations of Ntim et al. (2013), this hypothesis was rejected. This could be a consequence of the fact that the change over time in risk reporting was low or because no such relationship exists. Moreover, the influence of the occurrence of an IC-failure is tested on these relationships. Against expectations (Reynolds et al., 2006), both hypotheses were rejected. This could be explained by the fact that many of the firms in the sample face multiple years in a row an IC-failure, therefore keeping their risk reporting strategy the same.
23 In conclusion, none of the hypotheses in this study were significant. However, due to the recent accounting scandals, increase in regulation, and growing need among stakeholders for additional non-financial information, risk reporting is a trending subject. Firms could use this in their benefit by composing an appropriate board, which provides additional risk disclosure, in order to manage the stakeholders. Therefore, risk reporting is a relevant topic for both companies and their stakeholders, regardless of whether an IC-failure occurred.
5.2 Suggestions and limitations
This paper leaves ample opportunities for further research. First, other board characteristics could be studied, such as education, age, and nationality of board members, to see if these variables influence the change over time in risk reporting. Moreover, the method used to measure risk reporting is disputable. In this study, the focus is on certain parts of the annual report, while risk information could also be disclosed in other parts of the annual report. Using qualitative instead of quantitative methods, such as classifying the risks in different categories, for example operational, financial, environmental etc., it is possible to say something about the nature of the risks. Even though firms disclose a significant amount of risk information, it might be that the information is irrelevant to the stakeholders, so a high quantity of risk disclosure says nothing about the quality.
The sample of this study consists of firms that are listed on the U.S. stock exchange, which makes the findings not generalizable to the public sector. Public organizations have different motivations for risk disclosure, since they fulfil another role in the society, it is therefore logical that they report more risk information. In addition, the firms in the sample are from all across the world: the perspectives on the importance of risk reporting can differ between countries. While some countries may just follow the rules of SOx, stakeholders could have more power in other countries and demand additional risk information of firms. In conclusion, this study has failed to identify significant relationships between board characteristics, the occurrence of an IC-failure, and the change over time in risk reporting. Although it is possible that these relationships do not exist, it is conceivable that the sample size of this study was too small to detect them. Future studies may therefore consider expanding the sample size to obtain more conclusive results.
24 References
Abraham, S. & Cox, P. (2007). Analysing the Determinants of Narrative Risk Information in UK FTSE 100 Annual Reports. The British Accounting Review, 39, 227-248.
Abraham, S. & Shrives, P. J. (2014). Improving the Relevance of Risk Factor Disclosure in Corporate Annual Reports. The British Accounting Review, 46, 91-107.
Adams, R.B. & Ferreira, D. (2009). Women in the Boardroom and Their Impact on Governance and Performance. Journal of Financial Economics, 94, 291-309.
Adams, R. B., & Funk, P. (2012). Beyond the glass ceiling: does gender matter?. Management science, 58(2), 219-235.
Adams, R. B., de Haan, J., Terjesen, S. & van Ees H. (2015). Editorial Board Diversity: Moving the Field Forward. Corporate Governance: An International Review, 23(2), 77-82.
Andres, P. d., Azofra, V. & Lopez, F. (2005). Corporate Boards in OECD Countries: size, composition, functioning and effectiveness. Corporate Governance: An International Review, 13(2), 197-210.
Bamber, L.S., Jiang, J. & Wang, I.Y. (2010). What's my style? The influence of top managers on voluntary corporate financial disclosure. The Accounting Review, 85(4), 1131-1162.
Bear, S., Rahman, N. & Post, C. (2010). The impact of board diversity and gender composition on corporate social responsibility and firm reputation. Journal of Business Ethics, 97(2), 207-221. Beck, T., Demirgüç-Kunt, A. & Levine, R. (2000). A new database on the structure and development of the financial sector. The World Bank Economic Review, 14 (3), 597–605.
Becker, C. L., DeFond, M.L. & Jiambalvo, J. and Subramanyam, K.R. (1998). The effect of audit quality on earnings management. Contemporary Accounting Research, 15(1), 1-24.
Belot, F., Ginglinger E., Slovin M.B. & Sushka, M.E. (2014). Freedom of choice between unitary and two-tier boards: An empirical analysis. Journal of Financial Economics, 112, 364–385. Bennett, R. J., & Robson, P. J. A. (2004). The role of boards of directors in small and medium- sized firms. Journal of Small Business and Enterprise Development, 11(1), 95-113.
Beretta, S. & Bozzolan, S. (2004). A Framework for the Analysis of Firm Risk Communication. The International Journal of Accounting, 39: 265-288.
Bhagat, S. & Black, B. (2002). The Non-Correlation Between Board Independence and Long- Term Firm Performance. The Journal of Corporation Law, 27(2), 231-273.
25 Blumberg, B., Cooper, D.R. & Schindler, P.S. (2005) Business Research Methods, Maidenhead, McGraw Hill Education.
Boulouta, I. (2013). Hidden connections: The link between board gender diversity and corporate social performance. Journal of Business Ethics, 113(2), 185-197.
Carter, D. A., Simkins, B. J., & Simpson, W. G. (2003). Corporate governance, board diversity, and firm value. Financial review, 38(1), 33-53.
Chen, C.J.P. & Jaggi, B. (2000). Association between independent non-executive directors, family control and financial disclosures in Hong Kong. Journal of Accounting and Public Policy, 19 (4-5), 285-310.
Cheng, S. (2008). Board size and the variability of corporate performance. Journal of Financial Economics, 87, 157-176.
Coles, J. L., Daniel, N. D., & Naveen, L. (2008). Boards: Does one size fit all? Journal of Financial Economics, 87, 329-356.
Committee of Sponsoring Organizations of the Treadway Commission – COSO. (2004). Enterprise Risk Management: Integrated Framework.
Dalton, C. M. & Dalton, D. R. (2005). Boards of Directors: Utilizing Empirical Evidence in Developing Practical Prescriptions. British Journal of Management, 16, 91-97.
Deegan, C. (2002) The legitimacy effect of social and environmental disclosure – a theoretical foundation’, Accounting, Auditing & Accountability Journal, 15 (3), 282-311.
DeFond, M.L., Hann, R.N. & Hu, X. (2005). Does the market value financial expertise on audit committees of boards of directors? Journal of Accounting Research, 43, 153-193.
Deumes, R. & Knechel, W.R. (2008). Economic Incentives for Voluntary Reporting on Internal Risk Management and Control Systems. Auditing: A Journal of Practice & Theory, 27, 35-66.
Durnev, A., Errunza, V. & Molchanov, A. (2009). Property rights protection, corporate transparency, and growth. Journal of International Business Studies, 40(9), 1533-1562.
Eisenhardt, K. M. (1989). Agency theory: An assessment and review. Academy of Management Review, 14, 57–74.
Elshandidy, T., Fraser, I. & Hussainey, K. (2013). Aggregated, voluntary, and mandatory risk disclosure incentives: Evidence from UK FTSE all-share companies. International Review of Financial Analysis, 30, 320-333.
26 Emanuels, J.A., de Munnik, W.G. & Judge, A. (2010b). International Risk Management. In W. Westerman and J. Meer-Kooistra (eds), International Management Accounting and Control, Chapter 8, 149-169. London: McGraw-Hill.
Eng, L.L. & Mak, Y.T. (2003). Corporate governance and voluntary disclosure. Journal of Accounting and Public Policy, 22, 325-345.
Erickson, J., Park, Y. W., Reising, J. & Shin, H. H. (2005). Board composition and firm value under concentrated ownership: the Canadian evidence. Pacific-Basin Finance Journal, 13(4), 387-410.
Fama, E., & Jensen, M. (1983). Separation of ownership and control. Journal of Law and Economics, 26, 301–325.
Francis, J., Lafond, R., Olsson, P.M. & Schipper, K. (2004). Cost of equity and earnings attributes. Accounting Review, 79(4), 967–1010.
Frias-Aceituno, J.V., Rodriguez-Ariza L. & Garcia-Sanchez I.M. (2013). The Role of the Board in the Dissemination of Integrated Corporate Social Reporting. Corporate Social Responsibility and Environmental Management, 20(4), 219-233.
Hambrick, D. & Mason, P. (1984). Upper echelons: the organization as a reflection of its top managers. Academy of Management Review, 9(2), 193-206.
Harrison, D., & Klein, K. (2007). What’s the difference? Diversity constructs as separation, variety, or disparity in organizations. Academy of Management Review, 32, 1199-1228.
Hermalin, B. E. & Weisbach, M. S. (1991). 'The Effects of Board Composition and Direct Incentives on Firm Performance', Financial Management, 20, 101-112.
Hill, C. W., & Jones, T. M. (1992). Stakeholder-agency theory. Journal of management studies, 29(2), 131-154.
Hillman, A. J., Cannella, A. A. & Harris, I. C. (2002). Women and racial minorities in the boardroom: How do directors differ?. Journal of Management, 28(6), 747-763.
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.
Hoitash, U., Hoitash, R. & Bedard, J.C. (2009). Corporate Governance and Internal Control over Financial Reporting: A Comparison of Regulatory Regimes. The Accounting Review, 84, 839-867.
27 Hunton, J.E., Hoitash, R. & Thibodeau, J.C. (2011). The Relationship between Perceived Tone at the Top and Earnings Quality. Contemporary Accounting Research, 28, 1190–1224.
Institute of Chartered Accountants in England and Wales. (1997). Financial reporting of risk- proposals for a statement of business risk. London.
Iyare, S. & Moore, W, (2011). Financial sector development and growth in small open economies. Applied Economics, 43, 1289-1297.
Jensen, M.C., & Meckling, W.H. (1976). Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305–360.
Koelewijn, J, (2012). Regulering van de financiële sector: gevolgen voor de welvaart en de maatschappelijke verhoudingen. VBA beleggingsprofessionals journaal, 111, 26-29.
Krishnan, G. & Parsons, L. (2008). Getting to the Bottom Line: An Exploration of Gender and Earnings Quality. Journal of Business Ethics, 78, 65-76.
Larcker, D.F., Richardson, S.A. & Tuna I. (2007). Corporate governance, accounting outcomes, and organizational performance. The Accounting Review, 82(4), 963–1008.
Laurenceson, J, (2007). Financial sector regulation, bank franchise values and savings mobilization. Applied Economics, 519, 519-525.
Linsley, P.M. & Shrives, P.J. (2006). Risk reporting: a study of risk disclosure in the annual reports of UK companies. The British Accounting Review, 38, 387-404.
Lipton, M. & Lorsch, J. W. (1992). A Modest Proposal for Improved Corporate Governance. Business Lawyer, 48(1), 59-77.
Maines, L.A., Bartov, E., Fairfield, P.M., Hirst, D.E., Iannaconi, T.E., Mallett, R., Schrand, C.M., Skinner, P.J. & Vincent, L. (2002). Recommendations on disclosure of non financial performance. Accounting Horizons, 16(4), 353-62.
Miihkinen, A. (2012). What Drives Quality of Firm Risk Disclosure? The Impact of a National Disclosure Standard and Reporting Incentives under IFRS. The International Journal of Accounting, 47, 437-468.
Milliken, F.J. & Martins L.L. (1996). Searching for Common Threads: Understanding the Multiple Effects of Diversity in Organizational Groups. The Academy of Management Review, 21(2), 402-433.
28 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.
Power, M. (2004). The risk management of everything. The Journal of Risk Finance, 5(3), 58 - 65.
Prado-Lorenzo, J-M. & Garcia-Sanchez, I.M. (2010). The Role of the Board of Directors in Disseminating Relevant Information on Greenhouse Gases. Journal of Business Ethics, 97(3), 391-424.
Reynolds, S. J., Schultz, F. C. & Hekman, D. R. (2006). Stakeholder theory and managerial decision-making: Constraints and implications of balancing stakeholder interests. Journal of Business Ethics, 64(3), 285-301.
Salmon, W.J. (1993). Crisis Prevention: How to Gear up Your Board. Harvard Business Review,71(1), 68-75.
Solanas, A., Selvam, R. J. M., Navarro, J. & Leiva, D. (2012). Some common indices of group diversity: upper boundaries. Psychological Reports, 111(3), 777-796.
Van Ees, H., Hooghiemstra, R. & Van der Laan, G. (2007). Diversiteit binnen de raden van commissarissen van Nederlandse beursgenoteerde vennootschappen. Research report for Monitoring Committee Corporate Governance.
Waard, D.A., De (2014). De bomen en het bos. University of Curacao.
Wang, K., Sewon, O.M. & Claiborne, C. (2008). Determinants and consequences of voluntary disclosure in an emerging market: evidence from China. Journal of International Accounting, Auditing and Taxation, 17, 14–30.
Xiao, J.Z., He, Y. & Chow, C.W. (2004). The determinants and characteristics of voluntary internet-based disclosures by listed Chinese companies. Journal of Accounting and Public Policy, 23, 191–225.
