Low quality strategic risk disclosure in U.S. and U.K. companies: a challenge for the Board of Directors

(1)

Low quality strategic risk disclosure in U.S. and U.K. companies:

a challenge for the Board of Directors

University of Groningen, Faculty of Economics and Business Master’s Thesis MSc Accountancy

___________________________________________________________________________

Abstract:

This research investigates the impact of Board of Directors’ composition on strategic risk disclosure quality. The strategic risk disclosure quality in annual reports of 71 U.S. and U.K. non-financial companies from the Global Fortune 500 in the 2015-2017 reporting period is measured by a self-constructed disclosure index. As emphasized by prior research and regulators, the findings show that strategic risk disclosure quality is generally low. U.S. firms are especially associated with lower quality. A significant positive relationship is found between board size and the quality of strategic risk disclosure. However, no association is found between the proportion of Strictly Independent Directors or CEO duality and strategic risk disclosure quality. Additional findings show that firm size and profitability respectively decrease and increase strategic risk disclosure quality. These findings have implications for companies and their boards, and suggest a way forward for standard-setters and regulatory bodies.

Keywords: strategic risk disclosure; disclosure quality; Board of Directors

__________________________________________________________________________________

Personal information:

Author: Dennis Bijman

Student number: s2965798

General information:

Supervisor: prof. dr. J.A. Emanuels

Date: 24-06-2019

(2)

1

1. Introduction

“We have evolved in our thinking about strategic risk.” “I think companies that figure out both the value protection and value creation part of risk are going to set themselves up for

success.”

- Sandra G. Carson, VP, Enterprise Risk Management (ERM) and Compliance, Sysco Corporation (Deloitte, 2013, p. 5/6)

The above citation is part of the survey: Exploring Strategic Risk of Deloitte (2013), that investigated how corporations like Sysco and their ERM directors can manage strategic risk more effectively, both now and in the future. The citation is part of a renewed emphasis on strategic risk and strategic risk management of businesses around the globe (Frigo and Anderson, 2009). The evolution of understanding strategic risks was first highlighted by the major corporate collapses in Western countries from the 1990s/2000s, and was intensified by the 2007/2008 global financial crisis (Frigo and Anderson, 2009; Elshandidy et al., 2018). These events created a significant need for better Enterprise Risk Management (ERM) and a demand for greater corporate governance through more transparency, corporate accountability, and disclosure practices (Mallin, 2002; Frigo and Anderson, 2009; Ntim et al., 2013). Among these are strategic risk management and strategic risk disclosure.

Strategic risks are defined as risks that affect, or are created by, an organization’s business strategy and strategic objectives (Deloitte, 2013). The Institute of Directors (2012) specifies that strategic risks involve the most material risks and uncertainties, often as a result of external orientation or foundation, that impact most of senior management decisions. Strategic risks can strike more quickly than ever before (Institute of Directors, 2012; Deloitte, 2013). Failing to recognize and manage strategic risks, such as rapid-fire business trends and technological innovation, can result in declining revenues, decreasing liquidity and even a disruption of business (ICAEW, 2011; Deloitte, 2013) that ultimately affects shareholder wealth (Miihkinen, 2012). Strategic risks can take the shape of ‘black swan’ events, which are rare catastrophic external factors outside of the company’s control (Institute of Directors, 2012). Therefore, overseeing strategic risks management and identifying, assessing, and disclosing strategic risks has become a high priority for the companies’ Board of Directors (Frigo and Anderson, 2009; Institute of Directors, 2012; Deloitte, 2013). As a corporate governance component, the Board of Directors has a significant role and responsibility in

(4)

3

monitoring the implementation of strategic risk management and in disclosing strategic risks with a quality that satisfies the informational needs of shareholders (Frigo and Anderson, 2009; An et al., 2011; ICAEW, 2011). Although the quality of strategic risk disclosure of U.S. and U.K. firms is generally poor (Beretta and Bozzolan, 2004: ICAEW, 2011; Abraham and Shrives, 2014) and legislation concerning this type of disclosure is ill-defined in these countries, its importance is not compromised (ICAEW, 2011; Institute of Directors, 2012). Because of this, examining what influences corporate boards into engaging in strategic risk disclosures between the lines of poor and good quality is most prominently valuable to our understanding of their functionality as governance mechanisms regarding corporate disclosure and the challenges they face in conducting this task. As indicated by the Institute of Directors (2012) and prior research (Abraham and Cox, 2007; Elzahar and Hussainey, 2012), board composition is important to the corporate governance structure in regards to risk reporting. Therefore, different board composition factors will be discussed and examined in relation to strategic risk disclosure quality in this research that follows in this research question:

To what extent does board composition influence the quality of strategic risk disclosures in U.S. and U.K. companies?

As of today, a large portion of accounting research has devoted its attention towards corporate disclosure (O’Sullivan, 2000; Adams, 2002; Watson et al., 2002; Elshandidy et al., 2013). However, the literature about corporate information on risk remains relatively unexplored (Miihkinen, 2012). Existing literature has focussed on specific aspects of risk reporting, in particular the disclosure on market-based risk in relation to financial instruments (Beretta and Bozzolan, 2004; Linsley and Shrives, 2006; Abraham and Cox, 2007). A further aspect of risk reporting involves examining disclosures on risk from a broad perspective and disaggregating the construct into various categories (Beretta and Bozzolan, 2004; Cabedo and Tirado, 2004; Abraham and Cox, 2007; Yang and Datta, 2014).

Whilst part of prior research has focussed on the drivers of risk reporting and the extent of information on risks (Solomon et al., 2000; Ahmed et al., 2004), they suffer from limitations. First of all, there is little if any research is focussed on strategic risk disclosure quality, despite the significant importance of strategic risks to modern-day businesses and their shareholders (Deloitte, 2013), the prioritization of strategic risks and strategic risk management on the board’s agenda (Frigo and Anderson, 2009), and concerns regarding poor

(5)

4

quality strategic risk disclosers in countries such as the U.S. and U.K. (Beretta and Bozzolan, 2004; ICAEW, 2011; Abraham and Shrives, 2014). Secondly, existing studies have mainly examined the influence of general firm characteristics, such as size, profitability, leverage and industry, as drivers for corporate risk reporting (Chavent et al., 2006; Linsley and Shrives, 2006; Raj and Handley-Schachler, 2009; Ntim et al., 2013), whereas studies examining the effect of board composition on corporate risk reporting are quite scarce (Abraham and Cox, 2007; Elzahar and Hussainey, 2012; Ntim et al., 2013) despite suggestions that corporate disclosures, such as strategic risk disclosure, are largely at the discretion of corporate boards (Beretta and Bozzolan, 2004). This notably limits our understanding of the influence of board composition on strategic risk disclosure quality.

Given the above, this research attempts to contribute to and expand on existing literature on corporate disclosure by examining the influence of board composition on strategic risk disclosure quality. The contribution is apparent in two ways. Firstly, this study attempts to construct a unique disclosure index to measure strategic risk disclosure quality in the annual reports of U.S. and U.K. corporations. For even though other studies have constructed various indices concerning corporate risk reporting before (Pérignon and Smith, 2010; Barakat and Hussainey, 2013; Al-Hadi et al., 2016), none have measured quality of strategic risk disclosure. Secondly, by specifically examining the extent to which board composition factors may affect strategic risk disclosure quality, this study departs from prior research that only investigated the influence of general firm characteristics on corporate risk reporting. For this purpose, the board composition factors that this research takes into account are: board size, the proportion of Strictly Independent Directors (SIDs) and CEO duality.

The results will benefit companies and their boards that want to improve their strategic risk disclosure quality, by demonstrating how board composition can influence this type of disclosure. Furthermore, the findings have implications for the FASB, ICAEW and U.S. and U.K. regulatory bodies such as the SEC and FRC by showing that current regulation on strategic risks is accompanied by low strategic risk disclosure quality.

This paper proceeds as follows: in the next section the current state of strategic risk disclosure quality and associated regulation is discussed. Section 3 reviews relevant literature and theories. In section 4, the hypotheses are developed. Then the methodology is described in section 5, including sample selection, timeframe, data collection and measurement of the variables. The results are presented in section 6. Finally, section 7 describes the discussion, conclusion, practical implications, and limitations of this research and presents recommendations for future research.

(6)

5

2. The current state of strategic risk disclosure quality and associated

regulation

Disclosure quality of strategic risks, as part of business risks (Deloitte, 2013), is broadly formulated and defined as: “The risks and uncertainties described in the business review are genuinely “…the principal risks and uncertainties that the board are concerned about. The descriptions are sufficiently specific that the reader can understand why they are important to the company” (FRRP, 2012 Annual Report, p. 3). Concerns regarding the quality of corporate risk reporting, such as strategic risk disclosure, have gotten more attention (ICAEW, 2011). Strategic risk disclosure, especially those of U.S. and U.K. companies, is deemed to be unclear, symbolic, too general and does not seem to alter between reporting years, resulting in a form of ‘box ticking’ and ‘window dressing’ (Beretta and Bozzolan, 2004; Linsley and Shrives, 2006; ICAEW, 2011; Miihkinen, 2012; Abraham and Shrives, 2014). Users, such as shareholders, have to read between the lines in order to identify the risks and to avoid surprises (Abraham and Shrives, 2014). They agree that directors need to provide more governance and accordingly more detailed risk disclosure in an organized fashion (Linsley and Shrives, 2006; ICAEW, 2011; Institute of Directors, 2012). This demand has motivated a substantial body of risk-reporting regulations (e.g., SEC, 1997, 2010; ICAEW, 1997, 2011) to ensure better quality of corporate risk reporting (Linsley and Shrives, 2006; Elshandidy et al., 2018).

However, in terms of reporting standards and legislation, U.S. and U.K. strategic risk disclosure requirements are short, vague and lack illustrative examples (Beretta and Bozzolan, 2004; Linsley and Shrives, 2006; Miihkinen, 2012). For example, the U.S. requires SEC registrants to disclose ‘risk factors’ under Item 1A of the Form 10-K and under the Financial Reporting Release No. 48 (the FRR 48 on market risk), but gives no specifications on disclosing strategic risks. The same is true for U.K. companies which are only required to describe ‘principal risks and uncertainties’ according to The Companies Act 2006 part 414C(2)(b). As it appears, one clause of the law concerning strategic risks may not be effective enough to ensure strategic risk disclosure quality (Miihkinen, 2012). This also becomes apparent in comparison to other countries from continental Europe who provide better legislation regarding (strategic) risks. The Dutch ‘Guideline RJ 400 Directors’ Report’ and the Finnish standard on risk disclosure from 2006 are good examples. They describe how firms should assess significant (strategic) risks and provide a comprehensive view on the

(7)

6

expected quality of risk reporting whilst including illustrative disclosure examples (Miihkinen, 2012).

It seems that U.S. and U.K. regulation on strategic risk disclosure is ill-defined (ICAEW, 2011, Miihkinen, 2012). Still, most U.S. and U.K. firms nevertheless internally identify and assess strategic risks as part of the eight components of the widely adopted Enterprise Risk Management (ERM) - Integrated Framework of the Committee of Sponsoring Organizations of the Treadway Commission (COSO-ERM), and strategic risk management (Frigo and Anderson, 2009; Deloitte, 2013; Whitehouse, 2015). Disclosing strategic risks among operational, financial (reporting), and compliance risks is in line with the four objectives of this framework. However, the COSO-ERM framework gives no mandatory obligation for U.S. and U.K. companies in disclosing strategic risks. All in all, strategic risk disclosure is left at the discretion of U.S. and U.K. Boards of Directors (Beretta and Bozzolan, 2004; Abraham and Cox, 2007). These circumstances provide this study with the perfect opportunity to examine the influences that lead corporate boards from the U.S. and U.K. to disclose strategic risks between the lines of poor and good quality.

3. Board engagement in strategic risk disclosure and its theoretical

background

Boards of Directors have key responsibilities in terms of strategy (Frigo and Anderson, 2009). As described by the Institute of Directors (2018), their role is to set strategic direction and to ensure that the organizational mission remains responsive to changes in the environmental context and organizational realities. Their responsibility is to ‘take the temperature’ and to ensure that strategic risk management practices, risk culture and risk appetite levels are appropriate (Frigo and Anderson, 2009; Pugliese et al., 2009; Institute of Directors, 2012 and 2018). Boards do not ‘manage’ strategic risks, rather, they ‘govern’ strategic risks as a distinct task (Institute of Directors, 2012). As such, the Board of Directors exists as a distinct layer of governance, located between shareholders and the company’s management at the apex of company structure (Institute of Directors, 2012). As part of corporate governance, the Institute of Directors (2018) describes that boards should monitor and evaluate performance, ensure transparency in a form of communication both to and from shareholders and relevant stakeholders, and protect the interests of these shareholders and stakeholders. More generally, boards act as a source of reassurance to all stakeholders that the

(8)

7

company’s strategic activities are subject to objective challenge and risk-analysis (Institute of Directors, 2012). In this regard, boards promoting and ensuring sufficient strategic risk disclosure serve shareholders and stakeholders in understanding strategic risks faced by businesses (Frigo and Anderson, 2009). Investors – from small individual investors to major institutional investors – can also use this information to make informed decisions to approximate the utility maximising decision model of financial theory (Abraham and Cox, 2007). Furthermore, strategic risk disclosure serves companies themselves as well. It helps them to identify and to assess strategic risks in order to implement appropriate strategic risk management, to manage change, to determine future trajectory, and to lower cost of capital (Cabedo and Tirado, 2004; Linsley and Shrives, 2006; Abraham and Cox, 2007; Frigo and Anderson, 2009; ICAEW, 2011). This simultaneously increases profitability, which in turn maximizes shareholder wealth (Miihkinen, 2012). Accordingly, it is incumbent upon boards to meet their responsibility in fulfilling their role in corporate governance and to also create value for their business by disclosing strategic risks in high quality (John and Senbet, 1998; Linsley and Shrives, 2006; Institute of Directors, 2012).

However, a comprehensive and uniform theoretical framework for examining and explaining motivations of boards to engage in risk disclosure is yet to emerge (Ntim et al. 2013; Abraham and Shrives, 2014). At the moment, most prior studies have relied on different economic (i.e., the agency theory and signalling theory) and socio-political (i.e., stakeholder theory) theories (Abraham and Cox, 2007; Barakat and Hussainey, 2013; Elshandidy et al., 2013; Ntim et al., 2013). Adopting multiple theories mitigates the limitations in each individual theoretical perspective and enhances their ability to explain the quality of risks disclosure in relation to board composition (Ntim et al., 2013; Abraham and Shrives, 2014). Following prior suggestions (Deegan, 2002; Ntim et al., 2013), combining a number of theories will provide a richer foundation for understanding the possible effects of board composition on strategic risk disclosures quality, these are: agency theory, stakeholder theory and signalling theory.

Agency theory defines corporations as a nexus of contracting relationships between individuals (Jensen and Meckling, 1976), with shareholders being the principals delegating decision-making authority to the managers whom in turn function as the agents (the principal-agent relationship). As a result of the separation between ownership and control in this agency setting (John and Senbet, 1998), information asymmetry arises as one party has an information advantage over another party (An et al., 2011). If both the principals and agents maximize utility, agents will take advantage of information asymmetry and not always

(9)

8

act in best interest of the principals (Jensen and Meckling, 1976). Therefore, information asymmetry is one of the key factors leading to agency problems (An et al., 2011). Corporate governance deals with mechanisms by which shareholders exercise control over corporate insiders (managers) such that their interests are protected (John and Senbet, 1998; Institute of Directors, 2012). Boards are seen as one of the most important elements of corporate governance in overseeing and monitoring the conduct of a company’s business is being properly managed (Said et al., 2009). As a basic principle of corporate governance, corporate disclosure policy and the preparation of the annual reports emanates from the board (Abraham and Cox, 2007). More specifically, better strategic risk disclosure quality by boards could reduce information asymmetry between the principal and the agent and as a consequence eliminate related agency problems (Rhodes and Soobaroyen, 2010; An et al., 2011; Elshandidy et al., 2013).

In addition, boards also have other motives to engage in strategic risk disclosure. Firstly, from a stakeholder theory perspective, engaging in comprehensive corporate risk reporting (Holm and Laursen, 2007; Elzahar and Hussainey, 2012) can help gain support of large stakeholders, such as investors, employees and regulators (Freeman, 1984; Ntim et al., 2013). An improved relationship between stakeholders and companies could be beneficial to sustainable survival and success of companies in society (An et al., 2011). Similarly to the agency theory, corporate risk reporting can be seen as a tool to manage powerful stakeholders within the managerial stakeholder theory in order to gain approval of business operations by satisfying their informational demand (Freeman, 1984). As a matter of fact, information regarding strategic risks is increasingly demanded by various stakeholders (ICAEW, 2011). Hence it can be expected that disclosure of strategic risks reduces information asymmetry between corporations and their stakeholders, and as a consequence improve the relationship between them (An et al., 2011). Boards should represent all relevant stakeholders and play a vital role in this process (Institute of Directors, 2018). Secondly, signalling theory is likewise underpinned by information asymmetry and explains directors’ incentives to disclose more and better information on risks (Elshandidy et al., 2013). By signalling their ability to identify, measure, and manage strategic risks, directors distinguish themselves from other directors on the job market that may be less effective in those practices. Additionally, one way in which better governed corporation can promote themselves is to engage in higher levels of strategic risk disclosures quality by the Board of Directors (Beekes and Brown, 2006). Apart from signalling the board’s skills, disclosing corporate information can also

(10)

9

improve corporate image, attract potential investors and lower cost of capital (An et al., 2011).

In short, boards are essential governance mechanisms (Institute of Directors, 2012). By reducing information asymmetry they can solve agency problems and improve stakeholder relationships (An et al., 2011). Better strategic risk disclosure is part of this practice. Boards also benefit themselves from this practice by improving their prospects on the job market (Elshandidy et al., 2013).

4. Hypothesis development

The board composition factors that are relevant in answering the question as to what extent board composition influences the quality of strategic risk disclosure in U.S. and U.K. companies are: board size, the proportion of Strictly Independent Directors, and CEO duality. The selection of these factors is both based on their importance in the corporate governance structure regarding corporate risk reporting (John and Senbet, 1998; Institute of Directors, 2012) and on prior studies investigating these factors in relation to other types of disclosure (Abraham and Cox, 2007; Elzahar and Hussainey, 2012; Ntim et al., 2013). The discussed theoretical background provides the foundation of the hypotheses developed in this section and is supported by argumentation and findings from prior research. The conceptual model that results from the hypothesis development can be found in figure 1 at the end of this section.

4.1 Board size

A major challenge faced by boards in disclosing high quality strategic risk disclosure arises from limitations in skills, competences, and from the possible inability of directors to effectively monitor and, as necessary, control the executive arm of a company (Institute of Directors, 2012). This can result in a defective flow of important strategic risks information up to the board level, resulting in lower levels of strategic risk disclosure. This phenomenon described as a ‘glass ceiling’ hinders communication between internal monitoring departments (such as risks management units) and directors (Institute of Directors, 2012). The ‘glass ceiling’ effect causes a form of information asymmetry between directors and managers, ultimately challenging the board in ensuring adequate disclosure practices.

(11)

10

Larger boards are believed to increase managerial monitoring which is positively influences corporate disclosures and performance (John and Senbet, 1998; Beiner et al., 2004; Abeysekera, 2010; Ntim et al., 2013). Larger boards encompass a bigger knowledge base, more experience, larger stakeholder representation (stakeholder theory), and more diversity in terms of expertise (Hou and Moore, 2010). This offers a better possibility for boards to establish appropriate communication with internal risks management units in order to obtain and interpret strategic risk information for disclosure practices (Institute of Directors, 2012). Board size can also add value by diversifying perspectives, providing greater choice among solutions and more decision criteria to achieve objectives of boards and shareholders (Abeysekera, 2010). These factors contribute to the board’s ability to ensure transparency as part of corporate governance (Institute of Directors, 2018). This is where having a strong knowledge base, expertise, and an understanding of the business is important and will help the board tackle challenges created by the ‘glass ceiling’ (Institute of Directors, 2012).

It is on this basis that, following signalling theory, proportionally larger boards are better at distinguishing themselves on the outside job market by disclosing strategic risks of high quality (Elshandidy et al., 2013). As their ability to disclose strategic risks increases, their performance and chances on the job market will enhance as well (Beekes and Brown, 2006). However, the benefit of more directors on boards may be outweighed by the incremental cost associated with larger groups (John and Senbet, 1998). Specifically, when boards increase too much in size, agency problems increase as well. It is likely that boards neglect monitoring and controlling duties as they become more symbolic and fall victim to director free-riding or groupthink (Beiner et al., 2004). Increase in board size makes it also easier for top management to exercise control over the directors and, therefore, directors cannot criticize the policies of top management or truthfully discuss the performance of the company (Al-Janadi et al., 2013).

Following Jensen (1993), it can be reasoned that a large board size may lead to less effective coordination, communication and decision-making, which can have a negative impact on (strategic) risks disclosure quality. Smaller boards are instead believed to effectively form discussions that have positive effects on disclosure practices (Ntim et al., 2013). Furthermore, increasing board size may affect the board’s ability to be an effective monitor of management since larger boards can influence the impact of insiders and block ownership on corporate performance by acting as either a complement or substitute for ownership structure (Singh and Davidson III, 2003). Thus, limiting the size of the board may

(12)

11

improve efficiency within the board necessary to fulfil their roles in corporate governance (John and Senbet, 1998; Said et al., 2009). Ultimately, board size is a trade-off between firm specific benefits and costs of monitoring (Boone et al., 2007).

As a result, studies that investigated the impact of board size on corporate risk reporting, disclosure, and performance show conflicting findings (Cheng and Courtenay, 2004; Hou and Moore, 2010). For instance, Hussainey and Al-Naijar (2011), Elzahar and Hussainey (2012) and Al-Janadi et al. (2013) find positive associations between board size and corporate risk reporting, and other types of disclosure. In contrast, Singh and Davidson III (2003) and Byard et al. (2006) find negative associations with board size. The contradicting and mixed results from prior research result in the following non-directional hypothesis:

Hypothesis 1: There is an association between board size and the quality of strategic risk disclosures.

4.2 Proportion of Strictly Independent Directors

A major theme in the corporate governance literature is the need for non-executive directors on boards to serve as a monitoring and control function on management in the interest of the shareholders (Jensen and Meckling, 1976; John and Senbet, 1998; Carter et al., 2003). Non-executive directors are part of the shareholders solution in solving agency problems, but will only be successful when they are completely independent of management (Carter et al., 2003). As the proportion of independent directors increases, boards will be able to better perform their duty as representatives of the owners and bring more balance to board (John and Senbet, 1998; Abraham and Cox, 2007; Barakat and Hussainey, 2013; Ntim et al., 2013). It is often times even recommended that at least half of the board consists of independent directors (Instituted of Directors, 2012). Policymakers in several countries (including the U.S. and the U.K.) suggest that independent directors are an important element of legal and policy reform in the field of corporate governance. The New York Stock Exchange (NYSE), for example, requires that independent directors constitute a board majority in domestic companies since 2004 (Carter et al., 2003; Institute of Directors, 2012).

However, different jurisdictions and corporate governance norms use various definitions for independent directors (e.g., ‘non-interested’, ‘ outside’, ‘non-executive’ or ‘non-employee’). The different roles for each described director are nonetheless frequently discussed together (Carter et al., 2003). This study tries to be as comprehensive as possible in

(13)

12

its definition by classifying a director as ‘strictly independent’ in case of meeting seven requirements: (1) not employed by the company; (2) not representing or employed by a majority shareholder; (3) not accepting any compensation other than compensation for board service; (4) not a reference shareholder with more than 5% of holdings; (5) not served on the board for more than ten years; (6) no cross-board membership; and (7) no recent, immediate family ties to the corporation (Carter et al., 2003; and defined within the Thomson Reuters Datastream). The general term used in this study to describe an independent director is Strictly Independent Director (SID).

Agency theory and stakeholder theory indicate that SIDs resolve agency problems between managers and shareholders, improve the effectiveness of a board in advising, monitoring, disciplining, and challenging management (Oliveira et al., 2011; Institute of Directors, 2012; Ntim et al., 2013). In comparison to dependent (non-)executive directors, SIDs are associated with increased independence and objectivity, which enhances corporate response to stakeholder concerns and interests (Abraham and Cox, 2007; Prison and Turnbull, 2011; Barakat and Hussainey, 2013; Elshandidy et al., 2013), such as regarding the quality of strategic risk disclosures (ICAEW, 2011). Business relationships of dependent non-executives could interfere with the exercise of independent judgement (Abraham and Cox, 2007). SIDs on the other hand, are an important corporate governance structure since they are more able to respect the obligations of firms and encourage greater transparency and accountability whilst bringing external viewpoints and experience to the boardroom (Abraham and Cox, 2007; Institute of Directors, 2012; Michelon and Parbonetti, 2012). Abraham and Cox (2007) find that SIDs reduce agency costs and strengthen the motivation for corporate risk reporting. In line with signalling theory, SIDs have greater incentives to demand greater transparency and disclosures, since, by doing so, they may improve their personal prestige and social standing (Michelon and Prbonetti, 2012). All in all, SIDs are found to improve corporate governance and enhance corporate transparency in terms of disclosure practices (Abraham and Cox, 2007; Prison and Turnbull, 2011; Ntim et al., 2013).

Consistent with the theoretical predictions, many prior studies report positive effects of SIDs on different corporate outcomes, such as performance (Beiner et al., 2006; Mahadeo et al., 2012), reducing the incidences of corruption and fraud (Hou and Moore, 2010), CSR disclosure (Michelon and Parbonetti, 2012; Ntim and Soobaroyen, 2013), intellectual capital disclosure (Li et al., 2008), voluntary disclosure (Cheng and Courtenay, 2004; Al-Janadi et al., 2013), and general corporate risk disclosure (Abraham and Cox, 2007; Elzahar and Hussainey, 2012; Barakat and Hussainey, 2013; Ntim et al., 2013). The presence of more

(14)

13

SIDs is thus expected to positively influence strategic risk disclosure quality. Therefore, the second hypothesis follows as:

Hypothesis 2: There is a positive association between the proportion of Strictly Independent Directors on boards and the quality of strategic risk disclosures.

4.3 CEO duality and the proportion of Strictly Independent Directors

An important corporate governance issue is the Board of Directors’ leadership structure (Ntim et al., 2013). One structure form contains a dual board whereby the roles of Chief Executive Officer (CEO) and chairman of the board are performed by separate individuals. The CEO is responsible for the day-to-day management (including strategic risk management) of the business, which also involves implementing board decisions. The board chairman is, among other tasks, responsible for managing the board, monitoring and reviewing performance of senior management, and settling possible conflicts within the board (Elzahar and Hussainey, 2012). In contrast, CEO duality is a board structure in which the CEO also holds the position of the chairman of the board. CEO duality can cause CEO domination over directors, including SIDs, because of the creation of a central individual power base, provision of crucial information on which the board relies, and overall concentrated decision making power (Gul and Leung, 2004; Pearce and Zahra, 1991). Grounded in agency theory, CEO duality may constrain board independence and impair board oversight and governance tasks, including corporate disclosure practices (Gul and Leung, 2004), such as strategic risk disclosures. Strategic risk disclosure is primarily at the discretion of boards (Beretta and Bozzolan, 2004), but CEO duality may signal an absence of separation of decision control, making the board an ineffective device in publishing corporate information to stakeholders (Fama en Jensen, 1983). Despite the number of SIDs encouraging greater transparency and disclosure (Michelon and Parbonetti, 2012), these factors may weaken the board as a crucial instrument of corporate governance (Pearce II and Zahra, 1991). By dominating over SIDs and the entire board, charismatic or overbearing CEOs in dual position may mitigate any positive effects that SIDs might have on strategic risk disclosure quality (Institute of Directors, 2012). CEOs are averse to report any corporate information to the outside that shows poor performance or that is likely to diminish his or her power (Belasen, 2008). As a result, CEOs might try to selectively disclose strategic risks to the public (Belasen, 2008), which increases the ‘glass ceiling’ effect towards directors, and ultimately leads to poor quality of such disclosures (ICAEW, 2011). For these reasons, it is

(15)

14

not surprising to see that corporate governance codes, like that of the Financial Reporting Council (FRC), recommend that the posts of chairman and CEO are kept separate.

Consistent with the indications from theory, empirical evidence from prior studies largely suggest that dual boards have a positive effect on several types of corporate disclosure, whilst CEO duality has a negative effects (Gul and Leung, 2004; Barako et al., 2006; Ntim et al., 2013). For instance, Ntim and Soobaroyen (2013) find positive relationships between dual boards and corporate social responsibility (CSR) disclosure. Likewise, Gul and Leung (2004) find that CEO duality is associated with lower levels of voluntary corporate disclosure. Similarly, Elzahar and Hussainey (2012) find that dual boards have positive effects on corporate risk reporting. These result imply that CEO duality has negative effects on corporate disclosure.

Overall, CEOs in dual position have contrary objectives compared to SIDs in regards to providing corporate information, such as strategic risk disclosure, to the outside (Belasen, 2008). Their dominance over the entire board may nullify the corporate governance intentions and practices of SIDs all together (Gul and Leung, 2004). Therefore, it is argued that CEO duality negatively moderates the positive association expected between the proportion of SIDs and the quality of strategic risk disclosure, resulting in the third hypothesis:

Hypothesis 3a: CEO duality will negatively moderate the relationship between the proportion of Strictly Independent Directors on boards and the quality of strategic risk disclosures.

Based on agency theory, an independent chairman provides strong power to the board in effectively monitoring and demanding sufficient disclosure (Al-Janadi et al., 2013). The reverse is expected to be true in case of CEO duality (Fama en Jensen, 1983). Either way, following prior arguments as well, a direct association between CEO duality and strategic risk disclosure quality may exist. The fourth and final hypothesis follows:

Hypothesis 3b: There is a negative association between CEO duality and the quality of strategic risk disclosures.

(16)

15 Figure 1. The conceptual model representing hypothesis 1, 2, 3a and 3b.

5. Data and methodology

5.1 Sample selection

This research was comprised of a sample of 71 non-financial companies from the U.S. and the U.K., listed, as of 2018, on the Fortune Global 500 Index; an index in which the worldwide top 500 corporations are annually ranked by revenue. Sample-items were randomly selected from this index after applying different filters and criteria that will be discussed below.

First of all, companies that were included in the sample were either from the U.S. or the U.K. The reasons for this were that, firstly, Anglo-Saxon countries such as the U.S. and the U.K. commonly use one-tier board structures, whilst Continental European countries, such as the Netherlands and Germany, mostly use two-tier board structures (Hayes et al., 2014). The former board structure allows for CEO duality, whilst the latter does not. Secondly, U.S. and U.K. companies make for an interesting sample. Measured in GDP, the U.S. and U.K. are among the world’s largest economies (Bajpai, 2019). As big players in the global economic market, U.S. and U.K. multinationals face a wide range of potential dangers too (Institute of Directors, 2012). This offers the opportunity to examine boards and their corporate governance functionality in relation to strategic risk disclosure quality of economically powerful nations with lots at stake.

The second filter concerned the sectors in which companies operate, which excluded financial firms from the sample. Financial firms are risk management entities and can be expected to make significantly different types of risk (Linsley and Shrives, 2006).

(17)

16

Furthermore, in contrast to other sectors, financial firms also have distinctive regulation and accounting practices (Elshandidy et a., 2013). For these reasons, they need to be studied independently. Apart from this constraint, the sample included companies from different sectors as defined by the Fortune Global 500 Index (e.g., Food Production, Information Technology Services and Energy).

As a final note, many U.S. and U.K. companies listed on the Fortune Global 500 are also cross-listed on one or more foreign stock exchange(s). Cross-listed firms are often excluded from samples in accounting research because of possible regulatory influences of other countries on the sample companies (Abraham and Cox, 2007; Elshandidy et al., 2013), especially when foreign regulation is stringent (Roosenboom and Van Dijk, 2009). However, as will follow from section 5.4.4 (dependent variable: disclosure index items), the index developed to measure quality of strategic risk disclosure was constructed to be tested on many companies around the world. As to the knowledge on which this research is based, no legislation on earth has mandatory rules on (strategic) risk disclosure in line with the strategic risks items and criteria included in this index. Therefore, foreign regulatory influences are likely to have little influence on the measured quality of strategic risk disclosure, since no legislation has shown to be this precise and inclusive in its definition of strategic risks or its rules on strategic risk disclosure. In short, filtering regulatory interference of other exchanges from the sample is not necessary because of the disclosure index’ overruling nature.

5.2 Timeframe

The timeframe of this research consisted of the reporting years 2015, 2016 and 2017. Because several U.S. and U.K. regulatory institutions (e.g., SEC, 1997, 2010; ICAEW, 1997, 2011) have been trying to ensure better quality of corporate risk reporting over many years (Linsley and Shrives, 2006; Elshandidy et al., 2018), choosing a three-year time span offers these institutions better insights into the present state of strategic risk disclosure quality of companies from their jurisdictions. Additionally, taking three consecutive years into account allowed to control for year specific influences. All in all, the sample of 71 companies and the timeframe of three reporting years resulted in a final sample of 213 firm-years.

5.3 Data collection

Data was collected through the annual reports of the sample companies, since annual reports are the primary source of financial and non-financial information for investors (Elshandidy et al., 2013; Scaltrito, 2015) and also the main vessel for boards to disclose risks (Abraham and

(18)

17

Cox, 2007). For U.S. companies, the disclosure of risk information was found under Item 1A of the Form 10-K, while for U.K. companies this information was reported under the risk section of the Strategic Report. The strategic risk information in the annual reports was collected and coded by using the four-eyes principle. All data was individually collected by two research practitioners, after which differences were discussed in order to increase reliability of the collected data and to ensure consistency in coding (Linsley and Shrives, 2006; Abraham and Cox, 2007; Hassan and Marston, 2010). In doing so, every coder was familiar with the relevant literature and the research objective. The financial data and board composition data used in the empirical analysis were collected from the Thomson Reuters Datastream.

5.4 Variables: measurement and description

The following section will present the description and measurement of the variables relevant for this research. The definition and operationalization of the dependent variable, the moderating variable, the control variables and the independent variables are summarized in table 1.

5.4.1 Independent variables

As a function of hypothesis 1, 2, and 3b, this research includes three independent variables in relation to strategic risk disclosure quality:

1. Board size (BSIZE), which is measured by the natural logarithm of the total number of directors of a company. This approach is in line with methods used in prior research (Carter et al., 2003; Elshandidy et al., 2013; Ntim et al., 2013).

2. The proportion of strictly independent directors (PSID), which is measured as the percentage of strictly independent directors of the total number of directors on the board of a company, in accordance with research done by Elshandidy et al. (2013), and Ntim et al. (2013). Directors are classified as Strictly Independent Directors when they meet the seven requirements formulated in section 4.2 (Proportion of Strictly Independent Directors). Data on the proportion of Strictly Independent Directors is directly available from the Thomson Reuters Datastream which presents data on this item based on the same requirements for SIDs used in this study.

3. CEO duality, of which the measurement is described in section 5.4.2 (moderating variable).

(19)

18

5.4.2 Moderating variable

CEO duality is expected to negatively moderate the relationship between the proportion of Strictly Independent Directors and strategic risk disclosures quality. Following prior research concerning CEO duality (Gul and Leung, 2004; Elshandidy et al., 2013; Ntim et al., 2013), this moderating variable (CEODM) is measured using a dummy variable, valued ‘1’ if the CEO is also the chairman of the board, and valued ‘0’ if the two positions are separated by different individuals.

5.4.3 Dependent variable: disclosure index development

The dependent variable of this research is strategic risk disclosure quality. As part of a collaboration of Accountancy and Controlling master students of the University of Groningen (see Acknowledgements), a unique disclosure index was developed to measure this variable. The disclosure index is based on other studies that constructed various disclosure indices for corporate risk disclosures as well (Pérignon and Smith, 2010; Barakat and Hussainey, 2013; Al-Hadi et al., 2016). The disclosure index of this particular study is presented in appendix A.

For the development of the disclosure index, it was useful to further specify the definition of strategic risk disclosure quality, given at the start of section 2. This was done by two concepts of disclosure quality defined by Abadi and Janani (2013), who suggest that corporate information should be (1) adequate, meaning that the disclosed information must be minimal and non-misleading, and (2) fully disclosed/complete. This means that all relevant information should be provided to users (Abadi and Janani, 2013), and that excessive unimportant information should be omitted from disclosure to not interfere with fair interpretation of statements. In general, quality can be further defined as the degree to which a set of inherent characteristics fulfils requirements (Wicks and Roethlein, 2009).

In the past, mainstream accounting literature on corporate disclosure has emphasized that quantity can be used as a sound proxy for quality (Beretta and Bozzolan, 2004) by using number of words and sentences for instance (Linsley and Shrives, 2006; Abraham and Cox, 2007; Miihkinen, 2012). As far as quantity of disclosure is concerned, two dimensions have to be balanced (Beretta and Bozzolan, 2004): (1) the absolute number of pieces of information disclosed as a proxy of the amount disclosed (quantity), and, (2) the relevance assumed by strategic risks-related information weighted in the overall communication (density). The second dimension is in line with the definition given to strategic risk disclosure quality in this research. Density of disclosure is related to relevance (full disclosure) and minimality (adequate disclosure). Also, in regards of the density dimension, it is evident that

(20)

19

the writing style and quantity of information strongly influences the effectiveness of disclosure (Beretta and Bozzolan, 2004). Reporting strategies like ‘hiding the needle in a haystack’ could dilute a limited number of relevant strategic-risks-related pieces of information in the masses created by hundreds of pages of strategic risk disclosure. Quantity makes it hard to find the quality within. Following Beretta and Bozzolan (2004), and with respect to the definition given to strategic risk disclosure quality, this research contends that quantity is not a satisfactory proxy for the quality of disclosure.

Although content analysis is an extensively employed measurement of disclosure quality in annual reports (Kavitha and Nandogopal. 2011; Scaltrito, 2015), this study proposes a self-constructed disclosure index as a measure of strategic risk disclosure quality. The reasons for this are that content analysis is labor-intensive and time consuming, and it relies heavily on subject knowledge and the linguistic ability of coders which is inevitably subjective (Abraham and Cox, 2007; Kavitha and Nandagopal, 2011; Scaltrito, 2015). Moreover, content analysis can erroneously assume that quantity is a valid proxy for quality (Linsley and Shrives, 2006). By using this disclosure index, the possible bias in the quality score that might result from extensive amounts of information on strategic risks, whilst totally failing to disclose on other crucial aspects, is alleviated (Barakat and Hussainey, 2013). Although disclosure indices can have reliability and validity issues as well (Kavitha and Nandagopal, 2011), considering the specific disadvantages of content analysis, a disclosure index was considered the best measurement of strategic risk disclosure quality in this research.

With this in mind, self-constructed disclosure indices in the accounting literature can be either unweighted or weighted (Beretta and Bozzolan, 2008; Scaltrito, 2015). Following Barakat and Hussainey (2013) this study employed an unweighted disclosure index and used a binary-coded scoring procedure that attributes ‘0’ for non-disclosure and ‘1’ for disclosure of specific pre-selected items, known as a dichotomous approach (Scaltrito, 2015). In total, eight main-items were identified, each containing four sub-items that provided coverage to typical strategic risks and their characteristics (Linsley and Shrives, 2006; Institute of Directors, 2012; Miihkinen, 2012). Therefore, the strategic risk disclosure quality of a company could reach a maximum score of 32 points (all items disclosed) and a minimum score of zero points (no items disclosed).

Similar to Taylor et al. (2010), a strategic risk disclosure quality score (DSCLQscore) was computed for all sample-items and divided by the maximum number of items that should or could be disclosed. This is represented in equation 1.

(21)

20

=

Equation 1: Mathematical representation of the strategic risk disclosure quality score

Even though unweighted indices in essence only measure quantity and giving equal weight to all criteria is arbitrary (Pérignon and Smith, 2010), assessing quality by attaching weight still remains dependent on the quantity of disclosure (Beretta and Bozzolan, 2008), and coming up with different weights to all criteria is even more arbitrary (Pérignon and Smith, 2010; Kavitha and Nandagopal, 2011). Moreover, quality of disclosure is still mono-dimensional and strictly based on the extent of the information disclosed (Beretta and Bozzolan, 2008). The proposed disclosure index is therefore still regarded as a better measurement of disclosure quality than direct measures of quantity (Beretta and Bozzolan, 2004).

Lastly, the disclosure index was tested for reliability. Reliability was addressed by calculating the Cronbach’s alpha, which tests the internal consistency of the 32-item scale (Elshandidy et al., 2013; Hooghiemstra et al., 2015). The Cronbach’s Alpha for the computed strategic risk disclosures scores is 0.664. Since a score of > 0.7 is generally deemed acceptable (Gliem and Gliem, 2003), the internal consistency of this research’s measure is questionable, but not poor or unacceptable for research (Gliem and Gliem, 2003).

5.4.4 Dependent variable: disclosure index items

The main-items included in the strategic risks quality index were mostly selected from strategic risk topicspresented by Linsley and Shrives (2006), the FRC (2012), the Institute of Directors (2012), and Miihkinen (2012). The sub-items were selected on the grounds of discussions of the Institute of Directors (2012) and other literature. Each disclosure item is considered to be in line with the definition of strategic risks given in the introduction. Each item is also found to cover strategic risks associated with businesses from different sectors (excluding the financial sector). Because every company is unique and faces its own specific strategic risks (Institute of Directors, 2012; Chen, 2018), it could be reasonably expected that each sample-item defines, describes and reports more types of strategic risks than others that are included in this disclosure index. It has to be noted that this disclosure index does not measure completeness of the overall strategic risk statements of companies, but only

(22)

21

measures the quality of the specific strategic risk items that are of interest for this study and thus included in the disclosure index (Scaltrito, 2015). It is impossible to measure all the items disclosed by companies (Kavitha and Nandagopal, 2011). Additionally, even though many strategic risks interface with each other, all items are unique in nature and were therefore separately evaluated in the annual reports.

Appendix A presents examples and a description of the eight main-items included in the disclosure index. The main-items are all a different type of strategic risk that companies are expected to report. The four sub-items included in the strategic risk disclosure quality index are further discussed in this section because of their complex nature and their contribution to the practical implications of this study.

Firstly, sub-item 1 (description of the strategic risk) takes into account the description of the concerned strategic risks. A company should provide an explanation as to why the strategic risk is added or removed from the risk paragraph and why the strategic risk is relevant to the company (Abraham and Shrives, 2014). If one strategic risk from the eight main-items is not sufficiently described in the annual report, this most likely excludes the subject from scoring any points on other sub-items as well. Furthermore, a point for disclosure is only given for a fairly stated, minimal and relevant description of the concerned strategic risk and is not measured in length or quantity. As discussed, the disclosure index gives no weight to the description, but only measures non-disclosure or disclosure.

Secondly, sub-item 2 (likelihood of the strategic risk) and sub-item 3 (impact of the strategic risk) are part of the level of risk appetite of organizations. Risk appetite is defined as the amount of risk that an organization is prepared to accept or tolerate (Institute of Directors, 2012). This can be illustrated in a risk matrix with on the y-axis ‘likelihood’ and on x-axis ‘impact’ and a line drawn somewhere on the matrix as a boundary between acceptable risks and those deemed unacceptable. It is the responsibility of boards to set an appropriate strategic risk appetite level as part of strategic risk management (Frigo and Anderson, 2009; Pugliese et al., 2009). The amount of risk tolerance thus requires boards to consider quantitative terms to determine materiality of strategic risks in order to help stakeholders evaluate them (Institute of Directors, 2012). Although these items are in conjunction with each other, companies might report them independently in a risk matrix or in a stand-alone manner. Therefore likelihood and impact (Beretta and Bozzolan, 2004) of strategic risks are separate items in the disclosure index and points for disclosure are given separately.

The fourth and final sub-item is year-on-year risk development. Companies need to revisit risk disclosure regularly and should decide which disclosures are still relevant and

(23)

22

which need to be replaced with new information (Abraham and Shrives, 2014). Strategic risks can be expected to have an impact on future performances of firms or to have occurred in former reporting years with subsequent reactions by management and boards to face them (Deloitte, 2013; Institute of Directors, 2012; Miihkinen, 2012). Annual repetition of increases and decreases of strategic risks provides users with a feeling that effort is put into the report (Abraham and Shrives, 2014). A disclosure point is given when firms disclose year-on-year development.

Lastly, because of the global lack of mandatory rules on (strategic) risk disclosure in line with the strategic risk items (main-items or sub-items) included in this disclosure index, the annual reports of companies from many different countries can be tested against the criteria in this disclosure index. Even though this research focuses on the ill-defined regulation regarding strategic risks from the U.S. and the U.K., the disclosure index can also be used in countries where strategic risk disclosure legislation is better defined in guidelines and standards, or in countries where strategic risk disclosure is classified in terms of internal risk management, as is the case with the German GAS5. After all, the disclosure index, as part of the collaboration, is constructed to be used in other research as well, and it allows to examine the determinants of strategic risk disclosure quality free of regulatory interference.

5.4.5 Control variables

In this section the control variables and their measurements are discussed. The control variables were included for the purpose of controlling for potential omitted variables bias. This research has controlled for the following eight firm characteristics that are frequently controlled for in prior research regarding general disclosure and corporate risk disclosure practices: firm size, profitability, leverage, firm growth, dividend yield, reporting year, country and industry sector.

5.4.5.1 Firm size.

A positive relation was expected between firm size and strategic risk disclosure. Costs of providing risk information in larger firms is likely to be lower relative to small firms (Elshandidy et al., 2013), making it more likely that boards disclose better strategic risk information when their companies are larger. Similarly, smaller firms are more vulnerable to their competitors when disclosing sensitive strategic risk information (Elshandidy et al., 2013), whilst larger firms can more easily afford to lose some advantages concerning inside corporate information. Empirical evidence from prior studies suggests that firm size

(24)

23

positively influences risk reporting (Lindsey and Shrives, 2006; Abraham and Cox, 2007; Elshandidy et al., 2013; Mokhtar and Mellett, 2013). Firm size (FSIZE) was measured using a natural logarithm of the book value of firm assets (Carter et al., 2003; Moeller et al., 2004; Almeida and Campello, 2007). Further, to reduce the influence of outliers in the statistical data, total assets of firms were winsorized. Outliers are increased to three times the standard deviation below the mean or are decreased to three mines the standard deviation above the mean.

5.4.5.2 Profitability

Signalling theory suggests that boards are likely to reveal good news to the market to avoid any undervaluation of company shares (Giner, 1997). High profitability is associated with adequate risk management (Elshandidy et al., 2013). By disclosing profitability and strategic risks, boards can signal their ability to identify strategic risks as a first step in strategic risk management (Frigo and Anderson, 2009) and the associated performance quality. The expected positive association between profitability and risk disclosure is supported by empirical research (Chavent et al., 2006; Miihkinen, 2012). Profitability (PROF) was measured by Return on Assets (ROA), defined as the ratio of income before taxes to the average of total assets (Aupperle et al., 1985; Greene and Segal, 2004).

5.4.5.3 Financial leverage

Agency theory is concerned with the agency problem of information asymmetry between management and shareholders (An et al., 2011). Financial leverage is also an agency problem, but in this case between shareholders and debtholders (Deumes and Knechel, 2008). This problem arises because shareholders may have incentives to take excessive risks, whilst debtholders are reluctant in risk-taking. This can result in wealth transfers from debtholders to shareholders (Jensen and Meckling, 1976). Disclosure of strategic risks and strategic risk management by leveraged firms might inform debtholders about this threat (Deumes and Knechel, 2008). Prior research has found positive associations between risk disclosure and the degree of financial leverage (Deumes and Knechel, 2008; Taylor et al., 2010). Financial leverage (LEV) was calculated as the percentage of the total debt to total assets (Aivazian, 2005; Huynh and Petrunia, 2010; Ntim et al., 2013).

(25)

24 5.4.5.4 Firm growth

Growing firms that depend on external financing have incentives to reduce information asymmetry between them and investors and creditors (Khurana et al., 2006). By disclosing strategic risks, high-growth firms can demonstrate their ability to identify, assess, and manage these risks as demanded by these investors and creditors (Elshandidy et al., 2013; Frigo and Anderson, 2009). Empirical evidence shows positive associations between firm growth and corporate disclosure in support of this argument (Chavent et al., 2006; Khurana et al., 2006; O’Sullivan et al., 2008). Firm growth (FGROWTH) was measured as the difference of the logarithm of firm sales between two years ( ) as done by Huynh and Pretrunia (2010). Sales growth is a firm growth proxy that translates easily across countries and industries (Delmar et al., 2003; McKelvie and Wiklund, 2010).

5.4.5.5 Dividend yield

Firms disclosing strategic risks may be motivated to pay higher levels of dividend to compensate their shareholders for these risks (Elshandidy et al., 2013). Higher levels of disclosure are generally positively associated with higher levels of dividends paid as well, as evidenced by Khang and King (2006), and Elshandidy et al. (2013). Dividend yield (DIY) was measured by a natural logarithm of the ratio of the most recent full-year dividends divided by the current share price (Elshandidy et al., 2013).

5.4.5.6 Reporting year

To control for year specific events (Lopes and De Alencar, 2010) and possible increasing quality of corporate risk disclosure between years (Miihkinen, 2012), a control variable for reporting years was included. Reporting year (YEAR) was measured as a set dummy variables for the years 2015, 2016 and 2017. The year 2017 was used as the comparator factor.

5.4.5.7 Country

This research controls for the influence of the country in which a firm is headquartered. Some countries can have a more positive influence on the quality of strategic risk disclosures than others. As explained by Roosenboom and Van Dijk (2009), specific country jurisdictions can improve information disclosure by companies, such as strategic risk disclosure, and better protect investors. They argue that jurisdictions differ on country-level, which can result in poor investor protection policies in one country and better investor protection policies in

(26)

25

= + + + + ( * ) + + + + + + + + +

another. Country (COUNTRY) was measured as a dummy variable, with a value of ‘1’ for companies headquartered in the U.S., and a value of ‘0’ for companies headquartered in the U.K.

5.4.5.8 Industry sector

Industry sector is found to have a positive relation with disclosure (Barako et al., 2006; Al-Janadi et al., 2013). A specific industry might face particular circumstances, relationships, or regulation that may influence their (risk) disclosure practices (Barako et al., 2006; Abraham and Cox, 2007). This study has categorized each sample item in one of the following fifteen industry sectors defined by the Fortune Global 500: (1) Retailing; (2) Energy; (3) Wholesalers; (4) Health Care; (5) Technology; (6) Telecommunications; (7) Motor Vehicles & Parts; (8) Food & Drug Stores; (9) Industrials; (10) Aerospace & Defense; (11) Transportation; (12) Household Products; (13) Food, Beverages & Tobacco; (14) Media; (15) Business Services. Industry sector (IND) was measured as a set of dummy variables with the Energy sector as the comparator factor.

5.5 The empirical model

This research used multiple regression analysis to regress strategic risk disclosure quality scores on the explanatory variables and other variables. The empirical model follows in equation 2, where DSCLQ is the strategic risk disclosure quality score for firm i, in reporting year t. is the intercept of the regression line and is the standard error of residuals. YEAR refers to the set of dummy variables for the three reporting years: 2015, 2016 and 2017. IND refers to the set of dummy variables for the fifteen industry sectors defined in section 5.4.5.8 (industry sector) and table 1.

(27)

26 Table 1

Summary of definitions and operationalization of variables.

Variable Description Measurement/proxy

Dependent variable

Strategic risk disclosure quality score (DSCLQscore)

The percentage of strategic risk disclosure quality points scored out of the maximum number of possible points.

An unweighted disclosure index consisting of eight main-items, each divided into four sub-items. A possible 32 out of 32 points can be scored (so a percentage between 0 and 1).

Independent variables

BSIZE The number of board

members.

Natural logarithm of the total number of directors on the board of a firm.

PSID The number of Strictly

Independent Directors.

Percentage of Strictly Independent Directors to the total number of directors on the board of a company.

CEOD The dual position/separation

between the CEO and the chairmanship of a board.

Dummy variable: 1, if the CEO is also the chairman of the board, 0 if otherwise.

Moderating variable

CEODM The dual position/separation

between the CEO and the chairmanship of a board.

Dummy variable: 1, if the CEO is also the chairman of the board, 0 if otherwise.

Control variables

FSIZE Firm size. Natural logarithm of the book value of firm

assets.

PROF Profitability of the firm. Return on Assets (ROA), defined as the ratio

of income before taxes to the average of total assets.

LEV Financial leverage of the

firm.

Percentage of the total debt to total assets.

FGROWTH Firm growth. Difference of the logarithms of firm sales

between two years: .

DIY Dividend yield of the firm. Natural logarithm of the ratio of the most

recent full-year dividends divided by the current share price.

YEAR Reporting years 2015, 2016

and 2017.

Set of dummy variables for the years 2015, 2016 and 2017, with 2017 as the comparator factor.

COUNTRY Country in which the firm is

headquartered.

Dummy variable: 1, if a company is

headquartered in the U.S., 0, if a company is headquartered in the U.K.

IND Industry sector of the

company categorized by the Fortune Global 500.

Set of dummy variables for the sectors: (1) Retailing; (2) Energy; (3) Wholesalers; (4) Health Care; (5) Technology; (6)

Telecommunications; (7) Motor Vehicles & Parts; (8) Food & Drug Stores; (9) Industrials; (10) Aerospace & Defense; (11)

Transportation; (12) Household Products; (13) Food, Beverages & Tobacco; (14) Media; (15) Business Services, with the sector Energy as the comparator factor.

Low quality strategic risk disclosure in U.S. and U.K. companies: a challenge for the Board of Directors