1
IFRS 15 disclosure quality of construction companies:
The impact of CFO and audit committee characteristics
on disclosure quality
Master thesis Accountancy
University of Groningen, Faculty of Economics and Business
Martijn de Vries
S2479737
Supervisors:
Prof. dr. R.L. ter Hoeven
MSc. R. van Duren
Word count:11893
2
Abstract
The main goal of this research is to examine if CFO career horizon, age and gender are determinants of IFRS 15 disclosure quality for construction companies, and if audit
committee characteristics strengthen the negative relationship between CFO career horizon and IFRS 15 disclosure quality. IFRS 15 is the revenue recognition standard for all
European listed companies. Beside the main goal of this study, I compared the 2018 and 2019 revenue recognition disclosures of the companies in the sample to identify best practices and pear learning effects. To measure the reporting quality of revenue we created an IFRS 15 disclosure index. The index strongly focusses on the voluntary character of IFRS 15 disclosure. The outcome of this assessment resulted in best practices, but no significant pear learning effects were visible among the companies in the sample. The outcomes showed that shorter CFO career horizons result in higher quality IFRS 15 disclosure quality and that audit committee characteristics strengthen this relationship.
3
I. Introduction
On the 6th of February the London Business School published an article in which it extensively discusses the reasons for the liquidation of Carillion, a British multinational active in the construction and service industries (Higson, 2018). The bankruptcy of the company resulted in a shock for the United Kingdom. One of the many ways that Carillion used to cover up poor performance was by using their long term contracts in a very flexible way (Higson, 2018). This became visible with the introduction of IFRS 15, which resulted in a one-time write down of over £100 million pounds of earnings (Higson, 2018). This indicates the effect of IFRS 15 revenue recognition on the construction industry. The construction industry is with approximately 9% of the gross domestic product of the European Union, one of Europe’s most important industries. Besides the financial
importance, the construction sector is vital for society through the building of new homes, offices, infrastructure and by facilitating approximately 18 million direct jobs and many more indirect jobs (EU, 2020). The construction industry is also Europe’s largest polluter of greenhouse emission (EU, 2020). The continued pressure on companies to perform in combination with the industries sensitivity for changes in the economic environment, the current housing shortage, and the growing pressure and measures by the European Union and its member states to reduce greenhouse emissions, shows that the construction industry is up to an interesting and challenging time (Appun, Eriksen, and Wettengel, 2020; COE, 2020). Especially in complex situations like this, investors benefit from extensive and qualitative revenue disclosure to estimates companies true performance and risks. Beside the investors, regulatory bodies continually try to improve and introduce laws and
standards to protect investors, stakeholders and make the financial markets more efficient. This study about revenue recognition might help regulators to further understand the application of IFRS 15 by companies and might give new insights that can lead to improvements of the IFRS 15 standard that is effective since 1 January 2018.
IFRS 15 is an accounting standard introduced by the IASB with the aim to standardize European revenue reporting and improve comparability between companies, which makes revenue reporting more transparent and useful (Aarab, Bissessur, and Ter Hoeven, 2015; Devalle, Onali, and Magarini, 2010). This is done by establishing reporting principles about the timing, nature, amount and uncertainty of revenue and cash flows from contracts with customers (IASPLUS, 2020). Construction companies, like Carillion, BAM, Eiffage and Vinci, cope with a lot of complex long term contracts with multiple variable
4 (Albert, 2017; IASPLUS, 2020). These contract conditions makes the construction industry unique. Therefore, the timing and recognition of revenue requires judgement, because the exact amount of consideration that a construction company receives is contingent on future events (Albert, 2017). Transparent voluntary project- or company specific information probably increases the understanding of investors and stakeholders regarding the recognition of revenue of these complex long term contracts.
There is no unifying theory or definition of corporate governance (Brown, Beekes, Verhoeven, 2011), but the Cadbury committee (1992) defined corporate governance in their report “as the system that directs and controls the company” which is a
comprehensive, but representative description of corporate governance. From an accounting perspective corporate governance plays an important role in companies
compliance with the general accepted accounting standards and in taking responsibility for the maintenance of the credibility of the corporate financial statements (Lin, and Hwang, 2010). From this perspective, corporate governance influences disclosure. Examples of corporate governance bodies are the executive board, the board of directors and its subcommittees, like the audit committee.
The CFO is the financial expert in the executive board of a company due to his superior knowledge and capabilities on financial topics (Mian 2001; Geiger and North 2006). The CFO is responsible for the financial reporting process and in this capacity he performs a crucial stewardship role in relation to financial reporting quality, which includes IFRS 15 disclosure (Aier, Comprix, Gunlock, and Lee, 2005; Ge, Matsumoto, and Zhang 2011; Geiger and North 2006; Graham, Harvey, and Rajgopal 2005). Another corporate
governance body that impacts the quality of disclosure is the audit committee (Abbott and Parker, 2000). The main role of the audit committee is to represent and advice the board of directors on finance and financial reporting topics (Al-Baidhani, and Mohsen, 2016). To fulfill this task they, among other things, monitor the audit process. Menon and Williams (1994) and Abbott and Parker (2000) argue that independent, well-functioning and active audit committee reduces agency costs and information asymmetry, which in turn is a sign of increased audit quality.
This study investigates the impact of audit committee characteristics on the relationship between the CFO’s career horizon, which is a combined measure of CFO age and CFO tenure, and IFRS 15 disclosure quality. A long CFO career horizon indicates that a CFO is relative young and has a shorter tenure than average in the sample. A shorter CFO career horizon indicates that a CFO is older and/or has on average more tenure years than the
5 average CFO in the sample. Previous studies argue that shorter tenured CFOs protect their reputation by protecting shareholders interest, through higher quality disclosure (Muttakin, Khan, and Tanewski, 2019). From an Upper Echelon perspective, age is associated with more conservativeness and experience which is useful for recognizing revenue (Muttakin, Khan, and Tanewski, 2019). The moderating effect is explained from an agency
perspective. It is likely that the presence of the audit committee as representative of the board of directors reduces the freedom of the CFO, therefore I expect that they are more willing to disclose fairly and qualitative IFRS 15 revenue recognition disclosure which leads to the following research question:
What is the impact of CFO career horizon on the revenue recognition disclosure quality based on the IFRS 15 standards for European listed construction companies and do audit committee characteristics influence this impact?
In the remainder of this introduction I will address the scientific contribution of the subject. After the scientific contribution, the literature review, hypothesis development, research methodology, results, discussion and conclusions will be addressed.
Scientific contribution
This study adds to the corporate governance and voluntary disclosure literature by
addressing the impact of audit committee characteristics on the relationship between CFO career horizon and IFRS 15 disclosure quality. IFRS 15 only recently became effective (KPMG, 2014) which means that there companies had no precedents on how to disclose IFRS 15 information, except for the IFRS 15 requirements provided by the IASB (KPMG, 2014). The uncertainty regarding the amount, quality and style of disclosure gives this study a voluntary nature. My own developed disclosure index score is unique because it considers both mandatory and voluntary elements. Voluntary company or project specific information helps shareholders and stakeholders understanding of company’s policies regarding revenue recognition, therefore voluntary information is equally important as mandatory information when allocating scores. Another unique feature of the study and the disclosure index is that it’s the first time that IFRS 15 disclosure information is assessed for two consecutive years, which makes it possible to identify pear learning effects and best practices. Both pear learning effects and best practices offer standard setters and local regulatory bodies insights in the effectiveness and differences in the application of IFRS 15 by companies in the construction industry. They can use this information to increase
overall revenue disclosure with the aim to increase the creditability and stability of financial markets and thus protect shareholders and investors.
6
II. Literature review
The main aim of this chapter is to provide context and background information for the explanatory theories underlying this study. Since IFRS 15 is a new standard no IFRS 15 disclosure examples are available which increases uncertainty regarding the amount, quality and style of IFRS 15 disclosure. The uncertainty makes a mandatory disclosure like IFRS 15 have a voluntary character. Therefore, I address the voluntary disclosure
background literature in which is explained what voluntary disclosure is, why it is provided by companies and which effects voluntary disclosure has on organizations and
stakeholders. Based on the research question discussed in chapter 1 multiple theory’s that underlie the expected relationship between CFO career horizon, audit committee
characteristics and IFRS 15 disclosure quality will be discussed.
Voluntary disclosure
This study will not only focus on the mandatory disclosure, but it will also pay attention to useful voluntary revenue disclosure in the construction industry. IFRS 15 was issued in 2014 and became effective as per 1 January 2018 (IASPLUS, 2020). All European listed companies need to apply to the IFRS standards for European entities which is enforced under regulation (EC) 1606/2002 by the European parliament and the council of the European Union (European Commission, 2002), for more information on how to apply IFRS 15 I refer to appendix D. Voluntary disclosure is especially relevant for this study, because for the year 2018 companies had no precedent on how to disclose IFRS 15 information in their annual report, except for the IFRS 15 requirements provided by the IASB (KPMG, 2014). Therefore, companies have some discretion on how to present the IFRS 15 requirements in their disclosure. IFRS 15 is mandatory for all companies who apply IFRS standards, but the uncertainty surrounding the presentation of IFRS 15 requirements and the discretion companies have in explaining their IFRS 15 policies, makes the disclosure have something of a voluntary nature. Therefore I decided to treat both mandatory and voluntary disclosures as equally important, which is reflected in the distribution of points in the disclosure index.
Voluntary disclosure can be described as extra information provided by the manager on top of mandatory disclosure with the aim and intention to inform stakeholders with
additional relevant information (Meek, Roberts, and Gray, 1995). Company specific, project specific and regional specific information are all examples of voluntary
information. With the help of my own developed IFRS 15 disclosure index score I assess if companies comply with the disclosure requirements mentioned in IFRS 15.113 until
7 15.129 and if they provide extra information in the forms discussed above. To illustrate this with an example: Index item 15.123A can only earn one (mandatory) point when generic information is provided about the use of significant judgement for determining the timing of revenue. But if the company explains an IFRS disclosure index item per segment, region, product group or by using examples, to make it easier to understand or to give more detailed information, the company can be rewarded with an additional (voluntary) point.
Silence is often seen as bad news by investors, therefore one of the main reasons of companies in the past to disclose additional information was to raise money from investors (Milgrom, 1981). Companies that failed to provide sufficient information found it hard to attract investors. In recent years, managers are put under pressure by the society to disclose information about firm performance that impacts stakeholders due to the growing
popularity of the stakeholder theory (Boesso, Kumar, 2007). This development has increased the reporting quality of companies over the years, but there are also other incentives for companies to provide high quality voluntary disclosure.
Disclosure for example, is an important source of information for investors when monitoring managers’ compliance to contractual agreements (Healy and Palepu, 2001). Higher disclosure levels will positively influence the ability of shareholders to monitor the managers, and thus disciplining them (Hooghiemstra, Hermes, Emanuels, 2015).
Therefore, disclosure is a tool for reducing agency problems (Edelen, Evans, and Kadlec, 2012; Dumontier and Raffournier, 1999). Healy and Palepu (2001) further argue that the amount of disclosure is based on a tradeoff between the benefits and costs of additional disclosure.
Managers profit in multiple ways from providing additional information. First of all, voluntary disclosure can improve the knowledge of investors regarding future company payoffs. This reduces uncertainty and thus decreases the cost of capital (Barry and Brown, 1985; Coles and Lowenstein, 1988; Coles, Lowenstein, and Suay, 1995). Providing voluntary disclosure positively influences manager’s reputation as credible reporter which improves manager’s reputation in the labor market (Kothari, Shu and Wysocki, 2009). A CFO who is keen on his career development is incentivized to report not only positive, but also negative information to improve his personal reputation (He, Plumlee, and Wen, 2019; Campbell, Chen, Dhaliwal, Lu, and Steele, 2014; Skinner, 1994), this is in line with the agency theory in which the manager maximizes its own utility (An, Davey, and Eggleton, 2011).
8 Agency theory
Agency Theory is one of the main theories in the economics and corporate governance literature. Both the CFO as a member of the executive board and the audit committee as representative of the board of directors have important roles in corporate governance. The main focus in this paragraph is to make use of the agency theory to explain the expected impact of audit committees on the expected relationship between the CFO and IFRS 15. Agency Theory can be explained as a separation of ownership and control which leads to problems due to different risk appetites and interests (Fama and Jensen, 1983). Fama and Jensen (1983) explain agency as a contractual relationship in which the shareholders of a company (principal) hire directors or managers (agent) to take care of their interests. The agent is considered a rational person and therefore motivated by self-interest. The agency problems can be divided in two groups: (1) information a-symmetry occurs when it is hard for the principal to identify what the agent is actually doing and (2) the interests of the principal do not align with the interests of the agent (Eisenhardt, 1989). Both situations impact the reasoning of the CFO to disclose more or less information. Therefore, both types of agency problems play a central role in this study. In the first situation information asymmetry leads to adverse selection. Information asymmetry occurs often in
organizations with complex structures or complex industries, like the construction industry in which the long term contracts with multiple variable considerations increase the
complexity of revenue recognition, which makes it hard for shareholders to identify the current financial position and future profits of an organization. The second agency problem type, also called moral hazard, might exists when information a-symmetry is exploited by the CFO to strive for personal gain instead of protecting the interests of the principal (Boučková, 2015). CFOs are responsible and closely involved with the accounting process and are therefore able to influence the accounting quality providing limited disclosure or by influencing the style of reporting. CFOs are often tempted to influence accounting when they have something to conceal from shareholders or investors, like aggressive revenue accounting which is done to conceal bad performance and prevent stock price declines. Performance based remunerations increase the likelihood that CFOs are lured into aggressive revenue accounting. Other reasons for CFOs to provide less or lower quality disclosure is pressure of the CEO to disclose less information, this is in line with Beasley, Carcello, Hermanson and Neal (2009) who found that CFOs participate with CEOs in accounting manipulation with a self-serving purpose. An independent, well-functioning and active audit committee can reduce these agency costs and reduce information
9 asymmetry through monitoring the audit process and actively influencing and steering the internal audit committee and the financial reporting process led by the CFO (Menon and Williams, 1994; Abbott and Parker, 2000). In this role the audit committee increases the quality of disclosure which decreases information a-symmetry between the CFO and the shareholders, which in turn prevents self-serving behavior of the CFO. Information a-symmetry on his turn reduces agency costs (Easley and O’Hara, 2004). Blanco, Lara and Tribo (2014) found prove that segment disclosure, especially voluntary segment disclosure decreases agency costs.
Upper Echelons theory
In addition to the Agency Theory, Upper Echelons Theory is used to explain the
relationship between CFO career horizon and IFRS 15 disclosure quality. Lee, Park and Folta (2018) inspired me to come up with an equation for CFO career horizon. In short, CFO career horizon is based on two important components: age and tenure. Age is recognized by Hambrick and Mason (1984) as a managerial characteristic. Managerial characteristics are used as proxies for cognitive base and managerial values. Therefore managerial characteristics can be used to predict choices of top managers (Hambrick, and Mason, 1984). The CFO is a top manager, because the CFO is responsible for the daily financial operations of the company and has a seat on the executive board.
The Upper Echelon theory recognizes that CFOs make have a strong impact on the outcome of companies (Hambrick, and Mason, 1984). Among other things, CFO’s decide on the amount of disclosure and the form and tone in which the disclosure is presented. Hambrick and Mason (1984) also recognize that top managers characteristics, like age play a substantive role in recognizing strategic choices. The theory recognizes that groups of top managers with the same characteristics will behave and make decisions in the same way. This indirectly means that the amount and quality of IFRS 15 disclosure is influenced by characteristics of the CFO, like age. A previous study found evidence that older people are more conservative and take less risk than younger people (Vroom, and Palh, 1971).
Besides that, older people are often more experienced which positively impacts the quality of disclosure. Therefore it is likely that CFOs who are older and have more experience, have less to prove and take less risks, and are more willing to comply with IFRS 15 standards and provide voluntary disclosure.
10
Stakeholder Theory
Stakeholder Theory covers not only the shareholders perspective which is addressed by the Agency Theory, but includes any group that is affected by a company trying to achieve its goals (Freeman, 1984). Ansoff (1965) was the first to introduce the term “stakeholder theory” in combination with company objectives. Stakeholders control critical resources of companies which puts them in a position of power (Ullmann, 1985), which in turn
challenges companies to find a balance in the demands of all stakeholders. Freeman, Phillips and Sisodia (2018) recently defined the concept of stakeholder theory as: “knowing how to engage shareholders and create value for them”. Examples of
stakeholders are shareholders, suppliers, customers, government and creditors (Roberts, 1992).
Investors demand performance information from companies to assess if their
investments are worth the risk. Especially for companies with long term complex contracts with uncertainties regarding the recognition of variable considerations, company, segment and project specific information about IFRS 15 revenue recognition disclosure can help investors better assess the project specific or company specific risks and financial
performance. This helps investors make calculated and motivated investor decision. This is in line with stakeholder theory, which advocates that financial, environmental and social disclosure information is important in the relationship between companies and
shareholders and legitimize companies behavior and educates, informs and change perceptions and expectations (Adams, LarrinagaGonzelez, 2007; Adams, McNicholas, 2007).
Other important stakeholders for construction companies are standard setters, suppliers, subcontractors, customers, banks, employees, labor unions, and regulators, although not all of them are equally interested in revenue disclosure. Stakeholders that are interested in the revenue recognition disclosure are standard setters and international and national
regulatory bodies. These stakeholders have different motives than investors. Regulatory bodies like the ESMA are interested in how companies applied IFRS 15. They are mainly interested in the explanations companies provide for judgement made to recognize revenue in accordance with the IFRS 15 requirements. Multiple requirements of IFRS 15 demand judgement from companies, like the timing of satisfaction of performance obligations, the timing for the recognition of revenue, and the determination of the transaction price (IFRS 15, 2014. International and local regulatory bodies control if companies properly used judgement and explain this accordingly in their IFRS 15 disclosure. The ESMA and other
11 local regulator bodies want companies to provide transparent and qualitative IFRS 15 revenue recognition disclosure that faithfully depict reality to increase the credibility of the financial market and protect investors from incomplete or inaccurate information (ESMA, 2020).
III. Hypothesis development
In this chapter the literature background and theories are transformed into multiple
hypothesis. The first hypothesis focusses on the influence of CFO career horizon on IFRS 15 disclosure quality. The second hypothesis is more complex and investigate the
moderating effect of audit committee characteristics on the relationship between CFO career horizon and IFRS 15 disclosure.
CFO career Horizon
The CFO as member of the executive board of a company is responsible for reporting financial results in a fair way. The capabilities and experience of the CFO to interpreted financial reporting information makes him most suitable of all board members to coordinate the financial reporting process and take responsibility for the final version of the financial report (Mian, 2001; Geiger and North 2006). In this role he performs a crucial stewardship role to ensure quality of disclosure (Aier et al., 2005; Ge, Matsumoto, and Zhang, 2011; Geiger and North, 2006; Graham, Harvey, and Rajgopal, 2005). From the Agency and Stakeholder Theory perspective, governance bodies, shareholders and investors demand voluntary disclosure which put the executive board, and thus the CFO, under pressure to provide qualitative voluntary disclosure. The complex long term construction contracts with variable considerations provide useful information for
investor’s risk and investment decisions. Therefore investors demand companies to provide high quality company or project specific revenue disclosure. Besides that, regulatory bodies and standard setters are interested in how companies applied and disclosed judgement for the new IFRS standard.
The importance of the CFO in the financial reporting process is clear and the demand for disclosure as well, but does his age influence disclosure quality? Said, Omar and Abdullah (2013) found evidence that younger chairpersons are associated with less extended and lower quality voluntary disclosure due to a lack of experience and knowledge. Vroom and Palh (1971) found that older people are more conservative and more risk averse than younger people. From an Upper Echelon perspective age is seen as a top management characteristic. Top management characteristics influence companies outcomes. Therefore,
12 it is expected that older CFOs make more conservative and risk averse accounting
decisions. As a result of this, older CFOs are more aware of the risk for not providing enough disclosure and therefore violating the regulations. To limit this risk, they increase the quality of mandatory disclosure and provide extra voluntary disclosure. IFRS 15 recently became effective which results in uncertainty by CFO’s on how to disclose IFRS 15 revenue information. This makes it likely that older CFO’s fear for not providing sufficient disclosure stimulates the level of revenue disclosure quality. In addition to that, Reeb and Zhao (2013) studied the relationship between board of directors and disclosure quality and found prove that work experience of the board positively influences disclosure quality. They argued that work experience of inside managers able them to gather, filter and process various forms of information better than inexperienced managers (Reeb, and Zhao, 2013). The CFO, as an inside manager of the board, is responsible for the financial reporting process, therefore it is likely that CFOs who are better able to gather, filter and process data should positively influence disclosure quality. Beside that they argued that experienced inside managers are better capable of communicating with outside directors and include their views in future operations (Reeb, and Zhao, 2013). Experienced CFOs who are better able to process audit committees recommendations in the financial report should probably provide a higher levels of revenue disclosure quality. Therefore it is expected that CFO age and CFO tenure will positively influence audit quality disclosure.
Based on the above literature and arguments I expect that companies will provide higher IFRS 15 disclosure quality when CFO career horizon is shorter. To conclude, I have come to the following hypothesis:
H1: Shorter CFO career Horizons are positively associated with IFRS 15 disclosure quality for European listed construction companies.
Role of audit committee characteristics
In the past decades audit committees have become indispensable in the accounting process, especially as quality protector of the financial statements and oversight body of the audit process (Deloitte, 2018). The audit committee is entitled by EU audit directive 39.6 to take responsibility for the auditors’ selection procedure, monitoring of auditors’ performance and to inform the supervisory board on the outcomes of the audit (KPMG, 2016). They also monitor the efficiency of internal control, risk management systems and internal audit in relation to the audit (KPMG, 2016). But most importantly, the audit committee has a monitoring role in the financial reporting process and should in that role protect the
13 financial reporting integrity by making recommendations. Besides that, they should
oversee the CFO in his daily work regarding the financial reporting process and address which disclosures need priority and help develop and approve a working plan for the internal audit department (CIIA, 2019). The above information illustrates audit committees importance from an agency perspective. The audit committee represents the board of directors and indirect the shareholders of the company. In this role, the audit committee supervises the financial reporting process and corrects the CFO when he places his own interests above those of the organization. IFRS 15 is a relatively new standard, therefore it is expected that the audit committee will actively share opinions with the CFO about which IFRS 15 topics should be disclosed more elaborately and which topic should be disclosed less extensive. This all, is critical for the credibility of the financial report (KPMG, 2016; Crowe, 2019). By doing so, the committee makes sure that there is enough relevant and credible disclosure for stakeholders to assess top management performance and behavior (Allegrini, and Grecco, 2013). This is in line with agency theory and Abbott, Parker and Peters (2004), the latter one argues that audit committees monitoring role reduces
opportunistic behavior of the CFO and CEO and aligns shareholders and managements interest. This all, is critical for the credibility of the financial report (KPMG, 2016; Crowe, 2019).
From this paragraph onward I will address specific characteristics of the audit
committee that positively influence the performance of the audit committee’s agency role to provide higher quality disclosure. Previous studies found evidence that frequency of meetings plays an important role on the functioning of the audit committee (Beasley, Carcello, Hermanson, Lapides, 2000; Vafeas, 2005). Audit committees that meet more often are better able to reduce management’s freedom to manage earnings, and more frequent audit committee meetings are associated with less fraudulent activities by companies (Xie, Davidson, Dadalt, 2003). Both findings indicate that more frequent meetings reduce management’s freedom regarding financial reporting and thus increase audit quality. This is in line with the American study of Abbott and Parker (2000), in which they find that audit committees that meet at least twice a year have less financial statement problems. This sounds logical from my studies perspective, because I expect that audit committees actively influence material decisions made by the CFO for an important new standards, like IFRS 15. More audit committee meetings would probably increase the influence of the audit committees on the structure and topics explained in the IFRS 15 disclosure paragraph. Other studies confirm the positive influence that Abbot and Parker
14 (2000) found between audit committee meetings and disclosure issues. McMullen and Raghunandan (1996), Abbot et al. (2004) and Vafeas (2005) all associate less frequent audit committee meetings with an increase of financial reporting problems, Krishnan and Visvanathan (2007) go one step further and found evidence that audit committees who meet regularly are more likely to identify internal control weaknesses. All literature results indicate that audit committee frequency positively affect disclosure quality. This leads to the following hypothesis:
H2A: Frequency of audit committee meetings positively impact the association between a shorter CFO career horizon and higher quality IFRS 15 disclosure.
Does size matter? Larger audit committees enable companies to establish a more divers audit committee with members of different background, expertise, experience and gender (Sultana, Singh, Van der Zahn, 2015). Previous studies were skeptical about the extension of audit committee size (Mintzberg, 1983; Beasley, 1996). Mintzberg (1983) argues that audit committees that grow in size are less unified which negatively influences there decision making and might lead to impairment of control and monitoring functions (Collier, Gregory, 1999; Hillman, and Dalziel, 2003). But more recent literature of
DeZoort, Hermanson and Houston (2002) and Turley and Zaman (2007) argues that larger audit committees who view issues from different angles due to differences in expertise and background between members are better able to assess the activities, responsibilities and performance of the auditor (Sultana et al., 2015). Dezoort, Hermanson and Houston (2003) argue that audit committees with subcommittees possess wider set of skills which are valuable when conflict or issues occur in the audit report. Dezoort et al. (2003) do not focus on audit committee size, but it illustrates and confirms the view of resource dependence advocates like Dezoort et al. (2002) that larger audit committees have wider sets of skills which positively influences audit committees functioning. Therefore I expect that larger audit committees possess more expertise, skills and broadmindedness which improves their ability to understand the CFO’s daily activities. Besides that, it increases the audit committee’s capability to coordinate the working plan for the internal audit
department and supervise the financial reporting process. Only when audit committees become too big, I expect in line with criticizers like Mintzberg (1983) and Beasley (1996) that the advantages that come with size do not weight up to the disadvantages. But audit committee rarely consist of more than five members, which is in my opinion also the
15 tipping point between the advantages and disadvantages of size. Taking all of this in
account, I have come to the following hypothesis:
H2B: Audit committee size strengthens the association between a shorter CFO career horizon and higher quality IFRS 15 disclosure.
Previous studies found evidence that women are more risk averse, ethical and
communicate more effectively than man (Wood, Polek, and Aiken, 1985; Powell and Ansic, 1997; Byrnes, Miller, Schafer, 1999). Previous studies found evidence that women think differently than men and are therefore able to look at issues from a different angle, which reduces “group thinking” and errors in the audit committee activities, and thus increases financial reporting quality (Pucheta-Martinez and Fuentes, 2007; Srinidhi, Gul, Tsui, 2011). This is in line with the findings of Adams and Ferreira (2009) that groups with a mix of women and men perform better and more effective than single gender groups. Women are also more ethical than men (Ambrose, and Schminke, 1999). As a result of their ethicality, women might be more willing to provide information about loss-making contracts or sensitive claims that would negatively influence revenue. In addition to that, it also makes women less vulnerable for the tricks of the CFO to manage earnings or mask the true performance of the company. Besides that, Virtanen (2012) found that women are also more active on boards and committees than men, which means that it is likely that they influence the monitoring function with their moral and ethical qualities. Other qualities of women that positively influence audit committees functioning are their consciousness and the fact that women take their task more serious than men, these qualities indicate that women would probably demand more detailed and extensive audits (Schmitt, Realo, Voracek, and Allik, 2008; Fondas and Sassalos, 2000; Aldamen,
Hollindale, Ziegelmayer, 2018). In sum, all above reasons suggest that female audit committee members positively affect audit committee functioning. If audit committees function better, it is expected that financial reporting quality increases, and thus IFRS 15 disclosure quality. Therefore we expect the following:
H2C: More audit committee women positively impact the association between a shorter CFO career horizon and higher quality IFRS 15 disclosure.
16
IV. Methodology
This chapter describes the research methodology of this study. First the dependent variable is explained after that the independent variable and moderator will be addressed. In the last phase of this chapter time is spent on the relevant control variables that need to be
controlled to execute the study properly.
Sampling procedure
The sample consist of European listed firms who fall in the following NACE Rev. 2 classification of Eurostat (2008) who are hereafter named as ‘Construction companies’: construction of buildings (41), civil engineering (42), and specialized construction activities (43). The 35 European listed companies with the largest operating incomes are selected for this study. Those 35 companies are complemented with 2 large Dutch
construction companies to create a Dutch focus. The selected sample is relevant since the construction industry is affected by IFRS 15 and the magnitude of the companies severely impacts society. Especially the complex long-term contracts with customers, which contain multiple variable items and require loads of judgement, are interesting to study from an IFRS 15 perspective (Roozen, and Pronk, 2018; KPMG, 2014). The hand collection of data, which is labor-intensive work, in combination with a limited amount of resources, forced us to limit the sample size. In total, information from 63 annual reports from the years 2018 and 2019 is hand collected. 11 companies did not yet publish their 2019 annual report at the time of writing, therefor only 2018 data of those companies is included. Information about CFO age and CFO tenure was missing in a few annual reports. In those cases, I used the company’s website or the Orbis database, which contains data of top executives, as alternative sources. I compared CFO age and tenure data from Orbis with data from the annual report to secure the credibility of the extracted data.
Dependent variable: IFRS 15 disclosure quality
The data to measure the dependent variable IFRS 15 disclosure quality is derived from the annual reports of the years 2018 and 2019, published by the companies in the sample on their official websites. I took a qualitative approach to make the data quantitative which was necessary to assess the IFRS 15 disclosure quality. IFRS 15 became effective in 2018 which explains the lack of precedent for companies on how to apply and disclose IFRS 15. To assess IFRS 15 disclosure quality an objective self-developed disclosure index is used based on the requirements set in IFRS 15.111 to 15.129. The index is included in appendix A. A panel of master students, including myself, transformed after extensive discussions,
17 the legislation items into index values. To stimulate the objectivity of this study I
thoroughly discussed with fellow master students how to allocate points to the index values and how to control for the subjectivity of the assessors when assessing the disclosure indexes. An overview of the requirements to allocate points is explained per index item, as illustrated in appendix A. In short, if an index item is properly mentioned, conform the regulations, than one point is given, otherwise none. A score of two is awarded when voluntary company specific or project specific information is included in the financial statements. If a question is not applicable, no points will be allocated. The index is inspired by previous voluntary disclosure studies (Rietema, 2019; Boubaker, Hamrouni, Liang, 2015). In line with previous studies of Rietema (2019) and Boubaker et al. (2015) quality scores were based on the number of items disclosed, divided by the total number of points that could be achieved, the latter can be different for every annual report in the sample due to the fact that not all questions are applicable to all companies. The assessment of the company’s disclosures takes place by a panel of master students. To control for
inconsistency in the assessment of disclosures, overlap between the assessors is created. The missing of any precedent regarding the reporting of IFRS 15 information gives the assessment of companies’ revenues disclosures a voluntary nature. Guey, Samuels and Taylor (2016) find that voluntary disclosure volume shows a positive relation with firm complexity. This suggests that voluntary disclosure is used as a tool to improve the quality and transparency of complex mandatory disclosure, like IFRS 15 disclosure.
The following attention points help better understand the index scores and its limitations. All index items are assessed as if they were equally important. The main reason for doing so, is that weighting items is subjective and it heavily depends on what the researcher considers important. Besides that, Jones (1973) found that index term weighting did not result in substantive better scores, but it does result in additional trouble. Therefore I chose to leaf out index weighting. Another point of attention in the index model is that not every annual report is assessed by the same assessor and assessments are never free of subjectivity. To minimize inconsistency between the assessments of the assessors we took a few measures. First we had a few meetings in which we discussed how the assessment of disclosure quality should take place. We formulated a plan when and how to allocate scores per index item. Besides that, overlap was created among the assessors to reduce inconsistencies in scores of the different assessors and reduce subjectivity. Deviating results are discussed extensively by the assessors involved.
18 Independent variables
Previous researchers studied the relationship between CFO tenure and accounting (Jenkins and Velury, 2008; Basu, 1997; Watts, 2003i, 2003ii). Basu (1997) and Jenkins and Velury (2008) found evidence that auditor tenure resulted in more conservative accounting. Which implies that auditors with more experience as CFO, better comply with regulations and use less aggressive accounting methods. This leads to higher accounting quality and it is likely that this would also lead to higher IFRS 15 disclosure quality and more voluntary
disclosure. From an Upper echelon perspective people with the same characteristics will act in the same way, age is considered a personal characteristic by the Upper echelon theory (Hambrick, and Mason, 1984). A study from the USA found that age positively influences conservatism (Truett, 1993). Older people are more risk averse and would therefore be more willing to comply with regulations and provide additional voluntary disclosure. For these reasons, both age and tenure seem to influence disclosure quality. From the hypothesis development it is to be expected that the relative CFO career horizon will negatively affect IFRS 15 disclosure quality. Therefore, I need an equation that captures both CFO age and CFO tenure. Lee et al. (2018) inspired me to come up with the following equation:
CFO CAREER HORIZON = [CFOTENURE,t – CFOTENUREi,t] +[CFOAGE,t –CFOAGEi,t] CFO TENUREi,t is defined as the number of years (t) that the CFO is fulfilling his function as CFO for company i. CEO AGE is measured by the age of the CFO in year t when working at company i. CFOTENURE,t and CFOAGE,t are based on the average of the selected sample, in this case the construction industry. A positive CFO career horizon indicates that a CFO is relative young and has a shorter tenure than average in the industry. A negative value for the CFO career Horizon equation indicates that a CFO is older and/or has on average more tenure years than the average CFO in the industry. A relative CFO career horizon is measured, because CFOTENURE,t and CFOAGE,t have certain
shortcomings that need to be taken in consideration.
This sample consist of the largest European listed construction companies. Therefore the average age and tenure of the sample might not be representative for the entire European construction industry, which limits the conclusions that can be made based on the outcomes of this study for the entire construction industry. The advantage of using the average CFO age tenure is their representativeness for large European construction companies on who IFRS 15 has the largest impact.
19 Moderator
Larger audit committees often consist of people with different expertise or backgrounds which stimulates different ways of thinking and therefore improve the assessment of auditors activities, responsibilities and performance (DeZoort, Hermanson, Archambeault, and Reed, 2002; Turley and Zaman, 2007). Beside size, gender is also an important factor in increasing audit committee’s functioning. Their ability to look differently due to the fact that they have different qualities than men, reduces the possibility of group thinking, which instead reduces the amount of errors made during audit committee activities (Schmitt, Realo, Voracek, and Allik, 2008; Fondas and Sassalos, 2000; Aldamen et al., 2018). Lastly, audit committees that meet more often, are better able to reduce management’s freedom to manage earnings and more frequent audit committee meetings are associated with less fraudulent activities by companies, which should probably lead to more revenue and higher quality revenue disclosure (Xie et al., 2003). For this reasons size, female members and frequency of audit committee meetings are chosen as moderator effects for the relationship between CFO career horizon and revenue disclosure quality. To measure for size, female members and frequency of audit committees we used the number of audit committee members, the amount of female members on the audit committee and the amount of audit committee meetings in a year.
Control variables
In this paragraph, control variables are discussed that might influence this study. Based on previous literature I control for size, debt and profitability. All data for the control
variables can be retrieved from the annual reports.
Size
Previous studies found that size is positively related to audit quality disclosure (Chow and Wong-Boren, 1987; Lang and Lundholm, 1993). In 2002 Watson, shrives and Marston focused more on the relation between voluntary disclosure and size and found a positive relation as well. Therefore, it is reasonable to expect that company size impacts the quality of IFRS 15 revenue recognition. Revenue will be used to control for this variable. Because the revenue values of our sample are extremely big and divers a logarithm of revenue is used in the ordinary least squared regression.
Debt
In line with agency theory, organizations with high debt ratios can lower their cost of capital by declining information asymmetry (Watson et al., 2002). One way to decline information asymmetry is by providing additional voluntary disclosure (Jensen, and
20 Meckling, 1976). Broberg et al. (2010) find evidence of a significant relation between debt ratio and the amount of voluntary disclosure. In line with Broberg’s study we use debt ratio to control for debt.
Profitability
Firms with better news are more willing to disclose qualitative and precise disclosure quality (Newman and Sansing, 1993). Lang and Lundholm (1993) and Wallace and Naser (1995) found a positive relation between profitability and disclosure. Lang and Lundholm argue that this is in line with the information a-symmetry concept. It is not in managers’ interest to provide negative performance information, instead they try to cover up poor performance by earnings smoothing for which less and more vague disclosure is necessary (Imhoff, 1992; Newman and Sansing, 1993). In line with Erhardt, Werbel and Shrader (2013) we use ROA to control for profitability.
Empirical model
In order to investigate if the variables in the hypothesis development are associated, the ordinary least squared regression model in the statistical software program SPSS is used. The independent variable CFO career horizon, the moderating variables, and the control variables size have all been included in the regression model. To control for size we used the logarithm of revenue. Winsorizing is used to reduce the effect of outliers on the
samples outcomes. Audit committee frequency, profit, debt ratio and logarithm of revenue are winsorized at the 95th percentile, while CFO career horizon, profit and debt ratio are winsorized at the 5th percentile. All variables together in one model leads to the following formula:
SCORE= α1+β1*CFO career horizon + β2*FEMALE AC + β3* ACMEETING + β4*AC SIZE + β5* AC FEMALE MEMBERS*CFO Career Horizon+ β6* ACSIZE*CFO Career Horizon + β7*AC MEETING FREQUENCY*CFO Career Horizon + β8*PROFIT% + β9*SIZE + β10*DEBT + ε
21 Kind of
variable
Variable Definition Source
Dependent SCORE Total score on the assessment divided by the total amount of points that could be collected with exception of N/A questions.
Disclosure index Independent Moderator Interaction term CFOCAREER Horizon AC FEM. Members ACSIZE AC MEETING FREQUENCY AC FEM. MEMBERS*CFO Career Horizon ACSIZE*CFO Career Horizon AC MEETING FREQUENCY* CFO Career Horizon
Equation that captures both age and tenure
Number of female AC members.
Number of AC members on the audit committee
Frequency of AC meetings. Measured by the number of meetings in a year. Annual report ORBIS Annual report Annual report Annual report Computed in spss Computed in spss Computed in spss Control variable
PROFIT% Return on assets. Measured by net income divided by total assets.
Annual report
SIZE DEBT
The natural logarithm of revenue Total assets/ total liabilities
Annual report Annual report
22
V. Results
In this chapter the outcomes of the disclosure index scores, correlation analyses, regression analyses and best practices are discussed.
Descriptive statistics
This paragraph consist of descriptive statistics of the sample sorted by year. In table 2 the disclosure index scores of the sample are presented. For a more detailed overview of the disclosure scores sorted by company and year I refer to appendix B. The sample that is investigated consist of 63 annual reports of which 34 are 2018 annual reports and 29 are 2019 annual reports. The minimum score in 2019 was slightly lower 0,26 than in the previous year 0,3 and the maximum score is with 0,75 slightly higher than in 2018. No company came close to receiving the maximum score. The mean score for the year 2019 is with 0,515 slightly higher than for the year 2018 which has a mean of 0,502. The average scores for both years are relatively similar, although a slight raise in the average score of 2019 is visible. In the year 2019 companies had the availability to referential material from 2018 IFRS 15 disclosures, this could very well explain the small raise in IFRS 15 scores.
YEAR N Minimum Maximum Mean Std. Deviation SCORE 2018 34 0,3 0,7 0,502 0,102 SCORE 2019 29 0,26 0,75 0,515 0,111 total 63 0,26 0,75 0,508 0,105
Table 2: IFRS 15 SCORE Table 3 shows the descriptive statistics of the independent variable CFO career horizon and its components age and tenure, and the three moderator variables. The descriptive statistics show that for both years audit committee size was between the 3 and 5 members. In 2018 the average number of audit committee members was slightly higher than in 2019 3,85 to 3,69 respectively. Companies in the sample do have 0 to 3 women on the audit committee board. On average in 2018 there were 1,41 women on the audit committee board which is slightly less than in 2019 when 1,52 women had a seat on the board. This increase in women on the board, might be explained by a demand from society to stimulate the amount of women on the board. Although there are still more men than female on the audit committees as is shown by the two variables. Of the 3,69 members on an audit committee board in 2019, only 1,52 where women. Every audit committee organized at least 3 meetings a year with a maximum of 20 in 2018 and 7 in 2019. On average the frequency of audit committee meetings declined from 5,53 to 4,48 which is remarkable, but could not explained by a logical reason. The change in minimum and maximum age of
23 CFO’s between 2018 and 2019 is relatively low. Overall, the average minimum age is 39 and the maximum age is 70. The mean shows a small decline from 52,44 to 52. The relatively high age of the CFO is explainable, the function is vital for the company and therefore requires someone with experience to fulfil the task. The CFO tenure has a large range. The minimum CFO tenure in 2018 and 2019 was 1 year and the maximum was 21 respectively 22 years. On average tenure of CFO’s increased from 6,71 to 7,21. CFO career horizon has a range between -32,683 and 16,317 and a mean of zero.
YEAR N Minimum Maximum Mean Std. Deviation
AC Size 2018 34 3 5 3,85 0,784 AC Size 2019 29 3 5 3,69 0,66 Total 63 3 5 3,78 0,728 AC Female members 2018 34 0 3 1,41 0,925 AC Female members 2019 29 0 3 1,52 0,871 Total 63 0 3 1,46 0,895 CFO Age 2018 34 42 69 52,62 5,934 CFO Age 2019 29 39 70 52,1 6,966 Total 63 39 70 52,38 6,381 CFO tenure 2018 34 1 21 6,71 5,870 CFO tenure 2019 29 1 22 7,21 6,366 Total 63 1 22 6,94 6,059 Ac meeting freq. 2018 34 3 20 5,53 3,184 Ac meeting freq. 2019 29 3 7 4,48 1,214 Total 63 3 20 5,05 2,517
CFO career horizon 2018 34 -30,683 16,317 -0,006 10,803 CFO career horizon 2019 29 -32,683 16,317 0,007 11,899
Total 63 -32,683 16,317 0 11,228
24 Correlation analyses
Correlations can influence the regression analyses, therefore a correlation analyses is performed to identify variables that correlate. Variables which have high correlation might bias the regression analyses. The outcomes of the correlation matrix are shown in table 4.
1 2 3 4 5 6 7 8
1 SCORE 1
2 CFO CAREER HORIZON -0,153 1
3 ACSIZE -0,054 0,037 1 4 AC FEM. MEMBERS -0,178 ,320* ,357** 1 5 AC MEETING FREQ. 0,124 0,127 0,107 0,081 1 6 % PROFIT ,439** 0,037 0,018 0,245 -,278* 1 7 DEBT RATIO ,295* 0,215 -0,026 -0,045 ,560** ,458** 1 8 LOG SIZE 0,088 0,206 0,223 0,261* 0,232 -0,002 0,208 1 ** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
Table 4: Correlation among variables The outcomes of the correlation analyses resulted in logical and unexpected correlations among variables, as can be seen in table 4. In this paragraph only the moderating and independent variables are explained. The First unexpected correlation is the positive relation (r = .320, p = .011) between CFO career horizon and audit committee female members (AC FEM. MEMBERS). No scientific articles were available to back up this relation between a CFO’s career horizon and the amount of female members, but it might be possible that companies with younger or less experienced executive boards are also more progressive in hiring female members for the audit committee. In addition to the correlation between the independent variable CFO career horizon, audit committee female members is also correlated to audit committee size (r = .357, p = .004). There is a logical reason for the relation between audit committee size and the number of audit committee female members. If audit committee size increases it is likely that the amount of female members would increase as well. Besides that, companies are under pressure of society and governments to increase the amount of women on boards and committees. Therefore, it is logical that companies that increase the amount of board members, or hire new audit committee members, choose for a female instead of a men. Other variables that showed remarkable correlations were profit (PROFIT%) and debt ratio. Profit is strongly correlated (r = .439, p
25 = .000) with the disclosure quality score (SCORE) and debt ratio does also correlate (r = .295, p = .019) with audit quality score (SCORE).
Multicollinearity
All correlations in table 4 do not exceed the 0.7 which could cause collinearity (Hair et al., 2010). In table 5 the results of the multicollinearity analysis are shown to confirm that there is no collinearity. Multicollinearity is measured by VIF. Increases of the VIF level result in more multicollinearity, which indicates the impact of other independent variables on the standard error (Menard, 1995; Marquardt, 1970; Neter, Wasserman, Kutner, 1989). The VIF should be reasonably low, which is below 10 according to Neter et al. (1989). In table 4 all variables are tested on multicollinearity and the results show that none of the variables show values above 2. This means that a regression analyses with a reliable error term is guaranteed.
Variable Tolerance VIF
AC SIZE 0,808 1,238
AC FEMALE MEMBERS 0,680 1,471 AC MEETING FREQ. 0,657 1,522 CFO CAREER HORIZON 0,807 1,240
PROFIT% 0,728 1,374
DEBT RATIO 0,552 1,812
REVENU 0,841 1,189
Table 5: VIF with all variables Regression analyses
The ordinary least squared regression is used to examine if the hypothesis developed have significant relationships. Multiple models are used to investigate all hypothesis separately. The results of these analyses are shown in table 6. Model 1 which will further be referred to as the basic model, tests the relation between all control variables in relation to the dependent variable (SCORE). Model 2 adds the independent variable CFO career horizon to the basic model with the aim to test hypothesis 1 “Shorter CFO career Horizons are positively associated with IFRS 15 disclosure quality for European listed construction companies”. The following three models are composed with the aim to examine the moderating effects of audit committee characteristics.
A moderating relationship is created by standardizing the independent and moderating variables involved. After the variables are standardized they can be multiplied with each other to create a new variable ‘the interaction term’. For model 3 this means that the
26 independent variable CFO career horizon, the moderating variable ACSIZE and the
interaction term ‘CFO career horizon*AC Size’ are added to the basic model to test if hypotheses 2b ‘Audit committee size strengthens the association between a shorter CFO career horizon and higher quality IFRS 15 disclosure’ has a significant relationship. Model 4 is composed to test if hypothesis 2c ‘More audit committee women positively impact the association between a shorter CFO career horizon and higher quality IFRS 15 disclosure’ is significant. To examine this, the independent variable CFO career horizon, the moderator variable number of audit committee women and the interaction term ‘CFO career horizon*AC female members’ are added to the basic model. In model 5 hypothesis 2a ‘Frequency of audit committee meetings positively impact the association between a shorter CFO career horizon and higher quality IFRS 15 disclosure’ is tested. In order to test hypothesis 2a the independent variable CFO career horizon, the moderator variable AC meeting frequency and the interaction term ‘CFO career horizon*AC meeting frequency’ are added to the basic model. The final model, which is named model 6, tests the entire model and consist of all independent variables, moderating variables and control variables.
27
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Intercept 0,369 0,417 0,729 0,365 0,387 0,738
(0,198) (0,154) (0,65) (0,242) (0,199) (0,038)
CFO CAREER HORIZON (-) -0,001 -0,012 -0,015 -0,012 -0,027
(0,357) (0,324) (0,297) (0,373) (0,076)* AC SIZE (+) -0,003 0,012 (0,811) (0,435) AC FEMALE MEMBERS (+) -0,016 -0,024 (0,247) (0,117) AC Meeting FREQ. (+) -0,008 0,013 (0,609) (0,428) PROFIT -0,745 -0,757 -0,555 -0,679 -0,766 -0,343 (0,004)*** (0,004)*** 0,039** (0,011)** (0,004)*** (0,245) DEBT RATIO 0,005 0,004 0,003 0,003 0,006 0,007 (0,452) (0,569) (0,662) (0,606) (0,487) (0,401) SIZE 0,007 0,005 0,007 0,006 -0,01 (0,583) (0,696) (0,605) (0,645) (0,523) CFOCarHorizon_AC Size (-) -0,03 -0,048 (0,042)* (0,014)** CFOCarHorizon_AC Femmember (+) -0,005 0,016 (0,714) (0,302) CFOCarHorizon_ACMeetFreq. (+) -0,001 0,013 (0,928) (0,391) Observations 63 63 63 63 63 63 R-Square 0,208 0,22 0,278 0,241 0,225 0,327 Adj. R-Square 0,168 0,166 0,201 0,16 0,157 0,198 MAX VIF 1,338 1,384 1,608 1,387 1,996 1,812 F-value 5,174*** 4,087*** 3,602*** 2,97** 2,704** 2,526** ***. significant at the 0.01 level (2-tailed)
**. significant at the 0.05 level (2-tailed) *. significant at the 0.10 level (2-tailed)
28 Control variables
The first model in table 6 tests the relationship between the control variables Profit, Size, Debt ratio and the dependent variable SCORE. In model 1a significant result are found between Profit% and the dependent variable SCORE (β= -0.745, P=0.004). This relationship implicates that a higher return on assets leads to a decrease of the IFRS 15 disclosure index score. Other control variables, Debt ratio (β= 0.005, P=0.452) and Size (β= 0.007, P=0.583) do not have significant relationships with IFRS 15 disclosure index scores.
CFO career horizon
Model 2 in table 6 illustrates the relationship between CFO career horizon and IFRS 15 disclosure index score. The outcomes indicate a minimal negative non-significant relationship between CFO career horizon and IFRS 15 disclosure index (β= -0.001, P=0.357). Although the relationship is slightly negative, the relationship is not significant. But, in the complete model (6) we found a significant relationship for CFO career horizon and disclosure quality. Based on the information in model 6, hypothesis 1 is accepted.
Moderating role of audit committee size
Model 3 of table 6 examines the moderator effect of audit committee size on the relationship between CFO career horizon and IFRS 15 disclosure index score. The regression analyses did not result in a significant positive relationship between audit committee size and the index score (β= -0.003, P=0.811) or between CFO career horizon and the index score (β= -0.012, P=0.324). The second step was introducing the interaction term (CFO career horizon*audit committee size). The results show a significant
relationship between the interaction term ‘CFO career horizon*audit committee size’ and IFRS 15 disclosure index score (β= -0.03, P=0.042). This result indicates that there is a significant negative relationship between the interaction term and IFRS15 disclosure index score. To conclude, CFOs with a shorter career horizon and a relative large audit
committee size have higher IFRS 15 disclosure scores than CFOs with a longer career horizon with a smaller audit committee size. Based on the outcomes of model 3 I accept hypotheses 2b.
Moderating role of audit committee female member
The results from the relationship between independent variable CFO career horizon and the moderating variable audit committee female members are discussed in model 4. Both variables, CFO career horizon (β= -0.015, P=0.297) and AC female members (β= -0.016, P=0.247) do not show significant relationship with the dependent variable. In line with the
29 independent and moderator variable, the interaction term ‘CFO career horizon*AC female members’ does not show a positive significant relationship, but a slightly negative
insignificant relationship instead (β= -0.005, P=0.714). Therefore hypothesis 2c is rejected, there is no evidence that the amount of female members significantly strengthens the negative relationship between CFO career horizon and IFRS 15 disclosure index score.
Moderating role of audit committee meeting frequency
Model 5 of table 6 examined the relationship between the moderator (CFO career horizon*ACMeetFreq) and the dependent variable. The moderating variable Audit committee meeting frequency (β= -0.008, P=0.609) and the independent variable CFO career horizon (β= -0.012, P=0.373) do not show significant relationship with IFRS15 disclosure index score. When the interaction term CFO career horizon*AC meeting frequency was added to the regression analyses an insignificant negative relation is found between the moderator CFO meeting frequency and the dependent variable score (β= -0.001, P=0.928). Therefore hypotheses 2a is rejected, because there is no evidence that the audit committee meeting frequency strengthens the negative relationship between CFO career horizon and the dependent variable IFRS 15 disclosure index score.
Complete model
Finally, all variables are tested together in model 6. A significant negative relationship at the 10 percent level is found between CFO career horizon and IFRS 15 disclosure quality (β= -0.027, P=0.076). Although this relationship was not significant in model 2, I can accept hypothesis 1 based on the complete model. Shorter CFO career horizons positively influence IFRS 15 disclosure quality. In line with my results from model 3 a significant negative relationship was found for the interaction term ‘CFO career horizon*AC size’ (β= -0.048, P=0.014). Based on the outcomes of model 3 and 6 I accept hypothesis 2b. Audit committee size strengthens the association between a shorter CFO career horizon and higher quality IFRS 15 disclosure. Model 6 can be seen as the base model because it includes all variables and has the highest explanatory power. Approximately 32.7% of the disclosure quality score can be explained by the characteristics of the complete model. Other models in this study show a lower R-Squared which means that all variables together have a positive influence on the explanatory power of the model.
Sensitivity analyses
Sensitivity analyses are performed to examine the robustness of the components age and tenure of the CFO career horizon variable. The main goal of this analyses is to investigate
30 if the findings in table 6 are not a result of the composition of CFO career horizon. To exclude this possibility, Lee et al. (2018) inspired me to create modified career horizons based only on CFO age or tenure. In addition to that, I adjusted CFO age and tenure by the average age and tenure of the sample. Plenty of studies presented the importance of both CEO age and tenure in career horizon (Lee et al., 2018; Krause, Semadeni, 2014). But none did this for the CFO. Therefore, I have examined if one of the components has a stronger influence on CFO career horizon. The results in table 7 show that career horizon based only on age has significant negative relationships with audit quality (β= -0.034, P=0.027) and the interaction term ‘CFO career horizon (only age)*audit committee’ (β= -0.044, P=0.016), which is in line with the outcomes in table 6. I only found 1 significant result for CFO career horizon based solely on tenure. The interaction term ‘CFO career horizon (only tenure)*audit committee’ had a significant negative effect (β= -0.04, P=0.038). Overall, the outcomes of alternative measure CFO career horizon (only age) supports the results of the main analyses. The alternative measure only CFO tenure does also support all significant relationships of the main analysis, except for the relation between CFO career horizon and audit quality.
31
Only CFO AGE ONLY CFO TENURE
Intercept 0,715 0,679
(0,039) (0,062)
CFO CAREER HORIZON (-) -0,034 -0,012
(0,027)** (0,449) AC SIZE (+) 0,009 0,007 (0,511) (0,642) AC FEMALE MEMBERS (+) -0,025 -0,017 (0,093)* (0,26) AC Meeting FREQ. (+) -0,015 -0,011 (0,362) (0,5) PROFIT -0,256 -0,437 (0,385) (0,144) DEBT RATIO 0,006 0,008 (0,446) (0,349) SIZE -0,009 -0,007 (0,552) (0,643) CFOCarHorizon_AC Size (-) -0,044 -0,04 (0,016)** (0,038)** CFOCarHorizon_AC Femmember (+) 0,017 0,009 (0,263) (0,561) CFOCarHorizon_ACMeetFreq. (+) 0,013 0,011 (0,396) (0,497) Observations 63 63 R-Square 0,349 0,292 Adj. R-Square 0,224 0,156 MAX VIF 2,449 2,563 F-value 2,789*** 2,149**
***. significant at the 0.01 level (2-tailed) **. significant at the 0.05 level (2-tailed) *. significant at the 0.10 level (2-tailed)
32 Best practices
In this chapter a summary of the identified best practices is given. For the extensive version with illustrations I refer to appendix C Best practices. YIT OYJ transaction price and timing of revenue disclosure is identified as best practice. Their approach stands out through the use of company and project specific information. Bouygues only explained non-specific standardized information based on the IFRS 15 requirements, while YIT OYJ mentions how IFRS 15 affects the transaction price and timing of revenue for their
segment Housing Finland. Among other things they mention that the transaction price includes the share of housing corporation loans allocated to the apartment, that every apartment is recognized as a performance obligation and that payments are received in advance of delivery.
Hochtief’s approach towards IFRS 15.120A ‘unrecognized performance obligations’ is also considered a best practice, because they provide a clear table with the amounts of unrecognized performance obligations sorted by regional divisions and type of activies. CFE on the other hand is focussed on the short term, and provides only the total amount of remaining performance obligations and the part of the remaining performance obligations that they expect to recognize in the coming year. Hochtief’s table makes it possible for shareholders to assess the expected future revenue of a company per region and segment in a blink of the eye. This helps investors assess the sustainability of future profit by specific regions or activities.
The third best practice is Salini Impregilo who extensively motivate their choice to use the percentage of completion method for the transfer of goods and services. Most
companies, like Colas shortly describe that the percentage of completion method is used to recognize the transfer of goods or services. A method can severely change a company’s revenue outcomes, therefore it is vital for shareholders to understand what the motives of the executive board are when deciding on the method to choose.
VI. Discussion
In this chapter I will elaborate on remarkable and interesting results, the limitations of this study and the opportunities for further research. In accordance with the expectations from previous literature and in line with hypothesis 1 I found a significant negative relationship between CFO career horizon and IFRS 15 disclosure quality. Interesting is that the
relationship between CFO career horizon and disclosure quality was insignificant when the moderating variables were excluded. Therefore I conclude that all variables combined
