MSc-Controlling University of Groningen Date of submission: 24-6-2019

MSc-Controlling University of Groningen Date of submission: 24-6-2019

Chiel Bruins S3541894 Goudplevier 118 8271 GC IJsselmuiden

31641334934

Word count: 10 330 Supervisors:

Prof. dr. R.L. ter Hoeven Msc. R. van Duuren

IFRS 15 in the Telecommunications, Technology,

Utilities and Construction & Materials Industry

IFRS 15 in the Telecommunications, Technology, Utilities and Construction & Materials Industry

Abstract: This study investigates the effect of three corporate characteristics (leverage,

profitability and free cash flow) on IFRS 15 disclosure quality. The sample size consists of data

gathered from 52 financial statements starting from the 1

of January 2018 until 31 December

2018, since IFRS 15 is effective for European listed companies starting January 1 2018. The

selected industries consist of the following: Telecommunications, Technology, Utilities and

Construction & Materials, since it is expected that IFRS 15 has the largest impact on these

industries. This study is the first to investigate the potential impact of corporate characteristics

on IFRS 15 disclosure quality. Together with three other researchers, a disclosure index has

been constructed in order to measure disclosure quality. Using a linear regression model, this

study was not able to find any significant findings on all hypotheses. However, the results of

this study do show that the industry Utilities has a negative impact on IFRS 15 disclosure

quality.

Table of content

1. Introduction ... 4

2. Literature review ... 6

2.1 Background literature ... 6

History of IFRS ... 6

Transition to IFRS ... 6

Mandatory and voluntary disclosure ... 7

Past research on the effect of determinants on disclosure quality ... 8

2.2 Theoretical framework ... 8

Agency Theory ... 9

Signaling Theory ... 9

Stakeholder Theory ... 10

2.3 Hypotheses development ... 10

Leverage ... 10

Profitability ... 11

Free Cash Flow ... 11

3. Research Methodology ... 12

3.1 Sample ... 12

3.2 Dependent variable ... 13

3.3 Independent variables ... 14

Leverage ... 14

Profitability ... 14

Free Cash Flow ... 15

3.4 Control variables ... 15

Firm size ... 15

Industry type ... 15

3.5 Data adjustments ... 15

3.6 Statistical model ... 16

4 Results ... 16

4.1 Descriptive statistics ... 16

4.2 Validity model ... 17

4.3 Correlation matrix ... 18

4.4 Regression analysis ... 18

5 Conclusion and Discussion ... 21

5.1 Findings ... 21

5.2 Theoretical and practical implications ... 21

5.3 Limitations and future research ... 22

References ... 23

Appendix ... 27

4

1. Introduction

The world around financial reporting has been constantly changing. One of the reasons for this continuous change, which has been addressed in many papers and news articles, is due to severe financial scandals in the past. Therefore, the number of regulations has increased drastically the past few years. One large regulation standard which has become mandatory for listed European firms in 2005 is the International Financial Reporting Standards (IFRS). IFRS has two main objectives: improve the comparability and quality of financial reports (De George et al., 2016).

Until now, the recognition of revenue has been one of the biggest gaps between IFRS and the US Generally Accepted Accounting Principles (GAAP). According to Aarab, Bissessur and ter Hoeven (2015), revenue is one of the most important indicators to measure the functioning of an organization. To recognize revenue under IFRS, there were two specific standards:

International Accounting Standards (IAS) 18 Revenue and IAS 11 Construction Contracts.

However, IAS 18 contains broad principles for organizations how to recognize revenue, whereas IAS 18 left room for interpretation and offered organizations little to no guidance for more complex contracts on how to recognize revenue. This resulted in low comparability of financial reports. On the contrary, US GAAP is more detailed about the recognition of revenue.

US GAAP contains approximately 100 documents on revenue recognition for specific situations, which is substantially more detailed than IFRS. To tighten the gap, the International Accounting Standards Board (IASB) and the Financial Accounting Standards Board (FASB) worked together in a joint project and introduced another Standard to set aside their differences in revenue recognition.

On the 28th of May 2014, the IASB issued IFRS 15 Revenue from Contracts with Customers.

This new Standard is effective for financial statements on or after the 1

of January 2018. In short, IFRS 15 indicates how and when revenue will be recognized by an IFRS reporter. One of the goals of IFRS 15 is to provide more informative and relevant information to users of financial statements. To this day (to my knowledge), there is no available research on the effectiveness of IFRS 15 on the disclosure quality of listed European organizations. There are studies which address IFRS adoption and disclosure quality, but not of IFRS 15. Hence, the implementation of IFRS, which is effective on or after the 1

of January 2018, is an interesting setting to investigate. Because the goal of IFRS is to increase the comparability and quality of financial reports, I expect that the implementation of IFRS 15 will positively affect disclosure quality.

According to Ahmed and Courtis (1999), a considerable amount of international literature has been developed which investigates the association between corporate characteristics and disclosure levels in annual reports. Marston and Shrives (1991) noted, for example, that firm size, leverage, profitability and audit firm size were the most common corporate characteristics which are examined with regard to disclosure level (Ahmed & Courtis, 1999). In this research, I will focus on three corporate characteristics which are treated in relation to IFRS 15. The three corporate characteristics include leverage, profitability and free cash flow.

According to Jensen and Meckling (1976) and Smith and Warner (1979), organizations with more debt in their capital structure incur higher agency costs. These agency costs can be mitigated by disclosing more information to shareholders, because financial information assists debt holders to monitor management behavior (Chung & Jung, 2016).

In studies which have been performed in the past, profitability has been positively associated

with disclosure (Wallace et al., 1994; Owusu-Ansah, 1998). From a corporate perspective, an

argument behind this association is that profitability as a measure of management performance,

5

managers of profitable organizations are more inclined to disclose information due to management compensation (Palmer, 2006). From a stakeholder perspective, stakeholders are also interested in profitability, as more profitable organizations are more likely to hand out dividends and profitable organizations are also more interesting for investors to invest in.

One corporate characteristic which is not commonly seen together with disclosure level studies is Free Cash Flow (FCF). According to Zhang (2009), firms with free cash flow tend to have higher agency costs due to conflicting interests between shareholders and managers. In the study of Jensen (1986), he mentioned that companies with high FCF prefer to invest in negative net present value projects rather than to pay dividends. Studying the different levels of FCF could, therefore, be an interesting corporate characteristic for shareholders as high FCF decreases dividend pay-out of organizations. Based on the agency theory, if there is a conflict between shareholders and managers, this will negatively impact disclosure quality as agents have no interest to disclose private information of the firm to principals if their interests misalign.

As follows, these three corporate characteristics could be used in conjunction with IFRS 15 disclosure quality because as of right now, there is no available research which examines the effect of these three characteristics on IFRS 15 disclosure quality. The goal of this study is to investigate the effect of leverage, profitability and FCF on IFRS 15 disclosure quality. More specifically, I expect that listed European organizations which have implemented IFRS 15 and are more leveraged and profitable have higher disclosure quality. Furthermore, I expect that firms with higher levels of FCF will have lower disclosure quality. Hence, my research question is formulated as such:

What is the effect of leverage, profitability and FCF on IFRS 15 disclosure quality?

In order to measure disclosure quality, I have developed together with three other researchers a disclosure index. The goal of this index was to determine if there is a difference in disclosure quality between organizations in the Technology, Telecommunications, Utilities and Construction & Materials industry.

Starting on or after the 1

of January of 2018, listed European organizations mandatorily have to disclose their financial statements under IFRS 15 for the first time. Many studies try to find a relation between the effect of regulations on disclosure (Barth et al., 2008; Yurisandi &

Puspitasari, 2015; Gastón et al., 2010). Therefore, I will contribute to this research strain by examining the effect of IFRS 15 on disclosure quality under certain corporate characteristics.

More specifically, the setting in which this study takes place is quite unique since 2018 is the first year that IFRS 15 is effective. Therefore, this will be the first study on the effect of three corporate characteristics (leverage, profitability and FCF) on IFRS 15 disclosure quality, whereas disclosure quality is measured through a disclosure index which is based on the disclosure requirements of IFRS 15. Although there is a lot of research on regulation and disclosure quality, the regulation around financial disclosure constantly changes, which makes this area still relevant to investigate.

This research is organized as follows. The second chapter will contain a literature review, which

will result in the development of the hypotheses. The third chapter will contain the methodology

used for this research. The fourth will contain the results of the research. In the final chapter, I

will include the conclusions, limitations and possibilities for future research in this area.

6

2. Literature review

The first paragraph of this section will cover the background literature, whereas the second paragraph covers the theoretical framework. Lastly, based on the background literature and theoretical framework, I will discuss my hypotheses which include three corporate characteristics.

2.1 Background literature

First, I will discuss the history of IFRS to provide some background information why the IFRS is created in the first place. Secondly, the general transition to IFRS is discussed, as the transition to IFRS can cause major consequences in disclosure quality. Next to this, the transition to IFRS 15 is discussed, as the application of IFRS 15 can be difficult for organizations since they have no indication whether or not they are applying IFRS 15 sufficiently. Fourthly, the mandatory and voluntary aspects of IFRS 15 are provided and explained why this partition is important to make for this research. The last part contains information on the effect of determinants on disclosure quality, as results of past research indicate that disclosure quality can be affected positively or negatively by introducing IFRS.

History of IFRS

Professional accounting bodies consisting of the following countries: Australia, Canada, France, Germany, Japan, Mexico, Netherlands United Kingdom, Ireland and the United States formed in 1973 the International Accounting Standards Committee (IASC), in which they agreed to adopt International Accounting Standards for cross-border listings. Starting from 1973 until 2001, the IASC has issued frameworks for organizations to present financial statements and have also completed projects to reduce the range of accounting policy choices under international standards. However, the IASC had certain problems with guaranteeing their independence, so the IASC became in 2001 the International Accounting Standards Board (IASB). The IASB Chairman, Sir David Tweedie, stated the following:

(IASB, 2001, p. 1). After 2001, the IASB has seen some international convergence of accounting standards. For example, the ‘Norwalk Agreement’ was signed in 2002 to improve convergence between IFRS Standards and the US GAAP. In 2005, the IFRS became mandatory for European listed organizations and these organizations have to report according to IFRS. The application of IFRS has to be complete, meaning that “cherry picking” is not permitted. This implies that an organization cannot choose which elements of IFRS they want or do not want to implement. Organizations have to apply IFRS completely.

Since IFRS became mandatory in 2005, the reporting standards had two main objectives. The first objective is to increase comparability of financial reporting between countries, whereby the second objective is to enhance the quality of financial reporting (De George et al., 2016).

Transition to IFRS

The transition to IFRS could cause a major fundamental change in the business environment, because prior to 2005, companies followed a variety of US GAAP (Soderstrom, & Sun, 2007).

In light of this information, Doupnik and Perera (2012) identify four major issues which can be

the consequence of diversity in accounting standards internationally. Consolidation, access to

foreign markets, comparability of financial reports and the quality of accounting standards are

consequences which the authors identify in their research. Consequently, the goal of the IASB

is to tackle these problems by publishing new standards and to eventually reduce the diversity

experienced internationally. As of this day, the IASB is still publishing new standards to

improve disclosure quality and comparability of financial reports. On the 28

of May 2014, the

IASB introduced IFRS 15 Revenue from Contracts with Customers.

7

IFRS 15 Revenue from Contracts with Customers, in short, indicates how and when revenue will be recognized by an IFRS reporter. As mentioned before, organizations which have to apply IFRS have to apply it completely, whereas “cherry picking” is not permitted. This principle also applies for IFRS 15. If an entity applies IFRS 15, all aspects of IFRS 15 need to be included in the financial statements of an entity. However, nowadays, organizations have no measure to indicate whether or not they are disclosing sufficiently, since in practice organizations mainly benchmark how other organizations disclose their financial statements compared to their own. Since 2018 is the first year that IFRS 15 became effective, organizations do not have the resources readily available to them to benchmark their financial statements to others. This setting makes this study quite unique as it investigates if there is a difference in disclosure quality of IFRS 15 between listed European organizations without the opportunity for benchmarking in different industries. In order to differentiate disclosure quality between organizations, first a distinction has to be made between mandatory and voluntary disclosure.

Mandatory and voluntary disclosure

The main difference between mandatory and voluntary disclosure is that mandatory disclosure is based on requirements which are set by regulators in a certain country (Charumathi, &

Ramesh, 2015). Voluntary disclosure is more frequently used by organizations to differentiate themselves with other organizations on the market (Zhuang et al., 2014). Also, Cheynel (2013) indicates that firms which voluntarily disclose have a lower cost of capital which, from a corporate perspective, is beneficial for the organization as a lower cost of capital increases firm value. In another study, Deegan (2002) indicates several reasons why organizations disclose information voluntarily: Economic rationality, to comply with community expectations, the result of threats to the legitimacy of the organization, managing particular stakeholder groups or to attract investment funds. However, if an organization chooses to voluntarily disclose information, they should also consider the types of costs which may arise from disclosing voluntary information such as agency-, proprietary-, or political costs (Shehata, 2014). These costs can withhold certain organizations to disclose voluntary information, as the benefits are insufficient compared to the costs associated with disclosing voluntary information. Past research shows that these benefits can include an improvement in earnings quality (Francis et al., 2008) or a decrease in information asymmetry between investors and management (Rezaee

& Tuo, 2017) for which the research of Francis et al. (2008) can be interpreted as higher disclosure quality.

IFRS 15 consists out of two components: mandatory and voluntary disclosure. The study of Roozen and Pronk (2018) highlights some major differences between the old and new regulation on revenue recognition (IAS 11 and 18 versus IFRS 15). One particular highlight is important to mention, which is that IFRS 15 has additional requirements regarding disclosure on revenue from contracts with customers. These additional requirements can be found in IFRS 15.110 up to and including 15.129. Since it is required to disclose additional information, it is mandatory of nature and not voluntary. As mentioned before, if an organization applies IFRS, they have to apply it completely, whereas “cherry picking” is not permitted.

However, if an organization discloses information in their financial statements which is based

on a disclosure requirement, they can choose to either 1) disclose sufficient information just to

meet the mandatory disclosure requirement at a minimum level or 2) provide additional

information regarding this disclosure requirement since, as was mentioned in the research of

Zhuang et al. (2014), organizations may want to distinguish themselves with other organizations

in the market to, for example, mark their superior performance or to manage particular

stakeholder groups. Organizations can also choose not to disclose additional information

because, for example, the benefits of disclosing additional information are insufficient

8

compared to the costs. By determining this quality difference in financial statements, a distinction can be made of the level of disclosure quality between organizations. However, this distinction is subject to subjectivity since myself and three other researchers have to determine the level of disclosure quality. This subjectivity risk is further discussed in the chapter on Research Methodology.

Past research on the effect of determinants on disclosure quality

One of the main objectives of IFRS was to enhance the quality of financial reporting. Most researchers do agree that the implementation of IFRS has increased the quality of financial reporting. For example, Barth et al. (2008) argue that firms which adopt IFRS have less earnings management, more value relevance and more timely loss recognition, which is seen as higher accounting quality. Yurisandi and Puspitasari (2015) found that the IFRS adoption increased the quality of financial reporting. Iatridis (2010) also researched IFRS implementation and accounting quality. He explains that by providing more informative and relevant information to users of financial statements, the implementation of IFRS generally supplements accounting quality (Iatridis, 2010). Gastón et al. (2010) found that the quantitative impact of mandatory IFRS adoption was significant in both the UK and Spain. However, Ball et al. (2003) argue that high-quality financial reporting is not generally the result of high-quality accounting standards.

In line with their research, Jeanjean and Stolowy (2008) argue that the improvement of financial reporting is not the result of accounting standards alone. Management incentives and national institutional factors are also necessary to improve financial reporting (Jeanjean, & Stolowy, 2008). Guay et al. (2016) argue that the growing complexity of financial statements negatively affects the information environment of organizations, because the complexity makes it difficult for management to provide information to investors. Consistent with this reasoning, You and Zhang (2009) and Callen et al. (2013) also document that the complexity of financial statements negatively affects the information environment, and that complex financial statements reduce price efficiency and increases uncertainty (Guay et al,. 2016).

Although the studies of Yurisandi and Puspitasari (2015), Iatridis (2010) and Gastón et al.

(2010) find a positive relation between IFRS adoption and disclosure quality, they involve a pre- and post IFRS period. This study only involves the post IFRS period, in which IFRS 15 replaces IAS 11 and 18. Therefore, other studies have to be included which contain a similar transition. As in the study of Aleksanyan and Danbolt (2015), they document on the transition from IAS 14R Segment Information to IFRS 8 Segment Reporting. In their study, they include that although IFRS 8 may bring reduced proprietary costs, the usefulness of segment reports to investors may decrease with this transition. Gross and Königsgruber (2012) even mention that IFRS 15 will bring an excess of regulations into the market. Therefore, it can be argued that the implementation of IFRS 15 will negatively affect disclosure quality, because the new Standard can be difficult for management to implement due to the overwhelming amount of additional disclosure requirements.

2.2 Theoretical framework

This paragraph contains an introduction to three theories which I will use to develop my

hypotheses. The first theory is the agency theory, which helps to explain why owners of

economic resources demand relevant information from the organization through financial

statements. The second theory is the signaling theory, which provides insights in the behavior

of the organization on disclosing financial statements. The third theory is the stakeholder

theory, which tries to explain why the organization will provide relevant external information

in their financial statements.

9

Agency Theory

The agency theory implies that the firm consists of a set of contracts between the owners of economic resources (principals) and managers (agents) who are in charge of using and controlling those resources (Adams, 1994). The agency theory implies that agents have more information than principals and that this information asymmetry causes agents to act on their own interests instead of those of the principals. This agency problem arises because (a) there is a misalignment of goals between the agent and the principal and (b) the principal has no indicator if the agent has behaved appropriately (Eisenhardt, 1989). If information asymmetry exists, managers will have an incentive to act on their own interests instead of those of the owners, because they have the possibility to do so (Fama, & Jensen, 1983). To address this issue, Healy and Palepu (2001) argue that due to information asymmetry and agency conflicts, there is a demand for financial reporting and disclosure. Also, as was discussed by Frankel and Li (2004), by disclosing timely and relevant information from financial reports, this will reduce information asymmetry. One of the functions of financial reporting is to align the interests and goals of the principals and the agents by providing private information of the organization to the principal (Healy, & Palepu, 2001). Financial statements is a part of financial reporting.

Therefore, it can be argued that by issuing financial statements, this will reduce agency costs as investors receive private information from the organization, compared to when this information would not have been issued. These financial statements which are disclosed by organizations have to comply with certain regulations, such as IFRS 15. The goal of IFRS 15 is to ensure more consistent revenue recognition, better comparability between organizations and to provide more information (Roozen & Pronk, 2018). From an agency perspective, if information asymmetry between the principals and agents exists, providing more information to the principal could alleviate information asymmetry as more information enables the principal to monitor the behavior of the agent more effectively. Also, voluntarily disclosing information of the organization 1) improves the relationship between the agent and principal (An et al., 2011) and 2) reduces information asymmetry (Rezaee & Tuo, 2017). Therefore, voluntarily disclosing additional information on IFRS 15 can be explained because it reduces information asymmetry and improves the relationship between the agent and principal.

Signaling Theory

Another theory which is also concerned with information asymmetry is signaling theory.

Signaling theory helps describing the behavior of two parties (individuals or organizations) when both parties have access to different information (Connelly et al., 2011). The sender of the information has to choose how to signal the information to the receiver and the receiver, in turn, has to choose how to interpret the signal (Connelly et al., 2011). To illustrate the use of signaling theory, Kirmani and Rao (2000) provide an example of a basic signaling model. The authors use two entities: high-quality and low-quality firms. For high-quality firms, if strategy A (signaling) has a higher payoff then strategy B (nonsignaling), they will signal. In contrast, low-quality firms in which strategy C (signaling) is lower than strategy D (nonsignaling), they will not signal their information to outsiders. In this scenario, high-quality firms are inclined to signal, whereas low-quality firms are inclined not to signal. This scenario results in a separating equilibrium, in which firms select the most profitable strategy and it also indicates for outsiders that a signaling firm is of high-quality (Kirmani, & Rao, 2000). However, if A > B and C > D, both firms benefit from signaling, whereas outsiders cannot distinguish between high and low- quality firms, which results in a pooling equilibrium. Whether to signal or not to signal, both choices affect the level of information asymmetry as one could say that signaling reduces and nonsignaling increases information asymmetry.

There have been several examples developed to demonstrate these general relationships

(Connelly et al., 2011). According to Ross (1977) and Bhattacharya (1979), firm debt and

10

dividends represent signals of the quality of the firm, respectively. Firms who are able to pay their interests and dividends are considered high-quality firms. Conversely, firms who cannot afford such payments are considered low-quality. These examples could potentially be used to investigate whether or not firms with higher levels of debt have higher disclosure quality as, according to the signaling theory, these firms will signal their own quality to outsiders.

Stakeholder Theory

The stakeholder theory is one of the theories that try to explain the reasons for organizations to disclose information externally. Freeman (1984, p. 25) defines a stakeholder as ‘any group or individual who can affect or is affected by the achievement of an organization’s objectives’.

These groups usually include employees, shareholders, creditors, customers, suppliers, governmental bodies, public interest groups and the community (Chiu & Wang, 2015). An et al. (2011) refer in their research to accountability, in which organizations are responsible towards their stakeholders to help them making the right decisions and to protect their rights.

Furthermore, since stakeholders control the resources of a firm, they also have a certain influence on the organization (Maltby, 1997), whereas she argues further that organizations who acknowledge their stakeholders more, perform better. Therefore, it can be reasoned that an organization should take into account the interests of all its stakeholders (Freeman, 1984).

2.3 Hypotheses development

In this section I will introduce three corporate characteristics which I will use in order to develop my hypotheses and to answer the following research question:

What is the effect of leverage, profitability and FCF on IFRS 15 disclosure quality?

The first and second corporate characteristics includes leverage and profitability. Based on previous studies, leverage (Hossain et al., 1995) and profitability (Wallace & Naser, 1995) have been used together with the level of disclosure and are therefore potentially relevant for this study. The last characteristic, FCF, has not been used frequently together with the level of disclosure. However, Zhang (2009) mentioned that FCF could create a conflict of interest between shareholders and managers if managers were to misuse FCF for their private benefits.

Based on the agency theory, FCF could impact disclosure quality as managers have no interest to disclose information regarding FCF. Therefore, FCF is relevant to include as a corporate characteristic in this study.

Leverage

Leverage is a measure which provides insight in the capital structure of the firm. If a firm is more leveraged, this indicates that the firm has more debt than equity in their capital structure.

The amount of debt in one’s capital structure is important, because a high amount of debt indicates that an organization is not generating sufficient money to pay its debt obligations.

Leverage has been used in many past research studies together with disclosure level. Some studies do recognize a positive association between leverage and disclosure level. One reason which explains the positive association is mentioned by Jensen & Meckling (1976), who state that more leveraged firms incur more monitoring costs, which is an incentive for these firms to reduce the monitoring costs by disclosing more information. Also, as was mentioned in the research of Meek et al. (1995), voluntary disclosure is expected to increase with leverage.

Hossain et al. (1994) found a positive association between leverage and disclosure level.

However, other studies do not find a positive association (Ahmed & Nicholls, 1994;

Raffournier, 1995). The research of Iatridis (2008) goes in on the relationship between

accounting disclosure and firms’ financial attributes. In his research, he found that firms with

extensive accounting disclosures tend to have more debt than equity to finance their operations

(Iatridis, 2008). Furthermore, firms who have more debt than equity in their capital structure

11

tend to have higher monitoring costs which, according to agency theory, indicates there is an agency conflict between the owners of the firm and the managers.

Based on the research of Meek et al. (1995) who indicate that leverage has a positive effect on voluntary disclosure and on the research of Iatridis (2008), who indicate that more debt in the capital structure of an organization will increase accounting disclosures, I expect that leverage will positively affect the disclosure quality of IFRS 15 compared to organization which are less leveraged. Hence, based on what is reasoned above, I formulated the following hypothesis:

H1: Leveraged organizations have higher disclosure quality compared to organizations which are less leveraged under IFRS 15.

Profitability

Profitability indicates the performance of the firm. Based on the stakeholder theory, profitability is one obvious corporate characteristic which stakeholders use to determine the performance of an organization. Therefore, profitability is an important measure for organizations. The profitability of a firm has also been used in many past studies together with disclosure level.

Wallace et al. (1994) and Wallace and Naser (1995) argue that a profitable firm is more inclined to disclose more information in the annual report. In the study of Wallace et al. (1994), there was a significant positive relationship between profitability and disclosure level. However, McNally et al. (1982) did not find such relationship. In the study of Wallace & Naser (1995), there even was a significant negative relationship between profitability and disclosure level. In theory, the organization is more inclined to disclose more information to the outside when it is performing well than when it is performing poorly (Wallace et al., 1994). Consecutively, signaling theory indicates that firms with high profitability levels are inclined to report more information in their annual report to distinguish themselves from other organizations by marking their superior performance (Wallace et al., 1994; Archambault & Archambault, 2003).

Campbell et al. (2001) extends this reasoning by indicating that voluntary disclosure is a signal, where organizations will disclose more information compared to the mandatory ones in order to signal that they are better (Shehata, 2014). Verrecchia (1983) mentioned that as a result of information asymmetry, organizations signal certain information to investors to indicate that they are performing better compared to other companies to attract investments and to enhance a more favorable position. Therefore, it can argued that more profitable organizations are more eager to share their good performance to the outside, as this distinguishes more profitable organizations from less profitable ones. Based on this reasoning and on signaling theory, I expect that firms which are more profitable will have higher levels of disclosure quality, because they are more inclined to disclose more information to the public if they are performing well and to mark their superior performance compared to other organizations. Therefore, I formulated the following hypotheses:

H2: Profitable organizations have higher disclosure quality compared to organizations which are less profitable under IFRS 15.

Free Cash Flow

FCF regards the last part of this research. According to Jensen (1986), FCF is the excess cash

flow over what is required to fund all projects with a positive net present value. FCF can affect

firm value by either increasing or decreasing it, depending on the use of FCF (McCabe & Yook,

1997). FCF is therefore important financial information, as, based on stakeholder theory,

misusing FCF could lead to a decrease in firm value, which is not in line with the interests of

stakeholders. The value of the firm contains of many components, such as Property, Plant and

Equipment (PPE), brands and cash. Of these items, cash is the only thing of which managers

can freely use. However, not all cash can be ‘freely’ used (Yeo, 2018). The part of which

12

managers can use is FCF. Effective asset utilization or investment will result in an increase of firm value, whereas a decrease in firm value is the result of ineffective asset utilization (Yeo, 2018). Jensen (1986) states that FCF can incentivize managers to use FCF for various activities which do not necessarily result to an increase in firm value. FCF can also result in a situation where managers retain cash in the firm and do not distribute this cash as dividends (Jensen, 1986). Based on the agency theory, managers may invest in nonvalue-maximizing investment decisions with FCF by investing in projects which have a negative net present value (Jensen, 1986). This situation can be described as an overinvestment. Together with the FCF agency problem, this creates the conflict of interest between shareholders and managers (Zhang, 2009).

According to Joseph and Richardson (2002), free cash flow tempts managers to expand the size of the firm, thereby increasing managers’ control and personal remuneration even though such action may result in a decrease of overall firm value as growth is typically associated with more managerial power over the firm by increasing the resources under their control (Jensen, 1986).

The signaling hypothesis states that if owners and managers of the firm have different sets of information available, an organizations dividend policy may provide signals regarding the organizations current performance (Fairchild, 2010). If a manager has excess FCF to spend on negative net present value projects, this tends to cause the agency problem (Jensen, 1986).

However, dividends tends to alleviate this problem by reducing the FCF available to managers (Fairchild, 2010). Furthermore, as mentioned in the research of Bhattacharya (1979), dividends represent a signal of firm quality. If organizations are able to afford their dividend payments, they are considered as high-quality firms whereas low-quality firms cannot afford such payments. As we have seen, dividends tend to reduce the FCF available to managers, which eventually reduces the agency conflict. To conclude, I expect that firms with high levels of FCF will disclose lower levels of information in their financial statements to the outside as compared to organizations which have lower levels of FCF, as organizations with a high level of FCF tend to have an agency conflict between the owners and managers of the firm, whereas voluntary disclosure reduces information asymmetry (Rezaee & Tuo, 2017). Hence, the following hypothesis is formulated as such:

H3: Organizations which have higher levels of free cash flow have lower disclosure quality compared to organizations which have lower levels of free cash flow under IFRS 15.

3. Research Methodology

This section of the research will contain the methodology to perform this research. The first part contains information about the selected sample and its size. In the second and third section, I will explain how to measure the dependent and independent variables, respectively. Fourth, the control variables are included in order to limit the effect of these variables on the statistical analyses. The fifth section contains some data adjustments, whereas the last section includes the statistical model.

3.1 Sample

This research will consist of an analysis of financial statements from the financial year of 2018 of listed European firms who mandatorily have to apply IFRS 15 on January 1

2018. The data collection period starts from spring 2019 and oversees the financial statements of 2018. The total amount of financial statements analyzed consist of 52. At first, the goal was to compare the financial statements of one industry. However, due to the low amount of financial statements available to compare, three other industries are included. The four selected super sectors are:

Telecommunications, Technology, Construction & Materials and Utilities. All financial

statements are collected from the Stoxx 600. The Stoxx 600 is well suited to use for this

research, as it contains data from 600 listed European firms spread over 17 different European

13

countries. Furthermore, every firm in the Stoxx 600 was required to issue their financial statements which makes selecting suitable firms effortless. Next, from the Stoxx 600, organizations which had revenue recognition as a Key Audit Matter (KAM) in their financial report were selected to investigate as KAM require significant auditor attention. This ensures that IFRS 15 will have an impact on all the sampled organizations. Table 1 consists of the composition of the studied population in total and in percentages by super sector. Appendix III consists of the analyzed firms by country, by industry and the disclosure score in percentages.

Table 1: Composition research population by super sector

Super Sector N(#) N(%)

Construction &

Materials 8 15%

Technology 18 35%

Telecommunications 15 29%

Utilities 11 21%

Total 52 100%

In the next paragraphs, I will discuss the variables that will be used in this research.

3.2 Dependent variable

In order to determine disclosure quality, a disclosure index has been constructed. This disclosure index has been constructed in part of a research group together with three other researchers. Several workshops and meetings with other researchers were held to increase the reliability of the results by reviewing various annual reports and ultimately determining why an organization will or will not receive a point for mandatory and/or voluntary disclosure aspects.

The basis of the disclosure index are the disclosure requirements of IFRS 15. However, the disclosure index also includes the voluntary disclosure aspects of IFRS 15, since the difference between these two disclosures will determine disclosure quality. Through these workshops and meetings, it became clear that some organizations prefer to fulfill disclosure requirements by presenting tables in their financial statements. Therefore, through mutual consultation, we concluded that organizations can receive points for either 1) quantitative information which includes tables with relevant information or 2) qualitative information which includes plain text. The disclosure index has been included in Appendix II and particularly aims at the following items: contracts with customers, significant judgements made in applying IFRS 15 and assets recognized from the costs to obtain or fulfil a contract.

There are a number of advantages and disadvantages of using content analysis as a research method. An advantage is that by analyzing financial statements based on the disclosure index, every organization will get judged equally. The analyzed results can thus be used to compare firm scores using statistical analyses and eventually draw conclusions based on these results.

By means of the disclosure index, every organization has been given a score expressed in

percentages. However, a disadvantage of using this method is the subjectivity risk. Judgements

have to be made by every researcher in order to determine whether or not an organization will

receive a point for a disclosure requirement. Different judgements could be made if an

organization were to receive a point. Therefore, to limit the potential subjectivity of this

research method, all researchers performed a control analysis. This control analysis includes a

comparison of two analyzed financial statements for every researcher. The results of the control

analysis brought forth some differences in them. These differences were resolved through a

pilot, several workshops and meetings with other researchers.

14

The disclosure requirements are included in IFRS 15.110 et sequel. From 15.110, organizations can receive points by meeting the disclosure requirements. However, as mentioned before, organizations can either 1) disclose sufficient information just to meet the mandatory disclosure requirement at a minimum level or 2) provide additional information regarding this disclosure requirement. Therefore, it is expected that starting from IFRS requirement 15.110, differences will arise between disclosure levels.

Besides the analysis of the financial statements, certain corporate characteristics were collected in order to determine what the effects of these characteristics are on disclosure quality under IFRS 15. In order to perform the statistical analyses, the items of the disclosure index have been quantified. This quantification represents the disclosure quality. As not all organizations can receive equal points due to the not applicableness, the included organization will not receive any points for this. However, the value of this item will not be included in the total number of points. Every researcher had to analyze an equal amount of financial statements in order to determine the disclosure quality and corporate characteristics per company. Once a company has received points for the applicable disclosure requirements, the total of these points were converted to percentages. A higher percentage indicates that the firm has higher disclosure quality and lower vice versa.

3.3 Independent variables

In this section, I will explain the three independent variables which include leverage, profitability and FCF.

Leverage

𝐿𝐸𝑉 = 𝑇𝐿 𝑇𝐴

LEV = Leverage

TL = Total Liabilities

TA = Total Assets

Leverage indicates how much debt a firm has in its capital structure. Leverage is calculated quantitively, whereas this number can be between 0 and 1. If the value is closer to 1, this indicates that the firm has more debt in its capital structure and if the value is closer to 0, the firm has less debt in its capital structure. As indicated earlier in this research, debt can play a major part in a firm’s capital structure. Therefore, the usefulness of financial information increases with more debt, because it assists debt holders to monitor management behavior (Chung & Jung, 2016). This makes leverage very suitable to investigate for this research.

Profitability

𝑅𝑂𝐸 = 𝑁𝑃 𝑇𝐸

ROE = Return On Equity

NP = Net Profit

TE = Total Equity

Return On Equity (ROE) is in this study a measure for profitability as it is a popular ratio for

measuring profitability (Ahmed & Courtis, 1999). ROE indicates how much profit an

organization generates compared to its total equity. Net profit is an accounting number from

the income statement which indicates how much money an organization is generating, less any

costs of goods sold, operating and/or other expenses, interest and taxes. Total equity is an

accounting number which can be found in the balance sheet and represents the total value of

15

assets less liabilities. ROE is a well-suited measure for this research, as ROE compares the profit of an organization to the capital of shareholders.

Free Cash Flow

𝐹𝐶𝐹 = 𝑁𝐶𝐹𝑂𝐴 + 𝐼 − 𝑇𝑆𝐼 − 𝐶𝐴𝑃𝐸𝑋

FCF = Free Cash Flow

NCFOA = Net Cash Flow from Operating Activities

I = Interest expense

TSI = Tax Shield on Interest expense

CAPEX = Capital Expenditures

FCF provides a basic tool for the valuation of the firm. Investors find FCF a useful indicator, as FCF helps them to estimate the value of the firm and its individual projects (Yaari et al., 2016). Net Cash Flow from Operating Activities is an accounting number which indicates the amount of money an organization generates from its operating activities. One could also calculate FCF by using Earnings Before Interest and Taxes (EBIT), depreciation and changes in working capital as well. However, using Net Cash Flow from Operating Activities saves time as this number will already have adjusted earnings. Interest expenses less the tax shield on interest expenses are added to this number, because this money is readily available for payments to the debt holders. Finally, capital expenditures are deducted from the total number as this money is required to maintain, upgrade and acquire assets such as Property, Plant and Equipment (PP&E).

3.4 Control variables Firm size

Karim, Pinsker and Robin (2013) mention that firm size may play a part in disclosure, ‘because large capital providers (e.g. banks) may encourage more systematic and timely disclosure; the larger the firm, the greater the possibility that large capital providers are involved’ (Karim et al., 2013: 868). In order to measure firm size, the natural logarithm of total assets will be used and is expected to have a positive effect on disclosure quality.

Industry type

Another control variable for which has been shown that it could impact disclosure is industry type, as the relevance of disclosure items could vary across industries (Meek et al., 1995).

Stanga (1976) also described that the industry variable plays a relatively important role in describing the differences in the extent of annual report disclosure among large industrial firms.

Roozen and Pronk (2018) included the expected material impact of IFRS 15 on different industries. They specifically mention that telecommunications and utilities are most likely to be affected by IFRS 15 and are therefore included in this study. The last two industries include utilities and construction & materials, as these industries had the highest percentage (after telecommunications and utilities) of expected material impact of IFRS 15. Industry type will be used as a dummy variable and divided into four different dummy variables in this study, whereas each dummy variable represents one industry.

3.5 Data adjustments

As not all firms use the same currency in their financial statements Appendix I consists of the

used exchange rates. Due to the fact that all financial statements end at 31 December 2018, I

will use the following exchange rates: Every balance sheet item is exchanged from their

respective currency to euros using the closing rate at 31 December 2018, and every profit- and-

loss item is exchanged using the average year rate. Two additional adjustments were made to

16

one independent variable (leverage) and one control variable (firm size). Since the unstandardized coefficient of these two variables where considerably too high, the natural logarithm for both the variables is computed and used in the regression analysis. Lastly, an adjustment has been made to the natural logarithm of FCF. The results indicate that some FCF’s were considerably negative and since one cannot take the natural logarithm of a negative number, the following formula has been applied to the natural logarithm of FCF: LOG(FCF+

maximum negative value+1). By adding a constant value which is the maximum negative value and adding one extra value, the natural logarithm of FCF will always be positive.

3.6 Statistical model

In order to test the different hypotheses, a linear regression model has been used. The model consists of the independent variables (leverage, profitability and FCF) and the control variables (firm size and industry type). The dependent variable in the model is disclosure quality score of IFRS 15. The statistical model for including all variables is as follows:

Score disclosure quality IFRS 15= α

+ β

*LEV+ β

*ROE+ β

*LOGFCF+ β

*LOGSIZE+

β

*UT+ β

*CM+ β

*TC+ β

*TE

Where α and β are parameters, LEV is leverage, ROE is return on equity, LOGFCF is the natural logarithm of FCF, LOGSIZE is the natural logarithm of firm size, UT is the utilities industry, CM is the construction & materials industry, TC is the telecommunications industry and TE is the technology industry.

4 Results

This chapter is divided into three paragraphs. The first paragraph contains the descriptive statistics which indicates, among other things, the means of the variables. The second paragraph is dedicated to the validity and reliability which is measured through the adjusted R-squared model and Cronbach’s alpha, respectively. The third paragraph contains the correlation matrix, which indicates whether variables are positively or negatively correlated and if variables have multicollinearity. The fourth and last paragraph contains the regression model which provides the basis for interpreting the outcomes.

4.1 Descriptive statistics

Table 2 consists of the descriptive statistics. The descriptive statistics indicate the sample size (N), minimum, maximum, mean and standard deviation. The most important statistic from table 2 is SCORE. SCORE represents the score in percentages of all firms. For example, a value of 0,60 indicates that an organization scored 60% out of 100%. The lowest score was 0,13 whereas the highest score was 0,76, which indicates a large gap between the disclosure quality of organizations. Furthermore, the average showed a value of 0,48. This means that, on average, organizations did not score above 50%, which can be interpreted as low. Reasoning behind this could be that organizations cannot compare their financial statements with others, since 2018 is the first year IFRS 15 is effective. Therefore, they have no indication whether or not their financial statements from 2018 are either good or bad compared to others. In other words, there are no ‘best practices’ available at the moment for organizations to guide them through disclosing their financial statements.

The minimum of LEV is 0,11 and the maximum is 0,84. This indicates that there is a

considerable difference in capital structures between organizations, where the organization with

a leverage of 0,11 has a low amount of debt in its capital structure and the organization with

0,84 has a high amount of debt in its capital structure. The average of LEV is 0,6123, which

indicates that on average, organizations have more debt in their capital structure then equity.

17

The minimum of ROE is -0,17 (-17%) and the maximum is 0,58 (58%). This indicates that the profitability of organizations varies considerably. An organization with a ROE of -0,17 did not generate net profits in the year of 2018, since it is practically impossible that total equity is negative. The average shows a ROE of 0,1364 (13,64%). This indicates that on average, organizations can be considered profitable.

LOGFCF shows a minimum of ,00 and a maximum of 23,06. This indicates that FCF among organizations vary considerably. As described in data adjustments, FCF has had some adjustments in order to use FCF in the analysis, since one cannot take the logarithm of a negative value, hence the minimum is ,00.

The dummy variable Industry type is split up in four control variables. The dummy variable Telecommunications indicates 1 if a company is a Telecommunications firm and 0 otherwise.

The dummy variable Technology indicates 1 if a company is a Technology firm and 0 otherwise. The dummy variable Utilities indicates 1 if a company is a Utilities firm and 0 otherwise. The dummy variable Construction.and.Materials indicates 1 if a company is a Construction & Materials firm and 0 otherwise. As indicated by the mean of 0,35, the sample mainly consists of Technology firms, whereas the lowest amount of analyzed organizations consist of the industry Construction & Materials.

Table 2

Descriptive Statistics

N Minimum Maximum Mean Std. Deviation

SCORE 52 ,13 ,76 ,4816 ,15194

LEV 52 ,11 ,84 ,6123 ,15861

ROE 52 -,17 ,58 ,1364 ,12996

LOGFCF 52 ,00 23,06 21,9091 3,11730

LOGSIZE 52 19,41 26,37 23,3373 1,51133

Telecommunications 52 0 1 ,29 ,457

Technology 52 0 1 ,35 ,480

Utilities 52 0 1 ,21 ,412

Construction.and.Materials 52 0 1 ,15 ,364

4.2 Validity model

Table 5 consists of the regression results and also the adjusted R-squared model. The adjusted

R-squared model adjusts for the number of variables included in the model. The most important

model to interpret for validity is model five, which consists of the dependent variable and all

the independent and control variables. As indicated in model one, which only consists of the

control variables, R-squared is 0,148 and adjusted R-squared is 0,022. However, if we include

all the independent variables in the model, R-squared increases to 0,154 but adjusted R-squared

decreases to 0,020. This indicates that be adding more variables, the explanatory power of

adding extra variables to the model decreases. Therefore, by adding the independent variables

18

LEV, ROE and LOGFCF, the explanatory power of these variables on the dependent variable SCORE decreases.

Of the collection of all variables in this study, there is one variable which is bound to subjectivity. The percentage level of score is edified on how well an organization scores on the disclosure requirements of IFRS 15. An organization can receive points for every disclosure requirement from IFRS 15.110 up to and including 15.129. In order to increase the internal consistency of the items, which will eventually lead to the percentage level of score, the Cronbach’s alpha is calculated. Since some organizations have ‘Not Applicable’ as a result for disclosure requirements, the average of the specific column will be taken and replace ‘Not Applicable’ with the average. Cronbach’s alpha showed a value of α= 0,790. Values above α=0,7 are considered reliable. The underlying variables are therefore considered reliable.

4.3 Correlation matrix

Table 4 consists of the Pearson correlation matrix. The Pearson correlation matrix is used to determine whether or not variables are positively or negatively correlated and if variables have multicollinearity. According to the correlation matrix, none of the independent variables are significantly correlated with score. One control variable, Utilities, is significantly negatively correlated with score (-0,276*). Furthermore, LOGSIZE is significantly negatively correlated with the variable ROE (-0,338*). This indicates a negative correlation with total assets and profitability. Furthermore, every variable is tested for multicollinearity. The results indicate no sign of multicollinearity as most VIF-values are approximately 1.

Table 4

4.4 Regression analysis

Table 5 consists of the regression analysis. The regression analysis is used in order to determine if there is a relationship between the dependent, independent and control variable(s). In order to prevent perfect collinearity, the dummy variable Technology has not been included in the regression analysis, as it has the most observations with a mean of 0,35 indicated in table 2 and has the highest correlation with other variables. The first column represents the variable name.

The second column consists of the expected relation between the variables and the dependent variable (+) (-). Model one only includes the control variables and no independent variables

Correlations

(1) (2) (3) (4) (5) (6) (7) (8) (9)

SCORE (1) 1

LEV (2) -0,107 1

(0,451)

ROE (3) -0,114 0,129 1

(0,421) (0,363)

LOGFCF (4) 0,026 -0,091 0,111 1

(0,856) (0,519) (0,433)

LOGSIZE (5) 0,122 0,404** -0,338* -0,183 1

(0,389) (0,003) (0,014) (0,194)

Telecommunications (6) 0,141 0,040 -0,079 0,102 0,074 1

(0,319) (0,777) (0,578) (0,473) (0,600)

Technology (7) 0,08 -0,458** 0,142 0,099 -0,423** -0,463** 1

(0,571) (0,001) (0,315) (0,486) (0,002) (0,001)

Utilities (8) -0,276* 0,358** -0,011 -0,262 0,414** -0,330* -0,377** 1

(0,048) (0,009) (0,937) (0,061) (0,002) (0,017) (0,006)

Construction.and.Materials (9) 0,029 0,148 -0,076 0,038 -0,003 -0,271 -0,310* -0,221 1 (0,837) (0,295) (0,595) (0,788) (0,982) (0,052) (0,025) (0,116)

*. Correlation is significant at the 0.05 level (2-tailed).

**. Correlation is significant at the 0.01 level (2-tailed).

19

and indicates the relationship between the dependent variable and the control variables. Models two up to and including four each include the regression analysis regarding one of the independent variables with the control variables. Hence, model two indicates the relationship between SCORE and LEV. Model three indicates the relationship between SCORE and ROE.

Model four indicates the relationship between SCORE and LOGFCF. Finally, model five indicates the relationship between SCORE and all of the independent variables included together with the control variables. All models indicate whether there is a positive or negative relationship between the dependent variable and the assigned variables. Furthermore, the R- squared values are given which explain the variance of the given model on the dependent variable.

Table 5

Regression results

The first regression model analyses the effect of the control variables on the percentage level of score. The results show a significantly positively coefficient (0,031) with a P-value of (0,057) at the ten percent level for the control variable LOGSIZE. This indicates that the level of total assets has a significant positive effect on the percentage level of score, which is in line with the expected positive relation. Furthermore, the dummy variable Utilities has a significantly negatively coefficient (-0,160) with a P-value of (0,017) at the five percent level. This indicates

Regression model 1 2 3 4 5

Assumption

LEV (+) -0,082 -0,081

(0,597) (0,631)

ROE (+) -0,034 -0,004

(0,841) (0,984)

LOGFCF (-) -0,001 -0,001

(0,853) (0,867)

LOGSIZE (+) 0,031* 0,033** 0,030* 0,030* 0,032*

(0,057) (0,050) (0,087) (0,062) (0,084)

Telecommunications -0,015 -0,008 -0,016 -0,015 -0,008

(0,776) (0,882) (0,775) (0,784) (0,887)

Utilities -0,160** -0,147** -0,159** -0,162** -0,149**

(0,017) (0,041) (0,020) (0,018) (0,045)

Construction.and.Materials -0,033 -0,022 -0,033 -0,033 -0,022

(0,610) (0,750) (0,607) (0,614) (0,755)

Constant -0,191 -0,190 -0,159 -0,156 -0,155

(0,593) (0,597) (0,685) (0,701) (0,726)

F-Test 2,048 1,670 1,613 1,612 1,146

(0,103) (0,161) (0,176) (0,176) (0,353)

R² 0,148 0,154 0,149 0,149 0,154

Adjusted R² 0,022 0,062 0,057 0,057 0,020

Significant at the 0,10 (*) and 0,05 (**) level (2-tailed).

20

that an organization which operates in the Utilities industry has a significant negative effect on the percentage level of score.

The second model analyses the effect of leverage on the percentage level of score and also tests hypothesis 1. Results indicate a negative coefficient (-0,082) with a P-value of (0,597) for the variable LEV. This result implies that leverage has a negative effect on the percentage level of score, which is contradictory to the expected assumption. Although leverage has a negative effect on the percentage level of score, the P-value indicates no significant relation at the 10%

level. Given that leverage has a negative relationship with the percentage level of score, I reject hypothesis 1 which claimed a positive relationship.

The third model analyses the effect of profitability on the percentage level of score. As indicated, the coefficient of ROE (-0,034) is negative with a P-value of (0,841) which implies that profitability has a negative effect on the percentage level of score. This relation, however, is not significant. Given that profitability has a negative relationship with the percentage level of score, I reject hypothesis 2 which claimed a positive relationship.

The fourth model analyses the effect of FCF on the percentage level of score. The results indicate a negative coefficient of (-0,001) and a P-value of (0,853). This implies that FCF has a negative effect on the percentage level of score. However, this relation is not significant.

Given the negative relationship between FCF and percentage level of score which is not statistically significant, hypothesis 3 is not supported.

In summary, no significant results are found for all hypotheses. Hypothesis 1, 2 and 3 are all

not supported. This implies that more leveraged and profitable organizations score at a lower

rate and firms with high levels of FCF score at a higher rate with respect to the IFRS 15

disclosure requirements. As was indicated in most studies, firm size has a positive effect on

disclosure quality, which was also the result in this study. Furthermore, as indicated by the

significant negative coefficient (-0,160) with a P-value of (0,017), organizations who operate

in the Utilities industry score at a lower rate compared to organizations who operate in other

industries.