Master Thesis June 22th, 2020
MSc A&C University of Groningen Faculty of Economics and business
ABSTRACT: In this research the relationship between national culture and IFRS 15 disclosure quality of EU-listed construction firms is examined. Both the direct and indirect influence, through investor protection and capital market development, of national culture on IFRS 15 disclosure quality are assessed. In this research data of a self-constructed disclosure index of 37 firms from 10 different EU-countries is included for the years 2018 and 2019. This research is unique using all 6 national cultural dimensions of Hofstede in measuring national culture. It is found that national culture does influence IFRS 15 disclosure quality, as expected. Furthermore, it is found that the level of investor protection positively influences IFRS 15 disclosure quality. No significant indirect effect of national culture through investor protection and capital market development is found. Furthermore, best practices are identified. This research contributes to existing literature of disclosure quality and provides insights to practitioners and regulators regarding the quality of IFRS 15 disclosures in a cross-country setting.
IFRS 15 Disclosure Quality and
Comparability: The Influence of Culture in
a European Union Cross-Country Analysis
Riemer Akkerman
University of Groningen
Keywords: IFRS 15; Revenue Recognition; Disclosure Quality; Voluntary Disclosure;
Agency Theory; Signaling Theory; Cultural Influence Theory; Moral Hazard; Adverse Selection
Data availability: Data available upon request. Please contact r.akkerman.3@student.rug.nl. Word count: 11996
2
TABLE OF CONTENTS
i. INTRODUCTION ... 3
ii. THEORATICAL FRAMEWORK ... 6
IFRS 15 Content ... 6
Theories ... 7
Agency theory ... 7
Signaling theory ... 7
Theory of cultural influence ... 8
Disclosure quality determinants and hypothesis development ... 8
National culture ... 8
Investor protection ... 15
Capital market development ... 17
iii. METHODOLOGY ... 18 Sample selection ... 18 Dependent variable ... 19 Independent variables ... 20 Explanatory variables ... 20 Control variables ... 22 Model of analysis ... 23 Variable overview ... 24 iv. RESULTS ... 25 Descriptive statistics ... 25
Best practices and learning effects ... 27
Correlation analysis ... 28 Regression analysis ... 31 Robustness ... 36 v. DISCUSSION ... 37 vi. CONCLUSION ... 38 vii. REFERENCES ... 40 viii. APPENDICES ... 47
3
i.
INTRODUCTION
he IASB issued IFRS 15 “Revenue from contracts with customers” in 2014 which became mandatory applicable for listed firms in the EU from the 1st of January 2018, therefore, EU-listed firms are required to apply IFRS 15 onwards from financial year 2018. The European Securities and Markets Authority (ESMA) recognized the efforts issuers have taken by implementing IFRS 15, however the ESMA believes that further improvement is necessary to improve disclosure quality, especially for industries where significant judgment is necessary to achieve the goal of IFRS 15: improve the disclosures regarding revenues and increase
comparability of the financial statements globally (ESMA, 2019). An industry that is interesting
regarding IFRS 15 is the construction sector, due to complex long-term contracts with their customers. Since its issuance the implications of IFRS 15 for firms’ accounting practices was subject of publications (Grant Thornton, 2017; PwC, 2017; Pronk & Roozen, 2018; Deloitte, 2019; KPMG, 2019; EY, 2019) since revenues are an important measurement of financial performance and prospects (Wagenhofer, 2014; Mattei & Paoloni, 2019; ESMA, 2019). In a public statement the ESMA stresses the importance of consistent and high-quality implementation of IFRS 15 to be useful for users of annual reports (ESMA, 2016). Nevertheless, in the application of IFRS 15 a significant amount of judgment is present. Therefore IFRS 15 requires firms to disclose significant judgments to provide useful information about the nature, amount, timing and uncertainty of revenue and cash flows (IFRS15.110-15.110b). In this research the current state of IFRS 15 disclosure quality and comparability will be assessed by using a unique self-constructed disclosure index to measure the quality of IFRS 15 disclosures in a cross-country setting.
As mentioned earlier, the construction industry is interesting regarding IFRS 15 disclosures due to presence of complex, multiple elemental, long-term contracts between them and their clients (KPMG, 2014). It is not surprising that audit/consultancy firms assessed implications of IFRS 15 application for disclosure practices for construction firms (KPMG, 2014; NavarroAmper&Co., 2014; EY, 2015; PwC, 2017; BDO, 2018). IFRS 15 superseded IAS 11 and IAS 18 for which in particular IAS 11 “construction contracts” has been of relevance for construction companies (PwC, 2017)..
The importance of construction firms for society in the construction of houses, hospitals, schools and infrastructure is evident. The construction firms were severely hit by the global financial crises of 2008. In the Netherlands for example, construction business decreased by 30% and led, among others, to a shortage of affordable and sustainable houses as of today
T
4 (Financieele Dagblad, 2020). From a European perspective, the construction sector generated over €1.500.000 million of revenues and employed roughly 13 million people in the EU in 20161 (European Commission, 2016). The importance of high-quality disclosure can be further illustrated by looking at the impact a construction firm bankruptcy can have on a country. UK construction firm Carillion collapsed after, among other things, overestimating revenues. Costing UK taxpayers around £148m and the loss of around 2300 jobs (BBC, 2018).
As the goal of IFRS 15 is to improve disclosure quality and comparability globally (IFRS Foundation, 2014) one would expect convergence of disclosure quality among listed construction firms from different countries. Nonetheless, disclosure quality is not only determined by accounting standards but also by many other determinants (Soderstrom & Sun, 2007; Hooghiemstra, et al., 2015, Akman & Al-Attar, 2017). Which suggests that there will remain differences in IFRS 15 disclosure quality, for example regarding earlier mentioned significant judgments in IFRS 15. In literature a variety of determinants of disclosure quality can be found, among others, culture, religion, national systems (political and economic) and company specific factors (Archambault & Archambault, 2003). In this research the focus is on the countries’ specific cultural values and national institutions which influence disclosure. The following three country specific determinants of disclosure quality are assessed that are most relevant for this research.
First, multiple papers show the direct influence of national culture on reporting practices (Gray, 1988; Asiyaban & Abdoli, 2012; Hooghiemstra, et al., 2015; Houqe, et al., 2016). Aggerwal & Goodell (2014) stress the importance of research to national culture in accounting studies since this is not done enough in top accounting journals they argue. According to Archambault & Archambault (2003) accounting is a social system which exists along with other systems in a country (e.g. political, economic) which are all influenced by national culture. Aggerwal & Goodell (2014) underline this by stating that national culture is the basic force that influences national institutions which in turn also influence accounting practices, besides the direct influence of national culture. In this research the national institutions investor protection and capital market development are considered as these are found to be associated with disclosures in prior research (La Porta, et al., 2008; Glaum, et al., 2013; Hooghiemstra et al., 2015; Dayanandan, et al., 2016;).
5 The determinants of disclosure quality can be explained using agency theory, signaling theory and cultural relevancy theory. Agency and signaling theory are economics-based theories where agency theory primary addresses moral hazard and signaling theory adverse
selection (Cotter, et al., 2011). Both theories deal with the issue of information asymmetry
problems, which can be useful in explaining differences in disclosure quality the incentives construction firms face (Cotter, et al., 2011). The theory of cultural influence is a widely used theory in the process of forming expectations about the influence of national culture on disclosure quality and the underlying mechanisms (Maali & Al-Attar, 2017).
In conclusion, it is expected that even with the same reporting standards, IFRS 15 disclosure quality differs between firms from different countries. This is accordance with earlier research on IFRS disclosures (De George et al., 2018; Dayanandan, et al., 2016; Soderstrom & Sun, 2007). The aim with this research is to gain insight into the current state of the quality and comparability of IFRS 15 disclosures across the EU and if cultural and institutional factors influence the IFRS 15 disclosure quality. So, the research question will be:
Do national culture, investor protection and capital markets influence the quality of IFRS 15 disclosures of listed European Union construction firms?
This research will contribute to existing literature by giving insight in the quality of IFRS 15 disclosures in the EU for the years 2018-2019 and influence of national culture and institutions on IFRS 15 disclosure quality. When expected national effects are found this can give valuable insights to regulators, accounting bodies, governments and other practitioners on IFRS disclosures. Furthermore, this research contributes to prior literature by including all 6 national cultural dimensions of Hofstede together with institutional factors in relation with IFRS 15 disclosure quality. Finally, best practices will be identified which could serve as example for other firms. This paper will be structured as follows: first the theoretical framework and methodology will be addressed. Followed by the results, discussion and conclusion.
6
ii.
THEORATICAL FRAMEWORK
This section consists of the following components: IFRS 15 content, theories, disclosure quality determinants and hypothesis development.
IFRS 15 Content
EU-listed firms must prepare annual reports in accordance with EU-endorsed IFRS (EC, 2002). The IASB and FASB issued IFRS 15 “Revenue from contracts with customers” in 2014 which became mandatory applicable for listed firms in the EU from the 1st of January 2018. The goal of IFRS 15 is to improve the disclosures regarding revenues and increase
comparability of the financial statements globally (IFRS Foundation, 2014). IFRS 15 replaced
IAS 11; IAS 18; IFRIC 13; IFRIC 15; IFRIC 18; and SIC 31 (IFRS 15, 2018).
The basic principle of IFRS 15 is that firms recognize revenue to the extent that they expect they are entitled to in exchange for their products or services to a customer (Pronk & Roozen, 2018). To execute this principle firms must follow a five-step approach2 (IFRS 15, 2018). As shown in appendix 1, in the process of recognizing revenues, a significant amount of judgement is present. This contradicts with the goal of IFRS 15 that wants to increase
comparability of revenues globally (IFRS Foundation, 2014). IFRS15.110 et seq. provides
disclosure requirements for disclosing information and those judgments.
Next to mandatory disclosure requirements, firms have the freedom to disclose more than is strictly necessary. This would result in voluntary disclosed information and could increase the quality of disclosures (Francis, et al., 2008). In this research the focus will be on voluntary disclosures since all annual reports are audited. Nevertheless, in our analysis also mandatory requirements will be considered because of the varying degree of compliance3.
Disclosure quality plays an important role in the field of accounting and business research. A wide range of literature is published regarding the quality of corporate disclosures (Healy & Palepu, 2001). A similar accounting standard is just one of the determinants of influence on disclosure quality (Holthausen, 2009). In the following paragraph the main theories will be addressed.
2 1) Identifying the contract 2) Identifying performance obligations 3) Determining the transaction price 4) Allocating the transaction price to performance obligations 5) Recognize revenue when performance obligation is fulfilled.
7 Theories
We discuss the main theories in this research: agency theory, signaling theory and cultural influence theory.
Agency theory
Since its publication by Jensen & Meckling (1976) agency theory has provided theorical substance in many lines of research. In agency theory a contract is present between the agents and the principals wherein the agents perform services on behalf of the principals. The agency theory assumes agents and principals have different interests due to the separation of ownership and control wherein the presence of information asymmetry enlarges this agency problem (Jensen & Meckling, 1976; Fama & Jensen, 1983). Information asymmetry is caused by the possession of more knowledge about the firm by managers (agents) than the owners/debt holders (principals). This can result in adverse selection and moral hazard problems. By aligning the interests of the managers and the owners these problems can be mitigated. Also, monitoring voluntary disclosures by managers, which indicates focus on moral hazard, is a way to reduce information asymmetry (Cotter, et al., 2011). Voluntary disclosure can mitigate the agency problem because voluntary disclosure reduces agency costs and convinces principals that agents are acting in the best interest for the firm (Shehata, 2014).
Agency theory can help explain that stronger incentives for managers to reduce information asymmetry will result in higher disclosure quality (Healy & Palepu, 2001). This suggests that when incentives to reduce information asymmetry in country X are higher than in country Y also the level of voluntary disclosure could be higher in country X. For this research IFRS 15 voluntary disclosures can be seen as the way of management to reduce the information asymmetry. The focus will be on voluntary disclosure since the dependent variable ‘IFRS 15 disclosure quality’ is derived from audited annual reports which should in general be compliant with mandatory disclosure requirements.
Signaling theory
Similar to agency theory, signaling theory deals with information asymmetry problems (Cotter, et al., 2011). Signaling theory primarily addresses adverse selection issues. Signaling theory states that information asymmetry can be reduced by disclosing more information which is difficult to mimic by other firms. In this way an organization can signal its firm quality to stakeholders. An often-used channel to communicate this information is by voluntary disclosures. If a construction company performs well, they could signal this performance to the
8 market by providing additional revenue disclosures for example. This is relevant for users of annuals reports in assessing current performance and form expectations about the future performance. By signaling their excellence to the market high-quality firms improve their disclosure quality. At the same time signaling theory research found that low-quality firms, to some degree, ‘pre-empt’ bad news to avoid reputation and litigation costs (Skinner, 1994). This could be the case when a construction firm faces significant setbacks at a project and could therefore choose to disclose addition information to the market to avoid reputation and litigation costs. Signaling theory can be used to explain why firms in certain countries face stronger incentives to show excellence or to prevent litigation and reputation costs.
Theory of cultural influence
Hofstede introduced four cultural dimensions in 1980 upon which cultures can be characterized. Two dimensions have been added to the model later4. Gray (1988) extended Hofstede’s model by adding the presence of an accounting subsystem which is influenced by the primary cultural values, the theory of cultural influence (Salter & Niswander, 1995). The theory of cultural influence consists of four accounting values which are linked to Hofstede’s cultural dimensions5 (Salter & Niswander, 1995). The most relevant accounting value for our research is secrecy. The level of secrecy in a country is directly related to the amount of disclosure (Gray, 1988). When the level of secrecy is high people are prone to keep more information to themselves instead of disclosing information. The other accounting values could also indirectly influence the amount of disclosure. Because of the direct relation between secrecy and amount of disclosure, Gray’s predictions (1988) regarding the accounting value secrecy are used in hypothesis development to formulate expectations about influence of national culture on IFRS 15 disclosures and the underlying mechanisms.
Disclosure quality determinants and hypothesis development
In this section the determinants of IFRS 15 disclosure quality and relating hypothesis development is addressed.
National culture
Multiple papers show the influence of national culture on accounting practices (Gray, 1988; Archambault & Archambault, 2003; Soderstrom & Sun, 2007; Asiyaban & Abdoli, 2012; Glaum et al., 2012; Hooghiemstra, et al., 2015; Houqe, et al., 2016) which is relevant for this
4 See appendix 2 for a complete description of all six cultural dimensions of Hofstede. 5 Professionalism, uniformity, conservatism and secrecy.
9 research since national cultures within the EU differ significantly (see table 4). Common way to identify cultural differences is by using Hofstede’s (1991) cultural dimensions. Hofstede’s cultural model consist of six dimensions: power-distance index; uncertainty-avoidance index; masculinity vs. femininity; individualism vs. collectivism; short-term vs. long-term orientation6; Indulgence vs. restraint7. By providing dimension scores per country Hofstede provides insights in the cultural characteristics of countries. There has been criticism on Hofstede’s model. Nevertheless, researchers widely used Hofstede’s model on the influence of culture on management behavior and diversity of financial systems (Hooghiemstra, et al., 2015).
Much accounting research that assessed the influence of culture on disclosures used the theory of cultural influence to relate national cultural characteristics to accounting practices (Hooghiemstra, 2015). As earlier introduced, Gray (1988) developed the theory of cultural influence which consists of four accounting values. Of which secrecy vs transparency (secrecy) is the most relevant for this research since it is directly related to the amount of information disclosed (Gray, 1988; Maali & Al-Attar, 2017). Gray (1988) defines secrecy as “a preference
for confidentiality and the restriction of disclosure of information about the firm only to those who are closely involved with its management and financing as opposed to a more transparent, open and publicly accountable approach”. Archambault & Archambault (2003) state that as
secrecy increases, the amount of disclosure decreases and vice versa. Gray did not empirically test his published theoretical framework about the cultural influence on accounting practices. Multiple studies examined Gray’s accounting values and found that it serves best in explaining reporting practices and weak in explaining legal and professional structures (Houqe, 2016). Since this research examines the IFRS 15 reporting practices, it is used for explaining national cultural influence on IFRS 15 disclosures.
Culture of the headquarters’ country influences the whole organizational culture (Du, et al., 2017). Therefore, cultural values are assigned to the firm where the headquarter is located, in accordance with prior research (Thanetsunthorn, 2015). The degree to which the cultural dimensions are found to be significantly associated with corporate disclosure varies in previous literature. Nevertheless, all six cultural dimensions represent a country’s national culture. Therefore, all of Hofstede’s cultural dimensions are part of this research which is a unique
6 Added in 2001. 7 Added in 2010.
10 characteristic of this research. In the following sections the influence of the individual national cultural dimensions on IFRS 15 disclosure quality is further elaborated upon and hypotheses are formulated. In hypothesis development individualism, uncertainty-avoidance and power-distance are combined into the construct secrecy, since they are considered to be the most significant for the level of secrecy according to Gray (1988). For the other dimensions separate hypotheses are developed.
Individualism vs. collectivism
The first national cultural dimension is individualism vs. collectivism, which refers to the degree to which people prefer a loosely- versus tightly-knit social framework (Hofstede, 1980). In societies characterized as individualistic, people take care for themselves and the environment is more competitive and less secretive (Hope, 2008). Gray’s theory of cultural influence (1988) states that firms from a country with an individualistic orientation disclose more information. Because in those firms, people feel less responsible to insiders as opposed to outsiders and therefore tend to disclose more information (Maali & Al-Attar, 2017). This could also be explained signaling theory which states that people want to signal their excellence which can be done by voluntary disclosing information. The opposite mechanism is applicable for firms from a country where society is collectivistic oriented. Furthermore, individualistic societies are more demanding for accountability than collectivistic orientated societies and therefore require more disclosure, possibly resulting in more voluntary disclosed information (Akman, et al., 2011). Signaling theory could explain that in order to fulfill the demand for accountability in individualistic societies, people disclose more information to signal they comply to the demand for accountability.
In general, prior accounting research support the relationship Gray (1988) states between individualism and secrecy (Hooghiemstra et al., 2015). In recent research, Akman et al. (2011) also found this relationship in their cross-country research on IFRS implementation in Europe and Australia. They find that there remains a difference in disclosure quality between firms from different countries which is partly explained by the cultural dimension individualism (positive). In other words, the higher the level of individualism, the higher the amount of disclosure of firms (opposite of secrecy). So, this suggests that after IFRS implementation individualism still could influence disclosures in Europe, and possibly also for disclosures regarding the newly introduced IFRS 15 standard in this research. Furthermore, Hooghiemstra et al. (2015) found a positive relation between individualism and the amount of voluntary disclosed information on internal controls. Based on their findings a similar relationship could
11 also be expected for other forms of voluntary disclosure. Since focus in our research is also on voluntary disclosed information, a similar influence of individualism is expected for IFRS 15 disclosures.
Uncertainty-avoidance
The second cultural dimension is uncertainty-avoidance, which refers to the extent to which persons feel uncomfortable with uncertainty and ambiguity (Hofstede, 1980). In other words, the extent to which persons feel threatened by uncertain or unknown situations (Braun et al., 2008). According to the theory of cultural influence uncertainty-avoidance is negatively associated with disclosures. Because in societies where the level of uncertainty-avoidance is high the amount of disclosure is less to avoid potential conflict and competition with external parties (Hope et al., 2008). So, firms that are situated in a country with a high level of uncertainty-avoidance tend to disclose less information to protect their own position. This way of protecting their own position by firms makes them less transparent and consequently could lead to lower the quality of disclosures (Maali & Al-Attar, 2017).
The relationship that Gray (1988) states between uncertainty-avoidance and secrecy is mostly supported in previous accounting research (Aggerwal & Goodell, 2014; Hooghiemstra et al., 2015). Hooghiemstra et al. (2015) find that the amount of disclosures related to internal controls is negatively associated with uncertainty-avoidance. Which suggests that firms which are located in a country with relatively low uncertainty-avoidance are expected to disclose more investor-related information which could be useful for all users of annual reports (Hooghiemstra et al., 2015). This could indicate that a similar relationship could be expected for revenue disclosures since revenue is an important measure of financial performance and prospects (Mattei & Paoloni, 2019; ESMA, 2019; Wagenhofer, 2014). Furthermore, Akman et al. (2011) also find evidence for a relationship between uncertainty-avoidance and financial disclosures. They find that firms disclose more information on IFRS disclosures when firms are located in low uncertainty-avoidance countries. Since uncertainty-avoidance is found to have an impact on IFRS disclosure levels between countries it is expected this is also the case for the more recently implemented IFRS 15 disclosures.
Power-distance
The third national cultural dimension is power-distance, which refers to the extent to which less powerful people expect that power is distributed unequally (Maali & Al-Attar, 2017). This means that in a country with high power-distance, persons more easily accept
12 hierarchical differences in society compared to low power-distance countries (Akman et al., 2011). According to the theory of cultural influence a negative relationship between power-distance and disclosures is expected. Gray (1988) argues that persons in high power-power-distance societies are more likely to restrict information to protect their own comfortable position. Because of their powerful position, which gives them more freedom and security, they have less incentive to disclose information to other parties which could pose a threat for their powerful position (Maali & Al-Attar, 2017).
Maali & Al-Attar (2017) find evidence for the negative relationship between power-distance and disclosures of multi-nationals as predicted by the theory of cultural influence of Gray (1988). An addition explanation of the mechanism behind this relationship is that managers in high power-distance countries are better able to conceal unfavorable disclosures as their subordinates have less power over them to be transparent (Maali & Al-Attar, 2017). This is in line with the reasoning in signaling theory, the managers have less incentives to signal their excellence since they are in a powerful position. This is relevant for our research since within our sample there are considerable differences in the level of power-distance (see table 4). A study by Aggarwal & Goodell (2015) about disclosures of large multination corporations provide similar evidence for the negative relationship between power-distance and disclosures, although Aggarwal & Goodell (2015) acknowledge that the impact of power-distance was not as significant as that of uncertainty-avoidance and individualism on disclosures. There is also research which finds inconclusive results or the opposite relationship between power-distance and disclosures (Jaggi & Low, 2000; Hope, 2003). Since most of the recent research provides support for a negative relationship between power-distance and disclosures, a negative relation between power-distance and IFRS 15 disclosures is expected.
Since Gray (1988) argues that individualism, uncertainty-avoidance and power-distance are the most important national cultural dimensions with regard to secrecy, they are combined to assess the relation between secrecy and IFRS 15 disclosures. This is in line with prior research (Hope et al., 2008). Therefore, the following hypothesis is tested.
Hypothesis 1a: The countries’ level of secrecy8 is negatively associated with the quality of IFRS 15 disclosures.
8 SEC = score on power-distance index + score on uncertainty-avoidance index - score on the continuum of individualism vs. collectivism.
13 Next, the other dimensions are assessed separately.
Masculinity vs. femininity
Masculinity vs. femininity is the last cultural dimension from the original set cultural dimensions that Hofstede introduced in 1980. Masculinity refers to the extent that a society is more assertive and gender differences are more emphasized. In other words, in masculine oriented societies individuals are more success and material oriented. On the contrary, in feminist societies care for others and social welfare are of bigger importance (Akman et al., 2011). According to the cultural influence theory masculinity is negatively associated with secrecy (i.e. positively associated with disclosures). Gray (1988) argues that masculine societies have a bigger focus on progress, development and competition than societies that are more feministic oriented. Feminist societies are expected to put more emphasis on work-life balance, people and environment (Gray, 1988). Therefore, firms from countries with masculine societies are expected to produce more disclosures according to the theory of cultural influence. Borker (2012) introduced differentiation in strength of the expected relationships. For the accounting value secrecy this resulted in an expected positive relationship between masculinity and disclosures, but of less impact than previous dimensions. Additionally, in economics and finance literature masculinity is often associated with more risk seeking behavior (Nelson, 2015). Using signaling theory, risk seeking behavior comes along with additional disclosures to signal excellence when performance is good but also when with bad performance to prevent litigation or reputational costs.
There is less research conducted on the influence of masculinity on disclosures compared to previous dimensions. Findings in prior research about the relationship between masculinity and disclosures provide diverse results (Aggerwal & Goodell, 2014). Santema et al. (2005) find that firms from masculine societies include more economic and financial information in their corporate disclosures, as Gray (1988) predicted. Since masculine societies are more concerned with the position of firms compared to other firms, resulting in more incentives to disclose economic, financial and performance information (Aggerwal & Goodell, 2014). Since this research concerns revenue disclosures this could also be applicable to IFRS 15 disclosures. However, there is also research that does not support this relationship (Hope, 2003; Archambault & Archambault, 2003; Maali & Al-Attar, 2017). Maali & Attar (2017) find an opposite relationship between masculinity and disclosures. They explain this unexpected relationship with the type of disclosures examined. In feminist societies there is more emphasis on people, society and environment which could lead to more disclosure on CSR,
14 environmental and social reporting instead of on financials (e.g. revenue) that could lead to higher overall scores of information disclosure in their research (Maali & Al-Attar, 2017). This is also found in recent research to environmental disclosures (Pucheta-Martínez & Gallego-Álvarez, 2019). Since our research focusses on revenue disclosures the following hypothesis is tested:
Hypothesis 1b: The countries’ level of masculinity is positively associated with the quality of IFRS 15 disclosures.
Short- vs. long -term orientation
Long-term orientation refers to the extent to which a society is future-oriented, dedicated and thrift9. On the contrary, short-term orientation refers to a society which is more oriented on past (Hofstede et al., 2010). Short-term orientation focusses on current results whereas long-term orientation emphasizes long-long-term relationships and reputation in order to achieve sustainable performances (Hofstede et al., 2010). IFRS 15 Disclosures requirements (IFRS15.110ev) require firms to disclose information, besides current performance, also about future-oriented information (f.e. IFRS116/117/118/119). Therefore, firms from long-term oriented countries could be expected to provide better disclosures on those specific IFRS 15 disclosure requirements.
This cultural dimension was added in 1991, so no expected relationship was included in the original study in 1988 of Gray (1988). Radebaugh & Gray (2002) suggest that long-term orientation of a society is negatively related with disclosures. Since long-term orientation is consistent with a need to keep funds inside the firm to invest for example, instead of paying high dividends or raise wages of employees. In order to prevent cash from flowing out of the firm, firms are expected to disclose less information to their stakeholders. On the contrary, Hooi (2007) suggests that long-term orientation is positively related to disclosures. Firms from a country with strong long-term orientation are expected to prefer sustainable relationships with their stakeholders which require are more transparent disclosure practices which increases disclosure quality. Hooi (2007) does not find support for relationship among financial disclosures of banks. Similar inconclusive results have been found for CSR disclosures (Orij, 2010). Prior research does not provide an unambiguous theoretical basis (Hooi, 2007; Orij,
15 2010; Maali & Attar, 2017). Therefore, no expectation about the direction of the relationship is formulated. The following hypothesis is tested:
Hypothesis 1c: The countries’ level of short- long-term orientation is associated with the quality of IFRS 15 disclosures.
Indulgence vs. restraint
In 2010, Hofstede added a sixth dimension: indulgence vs. restraint. Indulgence refers to the extent to which a society allows human desires of freedom, health, enjoying life and having fun. On the contrary, restraint refers to opposite where those desires are curbed and regulated by strict social norms (Hofstede et al., 2010). To the best of our knowledge Maali & Al-Attar (2017) is the only study who empirically tested the relation between indulgence and disclosures. They argue that indulgence is associated with freedom of expression which could indicate higher disclosure quality, although no statistical evidence is found. According to Borker (2013) indulgence is negatively related to secrecy (i.e. positively related to disclosures) since high indulgence complements the notion of freedom of judgment. Borker (2013) does not empirically test his proposition. Since IFRS 15 contains multiple disclosure requirements which require disclosure of significant judgment information (IFRS15.110/123/125) it is expected that firms from high indulgence countries are better able to freely disclose their IFRS 15 judgments. In addition, our research sample compiled by firms from which it is plausible they apply significant judgments. Therefore, the following hypothesis is tested:
Hypothesis 1d: The countries’ level of indulgence is positively associated with the quality of IFRS 15 disclosures.
Besides the direct effect of national culture on disclosures Gray (1988) states that national culture also has an indirect influence on disclosure practices because of the influence of national culture on institutional determinants of disclosure quality (Aggerwal & Goodell, 2014). In this research investor protection and capital market development are assessed because these country specific factors are often found to be associated with disclosure quality and are regularly measured.
Investor protection
Investor protection is shown to be an important institutional factor in the firm policy practices (Emanuels, et al., 2015). Hooghiemstra et al. (2015) summarize findings from other research on investor protection. The basis of the reviewed papers by Hooghiemstra et al. (2015)
16 are the papers of La Porta et al. (2008, 1998) which state law and finance are related to each other. Other studies extended this research by showing, among other findings, that investor protection is positively related to the amount of voluntary disclosure (Hooghiemstra et al., 2015; Dayanandan, et al., 2016). Hooghiemstra et al. (2015) state that agency problems are present in all countries, but the nature of agency problems depends on the strength of investor protection. By applying agency theory, it is expected that that firms face stronger incentives to reduce agency problems when high investor protection is present. Which could result in more voluntary disclosure since agents want to show that they act in the interest of the firm since they want to reduce litigation and reputational costs.
Multiple papers state that not only the presence of investor protection regulation has influence on disclosures but also its enforcement (La Porta, et al., 2008; Glaum, et al., 2013; De George, et al., 2016). Reporting practices are mainly determined by a firms’ reporting incentives, enforcement regime is one of the institutional factors who shape these incentives (Brüggemann, et al., 2013). In IFRS adoption research is found that IFRS only has positive effects on disclosure quality when firms have incentives to be transparent and face strong regulatory enforcements (Daske, et al., 2008; Holthausen, 2009). Enforcement of reporting standards is an important topic for the ESMA (European Securities and Markets Authority). From 2012 onwards the ESMA identifies ECEPs (European Common Enforcement Priorities) to promote consistent application of IFRS since countries must enforce reporting regulation themselves which is likely to result in differences in enforcement practices. In the ECEP of 2019, ESMA underlines the importance to enforce application and understandability of IFRS 15 disclosures. Furthermore, they highlight the need to further improve quality, consistency and coherence of information provided in the annual reports for 2019. Which is relevant for the EU-listed construction firms in our research as most of their revenues come from long-term complex construction contracts. The differences across the EU in regulatory enforcement power are illustrated by Kaufmann & Kraay (2019) and are expected to influence IFRS 15 disclosure practices. This relationship can be explained by agency theory. Firms face stronger incentives to reduce information asymmetry with investors when enforcement of investor protection is high, because of higher risk of litigation or reputation costs for example.
In accordance with Hooghiemstra et al., (2015) investor protection is assessed as a combination of the investor protection regulation and its enforcement. Han et al. (2010) found that investor protection in return in influenced by the national culture. Which is in line with Aggerwal & Goodell’s (2014) statement that culture does not only have a direct influence on
17 accounting practices but also an indirect effect through institutional factors, as predicted by Gray (1988). Therefore, the following hypotheses will be tested:
Hypothesis 2a: The countries’ level investor protection is positively associated with the quality of IFRS 15 disclosures.
Hypothesis 2b: There is significant mediating effect of national culture on IFRS 15 disclosure quality through investor protection.
Capital market development
Capital market development is the second institutional factor in this research (Gray, 1988) 10. Economic systems have an influence on the relationship between firms and investors and therefore have an influence on the information that needs to be disclosed (Archambault & Archambault, 2003; Glaum et al., 2013). Because of the listed status of our sample firms they are partly financed by their shareholders11. Shareholders are expected to have stronger incentives to monitor firm performance compared to for example banks, because shareholders experience both downside risk and upside potential where credit providers only face downside risk. For shareholders revenue disclosures are an important source of information to form judgments and expectations about firm performance. As mentioned earlier, “demand for information results from market participants' need to reduce information asymmetry” (Soderstrom & Sun, 2007). If market participants cannot determine good and bad firm performance, adverse selection can occur which could lead to high finance costs. Credible signaling can reduce this problem (Soderstrom & Sun, 2007). For example, it is found that public firms in high developed capital markets produce more high-quality accounting information due capital markets providing incentives to produce informative information (Archambault & Archambault, 2003; Burgstahler, et al., 2006). By applying signaling theory it can be shown that firms have more incentives to signal their excellence to market when the capital market they are in is more developed. For firms this could mean that in more developed capital markets incentives to disclose more information on revenues, as important measure of performance, could lead to higher disclosure quality. According to the World Bank (2019) considerable differences in capital market development are present in Europe. EU-construction firms could react to pressure from capital markets to disclose additional revenue information on specific projects for example which could be useful information as a performance indicator
10 See appendix 4: Gray’s (1988) Model, extended by Doupnik and Tsakumis (2004). 11 In our sample on average 31% of firms’ capital consists of equity.
18 for investors. On the other hand, when firms perform less, they could decide to disclose additional information to avoid reputation or litigation costs according to signaling theory.
Since capital market development could be seen as an institutional consequence of national cultural values this indicates there is not only a direct relationship with disclosures, but it could also act as mediator in the relationship between national culture and IFRS 15 disclosures (Gray, 1988). Therefore, the following hypotheses will be tested:
Hypothesis 3a: The countries’ level of capital market development is positively associated with the quality of IFRS 15 disclosures.
Hypothesis 3b: There is significant mediating of national culture on IFRS 15 disclosure quality through capital market development.
The freedom for firms to voluntary disclose additional information creates opportunities for cultural and institutional factors to influence the quality of IFRS 15 disclosures. In conclusion, it is expected that even with the same reporting standards, IFRS 15 disclosure quality will differ between countries. This is in accordance with earlier research on IFRS disclosures (Soderstrom & Sun, 2007; Glaum et al., 2013; Dayanandan, et al., 2016; De George et al., 2018).
iii.
METHODOLOGY
In this section the methodology is addressed in the following order: sample selection, dependent variable and independent variables.
Sample selection
Data is collected from the annual reports of listed firms from the EU. The annual reports are reported in accordance with EU-endorsed IFRS standards. Since EU-endorsed IFRS application is mandatory, these annual reports are relevant for this research. The sample consists of 35 European listed firms which are selected from the Orbis database. The selection consists of active listed firms from the EU which fall in the following NACE Rev. 2 classifications which will thereafter generally be referred to as ‘construction firms’: construction of buildings (41), civil engineering (42) and specialized construction activities (43). The 35 biggest construction firms are selected, sorted by operating revenue, due to research capacity limitations in this research. This research sample is relevant for this research since they are affected by the
19 IFRS 15 standard12 and have a large impact on society and their sector because of their size. The annual reports of sample firms for the years 2018 and 2019 are used to collect data. For 11 firms only the annual report of 2018 is included in analysis since they did not yet publish their annual report of 2019 at the time of writing or their financial year runs to the 30th of June 2020. Dependent variable
IFRS 15 Disclosure quality
IFRS 15 Disclosure quality (DC) is the dependent variable in this research. Annual reports are one of the most important sources of voluntary disclosure (Shehata, 2014). Furthermore, the annual reports are easily accessible and available for research (Catasús, 2008). The value of DC will be determined on data from the annual reports of sample firms.
The quality of IFRS 15 disclosures is measured using a disclosure index. Disclosures indices are widely used in previous studies to measure disclosure quality (Scaltrito, 2015; Akman, 2011). In research both self-constructed and existing disclosure indices are used. For this research a self-constructed unique disclosure index is used since, to the best of our knowledge, there is no IFRS 15 disclosure index with a focus on construction firms available. The index is based on IFRS 15 disclosure requirements (IFRS15.110 et seq.) and each element is weighted equally.
Both voluntary and mandatory disclosures are part of the disclosure index13. Construction firms face certain requirements regarding quantitative disclosures which they could choose to provide further qualitative or quantitative disclosures on. Therefore, the voluntary disclosure aspect is included in this disclosure index. On each element of the index a maximum of 2 points can allocated when a firm discloses both mandatory and voluntary information. 1 Point is allocated when a firm only discloses generic mandatory information. No points are allocated when required information is not disclosed. When an index element is not applicable this will be marked as ‘N/A’.
The team of researchers consists of 5 master students of the University of Groningen. The allocation a disclosure index score requires judgment of the researchers which could be weakness in this research (Akman, 2011). Multiple meetings have been organized to align the method of interpretation and judgment. When differences in interpretation occurred, this is discussed within the team to safeguard consistency in the disclosure scores. To strengthen the
12 IFRS 15 has replaced IAS 11 “construction contracts”. 13 For the complete disclosure index, see appendix 6.
20 reliability of DC further, an overlap in the distribution of annual reports between researchers is applied. This ensures that differences in interpretation are noticed and are discussed which further increases the consistency in the allocation of points. Insights in differences in interpretation found in overlapping annual reports are also applied to the non-overlapping annual reports.
The DC score is calculated by dividing the total amount of points by the number of elements that are judged as relevant, which is in line with prior research14 (Scaltrito, 2015). Independent variables
In this section the methodology of the independent, mediating and control variables is addressed.
Explanatory variables
National cultural dimensions & secrecy
National culture is measured by Hofstede’s cultural dimensions. Hofstede’s cultural dimensions have been subject of criticism other studies, nevertheless researchers widely used Hofstede’s model on the influence of culture on management behavior and diversity of financial systems (Hooghiemstra, et al., 2015). Hofstede’s data on cultural dimensions is derived from his website, which is freely available for research purposes. The use of this database does not restrict the number of firms in our analyses. The most recent cultural dimension data is from 2015. Hofstede (2015) states that "Since culture only changes very slowly, the scores can be
considered up to date". This is in accordance with recent research on national cultural dynamics
(Matei & Abrudan, 2018). They use a tree as analogy to illustrate national cultural change:
Every tree is unique in its own way, and just as a tree changes leaves but a part of it remains intact, so do cultures change on the surface, with their essence is left unchanged. Therefore,
this data is used in testing the hypotheses. The firms are assigned to the country where the headquarter is located since culture of the headquarters’ country influences the whole organizational culture (Thanetsunthorn, 2015; Du, et al., 2017). Another line of reasoning could be that the tone at the top (i.e. directors) has an influence on the culture within a firm (Schwartz, et al., 2005). In our firm sample, mostly the majority of the boards consist of directors with the nationality of the country where the headquarter is located. Therefore, also within this line of
21 reasoning firm headquarter location as determinant for dominant cultural values of a firm can be seen as legitimate.
The national cultural dimensions individualism, uncertainty-avoidance and power-distance are combined into one construct as they are the most relevant for the quality of disclosures according to Gray (1988): secrecy (SEC). This is in line with prior research of Hope et al. (2008). An advantage of using this construct is prevents the occurrence of multicollinearity among the national cultural dimensions. SEC is constructed based on the hypothesized direction of the relationship between the respective dimension and IFRS 15 disclosure quality and weighted equally, in line with Hope et al. (2008)15.
Investor protection
Djankov et al. (2008) developed a measure of investor protection, named the anti-self-dealing index (ASDI). Djankov’s (2008) methodology captures three dimensions of investor protection: transparency of related-party transactions, liability for self-dealing and shareholders’ ability to sue managers for improper behavior. Based on prior studies, the predecessor of ASDI, ADRI is a common method to determine a country’s level of investor protection (Houqe, et al., 2016; Hooghiemstra, et al., 2015). There has been considerable criticism on the use of the ADRI because of possible coding mistakes and conceptual ambiguities (Houqe, et al., 2016). Spamann (2010) argues that many studies who have used the
ADRI may not be replicable with his revised version of the ADRI. The ASDI index generally
worked better than the prior ADRI (Haidar, 2009). ASDI is stated to be: “a superior alternative
for ADRI (..) and the most accepted numerical measure of the quality of corporate law”
(Cheffins, et al., 2016). Therefore, in this research ASDI is used as measure of investor protection. The World Bank (i.e. Doing Business Project) annually provides an updated investor protection index based on Djankov et al.’s (2008) methodology (World Bank, 2018). The index provides a range of 1 to 10, where a score of 10 indicates the highest form of investor protection and 1 the lowest. The World Bank uses survey data from corporate lawyers about court rules of evidence, company laws and securities regulations (Rachisan, et al., 2017)16. The scores for the years 2018 and 2019 are used in analysis. Furthermore, in the this research a country’s enforcement power in considered in the protection of investors. As stated by Burgstahler (2006): ‘without proper enforcement legal rules remain largely ineffective’. In accordance with prior research the rule of law component (ROL) of the WGI index (Kaufmann & Kraay, 2019)
15SEC = score on power-distance index + score on uncertainty-avoidance index - score on the continuum of individualism vs. collectivism.
22 will be used to account for the enforcement aspect of investor protection (Emanuels, et al., 2015). The level of enforcement power is measured annually. Data for 2018 and 2019 are available for research. The combination of investor protection and enforcement score, weighted equally, are used in this research as measure of investor protection, in line with Hooghiemstra et al. (2015).
Capital market development
The third independent variable is the level of capital market development (CMD). A common method to measure capital market development is the ratio of market capitalization plus domestic credit to gross domestic product (GDP) (Hsu, et al., 2014). Where market capitalization is measured as stock price times the number of outstanding shares per company listed on the domestic stock exchange. Our research is mainly concerned with the incentives provided by equity markets regarding disclosing information about IFRS 15 disclosures since shareholders experience downward risk and profit form upside potential where banks only experience downward risks. This could indicate that equity markets have a stronger influence on IFRS 15 disclosure quality as they profit more from it. Therefore, the credit provided by banks is not included in the measurement of capital market development in our research. Capital market development is measured as market capitalization / GDP (Hsu, et al., 2014). The data is derived from CEIC and is publicly accessible on the website of CEIC for the years 2018 and 2019, composed from information provided by local stock exchanges17 (CEIC, 2020).
Control variables
Since above mentioned dependent variable are of country specific nature there must be controlled for firm specific determinants of disclosure quality (Soderstrom & Sun, 2007).
Profitability (PROFIT)
In prior voluntary disclosure quality research PROFIT is a widely used control variable (Dhaliwal, et al., 2011). PROFIT is used as measure of performance for firms. Poor performance is associated with less disclosure (Francis, et al., 2008). Firms with low profitability are expected to have less resources to provide voluntary disclosures, vice versa holds for firms with a high profitability. PROFIT is measured by firms’ return on assets. Calculated as net result / total assets, derived from the annual reports.
17 Equity market capitalization data for the Austrian stock exchange was only available till year 2018 (CEIC, 2020). Therefore, more recent data from 2019 for the Austrian stock exchange is included, derived from the World Bank database, which shows a similar value compared to prior years from the CEIC database (World Bank, 2019).
23
Firm size (SIZE)
SIZE is a common determinant of disclosure quality in literature (Archambault &
Archambault, 2003). There are multiple lines of reasoning why SIZE influences disclosures. Archambault & Archambault (2003) summarize serveral explanations. Large firms have more resources to disclose more information. Also a bigger public demand for disclosures from large firms opposed to smaller firms is seen as explanation for higher disclosure quality. While the theoratical reasonings is conflicting, large firms are expected to deliver a higer level of disclosure quality than small firms. SIZE is measured as logarithm of total revenue since total amount show large differences between firms, which is in line with prior research (Dhaliwal, et al., 2011). Total revenue is derived from the annual reports.
Financial leverage (LEV)
The relationship between LEV, as firm specific factor, and disclosure quality is often found to be a significant relationship in studies. Debt provided by banks is not included in the country specific institutional variable CMD as explained. Nevertheless, differences in leverage on firm level could influence disclosure quality and therefore needs to be controlled for. In the article of Archambault & Archambault (2003) about the cultural, national and corporate factors of influence on corporate disclosures multiple studies are metioned which find conflicting results regarding the direction of the relationship between LEV and disclosure quality. Archambault & Archambault (2003) find a significant positive relationship between LEV and disclosure quality which can be explained using agency theory. Firms with high LEV disclose more information to reduce the agency costs of debt (Archambault & Archambault, 2003). LEV is measured by total liabilities (current and non-current) / total equity, which are derived from the annual reports.
Year dummy (DYEAR)
To control for year effects a year dummy is included. The dummy is 0 for 2018 and 1 for 2019 in analysis.
Model of analysis
For our research the ordinary least squares method is used in regression analyses. A complete list of all variables is shown in the table 1. Hypotheses are tested separately, and the full empirical model of this research is as follows:
DC = β0 + β1*SEC + β2*MA + β3*ST + β4*INDUL + β5*IP + β6*CMD + β7*PROFIT + β8*SIZE + β9*LEV + β10*DYEAR + ε
24 Variable overview
Type of variable Name Definition Source
Dependent variable DC Final score in the IFRS
15 index calculated as total allocated points / (maximum amount of points to be earned less N/A)
Hand collected data from annual reports by thesis group
Independent variables (hypotheses 1a: d)
SEC Score on power-distance
index + score on uncertainty-avoidance index - score on the continuum of individualism vs. collectivism.
Hofstede database
MA Score on the continuum
of masculinity vs. femininity
Hofstede database
ST Score on the continuum
of short-term vs. long-term orientation
Hofstede database
INDUL Score on the continuum
of indulgence vs. restraint
Hofstede database Mediating variables
(hypotheses 2a: b, & 3a: b)
IP Score on the Investor
Protection Index
Doing Business website (World Bank)
CMD Market capitalization /
GDP
CEIC website
Control variables DYEAR Dummy variable for year,
0 = 2018; 1 = 2019
-
PROFIT Net results / total assets Annual report
SIZE Natural logarithm of total
revenue
Annual report
LEV Total liabilities / total
equity
Annual report
25
iv.
RESULTS
In this section the results of our research are discussed. The structure is as follows: descriptive statistics, best practices, correlation analysis, regression analysis and robustness tests.
Descriptive statistics
In this section general information is provided about the research sample and countries they originate from. Information is provided about average numbers, frequency, minimum versus maximum values and standard deviations.
Frequency Percentage Austria 2 5% Belgium 2 5% Finland 2 5% France 4 11% Germany 2 5% Great-Britain 10 27% Italy 1 3% Netherlands 4 11% Spain 5 14% Sweden 5 14% Total 37 100%
Table 2: Distribution of sample firms between EU countries
Table 2 shows the distribution of firms among EU countries. The firms are based in Southern Europe (France, Italy and Spain), Western Europe (Austria, Belgium, Germany, Great-Britain and The Netherlands) and Northern Europe (Finland and Sweden). No firms from Eastern European countries are included in the sample. A large proportion of the firms originates form Great-Britain (27%), which could be explained by the large number of stock listings in Great-Britain.
26
N Minimum Maximum Mean Std.
Deviation
IFRS 15 Disclosure score 2018
34 0,30 0,70 0,50 0,10
IFRS 15 Disclosure score 2019
29 0,26 0,75 0,51 0,11
Table 3: IFRS 15 Disclosure score by year
Table 3 shows IFRS 15 disclosure index scores. The number of firms of which IFRS 15 disclosures are assessed differs between 2018 and 2019 by 5 firms, mainly due the fact that a number of firms did not yet publish their annual report of 2019 at the time of this research or use a different fiscal year period. The average score and standard deviation for 2018 and 2019 are similar (0.50 versus 0.51; 0.10 versus 0.11). The average scores are not very high, also no firm has obtained the maximum score of 1.00. Since half of maximum amount of points could be scored by providing voluntary disclosure, it is no surprise the maximum score is not reached in this research. The distance between the minimum and maximum score in 2019 (0.26 versus 0.75) is slightly higher than for 2018 (0.30 versus 0.70). This is also reflected is the slightly higher standard deviation for 2019 (0.11) compared to 2018 (0.10). Based on the descriptive statistics in table 3 it can be concluded there is no noteworthy increase in average score in 2019 compared to 2018 which might be expected since firms could have learned from best practices of other firms from 2018 disclosures.
N Minimum Maximum Mean Std. Deviation
Power-distance 10 11 68 42 18
Individualism vs. collectivism 10 51 89 70 11
Masculinity vs. femininity 10 5 79 47 25
Uncertainty-avoidance 10 29 94 65 22
Short- vs. long-term orientation 10 38 83 61 14
Indulgence vs. restraint 10 30 78 55 15
Table 4: Independent variables
Table 4 shows the 6 cultural dimensions values, derived from Hofstede’s database (Hofstede, 2015). Data is available for all countries included in the research sample (N=10). In this table each country is weighted equally. It is obvious that there are considerable differences in values for the different dimensions. The difference between the minimum and maximum of
27 masculinity vs. femininity (5 versus 89) is the highest which is also reflected in the high standard deviation (25). This is mainly caused by high masculinity scores for Austria (79) and Italy (70) and low masculinity score (i.e. high feminist score) in The Netherlands (14) and Finland (5)18.
N Minimum Maximum Mean Std. Deviation
Investor protection index 2018 10 58 84 68 7
Investor protection index 2019 10 58 84 68 7
Rule of law index 2018 10 55 91 80 11
Rule of law index 2019 10 65 87 79 7
IP 2018 19 10 60 83 74 6
IP 2019 10 66 82 73 5
CMD 2018 10 26 124 79 36
CMD 2019 10 26 154 84 39
Table 5: Mediating variables
In table 5 the mediating variables are shown. Values for the investor protection index for 2019 were the same as 2018 which is in line with earlier mentioned limited year-to-year changes for institutional factors20. Capital market development shows a slight increase in average number (79 in 2018 to 84 in 2019). Furthermore, the maximum value increased to 154 in 2019 due to increased market capitalization in Sweden (124 in 2018 to 154 in 2019). Differences in value for CMD are higher than for IP as can be seen by the larger standard deviation.
Best practices and learning effects
In order to highlight some of the high-quality disclosures in IFRS 15 disclosures we identified best practices and learning effects.
The first best practice is found in the annual report of Balfour Beatty PLC for the 2019 annual report. They very clearly disclosed when performance obligations are typically satisfied, in accordance with IFRS15.119a. This information is specified per type of asset and qualitatively explained (see appendix 11). This presentation makes it easy for the reader to get
18 See appendix 7 for a complete overview of all values of sample countries regarding cultural dimensions. 19 IP is constructed of two indices: investor protection and rule of law.
28 a thorough understanding of the nature, timing and satisfaction of performance obligations and significant payment terms per different type of asset. In other annual reports this information is often not that clearly and easily readable presented. Another important part of the IFRS 15 standard are the contract assets and liabilities (IFRS15.116a2/3). Best practice for this requirement is found in the annual reports of Vinci SA. Vinci SA separately presents contract assets and liabilities per business segment and discloses where change originates from, with additional qualitative information (see appendix 12). This enables readers to quickly assess the developments regarding contract assets/liabilities within the firm’s different business segments in one table.
In general, little learning effects can be observed by looking at the average disclosure scores for 2018 and 2019 (0.50 vs. 0.51). This small increase in averages disclosure score does not clearly indicate learning effects in 2019 disclosures. Nevertheless, on firm level learning effects can be observed. For example, Constain Group PLC improved their disclosures regarding judgments made in determining the amount of costs incurred (IFRS15.127a) by adding project specific information in 2019. Another example of a learning effect is the improvement of disclosures regarding IFRS15.116a2/3 of ACS S.A. They present contract assets/liabilities balances in a separate table in 2019 instead of within the net trade receivables table in the 2018 annual report.
Correlation analysis
In this section the correlations between the used variables are assessed. A high degree of correlation between variables could cause the regression model estimates of coefficients to become unstable and standard error for the coefficients could become inflated. In table 7 the existing correlations are presented, findings are subsequently discussed.
DC SEC MA ST INDUL IP CMD DYEAR SIZE21 PROFIT LEV DC 1 SEC -0,061 1 MA -0,197 0,005 1 ST -0,192 .245* 0,096 1 INDUL 0,012 -.819** -.321** -.243* 1 IP -0,073 -.800** 0,212 -.371** .766** 1 CMD 0,073 -.437** -.707** -.386** .645** .364** 1 DYEAR 0,056 0,000 0,000 0,000 0,000 -0,075 0,103 1 SIZE 0,072 .475** 0,038 0,212 -.423** -.342** -0,186 0,016 1 PROFIT -.409** -.376** 0,105 -0,172 .384** .478** 0,247 0,149 0,045 1 LEV .331** .319* -0,216 0,087 -.341** -.469** -0,128 -0,084 .317* -.388** 1
** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed). Table 6: Correlations table
