The Effect of Demographic Characteristics on the
Disclosure Quality of IFRS 15
Looking at Age and Gender
Master Thesis Accountancy
University of Groningen, Faculty of Economics and Business
Ine Rietema S2705222 Supervisors: Prof. dr. R.L. ter Hoeven
Msc. R. van Duuren
Word count: 11,280 24th of June 2019
2
ABSTRACT
The aim of this study is to investigate whether gender diversity and the average age of Board of Directors and Audit Committees, and the amount of analysts influences the disclosure quality of the first full adoption of the new revenue recognition standard: IFRS 15 Revenue
from Contracts with Customers. The quality of disclosure of IFRS 15 is determined by manually rating the disclosures of IFRS 15 for 52 European listed companies operating in the telecommunications industry, utilities industry, construction and materials industry, and
technology industry. Contrary to the expectations, a significantly negative relationship is found between the degree of women on the Board of Directors and on the Audit committee and the disclosure quality of IFRS 15. A significantly positive relationship is found between the average age of members of the Board of Directors and disclosure quality of IFRS 15. This
research provides a general indication of the level of compliance of the first full disclosure of IFRS 15, and gives implications for further research.
Keywords: Disclosure Quality – IFRS 15 – Revenue Recognition – Board of Directors –
3
TABLE OF CONTENTS
I. INTRODUCTION 4
II. ACADEMIC CONTRIBUTION 7
III. THEORETICAL FRAMEWORK 7
Content of IFRS 15 8
Different board structures 9
Agency Theory 11
Upper Echelons Theory 12
Effect of women 13
Effect of age 14
Role of analysts 15
IV. RESEARCH METHODOLOGY 17
Sample selection 17
Dependent variable: Disclosure Quality of IFRS 17
Independent variables 19 Control variables 21 Variable overview 23 V. RESULTS 24 Descriptive statistics 24 Correlation analysis 27 Multicollinearity analysis 31 Regression analysis 35 VI. DISCUSSION 40 VII. CONCLUSION 42 REFERENCES 43
APPENDIX I: DISCLOSURE INDEX 48
APPENDIX II: OVERVIEW OF COMPANIES INCLUDED IN THE SAMPLE 55
APPENDIX III: AWARDING POINTS 56
4
I. INTRODUCTION
Since the beginning of the financial year of 2005, all European listed firms have to apply the International Financial Reporting Standards (IFRS 15, 2018). In 2016, firms in over 100 countries reported under, or linking their local accounting standards close to IFRS (de George et al., 2016). Financial statements of European listed firms beginning on or after the first of January 2018 have to comply with new IFRS 15 regulation. IFRS 15 replaces the former standards IAS 11 Construction contracts and IAS 18 Revenue, as these were no longer sufficient due to increasing complexity of business models and complex contracts with customers. The increased complexity of business models and contracts resulted in less comparability and less information value of annual reports for their users. This was the main reason for replacing IAS 11 and IAS 18 and interpretations related to these standards by IFRS 15 Revenue from contracts with customers. Another reason for the introduction of IFRS 15 was the increased attention for revenue recognition due to accounting scandals like the scandal of Enron (Araab et al., 2015). Furthermore, IFRS 15 was developed because there existed many revenue recognition standards with different opportunities for interpretation, which resulted in inconsistencies in accounting for contract revenues (Almohashi, 2017). With the adoption of IFRS 15, regulators strive for one revenue recognition model that can be used in all industries (Nicolae and Ionela-Claudia, 2014). Holzmann and Munter (2014) also issued a paper about the new standard. In addition to the previous objectives given, they add that the new revenue recognition model should provide a more robust framework for addressing issues regarding revenue recognition, provide more useful information to investors, and simplify the preparation of financial statements (Holzmann and Munter, 2014). The introduction paper issued by EY regarding the regulation of IFRS 15 provide a more general objective of the new reporting standards, focusing only on the users of the financial statements: ‘The objective of the new
disclosure requirements in the new standard is to provide sufficient information to enable users of financial statements to understand the nature, amount, timing, and uncertainty of revenue and cash flows arising from contracts with customers’ (EY, 2018).
A study of Roozen and Pronk (2018) indicates that companies that are operating in the telecommunications industry and utilities industry will mainly be affected. The reasons for the material impact of IFRS 15 on the telecommunications industry are the change of identification of performance obligations and the change of the processing of costs for obtaining contracts. The last one is also important for the utilities industry. Furthermore, the construction and
5 materials industry is likely to be affected as their standard IAS 11 construction contracts is replaced by IFRS 15 (BDO, 2018).
Moreover, IFRS 15 contains more requirements regarding the disclosure of accounting methods. Requirements are for example about the disclosure of judgement, so the users of the financial statements should be better able to understand the judgements made concerning revenue recognition (Corbi and Maithripala, 2019). Due to the new requirements for disclosure, more information will become available and a decrease in the differences between accounting methods will lead to a better way of comparing companies and will improve investors’ predictions of future cash flows (Core et al., 2015).
Different research has been conducted regarding compliance with IFRS. Results show that there is not always full compliance to IFRS disclosure requirements (e.g. Tsalavoutas, 2011; Erkens, 2016). This could be explained by low enforcement mechanisms, a lenient approach of regulators regarding compliance in the first year of IFRS, the tendency of companies to not provide high levels of disclosure, and low familiarity of (new) IFRS standards. But what factors have an influence on the quality of disclosure? It will be interesting to take a corporate governance approach to this issue, as the aim of corporate governance is to protect the interests of shareholders and other stakeholders and ensure transparency and integrity in the communication and to provide available, accurate and clear disclosures (Madhani, 2017). In the unique setting of the first year of full adoption of IFRS 15, the research will focus on characteristics of the members of the Board of Directors and the members of the Audit Committee. Board of Directors are seen as the major internal corporate governance system to monitor management. Another reason for including the Board of Directors, is the increasing attention for the composition of Board of Directors due to for example gender quota regulations and Corporate Social Responsibility. Gender diversity can affect the functioning of corporate boards and their committees (Adams and Ferreira, 2009). The Board of Directors is regarded as the non-executives in an one-tier board structure and the Supervisory Board members in a two-tier board structure. this topic later will be adressed later. Furthermore, the Audit Committee is included in this research. The Audit Committee is considered as a sub-committee of the Board of Directors and is concerned with the oversight of financial reporting. The Board of Directors and the Audit Committee are considered to be the monitoring and advising committees within companies.
The article of Ball et al. (2003) states that standards alone are not sufficient because financial reporting practices are not only subjected to standards, but also subjected to incentives of managers who are responsible for preparing the financial statements. As boards have an
6 active role in the disclosure process (Holland, 2005; Ettredge et al., 2011), they can influence the disclosure quality. These days, boards are changing. Countries more often adopt regulation about the minimum percentage of women on the board. In 2013, already ten countries had adopted legislation for listed and/or state owned enterprises concerning a quota for feminine board members (Terjesen et al., 2015). Among the largest European listed companies, the percentage of women on the Board of Directors has more than doubled, from 10% in 2005 to 22% in 2015 (Boffey, 2017). In 2017, the average percentage of women on the Board of Directors in the largest publicly listed companies registered in EU Member States was around 25% (European Commission, 2018). Due to the quotas applied in some countries, the percentage of women on the Board of Directors was far higher, for example in France (43.4%). In addition to the fact that more women enter higher positions, the composition of age in firms is changing as younger people are entering in higher functions (Tanikawa et al., 2017). Due to gender legislation the percentage of women on Boards of Directors and the average age of members of corporate governance committees change. Age can affect the values of a person and therefore decisions of people can be affected (Hambrick and Mason, 1984). This research will examine the role of these characteristics on the disclosure quality of IFRS 15.
Furthermore, the effect of analysts is included in this research. Analysts are one of the most important users of financial statements of companies. They use financial statements as source for their forecasts. For analysts, one of the most important items of the annual report is the reported earnings. Complying with the requirements of investors and analysts is mentioned as one of the main reasons for companies to provide information (Fuertes-Callén et al., 2014). Regarding this explanation, analysts can put pressure on companies to disclose information. As this research is about the disclosure of revenue, it will be interesting to examine the influence analysts can have on the disclosure process.
In sum, this research considers the influence of age and gender of the Board of Directors and the Audit Committee of listed European firms on the level of disclosure quality and the role of analysts concerning the first full application of IFRS 15. This results into the research question:
To what extent does the degree of women and the age of members of the Board of Directors and the members of Audit Committee influence the quality of the disclosure of IFRS 15 and what is the role of the analysts?
7 This paper is organized as followed: Section II extends the introduction with discussing the academic relevance of this research. Section III provides the theoretical framework. Section IV addresses the methodology used in this research. Section V provides the descriptive statistics and the results of the study. Section VI addresses the discussion of the research and provides further research possibilities. Section VII provides the conclusion of this research.
II. ACADEMIC CONTRIBUTION
The research is relevant, because the research will be executed in an unique setting, namely the first year of full adoption of IFRS 15. The annual reports that will be used for this research are the first fully containing disclosures of the IFRS 15 standard. This research will be interesting to legislators, investors, companies, and researchers because new regulation will be examined. The research will provide a general picture of the extent to which the new regulation has been applied. Beside that, legislators can use the findings of this research when considering new or additional regulation, and evaluating existing regulation with respect to disclosure. Research concerning this topic has not yet been executed as the data of full adoption became available during the research process. Therefore, it extends previous research on the influence of board characteristics on disclosure quality. The results of this research and the possible implications of this research can provide researchers with new research possibilities. Furthermore, the disclosure index that is constituted could be used as a starting point for further research.
III. THEORETICAL FRAMEWORK
First, the content of IFRS 15 will generally be discussed. Second, we will elaborate on the different board structures that are included in this research. Third, several theories that are applicable for this research are discussed. Fourth, we will elaborate on the variables included in this research and come to the hypotheses.
8
Content of IFRS 15
IFRS 15 Revenue from contracts with customers is applicable for financial statements of European listed firms beginning on or after the first of January 2018. This new standard is a result from the collaboration between the two major standard setters: the US Financial Accounting Standards Board (FASB) and the International Accounting Standards Board (IASB). The new regulation of IFRS 15 will be applicable for all revenue from contracts that are closed with customers with exception of revenue within the scope of IFRS 16 Leases, IFRS 17 Insurance Contracts, IFRS 9 Financial Instruments, IFRS 10 Consolidated Financial
Statements, IFRS 11 Joint Arrangements, IAS 27 Separate Financial Statements, IAS 28 Investments in Associates and Joint Ventures, and non-monetary exchanges between entities in the same line of business to facilitate sales to customers or potential customers (IFRS 15,
2018). IFRS 15 contains standards for when revenue from contracts with customers should be recognized, the way revenue should be measured and which disclosures need to be in place (Khamis, 2016).
Firms have to apply a five-step approach during the process of revenue recognition. In each step of this model, a certain kind of judgement is needed. Accordingly, disclosures have to be made concerning these judgements. This approach to revenue recognition will be discussed and some examples of judgements are given to present an overall impression of the content of the new standard. The judgements that are made, have to be disclosed. So, presenting some possible judgement issues will also show what kind of disclosure have to be presented in the annual report.
First, the contract with the customer has to be identified. This first step often requires certain judgement because it could be unclear if there is a contract with a customer or not, for example if there is a contract when two parties made an oral agreement. Second, the performance obligations of the contract have to be identified. In this step is also some judgement applicable with respect to the identification of performance obligations, for example in determining when there are two separate performance obligations and when there only is one performance obligation. Third, the transaction price has to be determined. The transaction price is the amount an entity expects to be entitled to in exchange for the transferring the goods or services. How to account for variable consideration? The fourth stage of the approach is to allocate the transaction price to each performance obligation separately. The last stage of the approach is to recognize revenue when the performance obligation is met by the transfer of the promised goods or services (IFRS 15, 2018). The moment the revenue is recognized has to be central to the moment of transfer of control over goods or services delivered (AFM, 2018).
9 IFRS 15 can differ from the former revenue recognition practices of a firm with respect to the moment of revenue recognition, and the amount of recognized revenue. Under this new standard, revenue can be recognized in time and over time. This depends on the moment a performance obligation is met. As can be concluded from above, firms have a lot of freedom in applying their own judgement. Firms have to disclose the judgements and the changes in judgements that they had to make in the application of this standard that have a significant impact on the determination of the amount and timing of revenue from contracts with customers (IFRS 15.123). With the high extent of judgement firms can exercise, the objective of comparability and consistency (Yeaton, 2015) seems not to be achieved.
Gibbins et al. (1990) defines financial disclosure as ‘Any deliberate release of financial information, whether numerical or qualitative, required or voluntary, and via formal or informal channels’. The article of Fekete et al. (2008) points out that there are companies that comply with IFRS on a mandatory basis and there are companies that apply IFRS on a voluntary basis. In addition, the content of disclosed information has these levels as well, information disclosed mandatory (within the scope of IFRS) and information disclosed voluntary (outside the scope of IFRS) (Fekete et al., 2008). For this research, we look at the information European listed firms disclose with regard to IFRS 15, so firms complying with IFRS mandatory with mandatory disclosed information. However, the extent to which companies provide disclosures can vary. Next to the required disclosures, a company can choose to disclose the required disclosures to a minimum, very extensive or more than required. When disclosing more than required, this can be seen as disclosing on a voluntary basis. Providing more information than required will enhance the quality of disclosure of IFRS 15. How the disclosure quality regarding to the voluntary and mandatory aspect of IFRS 15 is measured will be discussed further in the methodology section. As the first full adoption of IFRS 15 is taken into account, it should be assumed that there is not yet a general practical implementation among firms. That could result in differences between companies in the first application of the standard.
Different board structures
After having discussed the content of IFRS 15, we will continue with describing the different board structures that exist among European countries. As this research takes a corporate governance perspective, it is relevant to elaborate on board structures that exist among European firms, because there are differences in the structure of boards among European listed firms. Generally, a distinction can be made between one-tier boards and
two-10 tier boards. This paragraph will discuss the main differences and main similarities of the two board structures and will introduce the Audit Committee as a sub committee of the Board of Directors.
The control of managing directors of companies can be included in the board or can form a different committee: the Supervisory Board (Jungmann, 2006). When the monitoring role and the management role are together in one board, it is called a one-tier board. When the Management Board and the Supervisory Board are separated, it is called a two-tier board. The United Kingdom is an example of a country adopting the one-tier board system, and Germany and the Netherlands are examples of countries adopting a two-tier system (Jungmann, 2006). There are also some countries that give companies the opportunity to choose for themselves. Examples of these countries are France and Italy. The article of Jungmann (2006) equates non-executives in the one-tier system and Supervisory Board members in the two-tier system. These members have a comparable monitoring task. Xie et al. (2003) and Payne et al. (2009) mention that boards are charged with monitoring management to protect their shareholders’ interest, respectively are not involved in the day-to-day operations and have a monitoring and advisory role. A difference between non-executives and the members of the Supervisory Board is that the members of the Supervisory Board can appoint the members of the Management Board, while the shareholders of an one-tier board structure appoint executives and non-executives. An advantage of the two-tier system is that the control and management tasks are separated strictly, which lead to a more independent Supervisory Board. A disadvantage is that the Supervisory Board is not involved in decision-making and is not present at meetings of the Management Board, this results in information asymmetry between the Management Board and the Supervisory Board (Jungmann, 2006). For this research, only the monitoring function of the board is taken into account. Therefore, Board of Directors is defined as the members of the Supervisory Board of the two-tier system and the non-executive directors of the one-tier system.
A sub-committee of the Board of Directors is the Audit Committee. The Audit Committee is widely accepted in countries with one-tier boards and two-tier boards. (Collier and Zaman, 2005). The Audit Committee has to take an active role in monitoring the accounting and financial reporting policies and practices of the firm. Because of this, the Audit Committee is responsible for providing reliable information regarding accounting and reporting, and is responsible for the compliance of disclosure requirements (Berezinets et al., 2017). The European Commission states that Audit Committees have a key role in ensuring
11 high standards in financial reporting (European Commission, 2003). In addition to the Board of Directors which is already discussed in the previous paragraph, this research will also include the Audit Committee. The Audit Committee is together with the Board of Directors the main two monitoring committees within companies regarding the financial statements.
Agency Theory
We will continue with elaborating on the Agency Theory and the Upper Echelons theory as these theories can help explain why the variables can influence the disclosure quality of IFRS 15.
Agency Theory is concerned with the relationship between an agent and a principal, respectively a manager and an investor/owner (Eisenhard, 1989). The Agency Theory consist of two problems. First, the desires of a principal and an agent can conflict and it could be difficult or expensive for a principal to verify what the agent is doing. Second, there could be a different attitude towards risk taking by the principal and the agent (Eisenhard, 1989). The agent is able to make decisions concerning risks, while the principal actually bears the risks. Because the agent is not responsible for the risks, he is probably more risk-seeking to meet goals, while the principal is more risk-averse. Because it is expensive and difficult to verify the behavior and actions of the manager and information is not shared entirely, there exists information asymmetry between the agent and the principal. The existence of information asymmetry between managers and investors results in a demand for financial reporting and disclosure (Healy and Palepu, 2001; Kothari et al., 2009). The increased regulation concerning disclosure will diminish the information asymmetry existing between the company and the investors (Shi et al., 2013). Furthermore, the quality of disclosure is an important factor for investors to be able to make good decisions. The article of Brown and Hillegeist (2007) found that firms’ overall quality of disclosure is negatively associated with the average level of information asymmetry. Therefore, disclosures of good quality will reduce the incentives of investors to search for private information (Brown and Hillegeist, 2007). Corporate governance systems are developed to reduce agency problems (Islam et al., 2010). By corporate governance systems, monitoring mechanisms and evaluation procedures are meant. These systems help align the behavior of organization’s agents and the interest of the investors and shareholders. The Board of Directors and the Audit Committee, as discussed above, are regarded as corporate governance systems because these are the monitoring functions of companies. By proper monitoring and ensuring proper disclosures, they aim to reduce the information gap between
12 the company and investors and try to align the activities of the Board of Management with the preferences of the investors. Agency Theory stresses the importance of disclosures, as there is a gap between the information owned by managers and the information owned by owners/investors. Disclosures have the function to inform (potential) investors and thereby reduce the information asymmetry existing between firms’ management and investors. Dechow and Skinner(2000) found that firms are trying to avoid losses, tend to report increases in earnings, and try to meet analysts’ forecasts by making use of earnings management. High quality disclosures regarding revenue recognition are relevant to investors and other stakeholders as reported revenue can influence the stock price and will help (potential) investors to assess the financial performance of the entity. The Board of Directors and the Audit Committee are mechanisms in between investors and managers that can influence the information asymmetry by their monitoring and advising function.
Upper Echelons Theory
The Upper EchelonsTheory is the second theory included in this research and will help to address the relationship between age and disclosure quality. The Upper Echelons theory considers the organization as a reflection of its managers. The theory consist of two interconnected parts: personal interpretations influence the way executives act and these personal interpretations are a function of the experiences, values and personalities of executives (Hambrick, 2007). Therefore, the decision making process and their outcomes are influenced by the characteristics of managers (Hambrick & Mason, 1984). The theory suggests that the managers with the same personal characteristics have the tendency to make the same decisions because the characteristics they have in common will influence these values. They define this as a cohort: a group of individuals that have something relevant in common (Hambrick and Mason, 1984). Examples of the personal characteristics are age and gender. Following this theory, people of the same age will respond on issues more similar compared to people that are older or younger because, for example, they have different experiences and values. An article of Truett (1993) shows the relationship between age and conservatism in the US. The article found, in general, that older people are more conservative and risk-averse compared to younger people. Risk averse people are more likely to comply with regulation and therefore older people will be more intended to comply to regulation. This theory, as described above in the context of this research, contributes to the research because it provides the underlying framework for different decisions or outcomes by different managers. Managers with a different experience
13 or personal characteristics can respond different on circumstances, therefore they can influence the extent and quality of disclosure of IFRS 15. After having discussed the content of IFRS 15 and the two theories that are used in this research, we will continue with addressing the independent variables of this research.
Effect of women
As Board of Directors and the Audit Committee fulfil the monitoring function in companies, they try to diminish the information asymmetry existing between managers and owners of the companies. The article of Fernandez-Feijoo et al. (2012) finds that the more women on boards, the higher the disclosure on Corporate Social Responsibility (CSR). The reason they provide for this relationship is the fact that women improve monitoring by being more active on corporate boards and being more independent. Next to better monitoring, women on boards also improve the communication. A paper examining the relationship between gender diversity, measured by the percentage of women on the Board , and voluntary disclosure of Greenhouse gas also finds a significant relationship (Liao et al., 2015). Pucheta-Martínez et al. (2016) found a significant relationship between the percentage of female directors on the Audit Committee and the financial reporting quality of annual reports. They explain this relationship by the fact that a higher amount of women on the Board of Directors leads to a broader perspective and a more extensive decision making process (Terjesen et al., 2009). By these findings, it becomes clear that women on the Board of Directors can have an influence on different kind of disclosures. Explanations of this relationship can be found in the article of Kirsch (2018). Kirsch (2018) found that female directors are associated with among other things, more comprehensive disclosure activities. Furthermore, women seem to improve monitoring because they are more independent than men, as they are not part of ‘the old boys network’. Women also have other resources than men have, as they are a more ethical, are more risk-averse, and have a better long term view (Kirsch, 2018). The article of Bilimoria & Huse (1997) found that women may ask questions more freely, and therefore will speak up earlier when they are concerned. The article of Abad et al. (2017) found evidence that the participation of women on the Board of Directors also diminishes the information asymmetry. Regarding the previous literature, we can assume that the degree of women will positively influence the disclosure quality of IFRS 15. Although, there was some contradictory evidence found in previous literature. Social scientists found evidence that women who are attracted to higher positions, do have more similarities with male characteristics (Kirsch, 2018). However, the
14 positive effect of femininity on the Board of Directors and the Audit Committee is taken into consideration because the previous articles predominantly found a positive relationship between gender diversity and disclosure practices. As women are improving monitoring, have better communication skills, more freely ask questions, are more risk-averse, and are more independent, it can be assumed that they will strive for more compliance with legislation and a better communication with outsiders. As revenue is an important measure to assess the performance of companies, women on the Board of Directors will strive for a good disclosure of revenue recognition. Therefore, we will come to the following hypotheses for both the Board of Directors and the Audit Committee:
H1: The degree of women on the Board of Directors is positively associated with the disclosure quality of IFRS 15
H2: The degree of women on the Audit Committee is positively associated with the disclosure quality of IFRS 15
Effect of age
As women tend to be more risk-averse compared to men, older people tend to be more risk-averse compared to younger people (Vroom and Palh, 1971). They gave two explanations for this: development and socio-cultural change. An example of the development explanation is that the older people get, the more responsibilities they receive. The socio-cultural explanation is that for younger people it is more likely that they have been raised in an atmosphere of economic stability and welfare (Vroom and Pahl, 1971). The underlying theory could be the Upper Echelons Theory. Age is an observable Upper Echelon characteristic (Hambrick and Mason, 1984). Therefore, the age of a manager can influence his/her decisions concerning disclosure. Taking into account that older people will be more risk-averse and feel more responsible, older people will strive for more compliance with the regulation. Therefore, older members of the Board of Directors or the Audit Committee will strive for compliance with the mandatory IFRS 15 items and strive for additional (voluntary) disclosures. This results in more extensive disclosure, which will lead to better disclosure quality of IFRS 15. As this is the first year of adoption of IFRS 15 practical implementation of the requirements of IFRS 15 are not yet defined. As managers are not familiar and sure about the way of reporting, risk averseness of members of the Board of Directors or members of the Audit Committee will lead
15 to disclosures that are more comprehensive. The article of Said et al. (2013) found a significant relationship between the age of a chairperson and the extent of environmental disclosure. Another reason that age is expected to have a positive relationship with disclosure quality is that younger people have less experience regarding the provision of disclosure compared to older people. Another possible explanation is that the quality of disclosures could be lower when young board members are involved because of their limited experience; they could be less informed about what has to be incorporated and how disclosures have to be reported. As this is the first year of adoption of IFRS 15, more experienced directors could apply their knowledge and experience of former disclosure requirements and know better what they have to do. Applying these two theories on disclosure quality, older board members would be more willing to comply to the requirements of IFRS 15, as they are more risk-averse. Because of this risk-averseness they are unwilling to bear the risk of not complying with the new regulation and therefore striving to enhance the quality of the disclosures. Furthermore, they are more experienced and therefore have more knowledge of disclosures.
H3: The average age of the members of the Board of Directors is positively associated with the disclosure quality of IFRS 15
H4: The average age of Audit Committee members is positively associated with the disclosure quality of IFRS 15
Role of analysts
Next to the role femininity and age of the Board of Directors and the Audit Committee, the role of analysts is taken into account in this study. Healy and Palepu (2001) define financial analysts as information intermediaries in financial markets who collect information from public and private sources, evaluate firms’ performance, and make forecasts. So, financial analysts use information of annual reports to make earnings forecasts to serve investors. Gathering information for investors themselves could be very time consuming when all potential investors have to collect and analyze the information available to make a good forecast for the future earnings of a potential firm to invest in. Therefore, Analysts’ jobs are created due to the information asymmetry existing between the company and the (potential) investors. Their forecast quality is dependent on the quality of the financial reporting of firms (Hamrouni et al., 2017). More requirements regarding disclosure of judgement regarding revenue recognition should lead to a better understanding of the company’s revenue for investors. Analysts need
16 reliable information to make good forecasts and to provide investors with recommendations. The article of Fuertes-Callén et al. (2014) states that analysts can put pressure on companies to disclose information. When more analysts are involved in the company, it can be expected that this lead to an increase in the disclosure quality of IFRS 15 because the higher the analysts’ coverage, the more they can put pressure on the company to provide an annual report of high quality. Members of the Board of Directors who are more risk averse are likely to be more sensitive to the pressure of the analysts. The article of Hamrouni et al. (2017) recommends managers to improve the quality of their annual reports to meet expectations of the market, then they can attract analysts who can convince investors to invest. Bhat et al. (2006) finds that analysts consider corporate governance characteristics as important because these characteristics enables them to consider the credibility of the firm’s disclosures. We expect that the number of analysts following the firm can strengthen the relationship between the degree of women on the Board of Directors because as stated earlier, women want to comply to rules and expectations, as they are more risk-averse. This reason also applies to the strengthening impact of analysts’ coverage on the relationship between the average age of the members of the Board of Directors and the Disclosure Quality of IFRS 15.
H5: Analyst coverage is positively associated with the disclosure quality of IFRS 15
H6: Analyst coverage strengthens the relationship between the degree of women in the Board of Directors and the disclosure quality of IFRS 15
H7: Analyst coverage strengthens the relationship between the average age in the Board of Directors and the disclosure quality of IFRS 15
17
IV. RESEARCH METHODOLOGY
In this chapter we will discuss the method which is used to examine the influence of gender, age and analysts’ following on the disclosure quality of IFRS 15. First, we start with discussing the sample selection. Secondly, the dependent and independent variables are addressed. Thirdly, the control variables are explained. Thereafter, the statistical model is presented. Finally, an overview of the variables included in this research is presented.
Sample selection
This research is executed by taking a qualitative approach to quantitative research. We first assessed the quality of disclosures of IFRS 15. Consequently, we performed statistical analyses on that data. The sample consist of 52 European listed companies who are under IFRS regulation mandatory. The companies are derived from the STOXX600. This index is relevant to use as it contains large, medium and small sized firms among 17 European countries. As different sized companies among different European countries are included in the sample, the research strives to be able to form generalizable results. To select a company, we had a look at the Key Audit Matters (KAM). A condition to be included was that ‘IFRS 15’ or ‘revenue
recognition’ had to be mentioned as a KAM. The reason for this condition is to be certain that
the companies are affected by the adoption of IFRS 15 and thus are relevant for this research. Furthermore, we choose to only include companies operating in the telecommunications industry, utilities industry, construction & materials industry, and technology industry. Roozen and Pronk (2018) examine the potential impact of IFRS 15 on different industries and found that the telecommunications industry and utilities industry are the main industries affected by IFRS 15. These industries are followed by the technology sector. Unfortunately, they do not distinguish the construction and materials industry but given the nature of their revenue and their changing revenue standard (IAS 11), we can assume that this sector will also seriously be affected by IFRS 15.
Dependent variable: Disclosure Quality of IFRS
This section will elaborate on the dependent variable, disclosure quality of IFRS 15, the following section will discuss the independent variables. A systematic overview of the variables is presented in table 1.
18 The data regarding the disclosure quality of IFRS 15 is extracted from annual reports of European listed firms as the corporate annual report is the primary source of corporate information disclosure (Adina and Ion, 2008; Healy and Palepu, 2001). These annual reports are extracted from the official websites of the companies.
We had to compose our own disclosure index to determine the quality of the disclosure regarding IFRS 15 because there was no database available with the right information. Although the topic of IFRS 15 is new and no disclosure index concerning disclosure of IFRS 15 was composed earlier, a lot of research is done regarding disclosure indexes. Many of these studies used a disclosure index and award a score by dividing the number of items disclosed on the applicable total number of possible variables as in the article of Camfferman and Cooke (2002), Boubaker et al. (2015), and Eng and Mak (2003) did that as well. This a very quantitative approach to disclosure. We composed a checklist derived from translating the legislation items into index items. The standard was the basis for making the disclosure index. Almost every item of the standard is included in the disclosure index, which makes it a comprehensive index. In many cases, the required items could be presented in a quantitative way and a qualitative way. For the quality of the disclosure of IFRS 15, it would be beneficiary to include both ways of disclosure. Some items do require to disclose both. For applicable index items, we decided to award points when a company disclose the item quantitatively and when a company disclose the item qualitatively, although it could be that it was not required. Because, a qualitative explanation of a quantitative amount would enhance the quality of disclosures. This add the voluntary aspect to the disclosure index. We have to consider this when interpreting the score of IFRS 15.
When a topic was clearly mentioned in the annual report, it received one point for that part, otherwise zero. Therefore, for some items, a company could receive two points for one index item, one for the quantitative aspect and one for the qualitative aspect. Some items contain multiple items that must be disclosed. When this was the case, we did not take into account a quantitative and a qualitative aspect, as these were relatively small disclosures. When a part is not applicable for the company and the company mentions that, it will be awarded a ‘not applicable’. The maximum score will be reduced by the value of these non applicable items.
To make a clear distinction, there is a difference between rating and weighting. Rating is the process where an item receives points for the extent to which a firm disclose an item in line with the prescribed extent. The weighting process is about the importance of items, some
19 items are more important to users of the annual report. There is some discussion about weighting certain items. The weighting process of index items is a subjective process as the determination of weight is based on the perceived relative importance by the researcher. The article of Robbins and Austin (1986) finds that not applying the weighting of index items does not materially affect the results of the study compared to weighting index items. Therefore, we will not include a weighting system. To award the final score, the number of points is divided by the maximum points a company could be given, the non-applicable items are deducted from the maximum score. This measurement of the disclosure score is used in several articles (i.e. Shalev, 2009). The Disclosure Index that is used is added as an appendix (Appendix I).
All the data was collected by four students of the University of Groningen writing their Master Thesis in Accountancy or Controlling. A potential danger of different persons collecting data, as already mentioned as a disadvantage of different analysts, is a different interpretation of information. This risk is mitigated by establishing several meetings with regard to the data collection process, like testing on beforehand if the four students awarded points in the same way by using pilot annual reports, and peer evaluations. Furthermore, we created a data collection manual, and made clear agreements regarding the measurement of variables and the awarding of points. We also created overlap between companies that were divided among us. Differences in the assignment of scores was discussed extensively. We made use of the latest available data concerning disclosure quality of IFRS 15 since this regulation will be applicable for the annual year beginning on or after the first of January 2018, information became available on/after the 1st of April 2019.
Independent variables
Below, the independent variables are addressed separately. Table 1 presents an overview of all variables included in the research.
20 Gender
The degree of women on the Board of Directors (WOMSB) is measured by a percentage, dividing the amount of women participating in the Board of Directors by the total amount of members of the Board of Directors. The amount of women on Board of Directors is determined in the same way. The degree of women on the Audit Committee (WOMAC) is measured by dividing the amount of women participating on the Audit Committee by the total amount of members of the Audit Committee For both variables, the information was extracted from the governance paragraph in the annual report. We started collecting information about gender in the Board of Directors and Audit Committee by making use of the annual report. When it could not be distracted from the annual report, we used public information (i.e. LinkedIn).
Age
For measuring the variable age, we made use of the average age of members of the Board of Directors (AGESB) and the Audit Committee (AGEAC). The article of Truett based the age on the year of completion of the article. As we examine the influence of age on the disclosure quality of IFRS 15 of the annual year 2018, the age is also based on 2018. We did not exactly look to the birthday of everyone, because it was not always given and it would be very time-consuming and of less relevance. When collecting the information, we preferred the date of birth above the age of a person. This was because the date of birth is more reliable as it does not change during the years. We tried to find the year of birth or the age of Board members and Audit Committee members in the annual report, often it was listed in the governance paragraph. When it was not available, we used the database of Compustat. When the year of birth or the age was not applicable, we took the average age of the other board/committee members. This was the case for three companies. We did not apply a requirement of a minimum amount of observations for a company but we had a look at the average of that company and the average age of the sample to assess if it was a realistic average. As there was little difference in the average age between Board of Directors and Audit Committees, no further adaptions had to be made. This was the case for all three companies.
21 Analysts
The number of analysts (ANALYSTS) is extracted from a database. The Institutional Brokers’ Estimate System (I/B/E/S) will be used for this. This database collects opinions of individual analysts and gives a summary of the opinions of the analysts. The number of unique analysts following a firm can be extracted from this database. The number of analysts is more often used as a proxy for analyst coverage (i.e. Allen et al., 2016). When making use of this database, a long list of analyst codes became available for each company of our sample. Because some analysts make several forecasts during the year, some analysts’ codes existed several times for one company. To extract the number of unique analysts during 2018 we counted the amount of unique analysts’ codes for each company for 2018.
Control variables
In this research, we control for firm size, debt, and profitability and industry because prior studies (i.e. Eng and Mak, 2003) found that these variables could influence the level of disclosure quality. Control variables have been added as they can unintentionally influence the dependent variable so there have to be controlled for that influence. They can all be derived from the annual report. Below, the control variables will be discussed briefly.
Size
Contrary to the article of Eng and Mak (2003), size is measured by the log of the total sales instead of the log of the market value of the firm. There have to be controlled for size as size is found to positively influence the quality of disclosure by a company (Chow and Wong-Boren, 1987; Lang and Lundholm, 1993). A logarithm is used for this variable, as the sales of companies can be enormous.
22 Debt
The article of Eng and Mak (2003) found that the amount of debt is negatively associated with disclosure. As debt can influence the disclosure of a company, there have to be controlled for it, therefore it is included as a control variable. In this research, as in the research of Eng and Mak (2003), debt is measured by the leverage of the company. Leverage is measured by dividing total liabilities by total assets.
Profitability
The ability of a firm to comply with the disclosure requirements would be lower when that firm is less profitable (Ettredge et al, 2011). This control variable is also included in the article of Ettredge et al. (2011). Although they measure profitability in terms of chance on bankruptcy, we take a more simple approach and measure profitability by dividing the net profit by the total assets.
Industry
Furthermore, there is controlled for industry type, as there can be differences between the disclosure rates between industries that have other causes. As there are four industries included in this research, we had to work with dummies. When a company is in the certain industry, it is rewarded a one, otherwise a zero. Only three of the industry dummies are included in the regression analysis to prevent multicollinearity problems.
Empirical model
The ordinary least squares method is used to test the different hypotheses. Model 1 is the basic model and contains only the dependent variable and the control variables. Thereafter in each model, hypotheses are tested separately by adding the corresponding variables. Including all variables into one model (model 10) results in the following statistical formula:
SCORE = α + 𝛽1* WOMSB + 𝛽2* WOMAC + 𝛽3* AGESB + 𝛽4* AGEAC + 𝛽5* ANALYSTS + 𝛽6* LOGSALES + 𝛽7* LEVERAGE + 𝛽8 * ROA + 𝛽9* INDtel + 𝛽10* INDuti + 𝛽11* INDcon + 𝛽12* INDtec + 𝛽13*(WOMSBcentr*ANALYSTScentr) + 𝛽14* (AGESBcentr*ANALYSTScentr) + ε
23 Variable overview Kind of variable Name of variable Definition Source Dependent Variable
SCORE Total scores of the themes divided by the possible score (total score minus scores that are not applicable)
Hand collected data form annual reports extracted form official websites. Independent
Variable
WOMSB Total number of women on the Board of Directors / total amount of Board of Directors members.
Annual report
WOMAC Total number of women on the Audit Committee / total amount of Audit Committee members
Annual Report
AGESB Average age of Board of Directors.
Annual Report, if not applicable;
Compustat Database AGEAC Average age of Audit Committee
members
Annual Report, if not applicable;
Compustat Database ANALYSTS Number of different analysts
following the firm during 2018
I/B/E/S database Moderating variables WOMSB* ANALYSTS WOMSBcentralized* ANALYSTScentralized Computed in SPSS AGESB* ANALYSTS AGESBcentralized* ANALYSTScentralized Computed in SPSS Control variables
LOGSALES Log of total sales Annual Report LEVERAGE Total assets / Total liabilities Annual Report PROFIT% Net profit / Total sales Annual Report INDtel Company operating in
telecommunications industry (Dummy variable)
STOXX600
INDuti Company operating in utilities industry (Dummy variable)
STOXX600 INDcon Company operating in
construction and materials industry (Dummy variable)
STOXX600
INDtech Company operating in technology industry (Dummy variable)
STOXX600
24
V. RESULTS
We will continue with discussing the results of this research. First, we will address the descriptive statistics. Second, we will elaborate on the best practices found in the annual reports during the phase of gathering the data. Third, we will introduce the correlation analysis, Fourth, we will move on with the multicolineairity analysis. Finally, we will come to the regression analysis.
Descriptive statistics
We will start with the general characteristics of the sample to get an impression of the countries, industries and companies involved in this research. In this section, we will address the average numbers and the minimum and maximum values of the dependent variable and independent variables. The total sample consist of 52 companies. More information of the companies and their disclosure scores is provided in Appendix II.
Frequency Percentage
Telecommunications 15 28.8%
Technology 18 34.6%
Construction & Materials 8 15.4%
Utilities 11 21.2%
Total 52 100%
Table 2: Sample by Sector
In table 2, the sample is divided in different industries. This figure gives an overview of the distribution of the sample amongst industries. The two major industries in the sample consist of companies operating in the technology industry and telecommunications industry.
25
Figure 1: Sample by Country
Figure 1 shows the sample divided amongst countries. Most of the firms are based in Western Europe (Germany, the Netherlands, Great Britain, and Ireland) and Southern Europe (France, Italy, and Spain). Some firms find their origin in northern part of Europe. There are no firms included in the sample that are located in the Eastern part of Europe.
N Minimum Maximum Mean Std. Dev.
IFRS Score 52 13.3% 75.5% 48.153% 15.196%
Table 3: IFRS score
Table 3 presents the disclosure score of IFRS 15. The average score of the disclosure quality of IFRS 15 is 48.15%, the minimum score is 13.3% and the maximum score is 75.5%. The average compliance with the new IFRS 15 legislation is 48.15%, which seems to be low. Although we have to take into account that there is also a voluntary part included in the index. For some legislation items we provided the opportunity to award one point for quantitative disclosure and one point for qualitative disclosure while disclosing one of both is sufficient. The maximum score is 75.5%, so there are no companies that received the full score.
Industry IFRS Score
Telecommunications 51.6%
Utilities 40.1%
Construction & Materials 48.9%
Technology 49.8% 0 2 4 6 8 10 12 14 CH DE DK ES FI FR GB IE IT NL NO SE
Sample by Country
26
Table 4: IFRS score by industry
Table 4 presents the average disclosure score of IFRS 15 by industry. The average score of disclosure quality of IFRS 15 is the lowest in the utilities sector and the score of disclosure quality of IFRS ranks the highest in the telecommunications industry.
N Minimum Maximum Mean Std. Dev.
%WOMSB 52 0.00 66.67 33.29 12.64
%WOMAC 52 0.00 75.00 37.12 18.41
AGESB 52 51.5 69.8 59.3 4.10
AGEAC 52 50.3 67.3 58.9 4.30
ANALYSTS 52 6 36 20.3 7.95
Table 5: Independent variables
In table 5, information about the independent variables is presented. In the sample, the Board of Directors consist of 33.3% of women on average. The average age of directors is 59.3 years. The Audit Committee has on average more women and slightly younger members, the means are respectively 37.1% and 58.9 years. The average number of analysts following a firm of the sample is 20.3 for 2018 and has a range between 6 and 36.
Board of Directors Audit Committee
Average amount of members 10.5 4.3
Average amount of women 3.7 1.7
Average percentage of women 33.3% 37.1%
Average Age 59.3 58.9
27 In table 6, there is a comparison between the number of members, the number of female members, the percentage of females, and the average age in the Board of Directors and the Audit Committee. The difference in age between the members of the Board of Directors and members of the Audit Committee is relatively small.
Best practices
Next to a statistical analysis, we include the best practices that we found during the research process. We noted the best practices when assessing the disclosure quality of IFRS 15 in the annual reports. For each company mentioned in this paragraph, an example is given in appendix IV. Overall, ATOS received the highest score (75.5%). They establish a very direct relationship with the standard. The required disclosure items are clearly reflected in the annual report. This makes the information regarding disclosure of IFRS 15 easy to find. KPN also has a high disclosure quality score of 73.9%. KPN is one of the few firms that made a clear distinction between revenue from contracts with customers and other revenue. Hochtief had a disclosure score of 65.2%, a noteworthy best practice of Hochtief is their extensive division in different kind of remaining performance obligations. Screenshots of examples of best practices are added in Appendix IV. It can be concluded that there are various practical implementations of the standard among companies.
Correlation analysis
To verify if there exists correlation between the variables, a correlation analysis has been performed. When there is high correlation between two variables, this could bias the results of the regression analysis. The correlation coefficients are presented in the correlation matrix in table 7a and table 7b. After the table, we will elaborate further on the implications of the correlation table.
28
SCORE %WOMSB %WOMAC AGESB AGEAC ANALYSTS
SCORE 1 %WOMSB -.164 1 %WOMAC -.147 .460** 1 AGESB .194 -.344* -.302* 1 AGEAC .098 -.087 -.095 .751** 1 ANALYSTS .074 .239 .181 -.337* -.236 1 ROA -.124 -.094 -.111 .092 .142 -.177 LOGSALES .188 .359** .153 -.090 -.005 .525** LEVERAGE -.077 .206 -.043 -.032 .041 .192 INDtel .147 .069 -.004 -.323* -.211 .347* INDuti -.276* .003 -.128 -.020 -.100 .165 INDcon .022 .168 -.199 .099 .222 -.247 INDtech .080 -.196 -.037 .250 .118 -.285*
**. Correlation is significant at the 0.01 level (2-tailed) *. Correlation is significant at the 0.05 level (2-tailed)
29
ROA LOGSALES LEVERAGE INDtel INDuti INDcon INDtech
SCORE %WOMSB %WOMAC AGESB AGEAC ANALYSTS ROA 1 LOGSALES -.389** 1 LEVERAGE -.310* .406** 1 INDtel -.112 -.026 .049 1 INDuti -.108 .332* .338* -.330* 1 INDcon -.190 .155 .155 -.271 -.221 1 INDtech .344* -.337** -.454** -.463** -.337** -.310* 1
**. Correlation is significant at the 0.01 level (2-tailed) *. Correlation is significant at the 0.05 level (2-tailed)
30 Some logical correlations can be extracted from the correlation matrix in table 7a and table 7b. The significant correlations that are discussed here are limited to the correlations between the independent variables. There is a significant relationship between the percentage of women on the Board of Directors (%WOMSB) and the percentage of women on the Audit Committee (%WOMAC) ( r = .460, p = .001). This can be explained by the fact that the Audit Committee is a subcommittee of the Board of Directors, members who are in the Board of Directors could also be part of the Audit Committee. As the percentage of women on the Board of Directors is high, it is likely that the percentage of women on the Audit Committee is also high. The same explanation can be given for the variable age, when the average age of the Board of Directors (AGESB) is high, it is likely that the average age of the members of the Audit Committee (AGEAC) is also high ( r = .751, p = .000). There exist a significant negative relationship between the percentage of women on the Board of Directors (%WOMSB) and the average age in the Board of Directors (AGESB) ( r = -.344, p = .013). This can explained by the fact that more women enter the boardroom during the last years due to several reasons like amongst others, the gender quota. These women are often young. Therefore, the more (young) women on the Board of Directors (%WOMSB), the lower the average age of member of the Board of Directors (AGESB). Furthermore, there is a significant correlation found between the average age of the members of the Board of Directors and the percentage of women on the Audit Committee ( r = -.302, p = .030). Contradictory, there is no significant relationship found between the number of women on the Audit Committee and the average age of the Audit Committee members. Another significant negative correlation is found between the average age of Board of Directors (AGESB) and the amount of analysts following the firm (ANALYSTS) ( r = -.337, p = .014). In the first instance, it seems to be a strange relationship. One possible explanation could be that young firms possibly have younger members on the Board of Directors, and young firms are more attractive to analysts.
31
Multicollinearity analysis
A multicollinearity analysis has been performed because in the correlation matrix, there is a correlation coefficient that exceeds 0.7. Two independent variables could cause multicollinearity as the correlation coefficient exceeds 0.7 (Hair et al., 2005). This is the case for the correlation between AGESB and AGEAC (correlation coefficient is 0.751). As the existence of multicollinearty may influence the results of the regression, a multicollinearty test is executed. The Variance Inflation Factor (VIF) is output of the multicollinearity test and is an indicator of the effect that the other independent variables have on the standard error of a regression coefficient (Hair et al., 2005). A large VIF indicates high multicollinearty. A generally accepted level of multicollinearty is a VIF below 10 (Hair et al., 2005). The results of this multicollinearity analysis were presented in table 8a and table 8b. As can be concluded from these tables, there are no serious problems with multicolinearity. The maximum found VIF value is 3.291, this value indicates multicollinearty but is not problematic. Consequently, all variables can be included in one model.
32
VIF: %WOMSB %WOMAC AGESB AGEAC ANALYSTS
%WOMSB 1.437 1.509 1.596 1.691 %WOMAC 1.303 1.420 1.462 1.433 AGESB 3.042 3.158 1.431 3.176 AGEAC 2.687 2.725 1.200 2.659 ANALYSTS 2.035 1.994 2.010 2.042 ROA 1.364 1.358 1.357 1.315 1.374 LOGSALES 2.107 2.258 2.253 2.261 1.614 LEVERAGE INDTEL INDUTI INDCON 1.466 1.868 1.874 1.718 1.465 1.822 1.858 1.699 1.496 1.752 1.868 1.680 1.485 1.847 1.878 1.668 1.497 1.729 1.879 1.590
33
VIF: ROA LOGSALES LEVERAGE INDTEL INDUTI INDCON
%WOMSB 1.613 1.519 1.600 1.633 1.627 1.632 %WOMAC 1.457 1.476 1.450 1.445 1.463 1.465 AGESB 3.236 3.275 3.291 3.088 3.270 3.219 AGEAC 2.629 2.754 2.739 2.729 2.755 2.680 ANALYSTS ROA 2.034 1.455 1.273 2.042 1.345 1.891 1.329 2.041 1.377 1.890 1.334 LOGSALES 2.089 2.233 2.244 2.149 2.149 LEVERAGE INDTEL INDUTI INDCON 1.458 1.799 1.876 1.660 1.475 1.850 1.785 1.630 1.788 1.652 1.636 1.433 1.404 1.520 1.315 1.394 1.426 1.425 1.653 1.561
34
Regression analysis
After performing a correlation test and a multicollinearity analysis, a regression analysis is performed. As already mentioned in the method section, the regression analysis is executed by making use of the ordinary least squares method. The hypotheses are divided in models. A regression analysis is performed for all models separately. Model 1, the basic model, tests the relationship of the control variables with the dependent variable (SCORE). In the second model, the percentage of women on the Board of Directors (WOMSB) is added to the basic model. Model 3 consists of the basic model and the percentage of women on the Audit Committee (WOMAC). In the fourth model, the basic model is supplemented with the average age of the Board of Directors (AGESB). The percentage of women on the Audit Committee (WOMAC) is added to the basic model in model 5. Model 6 consist of the basic model and the number of analysts (ANALYSTS).
The moderating relationships are included in Model 7 and model 8. To be able to include a moderating relationship, the independent variables had to be centralized first. This is done by subtracting the average of the variable from each observation. To calculate the moderating variable, the centralized variables had to be multiplied by each other. The basic model had to be complemented with the computed variable together with the centralized variables. Therefore, in model 7, the percentage of women on the Board of Directors (WOMSB), the number of analysts (ANALYSTS) and the corresponding new variable (WOMSBcentr*ANALYSTScentr) are added to the basic model. In model 8, the average age of the members of the Board of Directors (AGESB), the number of analysts (ANALYSTS) and the second computed variable (AGESBcentr*ANALYSTScentr), and the basic model are included in the model. In model 9, all independent variables are added to the basic model. Model 10 consists of the basic model, all independent variables and the moderating variables.
35
Model 1 Model 2 Model 3 Model 4 Model 5
Intercept -24.409 -42.284 -26.928 -86.352 -46.129 (.526) (.268) (.472) (.090) (.332) %WOMSB (+) -.306** (.037) %WOMAC (+) -.262* (.062) AGESB (+) .263* (.067) AGEAC (+) .115 (.430) ANALYSTS (+) ROA -.087 -.055 -.102 -.093 -.108 (.569) (.710) (.494) (.532) (.490) LOGSALES .343** .457*** .395** .360** .337** (.039) (.008) (.017) (.027) (.043) LEVERAGE -.085 -.057 -.117 -.107 -.100 (.598) (.713) (.459) (.500) (.539) INDtel -.018 .010 -.018 .085 .008 (.914) (.949) (.910) (.620) (.964) INDuti -.404** -.430** -.438** -.364** -.383** (.030) (.017) (.017) (.046) (.042) INDcon -.129 -.092 -.090 -.118 -.144 (.438) (.569) (.579) (.466) (.393) WMNSBANALYSTS (+) AGESBANALYSTS (+) Observations 52 52 52 52 52 R-Square .188 .266 .251 .248 .200 Adj. R-Square .080 .149 (.131) .128 .072 F-value 1.736 2.273 2.103 2.074 1.567
***. significant at the 0.01 level (2-tailed) **. significant at the 0.05 level (2-tailed) *. significant at the 0.10 level (2-tailed)
36
Model 6 Model 7 Model 8 Model 9 Model 10
Intercept -36.762 -58.613 -34.041 -73.403 -.82.561 (.396) (.330) (.464) (.171) (.127) %WOMSB (+) .333* -.178 -.198 (.032) (.295) (.243) %WOMAC (+) -.121 -.154 (.453) (.342) AGESB (+) .289* .235 .477* (.076) (.332) (.090) AGEAC (+) -.105 -.248 (.634) (.297) ANALYSTS (+) -.120 -.066 -.047 -.019 .039 (.521) (.720) (.804) (.922) (.839) ROA -.083 -.046 -.097 -.061 -.046 (.592) (.759) (.524) (.695) (.764) LOGSALES .412** .476** .388** .464** .415** (.040) (.020) (.050) (.024) (.049) LEVERAGE -.088 -.075 -.118 -.090 -.138 (.587) (.637) (.467) (.582) (.398) INDtel .020 .068 .108 .072 .172 (.909) (.701) (.553) (.689) (.364) INDuti -.400** -.409** -.384** -.416** -.430** (.033) (.026) (.043) (.027) (.030) INDcon -.157 -.060 -.147 -.071 .017 (.366) (.740) (.401) (.684) (.930) WMNSBANALYSTS (+) -.121 (.420) AGESBANALYSTS (+) .089 (.570) Observations 52 52 52 52 52 R-Square .196 .281 .255 .307 .360 Adj. R-Square .068 .127 .095 .116 .140 F-value 1.529 1.825 1.595 1.609 1.641 ***. significant at the 0.01 level (2-tailed)
**. significant at the 0.05 level (2-tailed) *. significant at the 0.10 level (2-tailed)
