The influence of Integrated Reporting on Firm Value
An Application to the European Banking industryName: Levi Bijlmakers Student number: 11421347
Thesis supervisor: ir. drs. A.C.M. de Bakker Date: 28 January 2018
Word count: 12643
MSc Accountancy & Control, specialization Accountancy Faculty of Economics and Business, University of Amsterdam
Statement of Originality
This document is written by Levi Bijlmakers, who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
In this paper, the effects of the issuance of an integrated report on firm value are examined for a selection of 50 European banking groups for the period 2010 until 2016. As current literature argues that earnings quality influences the positive relationship between integrated reporting and firm value, discretionary accruals (as a proxy for earnings quality) are added as a moderating variable, measured using the Modified Jones Model. The results show that integrated reporting is not significantly related to firm value. Earnings quality strengthens the positive relationship between integrated reporting and firm value, but this result should be handled with caution because of its insignificance. Further evidence is provided in this paper for the positive relationship between profitability and firm value. Finally, there is a higher probability found that large organizations issue integrated reports than small organizations.
Table of Contents 1 Introduction ... 6 1.1 Background ... 6 1.2 Research Objective ... 7 1.3 Research Question ... 8 1.4 Relevance ... 8 1.5 Research Design ... 9 2 Theoretical framework ... 10 2.1 Integrated Reporting ... 10 2.2 Agency Theory ... 11 2.3 Firm Value ... 12
2.3.1 Definition of firm value ... 12
2.3.2 Measuring firm value ... 13
2.4 Determinants of firm value ... 14
2.4.1 Earnings Quality ... 14
2.4.2 Other determinants of firm value ... 16
2.5 Hypotheses Development ... 16
3 Methodology ... 18
3.1 Conceptual Model ... 18
3.2 Regression Model for Hypothesis 1 ... 18
3.2.1 Dependent variable ... 19
3.2.2 Independent variables ... 19
3.3 Regression Model for Hypothesis 2 ... 20
3.3.1 Dependent variable ... 20
3.3.2 Independent variables ... 20
3.4 Regression Model for Hypothesis 3 ... 21
3.4.1 Dependent variable ... 21
3.5 Regression Model for Hypothesis 4 ... 21
3.5.1 Dependent variable ... 21
3.5.2 Independent variables ... 22
3.5.3 Moderating variable ... 22
3.6 Modified Jones Model ... 22
3.7 Summary of Variables ... 23
4 Data ... 25
4.1 Data Collection ... 25
4.1.1 Integrated reporting data ... 25
4.1.2 Other data ... 26
4.1.3 Discretionary Accruals ... 26
4.1.4 Balanced panel strategy ... 27
4.2 Descriptive Statistics ... 28 4.2.1 Core Statistics ... 28 4.2.2 Correlation Matrix ... 29 4.2.3 Multicollinearity ... 30 5 Results ... 31 5.1 Results regression 1 ... 31 5.2 Results regression 2 ... 32 5.3 Results regression 3 ... 34 5.4 Results regression 4 ... 36
6 Conclusions and limitations ... 38
6.1 Summary and conclusions ... 38
6.2 Limitations ... 40
6.3 Suggestions for further research ... 40
References ... 42
1 Introduction
1.1 Background
In the past few years, criticism on the way organizations report their information increased. Investors are not just interested in financial performance anymore; hence their interest in non-financial performance of companies has grown significantly. Merely reporting non-financial information does not meet the needs of all stakeholders, as it covers just a part of the business activities and ignores the social and environmental impact organizations make (Bernardi & Stark, 2015; Adams et al, 2011). As mentioned by Bouma et al (2001), “the existence of many companies will depend either on the continued availability of certain natural resources or their ability to adapt and reinvent themselves”, which covers the importance of the social and environmental ‘capitals’ for organizations.
Initially, the solution proposed by companies was to separately disclose non-financial information in sustainability or corporate social responsibility reports, in addition to their annual reports (Cheng et al, 2014). However, the link between the separate social and environmental reports and the financial reports was unclear. “To have a real impact, these separate reports need to be integrated with each other, thereby demonstrating that the company has a sustainable strategy based on a commitment to corporate social responsibility that is contributing to a sustainable society that takes into account the needs of all stakeholders, of which shareholders are one type” (Eccles & Krzus, 2010). This created the need for an integrated reporting system, a need that has been covered by an initiative of the IIRC (International Integrated Reporting Committee). The purpose of an integrated report is to combine all information relevant for stakeholders into one report, in which financial and non-financial information is interrelated (Ioana & Adriana, 2014).
Adoption of integrated reporting within Europe starts to increase. Currently, about 1,500 global companies started adopting it, including 36 top companies in Spain and a third of all listed companies in The Netherlands. France is even targeting to have all top listed companies adopting integrated reporting within three years (Howitt, 2017). However, some industries struggle with the implementation of their integrated reports, including the banking industry (IIRC, 2015).
Integrated reporting is not particularly different for banks than for other organizations. However, banks traditionally mainly aim to report on their financial performance (IIRC, 2015).
During the financial crisis, the market-to-book ratio (measure of firm value) decreased substantially, implicating a loss of confidence in banks by society (IIRC, 2015). By increasing efforts of banks to reinvent their business models, other ‘capitals’ became more important. The integration of those capitals and reporting on that led to a recovery of the market-to-book ratio (IIRC, 2015). Besides, the external environment puts pressure on banks to report on non-financial performance. Under CERCLA (Comprehensive Environmental Response, Compensation, and Liability Act) for example, banks could be held directly responsible for the environmental pollution of clients (Bouma et al, 2001). For these reasons, integrated reporting is very important for banks to satisfy all stakeholders, including shareholders. Within this paper, the efficient market hypothesis plays a central role, which states that a capital market reflects all information available (Malkiel, 1992). The assumption that is made in this paper states that integrated reports include new information and the information disclosed before the release of the integrated report is known to its users. Based on this efficient market hypothesis, an organization’s commitment to greater disclosure in the form of an integrated report is expected to decrease information asymmetry, which in turn decreases the cost of capital (Leuz & Verrecchia, 2000). Atkins & Maroun (2015) state the following about this causation (which basically aims at IC, meaning intellectual capital):
“Reducing IC information asymmetry leads to a lower weighted average cost of capital (WACC) and higher market capitalization, because IC information creates trustworthiness with stakeholders, promotes a long-term perspective, and has use as a marketing tool. In this way, preparing a " good " integrated report not only offers better quality information to the investor community, but actually signals that the respective organizations are taking steps to meet the information needs and expectations of stakeholder groups. Thus, reducing IC information asymmetry aligns with the goals of a prospectus – to influence investors to purchase the shares on offer and become owners.”
This would mean that generally, the demand of shareholders for stocks of banks using integrated reporting would increase, which would thus increase the firm value of those banks. Does the value of banks that adopt integrated reporting increase more than the value of banks that do not adopt integrated reporting? This relationship, in a balanced European environment, is the subject of study in this paper.
1.2 Research Objective
The objective of this paper is to provide insight into the impact of integrated reporting on firm value, in order to contribute to existing knowledge in the European banking industry.
1.3 Research Question
The research question formulated in this paper is the following:
Has the adoption of integrated reporting led to an increase in firm value in the European banking sector? In order to answer this research question, the following sub-questions could be asked: - What is integrated reporting?
- What is the underlying theory of integrated reporting? - What is firm value?
- How is firm value measured?
- What other factors influence firm value? These sub-questions are answered in chapter 2.
1.4 Relevance
Existing literature about this topic is discussed in chapter 2. While the relationship between integrated reporting and financial performance has been studied in existing literature, current research does not focus on the outcomes on capital markets as a result of integrated reporting. Churet and Eccles (2014) recommend to “investigate the possibility that the practice of integrated reporting could come to be viewed by investors as predictors of superior future financial performance” for further research. Also, Bernardi and Stark (2015) suggest to further investigate “the impact of integrated reporting on bid-ask spreads, trading volume, return volatility (as measures of information asymmetry)”. This paper aims to fill this research gap and explores the relationship between integrated reporting and firm value.
Investors and other users of financial statements increasingly express their need for disclosure of non-financial information. Richard Howitt, CEO of the International Integrated Reporting Council (IIRC), stated on 10 July 2017 that 6.000 European companies are going to change their way of reporting during the next 12 months (Howitt, 2017). It is even thought that integrated reporting will eventually become the international corporate reporting norm, according to Humphrey et al (2015). As the AFM (Dutch Authority for Financial Markets) states in their report ‘AFM in balans 2016’ of November 2016, the amount of organizations that start reporting integrated indeed increases, but most companies still do not do it. It also states that investors and analysts do not specifically ask for integrated reporting, but that they ask for important components of integrated reporting, like strategy, goals and opportunities. This paper adds value
from a societal point of view, as it combines the different components of integrated reporting to show how investors react on it, which is reflected in firm value.
Lastly, this paper focuses the banking industry. Today, sustainability is a focus point of many banks. ING, for example, has been announced by Sustainalytics as the most sustainable organization in the banking industry. Undoubtedly, reporting on this huge achievement definitely will not harm the organization. But will reporting on their accompanying sustainable strategy, their thoughts about the future of banking and their commitment to the UN’s sustainable development goals, as an addition to the financial performance of the organization, lead to a higher market value? As we do not know the answer yet, the answers on these questions are highly relevant for stakeholders.
1.5 Research Design
In this chapter, the research question has been formulated. In chapter 2, relevant literature is summarized, in order to clarify the most important concepts included in this paper. This eventually results in hypotheses to be tested. Subsequently, the methodology including relevant regression models to test the hypotheses formulated is described in chapter 3. Chapter 4 elaborates on the data needed for this paper and the way of testing these data, which leads to the results in chapter 5. Finally, chapter 6 includes the conclusion of this paper, providing the answer to the research question and giving recommendations for future research.
2 Theoretical framework
Current literature about integrated reporting mainly focuses on the IIRC (International Integrated Reporting Council) and its impact on the integrated reporting field, for example in Humphrey et al (2015). The current literature about the effects of integrated reporting is scarce. However, the research that exists predominantly covers the relevance or usefulness of integrated reporting for investors to base their decisions on. For example, Stubbs et al. (2014) discuss the perceptions of investors regarding integrated reporting and Churet and Eccles (2014) do research on the impact of integrated reporting on the quality of management. Churet and Eccles (2014) also investigate the impact of integrated reporting on financial performance, but use the ROIC (Return On Invested Capital) as dependent variable. In this paper, firm value is the dependent variable. To formulate the hypotheses for this paper, a literature review (consisting of a few of the most relevant articles) is performed. The results of this literature review are included in this chapter. Section 2.1 explains the concept of integrated reporting, followed by an explanation of the underlying agency theory in section 2.2. The definition and measurement of firm value are described in section 2.3. Subsequently, earnings quality as a moderator is clarified in section 2.4. Finally, based on the theoretical framework, the hypotheses are developed in section 2.5.
2.1 Integrated Reporting
According to the IIRC (2015), integrated reporting is about creating value for the organization itself (to enable financial returns for shareholders) and for other stakeholders over the short, medium and long term, integrating six ‘capitals’; financial, manufactured, natural, human, intellectual, and social and relationship. Capitals can be defined as “stocks of value on which all organizations depend for their success as inputs to their business model, and which are increased, decreased or transformed through the organization’s business activities and outputs” (IIRC, 2015). The most commonly used definitions of integrated reporting include the following: 1. “An integrated report communicates an organization’s strategy, governance, performance and prospects, in the context of its external environment, to show value creation over the short, medium, and long term” (IIRC, 2013a,b);
2. Integrated reporting is “organizational reporting for public disclosure, which includes both financial and important nonfinancial information” (Soyka, 2013);
3. “Integrated Reporting demonstrates the linkages between an organization’s strategy, governance and financial performance and the social, environmental and economic context within which it operates” (Thomson, 2015);
4. “Integrated Reporting provides companies with a means of credibly communicating the commitment of its top leadership to diffuse integrated thinking across the organization and build strong relationships with important external stakeholders” (Knauer & Serafeim, 2014). When comparing these definitions, it becomes clear that all definitions are focused on reporting both financial information and non-financial information, like strategy and governance. The definition of Soyka (2013) summarizes the most important concepts shortly, while Thomson (2015) specifically mentions that a linkage is created between the internal and external environment. Knauer & Serafeim (2014) consider building strong relationships with important external stakeholders. However, with the addition of the time framework, the definition of the IIRC (2013a,b) is considered to be the most complete and is therefore used in this paper.
The only country that obliges firms to adopt integrated reporting is South Africa (Atkins & Maroun, 2015). Therefore, it can be concluded that integrated reporting is voluntarily in other countries. According to Francis et al (2008), organizations generally disclose information voluntarily if they have good earnings quality. Voluntary disclosure leads to a lower cost of capital, but this relation disappears when a firm’s earnings quality is added as a moderator. It is therefore very important in this paper to moderate for the performance of the companies included in the data. The concept ‘earnings quality’ is explained further in section 2.4.1.
2.2 Agency Theory
As already discussed, the information provided by organizations in integrated reports combines financial information with non-financial information. The assumption made in this paper is that this leads to additional information and thus decreases the information asymmetry between investors and organizations. The underlying theory that elaborates on this information asymmetry between investors (principals) and organizations (agents) is called the agency theory. The agency theory assumes there is a relationship between principals and agents (see Table 1), calling this the agency relationship. Jensen & Meckling (1976) define this relationship as follows: “A contract under which one or more (principals) engage another person (the agent) to perform some service on their behalf which involves delegating some decision making authority to the agent’’
According to Eisenhardt (1989), the agency theory addresses two main problems in these agency relationships. The first problem is that agents and principals could have different goals, which creates a gap between the information that agents have and principals want. This gap is called information asymmetry. The second problem concerns the risk-sharing problem that is caused by a difference attitude towards risk of the agent and the principal.
In this paper, the focus is on the first problem of information asymmetry, as integrated reporting is expected to mitigate this.
Table 1. Overview of the most important concepts in agency theory (Eisenhardt, 1989, p.59)
2.3 Firm Value
2.3.1 Definition of firm value
Firm value is generally considered to be the result of the performance of companies. As Al-Matari et al (2014) state, the first thing investors look at when they assess a company is its performance. Where the performance of a firm can be defined as “the action’s efficiency and effectiveness” (Al-Matari et al, 2014), “the resulting firm value can be defined as the utility/benefits derived from the shares of a firm by the shareholders” (Rouf, 2011). Therefore, it
can be concluded that firm value is determined by the assessment of the performance of a firm by the company’s shareholders. These shareholders generally find their information in the financial statements of a company. However, Knauer & Serafeim (2014) suggest that shareholders increasingly find information in the ‘integrated reports’ of an organization and that longer-term investors are attracted by the issuance of integrated reports by those organizations. Attracting those longer-term investors could increase the confidence of management to successfully execute its strategy and enhance the value of the firm. This relationship between issuing integrated reports and firm value is further discussed in the hypotheses developed in section 2.5.
As Al-Matari et al (2014) state, from the organization’s perspective, a good performing company will be eager to disclose firm information (indicating a high earnings quality), as this further increases the value assessment by shareholders and thus of the firm. The concept of earnings quality is further described in section 2.4.1, whereas the relationship between earnings quality and firm value is included in the hypotheses formulated in section 2.5.
2.3.2 Measuring firm value
According to Al-Matari et al (2014), there are two ways to measure firm value: 1. Accounting-Based Measurements, and;
2. Market-Based Measurements.
Accounting-based measurement assesses the profitability of an organization on the short-term and is generally criticized for being backward-looking. Al-Matari et al (2014) state that accounting-based measurements are effective in measuring profitability of businesses, but that the only future-looking aspects of these measurement methods are depreciation and amortization. Examples of accounting-based measurements are ROE (Return on Equity), ROI (Return on Investment) and EPS (Earnings per Share). Market-based measurement takes previous or current performance as its basis, but focuses on shareholders’ long-term expectations of firm performance. Since the expectations of investors regarding future performance are crucial for this paper, market-based measurement is used to measure firm value. Examples of market-based measurements include Market Value Added (MVA), Market-to-Book Value (MTBV), Tobin’s Q and Abnormal Returns (RET). To measure firm value, this paper uses Tobin’s Q as this is the most common way to measure firm value (78% according to Al-Matari et al (2014)). According to Sucuahi & Cambarihan (2016), Tobin’s Q is considered as the best predictor of market correction. This could be useful for this paper, as firm value is influenced by many factors and only the issuance of integrated reports is regarded as the independent variable.
2.4 Determinants of firm value
2.4.1 Earnings Quality
As already discussed, integrated reporting is supposed to decrease information asymmetry and thus increase firm value. This assumption is based on the efficient market hypothesis, which states that a capital market reflects all information available (Malkiel, 1992). However, it is argued that firms performing well on financial as well as on non-financial ‘capitals’, are more likely to inform stakeholders by issuing integrated reports (Kuzey & Uyar, 2017). Francis et al (2008) also find that organizations with good earnings quality disclose more information voluntarily than organizations with poor earnings quality. Disregarding earnings quality, voluntary disclosure of information leads to a lower cost of capital. If earnings quality is used as a control variable between voluntary disclosure and cost of capital, the negative relation disappears. As the quality of information appears to influence the relationship, the moderating variable in this paper is ‘earnings quality’. Definitions of earnings quality include the following:
1. “By earnings quality, we mean the precision of the earnings signal emanating from the firm’s financial reporting system. Such imprecision affects the capital market’s demand for, as well as a firm’s motive to supply, disclosures that are useful to current shareholders and prospective investors in assessing firm value” (Francis et al, 2008);
2. “Higher quality earnings provide more information about the features of a firm’s financial performance that are relevant to a specific decision made by a specific decision-maker” (Dechow et al, 2010).
Three key features of earnings quality can be extracted from the definition as defined by Dechow et al (2010):
1. The relevance of information for decision-makers;
2. Earnings should provide more information about a firm’s financial performance;
3. Earnings quality is determined by both the relevance of information for decision-makers and the performance measurement as guided by the accounting system.
Determinants of earnings quality are firm characteristics, financial reporting practices, governance and controls, auditors, capital market incentives and external factors (Dechow et al, 2010). Dechow et al (2010) also mention the consequences of earnings quality: litigation propensity, audit opinions, market valuations, real activities including disclosure, executive compensation, labor market outcomes, a firm’s cost of equity capital, a firm’s cost of debt capital and analyst forecast accuracy. The findings that financial reporting practices determine earnings
quality and that variables as market valuations, cost of equity capital and analyst forecast accuracy are outcomes of earnings quality, support the use of earnings quality as a moderator in this paper.
The underlying assumption for earnings quality is the efficient market hypothesis, indicating that only new information influences firm value. This efficient market hypothesis is supported by existing literature, stating that “only the unexpected part of any announcement, the surprise, moves stock prices” (Pearce & Roley, 1984).
Current literature describes several methods to measure earnings quality. In this paper, a few of those methods are explained:
1. Earnings Response Coefficient (ERC), proposed by Dechow et al (2010)
2. A combination of accruals quality, earnings variability and absolute abnormal accruals, proposed by Francis et al (2008)
The first measurement method indicates how investors react to the value of information. Dechow et al (2010) mention several methods to measure earnings quality, including persistence, magnitude of accruals, residuals from accrual models, smoothness and timely loss recognition (TLR). However, ERC is most in line with this paper, as Dechow et al (2010) state that this measurement is directly linked to decision usefulness regarding equity valuation (which is related to firm value, the dependent variable in this paper). In the regression of ERC, β is higher when the information is more valuable. Higher R2 indicates that earnings are more value relevant. According to Dechow et al (2010), earnings quality is calculated using the following regression: Ret!= ∝ + β(EarningsSurprise!) + ε!
The second measurement method, proposed by Francis et al (2008), combines different methods used frequently in existing literature to measure earnings quality; accruals quality, earnings variability, absolute abnormal accruals and the combination of those three proxies. Accruals quality “separates accruals based on their association with cash flows by regressing working capital accruals on cash from operations in the current period, prior period, and future period, as well as the change in revenues and property, plant and equipment (PP&E). The unexplained portion of the variation in working capital accruals is an inverse measure of accruals quality; that is, a greater unexplained portion implies lower quality” (Francis et al, 2008). The second variable, earnings variability, is measured as the standard deviation of a firm’s earnings. According to Francis et al (2008), earnings are defined as earnings before extraordinary items. The higher the earnings variability, the lower the earnings quality. For the third variable, the Modified Jones
Approach is used to generate the absolute value of abnormal accruals (please refer to section 3.6). To combine the three variables mentioned, a factor analysis is applied to generate a common factor score.
2.4.2 Other determinants of firm value
In existing literature, many determinants of firm value have been relevant for researchers. Sucuahi & Cambarihan (2016) name the firm’s wealth, technology, organization structure, human resources, environmental factors, customer satisfaction, management understanding, technology usage, product quality, competition, sustainable growth and the firm’s financing. According to Glaum et al (2013) and Preiato et al (2015), firm value can be affected by factors like firm size, size of analyst following, leverage, return on assets, the sign of earnings, the book-to-market ratio and lagged accuracy. Purwohandoko (2017) identifies three main determinants of firm value, covering the basis of the previously mentioned variables: size, growth and profitability. In his article, Purwohandoko measures these as follows:
SIZE = Ln(total assets)
GROWTH = (annual total assets!− annual total assets!!!)/ annual total assets!!!
ROE = interest after tax total equity
According to Purwohandoko (2017), large companies tend to have a large market capitalization. Investors are interested more in firms that operate on a large scale, as large firms have a tendency to be in a more stable condition. Large companies thus tend to have a high firm value. Regarding growth, Purwohandoko (2017) states: “The existence of investment opportunities can provide a signal about the company’s growth in the future, so as to enhance shareholder value”. High growth rates are attractive for investors. Therefore, growing firms are expected to have a high firm value. Finally, Purwohandoko (2017) argues that the value of a company is enhanced when a firm is able to make a profit and investors thus expect a higher return. Profitable companies are therefore expected to have a higher firm value than loss-making companies.
2.5 Hypotheses Development
Based on the literature review, a few statements can be made. The first one is that integrated reporting is a voluntary exercise in Europe according to Bernardi & Stark (2015). Francis et al (2008) find that more voluntary disclosure leads to a lower cost of capital. This leads to the following hypothesis:
However, it is argued that firms performing well on financial as well as on non-financial ‘capitals’, are more likely to inform stakeholders by issuing integrated reports than firms performing worse (Kuzey & Uyar, 2017). Francis et al (2008) also find that organizations with good earnings quality disclose more information voluntarily than organizations with poor earnings quality. Based on these findings, the following hypothesis is proposed:
H2: Earnings quality is positively related to issuing integrated reports
Dechow et al (2010) state that financial reporting practices determine earnings quality, but they also mention the outcomes of earnings quality. Relevant outcomes are the findings that “firms that consistently meet or beat prior period earnings targets or analyst expectations are rewarded with higher valuations” and “evidence of a negative association between one or more earnings quality proxies and a firm’s cost of capital” (Dechow et al, 2010). These findings suggest a positive relationship between earnings quality and firm value, which leads to the following hypothesis:
H3: Earnings quality is positively related to firm value
After exploring the relationships between the three core concepts of this paper with the first three hypotheses, the last hypothesis is built on the influence of earnings quality on the relationship between issuing integrated reports and firm value. Existing literature suggests a positive relationship between issuing integrated reports and firm value, a positive relationship between earnings quality and issuing integrated reports and a positive relationship between earnings quality and firm value. Furthermore, Francis et al (2008) find that the relation between voluntary disclosure and cost of capital is affected by earnings quality. These findings support the use of earnings quality as a moderator in the relationship between issuing integrated reports and firm value. It is expected that a high earnings quality strengthens the positive influence of issuing integrated reports on firm value. Therefore, the last hypothesis reads:
3 Methodology
The literature review in the previous chapter helped in order to formulate the hypotheses. In this chapter, the regression models belonging to the corresponding hypotheses are described, including the measurement methods of relevant dependent, independent and moderating variables. Firm value is measured using the Tobin’s Q model (TOBQ), which is described in this chapter. The moderating variable (earnings quality) is measured using discretionary accruals (|DA|), as is described in this chapter. The chapter starts with an elaboration of the conceptual model of this paper. Then, the regression models for each hypothesis are described. Finally, a summary of the variables used is given.
3.1 Conceptual Model
Figure 1. Conceptual model of hypothesized relationships
Figure 1 shows the key variables of this paper: integrated reporting, firm value and earnings quality. This paper focuses on the relationship between integrated reporting and firm value and the influence of earnings quality on this relationship.
3.2 Regression Model for Hypothesis 1
In this section, the methodology behind the first hypothesis of this paper is described: H1: Issuing an integrated report has a positive effect on firm value
In this hypothesis, the key concepts are integrated reporting as the independent variable and firm value as the dependent variable. This hypothesis is tested using the following regression:
3.2.1 Dependent variable
To measure the dependent variable (the firm’s underlying value), the Tobin’s Q is used in this paper. Tobin’s Q is proposed by Chung and Pruitt (1994) and used by Sucuahi & Cambarihan (2016) in their research. They calculate Tobin’s Q using the following formula:
TOBQ = DEBT!"− ASSETS!" + DEBT!"+ No. of CS ∗ MP + PS TOTAL ASSETS
This formula includes the following variables: TOBQ = firm value
DEBT!" = short term debt
ASSETS!" = short term assets
DEBT!" = long term debt
CS = common shares
MP = market price of stocks
PS = liquidating value of preferred stock
Sucuahi & Cambarihan state that if the market value of company stocks equals the price of replacing an asset, the ideal situation is created and the TOBQ is equal to one. However, firms whose stock values are higher than its asset values are overvalued and have a TOBQ higher than one. A TOBQ lower than one indicates that a firm is undervalued, since stock values are lower than asset values.
Al-Matari et al (2014) suggest a simplified way to calculate Tobin’s Q: TOBQ = Market Capitalization + Total Liabilities
Total Assets
Since the data for the variables included in this simplified equation of Al-Matari et al can be more easily derived, this equation is used to calculate Tobin’s Q in this paper. The results can be interpreted the same as Sucuahi & Cambarihan (2016) state.
3.2.2 Independent variables
The independent test variable is the issuance of an integrated report (dIR). This variable is a dummy variable, meaning that the variable only can have two values. The value of the dummy variable dIR is one when the concerning firm issued an integrated report in a certain fiscal year, the value is zero when this firm did not issue an integrated report for the fiscal year.
As Purwohandoko (2017) clearly state the variables that determine firm value (see 2.4.2), size, growth and profitability are used in this regression as control variables.
3.3 Regression Model for Hypothesis 2
In this section, the methodology behind the second hypothesis of this paper is described: H2: Earnings quality is positively related to issuing integrated reports
In this hypothesis, the key concepts are earnings quality as the independent variable and the issuance of integrated reports as the dependent variable. This hypothesis is tested using the following logistic regression:
dIR = β!+ β!|DA| + β!SIZE + β!GROWTH + β!SIZE + ε
This is a logistic regression, since a binary dependent variable (dummy variable) is tested with outcomes of either one or zero.
3.3.1 Dependent variable
For an elaboration on the dependent dummy variable (dIR), please refer to section 3.2.2. 3.3.2 Independent variables
Since current literature suggests that the decision to issue an integrated report can be affected by earnings quality (see 2.4.1), this variable is added as independent variable. Earnings quality has been a topic of research for several authors, leading to a wide variety of measurement methods. As described in 2.4.1, Dechow et al (2010) use the ERC model and Francis et al (2008) use a combination of accruals quality, earnings variability and absolute abnormal accruals. Although the ERC model of Dechow et al could potentially lead to some interesting results (as it is intended to measure the same as this paper wants to measure), the authors do not really elaborate on the method and the variables used in their equation. For this reason, this equation is not used in this paper.
The model of Francis et al (2008) is a quite complex model, taking a high amount of variables into consideration and combining factor scores in a factor analysis. For efficiency purposes, this paper takes one of the variables included in the model to explain the presence of accrual-based earnings management; the absolute value of discretionary accruals (|DA|). The discretionary accruals are measured using the modified Jones Model (see section 3.6).
As hypothesis 2 entails that organizations are intended to issue an integrated report when overall results are good and the issuance of an integrated report depends on the size of the firm as well,
we use the same control variables as for firm value (see section 3.2.2) in this regression: size, growth and profitability.
3.4 Regression Model for Hypothesis 3
In this section, the methodology behind the third hypothesis of this paper ia described: H3: Earnings quality is positively related to firm value
In this hypothesis, the key concepts are earnings quality as the independent variable and firm value as the dependent variable. This hypothesis is tested using the following regression:
TOBQ = β! + β!|DA| + β!SIZE + β!GROWTH + β!ROE + ε
For an elaboration on the variable discretionary accruals (as a proxy for earnings quality), please refer to section 3.3.2.
3.4.1 Dependent variable
For an elaboration on the dependent variable (TOBQ), please refer to section 3.2.1. 3.4.2 Independent variables
For an elaboration on the independent variable (|DA|) and control variables (SIZE, GROWTH and ROE) used in this regression, please refer to section 3.3.2 for the independent variable and section 3.2.2 for the control variables.
3.5 Regression Model for Hypothesis 4
In this section, the methodology behind the fourth hypothesis of this paper is described: H4: Earnings quality strengthens the positive relationship between issuing integrated reports and firm value In this hypothesis, the key concepts are earnings quality as the independent variable and firm value as the dependent variable. This hypothesis is tested using the following regression:
TOBQ = β! + β!dIR + β!|DA| + β! |DA| ∗ dIR + β!SIZE + β!GROWTH + β!ROE + ε
3.5.1 Dependent variable
3.5.2 Independent variables
For an elaboration on the independent variables (dIR and |DA|) and control variables (SIZE, GROWTH and ROE) used in this regression, please refer to section 3.3.2.
3.5.3 Moderating variable
Since current literature suggests that the relationship between the issuance of integrated reports and firm value is moderated by earnings quality (see 2.4.1), this variable is added to the regression as a moderating variable. As already described in 3.3.2, earnings quality is measured by discretionary accruals (|DA|). To measure discretionary accruals, the modified Jones model is used (see section 3.6).
Standalone, the variables proposed are expected to have a moderating impact on the relationship between the issuance of integrated reports and firm value. However, to answer the research question in this paper, the influence of integrated reporting in combination with earnings quality on firm value is measured. This moderating effect is measured by using cross variable |DA|*dIR.
3.6 Modified Jones Model
To measure earnings quality, discretionary accruals (DA) are tested in this paper. Discretionary accruals are measured using the modified Jones model, as described by Dechow et al (1995). The model is created in order to measure the amount of earnings management, which causes earnings manipulation and has an effect on firm value. A high value for discretionary accruals would lead to a low value for earnings quality, whereas a low value for discretionary accruals would lead to a high value for earnings quality. Within the modified Jones model, a separation is made between discretionary accruals (DA) and non-discretionary accruals (NDA). The latter is not included as a variable in this paper, as it is difficult to manipulate earnings using non-discretionary accruals. Total accruals (TA) in fiscal year t are calculated as follows, according to Dechow et al (1995): TA! = ∆CA!− ∆CL!− ∆Cash!+ ∆STD!− DEP ! (A!!!)
The equation that Dechow et al (1995) formulate to measure total accruals (TA) in order to estimate the parameters on a yearly basis is:
TA! = α! (1 A!!!) + α! ∆REV!/A!!! + α! PPE!/A!!! + ε
As the regression shows, the parameters on a yearly basis should be calculated excluding the delta of net receivables. The non-discretionary accruals, as the following regression shows, need
to be calculated including the delta of net receivables. The regression formulated by Dechow et al (1995) to measure the non-discretionary accruals is the following:
NDA!= α! (1 A!!!) + α! (∆REV!− ∆REC!)/A!!! + α! (PPE!/A!!!)
Total discretionary accruals per year are calculated extracting the non-discretionary accruals from the total accruals, both of which are measured by Dechow et al (1995). This leads to the following calculation of discretionary accruals:
DA! = TA!− NDA!
These formulas include the following variables:
DA = discretionary accruals current FY TA = total accruals current FY
NDA = estimated non discretionary accruals current FY ∆CA = change in current assets
ΔCL = change in current liabilities
ΔCash = change in cash and cash equivalents
ΔSTD = change in debt included in current liabilities DEP = depreciation and amortization expense A = total assets
ΔREV = change in revenues
PPE = gross property, plant and equipment ΔREC = change in net receivables
α!, α!, α! = regression coefficients α!, α!, α! = estimates for coefficients
3.7 Summary of Variables
Table 2 provides an overview of all variables included in this paper. Furthermore, the expected correlations between the dependent and independent variables used and their values are explained.
Variable Description Explanation
TOBQ Market capitalization plus total liabilities divided by total assets
Firm value is measured in this paper by using Tobin’s Q. Values higher than one indicate that a firm is overvalued, as the market value is higher than the book value. Values lower than one indicate that a firm is undervalued, as the market value is lower than the book value.
dIR Dummy variable with a value of 1 when an integrated report is issued and a value of 0 when an integrated report is not issued
The issuance of an integrated report is a voluntary exercise in Europe according to Bernardi & Stark (2015). Francis et al (2008) find that more voluntary disclosure leads to a lower cost of capital. Therefore, it is expected that data with a value of 1 on dIR have a higher firm value than data with a value of 0 on dIR.
SIZE The natural logarithm of the total assets Purwohandoko (2017) states that large companies tend to have a large market capitalization. Therefore, it is expected that firms with a high coefficient on the variable SIZE have a higher firm value.
GROWTH The total assets of the current fiscal year minus the total assets of the previous fiscal year divided by the total assets of the previous fiscal year
High growth rates are attractive for investors an thus for firm value, as Purwohandoko (2017) says in his research paper. High growth rates are thus associated with high firm values in this paper.
ROE The net result after tax divided by the
shareholders’ equity According to Purwohandoko (2017), the firm value of the banking groups analyzed in this paper is higher when those banking groups make a profit and add value for shareholders with this profit, measured by the ROE.
|𝐃𝐀|𝐭 Absolute value of the discretionary
accruals Earnings quality is expected to be higher when discretionary accruals are lower. Earnings quality is expected to be positively related to the issuance of integrated reports and to firm value.
dIR*|DA| Cross variable of dIR and |DA| Firm value is expected to rise when integrated reports are issued with a high earnings quality.
4 Data
In order to test the hypotheses formulated in this paper, a quantitative research method is executed. For this paper, public data are used. The amount of organizations that adopted integrated reporting are derived from the IR reporting database created by IIRC and Black Sun Plc. The integrated reports themselves are readily available on the organizations’ corporate websites. Quantitative data like earnings, firm value, firm size, growth and profitability are derived from Compustat Global in WRDS and Datastream. In this chapter, the method of collecting data is described and the descriptive statistics of these data are provided.
4.1 Data Collection
The objective of this paper is to provide insight into the impact of integrated reporting on firm value, in order to contribute to existing knowledge in the European banking industry. This paper focuses on the European banking industry for the following reasons. First, the interest for integrated reporting is increasing in Europe, with many organizations currently adopting it. Second, there is a pressure from the government on the banking industry to report on non-financial performance, which makes this a relevant paper for banks (Bouma et al, 2001). Finally, this paper makes a comparison between the firm value of organizations that issue integrated reports and of organizations that do not issue integrated reports. This creates the need for a balanced panel strategy. Therefore, only the banking industry is chosen to be the scope of this paper, focusing solely on Europe. Because the dataset is limited to the banking industry and the geographical area of Europe, the companies in the dataset can be considered as comparable. For this paper, the 56 banking groups headquartered in Europe and stated on thebanks.eu, an independent website providing information about European banking, were supposed to be the dataset of this paper. In this chapter, the choices that have led to the ultimate dataset of 50 European banking groups are described. Another possible data set to choose were the banks included in the S&P Global Market Intelligence rankings. However, S&P only ranks the largest European banks. As a balanced panel strategy is used and ‘size’ is one of the control variables in the regressions of this paper, the dataset of thebanks.eu is chosen as it includes small banking groups as well.
4.1.1 Integrated reporting data
The data in this paper are extracted for a period from fiscal year 2010 until fiscal year 2016. This period of seven years covers a period in which integrated reporting was established (with the
issuance of the final draft version by the IIRC in December 2013), as well as a period in which banks started issuing integrated reports.
It is quite hard to determine whether an organization compiles an integrated report or not, as integrated reporting is still in an early phase. Many organizations apply some of the guiding principles or fundamental concepts, but when is an annual report an integrated report? Is an annual report an integrated report when it is called an integrated report? In this paper, an organization is supposed to issue integrated reports if it is included in the IR reporting database of the IIRC and Black Sun Plc. Because of its independence from the organizations included in the database, it is believed to be a reliable source of information.
4.1.2 Other data
For all other data needed, WRDS and Datastream are used. WRDS (Wharton Research Data Service) is used to retrieve specific firm data that are important for the calculation of the variables |DA|, SIZE and GROWTH. Data for these variables are retrieved for fiscal years 2010 until 2016, using the tool ‘Financial Statements: BS, IS and CF Data Items’ in ‘Compustat Global Fundamentals’. Of the 56 banking groups headquartered in Europe according to thebanks.eu, 53 banks could be found in WRDS. The three remaining banks (Rabobank, BPCE & Sparkassen-Finanzgruppe) are not listed and therefore, there are no data available for these banks in WRDS. For this reason, these organizations are excluded from the dataset.
In Datastream, a few specific variables are readily available, like ROE and market capitalization (which is needed to calculate TOBQ). These data are not available for DZ Bank and KfW, which are excluded from the dataset for this reason. Because of its liquidation in 2014, Banco Espirito Santo also lacks availability of data and is excluded from the dataset. These exclusions lead to a final dataset of 50 European banking groups. The list with these 50 European banking groups can be found in Appendix 1.
4.1.3 Discretionary Accruals
The proxy for earnings quality in this paper is discretionary accruals (DA). As already discussed in section 3.6, the DA are calculated using the Modified Jones Model. To calculate DA, the total accruals and nondiscretionary accruals need to be calculated first. Then, distracting the nondiscretionary accruals from the total accruals results in DA. The absolute value of those discretionary accruals (|DA|) is used to determine the relationship with firm value. First, total accruals (TA!) are calculated using the formula given in section 3.6. Data needed are extracted
variables that need to be created) and used as input for STATA. The delta for the concerning variables are calculated in STATA.
After the total accruals are calculated, the parameters need to be estimated on a yearly basis. This is done in STATA, running the regression stated in section 3.6 per year from 2010 to 2016. The resulting coefficients are included in table 3. Based on these results, the nondiscretionary accruals (NDA!) can be calculated. After distracting the nondiscretionary accruals from the total accruals, the discretionary accruals are calculated.
𝛂𝟏 𝛂𝟐 𝛂𝟑 2010 149,071 10,908 0,827 2011 -623,242 7,077 12,876 2012 -144,300 1,499 -0,856 2013 -285,118 -0,730 4,996 2014 -317,304 9,415 -0,645 2015 9,205 -0,528 3,803 2016 45,692 0,263 -1,097
Table 3. Yearly estimated coefficients of the Modified Jones Model 4.1.4 Balanced panel strategy
In this paper, a balanced panel strategy is used, proposed by Pope and McLeay (2011). “If it can be assumed that relevant firm characteristics, and their impact on the outcome variable, do not change over time, the inclusion of firm fixed effects in the model can control for these effects. Further, if there are time effects that are constant across firms in their impact on the outcome variable, they can be controlled for via the introduction of time fixed effects” (Bernardi & Stark, 2015). The balanced panel strategy makes sure that effects that might influence firm value over time are controlled.
In order to only measure changes in firm value as a result of the issuance of integrated reports in European banks, a comparison is made between the firm value of organizations that started issuing integrated reports in the period between 2010 and 2016 and the firm value of organizations that did not start issuing integrated reports in this period. It is expected that the
firm value of a bank issuing an integrated report in the fore mentioned period increases more than a bank not issuing an integrated report (see figure 2). By including comparable organizations in the dataset (balanced panel strategy) and using the control variables set out in chapter 3, chances that other determinants of firm value affect the relationship studied in this paper are decreased.
Figure 2. Expected Research Results (example)
4.2 Descriptive Statistics
This section includes the descriptive statistics. First, the core statistics of the used variables are described. Second, the Pearson correlation matrix is included to show associations between variables. Finally, the multicollinearity is tested using the Variance Inflation Factor (VIF).
4.2.1 Core Statistics
Core statistics are calculated using STATA. For the results of the core statistics, please refer to table 4 and table 5. The dependent variable TOBQ has a mean of 0,988 with a standard deviation of 0,029. This means that overall, firms in the dataset used are slightly undervalued, since stock values are lower than asset values (the mean is below 1). The median of 0,983 is close to the mean and falls within one standard deviation of the mean. The difference between the mean and median of all other variables is also smaller than one standard deviation.
In table 5, the core statistics of dummy variable dIR are described. The statistics state that of the 50 banking groups in the dataset, 18% issued an integrated report between 2010 and 2016.
Variables Observations Mean deviation Standard Median Minimum Maximum TOBQ 350 0,988 0,029 0,983 0,902 1,098 |DA| 350 0,064 0,099 0,042 0,001 1,169 dIR*|DA| 350 0,014 0,086 0,000 0,000 1,169 SIZE 350 12,234 1,565 12,335 7,705 14,628 GROWTH 350 0,025 0,268 0,000 -0,870 4,304 ROE 350 0,183 0,212 0,056 -1,637 0,981
Table 4. Descriptive statistics of the continuous variables dIR Frequency Percentage
0 287 82
1 63 18
Total 350 100
Table 5. Descriptive statistics of the dummy variable 4.2.2 Correlation Matrix
Pearson’s correlation matrix is created in STATA (please refer to table 6). With this matrix, the association between the variables is shown. The closer the absolute value of the correlation is to 1, the more two variables correlate to each other. Table 6 shows the cross variable dIR*|DA| correlates with dIR, |DA| and GROWTH. There is also an indication that |DA| and GROWTH correlate, since the correlation matrix shows a value of 0,5795. These findings suggest that a growing company is considered to do earnings management more than a non-growing company. Since the pressure to execute earnings management is considered to be higher when an organization wants to hold on to their growth, this is in line with expectations. Furthermore, a strong positive correlation exists between SIZE and dIR, suggesting there is a higher probability that large organizations issue integrated reports than small organizations. As large organizations have more aspects to integrate than small organizations and large organizations have more resources to actually create an integrated report than small organizations, the positive relationship between SIZE and dIR appears reasonable.
TOBQ dIR |DA| dIR*|DA| SIZE GROWTH ROE TOBQ 1,0000 dIR -0,0015 1,0000 |DA| -0,0400 0,0647** 1,0000 dIR*|DA| -0,0357 0,3476* 0,7887* 1,0000 SIZE 0,0343 0,3657* -0,1611* -0,0300 1,0000 GROWTH 0,0793** 0,0514** 0,5795* 0,6391* -0,0399 1,0000 ROE 0,1573* -0,0012 -0,0468 0,0418 0,0306 0,1525* 1,0000 * : Statistical significant at 10% ** : Statistical significant at 5% Table 6. Correlation matrix
4.2.3 Multicollinearity
Multicollinearity is caused by a high correlation between variables used in a regression model. To measure multicollinearity, the Variance Inflation Factor (VIF) is a widely used measurement method (Marquardt, 1970). Marquardt (1970) states that a VIF greater than 10 is an indicator of multicollinearity. For this paper, VIF is calculated in STATA (for the outcomes, refer to table 7). The highest VIF is 4,17, which implies no multicollinearity exists in the regressions used.
Table 7. Multicollinearity Variable VIF dIR 1,50 |DA| 3,27 dIR*|DA| 4,17 SIZE 1,21 GROWTH 1,86 ROE 1,06
5 Results
The four hypotheses formulated in section 2.5 have been analyzed using the methodologies described in chapter 3 and the data described in chapter 4. In this chapter, the results of the regression analyses performed on these data are described and are displayed in tables 8 to 11. The results are discussed per regression, starting from a fixed effects model stating that the differences between the banking groups are due to all independent variables used.
5.1 Results regression 1
The first hypothesis tested is stated as follows in section 2.5: H1: Issuing an integrated report has a positive effect on firm value This hypothesis is tested using the following regression:
TOBQ = β! + β!dIR + β!SIZE + β!GROWTH + β!ROE + ε For the results of the regression analysis, please refer to table 8.
Number of observations 350 F-value 2,60 p-value 0,0363** R-squared 0,0292 Adj R-squared 0,0180 * : Significant at 1% ** : Significant at 5%
Table 8. Results regression 1 with dependent variable TOBQ
TOBQ Coefficient Standard error t-value p-value
Constant 0,978 0,013 75,86 0,000*
dIR -0,001 0,004 -0,33 0,744
SIZE 0,001 0,001 0,68 0,495
GROWTH 0,006 0,006 1,10 0,271
The F-statistic has a value of 2,60 with an associated p-value of 0,0363. This means that the variables used in this regression model have significant explaining power. The R-squared of the regression is 0,0292, which means that at least 2,92% of the variance in the dependent variable (firm value) can be explained by the independent variables (issuance of integrated reports, size, growth and profitability).
As Purwohandoko (2017) states (which is included in section 2.4.2), all three control variables (SIZE, GROWTH and ROE) individually are expected to enhance firm value. The variables SIZE and GROWTH are insignificant with p-values of respectively 0,495 and 0,271 and relatively low coefficients (β = 0,001 and β = 0,006). The only significant variable is ROE (β = 0,020 and p = 0,007). The three control variables are expected to have a positive influence on firm value, the coefficients show that these positive relationships between SIZE, GROWTH and ROE on the one hand and TOBQ on the other hand exist.
The association between the dependent variable (TOBQ; firm value) and the independent variable of IR (issuance of an integrated report) is the topic of the hypothesis formulated. It is expected that the issuance of an integrated report has a positive effect on firm value (as stated in the hypothesis formulated in section 2.5), but the results state that this expectation is incorrect (β = -0,001). While we expect a positive influence of integrated reporting on firm value, table 8 shows a negative relation. However, the coefficient is insignificant (p = 0,744).
Based on the above information, hypotheses 1 is rejected.
5.2 Results regression 2
The second hypothesis tested is stated as follows in section 2.5: H2: Earnings quality is positively related to issuing integrated reports This hypothesis is tested using the following regression: dIR = β!+ β!|DA| + β!SIZE + β!GROWTH + β!SIZE + ε
As the dependent variable is a binary (dummy) variable, this regression is treated as a logistic regression. For the results of the regression analysis, please refer to table 9.
Number of observations 350 LR chi2 67,08 p-value 0,0000* Pseudo R2 0,2033 * : Significant at 1% ** : Significant at 5%
Table 9. Results regression 2 with dependent variable dIR
As already discussed in section 3.3, this is a logistic regression with a binary dependent variable (dummy variable) as an outcome. The likelihood ratio chi-square (comparable with an F-value in a linear regression) has a value of 67,08 with an associated p-value of 0,0000. This means that the variables used in this regression model have significant explaining power. The pseudo-R-squared of the regression is 0,2033, which means more or less that at least 20,33% of the variance in the dependent variable (the issuance of an integrated report) can be explained by the independent variables (earnings quality, size, growth and profitability).
The variable SIZE is significant (p = 0,000) with a coefficient of 1,017, which supports the expectation that the probability that large organizations issue an integrated report is higher than in small organizations. The variables GROWTH (β = -0,530) and ROE (β = -0,131) show negative relationships with dIR, which is contradictory to the expectations that the probability of issuing integrated reports is higher in growing organizations and profitable organizations. However, those coefficients are insignificant, with p-values of respectively 0,448 and 0,870.
dIR Coefficient Standard error z-value p-value Constant -14,87 2,086 -7,13 0,000*
|DA| 4,910 1,963 2,50 0,012**
SIZE 1,017 0,154 6,61 0,000*
GROWTH -0,530 0,699 -0,76 0,448
This logistic regression aims to understand whether the earnings quality of organizations predicts the probability that those organizations issue an integrated report. The second hypothesis assumes a positive relationship between the dependent variable (dIR; the issuance of an integrated report) and the independent variable (earnings quality). It is expected that the lower the discretionary accruals are, the higher the earnings quality is and the higher the chance is that an organization decides to issue an integrated report. This relationship is confirmed by prior research by Francis et al (2008), who state that organizations with high earnings quality disclose more information voluntarily than organizations with low earnings quality. However, this statement is not supported by the results of the logistic regression. Those results give a β of 4,910 with a p-value of 0,012, indicating that the presence of discretionary accruals gives rise to the probability of issuing an integrated report.
Based on the above information, hypothesis 2 is rejected.
5.3 Results regression 3
The third hypothesis tested is formulated as follows in section 2.5: H3: Earnings quality is positively related to firm value
This hypothesis is tested using the following regression:
TOBQ = β! + β!|DA| + β!SIZE + β!GROWTH + β!ROE + ε
Number of observations 350 F-value 3,12 p-value 0,0152** R-squared 0,0350 Adj R-squared 0,0238 * : Significant at 1% ** : Significant at 5%
Table 10. Results regression 3 with dependent variable TOBQ
The F-statistic has a value of 3,12 with an associated p-value of 0,0152. This means that the variables used in this regression model have significant explaining power. The R-squared of the regression is 0,0350, which means that at least 3,50% of the variance in the dependent variable (TOBQ; firm value) can be explained by the independent variables (earnings quality, size, growth and profitability).
The control variables SIZE, GROWTH and ROE are expected to have a positive effect on firm value (for more information regarding this expectation, please refer to section 5.1). Although its coefficient (β = 0,001) shows a positive relationship with firm value (as expected), the variable SIZE is insignificant, with a p-value of 0,723. The other control variables GROWTH and ROE show significant (p = 0,082 and p = 0,014, respectively) positive relationships (β = 0,013 and β = 0,019, respectively) with firm value, in line with the expectations.
The association between the dependent variable (TOBQ; firm value) and the independent variable of earnings quality (|DA|) is the topic of study of this regression. It is expected that low discretionary accruals lead to a high earnings quality. This high earnings quality is expected to
TOBQ Coefficient Standard error t-value p-value Constant 0,984 0,013 77,97 0,000*
|DA| -0,029 0,020 -1,47 0,143
SIZE 0,001 0,001 0,35 0,723
GROWTH 0,013 0,007 1,75 0,082
result in a high firm value. The results of the regression analysis confirm this negative relationship between |DA| and TOBQ, as |DA| shows a negative coefficient of β = -0,029. However, the coefficient is insignificant (p = 0,143) and hypothesis 3 thus has to be rejected.
5.4 Results regression 4
The fourth hypothesis tested is formulated as follows in section 2.5:
H4: Earnings quality strengthens the positive relationship between issuing integrated reports and firm value This hypothesis is tested using the following regression:
𝑇𝑂𝐵𝑄 = 𝛽! + 𝛽!𝑑𝐼𝑅 + 𝛽!|𝐷𝐴| + 𝛽! |𝐷𝐴| ∗ 𝑑𝐼𝑅 + 𝛽!𝑆𝐼𝑍𝐸 + 𝛽!𝐺𝑅𝑂𝑊𝑇𝐻 + 𝛽!𝑅𝑂𝐸 + 𝜀 For the results of the regression analysis, please refer to table 11.
* : Significant at 1% ** : Significant at 5%
Table 11. Results regression 4 with dependent variable TOBQ Number of observations 350
F-value 2,42
p-value 0,0263** R-squared 0,0407 Adj R-squared 0,0239
TOBQ Coefficient Standard error t-value p-value
Constant 0,981 0,014 72,31 0,000* dIR 0,003 0,005 0,52 0,606 |DA| -0,001 0,028 -0,00 0,997 dIR*|DA| -0,052 0,037 -1,42 0,157 SIZE 0,001 0,001 0,49 0,624 GROWTH 0,017 0,008 2,16 0,032** ROE 0,019 0,008 2,56 0,011**
