Sustainability assurance and financial performance

(1)

Sustainability assurance and financial

performance

Student: Tamara Timár Student number: 10681256

Date of final version: 22 June 2015 Word count: 11297

Study: MSc Accountancy & Control, variant Accountancy

Faculty of Economic and Business, University of Amsterdam Supervisor: dr. A. Sikalidis

(2)

Statement of Originality

This document is written by student Tamara Timár 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.

(3)

Abstract

This study empirically examines the relationship between corporate social responsibility (CSR) and corporate financial performance (CFP), particularly focusing on assured sustainability reports in relation to markets in the Netherlands and the United Kingdom. There is no consensus regarding this association and causality, but the majority of prior research found a positive relationship with applied accounting-based performance measures (Orlitzky et al., 2003). Therefore, this study analyses the CSR- CFP links from both accounting- and market-based perspectives as well as testing causality in both cases. Therefore there are 732 firm-year observations gathered from companies listed on FTSE and AEX from the period of 2007 to 2014. Mixed results are found regarding the first hypothesis that assured sustainability reports have a positive association with companies’ financial performance. Assured sustainability reports are found to have a negative effect on the performance ratio Tobin’s Q. This model has been found to be the most significant, expressing namely twice as much variance than the models with other measures. On the other hand accounting-based measures ROA and net income are positively affected by assured sustainability reports which is in line with expectations and the majority of literature. The tests of the second hypothesis, that better performing companies are more likely to engage a third party to assure their sustainability report, showed similar results as the first hypothesis. The model with the indicator Tobin’s Q is found to be the most significant and negatively affects sustainability assurance engagements. Moreover the U.K. is found to produce and assure more stand-alone sustainability reports than the Netherlands.

(4)

1. Introduction

1.1. Background

In addition to economic turbulences like accounting scandals and the recent financial crises, different factors as environmental issues like climate change, environmental disasters (e.g. oil spills) effect a companies’ reputation and the stakeholders reliance on these companies. These events of the past decades have been gaining more attention in society and so grows the demand on more information concerning how organizations impact their surroundings. Not only concerning the environment but also society and the economy (Buhr et al., 2014).

The companies’ reaction to these events to restore their reputation was to provide other kind of aspects and richer view to stakeholders about their operations besides financial information. Therefore it was a trigger to the companies to also report non-financial (corporate social and environmental or sustainability) information regarding their business actions. So evolved sustainability reporting in accounting, next to financial reporting as society embraced the importance of sustainability (Buhr et al., 2014). The need for the reliability of these reports has also been increased, which means that the users of the reports demand more credibility and transparency regarding the sustainability information to gain a more comprehensive picture of an organizations overall performance (Cooper & Owen, 2014, p. 72). According to the latest KPMG’s International Survey of Corporate Responsibility Reporting (KPMG, 2013, p. 22), 93 percent of the worldwide 250 largest companies provide insight into their activities on corporate responsibility and create reports on this field. This type of reporting gives investors and stakeholders a good overview of the overall performance and impact that a company has on the environment, society and economy.

Sustainability reporting and the assurance of these statements are not mandatory. Guidelines, standards and scholars highlight and examine the appliance of corporate responsibility reporting which is despite its voluntary nature, quite widespread among the global largest 250 companies (Buhr et al., 2014, p. 56; Junior et al., 2014). Unlike the verification of the sustainability information by third parties. Scholars have been researching sustainability reporting and how diverse and widespread it is for a while. Substantial amount of studies have examined the connection between sustainability reporting and financial performance (Laan et al., 2008) but this relationship remains a topic for a debate (Perrini et al., 2011). A significant amount of research has found a positive relationship between corporate social and financial performance (Orlitzky et al., 2003; Margolis & Walsh, 2003).

(6)

This is also consistent with the outcome of KPMG’s survey (2011, p.18): “Almost half of the G250 companies report gaining financial value from their corporate responsibility (CR) programs, while a third of N100 companies report the same”.

Besides above mentioned, the auditing and verification of sustainability reports lag behind the reporting appliance due to its voluntary aspect. However Simnett et al. (2009, p. 961) found that companies that care to improve their credibility, assure the sustainability reports. This could lead to an increased demand on verification of these non-financial reports by an independent third party, as illustrated in the PwC’s (2007) report.

The Global Reporting Initiative (GRI) recommends the use of external assurance but it is not a requirement “to be ‘in accordance’ with the Guidelines” (Global Reporting Initiative, 2013). GRI argues in its guideline however why it is worth it to an organization to assure its sustainability report, and indicates: increased recognition, trust and credibility; reduced risk and increased value; improved Board and CEO level engagement; strengthened internal reporting and management systems and improved stakeholder communication (Global Reporting Initiative, 2013, pp. 6-7). As the GRI guideline shows the different dimensions of why an organization should assure the sustainability reports, it also declares that stakeholders are showing increased interest in these disclosures. From the firms’ viewpoint KPMG (2013, p. 14) states that 36 percent of the Global Fortune 250 reporting companies see social and environmental change as an opportunity to improve market position or growing market share which implies financial performance growth.

Above mentioned factors could indicate that these aspects strengthen the companies’ overall performance, therefore and consequently the financial performance. To look further into the matter of more reliable information regarding the firms’ sustainability information I am interested in examining whether sustainability assurance can also be associated with enhanced company financial performance. The purpose of this study is to investigate whether the assurance of sustainability reports contributes to companies success and survival in the long run and analyze the association between them if there is any. Therefore, the following question will be in the scope of this research: Does assurance of sustainability report have an impact on a firms’ financial performance?

1.2. Motivation of this study

Corporate environmental and social responsibility has become almost a standard part of business strategies and activities of companies and the external assurance of this report is

(7)

increasing. (Cooper & Owen, 2014; Jones & Solomon, 2010; KPMG, 2013). Organizations are not mandated to audit their reported sustainability information, although there are standards and guidelines available. However there is a significant growth observed by KPMG: whilst in the survey of 2011 less than 50 percent of the 250 largest international companies did undertake assurance of their sustainability information, by 2013 this percentage changed to 59 percent (KPMG, 2013, p. 12).

This phenomenon makes me wonder what the companies reason could be – besides enhanced transparency and reliability – that drives the emerging engagement in voluntary commissioning assurance services of sustainability reports. And this is even more interesting considering the fact that on the one hand sustainability assurance is voluntary and on the other hand obtaining assurance means extra costs for the firms next to the “conventional” financial statement audits. Therefore the assurance must also eventually deliver financial benefits for the companies otherwise they would not obtain assurance over sustainability reports. Since most studies have found a positive relationship between corporate social reporting and financial performance (Orlitzky et al., 2003; Margolis & Walsh, 2003) it could be argued that assuring sustainability reports enhance financial performance as well, since assurance can amplify this effect (Casey & Grenier, 2015). Therefore at the center of the this study is the question of whether applying assurance on sustainability reports affects the companies’ financial performance. In line with the research of Simnett et al. (2009, p. 938) this study will also concentrate on stand-alone, assured non-financial reports to be able to measure the effect of financial performance.

This research addresses the future research suggestion of Junior et al. (2014) to study the motivations and benefits of commissioning sustainability assurance services, moreover to examine how in two different countries and legal traditions within Europe, the assurance of sustainability reports have developed between 2007 and 2014. To draw upon the research of Ruhnke and Gabriel (2013) two European countries will be examined on how sustainability assurance and financial performance developed in the past years to draw a long term observation of this link. Simnett et al. (2009) found, among companies around the world, that the U.K. was one of the three countries that mainly produced sustainability reports.

The aim of this research is to contribute to the quantitative research approach of sustainability studies with examining how sustainability affects financial performance on a different level, more in-depth through assurance that adds extra credibility and transparency to stakeholders. Further, the findings are intended to give more insight to companies that

(8)

them if they operate in a more conscious (i.e. socially, environmentally and economically) way that is beneficial for everyone: stakeholders and the firms themselves.

This study is organized as follows. In the next chapter the relevant literature is discussed and the hypotheses are developed. Chapter 3 describes the data collection procedure and the research method. Chapter 4 presents the results of the analysis performed followed by discussion on the findings. Lastly the conclusion and some suggestions for future research are provided in chapter 6.

2. Literature review and hypotheses

This section will be concerned with the theoretical framework, afterwards an overview of prior research will be presented and subsequently the hypothesis of this paper will be developed, based on the main research question.

2.1. Theoretical framework

Two financial accounting theories, legitimacy and stakeholder theory, have been used commonly in the accounting literature to describe organizational strategy to voluntarily disclose corporate social responsibility information (Deegan & Unerman, 2006, p. 268). This research will also build on these two theories in order to answer the main research question.

Scholars are still researching the underlying reason for a companies’ motives for publishing standalone sustainability reports. Legitimacy theory is used to account for the foundation that clarifies the companies’ motivation to prepare voluntary sustainability reports. The concept of this theory is that companies have to take the society’s expectations, values and norms, into account in everyday operations (Deegan & Unerman, 2006). Therefore organizations report their operations to the society to legitimate themselves. In Suchman’s (1995, p. 574) interpretation “Legitimacy is a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions.”. Based on legitimacy theory if a company does not operate conform society’s expectations, latter will discipline the company in various ways, such as through changing demand for the organizations products, through supply or lobbying the government (Deegan & Unerman, 2006, p. 272). Another method of the society could be to deny financial capital from the organization that really could put the existence of the organization in danger.

(9)

These actions effect the companies’ operations and therefore their performance as well and hence can be linked to the relationship between the commission of sustainability assurance services and corporate financial performance.

Deegan and Unerman (2006, p. 285) describe the main difference between the two theories is that “… stakeholder theory focusing on how an organization interacts with particular stakeholders, whilst legitimacy theory considers interactions with ‘society’ as a whole.”

Stakeholder theory is derived from legitimacy theory and can be divided into two branches: ethical and managerial (Deegan & Unerman, 2006). The previous one considers stakeholders as one group with rights that has to be respected. These stakeholders have also rights to information regarding the impacts of the company that affects them (Deegan & Unerman, 2006). Stakeholder theory also studies how companies operations and existence are affected by the different stakeholder groups interests. These two aspects, the need for information and the latter an organization’s activities in the light of stakeholder interest can be reflected in the financial performance and sustainability assurance relationship.

2.2. Corporate social responsibility (CSR)

There are different definitions of corporate environmental and social responsibility or sustainability. This inconsistency arises from the voluntary and unregulated manner of reporting (Buhr, Gray, & Milne, 2014, p. 55). Thus the use of the terminology is also not consistent as neither practitioners nor scholars refer to it in the same way. Standalone sustainability reports are also referred to as “corporate social responsibility reports“ (CSR), corporate social and environmental responsibility (CSER) or corporate citizenship reports.” (Freundlieb et al., 2014, p.20). Despite these differences, three characteristics are common in these reports: social and environmental issues are presented, they are separated from the annual reports, and the subject matter is not mandatory or regulated (Thorne et al, 2014). This variety of usage of terminology is also showed with statistical numbers from the practitioners perspective in the KPMG triennial survey and it is revealed that among 250 of the world’s largest companies ‘sustainability’ reporting is the most applied term (KPMG, 2013, p. 6). In this thesis all these variation of the terminology will be used.

One of the most cited definitions of sustainability is given by Carrol (1999) that three elements should be considered by firms in the conceptualization of sustainability in strategic and decision making processes and these are social, environmental and economic issues

(10)

(García-Benau et al., 2013, p. 1531). Therefore in order for sustainable existence of a company and for economic benefits, social and environmental issues should be maintained as well. The Commission of the European Communities (2001) takes internal and external stakeholders into account and describes sustainability reporting as “a concept whereby companies integrate social and environmental concerns in their business operations … on a voluntary basis.”.

In the past decades many companies have implemented CSR into their business activities, voluntarily. CSR can be found at different levels and in different processes within a company. CSR policy implementations however, require investment and extra costs from the companies, in order to benefit from these efforts in the future. Wang & Choi (2013) showed that consistent performance in CSR has positive effects on firm value. In the following sections of the thesis this association will be discussed further. In some countries it is mandated by the government to report on non-financial issues. South Africa is among the first nations where CSR reporting is required. Other countries are for instance Denmark, France, the Netherlands, India and Australia (GRI website; KPMG, 2013). In 2014 The European Union (EU) legislated a new directive wherein large corporations are obliged to disclose non-financial information with regard to environmental, social and personnel matters, respect for human rights and the fight against corruption and bribery. A non-financial statement should be part of the management report. A separate report is also permitted if it is available simultaneously with the annual report or within six months after the balance sheet date. The EU member states have a couple of years to incorporate the directive into their national law (The European Parliament and the Council of the European Union, 2014).

The latest KPMG (2013) survey also reported that the Netherlands is the eleventh on the list of global N100 firms across 41 countries surveyed, despite being a small country. It was overtaken by U.K. for seventh place. The reporting rates did not change in the Netherlands. However, they dropped in the U.K. from 9% to 91% from 2011 to 2013. KPMG also reports the average corporate responsibility reporting quality among the global 250 firms the Netherlands scores were 69 percent while this ratio of U.K. companies was 76 out of 100 (KPMG, 2013). This makes the U.K. take third place for global large entities regarding sustainability reporting quality. These two countries will be further examined in this research regarding sustainability assurance tendencies. The U.K. and the Netherlands were selected because they have divergent legal systems, the former is considered a common law country

(11)

and the latter a code law country, consistent with Simnett et al. (2009) and built on Ruhnke and Gabriel’s (2013) research.

In the U.K. large and medium-sized enterprises are required to disclose non-financial information according to the performance by Companies Act.

Requirements for external reporting of Dutch companies are included in the Civil Code. This also provides insight into the contents of this report with regard to sustainability reporting, however to a very limited extent (Böhmer, Hoogendoorn, & Kruit, 2013).

2.3. Assurance on sustainability reports

This section will provide an overview of assurance practices of sustainability reports together with outlining the guidelines and standards.

Assurance on sustainability reporting is a relatively new field for practitioners and scholars, (Smith et al., 2011; KPMG, 2013) but despite of the current voluntary nature, the number of companies that employ social and environmental reports assurance (SERA) has been increasing since the last decade (KPMG, 2013, p. 33). It represents companies’ intention to provide an enhanced credibility and reliability of the CSR reports (Simnett et al., 2009; Pflugrath et al., 2011; Perego and Kolk, 2012, PwC, 2007; KPMG 2013). On the other hand The Global Reporting Initiative (GRI) report (2013) also describes that stakeholders’ demand is growing for assured sustainability disclosures and this call is also followed by the commission of assurance statements from companies as well as through the development of standards from regulators. With regard to the two countries of interest in this study, the U.K. and the Netherlands, the assurance of sustainability reports occurs on voluntary basis.

As Jones and Solomon (2010, p.22) point out there are many previous studies that have criticized SERA. They refer to Ball et al. (2000), for example, who question the independence and quality of SERA in the early stage of assurance practice (Jones and Solomon, p. 22). On the other hand Cooper and Owen (2014, p. 75) emphasize that these concerns were meanwhile addressed by the different guidelines and standards. Therefore in the next section these standards and guidelines will be presented and outlined.

There are multiple reasons that drive companies to obtain assurance over their sustainability reports, ACCA (2012, p.5) sums up:

• improve the credibility of the reporting among stakeholders (including investors) • ensure that what is addressed in the report is responding to the issues that are important and relevant

(12)

• improve the quality and reliability of their reported information

• identify areas where internal and external sustainability management systems could be improved, and

• improve operational and risk-management practices.

This is also in line with the KPMG (2013, p. 43) report results which showed that multinational companies’ general opinion for why they choose for assuring their sustainability reports was to augment credibility of their CSR data and to gain from it internally. Therefore the purpose of assurance statements serve a dual-purpose namely, to enhance the trust of the user regarding the reported non-financial information and to provide the preparer organization with an internal assessment of their reporting practice.

Simnett et al. (2009) investigated the companies’ motives behind the increasing tendency of voluntary assurance of sustainability reports. They make use of agency theory to describe the companies motivation for voluntary assurance of sustainability reports namely, to reduce the information asymmetry between the company and stakeholders. Simnett et al. (2009) also find that companies operating in stakeholder countries are more likely to have their sustainability reports assured, which is in line with stakeholder theory. They also found that in some industries (i.e., environmentally-sensitive industries) firms are more likely to commission assurance services in order to augment their credibility. Production companies were found as an exception from this. Furthermore they assert that companies that care to improve their credibility, assure their sustainability reports (2009, p. 961).

Ruhnke and Gabriel (2013) illustrate the voluntary demand of assurance services in three European countries (U.K., The Netherlands and Germany) and analyze the moderating effects of agency costs, applied reporting framework (GRI) and whether the firm has a sustainability department and how these components affect third party assurance endorsement.

Pflugrath et al. (2011) also argue that assurance adds more credibility to sustainability reports. They showed that financial analysts and investors perceived more credibility through audited reports provided by accounting firms. This credibility is also context specific and the need is influenced by industries and country specific factors. This is also in line with the KPMG survey findings (2011). The country of origin and ownership have an effect on CSR reporting and assurance practices are confirmed in most of the studies (Thorne et al, 2014, p. 691).

Casey & Grenier (2015) find that highly leveraged U.S. firms are less likely to commission assurance services on their CSR reports. They argue that in some industries high

(13)

regulation function as a substitute for assuring CSR reports, contradictory to the findings of Simnett et al. (2009).

2.3.1. Assurance guidelines and standards

The worldwide increasing stakeholder needs for accountability and credibility are reflected in the different CSR standards and guidelines (Golob & Bartlett, 2007, p. 8).

AccountAbility is an international membership-based organization established in 1995. AccountAbility conducts research, sets standards and also provides advisory services in the field of corporate responsibility and sustainable development. AccountAbility’s AA1000 series of standards are stakeholder centered and based on three principles, inclusivity, materiality and responsiveness, respectively (AccountAbility, 2008; Cooper and Owen, 2014) and focuses on broad firm performance. AA1000 distinguishes two levels of assurance, “high” and “moderate” (AccountAbility, 2008). As Cooper and Owen’s (2014, p. 79) findings inferred from surveyed FTSE 100 companies, the clear majority of the assurance statements referred to ISAE 3000 in 2010 and 2011.

The ISAE 3000 is the most commonly employed accountancy standard for SERA set by International Auditing and Assurance Standard Board (Cooper and Owen, 2014). This is considered a ‘business-centered’ and accountancy based guideline (Prego and Kolk, 2012). Under ISAE 3000 there are two assurance levels that can be provided, ‘reasonable’ and ‘limited’ assurance. These levels are concerned with in-depth, as well as in full range of tasks and procedures to produce an assurance statement (PwC, 2007).

According to the latest KPMG triennial survey, the GRI remains the leading sustainability reporting framework (KPMG, 2013, p. 30). GRI’s aim is to create an internationally accepted and applied framework and it provides “principles and guidelines” for reporting and for “ensuring report quality”. Furthermore companies can employ different levels of the GRI guidelines application (GRI, 2013).

Next to the international standards, country-specific ones have also emerged, for instance in Germany, France, Italy, Sweden, Spain and the Netherlands. The Dutch assurance standard for sustainability reports called COS 3410N was issued in 2007 and is based on ISAE 3000 (PwC, 2007).

(14)

2.4. Sustainability and financial performance

This has been an extensively researched area in the past couple of decades. For the last 40 years, previous studies have been extensively examining the relationship between sustainability reporting and corporate financial performance, but the findings are inconsistent and a consensus has not yet been reached (Margolis & Walsh, 2003; Orlitzky et al 2003, Wang & Choi 2013). There are negative associations reported and there are also findings that did not show an association. However most studies found a positive relationship between corporate social and financial performance (Orlitzky et al., 2003; Margolis & Walsh, 2003). Interestingly Orlitzky (2007) argues that the strength of this association is also affected by the researcher’s field of interest, whether it is economics, finance, accounting or social economics related. In a recent comprehensive review Peloza (2009) assess the literature regarding the link between corporate social and financial performance focused on metrics, between the years of 1972 till 2008. He also shows that the majority of studies (59%) found a positive CSP-CFP link, however the 22% of the examined studies reported mixed results. On the other hand they emphasize that the results of the examined practitioner reports showed a more positive relationship between CSP-CFP than academic literature did. Furthermore he asserts that the form of CSP has an impact on the relationship to CFP (Peloza, 2009, p. 1522). He further reports that financial performance has more impact on CSP than the other way around.

Clarkson et al. (2011) also found that environmental strategy enhances economic performance. However they argue that these environmental strategy findings can be mostly applied by economically well-resourced firms. This is also in line with the overall view that sustainability is a large company phenomena (e.g. Waddock & Graves, 1997; Simnett et al., 2009). Studies that found positive association, argue that sustainability creates competitive advantage to the firm in different forms, like enhanced reputation (e.g. Waddock & Graves, 1997; Clarkson et al, 2011). Harrison et al. (2010) emphasize the importance of integrating stakeholder perspective with strategic decision making, which can result in advantages for the companies which contributes to enhanced firm performance. Studies that found no association between sustainability performance and financial performance argue that the wide range of components and complexity give no reason to expect a relationship (Waddock & Graves, 1997). Those who found a negative relationship between CSP and CFP argue for competitive disadvantage because the costs of CSP outweigh the benefits. Therefore sustainability has a negative effect on a firms’ economic performance.

(15)

The direction of causation between sustainability and financial performance serves as an issue of discrepancy. Margolis and Walsh (2003) pointed out in their meta-analysis that the divergent findings regarding the relationship between CSR and financial performance was also effected by the choice of whether CSR was treated as an dependent or independent variable in empirical studies. In contrast with the aforementioned, Peloza (2009) observed that market-based metrics namely, share prices, are the most commonly used metrics. Consequently in this research, this contradiction will be taken into account and the relationship will be checked in both directions therefore, building on the study of Waddock and Graves (1997).

Because the majority of the studies found financially beneficial effects of CSR, probably because it helps establish better stakeholder relations, and considering the findings and opinions regarding the positive added value of sustainability assurance, I presume that the effects of CSR will be enhanced by obtaining assurance on sustainability reports. Therefore in the next section, I will present the hypotheses for this study.

2.5. Hypotheses

Based on the research question, indications and from the literature review I first examine whether there is a relationship between the assurance of sustainability reports and corporate financial performance. Then I will test whether this presumed association is positive namely, that assurance of sustainability reports has a positive effect on the corporate financial performance. Therefore the first hypothesis is as follows:

H1: An assurance statement on a sustainability report has a positive association with companies’ financial performance.

Furthermore prior research shows that more profitable companies are more likely engage in assurance on sustainability reports (Simnett et al., 2009; Ruhnke & Gabriel, 2013). Following the research approach of Waddock and Graves (1997, p. 307) that “corporate social performance is both a predictor and a consequence of a firm financial performance”, the direction of causation mentioned in H1 will be tested from both directions. The direction of causation between SERA and financial performance is also examined in both ways in this study because Orlitzky et al. (2003, p. 424) asserts that corporate social- and financial performance mutually affect eachother. Thus the following second hypothesis is stated:

(16)

H2: Companies with better financial performance are more likely to engage a third party to assure their sustainability report.

In the following section I will present the research methodology employed in testing these hypotheses.

3. Data and Research methodology

In this section I will show the research sample, the research method and the regression model used to obtain research results. The primary objective of this paper is to examine the relationship between assurance of sustainability reports and corporate financial performance. Therefore a quantitative research method is chosen to analyze this relationship, using archival data. The relationship will be tested by using empirical methods.

3.1. Data and sample

The observed data of publicly traded firms are collected between 2007 and 2014. This time frame is chosen to address Ruhnke and Gabriel’s (2013) suggestion to conduct panel analysis on assurance demand. Therefore the examined period of research has been expanded. The empirical analysis of Ruhnke and Gabriel (2013, p. 1087) considered three European countries for their comparison, namely Germany, the Netherlands and the United Kingdom. Based on their research and this study the largest publicly listed companies of two European countries will be compared, companies from the Netherlands listed on AEX AllShare and firms from the United Kingdom, listed on FTSE All-Share. The first 100 largest enterprises were considered from both lists based on their market capitalization. This resulted in a base of 200 companies.

To be able to examine the financial impact of assurance, stand-alone sustainability reports were searched in the Global Reporting Initiative database

(http://database.globalreporting.org/) and Corporate Register database

(http://www.corporateregister.com), following Simnett et al. (2009). Because of the missing

or inaccurate reports in the former database and the restricted access to the latter, corporate websites were visited and the available reports were searched to consider assurance statements. The financial data are retrieved from Compustat Global and for missing values Amadeus database was employed. Additional data was also gained from the financial statements and websites of companies to supplement the contingent absent data.

(17)

Companies that did not produce any stand-alone sustainability reports in the examined period are excluded from the sample. Following Ruhnke and Gabriel (2013) banks, insurance and real estate companies were not considered in the sample because of their unique manner of business activities and related regulations. After the reduction of the sample, a final sample of 732 firm-year observations remained for the period of 2007-2014.

To reduce the influence of extreme observations, the data were windsorized with the use of the mean of each continuous variable plus/minus two times their standard deviations. Values beyond this threshold are adjusted so the outliers are replaced with the next highest score that is not an outlier (Field, 2013).

3.2. Research methodology

The data are tested with regression analysis in order to examine the possible link between the assurance of sustainability reports and corporate financial performance. Consistent with prior literature this examination will be built on Waddock and Graves’ (1997, p. 308-309) research approach as a bases and the model will be applied differently with other variables based on other studies. In line with Waddock and Graves’ (1997) approach, in this study the relationship between the two variables will be investigated in both directions. Therefore financial performance will also be an independent variable in an additional test which is also calculated based on accounting and market measurements.

There are different views with regard to measuring firm financial performance and the opinions in the literature are mixed as to how to capture and reflect on it. Financial performance can be approached from an accounting- or market based viewpoint. Previous studies showed a positive association between CSR and firm performance in the case of using accounting measures (Margolis et al., 2007; Orlitzky et al., 2003; Peloza, 2009). Furthermore based on their meta-analysis Orlitzky et al. (2003) concluded that applied accounting- or market-based measures influenced the results of the association between corporate social reporting and financial performance. There were only a few studies that examined the association with both points of view. The majority employed one of the two; accounting or market perspective. In order to get a more compound view with regard to financial performance, both accounting-based and market based indicators will be applied. The mixed opinions, regarding the use of accounting- or market-based measures, also contributed to my decision to examine the relationship with both kinds of measures.

(18)

To examine the financial performance from an accounting perspective, return on assets (ROA) is calculated and furthermore an unscaled measure; net income is also tested as a dependent variable, following Barnett and Salomon (2012). They highlight that scaled measures can be influenced by independent variables and therefore distort the results. To address this concern the accounting-based perspective is measured alternatively as well. The market perspective is proxied by Tobin’s Q and market to book ratio (MTB), a parallel to prior research (García-Benau et al, 2013; Wang and Choi, 2013). The following models are composed and will be analyzed:

Tobin’s Qi(t) = β0 + β1 *ASSURi(t) + β2 *R&Di(t) + β3 *SIZEi(t) + β4 * LEVi(t) + β5 *COUNi(t)+

β6 *INDi(t) equation (1)

MTBi(t) = β0 + β1 *ASSURi(t) + β2 *R&Di(t) + β3 *SIZEi(t) + β4 * LEVi(t) + β5 *COUNi(t)+ β6

*INDi(t) equation (2)

ROAi(t) = β0 + β1 *ASSURi(t) + β2 *R&Di(t) + β3 *SIZEi(t) + β4 * LEVi(t) + β5 *COUNi(t)+ β6

*INDi(t) equation (3)

Net incomei(t) = β0 + β1 *ASSURi(t) + β2 *R&Di(t) + β3 *SIZEi(t) + β4 * LEVi(t) + β5 *COUNi(t)+

β6 *INDi(t) equation (4)

Based on the hypotheses, the following coefficients are expected: β1 >0, β2 >0, β3 >0, β4 >0

and β5 >0.

As previously mentioned to test whether the tested variables also effect the firms decision to get the sustainability reports assured, the following logit model will be tested, which is partially built on the model of Simnett et al. (2009):

ASSURi(t) = f (PROFITABILITYi(t) + SIZEi(t) + LEVi(t)+ R&Di(t)) equation (5)

To test the assurance probability, firm profitability measures are included in the models following the literature (Simnett et al., 2009; Ruhnke and Gabriel, 2013; Casey and Grenier, 2015). The same measurements were applied as in the tests of the first hypothesis: Tobin’s Q, MTB, ROA and net income/loss respectively. This implies 4 additional models based on equation 5. Furthermore size, leverage and research and development intensity are incorporated into the regression models.

(19)

Tobin’s Q is calculated as the ratio of firm sum of equity, long-term debt and net current liability divided by the replacement costs of tangible assets. The denominator is calculated as the sum of the book value of inventory and net value of plant, property and equipment (Dowell et al, 2000). It enables one to measure long term value of investments, argued by Surroca et al. (2010, p. 475). Another market-based ratio of the financial performance is measured by market to book ratio (MTB), that reflects how an enterprise’s shares are valued. Following Ruhnke & Gabriel (2013), ROA is calculated as earnings before interest and taxes divided by total assets and lastly net income is determined as earnings after interest, taxes, depreciation and amortization, given in or converted to million euros.

In equation 5, mentioned variable PROFITABILITY, is proxied by the above mentioned four financial performance measurements applied for the first four equations (i.e. Tobin’s Q, MTB, ROA and net income) that would be used separately to result in four more separate equations.

Assurance (ASSUR) is a dichotomous variable that takes the value of 1 if the stand-alone sustainability report is audited by a third party and takes the value of 0, otherwise. The stand-alone reports were considered as assured, where there were assurance statements available. Companies that have not produced stand-alone reports were excluded from the sample because most of these firms reported their CSR activities in their annual or integrated reports and therefore the effect of CSR reporting would not be clearly measurable.

Control variables

Following prior studies with controlling for variables that affect firm performance and sustainability reporting, company size, risk tolerance and the industry where the firm operates will be controlled for (Waddock and Graves 1997; Margolis and Walsh, 2003; Wang and Choi, 2013) together with year effects (Surroca et al., 2010; Barnett & Salomon, 2012). Firm size (SIZE) is proxied by the number of firm employees, measured in thousands (Waddock & Graves, 1997; Barnett & Salomon, 2012).

Risk tolerance by computing leverage (LEV) as long-term debt on total assets is additionally proxied (Dowell et al., 2000; Simnett et al., 2009). The model is expanded with an explanatory variable of “knowledge intensity” which is proxied by expenditure on research and development (R&D) (Wang and Choi, 2013, p. 426). R&D is considered as a factor that differentiates firms from others and which enhances performance, therefore this will be included in the model. Following Dowell et al. (2000) R&D intensity is calculated as

(20)

R&D expenses divided by total assets. Consistent with Ruhnke and Gabriel (2013) country-specific factors (COUNTRY) will be controlled. As Fortanier et al. (2011) summarize prior literature on the influence of country-of-origin on the firms’ sustainability reporting practice, the country in which a company is based determines it’s company culture and also the legal environment and politics results in differences among companies. Therefore this country of origin can influence the performance of companies and the variable is set to 1 if the firm is listed on the AEX, 0 if listed on FTSE.

To reduce any occurrence of significant autocorrelation arising from the use of panel data year and industry, dummies were added to the models (Wang & Choi, 2013). To filter out industry-level effects, given differences in stakeholder interest and industry-specified sustainability issues (Waddock & Graves, 1997), one digit SIC industry dummies were used. Based on the 4 digit SIC codes they were clustered based on the first digit into nine different industry categories.

4. Results

First I will present an overview of the sustainability reports, followed by descriptive statistics with correlations between the different variables. Lastly the results of the regression analyses are described.

Table 1 shows a summarized comparison of assured stand-alone sustainability reports within the listed companies in the U.K. and the Netherlands in the examined period. There are in total 543 sustainability reports published whereof 71,1% and 28,9% of them came from British and from Dutch companies respectively. The 62,6% of these sustainability reports in total were assured between 2007 and 2014, whereof the 72,9% of SERA derived from companies listed in London and 27,1% from firms listed on AEX. Up until 2013 the number of SERA increased, consistent with prior literature (i.e. KPMG, 2013) when a drop can be observed which decline is also in line with the number of published stand-alone sustainability reports. To gain a better insight into the data in the next section descriptive statistics and the correlations are discussed.

(21)

Table 1. Assurance onsustainability reports in Great Britain and the Netherlands between 2007-2014

United Kingdom Netherlands Totals

no SERA SERA no SERA SERA no assurance assurance

2007 22 26 8 8 30 34 2008 19 31 9 6 28 37 2009 21 29 7 12 28 41 2010 19 31 7 12 26 43 2011 16 38 10 12 26 50 2012 15 36 9 13 24 49 2013 13 36 9 15 22 51 2014 13 21 6 14 27 35 Totals 138 248 65 92 203 340 4.1.Descriptive statistics

Table 2(a) provides the descriptive statistics as well as the Pearson correlation matrix for the total sample regarding the dependent, independent and control variables employed in the regression tests. The table shows per variable the mean, standard deviation, the minimum and maximum windsorized values. The average value of Tobin’s Q is 4,53 which indicates that the market value of the companies are high above book value. The minimum and maximum values of this variable are -2,3 and 135,9 respectively. These exorbitant values are derived due to the fact that some firms did not have one of the components of the ratio, such as inventory, that caused this effect. This effect also applies to market-to-book ratio for the same reason. The average value of ROA for each of the firms in the sample lies about nine percent per annum. The highest net income amounted 11.047 million euros which is a windsorized value. However a mean value is merely a tenth of this amount, 1.251 million euros per annum, while the mean value amounted to 1.251 million euros, which explains the large sum of standard deviation in net income/loss. The largest company had almost 297 thousand employees whilst the average listed company from both of the countries counted employees during the examined period.

Moving on to the correlations, the correlations between the variables are weak since the values lie under 0,3. In contrast with the expectations, SERA is negatively correlated with both of the market-based performance measures to the same extent (Tobin’s Q and MTB, r=-0,14, p<0,01) however on a weak but significant level. On the other hand the

(22)

accounting-based measures have a positively and correlate significantly with SERA. Additionally, net income shows the strongest correlation with SERA (r=-0,17; p<0,01). This is also in line with the findings of Barnett and Salomon regarding sustainability and net income (2012, p. 1312).

Table 2 (a). Descriptive statistics and correlations for the whole sample

1. 2. 3. 4. 5. 6. 7. 8. 9. 1. Tobin’s Q 1 2. MTB -0,04 1 3. ROA -0,06 0,13** 1 4. Net income -0,08* -0,14** 0,32** 1 5. Assurance -0,14** -0,14** 0,11** 0,17** 1 6. R&D intensity -0,01 -0,02 0,21** 0,06 -0,12** 1 7. Size -0,07 -0,19** -0,08* 0,20** 0,10** -0,07 1 8. leverage 0,02 0,02 0,10** -0,09* 0,11** -0,22** 0,01 1 9. Country -0,06 0,31** -0,26** -0,16** -0,14** 0,03 -0,03 -0,13** 1 Mean 4,5285 0,0163 0,0928 1.251,03 0,46 0,0104 58,5619 0,2026 0,34 St. deviation 14,6821 0,049 0,0607 2.383,52 0,499 0,0204 71,0647 0,1206 0,475 Median 1,2077 0,0029 0,0817 421,5819 0 0,0002 29,8365 0,1906 0 Minimum -2,29 -0,1374 -0,082 -7.419,96 0 0 0 0 0 Maximum 135,85 0,3567 0,2464 11.047,32 1 0,0711 297,616 0,4771 1 ** Correlation is significant at the 0,01 level (2-tailed).

* Correlation is significant at the 0,05 level (2-tailed).

Moreover correlations are also used to identify whether there is any multicollinearity issue present. Multicollinearity is a phenomenon that can generate incorrect view and distorts the inferences drawn upon the regression model through highly correlated variables, because explanatory variables are weakly correlated and the correlation coefficients lie under 0,8. Furthermore the variance inflation factors (VIF) were examined as well and I have not found any multicollinearity problems in the data, since the VIF values are below the threshold of 10

(23)

(Field, 2013; Surroca et al, 2010). Hence it can be concluded that there is not a problem with multicollinearity.

Furthermore there are descriptive statistics calculated separately for the two countries that are shown in Table 2(b). During the eight year period all the four different financial performance measures are greater in the U.K. than in the Netherlands. One explanation for this difference lies in the fact that British companies are on average larger, as the size variable shows with a mean value of 61,6 thousand employees. This table also reinforces the statistics of Table 1 namely, that there are more assured sustainability reports for British companies.

Table 2(b). Descriptive statistics per country

4.2. Statistical analysis

The results of the regression analyses will be outlined in this paragraph. The first four equations were tested with hierarchical multiple regression and equation five and six are analyzed using logistic regression.

Tobin’s Q MTB ROA Net income Assur. R&D size leverage United Kingdom Mean 5,118 0,005 0,104 1.529,141 0,509 0,010 61,585 0,214 St. deviation 17,811 0,022 0,059 2.628,404 0,500 0,021 69,338 0,124 Minimum -2,290 -0,137 -0,036 -7.419,955 0 0 0 0 Maximum 135,853 0,142 0,246 11.047,319 1 0,071 297,616 0,477 The Netherlands Mean 3,432 0,037 0,072 733,924 0,359 0,011 55,685 0,182 St. deviation 5,018 0,073 0,058 1.733,331 0,481 0,020 74,649 0,112 Minimum 0,081 -0,012 -0,082 -2.837,392 0 0 0 0 Maximum 18,197 0,357 0,236 7.426,791 1 0,063 283,486 0,425

(24)

4.2.1. Linear regression results: financial performance and SERA

The results of the linear regression tests using the financial performance variables as the dependent variables, are provided in Table 3. The F-statistics of all the four models reported in the table are significant at the 0,001 level (FT=27,5; FMTB=13,1; FROA=10,1 and

Fni=11,070; p < 0,001 respectively), indicating that the explanatory variables of the models

are significant and explain well the different financial performance measures. However these explanatory powers are quite weak, the variables explain the model with Tobin’s Q the most, namely the 43,2% of variance in Tobin’s Q. The other three models’ goodness of fit are less appropriate and explain respectively 27,9% (MTB), 23% (ROA) and 24,7% (net income) of the variance in the different financial performance indicators.

Assurance has a negative association but the strongest explanatory power with Tobin’s Q and explains 16,8 percent variance in Tobin’s Q (β=-0,168; p<0,001). Out of the control variables R&D and leverage are significant. This model indicates that firms that assure their sustainability reports and are more R&D intensive have lower Tobin’s Q. This is in line with research implied market-based indicators to assess the CSR-financial performance relationship (Orlitzky et al, 2003). On the other hand enterprises with larger debt burdens have higher market than book value. Other control variables as size and country were not significant. These outcomes are against the Hypothesis 1 that assured sustainability reports results in improved financial performance.

However the model proxied by market-to-book ratio is significant, there is no association found with the main independent variable of interest (assurance) (β= -0,05; p>0,05). In this model the control variables size, leverage and country were significant, implying that larger firms have lower MTB ratio. An as it can be derived from Table 2(b) in this model Dutch companies have higher MB ratios than British firms. Furthermore country has the greatest explanatory power in MTB.

R&D has the most influence on ROA (β=0,344,; p<0,000), furthermore all the independent variables are significant in model 3. The results support Hypothesis 1, companies that are more R&D intensive, have larger debt burdens, are listed on FTSE and assure the CSR reports, have higher ROA, while larger firms have lower ROA. There exists a positive and significant link between financial performance and SERA (β=0344,; p<0,01). The results are in line with earlier research (e.g. Waddock & Graves, 1997; Orlitzky et al., 2003; Barnett & Salomon, 2012). As discussed above these variables explain 21% variance in ROA.

(25)

If a sustainability report is assured, it implies that net income will increase with around 398 million euros, which means a positive and significant effect between financial performance and SERA. Summarized, it can be stated that model 4 supports Hypothesis 1 as well. Size has the most effect on net income (β=0,273; p<0,000), so British, larger firms and companies with more “knowledge intensity” have higher net income.

Table 3. Linear regression results: financial performance and SERA

Dependent variable

Tobin’s Q MTB ROA net income

Constant 3,758 0,023 0,120 4.192,395 (1,077) (1,698) (7,070)*** (6,334)*** Independent variable: Assurance -4,944 -0,005 0,014 397,845 (-5,571)*** (-1,543) (3,179)** (2,363)* Control variables: R&D intensity -64,869 0,118 1,025 4.406,789 (-2,596)** (1,220) (8,269)*** (0,915) Size 0,002 -9,490E-5 0,000 9,140 (0,268) (-3,841)*** (-3,186)** (7,443)*** leverage 12,980 0,057 0,074 -2.615,471 (3,265)*** (3,785)*** (3,807)*** (-3,468)*** country (NL) -1,415 0,024 -0,024 -598,895 (-1,544) (6,834)*** (-5,414)*** (-3,444)*** Year and

industry dummies included included included included

No. of observations 732 732 732 732

R2 _0,448 _0,279 _0,230 _0,247

Adj. R2 0,432 0,258 0,207 0,224

F 27,488*** 13,096*** 10,103*** 11,070***

*** Correlation is significant at the 0,001 level. ** Correlation is significant at the 0,01 level. * Correlation is significant at the 0,05 level. t-statistics are in parentheses.

To sum up the linear regression tests, SERA had the most significant effect on the performance measure Tobin’s Q. Considering the fact that in the different models only the performance measures were changed and every other variables were the same, it can be concluded that SERA explained namely twice as much the variance in performance, measured by Tobin’s Q, than in other performance measures.

(26)

4.2.2. Logistic regression results: SERA and financial performance

Table 4 presents the results of the logistic regression using SERA as dependent variable. In all four models the key independent variable, financial performance indicators, varies. All four regression models for the whole sample proved to be statistically very significant (p< 0,001). Although the regression outcomes are very mixed. With respect to the main independent variables, performance measurements, Tobin’s Q, ROA and net income are significant (p<0,05). Tobin’s Q has a negative, ROA a positive and net income no effect on assurance, furthermore ROA has the most effect on assurance (β=4,702 , p<0,01). Regarding the control variables, R&D, size and country had significant effects in all models.

To determine the effect size, the odds ratios are studied. In model 5 all independent variables are significant (p< 0,05) which means that they make a significant contribution to the model. Overall the model highly significantly predicts whether external assurance is endorsed on a sustainability report (at p<0,001 level). This model also tells that if a sustainability report is assured than it is less probable that it is a Dutch company, in other words most of the assured reports are from the U.K. Financial performance proxied by Tobin’s Q lowers the chance by a factor of 0,88 of finding an assured report, this variable is negative and highly significant (p<0,001). R&D has the same effect on assurance, it is negative and highly significant (p<0,01 level). Moreover, companies endorse less likely external assurance on CSR reports if they are more R&D intensive. Size and leverage have also a highly significant but positive effects on assurance with the corresponding regression coefficient 0,006 and 2,062 (p<0,001 and p< 0,05) respectively. This outcome regarding size is in line with prior literature Ruhnke & Gabriel’s (2013) finding however they did not find leverage significantly affect assurance.

Model 6 includes the proxy MTB to measure firm performance. The variable MTB is negatively related to voluntary audit demand but not significantly (β = -3,371; p>0,05). Merely the control variables have significant influence on assurance in this setup. In this model leverage has the highest odds ratio (7,56) thus the highest effect on assurance which is positive, together with size. Taken together, this outcomes suggest that larger and higher debt burdened firms are more likely to seek external assurance. This is also logical considering the fact that high leveraged firms could try to provide more information regarding their operations to stakeholders in order to convince the investors about a good and stable operations.

(27)

The results of model 7 show that companies with higher ROA assure their CSR report more probably. This influence of the financial performance on variable Assurance is thus significant and positive (β = 4,702; p<0,01) therefore supports the assumption of Hypothesis 2 on a strong level. Except of leverage, the other control variables (i.e. R&D and size) are highly significant (p<0,001) although the influence of R&D is negative and size effects positively the audit demand.

Table 4. Logistic regression results: SERA and financial performance Dependent variable: SERA

Model 5. Model 6. Model 7. Model 8.

Constant 0,566 0,317 -0,331 -0,110 (1,762) (1,373) (0,718) (0,895) Independent variable: Tobin’s Q -0,126 (0,882)*** MTB -3,371 (0,034) ROA 4,702 (110,155)** Net income/loss 0,000 (1,000)* Control variables: R&D intensity -15,935 -14,533 -19,932 -15,511 (0,000)** (0,000)** (0,000)*** (0,000)** Size 0,006 0,005 0,006 0,005 (1,006)*** (1,005)*** (1,006)*** (1,005)*** Leverage 2,062 2,022 1,510 2,041 (7,862)* (7,557)** (4,527) (7,695)** country (NL) -0,531 -0,433 -0,395 -0,461 (0,588)** (0,648)* (0,673)* (0,631)** Year and

industry dummies included included included included

No. of observations 732 732 732 732

Log (pseudo likelihood) 852,989 891,530 884,315 889,145

Nagelkerke R2 0,259 0,201 0,212 0,205

Wald Chi-squared 157,791*** 119,250*** 126,466*** 121,635*** *** Correlation is significant at the 0,001 level.

** Correlation is significant at the 0,01 level.

(28)

The last model, with net income as proxy for performance measurement is also conform with the assumption that more profitable companies are more likely to endorse external third party assurance on sustainability reports. Similarly to the models 5 and 6 the control variables are significant as well. R&D influences negatively the assurance demand (β = -15,51; p<0,01) while size and leverage have a positive effect (β = 0,005; p<0,001 and β = 2,041; p<0,01 respectively) on SERA. Therefore in this case it can be also concluded that more R&D intensive firms are less likely to obtain SERA. However larger, highly leveraged and British companies are more likely to seek assurance to their sustainability reports. All four models show that Dutch companies are less likely to obtain assurance services on their sustainability reports (i.e. all regression coefficients are negative) and this effect is the strongest in model 7. This is also consistent with the results of Ruhnke and Gabriel (2013) where they found that the assured reports are at British companies assured to the highest extent irrespective the shareholder orientation to the country.

Lastly, the Nagelkerke R- square of the model of 0,259 reflects a satisfactory overall fit that makes model 5 with Tobin’s Q the best fitted among the 4 logistic regression models and that is the equation which predicted the best whether a sustainability report is assured.

To conclude the logistic regression analyses, the performance measure Tobin’s Q explained a slightly more (25,9%) variance in SERA than other performance measures (MTB 20,1%, ROA 21,2% and net income 20,5% respectively). However the differences between the performance measures are not so divergent as it is the case of the opposite direction of causation.

5. Discussion

In this study the relationship between sustainability assurance and financial performance has been empirically tested in both directions. This chapter discusses the regression results presented in the previous section.

First, this research examined whether assured sustainability reports enhance company’s financial performance, measured by two market-based and two accounting based measurements. However there was no relationship between Assurance and MTB ratio, the findings confirm a positive relationship in the case of financial performance measured by accounting-based indicators, ROA and net income, but indicate a negative relationship with market based ratios, Tobin’s Q. These empirical results are in line with prior literature as

(29)

Orlitzky et al. (2003) showed that the majority of the studies found a positive link between CSR and CFP. Furthermore the positive relationship of SERA on ROA and net income are in line with the findings of Barnett and Salomon (2012, p. 1314). However the adjusted R-square values of this study are lower than those of Barnett and Salomon (2012). I can be summarized that hypothesis 1 is accepted in the cases of equation 1, 3 and 4 and rejected based on equation 2. Moreover firms listed on FTSE are found to be significantly influencing the relation between sustainably assurance and financial performance, indicating that in case of obtaining SERA, British firms have higher performance. Concerning the control variables it is also found in this regression test that except R&D, size and leverage have significant effects on the financial performance measures, except leverage on Tobin’s and net income.

Regarding the four logit regression tests, how assurance demand is effected by financial performance, all models have also been found significant and valid. Moreover in accordance with the KPMG survey (2013) findings, firms listed on the London Stock Exchange are observed to assure their sustainability reports, which is in congruence with the research outcomes of Ruhnke and Gabriel (2013) and Perego and Kolk (2012). However this contradicts the findings of Simnett et al. (2009). They observed that firms from stakeholder oriented countries are more likely to produce sustainability reports.

Further results regarding the control variables are in line with the findings of Simnett et al., (2009) that larger and more highly leveraged companies are more likely to produce sustainability reports, in the case of this research, issue SERA. Moreover Assurance has the most impact and positive effect on financial performance, namely ROA in model 7. The associations with market based measures are in these regressions also negatively related to assurance. However, MTB ratio was again not significant. The negative association between SERA and Tobin’s Q is interesting because this ratio is a long term indicator of firm performance, and this result implies that when a sustainability report is assured it has a negative effect on firm performance and vice versa. This implication is against the majority of the findings regarding the relationship between sustainability and financial performance (Orlitzky et al., 2003; Margolis et al., 2007). On the other hand accounting-based measures have, in this setup a significant positive effect on voluntary external assurance, which is in line with Waddock & Graves (1997). Therefore hypothesis 2 is accepted based on model 5, 7 and 8, but rejected based on model 6.

Considering the direction of causation, it can be summarized thus that causation goes in both directions which is in line with the findings of Waddock & Graves (1997). However this

(30)

the case of Tobin’s Q as performance measurement. This causality is negative in both directions. This, in contrast to the findings of Peloza (2009), who reported that financial performance has more impact on sustainability. However, in the case of the other measures causality is about the same in both directions.

6. Conclusion

The non-financial aspects of organizations are becoming increasingly important in determining the value of an organization. The time that non-financial information is going to be as important as financial information is drawing nearer and nearer. Societal and environmental issues regarding companies operations have been increasing in importance over the past decades. Companies also began to address these issues and adjust their operations to a sustainable level as well as providing a wider range of information about these operations in addition to the traditional financial facts. The need for the reliability of these reports has also increased, which means that the users of the reports demand more credibility and transparency regarding sustainability information to gain a more comprehensive picture of an organizations overall performance (Cooper & Owen, 2014, p. 72).

In this study the driver of the company’s decision to seek voluntary external assurance on stand-alone sustainability reports was examined in the context of financial performance, proxied by accounting- and market-based measures. Furthermore the direction of causation was also examined, whereby SERA was tested as both a dependent and independent variable. Causation is found in both directions. The regression analyses revealed mixed results. Regarding the first hypothesis, that assured sustainability reports have a positive association with companies’ financial performance, assured sustainability report is found to have a negative effect on the performance ratio Tobin’s Q. This model was found to be the most significant, explained namely twice as much the variance than the models with other measures. Market-to-book ratio was not found to be associated with SERA. On the other hand accounting-based measures, ROA and net income are positively affected by assured sustainability reports, which is in line with the expectations and the majority of literature. The tests of the second hypothesis, that better performing companies are more likely to engage a third party to assure their sustainability report, showed similar results as the first hypothesis regarding Tobin’s Q and MTB ratio, except the fact that Tobin’s Q was slightly more significant than other performance indicators.

(31)

Altogether based on the regression results a positive relationship can be concluded between SERA and financial performance in both directions. Moreover the direction of causation in this relationship is found to exist in both directions. Furthermore, in line with prior literature, larger and more highly leveraged firms are more likely to engage a third party to assure their sustainability reports. Moreover it is also observed that firms listed on the London stock exchange are more likely to endorse third party assurance on their sustainability reports, despite the expectations of shareholder orientation. However the negative relationship between sustainability assurance and Tobin’s Q, is interesting because this ratio is a long term indicator of firm performance, and these results imply that when a sustainability report is assured it has a negative effects on firm performance and vice versa. This implication is against the majority of the findings regarding the relationship between sustainability and financial performance (Orlitzky et al., 2003; Margolis et al., 2007).

This study is also subject to several limitations, for instance firms operating in the financial and real estate sectors were eliminated from the sample. However there are studies that did not make this distinction such as (Simnett et al., 2009; Casey & Grenier, 2015). This offers future research possibilities to include those sectors in the sample as well or research them separately. Another limitation of this study is, that the sample focused only on large and very large companies. Also the sample was taken from two European countries causing bias and make, to some extent specific the inferences. However these limitations also make the particular point of view and verification of prior research. Future research could look at smaller, eventually not listed firms that are located outside of Europe.

Sustainability assurance and financial performance