MASTER
THESIS WITHIN: Business Administration
NUMBER OF CREDITS: 30 ECTS
PROGRAM OF STUDY: Civilekonom
AUTHOR: John Rydholm & Samuel Schultzberg Bagge
TUTOR: Fredrik Hansen & Toni Duras
JÖNKÖPING May 2020
ESG Ratings Subjectivity
and its Consequences
Rating Objectivity:
The Confusions in
Nordic ESG Ratings
i
Master Thesis within Business Administration
Title: Rating Objectivity: The Confusions in Nordic ESG Ratings - ESG Ratings Subjectivity and its Consequences
Authors: John Rydholm,
Samuel Schultzberg Bagge Tutors: Fredrik Hansen,
Toni Duras Date: May 2020
Key terms: ESG, CSR, Ratings, Rating agencies, MSCI, Thomson Reuters, Sustainalytics, RobecoSAM
Abstract:
Environmental, Social and Governance measurements have significantly increased in usage due to growing concerns for environmental and sustainability problems in today’s world. However, with no commonly agreed-upon criteria for ESG ratings, the scoring measure creates confusion both at the investor and company level. Besides, ESG agencies have different processes and parameters for measuring ESG compliance, which contributes to the problem. The study examines four ESG rating agencies’ rating models and ESG scores to get a better understanding of deviations in ESG scores among Nordic companies. By also studying the correlation amongst ESG scores and market capitalizations in firms, the paper hopes to shed light on if any relationships exist between them. Our results show that the four major ESG raters in the study showed a weak to a non-significant correlation against each other. The maximum correlation found was 0.419 between Thomson Reuters and MSCI. RobecoSAM and MSCI showed the lowest significant correlation at 0.291. Sustainalytics was detected not to show any significant correlation with the other raters. Correlation among market capitalization and ESG Raters was detected to not correlate to a greater extent. Only one ESG rater, RobecoSAM, showed a significant size to score-correlation at 0.278 with market capitalization. Thus, market capitalization does not seem to have any significant influence on ESG agencies’ decisions to set scores. Précising the study’s findings, the raters’ methods deviate from one another, but also how ESG raters make use of underlying factors.
ii
Acknowledgments
We would like to thank our supervisors Fredrik Hansen and Toni Duras, for the support they have given us throughout the process of producing this paper.
Jönköping University 2020-05-18
iii
Table of Contents
1
Introduction ... 1
1.1 Problem Background ... 1 1.2 Problem Discussion ... 2 1.3 Purpose ... 3 1.4 Research Question ... 3 1.5 Delimitations ... 42
Frame of Reference ... 4
2.1 Institutional Theory ... 4 2.2 Signaling Theory ... 5 2.3 Theory of Rating ... 62.4 Theory of Effective Regulation ... 7
3
Review of Related Literature ... 8
3.1 Definition of Environmental, Social and Governance ... 8
3.2 Definition of Corporate Social Responsibility ... 9
3.3 ESG ... 10
3.4 Rating Agencies ... 11
3.5 Divergences in ESG ratings ... 13
3.5.1 Rater-effect ... 14
3.6 ESG and its Effect on Firm Valuation ... 15
3.7 ESG Regulation within the European Union ... 16
3.8 Regulation of ESG Agencies ... 17
4
Methodology/Method ... 18
4.1 Collection of ESG Ratings ... 19
4.2 How ESG Raters Produce Ratings ... 22
4.2.1 MSCI ... 22
4.2.2 Thomson Reuters ... 24
4.2.3 RobecoSAM ... 24
4.2.4 Sustainalytics ... 25
4.3 Selection of Data ... 26
4.4 Management of ESG Data ... 28
4.5 Spearman Rank Correlation ... 29
4.6 Market Cap and ESG Score ... 29
iv
5
Empirical Findings ... 31
5.1 ESG Correlation Results ... 31
5.2 Market Capitalization and ESG Rating Correlation ... 32
5.3 Market Capitalization Differences ... 32
6
Analysis ... 34
6.1 Quantitative Analysis ... 34 6.1.1 Correlation ... 34 6.2 Qualitative Analysis ... 36 6.2.1 Qualitative differences ... 36 6.2.2 Input Sources ... 37 6.2.3 Peer Comparison ... 376.2.4 Rating Model Variables ... 38
6.2.5 Rating Methodologies ... 39
6.2.6 Change in ESG ratings ... 39
6.3 Regulation ... 41
7
Conclusion ... 42
8
Discussion ... 43
8.1 Practical Contribution of the Study ... 45
8.2 Further Research ... 45
9
Ethical Considerations ... 46
10
References ... 48
11
Appendices ... 53
11.1 Appendix A ... 53 11.2 Appendix B ... 54v
List of Tables
Table 1. Population Characteristics ... 27
Table 2. Common Sample characteristics ... 27
Table 3. Transformation of MSCI AAA-C Scale Into 1-100 ... 28
Table 4. Spearman Rank Correlation Test – ESG Raters ... 31
Table 5. Pearson Correlation – ESG Rating/Market Capitalization ... 32
Table 6. Spearman Rank Correlation – ESG Rating/Market Capitalization ... 32
Table 7. Descriptive Statistics over the data samples. ... 32
Table 8. Largest and Smallest Firms (Market Capitalization) in Overall and Common Sample ... 33
Table 9. Overview of Findings by Rating Firms ... 36
Table 10. Nordea Bank and Pandora’s ESG score changes over time 2015-2018. ... 39
Table 11. Thomson Reuters - Environmental, Social and Governance data, 2015-2018 ... 40
Table 12. Total population of companies ... 54
List of Figures
Figure 1. The Common Sample of ESG Scores ... 21Figure 2. Population and Common Sample, Visualization ... 28
Figure 3. Distribution of Firm Size in Common Sample and Population ... 34
Definitions
CSR = Corporate Social Responsibility
ESG = Environmental, Social and Governance SRI = Socially Responsible Investing
1
1
Introduction
1.1 Problem Background
With water levels, global temperatures, and wildfires all on rapidly rising levels, investors are trying to steer financial flows in order to do their part to solve the world’s problems or, at least, create the least harm possible (NBIM, 2019). Investors are now demanding legislative action and are pushing the frontier of Socially Responsible Investments forward (Eccles and Klimenko, 2020), which in turn drives further institutional investment growth and interest.
Showing the magnitude of the problem, the European Central Bank, which is the world’s second-largest central bank, controlling direct assets under management of €14.098 billion, stated that ignoring the environment is a risk to the future of the financial stability in the euro area as well as in the world (European Central Bank, 2020). The central bank ended the communication, stating that they were ready to stimulate new purchases of green bonds, which clearly sets the current situation and the level of attention to Environmental, Social and Governance (ESG) factors (Farand, 2020; Lagarde, 2020). The interest in ESG does not only come from the financial superpowers such as the ECB, but retail investors are actively pushing their capital into exchange-traded funds. ESG compliant funds saw a total inflow of €29 billion in the period Jan – Jun 2019, which compared to the previous year has already surpassed the full-year figure. The growth can be compared to the overall growth rate of 8% in fund inflows in the general fund industry, with 12% growth in ESG compliant alternatives (Lyxor, 2019).
The demands of investors and stakeholders created pressure on the industry and legislative powers to create some sort of measurement system in order to rank the suitability of environmental, social, and governance positive investments (Avetisyan & Hockerts, 2017; Hill et al., 2007). The ESG measurement became an industry standard to supply the needs of the market. The scoring system is used by financial institutions, central banks, and investors to try to grasp the level of good their investments do to the world.
The suppliers of ESG ratings range from niched ESG-research companies, clearly relying on their ratings as a primary source of income, to large banks with total ownership of its rating agency branch, where their ESG-rating services only counts for a part of their revenue stands for the total coverage in the market. The industry has been accused of over consolidation, where all capital flows are flowing through ESG ratings from a few rating agencies with profit maximization goals (Avetisyan & Hockerts, 2017).
2 In the current state, concern for climate change and the environment around the globe has escalated, making ESG a burning topic for investors that want to be at the forefront of sustainable investing. In a changing climate of business, the field of ESG and sustainable investing have become more mainstream with each passing day, thanks to the shifting opinion on the environment. The interest of the public, but also investors and stakeholders, has led to more green financial options to take part in the new era of green investments.
With both the retail investors and top financial powers allocating a large amount of capital towards ESG alternatives, the importance of the accuracy of the measurements, the objectivity, and motivations behind the rating agencies become ever important.
1.2 Problem Discussion
The current providers of ESG-ratings have divided the market into their specific corner, where often a specific provider serves a broader marketplace as a stock exchange or research firm with a greater publicity reach. It has led to a few agencies that have a great deal of control over ESG-specific information.
In the current state, there are no commonly agreed-upon criteria for ESG, which creates confusion at an investor level where one agency could have different processes and parameters for measuring ESG compliance. A study by Escrig-Olmedo, Muñoz-Torres, and Fernandez-Izquierdo (2010) have found that raters base their analysis and measurement on questionnaires, public reports, and news which in turn could pose threats to the objectivity in the measurement.
Another problem that arises is the lucrativeness of providing ESG ratings. Offering a paid service to investors and stakeholders may lead to more problems due to biases in how the ratings are set, mainly if there is an incentive for raters to sell their services. Hence, the question arises if raters want to provide correct ESG ratings or if it could pose a new problem, where the ESG market deviates away from its original purpose. If the industry sidetracks too much, the interest and trust in the score could deplete and make the rating obsolete in the worst case.
With an increase in ESG compliant equities and funds, the issue of possible delusion for investors, society, and companies (Berg, Kölbel & Rigobon, 2019) becomes more vital as time passes since there is more and more capital per ESG-rate.
3
1.3 Purpose
By examining the rating models, ranks, and conduct of the different ESG rating agencies, we intend to shed light upon how the models differ and what discrepancies in terms of correlation that can occur on Nordic companies’ ESG ratings. The specific aspect of the problem includes the rating agencies’ objectivity of rating models in terms of standardization of measurement and conflicting interest as a discussion point.
With more information and groundwork done on Nordic companies’ relationship to ESG-rating, we believe that institutions, policymakers, and investors can take a more informed stance. With rationality on their side, questioning ratings, and demanding transparency, it will generate an objective way of comparing companies in their environmental, social, and governance processes. It may also be helpful when evaluating the individual parts of ESG, environmental, social, and governance, to improve on the understanding and effectiveness when using ESG as a driver for investing regulation. However, it may be challenging to achieve a commonly agreed upon ranking from raters. With more research on the subject, we hope to contribute to ESG ratings being as objectively and as accurate as possible.
1.4 Research Question
This thesis sets out to answer three questions regarding the field of ESG. The first question:
- Does the ESG ratings for Nordic publicly traded companies correlate on a significant measure?
Answering the first question will make us examine our findings in order to deliver a conclusion, which naturally will lead us to our second research question:
- How do the rating agencies’ models differ?
Because of the evolving nature in this field, we will also set out to compare two companies ESG timeline data in order to see the progress of ratings and how the rating firms interpret the same company with its respective company event:
- Given two different companies' timeline and company events, how do they differ in terms of ESG rating over the time period 2015-2018?
With our three research questions being intertwined, our conclusion will provide an answer to if ESG scores diverge and if it is due to the ESG raters themselves. Furthermore,
4 we want to explore the field of ESG with these questions to provide a conclusion on how to move forward within the field of ESG.
1.5 Delimitations
Particular limitations of the study present itself in both the data collection, but also in the ESG rating market in itself. As demand for ESG is a rather new phenomenon, both data and ratings on the subject are hard to come by. ESG raters undoubtfully operate within a beneficial and profitable market, and thus, ratings, statistical data, and the underlying measurements are scarcely provided by raters without payment. With only incorporating public data by raters, it has somewhat hindered a larger common sample size from being collected. Hence, the given data sample is as good as it gets with public data without providing payment to ESG raters, and within the framework of the thesis.
2
Frame of Reference
This chapter contains theories surrounding the area of ESG. To build our foundation, which we argue from and use as our vantage point, we will use the combination of theories presented in this chapter to explain why and how discrepancies in the field through a theoretical perspective.
Since the ESG area of research is a newly developing policy-driven field with no precise frameworks of theory. We, as other authors in this field, such as Escrig-Olmedo et al. (2019), will incorporate theory from different fields of research to explain behavior and nature in line with previous ESG research.
Theories on corporate governance areas such as Effective Theory of Regulation combined with Institutional theory are viewpoints that are crucial in order to get an understanding of the motivations and structures in the field. To explain the differences of ESG ratings, we will use both Signalling theory as well as the Theory of Rating in order to explain the nature of rating models.
2.1 Institutional Theory
Institutional theory is a general theoretical framework, and with the assistance of more niched theories when analyzing the interconnections between the rating agencies, its models and the customs come into focus.
5 As stated by Scott (2005), institutional theory examines the main procedures under which rules and structures are established within the guidelines of social behavior. Institutional theory is not comprised of only one unified structure of propositions and assumptions. Instead, it is built upon a collection of theoretical approaches and perspectives. Peters (2016) argues these to be Sociological, Rational Choice, Historical, Empirical and Discursive Institutionalism. Furthermore, Scott (2005) explains that institutions are considered structures of governance, where they adhere to social conduct rules. Institutions are also reluctant to change due to inertia within their structure of operating. In a study on the subject conduct by Knetter (1989), the author found evidence that pointed to institutions reacting differently to similar challenges.
The Sociological part of the theory touches on the change in behavior and mindset in individuals joining new institutions (Peters, 2016). It also mentions that actors within the same area of policy typically share values in the structure of governing than in areas with more comprehensive policies. Furthermore, Béland and Cox (2011) state that ideas and norms play an essential part in steering on a governance level. Rational Choice works as an essential part of the implementation and formulation of policy rules and incentives relating to the neoclassical program in economics. Moreover, Peters (2016) continues to explain that Historical Institutionalism focuses on the development of theory and regulation within governance. Changes in public policy, displacement, and layering lead to new ways of implementing and adopting new measures within the governance field. The Empirical pillar might be less empirical driven, according to Peters (2016). However, it argues that structure within governance is highly important. Discursive institutionalism, according to Schmitt (2010), focuses on communicative and coordinative disclosures. The former is used to relate an institution to the environment it operates within, and how to explain the actions it portrays to an extensive range of stakeholders. Coordinative disclosure, instead, defines the purpose of an institution.
Hence, when all pillars of the institutional theory are combined, it describes the structure surrounding governance in institutions (Peters, 2016). Besides, it provides an insight into issues concerning governance types and regulations. We will try to incorporate institutional theory to explain the fundamentals of why ESG differs and how they adhere to social structures and rules within our society.
2.2 Signaling Theory
According to Connelly et al. (2011), the Signaling Theory successfully explains the behavior of two parties with different access to information and how they act and
6 influence each other. The theory will be used in at least two steps since this paper investigates the ratings of ESG-raters. Hence, ratings can be seen as a form of communication between the firm being rated and the rating agency. Further, this applies similarly to the communication of ESG rating information to the end-user by agencies and firms.
The Signaling theory is mainly based on neoclassical microeconomics, in particular, what is known as information symmetry. Information asymmetry exists when there is an asymmetric distribution of information among agents (Stiglitz, 2002). It is the base for Stiglitz’s work on information asymmetry for which he received the Nobel Prize in Economic Science in 2001 (Rosser, 2003). In the frame of this thesis, information asymmetry can arguably arise in three steps. The first step is when publicly traded companies communicate their information to rating agencies and the public through marketing, financial reports, and general communication. The second step, as defined by Stiglitz (2002), creates friction in the understanding of the raters, as in the third step, between raters and end investors or decision-makers.
Asymmetry in information is particularly important in two types, quality and intent. Elitzur and Gavious (2003) describe Quality Asymmetry as when one part is not fully aware of the complete characteristics of the information provider, as well as intent asymmetry is when the information receiver is concerned with the supplier’s intent. Signaling theory has, in the past, been used to explain information communication of high-quality firms and low-quality firms (Kirmani & Rao, 2000). However, as a suitable substitute, we will use the theory to explain communications by high-level ESG firms as well as level ESG firms. Hence, High-level firms profit from transparency, and low-level firms instead profit from low transparency according to the Signaling Theory.
2.3 Theory of Rating
Since this paper is based on the performance of the rating agencies and the objectivity of their models, we find that the theory of rating is a useful theory for explaining potential discrepancies in ratings between ESG agencies.
The theory is applicable to ratings in general (Wherry & Bartlett, 1982). According to the authors, Wherry and Bartlett (1982), rating theory is built upon three major components. The performance of the ratee, observation of the performance, and the recollection of the observations by the rater. We, however, deem that the first two components as important for our theoretical framework. Wherry and Bartlett (1982) state that the performance of a
7 rater can be linked to the capability to influence and affect the environment. The observations done by raters can be concluded as a combined formula. The formula may include a combination of observation biases, ratee performance, and rater observations. Moreover, the authors explains that rating theory and its development is due to the change in factors that influence the variables in their theory formula. However, in his original notes, Wherry (1951) concludes that the theory is not fully complete, and never will be. Nevertheless, in unison with other theories and frameworks, it will provide a good ground for further research.
By combining Wherry and Bartlett’s theory of rating with the theoretical framework provided by the rating agencies, we will examine their models with the academic theory.
2.4 Theory of Effective Regulation
Regulation is usually ineffective at its core, according to Llewellyn (1940). The problem lies in the combination of inferior policy choices in addition to bad design choices. Therefore, the construction of effective regulation to stem a problem is an essential step when passing new legislation. Theory on how law becomes effective provides a structure for how future legislation and standards in the environmental rating area can be improved. It holds true for both existing legislation, but also for new laws and standards.
Feaver and Benedict (2015) state that the fundamentals of regulation are to change behaviors by guiding, modifying, and prohibiting social practices. How the control will be applied will determine the burden on the regulatory systems in terms of costs and administration. The allocation of the burden, however, can be distributed over both the regulatory body, but also private entities, such as ESG raters and businesses, affected by the legislation. Furthermore, Feaver and Benedict (2015) explain that effective regulation requires coherence between psychological levers and the method to produce changes in behavior.
Moreover, The European Commission has presented ways of optimizing their way of passing legislation called “Better Regulation.” The objective of the initiative is to ensure a more transparent and inclusive decision-making process within the European Union. The report puts emphasis on strengthening their preparation, ensuring regulations adhere to their purpose, and better implementations of laws (European Commission, 2015). It also aims to make regulation more accessible to the public by further enhancing the transparency and ease of access. The European Commission (2015) also states that it must increase the clarity where unintended and disappointing results in economic, social, or
8 environmental areas have occurred. The intention of the initiative is not about the change in the amount of legislation, but instead to focus on better existing and future regulation. With the framework in place, we will try to identify how future regulation and standards might present itself to improve the ESG rating market going forward.
3
Review of Related Literature
3.1 Definition of Environmental, Social and Governance
ESG or Environmental, Social, and Governance, is commonly referred to as sustainable investing and a part of the upcoming field of sustainable finance, according to Cherneva (2012). With the use of an ESG analysis, Escrig-Olmedo et al. (2019) describe that agencies evaluate firms and calculate companies’ rate of sustainability in the form of aggregate scores. In the evaluation process of firms, ESG emphasizes the business model, but also how the organization’s services and products deal with progress in sustainability, as stated by Elbasha & Avetisyan (2018). Likewise, enough data collected on the environmental, social, and governance aspects in a firm, can be utilized to calculate an aggregate ESG-score. In turn, this gives stakeholders and investors an indication of how sustainable a company is. Furthermore, Duuren, Plantinga, and Scholtens (2016) explain that ESG creates a tool that will help stakeholders decide which sustainable bonds and equities to buy or not.
The Environmental aspect of ESG concerns an organization’s environmental footprint. According to Limkriangkrai, Koh, and Durand (2017), whether it is cutting greenhouse gas emissions, increasing their energy efficiency, or water use, all contribute to the environmental measure of ESG. Also, the environmental factor considers how a firm makes use of natural recourses in its production and supply chains. Thus, the attributes included in the environmental part of ESG is screened for efforts to decrease its impact on the environment (Hanson, 2013). However, ecological aspects are difficult to measure and can be very subjective, according to Montabon, Sroufe, and Narasimhan (2006), which will pose a problem when implementing a score.
The social factors, according to Limkriangkrai et al. (2017), focus on the human and labor aspects of a firm. One question that could arise within the social aspect according to the authors, is wheter a company adhere to labor standards and human rights issues or not. The authors continue to explain that with a more integrated way of operating in local communities and its social ecosystem, an organization can improve on their score. When it comes to the social spectrum of ESG, there are a couple of serious challenges firms face
9 both in the short and long term. New workforce requirements and laws might be demanding to apply, especially in the short run. Geopolitical attributes also contribute to the social score, concerning both the production and distribution phases, according to a report by S&P Global (2020).
Moreover, labor issues and disputes can also be listed as social attributes, according to Hillman and Keim (2001). However, problems are mostly present in the relationship between a firm and outside parts such as governments, banks, and investors (S&P Global, 2020). Branch (2019) also concludes that the quantification of social issues might be even harder to asses due to an array of different factors, which would indicate an increase in the divergence between raters.
Governance sets out to define the principles and rules regarding the way a firm is operated. It implies both the responsibilities and expectations of the firm by stakeholders (Hanson, 2013). An adequately constructed corporate governance system will work as an instrument and, in turn, will support an organization and its strategy in the long term. A governance system will also help with balancing the interest among stakeholders (Siebens, 2002). Governance applies to a significant part of the structure of a company; stakeholders, the board of directors, and managers. The rise of problems, such as board diversity, how the board of directors takes part in the firm, or how compensation and supervision of executives are conducted, has come to play a significant role in the structure of corporate governance (Khan et al., 2018).
3.2 Definition of Corporate Social Responsibility
Like ESG, Corporate Social Responsibility or CSR came to be, after an increase in demand for environmental action in firms, by investors but also other stakeholders such as suppliers and employees, according to McWilliams and Siegel (2001). However, with several different goals and ideas from stakeholders, the definition of CSR might be challenging to pinpoint. Jones (1980), on the other hand, states that Corporate Social Responsibility can be characterized as the obligations a corporation has towards the social and environmental aspects.
Sheehy (2015) highlights CSR as private regulation, thus supplying a framework for companies internationally. Corporate Social Responsibility and its political agenda offer a vital contribution to improving difficulties and obligations for companies on the environmental spectrum. Moreover, Lindgreen & Swaen (2010) provide information on what amounts to responsible behavior and describe that CSR often signifies a continuous devotion to improving and behaving ethically. All this, while growing on its economic
10 development in addition to social improvements. Rondinelli & Berry (2000) adds that CSR also includes a firm’s efforts to reach a broader range of ecological and social issues.
3.3 ESG
ESG became popular in the wake of the newfound concern for social and environmental issues. ESG stands to be a measurement of environmental, social, and governmental risks to an investment in a company (Avetisyan & Hockerts, 2017).
According to Galbreath (2012), stakeholders, from investors to governments, have reaped the advantages of the introduction of the rating through the requirement of truthful information regarding both financial performance and the individual parts of ESG, Environmental, Social, and Governance. Both ESG and Social Corporate Responsibility (SCR) rating firms can be viewed as a direct response to the growing demand for both environmental and social information of firms. The demand has derived from stakeholders wanting to apply a more sustainable way to invest, as stated by Avetisyan and Hockerts (2017).
According to Elbasha and Avetisyan (2018), the primary purpose of an ESG rating firm is to assess and dissect the performance of a company’s sustainability and environmental impact. The methodology of an ESG rating firm builds upon the research each rating agency has conducted. With the immense knowledge and data, ESG rating companies possess, has turned the ESG industry into a central part of the sustainability performance valuation for not only firms and financial markets, but also regarding academia.
Elbasha and Avetisyan (2018) continue to describe that the rating market for sustainability has grown exponentially in the last ten years. The growth has contributed to a change in the way that ESG companies are seen. ESG rating firms began with being projected as economic actors but have now emerged into social actors instead. This, in turn, has led to a shift in the behavior of other social entities.
However, Escrig-Olmedo et al. (2019) argue that if ESG rating firms do not try to apply and internalize the social impact and only function as economic entities, it could pose a risk to the industry. The information presented on what a sustainable firm and how its environmental performance is measured, could, therefore, be seen as misrepresented. Providing disingenuous data on the sustainability of firms could have a negative effect, including the deterioration of trust in both companies and ESG rating agencies, which could deteriorate the ESG industry’s social legitimacy as a whole. Hence, there is a solution to the problem, according to Escrig-Olmedo et al. (2019), if the level of expectation of both ESG-rating firms and society is at a comparable level.
11 Abdelkafi and Täuscher (2015) state that the current environment also mandates companies to strive towards advancement in their sustainable development. Corporations today are aspiring to accomplish a higher level of corporate sustainability through the use of business strategies to be able to apply new sustainable practices into their daily operations. The balance amongst environmental, social and economic requirements is therefore essential to corporate sustainability
In the wake of the financial crisis in 2008, the perception and attitude of corporate sustainability saw an optimistic improvement in the capital market (Lopatta & Kaspereit, 2014). Moreover, the ESG industry experienced a process of concentration in terms of the number of agencies in the decade to follow. The diminishing number of firms and growth of the sector can mainly be traced back to the investment decision-makers and financial stakeholders being the drive to the change (Saadaoui & Soobaroyen, 2018; Scott, Cocchi & Campbell-Gemmell, 2014). According to Scalet & Kelly (2010), this only shows that the growth and popularity of sustainable finance and green investing.
3.4 Rating Agencies
The influence of the ESG-rating sector has taken a more influential position over the financial market actors and their behaviors (Slager, Gond & Moon, 2012). Moreover, Elbasha and Avetisyan, 2018 describe that ESG has also had a significant impact on sustainability management in firms and its progress towards becoming more institutionalized. With the increase in the influence of agencies of the financial markets capital flows, Escrig-Olmedo et al. (2019) emphasize that companies are trying to change their ways of doing, for better or worse, and therefore generates an impact on our society. Furthermore, the authors clarifies the obstacles ESG-agencies have faced in their attempt to assess firms according to their methodologies. Since ESG-raters are corporate entities, and their service provided is ratings, the transparency and the lack of it create difficulties. According to Escrig-Olmedo, Muñoz-Torres, and Fernandez-Izquierdo (2010) and, Saadaoui and Soobaroyen (2018), agencies are reluctant to provide public information regarding how their process of assessing, and how an aggregated result is calculated. Thus, giving private entities, such as investors, fund managers, and organizations, a more difficult time forming comparisons between scores for the same firm. Hence, ESG-raters also has to consider the preference of stakeholders. ESG ESG-raters, explained by Windolph (2011) and Escrig‐Olmedo, Muñoz‐Torres, Fernández‐Izquierdo & Rivera‐ Lirio (2017), disregard the preferences of investor and stakeholder in their process of assessing and evaluating firms, and therefore, could affect both the usefulness and the
12 acceptance of ESG-industry in the long run. Another problem for agencies to tackle is the trade-offs in relation to criteria.
In their study on sustainability rating scoring, Escrig-Olmedo, Muñoz-Torres, Fernández-Izquierdo & Rivera-Lirio (2014), defined the practices ESG raters use to calculate the ESG-scores, where an attribute with a low score may be compensated by a factor with a better score. The authors performed an exploratory analysis to shed light on the correlations among four sustainability factors: operational performance, strategic intent, engagement, and governance and management. The four factors are a part of “The Accountability Rating.” The results from their study showed that a significant correlation was observed among the factors. However, Operational performance was found to keep a low correlation. Hence, the authors state the need for special treatment of operational performance in future assessment progressions of rating companies. Further, Atkinson (2000) stated that the development of sustainability is both an intricate concept, where interpretation plays a considerable role. Escrig-Olmedo et al. (2014) added that progression of sustainably becomes more complicated with the measurement of CSR and its scores.
Chatterji et al. (2016) examined the convergences in six Socially Responsible Investing (SRI) raters for three years, 2004, 2005, and 2006. In their study, the authors make use of both Pearson and Spearman correlation tests to unravel if Raters’ SRI scores correlate. Their findings presented evidence of dissimilar theorization and low commensurability. According to Chatterji et al. (2016), low commensurability implies that raters still disagree on ratings, or in other words, raters possess significant measurement errors. There stands a possibility for ESG raters to measure the same attribute in different ways depending on their methodology, hence commensurability. In the likelihood of an inconsistency in the assessment of ESG-scores by agencies, CSR may not truly be applicable, according to Chatterji et al. (2016). The authors continue to discuss the validity of SRI ratings, where they conclude that investors and stakeholders ought to be careful when basing their decisions on ratings that diverge. Chatterji et al. (2016), also concludes that different use of theory and methodologies among raters lay the ground for the low similarities in scores
Moreover, individual scores for each environmental, social, and governance category are provided by the majority of raters, as said by Liern & Pérez‐Gladish (2018). However, Liern and Pérez‐Gladish (2018) found that some evaluators did not provide an overall ESG-rating on the sustainability performance of a firm, thus not describing a corporation’s performance score. Therefore, the last challenge for some ESG raters is the absence of a total score.
13
3.5 Divergences in ESG ratings
With studies on the subject of ESG being scarce to our knowledge, the article by Berg et al. (2019), is an essential study for this paper.
In their paper on ESG ratings, Berg et al. (2019) found that correlation among raters ranges between 0.42 to 0.73, with an average correlation of 0.61 in their firms tested. The study was made up of ESG scores and underlying indicators on firms from five ESG agencies. The authors were also allowed access to measurement protocols and aggregation rules on how indicators were used. The population among the raters included between 1668 and 4551 companies over the five raters, with a mutual sample of 823 firms. Berg et al. (2019), concludes that ratings from ESG providers do diverge to a larger extent. In the analysis of the results, the authors clarify that 53% of the correlation can be explained by the variance in the measurement. 44% could be described by scope, and 3% by weight. The measurement deviation denotes how agencies, using indicators, measure the same features in different ways. Hence, labor practices of firms may be assessed based on an indicator such as labor cases against a firm or workforce turnover in an organization. The two different aspects can be assessed but will probably produce results not alike. Indicators may have their emphasis on processes or outcomes.
The data analyzed by agencies originate from numerous sources, for example, public data sources, media reports, or company reports. Additionally, Berg et al., (2019), mentions that they assume ESG-agencies are using indicators that they presume to be the most accurate, but the attribute measured is the same. Further, Berg et al. (2019) state that rankings can provide a better way of measure than using individual ESG scores. Disagreement on individual ESG scores would, therefore, be less relevant than a firm’s placement in relation to other companies. The consequences of diverging scores, in return, marks skewed information, on which decision-makers base their assessments.
There are three critical aspects to be aware of if the information is not reported correctly. The first problem that Berg et al. (2019) bring to light is how the performance of ESG, and its inability to be accurately reflected in bond and stock prices. Investors will, therefore, have a more complicated process when assessing and recognizing firms and bonds that lags behind the curve or out-performs the market. Besides, Fama and French (2007) concluded that the preference of investors could profoundly impact asset prices, but only if a significant portion of stakeholders and investors have a strong will to adopt a uniform non-financial preference. Moreover, Berg et al. (2019) conclude that when a majority of stakeholders prefer to adopt ESG performances, the discrepancy of the ratings will affect asset prices.
14 The second problem, deviation in ESG scores will have a negative impact due to the mixed signals it sends to a firm. A company will have a more difficult time attempting to understand what causes a negative or positive effect on its ESG performance. The last point, according to Berg et al. (2019), is how divergence has an impact on empirical research where the difference in ESG scores might impact the result of a study. When combined, the uncertainty regarding the different ESG scores will be a disadvantage when investors and stakeholders are trying to make an informed decision when making environmental and sustainable contributions. Combining the divergence of the three aspects will yield an aggregate rating that is the final ESG-score (Berg et al., 2019). We may observe that various agencies include different types of attributes, which may lead to different final scores by each ESG-agency (Berg et al., 2019).
Berg et al. (2019), however, concludes that three primary sources were found to contribute to discrepancies in ESG ratings. The divergence of weight, scope, and measurement. Scope divergence includes the sets of features that are utilized as the baseline to build scores. Greenhouse gas emissions, lobbying, and human rights are attributes that can be integrated into the scope of a rating. Weight divergence, on the other hand, states how ESG-agencies assessments the importance of features included, but also the relative performance of every element. Furthermore, it also considers how and if a characteristic compensates for another attribute (Berg et al., 2019).
According to Sadowski (2010), there are countless of rating schemes assessing comparable factors with the use of unique methods to calculate a final score. Hence, the problem may be the fact that the field of sustainability has not reached mainstream status yet, while the nature of the domain is still much complex.
3.5.1 Rater-effect
When trying to do an assessment and ranking of companies, agencies are subjected to something called the rater-effect. The phenomenon comes into effect when evaluating ESG attributes, where ESG-raters must make judgment calls regarding characteristics, such as human rights or labor practices (Berg et al., 2019). The implication of the rater-effect is that there is a correlation among the decisions of ESG-agencies. According to Berg et al. (2019), when the decisions of a firm were positive, there was also a high chance for other indicators to be positive too. An explanation of this phenomenon can be traced to the fact that ESG-agencies are coordinated by companies instead of indicators. The authors conclude that a company that is perceived as a respectable firm will tend to be looked at in a favorable way. The firm would then receive an indicator score that would tend to be better, which otherwise would not be the case, and vice versa.
15 Besides, Chatterji et al. (2016) clarify that the problem in rater bias, rater effect, and diverging scores, can be helped with transparency, to understand where ratings originate. Chatterji et al. (2016) continue to explain that divergence in scores in the social and environmental responsibility sector should be observed with care due to rater bias. In diverging scores, investors will have a challenge distinguishing the “correct” score among ratings, thus may allocate their capital in wrong equity.
Furthermore, Berg et al. (2019) enlighten that certain ESG-raters create an environment where it is tough to acquire a good score. If a firm chooses not to provide an answer to some questions collected by the rating agencies, the outcome would then be a lower score. Hence, the willingness to report and reveal company information to ESG-agencies will have an impact on the rating and, therefore, may clarify to some extent why the rater-effect exists. Therefore, the rater-rater-effect suggests that there is a positive correlation in the inconsistencies among categories within an ESG-rater.
3.6 ESG and its Effect on Firm Valuation
According to Fatemi, Glaum & Kaiser (2018), there are observable differences in that ESG’s strengths increase the value of a firm where the opposite is true for concerns in ESG. The data in their study is assembled from KLD Research, and Analytics on how to categorize CRS activates, in addition to Bloomberg, which supplied the data points on 403 US firms’ ESG scores. Further, data was collected over the years 2016-2011, which provided the study with 1640 firm-year observations. ESG disclosures were found to have a negative impact on firm valuation. Moreover, the disclosers of ESG can reduce the adverse valuation effects of ESG concerns. Fatemi et al. (2018), explains that it could be since disclosers provide a method for companies to legitimate their performance, the effectiveness of ESG policies, and their operations when addressing stakeholders. Also, Fatemi et al. (2018) state, it is because companies convince stakeholders that the firms have finalized reliable vows to make changes in their way of operating, and therefore, overcome the ESG weaknesses of a business.
Moreover, according to Giese et al. (2019), it might be difficult to differentiate whether ESG ratings lead to better valuations or vice versa. Besides, Krüger (2015) notes that the route of causality amid the correlation between firm valuation and ESG scores is somewhat diffuse. ESG ratings that are located on the higher end of the spectrum, with the help of both a lower cost of capital and a less systematic risk, thus influence valuation for the better.
16 Giese et al. (2019), continue to explain that ESG does affect both the performance of a firm and its valuation. Furthermore, the transmission between the characteristics of ESG to economic value is, therefore, considered a multi-channel process. Giese et al. (2019) emphasize that the intensity of ESG scores was indeed worse than factors considered more traditional, low volatility, and momentum, for example. Hence, ESG has a relatively small financial impact per unit of time, according to Giese et al. (2019, but lasts for several years, whereas momentum and volatility are better suitable for factor investing, which makes use of a straightforward one-dimensional strategy.
Additionally, a report by OECD (2017) provides evidence that ESG factors can have an impact on financial performance. An ESG analysis can, therefore, play an important role in addition to regular economic analysis. However, OECD (2017), also specify that it might not always be as straightforward to interpret the outcomes of ESG risk. This, in turn, makes it especially challenging to integrate ESG measures into financial models.
3.7 ESG Regulation within the European Union
With more pressure to implement the regulation in the field of ESG assessments, the European Union released newly revised laws in 2013, regarding transparency, in the form of a new directive, Directive 2013/50/EU. The new ruling is an update to the former directive, 2014/109/EC. The new 2013 law, in addition to the decisions from the law from 2014, includes a better emphasis on the disclosure of information regarding social and environmental matters. Moreover, ESG disclosures are required together with obligations to disclose financial reports. Hence, the new directive put a more considerable prominence on the transparency requirements of firms in the EU region (Camilleri, 2015). Furthermore, Camilleri (2015) explains that in 2014, the introduction of non-binding guidelines in the form of the Accounting Directive, 2013/34/EU, was initiated. The law touches upon the improvement of what non-financial information public companies need to disclose. It applies to organizations operating within the borders of the European Union. Also, it includes insurance companies and banks. These new guidelines on non-financial reporting are expected to bring light to subjects such as environment, corruption, and human rights. Besides, the law was introduced to encourage firms to make use of applicable non-financial performance measures, for example, air and water pollution, and greenhouse gas emissions.
Camilleri (2015) continues to convey that new regulation in the European Union is mostly designed for larger corporations and organizations. Hence, the European Union is highly influential in the way CSR and ESG policies are conducted.
17 In their pursuit to improve its sustainability policies, the European Union Commission issued three proposals for new regulations in May 2018. The measures include suggestions to; produce a sustainable finance taxonomy for the European Union, improve on the disclosing of sustainability risk involving sustainable investments, and set new benchmarks for carbon emissions. The proposition also puts a great emphasis on the new regulation in terms of disclosure. Hence, this was done to promote asset managers and stakeholders to incorporate environmental, social, and governance factors into their progress of decision making. Besides, increased transparency responsibilities for financial intermediaries when supplying information on sustainable risk and investments to end-investors. The proposal aims to decrease the cost of searching for both sustainable investments, but also enable a more straightforward way to compare sustainable finance products, according to Spinachi (2020).
In a report by the Official Journal of the European Union (2019), to be able to cope with the increase in ESG-ratings and sustainable finance, the European Parliament passed new laws in November 2019. The new regulations will take effect over the coming next coming year to bring further clarity to ESG benchmarks. The new legislation will improve on and ad to a number of existing regulations. According to the report, the Commission is to present findings to the Parliament of the European Union on the practicality of “ESG benchmarks” before December 2022. The report should consider approaches to sustainability and the development if its disposition in addition to proposals of new regulations.
3.8 Regulation of ESG Agencies
Quite remarkably, in contrast to new reporting-standards for non-financial information in the EU, ESG-agencies do not seem to be heavily regulated yet. In a news article on the future of ESG, Jones (2020) stated that in a speech at the European Financial Forum, in February 2020, Steven Maijoor, Chair of European Securities and Markets, talked about how the EU and governments need to improve its supervision since the current ESG-regulation was “far from optimal.” Maijoor also stated, “where ESG ratings are used for investment purposes, ESG rating agencies should be regulated and supervised appropriately by public sector authorities.”
Nevertheless, Maijoor is not the only lobbyist alerting about ESG regulation. According to a report by Riding (2019), the chairman of the European Securities and Markets Authority, Robert Ophèle, acknowledged that the growing appeal in ESG, in addition to the lack of regulation, could lead to a situation where EU member states start undercutting each other on ESG. Ophèle also stated that a need for a minimum requirement would be
18 “crucial to avoid the concept of ESG being watered down” (Riding, 2019). Furthermore, new ESG legislation is being implemented in the EU in the coming years. However, Ophèle further explained that he thinks the EU should speed up its efforts in implementing new regulations since the need is urgent (Riding, 2019).
With a boost in the number of new investment-related alternatives focusing on ESG measures have provided an easy way to make greener investments. OECD (2017), also claims that they face not just behavioral, but also operational and technical issues due to the difficulties in incorporating ESG strategies and its effects. Nevertheless, OECD (2017) explains that policymakers play a part in the future of making ESG accessible. With encouraging institutional investors to enhance their knowledge on both ESG issues but also setting guidelines for how ESG products are to be selected, some problems would be eliminated. Moreover, OECD (2017) clarifies that some regulatory measures have been taken to combat the difficulty of ESG measurements. Hence, many hurdles still exist not only regulatory but also behavioral and practical challenges.
4
Methodology/Method
This thesis will concern a quantitative, positivistic the explorative nature of the study is based upon preliminary second-degree data. To further improve our reasoning, we will also incorporate a qualitative analysis where we will make use of the rating agencies' own framework in order to form an appropriate assessment of the more qualitative parts in the rating agencies' methods.
The quantitative approach will include an analysis and a numerical measuring of ESG data collected. With the study, we want to identify any patterns that can be observed between ratings set by the ESG agencies and, thus, draw conclusions from the results. Data will be collected from reliable and applicable sources, and thereafter, processed and analyzed. The thesis will focus on studying the relationship between the scores ESG raters set and incorporate previous statistical data and experiments in the form of peer-reviewed articles and research in addition to our data and results. Consequently, we deem a qualitative approach (Saunders et al., 2016) to be suitable for this study.
The study will incorporate a positivistic approach and utilize scientific evidence of ESG data from raters. A positivistic approach is fit for a qualitative method of research (Saunders et al., 2016), which this thesis will build upon. According to Bryman (2011), a positivistic approach in research revolves around both quantitative and objective methods, in addition to data. By incorporating previous studies, theories, and gathered data, this study will perform statistical tests, attempt to analyze the outcomes, in order to
19 conclude the research question at hand. Furthermore, a study that contributes to already established research is viewed as a positivistic method of research (Bryman, 2011). Moreover, with our choice of a positivistic approach that builds upon previous research and data, which we hope to contribute to, we deem our choice of approach to be motivated.
Prior research on the subject, including transparency of financial information, ESG’s effect on firm value and performance, and current regulation regarding the issue, will play a role in what makes ESG scores differ. In addition to the ESG scores, we want to gather data and information on the current and future situation regarding regulation in the Nordics and European Union. This data will be collected from various applicable sources. Hence, previous research will support our interpretation of our results to conclude how the ESG market will move forward.
Analysis and interpretation of the data collected and processed will contain comparisons between the ESG raters. By comparing the ratings between raters, we will try to create a better picture of how the ratings differ and how they raters’ decisions correlate.
4.1 Collection of ESG Ratings
For the collection of data on how ESG data differ between companies, we set out to make observations among ESG rating firms to be able to gather essential data for our analysis. Our approach of collecting and analyzing data began with the process of gathering our central portion of data points. First, we started with researching and inventory what firms and companies set ESG ratings. In our findings on what firms set ESG ratings, we observed around ten major raters that provide ESG data. Among the ten firms, we only found four raters that provided data without payment, namely MSCI, Thomson Reuters, RobecoSAM, and Sustainalytics.
Our reasoning regarding the selection of these four rating firms can be concluded to the fact that these rating agencies provided information for free to the public in some form. Hence, the ESG service market is new and upcoming, and some agencies a reluctant to provide information regarding ratings (Escrig-Olmedo et al., 2019). Therefore, these ESG-raters were chosen for our portion of data. Hence, raters that offer ESG ratings for free might have an advantage in having their score more readily available. Their scores could potentially affect the market at a more significant level than firms with “hidden” ESG data.
20 In our study, we have set out to investigate the ESG ratings on the Nordic markets and have opted to collect data on the region of the Nordic countries. Since environmental issues have become a significant subject in the eyes of Nordic companies’ stakeholders (Rodrigue, Magnan & Boulianne, 2013), we believe that Nordic countries are a good ground for our data collection.
To reach a suitable number of observations, four countries within the Nordic region were selected. Our selected countries of data collection can be concluded to be Sweden, Denmark, Norway, and Finland. With data on Iceland being scarce, we decided to leave them out of our research.
According to Egeberg and Trondal (1999), even though Norway is not part of the European Union, it has agreed on a trade agreement to get access to the internal market of the EU, also known as the Agreement on the European Economic Area. Hence, in order to get access, Norway and EEA countries have to adhere to the internal market regulations, such as environmental protection, social policy, and consumer protection. When selecting our sample of firms in our research, we opted to keep our selections to the major stock markets in the four Nordic countries. Unlisted companies are not included in the study since data on such firms’ ESG data is harder to come by to our knowledge. For Sweden, companies selected for data retrieval were all included in the OMXS index, whereas the Norwegian companies were chosen from the OSEAX. The Danish stocks were selected from the OMXC, and the Finish chosen firms were listed on OMXH. Hence, data collection for ESG is a bit troublesome since transparency in ESG-ratings is quite undeveloped (Escrig-Olmedo et al., 2019), we opted to make use of more prominent firms listed on these stock exchanges. The firms we have come to select with all scores available is presented in Appendix A and illustrated in Figure 1.
The four rating firms providing ESG data can be viewed as reliable sources for financial data. The data for MSCI and RobecoSAM were collected from their respective ESG websites (MSCI, 2020; RobecoSAM, 2020). Data from Thomson Reuters was collected through Refinitiv DataStream. The information gathered for Sustainalytics was obtained through each firm’s respective page on the Yahoo Finance website (Yahoo Finance, 2020).
In Figure 1, we observe our common sample of firms and the four different raters. Each firm's ESG ratings in our common sample are plotted on the graph to present a visualization of the four different ESG ratings for the same firm. The specific scores are presented in Appendix A.
21 Besides, for the data collected via MSCI and Refinitiv DataStream, we were able to obtain historical data points in addition to the latest set of data. In the collected data, rating changes over time for two rating agencies are present, which can be observed in Table 10. To further grasp the differences in scores, the study will also try to examine changes
A P MOLLER MAERSK AKER BP ALFA LAVAL ASSA ABLOY ASTRAZENECA ATLAS COPCO BOLIDEN ORD SHS CARLSBERG CHRISTIAN HANSEN HOLDING COLOPLAST DANSKE BANK DNB DSV PANALPINA ELECTROLUX ELISA EQUINOR ERICSSON GENMAB GJENSIDIGE FORSIKRING H LUNDBECK HENNES & MAURITZ HEXAGON ICA GRUPPEN INDUSTRIVARDEN INVESTOR ISS KINNEVIK KONE LUNDIN PETROLEUM MILLICOM INTERNATIONAL CELLULAR MOWI NESTE NOKIA NOKIAN RENKAAT NORSK HYDRO NOVO NORDISK NOVOZYMES ORKLA ORSTED PANDORA SAMPO SANDVIK SCHIBSTED SECURITAS SKANDINAVISKA ENSKILDA BANKEN SKANSKA SKF STORA ENSO SWEDISH MATCH SVENSKA HANDELSBANKEN TELENOR TELIA COMPANY TRYG VESTAS WINDSYSTEMS VOLVO YARA INTERNATIONAL 0,00 20,00 40,00 60,00 80,00 100,00 Rater: MSCI RobecoSAM Sustainalytics Thomson Reuters
22 in scores for two individual firms, Nordea and Pandora. The two firms were selected due to the differences in their total score, but also how they change over time. The two firms operate in different business models and industries, which would provide a better overall picture. However, with some underlying data of the ESG score in Thomson Reuters and MSCI, the study will try to get an understanding of what could have caused the changes in scores between the raters, with the help of the two raters’ respective rating method. Through the collection of such information, we will be able to make a minor comparison on the development of the ESG ratings for the two firms.
All data collected for our sample were the latest available from each rater. Thus, the latest values were collected. In the data collection of timelines, for Thomson Reuters, Pandora’s score over 2019 was not available, hence the exclusion of 2019 from the timeline analysis. The data from Thomson Reuters were collected on 2020-02-07 from Datastream, MSCI, on 2020-02-10. RobecoSAM and Sustainalytics were gathered on 2020-02-14.
4.2 How ESG Raters Produce Ratings
Due to currently lacking standards in the rating process, a shattered workflow is apparent in the different rating agencies, since both business models and statistical models differ in their rating procedure. However, what is equal for each rating agency is that they must start with the given information, decide which parameters to analyze, quantify variables that are not typically quantifiable, analyze the data and come up with a final ranking. In this chapter, the study will try to explain the four raters’ method procedure for creating ESG scores.
4.2.1 MSCI
The method MSCI is using to rate a firm on ESG risks is created like a funnel, where they gather multiple thousands of data points on management, operations, and policies in order to have a data set for analysis. According to MSCI (2019), their ESG Rating Model seeks out to answer four questions:
First and foremost, MSCI wants to answer what the most substantial ESG opportunities and risks a company is encountering. The next part of their rating model concludes to what degree of exposure to the risks and opportunities the firm is. The third point tries to answer how well the company is dealing with the main opportunities versus the risk. Lastly, MSCI looks to the overall picture of the firm and how the firm measure against its industry competitors MSCI (2019).
23 To answer these questions, MSCI retrieved their data from governments as well as through company disclosures, such as annual reports and sustainability reports. Finally, to get event sources, they monitor local and global media channels. The data gathered goes through both an exposure metric and a management metric. It is done to decide what risks the firm is facing and how management is dealing with risk. These risks are industry-specific and are based on policy and performance metrics for each industry.
MSCI then agrees on 37 key issues for each industry and analyses how well each firm is dealing with these issues, where each issue is weighted according to their own mapping framework. In the final step, the rating agency ranks and gives relative scores compared to industry peers to arrive at an ESG rating, which makes this score relative and not applicable for analyses over different industries.
Material ESG risks and opportunities is a large part of their rating model, where the company tries to quantify how each industry is a victim of costly environmental, social, and governance issues. Where a risk is deemed as a substantial cost in connection with the business model and opportunity, the industry and company could capitalize on an ESG trend or issue. Material ESG risks are based on the classic pillars with a subset of themes for each pillar, such as environmental component has natural resources, climate change, and ecological opportunities as themes. Furthermore, the social pillar has human capital, product liability, and social opportunities as themes. The final pillar, governance, has corporate behavior and corporate governance as central themes.
From the themes, MSCI chooses applicable key issues and creates weights to weight them according to their importance and cost to the business. The conduct of deciding on weight is set up in a two-dimensional matrix on the Y-axis “Level of Contribution to Environmental or Social Impact” as well as on the X-axis “Expected Time frame for Risk/Opportunity to Materialize.” The largest weight is assigned to issues with a short timeframe and high risk where the highest score weighs three times more than a more extended timeframe with lower risk.
Corporate Governance is the only wide-spanning factor that is not changed per industry and is based on raw data that is filtered through key metrics in board metrics, pay metrics, ownership, and control as well as accounting metrics. The score in the parameters is then brought to a percentile rank, and the firm gets a total governance score.
From the scores governmental and material ESG risks, the firm assigns a company score on a level 1-10 whereas MSCI then groups and translates the 1-10 score into a letter rating AAA to CCC in seven letter-categories.
24
4.2.2 Thomson Reuters
Eikon by Thomson Reuters is the database where the company records its ESG ratings. The ESG ratings by Eikon was previously named ASSET4 but has been improved by reducing manpower to reduce the human error factor as well as automating the workflow in order to handle more quantitative data. By achieving a quantitative approach, the rating agency now controls for size and impact of each analyzed category.
However, their process is still mainly operated by a trained analyst that gathers ESG data in the forms of stock filings, CSR reports, news sources, company websites, and Non-Governmental Organization websites and annual reports. The data is collected in their database, where then a bottom-up approach is utilized to analyze each company. The databased is then tested with 400 built-in error checks in order to ensure quality data, according to Thomson Reuters (2017).
The ESG scores are calculated out of 400 data points, where the most relevant 178 data points are chosen to set a scoring. These measures are based on the type of industry, data accessibility, and specific material data drivers. The 178 data points are divided into ten categories, which form an ESG performance on the environmental, social, and governance factors. In order to give each category, the right amount of attention, the data points are weighted per pillar, category, and indicators. The most substantial category is under governance and management, whereas the lightest is the social category, human rights.
After having an overall score based on materiality and industry benchmarks, the company in question is then screened for controversy regarding ESG in 23 controversy measures, which are measures that negatively affect the company in media regarding the classic ESG parameters.
The rankings in their database Eikon by Thomson Reuters come in the form of scores between 0-100.
4.2.3 RobecoSAM
RobecoSAM’s ESG analysis consists of both quantitative and qualitative data, according to S&P Global Ratings (2019). The data, according to themselves, consists of short term and long-term risk analysis, which is gathered by interaction with the analyzed entities management. RobecoSAM does not reveal the specific method or methodology behind their rating, but they do provide an explanatory evaluation guide. According to the firm,
25 the measurement is a combination of two factors, “ESG Profile” and “Preparedness,” which combined becomes the ESG evaluation score on the 0-100 rank.
The ESG Profile is a firm’s observable risk in environmental, social, and governance factors. Each profile has factors specific to the industry, where the firm is operating, and different sectors are assigned different risk ratios between environmental and social risks. The rating firm categorizes risks per sector categories where different sectors are assigned different risk ratios between environmental and social risks. Oil and gas sectors rank in the highest risk in both social and ecological risk according to S&P Global; on the contrary, real estate operators assess the lowest amount of risk. The agency also includes country risk as a factor in their models, in which natural disasters and political uncertainty are the underlying factors. RobecoSAM arrives at a score on a scale from 0-100. Moreover, the ranks from the rating agency are a comparability rank within each firm’s industry, and the agency has the power to change the rating if they regard a firm work as positive or negative, whatever the past data gathering says, according to S&P Global (2019).
For the gathering of the observable data, RobecoSAM is sourcing their data from highly trusted sources which include Global Reporting Initiative, the Greenhouse Gas Protocol, Climate Disclosure Project, the Taskforce for Financial-Related Climate Disclosures recommendations, the US Occupational Safety and Health Administration reporting metrics, and the World Resource Institute’s definitions of water accounting according to S&P Global (2019).
4.2.4 Sustainalytics
The rating agency provides clear and explanatory data on their methodology and process in their rating method. Sustainalytics analyzing process stems from a quantitative standard where the scores, according to the agency, are directly correlated to which ESG risks the firm has in its business.
Sustainalytics rating method is built upon three pillars. Where the data they have collected through annual reports, official sources, and interviews are put into one of three components to support the ESG score. The first pillar is corporate governance and the most substantial explanatory factor for the general ESG score since governance is general and applicable to all industries. According to Sustainalytics (2019), unmanaged corporate governance risk accounts for 20% of the total ESG score.
26 The second pillar, also called Material ESG, is environmental, social, and governance risk that is due to managerial oversight as CO2, labor safety, and diversity among the
workforce. The specific material risk can be modified to fit the ranked firm’s business and sector. Sustainalytics explains that there are two components of the second pillar, exposure, and management. Where the company gets scored both on the ESG risks facing the firm, exposure, and how well management is managing the risk, management. The industry-specific exposure assessment process starts with data gathering through quantitative data in the form of incidents reports, water usage, and other industry-specific data points. After accessing the industry-specific exposure, the Corporate View is analyzed through the Global Reporting Initiative guidelines, which is a universal standard for ESG reporting (Global Reporting Initiative, 2016). Passing the Corporate View comes the Expert View where Sustainalytics sources expert insight in each industry and the Sustainability Accounting Standard Board, for industry insights.
The final, third pillar accounts for idiosyncratic risks. Unsystematic risks are, according to Sustainalytics (2019), risks that unpredictable for a business and its industry. When an idiosyncratic issue materialized, the event only affects the company in question and not the subsector industry. The third pillar weight is event-driven, which makes it hard to set a specific weight that it has on the ESG score.
Through the whole rating process, the models are built so that industry risk gets ranked in each parameter. Thus, the risk score is absolute, meaning that the rank of a company in one sector is comparable to any other industry. The rating arrives at a score on a scale of 0-100.
4.3 Selection of Data
To select quality data for our analysis, we set out with the objective to find the most relevant data for our research question. Since we want to answer how much the rating industries’ rankings correlate, we first ranked the most impactful and largest market actors in order to see how the consensus is formed.
