ESG and Risk-Adjusted Performance

ESG and risk-adjusted performance

An empirical and comparative analysis on ESG-fund performance relative to their conventional counterparts in Scandinavia

Cassandra Myhre

Master’s Thesis

University of Oslo

Department of Economics

Supervisor: Jin Cao

Date: May, 2021 Preface

This thesis marks the end of my Master’s degree in Economics at the University of Oslo.

I want to thank my supervisor, Jin Cao, for his excellent support and guidance throughout the process. I would also like to thank ﬁnance major, Edmundas Lapenas, for his invaluable insight in ﬁnancial data analysis. Our many in-depth discussions have made this thesis possible.

I hope that my humble contribution in analysing the ESG effects on risk-adjusted returns in the Scandinavian market will be considered as a valuable source and incentive for further research in this ﬁeld.

Cassandra Myhre, May 2020 Abstract

There is an accepted principle in the ﬁnancial industry that the bigger the risk, the higher the potential reward. The introduction of sustainability in the investment equation has however revolutionised the risk-return tradeoff. Not only with regards to creating value but also because it has ruptured the misconception that responsible funds provide lower returns. Companies who have incorporated Environmental, Social and Fair Governance (ESG) factors into their investment portfolios have shown lower volatility compared to their peers in the same industry. Furthermore, numerous studies have shown that ESG also generated higher returns in the long run, meaning that not only does it potentially minimise risk, but it is also a more proﬁtable choice. In this paper I will conduct an empirical and comparative study of the risk-return tradeoff where I will analyse ESG funds and conventional funds based in Scandinavia, with a sample developed specially for this thesis. The prime goal is to see whether ESG has any affect on risk-adjusted returns. By using traditional performance measurement tools, I will evaluate the portfolios and investigate their risk- adjusted performance. Table of Contents 1. Introduction 1

1.1 Structure of the thesis 2

2. Background 3

2.1 History 3

2.2 Case 4

2.3 Why the Nordics? 5

2.4 Scandinavian ESG Markets 7

2.4 Approaches 7

2.5 Short-termism 9

3. Literature review 9

3.1 Social Responsibility and Financial Performance 9

3.2 ESG factors and risk-adjusted performance 10

4. Theory 13

4.1 Shareholder Theory 13

4.2 Stakeholder Theory 13

4.3 Modern Portfolio Theory 13

4.4 The efﬁcient frontier and ESG 15

5. Empirical methodology 17

5.1 Treynor ratio 17

5.2 Sharpe ratio 18

5.3 Jensen’s Alpha 18

5.4 The Capital Asset Pricing Model (CAPM) 19

6. Data 20

6.1 Research design - sample selection 20 6.2 MSCI Nordic as benchmark 21

6.3 ESG dataset - Morningstar Sustainability Rating 22

6.4 Conventional portfolio dataset 22

6.5 Risk-free rates 22

6.6 Concerns about the dataset 23

7. Results and discussion 23

7.1 Non-ESG portfolio data analysis 23

7.2 ESG portfolio data analysis 25

7.3 Comparing the results 26

7.4 Discussion 27

7.4.1 Response to a market crisis: COVID-19 29

8. Conclusion and further research 30

9. References 32

Appendix 34 1. Introduction In the past decades there has been a growing demand from various investors to see sustainability issues reﬂected in their investment choices. Environmental, Social and (Fair) Governance (ESG) investment practices are particularly used as it has shown to potentially have better long-term performance (so far)1 compared to conventional investment opportunities in the same industry. Not only has it shown to be in some industries a more proﬁtable choice in the past years, but the ESG practices also lead to less reputational risk as it is less vulnerable to political and other controversial policies.

There has been an increasing amount of papers published on the performance of sustainable funds, however much less has been done to evaluate the actual link between ESG-investing and risk- adjusted performance. Furthermore, how can this be compared to the conventional, non-ESG funds in the same industry? Is it possible to demonstrate that ESG-practices lead to less risk while yielding the same or higher reward? This thesis attempts to investigate the link between ESG factors and risk in the Scandinavian fund market by analysing the following empirical questions:

1) Does ESG-investing have an affect on risk-adjusted returns? 2) Do ESG-funds outperform their conventional counterparts in the risk-return tradeoff in the Scandinavian market?

It is important to distinguish the difference between ESG and socially responsible investing (SRI), which should not be used interchangeably. SRI originally was developed to help investors build up a sustainable investment portfolio. This entails strategies such as negative screening and exclusions of companies that would not live up to the investor's ethical standards. Other strategies have been developed such as positive screening. The main difference between ESG and SRI lies in the prioritisation of return and values. While with SRI investors actively eliminate funds on a set of specific ethical standards, the main objective of ESG remains financial performance with an overlay of social consciousness. In other words, even an oil and gas company can be considered responsible if they are working to reduce emissions in its operations. Due to this core difference, it is more interesting to establish how much risk do ESG practices entail and if there is any statistically significant differences compared to non-ESG practices.

1 There is an extensive amount of literature published on the financial performance of ESG funds, however, as the topic is quite new, the findings remain somewhat ambiguous. 1 It also seems that the extent of research done so far suffers from the endogeneity problem, meaning that it is uncertain if a fund's choice of whether or not to invest in ESG is an endogenous decision. Therefore, most of the estimates from simple regressions are quite biased. However, the outbreak of COVID-19 gives us a chance to eliminate this issue. As the ESG characteristics were already fixed before the outbreak, by looking for differences in various variables, such as return, it is possible to measure the resilience of the funds to exogenous shocks. This particular approach could give us a better understanding of the ESG/non-ESG financial fragility and the fund resilience related to it. This suggests a third research question:

3) How do ESG funds react to exogenous shocks and what does it say about the link between ESG and risk-adjusted returns?

1.1 Structure of the thesis The thesis is categorised as follows: Chapter 2 discusses the background and history of ESG investing, followed by a brief case discussion about BP’s Deepwater Horizon 2010 oil spill in the Gulf of Mexico challenging the reader to think what could have been the alternative outcome if BP Petroleum would have considered ESG in their investing strategies. The Nordic ESG market is also briefly discussed in Chapter 2 explaining why this thesis focuses on the Scandinavian market. Chapter 3 discusses previous literature on ESG and risk-adjusted performance. Chapter 4 goes in depth about the theoretical framework and how these should be adjusted in lieu of ESG investing. Chapter 5 looks at the empirical methodology used to analyse the data set in this thesis, while chapter 6 describes the data and shows the findings after applying the methods described in the previous chapter. Chapter 7 is the discussion part where the results are being empirically and comparatively analysed based on previous publications and theoretical frameworks. An important part of chapter 7 is the discussion about COVID-19 and how this exogenous shock affected the ESG funds and their conventional counterparts, which is a key finding as it gives the reader a better understanding whether ESG investing is an endogenous decision. This is followed by a conclusion in Chapter 8. At the end of this paper there is a reference list and an appendix with more detailed data analysis.

2 2. Background The concept of sustainable development was for the first time officially defined in the Brundtland Report2 in 1987. The definition reads as follows: « sustainable development meets the needs of current generations without compromising the ability of future generations to meet their own needs». Companies, governments and other entities which actively lead a sustainability profile are becoming more and more recognised, hence rapidly changing the financial market.

2.1 History ESG investing is a relatively new topic, which quite recently has gained popularity among investors. As our world is becoming more globalised and inevitably smaller, investing into risky markets is simply not viable anymore. Policy makers, financial institutions, various investors and governments all share a common interest of decreasing risk and ensuring some stability. The financial crisis of 2008, a result of numerous irresponsible decisions, inefficiencies and lack of transparency in the financial sector, is a good reminder of how risky and uncertain our world actually is. Therefore, the incorporation of ESG factors has become an inevitable part of the investment processes as it has proven to have strong risk and return performance, especially throughout turbulent periods.

January 2004 marks the beginning of ESG investing as Kofi Annan, the former UN Secretary General, invited over 50 CEOs of large financial institution to be a part of a joint initiative, where they would try to find ways of how ESG can be incorporated into capital markets. An important milestone in this effort was marked when a report «Who Cares Wins» by Ivo Knoepfel was published followed by a conference hosted in Zurich on the 25th of August in 2005, where senior executives from the financial spectrum participated. This event revealed a great momentum showing how the investment value chain was rapidly changing. Another paper «Freshfield Report» was published at the same time which emphasised the importance of ESG. In 2006, Principles for Responsible Investment (PRI)3 was launched at the New York Stock Exchange, followed by the

2 Otherwise known as «Our Common Future», a report published in 1987 by the United Nations through the Oxford University Press. The name stems from the former Norwegian Prime Minister, Gro Harlem Brundtland, who was the Chair of the World Commission on environment and Development (WCED).

3 A set of principles that showcase possible actions for incorporating ESG into investment practices. By signing these principles, an investor commits to adopt and implement them. 3 launch of Sustainable Stock Exchange Initiative (SSEI)4 in 2007. Both of these initiatives have grown rapidly advancing the integration of ESG.

ESG investing was a niche investment strategy, which became more recognised after the implementation of the UN Sustainable Development Goals (SDGs) in 2015. This was proceeded by the signing of the UN Paris Agreement in 2016 and the release of the European Commission's Action Plan on Financing Sustainable Growth in 2018. These actions have tightened the ESG agenda by implementing mandatory minimum standards and teaching how active management plays an important role in contributing to a positive change. In the latest PricewaterhouseCoopers (PwC)5 report «2022 - The growth opportunity of the century», PwC assesses Europe’s dominance in the global ESG market. They expect the boundary between traditional and sustainable investments to disappear in 2022, where the Taxonomy6 and Sustainable Finance Disclosure Regulation7 will have been fully implemented in the EU.

2.2 Case «British Petroleum» (BP) is one of the most widely used case studies on risk. The BP´s Deepwater Horizon 2010 oil spill in the Gulf of Mexico is an example of how environmental risks can have major consequences. The explosion killed 11 people, injured 17 and triggered a devastating environmental crisis with deaths of as many as 105,400 sea birds, 7,600 adult and 160,000 juvenile sea turtles and a 51% decrease in dolphins in Louisiana’s Barataria Bay. Not only did the incident follow severe legal and reputational damage, but it also resulted in the company having to pay $20.8 billion - the largest environmental damage settlement in the Unites States history. The Financial Times estimated that the clean-up cost was around $60 billion and BP´s share price lost 50% between April and June 2010 and ever since the disaster, BP´s share price has been underperforming.

BP recently announced that they will be cutting in their oil production and moving towards green

4 A global platform showcasing how various instances can enhance the performance on ESG and encourage sustainable investment.

5 Global network of ﬁrms delivering Assurance, Tax and Advisory services. PwC ranks as the second-largest professional services network in the world.

6 Taxonomy Regulation introduces criteria for determining which economics actives can be considered environmentally sustainable.

7 A part of the European Commission’s Sustainable Finance Action plan which imposes an obligation on fund managers and other financial entities to disclose information on various ESG considerations to potential investors and on their websites. 4 energy. This could indicate that the company would be suitable for ESG as it serves a purpose, namely investing in clean energy research. Despite these alleged improvements, BP together with other oil giants such as Total, Shell and Repsol all fall short of aligning their CO2 output with the Paris Agreement. It does seem that BP has been trying to repair their reputation by becoming more environmentally friendly and aware. However, there has been without doubt a material effect on the company’s risk profile leading to weaker financial performance. One could therefore argue, that if BP was investing into sustainability and following ESG-practices from the very beginning, they might have avoided such reputational and financial damage.

The possibility of future accidents and the associated externalities has become a bigger factor in investment decisions after the Deepwater Horizon disaster. The global valuation of the oil sector has adapted a new cost and externalities structure, especially due to the increasing integration of ESG practices. The BP spill has proved to be a deﬁning moment in this shift of investment practices, advocating for sustainability and long-term value.

2.3 Why the Nordics? Some investors are looking for answers in the north. The Nordics can be described as having a strong ESG environment as it has become an integral part of all investment processes.

In Scandinavia, the concept of socially responsible investing (SRI) was adopted early on in the 1980s. This is perhaps reﬂected in the areas` politics, where equal rights has been an important issue early on. For example, already in 1862, unmarried Swedish women were the ﬁrst worldwide to be granted conditional right to vote in municipal elections. Furthermore, people have always had a close relationship to the nature, hence the societies strong need to preserve and nourish it. Whatever the reason may be, the fact remains that the Nordics has a high ESG score, which is why it is worth dwelling for some time and seeing what it actually entails.

Many investors look at the Nordic countries as a miniature copy of the overall global investment market. The countries in Scandinavia have a wide range of fields of expertise, which contribute to the broad diversification factor when investing in the Nordics. If we take a closer look at the Norwegian market, there are many investment opportunities in the energy sector and seafood. Sweden is known for its global brands such as IKEA. Finland and Denmark also have their areas of expertise, which helps to spread the investment risk without jeopardising the financial returns. Øyvind Fjell, who works with DNB Asset Management which is one of the largest asset managers

5 in the Nordic regions, suggests that the Nordic market should be highly considered when one wishes to invest. The data from Morgan Stanley Capital International (MSCI)8 Nordic index has proven to be more proﬁtable compared to the MSCI World. Over the past 20 years, the Nordic stock market has risen by 230%, while the MSCI World rose by 153% in the same period. Furthermore, MSCI Europe rose by 129%. Øyvind Fjell further discusses the overall ESG score of the Nordics and the global market. The ESG score in the Nordics is also higher compared to the global ESG score. Not only does the data of the past two decades show larger returns when investing in the Nordic region, it also is a more sustainable and ethical choice.

One of the leading examples of the institutional investors that practice sustainable investing is Norway's Government Pension Fund Global (Norwegian: Statens pensjonsfond Utland, SPU) commonly referred to as the Oil Fund (Norwegian: Oljefondet), being the world’s largest Sovereign Wealth Fund (SWF9) by assets as of august 2020. According to the guidelines of The Petroleum Funds Advisory Council on Ethics (established in 2004), SPU cannot invest money in companies that contribute to «serious or systematic human rights violations, such as murder, torture, deprivation of liberty, forced labour and the worst forms of child labour». Furthermore, it is unacceptable to invest in any activities resulting in «severe environmental damage» and «other particularly serious violations of fundamental ethical norms».

In summer 2020, the Ethics Committee led by Ola Mestad, professor at the University of Oslo and former chair of the SPU´s Council on Ethics, proposed changes to the ethical investment guidelines. This, according to Norges Bank, would result the fund to sell off NOK10bn of its holdings. Furthermore, Nicolai Tangen, CEO of Norges Bank Investment Management, told Financial Times that SPU should «use risk in a more clever way», especially when it comes to ESG. He said «that is an area were we can take a bit more risk». In 2019, the fund sold itself out of 42 companies for breaching ESG standards. Selling out for climate change reasons (mostly coal) resulted in a rise of returns to almost 0,3 percentage points. This is a good example of how a company might not necessarily decrease the risk but use it better. Yngve Slyngstad, the predecessor of Nicolai Tangen, said that handling responsible ownership and ESG was the biggest challenge he faced while being the chief executive. This shows that it is a difﬁcult task to incorporate ESG factors in an investment portfolio as it requires extensive work of not only ﬁltering out the companies that breach the ethical

8 MSCI ESG Indexes are designed to showcase the performance of the most common ESG investment approaches.

9 A state-owned investment fund that has a primary function to stabilise the country’s economy and to generate wealth for future generation. 6 guidelines, but also actively engage in promoting ESG values. SPU being the largest SWF in the world still remains the prime example of how ethical and sustainable investments do not result in lower ﬁnancial returns.

2.4 Scandinavian ESG Markets In the PwC’s published report «2022 - The growth opportunity of the century», they believe that the European ESG assets will reach between EUR 5.5 trillion and EUR 7.6 trillion, making up between 41% and 57% of total mutual assets in Europe by 2025. The major national markets in Europe include Luxembourg, France, Sweden and Ireland, accounting for a total of 48,1% of all ESG funds worldwide. Furthermore, according to RobecoSAM’s10 latest Country Sustainability Ranking, Norway has become the world’s most sustainable country, with Sweden, Finland and Denmark being in second, third and fourth place respectively. The strong sustainability proﬁle can be attributed to good leadership in governance, human capital, innovation and environmental indicators.

According to an article written by Øyvind Fjell, Norway, Sweden, Denmark and Finland have coped better with the corona crisis than the other European countries. The MSCI Nordic index has outperformed its international counterparts, thus continuing its traditional trend in long-term outperformance. Denmark was the leader of being the best performing country in Europe, as it has been for the last ten years. Even Sweden’s alternative strategy of dealing with COVID-19 was not reﬂected in the economic development. This resilience should be attributed to strong sector diversiﬁcation and strong focus on ESG. As Fjell concludes «Investors have understood that ESG factors can be seen as an additional risk parameter and improve the risk/reward ratio of an investment».

2.4 Approaches As the concept of ESG is quite new, there is still some confusion regarding the methodologies on how one should approach it. When evaluating the ESG parameters, there are mainly two different aspects one has to keep in mind. Firstly, how to identify true ESG companies and secondly, how to incorporate ESG factors when evaluating the risk-reward ratio of an investment. To identify an ESG fund can be a somewhat difﬁcult task. The most common method is known as

10 International investment company with specific focus on sustainability investments. 7 positive screening, where fund managers invest in firms with a high ESG score. This can also be referred to as «best in class» strategy. However, it poses some difficulties. For starters, there is no consistent terminology in defining the key concepts. Many terms are used interchangeably such as sustainable, ESG, responsible - and impact investing. Due to the lack consensus on what these definitions entail, it leaves room for misunderstandings and different interpretations. This issue raises a debate whether ESG inclusion is merely a marketing technique for some companies rather than a step forward to strengthening its sustainability profile. Therefore it is important to use ESG assessment tools (such as Morningstar Sustainability Screener) with vast databases helping to find funds that intentionally integrate ESG into their investment portfolio. This will be further discussed in section 6.3.

Morgan Stanley Investment Management (MSIM), a global investment manager, has outlined four main approaches of how to achieve sustainable and long-term returns:

1) Restriction Screening: intentionally avoid to invest in sectors which do not align with ESG values. By implementing such policy, MSIM reduces the exposure to risk related to ESG factors. As already mentioned previously, these risks can be both financial and reputational. 2) ESG Integration: an investor does not merely consider the financial returns, but also identifies both the risk and opportunities related to ESG factors. Like in the example of BP, an investor who was ESG aware, would not have invested in the company as the environmental risks already before the accident seemed too great no matter of the potential high financial returns. 3) Thematic investment: the focus is set on long-term global trends which are transforming the world and thus creating investment opportunities. The idea is that an investor tries to identify long-term trends early on so that its portfolio has growth opportunity. In other words, an investor can ask himself «what are the risks I might face in the future?» and thus try to mitigate them. The Government Pension Fund of Norway, which heavily relies on oil revenues, has invested in themes such as renewables and green energy as oil decline is a realistic threat in the future. Even though this investment strategy does not eliminate financial risk metrics like volatility, it is an effective method to mitigate potential long term external risks. 4) Impact Investing: a rapidly growing industry where investors are determined to generate positive measurable social and environmental impact alongside with financial returns. Many investor still believe that impact and financial returns do not go hand in hand and despite strong academic evidence (discussed more in depth in «Literature» section) still believes in this misconception.

8 These are merely policies which could contribute in effectively ﬁltering out companies that claim to have an ESG proﬁle, but perhaps are not entirely clear on how they approach this matter. Investors who wish to be ESG aware, can also partake in active ownership where they can encourage companies to improve their practices and policies. Investor can also govern the managerial performance and vote against their decisions if they are not in the companies best interest.

2.5 Short-termism A recurring theme in many publications is how well does ESG fit in a short-term perspective. The extensive data available on ESG financial performance indicates that short-termism does not fit in an ESG framework as it tends to affect returns over longer periods. It is important to have in mind that when considering financial performance of ESG, we consider the long-term perspective.

3. Literature review 3.1 Social Responsibility and Financial Performance As the topic of sustainability has become more popular, the body of literature on SRI/ESG investments has grown rapidly. According to a paper published by Friede, Busch & Bassen (2015), their estimates indicate that between 1970 and 2014 there «have been published more than 2000 empirical studies and several review studies» on the search for a relation between ESG criteria and corporate ﬁnancial performance (CFP).

The first monumental paper from Moskowitz (1972) launched a series of other publications in the field of socially responsible investing. He was the first to suggest that there may be a link between social responsibility and financial performance, but as the topic was so new, the data was inadequate for statistical purposes. Hamilton et al. (1993) building on Moskowitz`s conclusion, performed a study where they compared socially responsible mutual funds to conventional funds in the period between 1981 and 1990. Here they used Jensen´s alpha11 to compare excess returns and they found no statistical significance between investing in the two types of funds. This could be attributed to the fact that sustainability was not a «hot» topic in the 80s and thus there was no price on social responsibility. Furthermore, Bauer et al. (2007), by using a wide range of statistical models, compared the risk-adjusted performance of Canadian ethical mutual funds relative to their conventional peers and found no significant performance differences between them. In their concluding remarks, they acknowledge that their findings largely corroborate previous research on

11 Used to calculate an investment's risk-adjusted value. 9 ethical mutual fund performance and thus emphasises that one study does not support the claim that imposing ethical constraints leads to weaker investment performance.

Eccles, Ioannou and Serafeim (2014) published a paper where they investigated the effect of corporate sustainability on organisational process and performance using a matched sample of 180 US companies. They divide them into two groups, one with low and the other with with high sustainability performance based on the ESG Disclosure Scores from Bloomberg and Reuters12. Eccles et. al. (2014) finds that there is a statistically significant difference where the high sustainability firms have a score of 41.1% whereas the low sustainability firms have only 8.31%. They conclude that the firms with high ESG score significantly outperform their counterparts over the long-term perspective. In other words, the firms who have adopted environmental and social policies for a significant amount of years exhibit lower volatility thus indicating lower risk. Overall, the conducted research so far does provide statistically significant evidence for the empirical identification of ESG as a factor of risk premium when integrated in an investment portfolio (Pollard et. al. 2018).

It is important to have in mind that the SRI/ESG investors are a heterogenous group using various funds with different types of screens. Thus, comparing the results and putting equal weight on all of the ﬁndings seems somewhat unreasonable. Sandberg et. al. (2009) discussed the issue of standardisation of the SRI terminology and concluded that even though much diversity prevails, the heterogeneity may not be as problematic as it is often made out by academic writers.

The key in determining whether there is a correlation between sustainability and financial performance of a fund depends on the investment strategy. It is not enough to merely screen out the non-sustainable companies, but one should look for a company who actively works towards the incorporation of ESG. When a portfolio has been filtered out with «true» ESG funds, only then can we say something about the financial returns, thus eliminating the terminology bias mentioned previously.

3.2 ESG factors and risk-adjusted performance Even though there is a substantial amount of research performed on the relationship between sustainability and ﬁnancial performance, most studies seem to overlook the issue of risk (Kumar et.

12 A measure of how complete the company’s reporting is on a range of ESG topics based on a 0%-100% scale. 10 al. 2016). According to them, recent studies take the notion of ‘high ESG-less risk’ for granted and immediately go on demonstrating the improved ﬁnancial, portfolio and stock performance where ESG factors are analytically applied without considering the entailed risk. In other words, by focusing on the idea that the incorporation of ESG principles automatically lead to higher risk- adjusted returns, there is little research done on ESG impact on risk. By analysing the link between ESG and risk, investors can better identify strategies that are best suited for their desired social, environmental and ﬁnancial criteria.

Ick Jin (2018) published a paper called «Is ESG a systematic risk13 factor for portfolio managers?» where he discusses the importance of ESG factors in an investment portfolio and how to take it into consideration into a company’s investment analysis. His research is heavily based on Hayat and Orsdagh (2015) publication «Environmental, Social, and Governance Issues in Investing». The objective of both papers is to identify and properly evaluate investment risks while taking ESG issues into consideration as well. As mentioned previously, the classiﬁcation of ESG issues is a difﬁcult task, both in terminology and monetary value. Hayat and Orsdagh (2015) presented a table where they show what they mean by ESG issues (Table 1).

Table 1: Examples of ESG Issues; Source: Hayat and Orsagh (2015)

It is clear that ESG issues are not a simple matter. From a survey conducted in 2015 by the the CFA Institute14, 27% of the respondent do not consider ESG issues at all, whereas 73% consider at least one of the ESG factors (Graph 1).

13 Otherwise known as «volatility» or «market risk» which aﬀects the overall market. This type of risk cannot be eliminated by diversiﬁcation as it is unpredictable.

14 Global, not-for-proﬁt professional organisation providing investment professionals with ﬁnance education, while aiming to promote standards in ethics. 11 Graph 1: Consideration of ESG issues from respondents. Source: CFA Institute

According to Hayat and Orsagh (2015), the main reason why companies and investors consider ESG issues is «to help manage investment risks» (Table 2).

Table 2: Response to why investors consider ESG issues. Source: Hayat and Orsagh (2015)

The fact that 63% of the respondents consider ESG issues primarily because of «risk management» seems to be consistent with the existing literature on ESG. Referring back to the case of BP and the oil spill in the Gulf of Mexico in 2010, the ﬁnancial consequences were large, thus reinforcing the company to analyse ESG factors and how they would contribute to managing their risk proﬁle. This is only an example of how negative environmental externalities can affect a companies performance and reputation. It seems that ESG issues become particularly popular when something of extreme nature takes place, such as an environmental disaster . The key principle for an investment 12 professional would be to systematically take ESG issues into consideration for better-informed investment decisions.

4. Theory Some theories have been drafted in order to better understand the topic of investment responsibility, performance and associated risk. Shareholder and Stakeholder Theories, as well as the Modern Portfolio Theory, provides needed fundament to develop new theories that incorporate ESG into the equation, such as the recently developed ESG-efﬁcient frontier theory. All these theories create a base for the empirical methodologies which will be used to answer the research questions introduced at the beginning of this thesis.

4.1 Shareholder Theory Martin Friedman in 1970 stated that the firm’s primary responsibility lays with its shareholders, meaning that the only responsibility of managers is to act on the interest of shareholders. He describes social responsibility as a «fundamental subversive doctrine» which means that the only social responsibility a firm has is to increase its profits. If the shareholders would consider ESG as an unprofitable investment strategy, this according to the Shareholder theory, would violate the firm’s core responsibility. The opposing theory is the Stakeholder theory.

4.2 Stakeholder Theory The Stakeholder theory suggest that a firm or a corporation should be involved in activities beyond profit maximisation. R. Edward Freeman in his book «Strategic Management: A Stakeholder Approach» published in 1984 states that «…for any business to be successful it has to create value for customers, suppliers, employees, communities and financiers, shareholders, banks and others people with the money». All of these groups have some «stake» in the company thus all of their interests have to be considered in an equal manner. It seems that the Stakeholder theory has become more relevant in today’s world as the integration of ESG factors is becoming more popular.

4.3 Modern Portfolio Theory The modern portfolio theory (MPT) (otherwise known as the mean-variance theory) was developed by economist Harry Markowitz introduce in his publication under the title «Portfolio Selection» in the Journal of Finance in 1952. His research focused on investment portfolios and how return is related to risk. He especially emphasises the importance of diversiﬁcation which can prevent

13 unsystematic risk. The theory hypothesises that it is possible to design an optimal portfolio which maximises returns by taking a quantifiable amount of risk, in other words, if an investors uses a quantitative method, he can reduce his investment risk. According to MPT, in order to achieve a portfolio with low variance, investors should avoid investing in securities with high covariance. The goal is to eliminate any idiosyncratic risk15 associated to a particular asset and be solely exposed to the systematic risk. Evans & Archer (1968) discuss how large a diversified portfolio needs to be to significantly reduce risk, where they conclude that no more than 15 stocks are needed to achieve diversification. This number has been criticised where other researchers claim it to be around 40 to 50 stocks (Benjelloun, 2010; Campbell et al, 2001).

The set of optimal portfolios is introduced in Morkowitz’s (1952) publication as the efficient frontier. These portfolios would offer the highest expected return for a defined level of risk. The portfolios that lie below the efficient frontier, do not provide enough return for the level of risk. The efficient frontier is represented by plotting the expected returns of a portfolio and the standard deviation of returns. The y-axis represents the expected returns whereas the x-axis measures the standard deviation of returns, which is the measure of risk (Graph 2). When the portfolio is plotted into the graph, it is being compared to the efficient frontier.

Graph 2: Efﬁcient frontier. Source: Investing Answers

15 Idiosyncratic risk refers to uncertainties and potentials problems that are unique to individual assets. 14 Graph 2 shows the efficient frontier without the risk-free asset. When contrasting portfolios, investors tend to combine risky assets with risk-free assets to minimise risk. When referring to a complete portfolio, it is usually defined as a combination of both type of assets. The optimal risky asset will thus be at the point where the risk-free asset or capital allocation line (CAL) is tangent to the efficient frontier (Graph 3).

Graph 3: Efﬁcient frontier with the capital allocation line depicting the optimal portfolio. Source: CFI

The slope of CAL is called the Sharpe ratio, otherwise called as the reward-to-risk ratio. Sharpe ratio is the measure of risk-adjusted return of a ﬁnancial portfolio. It measures the excess portfolio return over the risk-free rate relative to its standard deviation. If Sharpe ratio is greater than 1.0, the fund is considered acceptable or good by investors. If the value is 2.0 and above, it is very good. Sharpe ratio is not enough in itself to tell us something about the performance of a portfolio and thus only by comparing different funds is it possible to understand the risk-adjusted returns of the various funds.

4.4 The efficient frontier and ESG As already discussed in the literature overview section, the awareness of ESG investments has grown rapidly over the past years, however there exists little guidance in how to practically incorporate ESG factors when assessing a portfolio. A major issue with the efficient frontier is that the theory is solely based on expected returns and standard deviation. ESG is not an investment strategy that only takes financial returns into account. Therefore, when evaluating a portfolio using the efficient frontier, one has to compliment it by using other screening methods of ESG evaluation.

15 Furthermore, the model was built on the idea of diversification. By only choosing ESG fund, the investment universe becomes restricted as particular investment opportunities are being filtered out. This will thus affect the diversification possibilities and thereby lead to the possibility of only sub- optimal diversification (Markowitz, 1952). If going back to the Shareholder theory, the incorporation of ESG would not lead to better performance than when having no constraints at all. Having this criticism in mind, Pedersen et al. (2019) introduced a new, augmented version of the efficient frontier model where they actively incorporate ESG.

The theory illuminated both the potential costs and beneﬁts of ESG-based investing. Pedersen et al. (2019) considers three types of investors. 1) Type-U (ESG unaware): this type of investor is unaware of the ESG scores and only wishes to maximise his unconditional mean-variance utility. 2) Type-A (ESG aware): this type of investor also has mean-variance preferences, but he is aware of the ESG scores and uses them to update his beliefs on risk and return. 3) Type-M (ESG motivated): this type of investor actively seeks out for ESG portfolios and has a preference for high ESG scores. The investors seeks an optimal trade-off between low risk, high ESG score and high expected returns.

The first two types of investors fit in the framework of the Markowitz efficient frontier solution, however the incorporation of ESG in the optimisation of a portfolio might seem more challenging. Pedersen et al. (2019) show that the investors problem can be reduced to a trade-off between ESG and Sharpe ratio. In their paper, they compute the highest attainable Sharpe ratio (y-axis) for each level of ESG (x-axis) (Graph 4).

Type-U investors may choose a portfolio below the frontier as they do not take into consideration any ESG factors. As mentioned previously, the higher the Sharpe ratio, the better the performance, thus Type-A investors would always choose the portfolio with the highest Sharpe ratio which in the graph is shown as the tangency portfolio using ESG information. Type-M investor would choose a portfolio shown on the ESG-efficient frontier as they have a preference for higher ESG. Pedersen et al. (2019) analyses what impact an increase in the ESG level has on the portfolio construction, and they find that a small increase leads to only a small reduction in Sharpe ratio, which means that ethical goals can be achieved at a small cost. This is in line with other publications where ESG slightly underperforms compared to non-ESG. However, when they impose «realistic constraints», this leads to a downward shift in the ESG-SR frontier. This is a natural cause of effect as imposing 16 Graph 4: ESG-efficient frontier: Source: Pedersen et al. (2019)

more constraints reduces the maximum Sharpe ratio.

5. Empirical methodology It is not enough to simply look at returns alone. In order to properly evaluate a portfolio, an investor has to quantify and measure risk with the variability of returns. In this chapter I will introduce the various performance measurements used for analysing the data. These are know as the Treynor, Sharpe and Jensen’s ratios as well as the Capital Asset Pricing Model (CAPM)

5.1 Treynor ratio The Treynor measurement was introduced by Jack L. Treynor who was the ﬁrst one in his ﬁeld to provide investors with a portfolio evaluation tool which also included risk. The formula for the Treynor ratio is as follows:

TreynorRatio = (rp − rf )/βp

where rp is the portfolio return, rf is the risk-free rate and βp is the beta of the portfolio which captures the systematic risk. The Treynor ratio only focuses on systematic risk, which means that it

17 can only be used with well-diversiﬁed portfolios, where all of the non-systematic risk has been diversiﬁed away. In other words, the beta measures the tendency of a portfolio’s return to change in response to changes in return for the overall market. One of the main limitations for this measure is that the accuracy of the Treynor ratio is highly dependant on the use of an appropriate benchmark to measure beta. The use of beta as the sole measure of portfolio risk is both the point and the criticism of the Treynor ratio.

5.2 Sharpe ratio Another common risk measurements used in ﬁnance is called the Sharpe ratio, introduced by William Sharpe in 1966. It is a measure of risk-adjusted return of a ﬁnancial portfolio. The formula for the Sharpe ratio is as follows:

SharpeRatio = (rp − rf )/σp

where rp is the portfolio return, rf is the risk-free rate and σp is the portfolio’s total risk measured by standard deviation. The higher the Sharpe ratio of a portfolio, the more optimal is the funds’ risk- return relationship. In other words, the higher the ratio, the better its returns are relative to the amount of investment risk it has taken.

5.3 Jensen’s Alpha Jensen’s alpha introduced by Michael C. Jensen in 1968 is a single factor model based on the CAPM. He introduces an additional alpha to the equation which measures the excess returns earned by the mutual fund compared to returns suggested by the CAPM model, in other words it tries to explain whether an investment has performed better or worse than its beta value would suggest. Jensen’s measure can be shown by the following formula:

α = E(Ri) − Rf + β[E(Rm) − Rf ]

A positive Jensen's alpha suggests that the funds have delivered superior risk-adjusted returns, and if an investor should choose between two funds with similar betas, he would choose the one with greater alpha as this would indicate higher returns at the same level of risk. Jensen’s alpha is an important measure when determining whether a manager has added value to a portfolio, as it is an attribute to the skill of the portfolio manager. In other words, the value of alpha indicates the value added to or subtracted from the portfolio as a result of portfolio management. The measure is 18 somewhat controversial as there are those who believe in the efﬁcient market hypothesis, which states that it is impossible to beat the market. If there exists a portfolio that has in fact outperformed the market, then the alpha is solely attributed to luck rather than skill.

5.4 The Capital Asset Pricing Model (CAPM) As discussed under the modern portfolio theory, no matter how much one diversifies, the problem of systematic risk remains unsolved. The Capital Asset Pricing Model (CAPM) was developed by Treynor (1962), Sharpe (1964), Lintner (1965) and Mossin (1966), and builds on Markowitz (1952) portfolio choice. CAPM assumes that there is a linear relationship between asset returns and market risks. If all investor would have homogenous expectations, then all investors would demand the same efficient portfolio of risky assets, what was previously shown as the tangency portfolio. This thus indicates that the tangency portfolio is the market portfolio and therefore, by using CAPM, it is possible to find the expected return of an investment by using the market portfolio as a benchmark. This can be expressed by the following formula:

E(Ri) = Rf + β[E(Rm) − Rf ]

E(Ri) is the expected return on asset i, Rf is the risk-free rate and E(Rm) is the expected market return. The difference between the expected market return and the risk-free rate (E(Rm) − Rf) indicates the market risk premium16. The β measures the systematic risk (volatility) of a fund compared to the entire market. A company with higher β has greater risk and also greater expected returns. A β that is equal to 1 is exactly as volatile as the market. If it is equal to 0, then it is completely uncorrelated to the market. Respectively, if β > 1 then it is more volatile to the market and if β < 0 then it is negatively correlated to the market. The β is calculated by the covariance between the asset i and the market (benchmark) divided by the variance of the market return. This can be calculated by the following formula:

2 β = COV(Ri, Rm)/σ (Rm)

Some researchers have criticised the CAPM theory, as it seems to not hold in practice. Some suggest that the anomalies that happen in the market cannot be explained by a simple linear relationship.

16 Excess return measures how much a fund has under- or outperformed the benchmark against which it is compared to. 19 6. Data This chapter provides a description of the process of how the dataset for this thesis was collected. An overview over the Morningstar database is presented along with an overview of the total number of ESG and conventional funds which constitute the basis of the analysis.

6.1 Research design - sample selection The data used in this thesis is collected from Morningstar and Financial Times databases. Firstly, by using the Morningstar database, I found 73 funds that go under the category of «Nordic Funds», which mens that these are funds that are solely based in Scandinavia. Out of these 73 funds, I found 12 funds that had above average or high sustainability rating17. Out of the remaining 61 funds, I found those that had below average or low sustainability rating. Out of these, I randomly selected 12 funds that would represent my non-ESG sample. Many funds were automatically eliminated due to data shortage. It was important for diversiﬁcation purposes to choose different funds with different management. As an example, the DNB funds have the same management, which is why they perform very similarly.

Most of the ESG funds are relatively new, thus my time period used which would incorporate the majority of the funds is from 14th of February, 2019, until 1st of March, 2021. One of the ESG funds only had data from late 2020, thus my sample had to be reduced to 11 ESG funds (which resulted in eliminating one fund from the non-ESG sample as well). The chosen funds are:

Non-ESG funds:

Delphi Nordic A DNB Grønt Norden A DNB Norden CDNB Norden C

DNB Norden R Evli Nordic B Lannebo Norden Hållbar SEK ODIN Norden A ODIN Norden B Pareto Nordic Equity B NOK Pareto Nordic Equity C NOK

17 The Morningstar Sustainability rating is based on the underlying fund holdings and company- level ESG Research and ratings from Sustainalytics, which is a leading independent provider of ESG ratings and research. 20 Skandia Idéer För Livet

ESG funds:

DNB Barnefond A Eika Norden Handelsbanken Norden Selektiv (A1 NOK) Handelsbanken Norden Selektiv (A10 NOK) Landkreditt Norden Utbytte A Landkreditt Norden Utbytte I Nordea 1 - Nordic Stars Equity Fund BC NOK Nordea 1 - Nordic Stars Equity Fund BI NOK Nordea 1 - Nordic Stars Equity Fund BP EUR Nordea 1 - Nordic Stars Equity Fund BP NOK

Swedbank Robur Nordenfond

After deciding on the funds and the time period, I used Financial Times database to get daily returns of each fund between 14th of February, 2019, and 1st of March, 2021. An important momentum after compiling the data was to go through the dates and check for inconsistencies. It was important to match all the dates so that it would be possible to compare the daily returns, so if a fund was missing data for one day, it was important to eliminate this date as to avoid any comparison errors in data analysis.

6.2 MSCI Nordic as benchmark The market proxy used in the main results is MSCI Nordic Index. It was chosen because it is one of the most commonly used benchmarks in the Scandinavian market. The index captures small-cap18 representation across Norway, Sweden, Denmark and Finland. It is based on the MSCI Global Investable Indexes (GIMI) Methodology, which is a consistent and comprehensive approach to index construction. The aim of such methodology is to allow for cross regional comparisons across all markets and provide an exhaustive coverage of the relevant investment opportunities.

18 Small-cap stocks are shares of companies with a market capitalisation between $300 million and $2 billion. 21 6.3 ESG dataset - Morningstar Sustainability Rating The ESG dataset19 in this thesis consists of 11 Scandinavian funds with a sustainability rating between 4 and 5 globes collected from Morningstar’s database. The globe system has a range from one to five, with five being the most sustainable. The worst 10% and the best 10% of funds received one and five globes respectively. According to Morningstar, if a fund declares a non-economic investment objective in its prospectus, it can be categorised as socially conscious. The rating was introduced in 2016 in order to help investors to incorporate ESG factors in their portfolio evaluation. The rating is a historical holding-based calculation which uses the ESG Risk Rating from Sustainalytics. Morningstar Sustainability Rating is a three-step process, which is updated once a month with the newest data. Firstly, the Morningstar Portfolio Sustainability Score is calculated for every portfolio reported within the trailing 12 months (TTM)20. Then these scores are used to calculate a portfolio’s Morningstar Historical Portfolio Sustainability Score. Lastly, the Morningstar Sustainability Rating is assigned for a portfolio based on its Morningstar Historical Portfolio Sustainability Score relative to its Morningstar Global Category21.

6.4 Conventional portfolio dataset As mentioned previously, the ESG portfolio for this thesis is created by looking at the highest rated Scandinavian funds, based on the Morningstar Sustainability Rating. A similar approach is used to create a dataset to represent conventional investing. Through Morningstar’s database, I found all the Scandinavian funds which has a sustainability rating of 1-2 globes. In order to avoid any errors, I randomly picked 11 funds in order to have the same amount as in my ESG dataset. A low sustainability score indicates that these companies do not take ethical issues seriously and may also be a sign of greenwashing.

6.5 Risk-free rates The risk-free rate represents what kind of interest an investor would expect from a risk-free investment. In most studies, the risk-free rate is represented by US 1-month T-bill, since many studies focus on the US. However, since this thesis solely focuses on Scandinavian funds, I decided to use the Norwegian Interbank Offered Rate (NIBOR) as a risk-free proxy. The data for NIBOR

19 The data in this thesis was last updated in February, 2021.

20 A term used to describe the past 12 consecutive months of a company’s performance data.

21 A category introduced in 2010 to help investors ﬁnd similar investments across the globe. The Morningstar Global Category is grouped into nine global categories: Equity, Allocation, Convertibles, Alternative, Commodities, Fixed Income, Money Market, Property, Miscellaneous. 22 was directly collected from the Norwegian Bank’s webpage. The rate is published as an annualised rate which has been converted to a daily rate22 to match the data collection from Financial Times for ESG and conventional portfolios.

6.6 Concerns about the dataset

There are several concerns about the dataset that should be brieﬂy discussed. First of all, the amount of funds in each portfolio is quite low, simply due to the fact that the amount of companies based solely in the Nordics is quite low. When comparing the results of this thesis to other publications that use for example US portfolios, some caution should be taken into consideration as the Scandinavian market can be much more sensitive to changes in a single stock as the portfolio consists of less funds. Another concern is the timespan of the dataset. Since ESG is a relatively new concept, the amount of available data is quite limited. For this reason, the dataset begins on the 14th of February 2019. Even though this thesis can be considered as a good starting point for further research in this area, the results can be misleading for the lack of historic data. Furthermore, the portfolio does not equally represent the Scandinavian market as the majority of the funds are either Norwegian or Swedish with a few Danish and no Finnish. Therefore, even though the results should represent the whole Scandinavian market, due to the lack of data, it might be a misleading representation of the Danish and Finnish ESG-market. An explanation for this is that these markets have broader portfolios and do not have funds solely based in Scandinavia, which does not exclude the possibility of them being ESG-oriented.

7. Results and discussion

After collecting and organising the data, the analysis is ready to be conducted. This chapter discusses the results received from the data analysis. The results covered are the Sharpe and Treynor ratios, followed by CAPM and Jensen’s alpha.

7.1 Non-ESG portfolio data analysis Table 3 presents the results of the non-ESG funds used in this thesis. From the mean of return, we see that all the funds have outperformed the market/benchmark. Furthermore, the Sharpe ratios serve as a measurement for comparing risk-adjusted returns. The higher the fund’s Sharpe ratio, the

22 After evaluating all the funds’ daily returns and organising so that all the days correlate, there is in total 246 trading sessions per year, which is the number I use to convert NIBOR’s annualised rate to daily rate. 23 better is the return relative to its risk. From table 3, we see that DNB Grønt Norden A has the greatest risk-adjusted returns.

Sharpe Jensen’s Mean StDev BETA CAPM Ratio/ alpha Delphi Nordic A 0,07 % 1,19 % 0,91 0,07 0,009 % 0,07 %

DNB Grønt Norden A 0,11 % 1,15 % 1,48 0,05 0,009 % 0,11 %

DNB Norden C 0,10 % 1,13 % 1,35 0,05 0,009 % 0,10 %

DNB Norden R 0,10 % 1,11 % 1,31 0,05 0,009 % 0,09 %

Evli Nordic B 0,07 % 1,48 % 0,73 0,13 0,012 % 0,06 %

Lannebo Norden Hållbar SEK 0,08 % 1,24 % 0,99 0,09 0,011 % 0,07 %

ODIN Norden A 0,10 % 1,07 % 1,40 0,05 0,009 % 0,09 %

ODIN Norden B 0,10 % 1,07 % 1,38 0,05 0,009 % 0,09 %

Pareto Nordic Equity B NOK 0,06 % 1,22 % 0,71 0,12 0,012 % 0,05 % Pareto Nordic Equity C NOK 0,06 % 1,22 % 0,73 0,12 0,012 % 0,05 % Skandia Idéer För Livet 0,08 % 1,20 % 0,99 0,06 0,009 % 0,07 % Xtrackers MSCI Nordic UCITS ETF 0,04 % 1,34 % 1D (in GBX) Table 3: Results calculated for all conventional non-ESG funds and the MSCI Nordic benchmark. The mean of expected returns are reported in daily percentage based on 24 months of data.

24 7.2 ESG portfolio data analysis Table 4 presents the results of the ESG funds used in this thesis. From the mean of return, we see that all the funds have outperformed the market/benchmark. Furthermore, based on the Sharpe ratio, we see that Nordea 1 - Nordic Stars Equity Fund BI NOK has the greatest risk-adjusted returns.

Sharpe Jensen’s Mean StDev BETA CAPM Ratio/ alpha

Nordea 1 - Nordic Stars Equity Fund BC NOK 0,08 % 1,04 % 1,16 0,47 0,022 % 0,06 %

DNB Barnefond A 0,079 % 1,05 % 1,07 0,56 0,025 % 0,05 %

Eika Norden 0,064 % 1,02 % 0,88 0,54 0,025 % 0,04 %

Handelsbanken Norden Selektiv (A1 NOK) 0,082 % 1,12 % 1,06 0,54 0,025 % 0,06 %

Handelsbanken Norden Selektiv (A10 NOK) 0,088 % 1,12 % 1,13 0,54 0,025 % 0,06 %

Landkreditt Norden Utbytte A 0,057 % 0,96 % 0,82 0,52 0,024 % 0,03 % Landkreditt Norden Utbytte I 0,059 % 0,96 % 0,85 0,52 0,024 % 0,04 % Swedbank Robur Nordenfond 0,056 % 1,25 % 0,61 0,66 0,029 % 0,03 % Nordea 1 - Nordic Stars Equity Fund BI NOK 0,084 % 1,04 % 1,17 0,47 0,022 % 0,06 %

Nordea 1 - Nordic Stars Equity Fund BP EUR 0,069 % 1,33 % 0,73 0,66 0,029 % 0,04 %

Nordea 1 - Nordic Stars Equity Fund BP NOK 0,081 % 1,04 % 1,12 0,47 0,022 % 0,06 %

Xtrackers MSCI Nordic UCITS ETF 1D (in GBX) 0,04 % 1,34 % Table 4: Results calculated for all ESG funds and the MSCI Nordic benchmark. The mean of expected returns are reported in daily percentage based on 24 months of data.

25 7.3 Comparing the results The empirical results from ESG and non-ESG portfolios are summarised in Table 5 by giving each fund equal weight.

Sharpe Treynor Jensen’s Mean StDev CAPM ratio ratio alpha

ESG portfolio 0,073 % 1,086 % 0,95 0,02 0,025 % 0,048 %

Non-ESG portfolio 0,088 % 1,188 % 1,07 0,17 0,01 % 0,078 %

Xtrackers MSCI Nordic UCITS ETF 0,04 % 1,34 % 1D (in GBX)

Mean difference 0,015 % 0,102 % 0,126

Statistical signiﬁcance t-stat −1,982 −2,127 −1,029 (two-tail test) Statistical 0,076* 0,059* 0,328 signiﬁcance p-value

Table 5: Summary of results; * 0,1 signiﬁcance level

Table 5 shows that the average return of the ESG portfolio is 0,073% and 0,088% for the non-ESG portfolio, which means that both portfolios outperform the market, but the conventional funds have better returns. The volatility of the ESG funds is lower compared to the conventional funds. The Sharpe ratios are higher for the non-ESG portfolio (1,07), which might indicate greater efficiency of the investment. The non-ESG portfolio also exhibits higher Treynor ratio (0,17) over the ESG portfolio (0,02). Furthermore, when comparing the results of CAPM, both portfolios satisfy the relationship between systematic risk and the expected return on the assets. The Jensen’s alpha also suggests that both portfolios are earning excess returns, where the conventional funds are doing slightly better in comparison with their ESG counterpart. Table 5 also shows the differences in means for the return measures, with t- and p-values for each difference. These values show whether the differences are statistically significant. The data indicates statistically significant difference in returns and standard deviation between the studied groups at a significance level of 0,1. The p- values of the t-test are 0,076 for mean of returns and 0,059 for standard deviation. The non-ESG portfolio has significantly higher mean of returns, but also significantly higher standard deviation

26 (volatility). This is why Sharpe Ratio is not signiﬁcantly different (0,328) because it is a volatility- adjusted return.

7.4 Discussion The notion of investing using ESG-practices is not about outperforming conventional investing, but rather, achieving the same returns with lower risk. The idea of having not only ﬁnancial gain but also sustainability in mind is challenging all traditional investment measures, as they do not incorporate ESG-factors in their equations. Using the results from both of the portfolios in this thesis, we see that the conventional funds outperform the ESG-portfolio with a difference of 0,015%. However, when looking at the volatility, the conventional funds seem to be more volatile, with a difference of 0,102%. This suggests that ESG-investment practices are less risky. This is in line with Eccles et al. (2014) ﬁndings who conclude that ESG exhibits lower volatility.

Previous publications that have been analysing the ESG market in the US, Europe and other regions have found that ESG portfolios tend to slightly underperform compared to non-ESG portfolios, albeit without statistical significance (Hamilton et al, 1993; Bauer et al., 2007). The results in this thesis seem to contradict those findings, showing that there is a statistically significant difference in returns and standard deviation in the Scandinavian fund market. After conducting a t-test analysis, the differences are statistically significant at a significance level of 0,1. The non-ESG portfolio has significantly higher returns as well as significantly higher volatility. One explanation to this could be that ESG is a long term investment and the timespan analysed in this paper is too short. Furthermore, the ESG screening process utilised in this paper has been very strict, choosing only the funds with highest ranking. This is in line with findings published by Pedersen er al. (2019) where they show that a higher ESG preference slightly reduces the funds’ performance. It should also be noted that funds incur management fees and transaction costs associated with purchasing and selling the stock. When analysing the costs associated with ESG and non-ESG funds, it seems that ESG funds incur a lower cost23, which could outweigh the difference in returns making ESG a better investment overall.

23 Average cost of ESG portfolio in this thesis is 1,25% while the average cost of non-ESG portfolio is 1,35%. 27 A crucial point when analysing and comparing the measures of risk-adjusted returns (such as Sharpe and Treynor ratios) is the difference between systematic and idiosyncratic risk. The Treynor Ratio compares the returns earned against the beta (which captures the systematic risk of a portfolio). This means that the lower the betas are of a portfolio (which is the case for the non-ESG portfolio in this thesis), the higher the Treynor Ratio will be. However, this measure can only be used when the portfolio is well-diversiﬁed and in theory, ESG does have a narrower investment universe making it more difﬁcult to diversify. Therefore, the Treynor Ratio might not be a solid representation of ESG funds’ risk return performance (Evans & Archer, 1968; Benjelloun, 2010; Campbell et al., 2001).

Sharpe Ratio on the other hand observes both systematic and idiosyncratic risks. After comparing the Sharpe ratios, it seems that the conventional funds have higher risk-adjusted returns. Going back to section 4.4, this is in line with the efficient frontier and ESG model introduced by Pedersen et al. (2019). When a manager actively incorporates ESG, this imposes constraints leading to a downward shift in the ESG-efficient frontier. This reduces the Sharpe value ratio. It should be noted that the comparison made is of the overall portfolio and not the specific industries. If a matched-pair approach was conducted based on the funds’ industries, perhaps the results would have been different.

After conducting linear regression as part of the data analysis, the difference in Jensen’s alpha for ESG and non-ESG portfolios had no statistical significance, with p-values being 0,124 and 0,113 respectively. This is in line with results published by Hamilton et al. (1993). Therefore, the management does not seem to add any value to the portfolios’ returns and has no affect on risk- adjusted performance. In order to check the significance of Jensen’s alpha, more accurate data premises should be introduced: have a larger pool of funds in the portfolios, as well as categorise each fund by industry and specific characteristics.

28 7.4.1 Response to a market crisis: COVID-19 Nofsinger and Varma (2013) found that socially responsible funds, during non-crisis periods, underperformed compared to their conventional counterparts. However, during an exogenous shock, the ESG funds would in fact outperform the non-ESG funds. COVID-19 crisis offers a unique opportunity to evaluate fund fragility ﬁrst-hand. If looking at the overall market (in this case, the MSCI Nordic benchmark), the index dropped by around 24% at the outbreak of the pandemic as portrayed in graph 5.

Graph 5: performance of MSCI Nordic benchmark

The ESG portfolio, fell on average by 25%, while the non-ESG portfolio fell at around 29%. This result is in line with Nofsinger and Varma (2013) conclusions, showing that ESG funds are less sensitive to exogenous shocks. Both of the portfolios had a larger drop than the market during the pandemic outbreak. This could be explained by the fact that the MSCI Nordic is better diversified. While the benchmark and ESG had a similar drop, the non-ESG portfolio had a larger drop by around 5%. This confirms the findings that ESG are less volatile and more resilient to exogenous shocks, yielding less financial fragility. As the ESG funds already had fixed characteristics before the outbreak, the affect on returns on both portfolios indicates that incorporation of ESG in an investment portfolio is in fact an endogenous variable.

29 Another interpretation is the expectation effect. Many seem to look at ESG almost as a new religion where everyone wants to buy it. As a result, the risk of ESG products’ become lower, since there is less disagreement on prices in the market. Furthermore, the returns may fall as well as a consequence of being over-valued. This could explain why ESG funds were more resilient to COVID-19 shocks as investors chose to invest in something with «quality» and less uncertainty compared to the conventional counterparts.

The location is another important aspect to consider. According to Ø. Fjell, during the outbreak of the COVID-19 pandemic, the Nordics performed better compared to other countries. Scandinavian funds are in general known to have strong sector diversiﬁcation and focus on ESG, making them less vulnerable to exogenous shocks. Therefore, the data used in this thesis could be an inaccurate representation of the the markets outside the Nordics and therefore one should be cautious when comparing these results with other publications.

8. Conclusion and further research

This thesis investigates whether ESG-investing has an affect on risk-adjusted returns and if there is a statistical signiﬁcance in the funds’ risk-return tradeoff in the Scandinavian market. ESG funds are sustainable and aim at long-term growth. It is an investment strategy that has gained momentum over the last years and has been incorporated in every recently adopted action plan regarding ﬁnancial sustainable growth. The emergence of new conventions and policies within sustainability indicates that the boundary between traditional and sustainable investments will disappear in the nearest future, therefore ESG should be incorporated as an additional risk parameter in every portfolio performance measurement tool.

The research question was chosen because the Scandinavian market has not previously been described in the academic literature focusing on ESG and the associated risk-return tradeoff. The Scandinavian market in general has very high ethical standards, meaning that even the conventional funds could be viewed as sustainable in other countries, due to strict screening policies. Therefore, to ensure comparability between regions, the sample was collected using Morningstar Sustainability Score, where the ESG funds in this thesis only have the highest score of 4-5 globes. The analysis was conducted on the basis of a total of 22 funds, consisting of 11 ESG and 11 non-ESG funds. This active ﬁltering has not been done before, which resulted in an unbiased ESG dataset of true sustainable funds.

30 In order to answer the research questions posed in the introduction, the thesis was divided into three main focus areas. Firstly, it was important to build an understanding of the concept of ESG and to provide enough background information in order to carry out a comparative analysis. This provided a good fundament for conducting an empirical analysis on the dataset speciﬁcally collected for this paper. The dominant indication was that ESG funds tend to exhibit a slight underperformance. The outbreak of the COVID-19 pandemic provided valuable insights to what extent these results gained both from the comparative and empirical analysis would hold. The ﬁndings showed that the ESG portfolio was less sensitive to the shock compared to the non-ESG portfolio, supporting the results found in the data analysis. Furthermore, as ESG funds incur on average lower costs, for those investors that are risk averse with a long-term perspective, ESG funds should be a preferable choice of investment.

Sustainable growth is key to the stable growth of economy. ESG is becoming a catalyst for companies to do something that resonates with them and has an impact on society at large, where not only do they share what they stand for, but what they stand up for. ESG investing extends beyond shareholders and short-term profits. By incorporating sustainable development, organisations are supportive both of their stakeholders and shareholders, leading to an economy that serves a broader constituency. The outbreak of COVID-19 was a breaking point for many, hastening the shift in mindset towards ESG investing. It proved that sustainability has a significant effect on risk, challenging the neoclassical economic theory where the sole focus has been maximising profits. High ESG-rated companies are better at managing idiosyncratic risk, therefore even early- stage startup companies are initially thinking about their impact. We find ourselves in the midst of a sustainability movement witnessing how companies tackle risk through the incorporation of environmental, social and governance investment strategies. Further research is needed of a broader Scandinavian portfolio in order to measure to what extent ESG affects risk-adjusted returns. However, both through comparative and empirical analysis it is possible to draw a conclusion that ESG has a significantly positive effect on risk-reward ratio of an investment.

31 9. References

“2022 - The Growth Opportunity of the Century.” PwC, 4 Nov. 2020, www.pwc.lu/en/sustainable- ﬁnance/esg-report-the-growth-opportunity-of-the-century.html.

Bauer, Rob, et al. “The Ethical Mutual Fund Performance Debate: New Evidence from Canada.” ResearchGate, Journal of Business Ethics , Nov. 2007, www.researchgate.net/publication/ 225148764_The_Ethical_Mutual_Fund_Performance_Debate_New_Evidence_from_Canad a.

Benjelloun, Hicham. Evans and Archer - Forty Years Later. Investment Management and Financial Innovations, 18 Mar. 2010, businessperspectives.org/images/pdf/applications/publishing/ templates/article/assets/3162/imﬁ_en_2010_01_Benjelloun.pdf.

Campbell, John Y., et al. Have Individual Stocks Become More Volatile? An Empirical Exploration of Idiosyncratic Risk. The Journal Of Finance, Feb. 2001, rady.ucsd.edu/faculty/directory/ valkanov/pub/classes/mfe/docs/Campbell_etal_JoF_2000.pdf.

Cornell, Bradford. “ESG Preferences, Risk and Return.” Wiley Online Library, John Wiley & Sons, Ltd, 17 Nov. 2020, onlinelibrary.wiley.com/doi/epdf/10.1111/eufm.12295.

“ENVIRONMENTAL, SOCIAL, AND GOVERNANCE ISSUES IN INVESTING.” CFA, CFA Institute, 2015, www.cfainstitute.org/-/media/documents/article/position-paper/esg-issues- in-investing-a-guide-for-investment-professionals.ashx.

ESG Approach and Principles. Morgan Stanley, 2019, www.morganstanley.com/im/publication/ resources/esg-approach-and-principles-us.pdf.

Fjell, Øyvind. “Hvorfor Er Det En God Idé å Kjøpe Nordiske Fond?” DNB Asset Management, DNB, 26 Nov. 2019, dnbam.com/no/ﬁnance-blog/nordiske-fond.

Global Sustainable Investment Review. Global Sustainable Investment Alliance, 2016, www.gsi- alliance.org/wp-content/uploads/2017/03/GSIR_Review2016.F.pdf.

Guidelines for Observation and Exclusion from the Government Pension Fund Global. Regjeringen, 2019, nettsteder.regjeringen.no/etikkradet3/ﬁles/2019/12/guidelines-for-observation-and- exclusion-from-the-gpfg-01.09.2019.pdf.

“Integrating Environmental, Social and Governance Value Drivers in Asset Management and Financial Research.” Scribd, OnValues Investment Strategies and Research Ltd., 25 Aug. 2005, www.scribd.com/fullscreen/16876744?access_key=key-mfg3d0usaiuaob4taki.

Kidd, Deborah. Measures of Risk-Adjusted Return: Let’s Not Forget Treynor and Jensen. CFA Institute, 2011, static1.squarespace.com/static/571a590286db43237523fc11/t/ 585168a1b3db2bd1b028b722/1481730218485/Measures+of+Risk-Adjusted+Return.pdf.

Kreander, Niklas, et al. “Evaluating the Performance of Ethical and Non-Ethical Funds: A Matched Pair Analysis.” ResearchGate, Journal of Business Finance & Accounting, Oct. 2005, www.researchgate.net.

Lantos, Geoffrey. (2001). The boundaries of strategic corporate social responsibility. Journal of Consumer Marketing. 19. 205-230. 32 Markowitz, Harry. Portfolio Selection. JSTOR, Mar. 1952, www.math.ust.hk/~maykwok/courses/ ma362/07F/markowitz_JF.pdf.

Maiti, Moinak. “Is ESG the Succeeding Risk Factor?” Journal of Sustainable Finance & Investment, 9 Feb. 2020, www.tandfonline.com/doi/pdf/10.1080/20430795.2020.1723380? needAccess=true.

“Morningstar Sustainability Rating.” Morningstar, Morningstar, 31 Oct. 2019, www.morningstar.com/content/dam/marketing/shared/Company/Trends/Sustainability/ Detail/Documents/SustainabilityRatingMethodology2019.pdf.

“MSCI Nordics ESG Universal Index (USD).” MSCI, 2021, www.msci.com/documents/ 10199/004d6325-8a99-d302-656b-af822cd2026a.

Nilsen, Heidi Rapp, et al. “The Norwegian Government Pension Fund Global: Risk Based Versus Ethical Investments.” EconStor, Leibniz-Informationszentrum Wirtschaft Leibniz Information Centre for Economics, 2019, www.econstor.eu/bitstream/ 10419/215469/1/10_3790_vjh_88_1_065.pdf.

Nofsinger, John, and Abhishek Varma. “Socially Responsible Funds and Market Crises.” Journal of Banking & Finance, North-Holland, 26 Dec. 2013, www.sciencedirect.com/science/article/ abs/pii/S0378426613004883.

Pollard, Julia. Establishing ESG as Risk Premia. The Journal Of Investment Management, Jan. 2018, www.researchgate.net/publication/327548430_Establishing_ESG_as_Risk_Premia.

Restriction Screening Policy. Morgan Stanley, Jan. 2021, www.morganstanley.com/im/publication/ msinvf/material/ msinvf_restrictionscreeningpolicy_sustainablestrategicbond_sustainablecorporatebond_en.p df.

Sandberg, Joakim, et al. “The Heterogeneity of Socially Responsible Investment.” Journal of Business Ethics, Springer Netherlands, 14 Oct. 2008, link.springer.com/article/10.1007/ s10551-008-9956-0.

Sustainable Reality. Morgan Stanley, 2019, www.morganstanley.com/pub/content/dam/msdotcom/ ideas/sustainable-investing-offers-ﬁnancial-performance-lowered-risk/ Sustainable_Reality_Analyzing_Risk_and_Returns_of_Sustainable_Funds.pdf.

33 Appendix

The following scatter plots of the ESG and non-ESG portfolios depict the performance of each fund based on their daily returns between 14th of February, 2019, and 1st of March, 2021. The sharp drops indicate the outbreak of COVID-19. The x-axis denotes the date in total amount of days analysed in the dataset.

Graph 6: Performance of the non-ESG fund Delphi Nordic A

Graph 7: Performance of the non-ESG fund DNB Grønt Norden A 34

Graph 8: Performance of the non-ESG fund DNB Norden R

Graph 9: Performance of the non-ESG fund DNB Norden C

35 Graph 10: Performance of the non-ESG fund Evli Nordic B

Graph 11: Performance of the non-ESG fund Lannebo Norden Hållbar SEK

36 Graph 12: Performance of the non-ESG fund ODIN Norden A

Graph 13: Performance of the non-ESG fund ODIN Norden B

37 Graph 14: Performance of the non-ESG fund Pareto Nordic Equity B NOK

Graph 15: Performance of the non-ESG fund Pareto Nordic Equity C NOK

38 Graph 16: Performance of the non-ESG fund Skandia Ideer For Livet

Graph 17: Performance of the ESG fund Nordea 1 Nordic Stars BC NOK

39 Graph 18: Performance of the ESG fund DNB Barnefond

Graph 19: Performance of the ESG fund Eika Norden

40 Graph 20: Performance of the ESG fund Handelsbanken Norden Selektiv (A1 NOK)

Graph 21: Performance of the ESG fund Handelsbanken Norden Selektiv (A10 NOK)

41 Graph 22: Performance of the ESG fund Landkreditt Norden Utbytte A

Graph 23: Performance of the ESG fund Landkreditt Norden Utbytte I

42 Graph 24: Performance of the ESG fund Swedbank Robur Nordenfond

Graph 25: Performance of the ESG fund Nordea 1 - Nordic Stars Equity Fund BI NOK

43 Graph 26: Performance of the ESG fund Nordea 1 - Nordic Stars Equity Fund BP EUR

Graph 27: Performance of the ESG fund Nordea 1 - Nordic Stars Equity Fund BP NOK

44 Summary statistics

Non-ESG portfolio (regression for Jensen’s alpha):

ESG portfolio (regression for Jensen’s alpha):

t-Test: Paired Two Sample for Means

45 Variable 1 Variable 2 Mean 0,000729851 0,000879425 Variance 1,48205E−08 3,17436E−08 Observations 11 11 Pearson Correlation −0,371006328 Hypothesized Mean 0 Difference df 10 t Stat −1,98181 P(T<=t) one-tail 0,037822287 t Critical one-tail 1,812461123 P(T<=t) two-tail 0,075644573 t Critical two-tail 2,228138852 Table 6: Difference in mean for return of both portfolios (t-stat for return)

t-Test: Paired Two Sample for Means stdv

Variable 1 Variable 2 Mean 0,01086232 0,011882718 Variance 1,34317E−06 1,28101E−06 Observations 11 11 Pearson Correlation 0,035714188 Hypothesized Mean 0 Difference df 10 t Stat −2,127473903 P(T<=t) one-tail 0,029636471 t Critical one-tail 1,812461123 P(T<=t) two-tail 0,059272943 t Critical two-tail 2,228138852 Table 7: Difference in mean for StDev of both portfolios (t-stat for standard deviation)

46 Variable 1 Variable 2 Mean 0,96317543 1,089463636 Variance 0,037007624 0,091027055 Observations 11 11 Pearson Correlation −0,324294176 Hypothesized Mean 0 Difference df 10 t Stat −1,029023622 P(T<=t) one-tail 0,163857468 t Critical one-tail 1,812461123 P(T<=t) two-tail 0,327714937 t Critical two-tail 2,228138852 Table 8: Difference in mean for Sharpe Ratio of both portfolios (t-stat for Sharpe Ratio)