Final Thesis
SUSTAINABLE INVESTMENTS
Looking for long-term shareholder value
Nico Rieske
1Faculty of Economics and Econometrics, University of Amsterdam, July 2007
Instructor: Ludovic Phalippou 1 Student number: 0360295
Abstract
More and more pension funds are feeling pressure from external organizations, and public opinion, to adopt a more social responsible role. For long the debate about Socially
Responsible Investing (SRI) for pension funds in Europe has centered on the main question: what are the costs, in terms of risk and return, of a more sustainable or ethical investment policy. This question relates to the fiduciairy duty of pension funds, which basically says that these funds have to find the highest return possible, within certain risk limits. A potential conflict between these two responsibilities has been the center of the debate.
Working for a pension fund in the Netherlands, DSM Pension Services2, the question about SRI or sustainable investments has intrigued me and finally resulted in this study. Where almost all historical studies focus on the environmental aspect of SRI exclusively, this study takes a broader definition of SRI. It analyses the effects of integrating a sustainability factor, consisting of an environmental and a social factor (both equally weighted), into the stock selection process. What are the implications of sustainable criteria entering the stock selection process, in terms of risk and return?
The study is done for a broad universe of European public companies over the 2000-2006 period. The statistical data suggests that a portfolio of stocks of sustainable companies underperformed a portfolio of non-sustainable companies. This difference was however not statistically significant. The risk-adjusted return of the sustainable portfolio, as measured by the Sharpe ratio, was also inferior to that of the portfolio of non-sustainable companies. The study shows that both the sustainable portfolio and the non-sustainable portfolio had a large cap bias, which is consistent with prior empirical research.
A potential conflict between SRI and the fiduciairy duty of pension funds is confirmed by this study. Therefore it is recommended that pension funds are careful with the implementation of sustainable factors into the investment process.
2 DSM Pension Services (DPS) is a full subsidiary of DSM N.V. DPS is an in-house pension organization working for DSM and DSM affiliated pension funds.
Acknowledgements
I would like to thank the following people for their help by accomplishing this study in various ways.
First I would like to thank Pieter van Stijn (Research Analyst) and Joos Grapperhaus (Director Institutional Sales & Marketing) at SNS Asset management. Without their database, this study would not have been possible.
Next I would like to thank my colleagues at DSM Pension Services who have been involved (one way or the other) in interesting discussions about this complex topic. Especially I would like to thank Ger van Neer, Ewout Gillissen, Wilfried Bolt and Daniel Rijs.
Last but not least I would like to thank my supervisor, Ludovic Phalippou, for providing me with useful comments.
Table of contents:
1. SUBJECT 6
2. INTRODUCTION 7
3. THEORY 11
3.1. Corporate Social Responsibility (CSR), value creation or destruction? 11 3.2. What are the implications of sustainability for investors? 14 3.2.1. The general belief: “SRI portfolios are suboptimal” 15 3.2.2. Proponents of Sustainable Investing: “Sustainable portfolios outperform” 17 3.2.2.1. Sustainable companies, a valuation premium! 18 3.2.2.2. Sustainable companies, trading at a valuation discount? 19
4. LITERATURE 21
5. HYPOTHESIS (RESEARCH QUESTION) 26
6. METHODOLOGY 28
6.1. Best-in-Class Method 29
8. EMPERICAL ANALYSES 33
8.1. Portfolio Performance in a CAPM framework 34
8.2. Portfolio Performance in a Multifactor framework 35
8.3. Portfolio Characteristics 38
9. CONCLUSION 39
REFERENCES 40
1.
Subject
The central question of this paper is about the risk-return profile of sustainable investment portfolios. Sustainable investments is part of SRI, but avoids ethical questions like for example investments in manufacturers of armaments, tobacco and gambling. Sustainable investments take into account social- and environmental factors, as well as the profitability of a company. It balances these factors, such that all stakeholders to a company are better off. As conflicts arise, SRI investors, like all other investors, prevail their own interests over that of competing stakeholders.
But what does sustainable investing mean for investors, like pension funds? What are the implications for a well diversified equity portfolio, in terms of risk and return, when a sustainability factor is introduced into the stock selection process? Does sustainability create risk-adjusted excess returns, or does sustainability lead to lower risk-adjusted returns on a portfolio level?
In order to answer this question two equally weighted equity portfolios were constructed, which differed based upon a sustainability factor. A relative approach towards sustainability is used, the so called best-in-class method. This method ranks companies, based on their social and environmental performance, against its industry peers. The highest ranking companies of each sector (top 20%), comprised the so called sustainable portfolio. The lowest ranking stocks of each sector (bottom 20%) comprised the non-sustainable portfolio. The Sharpe ratio is used to compare the risk-adjusted returns of each portfolio. The study covers circa 300 publicly listed European companies (including the UK) over the 2000-2006 period, using sustainability scores from SNS Asset Management in the Netherlands.
2.
Introduction
In recent decades the investment process of institutional investors has been influenced by social-, environmental and ethical factors. This trend is often referred to as Socially
Responsible Investment (SRI). SRI is a broad concept that encompasses investments based on sustainability and ethics. Some of the world’s largest pension funds have already adopted some form of SRI into their investment process. Pension funds, like for example ABP and PGGM in the Netherlands, are publicly3 communicating their commitment to SRI. The 2006 European SRI Study (Eurosif)4 shows, that the broad5 SRI market’s assets under management (AuM) is currently valued at over €1 trillion as of December 31st, 2005. The core6 SRI market is valued at €105 billion. They conclude that, across Europe, more and more institutional investors are demanding mandates based on SRI strategies. From December 2003 to March 2005, assets in SRI funds (broad) increased by 35 %, against 16% for the European market in total. In the US it is estimated that almost 12% of money under professional management is part of a social responsible portfolio (Social Invest Forum, 2001 Trends Report)7.
In this paper the question about ethics will be avoided. For years there have been debates about ethics, but it has remained a rather undefined area. This thesis exclusively focuses on the concept of sustainability, consisting of social and environmental factors. Sustainable investing takes as its basis the objective of balancing social, environmental and economic factors. The study compares the risk-adjusted return of a portfolio of stocks of sustainable companies to that of a portfolio of stocks of non-sustainable companies. Using a
comprehensive database of SNS Asset Management in the Netherlands, the consequences of a sustainability factor entering the stock selection process will be analyzed.
3 For more information, please check the respective websites; www.abp.nl and www.pggm.nl 4http://www.eurosif.org/publications/sri_studies
5
The broad SRI market consists of all AuM, invested under some form of SRI, ranging from ethical exclusions, positive screening (best-in-class) to engagement and other forms of SRI (see chapter 6)
6 The core SRI market consists of all AuM invested under ethical exclusions (negative screening) and positive screenings (best-in-class)
7
SRI is in a way the investors’ response to sustainable development. For a long time
companies were supposed to be profit maximizing entities. Milton Friedman once said “The business of business is business”, all other things lead to a less optimal focus of a company. Other performance measures, like the contribution to society as a whole or the natural environment, have been of inferior importance. More recently however, because of climate change and other environmental and social issues, companies are changing strategy to
embrace a more social responsible role. Sustainability or corporate social responsibility (CSR) will most probably affect the future way corporations run their business. The large number of companies in the S&P 500, disclosing social responsible information, can be seen as evidence of this changing role of companies (Bloomberg).
There is however still ongoing debate about the question, whether or not sustainability adds value to a company. Two theories about this subject still exist. People against sustainability, like for example Walley and Whitehead (1994), argue that it leads to a competitive
disadvantage for a company, that the definition is still vague and that it distracts business from their core focus of optimizing profits. However, Porter and Van der Linde (1995), Spicer (1978) and other proponents of corporate social responsibility (CSR), argue that a sustainable business has advantages. They argue that a sustainable company has competitive advantages, it faces less reputation risk, it has a better relationship with its stakeholder and, as a result, it has a lower cost of capital.
If true, these two theories should affect stock prices. If sustainability creates shareholder value, then these companies should trade at a premium multiple (like for example a higher price-to-earnings ratio) to their non-sustainable peers. If sustainability just leads to higher costs, then these stocks should trade at a discount to their sustainable peers. However to the extent sustainable criteria affect the valuation of stocks, and potentially even affect stock returns (as is claimed in several scientific papers) depends on the financial markets ability to factor potential benefits (or costs) of sustainability into share prices. This will be extensively discussed in Chapter 3.
The general belief is that incorporating sustainability into the investment process comes at a cost, in terms of risk and return. This belief relies on asset-pricing theory, based on the hypothesis of efficient markets. According to this theory (i) portfolios deliver returns
proportional to its associated risk and (ii) the optimal portfolio is a well-diversified one. Any risk-adjusted anomalies are attributable to deficiencies in performance attribution models, which try to explain them. According to this theory, SRI investors should report suboptimal risk-adjusted returns, because of self imposed limits to diversification.
Proponents of sustainable investing however argue that, market participants in general do not price sustainability correctly. As a result an incorrect pricing could exist, as also suggested by Moskowitz (1972). If sustainability influences asset pricing and if these factors are not (fully) priced in yet, it could be possible that an investor is able to create risk-adjusted excess return, or so called alpha8 to offset the additional risk of these screened portfolio. The economic return relating to investments in a more sustainable business model (if any) could very likely be further in the future then the horizon of the equity markets participants in general. Long term investors, like for example pension funds, are best equipped to exploit such potential incorrect pricings.
The study is based on and extends the research performed by Derwall et al (2005), who in their article “The Eco-Efficiency Premium Puzzle” investigated the effects of corporate environmental performance on portfolio return. However, their study, like most other studies, only focuses on the environmental aspect of sustainability and neglect the social dimension (Ziegler, 2002). To my best knowledge, with the exception of Ziegler, the social aspect of sustainable performance is only considered in the investigation of ethical mutual funds so far (see e.g. Statman, 2000, Kreander et al., 2000).
This thesis combines an environmental and a social factor into one quantitative score, because of the still ongoing debate about a possible conflict of sustainability with the fiduciary duty of pension funds.
8 Alpha is the abnormal rate of return on a portfolio in excess of what would be predicted by an equilibrium model like CAPM.
A sustainability score is determined by comparing companies within their respective sector, the so called best-in-class methodology. Two equally weighted equity portfolios, which differed based on sustainability, were constructed and analyzed. The top 20% of the
sustainable companies of each sector comprised the so called “sustainable portfolio”, whereas the bottom 20% comprised the “non-sustainable portfolio”. This relative method of
sustainability, enables an institutional investor to create a well-diversified portfolio
(consisting of stocks of all sectors), while incorporating a SRI policy. It was also analyzed whether or not sustainable portfolios are tilted towards large capitalization stocks and value or growth stocks, as is the claimed by current empirical research, for example Bauer et al.
(2005).
The remainder of this thesis will be as follows. First, in chapter 3, the theory concerning sustainable development and sustainable investments will be put forward. The current debate about value creation regarding sustainable companies is relevant with respect to a potential mispricing of sustainability by financial markets. Chapter 4 gives a review of the existing literature concerning SRI. The following chapters give: the research question (Chapter 5); the methodology used (Chapter 6); the dataset (Chapter 7); and the empirical analyses
3.
Theory
In this chapter the theory concerning sustainable investments will be discussed. This theory is closely linked to the current debate about corporate social responsibility (CSR). Therefore the ongoing debate about CSR will be described first in more detail in the next paragraph.
3.1. Corporate Social Responsibility (CSR), value creation or destruction?
The question whether or not CSR adds shareholder value to a company is still debated. This paragraph describes the current debate, which is relevant regarding a potential incorrect pricing of sustainable companies by financial markets in general. The ongoing debate could be a possible explanation for a potential valuation discount of sustainable companies, as suggested by various research papers (Guenster et al., 2005). Another possible explanation is that the horizon of equity markets in general is too short, to factor in the benefits of CSR. These explanations will be further discussed in the following paragraphs.
“…..creating value takes more than the acceptance of value maximization as the
organizational objective. As a statement of corporate purpose or vision, value maximization is not likely to tab into the energy and enthusiasm of employees and managers to create
value....A firm cannot maximize value if it ignores the interest of all its stakeholders…”, Michael Jensen (Journal of Applied Corporate Finance, 2001)
Proponents argue that sustainable companies have competitive advantages, reputation
advantages, and a better relationship with its stakeholders. The result of this is a lower cost of capital for a sustainable company. Also they argue that sustainable companies have superior management, which results in a more innovative company. This last argument although seems to be a causality. In “Green and Competitive”, Porter and Van der Linde (1995) describe a company that is able to do financially well, while, at the same time, doing good regarding sustainable issues. They speak of a win-win situation, where all stakeholders to a company are better off. According to them companies should be innovative, which eventually leads to a more efficient production process. Such innovations allow companies to use a wide range of inputs more efficiently – from raw materials to energy to labor - offsetting the upfront investments. As an example they describe the complete transformation of the Dutch flower
industry. The cultivation of flowers was devastating for the environment, because of the use of pesticides, herbicides and fertilizers. By completely reinventing the production process, the way flowers were cultivated was fundamentally changed. Now, because of the use of
greenhouses, the damage to the environment is much less. But also the quality of the flowers became better and labor was used more efficiently, making the industry more competitive and opening up new markets.
Opponents argue that the definition of sustainability is still undefined, that it distracts a company from their core focus of maximizing profits and that the revenues of these
investments are uncertain. Walley and Whitehead (1994) for example argue that a company should only comply with minimum standards, as required by law, with regards to their social- and environmental policy. The win-win situation, as described by Porter and Van der Linde, is according to them very rare. In general, investments in sustainability can be seen as sunk costs without any direct economic payoff. As a result these investments decrease shareholder value. They give the example of the American chemical company “DuPont”, which according to the authors had 35% of the value of the company invested in the protection of the environment. According to them these investments yielded no economic returns and, as a result, destroyed shareholder value.
According to Ullman (1995) the corresponding relevant hypotheses concerning the relationship between sustainable performance and economic performance9 are:
Hypothesis 1: Sustainable performance is positively correlated with economic performance. There are economic benefits to being a sustainable company and these benefits are greater than the costs associated with it (i.e. sustainability increases shareholder value)
Hypothesis 2: Sustainable performance is negatively correlated with economic performance Because of the upfront investments in sustainability by a company some studies posit a negative correlation between sustainability and performance (i.e. sustainability decreases shareholder value)
9
Hypothesis 3: Extreme levels of sustainable performance are associated with low economic performance. There exits an inverted U-shaped correlation. This suggests that there is an optimal level of allocation of resources to social and environmental factors.
Although the debate about value creation (or destruction) concerning CSR is still ongoing, it is personally thought that hypothesis 3 is most valid in relation to sustainability. A company that invests in the win/win situation, as described, will most likely increase shareholder value. However management and shareholders should be aware of an optimal level of these
investments. If they invest too much money into sustainability it will most likely decrease shareholder value. But also companies that don’t consider social- and environmental issues at all will most likely destroy value. The optimal level of investments in sustainability will vary from company to company. This level is determined by the point where the marginal benefits of sustainability equals the marginal costs.
This measurement problem is, according to my opinion, one of the main problem of databases that are used by investors to rank companies on their sustainability. These databases generally use absolute values of the social and environmental behavior of companies. Thus a company that is the most social or environmental responsible ranks the highest. However this does not mean this company creates the highest shareholder value, because of the U-shaped
correlation.
There are several problems affecting the debate between proponents and opponents of
sustainable development. Below two of these issues are described: First, the relation between sustainability and economic performance, as measured by for example market returns or accounting variables, is complex. A positive correlation between these two parameters could imply that the efforts of a company to do well are rewarded in terms of financial performance, i.e. sustainability adds value to a company. It could however also be argued that, only
companies that do financially well are able to invest in a more sustainable business, because of the additional costs (Alexander and Buchholz ,1978). In the latter case sustainable
companies create shareholder value because of accounting variables and not because of their more sustainable business model. Second, companies can display sustainable awareness by “end-of- pipe” pollution control as emphasized by Hart and Ahuja (1996) and Russo and Touts (1997). In stead of actively pursuing a win/win situation, companies clean up their
waste subsequent to the production process. For outsiders, like investors, this is very hard to distinguish.
Although these problems are very difficult to solve, it is believed that there are very likely economic benefits to a sustainable business model. A company that actively pursues
innovation of the production processes will most likely optimize these processes eventually. A reduced use of resources and/or a more efficient use of workforce will very likely lower the cost price per unit of product. If this company is able to keep its selling price constant, because its competitors are unable to keep up, it will report higher earnings. The question is whether or not these economic benefits (higher earnings) outweigh the up-front investment. The next paragraph describes the implications of CSR for investors.
3.2. What are the implications of sustainability for investors?
The theoretical debate relating to the potential virtues of sustainable investing, can be structured along several lines. Hamilton et al (1993) mention three:
1. The Value Relevance Hypothesis:
The first hypothesis posits that: information about the environmental- and social behavior of a company is relevant for the valuation of a company and its stock returns. The ability of the financial markets to factor in potential benefits of sustainable companies plays a crucial roll in this respect. Litigation risk for example can be of influence on the expected cash flows of a company and its cost of capital. An important question is whether or not this information is already factored into stock prices? When markets do not price sustainability correctly (yet), the expected risk-adjusted returns on portfolios of stocks of sustainable companies could be higher than the returns on portfolios of stocks of non-sustainable companies.
2. The Stock Neglect Hypothesis:
The second hypothesis posits that: there is a significant group of SRI investors with
preferences that could possible lead to imperfect markets10. SRI investors drive up prices of sustainable companies by driving down expected returns and a firm’s cost of capital.
Therefore, sustainable portfolios deliver lower risk-adjusted returns.
Hong et al. (2006) provide evidence of this by comparing stock returns of “sin” stocks, publicly traded companies involved in producing alcohol, tobacco and gaming. They find that these stocks have higher expected returns than other comparable stocks. According to them this is the result of these stocks being neglected by norm-constrained investors, like for
example pension funds. The study covers the US market (NYSE, AMEX, NASDAQ) over the period of 1962-2003.
3. The Irrelevance Hypothesis:
The third hypothesis posits that: information about the environmental- and social behavior of a company is not relevant for the valuation of a company and its stock returns. Sustainable information does not influence asset-pricing because there is no relation, or the accompanying risk is not priced (diversifiable risks). Sustainable portfolios are by definition suboptimal, because of the higher risk of these portfolios.
3.2.1 The general belief: “SRI portfolios are suboptimal”
The belief of many institutional investors is that the risk-return profile of a sustainable portfolio is suboptimal. In theory, sustainable funds (social, environmental and ethical screened funds) take on higher risk than unscreened funds, because they limit the number of stocks in which they can invest.
This line of thinking corresponds with modern portfolio theory that relies on the hypothesis of efficient markets, where all information is priced into share prices. Deviations from the
10 Meaning that again; information about the environmental- and social behavior of a company is relevant for the valuation of a company and its stock returns.
efficient portfolio (the Market Portfolio11) are by definition suboptimal, because it increases risk. This additional risk yields no abnormal return, because of market efficiency. As a result the risk-adjusted return of a sustainable portfolio is by definition suboptimal compared to “the Market Portfolio”.
According to the asset-pricing model of Sharpe (1964) and Litner (1965), the level of expected return of a stock is a function of its respective degree of systematic risk exposure, which is referred to as beta (β).
Ε(r
i) =r
f+ β
i(Ε(r
m)- r
f)
where
Ε(ri) = the expected return on a security as of the beginning of the holding period.
Rft = the risk free rate
β
i = the beta coefficient of the security with the market portfolioΕ(rm) = the expected return on the market
This theory states:
1. The optimal portfolio is a well diversified portfolio, known as the Market Portfolio. 2. Any deviation from this Market Portfolio results in a lower risk-adjusted return.
Anomalies are, according to Sharpe and Litner, a result of differences in methodology. A portfolio based on for example the size of a company or its book-value-to-price ratio is, as a direct consequence, less optimal then an investment in this so called Market Portfolio. If this theory holds it is impossible to create alpha (risk-adjusted excess return) by selecting stocks based on a sustainability factor (or any factor at all). A less optimal risk-reward ratio of a sustainable portfolio is a direct consequence of self imposed limitations (i.e. higher risk).
However, Fama and French (1993) show that return on common stock show little relation with Sharpe’s and Litner’s β. They identify two risk factors that show powerful predictors of
11
the market as a whole: (i) the size of a company (ii) its book-value-to-price ratio. These findings are an important challenge to the notion of efficient markets.
3.2.2 Proponents of Sustainable Investing: “Sustainable portfolios outperform”
Proponents of sustainable investing belief that in order to optimize shareholder value, companies are required to balance their economic, environmental and social responsibilities. Where conflicts arise companies have to find creative solutions to reduce the negative impact on any stakeholder group. However, where conflicts cannot be resolved, SRI investors, like all other investors, require companies to prioritize shareholder interests over competing demands. SRI investors are however inclined to take a longer view of a situation than other, more conventional investors. SRI investors belief in the described win-win situation, where all stakeholders to a firm are better off.
Proponents of sustainable investments suggest that sustainable screened funds could potentially outperform, because they use important information not well understood by the broader market. They argue that sustainable companies are currently undervalued. The reasons for this misunderstanding could be twofold: (i) the still ongoing debate centered on value creation of sustainability, and (ii) the belief that equity market participants in general have a shorter horizon than the potential benefits of sustainability. Hamilton et al. (1993), for example, mention the possibility that investors underestimate the likelihood of damage to a company’s reputation, because of negative news flow. Long term investors, like for example pension funds, are best equipped to exploit such potential incorrect pricings.
These arguments are often used to persuade pension funds to adopt a more social responsible role. If sustainable portfolios produce risk-adjusted excess return, there would be no conflict between sustainable investing and the fiduciairy duty of pension funds.
The potential valuation discount of sustainable companies will be further explained in the remainder of this chapter. First I like to make the following remarks: (i) for this argument to be true, it has to be assumed that: while in the past the value of sustainability has been misinterpreted by financial markets, now investors suddenly realize this value and act on it, and (ii) a situation of abnormal returns of sustainable portfolios will only last temporarily. Once this information is priced “correctly”, these stocks will trade in equilibrium again. The
return on these stocks is then again determined by its respective degree of systematic risk exposure (β). As a result, sustainable portfolios will again be suboptimal, because of its higher risk.
3.2.2.1 Sustainable companies, a valuation premium!
According to proponents of SRI, sustainability adds shareholder value to a company.
Therefore these companies should trade at a premium (higher multiple) compared to their not sustainable peers, Guenster et al. (2005). This premium valuation multiple can be best
explained by the following example:
Company C is a conventional company, which does not invest in sustainability. It has a choice to invest in its sustainable business model. This transforms company C, into a sustainable company, denoted as S. These companies only differ based on sustainability. Table 1, shows the net present value (NPV12)of both companies. It is assumed that the investments of company S, in its sustainable business model, reap economic returns13. Therefore company S creates more shareholder value than company C. This is reflected by a higher NPV value of company S (€ 55,1) versus company C (€ 52,1).
Table 1:
Conventional company C ; cash flows
Years 1 2 3 4 5 6 7
CF's € 10,0 € 10,0 € 10,0 € 10,0 € 10,0 € 10,0 € 10,0
PV of CF's € 9,3 € 8,6 € 7,9 € 7,4 € 6,8 € 6,3 € 5,8
NPV € 52,1
Sustainable company S ; cash flows
Years 1 2 3 4 5 6 7 CF's € 8,0 € 9,0 € 10,0 € 11,0 € 12,0 € 13,0 € 13,0 PV of CF's € 7,4 € 7,7 € 7,9 € 8,1 € 8,2 € 8,2 € 7,6 NPV € 55,1 t d FVt NPV N t
∑
= + = 0 (1 )12 It is assumed that these companies cease to exist after 7 years. 13
where
• NPV is the discounted cash flow of future cash flows;
• FV is the nominal value of a cash flow (CF) amount in a future period; • d is the discount rate (8%)
• n is the number op periods used
As can been seen from table 1, the structure of the cash flows14 of both companies is different. Because company S invests in its sustainable business model, its more current cash flows (first 2 years) are below that of company C. Its future cash flows are however higher because of these investments.
Much of the real-world discussion of stock market valuation concentrates on a firm’s price-earnings multiple, the ratio of price per share to price-earnings per share. If we value both
companies on a P/E multiple15, in a way a very simplified discounted cash flow analysis, we get a fair value P/E multiple of:
• Company C; P/E = € 52,1/ € 9,3 = 5,6 • Company S; P/E = € 55,1/ € 7,4 = 7,4
To be fairly valued, company S should trade at a premium P/E multiple of 7.4 compared to company C (5.6).
3.2.1 Sustainable companies, trading at a valuation discount?
Proponents of sustainable investments claim that the broader financial markets do not value the future benefits of sustainable companies correctly, because (i) this information is not well understood by the broader market, and (ii) the horizon of market participants in general is too short.
In the following example it is assumed that financial markets do not look at the benefits of sustainability at all (although this is found highly unlikely). Therefore although company C
14 The cash flows are at the end of every year, discounted at 8% annually 15
changes to the more sustainable company S, it is assumed to be valued on the same P/E multiple as the conventional company C (5.6).
Next it is assumed that the horizon of the market is one year. So it is assumed financial markets value companies on next years earnings. For company S this means that the costs of the sustainable investments in the first year, falls within the horizon of the financial market. The market value of company S, as denoted by S market :
• 5.6 (=P/E, company C) * € 7.4 (= cash flow year 1, discounted at 8%) = € 41.4
So in this particular case, the market underestimates the real value of company “S” (€ 55.1), by almost 25%.
Although this scenario is highly unlikely, it is believed that a potential undervaluation of sustainable companies by financial market is possible. This because of the reasons, as already described. The market will most likely value some of the benefits of sustainability, although not all: PE = C < S market < S. Statistical analyses by Guenster et al. (2005) suggest that
sustainable companies, measured by an environmental factor, trade at a premium compared to their peers and that this premium has increased strongly over time.
The ability of the financial markets to factor in potential benefits of sustainable companies plays a crucial roll in this respect. Proxies for sustainability, for example, are most likely subjective and questionable. Not the least because of the U-shaped correlation, as described. The most accurate way probably is, to ingrate social-, environmental and financial
performance into one discounted cash flow analysis, by identifying additional sources of risk and opportunity. Analysts until today however mainly focus on financial criteria to value a company.
4.
Literature
Empirical studies concerning the relationship between sustainable performance of a company and its financial performance, as measured by accounting measures as well as market based measures, are still inconclusive. Many studies have been done that focus on several aspects of the relation between corporate social and environmental responsibility (CSR) and corporate financial performance (CFP). Most empirical studies focus exclusively on an environmental factor as a proxy for sustainability, and neglect the social performance of a company.
Ullmann (1985), Griffin and Mahon (1997) and Orlitzky et al. (2003) give a review of the existing literature regarding SRI studies.
Ullman (1985) concludes that the findings regarding the social and environmental disclosure, social and environmental performance and economic performance for US corporations are inconsistent. He attributes the inconsistent findings to a lack in theory (i.e. not well defined), an inappropriate definition of key terms and deficiencies in the available data bases.
According to Ullmann, confusion is almost inevitable given conflicting hypotheses, as well as differences in models, methods, measurements and time periods considered.
The evidence that environmental factors are related to the financial performance of a company is inconsistent. According to Griffin and Mahon (1997) the main problem in prior research is methodological inconsistencies and differences in the choice of financial en environmental performance indicators.
Orlitzky et al. (2003) conducted a meta-analysis of 52 studies (almost 34.000 market based and accounting based observations). They conclude that the general belief, that prior SRI research is too fractured to draw any conclusions, is built on shaky grounds. According to their study there is a positive correlation between social responsibility and financial
performance. The evidence of a positive relation between environmental responsibility and financial performance is less convincing. The study shows that evidence on a positive relation between CSPR and CFP is best visible in the accounting- based measures of a firm. Evidence on market-based measures is less convincing. Sustainability indices are more correlated to CFP than any other measure of CSR.
According to Wagner (2001), SRI studies, based on financial market performance, can be divided into three groups: Analyses of fund performances, event studies (short observation periods) and econometric studies with longer observation periods.
Fund studies can be based upon analyses of existing mutual SRI funds16, or upon virtual funds (screened funds), so called portfolio studies, created by the researcher. Portfolio studies rank a universe of companies, based upon SRI criteria. These studies compare the performance of sustainable portfolios (or existing SRI funds) to the performance of non-sustainable portfolios (or conventional funds).
Studies on mutual funds are mainly focused on the US and UK markets. Hamilton et al. (1993) and Statman (2000) studied ethical mutual funds in the United States. After comparing risk-adjusted returns of these funds to the S&P500 and the Domini Social Index, they find no significant differences. Luther et al. (1992) find weak evidence of outperformance of ethical mutual funds in the UK market. Also it is found that these funds are tilted towards small cap stocks. After controlling for this small cap bias, Mallin et al. (1995), confirm the
outperformance of UK ethical mutual funds over conventional mutual funds. Bauer et al. (2005) looked at SRI indices in the US, Germany, and the UK and found no evidence of significant differences between SRI indices and conventional indices. The study controls for size, book-to-market and stock price momentum. Scholtens (2005) comes to the same conclusion for Dutch social responsible funds, after controlling for investment style. Scholtens finds that social responsible funds are tilted to value stocks. Studies based on existing SRI funds have two main weaknesses. First it is almost impossible to isolate the influence of management skill on fund performance. Second, no correction is made for other variables that are known to influence performance, like for example the size of a company or its valuation (Wagner, 2001).
A portfolio study by Diltz (1995) found not much evidence of a relation between returns and several ethical factors for US stocks. However, portfolios based on environmental factors only, outperformed on a risk-adjusted basis. These results were statistical significant during the 1989-1991 timeframe.
16
No significant environmental results were found in a US study done by Cohen et al. (1997). Their study examined the performance difference between low-polluter and high-polluter companies within industries (S&P 500). Contrary to Diltz, they conclude that there is neither a premium nor a penalty for investing in companies that are environmental leaders within their industry. Like Diltz, also Yamashita et al (1999) find support that selecting stocks, based on environmental factors, add value to the investment process. They show that a portfolio of environmental leaders performed significantly better than a portfolio of environmental laggards. Derwall et al. (2005) use “eco-efficiency” to measure the environmental
performance of a company. Companies are ranked within their own industry based on this eco-efficiency score, which is the ratio of the value added by a company (by producing products) to the waste it generates. By using a relative approach to sustainability they avoided that sustainable portfolios were skewed to sustainable sectors. The study shows that the sustainable portfolio performed substantially better than the non-sustainable portfolio. This performance difference could not be explained by market sensitivity, investment style or industry-specific factors. The study covers more than 1200 US companies over the 1997-2003 period.
Event studies show the most promising evidence of a relation between environmental factors and stock returns. The main weakness of event studies is its short term character. These short term abnormal returns may very well be compensated over time. Shane and Spicer (1983) study security price movements associated with external information, from the US Council of Economic Priorities (CEP) of a company’s environmental performance. The study indicates that these so called CEP stocks experience abnormal declines two days prior to the release of their environmental performance data. On the day of the release of the environmental
performance data, low ranking companies were found to have significantly lower returns than their higher scoring peers. Hamilton (1995) found significant evidence of negative abnormal returns on the release of pollution figures in the US. Klassen & McLaughlin (1996) find that companies, which receive an external reward for their environmental performance, measure significant excess returns. Negative environmental events (an environmental crisis), lead to significant underperformance of the company involved. These event studies are mostly based on the CAPM model.
To overcome the restrictions of fund studies and event studies, more recent research has been done on longer observation periods using econometric models. This category of research uses (multivariate) regression or correlation analyses to measure the effects of environmental responsibility on stock performance.
McWilliams and Siegel (2000) argue that in existing econometric studies an important control variable is missing. Most studies controlling for size, book-to-market price and market
sensitivity are missing one important variable, the amount a company spends on R&D. According to the authors, this misspecification leads to upward biased estimates of the effect of CSR on financial market performance. They conclude that the impact of CSR on CFP is neutral after controlling for R&D. Ziegler et al. (2002) analyze the effects of CSR of European companies on shareholder value, as measured by stock returns from 1996-2001. They find that an increasing environmental sector performance has a significant positive effect on shareholder value. In contrast, an increasing social sector performance has a
negative effect on shareholder returns. Within sectors, the effect of social and environmental factors, have no significant impact on shareholder value. Guenster et al.(2005) analyze the relation between the environmental performance of a company, measured by the
eco-efficiency rating, and Tobin’s q. Tobin’s q is the ratio of the market value of a firm’s assets to the replacement cost of the firm’s assets (Tobin, 1969). The study covers firms listed on the US stock market from 1992 to 2002. They conclude that sustainable companies trade at a premium compared to their peers and that this premium has increased strongly over time. Environmental leaders do not have a return on assets superior to that of a control group, but the low ranking companies (environmental laggards) have a significant lower return on their assets. The study extends the paper by Derwall et al. (2005), who performed a portfolio study based on the same eco-efficiency rating. They find that portfolios of environmental leaders performed superior on a risk adjusted basis compared to portfolios of environmental laggards.
Important is the mounting evidence that SRI portfolios are biased to certain investment styles. According to Bauer et al. (2005) size and value versus growth effects are explanatory of a significant portion of the performance of SRI portfolios. According to them ethical mutual funds in the US tilted are tilted towards large cap stocks, whereas in the UK and Germany these funds are heavily exposed to small caps. Also these funds tend to be more growth-oriented. Luther et al (1992) and Luther and Matatko (1994) show the existence of a small cap bias for UK ethical mutual funds. On the contrary, US studies show evidence of a large cap growth bias (for instance Guerard (1997) and Kurtz (1997). Garz et al. (2002) look into the Dow Jones Sustainability Index and find a large cap and value bias.
As already described, Fama and French (1993) show evidence of these risk factors (the size of a company and book-value-to-price ratio), to be powerful predictors of the market as a whole.
5.
Hypothesis (Research Question)
The research question of this thesis is:
What are the consequences, in terms of risk and return, of a more sustainable investment policy?
Does inclusion of sustainable criteria, into the stock selection process, create risk-adjusted excess return or does it come at a cost? This question relates to the fiduciairy duty of pension funds, which basically says that these funds have to find the highest return possible, within certain risk limits. A potential conflict between these two responsibilities has been the center of debate for pension funds.
Two equally weighted, well diversified (consisting of stocks of all sectors17), equity portfolios were created. These portfolios differed based on sustainability. To compare the results of both portfolios the Sharpe ratio will be used. The Sharpe ratio divides a portfolio’s return in excess of the risk-free rate by the portfolio’s standard deviation:
Sharpe ratio = (
R
p (t+1)–
R
f (t+1)) / σ
p (t+1)where
Rp (t+1) : is the realized return of portfolio p at time t+1
Rf (t+1) : is the risk free rate at time t+1
σp (t+1) : is the ex-post risk of the portfolio measured in term of its realized variance.
In order for a sustainable investment policy to be aligned with the fiduciairy duty of pension funds:
Sharpe Ratio (sustainable portfolio)≥ Sharpe Ratio (non-sustainable portfolio)
17
Because of the mounting evidence that sustainable portfolios are biased towards size- and value effects, several statistical tests were conducted to see whether or not this was true for this sample period.
It is my belief that sustainable stocks are potentially undervalued by financial markets. The reasons for this incorrect pricing are believed to be threefold: (i) the still ongoing debate centered on value creation of sustainability, (ii) the belief that equity market participants in general have a shorter horizon than the potential benefits of sustainability, and (iii) the ability by financial markets to factor in sustainability at all.
My expectation relating to the risk-return profile of sustainable portfolios is however neutral to negative. The reasons for this are, amongst others:
• There is an optimal level of sustainability, a U-shaped correlation. The database used, as most databases, use absolute measures of sustainability. Therefore it is likely to overstate the value of high ranking companies and understate the value of low ranking companies.
• Quantifying sustainability into one quantitative score seems to be almost impossible. • It would be highly unlikely that during this short sample period these stocks will
outperform, because investors suddenly are recognizing the undervaluation argument and act on it.
• If investors recognize the potential undervaluation and therefore sustainable funds outperform, this will only be temporarily. This abnormal return will exist only until the new equilibrium value is reached.
• The sample period is too short to draw any really meaningful conclusions in relation to this theory. The study is meant to give institutional investors, especially pension funds more insight in this field.
6.
Methodology
Environmental and social considerations can be introduced into the investment process in a number of different ways. A number of these approaches are summarized below:
• Ethical or negative screening, which avoids companies that undertake “unethical” activities, such as the manufacture of armaments, tobacco, gambling, etc
• Positive or best-in-class screening, which involves investing only in companies that lead their peer groups (i.e. a sector or industry).
• Constructive engagement, which involves communication between investors, and company management, whereby investors aim to encourage management to improve on their social and environmental performance.
• Shareholder Activism, which is a more confrontational form of engagement.
• Sustainable the investing, which is a thematic investing style based on identification of trends of sustainable development (renewable energy, education providers etc)
• Integrated analyses, which is an investment style in which analyses of environmental and social factors contribute to the overall analyses of a company by identifying additional sources of risks and opportunities.
Sustainable portfolios are derived from a universe using a selection process. This selection process can be very advanced, such as the “best-in-class” method. Also a more simple approach, such as the exclusion of companies or sectors can be used (negative screening). Importantly, each approach results in a different risk-return profile. While some of these (e.g. screening approaches or theme investing) have considerable impact on the number and diversity of stock available to the manager, others (e.g. engagement) have little to no impact on either the number or type of stocks that can be held. The most promising is integrated analyses. This issue of integrated sustainable and financial research has however always been in the background.
A representative study amongst pension funds in the Netherlands shows that, the majority of these funds use the best-in-class method (50% of respondents) to develop a more sustainable investment process (Hummels, 2007). Probably the main reason for this is the diversification possibilities of this relative approach of sustainability.
6.1. Best in Class Method
Companies are ranked within their respective sector/industry based upon a durability score. This method, known as best-in-class enables an investor to invest in a broad and diversified universe of stocks. Portfolio’s based on absolute sustainability tend to avoid certain sectors, like for example energy and mining companies. The risk of such a strategy is that it has the potential to increase the variance of the portfolio substantially. The best-in-class method does not exclude any sector. It invests in the most sustainable companies of each sector, whereas the least sustainable stocks are shorted.
The relative approach of the concept of best-in-class sustainability can best be explained by the following example. Companies in for example the energy or mining sectors are much more damaging for the environment then for example an internet or software company. This is called absolute sustainability. A sustainable portfolio, based on absolute values, would therefore avoid sectors like energy and mining. These portfolios tend to be skewed towards so called durable sectors, like for example renewable energy and education providers. Relative sustainability however ranks the companies within a certain sector. So a mining company is ranked in accordance with its mining peers and a software company is ranked within the software sector. A company that has a poor absolute environmental score can thus be a high scoring company on a relative basis.
7.
Dataset
The database obtained for this sustainability analysis is obtained from SNS Asset Management18 in the Netherlands. They analyzed a broad universe of publicly listed European companies (circa 300) based on its social- and environmental responsibility.
1. Environmental responsibility:
Notwithstanding the complexity of some of the science, a company role as an actor in the environment is relatively simple; it should aim to manage its resources efficiently and minimize pollution and waste that it causes. In so doing, it should have regard to the supply chain, production processes, product output and end-of-life disposal.
2. Social responsibility:
Assessing companies’ success in the societal dimension is more complex as it typically involves a wide range of different stakeholders, for each of which the determinants of success may be different. As a starting point, SRI analysts seek to understand whether companies themselves have a clear communicable idea of who their key stakeholder groups are, what constituents success and what constituents failure for each group, where they currently stand and how they plan to progress. Stakeholders to a firm are, amongst others:
Customers – Companies that do not understand their customers’ immediate needs are unlikely to remain in business for long. While most analysts keep a close eye on near-term product pipeline, SRI analysts will also look to the longer-term to understand how companies are preparing for long-term trends in customers expectations.
Suppliers – While price will clearly be the critical determinant of the relationship between a company and its suppliers, companies should have regard for factors as security of supply, reliability, quality, innovation, etc.
18
Employees – Whilst wages and conditions are the principal determinants, attracting, motivating and retaining good employees will also require considerations of wider factors such as training and development, clear
communication, aligning personal objectives with business ones, flexibility on work-life balance, etc.
Local communities – While failure to manage relationships with local
communities is manifested by demonstrations and campaigns, the best indicator of success may be silence. Transparency and engagement with local communities are tried and tested methods for building relationships with this stakeholder group.
Measuring social- and environmental responsibility:
The relevant social- and environmental issues of companies are grouped into so called thematic clusters:
Environment: Social:
• Environmental Strategy • Production process • Product / service use • Suppliers
• Human capital • Health & Safety • Audit & Reporting • External social policy • Strategy risk countries • Suppliers
Companies are grouped into sectors19. There are differences per sector of what is relevant and what not. For the energy sector, for example, environmental issues and health and safety of the employees are very important. These differences are reflected in different weightings of the thematic clusters comprising the overall environmental and social score of a company within its sector.
19 9 sectors are identified: Basic Materials, Communications, Consumer Cyclical, Consumer Non- Cyclical, Diversified, Energy, Financial, Industrial, Technology
To determine a rating for a company on these thematic clusters, so called indicators are used. A score for human capital, for example, is determined by several indicators: (i) career
opportunities and career development, (ii) training and education, (iii) equal rights for
employees, (iv) a transparent job classification system, (v) job satisfaction surveys and several others. An environmental score for the production process consists of, amongst others, the following indicators: (i) energy and water usage relative to output of products, (ii) active innovation of the production processes, (iii) environmental responsible waste disposal.
The environmental and social score of a company are determined by as well forward looking measures as backward looking measures. Not only is a company judged upon its own history, but also the actions by that company to improve their sustainable model are rewarded by this model. A backward looking measure is for example an accident at a plant, where a number of employees were injured. This influences the score of the cluster, health and safety, negatively. The actions a company undertakes to prevent this from happening again, a forward looking measure, also influences the score of the cluster. This can be both positive and negative.
The scores on the thematic clusters, together with the weight of the cluster within the overall score, determine the overall social and environmental score. The social and environmental score together determine the sustainability score, both weighting fifty percent. This score, ranging from zero to one hundred, compares the relative sustainability of a company to its respective sector peers. These rankings determine the best-in-class, and worst-in-class companies.
There are several channels, which are used, to obtain the necessary information. The first source of information comes from within the company. Examples of these are: annual reports, sustainability reports, company website, and interviews and questionnaires. The second source of information comes from outside the company, for example NGO’s, pressure groups and society organizations.
8.
Empirical Analyses
Two mutually exclusive equity portfolios were created. These portfolios differed based upon a sustainability score. The sustainability score compares the ranking of companies within their respective sector. The stocks are ranked semi-annually, which occurred at the end of June and the end of December of each year. The rebalancing of the portfolio occurred at these dates. The top twenty percent of companies in each sector make up the sustainable portfolio (High-Rank), whereas the bottom twenty percent of companies make-up the non-sustainable portfolio (Low-Rank). Each stock is equally weighted within the portfolio.
The sustainability database contains scores only from 2002-2006. However asset pricing requires sufficient data points. To obtain more meaningful results the ratings per end of December 2001 were extended backwards by two years. The sustainability scores tend to have a very low variability, therefore it is believed that extending the rankings backwards by two years is acceptable 20(see also Bauer et al.).
Graph 1, shows the performance of the portfolio of sustainable companies (High-Rank), the non-sustainable portfolio (Low Rank) and a portfolio of all companies (All), covered by SNS Asset Management. The portfolio consisting of all companies is used as a market proxy. The graph suggests that the β of the Low-Rank portfolio is smaller than the β of the High-Rank portfolio.
Graph: 1. Returns High, Low and All Stocks Portfolios
Returns High, Low and All Ranks
-60% -50% -40% -30% -20% -10% 0% 10% 20% 30% 40% 50% 2000_1 2000_2 2001_1 2001_2 2002_1 2002_2 2003_1 2003_2 2004_1 2004_2 2005_1 2005_2 2006_1 2006_2 Time Cum ul at ive R et ur ns
High Rank Low Rank All
20 I am aware that this procedure introduces a potentially look-ahead bias. The results using only real data are similar to those reported here.
Table 1: Descriptive statistics of the sustainable and non-sustainable portfolios, December 1999- December 2006
Portfolio Mean Std. Dev.
Sharpe
Ratio Maximum Return (semi-annually)
Minimum Return (semi-annually) High-Ranked 2.51% 15.54% -0.04 19.37% -18.49% Low-Ranked 5.47% 19.17% 0.12 18.98% -23.07% All-Stocks 5.67% 15.75% 0.16 17.20% -22.74%
Notes: The Sharpe ratio is the ratio of the mean excess return to the standard deviation. The mean return, the standard deviation and the Sharpe ratio are annualized.
These basic statistical results suggest that the risk-adjusted return, as measured by the Sharpe ratio, of the sustainable portfolio (-0.04) was inferior to that of the non-sustainable portfolio (0.12). The risk-adjusted return of the sustainable portfolio was even negative during the sample period. The “All Stocks Portfolio”, which was used as a proxy for the market (as in Bauer et al.), had the highest risk-adjusted return (0.16). The average yearly performance of the sustainable portfolio (2.51%) was below the average yearly performance of the non-sustainable portfolio (5.47%). The all stock portfolio had a yearly return of 5.67%.
The difference in returns (-0.87%), between the sustainable and non-sustainable portfolios, was not statistically significant21; after adjusting for market sensitivity and investment style (see appendix 1, table 3). Table 1 and 2 give the results of the sustainable and non-sustainable portfolios.
8.1. Portfolio Performance in a CAPM framework
To account for differentials in the portfolios’ market risk, the CAPM was used. A least-squares regression was used to estimate the following model:
Rit – Rft= αi+ βi(Rmt – Rft) + εit
where
Rit = the return on portfolio i (6 months).
Rft = the risk free rate (6 month Euribor).
Rmt =the return on a value-weighted portfolio (6 months).
21
εit = an error term
The value-weighted market proxy was provided by the Kenneth French Data Library22. Firms in the country portfolios are value weighted. Each country is weighted in proportion to its EAFE weight. The risk-free rate (Euribor, 6 month) is used from Bloomberg. The model beta (βi), measures the portfolios market-risk exposure. Jensen’s alpha (αi) represents the average
abnormal return in excess of the return of the market proxy. Table 2 shows the performance evaluation results obtained from the CAPM framework.
Table 2: Empirical Results of CAPM, from 12/1999- 12/2006
Portfolio α Rmt - Rft R 2 High-Ranked -1.91% 0.66 0.17 (-2.25) (10.28) Low-Ranked -0.70% 0.82 0.14 (-0.60) (9.34)
Note: the t-statistics are in parentheses.
The influence of sustainable screening on the investment performance is the difference between the alpha on the sustainable (High-Ranked) portfolio and the alpha on the non-sustainable portfolio (Low-Ranked), i.e. -1.21%23. As can be seen from the reported alpha estimates and the corresponding t-statistics, only the performance of the sustainable portfolio (High-Ranked) was significantly different from that of the market proxy. The Low-Ranked portfolio was more exposed to the market factor. The most important observation is that the difference of both portfolios is negative, which suggests that the risk-adjusted return of these screened portfolios is suboptimal.
8.2. Portfolio Performance in a Multifactor framework
Fama and French (1993) empirically establish the inefficiencies of the single-factor CAPM framework. They introduce a three-factor model that adds a capitalization based factor (small cap stocks minus large-cap stocks returns, SMB) and a BV/MV based factor (stock returns for companies with high BV/MV minus stock returns for companies with low BV/MV, HML).
22 Available at mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html 23
In this section the three-factor model of Fama and French is used to analyze the different portfolios. These additional control factors are particularly important because of the mounting evidence that SRI portfolios are biased to certain investment styles (size, value versus
growth); see for example Bauer et al., Luther and Matatko.
The performance assessment entailed estimating the following equation:
Rit – Rft= αi+ β0i(Rmt – Rft) + β1i SMBt + β2i HMLt+ εit
where
SMBt = return difference between a small-cap portfolio and a large-cap
portfolio in month t
HMLt = return difference between a value portfolio (high BV/MV) and a growth
portfolio (low BV/MV) in month t
HML data were obtained from the Kenneth French Data Library. Because SMB data for the European market was not available from this library, this factor was constructed using Bloomberg data. In relation to small-caps; the MSCI Europe Small Cap Index was used. In relation to large-caps; the MSCI Pan-Euro Index was used. These indices both include
publicly listed continental European and UK stocks. The next table (3) shows the performance evaluation results obtained from the multifactor regression.
Table 3: Multifactor Regression Results, from 12/1999- 12/2006
Portfolio α Rmt - Rft SMB HML R2 High-Ranked 2.76% 1.00 -0.68 -0.05 0.23 (2.21) (10.96) (-6.26) (-0.35) Low-Ranked 3.73% 1.15 -0.69 0.02 0.20 (2.16) (9.02) (-4.55) (0.11)
Table 3 shows the performance results from the estimation of the three factor model24. The increase of the R2 confirms the incremental explanatory power of the three factor model over the CAPM. The average factor adjusted return of the high ranking (sustainable) portfolio of 2.76%, was below that of the low-ranking (non-sustainable portfolio) of 3.73%. Both factors were statistically significant over the sample period. The additional determinant SMB was significant, whereas HML factor was not significant. For both the High-Ranked portfolio and the Low-Ranked portfolio, the coefficient on SMB is significant negative, which implies a bias toward large capitalization stocks. A significant value or growth bias was not found. As can be seen from table 3, the low-Ranked portfolio was more sensitive to the market (higher β). This can also be seen from graph 1; the low ranking portfolio underperformed the high-ranking portfolio when the market was going down, and vice versa.
Next the sustainability factor is added to the three factor model, to see whether or not this factor had any explanatory power during the sample period. The regression was performed on the portfolio containing all stocks. The statistical data suggests that the sustainability factor had no explanatory power at all during the sample period (table 4).
The performance assessment entailed estimating the following equation:
Rit – Rft= αi+ β0i(Rmt – Rft) + β1i SMBt + β2i HMLt+ β3i SUSTt+ εit
Where
SUSTt = Sustainability factor of a company in month t
24
Table 4, shows the performance evaluation results obtained from the multifactor (4) regression.
Table 4: Performance results from the estimation of the four factor model
SUMMARY OUTPUT Regression Statistics Multiple R 0,46 R Square 0,21 Adjusted R Square 0,21 Standard Error 0,20 Observations 2498 ANOVA df SS MS F Significance F Regression 4 26,83 6,71 166,13 0,00 Residual 2493 100,64 0,04 Total 2497 127,47 Coefficients Standard
Error t Stat P-value Lower 95% Upper 95%
Intercept 0,05 0,01 4,50 0,00 0,03 0,07 Rmt - Rft 1,10 0,05 22,65 0,00 1,00 1,19 SMBt -0,78 0,06 -13,56 0,00 -0,89 -0,66 HMLt -0,07 0,07 -0,95 0,34 -0,20 0,07 Sust. Score 0,00 0,00 -1,48 0,14 0,00 0,00 8.3. Portfolio Characteristics
Finally it was analyzed25 whether or not the portfolios (High-, and Low) differed on size and/or valuation (BV/MV), when compared to each other. Data on the
market-capitalization26 and BV/MV of the companies was obtained from Bloomberg. The results of these tests can be found in the appendix 4 (table 1 and 2). It was found that the sustainable portfolio had a statistically significant bias towards large-cap stocks over the non-sustainable portfolio. A difference between these portfolios in relation to BV/MV was not found. It is therefore concluded that, although both portfolios (High- and Low- Ranked) had a significant bias towards large-cap stocks (see multifactor model), this bias was even bigger regarding the sustainable portfolio (high-ranking portfolio).
25 Using statistical t-tests, assuming equal variances 26
9.
Conclusion
The general belief of many institutional investors is that the risk-return profile of a sustainable portfolio is suboptimal. This study confirms this belief.
The risk-adjusted return of the sustainable portfolio was below the risk-adjusted return of the non-sustainable portfolio. Using a more enhanced performance attribution model, to
overcome methodological concerns, it was shown that; the average factor adjusted return of the High Ranked (sustainable) portfolio, was below that of the Low-Ranked (non-sustainable portfolio). The sustainability factor seems to have no explanatory power at all, during the sample period. The hypothesis of proponents of sustainability that sustainable portfolios could potentially lead to higher risk-adjusted returns seems shaky, using this dataset. It was found that both the sustainable portfolios (High- and Low Ranked) were biased towards large capitalization stocks. Statistical tests suggest that the sustainable portfolio had an even bigger bias towards large-caps than the non-sustainable portfolio.
It is recommended that pension funds are very careful with respect to a more social
responsible investment policy. The integration of sustainable criteria into the stock selection process could potentially lead to a conflict with its fiduciairy duty.
References
Alexander, G.J., and Buchholz, R.A. (1978). Corporate Social Responsibility and Stock Market Performance. Academe of Management Journal, 21, 479-486
Bauer, Rob, Kees Koedijk, and Roger Otten (2005). “International Evidence on Ethical Mutual Fund Performance and Investment Style.” Journal of Banking and Finance.
Cohen, Mark A., Scott A. Fenn, and S. Konar (1997). “Environmental and Financial Perfromance: Are They Related?” Working Paper, Vanderbilt University.
Derwall, J., N. Guenster, R. Bauer, K. Koedijk (2005). The Eco-Efficiency Premium Puzzle. Financial Analyst Journal, Volume 61, 51-63
Diltz, David J. (1995). “Does Social Screening Affect portfolio performance?” Journal of Investing, vol. 4, no.1: 64-69
Fama, Eugene F., and Kenneth R. French (1993). Common Risk Factors in the Returns on Stocks and Bonds. Journal of Financial Econometrics, Vol. 33, no. 1 (January): 3-56
Garz, H., C. Volk, and M. Gilles (2002), More Gain than Pain – SRI: Sustainability Pays Off. Düsseldorf: WestLB Panmure
Griffin, Jennifer J., and John F. Mahone (1997). The Corporate Social Performance and Corporate Financial Performance Debate : Twenty-Five years of Incomparable Research. Business & Society, vol 36, no. 1 : 5-31
Guerard, J.B. (1997). Is there a cost to being socially responsible in investing? The Journal of Investing, 11-18.
