Is Socially Responsible Investment Values- or Profit-Driven?

Is Socially Responsible Investment Values- or Profit-Driven?

David Holwerda

University of Groningen, Faculty of Economics and Business, Department of Finance, Economics and Econometrics

__________________________________________________________________________________

Abstract

In this thesis we relate U.S. portfolio (excess) monthly returns over the period 2002-2013 to several Corporate Social Responsibility (CSR) dimensions. We construct 96 portfolios from a comprehensive sample of 945 U.S. companies by using CSR data from the Thomson’ Reuters Asset4 database. We find that Socially Responsible Investing (SRI) impacts risk-adjusted returns for the Environmental dimension by generating a significant negative abnormal return for best-in-class Environment portfolios. For Social and Corporate Governance dimensions, SRI impacts by having a lower size risk loading effect and not by generating superior returns. We identify the findings for the Social and Corporate Governance dimensions as profit-driven SRI, having a lower risk without the downside of a lower return. On the other hand, the findings for the Environmental dimension support the impact of values-driven SRI investors; investors who are willing to accept a suboptimal return for investing in companies with high Environmental scores.

Keywords: Socially Responsible Investing (SRI), Best-in-class performance, Financial performance, Corporate Social Responsibility

JEL codes: G01, G11, G12, M14

__________________________________________________________________________________

MSc. Thesis Finance David Holwerda, s1795392

1. Introduction

Nowadays, Socially Responsible Investing (SRI) is an important form of investing. The total U.S. assets under management using SRI strategies increased from $3.74 trillion at the beginning of 2012 to $6.57 trillion at the start of 2014, implying that SRI now accounts for one out of every six dollar under professional management in the U.S. (U.S. SIF, 2014). The idea of SRI is that shareholders should not only care about financial criteria such as cash flows and dividends, but also about how these cash flows are generated by taking into account nonfinancial aspects such as Environmental, Social, and Governance policies into their investment decisions (Dam, 2008; Renneboog et al. 2008). Is caring whether cash flows are generated responsibly however the primary reason for investors to engage in SRI?

What are the reasons for investors to engage in SRI? The article of Derwall et al. (2011) acknowledges the different reasons why investors might be interested in the field of SRI. The distinction between so called values- and profit-driven investors suggested by Derwall et al. (2011) is of interest because the effect of SRI may differ between these two groups. The first group, values-driven1 investors, choose their asset holdings regardless of the profit motive and the optimal risk and return as for instance suggested by Modern Portfolio Theory (Markowitz, 1952). These investors also take into account Corporate Social Responsibility (CSR) criteria into their investment decisions; where CSR refers to a firms’ consideration of, and response to, issues that go beyond narrow economic and legal requirements (Davis, 1973). The decision of values-driven investors to invest in certain companies is therefore grounded in social and personal values instead of sole financial considerations (Derwall et al., 2011). However, as the article of Sharfman (2008) argues, in general, high levels of CSR are associated with lower risk factors. CSR is therefore is not only attractive for values-driven investors but also for profit-driven investors. These profit-driven SRI investors do not use CSR for the same reason as values-driven investors but as a mere strategy to reduce overall portfolio risk (Kiymaz, 2012). The fact that many articles in the field of SRI are not aware or do not realize the impact of this distinction, is, according to Derwall et al. (2011), one of the main reasons why research papers have not yet reached consensus on the effect of CSR on the financial performance of companies. In this thesis we attempt to differentiate between values- and profit-driven investors by looking at risk-adjusted return of CSR portfolios. The main question throughout this thesis is: What is the effect of CSR on the financial performance of companies and how can we use potential differences in financial

1

3 performance to differentiate between values- and profit-driven SRI investors? By looking at the risk factors, we will try to provide quantitative evidence for the reason why investors engage in SRI: are they values- or profit-driven?

The effect CSR can potentially have on individual company and market returns is in an earlier stage identified by Fama and French (1997). According to Fama and French, preferences of investors, so called ‘tastes’, are the reason why returns differ from the expectation of traditional asset pricing models such as the CAPM model of Sharpe (1966). Fama and French identify diverse ‘tastes’ ranging from behavioral finance tastes, such as home bias investing and loyalty based investing, to CSR. Measuring the effect of this CSR ‘taste’ on individual companies and market returns is therefore of interest and is subject to growing attention from diverse papers. The article of Heinkel et al. (2001) for instance, states that exclusionary ethical investing leads polluting firms to be held by fewer investors and thus resulting in a higher cost of capital for these firms (Heinkel et al., 2001). The answer to our research question is therefore relevant for institutional and individual investors, both looking for the optimal allocation of assets and maximal performance of their portfolio. Besides, the answer to this question may also (partly) explain the tremendous growth of SRI. Moreover, empirical research has proved that investors who pursue values, i.e. nonfinancial goals, affect prices and returns differently compared to the profit maximizing investors (Derwall et al., 2011).

Throughout this thesis, the relationship between CSR and SRI is an important one to be aware of. Where CSR directly is connected to the daily practices of a company with regards to its Environmental, Social and Governance policies, SRI is focused on the investors’ decisions to value these ESG practices. SRI can take diverse forms to evaluate the performance of CSR in a company. The U.S. SIF foundation distinguishes five main forms: exclusionary, ESG integration, best-in-class performance, impact investing, and sustainability themed investing (USSIF, 2014). Each of these forms of SRI are narrowing the choice of available investment opportunities, and, by doing so, social responsible investors are implicitly willing to except a lower rate of return (Dam, 2008). The largest of these forms of SRI2 is exclusionary3, i.e. excluding certain controversial industries, such as the weapon industry and fur industry, from the investment universe. The main issue with the exclusion SRI policy is, however, what type of companies to exclude. There is a large consensus on excluding the truly controversial companies such as companies producing cluster munition. There is, however,

2

According to the U.S. SIF Report 2014

more debate on excluding for instance meat and adult entertainment companies4. In the U.S., 95% of the SRI funds use the exclusion SRI form when taking investment decisions (Renneboog, 2011).

This thesis focuses on the financial performance, as measured by stock market returns, of the best-in-class performing U.S. companies with regards to the three main CSR dimensions5 and compares this performance with that of their worst-in-class performing peers. Moreover, we will analyze this performance by looking at four common risk factors. By using the best-in-class SRI method, we refrain from the decisions on what companies to exclude. The best-in-class policy does not lead to an exclusion of all companies belonging to controversial business area, but rates all companies based on a set of criteria in the field of Environment, Social and Corporate Governance. This particular investment strategy is for this reason also embraced by several papers on the effect of CSR on financial performance (Galema et al., 2008; Lam et al., 2012; Kempf and Osthoff, 2007). The intrinsic values of SRI investors translate into certain standards with regards to the three CSR dimensions used in this thesis. The level of these standards can potentially differ per SRI investor. In this thesis we define high scoring companies as the top 20% of the total number of companies publishing data on the level of these three CSR dimensions. For reasons discussed above, this high level of CSR not only attracts values-driven investors but also the profit-driven investors. The latter investors are however not willing to sacrifice return for the “good cause” (Derwall et al., 2011). Following this line of reasoning, the differences between high and low CSR portfolios for profit-driven investors will only be visible in the risk factors. For values-driven investors, this effect will not only be visible in the risk factors but also in a significantly lower alpha; representing the loss in return for not investing in low CSR companies as suggested by Dam (2008). We will test the differences between the financial performance of the high and low CSR companies by taking a so called long-short position: going long in the worst-performing companies and taking a short position in the best-performing companies’ portfolios. The process of constructing the portfolios and the methodology for measuring these differences are discussed in respectively section three and four of this thesis.

The aim of this thesis is to provide a manner to mathematically differentiate between values- and profit-driven investors. We construct 96 U.S. company portfolios, 48 best-in-class performing portfolios and 48 worst-in-class performing portfolios, based on Thomson Reuters’Asset4 data over the period 2002 to 2013. The total sample of companies that are listed on either the NYSE, AMEX or NASDAQ index and are reported by the Asset4 database increased from 335 in 2002 to 945 in 2013.

4

For debates on the issue of exclusion see for instance the article Renneboog et al. (2008)

5

5 We find that best-in-class SRI impacts risk-adjusted returns for the Environmental dimension by generating a significant negative alpha for best-in-class Environment portfolios. For the Social and Corporate Governance dimensions, SRI impacts by having a lower size risk loading effect and not by generating alpha’s. We identify the findings for the Social and Corporate Governance dimensions as profit-driven SRI investors, having a lower risk without having the downside of a lower return. On the other hand, the findings for the Environmental dimension support the impact of values-driven SRI investors; investors who are willing to accept a suboptimal return for not investing in certain companies.

This is not the first paper to investigate the financial performance of the best-in-class SRI strategy. The use of the relatively new Asset4 database and using this database for creating top and bottom portfolios for each different CSR dimensions as a research method, however, is different from the more commonly used KLD database; embraced by the articles of for instance Galema et al. (2008) and Kempf and Osthof (2007). Moreover, the use of these outcomes to identify values- and profit-driven investors is, to our knowledge, not attempted before by other articles.

The remainder of the thesis is structured as follows: section two provides a background on the relevant literature and from there on states the hypothesis. Section three focuses on the data and research methods used in this thesis. The methodology used to test our hypothesis is described in the fourth section. The fifth section presents our results. The last section concludes.

2. Background

In this section we review the relevant literature that provides both theoretical and empirical evidence for stating our hypothesis. We will first discuss the various reasons for companies to engage in CSR. Moreover, the two main financial performance measurements for the effect of CSR and their theoretical expectations will be reviewed. Furthermore, the empirical evidence on the financial performance of the two main forms of SRI, exclusion and best-in-class performance, will be discussed. We will combine all this information for the purpose of stating our main hypothesis on the effect of CSR on the financial performance of our portfolios.

2.1 Why do companies engage in CSR?

or should not engage in CSR activities is one of the issues on which researchers have not reached consensus. In 1970 for instance, Friedman wrote an article in the New York Times stating that “The Social Responsibility of Business is to increase its Profits” (Friedman, 1970). According to Friedman, it is the responsibility of the firm to “conduct the business in accordance with the their desires (of the owners), which generally will be to make as much money as possible while conforming to their basic rules of the society, both those embodied in law and those embodied in ethical custom” (Friedman, 1970).

However, Bénabou and Tirole (2010) provide three visions on why a firm may engage in CSR activities: Win-win, delegated philanthropy and insider-initiated philanthropy. The first view, win-win, can be described as ‘doing well by doing good’ (Bénabou and Tirole, 2010). The second vision relates to the firm as a channel for expression of citizen values (Bénabou and Tirole, 2010). The article of Dam (2008) describes this incentive for engaging in CSR as vertical product differentiation, where having a strong CSR policy attracts new customers. The third vision is not motivated by stakeholders demands but reflects managements’ or the board members’ own desires to engage in philanthropy; identified as intrinsic motivation by Dam (2008). While the latter two visions are either harder to measure or occur less frequently, the first vision, doing well by doing good, has been investigated by numerous papers. From an instrumental standpoint, engaging in a CSR strategy is a form of investment, entailing initial costs for future financial benefits (Branco and Rodrigues, 2006). It may be that the impact on the long run future cash flows is positive, but short run cash flows are adversely affected.

7 that CSR-related shareholders proposals lead to superior financial performance. We use this outcome as support for our basic assumption that CSR is more likely the driver of (abnormal) financial performance than the result of it (Flammer, 2013). We shall first elaborate on the measurements of financial performance and the differences between these measurements.

2.2 Measurements of financial performance

The article of Dam and Scholtens (2015) focuses on one of the reasons why the empirical evidence between CSR and financial performance is mixed. According to Dam and Scholtens (2015), the reason for this is the fact that there are differences in behavior of the several financial performance measures used in studies. Measurements of financial performance can be divided in two main categories: accounting measures and the (stock) market returns method.

2.2.1 Accounting methods

Although accounting methods are used frequently, Gregory et al. (2013) are critical on the use of such methods. Gregory et al. (2013) argue that whilst such measures have their use, they are backward looking and their objectivity and informational value can be questioned, referring to the work of Benson (1982) on accounting and corporate accountability.

What can we expect from a theoretical point of view from accounting measures? To explore the effect of CSR on financial performance, Dam and Scholtens (2015) identify two types of corporate behavior: market value maximization and pure profit maximization. The latter can be seen as irresponsible behavior where the maximizing the market value can be seen as socially responsible behavior. The Market-to-book ratio, defined as total market value divided by installed capital, will for socially responsible companies always be higher than that of irresponsible companies. The reason for this is that socially responsible companies are priced above conventional or irresponsible firms for whose stock there is less demand (Dam and Scholtens, 2015).

Following the same line of reasoning, the Return on Assets (ROA), defined as profits divided by installed capital, should also be higher for high CSR firms, a finding that is rooted in decreasing marginal returns to production (Dam and Scholtens, 2015).

2.2.2 Market returns

non-changing level of CSR, return data would conclude that CSR does not affect financial performance (Gregory et al., 2013). Dam and Scholtens (2015) conclude that “whether the risk-adjusted stock market returns are higher for socially responsible firms or irresponsible firms is ambiguous”, depending on the damage per output. The effect for companies within the same industry, that is the same level of damage per output, the high scoring CSR companies would expect to have lower stock returns in comparison with their low scoring peers. On the aggregate level, i.e. not corrected for industry type, it is hard to predict whether stock market returns are higher or lower for high scoring CSR firms (Dam and Scholtens, 2015). Gregory and Whitaker (2013) claim that ESG information should not affect returns, as efficient markets would fully incorporate the publicly available and generally stable ESG information into stock prices. However, the article of Boutin-Dufresne and Savaria (2004) do not agree with this finding. Because of the fact investors’ expectations are not fully adapted to all responsibility-related information and its (potential) effect on the expected future cash flows, responsible investments are expected to have higher risk-adjusted returns (Boutin-Dufresne and Savaria, 2004). This finding of higher expected returns is supported by Galema et al. (2008) although be it for irresponsible companies. Galema et al. (2008) argue that the shortage of demand for irresponsible firms’ stock implies that the risk sharing opportunities for people investing in these stocks are limited and therefore command a return premium. Angel and Rivoli (1997) predict socially controversial stocks that are shunned by investors have a higher expected return, and the expected return increases with the proportion of socially responsible investors in the market. Heinkel et al. (2001) relate to this research by stating that the stock boycott by the “green”, i.e. eco-conscious, investors limits the risk-opportunities of those invested in environmentally controversial firms. As a result, shareholders of controversial companies receive a higher compensation for holding these stocks than they would if the markets were free of boycotts (Heinkel et al., 2001).

2.3 Empirical findings

9 CSR stock versus the market (Margolis et al., 2009). The effect is however very modest and should also according to Margolis et al. be handled with care. The problem with meta analyses is that they do not always take into account the contradicting effect diverse financial performance methods can have, as described in the previous section.

We will focus on the two forms of SRI for which most empirical evidence is available: exclusion and best-in-class performance. Because the latter is similar to the research method of this thesis, we will use this empirical evidence for stating our hypothesis. Important to note is that most articles use a different database than this thesis, namely the KLD database. Other than the Asset4 database, the KLD database identifies seven CSR characteristics: Corporate Governance, Community, Diversity, Employee Relations, Environmental, Human Rights and Products. These dimensions cannot conveniently be matched with our three Asset4 CSR dimensions. We will therefore only directly compare the Environment and Corporate Governance dimensions of KLD with that of Asset4 and will cluster the other five dimension of KLD with that of the Social dimension of Asset4. For all dimensions these comparisons should be handled with care since none of the dimensions of both databases exactly measures the same.

2.3.1 Exclusion / negative screening of sin stocks

The financial performance of certain so called ‘sin’ stocks is investigated by many articles. Hong and Kacperczyk (2009) are more specific about what kind of stock should be shunned. Their expectations are that sin stocks, such as tobacco, alcohol and gambling, are in conflict with societal norms and therefore are shunned by institutional investors vulnerable to public opinion, such as pension funds. They predict that by not buying sin stocks, these norm-constrained investors cause sin stocks to be cheaper than is suggested by their fundamental value, and to have higher expected returns (Hong and Kacperczyk, 2009). The conducted research confirmed their hypothesis; U.S. sin stock in the period of 1965 - 2004 significantly outperformed their comparable counterparts by 4.5% per year. Fabozzi et al. (2008) relate to this finding, using a diverse international sample of 267 sin companies over the period 1970-2007. Risk adjusted returns using the CAPM Model for these sin stocks are 13.70% higher than the market returns.

alcohol, gambling, controversial weapons, and adult entertainment, the 0,7% monthly alpha being significant at the 1% level (Trinks and Scholtens, 2015).

Looking at the empirical results of the three articles above, we can conclude that the exclusion SRI strategy results in an outperformance of shunned stocks. Using the exclusion strategy therefore requires a sacrifice in the form of a lower risk adjusted return, a sacrifice only values-driven or institutions vulnerable for public opinion are willing to make.

2.3.2 Best-in-class performance

We start with three important articles that use the KLD database, an U.S. sample and return data. Kempf and Osthoff (2007) use KLD data for constructing top 10% and bottom 10% U.S. portfolios on diverse CSR characteristics defined by the KLD database in the period of 1992 - 2007. Although the KLD database uses a different measurement method than Asset4, relevant results arise. When taking a long-short position, i.e. a long position in the high ranking portfolios and a short position in their low ranking counterparts, top 10% companies have a significant outperformance of 3,73% for the Employee relations dimension and 4,52% for the Community involvements dimension. We relate these factors to the Social dimension of Asset4. Galema et al. (2008) use the same KLD database to construct their top 20% and bottom 20% portfolios of U.S companies over the period of 1992 – 2006. They find less convincing results, only the top Community outperform there bottom counterpart, although this result is only significant at the 10% level. The differences may be partly due to the different research method conducted by Galema et al. (2008); the GMM method. The main conclusion is that ESG performance is priced in the book-to-market ratios and therefore there are no significant alpha’s (Galema et al., 2008). When using the Carhart (1997) model, the effect of this lower book-to-market ratio, implying that it is a growth stock, will become visible in the HML risk factor.

11 Derwall et al. (2005) embrace a different database, the eco efficiency data from Innovest. They report that a best-in-class portfolio containing the top 30% of U.S. stocks with the highest eco-efficiency scores relative to industry peers deliver a four-factor alpha of 3.98% per year over the period 1995-2003. In contrast, a portfolio consisting of firms with the lowest scores produces a negative but non-significant alpha of -1.08%. Taking a long-short position, results in a 5.06% outperformance for the top portfolio, a result that is significant at the 10% level.

Although Gregory et al. (2013) use a different research method6 than the above mentioned top and bottom portfolios, findings of this article are interesting for our thesis for reason Gregory et al. (2013) perform comprehensive research on the Environmental dimension of the KLD database. Instead of taking the top/bottom 20%, Gregory et al. (2013) identify four dimensions of Environmental firms: green, grey, neutral and Toxic. When looking at the four factor performance of a long-short position, i.e. long in the green companies and short in the toxic firms, no significant alpha is found. Significant differences can be found in the market risk factor, green companies having a lower exposure to this risk factor. Notable differences can also be found in a lower SMB risk factor for toxic firms and a higher HML risk factor for toxic firms (Gregory et al., 2013).

The theoretical expectations for return data are in line with the empirical findings for the exclusion SRI strategy. Most articles on the performance of so called ‘sin’ stocks find that these stocks outperform their counterparts (Hong and Kacperczyk, 2009; Trinks and Scholtens, 2015). However, some empirical results for the best-in-class performance contradict the theoretical expectations, high CSR companies having a higher return than their lower scoring counterparts (Derwall et al., 2005; Kempf and Osthoff, 2007). However, most articles on best-in-class SRI find no (significant) outperformance for most of the CSR dimensions (Galema et al., 2008; Statman and Glushkov, 2009). Nevertheless, we will focus on the theoretical expectations for market returns studies: we therefore state the hypothesis that low scoring CSR companies outperform their high scoring CSR counterparts (Dam, 2008; Galema et al., 2008; Angel and Rivoli, 1997). We will relate this outperformance to values-driven SRI investors; the negative alpha representing their loss in return for not investing in low CSR companies. With respect to the risk factors, we state that CSR portfolios have lower risk loadings (Gregory et al., 2013, Galema et al., 2008). In case of differences only visible in lower risk factor loadings, and thus not in a lower alpha, we will label them as profit-driven SRI investors: investors using CSR companies for lowering the risk of their portfolio without the downside of having a lower return. In this case our hypothesis of values-driven SRI investors will be rejected.

6

3. Data

This section describes the process of obtaining the necessary data for testing the hypothesis stated in the previous section. First, we describe the database used for obtaining the ESG data. Next, we describe the process of selecting the data. Lastly, we elaborate on the steps taken for constructing the portfolios.

3.1 The Asset4 database

This thesis uses the Asset4 database of Thomson Reuters to obtain the ESG scores of the U.S. companies. The Asset4 database is a relatively new database and measures the level of CSR on three dimensions: Social, Corporate Governance, and Environmental. Besides these dimensions, an Economic dimension is also part of the database, containing the level of several economic indicators of the company7. This Economic dimension however, is not taken into account in this thesis. On basis of the scores on the different dimensions, a company receives a rank8. The scores of the dimensions are, according to Thomson Reuters, updated on a weekly basis. The outcome of the scores, however, shows that they appear only to be updated on a yearly basis. A common problem using any CSR database is the (likely) delay in both company CSR practices and actual CSR reports. For this thesis, however, we do not regard this as an issue since new portfolios are being composited every year and it is not likely that a company will go from the bottom 20% to the top 20% in a short time span. This also solves the issue of non-changing or barely changing CSR scores raised by Gregory et al. (2013).

3.2 Data Selection

Our data consists of all companies listed on the NYSE, AMEX or NASDAQ index and for which Thomson Reuters publishes Asset4 data in the period of 2002-2013. This period is the longest period Asset4 data is available. The growing interest in CSR can clearly be seen in the number of companies for which Asset4 data is available: from 335 in 2002 to 945 in 2013. The companies for which no Asset4 data was available in the year of interest, have been deleted from the sample we used for constructing portfolios in that year. To prevent possible biases, such as survivor bias, companies having more than two consecutive zero return or zero market value observations have been deleted from the portfolios and were replaced by the next best- or worst-in-class performing company from the whole sample.

7

Economic performance is generally not regarded as a CSR factor and besides, measuring the effect of economic performance on financial performance does not make sense.

13 3.3 Constructing the portfolios

For constructing the portfolios, this thesis uses the so called best- and worst-in-class performance measure. This measure has the advantage that it does not suffer from any interpretation- or biased opinion issues for selecting the companies forming the portfolios. Individual portfolios have been constructed per dimension. For each dimension, the top and bottom 20% form two portfolios, representing the best- and worst-in-class performing companies per dimension based on their rank provided by the Asset4 database of that year. In total, 96 portfolios over the period of 2002 until 2013 have been constructed; 48 best-in- and 48 worst-in-class performing portfolios, equally divided over the three ESG dimensions and the Asset4 Aggregate score dimensions. The list of companies resulting from this selection have been matched with their end-of-the-month Return Index data9 and Market Value10 _{data, both obtained from DataStream.}

4. Methodology

This section describes the methodology that we use to test our hypothesis. We first describe the regression model used to adjust our monthly returns for risk. Furthermore, we elaborate on the methodology of both the Sharpe performance ratio and the Sortino ratio. In the last part, we describe the process of obtaining Fama-French factor data, the market portfolio, risk-free rate and distinguishing between values- and profit-driven investors.

4.1 Carhart Four-Factor model

To adjust our portfolio return data for risk, this thesis uses the Carhart Four-Factor model (Carhart, 1997). The Carhart Four Factor model is an adaptation of the more known Fama-French three factor model (Fama and French, 1993). Next to the three risk factors of the Fama-French model, the Carhart model adds the Momentum factor to the model. This momentum factor is designed to capture the risk due to the momentum found in the stock return by the article of Jegadeesh and Titman (1993). The Carhart Four -Factor model is defined as:

𝑅𝑖,𝑡− 𝑅𝑓,𝑡 = 𝛼 + 𝛽𝑀𝐾𝑇(𝑅𝑚,𝑡− 𝑅𝑓,𝑡) + 𝛽𝑆𝑀𝐵𝑆𝑀𝐵𝑡+ 𝛽𝐻𝑀𝐿𝐻𝑀𝐿𝑡+ 𝛽𝑀𝑂𝑀𝑀𝑂𝑀𝑡+ 𝜀𝑖,𝑡 (1)

According to the Carhart model, the excess return on a particular security or portfolio can be measured by four risk factors. The first factor, 𝑅𝑚,𝑡− 𝑅𝑓,𝑡 , the market risk premium, is represented

by the excess return on a value-weighted market portfolio. The second factor, 𝑆𝑀𝐵𝑡 is defined as the

9

The return of the stocks is measured as the Total Return index on t=0 divided by the index on t=-1. The Total Return index is created by DataStream to proxy a stock’s theoretical growth in value, making the assumption that dividends are reinvested.

difference in monthly return between a portfolio of small market capitalization stocks and a portfolio with large market capitalization stocks. The 𝐻𝑀𝐿𝑡 factor is defined as the difference in monthly

return between a high book-to-market portfolio minus a low book-to-market portfolio. The Momentum factor 𝑀𝑂𝑀𝑡 is defined as the difference in monthly return between a portfolio of

successful securities minus a portfolio of unsuccessful securities. The alpha 𝛼 represents the variable of interest for this thesis, namely the abnormal performance of the security or portfolio, which cannot be explained by the four mentioned risk factors. Bauer (2005) reports that socially responsible mutual funds differ from their conventional peers with respect to the loadings of these factors.

4.2 Sharpe and Sortino Ratio

The Sharpe Ratio (Sharpe, 1966) is a widely used measure for calculating risk-adjusted returns. The Sharpe ratio can be defined as the average return in excess of the risk-free rate per unit of volatility of risk. A higher Sharpe ratio is positive, implying that the chance is higher that the portfolio or security return will exceed that of the risk-free rate. The Sharpe ratio is defined as followed:

𝑆𝑝= 𝐸(𝑅𝑝_𝜎)−𝑅𝑓

𝑝 (2)

For the reason that investors are mainly interested in downside risk, Sortino and Van Der Meer (1991) proposed the Sortino ratio. In contrast to the Sharpe ratio, the Sortino ratio measures the downside risk of a security or portfolio and favors investments with the highest return for a given level of downside risk. The Sortino ratio is defined as followed:

𝑆𝑜𝑟𝑡_𝑝=𝐸(𝑅𝑝)−𝑅𝑀𝐴𝑅

𝛿𝑝 (3)

Where 𝛿𝑝= √_𝑇1

∑

𝜄

− (𝑅

− 𝑅

)

4.3 WRDS Fama-French Factor and Risk-Free Rate Data

15 French U.S. factors. The returns used in this thesis are value weighted. The use of market-value weighted returns, i.e. the price times the number of shares outstanding, gives a more realistic view of the returns of the portfolio because these are most feasible for institutional investors, which make up for the largest part of the SRI market (Trinks and Scholtens, 2015). Besides, the market portfolio used by Carhart (1997) is also market-value weighted. The use of market-value weighted portfolios is also embraced by the articles of Statman and Glushkov (2009) and Galema et al. (2008). Results of the equally-weighted portfolios for this thesis are used as a robustness check11.

4.5 Distinguishing between values- and profit-driven SRI investors

Within this thesis we embrace various investment strategies to grasp the effect of CSR on the financial performance. We first use the 96 best and worst performing company portfolios and take a long position in each of them. Although these portfolios will give a clear first image of the effect of CSR, we need a different investment strategy for testing the differences between the best and worst performing companies. We test these differences by taking a so called long-short position; going long in the worst-performing companies portfolios and taking a short position in the best-performing companies portfolios. By doing so, we test the differences on all the four risk factors and on the variable of interest: the alpha. The long-short method is common for testing differences between portfolios and is for instance also used by Galema et al. (2008), Kempf and Osthoff (2007). This translates in the following regression based on the Carhart (1997) model:

𝑅_{𝐵𝑜𝑡𝑡𝑜𝑚 𝑝𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜𝑠}− 𝑅_{𝑇𝑜𝑝 𝑝𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜𝑠}− 𝑅_𝑓,𝑡 = (4)

𝛼 + 𝛽_𝑀𝐾𝑇(𝑅_𝑚,𝑡− 𝑅_𝑓,𝑡) + 𝛽_𝑆𝑀𝐵𝑆𝑀𝐵_𝑡+ 𝛽_𝐻𝑀𝐿𝐻𝑀𝐿_𝑡+ 𝛽_𝑀𝑂𝑀𝑀𝑂𝑀_𝑡+ 𝜀_𝑖,𝑡

5. Results

This section presents our results. We start with the descriptive statistics and risk/performance measurements of our sample. Furthermore, we will discuss the risk-adjusted performance of both the top and bottom portfolios. Moreover, the results of the long - short portfolios of all ESG dimensions will be discussed. Thereafter, we discuss the possible arbitrage opportunities and alternative interpretations of the results. Lastly, we consider two robustness checks, the exclusion of the Great Recession and the use of equally-weighted portfolios.

5.1 Descriptive Statistics

Table 1 displays the descriptive statistics of the raw monthly returns for our various constructed portfolios by ESG dimension.

Table 1: Descriptive statistics and risk/performance-related measures of best- and worst-in-class portfolios monthly returns of U.S. firms by ESG dimension

Mean Median StDev Variance Min Max Kurt Skew Sharpe Sortino

Best-in-class portfolios Aggregate 0.97% 1.43% 4.02% 0.16% -14,0% 10.6% 1.306 -0.570 0.162 0.026 Environmental 0.49% 1.68% 6.36% 0.40% -28.3% 17.7% 6.161 -1.843 0.058 0.052 Social 1.04% 1.74% 4.13% 0.17% -14.8% 10.8% 1.455 -0.668 0.221 0.027 Corporate Governance 0.99% 1.60% 4.12% 0.17% -13.6% 11.0% 0.880 -0.572 0.209 0.027 Worst-in-class portfolios Aggregate 1.33% 1.75% 5.80% 0.34% -22.4% 21.3% 3.033 -0.349 0.209 0.037 Environmental 1.33% 1.68% 4.41% 0.19% -16.4% 14.4% 1.924 -0.401 0.274 0.034 Social 1.41% 1.64% 4.72% 0.22% -17.2% 16.8% 1.739 -0.383 0.272 0.029 Corporate Governance 1.33% 1.45% 4.82% 0.23% -21.3% 16.0% 3.223 -0.583 0.250 0.030

Descriptive statistics per market value-weighted portfolio. Best (worst)-in-class portfolios represent the top (bottom) 20% companies per CSR dimension in the period of 2002 – 2013.

17 support for our hypothesis, they may simply reflect a compensation for additional risk. In order to identify whether the differences in returns are value- or profit-driven, we need to investigate whether the worst-in-class portfolios have a larger loading on common identified risk-factors. If not, the difference in returns will be reflected in the alpha, supporting the values-driven hypothesis.

5.2 Risk-Adjusted Returns of Individual Portfolios

Panel B of table 2 displays our results for the separate Asset4, Environment, Social, and Corporate Governance portfolios. All portfolios, except top Asset4 (0,3%) and top Environment (-0,2%), show significant monthly outperformance. Thus, the stocks of the Asset4 universe performed better than our benchmark, the value-weighted returns of the market portfolio. Because almost all portfolios performed better, both top and bottom, we will not put much emphasis on the alpha’s of these portfolios and focus primarily on the alpha’s of the long-short strategies.

Table 2: Fund performance of U.S. companies portfolios selected on their CSR scores in the period of 2002 – 2013 using various investment strategies

α RM-RF SMB HML MOM Adjusted 𝑹𝟐

Panel A: Long – Short portfolios

Bottom minus Top Asset4 Aggregate scores 0.004

(1.61) 0.05 (0.92) 0.66*** (6.51) -0.17* (-1.82) -0.26*** (-5.33) 0.404

Bottom minus Top Environmental 0.009***

(2.69) -0.31*** (-3.62) 0.31** (2.04) 0.08 (0.58) -0.03 (-0.48) 0.072

Bottom minus Top Social 0.002

(1.43) 0.01 (0.12) 0.42*** (5.67) -0.05 (-0.68) -0.05 (-1.49) 0.212

Bottom minus Top Corporate Governance 0.002

(1.35) 0.07 (1.41) 0.23*** (2.83) -0.01 (-0.11) -0.07* (-1.77) 0.123

Panel B: Top and Bottom portfolios

Asset4 Aggregate Top 0.003

(2.62) 0.89*** (33.78) -0.25*** (-5.52) -0.01 (-0.13) 0.01 (0.24) 0.913

Asset4 Aggregate Bottom 0.006***

(2.86) 0.94*** (16.38) 0.41*** (4.22) -0.18* (-1.94) -0.25*** (-5.41) 0.808 Environmental Top -0.002 (-0.72) 1.21*** (15.03) -0.19 (-1.36) -0.00 (-0.02) -0.03 (-0.47) 0.687 Environmental Bottom 0.007*** (1.61) 0.89*** (34.14) 0.12*** (2.68) 0.08* (1.89) -0.07*** (-3.09) 0.932 Social Top 0.005*** (5.55) 0.92*** (39.34) -0.23*** (-5.78) 0.03 (0.72) -0.02 (-1.19) 0.937 Social Bottom 0.007*** (5.72) 0.93*** (27.58) 0.19*** (3.33) -0.02 (-0.38) -0.08*** (-2.77) 0.901

Corporate Governance Top 0.004***

(4.35) 0.93*** (36.62) -0.17*** (-3.81) -0.05 (-1.14) 0.02 (0.97) 0.925

Corporate Governance Bottom 0.007***

(5.20) 1.00*** (29.84) 0.06 (1.06) -0.06 (-1.03) -0.05* (-1.74) 0.906

5.3 Risk-Adjusted returns of Long-Short portfolios

Panel A displays the monthly market excess returns, SMB, HML, MOM, and alpha for the long-short portfolios. We will discuss the results per dimension below.

5.3.1 Asset4 Aggregate dimension

Aggregate results of CSR performance are problematic because combining the different effects might give a biased view on the effect of CSR (Galema et al., 2008). We will therefore only briefly discuss the outcomes of this long-short portfolio and not draw conclusions from these results. The long-short position in Asset4 Aggregate portfolios reveals a significantly lower SMB risk factor for the top 20% portfolios. This suggests a reduction in systematic risk attributable to the size risk factor. Besides this factor, the Momentum risk factor also reveals significant results implying that the bottom portfolios have a higher exposure to momentum risk.

5.3.2 Environment dimension

As shown by the descriptive statistics of table 1, the Environmental dimension exhibits the largest differences with regards to mean returns and Sharpe ratios. The risk-adjusted returns of the long-short portfolios for this dimension are therefore also interesting to analyze. When taking a long position in bottom and a short position in top Environmental portfolios, we observe a significant alpha of 0,9% per month. This significant alpha represent an outperformance of bottom portfolios in comparison to their top peers after having adjusted the returns for risk. This finding is empirical evidence in favor of the values-driven SRI investors hypothesis: by investing in top scoring Environmental companies, an investor is actually sacrificing risk adjusted returns. These findings contradict the hypothesis of Porter and van der Linde (1995) who predict that strong environmental management results in superior resource efficiency that has the potential to outweigh the associated costs. Furthermore, we observe the effect on systematic risk by looking at the risk factors. For the first risk factor, the excess return with regard to the market, we observe a significant lower loading for bottom companies. Significant results are also present for the SMB risk factor, only this time the bottom companies have a significant higher exposure to this risk factor. High environmental portfolios have a lower systematic risk exposure than their low peers, implying a risk reduction.

5.3.3 Social dimension

19 performance (Lam et al., 2012). These results fuel the expectations of profit-driven SRI investors. Investing in top Social companies provides these investors with a significant lower exposure to the SMB risk factor while not having to sacrifice return for this risk reduction. Not having a significant alpha contradicts the expectations of Chung et al. (2003), who expect a positive relationship between social performance and subsequent financial performance.

5.3.4 Corporate Governance dimension

The results of the long-short position in top/bottom Corporate Governance companies are in line with that of the Social dimension. Besides the significant lower SMB risk factor, we also observe a lower Momentum risk factor. This result however should not be valued too much since it is only significant at the 10% level. Investments in top Corporate Governance companies might therefore potentially also be attractive for profit-seeking investors.

In general, we find significant lower SMB risk factor loadings for the top portfolios. A lower sign of this risk factor indicates that investors face a lower systematic risk with regards to size. This finding is partly in line with that of Galema et al. (2008), who also find lower SMB risk factor loadings for diversity strengths, employee strengths and governance strengths. However, we do not observe the significant lower HML risk factor loadings their article finds. All dimensions, except the Environmental dimension, seem attractive for profit-driven SRI investors, having lower risk loadings without sacrifice of risk-adjusted return. This evidence does not support our hypothesis of values-driven SRI investors. For the Environmental dimension however, we do observe a significant negative alpha for top portfolios. Investing in best-class performing Environmental companies therefore demands a sacrifice in the form of a lower risk-adjusted return, a finding supporting our hypothesis that SRI is values-driven. What is also worth mentioning is the fact that the Environment dimension is the only dimension that displays significant differences between the market risk factor of the top and bottom portfolios, the latter being lower. This lower exposure to market risk should be visible in the performance during a financial crisis; an expectation we will elaborate more on in our robustness checks section12.

Using the definitions of both values- and profit-driven investors stated in the introduction, we conclude that SRI investors in top Social and Corporate Governance dimensions companies are profit-driven and investors in top Environmental companies are values-profit-driven. These results should however be handled with care. As described by the article of Derwall et al. (2011), the SRI community

is diverse, consisting of both profit-driven and values-driven investors. The quantifiable research method used in this thesis does not allow us to assess potential other intrinsic motives for SRI besides the effect on risk factor loadings and alpha’s. It is not likely that only profit-seeking investors invest in top Social and Corporate Governance scoring companies. Other motivations for investors to invest in the Social and Corporate Governance dimensions, such as material incentives and social- and self-esteem concerns identified by Bénabou and Tirole (2010), may arise in for instance a questionnaire.

5.4 Arbitrage opportunity?

The result of a significant alpha for the long-short position in the environmental dimension can at first sight easily be mistaken for an arbitrage opportunity. Taking a long position in bottom environment scoring companies and a short position in their top performing counterparts yield a positive alpha while it is in fact a zero sum investment. As discussed in the introduction, the SRI movement is rapidly growing (U.S. SIF, 2014). An increasing number of investors do take into account CSR performance of companies in the process of constructing their portfolios. Therefore, not all investors will be willing to invest in low scoring CSR companies. However, both green and non-green investors are willing to invest in high scoring CSR companies. The article of Dam and Scholtens (2015) elaborates on this issue using an adjusted CAPM model that takes into account the social damage premium. An investor with a stronger preference for environmental quality will hold less of the share if the firm pollutes more13. The rapidly growing SRI movement enlarges the social damage premium, implying that fewer and fewer investors are willing to take a long position in the bottom Environmental companies; companies that are likely to pollute more. When following the Asset Pricing Model of Merton (1973), the main assumption is a perfect market where the sum of all alphas is zero. The current market, involving many SRI investors, however, can be seen as an imperfect market. The observed significant alpha should therefore not be interpreted as an arbitrage opportunity.

5.5 Alternative interpretation of the results

Our results indicate that bottom Environmental companies are outperforming their top counterparts. However, when using a regression methodology, there is always the possibility of a missing risk factor. That is, always a chance for a type of risk that is not covered by the four factors of

13 _{The equation used by Dam and Scholtens (2015) can be perceived as an adjusted CAPM model with an additional term} 𝜆𝑗

𝑝𝑖𝐷

𝑖 _{added to the}

intercept, representing the social damage premium:

21 Carhart (1997). For instance, an event that occurs rarely is potentially not visible in one of the four risk factors of Carhart (1997). This type of risk is often labeled as “earthquake” risk or, as described by Taleb in his famous book, “Black Swans” (Taleb, 2008). A relevant example for Environmental companies is the recent Deepwater Horizon oil spill of BP in 201014. However, within this thesis we assume that the Carhart model (1997) is the correct asset pricing model and covers all relevant risk factors. By doing so, we can relate the observed alpha to values-driven SRI investors.

5.6 Robustness Checks

We employ two additional analyses to check the robustness of our results. The first robustness check relates to the effect of the Great Recession. The second test focuses on a different methodology comparing the returns value weighted portfolios with that of equally weighted portfolios.

5.6.1 The Great Recession

During the research timespan of this thesis, a large recession shocked the financial world. According to the National Bureau of Economic Research, this recession started in December 2007 and lasted till June 200915. The performance of sin stocks in time of recession is researched by the articles of Salaber (2009) and Nofsinger and Varma (2013). Although the time-span of Salaber (2009) is impressive, stretching from 1926-2005, it does not include the recession of 2008. Nonetheless, the findings are relevant for our thesis. The results indicate that U.S. sin stocks earn an abnormal return during recession periods (Salaber, 2009). Nofsinger and Varma (2013) use an U.S. data set of SRI mutual funds in the period of 2000-2011. Interesting, the performance of these SRI funds outperform that of non-SRI funds during the period of the Great Recession (Nofsinger and Varma, 2013). The results of our robustness check are as follows16: when excluding the Great Recession from our sample, no out- or underperformance of CSR portfolios is found. The highly significant outperformance of the bottom Environmental portfolios in the whole time-span, changes to a non-significant result when excluding the recession period. It can therefore be concluded that bottom Environmental companies significantly outperform their top peers in times of recession. For the other factors, no differences are found. This finding is in disaccord with the finding of superior returns for SRI funds during crisis periods, as found by Nofsinger and Varma (2012).

14

For more information on the Deepwater Horizon oil spill, see: http://www.britannica.com/event/Deepwater-Horizon-oil-spill-of-2010

15

Source: www.nber.org

5.6.2 Value weighted vs. equally weighted portfolios

In line with the articles of Galema et al. (2008) and Statman and Glushkov (2009), we conduct a robustness check of our findings by analyzing the returns of equally weighted portfolio returns17. When using equally weighted portfolios, none of the long-short portfolios display any significant differences with regards to the alpha. The findings for a lower risk loading for top Social and Corporate Governance portfolios with regards to the size risk factor remain intact. We do not consider these outcomes as problematic since we focus on the market-value weighted method, as suggested by Trinks and Scholtens (2015). Both research methods however are used frequently; the differences between the performance of market-value and equally weighted portfolios may therefore potentially explain why diverse papers have contradicting results.

6. Conclusion

This thesis relates CSR performance to financial performance for U.S. companies in the period of 2002 until 2013. Moreover, we attempt to differentiate between values- and profit-driven SRI investors by looking at risk-adjusted return data of CSR portfolios. Knowledge over motives of investors to engage in SRI is relevant because the effect of these investors on (market) returns of stocks differs between the two types of investors. We adopt the best-in-class SRI method for constructing our portfolios on three CSR dimensions: Environmental, Social and Corporate Governance based on the ranking provided by Thomson Reuters’ Asset4 data.

Our main findings are that top Environment portfolios display a negative four factor abnormal return. Top Social and Corporate Governance portfolios have a lower risk loading for the size risk factor. We identify investors in top Environment companies as values-driven; by not investing in certain companies scoring low on the Environment factor, these investors are sacrificing risk-adjusted return. We label SRI investors in top Social and Corporate Governance companies as profit-driven, having lower risk without the downside of a lower risk-adjusted return.

Our findings of lower risk factor loadings for top ESG companies are in line with existing literature. However, none of the articles known to us have found an outperformance for companies scoring low on the factor Environment, even though this is in line with theoretical expectations. The result of this significant alpha can be attributed to the impact of the Great Recession; low scoring Environment companies outperform their top scoring counterparts in time of recession, a finding in line with

23 theory. We therefore add to the existing literature quantitative prove of the existence of both values- and profit-driven investors per ESG dimension.

One limitation of this thesis is the relatively short period of the study from 2002 to 2013. Furthermore, our sample only consists of companies publishing ESG Asset4 data. All of these companies may therefore potentially already more responsible than the average of the market. Bottom 20% portfolios may for this reason not be representative as proxy for the actual bottom 20% CSR companies. Moreover, by attempting to mathematically differentiate between values- and profit-driven investors, we only analyze the financial motivation for engaging in CSR. Other motivations, such as material incentives and social- and self-esteem concerns, may also explain why certain investors engage in SRI.

There are several recommendations for further research regarding this subject. First, it would be interesting to extent the analysis from the U.S. to other parts of the world, in particular developing economies. At the moment however, the CSR data from these development countries is not reliable. Second, the effect of a recession on the returns of in particular Environmental companies is notable. It would be interesting to see if this effect is robust for past or new recessions using the same criteria as the Asset4 database, a database which is unfortunately only available since 2002. Lastly, qualitative research on the non-financial motivations for investors to engage in SRI has the potential to extent the knowledge on the effect of SRI investors.

References

Angel, J.J., Rivoli, P., 1997. Does ethical investing impose a cost upon the firm? A theoretical perspective. The Journal of Investing 4, 57-61.

Bauer, R., Koedijk, K., Otten, R., 2005. International evidence on ethical mutual fund performance and investment style. Journal of Banking & Finance 29, 1751-1767.

Benston, G. J., 1982. Accounting and corporate accountability. Accounting, Organizations and Society, 7(2), 87–105.

Boutin-Dufresne, F., Savaria, P., 2004. Corporate social responsibility and financial risk. The Journal of Investing 13(1), 57-66.

Branco, M.C., Rodrigues, L.L., 2006. Corporate Social Responsibility and Resource-Based Perspectives. Journal of Business Ethics 69, 111-132.

Carhart, M., 1997. On persistence in mutual fund performance. Journal of Finance, 52(1), 57–82.

Chung, S.Y., Eneroth,K., Schneeweis, T., 2003. Corporate Reputation and Investment Performance: The UK and US Experience. Research in international Business and Finance 17, 273-291.

Dam, L., 2008. Corporate Social Responsibility and Financial Markets. PhD thesis. University of Groningen: Groningen.

Dam, L., Scholtens, B., 2015. Toward a theory of responsible investing: On the economic foundations of corporate social responsibility. Resource and Energy Economics 41, 103-121.

Davis, K., 1973. The Case for and against Business Assumption of Social Responsibilities. The Academy of Management Journal 16 (2), 312-322.

Derwall, J., Koedijk, K., & Horst, J.T., 2011. A tale of values-driven and profit-seeking social investors. Journal of Banking & Finance, 35, 2137-2147.

Derwall, J., Guenster, N., Bauer, R., & Koedijk, K., 2005. The eco-efficiency premium puzzle. Financial Analysts Journal, 61 (2), 51-63.

Fabozzi, F.J., Ma, K.C., Oliphant, B.J., 2008. Sin stock returns. The Journal of Portfolio Management 35(1), 82-94.

Fama, E.F., French, K.R., 1993. Common risk factors in the returns on stocks and bonds. Journal of Financial Economics 33(1), 3-56.

25 Flammer, C., 2013. Does Corporate Social Responsibility Lead to Superior Financial Performance? A Regression Discontinuity Approach. Available at SSRN: http://ssrn.com/abstract=2146282

Friedman, M., 1970. The Social Responsiblity of Business is to increase its Profits. The New York Times Magazine, September 13, 1970.

Galema, R., Plantinga, A., Scholtens, B., 2008. The stocks at stake: Return and risk in socially responsible investment. Journal of Banking & Finance 32, 2646-2654.

Ghoul, El, S., Guedhami, O., Kwok, C. C. Y., Mishra, D. R., 2011. Does corporate social responsibility affect the cost of capital? Journal of Banking & Finance 35, 2388-2406.

Gregory, A., Tharyan, R., Whittaker, J., 2013. Corporate Social Responsibility and Firm Value: Disaggregating the Effects on Cash Flow, Risk and Growth. Journal of Business Ethics 124, 633-657.

Gregory, A, & Whittaker, J., 2013. Exploring the valuation of corporate social performance - A comparison of research methods. Journal of Business Ethics, 16(1), 1–20.

Heinkel, R., Kraus, A., & Zechner, J., 2001. The effect of green investment on corporate behavior. Journal of Financial and Quantitative Analysis, 35(4), 431–449.

Hong, H., & Kacperczyk, M. 2009. The price of sin: The effects of social norms on markets. Journal of Financial Economics, 93, 15–36.

Jegadeesh, N., Titman, S., 1993. Returns to buying winners and selling losers: implications for stock market efficiency. Journal of Finance 48, 65–91.

Kempf, A., Osthoff, P., 2007. The Effect of Socially Responsible Investing on Portfolio Performance. European Financial Management 13, 908-922.

Kiymaz, H., 2012. SRI Mutual Fund and Index Performance. Socially Responsible Finance and Investing 22, 425-442.

Margolis, J. D., Elfenbein, H. A., Walsh, J. P., 2009. Does it pay to be good…and does it matter? A meta analysis of the relationship between corporate social and financial performance. Harvard Business School Working paper.

Markowitz, H., 1952. Portfolio Selection. The Journal of Finance 7, 77-91.

Merton, R.C., 1973. An intertemporal capital asset pricing model. Econometrica 41, 867-887.

Nofsinger, J., Varma, A., 2013. Socially responsible funds and market crises. Journal of Banking and Finance 48, 180-193.

Porter, M.E., Van Der Linde, C., 1995. Toward a New Conception of the Environment-Competitiveness Relationship. Journal of Economic Perspectives 9, 97-118.

Renneboog, L., 2011. Is ethical money financially smart? Nonfinancial attributes and money flows of socially responsible investment funds. Journal of Financial Intermediation 20(4), 562-588.

Renneboog, L., Horst, J.T., Zhang, C., 2008. Socially responsible investments: institutional aspects, performance and investor behaviour. Journal of Banking & Finance 32 (9), 1723-1742.

Salaber, J., 2009. Sin Stock Returns over the Business Cycle, Unpublished Working Paper, University of Bath School of Management, Retrieved from: http://dx.doi.org/10.2139/ssrn.1443188 (Accessed: May 11, 2015).

Scholtens, B., 2008. A note on the interaction between corporate social responsibility and financial performance. Ecological economics 68, 46-55.

Sharfman, M. P., & Fernando, C. S. (2008). Environmental risk management and the cost of capital. Strategic Management Journal 29, 569–592.

Sharpe, W.F., 1966. Mutual Fund Performance, The Journal of Business 39 (1), 119-138.

27 Sortino, F.A., Van der Meer, R., 1991. Downside Risk, The Journal of Portfolio Management 17(4), 27- 31.

Statman, M., Glushkov, D., 2009. Wages of Social Responsibility, Financial Analysts Journal 65(4), 33-46.

Taleb, N.N., 2008. The Black swan: The Impact of the Highly Improbable. Penguin; Re-issue edition, February 28th.

Trinks, P.J., Scholtens, B., 2015. The Opportunity Cost of Negative Screening in Socially Responsible Investing. Journal of Business Ethics, published online 15 may 2015.

Appendix A – Definitions of the Asset4 Factors

Table 3: Definitions of the ESG factors used by Thomson Reuters’ Asset4 Database

Factor Description

Equal-Weighted Rating

The Equal Weighted Rating reflects a balanced view of a company's performance in all four areas, economic, environmental, social and corporate governance

Corporate Governance

The corporate governance pillar measures a company's systems and processes, which ensure that its board members and executives act in the best interests of its long term shareholders. It reflects a company's capacity, through its use of best management practices, to direct and control its rights and responsibilities through the creation of incentives, as well as checks and balances in order to generate long term shareholder value.

Economic* The economic pillar measures a company's capacity to generate sustainable growth and a high return on investment through the efficient use of all its resources. It is reflection of a company's overall financial health and its ability to generate long term shareholder value through its use of best management practices.

Environmental The environmental pillar measures a company's impact on living and non-living natural systems, including the air, land and water, as well as complete ecosystems. It reflects how well a company uses best management practices to avoid environmental risks and capitalize on environmental opportunities in order to generate long term shareholder value.

Social The social pillar measures a company's capacity to generate trust and loyalty with its workforce, customers and society, through its use of best management practices. It is a reflection of the company's reputation and the health of its license to operate, which are key factors in determining its ability to generate long term shareholder value.

29

Appendix B – Process steps Asset4

Table 4: Process steps as described in Rules and Methodologies book of Thomson Reuters Corporate Responsibility Ratings (TRCRR)

Step Description

1. Raw Score Every company with at least on reported KPI in a given year is scored from 0 to 1 for each pillar. These scores are driven by ASSET4 data, which in turn is driven by company financial reporting. For current scores, the most recent year available is used with the fiscal year clearly delineated. The scores are calibrated to be robust over time while also be relative to each company’s peer group.

2. Ratings The raw scores are normalized and adjusted for skewness and the differential between the mean and the median, then fitted to a bell curve to derive ratings between 0 and 100 for each company. The ratings are centered and comparable across pillars. The result is a consistent, objective and finely calibrated standard of rating every company’s environmental, social, governance and combined ESG practices.

3. Percentile Rank

Based on a company’s raw scores as defined above, percentile ranks are calculated for all companies screened.

Appendix C – Measuring the effect of the Great Recession

Table 5: Descriptive statistics and risk/performance-related measures of best- and worst-in-class portfolios monthly returns of U.S. firms by ESG factor excluding the Great Recession

Mean Median StDev Variance Min Max Kurt Skew Sharpe Sortino

Best-in-class portfolios Aggregate 1.00% 1.01% 3.49% 0.12% -10.4% 10.6% 1.222 -0.276 0.255 0.021 Environmental 1.33% 1.74% 3.68% 0.14% -9.3% 10.9% 0.501 -0.289 0.327 0.021 Social 1.41% 1.75% 3.79% 0.14% -10.5% 11.5% 0.777 -0.264 0.393 0.021 Corporate Governance 1.30% 1.76% 3.72% 0.14% -9.9% 11.0% 0.653 -0.332 0.316 0.022 Worst-in-class portfolios Aggregate 1.44% 1.89% 5.52% 0.31% -22.4% 21.3% 3.904 -0.695 0.237 0.036 Environmental 1.63% 1.83% 3.76% 0.14% -9.1% 13.1% 0.903 -0.100 0.400 0.011 Social 1.71% 2.11% 4.02% 0.16% -9.6% 13.0% 0.377 -0.199 0.320 0.021 Corporate Governance 1.70% 1.76% 3.94% 0.16% -7.8% 13.0% 0.452 0.100 0.398 0.020

31

Table 6: Fund performance of U.S. companies portfolios selected on their CSR scores in the period of 2002 – 2013 excluding the Great Recession.

α RM-RF SMB HML MOM Adjusted 𝑹𝟐

Panel A: Long – Short portfolios

Bottom minus Top Asset4 Aggregate scores 0.002

(0.74) 0.20*** (3.09) 0.64*** (6.59) -0.04 (-0.40) -0.34*** (-5.54) 0.516

Bottom minus Top Environmental 0.002

(1.29) -0.04 (-0.95) 0.34*** (4.48) 0.06 (0.71) 0.00 (0.01) 0.134

Bottom minus Top Social 0.003

(1.46) -0.01 (-0.21) 0.43*** (5.42) -0.04 (-0.45) -0.01 (-0.10) 0.199

Bottom minus Top Corporate Governance 0.003*

(1.70) 0.03 (0.60) 0.19** (2.33) -0.15 (0.05) 0.05 (0.90) 0.041

Panel B: Top and Bottom portfolios

Asset4 Aggregate Top 0.001

(1.04) 0.94*** (30.36) -0.30 (-6.40) -0.07 (-1.46) 0.06** (1.04) 0.898

Asset4 Aggregate Bottom 0.003

(1.31) 1.13 (18.77) 0.34 (3.67) -0.11 (-1.17) -0.28*** (-4.74) 0.846 Environmental Top 0.004*** (3.60) 0.96*** (31.24) -0.26 (-5.43) 0.03 (0.51) -0.01 (-0.49) 0.910 Environmental Bottom 0.006*** (5.62) 0.92*** (29.41) 0.08 (1.76) 0.08 (1.63) -0.01 (-0.47) 0.912 Social Top 0.004*** (4.23) 0.96*** (31.28) -0.28 (-6.52) -0.01 (-0.19) -0.01 (-0.19) 0.923 Social Bottom 0.007*** (4.97) 0.95*** (24.33) 0.15 (2.53) -0.05 (-0.73) -0.01 (-0.27) 0.879

Corporate Governance Top 0.003***

(3.12) 0.96*** (30.10) -0.18 (-3.66) -0.02 (-0.52) -0.01 (-0.20) 0.905

Corporate Governance Bottom 0.006***

(5.23) 0.99*** (27.70) 0.01 (0.11) -0.04 (-0.68) 0.04 (1.12) 0.894

Appendix D – Equally weighted portfolios

Table 7: Descriptive statistics and risk/performance-related measures of best- and worst-in-class portfolios monthly returns of U.S. firms by ESG factor

Mean Median StDev Variance Min Max Kurt Skew Sharpe Sortino

Best-in-class portfolios Aggregate 1.05% 1.68% 4.70% 0.22% -19.4% 16.2% 2.62 -0.612 0.196 0.031 Environmental 1.10% 1.82% 5.24% 0.27% -20.8% 19.9% 2.681 -0.483 0.157 0.036 Social 1.10% 1.79% 4.96% 0.25% -21.0% 19.2% 3.309 -0.590 0.196 0.033 Corporate Governance 1.08% 1.67% 5.24% 0.27% -20.2% 19.1% 2.428 -0.584 0.182 0.035 Worst-in-class portfolios Aggregate 1.44% 1.44% 8.15% 0.66% -24.9% 54.4% 12.44 1.809 0.162 0.044 Environmental 1.05% 1.44% 5.35% 0.29% -22.2% 17.7% 2.446 -0.660 0.173 0.014 Social 1.22% 1.77% 5.81% 0.34% -23.0% 20.4% 2.161 -0.500 0.189 0.038 Corporate Governance 1.35% 2.11% 5.89% 0.35% -24.9% 25.2% 3.780 -0.410 0.209 0.038