An empirical analysis on the impact of sustainability on stock returns.

(1)

0

University of Amsterdam, Amsterdam Business School

Master in International Finance

An Empirical Analysis on the Impact of

Sustainability on Stock Returns

September 2014

Abstract

Sustainable investing has attracted a lot of attention over the last decade, though the financial implication of sustainability is still inconclusive. This paper examines the relationship between sustainability and stock’s risk and return by analyzing the US sustainability rating data provided by Thomson Reuters ASSET4 for the period 2002-2011. No clear relationship is found between sustainability and stock’s abnormal return and negative relationships are observed between sustainability and market & idiosyncratic risks. After controlling for market capitalization, annual multi-factor alphas of 3.5% and 4.7% are observed in long-short portfolios created based on ESG ratings and corporate governance ratings respectively. The results indicate that sustainability ratings could provide valuable information to investors on stocks’ risk, and combining sustainability information with market capitalization can help improve investment performance.

Author: Ying Xiong Thesis Supervisor: Jeroen Ligterink

(2)

1

Acknowledgement

I would like to take this opportunity to express my gratitude to my supervisor Prof. Ligterink for his patience, motivation, and insightful comments. Besides my supervisor from the university, my sincere thanks also go to my supervisor from Robeco, Van der Grient, for his continuous support and valuable guidance for the writing of this thesis.

I would also like to thank Dr. Blitz for offering me the internship in his team so that I could have the great opportunity to learn from and work closely with professionals at Robeco.

Last but not the least, I would like to thank my family: my parents Guanghui Xiong and Xiaolan Chen, for giving birth to me at the first place and supporting me spiritually throughout my life.

(3)

2

1. Overview of the Research

1.1 Introduction

The concept of “Socially Responsible Investment” (SRI) can be dated back to the movement of Religious Society of Friends in the 17th_{century, which was a}

significant part of the movements for the abolition of slavery, to promote equal rights for women and peace. The members were informally called “Quakers” and they were prohibited from investing in slavery (Schueth, 2006). Till now it is a hot topic in the investment field; there has been a myriad of researches on the significance of SRI in the last 2 decades and it has been driven partly by the climate change issues and partly by the recent financial crisis. The modern era of SRI dates back to the 1960s, when many responsible investment funds started to appear and increased focus was placed on individual companies’ business choice and behavior (Eurosif-European SRI Study 2012). After decades of development, the notion of “sustainable development” was brought in by investors in the 2000s, expanding SRI into “sustainable and responsible investment” (Eurosif-European SRI Study 2012). SRI was defined negatively-avoiding to invest in “sin companies” or problematic industries engaged in businesses involving tobacco, alcohol, weapon etc. (Joseph F. Keef, 2007). However, now the meaning of SRI has been expanded. According to the definition of Euroif (the European Sustainable Investment Forum), sustainable and responsible investing (SRI) is “a generic term covering any type of investment process that combines investors’ financial objectives with their concerns about Environmental, Social and Governance (ESG) issues.” “ESG” might sound

(5)

4

irrelevant to investors, whose main goal is to generate higher monetary returns; however, if we do not take ESG factors into consideration, we are potentially ignoring major risks, such as legal risk, operating risk and reputation risk

(Schwesernotes 2013, CFA Level II Book 2). Instead of being defensive, sustainable and responsible investing bases its decisions on identifying companies that are better-managed, more forward-looking in the form of corporate social responsibility (CSR) ratings, so as to have better long-term financial prospects (Keef, 2007, Gillan, Hartzell, Koch, and Starks, 2010). CSR was defined by European Commission as “a concept whereby companies integrate social and environmental concerns in their business operations and in their interaction with their stakeholders on a voluntary basis”. A survey by the UN Global Compact in 2010 shows that 93% of the 766 CEOs believe that CSR issues will be critical to the future success of their business (Lacy, Cooper, Hayward, Neuberger, 2010).

Hence, the idea behind sustainable investing is not just to contribute to a better world with less pollution, fewer scandals and more harmonious society, but to lower the risk and generate higher financial returns. However, no consensus has reached regarding the financial implications of sustainability. More details are shown in Chapter 2, Literature Review. The objective of the research is to find out empirical evidence of whether following a sustainable investing strategy leads to lower risks and higher financial returns.

In this study, the ESG ratings as well as the separate ESG dimension ratings provided by Thomson Reuters ASSET4 US data are used to analyze the impact of

(6)

5

sustainability on stock returns and risks for the period of 2004-2013 (using the rating data from 2002 to 2011). Sample stocks are divided into quintile portfolios according to ratings and their financial returns are measured under Carhart (1997) model. In order to test the relationship between sustainability and risk, CAPM beta is used to measure market risks, while the standard deviation of Carhart regression residuals is used to measure idiosyncratic risks of portfolios.

No significant alphas are found for the long-short portfolios created by sorting on ratings under Carhart (1997) model in the current study. Portfolios with higher ratings are found to have lower market risk and idiosyncratic risk, and the results are consistent with earlier studies of McGuire et al (1988) and Mishra and Modi (2012). After

controlling for size factor, significant outperformance of 3.5% annually is found and the outperformance observed might be attributed to the biasness in ratings for large companies. This thesis contributes to the literature by analyzing the relationship between sustainability and risk at a portfolio level and by showing that combining sustainability information with market caps could help to improve the investment performance.

1.2 Thesis Structure

Chapter 2 reviews the literature and derives the hypotheses. Chapter 3 discusses the data.

Chapter 4 describes the methodology and this is followed by the empirical results and

discussion of those results in chapter 5. Chapter 6 is the robustness test. The conclusion of the thesis is in chapter 7.

(7)

6

2. Literature Review

2.1 Different Views on the Financial Benefits of Sustainable Investing One important obstacle that prevents sustainable investing from being included in the mainstream investing is the lack of theoretical agreement regarding whether it is consistent with the aim of maximizing shareholder wealth (Statman, 2000). There are different views regarding the relationship between sustainability and financial performance.

Opponents of SRI state that “do we really want people pooling their investing power for the avowed purpose of achieving some specific end, other than making more money?”(Clark, 1998). Langbein and Posner (1980) claimed that social responsible investing involves excluding attractive companies which are socially irresponsible. Based on modern portfolio theory, they conclude that though there is no reason to expect a socially responsible portfolio to perform worse, the reduced diversification may make social investing economically unsound. However, Kurtz (2005) found that the Domini Social Index outperformed the S&P 500 by an annualized 125 basis points per year for 15 years ended December 2004, and Barnett M. and Salomon R. (2006) hypothesize that the financial loss of SRI caused by reduced diversification can be offset because better-managed and more stable firms are selected and they find support for this hypothesis through empirical study.

Other opponents argue that “provision of CSR actions and attributes leads to ‘extra’ costs” (McWilliams and Siegel, 2001a). The extra costs associated with socially responsible activities might put these companies in a disadvantaged

(8)

7

competitive place compared to other firms, and therefore lowering the rates of return.

However, proponents argue that many benefits of sustainable investing should be taken into account instead of only looking at the “extra” costs involved. Acting in an ethical and socially responsible manner will bring about reputational benefits which are important to firms’ competitive success (Porter and Kramer 2006). In addition, the better the CSR reputation of a firm, the lower is its total market risk (measured by standard deviation of expected return) (Orlitzky and Benjamin, 2001); a high degree of sustainability may help to generate moral capital or goodwill which can serve as an “insurance-like” protection to preserve financial performance (Godfrey, 2005).

Companies that do not address ESG factors might be exposed to potential risk for lawsuits, which could have a material impact on a company (Schwesernotes 2013 CFA Level II Book 2). Hong and Kacperczyk (2007) conduct a study on “sin stocks” and they find that “sin stocks” are facing greater litigation risks.

In addition to that, ESG laggards may face higher cost of capital. Because the solvency of firms can be influenced by their environmental or social activities, and lenders are also concerned about their reputation, socially irresponsible corporations may have to pay a higher rate for their capital (Willard B., 2011) so as to negatively influence the value of the firm. By analyzing 3996 loans to US companies, Goss A. and Roberts G. (2011) find that firms behaving less socially responsible pay between 7 and 18 basis points more for bank debts than firms that are more responsible.

(9)

8

Ghoul, Guedham, Kwok and Mishra (2011) find that firms with higher CSR scores are paying less for equity financing.

Though proponents and opponents have been trying to find theoretical and driving influences of sustainable investing, there is still no consensus regarding the relationship between CSR and corporate financial performance.

2.2 Empirical Studies on Return of Sustainable Investments

Various studies analyze whether a sustainable investing strategy outperforms conventional investment strategy. Below I will discuss the most relevant ones. Kempf and Osthoff (2007) conducted a research on the effect of socially responsible investing on portfolio performance using the KLD ratings data for the period of 1992-2004. KLD ratings measure the social responsibility of a company and it covers all the stocks in S&P 500 and the DS 400 (Domini Social Index 400, around 650 stocks). On the basis of these ratings they created long portfolios in the highest KLD ratings and short portfolios in low-rated stocks. They then relate the returns to a multifactor model as in Carhart (1997) to test for outperformance of these portfolios. The result shows that under Carhart (1997) model, the best-in-class strategy leads to the highest alphas (up to about 8.7% per year).

Derwall J. Guenster N. Bauer R. and Koedijk K. (2005) constructed two mutually exclusive portfolios with distinctive “eco-efficiency” characteristics for the period between 1995 and 2003 based on the rating data of Innovest, which measures a company’s eco-efficiency on about 60 dimensions using both quantitative and qualitative sources. They only considered US companies in the database and the

(10)

9

number of companies in the sample was 180 in 1997 and increased to 450 in 2003. The result shows that after considering industry effect, and using a similar long short approach as in Kempf and Osthoff (2007), the spread of multi-factor alpha between the two portfolios reached 6.04%, statistically significant at 5% level.

Alex Edmans (2010) analyzed the relationship between employee satisfaction and long-run stock return by building up a value-weighted portfolio of the “100 Best Companies to Work for in America”, which earned an annual 4-factor alpha of 3.5% from 1984 to 2009.

Paul A. Gompers, Joy L. Ishiiand Andrew Metrick (2003) constructed a “Governance Index” to proxy for the level of shareholder rights at about 1500 large firms during the 1990s. By using the strategy of buying firms in the lowest quintile of the index (strongest rights) and selling firms in the highest quintile of the index (weakest rights), 8.5% of abnormal returns can be earned per year on the basis of a similar Carhart (1997) model.

From the studies mentioned above we can find that portfolios with better CSR performance (Kempf et al, 2007), better Environmental (Derwall et al, 2005), Social (Alex Edmans, 2010), and Corporate Governance performance (Grompers et al, 2003) were found to have better financial performance in terms of stock returns. However, some other studies fail to get a statistically significant alpha.

Van de Velde,Vermeir and Corten (2005) used companies covered by Vigeo database (about 300 stocks), which is an independent agency providing sustainability scores in the euro zone, to construct a sample. They built 4 portfolios based on the

(11)

10

Vigeo corporate social responsibility scores for the period between 2001 and 2003. They concluded that under the Fama French model, sustainable investments performed slightly better than traditional investments-but not enough to result in a statistically significant out-performance.

The research by Dinusha Peiris (2009) shows that for the period of 1991-2006, the average monthly returns of Dow Jones Sustainability (DSI) index have been consistently higher than the S&P 500 with the overall average difference of 0.3% and after adjusting for volatility; the returns of DSI were only slightly higher than the S&P 500. Under the multifactor framework, the abnormal return of DSI stocks lack statistical significance, no clear relationship between stock returns and rating criteria can be found.

By using different sustainability ratings and different samples, different results were found. It is still not conclusive regarding the relationship between sustainability ratings and stock returns. A summary of the literatures is in Table 1.

Table 1: Summary of Literature Review: Sustainability and Stock Return

Study Data Analysis method Findings

Kempf and Osthoff (2007)

KLD CSR Ratings from 1991 to 2003, around 650 US stocks in the database Carhart (1997) Model Annualized alpha of 8, 7% of long-short portfolio (significant at 1% level) was found using best-in-class approach.

Derwall et al (2005)

Innovest eco-efficiency rating data from 1995 to 2003, between 180 and 450 US stocks in the sample

Carhart (1997) Model

The spread in annualized alpha between high rating portfolio and low rating portfolio is 6, 04%, significant at 5% level.

(12)

11

Table 1 (Continued)

Edmans (2010)

The list of the ‘‘100 Best Companies toWork for in America’’

(Levering,Moskowitz,andKatz,1984; Leveringand

Moskowitz,1993)

Value weighted portfolio earned a four-factor alpha of 3.5% (significant at 1% level) annually from 1984 to 2009. Gompers et al (2003)

"Governance Index" created by authors for about 1500 large firms, sample period: 1990-1999

Abnormal return of 8.5% yearly can be earned by a long-short strategy

Van de Velde et al (2005)

Vigeo database, around 300 stocks in euro zone for the period from 2001 to 2003

Fama French model

No significant

outperformance can be found for sustainable investment

Dinusha Peiris (2009) DSI stocks for the period from 1991 to 2006

No significant

outperformance can be found for sustainable investment

2.3 Empirical Studies on Risk of Sustainable Investments

As to CSR and firm risk, most previous studies on the relationship between corporate social responsibility and risk have shown that there is a negative relationship between sustainability and risk.

McGuire, Sundgren and Schneeweis (1988) use beta and return volatility to measure risk and find that a significant portion of corporate social responsibility can be explained by firm risk. Using standard deviation of the residuals of Carhart (1997) regression as the measure of idiosyncratic risk Mishra and Modi (2012) find that CSR has a significant effect on corporates’ idiosyncratic risk with positive CSR (positive aspects of CSR) reducing risk and negative CSR (negative aspects of CSR) increasing risk. Dobler, Lajili and Zeghal (2012) base their study on 89 US firms from 4 sectors which are considered to have high pollution propensity and by using

(13)

12

multiple regression, they find a negative relationship between environmental performance and environmental risk. Bouslah, Kryzanowski and M’Zali (2013) conduct a research on the impact of single social performance dimensions on firm risk and they find that employee relations and human rights are negatively related to firm risk. Using implied volatility as the measure of risk, Li, Jahera Jr and Yost (2012) find that corporate risk is significantly inversely-related to corporate governance performance.

A summary of the literatures on the relationship between CSR and risk is shown in Table 2. Various studies have found significant negative relationships between CSR and firm risk; however, it is not clear whether better CSR performance leads to lower firm risk or companies with lower risks are performing better in terms of CSR.

Table 2: Summary of Literature Review: Sustainability and Risk

Study Data Analysis method Findings

McGuire et al (1988)

131 firms rated by Fortune for CSR for the period 1983-1985

Regressing CSR measure on various financial performance variables

Significant portion of

corporate social responsibility can be explained by firm risk (market risk & total risk)

Mishra and Modi (2012)

192 US stocks included in KLD database from 2000 to 2009

Idiosyncratic risk is regressed on positive and negative CSR measures and other firm specific characteristics.

Significant positive

relationship is found between negative CSR and

idiosyncratic risk; significant negative relationship is found between positive CSR and idiosyncratic risk.

Dobler, Lajili and Zeghal (2012)

Four sectors in S&P 500 listed firms (89 firms), TRI data is used to measure environmental performance

Regressing environmental performance on

environmental risk variable and other firm specific characteristics

Negative relationship between environmental risk and environmental performance was found

(14)

13 Table 2 (Continued) Bouslah, Kryzanowski and M’Zali (2013) Unbalanced panel of 16599 firm-year observations (US) over the period 1991-2007 in KLD database

Regressing risk measures on social performance measures and firm specific characteristics

Using whole sample, it is found that employee relations and human rights are

negatively related to firm risk

Li, Jahera Jr and Yost (2012)

6176 biannual firm-year observations spanning 1998-2006 are covered in the study; Gompers Index is used to measure corporate

governance performance.

Using implied volatility as a measure of risk, risk variables are regressed on Gompers Index and other CG measurements.

Corporate risk is significantly inversely-related to corporate governance performance.

3. Hypotheses and Data Analysis

3.1 Hypotheses

Flowing from the aims of the study, the hypotheses being tested and their underlying reasoning are summarized below.

H1: Higher sustainability rating (overall ESG rating, environmental rating,

social rating and corporate governance rating) of a stock is associated with higher stock return.

According to the literature review, there is no consensus regarding the relationship

between sustainability and financial performance, and the empirical testing results are not conclusive. However, many evidences show that there are some financial benefits brought by being socially responsible and it is expected that in the long run, being sustainable will have positive impacts on the financial performances of firms.

H2: Higher sustainability rating (overall ESG rating, environmental rating, social rating and corporate governance rating) of a stock is associated with lower risk.

(15)

14

Stakeholder theory holds that CSR would affect the relationship between firm and its key stakeholders, which reflects in firm-specific resources, so as to influence firm’s idiosyncratic risk (Mishra and Modi (2012)). And most of previous studies find a

negative relationship between sustainability and firm risk (idiosyncratic risk/ total risk/market risk). In this analysis, CAPM beta will be used to measure the market risk of portfolios and test whether portfolios with higher sustainability ratings will show a lower market risk. In addition, the standard deviation of the residuals of Carhart (1997) regression is used to measure the idiosyncratic risks of the quintile portfolios to test whether portfolios with higher ratings will have lower idiosyncratic risks.

3.2 Introduction to ASSET4 ESG Database

In order to rank firms on the basis of corporate responsibility, the ESG scores of Thomson Reuters ASSET4, which is a leading provider of Environmental, Social and Corporate Governance (ESG) data, will be used. ASSET4 is staffed with over 100 analysts1 who use their experience to collect information from only publicly available sources (e.g. annual reports, NGO websites and CSR reports).

At inception, the database covers 959 companies in the year 2002 globally, and now it contains more than 4300 global companies, including MSCI World, MSCI Europe, STOXX600, NASDAQ100, Russell1000, S&P500, FTSE100, ASX300 and MSCI Emerging market funds.

The overall ESG scores are calculated by comparing company’s performance

(16)

15

against benchmarks for more than 280 ESG indicators and over 750 CSR data points so as to outline the competitive advantage of a company. These ratings are z-scored and normalized to position the score between 0% and 100%2. The ratings of a company might change from year to year because of the expanding of the database (the benchmark it compares to changes); however, it will not influence the result of the analysis because the rank of the ratings matters.

Companies usually report their ESG information on an annual basis, so the sustainability ratings given by ASSET4 are also updated annually. In addition, because most companies publish their CSR reports with a considerable delay to their annual reports and the lag between CSR and annual report can be up to 9 months (for example JP Morgan published its 2007 CSR report in January 2009). Portfolios will be created only when new information is available.

An advantage of ASSET4 ESG database is it provides the breakdown of the overall sustainability ratings (Environmental Performance score, Corporate Governance Performance score), so that it is possible to look into the effects of the various components. Another advantage of ASSET4 ESG is that the database also includes the ratings of companies that ceased operation, so that the influence of survivorship could be controlled for.

One concern of using ASSET4 ESG ratings to measure company’s sustainability is that the ASSET4 ESG ratings also include some economic issues, such as accounting compliance, audit independence etc., so it is not a “pure” rating. However,

(17)

16

out of more than 280 indicators only about 50 of them belong to economic performance, and some of the economic indicators are actually related to employee satisfaction and shareholder rights. So the impact of economic elements is not very high and the overall ESG rating given by ASSET4 is still a good approximation of sustainability. The details of the ASSET4 ESG rating indicators are in Appendix 1.

3.3 ASSET4 US Sample Summary

For the current research, only US companies of the database are considered; the database is expanding year by year and 440 and 1024 US companies were in the database in the year of 2002 and 2011 respectively. In total 1197 companies are included in the research including those ceased operating after 2002 (Figure 1). The overall ESG ratings are obtained from DataStream, while the individual ratings for each pillar are calculated by taking the equal weighted scores of all the categories (obtained from DataStream) in the pillar.

The composition of the sample according to market cap is also shown in Figure 1; the majority (89%) of the sample consists of large and mid-caps and only about 10% companies are small caps. So the sample is slightly biased to large firms.

(18)

17

Note: The market capitalizations of sample are obtained from datastream as per end of 2013; for companies that ceased operation before 2013, the market capitalizations when they are out of business are used. The definition of cap is as follows: Mega Cap - Market cap of $200 billion and greater, Big Cap - $10 billion and greater, Mid Cap - $2 billion to $10 billion, Small Cap - $300 million to $2 billion, Micro Cap - $50 million to $300 million, Nano Cap - Under $50 million. The definition of cap is according to Rick Wayman, Understanding Small- And Big-Cap Stocks.

3.4 Financial Data

Financial variables of the companies are obtained from Thomson Reuters DataStream.

Monthly returns of stocks: Total Return Index is obtained from DataStream for each stock every month from 2004 to 2013. It shows a theoretical growth in value of a shareholding over a specified period, assuming that dividends are re-invested to purchase additional units of an equity or unit trust at the closing price applicable on the ex-dividend date3. The monthly returns of stocks are calculated according to equation (1):

r_i,t = TRIi,t

TRI_i,t−1− 1 (1)

Monthly Market Capitalization of stocks: Monthly market values of the stocks are also obtained from DataStream in order to calculate the monthly value weighted return of the portfolios and also for robustness test (double sorting). Market value on DataStream is the share price multiplied by the number of ordinary shares in issue. The amount in issue is updated whenever new tranches of stock are issued or after a capital change. For companies with more than one class of equity capital, the market value is expressed according to the individual issue4.

Historical Beta: Historical betas are calculated by regressing stocks’ past three

years’ monthly excess returns on market excess returns. The monthly returns of the

3_{The explanation of total return index is according to DataStream.} 4_{The explanation of market value is according to DataStream.}

(19)

18

stocks are obtained from DataStream and the market excess returns are obtained from Kenneth R. French-Data Library.

4. Methodology

4.1 Portfolio Construction Method

In order to compare the performance of stocks with different sustainability ratings, quintile portfolios are constructed based on the ranking of different ratings (overall ESG ratings, environmental performance ratings, social performance ratings and corporate governance ratings) following the methodology used in Gompers et al (2003) and Van de Velde (2005).

As the sustainability ratings are updated annually, portfolios are re-balanced every year based on new information. Further, since there is usually a one-year delay in the publishing of ratings, portfolios will only be constructed when new ratings are available. For example, portfolios constructed at the beginning of 2004 are based on the rating information of 2002. For the current research, portfolios are built for the years from 2004 to 2013 based on ratings from 2002 to 2011.

A statistical summary of the quintile portfolios created based on ESG ratings is shown in Appendix 2.

Monthly returns of the value weighted portfolios are calculated according to equation (2):

rt = ∑ 𝑀𝑎𝑟𝑘𝑒𝑡 𝐶𝑎𝑝 𝑊𝑒𝑖𝑔ℎ𝑡𝑡−1,𝑖× 𝑟𝑡,𝑖 𝑖

(2)

(20)

19 different ratings is shown in Table 3:

Table 3: Summary Statistics of Value Weighted Portfolios Created on Ratings

This table gives a statistical summary of the returns of quintile value weighted portfolios (2004-2013) created based on ESG, environmental, social, and corporate governance ratings provided by Thomson Reuters for the period of 2002-2011. D1 is the portfolios with highest rating and D5 is the portfolios with lowest rating.

D1 D2 D3 D4 D5

ESG

Max monthly return 9.86% 12.18% 14.04% 13.78% 14.93% Min Monthly return -15.26% -17.53% -21.35% -19.22% -22.35% Annualized average return 8.22% 10.29% 6.92% 9.04% 10.05% Annualized Std. Devn. Of Return 13.48% 16.15% 17.01% 16.84% 17.18% Environmental

Max monthly return 9.76% 13.09% 14.63% 16.31% 11.64% Min Monthly return -14.92% -19.46% -19.96% -23.38% -17.73% Annualized average return 8.32% 8.80% 10.19% 10.16% 7.90% Annualized Std. Devn. Of Return 13.77% 16.18% 17.29% 18.08% 15.24% Social

Max monthly return 9.56% 13.87% 11.85% 14.71% 16.25% Min Monthly return -15.07% -17.68% -21.60% -21.42% -19.97% Annualized average return 8.25% 7.98% 9.77% 9.46% 10.76% Annualized Std. Devn. Of Return 13.51% 16.04% 16.81% 17.90% 16.39% Corporate Governance

Max monthly return 9.81% 10.75% 13.00% 14.11% 14.77% Min Monthly return -15.21% -17.06% -18.12% -19.27% -20.28% Annualized average return 9.71% 8.79% 6.74% 7.54% 10.02% Annualized Std. Devn. Of Return 13.34% 15.23% 16.59% 16.20% 17.31%

4.2 Performance Measurement

Following the literature (e.g. Kempf et al. (2007), Derwall et al. (2005) etc.), the outperformances of portfolios are determined using Carhart (1997) four-factor model. It controls for the impact of market risk, the size factor, the value factor and the momentum factor on stock returns:

𝑅_𝑖𝑡− 𝑅_𝑓𝑡 = 𝛼_𝑖 + 𝛽_0𝑖(𝑅_𝑚𝑡− 𝑅_𝑓𝑡) + 𝛽_1𝑖𝑆𝑀𝐵_𝑡+ 𝛽_2𝑖𝐻𝑀𝐿_𝑡+ 𝛽_3𝑖𝑀𝑜𝑚_𝑡+ 𝜀_𝑖𝑡

(3)

(21)

20

(1997), which is used to examine the outperformance of portfolios; 𝑆𝑀𝐵_𝑡 is the difference in return between a small cap portfolio and a large cap portfolio at time t; 𝐻𝑀𝐿_𝑡 is the difference in return at time t between a portfolio containing value stocks (with a high book-to-market ratio) and one containing growth stocks (with a low book-to-market ratio); 𝑀𝑜𝑚𝑡 is the difference in return between a portfolio of

past 12 months winners and a portfolio of past 12 month losers at time t (Mark M. Carhart (1997)).

Using monthly data, the model helps determine outperformance of each portfolio. 𝛽_0𝑖 of portfolios are used to measure the market risk of portfolios.

The Carhart factors data are obtained from Kenneth R. French-Data Library. The market risk of the portfolios is measured using CAPM beta. CAPM is used to determine a theoretically appropriate required rate of return of an asset. In CAPM model, market premium is the only factor taken into account to determine the required rate of return:

𝑅𝑖𝑡− 𝑅𝑓𝑡 = 𝛼𝑖 + 𝛽𝑖(𝑅𝑚𝑡− 𝑅𝑓𝑡) + 𝜀𝑖𝑡 (6)

where 𝛼𝑖is the monthly excess return of the portfolio based on CAPM model, 𝑅𝑖𝑡 is

the return on the portfolio i in month t, 𝑅𝑓𝑡 is the return on one month T-bill in

month t, 𝑅_𝑚𝑡 is the return on relevant equity benchmark in month t and 𝜀𝑖𝑡 is an

error term.

Idiosyncratic risks are also tested for each portfolio. Following the study of McGuire, Sundgren and Schneeweis (1988), idiosyncratic risks are measured by the standard deviations of the residuals of Carhart (1997) regressions. The higher the

(22)

21

standard deviation, the higher is the idiosyncratic risk.

Different from earlier studies, which used panel data regression to test the relationship between CSR and risk measures, current study is on conducted on a portfolio level and might give some more insights to investors who diversify their investments.

5. Results

5.1 Sustainability and Return

The main regression results of the quintile portfolios are reported in Table 4 (Detailed information of the multi-factor regressions is in Appendix 3).

It shows in Table 4 that marginal outperformances are found in long-short portfolios (D1-D5) created based on ESG ratings, Environmental ratings and Corporate Governance ratings, but all the alphas lack statistical significance. For portfolios created based on social ratings, D1 (portfolio with highest ratings) underperforms D5 (portfolio with lowest ratings) by 1.15% annually, but also lacking statistical significance.

Though some studies find strong relationship between sustainability and stock return (e.g. Kempf and Osthoff (2007)), the results of the current study indicate that no clear relationship between sustainability (ESG ratings) and stock abnormal returns is found, same as the finding of Van de Velde et al (2005). The differences in conclusions are attributed to different rating data and different samples used, and that is why the financial impact of sustainability is still inconclusive.

(23)

22

As to the relationship between environmental performance and stock return, Derwall et al (2005) found 6.04% outperformance of high rated portfolio against low rated portfolio, while in current study, the outperformance is not significant. One possible reason is the difference in the environmental performance measurement method: in the study of Derwall et al (2005), a relative environmental performance measurement method is used (eco-efficiency5_{); while in current study “absolute”}

environmental performance measurement method is used.

Outperformance is not observed in portfolios with higher social ratings in the current study. Employee satisfaction is an important element of company’s social performance and Alex Edmans (2010) found that companies with higher employee satisfaction had better stock returns. But for current study, more aspects (e.g. community relationship) of social performance are included in the ratings, and it turns out that after combining together these aspects, better social performance cannot lead to better stock returns.

For portfolios created based on corporate governance ratings, no significant abnormal return is found in the long-short portfolio (D1-D5), but it is remarkable to see that D1 (portfolio with the highest corporate governance rating) earns an excess return of 1.7% annually, significant at 5% level, while other portfolios (D2 to D5) fail to generate positive and significant alpha. Though the relationship between corporate governance ratings and stock returns according to the results does not seem to be so obvious compared with the results got by Gompers et al (2003), it

5_{Eco-efficiency can be defined as the ratio of the value a company adds to the waste the company generates by}

(24)

23

shows investing in companies within the top quintile of corporate governance ratings could bring out higher abnormal return.

5.2 Sustainability and Risk

CAPM beta is used to measure the market risks of the portfolios. The spreads in betas between D1 and D5 are all significant (for portfolios created based on ESG ratings, Social ratings and Corporate Governance ratings, the differences in beta are significant at 1% level and for portfolios created based on environmental ratings, the difference in beta is significant at 5% level). The results suggest a negative relationship between sustainability and market risk-portfolio with higher sustainability ratings is associated with lower market risk. The finding confirms the results got by McGuire et al (1988).

Idiosyncratic risk is measured by the standard deviation of the Carhart (1997) regression residuals and the results are shown in Table 4. Earlier study by Mishraand Modi (2012) was done using panel data regression, while the current study is conducted on a portfolio level, but the results are consistent: current study shows that higher rating portfolios are with lower idiosyncratic risks and lower rating portfolios are with higher idiosyncratic risks; the differences in idiosyncratic risks between D1 and D5 are all significant.

(25)

24

Table 4: Main Results: Value Weighted Portfolios Created Based on Sustainability Ratings

This table summarizes the multi-factor alphas and CAPM betas of value weighted portfolios created based on ESG ratings, environmental ratings, social ratings and corporate governance ratings respectively. The portfolios are for the period of 2004-2013 based on ratings from 2002 to 2011. D1 is the quintile portfolio with highest ratings; D5 the portfolio with lowest ratings. D1-D5 is the long-short portfolio of D1 and D5. The t-statistics of CAPM-beta for D1 to D5 are to test whether betas are statistically different from 1. *, **, *** indicate significance at 10%, 5%, and 1% respectively.

D1 D2 D3 D4 D5 D1-D5 ESG Multi-factor Alpha 0.2947% 1.0456% -2.5358% -0.4947% 0.2488% 0.0458% (t-value) 0.4040 1.0944 -1.8487 -0.4086 0.1949 0.0279 CAPM-beta 0.8770*** 1.0526*** 1.0937*** 1.0895*** 1.1077*** -0.2306*** (t-value) -7.3751 2.6890 3.5217 3.7814 4.1749 -6.3950 Idiosyncratic risk 0.6463% 0.8465% 1.2154% 1.0728% 1.1313 -0.4850%*** (F-value) 3.0639 Environmental Multi-factor Alpha 0.2604% -0.4455% 0.4028% 0.2379% -0.7045% 0.9649% (t-value) 0.3604 -0.3748 0.3142 0.1698 -0.5373 0.5689 CAPM-beta 0.8944*** 1.0475** 1.1172*** 1.1588*** 0.9751 -0.0806** (t-value) -5.9133 2.0973 4.7072 5.4294 -0.9612 -2.3640 Idiosyncratic risk 0.6402% 1.0538% 1.1359% 1.2417% 1.1617% -0.5215%*** (F-value) 3.2928 Social Multi-factor Alpha 0.3142% -1.2691% 0.3987% -0.4009% 1.4674% -1.1533% (t-value) 0.4473 -1.3209 0.2724 -0.2610 1.0666 -0.6528 CAPM-beta 0.8787*** 1.0462** 1.0741** 1.1441*** 1.0481* -0.1693*** (t-value) -7.1966 2.4563 2.5921 4.7548 1.7566 4.4915 Idiosyncratic risk 0.6223% 0.8513% 1.2969% 1.3607% 1.2190% -0.5968%*** (F-value) -3.8378 Corporate Governance Multi-factor Alpha 1.6968%** -0.0112% -2.3994%** -1.4967% 0.3431% 1.3537% (t-value) 2.0227 -0.0109 -2.0154 -1.1123 0.2937 0.8267 CAPM-beta 0.8654*** 0.9881 1.0707*** 1.0396 1.1234*** -0.2580*** (t-value) -7.6010 -0.5813 2.9179 1.5276 5.2677 -7.7621 Idiosyncratic risk 0.7433% 0.9134% 1.0549% 1.0923% 1.0353% -0.2920%*** (F-value) 1.9401

6. Robustness Analysis

Different sorting methods and portfolio construction methods are designed to test the robustness of the results.

(26)

25 6.1 Controlling for Other Factors

6.1.1 Controlling for Sector

A sector structure summary of the sample is shown in Figure 26:

Note: All the US companies in the sample (including those ceased operation) are included in the analysis. The classification of the sector is according to Global Industry Classification Standard (GICS).

When looking at the ratings of companies in different sectors, it is clear that some sectors have higher ratings and others have lower ratings (Figure 3). For example, financials sector has the lowest average ESG rating of 36.45 while consumer staples sector has the highest average ESG rating of 63.10. It means high rating portfolio might bias towards stocks within consumer staples sector and low rating portfolio might bias towards stocks within financials sector.

6_{The classification of sectors is according to Global Industry Classification Standard (GICS).}

8% 7% 12% 16% 6% 10% 19% 15% 2% 5%

Figure 2: Sector Composition of Sample energy materials industrials consumer discretionary consumer staples health care financials information technology telecommunication services

(27)

26

Note: The average ratings are calculated using available ratings of companies in the US sample from 2002 to 2013.

So how the results will change if the sector factor is controlled for? Method of double-sorting is used to control for sectors-first the sample is grouped according to sectors7, then in each sector the stocks are sorted according to sustainability ratings. Quintile portfolios are formed by combining the stocks in each sector group. The results of long-short (long top 20% stocks, short bottom 20% stocks) portfolio are shown in Table 5-Control for Sector, detailed information is in Appendix 4.

After sector factor is controlled for, only long-short portfolio created based on corporate governance ratings shows positive alpha; alphas of other long-short portfolios are all negative. The results lack statistical significance.

The CAPM betas are still significant except for long-short portfolio formed based on environmental ratings and the differences in idiosyncratic risk between D1 and

7_{The classification of sector is according to Global Industry Classification Standard (GICS).}

43.74 58.00 55.61 43.22 63.10 44.88 36.45 48.79 41.58 59.79 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 energy materials industrials consumer discretionary consumer staples health care financials information technology telecommunication services utilities energymateri als indust rials consu mer discret ionary consu mer staples health care financi als inform ation techno logy teleco mmun ication service s utilitie s Average ESG ratings 43.74 58.00 55.61 43.22 63.10 44.88 36.45 48.79 41.58 59.79

(28)

27 D5 are also significant.

After controlling for sector, the results do not change much compared to the base case.

6.1.2 Controlling for Historical Beta

From the results presented above, it is found that high rating portfolios tend to have lower market risk (ex-post CAPM beta). So how the results will be if historical beta is controlled before sorting on ratings?

Here the method of double-sorting is used again to control for historical beta. The portfolios are rebalanced every year based on new information on ratings and historical beta. The results of long-short portfolios are presented in Table 5-Control

for historical beta. Detailed regression results are shown in Appendix 5.

Alphas are not significant except for portfolios created based on corporate governance ratings: an annual alpha of 3.54% (significant at 10% level) was found in long-short portfolio created based on corporate governance ratings. The ex-post CAPM beta spreads between D1 and D5 become smaller after double sorting on historical beta, but they are still significant, which indicates that sustainability gives information on market risk on top of historical beta. The idiosyncratic risk spreads are still significant, indicating that controlling for historical beta does not influence the results got regarding the relationship between sustainability and idiosyncratic risk.

6.1.3 Controlling for Market Cap/Sector and Market Cap

(29)

28

finding is that the average rating of stocks increases as the market cap of the stock increases, which means bigger companies tend to have higher scores. A summary of the average ratings of different market cap groups is shown in figure 5:

Note: The average ratings are calculated using the ratings from 2002 to 2012.

It suggests that high rating portfolio is biased towards large caps and low rating portfolio is biased towards small caps8. Bigger companies might have better CSR disclosures and they tend to invest more into improving their CSR performances, so it might be the case that rating companies with different sizes under the same standard is not fair. It is interesting to test how the results will be if the size factor is controlled before creating portfolios based on ratings.

Double-sorting method is used again to control for the factor of size; the portfolios are rebalanced every month according to the new information on company market capitalization and ratings. A triple-sorting method is used to control for both size and sector: stocks are first grouped into different sector groups, and then in each sector group, they are divided into five sub-groups according to market cap. Stocks are then sorted according to sustainability ratings and quintile portfolios are formed. The

8_{It is also indicated by the SMB coefficient of Carhart (1997) regression, in Appendix 3 & 4.}

71.56 48.27 44.09 38.68 34.83 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 Largest cap 2 3 4 Smallest Cap

Largest cap 2 3 4 Smallest

Cap Average ESG rating 71.56 48.27 44.09 38.68 34.83

(30)

29

results of the regressions are shown in Table 5-Control for market cap/Control for

sector and market cap. Detailed information of the results is in Appendix 6 and Appendix 7.

Significant alphas for long-short portfolios created based on ESG ratings (3.50%/3.45%) and corporate governance ratings (4.07%/3.28%) are found in both cases. The significant outperformance is unexpected because in Carhart model the effect of size is already corrected by the SMB factor. Possible explanation is that the sustainability rating is a “mixed” rating of company size and sustainability, which means by sorting on the ratings alone, bigger companies instead of companies which are really more sustainable relative to their sizes are selected 9. Using the method of double-sorting, the size effect can be reduced and the outperformance we get here is the “real” outperformance comes from sustainability.

The CAPM-beta spreads between D1 and D5 are still negative and are all significant, meaning controlling for size before event does not change the relationship between market risk and sustainability.

However, the spreads in idiosyncratic risks between top quintile portfolio and bottom quintile portfolio become insignificant after controlling for size. One possible explanation is that there is a negative relationship between firm size and idiosyncratic risk (Sorescu and Spanjol, 2008), so after the size effect is neutralized, the idiosyncratic risk differences become less prominent.

9_{The results present in Table 5-control for sector and market cap are got by grouping stocks into 5 size groups;}

if companies are grouped into 10 size groups (the size effect will be neutralized even more), the outperformance of D1 against D5 created based on ESG ratings will be 3.85% (significant at 10% level); if companies are grouped into 15 size groups, the outperformance will become 4.04% (significant at 5% level).

(31)

30

Table 5 : Main Results of Value Weighted Long-short Portfolios

This table summarizes the multi-factor alphas and CAPM betas of long-short (value weighted) portfolios created based on different ratings and controlling for different factors. The portfolios are for the period from 2004 to 2013 based on rating information from 2002 to 2011. *, **, *** indicate significance level at 10%, 5%, and 1% respectively.

ESG Environmental Social Corporate

Governance Base case Multi-factor alpha 0.0458% 0.9649% -1.1533% 1.3537% (t-statistic) 0.0279 0.5689 -0.6528 0.8267 CAPM beta -0.2306*** -0.0806** -0.1693*** -0.2580*** (t-statistic) -6.3950 -2.3640 -4.4915 -7.7621 Idiosyncratic risk -0.4850%*** -0.5215%*** -0.5968%*** -0.2920%*** (F-statistic) 3.0639 3.2928 3.8378 1.9401

Control for sector

Multi-factor alpha -1.2058% -2.0000% -1.5195% 0.3890% (t-statistic) -0.6289 -1.1647 -0.8826 0.2304 CAPM beta -0.1580*** -0.0317 -0.1085*** -0.2107*** (t-statistic) -3.9097 -0.9025 -3.0386 -6.4319 Idiosyncratic risk -0.7048%*** -0.4928%*** -0.5516%*** -0.3458%*** (F-statistic) 4.5520 3.2101 3.5958 2.2148

Control for market cap

Multi-factor alpha 3.5028%* 0.1295% 2.3454% 4.0657%** (t-statistic) 1.7912 0.0766 1.1200 2.1105 CAPM beta -0.0995*** -0.0804** -0.0799* -0.1423*** (t-statistic) -2.6254 -2.5097 -1.9296 -3.9028 Idiosyncratic risk -0.0302% -0.2243%*** -0.1351% -0.1234 (F-statistic) 1.0570 1.6204 1.2549 1.2502

Control for sector and market cap

Multi-factor alpha 3.4495%* 0.5169% 2.8229% 3.2801%* (t-statistic) 1.7891 0.3082 1.3961 1.7190 CAPM beta -0.1067*** -0.1179*** -0.1434*** -0.1301*** (t-statistic) -2.8315 -3.6699 -3.6453 -3.5709 Idiosyncratic risk -0.1694% 0.1921%* -0.0234% -0.0759% (F-statistic) 1.3404 1.4092 1.0397 1.1295

Control for historical beta

Multi-factor alpha 0.6980% 1.5236% -0.4100% 3.5404%* (t-statistic) 0.4452 0.9093 -0.2556 1.9482 CAPM beta -0.1433*** -0.0696** -0.0897*** -0.2230*** (t-statistic) -4.2832 -2.1326 -2.6485 -6.2555 Idiosyncratic risk -0.4475%*** -0.4587%*** -0.4882%*** -0.3199%*** (F-statistic) 2.9488 2.9889 3.1798 1.9650

(32)

31 6.2 Equal Weighted Portfolios

In the earlier analysis, portfolios are all constructed using value weighted method. In this section, the results for equal weighted portfolios are presented in Table 6 to test whether different weighting method will change the results.

The figures are different but the main results look similar: no significant outperformance can be found in long-short portfolios created by sorting on ratings, while after controlling for size factor, significant outperformances are observed. In addition, the CAPM beta spreads are still negative in most cases except after controlling for market cap. The spreads in idiosyncratic risks between top quintile portfolio and bottom quintile portfolio are less prominent compared to the results got when portfolios are created using value weighted method. One possible explanation is that when using equal weighted method to construct portfolios, the idiosyncratic risks are diversified more because each stock’s return will take up same weight in the portfolio return while in value weighted portfolio, companies with larger sizes could have bigger impacts on the portfolio’s return and the idiosyncratic risks cannot be diversified as fully as in equal weighted portfolios.

(33)

32

Table 6 : Main Results of Equal Weighted Long-short Portfolios

This table summarizes the multi-factor alphas and CAPM betas of long-short (equal weighted) portfolios created based on different ratings and controlling for different factors. The portfolios are for the period from 2004 to 2013 based on rating information from 2002 to 2011. *, **, *** indicate significance level at 10%, 5%, and 1% respectively.

ESG Environmental Social Corporate

Governance Base case Multi-factor alpha 0.1605% 2.1869% 0.4447% 1.8221% (t-statistic) 0.1002 1.4226 0.2913 1.1960 CAPM beta -0.2782*** -0.0788** -0.1731*** -0.1313*** (t-statistic) -7.4388 -2.4591 -4.9117 -4.3298 Idiosyncratic risk -0.6437%*** -0.1198% -0.3754%*** -0.1005% (F-statistic) 3.4496 1.2513 1.9944 1.1955

Control for sector

Multi-factor alpha 0.6745% -1.1721% -0.1184% -0.0175% (t-statistic) 0.4264 -0.9283 -0.0829 -0.0124 CAPM beta -0.2127*** 0.0108 -0.0844*** -0.0749*** (t-statistic) -6.4334 0.4213 -2.8051 -2.7443 Idiosyncratic risk -0.6414%*** -0.2250%*** -0.3164%*** -0.3406%*** (F-statistic) 3.6930 1.6262 1.8384 1.8826

Control for market cap

Multi-factor alpha 3.9231%** 3.2502%* 3.6092%** 2.5465% (t-statistic) 2.5848 1.8713 2.4159 1.5824 CAPM beta 0.0093 0.1597*** 0.0737*** 0.0026 (t-statistic) 0.3029 4.3651 2.4254 0.0837 Idiosyncratic risk -0.0642% 0.2274%** -0.0358% 0.0792% (F-statistic) 1.1163 1.5049 1.0609 1.1401

Control for sector and market cap

Multi-factor alpha 3.2774%** 3.0810%* 2.8934%* 1.4660% (t-statistic) 2.0705 1.8071 1.8955 0.9312 CAPM beta -0.0561* 0.0888** -0.0065 -0.0314 (t-statistic) -1.8607 2.5925 -0.2228 -1.0552 Idiosyncratic risk -0.1694% 0.1921%* -0.0234% -0.0759% (F-statistic) 1.3404 1.4092 1.0397 1.1295

Control for historical beta

Multi-factor alpha 1.0733% 2.9891%* 1.2290% 3.0414%** (t-statistic) 0.7917 1.8508 0.8600 1.9850 CAPM beta -0.1060*** -0.0005 -0.0510* -0.0728** (t-statistic) -3.6793 -0.0167 -1.6561 -2.4692 Idiosyncratic risk -0.3109%*** 0.0053% -0.0500% 0.0711% (F-statistic) 1.7966 1.0098 1.0920 1.1318

(34)

33

7. Conclusion

The lack of widely accepted rating data is a key issue for the analysis of the relationship between sustainability and stock returns, and earlier studies have got mixed results. In this study, using the database of Thomson Reuters ASSET4 US sample for the period of 2002-2011, under multi-factor model, no evidence is found that sustainability (ESG ratings) could improve stock returns, consistent with the finding of Van de Velde et al (2005) and Dinusha Peiris (2009). However, stocks within the

top quintile of corporate governance ratings generate a positive abnormal return of 1.7%

annually.

Lower market and idiosyncratic risks are found in portfolios with higher sustainability

ratings, even after controlling for historical CAPM beta before event, suggesting a negative

relationship between sustainability and risk. The results of this study confirm the findings of

McGuire et al (1988) and Mishra and Modi (2012).

A further analysis is done by controlling for the factor of market capitalization before creating portfolios based on ratings, and significant outperformances of 3.50% and 4.07% are observed in long-short portfolios created based on ESG ratings and corporate governance ratings respectively. Possible explanation is that the ratings are biased: bigger companies might receive better ratings without being more sustainable relative to their sizes, so the ratings have to be corrected for company size to be comparable. Double-sorting on size and sustainability is a rough way to control for the size factor, but it gives some insights for further research.

(35)

34

size, but the spreads in idiosyncratic risks become less significant, which might be partly attributable to the negative relationship between company size and idiosyncratic risk and partly to diversifications of risk.

The overall conclusion is that sustainability ratings give valuable information on stocks’ market and idiosyncratic risks while no impact of sustainability is found on stock returns; after controlling for size, basing investing decisions on sustainability ratings and corporate governance ratings can generate abnormal returns. Yet whether size is really a bias is not proven, it at least shows that using sustainability rating information together with market capitalization can help to improve the investment returns.

(36)

35

8. References

Barnett M. and Salomon R. (2006). “The Curvilinear Relationship between Social Responsibility and Financial Performance.” Strategic Management Journal 27:1101-1122.

Bouslah K., Kryzanowski L. and M’Zali B. (2013). “The impact of dimensions of social performance on firm risk.” Journal of Banking & Finance, 37 (2013) 1258-1273

Carhart M. (1997) “On Persistence in Mutual Fund Performance.” The Journal of

Finance, Vol. 52, No. 1 (Mar. 1997), 57-82

Clark, Robert. (1998)."Anti-Social Investing." Editor's Note in Dow Jones Investment Advisor (September)

Derwall J. Guenster N. Bauer R. and Koedijk K. (2009). “The Eco-efficiency Premium Puzzle.” Financial Analysts Journal, Vol. 61, No. 2 (Mar. - Apr., 2005), pp. 51-63

Dobler M., Lajili K. and Zeghal D. (2012). “Environmental Performance, Environmental Risk and Risk Management.” Business Strategy and the Environment, 23, 1-17(2014)

Edmans A. (2010). “Does the stock market fully value intangibles? Employee satisfaction and equity prices.” Journal of Financial Economics, 101(2011)621–640 European Commission (2011). “Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions.”

Eurosif (2012). “European SRI Study 2012”

Galema R., Plantinga A., Scholtens B. (2008). “The stocks at stake: Return and risk in socially responsible investmen”. Journal of Banking & Finance, 32 (2008) 2646– 2654

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

Gillan S., Hartzell J., Koch A., and Starks L. (2010). “Firms’ Environmental, Social and Governance (ESG) Choices, Performance and Managerial Motivation.” working

(37)

36

Godfrey, P. C. (2005). “The relationship between corporate philanthropy and shareholder wealth: A risk management perspective.” Academy of Management

Review, 30, 777-798

Gompers P., Ishiiand J., Metrick A. (2003). “Corporate Governance and Equity Prices.” Quarterly Journal of Economics, 118(1), February 2003, 107-155

Goss A., Roberts G. (2011). “The impact of corporate social responsibility on the cost of bank loans.” Journal of Banking & Finance, 35(2011) 1794-1810

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

Keefe J. (2007). “From SRI to Sustainable Investing.” Report.

Kempf A. and Osthoff P. (2007). “The Effect of Socially Responsible Investing on Portfolio Performance.” European Financial Management, Vol. 13, No. 5, 2007, 908–922

Kurtz L. (2005). “Answers to Four Questions.” The Journal of Investing, 14(3): 125-139

Lacy, P., Cooper, T., Hayward, R., Neuberger, L. (2010). “A New Era of Sustainability-UN Global Compact-Accenture CEO Study 2010”

Langbein J. and Posner R. (1980). “Social Investing and the Law of Trusts.” Faculty

Scholarship Series, Paper 490

Li H., Jahera Jr J. and Yost K. (2012). “Corporate Risk and Corporate Governance: Another View.” Managerial Finance, Vol. 39 Iss 3 pp. 204 - 227

McGuire J., Sundgren A. And Schneeweis (1988). “Corporate Social Responsibitliy and Firm Financial Performance.” Academy of Management Journal, 1988, Vol. 31, No. 4, 854-872

McWilliams, A., & Siegel, D. (2001a). “Corporate social responsibility: A theory of the firm perspective.” Academy of Management Review, 26: 117-127

Mishra S. and Modi S. (2013). “Positive and Negative Corporate Social Responsibility, Financial Leverage, and Idiosyncratic Risk.” J Bus Ethics (2013) 117:431–448

(38)

37

Risk: A Meta-Analytic Review.” Business and Society Review, 40(4): 369-396

Peiris D. (2009). “The Relationship between Environmental Social Governance Factors and US Stock Performance.” Working Paper, University of New South Wales

Porter M.E. and Kramer M.R. (2006). “Strategy & Society: The Link between Competitive Advantage and Corporate Social Responsibility.” Harvard Business

Review 84(12): 78-92.

Schueth S. (2006). “Social Responsible Investing in the United States.” Report,First Affirmative Financial Network, LLC

Schwesernotes (2013) CFA Level II Book 2: Financial Reporting and Analysis and Corporate Finance, Environmental, Social, and Governance Factors

Sorescu A. B. & Spanjol J. (2008). “Innovation’s effect on firm value and risk: Insights from consumer packaged goods.” Journal of Marketing, 72(March), 114-132 Statman M. (2000). “Socially Responsible Mutual Funds.” Financial Analyst Journal May/June: 30-39

Van de Velde E., Vermeir W. and Corten F. (2005). “Finance and accounting corporate social responsibility and financial performance.” Corporate Governance, VOL. 5 NO. 3 2005, pp. 129-138

Willard B. (2011). “3 Reasons Banks Fear ESG Laggards.” Sustainability Advantage, blog

(39)

38

9. Appendix

Appendix 1: Description of ASSET4 ESG Ratings10

Pillar Description Indicators

Environmental

The Environmental Performance 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.

Biodiversity Impact, Greenhouse Gas Emissions, Cement CO2 Emissions, CO2 Reduction, F-Gases Emissions, Ozone-Depleting Substances Reduction, NOx and SOx Emissions Reduction, VOC Emissions Reduction etc. environmental products, Energy Footprint Reduction, Environmental R&D Expenditures, Noise Reduction, Hybrid Vehicles, Renewable/Clean Energy Products, Water Technologies, Materials Recycled and Reused Ratio, Toxic Chemicals, Energy Use, Cement Energy Use, Renewable Energy Use, Green Buildings, Energy Efficiency Initiatives, Water Use etc.

Social

The social Performance 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.

Managers Female Male Ratio, Management Equal Opportunity, Work-Life Balance, Family Friendly, Diversity Controversies and Diversity Compliance employee salaries, bonus plan, Generous Fringe Benefits, Employment Awards, Salary Gap, Trade Union Representation, Net Employment Creation, Personnel Turnover, Announced Lay-offs, Key Management Departures, Strikes, Wages or Working Condition

Controversies and Wages or Working Condition Controversies, Injuries, lost days, HIV-AIDS Program, Health & Safety Controversies etc.

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.

Audit committee independence, Audit Committee Expertise, Compensation Committee Independence, Nomination Committee

Independence, Nomination Committee Processes, Nomination Committee Involvement, Board Meetings and Board Attendance.

size of board, background and skills, board diversity, specific skills, Experienced board, Non-Executive Board Members, Independent Board Members, Strictly Independent Board Members, CEO-Chairman Separation, Mandates Limitation Individual Compensation, Highest Remuneration Package, Board Member Compensation, Remuneration Structure, Stock Option Program, Stock Compensation, Long Term Objectives, Compensation Controversies, and Sustainability Compensation Incentives etc.

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.

Accounting Compliance, Accounting Controversies, Non-audit to Audit Fees Ratio, Auditor Independence, Insider Dealings Controversies, Profit Warnings, Pension Underfunding, Stock Option Dilution, Market

Leadership, Employee Satisfaction Improvements, Employee Cost, Employee Productivity, Inventories Management, Anti-competition Compliance, Anti-competition Controversy, Consumer Complaints etc.