The performance of European Socially Responsible and Environmental funds
during the 2008 financial crisis
“Developments such as the construction of the Dow Jones Sustainability Indices and the UN Global Sustainability Goals show an increasing interest to invest with an ethical or environmental focus. Socially Responsible and Environmental funds offer a platform to do this. Whether investors pay a price for these ethical investments is widely, yet inconclusively discussed. In my research, I compare European Environmental, Socially Responsible and Conventional fund groups during and surrounding the 2008 financial crisis. I find that Environmental funds perform best in the periods before and after the crisis, followed by respectively other Socially Responsible and Conventional funds. During the crisis an opposite effect is shown. I elaborate on the performance of Socially Responsible Investing by measuring the impact of a Negative, Positive or Best-in-Class screening method on performance, but find insignificant results.”
Name: Inga Petersen
Student number: 10366407
Track: Economics and Finance
Field: Finance - Socially Responsible Investing Number of credits thesis: 12
Statement of Originality
This document is written by Inga Petersen who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of Contents
1. Introduction 4
2. Literature review 7
2.1 Socially Responsible Investing 7
2.1.1 Definition 7
2.1.2 Positive, Negative and Best-in-Class screening 7
2.2 Socially Responsible fund performance 9
2.2.1 Disadvantages of Socially Responsible Investing 9
2.2.2 Advantages of Socially Responsible Investing 9
2.2.3 Socially Responsible and Environmental fund performance 10
2.3 Assessing a fund’s performance 12
3. Hypotheses 13 3.1 Fund performance 13 3.2 Screening strategies 14 4. Research Design 15 4.1 Methodology 15 4.1.1 Fund performance 15 4.1.2 Screening strategies 16 4.2 Data collection 17
4.2.1 Dataset construction for fund performance 17
4.2.2 Financial data 17
4.2.1 The 2008 financial crisis 18
4.2.1 Dataset construction for screening strategies 18
5. Results 19 5.1 Fund performance 19 5.1.1 Summary 19 5.1.2 Effects 20 5.1.3 Discussion 21 5.2 Screening strategies 23 5.2.1 Summary 23 5.2.2 Effects 25 5.2.3 Discussion 26 6. Conclusion 28
7. Discussion and Limitations 29
1. Introduction
Socially Responsible Investments (SRI) are used as an instrument to deploy private investments for the purpose of achieving a social goal. This way investors are able to combine the financial benefit of traditional investing, with the ethical benefit of investing in socially responsible organizations
(Holzhauer, 2013). However, several studies suggest that investors pay a price for these ethical investments. Amongst others, Renneboog, Ter Horst, & Zhang (2008a) argue that the ethical incentive of SRI investments ensure that investors seem prepared to hand in some financial performance in return. They also report a decreased sensitivity to past returns of the money that flows into SRI funds, compared to the money flows of conventional funds (Renneboog, Ter Horst, & Zhang, 2011).
On the other hand, several articles mention an outperformance or insignificant research results when comparing SRI funds to conventional ones. Shank, Manullang, & Hill (2005) find that SRI mutual funds seem to outperform conventional mutual funds in the long run (10 years), but find no difference in the short run (3-5 years). Statman (2000) describe an outperformance of SRI mutual funds compared to conventional ones in the United States, however this is not statistically significant. These results suggest that such a price for SRI investments doesn’t exist.
Most of the research revolving around Socially Responsible Investments is focused on funds classified as SRI in general, without making a distinction between the different categories. However, the past decade has seen a rising interest in Environmental or Green funds in particular. This rising interest is in line with the numerous climate change initiatives seen around the world, such as the ‘Paris Agreement’ initiated by the United Nations and the ‘European Climate Change Program’ initiated by the European Union. The United Nations’ ‘Environment Program’ includes global agreements concerning, for example, the containment of global warming by reducing CO2 emission,
providing clean energy, building sustainable cities and communities, and establishing sustainable consumption and production
(https://www.unenvironment.org/explore-topics/climate-change/about-climate-change/climate-change-initiatives-and-partnerships). As seen in, amongst others, the Dow Jones Sustainability Indices, investors are more driven to invest their money in sustainable firms and initiatives (http://www.sustainability-indices.com/). Environmental or Green funds provide a way to do invest ethically.
Since the research on Environmental funds and its financial performance is limited, my thesis will specifically target these funds. The purpose of my research is to compare the risk-adjusted returns of both Socially Responsible funds in general, as well as Environmental funds as a subcategory, to that of Conventional funds. My research will focus on Europe, meaning that a
distinction will be made on where the fund is situated, not on where the funds’ investments are made. An additional comparison will be made for the performance of all three fund groups in both a crisis- and a non-crisis period, to investigate whether this makes a difference.
Furthermore, this study elaborates on Socially Responsible Investment funds and distinguishes three portfolio composing strategies: negative, positive and best-in-class. These different strategies will be taken into account during my research, to find out whether there’s a difference in performance dependent on what strategy is implemented by the fund. This possible performance difference due to the various strategies will be extended to measure the impact of the 2008 financial crisis.
The research question set out to be answered in this thesis is the following: “How does the performance of Socially Responsible Investment funds, and in particular
Environmental funds, compare to that of Conventional funds in Europe?”. To elaborate further on the impact of the state of the economy, and on the different implementations of Socially Responsible Investing, two sub-questions are added: “What is the influence of the 2008 financial crisis on the performance of all three fund groups?”, and “What is the impact of the different screening strategies implemented by Socially Responsible Investment funds on its performance?”.
Previous research has been done on the performance of Socially Responsible funds compared to Conventional funds (Renneboog et al., 2008a; Shank et al., 2005; Statman, 2000). Additionally, Nofsinger and Varma (2014) researched SRI performance for US funds in times of crisis, and extended their research to include the different screening criteria and strategies. Furthermore, the performance of Green funds in particular has previously been compared to that of Conventional funds (Ibikunle & Steffen, 2017; Muñoz, Vargas, & Marco, 2014; Silva & Cortez, 2016). This relative performance has also been extended to crisis periods (Chung, Lee, & Tsai, 2012; Lesser, Rößle, & Walkshäusl., 2016; Muñoz et al., 2014).
My research adds to previous literature in four ways. Firstly, the above mentioned research on the performance of Socially Responsible funds and the influence of its strategy is all focused on US funds. Previous research incorporating the different strategies did not have a specific focus on Europe. Secondly, the sample of European Green funds used for this research is more recent than previous samples. The increased interest in and growth of Green funds has led to new
developments, that are incorporated into my sample. Furthermore, many of the Green funds used in previous research no longer exist. Thirdly, the samples used for European Socially Responsible and Conventional funds are larger than in former research, so that the comparison can be made based on a more extensive dataset. Lastly, the influence of the different strategies on SRI fund
extension, not only the question if Socially Responsible Investing pays off is answered, but also the question which implementation of SRI has the largest pay-off, both in good and bad states of the economy.
Because Socially Responsible and Green funds are designed to invest in programs that improve certain social difficulties, the research on these has a societal as well as a financial relevance. A comparable or even positive risk-adjusted return of these ethical funds, compared to conventional ones, would provide not only a social, but also a financial incentive for investors to fund these investments. This would make investing in an ethical manner beneficial to investors that are purely looking to maximize their profit, making it more attractive to follow such a strategy.
This thesis is constructed as follows: I will start by discussing the relevant literature on Socially Responsible and Environmental Investing and the screening criteria. Afterwards, I will review former research on their relative performance to Conventional funds. Subsequently, the different methods of assessing a fund’s performance will be explained, leading up to the Carhart Four Factor-Model used in this research. From the literature reviewed I will derive a three of hypotheses in line with the theory. The subsequent part starts with the data collection process and methodology, and continues by describing and interpreting the results. In the final part I will review the results, answer my research question and will discuss the shortcomings of the research.
2. Literature review
2.1 Socially Responsible Investing
2.1.1 Definition
Socially Responsible Investing (SRI), also referred to as Sustainable Investing or Responsible Investing, has risen substantially in the past few years. The Global Sustainable Investment Review (2016) states that the amount of Assets under Management (AuM) classified as socially responsible has increased with 25% between 2014 and 2016, resulting in a total of $22.89 trillion worldwide. This equals a quarter of all assets under professional management. Also stated in the Global Sustainable Investment Review (2016) is the 12% increase of the amount invested in Europe between 2014 and 2016, using socially responsible strategies. This rise has led to a total amount of $12.04 trillion, or 53% of all AuM. Environmental Investments have has seen a growth of +146% between 2013 and 2015 in Europe alone, to a total amount of assets worth €145 billion in 2015 (Eurosif, 2016). Eurosif offers the following definition for Sustainable and Responsible Investment, the term they use for Socially Responsible Investing:
“Sustainable and Responsible Investment (”SRI”) is a long-term oriented investment approach, which integrates ESG1 factors in the research, analysis and selection process of securities within an investment portfolio. It combines fundamental analysis and engagement with an evaluation of ESG factors in order to better capture long term returns for investors, and to benefit society by influencing the behavior of companies.” (Eurosif, 2016, p.9).
2.1.2 Positive, Negative and Best-in-Class screening
Mutual Fund managers can compose a Socially Responsible portfolio by using several screening strategies. Based on existing literature, the most common screening strategies are Positive, Negative and Best-in-Class (Kempf and Osthoff, 2007; Renneboog et al., 2008b). Negative or Exclusionary screening was used for 44% of all European Socially Responsible Investments in 2015, and is the most popular screening strategy (Eurosif, 2016). With this strategy certain industries or firms are excluded from a portfolio or fund, because of their negative environmental or societal impact. Some examples are companies involved in tobacco, arms, adult entertainment, nuclear power, or animal testing.
A positive strategy means that funds include firms that actively pursue a positive or neutral change, thus they only include the firms that score the highest in the field of sustainability or social responsibility. This performance is assessed by rating firms from most to least ethical, based on a number of ESG criteria (Eurosif, 2016; Kempf and Osthoff, 2007).
Best-in-Class is a form of Positive screening, with the difference that the best scoring companies in any industry are included in a Best-in-Class screened portfolio. It is therefore possible to see a company involved with for instance the tobacco, fossil fuel or arms industry included in a Best-in-Class screened fund (Eurosif, 2016; Renneboog et al., 2008b). The Dow Jones Sustainability Indices are composed using the best-in-class strategy (http://www.sustainability-indices.com/).
2.2 Socially Responsible fund performance
2.2.1 Disadvantages of Socially Responsible Investing
Former literature research reviews various reasons as to why Socially Responsible funds would both over- and underperform a benchmark. Becchetti, Ciciretti, Dalò, & Herzel (2015) mention three different costs associated with Socially Responsible funds, that don’t exist in Conventional funds. One of these costs is gathering the characteristics of stocks to examine whether they are qualified as SRI stocks. This requires additional work for the fund’s investment managers.
The following cost is that of lesser diversification advantages, especially for SRI funds that follow an exclusionary screening strategy. According to Berk & DeMarzo (2014, p. 332) firms are subject to both systematic and idiosyncratic risks. Systematic risks concern the whole economy, while idiosyncratic risks are firm-specific, and can be diversified away if a portfolio is sufficiently broad. An example is an airline where the employees participate in a strike, after which the airline loses profits. If a portfolio is diversified enough, the risk of this influencing overall portfolio returns is mitigated. When a fund however excludes entire industries from its portfolio, the idiosyncratic risks cannot fully be diversified, which leads to a higher total risk. The higher risk level of exclusionary SRI funds is not compensated by a higher return. This effect is broadly described in literature: amongst others, Renneboog et al. (2008a) and Barnett and Salomon (2006) both describe that the risk-adjusted return of an SRI fund in relation to a Conventional fund is lower.
The last cost described by Becchetti et al. (2015) occurs when a firm previously classified as Socially Responsible no longer meets the SRI criteria, and the SRI fund managers have to sell the firm’s stock, independent of the price or the market conditions at that time.
2.2.2 Advantages of Socially Responsible Investing
Several articles argue that firms with a high concern for Corporate Social Responsibility are more financially stable than firms that are not. Amongst others, Innes and Sam (2008) discuss how Environmental Firms are less likely to incur costs because of demonstrations, boycotts, lawsuits and regulations concerning environmental, governance or social practices. Nofsinger and Varma (2014) argue that firms with for example high environmental values are less likely to be affected by a catastrophe concerning pollution. This line of reasoning can be extended to, for instance, firms with high governance standards and the impact of a strike, or with high social values and a restriction on animal testing.
Renneboog et al. (2011) argue that investors in SRI funds are less sensitive to negative past returns than investors in Conventional funds, so the money flows are less impacted by the generated returns of the firm. However, for Environmental funds opposing results are found: here the money flows seem to be more sensitive to past returns. Renneboog et al. (2008a) already argued that investors not only value financial rewards, but have an extended utility function that includes ethical and social incentives. Additionally, Sauer (1997) describes that companies with high social and ethical values attract more loyal shareholders. Furthermore, high Corporate Social Responsibility standards seem to signal that a firm is under good management and is trustworthy (Fombrun and Shanley, 1990; Fisman, Heal and Nair, 2006). These findings could mean that investors in SRI funds are less likely to retrieve their funds when the returns are lower than usual.
2.2.3 Socially Responsible and Environmental fund performance
Previous research done in the field of Socially Responsible Investing (SRI) and its performance opposed to that of Conventional Mutual funds is not always conclusive. Belghitar, Clark and Deshmukh (2014) describe a significant underperformance of SRIs, while Clark, Feiner, and Viehs (2015) found mainly positive results. Both Cortez, Silva, & Areal, (2012). Basso & Funari (2014) found no significant difference between European socially responsible funds and a conventional as well as a socially responsible benchmark. However, in an article written by Friede, Busch & Bassen (2015), over 2000 empirical studies on the financial performance of firms with high SRI standards were combined to provide an overview of all the results. The firms with high SRI standards significantly outperformed firms with low standards financially in 62.6% of the studies. Only 10% of all studies reported an underperformance of the SRI firms. In addition, the studies with an Environmental focus were reviewed separately. Of these, 58.7% reported that firms with high Environmental standards outperformed regular firms. Only 4.3% of the studies showed a negative relation.
Other research investigated the performance of funds and firms with high environmental values in particular. Derwall, Günster, Bauer, & Koedijk (2004) performed a study on the ‘eco-efficiency’ of firms, or the value they add through the production of goods compared to pollution caused by that production. They found a significantly better performance of the composed portfolio with the most eco-efficient stocks. They found that the return of the portfolio selected using a best-in-class strategy performed 6% better than the portfolio that scored the lowest on eco-efficiency. Additionally, Yadav, Han, & Rho (2015) argue that when companies are concerned with enhancing the sustainability of their activities, this is perceived as good news by investors. The investors react to this in the same way as when other good news is communicated by a firm, leading to significantly
improved ‘standardized cumulative abnormal returns’ (SCARS). On the other hand, Silva and Cortez (2016) find that Environmental funds underperform Conventional funds in the United States, but are increasing in Europe. They however also find evidence that the underperformance decreases when the economy is in a bad state.
Nofsinger and Varma (2014) also investigated whether socially responsible mutual funds perform better than regular mutual funds in times of crisis. This especially seems to be the case for funds where a best‐in‐class screening is used, since an exclusionary screening method leads to a smaller number of represented industries and thus more exposure to idiosyncratic risk (Berk & De Marzo, 2014). They argue that the relatively good performance of SRI funds in times of crisis could be explained by the SRI components of these funds making them a more attractive and stable investment in times of crises. Also, investors looking for downside protection would prefer investing in mutual funds despite their usual underperformance, because they outperform conventional investments during recessions (Glode, 2011).
2.3 Assessing a fund’s performance
For this thesis, three methods of measuring a fund’s performance are discussed: Jensen’s alpha and the models proposed by Fama and French and Carhart.
In 1968, Jensen introduced a method of assessing a fund’s historical excess abnormal performance, based on the Capital Asset Pricing Model (CAPM). CAPM is used to estimate the betas and the returns associated with a stock, by assuming a stock’s return is only influenced by its sensitivity to excess market risk. Jensen offered an expansion, by adding a variable to measure the return that cannot be explained by its systematic risk. This is called Jensen’s alpha. If Jensen’s alpha has a positive value, it means that a stock performed better than it should have based on its systematic risk, and vice versa. When used for a portfolio, Jensen’s alpha measures a portfolio managers’ ability to ‘beat the market’.
𝐽𝑒𝑛𝑠𝑒𝑛′𝑠 𝑎𝑙𝑝ℎ𝑎:
𝑖 = 𝑟𝑖− [𝑟𝑓+(𝑟𝑀 – 𝑟𝑓) +].2
Fama and French (1995) argued that a stock is not only sensitive to the excess market risk, but also to two additional factor loadings. They proposed a three-factor model to include a stock’s sensitivity to two additional factors: a firm’s “market equity, stock price times shares outstanding” (p. 131), and a firm’s “book-to-market-equity …, the ratio of book equity to market equity” (p. 131). The first factor concerns the size of a firm. Fama and French argue that stocks of small corporations perform better than those of large ones. The second factor concerns a firm’s stock price relative to its book value, and makes a distinction between ‘growth stocks’ (low book-to-market value) and ‘value stocks’ (high book-to-market value). Value stocks seem to perform better than growth stocks.
The last expansion is introduced by Carhart (1997), who proposes a fourth risk factor to expand the three-factor model to a fourth-factor model. This is the “one-year momentum factor” (p. 61) proposed by Jegadeesh and Titman in 1993. This factor captures the phenomena that well-performing stocks in the past year will frequently perform well in the near future, and vice versa. The Carhart four-factor model, being the most elaborated model to assess a fund’s performance, will be used in this thesis.
𝐶𝑎𝑟ℎ𝑎𝑟𝑡: 𝑟𝑖− 𝑟𝑓 =+1(𝑟𝑀– 𝑟𝑓) +2𝑆𝑀𝐵 +3𝐻𝑀𝐿 +4𝑀𝑂𝑀 +.3
2
3. Hypotheses
3.1 Fund performance
I expect the risk-adjusted returns of both the European SRI and Environmental funds to be not significantly different than those of the conventional benchmark in times of non-crisis and higher in times of crisis. As discussed earlier in this thesis, the consensus in previous literature seemed to be that there is a non-negative performance associated to the SRI and Environmental firms (Friede, Busch & Bassen, 2015; Derwall, Günster, Bauer, & Koedijk, 2004). Therefore, my first two hypotheses follow:
H1a: The European Socially Responsible Investment funds will not significantly outperform or underperform the Conventional mutual funds in non-crisis periods.
H1b: The European Environmental funds will not significantly outperform or underperform the Conventional mutual funds in non-crisis periods.
A number of arguments as to why the SRI funds would perform relatively better in times of crisis were discussed. These include the premise that firms with high Corporate Social Responsibility values are more financially stable, and induce more loyalty under their investors. Furthermore, SRI investors seem to be less sensitive to negative past returns than regular investors. On top of the above, SRI firms are less prone to costs associated with for instance lawsuits, regulations, lobbyists and demonstrations. For Environmental funds many of the above arguments are also applicable, apart from the lesser sensitivity to past returns. However, an additional reason as to why
Environmental firms might outperform Conventional ones is offered: the improvement of firms’ sustainability is perceived by investors as ‘good news’, leading to higher returns. Hence, my second hypotheses:
H2a: The European Socially Responsible Investment funds outperform the Conventional mutual funds in crisis periods.
3
MOM (Momentum) measures the difference in the returns between a portfolio with badly performing firms and a portfolio with well performing firms. 4 measures stock I’s sensitivity to the Momentum factor (Carhart,
H2b: The European Environmental funds outperform the conventional mutual Funds in crisis periods.
My third hypothesis concerns the difference in performance between Socially Responsible Investment funds in general, and Environmental funds in particular. As mentioned above, there are arguments in favor of both. However, the past decade has seen a significant increase in mainly Environmental funds, as well as increased worldwide Climate Change initiatives. These could lead to restrictions on and increased costs associated with, amongst others, pollution, and extra incentives for firms to perform in an environmentally friendly way. Therefore, I expect the Environmental funds to outperform SRI funds focused on the other criteria, namely Social and Governance.
H3: The Environmental funds will outperform the other SRI funds.
3.2 Screening strategies
Since the different screening strategies seem to influence the performance of a SRI fund, as discussed in the literature, I expect to see a difference in performance for the funds following a different strategy. Both a Positive and a Negative screening strategy struggle from a loss in
diversification benefits, since they select or reject firms from certain industries. This disadvantage of lesser diversification is not as evident within firms that use a Best-in-Class screening strategy. In that case, the firms that perform the best socially and environmentally in any industry are incorporated in a fund, creating a more diversified portfolio. Therefore, I expect the Best-in-Class screened funds to perform the best for both Socially Responsible as for Environmental funds.
H4a: The Best-in-Class screened funds will outperform the Positively and Negatively Screened funds for the European Socially Responsible funds.
H4b: The Best-in-Class screened funds will outperform the Positively and Negatively Screened funds for the European Environmental funds.
4. Research Design
4.1 Methodology
4.1.1 Fund performance
For the first three hypotheses, I will follow the model proposed by Nofsinger and Varma (2014) to estimate the returns of the SRI and Environmental funds in both crisis and non-crisis periods. As mentioned in the literature, the Carhart four-factor model will be used as a basis, since this is the most extended model to estimate fund performance to date. I will estimate the alphas for the SRI funds and the Environmental funds separately, by use of the same model:
𝑟𝑡− 𝑟𝑓,𝑡 =+𝐶𝐷𝐶,𝑡+1(𝑟𝑀,𝑡– 𝑟𝑓,𝑡) +2𝑆𝑀𝐵𝑡+3𝐻𝑀𝐿𝑡+4𝑀𝑂𝑀𝑡+𝑡.4
This regression will be performed three times: first for the Environmental fund group, secondly for the general SRI fund group, and thirdly for the Conventional fund group. I will exclude the Environmental funds from the list of SRI funds to prevent autocorrelation, and in order to make a clear comparison between Environmental funds and different kinds of SRI funds. The variables of interest for my research are mainly the alpha and the crisis alpha. The alpha measures the excess return of the fund group, as explained while discussing the methods of measuring fund performance in 2.3, while the crisis alpha measures the excess return in times of crisis. When all three regressions are performed, the alphas can be compared to see whether there’s a performance difference between the three fund groups. Through the crisis alpha it can be shown whether the fund group outperforms or underperforms in times of crisis. Additionally, the crisis alphas of the three fund groups can also be compared to show which one of the fund groups performs best in a crisis period.
4 𝑟
𝑡= equally weighted average return on the fund group at time t; 𝑟𝑓,𝑡= risk-free rate; = excess return of
the fund group; 𝐶 = alpha in crisis periods; 𝐷𝐶,𝑡=a dummy variable with a value of 1 if time t is in a crisis
period, and a value of 0 if otherwise; 1= sensitivity to market risk (or systematic risk); 𝑟𝑀,𝑡= return on the
market; 2= sensitivity to the size factor (SMB); 3= sensitivity to the book-to-market factor (HML); 4= sensitivity to the momentum factor (MOM); 𝑡= error term (Berk & DeMarzo, 2014; Fama & French, 1995;
4.1.2 Screening strategies
The last hypothesis concerns the difference in performance of Environmental and SRI funds for the various strategies, namely Positive, Negative or Best-in-Class. The regression I will apply in order to test this hypothesis is used in various former articles (Renneboog, Ter Horst, & Chendi, 2008b; Capelle-Blancard and Monjon, 2012; Lee, Humphrey, Benson, & Ahn, 2010):
𝑟𝑖,𝑡− 𝑟𝑓,𝑡=0+ γ1𝑃𝑂𝑆𝑖+ γ2𝑁𝐸𝐺𝑖+ γ3𝐵𝐼𝐶𝑖+ ∑ γ𝑗𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖+ 𝑡.5
This second regression will be performed four times. First for the entire time span ranging from January 2004 until April 2018. Afterwards, three samples will be constructed, based on the time frame. One for the pre-crisis time span (January 2004 until September 2007), one during the crisis (October 2007 until March 2009), and one for the post-crisis time span (April 2009 until April 2018). This way, the differences in performance and the influence of the strategies can be measured in different states of the economy.
5In this regression 𝑃𝑂𝑆
𝑖, 𝑁𝐸𝐺𝑖 and 𝐵𝐼𝐶𝑖 are all dummy variables for the various strategies for fund i.
𝑃𝑂𝑆𝑖 assumes a value of 1 if fund i uses a positive screening method and 0 if otherwise; 𝑁𝐸𝐺𝑖 assumes a value
of 1 if fund i uses a negative screening method and 0 if otherwise; 𝐵𝐼𝐶𝑖 assumes a value of 1 if fund i uses a
Best-in-Class screening method and 0 if otherwise; if a fund uses neither of these strategies, they follow a Combined strategy, 𝐶𝑂𝑀𝑖. This dummy is excluded to prevent a dummy trap. ∑ 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖 is a set of control
variables to account for differences between funds, following Capelle‐Blancard and Monjon (2012). The first two variables concern the characteristics of a fund: 𝐴𝐺𝐸𝑖 for how long the fund has existed; 𝑆𝐼𝑍𝐸𝑖 for the total
amount of assets the fund has in euro’s. The last three variables are dummy variables, concerning the style the fund uses for its investments: 𝐺𝐿𝑂𝐵𝐴𝐿𝑖 assumes a value of 1 if the fund invests internationally (not just in
Europe); 𝐵𝑂𝑁𝐷𝑖 assumes a value of 1 if it concerns a bond fund; 𝐸𝑄𝑈𝐼𝑇𝑌𝑖 assumes a value of 1 if it concerns a
fund investing in equity. 1, 2 and 3 estimate the coefficients of respectively 𝑃𝑂𝑆𝑖, 𝑁𝐸𝐺𝑖 and 𝐵𝐼𝐶𝑖. 𝑗 with j =
4.2 Data collection
4.2.1 Dataset construction for fund performance
I collected my data on which funds are considered Socially Responsible or Environmentally focused by using several sources. To start, I followed the approach used by Silva & Cortez (2016) and used the website www.yoursri.com to collect all European funds with a focus on sustainability. This website explains their sustainability criterion as follows: “A fund which has adopted investment policies that are sensitive to ethical and social concerns, and bases its investment selections on criteria such as environmental issues, religious beliefs, inclusive employee policies, etc. This is based on Lipper’s classification.” (https://yoursri.com/). I included all asset classes in the selection
(Alternatives, Bond, Commodity, Equity, Mixed Assets, Money Market, and Other), focused on Mutual funds and selected all funds that were domiciled in Europe.
After extracting a list of all SRI funds located in Europe, I manually made a division between the funds that focused solely on Environmental Criteria, and funds that focused on SRI as a whole or on one of the other sub-categories (Governance or Social). I made this distinction by looking at the fund’s prospectus, and found a total of 35 fully Environmental funds. The number of SRI funds with a broader orientation ended up at 441. Furthermore, I added the fund’s strategy (Positive, Negative, Best-in-Class or Combined), provided that they were explained in the description. Lastly I added the fund’s Inception Date, Geographic focus, Total Assets and Asset Class.
To further expand the sample for European Environmental funds, I looked at the list provided by Muñoz et al. (2013). I updated their list by excluding all funds that died between 2013 and now, leading to an additional 47 funds. Therefore, the total number of Environmental funds in my sample adds up to 82.
To compose a representative dataset for the Conventional fund group, five well-represented domicile countries of the funds in the Socially Responsible and Environmental datasets were used as a benchmark (Germany, France, Austria, Denmark and Belgium). Together, these five countries made up more than 50% of the SRI and Environmental datasets. For these countries, the Mutual Fund universe was retrieved from DataStream. This led to a total of over 13.000 funds.
4.2.2 Financial data
The monthly returns for the general SRI funds, Environmental funds and Conventional funds were retrieved from DataStream. I selected the period between January 2004 and April 2018 as a time
frame. Prior to 2014 the sample for Environmental funds was smaller than 30, which would complicate making generalized statements about the sample. The most recent data on the SMB, HML and MOM-factors that could be retrieved was for April 2018; hence this became the end of the time period.
In order to perform Carhart’s Four-Factor model on all three datasets, I retrieved Fama and French’s SMB- and HML-factors and Carhart’s MOM-factor from the Kenneth R. French database. The excess market return and risk-free rates were also provided in this database. I selected the monthly European dataset, and used the same time period as for the funds’ returns. The risk-free rate provided in the dataset was used to transform the funds’ returns into excess returns.
4.2.3 The 2008 financial crisis
In order to determine the time span of the financial crisis, the approach of looking at a fall in the S&P 500 Index was followed (Nofsinger & Varma, 2014; Muñoz et al., 2013; Silva & Cortez, 2016). With this approach the financial crisis can be narrowed down to the period between October 2007 and March 2009. During the months within this time span the dummy variable 𝐷𝐶,𝑡 was given a value of
1, and a value of 0 in the remaining months. The crisis dummy was added to the return- and factor loading-data in Excel, after which the Excel file was uploaded to Stata to perform the first regression.
4.2.4 Dataset construction for screening strategies
For the second regression meant to test the influence of the various screening strategies on the Socially Responsible fund group, a smaller sample was created. As mentioned, the screening
strategies were added manually from the information provided on yoursri.com. This information was provided for a total of 144 funds, of which 3 were excluded because they originated in the past few months.
The monthly returns of these 144 funds were divided into three four frames. One covering the entire period between January 2004 and April 2018, a second covering January 2004 to September 2007 (pre-crisis), a third for the period between October 2007 and March 2009 (crisis), and a fourth for the period between April 2009 and April 2018 (post-crisis). The average excess returns of each fund were calculated for all four time frames, to provide its mean performance during that time frame. These average excess returns were supplemented with the fund’s strategy, type and characteristics in Excel. After collecting and merging all the data, the Excel File was uploaded to Stata in order to perform the regression.
5. Results
5.1 Fund performance
5.1.1 Summary
A summary of data and the results for the first regression, using the Four-Factor Model, is provided below. Table 1.1 shows the average monthly excess returns.
Table 1.1. Summary of the Monthly Excess Returns of the Environmental, SRI and Conventional fund groups
Variable Obs Mean Std. Dev Min Max
Renv 172 .52 4.09 -17.26 12.44
RSRI 172 .42 2.82 -10.93 9.17
Rconv 172 .31 2.26 -8.26 13.69
The variables shown in this summary are the weighted averages of the monthly excess returns of the Environmental, SRI and Conventional fund groups, denoted by Renv, RSRI and Rconv. The summary shows the
number of observations (months) for which the regression is performed, the mean and the standard deviation of the averaged returns, and the lowest and highest returns measured. The data is given in percentage points, and is rounded to two decimals.
The summary in Table 1.1 shows an evident difference in the average excess returns of the three fund groups. The Environmental fund group performs best, with an annualized excess return of 6.42%. The SRI fund group comes in second with an annualized excess return of 5.16%, and the Conventional fund group performs worst, with an annualized excess return of 3.78%. The differences between the fund groups are also tested. The null hypotheses stating that the difference between the average returns of the fund groups is zero can however not be rejected for any of the
differences. The values for the t-tests are .25 (Renv-RSRI), .59 (Renv-Rconv) and .42 (RSRI-Rconv). The
Environmental fund group shows the highest standard deviation (4.09%), and the biggest difference between the minimum and the maximum monthly excess return (29.70%). This indicates that SRI funds, and especially Environmental funds performed better than Conventional funds in the period between January 2004 and April 2018, but are also the most volatile.
Table 1.2. Summary of the remaining inputs used for the Four-Factor Model-regression
Variable Obs Mean Std. Dev Min Max
Crisis 172 10.47 30.70 0 1 RMKT 172 .64 5.21 -22.02 13.67 SMB 172 .20 1.80 -4.98 4.88 HML 172 .05 2.13 -4.35 7.42 MOM 172 .81 3.60 -26.24 10.26 RF 172 .10 0.14 0 0.44
This table summarizes the other variables used to perform the Four-Factor Model-regression, extended with the crisis dummy. The summary shows the percentage of the observations performed in a crisis period (18 months). RMKT is the excess market return, SMB, HML and MOM are respectively the size-, value- and momentum-factors,
and RF is the risk-free rate. For the crisis variable the minimum and maximum values are 0 and 1, since this is a dummy variable. The data is given in percentage points, and is rounded to two decimals.
5.1.2 Effects
Table 2. Results of the Four-Factor Model-regression on fund performance
Excess Returns Environmental SRI Conventional
Crisis -.0381*** (-3.88) -.0305*** (-4.60) -.0227*** (-4.18) RMKT .0119 (0.17) .0154 (0.33) .0088 (0.23) SMB .6394*** (4.02) .4332*** (4.03) .3364*** (3.83) HML .1955 (1.17) .0929 (0.81) .0440 (0.48) MOM .0082 (0.09) -.0257 (-0.41) -.0354 (-0.70) Alpha .0076** (2.37) .0066*** (3.02) .0050*** (2.78) R-squared .2015 .2341 .2058
In this table the results of the Four-Factor Model-regression with the added crisis dummy are shown. The dependent variable is the Monthly Excess Return, constructed from an equally weighted portfolio for the Environmental, SRI and Conventional fund groups. Crisis is the crisis alpha, and shows how the fund groups perform in a period of crisis as opposed to in a period of non-crisis. RMKT is the Excess Market Return, SMB, HML
and MOM are Fama-French and Carhart’s additional factor loadings. The coefficients (beta’s) for these factor loadings show the fund groups’ sensitivity to the Excess Market Return and the Size-, Value- and Momentum-factors. The Alpha is the constant, and shows the performance of the fund groups that is not explained by the factor loadings. The values for the t-tests for all coefficients are added between brackets, and the star denotes the significance of a coefficient on a 10% level (*), on a 5% level (**) or on a 1% level (***). The R-squared is
As seen in Table 3, all three fund groups perform worse during crisis periods than during non-crisis periods, with a statistical significance of 1%. The crisis alpha is highest for Environmental funds, indicating that their performance is the most affected by the Financial Crisis. The SRI fund group also seems to be impacted more by the crisis than the Conventional fund group.
In non-crisis periods the opposite effect is shown: the ‘regular’ alpha (constant) is the biggest for Environmental funds, followed by SRI funds and then Conventional funds. This is
consistent with the higher returns described in the data summary for Environmental and SRI funds, as opposed to that of Conventional funds. The alphas are significant at a 5% level for Environmental funds, and at a 1% level for SRI and Conventional funds, meaning that they all show a performance that cannot fully be explained by their sensitivity to the factor loadings.
The betas to the market return are positive for all three fund groups. The Size- and Value-factors are also positive for all three fund groups. The momentum-factor only has a positive coefficient for the Environmental fund group. However, only the Size-factor is significant (1%).
5.1.3 Discussion
The differences in the average excess returns as seen in the data summary are consistent with most of the literature. As reported in Table 1.1, the Environmental fund group has the highest excess returns, followed by the other SRI funds. The returns of the Conventional fund group were the lowest. Friede, Busch & Bassen (2015) compared over 2000 empirical studies on financial performance of firms with high socially responsible practices, and found that both Socially Responsible and Environmental firms performed better than conventional ones in the majority of the studies.
My findings could mean that the benefits of investing in firms with high socially responsible and environmental values outweigh their disadvantages. A possible cause is the higher financial stability and investor loyalty of firms with high SRI and Environmental values, as described by Sauer (1997), Renneboog et al. (2011). Another explanation could be the lower costs associated with boycotts, lawsuits, regulations and pollution incurred by these firms (Sam, 2008; Nofsinger and Varma, 2014). The lesser diversification for SRI and Environmental funds possibly explains the higher volatility of these fund groups. Since the funds are limited in the firms they can invest in, and in some cases exclude entire industries, they are more sensitive to idiosyncratic risk, thus have higher overall risk levels.
The underperformance of all three fund groups in times of crisis is consistent with the performance of the stock markets during the crisis, that all experienced a fall. Since many firms
incurred big losses, funds investing in those firms also experienced lower returns. However, that the Environmental and the SRI fund groups experienced a larger negative effect than the Conventional one is contradictory to other articles. This is contradictory to the literature found. Nofsinger and Varma (2014) described how SRI funds underperform Conventional Funds in non-crisis periods, but outperform in periods of crisis. This effect is also described by Silva and Cortez (2016). A possible explanation could be that the Environmental and SRI funds were prone to even higher losses because of their higher idiosyncratic risks. My results seem to indicate that the disadvantages of the low diversification outweigh the advantages of socially responsible investing during crisis periods.
The coefficients for the market return show the beta’s of the funds, or their sensitivity to the market risk. The positive coefficients indicate that they are all prone to swings in the market. The positive coefficients for the Size and Value factor imply that the stocks of small firms perform better than those of large firms, and that stocks with high book-to-market value outperform stocks with low book-to-market value. This is consistent with the proposed effects by Fama and French (1995). The negative coefficients of the Socially Responsible and Conventional fund groups to the
Momentum-factor are inconsistent with the effect described by Carhart (1997). This would mean that stocks performing well in the recent year will perform worse in the near future and vice versa. The insignificance of the results however prevents us from assuming this opposite effect.
5.2 Screening strategies
5.2.1 Summary
The following data summaries belong to the second regression, where the effect of different strategies (Positive, Negative or Best-in-Class) on the returns is measured. Table 2.1 shows a summary of the average excess returns of the SRI funds for which the strategy was provided on yoursri.com. The regression is done for four different time frames: the total period and before, during and after the 2008 financial crisis.
Table 3.1. Summary of the Average Excess Returns of the SRI Funds with known strategies
Variable Obs Mean Std. Dev Min Max
Rtotal 141 .43 .31 -.13 1.45
Rpre-crisis 73 .71 .67 -1.19 2.76
Rcrisis 80 -2.53 1.14 -4.48 .30
Rpost-crisis 141 .71 .35 -.04 1.55
In this table the average excess return of all SRI Funds with known strategies is shown, as well as the average return of the pre-crisis-, crisis- and the post-crisis sample separately (Rtotal, Rpre-crisis, Rcrisis and Rpost-crisis).
In order to estimate these averages, first the monthly excess returns per individual fund were averaged for the time frame used in that sample, to calculate it’s mean performance. The summary shows the number of observations (funds) for which the regression is performed, the mean and the standard deviation of the averaged returns and the lowest and highest returns measured. The data is given in percentage points, and is rounded to two decimals.
As seen in Table 2.1, the number of funds included in the four time samples differs. Of the 141 SRI funds for which the strategy was provided, only 73 originated before March 2007 and could be included in the pre-crisis sample. During the crisis period 7 additional funds were established and could be added to this time sample. Because there are no dead funds in the database, all 141 funds included in the total sample could also be included in the post-crisis sample. The returns shown in the summary are calculated for only the time period of interest. This means that for the pre-crisis sample, the average excess returns for the time span between January 2004 and September 2007 were calculated for each individual fund. The summary shows the mean of these average fund returns for the corresponding period.
The summary shows a significant difference between the average excess return in the financial crisis and that before and after the crisis. The funds’ monthly excess returns were on average 3.24% lower during the crisis in comparison to the samples including the pre- and post-crisis periods, equal to an annualized 46.61%. The t-test measuring the differences between the means
respectively had values of 21.64 (pre-crisis – crisis) and 24.75 (post-crisis – crisis). These are both significant on a 1% level, so we can reject the null-hypotheses stating that the differences are zero. This is consistent with the results found earlier in this research, showing that all funds significantly performed worse during the crisis period. Furthermore, it is notable that the average excess returns pre- and post-crisis are approximately equal.
Table 3.2 Summary of the remaining inputs used for the regression on screening strategy
Variable Total Pre-crisis Crisis Post-crisis
Strategy Positive 24.82% 21.92% 23.75% 24.82% Negative 26.95% 21.92% 22.50% 26.95% Best-in-Class 15.60% 19.18% 17.50% 15.60% Combined 32.62% 36.99% 36.25% 32.62% Type Bond 11.35% 9.59% 8.75% 11.35% Equity 70.92% 73.97% 75.00% 70.92% Mixed 17.73% 16.44% 16.25% 17.73% Characteristics Global 55.32% 52.05% 50.25% 55.32% Age 14.30 19.14 8.67 14.30 Size 404.80 431.27 582.26 404.80
This table provides a summary of all strategies, types and characteristics used to test on the excess returns for the various strategies. The returns of each fund were corrected for the risk-free rates, and averaged for the sample’s time span. This led to an average excess return per fund. The fund’s strategy, type and characteristics were added. The table shows the percentages of funds using the various strategies, investing in Bond, Equity or both, and the percentage of funds investing globally. The table also shows the average age of the funds in years since inception, and the average size in millions of Euro’s. The data for the dummy variables is given in percentage points, age and size in actual value. Everything is rounded to two decimals.
Since the Total and the Post-crisis sample include the same funds, the data in these two columns is the same. The proportion of funds following a Positive and Negative strategy increased over the years, while the proportion of funds investing via the Best-in-Class principle decreased. The proportions shown seem consistent with the overview of SRI funds in Europe shown in Figure 1, earlier in this research.
Another development seen in Table 2.2 is in the investment type of funds. The number of funds investing in equity increased slightly between the pre-crisis and the crisis-period, while it decreased after the crisis- period. The bond type funds show an opposing development.
5.2.2 Effects
In Table 4 the results of the regression measuring the influence of the different strategies on SRI fund performance are shown. The columns show the four samples, constructed by taking the total, pre-crisis, crisis and post-crisis time frames.
Table 4. Results of the regression on the effect of strategies on SRI performance
Excess Returns Total Sample Pre-Crisis Crisis Post-Crisis
Positive -.0009 (-1.53) .0009 (0.44) .0012 (.46) -.0008 (-1.36) Negative .0003 (.54) .0011 (0.55) -.0013 (-.58) .0004 (.79) Best-in-Class .0003 (.45) .0008 (.37) -.0025 (-1.13) .0007 (1.07) Bond -.0021*** (-3.05) -.0047** (-2.46) .0092** (2.28) -.0025*** (-3.16) Equity .0029*** (5.29) .0059*** (4.20) -.0156*** (-6.79) .0040*** (6.90) Global .0003 (0.62) -.0011 (-.66) .0036* (1.77) .0002 (.39) Age -.0000* (-1.68) .0002* (1.85) -.0001 (-1.36) .0000 (1.15) Size -.0000*** (-3.09) -.0000* (-1.79) -.0000 (-.35) -.0000*** (-3.47) Alpha .0034*** (3.78) .0002 (.02) -.0132*** (-3.87) .0045*** (5.60) R-squared .3947 .4132 .6035 .5393 Observations 141 73 80 141
This table provides an overview on the regression to measure the effect of the various strategies (Positive, Negative, Best-in-Class or Combination). As control variables the Fund’s type is indicated (Bond or Equity or Mixed). The Combination strategy and Mixed type variables are excluded from the regression to prevent a Dummy Trap. The Fund’s characteristics (Global, Age and Size) are also included as control variables. The coefficients are measured for four time frames: the full time span, pre-crisis, during crisis and post-crisis period. Alpha denotes the constant of the regression and the R-squared and number of Observations are added. The values for the t-tests for all coefficients are added between brackets, and the star denotes the significance of a coefficient on a 10% level (*), on a 5% level (**) or on a 1% level (***).
For the total and the post-crisis sample, the Positive strategy gives a negative coefficient, while for the Negative and Best-in-Class strategies it is slightly positive. In the pre-crisis sample all coefficients are positive. The crisis sample shows different results: here, the Positive strategy is the only strategy with a positive coefficient, while both the Negative and the Best-in-Class approach have a negative one. The coefficients are not statistically significant.
The type of fund (Bond or Equity) does show significant results for all three samples. During non-crisis periods the Equity funds have a positive coefficient, while the Bond fund has a negative one. In the crisis-period the opposite is shown. The coefficient of the Global variable is significantly positive on a 10%-level for the crisis sample.
The values of the other control variables (Age and Size) are very small. The Age variable is significant on a 10% level in both the total and the pre-crisis sample. The Size coefficient shows a significant effect in all samples except the pre-crisis one (1% or 5%). The alpha is also significant for all except the pre-crisis sample ( 1%).
5.2.3 Discussion
The decrease in funds investing mainly in equity after the crisis period as seen in Table 3.2 could be caused by the fall in the stock markets encountered during the financial crisis. After the crisis, funds invested a larger part of their assets in bonds, which are perceived as safer than equity. This could also explain the rise in funds investing in bonds after the crisis.
None of the coefficients for the screening strategies shown in Table 4 are significant. I expected that the Best-in-Class approach would perform best, because of the inclusion of firms from all strategies, thus leading to a lower loss of diversification. However, it seems that following any of the screening strategies has no impact on a fund’s performance.
The type of fund (Bond or Equity) does show significant results for all three samples. This means the assumption that there exists no difference in performance between funds investing in different asset types can be rejected. During non-crisis periods we see an outperformance of Equity funds as opposed to and underperformance of Bond funds, and vice versa. This indicates that investing in Equity positively affects a fund’s performance when the economy is in a good state, and investing in Bonds has a positive effect on performance in bad states of the economy. As mentioned while discussing the Summary in Table 2.2, this is coherent with the fall in the stock markets as evidenced during the Financial Crisis.
The coefficient of the Global variable is significantly positive for the crisis sample. A possible explanation is that investing in a broader region makes a fund less exposed to region-specific risk.
Certain regions were affected more by the Financial Crisis than others, so investing globally gives funds another layer of diversification. A fund that has invested in multiple regions is less likely to incur losses in all of them than a fund investing in a sole region.
The values of the other control variables Age and Size are very small. The Age variable is significant on a in both the total and the pre-crisis sample. This could have to do with the
survivorship bias seen with mutual funds. The fact that a fund has survived during a relatively large time span, could mean that it performs better than the average fund. This would also explain why the effect in during and after the crisis is not significant: these time spans are shorter, so the effect of the bias would be smaller.
The alpha is significant for all except the pre-crisis sample, which suggests that the SRI funds generate an excess abnormal return. This is consistent with what I found in the first regression.
6. Conclusion
The interest in and demand of investments with an ethical or environmental aspect has risen substantially in the past decade. Climate change initiatives and the increasing importance of Corporate Social Responsibility have added to this rise in interest. Previous research seemed inconclusive to whether investing with an ethical or environmental purpose means gaining lower financial rewards. A widely described trade-off that comes with SRI Investing involves the cost of decreased diversification benefits, and the rewards of increased financial stability, investor loyalty and sensitivity to past returns.
Additional topics of interest in past literature were the influence of the 2008 financial crisis, and of the various screening strategies on the performance of the fund groups. Socially Responsible funds were perceived to perform better in times of crisis, due to the benefits mentioned above. Furthermore, because of the smallest loss in diversification benefits, Best-in-Class screened portfolios were assumed to outperform Positive and Negative screened ones.
My research added to previous research by using a different geographical focus (Europe), composing more recent and bigger samples, and by zooming into the influence of the various strategies in a crisis period. The focus was on comparing the performance European Socially
Responsible, Environmental and Conventional fund groups. In line with my hypotheses, I found that the Environmental fund groups performed the best between January 2004 and April 2018, followed by the Socially Responsible fund group. However, the effect in the crisis period contradicted my expectations. The Conventional fund group performed best in the crisis period, while the
Environmental funds performed worst. My final hypothesis on the outperformance of Best-in-Class screened funds could also not be assumed to be true. None of the strategies reported a statistically significant effect on performance, both during and surrounding the 2008 financial crisis.
In conclusion, my findings are that investing ethically or environmentally pays off, and involves no trading in of financial rewards.
7. Limitations and Recommendations
During my research I encountered a number of limitations. Like many researches on Mutual Fund performance, several biases could have influenced my datasets. The most evident of these biases is perhaps the survivorship bias: since no dead funds are included in my dataset, it could be the case that the funds surviving are simply the best performing ones. Since the field of Socially Responsible and Environmental Investing is quite new, data on dead funds is hard to find. Additionally, the backfill bias could have impacted my dataset. This revolves around the fact that funds can choose to only report on their returns when these are positive.
Another limitation is the size of the Environmental fund group examined. This provided a restriction on the time frame that could be used for my research: since the sample of Environmental funds decreased significantly before 2004, the prior years could not be covered. It would have been interesting to incorporate the dot.com crisis into my research, to measure the effect of multiple crises.
Furthermore, It could be interesting to extend the influence of the screening strategies to Environmental funds. However, since the sample was quite small, such an extension didn’t lead to sufficient samples per strategy. This would have made it difficult to generalize any statement concerning the influence of the strategies on the fund groups’ performance.
An additional limitation I encountered concerned the classification of firms as Socially Responsible of Environmental. Since the definition of these terms differs throughout countries and markets, it is difficult to define which firms qualify. This also proves to be difficult for funds: a fund that invests 51% of its assets in Socially Responsible or Environmental firms qualifies is seen as a SRI fund. This makes it challenging to measure the impact on performance due to the SRI or
References
Barnett, M., & Salomon, R. (2006). Beyond dichotomy: The curvilinear relationship between social responsibility and financial performance. Strategic Management Journal, 27(11), 1101-1122.
Basso, & Funari. (2014). Constant and variable returns to scale DEA models for socially responsible investment funds. European Journal of Operational Research, 235(3), 775-783.
Becchetti, L., Ciciretti, R., Dalò, A., & Herzel, S. (2015). Socially responsible and conventional investment funds: Performance comparison and the global financial crisis. Applied Economics, 47(25), 1-22.
Belghitar, Clark, & Deshmukh. (2014). Does it pay to be ethical? Evidence from the FTSE4Good. Journal of Banking and Finance, 47, 54-62.
Berk, J., & DeMarzo, P. (2014). Corporate Finance (3rd ed.). Boston, United States of Amerika: Pearson.
Capelle‐Blancard, G., & Monjon, S. (2014). The Performance of Socially Responsible Funds: Does the Screening Process Matter? European Financial Management, 20(3), 494-520.
Carhart, M. M. (1997). On persistence in mutual fund performance. The Journal of Finance, 52(1), 57-82.
Chung, H., Lee, H., & Tsai, P. (2012). Are Green Fund Investors Really Socially Responsible? Review of Pacific Basin Financial Markets and Policies, 15(4), 1-25
Clark, G. L., Feiner, A., and Viehs, M. (2015). “From the Stockholder to the Stakeholder: How Sustainability Can Drive Financial Outperformance.”
Cortez, M., Silva, F., & Areal, N. (2012). Socially Responsible Investing in the Global market: the performance of US and European funds. International Journal of Finance & Economics, 17(3), 254-271.
Derwall, J., Günster, N.K., Bauer, R., & Koedijk, C.G. (2004). The Eco-Efficiency Premium Puzzle. 2004.
Eurosif. (2016). European SRI Study 2016. Retrieved from: http://www.eurosif.org/wp-content/uploads/2016/11/SRI-study-2016-HR.pdf.
Fisman, R., Heal, G., & Nair, V. (2006). A model of corporate philanthropy. Working Paper Columbia University.
Fombrun, C., & Shanley, M. (1990). What's in a Name? Reputation Building and Corporate Strategy. Academy of Management Journal, 33(2), 233-258.
Friede, G., Busch, T., & Bassen, A. (2015). ESG and financial performance: aggregated evidence from more than 2000 empirical studies. Journal of Sustainable Finance & Investment, 5(4), 210-233.
Global Sustainable Investment Alliance. (2016). Global Sustainable Investment Review. Retrieved from:
http://www.gsi-alliance.org/wp-content/uploads/2017/03/GSIR_Review2016.F.pdf#page=4
Glode, V. (2011). Why mutual funds "underperform". Journal of Financial Economics, 99(3), 546-559.
Holzhauer, H. (2013). Socially Responsible Investing. Portfolio Theory and Management (p. Portfolio Theory and Management, Chapter 30). Oxford University Press.
Ibikunle, G., & Steffen, T. (2017). European Green Mutual Fund Performance: A Comparative Analysis with their Conventional and Black Peers. Journal of Business Ethics, 145(2), 337-355.
Innes, R., & Sam, A. (2008). Voluntary Pollution Reductions and the Enforcement of
Environmental Law: An Empirical Study of the 33/50 Program. Journal of Law and Economics, 51(2), 271-296.
Jensen, M. C. (1968). The performance of mutual funds in the period 1945– 1964. The Journal of Finance, 23(2), 389-416.
Kempf, A., & Osthoff, P. (2007). The Effect of Socially Responsible Investing on Portfolio Performance. European Financial Management, 13(5), 908-922.
Lee, D. D., Humphrey, J. E., Benson, K. L., & Ahn, J. Y. (2010). Socially responsible investment fund performance: the impact of screening intensity. Accounting &
Finance, 50(2), 351-370
Lesser, Rößle, & Walkshäusl. (2016). Socially responsible, green, and faith-based investment strategies: Screening activity matters! Finance Research Letters, 16, 171-178.
Muñoz, F., Vargas, M., & Marco, I. (2014). Environmental Mutual Funds: Financial Performance and Managerial Abilities. Journal of Business Ethics, 124(4), 551-569.
Nofsinger, & Varma. (2014). Socially responsible funds and market crises. Journal of Banking and Finance, 48, 180-193.
Renneboog, Ter Horst, & Zhang. (2008a). Socially responsible investments: Institutional aspects, performance, and investor behavior. Journal of Banking and Finance, 32(9), 1723-1742.
Renneboog, Ter Horst, & Zhang. (2008b). The price of ethics and stakeholder governance: The performance of socially responsible mutual funds. Journal of Corporate Finance, 14(3), 302-322.
Renneboog, Ter Horst, & Zhang. (2011). Is ethical money financially smart? Nonfinancial attributes and money flows of socially responsible investment funds. Journal of Financial Intermediation, 20(4), 562-588.
Sauer, David A. (1997). The impact of social-responsibility screens on investment
performance: Evidence from the Domini 400 Social Index and Domini Equity Mutual Fund. Review of Financial Economics, 6(2), 137-149.
of Socially Responsible Firms' Short-Term versus Long-term Performance. Managerial Finance, 31(8), 33-46.
Silva, & Cortez. (2016). The performance of US and European green funds in different market conditions. Journal of Cleaner Production, 135, 558-566.
Statman, M. (2000). Socially Responsible Mutual Funds (corrected). Financial Analysts Journal, 56(3), 30-39.
Yadav, P., Han, S., & Rho, J. (2016). Impact of Environmental Performance on Firm Value for Sustainable Investment: Evidence from Large US Firms. Business Strategy and the Environment, 25(6), 402-420.
