The Financial and Social Performance of SRI Mutual Funds

Transcription

1 The Financial and Social Performance of SRI Mutual Funds Master Thesis Author: Charlotte van Beek Student number: i Track: MSc International Business Sustainable Finance School of Business and Economics Maastricht University Supervisor: Jeroen Derwall Date:

2 Abstract Over the last decade sustainable and responsible investing has gained popularity. Increasingly, people begin to realize that financial return can also be achieved through more sustainable practices and investing in companies that pay attention to ESG issues. SRI mutual funds are one vehicle through which people can invest more sustainably. Their popularity has also risen, both among investors as well as academics. The current literature is mostly focused on the financial performance of American funds. The main conclusion of research so far is that it is difficult for SRI mutual funds to outperform the market and their conventional peers. Some attention has been given to performance of SRI funds in times of financial downturns and funds outside of the US. There the results show that (American) SRI mutual funds may be less affected by financial downturns than the market or their conventional peers. Moreover, in some other countries the SRI funds manage to outperform the market, but in most countries they do not. Evidence is often mixed. This underwhelming evidence begs the question if SRI mutual funds are all that different from conventional funds. This research looks into both financial and social performance of these funds. Findings suggest that SRI mutual funds indeed do not manage to outperform the market or their conventional peers on a significant level. In most countries, the SRI mutual funds underperform. Moreover, no evidence is found that SRI mutual funds indeed protect investors from financial downturns. However, SRI mutual funds are indeed more sustainable than conventional funds. Their portfolios prove to have higher ratings on all important sustainability categories. 1

3 1. Introduction Today our world is changing faster than ever before economic, geopolitical and environmental challenges abound. A company must invest in the key ingredients of profitability: its people, communities and the environment. ~ Warren Buffett Sustainable and responsible investing has taken really taken off in the past decade and has especially become of increasing interest among investors and the general public since the financial crisis of Even Warren Buffett, who is generally known for investing from a pure business perspective, has recognized that companies and investors should pay attention to their surroundings in order to invest profitably (John Mansville, 2011). Numerous scandals with oil companies and car, apparel and mobile device manufacturers, fraudulent bankers and accountants, and pension funds investing in weapons and bomb manufacturers and companies that sustain a range of addictions, have increasing led to more awareness about what types of companies to invest in. Hence, also among mutual funds there is an increasing number of sustainable or socially responsible investment (henceforth SRI) funds. A mutual fund, according to the United States Securities and Exchange Commission (2005), is an investment vehicle that brings together money from many people and invests it in stocks, bonds or other assets. These assets are pooled together in a portfolio, in which people can invest their money (The United States Securities and Exchange Commission, 2005). SRI can be defined in many ways, depending on one s opinion, point of view and vocabulary. The United States Forum for Sustainable and Responsible Investment, for example, defines it as Sustainable, Responsible and Impact Investing, which values the importance of environmental, social and governmental (henceforth ESG) criteria in investment decisions, in order to combine a longterm financial profit with creating a positive influence on society (US SIF, 2015). Mutual funds are identified by the US SIF (2015) as the fastest growing and most dynamic group of sustainable and responsible investors. There has also been attention from the academic field, both for the financial results of companies and their stock that invest in enhancing ESG criteria for their company as well as mutual funds that pay attention to this in making their investment decisions. For individual companies, investing in ESG related issues appears to be quite profitable. It shows, for example, that corporate social responsibility (henceforth CSR) initiatives started with slack resources increase profits (Waddock & Graves, 1997; Ruf et al., 2001; Ameer & Othman, 2012). It is also consistently found that companies with stronger CSR policies and more emphasis on one or more of the ESG factors have better financial performance and are able to outperform the market (i.e. earn alpha) (E.g. Edmans, 2011; Brammer, Brooks, & Pavelin, 2

4 2009; Gompers, Ishii, & Metrick, 2003; Guenster et al., 2011; Statman & Glushkov, 2009). Increased transparency is also an advantage of CSR (Dhaliwal et al., 2012). However, for SRI mutual funds previous research suggests that it is difficult for SRI mutual funds to outperform their conventional counterpart and both have a difficult time outperforming the market (Hamilton, Jo, & Statman, 1993; Statman, 2000; Bello, 2005; Bauer, Derwall, & Otten, 2007), but there is some evidence that it is possible for SRI funds to outperform (Kempf & Osthoff, 2007; Gil-Bazo, Ruiz-Verdú, & Santos, 2010; Weber, Mansfeld, & Schirrmann, 2010; Climent & Soriano, 2011). Moreover, there appears to be evidence that during financial downturns in the market SRI funds perform better than the market and their conventional peers (e.g Nofsinger and Varma, 2013; Peylo & Schaltegger, 2014). Looking at international evidence of SRI and conventional mutual fund performance does not make for a better or clearer picture of their performance. Some evidence suggests that SRI mutual funds manage to outperform in certain geographic areas (Renneboog, Ter Horst, & Zhang, 2008; Kreander et al., 2000; Luther, Matatko, & Corner, 1992; Cortez, Silva, & Areal, 2009), while other suggest underperformance (Renneboog, Ter Horst, & Zhang, 2008; Kreander et al., 2005; Cortez, Silva, & Areal, 2012; Bauer, Otten, & Rad, 2006; Lean, Ang, & Smyth, 2014). In neither case the evidence appears to be very strong. Two reasons have been suggested for the lack of difference in terms of financial performance between SRI and conventional funds; either ESG issues are not related to returns or SRI funds and conventional funds are not as different in terms of their holdings as is assumed. Hence, it is interesting to take a look at the funds holdings and see whether there is any difference to be detected there. There is only little research done into this area. The evidence found so far suggests there is a difference in terms of the sustainability ratings of the holdings, but the difference is not a big as might be expected (Kempf & Osthoff, 2008; Utz & Wimmer, 2014). With this research I attempt to add to the current, highly divided literature on the capability of SRI mutual funds to earn alpha, shed some light on SRI mutual funds behavior during financial crises and their performance in countries other than the United States, as well as provide some insight on whether or not SRI mutual funds actually do what they promise in terms of sustainability and responsibility. In order to do so, I use samples of SRI and conventional mutual funds from the United States, Australia, Belgium, Denmark, France, Ireland, Italy, Japan, Korea, Norway, Sweden, Switzerland, The Netherlands, The United Kingdom, using their monthly returns. I use the Fama and French factors as available on their website and through WRDS local as well as global factors to perform a Carhart (1997) Four Factor analysis over the period January 2002 to December Moreover, I use a 3

5 sample of 1330 portfolios of American conventional mutual funds and 950 portfolios of American SRI mutual funds to analyze the sustainability and responsibility levels of these funds. The results show that in most cases, the SRI mutual funds do not manage to outperform either the market or their conventional peers. Only in Australia, Korea and Denmark there appears to be outperformance that is relatively robust to the use of different lmarket factors. However, these results are not significant. Moreover, during the financial crisis of 2008 the SRI mutual funds perform a little better compared to their performance before and after the crisis, but they still underperform the market, so there is no reason to assume that SRI mutual funds have mitigating properties when it comes to financial downturns. On the upside, at least for investors who do not mind giving up financial return for social return, SRI mutual funds are more sustainable than their conventional peers. This research contributes to the existing literature in several ways. Firstly, both the local as well as the global Fama and French factors are used. Previous research tended to be United States focused or use (self-composed) global factors. I use all versions published by Fama and French, which simultaneously serves as a robustness check and shows the usefulness local and global factors. Moreover, it is the first research to explicitly separate a pre-crisis, duringcrisis and post-crisis periods for the financial crisis of 2008, rather than separating between crisis and non-crisis periods, and also to do so for all factors and not simply for alpha. In this way, also the effects of financial downturn on the other market factors are investigated. Lastly, the holdings of mutual funds have been largely overlooked by researchers and the available literature is mostly not that recent. Hence, this research tries to close the gap between the difference in performance and holdings for SRI and conventional mutual funds and see if being an SRI mutual fund is of importance to either. The remainder of this article is organized as follows: First I will give an overview of the current literature in the areas I will investigate, next I will give a description of the data I use and the methodology I employ, after which I will present my results. I will end with a conclusion, and a discussion of the limitations and implications. 4

6 2. Theory 2.1 Literature review In this literature review I will present the current literature on sustainable, responsible investing. First I will introduce some literature on the impact of CSR and/or ESG initiatives on financial performance and stock returns of companies. Second I will present the current literature and perspectives regarding the performance of conventional versus SRI mutual funds, and their behavior in and after financial crises as well as in different regions in the world. Third, I will give an overview of the current literature on the holdings and the differences among holdings of conventional and SRI mutual funds. This all will serve to get an overview of the general consensus about SRI funds and to develop the hypotheses that will be tested in this paper ESG and stock returns Milton Friedman is arguably the biggest and earliest critic of companies engaging in any kind of activity that is not their core activity, including activities that have some stakeholder s beside the shareholder s interest at hart. According to Friedman, (1962), the responsibility of the business is to increase profit. The question is if ESG initiatives could not do that. There are two ways the impact of ESG initiatives on stock returns has been analyzed. Researchers either look at the performance of stocks of individual companies with high ESG ratings or they compose hypothetical portfolios of companies with high and low ratings. Statman and Glushkov (2009) do their analysis the first way, by looking at the KLD ratings and stock performance of companies. Their findings suggest that there are two sides to the story: socially responsible companies earn abnormal returns, but so do sin companies. Smart investors can enhance the performance of their portfolio by employing positive screens, but harm it by negative screens (Statman & Glushkov, 2009). In line with the finding that controversial companies yield abnormal returns, it is found that companies with a low corporate social performance perform better financially compared to companies with moderate corporate social performance all times (Barnett & Salomon, 2012) as well as companies with good social performance in the short run (Brammer & Millington, 2008). These findings are in line with the shunned-stock hypothesis of Derwall, Koedijk and Ter Horst (2011) and also with the findings of Hong and Kacperczyk (2009). Companies that can be considered controversial are shunned by a large group of investors and need to yield higher returns to compensate for the extra risk resulting from the small investor base. 5

7 However, it is also found that in the long run companies with excellent social performance outperform all other companies (Brammer & Millington, 2008; Barnett & Salomon, 2012). This is, along with Statman and Glushkov s (2009) finding that companies with high KLD ratings outperform, in line with the errors-in-expectations hypothesis of Derwall, Koedijk and Ter Horst (2011). This hypothesis suggests that the market is not always efficient in capturing the importance of CSR and ESG initiatives and thereby creates an opportunity for investors to earn abnormal returns (Derwall, Koedijk, & Ter Horst, 2011). This might be the result of the intangibility of the advantages of these initiatives. Porter and Kramer (2006; 2011) show that CSR and ESG initiatives can create shared value and could thereby actually improve a company s bottom line performance. If the initiatives are clearly linked to a company s core activities, it is possible for society to benefit but also for the company by, for example, improving production processes or creating new and educating existing customers pools. In this way, the company can enhance its image and customer loyalty, bring down production cost on the long(er) run, stay ahead of competition and possibly create a unique selling point or sustainable competitive advantage (Porter & Kramer, 2006; 2011). However, it remains intangible, it cannot be put on the balance sheet, which makes it difficult for many analysts and investors to include it properly in their expectations. They might either underestimate the financial benefits of the initiatives or overestimate the costs. These errorsin-expectation is suggested to be the basis for the abnormal returns found by among others Statman and Glushkov (2009). Even though analysts might find it difficult to fully capture the advantages of CSR and ESG, researchers have been more successful, as the preceding findings by among other Statman and Glushkov (2009), Brammer and Millington (2008), and Barnett and Salomon (2012) already show. Moreover, it is found that after implementing CSR initiatives and improving social performance, financial performance also improves (Waddock & Graves, 1997). This is supported by findings of, among others, Ruf et al. (2001) and Ameer and Othman (2012). Gompers, Ishii and Metrick (2003) find that companies with stronger shareholder rights also perform better in terms of firm value, profits and growth, and have lower cost of debt and equity. This is confirmed by, among others, Ammann, Oesch and Schmid (2011), and Jo and Harjoto (2011). Moreover, eco-efficient companies are valued higher in the market and have a better operating performance (Guenster et al., 2011), and companies with high employee satisfaction ratings have lower bankruptcy risk and higher credit ratings (Verwijmeren & Derwall, 2010). The latter is in line with the findings of Hong and Kacperczyk (2009). 6

8 Moreover, sustainable companies might actually make analysts work a little easier, as they usually publish CSR or sustainability reports. These give extra information about business practices, thereby possibly decreasing information asymmetries and thus have the potential to improve stock performance forecasts by analysts (Dhaliwal et al., 2012). Also, companies with more CSR disclosure in their reports have higher market values and are more profitable (Qiu, Shaukat, & Tharyan, 2014). There is also a group of researchers who judge the impact of sustainability of stock returns by forming hypothetical portfolios. Among these studies there is some evidence that holding a portfolio of less or even the least socially responsible gives a much better return for shareholders, compared to holding a portfolio of sustainable stocks (Brammer, Brooks, & Pavelin, 2006; Hong & Kacperczyk, 2009), however, most evidence indicates that sustainable portfolios yield alpha. One well-known research, that also assumes that there might be errors in expectations, is done by Edmans (2011). He shows that the stock market does not fully value the Human Resources initiatives that lead to companies being among the 100 Best Companies to Work For in America. Edmans portfolio, constructed of the companies on the Fortune Magazine list, earned a significant abnormal return during the period Companies that focus on environmental issues also manage to yield alpha. Derwall et al. (2005) show that a portfolio of more eco-efficient companies outperforms a portfolio of less eco-efficient companies and earn a significant positive alpha over their benchmark. Moreover, it is found that high-sustainability rated stock portfolios outperform low-sustainability rating portfolios (Van de Velde, Vermeir, & Corten, 2005). Eccles, Ioannou, and Serafeim (2014) have composed two portfolios, one with companies ranked high on sustainability and one with companies ranked low. The results indicate that the portfolio with the High Sustainability companies outperform the portfolio with the Low Sustainability companies in terms of stock market performance in the long run (Eccles, Ioannou, & Serafeim, 2014). This is in line with the findings of Brammer and Millington (2008), Barnett and Salomon (2012), and Van de Velde, Vermeir, and Corten (2005). The preceding discussion of literature gives a clear view of the scientific reasoning behind the importance of ESG factors and proper CSR initiatives for companies financial performance. It gives a first glimpse of the potential for SRI mutual funds to outperform at least their conventional counterparts and possible the market. Clearly, sustainable companies, addressing 7

9 one or more ESG-related issue, can outperform and give investors an additional return. Hence SRI mutual funds should be able to take advantage as their object should be to invest only in these types of companies. Moreover, investors will be more loyal to them Mutual funds This part of the literature review is devoted to the performance of mutual funds. Much research has been done on (the determinants of) performance of mutual funds in general. Whether or not mutual funds are able to outperform their benchmark and peers depends on how the fund is managed. An early study by Sharpe (1966) concludes past performance is an indicator of future performance. Carhart (1997) and Agnesens (2013) indeed conform that there is some level of persistence in the performance of mutual funds. Agnesens (2013) results also suggest that fund size, family size and past performance are the only relevant indicators of future performance. Also according to Sharpe (1966), diversification is important for mutual fund performance and fund managers should not try to identify mispriced stocks in order to earn higher returns. Funds that do not try to find those, perform better than funds that do try, at least in Sharpe s research. This is to some extend conformed by Cuthbertson, Nitzsche and O Sullivan (2008), who show that well-performing mutual funds are mostly lucky while bad performance of funds is usually the result of the (lack of) stock picking abilities of their managers. The notion of diversification is an important one when considering the performance of SRI mutual funds. SRI mutual funds are sometimes said to be under-diversified compared to conventional funds, because certain industries or types of companies might be disregarded, and thus might perform worse according to Sharpe s point considering the importance of diversification SRI and conventional mutual fund performance Now it is time to turn to the scientific evidence on the performance of SRI funds and the comparison with the performance of conventional mutual funds. Evidence is scattered and sometimes contradicting, but most of all neutral. Only in a few instances a performance difference between SRI and conventional funds is documented. A relatively early study by Hamilton, Jo and Statman (1993) shows that SRI funds have an average yearly excess return which is not statistically different from zero. Hence, SRI mutual funds do not manage to outperform the benchmark in this study. After also including conventional mutual funds to their analysis Hamilton, Jo and Statman (1993) find that their performance is similar and 8

10 hence that there is no performance difference between conventional and SRI mutual funds. Neither manage to generate a significant alpha (Hamilton, Jo, & Statman, 1993). A similar result is found by Statman (2000). Even though he finds that on the index level, the Domini Social Index outperformed the S&P500 index for the period , the performance difference on fund level between SRI and conventional mutual funds is not significant but positive for the SRI funds. Moreover, neither group of funds outperformed the S&P500 or the Domini Social Index over that period (Statman, 2000). In a broader investigation, which also includes the characteristics of both types of funds to which I will return later, Bello (2005) seems to confirm the previous findings by Hamilton, Jo and Statman (1993) and Statman (2000). Bello (2005) finds no performance difference between the two types of funds and both underperform the S&P500 and the Domini Social Index 400 over the period Similar results are also found by Bauer, Derwall and Otten (2007), over a slightly later time period and using Canadian funds and a Canadian benchmark. The funds did not significantly outperform each other but they underperformed the benchmark significantly. The preceding evidence is all a little disappointing form the SRI point of view, especially after finding such positive results on sustainable companies. There is some evidence that shows that SRI mutual funds do have the ability to outperform. Kempf and Osthoff (2007) for example investigate SRI mutual fund performance for different screening methods and find that funds employing the Best-in-Class method perform best and also earn a significant alpha. Moreover, from the range of portfolios they investigate it is found that portfolios with a high sustainability rank outperform portfolios with a low ranking (Kempf & Osthoff, 2007). Gil- Bazo, Ruiz-Verdú and Santos (2010) also find that SRI mutual funds outperformed their conventional counterparts over the period the period , even after controlling for the level of management fees. Weber, Mansfeld and Schirrmann (2010) find that between 2002 and 2009 their sample of SRI mutual funds significantly outperformed the MSCI World Index. Environmental or green funds also seem to catch up. Where they still underperformed conventional peers up to 2001, their performance starts to pick up after 2001 (Climent & Soriano, 2011). Moreover, it is found that among the best performing SRI mutual funds there is a strong level of persistence (Abdelsalam et al., 2014), so once a fund is doing well it is likely to continue to do so in the future. As a side step, Derwall and Koedijk (2009) found that for fixed income funds SRI pure bond funds overall performed slightly, although not significantly, better than conventional bond funds. SRI balanced funds outperformed their conventional counterparts by a significant 1.36% (Derwall & Koedijk, 2009). 9

11 In general it can be concluded that it is difficult for SRI mutual funds to outperform their conventional peers. In most cases, there is no or only a slight outperformance and usually neither type of fund manages to outperform indices, such as the S&P500. However, it is not impossible either. There are apparently SRI mutual funds that do manage to outperform on a significant level, which could for example be attributed to the chosen benchmark, the different screening practices among the SRI mutual funds, the time period under investigation or country specific performance. The latter two are subject of the next sections of this literature review SRI mutual funds and financial downturns So far I have considered mostly papers that do not distinguish between time periods or geographical regions. However, this is interesting to look at. Due to cultural and local market conditions it is very well possible that SRI mutual funds may perform quite differently when looking at different geographical areas. Moreover, due to the most recent financial crisis, stock markets worldwide suffered. Hence, it is likely that the performance of all types of investment vehicles have been influenced by this. However, the extent of the damage and the recovery pattern may be quite different for SRI and conventional mutual funds. To start with the latter, on the one hand, there is research which says diversification is more important during crises and more stocks are needed in times of low returns and/or high volatility (Hu, Chang, & Chou, 2014). By intuition SRI funds are in general considered less diversified, so this might indicate they perform worse. On the other hand, there is also research which suggests that the limiting riskiness and increasing conservatism of the portfolio is important during crises (Areal, Cortez, & Silva, 2010; Lean & Nguyen, 2014), which could mean that SRI mutual funds are actually favorable over conventional funds. Moreover, Bello (2005) shows that there is no difference in the level of diversification between SRI and conventional mutual funds. Hence the first point might not hold. Nofsinger and Varma (2013) have investigated the performance of SRI mutual funds during two crisis periods. The first period is the result of the technology bubble March 2000 to October 2002 and the second period is the result of the global financial crisis October 2007 to March Using a sample of US domestic SRI and conventional mutual funds up to 2011, Nofsinger and Varma (2013) find that during crisis periods the SRI mutual funds outperformed the conventional mutual funds. Over the full period, the average returns of the two types of funds do not differ significantly. The Carhart four-factor alpha is insignificantly negative for both, but the difference is positive insignificantly for the SRI mutual funds. 10

12 During non-crisis periods the SRI mutual funds have an insignificantly lower return compared to their conventional peers, while the difference in the Carhart four-factor alpha is negative for SRI mutual funds on a marginally significant level. Both types of funds underperform the S&P500 on a significant level (Nofsinger & Varma, 2013). When looking at the crisis periods, both types have significantly negative average returns, but the difference between them is insignificant. The Carhart four-factor alpha is insignificantly positive for both SRI and conventional mutual funds, while the difference is significantly positive for SRI mutual funds (Nofsinger & Varma, 2013). Although insignificant, the positive alpha signals that mutual funds are better at mitigating downside risk than the market, SRI mutual funds even more so than conventional funds. In terms of screening methods, it becomes clear that negative screens do not have a significant impact on returns in either of the periods, while positive screens lead to significantly lower alphas during non-crisis periods, while providing a positive alpha during crisis periods. Faith or religion based screening never contributes significantly (Nofsinger & Varma, 2013). Similar results, on the basis of other measurements, are found by Areal, Cortez and Silva (2010), who investigate mutual fund performance over different volatility regimes. During periods of low volatility, vice funds outperform the benchmark. However, they underperform during high volatility periods. SRI mutual funds underperform during low volatility periods, while they (insignificantly) outperform during high volatility periods (Areal, Cortez, & Silva, 2010). Funds employing religious screening methods never outperform. These results imply that SRI might be a better investment during crises, as those are usually characterized by higher levels of volatility (Areal, Cortez, & Silva, 2010). Peylo and Schaltegger (2014) also find that there is a positive relationship between the sustainability level of a stock portfolio and its financial performance, that is especially strong during crisis times. This also implies that sustainable portfolios should perform better at least during periods of crisis, compared to non-sustainable portfolios (Peylo & Schaltegger, 2014). Leite and Cortez (2014) suggest that the underperformance of SRI mutual funds found by Nofsinger & Varma (2014) and Areal, Cortez, & Silva (2010) during non-crisis periods is driven by the use of negative screens. They find that both conventional and SRI mutual funds tend to underperform during non-crisis periods, while they both insignificantly outperform during crisis periods. Moreover, they tend to differ little in investment style during crisis periods (Leite & Cortez, 2014). For France, Ducassy (2012) finds a significant positive relationship between the social and financial performance of companies and firms. Before the crisis, during the first months of 2007, and after about six months after the beginning, there is a positive relationship, however this is not very strong. Especially at the beginning of the 11

13 financial crisis in the second half of 2007 there is a significant positive relationship between sustainability and financial performance (Ducassy, 2012). Summing it all up, the preceding discussion indicates that there appears to a positive relationship between sustainability or sustainable investing and performance during market downturns. This suggests that SRI mutual funds are less risky than the market and also to some extent conventional mutual funds. Screening appears to be the most important reason for this SRI mutual funds worldwide Another distinction that can be made is between geographical regions or countries. Although the latest financial crisis has hit nearly all countries worldwide and nowadays (financial) markets become more global and integrated, there are always local conditions on the economy and equity markets. These local conditions can mitigate or worsen the effects of global conditions or can have effects beside any global condition, not applicable to any other country. Next to local economic conditions, there might also be a cultural aspect involved in the case of sustainable investments. Possibly investors in some countries appreciate sustainable companies more than in other countries, because of a higher emphasis on the importance of the ESG factors, because they have a more collectivistic culture or because of formal regulations. Even though the capital market is global, which might limit the importance of cultural aspects, there is quite some evidence that investors prefer to invest in their own market the so-called home bias (Coval & Moskowitz, 1999; Karlsson & Nordén, 2007; Van Nieuwerburgh & Veldkamp, 2009; Coeurdacier & Rey, 2013; Levy & Levy, 2014). When investors prefer to invest in local markets, the cultural aspects may play a role. Hence it is interesting to look at the scientific evidence for SRI mutual fund performance over different countries or geographic regions. Renneboog, Ter Horst and Zhang (2008) have so far been the only ones to involve multiple geographic regions in their research. They include 11 European countries (including the UK), the US and Canada as North-America and four countries for Asia-Pacific. Their conclusion is that in all regions the SRI mutual funds underperform the conventional funds. In general, mutual funds tend to underperform the local benchmark (Renneboog, Ter Horst, & Zhang, 2008). This is a harsh conclusion since the majority of the underperformance they find is insignificant. Of all the countries, only in Sweden and Japan the SRI mutual funds underperformance of the market and the (negative) difference between the SRI and conventional mutual fund performance is significant (Renneboog, Ter Horst, & Zhang, 2008). 12

14 In France, Ireland, the Netherlands, the UK, the US and Canada, SRI mutual funds significantly underperform the benchmark, but the (negative) difference with the conventional funds is not significant (Renneboog, Ter Horst, & Zhang, 2008). Even though the other underperformance and/or difference are not significant for all other countries, the central tendency still remains quite negative for the SRI mutual funds. In most countries they underperform both the benchmark as well as their conventional counterparts (Renneboog, Ter Horst, & Zhang, 2008). Renneboog, Ter Horst and Zhang (2008) may have been the only ones so far to incorporate the different geographic regions in one study, there is more research done involving one region at a time. Hence it is also useful to see what the results of different studies per region are and to what extent these confirm the findings by Renneboog, Ter Horst and Zhang (2008). Most literature I presented in previous sections is based on US data. For example, the articles by Hamilton, Jo and Statman (1993), Statman (2000), Bello (2005), Kempf and Osthoff (2007), Gil-Bazo, Ruiz-Verdú and Santos (2010), Climent and Soriano (2011) and Nofsinger and Varma (2014) all make use of data from mutual funds in the US. Bauer, Derwall and Otten (2007) is based on Canadian data, which also falls under the North-American continent. Hence the conclusion for North-America could be that mutual funds in general tend to underperform the benchmark while SRI mutual funds usually do not perform significantly better or worse than their conventional peers. However, a few studies do find outperformance and also during financial crisis SRI mutual funds tend to outperform. The conclusions are a little less strong compared to the conclusions of Renneboog, Ter Horst and Zhang (2008). There has also been quite some research done on the SRI mutual funds performance in one or more European countries. For example, Luther, Matatko and Corner (1992) find a weak outperformance for British ethical mutual funds. International diversification of the portfolios of such funds is not uniform and the ethical funds tend to invest in companies with lower market capitalization than the overall market and companies with a low dividend yield (Luther, Matatko, & Corner, 1992). Kreander et al. (2000) find that one fund from the Netherlands and one fund from the UK managed to significantly outperform the MSCI World Index. Of the other funds under investigation, the majority also achieved a positive alpha, but not on a significant level. The few negative alphas reported were also insignificant (Kreander et al., 2000). Bauer, Koedijk and Otten (2005) also find that British SRI funds outperform their conventional peers when introducing time variation in betas. German SRI funds never show significant under- or over performance, while older funds overall also perform better (Bauer, Koedijk, & Otten, 2005) 13

15 Kreander et al. (2005) study the performance of 60 European social and conventional mutual funds over the period 1995 to 2001 and find that there is no statistical difference among the performance of the two types and that neither is capable of outperforming the market. Similar results are found by Cortez, Silva and Areal (2009). They also find that there is not much of a difference between the performance of their sample of 88 European funds and both the conventional as well as the socially responsible indices used as benchmarks. Only the Belgium Global Equity funds and French European Equity funds significantly underperform both the conventional and SRI indices (Cortez, Silva, & Areal, 2009). The Dutch Global Equity funds outperform both benchmarks, while all European Balanced funds underperform the conventional indices and Belgium, French, German and Dutch balanced funds underperformed the SRI indices, where Austrian and Italian balanced funds outperformed the SRI indices all results are insignificant, though (Cortez, Silva, & Areal, 2009). Cortez, Silva, & Areal (2012) find no significant out- or underperformance for Belgium, France, Germany, Italy and the Netherlands. Austria appears to constantly significantly underperform, as does the US, which is included in this study for comparative reasons (Cortez, Silva, & Areal, 2012). For the Asia-Pacific region, Australia is the most studied country. Bauer, Otten and Rad (2006) document no performance difference in Australia between SRI and conventional mutual funds over the period 1992 to However, when breaking up that time window it shows that in the first four years of the sample there is significant underperformance on the part of the SRI mutual funds. In the later periods, this difference evens out and becomes insignificant (Bauer, Otten, & Rad, 2006). Jones et al. (2008) find a significantly different underperformance over the entire period they investigate, 1989 to 2005, but in contrast to Bauer, Otten and Rad (2006) the underperformance is worst during the last five periods of the sample, so from 2000 to Humphrey and Lee (2011) find no performance difference among the Australian SRI and conventional mutual funds. Lean, Ang and Smyth (2014) investigate a number of SRI funds from Asia-Pacific, including funds from Japan, Hong Kong and New Zealand. They find no significant performance difference for these funds as compared to the MSCI All Country Asia Pacific as a benchmark (Lean, Ang, & Smyth, 2014). In conclusion, more than 20 years of research has failed to identify real, persistent, and significant outperformance of SRI funds over their conventional peers, despite the fact that this would be expected based on the positive influence of sustainability on stock performance of individual companies. Even when differentiating for financial crises or geographic region, 14

16 it shows that during crises or in certain regions SRI mutual perform better than their conventional counterparts, but also here there is some contradicting evidence indicating that there is no performance difference. Hence, there is no overwhelming and conclusive that SRI mutual funds outperform or underperform for that matter the conventional mutual funds, under whichever condition Are SRI mutual funds really more sustainable? This lack of convincing evidence and the big difference with the positive results of the impact of sustainability on stock returns in general begs the question if SRI and conventional funds are all that different. Possibly, SRI funds invest by far not as sustainably as advertised and expected or maybe conventional funds are far more sustainable than everyone thinks. Bello (2005) shows that conventional and SRI mutual funds do not differ significantly in terms of characteristics such as of assets held, portfolio diversification, and the variable effects of diversification on investment performance. This might be another sign that conventional and SRI mutual funds do not differ much in terms of underlying holdings. Moreover, Koellner et al. (2007) find that sustainable and conventional mutual funds perform very similar in terms of environmental impact, based on their portfolios. In order to be able to answer that question, a comparison of the underlying holdings of both type of funds is necessary. This is something that has not been readily done yet by research, but there are a few scholars who have made an attempt to answer this question. Among the first to do so where Kempf and Osthoff (2008). Using a sample of US mutual funds analyzed over a period from 1991 to 2004 they find that SRI funds are more sustainable or ethical than the conventional funds in their sample. Over the full period under investigation the SRI funds ranked higher on the qualitative Community, Diversity, Employee Relations, Environment, Human Rights, Product and exclusionary Alcohol, Gambling, Military, Nuclear Power and Tobacco measures as provided by KLD (Kempf & Osthoff, 2008). Less convincing evidence is found by Utz and Wimmer (2014). As a matter of fact they find that SRI mutual funds do not necessarily shun unethical companies. Looking at a portfolio level it becomes apparent that within the sample of conventional funds there are a few rotten apples as Utz and Wimmer (2014) call them. These are funds that invest heavily in unsustainable companies. These rotten apples do not exist among the SRI funds. However, on average there does not appear to be a difference in terms of ESG ratings for SRI and conventional funds (Utz & Wimmer, 2014). 15

17 Based on these papers it can be concluded that SRI mutual funds and conventional funds do differ to some extent in terms of underlying holdings. On the one hand, SRI mutual funds do invest more in sustainable or ethical funds. It is surprising, however, that apparently they also tend to invest in less sustainable companies, although not to the same extent as conventional mutual funds. Conventional mutual funds, on the other hand, are maybe more sustainable than many people expect. Hence, there is a difference in terms of holdings, but not the big difference that might be expected considering the placement of especially SRI mutual funds in the market. 2.2 Hypotheses Based on the previously discussed literature, I have formed expectations about the outcome of my analyses. I will present these in the form of hypotheses SRI mutual funds in the United States As the literature review makes clear, evidence on American mutual funds is all in all neutral. The SRI mutual funds appear to yield zero alpha. Considering the fact that there is some evidence of outperformance, but there appears to be more evidence of underperformance or no significant performance difference at all, I expect for my sample that over the full period the SRI mutual funds will not outperform the market or their conventional peers. H1: American SRI mutual funds do not outperform the market Pre-, during- and after-crisis performance The performance evidence for American SRI mutual funds is not all too positive, even though there seems to be a slight shift in recent years. Considering this fact, I expect for the first part of my sample period that SRI mutual funds will not outperform their conventional peers or the market. H2a: SRI mutual funds did not outperform the market before the crisis. However, Nofsinger and Varma (2013), and Leite and Cortez (2014) indicate that SRI mutual funds are less affected by financial downturns. They suggest that SRI mutual funds are better investments during crises than their conventional counterparts. Other papers show results supporting this view, as shown in the literature review. Hence, I also expect SRI mutual funds to outperform during the crisis. It has never been explicitly investigated if SRI mutual funds 16

18 also recover faster from any hits they may have taken from financial downturns, or how conventional funds recover after a crisis. But following the logic that SRI mutual funds are less affected by financial downturns and hence have less recovering to do, and investors appear seek more sustainable investment opportunities after the crisis, I would expect that also then SRI mutual funds perform well. H2b: SRI mutual funds outperformed the market during and after the crisis SRI mutual funds worldwide For expectations about the outcomes the analysis of the countries across different geographic regions outside the United Stated, I will predominantly focus on the finding by Renneboog et al. (2008). This is the only study that looks at different geographic region in a similar way and most of the result are confirmed by research focusing on one of the geographic regions. Although their final conclusion was somewhat strongly formulated considering the number of insignificant results, I do expect to find results pointing toward a similar direction. Hence I do not expect the SRI mutual funds to outperform the market or their conventional peers: H3: European and Asian-Pacific SRI mutual funds do not outperform the market Are SRI mutual funds really more sustainable? I analyze the holdings to see if at least on that level there is performance difference between the conventional and SRI mutual funds, assuming that there will be little difference in terms of returns. The literature available in this areas suggests that indeed there might not be much difference between the two types of funds, however, the only study ever to be fully devoted to the holdings of mutual funds (Kempf & Osthoff, 2008) concludes that SRI mutual funds are indeed more sustainable than conventional mutual funds. H4: The underlying holdings of SRI mutual funds are fundamentally different from and more sustainable than those of conventional mutual funds. 17

19 3. Data Return data In this section I will describe the data I use to test the hypotheses. I use different types of data from different databases. First of all, I use CRSP (through Wharton Research Data Services) to collect the monthly return data for the American mutual funds. I use Morningstar to identify and differentiate between conventional and SRI mutual funds. I use only those funds that have reported between January 2002 and December All funds included have reported for approximately 30 consecutive funds or more. For the American SRI mutual funds, this leaves me with a sample of 111 funds. For the conventional funds, I have randomly selected 157 funds. To collect the return data for non-us funds, I use Datastream. To detect the SRI funds, I search for a number of terms signaling SRI investment, i.e. Green, Ethic(al), Environment, Social, Sustainable, Responsible, SRI and ESG, also translated to local language. For example, in French SRI becomes ISR and in German Nachhaltigkeit means sustainability. The conventional mutual funds can easily be detected by scanning through the Datastream list per country. I include Australia, Japan and Korea, Switzerland, France, Italy, Ireland, Belgium, the Netherlands and the UK, and Denmark, Sweden and Norway in my analysis for the international returns. Again, all funds in the sample have reported between January 2002 and December 2014, for at least 30 consecutive months and the return data is monthly. Table 1 shows the sample sizes per country. For comparability, the international returns have been converted to US dollars. Country No. of conventional funds No. of SRI funds Australia Belgium Denmark 52 6 France Ireland Italy Japan Korea Norway 48 4 Sweden Switzerland The Netherlands United Kingdom United States Table 1 Overview sample sizes (return analysis) 18

20 3.1.2 Descriptive Statistics Returns In this section I present descriptive statistics for the return data, such as average return and volatility. Graph 1 and 2 below show the average returns and volatility of the returns for the United States. Graph 1 shows that the average returns do not differ much. Taken over the whole period, the bars tend to move in the same direction, not always with the same magnitude but mostly they are not far apart. This might be a first indication that there is little to no performance difference between the two types of funds. Graph 1 - Average returns United States. Calculated monthly over the full sample period ( ) based on the 157 conventional and 111 SRI mutual funds in the US sample. When looking at the average volatility, measured as the standard deviation of the returns, it can be seen that SRI mutual funds tend to be systematically less volatile than conventional mutual funds. This can be seen in graph 2 below. I calculated the volatility for every month over the sample period on a rolling window basis for both the group of SRI as well as the group of conventional funds. Graph 2 Volatility returns United States. Calculated monthly over the full sample period ( ) based on the 157 conventional and 111 SRI mutual funds in the US sample. 19

21 What can also be seen from graph 2 is that upward or downward spikes in the volatility tend to be less strong for the SRI mutual funds than for the conventional funds. This might indicate that indeed during a crisis, SRI mutual funds are a more steady investment compared to their conventional counterparts. Graph 3 and 4 show the returns and volatility during the crisis. It shows that also during the crisis there is little difference between the two types of funds, both in terms of returns and volatility. Volatility for the SRI funds is still systematically lower and appears to responds a little less strong to the changing conditions in the market than the conventional funds. Graph 3 - Average return during the crisis (US funds). Calculated monthly over the crisis period (November 2007 March 2009) based on the 157 conventional and 111 SRI mutual funds in the US sample. Graph 4 - Average volatility during the crisis (US funds). Calculated monthly over the crisis period (November 2007 March 2009) based on the 157 conventional and 111 SRI mutual funds in the US sample. For the international returns, the picture is rather similar. For illustrative purposes, I present here the return and volatility graphs of two countries Japan and Denmark. The graphs for the other countries can be found in the appendix. The general picture is that average returns do not differ much between the two types of funds. Denmark is an exception though. The average returns of the conventional mutual funds are rather small and stable over the period, 20