Social responsibility in mutual funds

Transcription

1 Social responsibility in mutual funds The effect of screening activities per category on mutual fund performance. BACHELOR THESIS Name: Nanda Baars ANR: Faculty: Tilburg School of Economics and Management Program: International Business Administration Supervisor: J. Hu Date: In this thesis I researched the effect of screening activities per category on mutual fund performance. I used the categories environment, social, governance, product, and no screening. I measured mutual fund performance with abnormal portfolio return in , computed by Jensen s alpha using a multiple linear regression and the Fama-French-Carhart model. I found that over investing in socially responsible mutual fund is not beneficial for investors. To be more precise I found that on the environment category no abnormal return can be gained, and in the social, governance, and product categories a 0,1% negative abnormal return was found. Investing in mutual funds without screening does gain a positive abnormal return of 0,1 percent and is beneficial for investors.

2 Table of Contents Table of Contents Introduction Theoretical background Mutual fund performance measurements Jensen s α Capital Asset Pricing Model Fama-French-Carhart 4-factor model Efficient market hypothesis Corporate social responsibility (CSR) Socially responsible investing (SRI) Investor motives for investing in CSR Past research results on SRI mutual fund performance Data and Methodology Data Methodology Categories Hypotheses Category: Environment Category: Social Category: Governance Category: Product Empirical analysis Data Summary Fama French Carhart Regression results Findings Robustness check Conclusion and recommendations Conclusion Recommendations References

3 1. Introduction Nowadays, more and more companies are aware of the damage their business causes to society. Engaging in business activities which do not harm social, environmental and ethical topics is playing a bigger role in the current world. Corporate social responsible (CSR) activities are increasing within companies and are valued by the society. Besides, companies engaging in corporate social responsibility, also investors are getting more involved with the society and environment around them. The rising awareness among investors has caused mutual funds to offer socially responsible investment portfolios next to the already existing risk/return driven stock portfolios which were originally offered. The number of companies engaging in corporate social responsibility (CSR) is increasing rapidly. In the US, the US Social Investment Forum (2012) has found that socially responsible investing has increased by 13% from 2007 till 2010, 12.2% of total US investments under management are involved in sustainable and responsible investing in A mutual fund is an investment vehicle which collects money from smaller investors and manages their funds. A mutual fund's portfolio is structured and maintained to match the investment objectives stated in its prospectus. This leads to more diversification then smaller investors would be able to get when they would create their own investment portfolios. In a normal mutual fund only shareholder expected return is taken into account when making a portfolio. Socially responsible mutual funds build their portfolios not purely on expected return but also take into account investor commitment in social concerns according to (Haigh and Hazelton 2004). Socially responsible investing can take different types of social concerns into account, for example environmental, corporate governance, social and product concerns. Some mutual funds focus more on which industries are good for society and which ones are not, while other funds focus on environmental concerns or product type, like restricting investments on tobacco and alcohol. Socially responsible mutual funds can also decide to follow an ethnical viewpoint and choose not to invest in for example pork. (Haigh and Hazelton 2004) use the following definition: Socially responsible investment (SRI) is a term used to refer to the practice of directing investment funds in ways that combine investors financial objectives with their commitment to social justice, economic development, peace or a healthy environment. Mutual funds engaged in SRI have several methods of building a SRI portfolio. Most commonly used are screening activities. Screening activities actively add or remove socially responsible or not socially responsible companies to/from their portfolios. There are two types of screens commonly used in SRI portfolio management; either one or both screens are used in constructing a portfolio. Negative screens are applied to an initial assets pool from which sectors are excluded, examples of these sectors are gambling, pornography and nuclear. In positive screening companies are selected on superior standards in for example corporate governance or environmental protection. The use of positive screening often uses a best in class approach. Companies are ranked on CSR performance per sector and the best are included in the portfolio, this can also be exercised by using for example the least polluting companies as explained by (Renneboog, Ter Horst et al. 2008b). 2

4 In this thesis I will research if screening per category affects the fund performance. Since former research has found different results and has not taken screening categories into account, this is an opportunity for me to conduct research which contributes to literature. If my hypotheses about screening categories are correct, future fund managers should take screening category into consideration when determining SRI portfolios, and use this information to increase their portfolios performance. My research question is formulated as follows: To what extent does screening per category affect mutual fund performance? Screening per Category Fund performance I will structure this thesis as following; in the following chapter I will review literature written on mutual fund performance, corporate social responsibility, and socially responsible investing. Besides I will discuss relevant theories from which my hypotheses will be derived. I will proceed in chapter three by describing my dataset and clarifying my methodology. In chapter four I will discuss my empirical results and interpret these results to answer my research question. In the last chapter I will conclude my research and discuss contributions and limitations of my research as well as making recommendation for further research. 3

5 2. Theoretical background 2.1 Mutual fund performance measurements Jensen s α (Jensen 1968) was the first to use the α as a measure for mutual fund performance. Nowadays Jensen s α is still widely used as a measure for abnormal portfolio return in mutual funds. He uses α as a measure to calculate abnormal return, the performance of the mutual funds which is not explained by asset pricing models. If the model explains the market perfectly α is zero. In this way he calculates the performance of the mutual fund due to the efforts/knowledge of the fund manager and other fund specifications Capital Asset Pricing Model First determined by (Sharpe 1964), (Lintner 1965) and (Black 1972), the Capital Asset Pricing Model is a financial model which can be used to determine a required rate of return for a portfolio. The model calculates portfolio return taking the market sensitivity ( ), expected market return ( ) into account. The risk free rate ( ) is distracted from the (market)return since, this is the return which can be gained without taking risk. E(R t); Return of portfolio at time t. R f,t; Volatility free rate at time t. α 1; Jensen s alpha, abnormal profit. β MKT; Sensitivity of the market. E(R m,t); Return on the market at time t. The Capital Asset Pricing Model does have some assumptions about investors which causes the model to be inconsistent with the market in which investors trade. The Capital Asset Pricing Model assumes that investors are rational and risk averse, and their only aim is to maximize economic utilities. Furthermore it is assumed investors are price takers which can lend and borrow unlimited amounts under the risk free rate of interest and trade without transaction or taxation costs. Securities in the market are highly divisible into small parcels and all information is available at the same time to all investors Fama-French-Carhart 4-factor model Fama and French added multiple variables to the Capital Asset Pricing Model. They added a company size factor and a book-to-market value factor. They added these factors to improve the original model. (Carhart 1997) added a momentum factor to the equation of Fama and French. Momentum explains the tendency of stock prices which are rising to keep rising and the tendency of stock prices which 4

6 are falling to fall further. This does mean that stock prices which have strong past performance are likely to keep outperforming stocks which have weak past performance. SMB t; Small minus Big company size factor β SMB; Sensitivity of SMB. HML t; High mines Low book-to-market value factor β HML; Sensitivity of HML. MOM t; Momentum factor of Carhart, defines winner/loser bias. β MOM; Sensitivity of momentum Efficient market hypothesis The efficient market hypothesis is first mentioned by (Fama 1970) and implies that all information which is publicly available in the stock market is fully reflected in stock price. There are three different market forms with regard to the information set at which is implemented. A weak marketform implies only information set in historical prices are included in stock prices. A semi-strong market-form implies that publicly available information is implemented into the stock price while a strong market-form implies all available information will be implemented in the stock price within a very short time period after announcement, this also means it is impossible to make profit on actively managed funds. 2.2 Corporate social responsibility (CSR) The issue of corporate social responsibility is discussed often in business and literature. It has become more important in recent years due to the dramatic growth of financial institutes, funds and academic literature which encourage companies to improve their performance on the field of corporate social responsibility according to (El Ghoul, Guedhami et al. 2011). Corporate social responsibility can be defined as continuous implementing social and environmental concerns as well as other concerns by society into a company s strategy and behaviour. This can occur both voluntarily and under pressure of governmental fines or damage to a company s image as found by (Dahlsrud 2006). 2.3 Socially responsible investing (SRI) Socially responsible investing (SRI) has become a trend within the financial market. Since 1960 awareness about socially responsible investing has evolved and since the 21th century also mutual funds engage in SRI. Besides only taking corporate social responsibility into account also the aim at ethical, social and/or moral companies has increased remarkably according to (Ballestero, Bravo et al. 2012). Socially responsible investing is defined by them as following: An investment that is considered socially responsible because of the nature of the business the company conducts. This means there is only be invested in companies which comply with the standard set by the fund. Common themes for socially responsible investments include avoiding investment in companies 5

7 that produce or sell addictive substances (like alcohol, gambling and tobacco) and seeking out companies engaged in environmental sustainability and alternative energy or clean technology efforts. Socially responsible investments can be made in individual companies or through a socially conscious mutual fund or exchange-traded fund (ETF). (Haigh and Hazelton 2004) use the following definition: Socially responsible investment (SRI) is a term used to refer to the practice of directing investment funds in ways that combine investors financial objectives with their commitment to social justice, economic development, peace or a healthy environment. Socially responsible portfolios usually exist of two types of companies. Firstly they exist of companies engaging in corporate social responsibility and secondly of companies who do not. (Ballestero, Bravo et al. 2012) say it is also possible to solely invest in socially responsible companies. As mentioned in the introduction there are two types of screens possible in SRI portfolio management; with positive screening companies are included in the portfolio on a best-in-class method or superior standards in corporate social responsibility. In negative screening companies are excluded if they do not comply with certain CSR measures set by fund portfolio managers. (Dillenburg, Greene et al. 2003) point out that at first only negative screening was used by fund managers to set up a portfolio, filters were used to exclude certain industries and/or companies involved in unethical situations or environmental unfriendly practices. In 2003 negative screening, especially on sin stocks still drove the SRI industry growth. (Renneboog, Ter Horst et al. 2008b) found that by applying positive screening next to negative screening alone, investors have the opportunity to invest in funds with stricter screens. Nowadays there are also portfolios consisting of only positive screens. Nevertheless it is researched before that the number of screens applied negatively influences the portfolio return Investor motives for investing in CSR Investors can have different motives to engage in socially responsible investing next to only caring about risk en return. As mentioned above, (Haigh and Hazelton 2004) say that it is a way for investors to combine their objectives with regard to corporate social responsibility and their financial objectives. (Renneboog, Ter Horst et al. 2008a) also added loss of reputation and not acting as social roles expect as an investor motive. (Becchetti and Ciciretti 2006) analyzed a large sample of socially responsible stocks, defined by the Kinder, Lydenbergand Domini Research & Analytics relative to a control sample for 14 years. They found that the stock risk of companies engaging is corporate social responsibility is slightly lower than risk of comparable companies. This can also be a motive for investors to engage in socially responsible investing. Nevertheless, they found that the risk adjusted return of these companies is not significantly different. (Nilsson 2008) explored the effect of Social, environmental, and ethical (SEE) factors (Pro-social influence), Financial perceptions of SRI (Profit oriented influence) and Socio-demographic variables on investor behaviour and came to the conclusion that not only profit maximization is important for 6

8 investors but also other factors have an influence. Investors, who believe that SEE issues are important and believe they can make a change in business by investing in SRI mutual funds, tend to invest more money in these funds. Furthermore, investors who believe SRI portfolios are currently underpriced and will generate more profit in the long run put a bigger share of their investment into SRI portfolios. The attention of these investors has also changed from only social interference to including environmental concerns into their portfolio. 2.4 Past research results on SRI mutual fund performance The question of fund performance of SRI funds is often raised; several articles have been written on mutual performance. There are two main findings in research. I will first explain the viewpoint of underperformance and afterwards elaborate on outperformance. I will first give several motives for companies to engage in CSR and I will continue with results found in research about mutual funds specifically. (Renneboog, Ter Horst et al. 2008b) argue that SRI mutual funds underperform conventional funds since they limit diversification potential. Hence, mutual funds under invest in attractive industries which do not comply with CSR objectives stated in the fund prospectus. (Diltz 1995) found that adding extra screens has a negative influence on investor performance, a 1% decrease per screen added is found. (Ballestero, Bravo et al. 2012) looked at the percentage of the portfolio invested into green portfolios. More specifically, they studied the difference between strong green investors and weak green investors. They made portfolios in which strong green investors are investors invest 75% of their stocks into green portfolios. Weak green investors have only 25% invested in green stocks. Investors in this study are defined as individuals, companies or funds. The authors found in their research that the return on strong green portfolios is less than on a normal comparable portfolio which is built not taking SRI into account, so investing 0% of the portfolio in green stocks. Weak green investment portfolio return is not significantly different from a normal portfolio. The riskiness of the portfolio of strong green investors is higher than the riskiness of weak green investors, both are above normal. These results imply underperformance of strong green investment portfolios and slight underperformance of weak green portfolios. With regard to outperformance the first motive for companies to engage in CSR is a reduction of conflict costs. (Heal 2005) found that SRI funds outperform conventional mutual funds because CSR reduces conflicts between corporations and society and therefore, costs. In the long run this will lead to higher net present value s of stocks. (Brekke and Nyborg 2008) add to this that CSR in companies will attract and lead to more motivated workers, which will lead to a better company performance. (Porter and Linde 1995) argued that active policies toward corporate social responsibility leads to better shareholder value because resources are used more cost-efficient. (Derwall, Guenster et al. 2005) argue that the extent to which engaging in CSR increases investment returns is dependable on the ability of the market to incorporate the financial consequences into the share price. They mention it is common to think that SRI funds underperform regular funds by limiting diversification. They explain as well that proponents of SRI argue that implementing CSR into businesses can be a sign of managers views on the future performance of the company (longrun view). To conclude, they found in their research that large companies engaging in CSR activities clearly outperform large companies which do not. They also mention it is possible that the 7

9 performance difference can be caused by mispricing of stock issued by companies engaging is CSR activities. Another motive for companies to engage in CSR is eco efficiency, a term which is used to define the relation between pollution (waste) and value added of a company. (Derwall, Guenster et al. 2005) define eco efficiency as the value added of a company minus the waste produced to gain this added value. Also (Diltz 1995) found that environmental screens do increase portfolio performance significantly. This is in line with findings by (Renneboog, Ter Horst et al. 2008b). (Shane and Spicer 1983) found as well that companies which have poor records of pollution control have significantly decreasing stock prices prior to announced pollution reports being published by the Council on Economic Priorities in the United states. (Klassen and McLaughlin 1996), (Hamilton 1995) and (Rao 1996) also found that in general pollution and scoring high on environmental rankings has a positive effect on abnormal stock price. Nevertheless (Mahapatra 1984) found that this relation cannot be verified significantly in the long-run. (Heinkel, Kraus et al. 2001) used a sample of companies which have pollution control technologies and companies which do not, to explore the effect of exclusionary ethical investment on corporate behaviour. They found evidence that companies with high pollution have more difficulties raising capital. This difficulty is due to investors not willing to invest their money in company with low CSR scores. However, they found evidence that at least 25% of investors should stop investing in companies with low CSR scores in order it to have a significant effect on the cost of capital. In 2001 they found the level was around 10%. Therefore, there is no sufficient effect in the market yet. This cost of capital is another motive for companies. For the company the cost of capital should be as low as possible, borrowing from a bank is more costly then raising capital through shares. Also borrowing increases the debt ratio of the company and reduces profit through interest payments. (El Ghoul, Guedhami et al. 2011) researched a large sample of companies in the US from 1992 until 2007 to see if corporate social responsibility has an influence on the cost of capital for a company. They found that the mean cost of capital for companies is 56 basis points lower at a 1% significance level. They argue that engaging in CSR can be beneficial when it comes to cost of capital for the following three reasons. At first the riskiness of a company is perceived as lower because the cost of capital/cost of equity is often used as a measure of riskiness by investors. A lower riskiness of the company is also found by (Becchetti and Ciciretti 2006). Secondly, engaging in CSR reduces information asymmetry and thus agency cost, this lowers cost of equity. Thirdly, the cost of capital will influence investment decisions of companies in the long run, because if you have a lower cost of borrowing capital it is more profitable to invest in new projects which are in line with CSR practices. (El Ghoul, Guedhami et al. 2011)found evidence that companies with CSR-related actions in employee relations, environmental policies and product strategies have significantly lower cost of capital while companies related to tobacco and nuclear power have a higher cost of capital. As you can see from this not only engaging in CSR is important for investors but the industry the company is in influences an investors decision, (Hong and Kacperczyk 2009) found that sin stocks (tobacco, alcohol etc) are underpriced compared to comparable other stocks. The under pricing is 8

10 explained by being neglected by pension funds and other norm constrained institutions. They also found evidence that this affects the stock price and return negatively. Next to the motives for companies to engage in CSR it is important to see what happens when an investor construct mutual fund portfolios of these companies. Screens are used when computing a portfolio. (Diltz 1995) found that screening activities in mutual funds have a significant influence on risk-adjusted returns, especially community involvement increases return. (Velde, Vermeir et al. 2005) researched the effect of sustainability on mutual fund performance, using sustainability ratings from specialized rating agency Vigeo. They used the Fama-French model to determine abnormal portfolio return. They found that portfolios consisting of companies with higher than average corporate governance measures perform better than portfolios of worse performing corporate governance. These results are in line with earlier research conducted by (Bauer, Guenster et al. 2004) and (Gompers, Ishii et al. 2003). (Kempf and Osthoff 2007) researched if it is profitable to follow a long-short strategy (buying stocks with high CSR ratings and selling those with low ratings). They used stock information from the Standard & Poor s 500 and the Domini Social 400 index from 1992 until Ratings from Kinder, Lydenbergand Domini Research & Analytics are used to form portfolios. Abnormal return is measured with the Fama-French-Carhart 4-factor model using Jensen s alpha. They found that one can earn a high positive abnormal profit with this strategy (significant (1% level) alpha between 4% and 5%). Moreover, they also looked at the influence of screens used in this approach. They found that this strategy works when you use positive screens but not when you use negative screens. Using a combination of positive and negative screening does also have a significant alpha between 4% and 5%. This strategy works best when using the best-in-class positive screening approach, selecting only those companies which score very high on a certain value (significant alpha of 8.7% per year). 9

11 3. Data and Methodology 3.1 Data My database exists of 143 socially responsible mutual funds listed on the NASDAQ and by The Forum for Sustainable and Responsible Investment (US SIF). The US SIF was formerly known as the Social Investment Forum (SIF). These mutual funds in the dataset all invest in retail related companies. I will compare these mutual funds against 944 mutual funds which do not engage in screening activities. Funds are sorted into categories. The number of funds per category and which screening type is used per category is defined in Table 1. Also the average monthly return per category is added. Data summary CSR Category # Funds # Positive # Negative # Mixed Environment Social Governance Product No category 944 Table 1 I will use only retail funds in my research since these require lower initial investments and attract also individual investors instead of mostly institutional investors. Since my hypotheses are build on individual investor motives next to past research results this fits my research better. I did not remove funds which start later than January 1 st 2006 or which end before December 31 st 2011 to prevent for a survivorship bias. A survivorship bias is caused since you remove the data of mutual funds which did not perform well and thus stopped existing. If you remove this data your overall performance will be higher since you remove funds with low performance. For the no screening category I excluded all mutual funds which I used in the screening categories. Return data will be determined using the monthly return from the CRSP database over the years 2006 until 2011 to calculate fund performance. Fama-French-Carhart monthly factors are down loaded from Fama s public database. 3.2 Methodology I will use daily return from the CRSP database over the years 2006 until 2011 to calculate average excess market return for each fund using the 4-factor Fama-French-Carhart model as stated below. The 4-factor-model is at the moment the most widely used asset model in papers identifying abnormal portfolio return. In the theoretical background an explanation is given of the formula. US SIF divided social responsibility into four categories: environment, social, governance and product. For each of the four categories I will use the predominant screening type. The predominant screening type is the type used most commonly in this category, which means more 10

12 data can be found. In table 1 the number of funds per screening type is given. I will use the type with the most funds as predominant design. Funds can appear in more than one category. I will test the most relevant screening type per category using a multiple linear regressions using a t-test to test fund performance (Jensen s alpha). With a t-test you test if the regression coefficient is significantly different from 0, I will use a 5% significance level to conclude if the Jensen s alpha is significantly different from 0 and thus if abnormal return can be detected. Furthermore, I will compare the results of the categories including screening with the results of the category without screening. From these findings I will draw my conclusion per category and an overall conclusion. 3.3 Categories US SIF divided social responsibility into four categories: environment, social, governance and product. The first category environment is usually measured by pollution in research. The second category, social, can be described as the way a company impacts its social environment and its stakeholders. Thus it relates to the way a company deals with the community in which it operates, how employees are treated and if there is discrimination on the work floor. Also human rights issues are seen as important by investors. The third category, governance, aims at the implementation of CSR measures by the board. Last characterized is the product category which looks at the product type, for example sin stocks like alcohol, weapons, gambling and tobacco. This also takes into account the way product are manufactures, for example if animal testing is used. 3.4 Hypotheses From the literature review above I defined hypotheses which I will test for all four categories. Taking into account the efficient market hypotheses, which has been proved to be a good guideline several times, and former literature, it is not to expect that the difference between performances will be significant in most cases. As there are two theories about mutual fund performance in relation to socially responsible investing, namely underperformance and outperformance I will have to choose between these two results. Besides, I will determine in my research if using positive or negative screens do have an influence on mutual fund performance, and whether this causes the two results to both be likely Category: Environment In the environmental category the predominant screening type is positive screening. Therefore I will only research the effect of positive screening on fund performance. The research on eco-efficiency is useful for this topic. Pollution is often used as a measure for environmental screening in mutual funds. It implies a correlation between pollution and stock return of a company. Thus better performing companies on CSR should perform better. From these finding I expect the performance for funds screened on environmental topics to be higher than normal mutual funds. 11

13 Cost of capital theory implies that for companies engaging in environmental CSR the cost of capital will be lower than for companies who do not engage in environmental CSR. If the cost of capital is lower, more investment is possible as well as more return on investment. Therefore, these companies have more resources to make profit. A mutual fund which portfolio is build from companies engaging in environmental CSR (among other screens) should therefore be able to generate positive abnormal returns. Research found that positive screening on this subject can generate positive abnormal profit. Nevertheless it is also found that strong green portfolios underperform, so screening should not reduce portfolio diversity too much. Above information leads to the following hypothesis: 1. Positive screening increases fund performance Category: Social Within the social category the predominant screening type is positive screening. For this category the cost of capital also has implications as seen in former research. Companies engaging in bad labour relations and employee discrimination are less attractive for investors as determined by (Nilsson 2008). So less investors means more expensive capital borrowing rates for the company, and this indirectly leads to less profit and stock performance. Companies which have good employee relations should thus attract more investors, this means positive screening should increase fund performance Nevertheless, reducing/excluding companies with human right issues and community involvement is not proven to be beneficial by (Derwall, Guenster et al. 2005). This leads to the following hypothesis: 2. Positive screening increases fund performance Category: Governance In the category governance the predominant screening type is positive screening. (Kempf and Osthoff 2007) found in their research about long-short strategies in mutual funds that positive screens can lead to more abnormal return. Furthermore they found that using the best-inclass approach can generate even more abnormal return. In the governance category a best-in-class approach is often used to determine which companies try to implement CSR into their business strategy the best and the worst. (Velde, Vermeir et al. 2005) also found that using positive screening on governance leads to abnormal return, this is in line with past research by (Bauer, Guenster et al. 2004) and (Gompers, Ishii et al. 2003). This leads to the following hypothesis: 3. Positive screening increases fund performance Category: Product In the product category the only en thus predominant screening type is negative screening. In the product category both product type and product production are taken into account. Stock of companies with a product type which is seen as bad for society is called sin stock. (Hong and Kacperczyk 2009) found that return on sin stocks is less then return on comparative stock. 12

14 The cost of capital has shown that the cost of capital will be higher if companies produce product which are seen as bad for the community (tobacco, alcohol, etc.). In this case screening for these companies and removing them from your portfolio should lead to beneficial results since companies which have more difficulties raising capital have less opportunities to grow and higher cost of capital will decrease profit. Part of profit will eventually be seen in stock performance. Also product production which is not beneficial for society will reduce cost of capital, negative screening on this subject is expected to positively influence the stock performance. This leads to the following hypothesis: 4. Negative screening increases fund performance. 13

15 4. Empirical analysis 4.1 Data Summary Data summary Category Average monthly return Standard deviation Environment 0, ,05157 Social 0, ,05504 Governance 0, ,05126 Product 0, ,05179 No Screening 0, ,04955 Table Fama French Carhart Regression results Environment Beta St Error t-value Sign. t-value Jensen s a -,001,000-1,584,113 Rm-Rf,688,011 62,583,000 SMB,109,022 4,853,000 HML -,100,020-4,942,000 MOM -,043,009-4,602,000 Table 3 Social Beta St Error t-value Sign. t-value Jensen s a -,001,000-2,209,027 Rm-Rf,822,011 74,815,000 SMB,096,023 4,223,000 HML -,085,020-4,161,000 MOM -,056,009-6,065,000 Table 4 Governance Beta St Error t-value Sign. t-value Jensen s a -,001,000 -,001,000 Rm-Rf,811,009,811,009 SMB,109,018,109,018 HML -,073,016 -,073,016 MOM -,036,007 -,036,007 Table 5 Product Beta St Error t-value Sign. t-value Jensen s a -,001,000-2,922,003 Rm-Rf,814, ,142,000 SMB,114,016 7,239,000 HML -,082,014-5,762,000 MOM -,036,006-5,571,000 Table 6 14

16 No screening Beta St Error t-value Sign. t-value Jensen s a,001,000 11,810,000 Rm-Rf,733, ,833,000 SMB,012,006 2,144,032 HML -,112,005-21,674,000 MOM -1,014, ,199,000 Table 7 Category: Environment Social Governance Product No screening Table 8 Capital Asset Pricing model Jensen s α:,000 -,001 -,001 -,001,003 t-value: -,631-1,289-2,060-1,648 12,434 Sign. t-value:,528,198,039,099,000 R Square:,545,744,762,758,481 Fama French 3 factor model Jensen s α: -,001 -,001 -,001 -,001,004 t-value: -1,310-1,834-2,794-2,593 17,638 Sign. t-value:,190,067,005,010,000 R Square:,548,746,764,760,511 Fama French Carhart 4 factor model Jensen s α: -,001 -,001 -,001 -,001,001 t-value: -1,584-2,209 -,001-2,922 11,810 Sign. t-value:,113,027,000,003,000 R Square:,550,749,765,762, Findings I performed a multiple linear regression according to the Fama-French-Carhart model. The goal of this regression is to investigate if there is any existence of abnormal portfolio returns. These abnormal portfolio returns can point out that investments deliver more wealth to shareholders and that there is existence of market inefficiencies or competitive advantages. The results of these regressions are summarized in tables 5 until 7. Table 8 shows the abnormal portfolio returns, the R Square s of the capital asset pricing model and the significance of the t-test, the Fama-French model and the Fama-French-Carhart model. My hypotheses are as follows; 1. Environment: Positive screening increases fund performance. 2. Social: Positive screening increases fund performance. 3. Governance: Positive screening increases fund performance. 4. Product: Negative screening increases fund performance. 15

17 As can be seen in the data summary of table 2 the average monthly return of mutual funds using screening is lower than the average monthly return of mutual funds without screening. In table 5 until 8 the Fama-French-Carhart 4-factor regression can be found, abnormal return is identified and it can be seen if return significantly differs from 0. In the Fama-French-Carhart 4-factor regression results one can seen that the R square is not high on every category. The R square is the coefficient of determination, which ranges from 0 to 1 (0% to 100%) in a multiple linear regression. The R-Square is the proportion of variance in the dependent variable which can be predicted from the independent variables. The higher the R Square is the more the dependent variable (in my research ( ), the expected return on portfolio minus the volatility free rate) is explained by the independent variables. My first hypothesis about environmental screening is not true. I expected the abnormal return to be positive, but it is negative with 0,1% for the time period I used. Compared to the category of funds not engaging in screening activities, one can see underperformance. Nevertheless, the results of this T-test are not significant on a 5% level with 11,3%. This means that the difference from zero cannot be found. Also the R square of the environment category is not high, only 55% of the variance can be explained by the model. According to the efficient market hypothesis no significant abnormal return should be found and my findings with regard to the environmental category are in line with this. I found that mutual funds using screens in the social, governance, and product category underperform mutual funds without screening in any category. All three measures are significantly different from 0 on a 5% level and have a R square above 74,9%. In the category of social screening the result is significant at 2,7%. In the governance category the result of underperformance is significant at 0,2%. Negative screening on the product category underperforms no screening with a significance of 0,3%. The no screening category had a positive abnormal return of 0,001 and the t-test is significantly different from 0. Also the R square is high with 87%. Thus all the categories underperform the category without screening. Literature found underperformance as well as outperformance. Underperformance is researched to be due to limited diversification (Renneboog, Ter Horst et al. 2008b) and using too many screens (Diltz 1995). The effect of the number of screens used by each mutual fund is beyond the scope of my research. Nevertheless (Renneboog, Ter Horst et al. 2008b) and (Diltz 1995) both found outperformance when using environmental screens. My research shows no abnormal return on this category while on the other three categories negative abnormal return is found. Furthermore, earlier research by (Heinkel, Kraus et al. 2001) found that socially responsible screening only generates positive abnormal return if more than 25% of investors use screening activities on environment. As mentioned in the introduction, only 12.2% of investors engaged in socially responsible investing in 2010 according to the US Social Investment Forum (2012). This can explain why I did not find outperformance. I used existing mutual fund portfolios instead of 16

18 constructing portfolios with a fixed percentage in socially responsible mutual funds, also I have taken mutual funds which screen on one or more categories, while former research mostly took only one type of corporate social responsibility into account. (Hong and Kacperczyk 2009) found underperformance due to under pricing of sin stocks, in the product category this can be a reason why I found underperformance as well. To summarize my findings, I found underperformance on three categories. The environment category also gives negative abnormal return, but is not significant, so we cannot conclude any abnormal return in this category. However, the other three categories; social, governance, and product are significantly different from zero and thus a negative abnormal return can be concluded. This means that for the time period of SRI is not only not beneficial but in three out of four categories causes negative shareholder return. Investing in mutual funds without screening does gain a positive abnormal return of 1 percent and is this beneficial for investors. 4.4 Robustness check In order to be sure the Fama-French-Carhart model is robust to alternative models, I performed a robustness check. I performed the same regressions as I did in the Fama-French-Carhart model for the Capital Asset Pricing Model and the Fama-French original model. As can be seen in table 8, the Capital Asset Pricing Model also gives a -0,1% abnormal return but only the governance category is significant at a 5% level with 3,9%. The no screening category shows a higher abnormal return, 0,3% over 0,1% with the Fama-French-Carhart model. All the R squares of this model are lower than under the Fama-French-Carhart model, which means less of the variance is explained by the model. With regard to the Fama-French model, also a -0,1% abnormal return is found for the categories governance and product, while the other two are insignificant. The abnormal return of the no screening category is the highest of all models with 0,4%. All the R squares are higher than the Capital Asset Pricing Model but still lower compared to the Fama-French-Carhart model. As a result, the Fama-French-Carhart model explains variance best. In addition, results are (more) significant with this model. Therefore the results of the Fama-French-Carhart model are robust. 17

19 5. Conclusion and recommendations 5.1 Conclusion In this thesis I researched the effect of screening activities per category on mutual fund performance. I used the categories environment, social, governance, product, and no screening. I measured mutual fund performance with abnormal portfolio return computed by Jensen s alpha. In the theoretical background I discussed mutual fund performance measures and relevant former research to explain what motives there are for firms to engage in corporate social responsibility and to explain the implications of socially responsible screening on the abnormal return of a SRI portfolio. I built my hypotheses on this literature and I expected for all four categories the abnormal return to be positive. In my empirical analysis I used multiple linear regression and the Fama-French-Carhart model to investigate if there was abnormal portfolio return in when using screening activities to create socially responsible mutual fund portfolios. I compared the results of these regressions with the no screening category and I found that over investing in socially responsible mutual fund is not beneficial for investors. To be more precise I found that on the environment category no abnormal return can be gained, and in the social, governance, and product categories a 0,1% negative abnormal return was found. Investing in mutual funds without screening does gain a positive abnormal return of 0,1 percent and is beneficial for investors. I performed a robustness check to see if the Fama-French-Carhart model is robust. The check concluded that the model has the highest R square compared to the Capital Asset Pricing Model and the Fama-French model. Thus the model explains the variance of the variable best for all regressions. According to my findings investors should not invest in mutual funds with screening activities in the social, governance or product category since an abnormal return of -0,1% is found here. Socially responsible mutual funds who screen on the environmental category do not earn any significant abnormal return. By investing in mutual funds which do not engage in screening activities one can gain a positive abnormal return of 0,1%. 5.2 Recommendations I would recommend further research to focus on subcategories to see if subcategories can generate abnormal return. I focussed on categories in the broad sense, while screening can also take place on subcategories like for example community involvement, animal testing or the use of child labour. Especially in the environmental category screening on subcategories could lead to significant abnormal portfolio returns. This is important to be able to fully answer my research question, because if there is existence of positive abnormal portfolio returns, investors can earn profit by investing in socially responsible mutual funds. Furthermore, I limited my research to only retail mutual funds, which require lower initial investments, to limited the scope of my research. Therefore, I did not research the effect of large investment groups, like for example pension funds and insurance companies. This was beyond the 18

20 scope of my research but it can be interesting to research what difference large investment groups can make on this subject, and if mutual funds which require higher initial investments can outperform the market. In addition, also screening per industry can make a difference in abnormal portfolio returns. I did not take industry into account in my research, but industry specific screening can lead to different abnormal portfolio returns per industry. The implication on this subject is that mutual funds only investing in one industry limit their diversification potential even more then when only using screening. Moreover, it is important to research if abnormal portfolio returns are persistent. Since there are multiple results on abnormal portfolio return when taking screening activities into account it might be possible to earn abnormal profit. If a persistent abnormal portfolio return in for example and industry of subcategory can be found, the existence of this persistent abnormal portfolio return can lead to a competitive advantage for companies taking action to meet the requirements of this industry or subcategory. 19

21 6. References (2012). US Social Investment Forum. The Forum for Sustainable and Responsible Investment. Ballestero, E., M. Bravo, et al. (2012). "Socially Responsible Investment: A multicriteria approach to portfolio selection combining ethical and financial objectives." European Journal of Operational Research 216: Bauer, R., N. Guenster, et al. (2004). "Empirical evidence on corporate governance in Europe: The effect on stock returns, firm value and performance." Journal of Asset Management 5: Becchetti, L. and R. Ciciretti (2006). "Corporate Social Responsibility and Stock." CEIS Tor Vergata - Research Paper Series, Vol. 27, No. 79,. Black, F. (1972). "Capital Market Equilibrium with Restricted Borrowing." Journal of Business Vol (45), No.3: Brekke, K. A. and K. Nyborg (2008). "Attracting responsible employees: Green production as labor market screening." Resource and Energy Economics vol. 30: nr. 4 p Carhart, M. M. (1997). "On Persistence in Mutual Fund Performance." Journal of Finance, Vol 52, No.1: Dahlsrud, A. (2006). "How Corporate Social Responsibility is defined: an Analysis of 37 Definitions." Wiley InterScience. Derwall, J., N. Guenster, et al. (2005). "The eco-efficiency premium puzzle." Financial Analyst Journal vol. 61: nr. 2 p Dillenburg, S., T. Greene, et al. (2003). "Approaching Socially Responsible Investment with a Comprehensive Ratings Scheme: Total Social Impact." Journal of business ethics vol. 43 nr. 3 p Diltz, J. D. (1995). "The Private Cost of Socially Responsible Investing." Applied Financial Economics vol. 5: nr. 2 p El Ghoul, S., O. Guedhami, et al. (2011). "Does corporate social responsibility affect the cost of capital?" Journal of banking and finance vol. 35: nr. 9 p Fama, E. F. (1970). "Efficient Capital Markets: A Review of Theory and Empirical Work." Journal of Finance vol. 25: nr. 2 p Gompers, P., J. Ishii, et al. (2003). "Corporate Governance and Equity Prices." The Quarterly Journal of Economics 118 (1): Haigh, M. and J. Hazelton (2004). "Financial Markets: A Tool for Social Responsibility?" Journal of business ethics vol. 52: nr. 1 p

22 Hamilton, J. T. (1995). "Pollution as News: Media and Stock Market Reactions to the Toxic Release Inventory Data." Journal of Environmental Economics and Management 28: Heal, G. (2005). "Corporate Social Responsibility: An Economic and Financial Framework." The Geneva Papers on Risk and Insurance - Issues and Practice vol. 30 (2005) nr. 3 p Heinkel, R., A. Kraus, et al. (2001). "The effect of green investment on corporate behavior." Journal of financial and quantitative analysis 36: Hong, H. and M. Kacperczyk (2009). "The price of sin: The effects of social norms on markets." Journal of Financial Economics, Volume 93, Issue 1: Jensen (1968). "The Performance of Mutual Funds in the Period " The Journal of Finance, Vol. 23, No. 2: Kempf, A. and P. Osthoff (2007). "The Effect of Socially Responsible Investing on Portfolio Performance." European financial management vol. 13: nr. 5 p Klassen, R. D. and C. P. McLaughlin (1996). "The Impact of Environmental Management on Firm Performance." Management Science, 42(8): pp Lintner, J. (1965). "The Valuation of Risk assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets." Review of Economics and Statistics, Vol (47), No. 1,: Mahapatra, S. (1984). "Investor Reaction to a Corporate Social Accounting." Journal of Business Finance and Accounting 11(1), Spring: Nilsson, J. (2008). "Investment with a Conscience: Examining the Impact of Pro-Social Attitudes and Perceived Financial Performance on Socially Responsible Investment Behavior." Journal of Business Ethics vol. 83: nr. 2 p Porter, M. E. and C. V. D. Linde (1995). "Green And Competitive: Ending The Stalemate." Harvard business review : the magazine of thoughtful businessmen vol. 73 nr. 5 p Rao, S. M. (1996). "The Effect of Published Reports of Environmental Pollution on Stock Prices." Journal of Financial and Strategic Decisions, vol. 9, no.1: Renneboog, L., J. Ter Horst, et al. (2008a). "Socially responsible investments: Institutional aspects, performance, and investor behavior." Journal of Banking & Finance vol. 32: nr. 9 p Renneboog, L., J. Ter Horst, et al. (2008b). "The price of ethics and stakeholder governance: The performance of socially responsible mutual funds." Journal of Corporate Finance vol. 14: Shane, P. B. and B. H. Spicer (1983). "Market Response to Environmental Information Produced outside the Firm." The Accounting Review, Vol. 58, No. 3: pp Sharpe, W. F. (1964). "Capital Asset Prices: a Theory of Market Equilibrium under Conditions of Risk." Journal of Finance, Vol (19), No. 3: