An analysis of risk, return, and alpha between sustainable and conventional mutual funds

(1)

1

Faculty of Economics and Business

An analysis of risk, return, and alpha between sustainable and conventional

mutual funds

Jesse Kok 11387912

Date: 30th_{of June 2020} Supervisor: Richard Evers

(2)

2

Abstract

Performance differences between sustainable and conventional mutual funds have been researched in the past. Academic research shows that financial performance results differ per study. Renneboog et al. (2008) found a slight underperformance of socially responsible investment (SRI) funds compared to conventional funds. Bauer, Koedijk, and Otten (2005), Bello (2005), and Utz & Wimmer (2014) indicate no significant performance differences. However, recent data has shown that the nine largest ESG funds in the US have outperformed their market benchmark over the past five years (Benhamou et al., 2020). Therefore, there is a need for up-to-date research. This study has replicated three hypotheses from Utz & Wimmer (2014) regarding excess returns, risk, and market model alpha when comparing SRIs and conventional funds. Utz & Wimmer (2014) found significant differences in risk and alpha. They also found a slight underperformance of SRIs on excess returns, however not significant. This study was unable to find significant differences between any of the performance measures: excess returns, risk, and market model alpha. This study concludes that financially, it does not matter if investors invest in sustainable or conventional funds.

Statement of Originality

This document is written by Jesse Kok, who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of the completion of the work, not for the contents.

(3)

3

1. _Introduction

The process of incorporating Environmental, Social, or Governance concerns (ESG) into investment decisions is known as socially responsible investing (SRI) (Mǎnescu, 2011). Individuals or funds that invest in a socially responsible manner, also known as sustainable investing, are looking to create an impact on environmental, social, or governance matters in addition to generating a financial return. Friedman (1970) said that a business's responsibility is to increase profit and maximize firm value for the shareholders. Renneboog et al. (2008) observe that SRI's goals are not just financial but also social. Nowadays, socially responsible investing is gathering considerable attention in conventional media as well as in academic research. Sustainable investment assets are continuing to climb globally, reaching $30.7 trillion at the start of 2018. This increase is a 34 percent increase in two years, as, at the beginning of 2016, all global assets reached a total of $22.8 trillion (Global Sustainable Investment Alliance, 2018). The indicated increase in sustainable investment assets has been occurring for the past few decades as in 1997, the assets in socially responsible portfolios reached $1.2 trillion (Statman, 2000).

There are various studies about the returns of sustainable investment funds in comparison to conventional investment funds. This subject has been examined numerous times. For example, Statman (2000) compared the returns of these funds between 1990 and 1998. He found that the raw returns of the Domini Social Index, a socially responsible version of the S&P 500, were slightly higher than those of the S&P 500. However, the difference was not statistically significant. Also, Bauer, Koedijk, and Otten (2005), Bello (2005), and Utz & Wimmer (2014) conclude that their studies indicate no significant performance differences. These results differ from Renneboog et al. (2008), who discovered underperformance of SRI portfolios compared to conventional portfolios.

Nevertheless, Benhamou, Chasan, and Kishan (2020) wrote an article about how nine of the biggest ESG mutual funds in the US outperformed the S&P 500 index in 2019. In addition to their excellent performance in 2019, seven of these nine mutual funds beat their market benchmark over the past five years. These funds incorporate ESG, Socially Responsible, Religiously Responsible, Environmentally Friendly, and Clean Energy, or Climate Change in their investment strategies. The returns of SR funds in 2019 sparked the question of whether there is a difference in risks when comparing sustainable versus conventional investment funds.

Utz & Wimmer (2014) used data from 2002-2012 to examine the financial performance of socially responsible (SR) mutual funds compared to conventional mutual funds. Utz and Wimmer (2014) used 36 monthly returns from the US Social Investment Forum (US SIF, 2013). They examined multiple hypotheses. They documented a slight underperformance in the returns of SR funds, although not significant. They also proved that SR fund returns are riskier when compared to conventional funds. In addition to risk and return, Utz & Wimmer (2014) proved that SR funds perform worse according to

(5)

5

alpha. This study will replicate their hypotheses on returns, risk, and alpha with the latest data. This research will contribute to the existing academic research because from 2012 to 2018, the amount of money invested in SRIs worldwide rose from $13.57 trillion to $30.7 trillion (Global Sustainable Investment Alliance, 2018). This increase is a rise of 126%, and new data requires up to date academic testing.

The research question of this study is: will the risk, return, and alpha of SRIs in comparison to conventional funds be the same as when examined by Utz & Wimmer in 2014? This study will explore existing literature on SRI and the risk and returns of SR funds in chapter 2. Chapter 3 will explain the methodology used in this paper, and chapter 4 will contain the data analysis. This paper will conclude with the conclusion & discussion, which will take place in the respective chapters 5 and 6.

2. Literature Review

This chapter will examine previous academic research. Section 2.1 will look at the definition of SRI, what classifies as an SRI portfolio, the possible overlap in sustainable and conventional portfolios, and why different investors choose to invest in a sustainable portfolio instead of a conventional one. After that, chapter 2.2 will look at different ways financial performance measures have been assessed in previous academic research and at the low-volatility anomaly. In section 2.3, this study will review the work of Utz & Wimmer (2014), and to conclude chapter 2; it will describe the hypotheses of this paper.

2.1 SRI

First, we need to understand the definition of socially responsible investment (SRI). According to Sandberg et al. (2008), sustainable investing is to integrate particular non-financial concerns into the otherwise strictly financial-driven investment process. These non-financial concerns are called ethical, social, environmental, or corporate governance criteria. Firms direct their investment capital to gain positive financial returns and combat climate change, ecological destruction, and various social issues. Sandberg et al. (2008) also state that there is no consensus within the SRI movement on how its main characteristic, the integration of non-financial concerns in the investment process, should be defined.

Secondly, we need to understand what type of assets are considered socially responsible and what types are not? Renneboog et al. (2008) mention that SRI applies a set of investment screens based on ESG or ethical criteria to include or exclude assets. In conventional funds, screening does not take place since these funds are solely focused on financial returns. The screening of assets places these assets into two categories: positive and negative. A positive screen meaning the asset is suitable to be placed in

(6)

6

SRIs. When an asset is excluded from SRIs because of screening, the asset received a negative screen. A few screening classifications that SRIs use are visible in Table 1.

Table 1. SRI screens

Screens Definitions Type

Tobacco Avoid manufacturers of tobacco products. – Alcohol Avoid firms that produce, market, or otherwise promote the consumption of alcoholic

beverages. –

Gambling Avoid casinos and suppliers of gambling equipment. – Environment Seek firms with proactive involvement in recycling, waste reduction, and

environmental clean-up. +

Avoid firms producing toxic products and contributing to global warming. – Human rights Seek firms promoting human rights standards. + Avoid firms that are complicit in human rights violations. – Animal testing Seek firms promoting the respectful treatment of animals +

Avoid firms with animal testing and firms producing hunting/trapping equipment or

using animals in end products. –

Renewable

energy Seek firms producing power derived from renewable energy sources +

This table is an example of investment screens used by SRI mutual funds. In the third column, 'Type', a positive screen is indicated by ''+', whereas a negative screen is indicated by a ''-'. This table is a segment of the table created by Renneboog et al. (2008).

Two investment strategies can be concluded from screening. A negative screening strategy implies that an investment fund will remove negative screened assets from an initial asset pool, for example, the S&P 500. The second strategy is a strategy on which SRI portfolios are mostly based on nowadays. These portfolios are based on positive screens, which in practice, summarizes selecting assets that meet superior standards. The approach of positive screening is often combined with a 'best in 'class' method, meaning that firms are ranked within their respective industry or market sector, and only the best are selected (Renneboog et al., 2008).

If sustainable portfolios use a negative screening process, there could be much overlap in assets between a sustainable and conventional portfolio if they use the same asset pool as a base. An SRI will exclude firms, based on factors such as alcohol, tobacco, and animal testing, from an initial asset pool such as the S&P 500. If the conventional fund uses the S&P 500 as their initial asset pool as well, there could be an overlap between the assets of the funds. Koellner (2008) investigated the environmental impacts of conventional and sustainable investment funds and found that the overlap between these two

(7)

7

types of portfolios can be substantial. He further found that the sector allocation of both types of portfolios is more often than not very similar. This corresponds to Renneboog et al. study in 2008. If there is an overlap between the assets of the two portfolios, then there is a possibility that the risk will also be similar.

Investors who invest in a socially responsible manner find themselves in two often complementary categories, according to Schueth (2003). The first group is what Schueth (2003) calls the "feel good" investors. This group invests its money in an approach that is closely aligned with their values and priorities. The second group of investors is more focused on how their money can have a positive impact on society. They feel a strong need to put their investment capital into a manner that improves the quality of life. This need differs vastly from conventional investors who are focused solely on profits. Renneboog et al. (2008) examined that social investors are less likely to move investments from one fund to another compared with conventional investors. The eight-month economic downturn of 2001 proved this since there was a total drop of 94% in the money inflows into all US mutual funds. The money inflow into socially responsible portfolios, however, dropped only 54%. The asset will not get 'panic sold' as much as conventional mutual funds, and therefore will be less volatile in an economic downturn. This study will measure the risk of portfolios as the standard deviation of excess returns. Standard deviation is also a measure of volatility. Less volatility in a crisis would mean a lower standard deviation, and therefore, SRIs would have less risk than conventional funds according to the findings of Renneboog et al. (2008).

2.2 Low-volatility anomaly and financial measures

Contrary to basic finance principles, high-volatility and high-beta stocks have long underperformed low-volatility and low-beta stocks. This is what researchers call the low-volatility anomaly. This is an anomaly since basic finance principle says that high risk is compensated with higher expected return (Baker, Bradley, & Wurgler, 2011). Blitz and Van Vliet (2007) present global findings that stocks with low volatility earn high risk-adjusted returns. Risk-adjusted return is a term also used to describe excess returns. Stocks with higher risk should yield a higher return since investors want to be rewarded for the risk. When comparing two stocks, stock A can have a higher return, but stock B can have a higher risk-adjusted return, meaning that it gained more per unit of risk. Blitz and Van Vliet (2007) find that well-known effects such as size and value cannot explain the volatility effect. Renneboog (2008) proved that SRIs are less volatile. According to the studies of Blitz and Van Vliet (2007), SRIs could thus earn higher excess returns.

The first risk assessment of this study will examine volatility. Volatility is widely used as a proxy for risk. Volatility is often documented as the standard deviation of returns. It is the amount that an 'asset's return varies through successive periods. The uncertainty of an 'asset's return makes the asset

(8)

8

riskier because the value at the time when the investor wishes to sell is less predictable. A higher volatility means a broader range of potential future value (Bekaert & Wu, 2000).

Research finds that the standard deviation of a portfolio can be reduced by diversifying the portfolio. By adding more assets to a portfolio, the volatility reduces if assets are less than perfectly positively correlated. Diversification is a significant factor in the current portfolio theory. The logic behind this is that with enough assets, the portfolio will match the market, and the portfolio will only contain systematic risk.

The Sharpe-Lintner-Black model is referred to as the CAPM (Fama & French, 2004). It measures risk and the relation between expected return and risk. CAPM uses beta as a measure of the risk of a portfolio or security compared to the market. Universities have widely accepted CAPM since it created an opportunity for finance to be an appropriate subject for econometric studies (Dempsey, 2012). Dempsey (2012) also states that, in addition to universities using CAPM, industries have also accepted CAPM to review fund managers and value investments in firms. Beta is used in the CAPM, and this study will also use beta when calculating alpha. However, there have been critics who suggest that beta lacks efficiency and completeness as a measure of risk (Pettengil et al., 1995). Critics challenge the usefulness of beta by at least two arguments. The first argument came from Fama and French (1992), who found no relationship between beta and asset returns. The second argument is that beta does not measure risk and that there is no risk-return trade-off (Roll & Ross, 1994). Based on these findings, Fama & French (1992) propose their 3-factor model.

Fama and French (1992) first published their work on the three-factor-model in 1992 and have continued to work on it since. Besides market risk, they found that factors' size 'risk' and 'value 'risk' are significant for explaining the realized returns of stocks. Fama and French (1992) constructed two factors to account for these risks: SMB and HML. SMB stands for Small Minus Big and is created to account for the Size Premium. Small market cap stocks tend to outperform big market cap stocks. HML stands for High Minus Low and is designed to account for the Value Premium. HML accounts for the outperformance of high book-to-market ratios versus small book-to-market ratios. High book-to-market ratio companies tend to generate higher returns in comparison to the market. The SMB and HML measure the historic excess returns of small caps over big caps and of value stocks over growth stocks.

In response to the Fama and French three-factor model, Carhart (1997) produced the four-factor model. Carhart added a momentum factor. He found that the three-factor model cannot explain cross-sectional variation in momentum-sorted portfolio returns. Therefore, he added a fourth factor. According to Carhart (1997), estimates from the four-factor model frequently differ due to significant loadings on the one-year momentum factor and is, therefore, a better model. The Carhart (1997) four-factor model is

(9)

9

used in the study of Utz & Wimmer (2014) to calculate and compare alphas between SR and conventional funds.

Besides calculating alpha via the Carhart (1997) four-factor model, alpha can also be computed via Jensen's alpha (Jensen, 1968). This method uses fewer variables than the Carhart (1997) four-factor model. Jensen (1968) defined Jensen's alpha as the difference between monthly actual portfolio returns and the monthly benchmark returns. Utz & Wimmer (2014) use, besides the Carhart (1997) four-factor model, Jensen's alpha (Jensen, 1968) to calculate and compare alphas between SR and conventional funds. Jensen's alpha measures the excess return of a portfolio, above or below that predicted by the CAPM. An alpha of zero means that the stock moves entirely similar to the market benchmark. A positive alpha, therefore, states that the stock or fund outperforms the benchmark. This study will use Jensen's alpha (Jensen, 1968) to compare alphas between the two types of funds.

The Efficient Market Hypothesis (Fama, 1970) suggests that portfolio managers cannot outperform the market systematically. EMH refers to the idea that markets are efficient, and all information that could predict the performance of a stock is priced in immediately. Thus, according to EMH, mutual fund managers cannot earn a positive alpha consistently. Therefore, it is expected that both SR and conventional funds will not earn significantly different from zero positive alphas.

2.3 Hypotheses

This section will inspect the study of Utz & Wimmer (2014). It will describe their methods and results. Afterward, this section will describe the hypotheses that this study will be conducting.

This study will replicate a part of the study of Utz & Wimmer (2014). They examined a broad sample of SRIs in comparison to conventional funds regarding financial and ethical parameters. We will only investigate the financial parameters for this study. In their comparison of funds regarding the financial criterion, Utz and Wimmer (2014) presented five different hypotheses. They used data of the financial performance measures of 230 SR mutual funds and for 37,398 conventional mutual funds. Utz & Wimmer (2014) used at least 36 monthly returns from the US Social Investment Forum (US SIF, 2013), data from 2002-2012. The financial performance measure data from Utz & Wimmer (2014) can be seen in Table 2.

(10)

10

Hypothesis F1: Conventional and SR mutual funds do not differ in terms of return.

Utz & Wimmer (2014) concluded that although they documented a slight underperformance of SR mutual funds, this data was not significant. They tested plain excess returns and SR mutual funds underperformed by three basis points monthly. Excess returns were calculated by subtracting the risk-free monthly rate from the fund's monthly returns.

Hypothesis F2: Conventional and SR mutual funds do not differ in terms of risk.

Their study concluded that the SR fund returns are riskier than conventional fund returns. The results showed a 1.00% increase in standard deviation, which is highly significant. Utz & Wimmer (2014) rejected hypothesis F2 on a 1% significance level. They measured risk as the standard deviation of the excess returns. Utz & Wimmer (2014) compared different risk-adjusted performance measures such as the Sharpe ratio and the Treynor ratio. They concluded that both ratios suffer from the fact that negative values are difficult to interpret and hard to use in statistics when comparing a total of 37,628 funds.

Hypothesis F3a: Conventional and SR mutual funds do not differ in terms of market model alpha.

The market model alpha is a measure of excess returns earned by the fund compared to the returns of the fund predicted by the CAPM model. This measure is also called Jensen's Alpha. According to Utz & Wimmer (2014), Jensen's alpha (Jensen, 1968) showed that SR funds perform worse on a 1% significance level than their counterparts. This result is in line with the Efficient Market Hypothesis, which states that funds cannot consistently outperform the market.

Hypotheses F1, F2, and F3a are the hypotheses that this study will be replicating. Besides these three hypotheses, Utz & Wimmer (2014) had two more financial hypotheses:

Hypothesis F3b: Conventional and SR mutual funds do not differ in terms of Carhart four-factor alpha. Hypothesis F4: Conventional and SR mutual funds do not differ in terms of 𝑀𝑀2_.

The results of the Carhart four-factor alpha are almost identical to the results of the market model alpha in the study of Utz & Wimmer (2014), as shown in table 2, as both financial measures of alpha were significant on a 1% significance level. Therefore, due to time constraints, this study has

(11)

11

chosen only to replicate the hypothesis regarding the market model alpha and not also the Carhart four-factor alpha.

Additionally, due to time constraints, the hypotheses F3b and F4 will not be replicated. The following hypotheses will be tested in this study:

Hypothesis 1:

H0: Conventional and SR mutual funds do not differ in terms of return. H1: Conventional and SR mutual funds do differ in terms of return.

Hypothesis 2:

H0: Conventional and SR mutual funds do not differ in terms of risk. H1: Conventional and Sr mutual funds do differ in terms of risk.

Hypothesis 3:

H0: Conventional and SR mutual funds do not differ in terms of market model alpha. H1: Conventional and SR mutual funds do differ in terms of market model alpha.

Instead of researching one hypothesis, this study chooses to research three hypotheses. The choice to use these hypotheses and performance measures are reasonable because they are easily replicable in the future. Also, they have been used by Utz & Wimmer (2014) and therefore it is essential to use the same performance measures to compare the results and answer the research question: will the risk, return, and alpha of SRIs in comparison to conventional funds be the same as when examined by Utz & Wimmer in 2014?

From these hypotheses, it is expected that no differences in financial performance are found between SR funds and conventional funds. If the H0 hypotheses are rejected, these findings could give insights into the development of the SR funds of the last few years and could provide insights for sustainable investing in the future.

3 Methodology

This chapter will consist of two subsections. Section 3.1 will describe the datasets used and the time-period used for each type of fund. Section 3.2 will describe the financial performance measures that this study uses.

(12)

12

3.1 Data

Utz & Wimmer (2014) used data from several sources to obtain information on 230 SR mutual funds and 37.398 conventional funds. The data for their 230 SR mutual funds are gathered from the US Social Investment Forum (US SIF, 2013), and Utz & Wimmer (2014) used at least 36 monthly returns. They gathered the data for the conventional funds by accessing the CRSP Survivor-Bias-Free US Mutual Fund Database. This database is a database that contains monthly returns of more than 59.000 mutual funds in the US. Unfortunately, it is not possible to copy the exact data since Utz & Wimmer (2014) do not explicitly mention which funds they used. Also, this study does not have access to the CRSP Survivorship-Bias-Free US Mutual Fund Database. This study wants to compare the results of new data to the results of Utz & Wimmer (2014). Therefore, this study will also use the US SIF (2019) to select the SR mutual funds and use 36 monthly returns. For hypothesis 3, this study needs a market benchmark. The S&P 500 will be used as such.

US SIF (2019) ranks SR funds based on various factors. This study will use the ten largest ESG Focus Funds (Appendix 9.1). ESG Focus Funds, according to US SIF (2019), are the most significant sustainable funds that incorporate ESG criteria through their investment process. This study will also use the 10 Largest Impact Funds (US SIF, 2019) (Appendix 9.2). Impact Funds focus on bonds whose use of proceeds goes towards projects that build sustainable infrastructure. In addition to those 20 funds, this study will use the 10 Largest Sustainable Sector Funds (Appendix 9.3). These are funds that invest in "green economy" industries like water infrastructure, renewable energy, energy efficiency, and green real estate. All these funds combined, make up the 30 SR funds that this study will use.

The conventional funds are chosen by sorting US funds based on total asset size. This study will use the 50 largest mutual funds, with available data, based in the US. The data for these funds is acquired by downloading the adjusted close of each fund from Yahoo Finance. From there, the monthly returns are calculated. Instead of using at least 36 monthly returns as performed for the calculations of the SR funds, 60 monthly returns will be used when computing all the data for the conventional funds. The reason for this is that some SR funds are newer than conventional funds and therefore have fewer data available to them.

To test the hypotheses, this study needs the actual returns of the SR funds and conventional funds. The data for the 30 SR funds will be obtained by accessing Yahoo Finance. The adjusted-close for the time-period of each fund will be downloaded, and from there, the monthly returns are calculated. Besides the actual returns of the SR funds, this study also needs the market portfolio returns. The data for the S&P 500 will also be acquired from Yahoo Finance.

In addition, this study needs risk-free monthly rates since these rates are included in every hypothesis. The risk-free monthly rates in the study of Utz & Wimmer (2014) are acquired from Kenneth

(13)

13

French's website (http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html). This study will also use Kenneth French's website to gather and compute the risk-free monthly rates. These rates are calculated by the 1-month T-bill return, reported by Ibbotson and Associates, Inc. This method is the method Utz & Wimmer (2014) use as well, and therefore the results are comparable.

The risk-free monthly rate most recent data, gathered from Kenneth French's website, is from April 2020. Therefore, this study will use data of both the SR and conventional funds until April 2020. The period for the SR funds is from April 2017 – April 2020, making it 36 monthly returns. The period used for the conventional funds is from April 2015 – April 2020, making it 60 monthly returns.

3.2 Financial performance measures

This study will research three different hypotheses regarding SR and conventional funds. For hypothesis 1, this study will use the following performance measure:

𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑟𝑟𝑒𝑒𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 = 𝑟𝑟� − 𝑟𝑟𝑃𝑃 � 𝑓𝑓

where the bars on top of the variables indicate average returns over the sample period and: 𝑟𝑟𝑃𝑃 = 𝑓𝑓𝑟𝑟𝑟𝑟𝑑𝑑′𝑒𝑒 𝑚𝑚𝑚𝑚𝑟𝑟𝑟𝑟ℎ𝑙𝑙𝑙𝑙 𝑟𝑟𝑒𝑒𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑒𝑒

𝑟𝑟𝑓𝑓= 𝑚𝑚𝑚𝑚𝑟𝑟𝑟𝑟ℎ𝑙𝑙𝑙𝑙 𝑟𝑟𝑟𝑟𝑒𝑒𝑟𝑟 − 𝑓𝑓𝑟𝑟𝑒𝑒𝑒𝑒 𝑟𝑟𝑟𝑟𝑟𝑟𝑒𝑒

The excess return per month is calculated by taking the monthly return of each fund and subtracting the risk-free monthly rate of the according month. This method will be repeated for all 36 months for the SR funds and all 60 months for the conventional funds. Then, the average excess return per fund over the sample period will be calculated.

Hypothesis 2 states that SR and conventional funds do not differ in terms of risk. Risk is computed by calculating the standard deviation of the excess returns. This study uses the following measure:

𝑟𝑟𝑟𝑟𝑒𝑒𝑟𝑟 = 𝑆𝑆𝑟𝑟𝑑𝑑 �𝑟𝑟𝑃𝑃− 𝑟𝑟𝑓𝑓�

where:

𝑆𝑆𝑟𝑟𝑑𝑑 = 𝑒𝑒𝑟𝑟𝑟𝑟𝑟𝑟𝑑𝑑𝑟𝑟𝑟𝑟𝑑𝑑 𝑑𝑑𝑒𝑒𝑑𝑑𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑚𝑚𝑟𝑟

Hypothesis 3 states that SR and conventional funds do not differ in terms of market model alpha. The formula of Jensen's alpha calculates market model alpha:

𝑟𝑟𝑀𝑀𝑀𝑀= �𝑟𝑟� − 𝑟𝑟𝑃𝑃 �� − 𝛽𝛽𝑓𝑓 𝑃𝑃�𝑟𝑟�� − 𝑟𝑟𝑀𝑀 �� 𝑓𝑓

Where:

(14)

14

Each of these financial measures will be calculated per individual fund. After these results, a Welch t-test will be performed. The Welch t-test is a two-sample test assuming unequal variances. When using different sample sizes, Welch's t-test should be the default t-test used (Delacre et al., 2017). The Welch t-test is the t-test used in the study of Utz & Wimmer (2014) as well, making the results easily interpretable and comparable.

According to our hypotheses and literature review, this study will not find any difference when comparing SR and conventional funds regarding excess returns, risks, and Jensen's alpha (Jensen, 1968).

4. Data analysis

The performance measures, discussed in chapter 3, are applied to the monthly returns of all the funds. For each measure, the differences between the two types of funds are tested by a classical Welch t-test performed in Excel. The Welch t-test allows for different sample sizes and heterogeneous variances (Delacre et al., 2017). Table 3 summarizes the results of the financial measures.

Table 3: Financial performance measures based on monthly data.

corresponding hypothesis Excess return 1 Risk 2 𝒂𝒂𝑴𝑴𝑴𝑴 3 SR funds (n=30) Conventional funds (n=50) 0.00744 0.00998 0.05505 0.06778 0.00136 0.00338 difference -0.00255 -0.01273 -0.00202 Welch t-statistic -0.46106 -0.40180 -0.36595 P-value (Welch) 0.64668 0.68954 0.71589

This table shows the averages of the 30 SR mutual funds' financial performance measures and the 50 conventional mutual funds.

When examining the measure of excess return, this study's research has shown that SR funds underperform the conventional funds by 2.5 basis points. This difference is, however, not significant. The p-value of the t-test for excess return is too high. Therefore, we cannot reject H0 of hypothesis 1: Conventional and SR mutual funds do not differ in terms of return.

(15)

15

When considering the risk of the two types of mutual funds, this study shows that the standard deviation of SR funds is slightly smaller, although, again, not significant. Consequently, we cannot reject H0 of hypothesis 2: Conventional and SR mutual funds do not differ in terms of risk.

Next, this study calculated and compared the market model alpha between SR funds and conventional funds. SR funds perform worse according to market model alpha; however, these results are not significantly different once again. Therefore, we cannot reject H0 of hypothesis 3: Conventional and SR mutual funds do not differ in terms of market model alpha.

To summarize the results of this study, we find no significant financial performance differences in each of the hypotheses. Therefore, this research cannot document an out- or -underperformance of SR mutual funds when compared to conventional funds.

5. Conclusion

This study started with a research question: will the risk, return, and alpha of SRIs in comparison to conventional funds be the same as when examined by Utz & Wimmer in 2014? Utz & Wimmer (2014) used data from 2002-2012. With new academic research and the fact that the nine largest ESG funds, according to Bloomberg, outperformed conventional funds in 2019, the need for new research with up-to-date data was necessary.

The results of Utz & Wimmer (2014) showed that SR funds compared to conventional funds do not differ in terms of excess returns. This research, with the most recent data, also showed that SR funds do not differ in terms of excess returns when compared to conventional funds. These results are in line with the academic research that there are no significant differences in performance (see e.g., Bauer et al., 2005; Bello, 2005).

Also, both funds generate positive alpha. However, these alphas do not statistically differ from zero. Therefore, this study has shown that neither the SR nor conventional funds are gaining above normal excess returns. These results differ from the results gathered by Utz & Wimmer (2014). They did not find that SR or conventional funds were able to generate positive alpha. Instead, Utz & Wimmer (2014) found that both types of funds generated negative alpha. When comparing SR to conventional funds, they proved that SR funds perform worse on a 1% significance level. Utz & Wimmer (2014) showed that this is in line with academic research on mutual fund performance. Academic research says that most actively managed funds underperform a benchmark index (Carhart, 1997). This study found positive alphas, although not significantly different from zero. Therefore, these results are also in line with academic research and the Efficient Market Hypothesis, which states that funds cannot consistently outperform the market.

(16)

16

The last performance measure this research examined was the risk. Utz & Wimmer (2014) found a significant difference in risk when comparing SR funds to conventional funds. They proved that the returns of the SR funds are riskier than of their counterpart. They showed a highly significant increase in standard deviation on the 1% significance level. With more recent data, this study showed that the risks of the excess returns are not significantly different between the two types of funds. This study showed that conventional funds are riskier, although not significantly different.

Considering these results, unfortunately, this study cannot comment on the low-volatility anomaly. The low-volatility anomaly, as discussed in chapter 2.2, says that funds with a lower standard deviation can earn higher excess returns. This study did not find that SR funds or conventional funds have significantly lower volatility.

Although the results of this study are similar to the results of Utz & Wimmer (2014), they are not the same. The descriptive statistics of both studies are close together, and therefore, we can conclude that the results gathered in this research are reliable, no apparent errors can be concluded from the descriptive statistics and results.

This research aimed to examine the risk, return, and alpha of SR and conventional funds in comparison to when it was examined by Utz & Wimmer (2014). Based on statistical tests with more recent data, this study found no significant differences between SR funds and conventional funds in all financial performance measures. Whereas Utz & Wimmer (2014), found a significant difference in risk and alphas. Therefore, this study concludes that financially, it does not matter whether investors invest in sustainable or conventional funds.

6. Limitations and future research

However, some limitations to this research should be noted. Where Utz & Wimmer (2014) uses over 37.000 conventional funds, this study only uses 50. This is because there was limited access to data. Utz & Wimmer (2014) were able to use the CRSP Survivor-Bias-Free US Mutual Fund Database, and this study was not. In addition to the relatively lower amount of funds used, this study also used the largest SR and conventional funds. Therefore, the solution to both these limitations would be to use a higher number of observations. A higher number of observations could bring the p-value down of each performance measure that was tested. The recommendation for future research would be to use a larger number of observations in order to get a better representation of the population.

Also, Utz & Wimmer (2014) used, in addition to Jensen's alpha, the Carhart (1997) four-factor model. Due to time constraints, this study was not able to test the fourth hypothesis calculating alpha using Carhart's four-factor model. The results in Utz & Wimmer (2014) between the two

(17)

17

measures of alpha are consistent, and therefore, this study chose to test one of the alpha measures. However, the recommendation for future research would also be to include the other hypotheses tested in Utz & Wimmer's (2014) study to get a better representation of all performance measures between SR and conventional funds.

When conducting future research, it would be interesting to see if the low-volatility anomaly, researched by Blitz & Van Vliet (2007), will hold when using a better representation of the population of SR and conventional funds.

Future research could, instead of testing three hypotheses like this study, test all the financial hypotheses that Utz & Wimmer (2014) examined. In addition to testing all the financial performance measures in future research, future research could also include the ethical parameters that Utz & Wimmer (2014) examined. Utz & Wimmer (2014) tested different hypotheses on ESG scores when comparing SR and conventional funds. Future research might include these hypotheses.

This research showed that there is no difference in risk between SR and conventional funds. An interesting subject for future research could be to research why there is no difference in risk. As mentioned in the literature review, there may be no difference in risk because of the overlap between assets. Since SR funds generally use a negative screening process, if the same base asset pool is used as the conventional fund, an overlap between assets is highly likely. Future research could test if the risk is not different between the two types of funds because of overlap between assets.

(18)

18

7. Reference list:

Baker, M., Bradley, B., & Wurgler, J. (2011). Benchmarks as Limits to Arbitrage: Understanding the Low-Volatility Anomaly. Financial Analysts Journal, 67(1), 40–54.

Bauer, R., Koedijk, K., & Otten, R. (2005). International evidence on ethical mutual fund performance and investment style. Journal of Banking & Finance, 29(7), 1751–1767.

Bekaert, G., and G. Wu (2000). Asymmetric Volatility and Risk in Equity Markets. Review of Financial Studies, 13, 1-42.

Bello, Z. Y. (2005). SOCIALLY RESPONSIBLE INVESTING AND PORTFOLIO DIVERSIFICATION. Journal of Financial Research, 28(1), 41–57.

Benhamou, M., Chasan, E., & Kishan, S. (2020, the 29th of January). Bloomberg - The Biggest ESG Funds Are Beating the Market. Retrieved from https://www.bloomberg.com/graphics/2020-ten-funds-with-a-conscience/

Carhart, M. (1997). On Persistence in Mutual Fund Performance. The Journal of Finance, 52(1), 57-82. Delacre, M., Lakens, D., & Leys, C. (2017). Why Psychologists Should be Default Use Welch's t-test Instead of Student's t-test. International Review of Social Psychology, 30(1), 92-101.

Dempsey, M. (2012). The Capital Asset Pricing Model (CAPM): The History of a Failed Revolutionary Idea in Finance? Abacus, 49, 7–23.

Fama, E. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. The Journal

of Finance, 25(2), 383-417.Fama, E., & French, K. (1992). The Cross-Section of Expected Stock Returns. The Journal of Finance, 47(2), 427-465.

Fama, E., & French, K. (2004). The Capital Asset Pricing Model: Theory and Evidence. Journal of Economic Perspectives, 18 (3): 25-46.

Friedman, M. (1970). A Friedman Doctrine - The Social Responsibility of Business Is to Increase Its Profits. The New York Times Magazine, the 13th of September, 32-33, and 123-125.

Global Sustainable Investment Alliance (2018). 2018 Global Sustainable Investment Review. Retrieved from http://www.gsi-alliance.org/trends-report-2018/

Jensen, M. (1968). The Performance of Mutual Funds in the Period 1945-1964. The Journal of Finance, 23(2), 389-416.

Koellner, T., Suh, S., Weber, O., Moser, C., & Scholz, R. W. (2008). Environmental Impacts of Conventional and Sustainable Investment Funds Compared Using Input-Output Life-Cycle Assessment. Journal of Industrial Ecology, 11(3), 41–60.

Mǎnescu, C. (2011). Stock returns in relation to environmental, social and governance performance: Mispricing or compensation for risk? Sustainable Development, 19(2), 95–118.

Pettengill, G. N., Sundaram, S., & Mathur, I. (1995). The Conditional Relation between Beta and Returns. The Journal of Financial and Quantitative Analysis, 30(1), 101.

(19)

19

Renneboog, L., Ter Horst, J., & Zhang, C. (2008). Socially responsible investments: Institutional aspects, performance, and investor behavior. Journal of Banking & Finance, 32(9), 1723–1742.

Roll, R., & Ross, S. (1994). On the Cross-Sectional Relation between Expected Returns and Betas. The Journal of Finance,49(1), 101-121.

Sandberg, J., Juravle, C., Hedesström, T. M., & Hamilton, I. (2008). The heterogeneity of socially responsible investment. Journal of Business Ethics, 87, 519-533.

Schueth, S. (2003). Socially responsible investing in the United States. Journal of Business Ethics, 43(3), 189–194.

Statman, M. (2000). Socially Responsible Mutual Funds. Financial Analyst Journal, 56 (3), 30-39.

Utz, S., Wimmer, M. (2014). Are they any good at all? A financial and ethical analysis of socially responsible mutual funds. Journal of Asset Management, 15(1), 72-82.

(20)

20

8. Appendix

8.1 The 10 Largest ESG Focus Funds

(21)

21 8.3 The 10 Largest Sustainable Sector Funds

8.4 50 Largest Conventional Funds based in the US (source: investing.com)

Name Symbol Total Assets (in billions of $)

Vanguard 500 Index Admiral VFIAX 254.93

Fidelity 500 Index Fund FXAIX 224.22

Vanguard Total Stock Market

Index Fund Admiral Shares VTSAX 197.04

Index Institutional Plus Shares VSMPX 181.98

Vanguard Total International

Stock Index Fund Investor Shares VGTSX 133.25

Fidelity Contrafund K FCNKX 128.94

Fidelity Contrafund FCNRX 128.94

Index Fund Institutional Shares VITSX 127.61

Index Fund Investor Shares VTSMX 116.89

Vanguard Total International Stock Index Fund Institutional Plus Shares

VTPSX 116.37

Vanguard Total Bond Market

Index Fund Admiral Shares VBTLX 110.48

American Funds The Growth

Fund Of America Class A AGTHX 100.56

Vanguard Institutional Index

Fund Institutional Plus Shares VIIIX 100.33

Vanguard Institutional Index

Fund Institutional Shares VINIX 90.43

Vanguard Wellington Admiral VWENX 79.4

American Funds American

Balanced Fund Class A ABALX 77.7

American Funds Europacific

Growth Fund Class R-6 RERGX 73.14

(22)

22 Index Fund Institutional Shares

iShares Core S&P 500 ETF IVV 69.26

American Funds The Income

Fund Of America Class A AMECX 68.88

Vanguard Intermediate-term

Tax-exempt Fund Admiral Shares VWIUX 68.31

Stock Index Fund Admiral Shares VTIAX 65.57

Dodge & Cox Income Fund DODIX 60.98

American Funds Investment

Company Of America AIVSX 60.06

American Funds Capital Income

Builder Class A CAIBX 59.44

PIMCO Income Fund Institutional

Class PIMIX 59.2

American Funds Washington

Mutual Investors Fund Class A AWSHX 56.83

American Funds Fundamental

Investors Class A ANCFX 56.57

Dodge & Cox Stock Fund DODGX 51.77

PIMCO Total Return Fund

Institutional Class PTTRX 51.76

Bond Index Fund Admiral Shares VTABX 51.04

Fidelity US Bond Index Fund FXNAX 50.69

T. Rowe Price Value Chip Growth

Fund TRBCX 49.6

Vanguard Total Bond Market

Index Fund Institutional Shares VBTIX 49.5

American Funds Capital World

Growth And Income Fund Class A CWGIX 49.41

Metropolitan West Total Return

Bond Fund Class MWTIX 48.82

Vanguard Primecap Fund

Admiral Shares VPMAX 48.4

American Funds New Perspective

Fund Class A ANWPX 46.27

Fidelity Total Market Index Fund FSKAX 45.52

Doubleline Total Return Bond

Fund Class I DBLTX 44.01

Schwab S&P 500 Index Fund SWPPX 42.55

Vanguard Wellesley Income Fund

Admiral Shares VWIAX 42.1

Vanguard Short-term Investment-grade Fund Admiral Shares

VFSUX 41.15

Vanguard Total Retirement 2025

Fund Investor Shares VTTVX 39.7

American Funds Growth Fund Of

America Class R-6 RGAGX 34.41

Dodge & Cox International Stock

Fund DODFX 34.33

(23)

23 Retirement 2030 Fund

Institutional Shares

American Funds AMCAP Fund

Class A AMCPX 33.12

Franklin Income Fund Class A1 FKINX 31.76

Vanguard Institutional Total Stock Market Index Fund Institutional Plus Shares

An analysis of risk, return, and alpha between sustainable and conventional mutual funds