The costs of socially responsible investing : a comparison of a 2018 Global 100 stock portfolio and traditional indexes

(1)

The costs of socially responsible investing

A comparison of a 2018 Global 100 stock portfolio and traditional indexes

BSc Econometrics and Operations research

Bachelor thesis

ABSTRACT

In this paper the 2018 Global 100 and socially responsible investment are investigated. The costs of socially responsible investing are calculated by making a comparison between an equal share portfolio of all 2018 Global 100 stocks and the S&P 500 and the MSCI world index. Next to that, an optimal portfolio of 2018 Global 100 stocks is constructed by optimising the Sharp ratio of the portfolio which is then compared to the equal share portfolio and the S&P 500 and MSCI. The period on which the research is conducted is January 2015 to December 2017. The costs of socially responsible investment (SRI) turn out to be negative, which means the equal share portfolio outperformed the S&P in the period under review. The optimal portfolio performs even better and consists of a diverse set of twelve companies.

Megan Vollebregt (10606920)

Supervisor: Ms D. Güler

June 2018

(2)

Statement of Originality

This document is written by Student Megan Vollebregt who declares to take full responsibility for the contents of this document.

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

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

(3)

1 Introduction

Sustainable living can both cost and save a lot of money, but there is another way to contribute to sustainability without it changing your daily habits and consumption choices. Socially responsible investing might be the perfect option for people who want to invest their money and also support social responsibility. Every year Corporate Knights publishes the Global 100. The 2018 Global 100 was published on the 22nd of January 2018. It is a list of the 100 most sustainable corporations in the world. Those 100 corporations are listed on stock exchanges around the world. This makes it possible for everyone to invest in sustainability and make a profit at the same time.

Socially responsible investments are a big topic both in existing literature and daily news. Both the risk of sustainable stocks and their returns are comprehensively discussed. An example of this discussion is an interview with three sustainability experts on the Forbes website (MoneyShow, 2017). In this interview an expert explains that many things concerning sustainability have changed in the last few years and that nowadays socially responsible investment factors are even linked to above average performance. A study done by Bloomberg is reviewed in the interview later on and this shows that about 84 percent of millennials is interested in socially responsible investments. The prognosis is that this number will not change significantly while this generation ages. Geczy, Sambaugh and Levin (2005) show that back in 2005 sustainable funds underperformed a non-sustainable portfolio of funds. A more recent paper by Murgaia and Lence (2015) shows that being acknowledged as a socially responsible company has a positive effect on stock returns. Curto and Vital (2014) find a similar positive effect on returns from choosing sustainable indexes over traditional indexes.

While some earlier papers, like Murgaia and Lence (2015) use the event study method to research the effect of social responsibility on stock returns, this paper addresses the stock returns of the 2018 Global 100 companies and calculates an optimal portfolio of sustainable stocks, like Geczy, Sambaugh and Levin do, but from the most recent Global 100 list. Hence, the profitability of socially responsible investing is addressed in a broad sense, rather than addressing the effects on profitability on a single event. The portfolio is optimised using the Sharpe ratio. This is the ratio of the excess return divided by the standard deviation of a stock. The portfolio is compared in terms of risk and return to traditional indexes of which the Sharpe ratios are also calculated. In this way one of the two main questions of this paper is answered, namely what an optimal portfolio of sustainable 2018 Global 100 stocks looks like. The other

(5)

main question of this paper is what the costs or benefits of sustainable investing are compared to traditional investing. This question is answered by comparing an equally divided portfolio of all 2018 Global 100 stocks to the traditional indexes. This study contributes to the literature in two different ways. First, it researches the methods and data of the new 2018 Global 100. Second, it shows what the optimal way of sustainable investing in the Global 100 stocks is and if it is profitable.

Constructing a by the Sharpe ratio optimised portfolio is not trivial, because it is hard to say which combination of stocks is best and in which proportions. Stocks with high returns are not per definition in the portfolio, because high returns commonly come with high volatility. To lower volatility the portfolio will be well diversified. This leads to the hypotheses on the two main research questions which are also explained in more detail in subsection 4.3. The hypothesis on the optimal portfolio is that it is well diversified in all company characteristics, such as the country the companies are established in and the sectors the companies are in. The hypothesis on the costs of socially responsible investing is that these costs are either very low or negative, which means SRI gives higher returns for the same risk as non-sustainable investing.

The remainder of this paper is structured as follows. In section 2, the theoretical framework, relevant papers about investing in socially responsible mutual funds, stocks and indexes are discussed in detail. The theories about socially responsible investing and the recent developments in SRI are discussed as well in this section. In section 3 the data used for this study are discussed. In section 4, the methodology section, the ways of constructing portfolios and how to compare them are discussed and the hypotheses are explained in detail. When the methodology is clearly explained the results of the research are presented in section 5, the results section, and the optimal portfolio is chosen. Finally, section 6 summarises, concludes and gives recommendations for further research.

(6)

2 Theoretical framework

To further analyse socially responsible investments, it is important to know what makes an investment socially responsible and, for this paper in particular, what factors Corporate Knights uses for compiling their list of the 100 most sustainable corporations, the 2018 Global 100. Therefore, subsection 2.1 explains socially responsible investing and subsection 2.2 will point out the developments within SRI. The existing literature about socially responsible index performance, mutual funds and event studies on social responsibility will respectively be explained in subsections 2.3, 2.4 and 2.5. Subsection 2.6 answers the question how the 2018 Global 100 list is compiled. Subsection 2.7 answers the question what distinctions can be made between the Global 100 companies according to existing literature.

2.1 Socially responsible investing

Socially responsible investing (SRI) is the combination of an investment strategy that pursues financial returns as well as a strategy that considers social and/or environmental factors. There are many ways of impacting the world through social responsible investments, but the strategies can be divided in four main types of strategy. The first strategy according to Sparkes (2008) is negative screening and divestment. This strategy is not about choosing the most socially responsible companies to invest in, but about excluding and removing the least socially responsible companies from a portfolio with the means to not invest in them. This can be done based on (recent) news about a company not being socially responsible in any way, for example, the Cambridge Analytica data scandal in which data from millions of Facebook users was manipulated (Fanger, 2018). But it can also be done based on company characteristics, like companies operating in the tobacco or alcohol branch.

The second strategy explained by Sparkes is (shareholder) activism. This strategy is all about asserting social objectives. This can be done using the rights of share ownership at the annual shareholder meeting when the company's performance over the past year is reviewed and plans for the next year are established. Sparkes explains that the third strategy is dialogue or (shareholder) engagement. For this strategy every company invested in should be monitored extensively on their non-financial performance. In shareholder engagement dialogues, investees receive constructive feedback on how to improve their environmental, social and governance policies within their sphere of influence. Bugg-Levine and Emerson (2011) discuss the fourth strategy: positive and impact investing. This is the opposite of the exclusion strategy.

(7)

This strategy consists of choosing to invest in a stock mainly because of their socially responsible strategy and making an impact in this way. This impact is made, because if every investor did this, all companies would be forced to have a socially responsible company strategy. In this investment strategy the financial performance of a stock is taken into account, but these are not as important in the decision to invest in the stock as the sustainable performance of the company.

2.2 Developments in socially responsible investing

The developments in SRI ensure a continuous growth in scientific papers on this subject. These developments happen all over the world. For example, in the United States of America there was a growth of 33% in sustainable investing between 2014 and 2016 (The United States Forum for Sustainable and Responsible Investment 2016). This study concluded that this resulted in a share of over one-fifth of all investments made under professional management in the United States consisted of sustainable, responsible and impact investing in 2016. When investment managers were asked why they invest in socially responsible projects and firms, by the conductors of the SIF US 2016 study, 85% of them answered because of client demand. In 2016 the two top socially responsible priorities in the USA were political spending/lobbying and climate change (US SIF 2016). Eurosif, the Europe-based Sustainable Investment Forum, the European version of US SIF also conducted a study on developments in socially responsible investment in 2016. They found that growth in socially responsible investment is consistent across all the different investment strategies at a European level, with rates ranging from 30% for engagement and voting, up to 385% for impact investing. Impact investing is still confirmed as the fastest growing SRI strategy in Europe in 2016. In the Global Sustainable Investment Review published in 2017 the proportions of SRI of all investments made under professional management of all continents in 2016 are calculated (Global Sustainable Investment Alliance). Both Europe and Oceania have a SRI rate of over 50%. Asia without Japan is lagging behind. In this part of Asia 0.8% of all investments made under professional management is socially responsible. This is the exact same rate as in 2014. Japan has gone from 0 to 3.4% which distinguishes Japan from the rest of Asia that has shown no improvement. Canada has gone from 31.3% to 37.8% in two years. All continents except Europe had a higher rate in 2016 than in 2014. This is not because of less SRI in absolute amounts in Europe, but because the total

(8)

amount of investments went up in a higher pace than the amount of socially responsible investments in Europe.

2.3 Socially responsible index performance

Curto and Vital (2014) compared the performance of sustainable and traditional stock indexes. They studied the performance of four traditional and ten sustainable indexes in the period from 2001 to 2011. Their major finding was that over all periods under analysis the traditional indexes were outperformed by the sustainable indexes. Nonetheless the average stock index returns where not significantly different in a statistical sense. Another conclusion from this paper is that traditional and sustainable indexes influence each other. The last conclusion that is drawn from this study was that sustainable and traditional indexes do not have a long-run relationship and that they can diverge from each other without constraints.

Although the research in general was conducted well, there are two main critiques on this paper. The first one is the small number of indexes it is conducted on, especially the number of traditional indexes. These indexes consisted on average of more companies than the sustainable indexes did, which leads to higher volatility in the sustainable indexes. This already leads to the expectation of higher returns for the sustainable indexes. The second critique is the period of time that was studied. From the start of 2013 the markets have acted very different than in the period before 2012. In march 2013 the S&P broke the outbreak level of October 2007 and since then the market has been on a rise. The most recent maximum level of the S&P 500 is 2873 from January 2018, while the maximum in the period 2001 to 2011 was 1576. This choice of the period of time is not to blame on the writers of this paper, since this paper was published in 2014, but does raise the question if the conclusions in the paper still hold. Hence, this paper, which uses more recent data, contributes to the currently available literature.

2.4 Socially responsible mutual fund performance

Geczy, Stambaugh and Levin (2005) compared portfolios constructed from a list of socially responsible mutual funds to portfolios from a larger set of different kinds of funds including those on the list of socially responsible funds. In this way they try to discover the costs of purely socially responsible investing. For the construction of the portfolios they maximise the Sharpe ratio, which can be seen as the risk-weighted performance of an

(9)

investment. Their conclusion is that the costs of sustainable investment strongly depends on the investors choice of pricing model and their belief in manager skill. Investors who use the Capital Asset Pricing Model and disbelieve in manager skill have the lowest cost of sustainable investments. When the investor uses a model more like the Fama-French (1993) three-factor model or the Carhart (1997) four-factor model or believes in managerial skill, the costs of sustainable investing rise to a significant level.

Fama and French (1993) identify a total of five risk factors, split in two bond market factors and three stock market factors, in the returns on stocks and bonds. The stock market factors are the firm size, the book/market value ratio and an overall market factor. The bond factors are the maturity and the default risk of the bond. These five factors, introduces by Fama and French, turned into a model well-known as the Fama-French model. This model is used to explain returns on stocks and bonds. When used only for calculations on stock prices, this model is also referred to as the Fama-French three factor model. In 1997 Carhart added a fourth factor to the three factor model of Fama and French. This factor was the momentum of a stock. This means whether the stock price has been going up or down in the near past. These models, by Fama and French and Carhart, can be used by regressing the excess return of an asset on an intercept.

The method used by Geczy, Stambaugh and Levin is different from the method of most papers on determining the costs or profits of social responsibility. They construct an optimal portfolio of all possible funds from their dataset. So this optimal portfolio is not restricted to non-SRI funds. Then they constructed an optimal SRI portfolio. So the SRI portfolio is made in the same way as the other portfolio except for the extra constraint of social responsibility. This makes it impossible for the SRI portfolio to have a higher return on investment.

Jones, Van der Laan, Frost and Loftus (2008) compared the performance of 89 ethical Australian funds, which was almost the entire range of Australian socially responsible funds in 2008, to the Australian market performance. To do so they use the Fama-French three factor model. They found that ethical funds significantly underperformed the Australian market in the period of 1986 to 2005. The performance of the ethical funds in the last five years of this sample was the worst, with an average underperformance of the market of around 1.5%. Not all of the 89 funds considered existed from the start of their study, during the sample period the number of funds grew. Many of the SRI funds are less than 5 years old. This could be one of the reasons why the funds performed less in those last five years of the sample. It can be costly to start a new fund and the fund might not have the best fund managers yet. So it is understandable that on average new funds perform worse than market returns in general and not specifically that

(10)

SRI funds performed worse in the period 2000-2005. And in the paper they indeed find that underperformance is more likely for smaller, more recently established funds. This weakens the conclusion of the 1.5% underperformance of the SRI funds. When they controlled for size-effects, book/market value and momentum, the level of underperformance that resulted was statistically significant. The final conclusion of this paper is that socially responsible investing comes at a financial cost, though this may be moderate relative to the performance of the conventional funds.

2.5 Socially responsible investments events effects

The reaction of investors to the announcement of environmental performance and corporate sustainability was analysed several times over the last ten years with the use of event studies. Murguia and Lence conducted such a study in 2014 on the release of the Newsweek’s “Global 100 Ranking” also called the Green Ranking. This is a list on the sustainability performance of companies worldwide, which is comparable to the Corporate Knights Global 100 list. They found that investors do react on the release of the list by an increase in the buying of the stock on the list, but there is a difference between the different positions on the list: The higher a company is on the list, the bigger the increase of the relative stock price. They also found that the relative increase of the price of the US-traded stocks is smaller than of the non-US-traded stocks.

Cheung (2010) analysed the addition and exit of stocks from the Dow Jones Sustainability World Index. This process of adding and deleting stocks from the DJSI happens each year when the RobecoSAM (n.d.) corporate sustainability assessment is conducted. Companies that are no longer in the top ten percent most sustainable of their branch get eliminated from the index and the new most sustainable companies are added. Cheung did not find a significant change in the stock prices on the day of the announcement of the exit or addition, but there is a significant but temporary effect on the actual day of the exit from or addition to the Index. This effect on the stock returns is positive when the stock is added and negative when it is deleted. By the fact that the effect of inclusion and exclusion is only temporary, the conclusion that can be drawn from this paper is that either the event did not give investors new information about the company or that investors did not value corporate sustainability.

(11)

2.6 The 2018 Global 100

The article titled 2018 Global 100 methodology published in October of 2017 on the website of Corporate Knights shows that the 2018 list is compiled using a set of 17 environmental, social and governance factors. A few examples of factors used to determine whether or not a company deserves a spot in the Global 100 list are the percentage of women on the board of a company and CEO-average worker pay ratio. These two examples where social factors, for environmental factors carbon, water and energy productivity are reviewed and a good example of governance factor is the taxes paid ratio. All these factors are taken into account relative to their industry peers using publicly available information. The 2018 Global 100 consists of companies from 22 different countries and are from a wide variety of economic sectors. Although the companies on the list are from all over the world, the list consisted mostly of European companies, with a total share of 59% of the companies being European by origin. This is followed by 22% of North-American companies, 12% Asian, 5% South-American and 2% of the companies were Oceanian.

The 2018 Global 100 methodology article by Corporate Knights also mentions differences with the 2017 methodology. There are two big changes in methodology between the years: The first one is the incorporation of a clean revenue factor in determining the total score of a company. This factor measures the impact of the services and/or products of a company on the environment. The second change in methods is the weighting of the key performance factors to reflect the relative contribution of the industry of every company. This leads for example to a larger weight of the taxes paid ratio for the banking sector. These methodology updates resulted in the fact that the 2018 list featured 47 companies that had not been on a Global 100 list before, a number that is much higher than usual. The changes that Corporate Knights make are supposed to make it possible for every company that is socially responsible in many ways to make it to the Global 100 list regardless of their economic sector and country.

2.7 Distinction between Global 100 companies

The companies in the 2018 Global 100 have many things in common like them being publicly-listed companies with latest gross revenue of a minimum of one billion US dollars and their sustainability on multiple levels such resources, employees and financial

(12)

management. But there are also many differences: They come from different branches and countries and almost half of the companies had not been on a Global 100 list before. This last factor, being a newly sustainable company, has been researched in comparison with profits by López, Garcia and Rodriguez in 2007. They found that newly sustainable companies have lower profits and thus lower stock returns in their first year of being sustainable. This makes sense because in the short run corporate social responsibility can be very costly. They argued that the government should support companies that try to be more sustainable. These days, in 2018, many governments do support sustainability of companies. For example, the Dutch government, Rijksoverheid (n.d.), has several subsidies to promote sustainability for both citizens as well as companies. To check if, despite the subsidies, the theory of López, Garcia and Rodriguez is still accurate, in this study a dummy is added for newly sustainable

companies.

Another difference between the companies is the indicator with the highest score, the category in which a company is the most sustainable. Not every company is sustainable in the same way: some have high environmental scores while others have high social scores for example. Ziegler, Schröder and Rennings (2007) studied the separate effect of environmental and social performance on stock performance. They found that while environmental performance (like carbon and waste productivity) has a positive effect on stock prices, a company’s social concern (measured by for example the CEO-average/worker pay ratio) has a significantly negative effect on their stock price. To check if these conclusions still hold, the score of the companies in the 2018 global 100 is split up into an environmental, a social and a government score.

3 Data

In this section the way the data are obtained and the main characteristics of the data are discussed. First the characteristics of the companies in the global 100 are discussed, then the stock prices of the companies and their characteristics are discussed. Finally, some information about the benchmarks is given.

3.1 Company characteristics

The companies in the Global 100 are very diverse. This has also been discussed in subsection 2.7, but not many numbers were given. The first characteristic on which the companies can be

(13)

distinguished is the country where they are established. This is broken down in continents and countries in table A.1 in the appendix. In total 59 out of the 100 companies are established in Europe. This makes sense because the biggest part, over 50%, of all professionally managed SRI assets also is in Europe (Global sustainable investment alliance, 2017).

GICS SECTOR NUMBER OF COMPANIES

Financials 19 Information Technology 18 Consumer Discretionary 12 Health Care 11 Industrials 11 Consumer Staples 9 Utilities 7 Materials 5 Real Estate 3 Telecommunication Services 3 Energy 2

Table 1: Distribution of sectors of the 2018 Global 100 companies

When looking at the different sectors of the companies, some seem quite overrepresented (table 1, the full version of table 1 with industry information can be found in the appendix as table A.2). Within the sectors, some industries also seem quite overrepresented: 10 of the companies are in banking and 10 are in pharmaceuticals. This has three possible explanations, first it is easier for these kinds of companies to get into the Global 100, secondly there are really more companies in this industry socially responsible enough to make it to the top 100 or thirdly there are more companies active in these sectors globally. When looking further into the data on the 2018 Global 100 companies provided by Corporate Knights the banks have especially high scores on cash taxes paid ratio (all top 33) and as said in subsection 2.6, this ratio has a higher weight in the total score for banks. Whether this is fair or not is up to Corporate Knights who decide about this methodology, which is explained in more detail in subsection 2.6. While it seems to help the banks getting into the Global 100, there were 15 banks on the 2017 Global 100, so the new methodology contributes to the companies in the list being more evenly distributed over the sectors. The pharmaceutical companies all have in common that they have a high R&D/revenue ratio. This makes sense because pharmaceutical companies earn most money by discovering new medicine and cheaper ways of making currently available medicine. The R&D expenses are thus not spent on research on how to be more sustainable, but on how to make more money, this is quite controversial as well.

The final characteristic of the companies that is discussed is whether they have been in a Global 100 list before 2018. A total of 36 companies have not been in a global 100 list in the five years prior to the 2018 Global 100. From the 100 companies, 53 were also in the 2017

(14)

Global 100 and 37 were in the 2016 Global 100 list. Of the 100 companies on the 2017 list 65 were also on the list in 2016.

3.2 Stock prices data

The stock prices of all the Global 100 companies come from Yahoo Finance. For every stock, the prices on the exchange with the highest trading volumes were chosen. The prices used are the daily adjusted closing prices. This means that the prices have already been adjusted for dividends and stock splits that might have happened. The data that has been obtained is from the second of January 2015 until the thirtieth of April 2018. Not all companies were listed in this period, Amundi, Orsted A/S and Hewlett Packard Enterprise Co were not listed yet in January of 2015 and Aberdeen Asset Management and Syngenta AG were not listed anymore in April 2018. For this reason, the data of these five companies will not be used in this analysis. The main analysis is on the data from 2 January 2015 until 29 December 2017. The first four months of 2018 will be used as testing period. Also, not every stock had prices on all day that were under analysis. To resolve this issue, the missing data points were filled in by linear interpolation.

The statistics per company can be found in the appendix in table A.3. To give an overview of the standard deviations and returns of the companies, a histogram and a scatterplot are used. In figure A.1 the distribution of the average annual returns of all Global 100 companies is shown. The S&P had an average return of 11.6% in this period, so from the histogram it is visible that about half of the companies had a higher average return. The average return of all 95 companies from the 2018 Global 100 is 13.2% per year. In figure A.3 the risk is shown in relation to the return of the companies in a scatterplot. The companies that are chosen for the optimal portfolio are mostly in the right lower corner of this graph, because these companies have high returns and low risks, which is optimal for a high Sharpe ratio.

3.3 Benchmarks

Two extra time series sequences need to be added to the returns of the 95 stocks. The first one represents the market returns, which will be used as a benchmark for the portfolio returns and the second one represents the risk free rate. Since the portfolio is constructed from a set of 95 stocks from all over the world there is not one perfect benchmark. Funds are, even if there is

(15)

one as globally diversified as this set of stocks, very subjective to the fund managers choices and therefore do not incorporate as much of the market as is desirable. Therefore, an index is more useful. But most indexes have a common feature like country or sector, this makes finding a perfectly matching index for the same market as the global 100 data very hard. Other studies that conduct a similar research, like Curto and Vital (2014) use the Standard & Poor’s 500 index and the MSCI world index. Therefore the S&P is the market rate chosen as one of this study’s benchmarks. But because this benchmark is not as globally diversified as is desirable another index is added for comparison. Namely, the MSCI World Index, this is a broad global equity index that represents both large and medium market capitalisation (so all market capitalisations of $2 billion or more) equity performance across more than 20 countries. It covers approximately 85% of the market capitalisation, excluding locked-in shares such as those held by insiders, promoters and governments, in each of those countries. MSCI World Index does not offer any exposure to emerging markets. This leads to four countries that are in the 2018 Global 100 not being represented in the MSCI world index, Brazil, South-Korea, China and Taiwan.

Not only the market return needs to be chosen. Also the risk-free rate of return is a variable that does not have one specific number for every period of time. The risk-free rate is the theoretical rate of return of an investment with zero risk. This rate is theoretical because no investment is complete free of risk. Usually treasury bill rates are used for the calculation of the risk free rate variable. However, in times of economic crisis these rate can go below zero which makes not investing a better choice than risk free investments. The risk free rate used in this paper is the daily 13 week treasury bill rate of the United States (U.S. department of treasury, n.d.). As visible in the data, this rate indeed does drop below zero for some days in September and October of 2015, although these number are negligibly close to zero.

4 Methodology

This section explains all the models used in this research and gives the hypotheses with explanations. In subsection 4.1 the Sharpe ratio is discussed and it is explained how this ratio is used to construct portfolios. In subsection 4.2 CAPM is discussed and it is explained why this model is necessary for this research. In subsection 4.3 the hypotheses for the results are formed and their grounds are explained.

(16)

4.1 Sharpe ratio

In 1964 Sharpe introduced a measure to calculate the performance of mutual funds and proposed the term reward-to-variability ratio to refer to it. However, this term was not adopted by other experts in the field, instead they called it the Sharpe ratio. Finally, in 1994 Sharpe embraces the term Sharpe ratio in a new paper about the same subject.

The Sharpe ratio is a ratio that measures the risk weighted return for any type of investment product or portfolio of investments. The higher the Sharpe ratio, the better is the investment. For example, a Sharpe ratio of 2 means you get 2 units of return per unit of risk. Risk is measured with the price volatility of the investment. In general, investors seek the highest return with the lowest risk. This makes the Sharpe ratio a very useful tool for comparing investments. The formula for the ex-ante Sharpe ratio is given by equation 1 which was introduced by Sharpe in 1964.

𝑆 = 𝐸[𝑅−𝑅𝑓]

√𝑉𝑎𝑟[𝑅−𝑅𝑓] = 𝑅𝑝−𝑅𝑓

𝜎𝑝 (1)

In this equation R is the asset return and Rf the risk free rate. E[R - Rf] is the expected value of

the excess of the asset return over the benchmark, the risk free return, this is also known as the risk premium. So what the Sharpe ratio exactly measures is the excess return per unit of deviation in an investment asset. The main assumption when using the Sharpe ratio is that the returns of the asset are normally distributed. For calculating the weights of every asset in the portfolio equation 2, that calculates the Sharpe ratio of the optimal portfolio, needs to be maximised. In this equation γi is the weight of asset i, σi is the standard deviation of asset i, Xi

and Xj represent asset i and asset j and Ri is the return of asset i. The restrictions on this equation

are that all weights (γi) must be at least 0 and at most 1 and must sum up to 1.

𝑆_𝑝 = ∑95𝑖=1𝛾𝑖∗𝑅𝑖

√∑ 𝛾𝑖2∗𝜎𝑖2 95

𝑖=1 +2∗∑94𝑖=1∑95𝑗=𝑖+1𝛾𝑖∗𝛾𝑗∗𝐶𝑜𝑣(𝑋𝑖,𝑋𝑗)

(2)

The formula for an average of Sharpe ratios over a number of periods is given by equation 3. In this equation T is the number of periods and ST is the Sharpe ratio at period T. This equation

was given by Sharpe in his paper from 1994. 𝑆_𝐴 = 1 𝑇√𝑇𝑆1+ 1 𝑇√𝑇𝑆2+ ⋯ + 1 𝑇√𝑇𝑆𝑇 = 1 𝑇√𝑇 ∑ 𝑆𝑖 𝑇 𝑖=1 (3)

In this paper the portfolio is constructed by giving all of the 95 assets a positive weight, so at least 0 and at most 1 for every asset. These weights sum up to 1. Because for the Sharpe ratio it does not matter how big the portfolio is, a total sum of 100 would give the same ratio of assets as a results as a total weight of 1 and this would thus result in the same Sharpe ratio.

(17)

Sharpe ratio is commonly used by investors and mutual fund managers, because it is easy to use and all kinds of assets and combinations of assets can be compared as long as you have price data of the assets. The main assumption of the Sharpe ratio that makes this ratio controversial is that risk equals the standard deviation of an asset. Because an asset that exponentially goes up can have the same volatility as an asset that fluctuates around a certain price, this is not ideal. Going up should be less penalised than going down in an ideal ratio. But because of the ease of use and the advantages like only needing the price data, the Sharpe ratio is chosen to be the best way of constructing a portfolio for this study. In this study the assumption is set that a certain value of every asset has the same transaction costs, so this is the same for every the stocks of every company. Also this transaction cost is negligibly low. This makes it possible to leave the transaction costs out of the Sharpe equation, because although it affects the Sharpe ratio slightly the effect is too small to really affect the final results.

4.2 CAPM

For calculating the expected value of an asset’s excess return over the risk free rate, the CAPM, capital asset pricing model, is used. This model takes into account the asset's sensitivity to market risk, represented by the quantity β, as well as the expected return of the market and the expected return of a risk-free asset. The β of an asset measures how much the price of an asset fluctuates compared with how much the benchmark fluctuates. If an asset price moves exactly in line with the benchmark, then the asset's β is 1. Both the market return and risk-free asset are discussed in subsection 3.3. Under three distribution assumptions, the CAPM shows that the cost of equity capital is determined only by β. These three assumptions are: the α in the CAPM formula (equation 3) is 0, the excess return is independent of non-systematic risk and β times the excess market return yields the excess stock return. The CAPM also is not perfect. It fails the empirical test in the study of Fama and French (2004), and more modern approaches to asset pricing and portfolio selection exist. Still, the CAPM remains popular due to its simplicity and utility in a variety of situations. The formula for the calculation of excess return with CAPM is given by equation 4. This formula was used by Sharpe in his paper in 1964.

𝐸[𝑅] − 𝑅_𝑓 = 𝛼 + 𝛽(𝐸[𝑅_𝑚] − 𝑅𝑓) + 𝜀 (4)

In this equation R is the asset return and Rf the risk free rate. E[R] - Rf is the expected value of

the excess of the asset return over the risk free return. E[Rm] is the expected market return. This makes (E[Rm] - Rf) the expected excess market returns over the risk free return, also known

(18)

as the market premium. The β is the sensitivity of the expected excess asset returns to the expected excess market returns.

With ordinary least square regressions the β’s of all the companies have been computed and used to calculate all assets’ excess return over the risk free rate. CAPM is used because if it would not be used the ex-post Sharpe ratio is used instead of the ex-ante Sharpe ratio. This means that if the CAPM is not used, not the expected excess return is used for Sharpe ratio (ex-ante) but the realised excess returns (ex-post). Ex-post Sharpe ratio is also a commonly used ratio. William Sharpe himself prefers the ex-ante ratio over the ex-post ratio (1994) because of the biasedness of the ex-post ratio.

4.3 Hypotheses

First the hypotheses for the best portfolio are given, then the hypothesis for the cost of sustainable investing is given.

4.3.1 Best portfolio hypothesis

The hypotheses for the optimal portfolio on its return, size and company characteristics are as follows. The portfolio with the highest possible Sharpe ratio will consist of stocks with high returns and the portfolio as a whole will at least yield two times the return of the S&P. Since the portfolio is based on the expected excess returns of stocks which are based on their realised excess returns, the portfolio will have a much higher Sharpe ratio than a non-optimised portfolio like the S&P. Diversification can lower the variation in the portfolio so that it is most likely not optimal to have only one or two stocks in this portfolio. But since certain time series of stock prices, for example from companies in the same sector, correlate with each other, probably not too many stocks will be in the portfolio either. This is because the addition of correlating stocks does not add to diversification. Therefore, the hypothesis on the number of stocks in the optimal portfolio is about 11 to 22 different stocks. Because there are 11 sectors and 22 countries and different countries probably correlate more than different sectors. So if every sector is represented by at least one company in the portfolio, but not every country is the portfolio, the hypothesis is as follows.

𝑯𝒚𝒑𝒐𝒕𝒉𝒆𝒔𝒊𝒔 𝟏: 11 ≤ # 𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑡 𝑠𝑡𝑜𝑐𝑘𝑠 < 22

On the company characteristics the correlation of the stocks is very important. An important paper on portfolio selection and diversification is by Markowitz (1952). He explains that by combining assets that are not perfectly positively correlated, portfolio risk can be

(19)

reduced without lowering the expected return. So the expectation is that the stocks in the optimal portfolio are very diverse, so both the countries the companies are in, as well as the sectors they are in will be diversified. The hypothesis on this subject is that there are no more than three companies from the same country and every sector is represented in the portfolio, but not more than two times the same sector. Since the expectation is that companies in the same sector correlate more than in the same country. This is also because of all the companies in the portfolio are from the 2018 Global 100 list and thus sustainable companies of a certain size. These companies especially the ones in the same sector will correlate more than random companies from around the world.

𝑯𝒚𝒑𝒐𝒕𝒉𝒆𝒔𝒊𝒔 𝟐 ∶ 1 ≤ #𝑡ℎ𝑒 𝑠𝑎𝑚𝑒 𝑠𝑒𝑐𝑡𝑜𝑟 ≤ 2 𝑯𝒚𝒑𝒐𝒕𝒉𝒆𝒔𝒊𝒔 𝟑 ∶ 0 ≤ #𝑡ℎ𝑒 𝑠𝑎𝑚𝑒 𝑐𝑜𝑢𝑛𝑡𝑟𝑦 ≤ 3 4.3.2 Hypothesis on the cost of sustainable investing

The hypothesis on the cost of sustainable investing is that the costs of sustainable investing in the equal size all share portfolio is not significantly different from 0. So the returns of the equal all shares sustainable portfolio will be the same as the returns of the indexes. This is a H0 hypothesis because it can be tested stochastically. The test that is used to do this is the

t-test with a significance level of α = 0.05.

𝑯𝒚𝒑𝒐𝒕𝒉𝒆𝒔𝒊𝒔 𝟒 ∶ 𝐻₀: 𝑅_𝑝 = 𝑅_𝑚 => 𝑅_𝑝− 𝑅_𝑚 = 0 𝐻₁: 𝑅_𝑝 ≠ 𝑅_𝑚

This means that the sustainable stocks on average perform the same as the S&P and MSCI. This is mainly based on the research of Geczy, Stambaugh and Levin (2005) and Curto and Vital (2014). Geczy, Stambaugh and Levin studied best portfolios from a group of funds, first they constructed the best portfolio overall and then with a constraint on sustainability as explained in subsection 2.4. Constraining will always give the same or worse result, but the results were not even that bad after constraining. This is why, in this paper it is expected that the non-optimised portfolio of sustainable stocks will have the same returns as the S&P and MSCI, which are indexes that are not optimised on risk and return. Curto and Vital (2014) studied the returns of traditional indexes and sustainable indexes and found that over all periods under analysis the traditional indexes were outperformed by the sustainable indexes. This is why, in this paper it is expected that the non-optimised, equally distributed all share portfolio of sustainable stocks might even outperform the S&P.

(20)

5 Results

The following results are from an analysis using the ex-post Sharpe ratio. In subsection 5.1 the S&P 500 and MSCI data is analysed and the equal all shares portfolio is constructed, analysed and compared to the S&P and MSCI. In subsection 5.2 the optimal, maximum Sharpe ratio portfolio is constructed, analysed and compared. In subsection 5.3 the companies in the optimal portfolio are analysed. In subsection 5.4 the optimal portfolios for the separate years are analysed. In subsection 5.5 the results are compared to the results of previous researches.

5.1 S&P, MSCI and equal all share portfolio

The results of the analysis based on the S&P data are shown in table 2. The results of the analysis on the MSCI data is given in table 3. Finally the results of the equal size portfolio are shown in table 4.

Table 2: Descriptive statistics on the S&P

Table 3: Descriptive statistics on the MSCI World index

Equal portfolio Rp Rp - Rf σ (Rp-Rf) Sharpe ratio

2015 8.6% 8.6% 14.6% 0.59

2016 16.7% 16.3% 15.3% 1.07

2017 23.5% 22.6% 7.5% 3.00

Average 17.4% 17.0% 13.0% 2.69

Table 4: Descriptive statistics on the equal all shares Global 100 portfolio

S&P Rm Rm - Rf σ (Rm-Rf) Sharpe ratio

2015 1.4% 1.3% 15.6% 0.08

2016 11.8% 11.4% 13.3% 0.86

2017 21.6% 20.7% 6.8% 3.04

Average 11.3% 10.8% 12.3% 2.30

MSCI World Rm Rm - Rf σ (Rm-Rf) Sharpe ratio

2015 -0.3% -0.4% 16.1% -0.02

2016 8.2% 7.8% 15.2% 0.51

2017 23.1% 22.1% 6.6% 3.37

(21)

The returns of these three portfolio show the same upward trend in the period 2015 up to and including 2017. As shown in tables 2, 3 and 4 the returns of the sustainable portfolio are consequently higher than the S&P and MSCI over the period. But when looking at the volatility, the equal Global 100 portfolio also is more volatile than the S&P and the MSCI. The average Sharpe ratio of the equal all share sustainable portfolio is higher than those of the indexes, 2.69 compared to 2.23 and 2.30, this means that per unit of risk the sustainable portfolio gives 2.69 units of return while the S&P gives a return of 2.30 units and the MSCI gives 2.23 units of return per unit of risk.

To calculate the costs of sustainable investing in terms of returns the following H0

hypothesis is tested against the alternative hypothesis. A t-test is used to check if the sustainable portfolio yields higher returns than the S&P, the market return. The results of this test are shown in table 5.

𝐻₀: 𝐸(𝑅_𝑝− 𝑅_𝑚) ≤ 0 𝐻𝑎: 𝐸(𝑅𝑝) > 𝐸(𝑅𝑚)

Mean Standard deviation T- statistic P-value

Rp-Rm 0.68 0.44 1.54 0.06

Table 5: T-test on equal share portfolios returns - S&P returns

So from table 5 it is clear that with a significance level α of 0.10, the H0 hypothesis is rejected

and with a significance level α of 0.05 the H0 hypothesis is not rejected. Comparing this to the

hypotheses from the methodology section, the equal all shares portfolio exceeds the expectation. The costs of sustainable investing are less than 0 although not significant at a significance level α of 0.05. This means that in this case sustainable investing in an equal all share Global 100 portfolio gives the same (at α = 0.05) or better (at α = 0.10) returns as non-sustainable investing in the S&P.

5.2 Best portfolio over three years and its results

The results of the best portfolio analysis are shown in table 6. The distribution of the daily return of the optimal portfolio are shown in figure A.4. Figure 1 is a graph with the values of the S&P and indexed value to the S&P of the equal all share portfolio, the by the Sharpe ratio optimised portfolio and the MSCI. This means that the values of all of these investments have been set to 2058 on 1 January 2015, the value of the S&P. As expected the returns of the optimal portfolio are much higher than those of the equal all shares portfolio and the S&P, while the

(22)

volatility is not much higher. This leads to a better Sharpe ratio for this portfolio than the two other portfolios had. In the graph this is also shown: the by the Sharpe ratio optimised portfolio clearly outperformed the other two while maintaining a low volatile line. All portfolios have increasing returns and increasing Sharpe ratios, this is most likely not tenable and will not persist in the years that follow. In the first four months of 2018, the testing period, the portfolio indeed does not by far match its results in the prior years.

Since the main assumption of the Sharpe ratio is that the returns of the asset are normally distributed, this is tested with the Jarque-Bera test. When testing the returns of the optimal portfolio for normality with the Jarque-Bera test the skewness is -0.31 and the kurtosis is 3.07 which leads to a Jarque-Bera statistic of 0.17. This leads to not rejecting normality for this sample size (n=780) with a α = 0.05. So the Sharpe ratio gives a reliable result for the optimal three-year portfolio.

Best portfolio Rp Rp - Rf σ (Rp-Rf) Sharpe ratio

2015 34.0% 33.9% 13.6% 2.50

2016 47.0% 46.7% 14.5% 3.21

2017 58.8% 57.9% 11.9% 4.88

Average 46.2% 45.8% 13.4% 6.12

Table 6: Descriptive statistics on the optimal portfolio

Figure 1: Stock prices of the indexes and the constructed portfolios

5.3 Best portfolio companies and their characteristics

In table 7 the companies in the optimal portfolio are shown in order of share of the total portfolio. A total of twelve companies made it into the portfolio. The weights as calculated

(23)

using equation 2 in subsection 4.1 are shown in the third column of the table. The three companies with the smallest shares all have a share of less than 1%. While the company that forms the biggest part of the portfolio has a share of over 23%. As the hypotheses expected in the methodology section both the countries and the sector are very diverse. Though three of the companies are based in the United States and two are in semiconductors it is fair to say that the portfolio as a whole is quite diversified. Also the companies are very diverse in their returns and standard deviations. Short sales were prohibited in the construction.

place in 2018 Global 100 company part of optimal portfolio average return per year

country sector industry

33 Biomerieux SA 23.3% 41% France Health Care Health Care Equipment & Supplies

40 Svenska Cellulosa SCA AB 16.5% 52% Sweden Consumer Staples Household Products

23 McCormick & Company Inc 14.0% 14% United States Consumer Staples Food Products

95 Umicore SA 10.9% 41% Belgium Materials Chemicals

74 Taiwan Semiconductor Manufacturing Co Ltd

8.5% 22% Taiwan Information Technology

Semiconductors & Semiconductor Equipment

59 NVIDIA Corp 7.7% 115% United States Information Technology

Semiconductors & Semiconductor Equipment

2 Neste Oyj 7.6% 43% Finland Energy Oil, Gas & Consumable Fuels

92 Johnson & Johnson 7.3% 13% United States Health Care Pharmaceuticals

66 Chr Hansen Holding A/S 2.8% 30% Denmark Materials Chemicals

29 Halma PLC 0.7% 19% United Kingdom Information Technology

Electronic Equipment, Instruments & Components

76 Banco Santander Brasil SA 0.5% 31% Brazil Financials Banks

38 Storebrand ASA 0.2% 32% Norway Financials Insurance

Table 7: Statistics on the companies in the optimal portfolio

5.4 Best portfolios over 2015, 2016 and 2017 and characteristics

The best portfolios for all separate years have also been computed by optimising the ex-post Sharpe ratio. In which companies this resulted and their weights in the portfolios is presented in table A.4 in the appendix. The Sharpe ratios per year of all four optimal portfolios are shown in table 8. This table makes it clear that a good Sharpe ratio in one year is no guarantee that the portfolio will perform as well in the nest year. All companies from the optimal three-year portfolio are represented in at least one of the one year portfolios. The 2017 portfolio is the most diversified with 26 companies represented in it, the other portfolios have about 10 companies in it.

Sharpe ratio 2015 Sharpe ratio 2016 Sharpe ratio 2017 Average Sharpe ratio

2015 portfolio 3,85 2,39 2,95 5,31

2016 portfolio 0,24 5,24 2,45 4,58

2017 portfolio 1,14 1,14 7,46 5,62

Three year portfolio 2,50 3,21 4,88 6,12

(24)

5.5 Comparison of results to previous research

The equal all 2018 Global 100 shares portfolio outperformed the S&P and MSCI, this means that by the research in this paper it is concluded that the costs of sustainable investment are negative in as compared to the S&P and MSCI. This means sustainable investing does not even have to reduce the profit you make with your investment when you invest in the Global 100 shares. The research by Geczy, Stambaugh and Levin from 2005 concluded that for different models the costs of sustainable investment vary but there were always costs to sustainable investing. So this might have changed during the last few years. It is possible that during the research from 2005 the researched funds were newly sustainable and being a newly sustainable company, as researched by López, Garcia and Rodriguez in 2007, lowers profits and thus lower stock returns. Or the funds might have contained some stocks of newly sustainable companies which will have the same effect.

Jones, Van der Laan, Frost and Loftus (2008) compared the performance of ethical Australian funds to the Australian market performance. They found that ethical funds significantly underperformed the Australian market in the period of 1986 to 2005. Since this study was also conducted on data up to 2005 the same conclusions can be drawn for this research as for the research of López, Garcia and Rodriguez. The sustainable market was still developing, which is a costly process. As could be read in the theory section, many of the SRI funds in the sample were less than five years old by 2005.

So this study is more in line with the research of Curto and Vital from 2014, this makes sense since their research was done later and therefore in a more developed sustainable stock market. They found that traditional indexes were outperformed by sustainable stock indexes in all periods under analysis. Looking only at the returns this is also the case for this research, but when looking at the Sharpe ratio, the S&P performed slightly better in 2017. Curto and Vital also concluded that traditional indexes and sustainable indexes influence each other. The regression in this current research also shows a very high correlation between the S&P and the sustainable all shares portfolio.

5.6 CAPM

To check is the ex-post Sharpe ratio gives robust results for the optimal portfolio, the CAPM is used to calculate ex-ante Sharpe ratios and an optimal portfolio. In figure A.2 the distribution of the CAPM β’s of all companies is shown. These 95 β’s are obtained by regressing the excess

(25)

asset return on the excess market return. Here it is visible that about 20 companies have a β higher than 1 and only three companies have a β below zero. A β higher than 1 means that the volatility is higher than the market volatility. A β below zero means that the volatility is in the opposite direction from the market volatility. The results from the CAPM Sharpe optimisation are shown in table A.5. These results indicate that the ex-post Sharpe ratio is not completely robust. The optimal CAPM portfolio consist for over 23% of Nordea bank, which was not in any of the optimal portfolios before. Biomerieux, Umicore and Svenska do have a share of over 10% in both the optimal three year portfolio and the optimal CAPM portfolio, so the there is some robustness to the ex-post Sharpe ratio compared to the ex-ante Sharpe ratio.

(26)

6 Conclusion

What a by the Sharpe ratio optimised portfolio of 2018 Global 100 stocks is and what the costs of socially responsible investment are has been answered in this paper.

The first question is answered by allowing all shares of the companies in the portfolio to be at least 0 and at most 1 and all together sum up to 1, while maximising the Sharpe ratio of the portfolio. This leads to 12 stocks being part of the portfolio. In this portfolio are 3 United States companies and the companies are from 8 unique countries. In the portfolio even Asia and South-America are represented which shows how important diversification is to optimise a portfolio. Also besides 2 companies in Semiconductors and Semiconductor Equipment, all industries are unique. These findings are in line with the hypothesis on the number of companies in the optimal portfolio and their diversity in countries and sectors. These findings are also in line with the conclusion that sustainable investing is not bad for returns as long as the investor keeps choosing a well-diversified, fully or partly sustainable, portfolio.

The second question is answered by comparing an equal shares all stocks 2018 Global 100 portfolio with the MSCI over a period of three years, namely 2015 till 2017. From the results it is clear that using the ex post Sharpe ratio the equal Global 100 outperformed the MSCI. So the costs of SRI are negative. And expressed in average Sharpe ratio over three years the costs are 2.23-2.69 = -0.47 units of return for every unit of risk. This finding contradicts earlier research by Geczy, Stambaugh and Levin (2005) and Jones, Van der Laan, Frost and Loftus (2008) who found that for various methods the costs of sustainable investing vary, but there are always costs to sustainable investing. This might show that sustainable investing is not in their early costly stages anymore, but becoming a worthy form of investing for investors looking for the highest possible returns while minimising their risk.

In further research, because of the controversy surrounding the Sharpe ratio, the calculations could be done with different ratios and measures to test the optimal portfolio and possibly construct a more optimal portfolio. Some ratios and measures that could be used are Treynor ratios, Jensen's alphas and the Modigliani risk-adjusted performance measure. Also the expected returns could be calculated in different ways to test the robustness of the current optimal portfolio. Finally, more sustainable portfolios could be constructed by for example combining sustainable indexes and sustainable funds to make a very well diversified portfolio and optimising the Sharpe ratio or any ratio used even further.

(27)

7 References

2018 Global 100 methodology. (2017, October 1). Retrieved April 9, 2018, from http://www.corporateknights.com/reports/2018-global-100/2017-global-100-methodology-2-15068539/

2018 Global 100 results. (2018, January 22). Retrieved April 9, 2018, from http://www.corporateknights.com/magazines/2018-global-100-issue/2018-global-100-results-15166618/

Bugg-Levine, A., & Emerson, J. (2011). Impact investing: Transforming how we make money while making a difference. Innovations: Technology, Governance, Globalization, 6(3), 9-18.

Carhart, M. M. (1997). On persistence in mutual fund performance. The Journal of finance, 52(1), 57-82.

Cheung, A. (2011). Do stock investors value corporate sustainability? Evidence from an event study. Journal of Business Ethics, 99(2), 145-165.

Curto, J. D., & Vital, C. (2014). Socially responsible investment: a comparison between the performance of sustainable and traditional stock indexes. Journal of Reviews on Global Economics, 3, 349-363.

Eurosif. (2016). European SRI Study 2016. Retrieved April 29, 2018, from http://www.eurosif.org/sri-study-2016/

Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of financial economics, 33(1), 3-56.

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

Fanger, D. (2018, April 13). Facebook shares are not as 'socially responsible' as many

investors thought. Retrieved April 27, 2018, from

https://www.cnbc.com/2018/04/13/facebook-privacy-scandal-a-black-eye-for-sri-funds-that-own-the-stock.html

Geczy, C., Sambaugh, R., & Levin, D. (2005). Investing in Socially Responsible Mutual Funds. Wharton School Working Paper. University of Pennsylvania.

Global sustainable investment alliance. (2017) Global Sustainable Investment Review 2016.

Retrieved June 6, 2018, from

(28)

Jones, S., van der Laan, S., Frost, G., & Loftus, J. (2008). The Investment Performance of Socially Responsible Investment Funds in Australia. Journal of Business Ethics, 80, 181-203.

López, M. V., Garcia, A., & Rodriguez, L. (2007). Sustainable development and corporate performance: A study based on the Dow Jones sustainability index. Journal of Business Ethics, 75(3), 285-300.

Markowitz, H. (1952). Portfolio selection. The journal of finance, 7(1), 77-91.

MoneyShow. (2017, August 16). Socially-Responsible Investing: Earn Better Returns From

Good Companies. Retrieved April 9, 2018, from

https://www.forbes.com/sites/moneyshow/2017/08/16/socially-responsible-investing-earn-better-returns-from-good-companies/#4e551a91623d

Murgaia, J. M., & Lence, S. H. (2015). Investors’ Reaction to Environmental Performance: Global Perspective of the Newsweek’s “Green Rankings”. Environmental and Resource Economics, 60(4), 583-605.

Rijksoverheid. (n.d.). Overheid steunt groene groei economie. Retrieved April 29, 2018, from https://www.rijksoverheid.nl/onderwerpen/duurzame-economie/groene-groei

RobecoSAM. (n.d.). The Corporate Sustainability Assessment at a glance. Retrieved May 1,

2018, from

http://www.robecosam.com/en/sustainability-insights/about-sustainability/corporate-sustainability-assessment/index.jsp

Sharpe, W. F. (1964). Capital asset prices: A theory of market equilibrium under conditions of risk. The journal of finance, 19(3), 425-442.

Sharpe, W. F. (1994). The sharpe ratio. Journal of portfolio management, 21(1), 49-58. Sparkes, R. (2008). Socially Responsible Investment (Herz. ed.). Bognor Regis, United

Kingdom: John Wiley And Sons Ltd.

The Forum for Sustainable and Responsible Investment. (2016). Sustainable, Responsible and

Impact Investing Trends 2016. Retrieved April 28, 2018, from

https://www.ussif.org/files/SIF_Trends_16_Executive_Summary(1).pdf U.S. department of treasury. (n.d.) Resourse Center. Retrieved June 5, 2018, from

https://www.treasury.gov/resource-center/data-chart-center/interest-rates/Pages/TextView.aspx?data=billRatesYear&year=2015

Ziegler, A., Schröder, M., & Rennings, K. (2007). The effect of environmental and social performance on the stock performance of European corporations. Environmental and Resource Economics, 37(4), 661-680.

(29)

Appendix

Table A.1: Continent and country distribution of the 2018 Global 100 companies

Continent Country Number of companies

Europe France 15 United Kingdom 10 Germany 6 Finland 5 Sweden 5 Switzerland 4 Netherlands 3 Denmark 3 Belgium 2 Spain 2 Norway 2 Italy 1 Austria 1

North-America United States 18

Canada 4 Asia Japan 4 South Korea 3 Singapore 3 China 1 Taiwan 1 South-America Brazil 5 Oceania Australia 2

(30)

GICS SECTOR GICS INDUSTRY NAME NUMBER OF COMPANIES

Financials Banks 10

Insurance 5

Capital Markets 3

Thrifts & Mortgage Finance 1

Information Technology Semiconductors & Semiconductor Equipment 6

Communications Equipment 3

Electronic Equipment, Instruments & Components 3

Software & IT Services 3

Technology Hardware, Storage & Peripherals 3

Consumer Discretionary Automobiles 5

Media 3

Other 4

Health Care Pharmaceuticals 10

Health Care Equipment & Supplies 1

Industrials Electrical Equipment 4

Industrial Conglomerates 3

Machinery 3

Construction & Engineering 1

Consumer Staples Food Products 3

Beverages 2

Personal Products 2

Other 2

Utilities Electric Utilities 5

Other 2

Materials Chemicals 4

Metals & Mining 1

Real Estate Real Estate Management & Development 2

Equity Real Estate Investment Trusts (REITs) 1

Telecommunication Services Diversified Telecommunication Services 3

Energy Oil, Gas & Consumable Fuels 2

(31)

Name company Country Sector Industry Place in top 100 β (CAPM) Avarage yearly return

Dassault Systemes SE France Information Technology Software 1 0,468 21%

Neste Oyj Finland Energy Oil, Gas & Consumable Fuels 2 0,576 43%

Valeo SA France Consumer Discretionary Auto Components 3 0,862 30%

Ucb SA Belgium Health Care Pharmaceuticals 4 0,529 3%

Outotec Oyj Finland Industrials Construction & Engineering 5 1,034 17%

Cisco Systems Inc United States Information Technology Communications Equipment 7 1,110 15%

Autodesk Inc United States Information Technology Software 8 1,360 21%

Siemens AG Germany Industrials Industrial Conglomerates 9 0,628 11%

Samsung SDI Co Ltd South Korea Information Technology Electronic Equipment, Instruments & Components 10 0,740 32%

Aareal Bank AG Germany Financials Thrifts & Mortgage Finance 11 0,890 10%

Enbridge Inc Canada Energy Oil, Gas & Consumable Fuels 12 1,068 -3%

Merck & Co Inc United States Health Care Pharmaceuticals 13 0,869 3%

Natura Cosmeticos SA Brazil Consumer Staples Personal Products 14 0,533 6%

Pearson PLC United Kingdom Consumer Discretionary Media 15 0,853 -14%

Amadeus IT Group SA Spain Information Technology IT Services 16 0,565 25%

Bayerische Motoren Werke AG Germany Consumer Discretionary Automobiles 17 0,841 3%

Companhia Energetica de Minas Gerais CEMIG

Brazil Utilities Electric Utilities 18 1,061 -12%

Koninklijke Philips NV Netherlands Industrials Industrial Conglomerates 19 0,797 15%

Allergan plc United States Health Care Pharmaceuticals 20 1,041 -14%

Honda Motor Co Ltd Japan Consumer Discretionary Automobiles 21 0,958 8%

Sanofi SA France Health Care Pharmaceuticals 22 0,758 2%

McCormick & Company Inc United States Consumer Staples Food Products 23 0,656 14%

Commonwealth Bank of Australia

Australia Financials Banks 24 0,271 6%

Vivendi SA France Consumer Discretionary Media 25 0,639 11%

Intel Corp United States Information Technology Semiconductors & Semiconductor Equipment 26 1,169 12%

Itron Inc United States Information Technology Electronic Equipment, Instruments & Components 27 0,984 18%

Telefonaktiebolaget LM Ericsson Sweden Information Technology Communications Equipment 28 0,989 -13%

Halma PLC United Kingdom Information Technology Electronic Equipment, Instruments & Components 29 0,300 19%

Deutsche Boerse AG Germany Financials Capital Markets 30 0,799 21%

Kesko Oyj Finland Consumer Staples Food & Staples Retailing 31 0,543 20%

Television Francaise 1 SA France Consumer Discretionary Media 32 0,567 5%

Biomerieux SA France Health Care Health Care Equipment & Supplies 33 0,141 41%

AstraZeneca PLC United Kingdom Health Care Pharmaceuticals 34 0,851 7%

Nokia Oyj Finland Information Technology Communications Equipment 35 0,747 -14%

BNP Paribas SA France Financials Banks 36 1,001 8%

Eli Lilly and Co United States Health Care Pharmaceuticals 37 0,886 9%

Storebrand ASA Norway Financials Insurance 38 0,761 32%

ABB Ltd Switzerland Industrials Electrical Equipment 39 0,942 12%

Svenska Cellulosa SCA AB Sweden Consumer Staples Household Products 40 0,518 52%

Intesa Sanpaolo SpA Italy Financials Banks 41 1,119 9%

Analog Devices Inc United States Information Technology Semiconductors & Semiconductor Equipment 42 1,249 20%

Applied Materials Inc United States Information Technology Semiconductors & Semiconductor Equipment 43 1,403 29%

Takeda Pharmaceutical Co Ltd Japan Health Care Pharmaceuticals 44 0,253 12%

Schneider Electric SE France Industrials Electrical Equipment 45 0,862 8%

Shinhan Financial Group Co Ltd South Korea Financials Banks 46 0,254 7%

Kering SA France Consumer Discretionary Textiles, Apparel & Luxury Goods 47 0,831 39%

Ingersoll-Rand PLC United States Industrials Machinery 48 1,129 14%

The costs of socially responsible investing : a comparison of a 2018 Global 100 stock portfolio and traditional indexes