firm is added to the Dow Jones
Sustainability Europe Index?
Sven Hiskemuller van der Zijden 10433333
Amsterdam, 1st February 2016
Bachelor thesis finance and organization Academic year: 2015-2016
Supervisor: Rob Sperna Weiland Semester 1, period 2 and 3
2
Statement of Originality
This document is written by Student Sven Bafo Hiskemuller van der Zijden 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 completion of the work, not for the contents.
3
Abstract
In this research I investigate whether investors appreciate the inclusion of a firm to the Dow Jones Sustainability Europe Index by doing an event study on the share prices. For this research I use the additions of 2015, a total 25 firms were added to the fund. The results of the event study shows that investors do not value an addition to the Dow Jones Sustainability Europe Index, the results show a small negative response to the inclusion, which is not in accordance with the existing literature.
4
Table of content
Statement of Originality ...2 Abstract ...3 Table of content ...4 1. Introduction...5 2. Literature...6 2.1 Sustainability...6 2.2 Sustainable funds ...82.3 Comparable research and the added value to the existing literature...9
3. The Dow Jones Sustainability Europe Index...11
4. Methodology and data ...13
4.1 Data...13 4.2 Methodology...13 4.2.1 The CAAR ...13 4.2.2 The CAR ...16 4.3 Hypothesis ...16 5. Results...18
5.1 The cumulative average abnormal return...18
5.2 The cumulative abnormal return ...20
6. Conclusion and discussion ...21
6.1 Conclusion ...21
6.2 Discussion ...21
Reference list...23
Appendix 1, the additions to the Dow Jones Sustainability Europe Index in 2015 ...26
Appendix 2, market model regression output ...27
Appendix 3, the average abnormal return per day ...28
5
1. Introduction
Sustainability is the general term for different aspects like corporate social responsibility, corporate governance and corporate responsibility (Robinson et al, 2011). Firms with high levels of sustainability tend to perform better than companies with low level of sustainability (Eccles et al. (2014), Artiach et al. (2010) and Hong et al. (2012)). Reasons for this are that sustainable firms seek to meet and balance the needs of current and future stakeholders (Artiach et al., 2010), they can secure loyal customers, gain a dedicated workforce and avoid different forms of costs Fatemi et al. (2015). All these positive effects are maybe also valued by investors.
Investors also value a sustainable reputation. According to Robinson et al. (2011) it is, however difficult for investors to determine whether a firm is sustainable or not. They show that one way to gain a sustainable reputation is an addition from a firm to a sustainable fund. According to Fowler and Hope (2007) there are a lot of sustainable funds. They show that the Dow Jones Sustainability Index is a logical index to use in a research because it is one of the oldest and biggest sustainable funds, has a global reach and is worldwide available.
Previous studies have shown different responses from investors to an addition of a firm to a sustainable fund. Robinson et al (2011) show that investors do value an addition to the Dow Jones Sustainability Index from North America firms. Nakai et al. (2012) found the same result on the Japanese market. Stellner et al. (2015) find more general evidence by looking into the level of sustainability of different countries. They find strong support for better ratings and lower z-spreads when a country is a superior sustainable performer. Arias Fogliano de Souza Cunha and Samanez (2012) did research on the emerging Brazilian market. They did not find significant results.
There is no research on this effect on the mature European market although it has been a far more sustainable market than any other market in the past decade (Bloomberg, 2016). This paper therefore investigates the value for an investor when a firm is added to the Dow Jones Sustainability Europe Index, in 2015. The result of this study show that investor do not necessarily value the additions of a firm to this fund but show a small negative response. However, this research has some shortcomings and therefore further research is recommended.
The structure of this paper is as follows. Chapter two describes the previous literature and combines it into the underlying theory. Chapter three describes the Dow Jones Sustainability Europe Index in detail. Chapter four describes the methodology and data, followed by the results in chapter five. Finally, the conclusion and discussion are set out in chapter six.
6
2. Literature
2.1 Sustainability
Sustainability has never been more important in our modern world as it i s today, for example for the first time all the countries in the world signed a climate accord (United Nations, 2015). Climate change and global warming are however not the only parts of sustainability. RobecoSAM is the organization behind the Dow Jones Sustainability Europe Index and according to their Corporate Sustainability Assessment Companion, Eco efficiency and reducing the environmental footprint is just a small part of sustainability. Robecosam identifies a total of ten other factors like tax strategies, human capital development and risk & crisis management which combined determine the level of sustainability.
According to Robinson et al. (2011) firms face higher expectations on corporate social responsibility (CSR). In their paper they refer to previous literature to show the positive relationship between CSR and financial performance (FP). For example, Ortilzky et al (2003) showed that CSR and FP are positively correlated. They did a meta-analysis using data of 30 years and concluded that studies that did not show a brief correlation were to narrow or ignoring important cross-study differences. Robinson et al. (2011) also show that the costs of CSR are offset by the reduction of other costs because of the CSR of the organization. Also, because of the good reputation of the company it is more likely that consumers are willing to pay a higher price. Furthermore they show that CSR decreases the variability of cash flows and protects them from shareholder losses.
Eccles et al. (2014) add an extra dimension to the literature by investigating not only the financial outperformance of a sustainable organization, but also looking into the organizational processes. They find that boards of corporations that voluntarily adopted a high level of sustainability in the period 1993 till 2009 are more likely to be formally responsible for sustainability. Also, the incentives for the board members depend on sustainable factors. Furthermore, these organizations are more long-term orientated and have established processes for stakeholder engagement. Eccles et al. (2014) show in their research that these processes lead to a financial outperformance of organizations that did not voluntary implied a high level of sustainability.
The theory that a high sustainable organization does outperform a low sustainable organization has different reasons. According to Artiach et al. (2010) corporate sustainability is a business and investment strategy that seeks to meet and balance the needs of current and future stakeholders. They point out that the previous literature on the relationship between sustainability and financial performance is not always consistent because of different research methodologies. In their paper Artiach et al. (2010) name the different perspectives on sustainability. One of them
7 argues that sustainability and financial performance are positively related, the other two clamed no association and a negative association.
There are three arguments for a positive relation between sustainability and financial performance. At first, the financial benefits of sustainability exceed its costs. Second, sustainable investments generate positive financial benefits by managing stakeholders. At last, sustainable organizations have superior resources over non sustainable organizations Artiach et al. (2010). In their research Artiach et al. (2010) do not focus on the existing evidence, but they provide insight into the incentives of sustainability for managers and different types of firms. Also, they provide a better understanding in the likely benefit of sustainability.
Artiach et al. (2010) found that the sustainable leaders are most likely the biggest firms in their industry, because they have the most stakeholders. Sustainable leaders are also better in realizing economies of scale regarding sustainability. Artiach et al. (2010) also found that the sustainable leaders are most likely they have a capacity to grow; they therefore have more opportunities to implant sustainability in their expanding operations. At last, Artiach et al. (2010) found that sustainable organizations have a positive relationship between CSP and profitability measured as a ratio on equity only, which is consistent with the stakeholder theory.
Fatemi et al. (2015) analyzed the effect of CSR engagement on the value of an organization. They combine the effect of the expenses of CSR on cost of capital and the probability of survival and the impact on the share prices. In their research they show different positive effects of CSR, for example, firms can secure loyal customers, gain a dedicated workforce and avoid different forms of costs. They show that only CSR companies gain benefit from these effects. Fatemi et al. (2015) also discuss some previous literature regarding the level of CSR activities and the performance of a company. For example, Erhemjamts et al. (2013) which report that firms with better performance are more likely to be sustainable. They show that organizations with a high level of R&D and organizations in new industries also tend to more sustainable since it is easier for these organizations to impend sustainability. They also show that the most sustainable firms are most likely the smallest or the biggest firms in the industry. The smallest firms tend to differentiate more, while the biggest firms have the most stakeholders witch both result in a higher level of sustainability. Albuquerque et al. (2012) found that CSR firms have lower systematic risk and higher expected returns.
In their paper, Fatemi et al. (2015) found that under certain circumstances social responsible organizations create value for the shareholders of a firm and its investors. However, they also point out that resources which are used for CSR are no longer available for other usage. They simulate this as well by using a simulation with decreasing marginal benefits and constant marginal costs for CSR expenses and found that the optimum level of CSR depends on the level of CSR of the competitors of an organization. Hong et al. (2012) also show that the level of CSR is higher when a firm performance
8 is higher and that the level of CSR only rises after the financial performance rises and not the other way around. This suggests that a sustainable organization must outperform a non sustainable organization otherwise it would not be a sustainable organization at all.
2.2 Sustainable funds
The different papers all suggest that sustainable organizations do outperform non-sustainable organizations. However, it is difficult for investors to know which organizations are sustainable an d which are not. According to Robinson et al. (2011) a reputation of CSR increases the value of a firm; the question is how firms gain such a reputation. In their paper, Robinson et al. (2011) refer to previous literature to explain how companies improve their reputation. Their main finding is that one way to gain such a reputation is an inclusion in a sustainable fund because of the fact that indexes serve as intermediaries for stakeholders of a company. It is therefore interesting to investigate the effects of an inclusion of a stock to a sustainable fund since this is a signal to investors that a company is sustainable.
Fowler and Hope (2007) claim that social screening of an investment has been around for over 100 years. Sustainable indices are however not that old, the first big sustainability index, the Domini 400 Social Index was launched in May 1990. The Dow Jones Sustainability Index was launched only a couple of months later, in September 1990 and is therefore one of the oldest sustainable indices. Furthermore, the index has a global reach and availability of the index for licensing. According to Fowler and Hope (2007) this makes the Dow Jones Sustainability Index a logical index to use in a research.
Literature shows that there are a couple of inclusion effects after a firm is added to a fund. Elliott et al. (2006) show that after an inclusion of a stock in the S&P 500 there are five explanations for a price change. The first effect is the price pressure effect, after an inclusion the fund sets a big order to add enough stocks the fund resulting in a short term price change due to liquidity constrains, this is also find by Chen et al. (2004). The second effect is the downward sloping demand curves which appear when an investor has perfect substitutions for a stock. If there are no perfect substitutions, an investor must be compensated to change his portfolio, therefore the downward sloping demand curve effect depends on the availability of perfect substitutions when a stock is added to a fund. If there is no perfect substitution the investor is paid the premium to compensate his portfolio change resulting in a higher price, this is also find by Blume and Edelen (2001). The third effect is the improved liquidity of a stock after it is included in a fund, due to the ownership of the stock by the fund the liquidity of the stock is improved resulting in lower asymmetric information and bid-ask spreads and therefore higher expected returns, this is also find by Hegde and McDermott (2003). The fourth effect is the improved operating performance, due to the inclusion a firm is better
9 monitored resulting in better performance and higher prices. The fifth effect is the increased investor awareness which means that after the inclusion in a fund investors are more aware of a stock. This results into lower costs due to incomplete information and therefore higher returns, this is also find by Chen et al. (2004).
In their paper Robinson et al. (2011) found evidence that there is an increase in the firm value after it is added to the Dow Jones Sustainability Index. They did not find a significant change in its value after it was removed from the same index. These results are the same in previous literature investigating the effect of an inclusion or removal of the S&P500. This suggests that the increased firm value is not because of the sustainability of the companies on the index but because of the extra awareness of investors. However, Robinson et al. (2011) found that the effect of an addition to the S&P500 are almost immediately after the announcement while the effect of an addition the DJSI are more likely to happen in the first few months after the announcement. Their results indicate that this higher value is due the better reputation of the added company. This suggests that the main effect of the inclusion in the Dow Jones Sustainability Europe Index is the increased investor awareness.
2.3 Comparable research and the added value to the existing literature
Nakai et al. (2012) did a research on the Japanese market using the Morningstar-SRI index. They investigated the effects of additions and deletions of firms and the effect on its share prices. They found a positive effect when a company was added and no significant effect when a company was deleted. Robinson et al. (2011) found evidence that there is an increase in the firm value of North-American firms after it is added to the DJSI. They also found evidence that this is not just because of the extra attention a firms gain after this addition but it is because of the sustainable reputation of the firm. Consolandi et al. (2008) found evidence that the CSR performance of a firm is a significant criterion for asset allocation activities. They used the Dow Jones Sustainability Stoxx Index using data over the period 2001-2006 and focused on European companies. Stellner et al. (2015) find strong support that companies from countries with a superior performance in CSR have better ratings and lower z-spreads. This implies that the sustainable leaders from the Dow Jones Sustainability Europe Index, where Europe is a superior performer, should have better ratings as well. Arias Fogliano de Souza Cunha and Samanez (2012) did a comparable research on the emerging Brazilian market. They did not find significant financial performance in the analyzed period. This implies that investors on a mature market value a sustainable investment more than investors in an emerging market.The previous literature on this topic does not include research on the inclusion effect of a company on a European sustainable fund although Europe has been more sustainable than the analyzed regions. During the last decade the European market has invested billions of dollars more in sustainability than any other region according to Bloomberg (2016). It is only since 2015 that China
10 does invest more in sustainability while America is reaching the same level as Europe. Furthermore, : America refused to sign the Kyoto-protocol in 1997 and placed their own financial interests above the sustainable interest of the world. Europe signed the protocol and actively participated in the composing of some strict principles..
On Monday, 14th September 2015, after the annual review was completed, 26 companies
where added to the DJSEI, effective from Monday, 21st September. This research investigates the
effect of these 26 inclusions on the share prices of the added companies to see whether investors value such inclusion on the mature European market. Theory suggests that investors do since sustainability leads to outperformance, Europe is a superior sustainable performer and the fund has a signaling function for investors.
11
3. The Dow Jones Sustainability Europe Index
According to Cavaco and Crifo (2014) investors value sustainable business models but they point out that investing in CSR is complex. There are complementary- and substitutable components of CSR and they both need a different approach. The Dow Jones Sustainability Europe Index does not distinguish between these different types of CSR and might therefore miss the synergy and the trade-off between CSR practices which might be an underlying reason for financial outperformance. However, there is no other fund that distinguish between these different components and I therefore choose to use the Dow Jones Sustainability Europe Index
RobecoSAM, the organization behind the Dow Jones Sustainability Index together with the S&P Dow Jones Indices, describes sustainability as a business approach that uses embracing opportunities and managing risk deriving from economic, environmental and social developments to create long-term shareholder value (RobecoSAM, 2015). RobecoSAM provides investors with indices which contain sustainable business practices and these, according to RobecoSAM, may lead to shareholder value at the long term.
The Dow Jones Sustainability World Index was launched in 1999 and was one of the first sustainable indices in the world (Fowler and Hope, 2007). According to the website of RobecoSAM the Dow Jones Sustainability World index was even the first global sustainability benchmark. The index tracks the performance of the world's biggest companies in economical performance, environmental and social criteria and serves as sustainability benchmarks for investors.
In compliance with the Dow Jones Sustainable World Index, RobecoSAM and S&P Dow Jones Indices launched the Dow Jones Sustainable Europe Index on August 04, 2010. This index differentiates from the existing indices by looking only into the sustainability of the leading firms in Europe. In short, the index focuses on the 600 largest developed European markets companies in the S&P Global BMI. The S&P Global BMI is an index that measures global stock performance and covers around 10,000 firms from almost 50 countries. The S&P Global BMI index is part of the S&P Dow Jones Indices just as the Dow Jones Sustainability Europe Index (S&P Dow Jones Indices, 2013). The Dow Jones Sustainability Europe Index is screened annually and contains 20% of the 600 biggest European firms. Which components are selected is based on a systematic assessment that identifies the leaders in 57 industry groups (RobecoSAM, 2015).
RobecoSAM and S&P Dow Jones Indices calculate the components of an index by using the so called ‘Best-in-Class Approach’. They calculate these best-in-class companies by doing a Corporate Sustainability Assessment (CSA). This starts with selecting the companies that are invited for the assessment, the so called invited universe (Robecosam, 2015). The selected companies ensure that every index is representative as a benchmark (RobecoSAM, 2015). The selected companies are asked to response to the assessment, however not every company does so. When a company is big enough
12 and chooses not to respond, RobecoSAM does the assessment by itself using publically available information (RobecoSAM, 2015). According to the invited universe of 2015, 3449 European companies were invited for the assessment. RobecoSAM and S&P Dow Jones Indices identify 57 different industries. The sustainable leaders of each industry become the components of the Dow Jones Sustainability Europe Index (RobecoSAM, 2015).
The assessment consist of a survey asking multiple questions regarding the following eleven general criteria sustainable subjects: Corporate governance, risk & crisis management, codes of conduct/compliance/corruption/Bribing, supply chain management, tax strategy, environmental & social reporting, operational eco efficiency, labor practice indicators & human rights, human capital management, talent attraction & retention and corporate citizenship & philanthropy (RobecoSAM, 2015). Besides this survey, a big component is the monitoring of media, stakeholders, public information of consumer organizations, NGO’s, governments and international organizations to identify involvement in sustainability. This resulted recently in the deletion of the Volkswagen group
from the Dow Jones Sustainability Europe Index on October 4the being effective on October 5th
(RobecoSAM, 2015).
RobecoSAM and the S&P Dow Jones Indices only select companies that score a minimum of 40% of the highest scores. After that, they select the best 20% per industry to compose the index. After this selection they implement an error margin of 30% which means that companies outside the best-in-class interval, within a 30% distance of the last company selected in the industry, are selected as well. As a result, since September 14th, the Dow Jones Sustainability Europe Index contains 161 components, in 51 industries from 13 countries (RobecoSAM, 2015). The index has a full value of 4,232.9 billion dollar. The annualized total return of the index for the past five years is 6,2% (RobecoSAM, 2015)
The Dow Jones Sustainability Europe Index is annually reassessed and furthermore the index is reviewed every quarter meaning that the shares outstanding are reviewed and updated. On
September 14th 2015 the changes as a result of the annual reassessment were announced, they are
effective on September 21st 2015. During the week in between, the announcement effects the
trading starting on the following money, September 17th. During the reassessment of 2015, 26
13
4. Methodology and data
4.1 Data
The data that is used for this paper are the 26 additions to the Dow Jones sustainability Europe Index from September 14th 2015. The firms which were added to this fund are listed in appendix 1. BHP Billiton is excluded from the data since it is a big outlier. BHP Billiton struggles with the decreasing prices of raw material resulting in an abnormal return of minus 32,7%. This exclusion leaves 25 firms. To run the regression which determines the normal return, the historical stock prices of the firms are used, these prices were provided by Compustat. The market model is represented by the by the S&P Europe 350 index (Junttila and Korhonen, 2011 and Dangenidou et al., 2011).
Unfortunately it is not possible to run a regression on the additions of previous years such as 2014 and 2013. RobecoSAM does not provide this information publicly but only has the most recent year available. RobecoSAM provides these sorts of information to students who fill in an academic request but they point out at their website that they only provide master students this sort of information. There it is also not possible to investigate the effects of deletions from the Dow Jones sustainability Europe Index. This information is also not publicly available.
4.2 Methodology
4.2.1 The CAAR
To investigate the effects of an addition of a company to the Dow Jones Sustainability Europe index and the value of this addition to investors I use an event study in accordance with the previous literature (Robinson et al. (2011) and Nakai et al. (2012)). This event study is based on abnormal return of a share after the event, which is in the general approach (MacKinlay, 1997). To calculate an abnormal return, there must be generated a normal return first, based on data from before the event, the estimation window. In this estimation window, there cannot be any overlap in events (MacKinlay, 1997) otherwise the estimation window would be biased. Since the event only happens
once a year it is possible to use a long estimation window, starting on October 5th 2014 until
September 13th 2015, containing 239 trading days. The timeline is shown in figure 1.
14 The market model is used to calculate the normal return by doing a linear regression on the daily return of the 25 companies which are added to the Dow Jones Sustainability Europe Index on the market return. The market model is as follows:
Where is the expected return of firm at time t, is the intersect of firm at time t, is the
sensitivity to the market return of firm , is the return of the market at time t and is the error
term of firm at time t. is expected to be zero.
After the normal return is calculated the abnormal return can be calculated as well. However,
in this case there is an announcement date and an effective date, respectively September 14th and
September 21st. The abnormal return will be indexed in event times. At first the event window, which
starts at the announcement date (AD) and ends one day prior the effective date (ED). By this event window the effects of the announcement can be measured. The second event time is the post-event window (Robinson et al. (2011)) which starts at the effective date and ends 60 trading days later on
Friday 11th, December 2015. The event window and the post-event window are also shown in figure
1. The abnormal return (AR) is calculated by subtracting the expected normal return from the realized return:
Where is the abnormal return of firm at time t and is the realized return of firm at time
t. The abnormal return will be calculated for the event window itself and the post-event window. Under the null hypothesis, the abnormal returns will be jointly normally distributed with a mean of zero and conditional variance as follows (MacKinley, 1997):
Where VAR is the variance of the abnormal return of firm at time t, is the variance of the
error term of firm , the average of the error term is expected to be equal to zero, L1 is the length of
the estimation window in days, is the mean of the market return and is the variance of the
market return. L1 is in this case large since it is possible to use a long estimation window, therefore
the second term disappears.
According to MacKinlay (1997) the abnormal return must be aggregated in two dimensions, time and across securities, to assure overall inferences for the event. Because of the fact that the addition of firm A is independent of the addition of firm B, the abnormal returns and the cumulative abnormal returns are independent across securities (MacKinlay, 1997). Therefore, it is possible to
15 calculate an average abnormal return on time t, across all the individual returns of the N firms, so the event can be analyzed as a whole. The average abnormal return (AAR) is calculated as follows:
Where AARt is the AAR at time t and N is 25 (all the 25 additions to the Dow Jones sustainability
Europe Index). The variance of the AARi is defined as follows when L1 is large:
Where is the variance of the AAR at time t.
According to MacKinlay (1997), after this it is possible to accumulate al the average abnormal returns for the two different timeframes, the event window and the post event window. This is done by using the same approach as used to calculate the CAR for each firm , using the following calculation:
Where is the cumulative average abnormal return for t1 until t2 where t1 = AD and t2=1 day
prior ED or t1=ED and t2=60 trading days after ED. The variance of the is calculated as follows:
After this the H0 can be tested by using the parametric n-test for every timeframe:
The outcome of this n-test determines whether the is indeed significant. According
to Stock and Watson (2012), if the outcome of t is absolute bigger than 1,96 the outcome is
significant at a level of 5%. To generate this outcome the is needed to estimate this
significance. H0 implies that this return is zero, however this is unknown. Therefore the estimated
variance of the market model regression is used to calculate the variance of the average
abnormal return (MacKinlay, 1997). The variance is the square of the standard error of the regression output.
16
4.2.2 The CAR
The results of the cumulative average abnormal return analysis could be influenced by a small number of firms lowering the average of all the firms. For example, Burberry has a negative
abnormal return of almost 10 percent on 15th October, 2015. To exclude the possibility that the
cumulative average abnormal return is lowered by a couple of firms, the firms are analyzed individually as well during the event window and the post-event window. This analysis uses the same approach as the CAAR as described in the methodology but does not aggregate across securities.
At first the abnormal return is calculated per day for each firm:
, With variance:
Second the cumulative abnormal return (CAR) is calculated for the event window and the post event window for each firm:
With variance:
The cumulative abnormal return per firm can be tested by using the parametric n-test for every timeframe:
Again, the estimated variance of the market model regression is used to calculate the
variance of the abnormal.
4.3 Hypothesis
As pointed out in the methodology, the cumulative average abnormal return is calculated for two separate time frames, the event window and the post-event window. This is because of the fact that RobecoSAM uses an announcement (AD) date and an effective date. Literature suggests that the announcement leads to a demand effect which leads to a higher stock price of a firm (Robinson et al., 2011). However, the literature on sustainability suggests that the reputation effects of the inclusion
17 in a sustainable fund are more likely to be seen on the longer run. Therefore this paper has separate hypotheses for these two time frames.
For the event window, starting on the announcement date until one trading date prior the effective date the hypotheses are as follows:
H0: Share prices of firms that were added to the Dow Jones Sustainability Europe Index experienced
no positive price change in the event window, is not significant.
H1: Share prices of firms that were added to the Dow Jones Sustainability Europe Index experienced a positive price change in the event window, is positive and significant.
For the post-event window starting on the effective date until 60 trading days after the effective date the hypothesis are as follows:
H0: Share prices of firms that were added to the Dow Jones Sustainability Europe Index experienced
no positive price change in the post-event window, is not significant.
H1: Share prices of firms that were added to the Dow Jones Sustainability Europe Index experienced positive price change in the post-event window, is positive and significant.
18
5. Results
5.1 The cumulative average abnormal return
The key factors of regression output for all the 25 firms that were added to the Dow Jones Sustainability Europe index is presented in appendix 2. The appendix shows the alpha and beta which combined form the market model. The results show that none of the alpha’s (the alpha correspondents with the intersect) is anyhow significant while all the beta’s are significant at the 1% level.
This result implies that none of the companies does outperform the market, in that case the
alpha would be significant. The firms tend to move ‘with’ the market at Beta * Rm,i, this finding is in
accordance with previous literature (Berk and DeMarzo, 2011). Also the estimated standard error of
the error term ( ) and the variance of the error term ( ) are shown because the variance of the
error term is needed to calculate the .
Although none of the alphas is significant they are part of the market model which means that the insignificant alpha is used to calculate the normal return. The reason for this is the underlying theory of a regression, the OLS estimator. The OLS estimator chooses the regression coefficient so that the regression line is as close as possible to the data and the sum of squares residuals is minimized (Stock and Watson, 2012). When the Alpha is left out, the results become less reliable because in that case, the regression model starts at a return of zero which increases the sum of squares residuals and therefore decreases the quality of the regression and its output.
The market model provides the possibility to calculate the abnormal return at time t for firm . The average abnormal return at time t is shown in appendix 3. This result provides the possibility to evaluate the result day by day. After the announcement the average abnormal return is 0,05%, followed by two days with an negative average abnormal return of over 0,6%. The rest of the event window shows again positive abnormal returns. This is shown in figure 2.
Figure 2, the daily abnormal return of the event window
-0.80% -0.60% -0.40% -0.20% 0.00% 0.20% 0.40% 0.60% 0.80% 14/09/2015 15/09/2015 16/09/2015 17/09/2015 18/09/2015
19 This results in a cumulative average abnormal return for the event window which is slightly negative and close to zero. This cumulative average abnormal return and the corresponding variance, are the input for the n-test. The results of the t-test for the event window are shown in table 1.
The first week of the post-event window shows a positive cumulative average abnormal return of 0,62% followed by four weeks of negative average abnormal return. This could be the demand effect as described in the theory. The daily abnormal return for the post-event window is shown is figure 3.
Figure 3, the daily abnormal return of the post-event window
The total cumulative average abnormal return for the post event window is -0,6998%. This negative return makes it impossible to find a significant result in the t-test since the return is the numerator in the formula for the t-test. The results of the n-test for the post-event window are also shown in table 1.
Figure 2 and 3 show the average abnormal return for the event window and the post-event window graphically. Both graphs show no clear jump in average abnormal return overall, although two of the three highest positive returns are in the first week of the post-event window. Figure 4 shows the cumulative average abnormal return. Again, the graph mainly shows the demand effect as described in the theory. It seems to be that in the end of the post-event window the cumulative abnormal return rises to a level of 0% which could be the reputational effect.
-1.50% -1.00% -0.50% 0.00% 0.50% 1.00% 1.50% 2.00% 2 1 /0 9 /2 0 1 5 2 4 /0 9 /2 0 1 5 2 9 /0 9 /2 0 1 5 0 2 /1 0 /2 0 1 5 0 7 /1 0 /2 0 1 5 1 2 /1 0 /2 0 1 5 1 5 /1 0 /2 0 1 5 2 0 /1 0 /2 0 1 5 2 3 /1 0 /2 0 1 5 2 8 /1 0 /2 0 1 5 0 2 /1 1 /2 0 1 5 0 5 /1 1 /2 0 1 5 1 0 /1 1 /2 0 1 5 1 3 /1 1 /2 0 1 5 1 8 /1 1 /2 0 1 5 2 3 /1 1 /2 0 1 5 2 6 /1 1 /2 0 1 5 0 1 /1 2 /2 0 1 5 0 4 /1 2 /2 0 1 5 0 9 /1 2 /2 0 1 5
20
Figure 4, the cumulative average abnormal return
To do the n-test the variance of the cumulative average abnormal is needed. This is calculated as follows, under the assumption that there is no correlation across securities:
where:
Table 4 shows the cumulative average abnormal return, the corresponding variance and standard error and the outcome of the n-test for the event window and the post-event window.
Timeframe CAAR VAR(CAAR) n-test
Event window -0.000377 2.96384E-05 0.0054 -0.06918
Post-event window -0.006998 0.00035566 0.0189 -0.37109
Table 1, n-test for the two timeframes
5.2 The cumulative abnormal return
The results of the n-test per firm individually are shown in appendix 5. The results of the n-tests for all the firms are in accordance with the results of the cumulative average abnormal return. During the event window there are three firms with significant results, one of them was positive while two were negative. All the other 22 firm did not show significant results. During the post-event window one firm showed negative significant results while all the other 24 firm did not. Therefore there is no evidence that the cumulative average abnormal return is influenced by a couple of firms with high negative abnormal return.
-5.00% -4.00% -3.00% -2.00% -1.00% 0.00% 1.00% 2.00% 1 4 /0 9 /2 0 1 5 1 7 /0 9 /2 0 1 5 2 2 /0 9 /2 0 1 5 2 5 /0 9 /2 0 1 5 3 0 /0 9 /2 0 1 5 0 5 /1 0 /2 0 1 5 0 8 /1 0 /2 0 1 5 1 3 /1 0 /2 0 1 5 1 6 /1 0 /2 0 1 5 2 1 /1 0 /2 0 1 5 2 6 /1 0 /2 0 1 5 2 9 /1 0 /2 0 1 5 0 3 /1 1 /2 0 1 5 0 6 /1 1 /2 0 1 5 1 1 /1 1 /2 0 1 5 1 6 /1 1 /2 0 1 5 1 9 /1 1 /2 0 1 5 2 4 /1 1 /2 0 1 5 2 7 /1 1 /2 0 1 5 0 2 /1 2 /2 0 1 5 0 7 /1 2 /2 0 1 5 1 0 /1 2 /2 0 1 5
21
6. Conclusion and discussion
6.1 Conclusion
This paper tested whether investors value an addition of a share to the Dow Jones Sustainability Europe Index. According to previous literature, investors in North-America and Japan do while investors in emerging markets do not. In this paper, 25 of the 26 additions in 2015 are tested by doing an event study. The results of the event study show that the abnormal return of the event window and the post-event window are both slightly negative. Giving the corresponding variance, this result in two insignificant n-tests.
The two insignificant n-tests imply that there is no evidence found to reject the H0
hypothesis. This means that there is no positive price change in the event window and in the post-event window. This paper tested the value for an investor when a firm was added to the Dow Jones
Sustainability Europe Index. Given the results of the H0 hypothesis, the conclusion is those investors
do not value the addition of a firm to the Dow Jones Sustainability Europe Index.
6.2 Discussion
The results are not in accordance with the previous literature which suggests that investors do value an addition to a sustainable index. This could be due to clustering in the event window. In this paper the additions of the 26 firms observed as individuals events since these events are independent from each other. This means that company A is added to the Dow Jones Sustainability Europe Index whatever might happen to company B. Thanks to this assumptions the variance of the average abnormal return can be calculated since the covariance across securities is zero (MacKinlay, 1997). However, since the firms are all added on the same date, one could argue that there is in fact a covariance across the securities.
The market model is based on the daily return of the European market. Since all the firms have exactly the same estimation window, event window and post-event window there might be a covariance across securities. If two companies have a similar beta, they tend to move with the market at the same level. Since all the timeframes are the same for every company, one could expect similar movement in the return of both companies resulting in a covariance. However, even if the variance was bigger, there is still a negative cumulative average abnormal return for the event window and the posevent window. This needs to be positive to have a positive and significant t-test.
This paper assumes that investors are not capable to determine the level of sustainabili ty of a firm and therefore need intermediaries such as the Dow Jones Sustainability Europe Index to signal the investors. However, firms often communicate about sustainability since it is not only mandatory
22 in some cases (Thomsen and Conyon, 2012) but it is also a marketing tool (Robbins and Coulter, 2014). Besides, RobecoSAM tend to do assessments by itself only based on publically available information when a company does not respond to the request to do the assessment. The information RobecoSAM uses to determine the level of sustainability is available for every investor. Therefore the reputation effect of the inclusion might not appear since the firms already have a sustainable reputation. The financial outperformance due to the level of sustainability of a firm are also know by investors before the inclusion in the Dow Jones Sustainability Europe Index since firms frequently communicate about the financial performance. Therefore the addition of a firm to the index might come after the financial positive effects of sustainability are already reflected in the share price of a firm. This could explain why investors do not value an addition to the Dow Jones Sustainability Index as expected since the signaling effect is not as big as expected in the literature.
It could be interesting in further research to use a different approach. MacKinley (1997) provides an approach which is commonly used when there is an event on the same day for a number of firms. This approach analyses the abnormal returns without aggregation but does it security by security data. According to MacKinley (1997) this is an application of a multivariate regression model with dummy variables for the event windows. This approach provides the opportunity to analyze each addition individually which might be interesting as well since there is a possibil ity to make a distinction between different industries. However, this approach has some strong drawbacks as well (MacKinley, 1997).
It could also be interesting to look into other sustainable funds that are not annually reviewed but are reviewed at a daily basis. When firms could be added at a daily basis, the clustering effect and the non-zero covariance are no longer applicable.
23
Reference list
Artiach, T., Lee, D., Nelson, D. and Walker, J. (2010), The determinants of corporate sustainability Performance, Accounting and finance, Vol. 50, pp. 31-51.
Arias Fogliano de Souza Cunha, F. and Patricio Samanez C. (2012), Performance Analysis of Sustainable Investments in the Brazilian Stock Market: A Study About the Corporate Sustainability Index (ISE), Journal of business ethics, Vol. 117, pp. 19-36.
Berk, J. and DeMarzo, P. (2011), Corporate Finance, Pearson Education Limited, pp. 330-376.
Blume, M.E. and Edelen, R.M. (2004), S&P 500 Indexers, Tracking Error and Liquidity: A complex answer to profiting, The Journal of portfolio management, spring 2004, pp. 37-46.
Castelo Branco, M. and Lima Rodrigues L. (2006), Corporate Social Responsibility and Resource-Based Perspectives, Journal of Business Ethics, Vol. 69, No. 2 (Dec., 2006), pp. 111-132.
Cavaco, S and Patricia C., CSR and financial performance: complementarity between environmental, social and business behaviours, Applied Economics,Vol 46:27, pp. 3323-3338.
Chen, H., Noronha, G. and Singal V. (2004), The Price Response to S&P 500 Index Additio ns and Deletions: Evidence of Asymmetry and a New Explanation, Journal of Finance, Vol. 59. No 4, pp. 1901-1929.
Craig MacKinley (1997), Event Studies in Economics and Finance, Journal of Economic Literatur,e Vol. XXXV (March 1997), pp. 13–39.
Consolandi, C., Jaiswal-Dale, A., Poggiani E. and Vercelli, A. (2009), Global Standards and Ethical Stock Indexes: The Case of the Dow Jones Sustainability Stoxx Index, Journal of business ethics, Vol . 87, pp. 185-197.
Dargenidou, C., McLeay, S. and Raonic, I. (2011), Accruals, Disclosure and the Pricing of Future Earnings in the European Market, Journal of Business Finance & Accounting, Vol. 38 (5) & (6), pp. 473–504.
24 Eccles, R.G., Loannou, L. and Serafeim, G. (2014), The Impact of Corporate Sustainability on OrganizationalProcesses and Performance, Management science, Vo.l 60(11), pp. 2835-2857.
Elliot, W.B., Van Ness, B.F., Walker, M.D. and Warr, R.S. (2006), What drives the S&P 500 inlcusion effect? An analytical survey, Financial Management, Vol. 35, pp. 31-48.
Fatemi, A., Fooladi, I. and Tehranian, H. (2015), Valuation effects of corporate social responsibility, Journal of Banking and Finance, Vo.l 59, pp. 182-192.
Fowler, S.J. and Hope, C. (2007), Critical Review of Sustainable Business Indices and their Impact, Journal of business ethics, Vol. 76, pp. 243-252.
Junttila, J. and Korhonen, M. (2011), Utilizing financial market information in forecasting real growth, inflation and real exchange rate, International Review of Economics and Finance, Vol. 20, pp. 281–301
Bloomberg new energy finance (2016), Clean energy defies fossil fuel price crash to attract record
$329BN global investment in 2015, 14th January 2016.
Nakai, M., Yamaguchi, K. and Takeuchi K. (2012), Sustainability membership and stock price: an empirical study using the Morningstar-SRI Index, Applied Financial Economics, Vol. 23:1, pp. 71-77.
Robinson, M., Kleffner, A. and Bertels, S. (2011), Signaling Sustainability Leadership: Empirical Evidence of the Value of DJSI Membership, Journal of business ethics, Vol . 101, pp. 493-505.
RobecoSAM (2015), Dow Jones Sustainability Indices Methodology, November 2015.
RobecoSAM (2015), Measuring intangibles ROBECOSAM’s Corporate Sustainability Assessment Methodology, September 2015.
RobecoSAM (2015), Volkswagen AG to be Removed from the Dow Jones Sustainability Indices, 29th
September 2015
25 S&P Dow Jones Indices (2015), S&P Global BMI Equity Indices, 29 March 2013.
Stellner, C., Klein, C. and Zwergel, B. (2015), Corporate social responsibility and Eurozone corporate bonds: The moderating role of country sustainability, Journal of Banking and Finance, Vol . 59, pp. 538-549.
Stock, J.H. and Watson, M.M. (2012), Introduction to Econometrics, Pearson Education Limited, pp. 83-84 and pp 149-185.
Thomson S. and Conyon M. (2012), Corporate governance mechanisms and systems, McGraw -hill, pp. 67-88.
Index in 2015
# Firm
# Firm
1 BNP Pariba SA 14 Delhaize Group PLC
2 DNB ASA 15 Heineken NV
3 Societe General 16 SCA AB
4 Bunzl PLC 17 BHP Billiton PLC (Excluded)
5 Koninklijke Philips 18 Sanofi
6 Sandvik AB 19 Wereldhave NV
7 Thales AB 20 Hennes & Mauritz AB
8 Vinci SA 21 ARM Holding PLC
9 ISS A/S 22 Atos
10 Randstad Holding NV 23 Deutsche Telekom AG
11 Burberry Group PLC 24 Aeroports de Paris
12 Accor SA 25 Engie
13 Henderson Group PLC 26 Red Electrica Corp SA
27
Appendix 2, market model regression output
Firm Alpha Beta Standard error Variance
BNP Pariba SA 0.000197 1.201064*** 0.0110 0.0001219 DNB ASA -0.000030 0.715412*** 0.0131 0.0001725 Societe General 0.000122 1.301496*** 0.0119 0.0001415 Bunzl PLC 0.000278 0.593503*** 0.0082 0.0000674 Koninklijke Philips -0.000440 0.938019*** 0.0098 0.0000955 Sandvik AB -0.000164 0.886511*** 0.0161 0.0002595 Thales AB 0.001514 0.778668*** 0.0114 0.0001301 Vinci SA 0.000924 0.990326*** 0.0108 0.0001168 ISS A/S 0.001562 0.538902*** 0.0130 0.0001699 Randstad Holding NV 0.001562 1.130121*** 0.0128 0.0001632 Burberry group PLC -0.000366 0.7903*** 0.0112 0.0001256 Accor SA 0.000667 1.102015*** 0.0108 0.0001164 Henderson group PLC 0.000908 1.095213*** 0.0125 0.0001551 Delhaize Group PLC 0.001710 0.995598*** 0.0177 0.0003118 Heineken NV 0.000615 0.657189*** 0.0128 0.0001632 SCA AB 0.001363 0.802329*** 0.0123 0.0001511 Sanofi -0.000050 1.109225*** 0.0118 0.0001392 Wereldhave NV -0.001038 0.581377*** 0.0140 0.0001966
Hennes & Mauritz AB 0.000270 0.716478*** 0.0094 0.0000875
ARM Holding PLC 0.000219 0.792553*** 0.0162 0.0002629
Atos 0.000561 0.908338*** 0.0117 0.0001359
Deutsche Telekom AG 0.001104 1.23058*** 0.0106 0.0001133
Aeroports de Paris 0.000092 0.681968*** 0.0098 0.0000953
Engie -0.001133 1.146454*** 0.0100 0.0001005
Red Electrica Corp SA 0.000243 0.763188*** 0.0106 0.0001119
Total: 0.0037048
Output regression market model ***significant at the level of 1%
28
Appendix 3, the average abnormal return per day
Date(t) AARt Date(t) AARt Date(t) AARt
14/09/2015 0.0005 14/10/2015 0.0014 13/11/2015 -0.0018 15/09/2015 -0.0065 15/10/2015 -0.0082 16/11/2015 -0.0042 16/09/2015 -0.0064 16/10/2015 -0.0060 17/11/2015 -0.0063 17/09/2015 0.0052 19/10/2015 -0.0003 18/11/2015 -0.0023 18/09/2015 0.0068 20/10/2015 0.0059 19/11/2015 -0.0007 21/09/2015 -0.0026 21/10/2015 0.0041 20/11/2015 -0.0003 22/09/2015 0.0100 22/10/2015 -0.0122 23/11/2015 -0.0006 23/09/2015 0.0010 23/10/2015 -0.0004 24/11/2015 -0.0012 24/09/2015 0.0047 26/10/2015 0.0006 25/11/2015 -0.0034 25/09/2015 -0.0069 27/10/2015 0.0046 26/11/2015 -0.0031 28/09/2015 0.0086 28/10/2015 -0.0011 27/11/2015 0.0044 29/09/2015 -0.0011 29/10/2015 0.0020 30/11/2015 0.0003 30/09/2015 -0.0055 30/10/2015 0.0016 01/12/2015 0.0043 01/10/2015 -0.0003 02/11/2015 -0.0030 02/12/2015 0.0006 02/10/2015 -0.0061 03/11/2015 -0.0071 03/12/2015 0.1043 05/10/2015 -0.0041 04/11/2015 0.0018 04/12/2015 0.0000 06/10/2015 -0.0044 05/11/2015 0.0085 07/12/2015 0.0031 07/10/2015 -0.0067 06/11/2015 0.0005 08/12/2015 0.0051 08/10/2015 -0.0012 09/11/2015 0.0021 09/12/2015 -0.0033 09/10/2015 -0.0022 10/11/2015 -0.0066 10/12/2015 -0.0002 12/10/2015 -0.0036 11/11/2015 0.0007 11/12/2015 0.0067 13/10/2015 0.0050 12/11/2015 0.0084 AAR (t)
CAR per firm during event window
Firm Return event window Standard error Variance error term Variance event window n-test
BNP Pariba SA -0.0400 0.0110 0.0001219 0.0006 0.0246842 -1.6193 DNB ASA -0.0115 0.0131 0.0001725 0.0009 0.0293723 -0.3912 Societe General -0.0446 0.0119 0.0001415 0.0007 0.0266013 -1.6759 ** Bunzl PLC -0.0089 0.0082 0.0000674 0.0003 0.0183574 -0.4842 Koninklijke Philips -0.0425 0.0098 0.0000955 0.0005 0.0218544 -1.9441 ** Sandvik AB -0.0120 0.0161 0.0002595 0.0013 0.0360197 -0.3320 Thales AB 0.0010 0.0114 0.0001301 0.0007 0.0255096 0.0394 Vinci SA -0.0002 0.0108 0.0001168 0.0006 0.0241697 -0.0067 ISS A/S 0.0047 0.0130 0.0001699 0.0008 0.0291482 0.1623 Randstad Holding NV -0.0182 0.0128 0.0001632 0.0008 0.0285679 -0.6385 Burberry group PLC 0.0052 0.0112 0.0001256 0.0006 0.0250633 0.2070 Accor SA 0.0157 0.0108 0.0001164 0.0006 0.0241266 0.6500 Henderson group PLC 0.0423 0.0125 0.0001551 0.0008 0.0278511 1.5186 Delhaize Group PLC -0.0181 0.0177 0.0003118 0.0016 0.0394814 -0.4582 Heineken NV 0.0739 0.0128 0.0001632 0.0008 0.0285700 2.5867 *** SCA AB -0.0231 0.0123 0.0001511 0.0008 0.0274844 -0.8416 Sanofi 0.0004 0.0118 0.0001392 0.0007 0.0263797 0.0152 Wereldhave NV 0.0129 0.0140 0.0001966 0.0010 0.0313522 0.4127
Hennes & Mauritz AB -0.0131 0.0094 0.0000875 0.0004 0.0209158 -0.6283
ARM Holding PLC 0.0291 0.0162 0.0002629 0.0013 0.0362584 0.8035
Atos 0.0294 0.0117 0.0001359 0.0007 0.0260706 1.1274
Duetsche Telekom AG 0.0077 0.0106 0.0001133 0.0006 0.0237970 0.3216
Aeroports de Paris -0.0205 0.0098 0.0000953 0.0005 0.0218233 -0.9394
Engie -0.0119 0.0100 0.0001005 0.0005 0.0224168 -0.5321
Red Electrica Corp SA 0.0328 0.0106 0.0001119 0.0006 0.0236542 1.3882
30
CAR per firm during post-event window
Firm Return post-event Standard error Variance error term Variance post-event window n-test
BNP Pariba SA -0.0419 0.0110 0.0001219 0.0073 0.0855087 -0.4897 DNB ASA -0.0741 0.0131 0.0001725 0.0104 0.1017485 -0.7285 Societe General 0.0033 0.0119 0.0001415 0.0085 0.0921498 0.0355 Bunzl PLC 0.0341 0.0082 0.0000674 0.0040 0.0635919 0.5357 Koninklijke Philips 0.1164 0.0098 0.0000955 0.0057 0.0757057 1.5379 Sandvik AB 0.0445 0.0161 0.0002595 0.0156 0.1247761 0.3570 Thales AB 0.0066 0.0114 0.0001301 0.0078 0.0883679 0.0745 Vinci SA -0.0625 0.0108 0.0001168 0.0070 0.0837263 -0.7461 ISS A/S -0.0464 0.0130 0.0001699 0.0102 0.1009722 -0.4598 Randstad Holding NV -0.0499 0.0128 0.0001632 0.0098 0.0989620 -0.5038 Burberry group PLC -0.1574 0.0112 0.0001256 0.0075 0.0868218 -1.8126 ** Accor SA -0.1287 0.0108 0.0001164 0.0070 0.0835770 -1.5398 Henderson group PLC -0.0017 0.0125 0.0001551 0.0093 0.0964792 -0.0175 Delhaize Group PLC 0.0078 0.0177 0.0003118 0.0187 0.1367675 0.0572 Heineken NV 0.0406 0.0128 0.0001632 0.0098 0.0989694 0.4100 SCA AB -0.0818 0.0123 0.0001511 0.0091 0.0952087 -0.8596 Sanofi -0.1294 0.0118 0.0001392 0.0084 0.0913820 -1.4160 Wereldhave NV 0.0783 0.0140 0.0001966 0.0118 0.1086072 0.7207
Hennes & Mauritz AB -0.0624 0.0094 0.0000875 0.0052 0.0724544 -0.8616
ARM Holding PLC 0.1047 0.0162 0.0002629 0.0158 0.1256029 0.8337
Atos 0.0751 0.0117 0.0001359 0.0082 0.0903110 0.8312
Duetsche Telekom AG -0.0512 0.0106 0.0001133 0.0068 0.0824352 -0.6213
Aeroports de Paris 0.0426 0.0098 0.0000953 0.0057 0.0755980 0.5641
Engie 0.1268 0.0100 0.0001005 0.0060 0.0776540 1.6323
Red Electrica Corp SA 0.0317 0.0106 0.0001119 0.0067 0.0819404 0.3869
