The effect of inclusion in the Dow Jones Sustainability Index Asia
Pacific on stock value: an event study
Martin Thomas 10846670
Supervisor: dr. L. (Liang) Zou
BSc Economics and Business Track: Finance & Organization
Amsterdam, Netherlands 31 January 2018
2 Statement of Originality
This document is written by Martin Thomas 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 Abstract
As environmental awareness among consumers and investors is rising, this paper reviews the effect of corporate social responsibility (CSR) on corporate financial performance (CFP) and adds to limited existing research in the Asia Pacific. The Dow Jones Sustainability Index Asia Pacific (DJSIAP) is used as a measure for CSR to test, by means of an event study for the period 2010-2017, whether inclusion in the index has a positive impact on stock value upon announcement. The results do not support the hypothesis as the cumulative average abnormal returns (CAARs) on announcement day are insignificant. However, the findings reveal significant CAARs for the 10 trading days prior to the announcement, which could indicate an anticipation by investors. Furthermore, abnormal returns are found to be
predominantly positive until two days after the announcement when the trend changes. While insignificant, the development may indicate that the price impact is temporary.
4 Table of Contents 1. Introduction ... 5 1.1 Definitions ... 6 2. Literature review ... 7 2.1 CSP-CFP relationship ... 7
2.2 Effect of index inclusion ... 8
3. Research question and Hypothesis ... 11
4. Methodology ... 11
4.1 DJSI Asia Pacific ... 11
4.2 Event study methodology ... 13
4.3 Estimation and event periods ... 15
4.4 Data ... 16
5. Results ... 16
6. Conclusion ... 21
5 1. Introduction
Recent years have shown that consumers are increasingly concerned about the
environment and may look for products that are more sustainable than others (Nielsen, 2015). Companies who would like to appeal to such consumers and meet the growing demand are likely to pursue a strategy that improves their Corporate Social Responsibility (CSR).
Existing research has explored whether an increase in corporate social activities increases the value of the stock affected. Indices that focus on sustainability assess CSR and include the companies with the highest scores. Assuming such valuations are credible these inclusions and deletions from a stock index can signal to the market how strong a company scores on relevant criteria. Investors who value this characteristic would react by buying or selling the stock and thus affect its value. What holds for consumers also matters to investors: Socially responsible investments (SRI) have increased significantly in the period 2013-2015 as shown in Europe (Eurosif, 2015) and the United States (Ussif, 2014). Both consumers and investors are also not held back by potentially higher costs or lower financial performance when purchasing the environmentally friendly option (Riedl & Smeets, 2017).
This thesis follows the results of the Nielsen Global Survey on Corporate Social Responsibility (2014), which suggests that consumers from the Asia Pacific region are more attracted to sustainable products than the global average, and adds an analysis of seven more years to the event study by Cheung & Roca (2012), who have found significant abnormal returns after the announcement of stock inclusion. Given the recent developments in consumer and investor interests, this research focuses on whether an addition to the Dow Jones Sustainability Index Asia Pacific (DJSIAP) leads to a positive price impact on stock value. While the effect of Corporate Social Performance (CSP), which is the preferred term for CSR used in many papers, on Corporate Financial Performance (CFP), as measured by stock returns for example, is found to be positive in most of past research, analyzing the relationship using a sustainability index as a measure for CSP has led to inconsistent results.
The first section consists of a literature review on both the CSP-CFP relationship and the effect of index inclusion, followed by research question and hypothesis. Section 2
specifies the methodology of the chosen sustainability index DJSI Asia Pacific, the event study and data used. Results are presented in section 3 after which the conclusion follows. The introduction finishes with a brief definition of CSR, corporate sustainability, and corporate social performance.
6 1.1 Definitions
This section aims to give an overview of terms that are used in existing literature to point out distinctions and similarities.
Corporate social responsibility (CSR) is a term that can be broadly defined as follows: ‘The social responsibility of business encompasses the economic, legal, ethical, and
discretionary [later referred to as philanthropic] expectations that society has of organizations at a given point in time’ (Carroll, 1979, p. 500).
Although very similar to the definition of CSR Wilson (2003) describes how
corporate sustainability (CS) demands initiative in that it ‘‘recognizes that corporate growth and profitability are important, [but] it also requires the corporation to pursue societal goals, specifically those relating to sustainable development— environmental protection, social justice and equity, and economic development’’ (Wilson, 2003), which is in line with the definition adopted by RobecoSAM who manages the Dow Jones Sustainability Indices, of which the Asia Pacific benchmark is analyzed in this thesis: “Corporate Sustainability is a business approach that creates long-term shareholder value by embracing opportunities and managing risks deriving from economic, environmental and social developments” (Robeco Sustainable Asset Management).
Van Marrewijk (2003) notices how both terms are often considered as synonyms nowadays, while CSR used to refer to social aspects such as human rights only and CS to the environment. He, therefore, suggests a distinction that describes CSR as related to the
transparent dialogue with stakeholders and within organizations, while CS encompasses the agency principle that includes value creation, and management of human capital and
environment (Van Marrewijk, 2003).
Lastly, a related term used in existing literature includes corporate social performance (CSP), which “refers to the principles, practices, and outcomes of businesses’ relationships with people, organizations, institutions, communities, societies, and the earth, in terms of the deliberate actions of businesses toward these stakeholders as well as the unintended
externalities of business activity” (Wood, 1991) and therewith offers the perhaps most neutral description that could also suit a socially passive company that merely bears the risks
associated with conducting business. The emphasis is also on outcomes though, which result from CSR activities (SAGE Publications, 2012). Despite the similarity between the terms defined above the thesis aims to adopt the respective term used in the discussed literature while definitions may be understood interchangeably.
7 2. Literature review
2.1 CSP-CFP relationship
The discussion about an effect of CSP (corporate social performance) on CFP (corporate financial performance) has existed since the 1960s (Cochran & Wood, 1984). In times in which the shareholder theory (Friedman, 1962), which supports the idea that managers should act in the interest of shareholders only and thus focus on maximizing firm value, was confronted by the stakeholder theory (Freeman, 1984), according to which
stakeholders consist of shareholders but also external environmental parties, which managers have to liaise with and satisfy, the investigation of a relationship between CSP and CFP became of interest.
Across the number of academic papers that analyzed this relationship since then, different measures for both CSP and CFP have been used, while the correlation for both is found to be positive in the majority of papers (Cochran and Wood, 1984; Waddock and Graves, 1997; Ruf et al., 2001; Orlitzky et al., 2003; Garcia et al. 2007; Beurden & Gössling, 2008). Measures for financial performance can be divided into accounting returns, such as EPS or P/E ratio, and investor returns like price per share (Cochran & Wood, 1984).
Moreover, according to Orlitzky et al. (2003), who conducted a meta-analysis of 52 studies, accounting measures, among which return on equity (ROE), return on assets (ROA) and return on sales (ROS) are common, have a higher correlation with CSP than investor returns. A different preference is expressed by Davidson and Worrell (1990) who suggest that
investor returns are the appropriate measure when conducting large cross-sectional analyses as accounting-based measures can vary with time among different industries and countries.
The social performance measures are broader and can be categorized into four types of measurement according to Orlitzky et al. (2003): CSP disclosures, CSP reputation ratings, social audits and managerial CSP principles. Thus, choice varies across research but often includes ratings by KLD (Graves & Waddock, 1997; Ruf et al., 2001) or a sustainability index such as the DJSI (Lopez et al., 2007). While disclosures are internally generated, ratings and audits are made by third-parties and review a company's social performance, a topic that is also relevant for index inclusions dealt with in the next section of this thesis. Orlitzky et al. (2003) claim that the correlation with CFP is highest with such reputation ratings.
Lopez et al. (2007) are among the first to study the effect within European rather than U.S. firms. They found that the relationship between corporate social responsibility (CSR),
8 which describes the actions resulting in CSP, and CFP are negative for the first years after applying the CSR strategy. This, however, is only relevant in the short term as firms have not put aside a budget for responsible investments such that their expenses adversely impact their financial performance, which then self-corrects in the long run (Lopez et al., 2007).
A finding that is in line with the results by Ruf et al. (2001) who observe the statistically strongest positive effect of a change in CSP on ROE and ROS in year 3.
While there is no consensus on how to measure the CSP-CFP relationship, the overall effect of an increase in CSR strategies on CFP is predominantly found to be positive, as shown in the analysis of past papers by Beurden & Gössling (2008) for mainly European countries and the US.
Past research, with a focus on investor returns as a measure of CFP, does not support a definite view on whether SRI funds generally outperform or match returns of conventional funds as other existing literature also suggests that investing socially responsible leads to a financial disadvantage (Riedl & Smeets, 2017). A reason for this may be, that sustainability can be another criterion that distracts from portfolio optimization, that focuses on criteria such as risk to maximize financial value not taking into account whether the investor gains utility from satisfying non-financial social preferences of intrinsic nature (Cheung & Roca, 2012). While Cheung & Roca (2012) describe Asian Pacific investors as perceived to be more “moralistic than those in Western countries” (Cheung & Roca, 2012, p. 54) the countries credited are not represented in both their sample and the one used in this thesis. Thus, further research on the mentality towards SRI in the Asia Pacific may be of interest for future studies. Renneboog et al. (2008) have a consistent finding with the hypothesis that investors in the Asia Pacific behave differently towards SRI than those from Europe and the US as SRI funds in countries within the Asia Pacific underperform their domestic
benchmarks by annually 5%. While Asia Pacific has not caught up on SRI as much as other countries, the amount of projects that can prove otherwise is increasing and Japan, who is among the leading countries in this movement (Cheung & Roca, 2012), also dominates the constituents list of the Dow Jones Sustainability Index Asia Pacific analyzed in this thesis.
2.2 Effect of index inclusion
One way to analyze the effect of CSP on CFP is by means of an event study on financial stock returns. Selected stocks constitute a stock index, which exists for different countries or markets but can also focus on sustainability. Such ethical stock indices select companies that fulfill certain requirements related to CSP. This can be determined by surveys
9 that are filled out by invited companies. An inclusion into an ethical stock index can thus signal to the market that the added company performs high in corporate social activities. Based on research on the CSP-CFP relationship these measures, namely stock index and stock returns, correspond to CSP reputation ratings and investor returns respectively. Thus, the event of inclusion is expected to have a positive effect on the stock price concerned. Research on the index inclusion effect is reviewed in this section, which analyses initial assumptions, methodologies and subsequently results.
Various theories that are identified in existing literature explain the effect of index additions and deletions on stock prices, among which at least two hypotheses claim that any change to an index does not affect share prices, while three other hypotheses suggest a price impact. Both the downward sloping demand curve hypothesis (Shleifer, 1986) and the price pressure hypothesis (Harris & Eitan, 1986) suggest that these changes do not contain information. While Shleifer claims that a significant impact on price is merely due to an increase in demand that is attributable to portfolio allocation, which is not based on
information, and both demand and price impacts are permanent, Harris & Eitan support the theory that these impacts are temporary. In contrast, the information cost hypothesis (Merton, 1987) describes that index composition changes can provide the investor with information and reduce the asymmetry problem and thus searching cost. This is valuable to investors and explains why a positive price impact is expected after an addition to the index. Similarly, the signaling hypothesis (Denis, McConnell, Ovtchinnikov, & Yu, 2003; Dhillon & Johnson, 1991; Jain, 1987) supports the theory that these events reveal firm relevant information that act as a signal to investors and thus affect their expectation of the firm's future value. If the event increases expected firm value, the stock price will rise.
A further theory on why these events impact prices significantly is described by the liquidity hypothesis (see Beneish et al., 2002; Hegde & McDermott, 2003), that inclusion causes trading volumes to increase which reduces the bid-ask spread among traders and increases liquidity. A more liquid market reduces stock volatility and discount rates, positively impacting prices.
An event study can be an appropriate measure to determine which theory applies if certain assumptions are given as Oberndorfer et al. (2003) point out. According to their paper, it is critical to assume that markets are sufficiently efficient to respond to these events and that the timing of events cannot be influenced by firms. Such exogeneity is given as stock indices reevaluate their constituents regularly once a year. Also, it should be noted that the aforementioned theories are not specific to ethical indices but have been formed with a view
10 on conventional funds. Thus, results related to sustainability indices may reveal different outcomes to the ones expected. Past research has investigated the effect of inclusion in a sustainability index on abnormal stock returns of added companies. The focus is on different indices such as the Dow Jones Sustainability World Index (DJSI World), their regional benchmarks such as the DJSI Europe, the Dow Jones Sustainability Stoxx Index (DJSSI), with a focus on European firms, the Domini 400 Social Index and the FTSE4GOOD. In most research, an event study on the DJSI is performed (Chakarova & Karlsson, 2008; Ziegler, 2011; Oberndorfer at al., 2013; Chatterji et al., 2014; Chen & Wang, 2015; Van Stekelenburg et al., 2015; Chatterji et al., 2017).
However, results vary as there is no distinct outcome on both the sign and
significance of returns. Consolandi et al. (2008) find significant positive cumulative average abnormal returns (CAARs) in their event study on inclusion to the DJSSI, starting several days before announcement and diminishing shortly after the implementation of changes, while a deletion has a permanent negative effect which can be interpreted as in accordance with the information cost hypothesis. Also, Oberndorfer et al. (2013) study the DJSSI, though only find insignificant results, in line with Becchetti et al. (2012) for inclusion to the Domini 400 Social Index, while in the case of the DJSI World strongly negative returns are found for inclusion. Other literature on the latter index also report significant negative returns for the period 1994-2004 but insignificant thereafter (Chatterji et al., 2017) and significant positive CAARs (Chen & Wang, 2015) on the other hand. Ziegler (2011) also finds positive price impacts for continental European companies, though insignificant for firms from Anglo-Saxon European countries. Chakarova & Karlsson (2008) confirm that significance differs between countries but CAARs are overall insignificant. With a focus on the DJSI Europe, Van Stekelenburg et al. (2015) find contrary results to Consolandi et al. (2008) as the
announcement causes insignificant price impacts and temporary significance only around the day of implementation, which supports the price pressure hypothesis. Finally, the only study that focuses on the Asian Pacific market reports significant negative CAARs for both
inclusion and deletion over the period 2002-2010.
Overall, according to Chatterji et al. (2017) who conducted a large-scale longitudinal event study, these inconsistent results are due to several reasons which include a varying focus on geography and time periods as well as not accounting for the effect on similar firms that are not included in the index. They also stress the importance of taking firm
11 been raised by Ziegler (2010). The overall net CAAR is found to be -0.17% on inclusion after adjusting the result using control variables (Chatterji et al., 2017).
3. Research question and hypothesis
This thesis adopts the theories that index events carry information and thus affect price. Given that sustainability awareness has been on the rise in recent years with an indication towards a stronger growth in the Asian Pacific region (Nielsen, 2014) this thesis aims to add to research that is lacking in this geographical region, particularly for the time period 2010-2017. Following the popular choice of the DJSI, which introduced the DJSI Asia Pacific in 2009, and taking into account previous sections of this thesis, lead to the
investigation of the following research question and hypothesis for the remaining sections: Does the inclusion of a stock in the DJSI Asia Pacific between 2009 and 2017 affect stock value?
The underlying hypothesis is that the announcement of the inclusion of a stock in the DJSI Asia Pacific has a positive effect on the stock value.
4. Methodology
The first part of this section explains the choice and methodology of the DJSI Asia Pacific, while the second part elaborates on the event study traits.
4.1 DJSI Asia Pacific
The Dow Jones Sustainability Indices appear a reasonable choice among similar indices, given their global reach and impact of the Dow Jones brand (Fowler & Hope, 2007). Thus, this study chooses the DJSI Asia Pacific, which is one out of seven regional
benchmarks that lists the companies according to domicile and focuses on the region in question. To be selected as a constituent of the index, first, S&P Dow Jones and
RobecoSAM, who manage the indices, conduct an annual Corporate Sustainability
Assessment (CSA) with invited companies (“the invited universe”). The 600 largest Asian Pacific companies (based on market capitalization) within the S&P Global Broad Market Index with 11303 constituents are invited to complete a comprehensive questionnaire that determines a company's total sustainability score (TSS). Finally, 20% of the companies with the highest TSS are added to the index.
A constituents list was made available by RobecoSAM for research purposes. The spreadsheet contains the names of companies included in the DJSI Asia Pacific and the
12 respective effective date for every year since its launch in 2009. Additions have been
identified by highlighting the differences in composition for every year, which is composed of roughly 150 constituents in any given year, and make up the sample of 172 companies for 2010-2017 (Table 1). Table 2 shows that Japanese companies form half of the sample and the largest three stock domiciles are Australia, Japan, and South Korea who account for 89% altogether. Furthermore, announcement dates and effective dates presented in Table 3 have been sourced from the official press release archive.
Table 1: Number of firms included and excluded by DJSI Asia Pacific
Year in
September Additions Deletions Total
Total number of constituents 2010 36 25 61 143 2011 27 10 37 156 2012 13 15 28 154 2013 26 26 52 152 2014 14 16 30 148 2015 18 17 35 145 2016 18 15 33 146 2017 20 14 34 152 Total 172 138 310
Source: Additions: own calculations, Deletions and total number of constituents: RobecoSAM
13
Table 2: Added companies over period 2010-2017 by country
Country Number Percent
Australia 22 12.79 Hong Kong 10 5.81 Japan 87 50.58 New Zealand 3 1.74 Singapore 6 3.49 South Korea 44 25.58 Total 172 100
Source: own calculations
4.2 Event study methodology
To determine whether stock inclusion in the index causes a significant positive price impact on announcement day a general event study approach is adopted. In line with past research (Consolandi et al., 2008; Cheung & Roca, 2013; Van Stekelenburg et al., 2015) two event days for each year are determined, namely the day of announcement (AD) and the therein mentioned effective day of composition change (ED). De Jong (2007) suggests the necessary steps to determine the cumulative average abnormal return (CAAR), which represents the average difference, around all event dates, between actual stock return and expected stock return based on a market benchmark:
First, as events are expected to cause an increase or decrease in returns, the actual stock returns are compared to expected returns, which is the return expected if the event had not happened, to determine the abnormal returns for each day within the event window:
𝐴𝑅#,% = 𝑅#,%− 𝐸(𝑅#,%)
To estimate the expected returns for this event window, also an estimation window is defined, which is necessary to determine the intercept (𝛼) and slope (𝛽) for each stock individually. Next to a CAPM type model, the market model is the adequate approach for short-horizon events. Alternatively, a multi-factor model by Fama and French can be used. However, necessary variables are not available for the region analyzed and the model is more beneficial to long-horizon studies of several years (De Jong, 2007). Additionally, gains with a
14 multifactor model over the market model are limited for event studies with firms across different industries according to MacKinlay (1997).
𝐸(𝑅#,%) = 𝛼#+ 𝛽#𝑅.% + 𝜖#,%
The variables 𝛼 and 𝛽 are determined through an OLS regression, while 𝑅.% represents the total return of the market benchmark, chosen as the S&P Asia Pacific Broad Market Index (BMI), launched in 1997, given its regional focus and its large number of constituents (3043) which includes all companies from the sample used. Popular choices for a broad based market index include the S&P 500 Index and the CRSP Value Weighted Index (MacKinlay, 1997), of which both focus on the US market though.
Secondly, the cumulative abnormal return (CAR) is calculated by summing all abnormal returns in the event window to return a value that is relevant to the specific company and event window:
𝐶𝐴𝑅(%2,%3) = 𝐴𝑅#,% %3
%4%2
Thirdly, the cross-sectional average abnormal return is calculated per event window by averaging the CARs. This also contributes to a more informative value compared to what results from an analysis of each individual firm as stock returns can be affected by event-unrelated movements that are specific to each firm and deteriorate the effect of the event (De Jong, 2007). 𝐶𝐴𝐴𝑅(%2,%3) = 1 𝑁 𝐶𝐴𝑅(%2,%3) 7 #48
Finally, the CAAR for the specific event window is tested for significance using a standard deviation
𝑠 = 1
𝑁 − 1 (𝐶𝐴𝑅(%2,%3)− 𝐶𝐴𝐴𝑅)² 7
#48
and a t-test value of
15 The t-statistic assumes a normal distribution across sections, the row of abnormal returns for a one-time period, which can be a poor approximation for daily stock returns (De Jong, 2007). A non-parametric test, the sign test by Corrado & Zivney (1992) is thus used alongside.
4.3 Estimation and event windows
Figure 1: Estimation and event windows
This study adopts an estimation window of 150 trading days, which is also used by Van Stekelenburg et al. (2015). It typically ends several trading days, in this case, 50, before the event window (De Jong, 2007). To capture any abnormal return in anticipation of the announcement day (AD) the total event window starts 10 trading days before AD and also takes into account movements after the effective change day (ED) until 5 trading days after. Thus, variables in figure 1:
𝑡 = 𝐴𝐷; 𝑡8 = 𝑡 − 10; 𝑡D = 𝐸𝐷 + 5; 𝑇8 = 𝑡8− 200; 𝑇D = 𝑡8− 50.
In line with Cheung (2011) and Consolandi et al. (2008) the total event window from AD-10 to ED +5 is split into several smaller windows to account for the different aspects that may have an impact and are specified as follows:
● The Pre-announcement window captures any abnormal returns that may have been caused by information reaching investors early and lies between AD-10 and AD-1. ● A Post-announcement window that lies between AD+1 and ED-1.
● A Post-effective window from ED+1 to ED+5 to cover the temporary price impact
The chosen event windows, shown in Table 3, include announcement and effective days for every year 2010-2017 as sourced from the fund manager RobecoSAM.
16
Table 3: Event windows of the study
Year Pre-Announcement Announcement Post-announcement Effective Post-effective
2010 26/08-08/09 09/09 10/9-17/9 20/09 21/09-27/09 2011 25/08-07/09 08/09 09/9-16/9 19/09 20/09-26/09 2012 30/08-12/09 13/09 14/9-21/9 24/09 25/09-01/10 2013 29/08-11/09 12/09 13/9-20/9 23/09 24/09-30/09 2014 28/08-10/09 11/09 12/9-19/9 22/09 23/09-29/09 2015 27/08-09/09 10/09 11/9-18/9 21/09 22/09-28/09 2016 25/08-07/09 08/09 09/9-16/9 19/09 20/09-26/09 2017 24/08-06/09 07/09 08/9-15/9 18/09 19/09-25/09 Source: RobecoSAM 4.4 Data
Necessary data for the event study include total stock returns for both the sample companies and the S&P Asia Pacific BMI, which were sourced from Datastream for the beginning of estimation window to the end of the total event window. To do so the identified company names that make up the sample have been matched with the relevant stocks, respective to time period and country, using Datastream. Data on the DJSI Asia Pacific is specified in section 4.1.
5. Results
The thesis aims to present the effect of a stock addition to the Dow Jones Sustainability Index Asia Pacific (DJSIAP) on the stock value of the companies added. By means of an event study with yearly announcement dates, on which the updated composition is revealed, the affected stocks are analyzed for abnormal returns in relation to its regional broad market index (BMI) by S&P. First, an overview of the performance of both indices is given after which the cumulative average abnormal returns (CAARs) are presented.
Results in table 4 show comparable average daily returns for both the DJSIP and BMI Asia Pacific, though the former tends to underperform slightly in comparison to its broad market index returns (a simple OLS regression reveals an insignificant alpha, p-value = 0.95) and risk, as measured by the standard deviation, which is overall similar however. It should be noted that the BMI includes all stocks of the DJSIAP. With around 3000 and 150
constituents respectively the close correlation may not seem to be explained by this fact at first, though as described in the methodology section, only the 600 biggest companies are
17 invited to the assessment which explains the similarity in returns. Nevertheless, the
underperformance of 0.005% per day on average, which translates to 1.84% annually, may be explained by Renneboog et al. (2008) who report that socially responsible funds in the Asia Pacific strongly underperform their benchmarks by 5% per year as additions to an index are expected to cause negative price impacts, in contradiction to evidence for funds in the US and Europe (Cheung & Roca, 2012). For US investors with a strong emphasis on value
maximization, the DJSIAP may not be a preferred choice as an insignificant alpha and thus underperformance is also estimated in a regression with returns of the S&P 500
(p-value=0.75). While skewness is in its normal range, a moderately high kurtosis may indicate non-normal data which justifies the additional use of the non-parametric sign test.
Table 4: Performance statistics of the indices 2009-2017
Average daily return (%) Standard deviation
DJSIAP BMI ASIA DJSIAP BMI ASIA
2009 0.103 0.101 0.0165 0.0146 2010 0.050 0.062 0.0115 0.0103 2011 -0.061 -0.053 0.0146 0.0136 2012 0.058 0.046 0.0093 0.0084 2013 0.050 0.053 0.0095 0.0092 2014 -0.025 -0.018 0.0077 0.0077 2015 -0.007 0.008 0.0100 0.0093 2016 0.014 0.014 0.0120 0.0113 2017 0.068 0.088 0.0059 0.0053 Total 0.028 0.033 0.0002 0.0002 Kurtosis 2.475 2.567 Skewness -0.129 -0.233
Note: Daily prices sourced from S&P Dow Jones Indices converted to returns
Table 5 concentrates on the performance within the defined estimation and event windows for the BMI and DJSIAP respectively. While there is no inconsistency in standard deviation across all four cases, the average daily returns are similar within the estimation window only and returns within the event window are higher for the DJSIAP (0.11%) than the BMI (0.083%). This finding is in line with the research hypothesis as the price impact from inclusion appears to be positive. The distribution has a moderately high kurtosis in the BMI event window only, which is taken into account as the sign test is used but can also be explained by the rather small sample size of 19 trading days in this window.
18
Table 5: Performance statistics of the sample for event and estimation window
Avg. daily return in % Standard deviation Kurtosis Skewness BMI estimation window
0.026 0.004 0.120 -0.045
BMI event window
0.083 0.003 -0.945 0.213
DJSIAP estimation window
0.023 0.004 0.265 -0.029
DJSIAP event window
0.110 0.004 -0.066 -0.106
Note: Daily prices sourced from S&P Dow Jones Indices converted to returns in percent. Results are averages in the estimation window of 150 trading days and total event window of 19 trading days respectively across 173 companies
The effect of inclusion in the index is insignificant for the day of the announcement and the effective day, as well as the window in between both days and the five trading days after the change has become effective, according to the CAARs in Table 6. Only the CAAR for the pre-announcement window is significant (at 5%) and indicates that investors
anticipate additions before the announcement day. As only large companies are assessed in the DJSIAP it seems reasonable to assume that investors note efforts to increase CSR related activities, given the “invited universe” includes around 600 companies and is released months before announcement, and likely draw early conclusions on composition changes from
information received in the weeks prior to announcement. The sign of CAAR turns negative on ED and afterward which may indicate price pressure on the market, though insignificantly. The total event window shows weak significance for a positive CAAR, which is mainly due to stronger significance in the pre-announcement window.
Table 6: CAARs of the sample for total and sub windows
Event window CAAR in % t-test sign test
(AD-10, AD-1) 0.905 2.284** AD 0.249 1.257 1.284 (AD+1,ED-1) 0.005 0.018 ED -0.082 -1.050 1.173 (ED+1, ED+5) -0.020 -0.055 (AD-10, ED +5) 1.057 1.786*
19 To capture any significance around the event days Table 7 includes further sub
windows tested for significance. While CAARs are positive for all days prior and after AD and ED respectively, weak significance is only given for (AD-1, AD+1), which stands out with a relatively high CAAR, and could represent a realistic window if considering short response delays or time differences.
Table 7: CAARs of the sample for additional subwindows
Event window CAAR in % t-test sign test
(AD-1, AD+1) 0.450 1.882* AD+1 0.070 0.616 1.443 AD-1 0.130 0.843 1.411 (ED-1, ED+1) 0.085 0.580 ED+1 0.155 1.520 1.348 ED-1 0.013 0.091 1.379
Note: * denotes significance at 10%, ** at 5% and *** at 1%
Figure 2 visualizes the increasing abnormal returns until two days after the
announcement day when the CAAR curve flattens as abnormal returns turn negative. A little dip can be noted on ED=7, followed by an increase in abnormal returns to its maximum CAAR on ED+4 and a stronger decrease in abnormal returns after the event window which can indicate a temporary impact only (price pressure). However, it should be noted that CAARs remain overall positive, which is also true if CAARs are presented relative to the AD as in Figure 3.
20
Figure 2: CAARs (%) for total event window and two additional days (AD-10, AD+12)
Note: the x-axis represents days with AD=0, cumulative average abnormal returns in % are represented on the y-axis
Figure 3: CAARs (%) for total event window and two additional days (AD-10, AD+12) relative to AD
Note: the x-axis represents days with AD=0, cumulative average abnormal returns in % are represented on the y-axis relative to AD
21 6. Conclusion
This research aims to find the effect of inclusion to the sustainability index DJSI Asia Pacific on stock prices and forms the hypothesis of a positive price impact based on past research that indicates growing environmental awareness. The findings provide weakly significant positive abnormal returns for the total event window with significance in only one sub-window, before the actual announcement day, which indicates that information may have leaked early or, more likely, has been derived from investors’ analyses of readily available information such as the list of companies invited to the assessment. The results may be explained by the information cost hypothesis (Merton, 1987), as investors may value the event related information according to growing abnormal returns after the announcement and effective day, as well as the signaling hypothesis because the inclusion seems to
communicate a higher firm value than expected by some investors. On the other hand, the period past the event window has not been studied sufficiently to infer permanent changes. A more thorough analysis can also result in evidence for price pressure (Harris & Eitan, 1986) as abnormal returns tend to slope downwards after the event window.
Cumulative average abnormal returns (CAARs) are overall positive and thus contradict evidence, that abnormal returns are significantly negative on day of effective change, as found by Cheung & Roca (2012), which may indicate that investors in the Asia Pacific value the positive CSR assessment of companies by the sustainability index and, given the significant abnormal returns pre-announcement, may seek out possible new
additions beforehand in anticipation of high returns. Whether these findings are evidence for a shift in investors’ mentality towards sustainability index additions, as Cheung & Roca (2012) find predominantly (insignificant) negative abnormal returns, remains questionable.
The significant CAAR in this thesis is twice as high as found in the closely related study by Cheung & Roca (2012) for 2002-2010 (0.52% vs 1.16% for (AD-10, AD)). A difference that, while other results between both analyses are similar though predominantly (insignificantly) positive in this thesis, may indicate that the use of local country benchmarks, as in the study by Cheung & Roca (2012), can contribute to more accurate abnormal returns, given that market returns differ across countries represented in the index.
More generally, a further limitation of this study may consist in the fact that companies have the opportunity to invest in CSR strategies because of their high financial performance and spare funds, according to the slack resource theory (Ziegler, 2010). With only the biggest companies in the benchmark chosen for inclusion in the index smaller companies who lack funds to convey their strategy to investors efficiently or invest into new
22 assets, which would inevitably increase sustainable performance, are disregarded even though their CSR efforts may be on par with companies that achieve a higher market capitalization. Furthermore, the index as a measure for CSR is linked to the assumption that inclusion is a credible signal of higher environmentally and socially responsible activities (Ziegler, 2011). A questionnaire, though extensive with around 100 pages, equates to self-reported activity rather than observed actions which may not convince an investor that distrusts such a methodology and prefers to rely on own judgment or (limited) other measures of CSR.
Future research could thus include a similar event study, to test for consistency, using local country benchmarks and a longer post-effective window to test the price pressure hypothesis.
23 References
Becchetti, L., Ciciretti, R., Hasan, I., & Kobeissi, N. (2012). Corporate social responsibility
and shareholder's value doi:https://doi.org/10.1016/j.jbusres.2011.10.022
Beneish, M. D., & Whaley, R. E. (2002). S&P 500 index replacements. The Journal of
Portfolio Management, 29(1), 51-60.
Carroll, A. B. (1979). A three-dimensional conceptual model of corporate performance.
Academy of Management Review, 4(4), 497-505.
Cheung, A. (. K., & Roca, E. (2013). The effect on price, liquidity and risk when stocks are
added to and deleted from a sustainability index: Evidence from the asia pacific context
doi:https://doi.org/10.1016/j.asieco.2012.08.002
Cochran, P. L., & Wood, R. A. (1984). Corporate social responsibility and financial
performance. The Academy of Management Journal, 27(1), 42-56. doi:10.2307/255956 Consolandi, C., Jaiswal-Dale, A., Poggiani, E., & Vercelli, A. (2009). Global standards and
ethical stock indexes: The case of the dow jones sustainability stoxx index. Journal of
Business Ethics, 87(1), 185-197. doi:10.1007/s10551-008-9793-1
Corporate sustainability. (2017). Retrieved from http://www.sustainability-indices.com/sustainability-assessment/corporate-sustainability.jsp
Corrado, C. J., & Zivney, T. L. (1992). The specification and power of the sign test in event study hypothesis tests using daily stock returns. The Journal of Financial and
Quantitative Analysis, 27(3), 465-478. doi:10.2307/2331331
Davidson, W. N., & Worrell, D. L. (1990). A comparison and test of the use of accounting and stock market data in relating corporate social responsibility and financial
performance. Akron Business and Economic Review, 21(3), 7.
24 Denis, D. K., McConnell, J. J., Ovtchinnikov, A. V., & Yu, Y. (2003). S&P 500 index
additions and earnings expectations. The Journal of Finance, 58(5), 1821-1840. doi:10.1111/1540-6261.00589
Dhillon, U., & Johnson, H. (1991). Changes in the standard and poor's 500 list. The Journal
of Business, 64(1), 75-85.
European SRI study 2016. (2015). Eurosif.
Fowler, S. J., & Hope, C. (2007). A critical review of sustainable business indices and their impact. Journal of Business Ethics, 76(3), 243-252. doi:10.1007/s10551-007-9590-2 Friedman, M. (1962). Capitalism and freedom (university of chicago press, chicago). IL
Google Scholar,
HARRIS, L., & GUREL, E. (1986). Price and volume effects associated with changes in the S&P 500 list: New evidence for the existence of price pressures. The Journal of Finance,
41(4), 815-829. doi:10.1111/j.1540-6261.1986.tb04550.x
Hawn, O., Chatterji, A., & Mitchell, W. (2014). How firm performance moderates the effect of changes in status on investor perceptions: Additions and deletions by the dow jones sustainability index. Boston U.School of Management Research Paper, 2418300 Hawn, O., Chatterji, A. K., & Mitchell, W.Do investors actually value sustainability? new
evidence from investor reactions to the dow jones sustainability index (DJSI). Strategic
Management Journal, , n/a-n/a. doi:10.1002/smj.2752
Hegde, S. P., & McDermott, J. B. (2003). The liquidity effects of revisions to the S&P 500
index: An empirical analysis doi:https://doi.org/10.1016/S1386-4181(02)00046-0
Jain, P. C. (1987). The effect on stock price of inclusion in or exclusion from the S&P 500.
Financial Analysts Journal, 43(1), 58-65. doi:10.2469/faj.v43.n1.58
Karlsson, J., & Chakarova, Y. (2008). Does corporate social responsibility pay off?-an event study of the impact of corporate entry and exit from the dow jones sustainability world index on the market value of a company. Rapport Nr.: Industriell Och Finansiell
25 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.
MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of Economic
Literature, 35(1), 13-39.
MERTON, R. C. (1987). A simple model of capital market equilibrium with incomplete information. The Journal of Finance, 42(3), 483-510.
doi:10.1111/j.1540-6261.1987.tb04565.x
Oberndorfer, U., Schmidt, P., Wagner, M., & Ziegler, A. (2013). Does the stock market value
the inclusion in a sustainability stock index? an event study analysis for german firms
doi:https://doi.org/10.1016/j.jeem.2013.04.005
Renneboog, L., Ter Horst, J., & Zhang, C. (2008). Socially responsible investments:
Institutional aspects, performance, and investor behavior. Journal of Banking & Finance,
32(9), 1723-1742.
Report on socially responsible investing trends in the united states. (2014). ().Social
Investment Forum (SIF).
RIEDL, A., & SMEETS, P. (2017). Why do investors hold socially responsible mutual funds? The Journal of Finance, 72(6), 2505-2550. doi:10.1111/jofi.12547
Ruf, B. M., Muralidhar, K., Brown, R. M., Janney, J. J., & Paul, K. (2001). An empirical investigation of the relationship between change in corporate social performance and financial performance: A stakeholder theory perspective. Journal of Business Ethics,
32(2), 143-156.
SAGE Publications. (2012). SAGE brief guide to corporate social responsibility SAGE.
SHLEIFER, A. (1986). Do demand curves for stocks slope down? The Journal of Finance,
41(3), 579-590. doi:10.1111/j.1540-6261.1986.tb04518.x
26 Van Beurden, P., & Gössling, T. (2008). The worth of values–a literature review on the
relation between corporate social and financial performance. Journal of Business Ethics,
82(2), 407-424.
Van Marrewijk, M. (2003). Concepts and definitions of CSR and corporate sustainability: Between agency and communion. Journal of Business Ethics, 44(2), 95-105.
Van Stekelenburg, A., Georgakopoulos, G., Sotiropoulou, V., Vasileiou, K. Z., & Vlachos, I. (2015). The relation between sustainability performance and stock market returns: An empirical analysis of the dow jones sustainability index europe. International Journal of
Economics and Finance, 7(7), 74.
Wai, K. C. (2011). Do stock investors value corporate sustainability? evidence from an event study. Journal of Business Ethics, 99(2), 145-165. doi:10.1007/s10551-010-0646-3 Wang, Y., & Chen, Y. (2017). Corporate social responsibility and financial performance:
Event study cases. Journal of Economic Interaction and Coordination, 12(2), 193-219. doi:10.1007/s11403-015-0161-9
Wilson, M. (2003). Corporate sustainability: What is it and where does it come from? Retrieved from https://iveybusinessjournal.com/publication/corporate-sustainability-what-is-it-and-where-does-it-come-from/
Wood, D. J. (1991). Corporate social performance revisited. Academy of Management
Review, 16(4), 691-718.
Ziegler, A. (2012). Is it beneficial to be included in a sustainability stock index? A panel data study for european firms. Environmental and Resource Economics, 52(3), 301-325. doi:10.1007/s10640-011-9529-z
Ziegler, A., & Schröder, M. (2010). What determines the inclusion in a sustainability stock
index?: A panel data analysis for european firms
