• No results found

Corporate Social Responsibility in the Netherlands

N/A
N/A
Protected

Academic year: 2021

Share "Corporate Social Responsibility in the Netherlands"

Copied!
48
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Corporate Social Responsibility in the

Netherlands

The influence of media exposure on Stock

Value

(2)

Corporate Social Responsibility in the Netherlands:

the influence of media exposure on Stock Value

Master Thesis

Wouter van Os Groningen

Supervisors: L. Dam

B. Scholtens

(3)

Preface

This master thesis investigates the influence of media exposure about corporate social responsibility (CSR) on corporate financial performance (CFP). Media exposure about CSR is categorized on the basis of the nature of the CSR topic. This makes it possible to identify whether certain aspects of CSR had more impact on CFP than others. It is possible to conclude that there is indeed certain media exposure that has significant influence on financial performance. News articles reporting on risk related topics appear to have a strong impact on financial performance. A natural economic explanation is that investors are simply risk averse. New information about riskiness of an organization influences investor appetite for that organization. It also becomes clear that news with the lowest value judgment had the largest impact on the market. This type of news probably brings new information; more shocking news reaches the market sooner and does not bring a shock when the news is in the paper afterwards. In a sense, very bad news has been anticipated.

The dataset of this master thesis is based on news articles printed in the newspaper „Het Financieel Dagblad‟ during the past five years. News articles about the CSR of organizations listed on the Amsterdam Stock Exchange were included in the database.

(4)

Index

Introduction ... 5

1: Literature review ... 7

Defining Corporate social responsibility ... 7

History of Corporate social responsibility ... 8

Determinants of Corporate social responsibility ... 9

Corporate social responsibility, reputation and corporate financial performance .... 9

Defining corporate financial performance and corporate reputation ... 10

Media exposure and CSR ... 11

2: Hypotheses ... 12

3: Data description ... 14

4: Methodology ... 18

5: Results ... 23

Abnormal return positive/negative news ... 23

Abnormal Return & areas of concern ... 25

6: Conclusion ... 37

References ... 39

Appendix 1 EIRIS Framework ... 41

Appendix 2 MacKinlay method ... 42

(5)

Introduction

Corporate social responsibility is a hot topic in the Netherlands today and receives quite some attention in society. There are various examples that support this statement. For example, many organizations publish an annual corporate social responsibility report, investors use the topic as an investment strategy, and the fact that the cancer foundation invested in tobacco companies received headlines in the newspapers. The discovery that the pension fund PGGM invested in companies that were involved with the production of land mines even caused political attention. Dutch militaries were removing such land mines at the same time in Afghanistan

Corporate social responsibility (CSR) is an umbrella term which captures various social aspects and consequences of activities. A popular and frequently used slogan to illustrate the actual meaning of CSR is: “People, Planet and Profit”. The main goal of CSR is to encourage organizations to hold itself accountable for the impact of its actions on society. Not only the shareholders, but all the organizations‟ stakeholders are important for that matter. The environment, social concerns, governance and ethical concerns are central to the topic CSR and are also used as point of departure of this thesis. There is explicitly chosen to zoom in on the various aspects of CSR since it is such a broad subject. We wanted to identify whether some aspects had more influence than others.

The basics of CSR do not seem to be in line with the view of traditional economics. Traditional economics emphasize the importance of shareholder wealth. A widely spread and accepted opinion of what the main goal of an organization should be, is to maximize shareholder wealth by increasing the value of their common stock. This is therefore also the sole responsibility of the management and the main reason of

existence for an organization (Friedman, 1970). In Friedman‟s view there does not

(6)

The impact of CSR on the performance of organizations has been a popular subject in economic writing and although it is a relatively new subject a rich literature emerged on the topic. The influence of CSR on corporate financial performance (CFP) is ambiguous and remains to be interesting to investigate. Orlitzky (2003) performed a large scale meta analysis in which he involved 52 studies on the topic CSR. The main conclusion of this analysis states that there is a positive correlation between CSR and CFP. Furthermore, corporate reputation seems to be a major facilitator in this positive relationship. This thesis builds on the conclusions of Orlitzky and also wants to test whether there is a correlation between CSR and CFP and corporate reputation. An important assumption in this analysis is that media exposure of an organization has influence on its corporate reputation.

The main research question of this thesis is whether there is an influence of media exposure about CSR on the financial performance of the organization. To answer this question an event study is conducted. An event study is a commonly used tool to research the influence of a particular event on the financial performance of an organization. In this particular case, such an event is media exposure about CSR. To map media exposure, we analyzed newspaper articles with the subject CSR that were published during the past five years. Each news article was subsequently classified on the basis of their specific topic and tendency.

This thesis contributes to the existing literature on CSR and financial performance by its unique focus on the influence of media exposure. Because each article is classified, it is also possible to search for differences in the impact of several areas of CSR. The scope of this research is restrained to organizations which are denoted on

the Amsterdam Stock Exchange and to the news articles printed in „Het Financieel

Dagblad‟. These restrictions are chosen to enlarge the uniformity of the data, although they naturally limit the amount of data.

(7)

1: Literature review

The following chapter will be dedicated to the existing literature on the relation between CSR, CFP, corporate reputation and the influence of media exposure.

Defining Corporate social responsibility

CSR is a comprehensive concept and therefore difficult to summarize in a simple phrase. This is probably the reason that there is not one unanimously accepted definition of CSR, there are small differences in the definitions used in the literature. Wood (1991) stated that there was no completely satisfactory definition of the concept at that time. Wood took the existing definitions and points of departure a step

further and formulized the definition: “a business organization‟s configuration of

principles of social responsibility, processes of social responsiveness, and policies, programs, and observable outcomes as they relate to the firm‟s societal relationships. Despite the attempt of Wood to introduce the future standard definition, other definitions based on different views on the topic kept emerging. Frequently used definitions focus on voluntary firm actions designed to improve social or environmental conditions (Aguilera, Rupp, Willeams and Ganapathi, 2007; Mackey and Mackey, 2007). CSR can also be seen as a form of soft law. It provides guidelines that are beyond the range of regular law. Some scholars claim that it is impossible to formulate the law in such a way that there is no need for soft law. McWilliams and Siegel (2001) say that the purpose of CSR is beyond the interest of the firm and not required by law. CSR exists to fill up the gaps in regular law. Heal (2005), developed another view on CSR and gives a simpler and more economic definition of CSR. Heal analyses the subject thoroughly and formalises the CSR into

the definition: „a program of actions taken to reduce externalized costs or to avoid

(8)

History of Corporate social responsibility

Corporate social responsibility originated in the beginning of the 20th century in the

United States and attracted attention during the 1940‟s when investing in

organizations with unfair labour practices was rejected. During the socially engaged 1960s and 1970s, corporate social responsibility became even more a point of discussion (Shank, 2005). Friedman (1970) reacts on the developments in the view of the society toward business management with his article „The social responsibility of business is to increase its profits‟. He emphasizes in this article that only a person can have something like a responsibility, an organization on the other hand cannot. The responsibility of people working inside an organization is to make the best possible choices for the sake of the organization. Employers of an organization cannot spend the money of their stockholders for a suboptimal choice, whether it is or is not in the interest of a social goal. The statements of Friedman caused many counter reactions of people opposing his opinion. Their main critiques attack his simplicity, authority and finality. According to these authors the subject should be seen in a much broader perspective (Kolstad, 2007; Davis, 2005; Grant, 1991). The Friedman opposition frequently uses the argument that CSR does not even have to stand in the way of financial performance but could contribute to it.

McGuire (1988) developed the „slack resource theory‟ in which he states that financial results enable organizations to develop a certain corporate responsibility. McGuire supports a minority opinion with this statement, since more evidence emerged that there is opposite causality: corporate responsibility supports financial performance.

(9)

Not only has the economic literature catered to the topic of CSR. Some major financial institutions are investigating the opportunities of the subject. The major question is whether it pays off to invest in social responsible organizations. The research firm KLD research & Analytics (www.kld.com) has developed the Domini 400 Social Index with publicly-traded companies which have met certain standards of social and environmental performance. The Financial Times Stock Exchange (FTSE 100), the index of the 100 largest assets traded at the London Stock Exchange, developed the FTSE4good (www.ftse.com). The intention of this index is to identify those companies who operate according to certain principles of CSR.

Determinants of Corporate social responsibility

The influence of CSR on the performance of an organization can only be investigated when there is decided upon the actual determinants of CSR. The definition of CSR written by Heal (2005) was previously chosen as the most applicable one: „a program of actions taken to reduce externalized costs or to avoid distributional conflicts‟. But what does it actually imply for an organization. Which externalized costs and distributional costs are taken into account? In other words, which factors determine the actual level of CSR of an organization?

Ethical Investment Research Services (EIRIS), is an authoritative research organization in the field of ethical investment which developed a framework to assess this problem. EIRIS recognizes four areas of concern in the field of CSR: social, environment, governance and ethical concerns. These four areas can be aggregated into 17 CSR topics, which are to a certain extent measurable. This framework is used by EIRIS to make a certain classification on the level of CSR of organizations. This framework is also used for this thesis and shall be further elaborated in chapter 4.

Corporate social responsibility, reputation and corporate

financial performance

(10)

correlation between CSP and CFP. In this relation, reputation seemed to be the most important mediator.

The relationship between corporate reputation and CSR is also emphasized by the Harris-Fombrun Reputation Quotient (Fombrun, 1996). This quotient is built on six „pillars‟ which measure reputation. CSR is an important component of the quotient and is one of the six „pillars‟. Schwaiger (2004) concludes in his paper „Components and parameters of corporate reputation‟ that the Harris-Fombrun Reputation Quotient is the most advanced measurement model of reputation. Schwaiger underlines the fact that for designing the quotient, a survey under a broad range of stakeholders was held and not just under experts. This construction enlarges the reliability.

The relationship between corporate reputation and CFP is also discussed by Schwaiger. He summarizes some positive effects of a good corporate reputation. A good corporate reputation could for example have a positive effect on the recruitment department; one can „win the battle for talent‟. Customer confidence is affected in several ways on a positive manner and the relationship with other stakeholders is boosted. Eventually, corporate reputation could increase profits and therefore financial performance. Podnar (2007) links CSR, corporate reputation and CFP even more obvious. CSR is seen in his paper as a form of corporate marketing. This kind of marketing stands above the product/service level of an organization and would be necessary to obtain long-term business survival.

This paper contributes to this by researching the effect of media exposure about CSR on an organization. CSR is further aggregated into sub-topics to find which particular topic has the greatest impact.

Defining corporate financial performance and corporate

reputation

There are several methods to determine the financial performance of an organization. One can look for example at the market to book value (Tobin‟s Q). Another possibility is to look at several accounting profit ratios, like Return on Assets (ROA) or Return on Equity (ROE). Another possibility for determining financial performance is looking at stock market returns (Ross, Westerfield, Jaffe, 2005). Market returns are used in this thesis as a proxy of financial performance.

(11)

definition of Hall (1992) appeared to be the most usable one: a corporate reputation consists of the knowledge and the emotions held by individuals. This is also the definition that is used for this thesis.

Media exposure and CSR

Orlitzky et. al. (2003) concludes in their meta-analysis that there is a relationship between CSR and CFP, where corporate reputation is the major facilitator. Swapping these concepts with the definitions that we use in this thesis gives the following statement:

The knowledge and emotions held by individuals over the a program of actions used by an organization, taken to reduce externalized costs or to avoid distributional conflicts has an impact on stock value of the organization.

(12)

2: Hypotheses

The connection between CSR, CFP and corporate reputation is taken as point of departure for this thesis. The influence of media exposure on corporate reputation is used in this thesis as a method to measure corporate reputation. This thesis investigates whether it is possible to find influence of media exposure about CSR on the financial performance of the organization. This is formulized into two hypotheses.

Media exposure about CSR is classified in two separate manners for this thesis. The fist classification is made on the basis of their positive or negative tendency. The tendency is interpreted as a value judgment for the level of CSR of the organization. It could be expected that this has consequences for the reputation of the organization and therefore on the CFP. It could be questioned whether media exposure concerning CSR has a steering effect on the level of CFP. In other words: is there a positive (negative) deviation in CFP when there is positive (negative) media exposure of a particular organization? This would be very interesting to know since this suggest that one could make certain assumptions about the direction of the CFP of an organization after reading about its CSR. This question is translated into hypothesis 1.

Hypothesis 1:

H0: Media exposure on corporate social responsibility of an organization has a significant relationship with the corporate financial performance of that organization.

H1: Positive (negative) media exposure on corporate social responsibility of an organization has a significant positive (negative) effect on the corporate financial performance of that organization.

(13)

about CSR is classified in these CSR topics. This is done because it is expected that there are differences between the impact of these topics on CFP.

Hypothesis 2:

H0: There are no differences in the impact on corporate financial performance between areas of corporate social responsibility.

(14)

3: Data description

This section presents the data that is collected for this thesis. We collected data on

media exposure about CSR from the archives of the Dutch newspaper „Het

Financieel Dagblad‟. ´Het Financeel Dagblad´ is one of the leading financial newspapers in the Netherlands. Data on CFP of the organizations that were mentioned in the news articles is collected from DataStream. The daily return indexes calculated the performance of a stock assuming that all dividends and distributions are reinvested.

We used all the news articles that were printed from November 2002 till September 2007 concerning the topic CSR of companies listed on the AEX for analysis. This consists in total of 104 news articles. We subsequently analysed and classified the articles on the basis of their positive/negative impact and on the CSR area that is addressed. Impact can differentiate between ´very negative´ to ´very positive´. The five possible scores on impact and the number that they occurred in the data are summarized in table 1. The events are not evenly distributed over the impact categories. There is only one event that is classified under „Very Negative‟ and only three events under „Neutral‟. This has negative consequences on the statistical reliability of the outcomes. For example: with such a small number of events, the impact of large outliers in the data is greater.

Table 1

Impact news articles

Impact Number occurred

Very Negative 1

Negative 31

Neutral 3

Positive 62

Very Positive 7

Source: Het Financieel Dagblad. Impact describes the value judgement in the news articles about corporate social responsibility.

(15)

on the area of concern of the news articles. The complete EIRIS framework is in the appendix, alternatively at www.eiris.com. The EIRIS topics that occurred in the news articles and their number of occurrence are summarized in table 2. The data is again

not evenly distributed over the categories. Both the categories „Ethical concerns;

military‟ and Ethical concerns; genetic engineering‟ merely have one event. „Governance; risk management‟ on the other hand contains 43 categories.

Please note that an article can have more than one CSR area. The total number that the areas occur does not correspond with the total number of news articles because of that reason.

Table 2

Area of concern news articles

Source: Het Financieel Dagblad. CSR areas describe the CSR topic that is addressed in the news articles. The topics correspond to the CSR topics formulated by EIRIS.

All of the news articles were marked with a certain level of positive/negative impact and with one or more codes. A cross table of the impact and codes is presented in table 3. This gives some more insight in the distribution of the data.

Only the newspaper articles concerning organizations listed on the AEX were included in the analysis. This restriction was chosen to narrow down the scope of this thesis and to increase the possibility to make comparisons between the results of the investigation. Eventually 22 organizations qualified for this research. All of this 22 AEX listed organizations emerged at least once in the newspaper during the period November 2002 - September 2007 and for example in the case of Royal Dutch Shell even up to 15 times.

CSR area Number occurred

Environment; policy 16

Environment; performance 4

Social; stakeholders 15

Social; human rights 15

Social; sourcing 10

Governance; risk management 43

Governance; risks 12

Ethical concerns; military 1

Ethical concerns; genetic engineering 1

(16)

Table 3

Type and nature of news articles

CSR AREA

Very

Negative Negative Neutral Positive

Very Positive

Environment; policy 29,41% 0% 0% 58,82% 11,76%

Environment; performance 0% 0% 25,00% 75,00% 0

Social; stakeholders 0% 13,33% 6,67% 60,00% 2,61%

Social; human rights 0% 0% 0% 92,86% 7,14%

Social; sourcing 0% 80,00% 0% 20,00% 0%

Government; risk management 2,33% 25,58% 2,33% 58,14% 11,63%

Government; risks 0% 25,00% 0% 50,00% 25,00%

Ethical Concerns; military 100,00% 0% 0% 0% 0%

Ethical concerns; genetic engineering 0% 100,00% 0% 0% 0%

Ethical concerns; breaches of certain

labour standards 0% 100,00% 0% 0% 0%

Source: Het Financieel Dagblad. CSR areas describe the CSR topic that is addressed in the news articles. The topics correspond to the CSR topics formulated by EIRIS. Positive/Negative describes the value judgement in the news articles on the CSR areas. The methods used to measure CSR are described above. The measurement of financial performance on the other hand is done by collecting the return indexes of

the 22 organizations that are included in this thesis. The return index1 is the most

appropriate measurement of financial performance since it includes dividends. The return indexes are all collected from DataStream.

Some descriptive statistics of the return indexes of the organizations are presented in

table 4. The statistics are based on data from the period November 2002 –

September 2007. The mean, standard deviation, minimum and maximum are

expressed in basis points which are equal to 1/100th of 1%. The expressions describe

respectively the average return, the dispersion of the data from the mean, the minimum value and the maximum value. The skewness is a measure for the asymmetry of the data relative to the normal distribution.

The results of ordinary least squares regression (OLS) are presented in the last four columns. Alpha (α) describes the risk-adjusted returns, which is the intercept of the

regression. Beta (β), systematic risk, measures the volatility of a stock relative to a

1

)

100

1

(

*

*

1 1

n

DY

PI

PI

RI

RI

t t t t

 

Where

RI

= Return index

PI = Price Index

DY = Dividend yield

(17)

benchmark, and is the risk that is priced in the market (Ross, Westerfield, Jaffe, 2005). The AEX, Euronext and FTSE100 are the three indexes used as benchmark.

Table 4

Descriptive statistics of the daily returns of the included companies.

This table presents descriptive statistics of the daily returns of the Companies included in this analysis between November 2002 – September 2007. All absolute measures are expressed

in basis points, alphas and betas are calculated by estimating

R=α+βaex*AEX+βeuronext*Euronext+βftse*FTSE using OLS.

The mean daily average returns vary between -14 and 11.2 basis points., The standard deviations show common daily fluctuations of about 2 to 3 percent (200-300 basis points) returns are somewhat skewed, which is also underlined by the large minimum and maximum values that are observed. With an alpha of zero, none of the organizations did outperform the benchmark. The three betas are different per organization, more than 1 indicates that the stock is more volatile than the benchmark. The betas smaller than 1 indicate that the stock is less volatile than the market.

Organizations Mean St. Dev. Skewness Minimum Maximum α βAEX βEuronext βFTSE

(18)

4: Methodology

The following section presents the methodology that is used for this thesis. The methodology is developed to search for possible abnormal deviations in the return index during the period of media exposure.

Event study

Event studies are one of the most frequently used methods in economic research. An event study measures the influence of an event on the value of a firm (MacKinlay, 1997). This influence is measured on de basis of abnormal returns during the period of the event relative to expected returns. Expected returns are called “normal returns” and determining these normal returns are therefore an important part of an event

study. When “normal returns” are determined the “abnormal return“ can be

determined.

The period prior to the event is used for estimation of the normal return in the period surrounding the Event. The period prior to the event is therefore very suitable called the „estimation window‟. The period surrounding the vent is called the „event window‟. This is depicted in figure 1.

Figure 1 Lay out event study2

There are several methods for determining the normal return. Brown (1985) describes and explains three procedures: using mean adjusted returns, market adjusted returns and the ordinary leas squares (OLS) market model. MacKinlay (1997) states that the market model generates the most reliable outcomes. This thesis uses a three factor model, which is a natural extension of the market model. The three factors represent respectively the AEX, Euronext 100 and the FTSE world

2 Source: MacKinley (1997)

T0

T1

0

T2

T3

(19)

index. The organizations that are included in this thesis are all listed on the AEX. The AEX represents the 25 largest stock funds of the Amsterdam stock exchange and is chosen as the first benchmark for the data (http://www.aex.nl/aex.asp). The second benchmark is Euronext 100 , which is the index of the 100 largest funds traded on Euronext (www.euronext.nl). The third and final benchmark used in this thesis is the FTSE world index. This index covers practically all the investable market capitalizations in the world (www.ftse.com).

The three benchmarks are used during the estimation period of each individual sample to calculate the normal return of the event window. For each event, the organizations‟ return index, and the return indexes of the AEX, Euronext 100 and the FTSE world index are used as input of a factor analysis determining the normal return of the organization during the event window. Equation (1) shows the three factor model that is used for estimating normal return.

OLS was used in estimating equation (1). By doing this, the

and the

of AEX,

Euronext and FTSE are estimated for each event. For each stock the estimation window is a period of 100 days prior to the event window.

t i FTSE F i Euronext E i AEX A i i it

R

t

R

t

R

t

R

,

(1)

Where Rit is the daily stock return of company I at date t, RAEXt the daily return of the

AEX index at date t, REURONEXt the daily return of the Euronext 100 at date t, RFTSEt

the daily return of the FTSE world index and the βi are the factor betas.

m i i i

ˆ

ˆ

ˆ

ˆ

   

1 0 1 0 1 2 1

)

ˆ

(

)

ˆ

)(

ˆ

(

T T m m T T m m i i i

R

R

R

    

(20)

   1 1 1 0 1 ˆ T T i i R L  

   1 1 1 0 1 ˆ T T m m R L  

The estimated coefficients based on of equation (1) are used to make predictions for the event window. Abnormal return is calculated by subtracting normal return from the actual return:

it it

R

R

AR

ˆ

(3)

The abnormal returns explain the actual impact of the event and are therefore the most important ingredients of the analysis. This impact becomes especially clear when the abnormal returns are aggregated through time or cross sectional across events. To test whether the abnormal returns are significant or not, Student‟s t-test is performed. The standard deviations of the data and the abnormal return are input of the t-Statistic but there are two slightly different methods in use. Both methods are used for the analysis and are labelled in this thesis respectively the MacKinlay method and the Khotari method.

The MacKinlay method

This method is used in the paper of MacKinlay in which he elaborates the methodology of the event study. The mean abnormal return of several events is simply calculated by:

(21)

The standard deviation is the error term of equation 1, used on an out of sample basis. The standard deviation of the mean abnormal return is calculated by taking the square root of the variance of the mean abnormal return:

)

var(

AR

(5) Where:

  N i i N AR 1 2 2 1 ) var(

The outcome of the Students t-test, the t-statistic, is calculated by dividing the AR by the standard deviation of the data.

AR statistic T 

(6)

The t-statistic explains in combination with a certain probability whether the data is distributed in a certain way. When the t-statistic indicates data is distributed in that way witch a certain probability, than the significance of the event is proven.

Khotari method

The Khotari method is slightly different from the MacKinlay method since it incorporates the standard deviation of each separate event in its abnormal return. The standard deviation of each event is than normalized and for each event equal to one. The mean abnormal return is subsequently calculated using equation (4). The standard deviation of the mean abnormal return is also unchanged and calculated using equation (5). The variance is normalized and for each separate event equal to one and can now simply be calculated using:

(22)

considered to be “normal” average fluctuations. In effect, this method puts less weight on very volatile stocks.

Cumulative abnormal return

The cumulative abnormal return (CAR) is calculated subsequently. The CAR is nothing more than the sum of abnormal returns and is calculated as follows:

  2 1 ) , ( 1 2    

i i AR CAR (8)

(23)

5: Results

The following section presents the results of the research. The design of the results has three main components. First the impact of negative/positive news on financial performance is analyzed by looking at abnormal returns, the results are presented in table 5. Secondly the impact of the several areas of concern of CSR are analyzed on the basis of their abnormal return, the results of this are presented in table 6-9. Finally, cumulative abnormal return (CAR) is analyzed for both positive/negative news and the CSR areas of concern, the results are presented in table 10-14.

Both the MacKinlay as the Khotari method are used for this study but the method proposed by Khotari delivered more convincing results. The results that are presented in this chapter are therefore the result of the method proposed by Khotari (2006). The results generated using the method proposed by MacKinley can be found in appendix 2 for the interested reader.

The abnormal returns and the CAR‟s are consequently presented in combination with a t-Statistic. The t-Statistic explains the significance of the results and is associated with a certain probability. The significant results are marked with one, two or three stars, dependent on their level of probability, respectively 10%, 5% and 1%.

Abnormal return positive/negative news

(24)

possible explanation for the negative abnormal returns after positive media exposure is that the market believes that CSR is expensive. When there is news that an organization operates in a social responsible manner, the market could perceive this as a sub-optimal way to operate. When this is new negative information for the market this would be translated in a negative deviation in the stock price.

Table 5

Abnormal Returns for event study after the impact of positive/negative news on financialperformance

Source: DataStream

* Significant at the 10% level (two-tailed). ** Significant at the 5% level (two-tailed). *** Significant at the 1% level (two-tailed).

AR = Abnormal Returns, t –Statistics = Student‟s t-statistics, AR values are calculated using

the equation: AR = Rit -Rnormal where Rit is the actual return at day t and Rnormal is the normal

return at day t. Rnormal is the result of a three factor model were respectively the AEX, Euronext and FTSE 100 were included as market indices. Estimation window = [-120, -20].

Neutral media exposure about CSR has a clear positive impact on financial performance. A possible explanation for the relative strong influence of neutral news is that it probably contains new information for the market. This news is categorized as neutral since it does not contain shocking new information. Shocking new

Very Positive Positive Neutral Negative Very Negative

Event

day AR t-Statistics AR t-Statistics AR t-Statistics AR t-Statistics AR t-Statistics

(25)

information would have reached the market in one way or another and this news is new. The market reacts positive at this extra information.

Abnormal Return & areas of concern

The second and more thorough segmentation in the set of news articles was made on the basis of their area of concern. Chapter 4, data description, already dealt with the segmentation of the areas of concern. In short: there are four possible areas of concern identified, namely ethical concerns, governance, social and environment. These areas are further split up into in total 17 CSR topics of which 10 actually appeared in the news articles.

Table 6

Abnormal Returns for event study after the impact of media exposure about the CSR area ‘Environment’ on financialperformance

Source: DataStream

* Significant at the 10% level (two-tailed).

Policy Performance

Negative Neutral Positive Negative Neutral Positive

Event

day AR t-Statistics AR t-Statistics AR t-Statistics AR t-Statistics

(26)

The categories with CSR areas of concern were all investigated for possible abnormal returns. This was done using two different methods. The first method only made a differentiation between CSR topics and did not take the positive/negative tendency into account. The second method used both the differentiation in CSR topics and the differentiation on the basis of positive/negative tendency. The results of the first method can be found in appendix 3. The results of the second method are presented in this chapter in table 6-9.

The surplus value of taking the CSR topics as well as the positive/negative tendency into account is that it gives a clear and detailed overview of the impact. Another advantage is that the contradicting results of negative and positive news articles cannot cancel each other out. A possible disadvantage of this method can be

summarized in the saying: „were the hogs are many, the wash is poor‟. In other

words: the more deviations there are made in the data, the fewer events per category. This has some consequences on the statistical significance of the results. The reader could for that reason also take a good look at the results of the first method which are presented in appendix 3.

Table 7

Abnormal Returns for event study after the impact of media exposure about the CSR area ‘Social concerns’ on financialperformance

Stakeholders Human Rights

Negative Neutral Positive Negative Neutral Positive

Event

day AR t-Statistics AR t-Statistics AR t-Statistics AR t-Statistics

(27)

Continuation table 7

Sourcing

Negative Neutral Positive

Event

day AR t-Statistics AR t-Statistics

-10 0.261 0.738 N.A. 0.094 0.132 -9 -0.334 -0.944 N.A. 0.077 0.109 -8 0.632 1.787* N.A. -0.578 -0.818 -7 0.126 0.356 N.A. 0.815 1.153 -6 0.343 0.970 N.A. -1.471 -2.081** -5 -0.040 -0.114 N.A. 0.080 0.113 -4 0.213 0.603 N.A. -0.824 -1.165 -3 0.222 0.627 N.A. -0.206 -0.291 -2 0.466 1.318 N.A. 0.182 0.257 -1 0.061 0.173 N.A. 0.380 0.538 0 -0.305 -0.863 N.A. 1.187 1.679* 1 0.009 0.026 N.A. -0.120 -0.170 2 0.034 0.096 N.A. 0.829 1.172 3 0.264 0.748 N.A. 0.821 1.161 4 0.136 0.385 N.A. 0.435 0.615 5 -0.342 -0.966 N.A. 0.085 0.120 6 0.368 1.040 N.A. 0.053 0.075 7 -0.220 -0.622 N.A. -0.302 -0.428 8 0.145 0.410 N.A. -0.349 -0.493 9 0.138 0.391 N.A. 0.217 0.308 10 -0.176 -0.497 N.A. 0.069 0.097 Source: DataStream

* Significant at the 10% level (two-tailed). ** Significant at the 5% level (two-tailed). *** Significant at the 1% level (two-tailed).

AR = Abnormal Returns, t –Statistics = Student‟s t-statistics, AR values are calculated

using the equation: AR = Rit -Rnormal where Rit is the actual return at day t and Rnormal

is the normal return at day t. Rnormal is the result of a three factor model were respectively the AEX, Euronext and FTSE 100 were included as market indices. Estimation window = [-120, -20]

(28)

results at all for the second topic, human rights. The third topic, sourcing, has more convincing results. There is a significant abnormal return (10%) at the event day. The third category of CSR areas, governance, gives some interesting results. There are two topics of this category addressed in the news: risk management and risks. Both topics give significant abnormal returns. The results on risk management, are convincing since there are highly significant results on the days surrounding the event day. Neutral news causes a positive deviation in the return index on the day of the event. This outcome is significant even on the probability level of 1%. The development of the abnormal returns prior to the event day of neutral news on risk management also show strong significant results. Positive news on risk management also seems to have coherence with abnormal return. There is significant abnormal return (1%) one day prior to the event day.

Table 8

Abnormal Returns for event study after the impact of media exposure about CSR area ‘Governance’ on financialperformance

Risk management Risks

Negative Neutral Positive Negative Neutral Positive

Event day AR t-Statistics AR t-Statistics AR t-Statistics AR t-Statistics AR t-Statistics -10 -0.369 -1.280 0.305 0.305 -0.013 -0.073 0.329 0.570 N.A. -0.196 -0.588 -9 0.408 1.412 1.663 1.663* -0.021 -0.118 -0.956 -1.655* N.A. 0.706 2.119** -8 -0.031 -0.107 0.209 0.209 0.040 0.217 0.503 0.872 N.A. -0.326 -0.977 -7 -0.166 -0.576 7.070 7.070*** -0.327 -1.792 -0.397 -0.687 N.A. -0.398 -1.195 -6 0.139 0.481 1.257 1.257 0.184 1.005 -0.419 -0.726 N.A. 0.057 0.172 -5 0.153 0.529 0.111 0.111 -0.009 -0.047 0.097 0.169 N.A. -0.236 -0.708 -4 0.110 0.380 -2.253 -2.253** -0.037 -0.203 -0.698 -1.209 N.A. 0.187 0.560 -3 0.132 0.457 -1.258 -1.258 -0.112 -0.614 -0.149 -0.258 N.A. -0.135 -0.406 -2 -0.167 -0.578 1.674 1.674* 0.019 0.103 -0.161 -0.279 N.A. 0.269 0.807 -1 -0.202 -0.699 1.545 1.545 -0.557 -3.049*** -0.368 -0.638 N.A. -0.163 -0.490 0 -0.132 -0.457 3.448 3.448*** 0.154 0.846 -1.104 -1.912* N.A. -0.008 -0.024 1 0.225 0.779 0.926 0.926 -0.159 -0.870 -0.729 -1.263 N.A. -0.201 -0.604 2 -0.295 -1.023 -1.441 -1.441 0.053 0.290 0.098 0.170 N.A. 0.176 0.528 3 -0.023 -0.079 0.431 0.431 0.172 0.941 -0.264 -0.458 N.A. -0.128 -0.384 4 -0.068 -0.237 0.202 0.202 -0.097 -0.532 0.271 0.469 N.A. -0.079 -0.238 5 -0.084 -0.292 0.684 0.684 0.049 0.267 0.137 0.237 N.A. -0.116 -0.349 6 0.293 1.016 0.070 0.070 -0.112 -0.615 -0.081 -0.140 N.A. -0.261 -0.783 7 0.165 0.570 -0.196 -0.196 0.001 0.004 -0.090 -0.156 N.A. 0.654 1.963** 8 -0.292 -1.012 -0.313 -0.313 0.117 0.639 -0.125 -0.216 N.A. 0.101 0.303 9 0.144 0.499 0.122 0.122 -0.158 -0.865 -0.580 -1.005 N.A. -0.013 -0.039 10 0.163 0.566 -0.594 -0.594 0.116 0.636 -0.272 -0.471 N.A. 0.090 0.271 Source: DataStream

* Significant at the 10% level (two-tailed). ** Significant at the 5% level (two-tailed). *** Significant at the 1% level (two-tailed).

AR = Abnormal Returns, t –Statistics = Student‟s t-statistics, AR values are calculated using

the equation: AR = Rit -Rnormal where Rit is the actual return at day t and Rnormal is the normal

(29)

The second topic, risks, also causes significant abnormal return (10%) on the event day. Negative news seems to have a negative impact on the return. Positive news on risks does not cause any convincing abnormal returns on the other hand. There are only quit long before and after the event some significant results, but it is difficult to assign these results to the event day.

Ethical concerns is the fourth category of CSR areas. Three topics of this category appeared in the news: military, genetic engineering and breaches of certain labour standards. Military, and genetic engineering did not show coherence with return. Negative news on code 15, breaches of certain labour standards, give very strong significant results. There are mainly negative abnormal returns(10% and 1%) prior to the event day. These changes one day before the event day in positive abnormal returns. This positive abnormal return keeps appearing till nine days after the event day. There are strong positive reactions on negative news on this topic.

Table 9

Abnormal Returns for event study after the impact of media exposure about the CSR area ‘Ethical concerns’ on financialperformance

Military Genetic Engineering

Negative Neutral Positive Negative Neutral Positive

Event

day AR t-Statistics AR t-Statistics

-10 -0.006 -0.006 N.A. N.A. 0.410 0.410 N.A. N.A.

-9 0.392 0.392 N.A. N.A. -2.225 -2.225** N.A. N.A.

-8 -0.094 -0.094 N.A. N.A. 0.977 0.977 N.A. N.A.

-7 0.082 0.082 N.A. N.A. -0.356 -0.356 N.A. N.A.

-6 -0.023 -0.023 N.A. N.A. -0.033 -0.033 N.A. N.A.

-5 -0.199 -0.199 N.A. N.A. -0.758 -0.758 N.A. N.A.

-4 -0.215 -0.215 N.A. N.A. 0.167 0.167 N.A. N.A.

-3 0.056 0.056 N.A. N.A. 0.250 0.250 N.A. N.A.

-2 0.492 0.492 N.A. N.A. -0.414 -0.414 N.A. N.A.

-1 -0.173 -0.173 N.A. N.A. 0.264 0.264 N.A. N.A.

0 -0.113 -0.113 N.A. N.A. 0.781 0.781 N.A. N.A.

1 0.772 0.772 N.A. N.A. -0.411 -0.411 N.A. N.A.

2 -0.564 -0.564 N.A. N.A. -0.369 -0.369 N.A. N.A.

3 -0.047 -0.047 N.A. N.A. -0.175 -0.175 N.A. N.A.

4 -0.515 -0.515 N.A. N.A. 0.530 0.530 N.A. N.A.

(30)

Continuation table 9

Breaches of certain labour standards

Negative Neutral Positive

AR t-Statistics 0.083 0.166 N.A. N.A. -0.830 -1.659* N.A. N.A. -1.613 -3.225*** N.A. N.A. -0.413 -0.826 N.A. N.A. 0.423 0.845 N.A. N.A. 0.458 0.916 N.A. N.A. -0.559 -1.118 N.A. N.A. 0.002 0.004 N.A. N.A. -1.182 -2.364* N.A. N.A. 1.542 3.083*** N.A. N.A. -0.161 -0.321 N.A. N.A. 0.142 0.284 N.A. N.A. 0.225 0.450 N.A. N.A. -0.394 -0.788 N.A. N.A. 0.451 0.902 N.A. N.A. 2.839 5.678*** N.A. N.A. 0.545 1.090 N.A. N.A. 0.868 1.735* N.A. N.A. 1.030 2.060** N.A. N.A. 1.636 3.272*** N.A. N.A. -1.191 -2.382** N.A. N.A. Source: DataStream

* Significant at the 10% level (two-tailed). ** Significant at the 5% level (two-tailed). *** Significant at the 1% level (two-tailed).

AR = Abnormal Returns, t –Statistics = Student‟s t-statistics, AR values are calculated using the equation: AR = Rit -Rnormal where Rit is the actual return at day t and Rnormal is the normal return at day t. Rnormal is the result of a three factor model were respectively the AEX, Euronext and FTSE 100 were included as market indices. Estimation window = [-120, -20].

(31)

There are also noticeable abnormal returns for the category ethical concerns for the topic „breaches of certain labour standards‟. The market reacts very convincingly positive on negative news about this topic. A possible explanation for this could be that the market expects that low labour standards are cost-efficient. The news does not seem too interpreted in a socially engaged manner but rather from a low-cost perspective.

Cumulative abnormal return

Cumulative abnormal returns (CAR) are based on the abnormal return. The sum of the abnormal returns is taken for several event windows. This gives a better view of the development of the abnormal return through time. The CAR‟s are calculated for a

period of one, two, five and ten day‟s prior till after the event. The CAR‟s of

positive/negative news are presented in table 10. The CAR‟s of positive/negative

news does not give any significant results. The abnormal returns do seem to sustain during the various event windows.

Table 10

Cumulative abnormal return for event study after the impact of positive/negative news on financialperformance

Very Negative Negative Neutral Positive Very Positive

Event

Window CAR t-Statistics CAR t-Statistics CAR t-Statistics CAR t-Statistics CAR t-Statistics

[0] -0,002 -0,002 -0,236 -1,314 0,066 0,270 0,187 1,476 0,066 0,270

[-1,1] 0,007 0,004 -0,089 -0,285 0,357 0,850 -0,065 -0,294 0,357 0,850

[-2,2] 0,006 0,003 -0,256 -0,636 0,700 1,290 0,145 0,511 0,700 1,290

[-5,5] -0,004 -0,001 0,198 0,333 0,217 0,269 0,185 0,439 0,217 0,269

[-10,10] 0,002 0,0004 0,623 0,7574 0,927 0,8336 -0,222 -0,3809 0,927 0,8336

CAR = Cumulative Abnormal Returns, t –Statistics = Student‟s t-statistics, CAR values are the calculated as the sum of the AR‟s of the event window. AR = Rit -Rnormal where Rit is the actual return at day t and Rnormal is the normal return at day t. Rnormal is the result of a three factor model were respectively the AEX, Euronext and FTSE 100 were included as market indices. Estimation window = [-120, -20].

This is different for the CAR‟s calculated for the various CSR areas of concern which are presented in table 11-14. News on the topics of the category environment does

(32)

The category governance contains some interesting results. Neutral news on risk management has a significant positive CAR (1%) for the event window [0]. Neutral news on risk management also has significant CAR‟s (1%) for the event windows [-1,1], [-2,2] and [-10,10]. Also, positive news on risk management has a significant impact. This news gives a negative CAR (10%) for the event window [-1,1]. Negative news on the second topic of governance, risks, give significant negative CAR‟s (5% and 10%) for all the event windows except [-5,5].

The fourth category gives again significant results for the topic breaches of certain

labour standard. This topic gives positive CAR‟s for negative news. For the event

window 1, 1] and 10,10] at a significance level of 10% and for the event window [-5, 5] at a significance level of 5%.

Table 11

Cumulative abnormal returns for event study after the impact of media exposure about the CSR area ‘Environment’ on financialperformance

CAR = Cumulative Abnormal Returns, t –Statistics = Student‟s t-statistics, CAR values are the calculated as the sum of the AR‟s of the event window. AR = Rit -Rnormal where Rit is the actual return at day t and Rnormal is the normal return at day t. Rnormal is the result of a three factor model were respectively the AEX, Euronext and FTSE 100 were included as market indices. Estimation window = [-120, -20].

The category governance gives the most interesting results, there are significant outcomes for several event windows. Risk, and especially risk management show coherence with financial performance. This result also became clear after analyzing the AR. Investors seem to be sensitive for information about risks. However, as we mentioned before, this relation could also be due to sample size. Neutral news is could be news that is new to the market and therefore give the strongest reaction. Very positive or negative news for example would have reached the market already in another way.

Another interesting result is that „sourcing‟ and „breaches of certain labour standards give contradicting results. The latter gives positive CAR‟s for negative news, this is the other way around for sourcing which gives a positive CAR for positive news. This means that the market wants certain legislation on labour standards but does not

Policy Performance

Negative Neutral Positive Neutral Neutral Positive

Event

Window CAR t-Statistics N.A. CAR t-Statistics CAR t-Statistics N.A. CAR t-Statistics

[0] 0,110 0,246 N.A. 0,047 0,163 0,267 0,267 N.A. -0,043 -0,074

[-1,1] 0,287 0,371 N.A. 0,386 0,772 -0,099 -0,057 N.A. 0,305 0,305

[-2,2] 0,350 0,350 N.A. 0,846 1,310 -0,448 -0,200 N.A. 0,578 0,448

[-5,5] -0,097 -0,065 N.A. 0,347 0,362 -1,265 -0,381 N.A. 1,014 0,529

(33)

mind if the rules are sometimes broken. The rules give some security but breaking these rules may be cost efficient.

Table 12

Cumulative abnormal returns for event study after the impact of media exposure about the CSR area ‘Social’ on financialperformance

Sourcing

Negative Neutral Positive

Event

Window CAR t-Statistics N.A. CAR t-Statistics

[0] -0,305 -0,863 N.A. 1,187 1,679*

[-1,1] -0,235 -0,384 N.A. 1,447 1,182

[-2,2] 0,265 0,335 N.A. 2,458 1,554

[-5,5] 0,718 0,613 N.A. 2,849 1,215

[-10,10] 2,001 1,235 N.A. 1,473 0,455

CAR = Cumulative Abnormal Returns, t –Statistics = Student‟s t-statistics, CAR values are the calculated as the sum of the AR‟s of the event window. AR = Rit -Rnormal where Rit is the actual return at day t and Rnormal is the normal return at day t. Rnormal is the result of a three factor model were respectively the AEX, Euronext and FTSE 100 were included as market indices. Estimation window = [-120, -20].

Table 13

Cumulative abnormal returns for event study after the impact of media exposure about the CSR area ‘Governance’ on financialperformance

Stakeholders Human Rights

Negative Neutral Positive Negative Neutral Positive

Event

Window CAR t-Statistics CAR t-Statistics CAR t-Statistics N.A. N.A. CAR t-Statistics

[0] -0,671 -0,949 0,715 0,715 0,174 0,604 N.A. N.A. 0,384 1,438

[-1,1] -1,843 -1,505 -0,036 -0,021 0,199 0,398 N.A. N.A. 0,525 1,133

[-2,2] -1,333 -0,843 0,676 0,302 0,603 0,934 N.A. N.A. 0,458 0,766

[-5,5] -1,526 -0,651 3,938 1,187 0,643 0,671 N.A. N.A. 1,899 2,143**

[-10,10] -4,067 -1,255 8,895 1,941* -0,215 -0,162 N.A. N.A. 0,543 0,443

Risk management Risks

Negative Neutral Positive Negative Neutral Positive

Event

Window CAR t-Statistics CAR t-Statistics CAR t-Statistics CAR t-Statistics N.A. CAR t-Statistics

[0] -0,132 -0,457 3,448 3,448*** 0,154 0,846 -1,104 -1,912* N.A. -0,008 -0,024

[-1,1] -0,109 -0,217 5,919 3,417*** -0,561 -1,774 -2,201 -2,201** N.A. -0,373 -0,645

[-2,2] -0,571 -0,885 6,152 2,751*** -0,489 -1,199 -2,264 -1,754* N.A. 0,072 0,097

(34)

Table 14

Cumulative abnormal returns for event study after the impact of media exposure about the CSR area ‘ethical concerns’ on financialperformance

Military Genetic Engineering

Negative Neutral Positive Negative

Event

Window CAR t-Statistics N.A. N.A. CAR t-Statistics N.A. N.A.

[0] -0,113 -0,113 N.A. N.A. 0,781 0,781 N.A. N.A.

[-1,1] 0,486 0,281 N.A. N.A. 0,635 0,366 N.A. N.A.

[-2,2] 0,414 0,185 N.A. N.A. -0,148 -0,066 N.A. N.A.

[-5,5] -0,270 -0,081 N.A. N.A. 0,245 0,074 N.A. N.A.

[-10,10] 0,112 0,024 N.A. N.A. -1,512 -0,330 N.A. N.A.

Breaches of certain labour standards

Negative Neutral Positive

Event

Window CAR t-Statistics N.A. N.A.

[0] -0,161 -0,321 N.A. N.A.

[-1,1] 1,523 1,759* N.A. N.A.

[-2,2] 0,566 0,506 N.A. N.A.

[-5,5] 3,364 2,028** N.A. N.A.

[-10,10] 3,902 1,703* N.A. N.A.

CAR = Cumulative Abnormal Returns, t –Statistics = Student‟s t-statistics, CAR values are the calculated as the sum of the AR‟s of the event window. AR = Rit -Rnormal where Rit is the actual return at day t and Rnormal is the normal return at day t. Rnormal is the result of a three factor model were respectively the AEX, Euronext and FTSE 100 were included as market indices. Estimation window = [-120, -20].

(35)

Graphs 1-5 Positive Negative Graphs 6-15 Areas of CSR. Very Positive -3 -2 -1 0 1 2 3 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

C u m u la ti v e a b n o rm a l R e tu rn CAR 10 % Significance Positive -1,5 -1 -0,5 0 0,5 1 1,5 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

C u m u la ti v e A b n o rm a l re tu rn CAR 10 % Significance Neutral -3 -2 -1 0 1 2 3 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

C u m u la ti v e a b n o rm a l re tu rn CAR 10 % Significance Negative -2,5 -2 -1,5 -1 -0,5 0 0,5 1 1,5 2 2,5 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

C u m u la ti v e a b n o rm a l re tu rn CAR 10 % Significance Very Negative -15 -10 -5 0 5 10 15 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

(36)

Stakeholders -0,06 -0,04 -0,02 0 0,02 0,04 0,06 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

C u m u la ti v e A b n o rm a l R e tu rn CAR 10 % significance Human Rights -4 -3 -2 -1 0 1 2 3 4 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days Relative to m edia exposure

C u m u la ti v e A b n o rm a l R e tu rn CAR 10% Significance Sourcing -4 -3 -2 -1 0 1 2 3 4 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

C u m u la ti v e A b n o rm a l R e tu rn CAR 10 % Significance Risk management -0,06 -0,04 -0,02 0 0,02 0,04 0,06 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

C u m u la ti v e A b n o rm a l R e tu rn CAR 10 % Significance Risks -4 -3 -2 -1 0 1 2 3 4 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 1214 16 18 20

Days relative to m edia exposure

C u m u la ti v e A b n o rm a l R e tu rn CAR 10 % Significance Military -15 -10 -5 0 5 10 15 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

C u m u la ti v e A b n o rm a l R e tu rn CAR 10% Significance Genetic Engineering -15 -10 -5 0 5 10 15 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

C u m u la ti v e A b n o rm a l R e tu rn CAR 10 % Significance

Breaches of certain labor standard

-6 -4 -2 0 2 4 6 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Days relative to m edia exposure

(37)

6: Conclusion

In this thesis we examine the influence of media exposure about CSR on financial performance of the organization. Media exposure is categorized for that matter on the basis of their positive/negative tendency and on the CSR area that the news article addressed. Some of the categories indeed have a significant impact on the financial performance of an organization, other categories do not.

The results show that media exposure indeed has a significant relationship with corporate performance, at least for some categories. For the categorization positive/negative media exposure we concluded that neutral news has the strongest impact. Neutral media exposure probably contains new information for the market. More shocking news would already have reached the market in another way and not be perceived as new information in the news.

Positive news has a negative influence on financial performance. This indicates that the market does not appreciate social responsible behaviour of an organization. CSR can be associated with extra costs and such as a suboptimal way to operate.

The second classification that is made in the news articles was on the basis of the particular area of CSR that was addressed. This was done to analyze if some areas would have more influence that another. It appears that indeed some areas of CSR have more impact on CFP than others. The results show in fact that risk related topics have relatively the strongest influence. Media exposure on the topic Risk management with a neutral tendency show a highly significant positive impact in the period surrounding the event day. There were also significant negative abnormal returns on the event day after positive news. Neutral news is again supposed to contain the highest information ratio. The market reacts heavily since it gets new information. Positive news on this topic gives negative abnormal returns. The market interprets the positive news on the topic as bad news for an investor. This could be explained from the cost-efficiency point of view. The CSR of operating is perceived by the investor as not cost efficient and therefore sub-optimal.

(38)
(39)

References

Aguilera, R.V., Rupp, D.E., Williams, C.A., Ganapathi, J., 2007, Putting the S back in corporate social responsibility: A multilevel theory of social change in organizations,

Academy of management review, Vol32, Issue 3, p836-863

Brown, S.J., Warner, J.B., 1985, Using daily stock returns, the case of event studies,

Journal of financial economics, Vol. 14, p. 3-31

Davis, I., 2005, The biggest contract, Economist, Vol. 375, Issue 8428, p. 69-71

Fombrun, CJ, 1996, Reputation,: Realizing value from the corporate image, Harvard Business School Press, Boston.

Friedman, M., 1970, the social responsibility of business is to increase its profits,

New York Times magazine, 13 September: 33ff

Grant, C., 1991, Friedman Fallacies, Journal of business ethics, Vol. 10, Issue 12, p. 907-914

Hall, R., 1992, The strategic analysis of intangible resources, Strategic management

journal, vol 13, p135-144

Heal, G., 2005, Corporate social responsibility: An economic and financial

framework. Geneva Papers on Risk and Insurance – Issues and Practice, Vol. 30,

Issue 3, p387-409

Khotari, S.P., Warner, J.B., Econometrics of event studies, Working paper, p2-42

Kolstad, I., 2007, Why firms should not always maximize profits, Journal of business

(40)

McGuire, J.B., Sundgren, A., Schneeweis, T., 1988, Corporate social responsibility and firm financial performance, Academy of Management Journal, Vol. 31, Issue 4, p854-872

MacKinlay, A.C., 1997, Event studies in economics and finance, journal of economic

literature, Vol. 35, p 13-39

McWilliams, A, Siegel, D,, 2001, Corporate social responsibility: a theory of the firm perspective, Academy of Management Review, Vol. 26, Issue 1, p117-127

Meijer, M.M., Kleinnijenhuis, J., 2006, News and corporate reputation: empirical findings from the Netherlands, Public Relations Review, Vol. 32, Issue 4, p341-348

Orlitzky, M, Schmidt, F.L., Rynes S.L., 2003, Corporate Social and Financial Performance: A Meta-analysis. Organization Studies, Vol. 24 Issue 3, p403-441

Podnar, K., Golob, U., 2007, CRS expectations: the focus of corporate marketing,

Corporate communications: An international journal, vol.12, issue 4, p326-340

Ross, S.A.,Westerfield, R.W. and Jaffe, J.F., 2005, Corporate Finance, McGraw-Hill, Ney York, US.

Sauer, D.A., The impact of social-responsibility screens on investment performance: Evidence from the Domini 400 social index and Domini Equity Mutual Fund. 1997,

Review of Financial Economics, Vol. 6, Issue 2, p137, 13p

Shank, T., Manullang, D.K., Hill, R.P., 2005, Is it better to be naughty or nice?,

Journal of Investing, Vol. 14, Issue 3, p82-87

Schwaiger, M., 2006, Components and parameters of corporate reputation – an

empirical study, schmalenbach business review, Vol. 56, p46-71

Wood, D.J., 1991, Corporate social performance revisited, academy of management

Referenties

GERELATEERDE DOCUMENTEN

To be able to critically analyse this complex context and how the (poorly) implemented ESCP may or may not influence the agency of students to contribute to social cohesion, I will

Door gebruik te maken van de wereldwijde uitgebreide Thomson Reuters ASSET4 database bestaande uit een steekproef van 18.383 beursgenoteerde ondernemingen uit

In order to examine the intervening effects of exploitation efforts on the relationship between corporate social responsibility and a firm’s financial performance,

We also found that Supplier Orientation plays an important role, both in the short-term through a direct effect and interaction with local community orientation,

In line with earlier research I also find evidence for a positive correlation between female representation in a board and CSR pillar scores at a 5% level for Environmental

In order to test if the impact of environmental and social dimension on CFP varies across industries, a model containing all interaction effects between the dimensions and

As the results show mixed results with different environmental performance measurements, it implies that only some aspects (underlying variables) of the environmental

Foreign MNCs could better capitalise on their greatly successful/high impact CSR initiatives by making more information available on their corporate website as well as