The effect of non-compliance with the Dutch corporate governance code on shareholder activism

(1)

MSc Accountancy & Control, variant Control

Faculty of Economic and Business, University of Amsterdam

MASTER THESIS

The effect of non-compliance with the Dutch corporate governance code on

shareholder activism

Name: Karlijn Dirks

Student number: 10674837

Date: 17 August 2015

Word count: 10.219

Supervisor: prof. dr. Victor S. Maas MSc Accountancy & Control, variant Control Amsterdam Business School

(2)

Statement of Originality

This document is written by student Karlijn Dirks who declares to take full responsibility for the contents of the document.

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

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

(3)

Abstract: This thesis explores whether non-compliance with the Dutch corporate

governance code has an effect on shareholder activism. I analyze the annual reports and minutes of the annual general shareholders meeting from the 25 companies listed at the AEX in 2012. The annual reports give me the measures for compliance, COMP and NONCOMPEX. COMP is the percentage of compliance and NONCOMPEX is non-compliance with an explanation divided by total non-non-compliance. The minutes of the annual general shareholders meeting gives me the measures for shareholder activism, TOTAL, ATTENDANCE, ANNUAL and END. Using a multiple regression analyses I try to establish a relationship between compliance and shareholder activism. The results of my analyses do not support my research question that there is a relationship between non-compliance and shareholder activism. These results warrant additional research with a bigger sample size.

(4)

1. Introduction

In this thesis I examine whether non-compliance with the Dutch corporate governance code (CGC) has an effect on shareholder activism. In recent years many different companies have been in the news because of the bonuses they paid to their top management. The public cannot rhyme these bonuses with failing companies and the economic crisis. The big accounting scandals (Ahold, Enron) contribute to this feeling of unhappiness and distrust that the public has. To restore the trust of shareholders the CGC was implemented. “The corporate governance code is the system of management and supervision” (Rijksoverheid, n.d.). Unfortunately research has shown that a lot of Dutch companies do not comply with the entire corporate governance code (Hooghiemstra and Van Ees, 2011). The companies that do not fully comply with the CGC use the comply-or-explain principle in the CGC to comply-or-explain away their non-compliance. The comply-or-comply-or-explain principle entails that a company that does not fully comply with the CGC may give an explanation for this non-compliance and than that is also seen as a form of compliance. Even now, many years after implementation of the CGC the papers are still full with stories of failing companies that still pay their top managers big bonuses. Just last year an article appeared in the newspaper where politicians where outraged by the amount of

compensation managers were receiving (CDA en PvdA boos over bonussen, 2013). In this article they stated that the financial sector is doing everything they can to not win back the trust of citizens. This leads me to believe that the implementation of the CGC did not restore the trust of the shareholders. Shareholders that do not trust the company that they invest in will make their concerns known to the company. They will do this by voicing their concerns at the annual general shareholders meeting and try to inspire change in that way. I call this shareholder activism. I think that there is a relationship between compliance with the Dutch corporate governance code and shareholder activism. More formally I have developed the following research question: Does compliance with the Dutch corporate governance code have an effect on shareholder activism?

To the best of my knowledge no scholarly research has been done to answer this research question. In my opinion this is a gap in the current management accounting literature. The provisions in the CGC are considered to be “best practice”. Executive pay is one of the topics which is being addressed in the Dutch corporate governance code. Research shows that if executive pay goes up shareholder activism increases (Ertimur et al., 2011). So I wonder why that would not be the case with all the other provisions in the CGC. Intuitively, the relationship between complying with the CGC and shareholder

(6)

activism is a negative one. So if compliance rates go up shareholder activism will

decrease. Some other scholars have found compliance with the CGC an interesting topic and have conducted research into the topic. Studies have related compliance with the CGC to share price (MacNeil & Li, 2006), ROA and ROE (Gavrea et al., 2013) but none of them have related compliance to shareholder activism. As far as I can tell the relationship between compliance with the CGC and shareholder activism has never been established by empirical research. The answer to this research question can not only further the current management accounting literature but may also be a contribution to practice. The negative attention that shareholder activism brings and the associated costs may be reduced by the outcomes of this research paper. Putting all this together makes answering this research question an important task.

As I said in the previous paragraph, this study is original and to my knowledge no other study has examined the same thing. This makes the literature on this exact subject non-existing. So to examine the literature I looked at compliance with the corporate governance code and shareholder activism separately. The literature relating to

compliance with the corporate governance code has found several relationships. These relationships are between compliance and firm size (Ettredge et al, 2011), compliance and share price (MacNeil & Li, 2006), compliance and financial performance (Tariq & Abbas, 2013), compliance and legitimacy (Seidl, et al., 2012) and compliance and organizational performance (Gavrea et al., 2013). All these positive relationships state a clear

relationships between compliance and different measures of performance. On the other hand the literature considering shareholder activism has found relationships between shareholder activism and firm size (Rehbein et al., 2004), CEO pay (Ertimur et al, 2011), costs (Gantchev, 2013) and the effect it has on companies (Karpoff, 2001). Looking at the previous literature we see a positive relationship between firm size and compliance and between firm size and shareholder activity. It is therefore logical that there is also a

relationship between compliance and shareholder activity. But as to my knowledge the two have never been related to each other in an empirical study.

To answer the research question I analyzed the annual reports and the minutes of the annual general shareholders meeting (AGM) of the 25 companies listed at the AEX in 2012. This resulted in 99 records of deviation from the CGC. I first studied the compliance rates. I divided compliance into two variables. The first is outright compliance with the provisions in the CGC (COMP) and the second is when a firm does not comply with the provisions in the CGC but provides an explanation for this non-compliance

(7)

(NONCOMPEX). After this I measured shareholder activism at the AGM. I used four different measures of shareholder activism namely. The first is attendance rates at the AGM (ATTENDANCE), second is questions asked during the presentation of the annual report (ANNUAL), third is questions asked at the end of the AGM (END) and finally I

combine all these variables (TOTAL). I tried to establish a relationship between the level of compliance and my different measures of shareholder activism. To find this relationship I formulated two hypotheses. H1: Companies that comply with the Dutch corporate

governance code have less shareholder activity than companies that do not comply and H2: Companies that do not comply with the Dutch corporate governance code but have explanations for this non-compliance have less shareholder activity than companies that do not comply and have no explanation for this non-compliance. To test these hypotheses I used a multiple regression analyses.

After running my multiple regression analyses I found no significant results. I did find significant correlations between dependent variable ANNUAL and independent variable NONCOMPEX and a measure of compliance called non-compliant with an explanation. Another significant correlation exists between dependent variable END and the measure of compliance, non-compliant. These correlations can not be found in the analyses of my hypotheses. With my multiple regression analyses I found no significant evidence to support H1 nor H2. So I can not say that there is a relationship between compliance with the Dutch corporate governance code and shareholder activism.

My paper contributes to the literature in several ways. My results do not support the intuitive relationship between compliance with the corporate governance code and

shareholder activism. However my results suggest that there are relationship between certain aspects of shareholder activism and (non-)compliance with the CGC. The positive correlation between the number of provisions that a firm does not comply with and the dependent variable END demonstrates that shareholders do speak up if the firm has relatively high compliance. The dependent variable ANNUAL shows a positive parametric Spearman correlation with explained compliance and with explained non-compliance in relationship to total non-non-compliance. These positive relationships gives reason to think that shareholders do pay attention to the explanations given for non-compliance and want to ask questions about this during the presentation of the annual report. So even though my overall hypotheses are not supported, this paper does

contribute to the current debate in linking compliance to the corporate governance code to shareholder activism.

(8)

The remainder of the thesis is structured in 4 sections. In chapter two the history of the CGC and the concept of the comply-or-explain principle is explained in more detail and a definition of shareholder activism is given. After this legitimacy theory is introduced. These are all the theoretical constructs and their relationship will be established after this. All this will lead to the development of my hypothesis. In chapter three the methodology is explained. After information about sample selection and data collection I define the

measures for my analyses. Chapter three ends with the model with which I will test my hypotheses. In chapter four I present my findings. Finally in chapter five I discuss the results and their contribution to existing literature.

2. Theoretical background

In this chapter I explain the theoretical background related to my research question. I start with some history about the Dutch corporate governance code and the comply-or-explain principle that it is based on. After that I give my definition of shareholder activism

supported by relevant literature. Next I explain legitimacy theory and its relationship with this thesis. I then link all this theoretical concepts together and establish their relationship in this paper. Finally I develop my hypotheses based on the literature in this chapter.

2.1 Comply-or-explain in the corporate governance code

In 2004 the Dutch corporate governance code (CGC) in the form that we know now was implemented. After accounting scandals such as Enron and Ahold the corporate

governance code had to restore the trust of stockholders.

Corporate governance codes are widespread around the globe since the

publication of the Cadbury report in England in 1992. In the Netherlands the discussion about corporate governance code started with the Peters Committee in 1997. This committee issued the first Dutch corporate governance code consisting of forty

recommendations (Akkermans et al, 2007). The compliance with these recommendations was evaluated by two studies (De Jong and Roosenboom, 2002; De Jong et al, 2005) and both studies found that compliance was low and that the shareholders did not value

corporate governance changes. In the early years of the new millennium a series of accounting scandals came to light Enron (2001), Worldcom (2002) and Ahold (2003). Due to these scandals shareholders now wanted more guidance for corporate governance. In 2003 a new corporate governance committee was formed, the Tabaksblat Committee. The task of the committee was to formulate a new corporate governance code taking into

(9)

account the recommendations of the Peters committee and international developments (Akkermans et al, 2007). In December 2003 the final version of the Tabaksblat Code was published. In 2008 the Tabaksblat Code was revised by the Frijns committee. The revised code was applicable as of the 1st of January 2009 (http://

commissiecorporategovernance.nl/corporate-governance-code). This is the corporate governance code that is in effect today.

The Code is divided into five chapters which all have their own principles and best practice provisions. The five chapters are: compliance with and enforcement of the code, the management board, the supervisory board, the shareholders and the general meeting of shareholders, the audit of the financial reporting and e the position of the internal audit function and the external auditor. The most recent CGC is based on the comply-or-explain principle (http://commissiecorporategovernance.nl/corporate-governance-code). This means that companies are required to follow the rules set in the CGC or explain their non-compliance. The comply-or-explain principle has to make sure the CGC can be interpreted flexible and to avoid a “one size fits all” approach (Seidl et al, 2012).

An interesting characteristic of the comply-or-explain principle is that even if a company does not comply, if it gives a proper explanation this non-compliance is also a form of compliance (Seidl et al, 2012). In theory shareholders check to see in how far a firm is compliant with the CGC and are firms obligated to disclose non-compliance (Seidl, 2007). Research has shown that a lot of companies do not fully comply with the corporate governance code (Hooghiemstra and Van Ees, 2011). In this study I will analyze the

annual reports of companies listed at the Amsterdam Exchange IndeX (AEX) from the year 2012. This analyses of the annual reports of the companies listed at the (AEX) for

compliance with the CGC can have varying results. In this study a company can have three different levels of compliance with a provision in the corporate governance code, a firm can be compliant, a firm can be non-compliant but with an explanation or a firm can be non-compliant.

2.2 Shareholder activism

In this section I define the concept of shareholder activism. Many different definitions of shareholder activism exist. Shareholder activism is a way in which shareholders try to influence a corporation’s behavior by exercising their rights as owners (Karpoff, 2001). Although shareholders do not run the company, they do own part of the company and there are ways for them to influence the supervisory board and the board of directors.

(10)

Gillan & Starks (2007) state that shareholder activism can be a continuum of possible responses to corporate governance decisions.

Research on shareholder activism is a relatively new field of research. Two conflicting perspectives on shareholder activism exist. On the one hand there is the financial shareholder activism perspective that believes in shareholder primacy. This perspective states that the interest of the shareholders go above all other stakeholders (Thomas & Cotter, 2007). On the other hand there is the social shareholder activism perspective. This perspective focusses on shareholder activists raising social and environmental issues at the annual general meeting (AGM)(Rehbein et al., 2004).

Ryan & Schneider (2002) argue that there are three types of antecedents of

shareholder activism: firm, activist and environmental. Regarding the first antecedent, firm, research shows that shareholder activists mainly focus on large companies (Ertimur et al, 2011). A variety of explanations can be given for this. From the financial shareholder activism perspective an explanation is that activist could create more value by targeting large firms (Strickland et al., 1996). From a social shareholder activism perspective it could be argued that large firms have greater visibility and are therefor more attractive (Rehbein et al., 2004). For the antecedent of activist it needs to be said that not all shareholders become activists. There are different factors that influence if a shareholder will become an activist. Goranova & Ryan (2014) name money as an important factor for a shareholder to become an activist. Putting together a campaign to inspire real change in a firm costs a lot of money. Seeing as how I measure shareholder activism at the AGM this is not

applicable. The other two factors that Goranova & Ryan (2014) suggest are self-selection and moral legitimacy, these factors are applicable to my study. Self-selection refers to the confidence of a shareholder that he or she can get the support from other shareholders. If a shareholder knows that his activism has no influence what so ever he will be likely to give up before he even started. Moral legitimacy comes from within the activist. A moral compass may be the reason for an activist to fight for a seemingly lost cause. The

environmental antecedent states that shareholder activism not only focusses on individual firms but on the environment as a whole. As stated above, the corporate governance code came into effect after scandals as Enron and Ahold. The CGC had to restore trust in shareholders. Over the last few years a lot of firms have been in the news as a result of huge bonuses awarded to their top managers. This is in contradiction with the provisions of the CGC. This created a lot of negative responses from shareholders. This study looks if that non-compliance with the CGC translates into higher shareholder activism.

(11)

Previous studies that looked into shareholder activism focused mainly on how institutional investors put pressure on the board of a company by submitting proposals for het annual meeting (Black, 2007; Ertimur et al, 2011). I too look at this form of shareholder activism but I also want to take the small shareholder into account, this is why I look at the minutes of the annual meetings. At the AGM all shareholders get a chance to speak up and voice their concerns. There is a very low threshold to ask questions and address the board of directors.

Using the minutes of the AGM of the companies listed at the AEX in 2012 I define shareholder activism using four variables, the attendance of the AGM (ATTENDANCE), the questions asked by stockholders during the presentation of the annual report

(ANNUAL), questions asked at the end of the meeting (END) and a total of the other three variables (TOTAL).

2.3 Legitimacy theory

I now give an explanation of legitimacy theory and link it to this study. Legitimacy has been defined by Lindblom (1993, p.2) as:

“a condition or status which exists when an entity’s value system is congruent with the value system of the larger social system of which the entity is part. When a disparity, actual or potential, exists between the two value systems, there is a threat to the entity’s legitimacy.”

It is important to note that legitimacy is not about the actual conduct of an

organization but about the perceived conduct of an organization (Bebbington et al, 2014). If an organization is perceived by society as complying with expectations of the society it is said to fulfill its social contract. This social contract is a theoretical construct and therefor different people can have different perceptions about how society expects the organization to behave. The social contract is made up of several terms some of these terms are

explicit others are implicit.

If an organization does not live up to the expectations of society there is a

legitimacy gap, this is stated in the citation above as: “When a disparity, actual or potential, exists between the two value systems, there is a threat to the entity’s

legitimacy.” (Lindblom, 1993, p.2) This gap (or disparity) may start to exist because of changing expectations of society or because new information is made public, A company must close the legitimacy gap in order to maintain their legitimacy. Societies have a lot of power to influence organizations. If the legitimacy gap is not closed stakeholders can effectively shut an organization down (Bebbington et al, 2014).

(12)

Hooghiemstra & Van Ees (2011) state that the CGC does not give any legitimate arguments for deviation. Deviations are accepted under the comply-or-explain principle but the question is if shareholders will accept the deviations. In order for shareholders to

accept deviations a proper explanation for non-compliance must be given. As mentioned before a company can have three different levels of compliance with provision in the CGC, a firm can be compliant, a firm can be non-compliant but with an explanation or a firm can be non-compliant. The two forms of non-compliance can be divided into a legitimate form and a non-legitimate form. Non-compliance with an explanation will be considered to be legitimate and non-compliance will be considered non-legitimate. In this study I will be looking at two different things. I will look whether shareholder activity is lower at firms that comply with the Dutch corporate governance code in comparison with firms that do not comply and I will look if shareholder activity is lower in a firm that has legitimate non-compliance in comparison to firms that has non-legitimate non-non-compliance.

2.4 Relationship

In this section I link all the theoretical concepts explained earlier in this chapter together and establish a relationship between these concepts. As I said in section 2.2 shareholder activism is a way in which shareholders try to influence a corporation’s behavior by exercising their rights as owners (Karpoff, 2001). If shareholders try to influence a firms behavior they are obviously not completely satisfied with the way this firm is being run at the moment. In theory a firm that is not fully complying with the CGC this would have poorer performance. Seeing as how shareholders want to optimize their profit (Stout, 2002) poorer performance will lead to unsatisfied shareholders. And in my study I want to see if this these unsatisfied shareholders will be more vocal in expressing their concerns.

The CGC states that a firm that does not comply with a provision can still be compliant if there is an explanation for this non-compliance therefor there is such a thing as legitimate non-compliance. I am hypothesizing that this legitimate non-compliance leads to less dissatisfied shareholders. In my study I will be looking if there is a significant difference in shareholder activity in companies that have legitimate non-compliance as opposed to firms that have non-legitimate non-compliance.

2.5 Hypotheses

Based on all the theoretical concepts explained earlier in this chapter I now develop my hypotheses. Previous studies have looked at shareholder activism differently than I do. Ertimur et al. (2011) look at vote-no campaigns and proposals submitted, Gillan & Stark

(13)

(2007) only look at proposals submitted. And all of these studies only focus on institutional investors. Only Black (1998) is looking at the possibility of jawboning but he too is only focusing on institutional investors. Furthermore his paper is from before the crisis so things may have changed.

Several studies have examined the relationship between compliance with the corporate governance code and firm performance (Tariq & Abbas, 2013; Gavrea et al, 2013). Firms that have a high compliance rate are valued higher by investors than firms that do not comply or comply less according to Klapper and Love (2004). These studies show that a higher compliance rate leads to better firm performance and higher valuation by investors. Firm performance can relate to the financial side of things but can also be seen as the social contract which is explained in the legitimacy theory paragraph (2.3). According to this theory firms that have a legitimacy gap must close this gap in order to maintain there legitimacy. If the legitimacy gap is not closed stakeholders can effectively shut an organization down (Bebbington et al, 2014). The stakeholders include the

shareholders. A way for shareholders to show that, in their opinion, there is a legitimacy gap is shareholder activism. Taking all this into account I think that compliance with the CGC leads to less shareholder activism. Formally put, I hypothesize the following:

H1: Companies that comply with the Dutch corporate governance code have less shareholder activity than companies that do not comply.

This first hypothesis will tell us if outright compliance leads to less shareholder activism. But in addition to this outright compliance there is a second form of compliance. The CGC has the comply-or-explain principle for a reason, it makes sure that the CGC avoids a “one size fits all” approach (Seidl et al, 2012). This means that non-compliance with an explanation for this compliance is still seen as compliance (legitimate non-compliance). This brings me back to the fact that higher compliance rates will lead to better firm performance which in turn will lead to less shareholder activity. Non-compliance with an explanation can be seen as compliance which leads me to my second hypothesis:

H2: Companies that do not comply with the Dutch corporate governance code but have explanations for this non-compliance have less shareholder activity than companies that do not comply and have no explanation for this non-compliance.

(14)

In this chapter I have explained the theoretical constructs that are relevant to this thesis. I have linked all these constructs together and with those relationships in mind I develop the hypotheses to answer my research question. In the next chapter I describe my methodology for answering these hypotheses. In the chapter after that I show the results from my analyses. In the final chapter I definitively show whether or not my hypotheses are supported.

3. Methodology

In this chapter I describe the methodology that I used to answer my research question and my hypotheses. First I describe how and why I chose my sample and how I collected all my data. Second, I a definition of my measures for compliance with the Dutch corporate governance code. Third, I give this same clear and concise definition but now for my measure of shareholder activity. At the end of this chapter I show my model with which I will answer my research question and hypotheses.

3.1 Sample selection and data collection

To study the effect of non-compliance with the Dutch corporate governance code on shareholder activism, I analyze the annual reports and annual general shareholders meeting of the 25 largest companies in the Netherlands. The annual report is the

document in which they are expected to declare compliance and justify deviations from the CGC. For each of the companies listed at the AEX I will read the annual report and count how many instances of compliance a firm has. I divide this compliance into non-compliance with an explanation and non-non-compliance without an explanation. This way I measure compliance. With the annual general shareholders meeting I measure

shareholder activism. By reading the minutes of each of the AGMs of the companies listed at the AEX I determine shareholder activism by counting the three different measures I use for shareholder activity. Putting all these measures together gives me my total shareholder activism. My sample consists of the companies listed at the AEX at the beginning of 2012. My data is based on the year 2012 because for this year all the necessary reports and minutes are available. I have excluded Shell because the minutes of the annual general meeting for 2012 were not available. I consciously limited my sample size due to the time consuming nature of collecting this data. The companies listed at the AEX are the 25 largest companies in the Netherlands. They have to provide a lot of information to a lot of different stakeholders and they do this through their company website. This information is

(15)

available for everyone interested. I have chosen the AEX because of this easy access to information. The companies at the AEX are a mixture of different industries. I have chosen companies in the Netherlands because the minutes of the AGM are not always translated into English but are usually only available in the language of the country where the

headquarter of a company is. Seeing as how I only speak Dutch and English I chose to focus on Dutch companies.

The Dutch corporate governance code consist of 121 best practice provisions divided into five different chapters. The fourth chapter is about the shareholders and the general meeting of shareholders. Seeing as how the dependent variables are about the annual general meeting of shareholders and the participation of these shareholders I found it conflicting to test compliance of this chapter. Chapter four is about the

shareholders and not so much about the corporate governance of a company, this is why I excluded chapter four from my analyses. This leaves 90 best practice provisions.

I found my data on the websites of the different companies. The data on compliance with the CGC I collected by looking at the annual reports of al the AEX companies. The data about shareholder activism is defined by four different measures. These measures are: the attendance at the AGM, the questions asked by stockholders during the

presentation of the annual report, questions asked at the end of the meeting and

proposals that are put on the agenda of the meeting. I have collected this data by studying the AGMs of all 24 companies.

3.2 Measurement of compliance

In this section I give a definition of my measure of compliance. To measure compliance I analyze the annual reports of the 24 companies in my sample and look if they comply with the 90 best practice provisions of the Dutch CGC that I test for. For each best practice provision there are four possible outcomes: N/A, compliant, non-compliant with

explanation, non-compliant. N/A means that this provision is not applicable for this

particular firm. Compliant means that the firm does live up to the standard set in the CGC. Non-compliant with explanation means that a firm at first glance does not (entirely) comply with the CGC but they give an explanation for this non-compliance. Finally there are those firms that do not comply with a provision of the CGC and also do not give an explanation for the reason of non-compliance, this means they are non-compliant. I gave each of these outcomes a score. These scores can be found in table 1.

(16)

By choosing these measures I create a divide between compliant and both forms of non-compliant but also between non-non-compliant with explanation and non-non-compliant without explanation. This is an important difference because non-compliant with an explanation can also been seen as a form of compliance.

After analyzing all the annual reports I looked at the percentage of compliance per company. I calculated this by dividing the number of best practice provisions that a firm is compliant with by the total number of best practice provisions minus the provisions that are not applicable. The outcome of this is my first independent variable named COMP.

The other variable consists of number of best practice provisions that a company does not comply with but does provide an explanation for divided by total non-compliance. This variable is called NONCOMPEX. This variable gives the percentage of legitimate non-compliance versus non-legitimate non-non-compliance. I expect the variables COMP and NONCOMPEX to be negatively associated with the different measures for shareholder activity.

3.3 Measurement of shareholder activity

This section defines my measures of shareholder activity. To measure shareholder activity I analyzed the annual general shareholders meeting of the 24 companies and looked at several different measures. I looked at the attendance at the AGM, the questions asked by stockholders during the presentation of the annual report, questions asked at the end of the meeting and proposals that are put on the agenda of the meeting. After analyzing my data I found that no proposals were put on the agenda for any of the companies so I eliminated that variable. In order to make the remaining variables comparable I created Z-scores for those variables. I named these Z-Z-scores ATTENDANCE for attendance at the AGM, ANNUAL for questions asked by stockholders during the presentation of the annual report and END for questions asked at the end of the AGM. To look at total shareholder

Table 1. Compliance scores

Compliance scores N/A 0 Compliant 1 Non-compliant with explanation 2 Non-compliant 3

(17)

activity I created a variable which combines these three variables and named it TOTAL. I chose these measures to include the smallest form of shareholder activity.

Others studies looked at shareholder activism mainly from an institutional

shareholders’ point of view (Ertimur et al, 2011; Ryan & Schneider, 2002) but I also want to take the individual shareholder into account. I selected the AGM as my measure for

shareholder activity because it is easily accessible for all shareholders. I created the four measures because in my opinion their is a difference between the four selected forms of shareholder activism. For example, showing up at the AGM might be a sign that that shareholder is dissatisfied but it may also be that this shareholder is just interested in the well being of the company. In my opinion both of these scenarios signal that the

shareholder wants more information which is the very basis of shareholder activism. I test for all of the measures individually as well as all four combined, this is because the four measures might cancel each other out. I expect that the different variables for shareholder activity are negatively associated with the variables that measure compliance.

3.4 Model

In this section I give the models that I use to answer my research question and my hypotheses. To answer my research question and my hypotheses I have developed models that will answer a specific part of my hypotheses and all together will help me formulate an answer to my research question.

To determine whether the effect of non-compliance with the Dutch corporate governance code leads to more shareholder activism as a whole I use the following model:

(1) TOTAL = b0 + b1COMP + b2NONCOMPEX + Ɛ

To determine whether the effect of non-compliance with the Dutch corporate governance code leads to a higher attendance rate at the annual general shareholders meeting I use the following model:

(2) ATTENDENCE = b0 + b1COMP + b2NONCOMPEX + Ɛ

To determine whether the effect of non-compliance with the Dutch corporate governance code leads to more questions during the presentation of the annual report I use the following model:

(18)

To determine whether the effect of non-compliance with the Dutch corporate governance code leads to more questions at the end of the annual general shareholders meeting I use the following model:

(4) END = b0 + b1COMP + b2NONCOMPEX + Ɛ

TOTAL Measure of total shareholder activism measured as the Z-scores for ANNUAL + END + ATTENDANCE summed up ANNUAL Z-score of questions asked at the presentation of the annual

report

END Z-score of questions asked at the end of the annual shareholders meeting

ATTENDANCE Z-score of the attendance at the annual shareholders meeting

COMP Percentage of compliance

NONCOMPEX Non-compliance with an explanation divided by

non-compliance with an explanation plus non-non-compliance without an explanation

In this chapter I have explained my sample selection and how I got my data. After this I defined my independent and dependent variables. Finally I developed the models that I will need to answer my hypotheses. The next chapter shows the results that I found after my analyses.

4. Analysis and findings

In this chapter of my thesis the results of my analyses will be shown. I start this chapter with some descriptive statistics about the research that I have done. In the next section the answer to hypothesis 1 can be found along with all the data to support these findings. The same can be found in section 4.3 but than considering hypothesis 2.

4.1 Descriptive statistics

In total the 24 firms report 99 deviations from the 90 provisions of the CGC that I test for. Five firms are completely compliant with all the provisions of the CGC and one firm did not comply with 21 provisions of the CGC. On average a firm does not comply with four

(19)

In panel A of Table 2 the descriptive statistics of the independent variables are shown. The independent variables COMP and NONCOMPEX are calculated as stated in paragraph 3.2. The independent variables compliant, non-compliant with explanation and non-compliant are the original values as I measured them. The 24 companies in my

sample are compliant with 95,66% of all the best practice provisions in the CGC. If they do not comply with a provision this is explained 68,37% of the time.

Table 2. Descriptive statistics

Panel A: Descriptive statistics (independent variables)

Independent variable Minimum Maximum Mean Std. Dev.

COMP 0,744 1,000 0,9568 0,0641 NONCOMPEX 0,000 1,000 0,6837 0,4638 compliant 64 88 82,58 5,540 non-compliant 0 9 2,33 2,745 with explanation non-compliant 0 18 1,42 4,442

Panel B: Descriptive statistics (dependent variables)

Dependent variable Minimum Maximum Mean Std. Dev.

TOTAL -1,8314 2,9216 0,0670 1,3293 ATTENDANCE -1,8484 2,0879 0,0000 1,0000 ANNUAL -1,0176 4,2564 0,4240 0,9983 END -0,5899 4,2255 0,0246 1,0138 attendance 35% 98% 64,58% 16,005 annual 0 153 30,75 28,9606 end 0 16 2,0417 3,3683

In panel B of Table 2 the descriptive statistics of the dependent variables are shown. The dependable variables in capital letters represent the z-scores as explained in paragraph 3.3. The dependable variables in small letters are the original scores as I measured them. Looking at dependent variable attendance a big difference in attendance rates at AGMs can be seen. With an average of 64,58% and a standard deviation of 16,005 the

attendances fluctuate tremendously. The same is true for dependent variable annual, the maximum amount of questions asked during a presentation of the annual report is 153 as

(20)

apposed to a minimum of 0 which means that none of the shareholders present at that particular presentation had a question during the presentation of the annual report.

In Table 3 the Pearson correlations are shown above the diagonal. Because of the small sample size (24) I also looked at the non-parametric Spearman correlations which can be found below the diagonal.

p-value in brackets

* Correlation is significant at the 0,05 level (2-tailed) ** Correlation is significant at the 0,01 level (2-tailed)

Pearson correlations appear above the diagonal, non-parametric Spearman correlations appear below the diagonal

Before the I test my hypotheses I want test for multicollinearity. Multicollinearity occurs when two or more predictors are highly correlated. After testing for multicollinearity I found a tolerance of 0,991 and a VIF of 1,009, this indicates that there is no

multicollinearity.

To test my hypotheses I look at the coefficients B COMP and B NONCOMPEX and not at the standardized coefficients Beta COMP and Beta NONCOMPEX. This because I used z-scores in my regression which are already a standardized measure.

Table 3. Correlations (1) (2) (3) (4) (5) (6) (7) (8) (9) COMP (1) 0,097 (0,654) 0,982** (0,000) -0,600** (0,002) -0,873** (0,000) -0,017 (0,939) 0,288 (0,172) 0,055 (0,799) -0,360 (0,084) NONCOMPEX (2) -0,341 (0,103) 0,153 (0,474) 0,423* (0,039) -0,378 (0,068) 0,009 (0,966) -0,042 (0,847) 0,323 (0,124) -0,264 (0,212) Compliant (3) 0,920** (0,000) -0,196 (0,359) -0,537** (0,007) -0,888** (0,000) -0,106 (0,622) 0,202 (0,343) 0,048 (0,824) -0,386 (0,063) Non-compliant with explanation (4) -0,745** (0,000) 0,701** (0,000) -0,628* (0,001) 0,134 (0,532) -0,272 (0,199) -0,279 (0,187) 0,229 (0,281) -0,307 (0,144) Non-compliant (5) -0,478* (0,018) -0,473* (0,019) -0,496* (0,014) -0,111 (0,605) 0,182 (0,394) -0,191 (0,372) -0,211 (0,321) 0,635** (0,001) TOTAL (6) 0,025 (0,909) 0,061 (0,776) -0,136 (-0,528 -0,136 (-0,528) -0,123 (0,526) 0,381 (0,066) 0,329 (0,116) 0,611** (0,002) ATTENDANCE (7) 0,269 (0,205) 0,017 (0,938) 0,137 (0,522) -0,196 (0,359) -0,264 (0,213) 0,468* (0,021) -0,425* (0,038) -0,068 (0,751) ANNUAL (8) -0,247 (0,244) 0,431* (0,036) -0,223 (0,294) 0,534** (0,007) -0,338 (0,106) 0,191 (0,372) -0,261 (0,219) -0,134 (0,534) END (9) 0,288 (0,173) -0,128 (0,552) 0,202 (0,345) -0,384 (0,064) -0,018 (0,933) 0,583** (0,003) 0,009 (0,967) -0,066 (0,759)

(21)

4.2 Hypothesis 1

In this section I try to find support for hypothesis 1: Companies that comply with the Dutch corporate governance code have less shareholder activity than companies that do not comply. Table 4 presents the results of the multiple regression analyses. To answer hypothesis 1 I looked at shareholder activity as a whole and I also looked at shareholder activity as three individual variables. I first looked at shareholder activity as a whole. With dependent variable TOTAL and independent variable COMP I find the b1 for model (1). In

table 3 we see that there is a very low, non-significant correlation between total

shareholder activity and the independent variable COMP. In the first row of table 4 we can find an AdR2 _{of -0,095, an F-ratio of 0,004 which is significant at p < 0,996. The B COMP}

for shareholder activity as a whole is -0,367 with a standard error of 4,546. This is the b1 in

model (1). These are non-significant results which means that there is no support for H1 if we look at shareholder activity as a whole as the measure of shareholder activity.

Next I test hypothesis 1 again but now I only look at the attendance rates at the AGM as a measure of shareholder activity. This will tell me whether non-compliance with the Dutch corporate governance code leads to a higher attendance rate at the annual general shareholders meeting. For this I use dependent variable ATTENDANCE and independent variable COMP, this way I will find the b1 for model (2). In table 3 there is a

very low, non-significant correlation between attendance at the AGM and the independent variable COMP. In the second row of table 4 we find an AdR2_{of 0,001, an F-ratio of 1,011}

which is significant at P < 0,381. The b1 for model (2) is the B COMP of ATTENDANCE

which is 4,599 with a standard error of 3,267. Based on this data there is no support for H1 if we only look at attendance at the AGM as the measure for shareholder activity.

I test hypothesis 1 again but now the measure for shareholder activity is questions asked at the presentation of the annual report. The dependent variable ANNUAL and independent variable COMP will find the b1 for model (3). Model (3) shows whether

non-compliance with the CGC leads to more questions asked at the presentation of the annual report. In table 3 there is a very low, negative, non-significant correlation between

questions asked at the presentation of the annual report and the independent variable COMP. In the third row of table 4 we find an AdR2_{of 0,019, an F-ratio of 1,228 which is}

significant at P < 0,313. The B COMP for ANNUAL is 0,373 and has a standard error of 3,231. This B COMP number is the b1 for model (3). Based on this data there is no support

for H1 looking at questions asked at the annual report as the measure of shareholder activity.

(22)

The last test for hypothesis 1 tests whether the effect of non-compliance with the Dutch corporate governance code leads to more questions at the end of the annual general shareholders meeting. For this I use dependent variable END and independent variable COMP. These variables will find the b1 for model (4). In table 3 no significant

correlation between questions asked at the end of the AGM and independent variable COMP is found. In the last row of table 4 we find an AdR2_{of 0,105, an F-ratio of 2,349}

which is significant at P < 0,120. The B COMP for END is -5,339 with a standard error of 0,433. Based on this data there is no support for H1 if questions asked at the end of the AGM is the measure for shareholder activity.

I have looked at shareholder activity as a whole and as the different individual variables, ATTENDANCE, ANNUAL and END but found no significant results for any of these measures. Taking all the previous into account I find no support for H1.

p-value in brackets

Std error in brackets and bold

4.3 Hypothesis 2

In this section I test hypothesis 2: Companies that do not comply with the Dutch corporate governance code but have explanations for this non-compliance have less shareholder activity than companies that do not comply and have no explanation for this

non-compliance. With H2 I used the same approach as for H1 the build up in this section is therefor the same as section in 4.2. I looked as shareholder activity as a whole and I also looked at shareholder activity as three individual variables. To start I look at shareholder activism as a whole. Dependent variable TOTAL and independent variable NONCOMPEX help me to find the b for model (1). In the correlation tables (tables 3 and 4) no significant

Table 4. Results

AdR2 F B COMP B

NON-COMPEX

Beta COMP Beta NON-COMPEX TOTAL -0,095 0,004 (0,996) -0,367 (4,546) 0,032 (0,628) -0,018 (0,936) 0,011 (0,960) ATTENDANCE 0,001 1,011 (0,381) 4,599 (3,267) -0,151 (0,451) 0,295 (0,174) -0,070 (0,741) ANNUAL 0,019 1,228 (0,313) 0,373 (3,231) 0,689 (0,446) 0,024 (0,909) 0,320 (0,137) END 0,105 2,349 (0,120) -5,339 (3,134) -0,507 (0,433) -0,338 (0,103) -0,232 (0,255)

(23)

correlations between the dependent variable TOTAL and the independent variable

NONCOMPEX was found. In table 4 the results of the multiple regression are shown. The AdR2_{and F-ratio for the variable TOTAL are the same as they were in section 4.2 (-0,095}

and 0,004 with a p-value of 0,996 respectively). The B NONCOMPEX for shareholder activity as a whole is 0,032 with a standard error of 0,628. This answers the b2 question for

model (1). These results are not significant and therefor I find no support for H2 if I look at shareholder activity as a whole.

Now I divide shareholder activism up into three variables and test H2 again for these three variables. To start I look if companies that have non-compliance without an explanation have higher attendance rates at the AGM than companies that have non-compliance with an explanation. Tables 3 and 4 show a very low, not significant, correlation between the dependent variable TOTAL and the independent variable

NONCOMPEX. The AdR2_{of 0,001 and the F-ratio of 1,011 with a p-value of 0,381 are the}

same as in section 4.2. Model (2) looks at the attendance rates at the AGM so I need b2

from model (2) to answer this question. With dependent variable ATTENDANCE and independent variable NONCOMPEX I find B NONCOMPEX of -0,151 with a standard error of 0,451. There is no support for H2 if we take attendance rates at the AGM as the

measure for shareholder activity based on this data.

The second variable for shareholder activity is questions asked at the presentation of the annual report. I test H2 again but now with this measure of shareholder activity. In table 3 above the diagonal we find a not significant correlation between questions at the presentation of the annual report and all measures of compliance. When we look at table 3 below the diagonal (non-parametric Spearman correlations) we find a significant

correlation between questions at the presentation of the annual report and the

independent variables non-compliant with an explanation and NONCOMPEX. For this variable the same applies as for the previous two, the AdR2_{and F-ratio are the same as in}

section 4.2 (0,019 and 1,228 with a p-value of 0,313 respectively). Model (3) shows whether non-compliance with the CGC leads to more questions asked at the presentation of the annual report. The dependent variable ANNUAL and independent variable

NONCOMPEX will find the b2 for this model. Table 4 reflects a B NONCOMPEX of 0,689

with a standard error of 0,446. This b2 for model (3) is not significant. Based on this data

there is no support for H2 taking questions asked at the annual report as the measure for shareholder activity.

(24)

The last variable for shareholder activity that we test H2 against is questions asked at the end of the annual general shareholders meeting. In table 3 above the diagonal a significant correlation between questions at the end of the AGM and the measure of compliance, non-compliant is found. Table 3 does not show this same correlation below the diagonal. Table 4 shows an AdR2_{of 0,105 and an F-ratio of 2,349 with a p-value of}

0,120 this is the same as in section 4.2. Model (4) tells whether the effect of

non-compliance with the Dutch corporate governance code leads to more questions at the end of the annual general shareholders meeting. To find the b2 of this model I use the

dependent variable END and the independent variable NONCOMPEX. Table 4 gives the B NONCOMPEX for variable END at -0,507 with a standard error of 0,433. Based on these numbers I find no support for H2 with questions asked at the end of the AGM as the measure of shareholder activity.

I have tested H2 with four different variables for shareholder activity. Although some correlations exist between these variables and some measures for compliance no

significant results were found when testing the hypothesis. Taking all this into account I find no support for H2.

In this chapter I have tested my hypotheses and found no support for them. In the next chapter I will draw my conclusions from this and try to think of a reason why my hypotheses were not supported.

5. Conclusion

In this final chapter of my thesis I recap on what I have done in my research and I

formulate a definite answer to my research question. In this thesis I have tried to answer the following research question: Does compliance with the Dutch corporate governance code have an effect on shareholder activity? To answer this research question I use the companies listed at the AEX in 2012. I look at their annual report to determine in how far the are compliant to the CGC on the one hand and on the other hand I analyze the

minutes of their AGM to measure shareholder activism. I then try to establish a relationship between these variables using a multiple regression analyses. As to my knowledge no prior research has been conducted into this exact subject. Research has shown that firm with higher compliance rates have better firm performance (Gavrea et al., 2013). Other research links shareholder activism to firm performance (Gillan & Starks, 2007; Karpoff, 2001). But making the direct link between compliance and shareholder activism has not been done yet. Taking previous literature into account it seems logical that compliance with

(25)

the Dutch corporate governance code would lead to lower shareholder activity. This negative relationship has lead me to hypothesis 1: Companies that comply with the Dutch corporate governance code have less shareholder activity than companies that do not comply. The Dutch CGC has the interesting characteristic that it enforces the comply-or-explain principle. This entails that not complying with a provision of the CGC can still be compliant if this compliance is properly explained. Taking this form of legitimate non-compliance into consideration leads to my second hypothesis: Companies that do not comply but have explanations for this non-compliance have less shareholder activity than companies that do not comply and have no explanation for this non-compliance.

To answer these hypotheses I have conducted a multiple regression analyses. I have tested the hypotheses several times using different measures for shareholder activity. I looked at shareholder activity as a whole (TOTAL) but also at the variables: attendance rates at the AGM (ATTENDANCE), questions asked at the presentation of the annual report (ANNUAL) and question asked at the end of the AGM (END). The

independent variable for hypothesis 1 is COMP for the percentage of compliance from the companies listed at the AEX and for hypothesis 2 I use NONCOMPEX as the independent variable, expressing the legitimate non-compliance divided by all non-compliance

(legitimate and non-legitimate). After running the regression I found no significant results for H1 nor H2. Looking at the correlations in table 3 there is a correlation between

dependent variable ANNUAL and between independent variable NONCOMPEX and the measure of compliance, non-compliant with an explanation. There is also a significant correlation between END and the measure of compliance, non-compliant. These correlations do not hold up with further analyses. So to answer my research question; I have found no evidence to support the assumption that compliance with the Dutch corporate governance code has an effect on shareholder activity.

As I have said in the previous section I have found no support for H1 nor H2. There are a variety of explanations for this. The first one which I consider to be true is that there simply is no relationship between compliance with the Dutch corporate governance code and shareholder activism. Seeing as how the relationship between compliance and firm performance (Gavrea et al., 2013) and between firm performance and shareholder activity (Gillan & Starks, 2007; Karpoff, 2001) has been established but I can not find a

relationship between compliance and shareholder activity it seems that firm performance is an intervening variable. Second, my sample size is very small. I looked only at the 25 companies listed at the AEX in 2012 and had to eliminate one because not all the

(26)

information was available. The 24 companies that I have analyzed are the 24 largest companies in the Netherlands. This brings me to my third explanation for why I have found no support for my hypotheses, sample bias. Research has linked shareholder activity to firm size (Rehbein et al., 2004) so when looking at the 24 largest companies in a country a certain degree of shareholder activism is to be expected.

As mentioned before, no prior research into this research question has been conducted, that in itself makes this research a valuable contribution to the academic literature. My literature review show that there is a relationship between compliance and firm performance and between firm performance and shareholders activity. Based on this the relationship between compliance and shareholder activism seems undeniable. My research has shown that this relationship does not exist unless the variable of firm performance is integrated.

Even though I found no support for my hypotheses some interesting things came up which could be a contribution to practice. The significant correlations that I have found between non-compliance and the dependent variable END shows that shareholders do speak up if the firm has relatively high non-compliance. The dependent variable ANNUAL shows a positive non-parametric Spearman correlation with explained non-compliance and with independent variable NONCOMPEX. These positive relationships give reason to think that shareholders do pay attention to the explanations given for non-compliance and want to ask questions about this during the presentation of the annual report. Executives from AEX firms should take this into consideration when preparing their presentation of the annual report.

This thesis has various limitations and there is room for improvement. The most important one is the limited sample that I have used. With a sample this small and firms that are relatively similar in terms of size, corporate governance structure and culture (they are all Dutch) a bias in the results will exist. The second limitation is in the measure of shareholder activism. I have found no study that uses the same measures for shareholder activism as I do. My measure of shareholder activism can be seen as controversial and other scholars may disagree that these are good measures for shareholder activism. The last limitation that I want to point out is my independent variable NONCOMPEX, this

variable is calculated by dividing the explained non-compliance by total non-compliance. In this measure no nuance is taken into account for the quality of the explanations.

My study has focussed on establishing a relationship between compliance with the Dutch corporate governance code and shareholder activism. Future studies can redo this

(27)

study but with a larger sample size. If a relationship between compliance and shareholder activism can be determined this would make a big contribution to the academic literature and would influence the decisions of managers and other professionals.

(28)

References

Aguilera, R., & Cuervo-Cazurra, A. (2004). Codes of good governance worldwide: What is the trigger? Organization Studies, 25, 415–444.

Akkermans, D., Van Ees, H., Hermes, N., Hooghiemstra, R., Van der Laan, G., Postma, T., et al. (2007). Corporate governance in the Netherlands: An overview of the

application of the Tabaksblat code in 2004. Corporate Governance: An International

Review, 15, 1106–1118.

Bebbington J., Unerman, J. & O’Dwyer B. (2014). Sustainability Accounting and Accountability. London. Routledge.

CDA en PvdA boos over bonnussen (2013, July). NOS news. Retrieved from http://nos.nl

Ertimur, Y., Ferri, F., Muslu, V. (2011). Shareholder Activism and CEO Pay. Review of

Financial Studies, 24(2), 535-592.

Ettredge, M., Johnstone, K., Stone, M. & Wang, Q. (2011). The effect of firm size,

corporate governance quality, and bad news on disclosure compliance. Review of

Accounting Studies, 16(4), 866-889.

Gavrea, C., Stegerean, R. & Marin, A. (2013). The relationship between the degree of compliance with the corporate governance code and organizational performance: A study on Romanian firms. Studia UBB Negotia, 1, 23-39.

Gantchev, N. (2013). The costs of shareholder activism: Evidence from a sequential decision model. Journal of Financial Economics, 107(3), 610-631.

Gillan, S., Starks, L.T. (2007). The Evolution of Shareholder Activism in the United States. Available at SSRN: http://ssrn.com/abstract=959670.

Goranova, M., Verstegen Ryan, L., (2014). Sharholder Activism: A Multidisciplinary Review. Journal of Management, 40(5), 1230-1268.

Hooghiemstra, R. & Van Ees, H. (2011). Uniformity as response to soft law: Evidence from compliance and non-compliance with the Dutch corporate governance code.

Regulation & Governance, 5(4), 480-498.

Jong, A. de & Roosenboom P.G.J. (2002). De praktijk sinds 1997. Corporate Governance

in Nederland 2002 - De stand van zaken. Amsterdam: NCGS.

Jong, A. de, DeJong, D.V., Mertens, G., Wasley C.E. (2005). The Role of Self-regulation in Corporate Governance: Evidence and Implications from the Netherlands. Journal of

(29)

Karpoff, J.M. (2001). The Impact of Shareholder Activism on Target Companies: A Survey of Empirical Findings. Available at SSRN: http://ssrn.com/abstract=885365.

Klapper, L.F. & Love, I. (2004). Corporate governance, investor protection, and

performance in emerging markets. Journal of Corporate Finance, 10(5), 703-728.

MacNeil, I. & Li, X. (2006). “Comply or Explain”: market discipline and non-compliance with the Combined Code. Corporate Governance: An International Review, 14(5),

486-496.

Monitoring Commissie Corporate Governance Code (2008). De Nederlandse corporate governance code.

https://docs.google.com/viewer?url=http://www.mccg.nl/download/?id%3D609

Rehbein, K., Waddock, S. & Graves, S.B. (2004). Understanding shareholder activism: Which corporations are targeted? Busisness & Society, 43(3), 239-267.

Rijksoverheid (n.d). Corporate Governance Code. Retrieved August 3 2015 from rijksoverheid.nl:

( http://www.rijksoverheid.nl/onderwerpen/corporate-governance/corporate-governance-code)

Ryan, L.V. & Schneider, M. (2002). The antecedents of institutional investor activism.

Academy of Management Review, 27(4), 554-573.

Seidl, D. (2007). Standard Setting and Following in Corporate Governance: An Observation Theoretical Study of the Effectiveness of Governance Codes.

Organization, 14(5), 705-727.

Seidl, D., Sanderson, P. & Roberts, J. (2012). Applying the ‘comply-or-explain’ principle: discursive legitimacy tactics with regard to codes of corporate governance. Journal

of Management & Governance, 17(3), 791-826.

Stout, L.A. (2002). Bad and not-so-bad arguments for shareholder primacy. Southern

California Law Review, 75(5), 1189-1209.

Strickland, D., Wiles, K.W. & Zenner, M. (1996). A requiem for the USA: Is small

shareholder monitoring effective? Journal of Financial Economics, 40(2), 319,338.

Tariq, Y.B. & Abbas, Z. (2013). Compliance and multidimensional firm performance: Evaluating the efficacy of rule-based code of corporate governance. Economic

(30)

Thomas, R.S. & Cotter, J.F. (2007). Shareholder proposals in the new millennium: Shareholder support, board response and market reaction. Journal of