Shareholder activism: the effect of shareholder proposals on target firms

Shareholder activism: the effect of shareholder

proposals on target firms

Master Thesis

Duco Rosier – S2522225

d.w.b.s.rosier@student.rug.nl

Supervisor: Prof. Dr. C.L.M. Hermes

Co-assessor: Dr. C. Laureti

MSc International Financial Management

Rijksuniversiteit Groningen, Faculty of Economics and Business

Date of submission: 08-01-2020

2

Abstract

The relationship between shareholder activism and firm performance is a much debated and researched subject in the field of corporate governance. However, to date, the empirical results remain inconclusive and show positive, negative, and even insignificant relationships, which may be explained by the heterogeneity of activists’ demands as well as the divergent

streams of theories used throughout literature. The key objective of this thesis is to contribute to this debate by using a multifaceted measure of shareholder activism in combination with a two-way distinction of firm performance. This is done by exploring the effects of filed executive compensation-related and corporate governance-related shareholder proposals on firm financial performance and by exploring the effects of filed social/environmental-related shareholder proposals on firm social performance of the 250 largest U.S. listed firms. Based on agency, institutional, managerial entrenchment, and business relationship theory, it is argued that the filing of shareholder proposals will lead to an increased financial and social performance at target firms. Additionally, a moderating role for foreign institutional ownership is assessed in this context. By conducting several multiple regression analyses it is found that all three categories of shareholder proposals have no or even negative effects on their respective performance dimension, which is contrary to this thesis’ predictions. Also,

against expectations, foreign institutional ownership does not moderate the relationship between the three categories of shareholder proposals and firm performance. The findings of this thesis thereby add to the shareholder activism-performance literature and have several theoretical and managerial implications.

Keywords: firm performance, corporate governance, shareholder activism, shareholder

3

Table of content

2.1.1 Agency theory and shareholder activism ... 11

2.1.2 Institutional theory and shareholder activism... 13

2.2 Antecedents of shareholder activism ... 14

2.2.1 Target firm antecedents ... 14

2.2.2 Activist antecedents ... 16

2.2.3 Environmental antecedents ... 17

2.3 Process and outcomes of shareholder activism ... 18

2.4 Hypotheses ... 20

2.4.1. Shareholder activism and firm financial performance ... 21

2.4.2. Shareholder activism and firm social performance ... 23

2.4.3. The moderating effect of foreign institutional ownership on the shareholder activism -firm performance relationship ... 24 2.5 Conceptual model... 26 3. Research methodology ... 27 3.1 Data collection ... 27 3.2 Sample ... 28 3.3 Measurement of variables... 29 3.3.1. Independent variables ... 29 3.3.2. Dependent variables ... 30 3.3.3. Moderating variable ... 32 3.3.4. Control variables ... 33

3.4 Empirical data analysis ... 34

3.5 Robustness tests ... 34

4. Results ... 36

4

4.2 Descriptive statistics ... 37

4.3 Correlations ... 38

4.4 Regression results ... 40

4.5 Robustness tests results ... 43

4.6. Regression results with a refined shareholder activism measure ... 45

5. Discussion... 48

6. Conclusion ... 53

6.1 Theoretical implications ... 53

6.2 Practical implications ... 54

6.3 Limitations and future research ... 55

Acknowledgments ... 57

References ... 57

Appendices ... 65

Appendix A: Glossary of shareholder proposals ... 65

Appendix B: Industry types ... 67

Appendix C: Preliminary analysis ... 68

Appendix D: Shareholder proposal and industry type frequencies ... 71

Appendix E: Descriptive statistics ... 73

Appendix F: Correlation matrix 2014 ... 75

Appendix G: Regression results robustness test 1 (2014 dataset)... 76

Appendix H: Regression results robustness test 2 (different ROA period 2015) ... 78

Appendix I: Regression results robustness test 3 (different ROA period 2014) ... 79

Appendix J: Regression results robustness test 4 (ROS 2015) ... 80

Appendix K: Regression results robustness test 5 (ROS 2014) ... 81

Appendix L: Regression results robustness test 6 (differentiation of proposal submitters 2015) ... 82

Appendix M: Regression results robustness test 7 (differentiation of proposal submitters 2014) ... 84

Appendix N: Regression results using a refined measure of shareholder activism (2015 shareholder vote %) ... 86

Appendix O: Regression results using a refined measure of shareholder activism (2014 shareholder vote %) ... 88

Appendix P: Regression results using a refined measure of shareholder activism (ROS shareholder vote %) ... 90

5

List of Tables

Table 1: Descriptive statistics Table 2: Correlation matrix

Table 3: Regression analysis financial performance Table 4: Regression analysis social performance

List of Figures

Figure 1: Conceptual model

List of Abbreviations

CSP Corporate social performance

FIO Foreign institutional ownership ROA Return on assets

ROS Return on sales

SD Standard deviation

SoF Say-on-Frequency

SoP Say-on-Pay

VIF Variance inflation factor

6

1. Introduction

Dating all the way back to 1622, in the time where the world’s first-ever formally publicly traded company, the ‘Vereenigde Oostindische Compagnie’ (Dutch East India Company, or

VOC), was being active, the concept of shareholder activism originates. At that time, the VOC dominated trade with the East Indies and its shares were being traded on the Amsterdam stock exchange. Although the shareholders of the VOC could easily sell their shares, they did not have any control rights at all as the VOC never held any shareholders’ meetings (Gelderblom & Jonker, 2004). The absence of control rights, however, did not hinder the VOC shareholders to articulate their concerns. Unsatisfied shareholders argued that the directors managed the company inefficiently and that they used their powers primarily for self-enrichment. In order to stop this, the complainants demanded more influence in the VOC (de Jongh, 2011).

7

sparked much research interest in corporate governance literature, but empirical evidence on the exact nature of this relationship remains inconclusive. Some studies have reported that shareholder activism improves performance (Clark & Crawford, 2012; Cunat, Gine, & Guadalupe, 2012), while others have found that it leads to negative performance (Johnson & Greening, 1999; Cai & Walkling, 2011), or not any significant reaction at all (Waddock, 2003; Becht et al., 2009). These mixed results may be explained by the heterogeneity of activists’ demands, as well as the divergent streams of theories used throughout literature (Karpoff, 2001). Although scholars have long been aware of these differences, they have had little consideration for their implications in the methodology of their studies, resulting in the different performance explanations (Goranova & Ryan, 2014). Individually these studies thus offer different performance explanations, however, collectively they suggest that shareholder activism is a complex phenomenon where a distinction should be made between different types of activism demands. It can, therefore, be argued that this heterogeneity of shareholder activism should be seen as a multidimensional construct, one in which shareholders voice their concerns through the filing of varying types of proposals. The study of Karpoff, Malatesta, & Walkling (1996) makes such a distinction by introducing different types of shareholder proposals: executive compensation-related proposals, corporate governance-related proposals, and social/environmental-governance-related proposals.

8

management more extensively (Kang & Shivdasani, 1995). Only large(r) shareholders possess these characteristics, especially institutional investors (e.g., banks, pension funds), as they own large blocks of stock within a firm (Chaganti & Damanpour, 1991). Compared to domestic institutional investors, foreign institutional investors have the relevant monitoring characteristics even to a stronger degree, as they are less likely to have business relations with the local firm (Desender et al., 2016). It can, therefore, be argued that the impact of shareholder activism on firm performance depends on activist characteristics, where foreign institutional ownership (FIO) has an especially important role within activism as it enables activists to realize the inherent performance effects of activism. However, only little research attention has been given on the possible interaction effects of FIO on the relationship between shareholder activism and firm performance.

Hence, in order to shine a better light on the mixed results for the relationship between shareholder activism and firm financial and social performance, this thesis uses a multidimensional construct of shareholder activism by taking into account three different types of shareholder proposals as introduced by Karpoff et al. (1996). Additionally, a moderating role for foreign institutional ownership will be examined with regard to the shareholder activism-performance relationship, since this has only received little research attention. This thesis thereby aims to bridge these research gaps by using the following main research question:

What is the effect of shareholder activism on the financial and social performance of firms and how is this relationship moderated by foreign institutional ownership of firms?

9

this thesis contributes to existing shareholder activism literature. Also, important practical contributions are made by this thesis. As financial performance, and increasingly so social performance, can be seen as the main goal of firms, new knowledge about how performance can be improved or influenced is highly useful for both shareholders and managers. This thesis contributes to this knowledge by showing how activism affects both financial and social performance.

In order to test the main research question, this paper will use multiple regression analysis on a sample of firms. In the first part, the existing literature on shareholder activism and firm financial and social performance is reviewed, as well as research on the influence of FIO. Based on these key underlying concepts, the hypotheses and the core conceptual constructs of this research will be elaborated on. In the second part, the research design and methodology will be explained, including a description of the data. Subsequently, the data is analyzed in-depth and discussed. Last, the limitations and contributions of the study, as well as directions for future research, will be examined.

2. Literature review

The concept of shareholder activism has been an evolving concept, especially over the last few decades. Until the 1970s, the main actors on the shareholder activism stage are individual investors, also called ‘corporate gadflies’ by the contemporary media (Gillan & Starks, 2007).

10

stream, the amount of shareholder-initiated proposals slowly starts to increase, but it is not until the mid-1980s that they really begin to surge (Karpoff et al., 1996). It is at this time that both the Institutional Shareholder Services and the Council of Institutional Investors is founded, thereby marking a turning point for institutional activism (Davis & Thompson, 1994; Lipton, 2007). These two associations offer the possibility to pool resources and use proxy votes to protect the interests of their member investors, including many public pension funds, and thereby pave the way for the second stream in shareholder activism: the so-called ‘financial activism’ field. This financial stream embodies shareholders' primacy and is

especially focused on governance-based financial activism, thereby attempting to improve governance structures, maximize shareholder value and making managers more accountable to the firm’s shareholders (Gillan & Starks, 2000, 2007).

11 2.1 Theoretical foundations

In order to explain the antecedents of the two streams of shareholder activism, two types of distinct underlying theoretical foundations can be identified from literature. Shareholder activism is thereby defined as “the use of ownership position to actively influence company

policy and practice. Shareholder activism can be exerted through letter-writing, dialogue with corporate management or the board, asking questions at open sessions in general meetings, and through the filing of formal shareholder proposals” (Sjöström, 2008: 142). In this sense,

shareholder activists aim to strive for active dialogue or confrontation when the costs of activism are outweighed by its benefits (Pozen, 1994). Activist identities can thereby range from a singly minority investor, frequently called a ‘blockholder’, to institutional investors with majority stakes.

As mentioned earlier, the emphasis of activist shareholders can be to focus on financially poorly performing firms, called financial activism, and/or to focus on socially poorly performing firms, called social activism. The roots of the financial activism stream can be traced back to the rise of agency theory in explaining the battle for corporate control in the 1980s and onward (Zajac & Westphal, 1995), while the social activism stream is an ideological successor of the 1960s civil rights movement and can be traced back to institutional theory (Reid & Toffel, 2009).

2.1.1 Agency theory and shareholder activism

12

ownership and control in publicly traded companies an agency problem arises between principals (the shareholders) and agents (the managers). This agency problem occurs because these two cooperating parties have different self-serving goals and division of labor. Specifically, agency theory focusses on the agency relationship, a relationship in which the principal delegates work and decision-making authority to a certain degree to the agent, who performs this work. So, in the case of publicly traded firms, the work that is delegated by the shareholders is the control of their ownership, the firm, to managers. The theory attempts to describe this relationship by using the metaphor of a contract (Jensen & Meckling, 1976). Agency theory is interested in resolving two problems that might occur in agency relationships: (1) a goal difference between the principal and agent where it is difficult or costly for the principal to actually verify what the agent is doing, resulting in an information-asymmetry, and (2) the problem of risk sharing that occurs when the principal and agent have different perspectives on risk. As the unit of analysis is the ‘contract’ governing the relationship between the principal and the agent, the theory focusses on deciding the most efficient contract that can regulate the principal-agent relationship (Eisenhardt, 1989). In the case of publically traded firms, the shareholders thus need to closely monitor and incentivize managers via structural mechanisms so that the managers will maximize shareholder value in order to avoid the principal-agent conflict (Beatty & Zajac, 1994). Agency theorists have presented different solutions for this alignment problem, resulting in many different options, such as board independence, shareholder right plans, and granting executive stock options, (Denis & McConnell, 2003).

13

requiring specific actions from managers to improve shareholder value (Gantchev, 2013). By conveying their dissatisfaction, these activists aim to reduce principal-agent problems.

2.1.2 Institutional theory and shareholder activism

Next to agency theory, another important theory within shareholder activism literature is institutional theory. Key in institutional theory is the necessity of organizational legitimacy as this ensures organizational survival. The extent to which organizations are deemed socially acceptable or legitimate thereby relates to societal norms and expectations (Ashforth & Gibbs, 1990). In order to achieve legitimacy, firms need to show certain behavior that adheres to these societal norms. In their search for organizational legitimacy, firms often end up with similar strategies as firms operating in the same environment due to three so-called ‘isomorphic processes’ (DiMaggio & Powell, 1983): (1) coercive pressure (political

influence, e.g., government regulations), (2) mimetic forces (standard responses to uncertainty, e.g., the copying of newly developed technology), and (3) normative standards (professionalization of an organizational field, e.g., adoption of ‘best practices’ in an industry). Each of these three forces thereby leads to an ever more increasing homogenization of organizational practices within a given organization field because firms that meet the environment’s expected characteristics receive societal legitimacy and gain access to

resources (Toma, Dubrow, & Hartley, 2005). If a firm, however, decides not to follow these isomorphic forces it could become illegitimate and would be denied access to resources that it needs in order to survive.

14

shareholder wealth because organizational survival depends on it (Rowley & Moldoveanu, 2003; Scott, 2011). As firms are ingrained within wider society through their operations, by, for example, their use of natural resources and employment of people, expectations are thus created on firms by social actors in their surrounding environment to respect shared norms. According to the institutional perspective, activists are thus seen as conveyers of dissatisfaction with the social performance of the target firm and require specific actions from managers to improve this (den Hond & de Bakker, 2007). By conveying their dissatisfaction, these activists aim to shape and/or change societal norms about corporate social performance (CSP), which in turn pressurizes firms via the three isomorphic forces to adhere to this.

2.2 Antecedents of shareholder activism

While the abovementioned theoretical foundations of shareholder activism can explain activists’ financial and/or social motivation to engage in activism, it is important to take a

fine-grained approach that goes beyond this as that helps to fully understand what triggers shareholder activism. The antecedents of shareholder activism are therefore also essential to take into account. These antecedents consist of firm, activist, and environmental characteristics (Ryan & Schneider, 2002).

2.2.1 Target firm antecedents

15

difficult for shareholders to monitor effectively and therefore more prone to agency problems (Strickland, Wiles, & Zenner, 1996; Del Guercio & Hawkins, 1999). Alternatively, according to the social activism stream, large firms are more sensitive to challenges to their social legitimacy than smaller firms due to their greater visibility (Rehbein et al., 2004). Activist campaigns aimed at large firms have a better chance of attracting public and media attention, thereby facilitating spillover effects to non-targeted firms that will adopt the reforms demanded by activists at industry leaders (Ferri, & Sandino, 2009). In terms of performance, shareholder activists tend to target firms whose stock market performance is suboptimal (e.g., Brav et al., 2008; Renneboog & Szilagyi, 2011). The agency problem of free cash flows theorizes that managers prefer to spend cash on value-decreasing investments instead of distributing it back shareholders (Jensen, 1986). Firms’ cash holdings draw activists’ attention, especially when they have lower distributions to their shareholders (Klein & Zur, 2009). In contrast to this, social activists choose to focus on targeting firms with good prior performance. Addressing social concerns diverts managerial time and firm resources away from short-term shareholder value creation to the interest of long-term value creation (Tudway & Pascal, 2006). This makes it rational for social activists to target firms that have slack resources to address social issues (Zietsma & Winn, 2008).

16

terms of ownership concentration, more diffused ownership leads to more constraints to address social concerns (Judge, Gaur, & Muller‐Kahle, 2010). Minority owners invest less money in their local communities than majority owners and are therefore inclined to show less ‘responsible’ ownership (Adams & Hardwick, 1998; O’Rourke, 2003). From the institutional perspective, this means that firms with less concentrated ownership are likely to be confronted with more diverse demands compared to firms with more concentrated ownership. When firms face multiple conflicting demands they are more prone to ignore some elements, one of these being, for example, social concerns (Child & Tsai, 2005). Therefore, firms with more concentrated ownership are likely to be targeted by shareholder activism. Lastly, firms with a misalignment between executive compensation and firm performance also cause shareholder activism because it represents a lost chance in alleviating the agency problem (Ertimur, Ferri, & Muslu, 2010; Cai & Walkling, 2011). If managers do not produce sufficient cash flows to justify their pay, they are likely to be targeted or pressured by activists to reduce their compensation (Murphy, 2002).

2.2.2 Activist antecedents

Despite the focus of most prior empirical research on firm-level antecedents, the own characteristics of shareholder activists are also theorized to have an influence in their engagement in activism (Ryan, & Schneider, 2002; Sikavica, & Hillman, 2008). Considering the capacity, interests, and salience of activists are therefore important in creating a full picture of shareholder activism. First, activists’ capacity and interests in engaging in activism could be decoupled from the target firm’s financial and governance situation. Activism costs

17

2013), and to justify its costs activists have to either derive an adequate improvement in overall shareholder wealth or look for benefits that are not shared by all other shareholders (Kumar & Ramchand, 2008). Hence, investors’ ability and desire to engage in activism can be affected by their fund size (a larger fund size allows activists to spread risks and expenses of activism; see Ryan & Schneider, 2002) and investment horizons (longer investment horizons allow the spreading of activist’ expenses; see Rubach & Sebora, 2009). Second, shareholders that possess superior salience and negotiating leverage relative to managers or that have the ability to gain other shareholders’ support could be more willing to engage in activism

because they expect greater returns on their activism investments (Kang & Sorensen, 1999). This self-selection would imply that these powerful investors may be more likely to become activists (Chandler, 2006; Chowdhury & Wang, 2009). Proposals by institutional investors, for example, have historically received much higher approval rates than those initiated by individual investors (Proffitt & Spicer, 2006).

2.2.3 Environmental antecedents

The third and last antecedent that provokes shareholder activism is the macro environment. In the macro environment, changing conditions can either have a direct or indirect effect on activism. The institutional environment can change due to altering societal norms and behavior, which can result in new or modified rules in the legal environment (Del Guercio, 1996). Additionally, technological changes, such as improved communication techniques, have lowered shareholder activists’ costs and ease of engaging in activism (Wessel, 2011).

18

shareholder activism (Giroud & Mueller, 2011). Managers may find it easier to observe and react to shareholder activism at identical firms in their environment, thereby assisting spillover effects and mimetic isomorphism (Lee & Park, 2009).

2.3 Process and outcomes of shareholder activism

Now that the motivation and triggers of shareholder activists are revealed, the actual process of activism and its effects on firm-level outcomes can be discussed. The process of activism starts when shareholders arrive at a point that they become increasingly dissatisfied with their firm’s policies, actions, and choices. At this point, they can opt for three choices, which have

traditionally been classified as loyalty (hold), exit (trade), or voice (activism) (Davis & Thompson, 1994). Voicing concerns can be done utilizing a variety of tactics, ranging from public activism, by for example shareholder proposals, media campaigns, and publicized letters (Song & Szewczyk, 2003; Reid & Toffel, 2009), to private activism, by for example private negotiations, phone calls, and letters (Becht et al., 2009; Logsdon & Van Buren, 2009). Especially filing shareholder proposals present shareholders with a cheap and relatively simple way to present matters to the company's owners (Thomas & Cotter, 2007). Filed proposals can thereby play an essential role in shaping firm practices by facilitating the fulfillment of activists’ demands or by sending powerful signals to managers and influence their actions (Ferri & Sandino, 2009; Hillman et al., 2011).

19

value-creating view, which sees activist efforts as helping to improve the incentive and control problems that characterize large diffusely held firms. According to this view, shareholder activism increases firm social performance (e.g., Guay, Doh, & Sinclair, 2004; Clark & Crawford, 2012), as well as financial performance (e.g., Brav et al., 2008; Cunat et al., 2012), the latter one often via a governance-based way. Second, there is the value-destroying view, which views shareholder activism as an impairment on firm management and operations because activists neither have the skills nor the experience to improve managers’ decisions. Due to shareholder interference firm financial performance (e.g., Bizjak

& Marquette, 1998; Cai & Walkling, 2011) and social performance (e.g., Johnson & Greening, 1999; Prevost & Rao, 2000) are lowered. Lastly, the third view on the effectiveness of shareholder activism outcomes considers activism as having a negligible impact on governance-based firm financial performance (e.g., Carleton, Nelson, & Weisbach, 1998; Becht et al., 2009) and social performance (e.g., Waddock, 2003).

The mixed results on the effectiveness of shareholder activism may be explained by the heterogeneity of activists’ demands, as well as the divergent streams of theories used throughout literature (Karpoff, 2001). Although scholars have long been aware of these differences, they have had only little consideration for their implications in the methodology of their studies. Consequently, most of the literature on shareholder activism arrives at varying explanations for the different performance implications of shareholder activism mentioned above (Goranova & Ryan, 2014). Individually these studies thus offer different performance explanations, however, collectively they suggest that shareholder activism is a complex phenomenon, and thus a distinction should be made between different types of activist’ demands. The process of shareholder activism (filing different shareholder proposals)

20

relationship, are the characteristics of activists. The characteristics of activists can namely influence whether shareholders actually choose to engage in the process of activism or not. Especially foreign institutional ownership as an activist characteristic may be effective in realizing the performance effects of shareholder activism because this allows activists to realize performance effects through salience and extensive monitoring and control (Desender et al., 2016).

2.4 Hypotheses

21

in different categories: (1) executive compensation, (2) corporate governance, and (3) environmental/social1.

2.4.1. Shareholder activism and firm financial performance

The first dimension of shareholder activism that influences firm performance is based on governance-related shareholder proposals. Governance-based types of shareholder proposals are thereby proposed by shareholders in order to improve the firm’s financial performance (Gillan & Starks, 2000, 2007). Using the categories introduced by Karpoff et al. (1996), the first two of these categories are, unsurprisingly, of the greatest interest to financially orientated shareholder activists (Romano, 2001).

The first category, executive compensation, relates to the alignment of executive incentives with firm objectives: mechanisms that motivate behavior. Agency theory predicts that managers are unlikely to act in maximizing shareholder wealth unless it is in their self-interest to do so (Ertimur et al., 2010). Allowing shareholders to express their opinions and have a say in executive compensation may reduce the agency costs between executives, directors, and shareholders, as it would result in increased accountability and communication between these parties (Davis, 2007). It puts pressure on the executives to improve performance, given that shareholder proposals with a negative content may have substantial consequences on the support for the executives within the firm (Fama, 1980). Consequently, shareholder proposals on executive compensation can result in increased shareholder incentivizing and monitoring (i.e. more efficient compensation contracts) which add value to

1

22

the firm by holding executives accountable (Cunat et al., 2015; Deane, 2007). Hence, I hypothesize:

Hypothesis 1: Filing executive compensation-related shareholder proposals will be positively related to a firm’s financial performance

23

When a board’s structure can be altered to a board that is more independent and better

qualified, it can promote the shareholders' interests (Brickley & Zimmerman, 2010) and also remove value-decreasing takeover defenses (Guo, Kruse, & Nohel, 2008). Consequently, shareholder proposals on corporate governance can result in the de-entrenchment of management, increased shareholder monitoring (i.e. stronger corporate governance mechanisms), which increase the target firm performance. Hence, I hypothesize:

Hypothesis 2: Filing corporate governance-related shareholder proposals will be positively related to a firm’s financial performance

2.4.2. Shareholder activism and firm social performance

The second dimension of shareholder activism that influences firm performance is based on social-related shareholder proposals. Social-based types of shareholder proposals are thereby proposed by shareholders in order to improve the firm’s CSP. The third category from

Karpoff et al. (1996), environmental/social, is therefore of the greatest interest to socially orientated shareholder activists. This category relates to raising awareness, directing attention, and challenging firms to improve their CSP (Wood, 1991). In comparison to the other dimension of shareholder proposals, social-based shareholder proposals are often filed with not only the intent to rally the shareholders of the target firm but also to reach and influence other stakeholders, such as the firm’s employees and customers. Current and potential employees may respond in a negative manner when they become aware of shortcomings in a firm’s CSP through proposals (Turban & Greening, 1997). Regulatory agencies may also be warned by activist’ proposals and may take them into consideration when designing

24

goods from the firm, or they may even start their own campaigns via boycotts (Hoyer & Maclnnis, 1997). Consequently, shareholder proposals on social/environmental issues can result in managers becoming alerted to the underlying CSP problems and the need for change, which leads them to evaluate the possibility of corrective action and to make changes to improve CSP. Hence, I hypothesize:

Hypothesis 3: Filing environmental and social-related shareholder proposals will be positively related to a firm’s social performance

2.4.3. The moderating effect of foreign institutional ownership on the shareholder activism -firm performance relationship

While the above reasoning provides insights into the performance impact of shareholder activism, it does not incorporate any ownership-level antecedents. However, it is after all the owners of the firm, the shareholders, who decide to engage in activism. When a firm has a highly diffused ownership, there is no incentive for any owner to monitor management because the individual shareholder would have to carry the total monitoring costs, while all the other shareholders would enjoy the benefits. Because of this, large(r), more salient shareholders are argued to have sufficient incentives to monitor firms (Kang & Shivdasani, 1995; Noe, 2002). In terms of salient/large shareholders, institutional investors have become an important participant in equity markets, giving them the potential to influence management’s activities directly via their ownership and indirectly by trading their shares.

25

investors has been growing, which in turn can have an important impact on the relation between shareholder activism and firm performance.

Institutional investors have the power to pressurize firms to comply with their needs regarding performance as they possess monitoring capabilities and negotiating leverage due to the fact that they commonly own large(r) blocks of stock within a firm (Chaganti & Damanpour, 1991). Like any other shareholder, they also can directly influence management using their ‘voice’, by voting on shareholder proposals. Gillan & Starks (2003) highlight the particular role that institutional investors, especially foreign institutional investors, play in inducing change in corporate governance worldwide. Foreign institutions often take a more active stance compared to domestic institutions. As domestic institutions have business relations with local corporations, they may feel compelled to be loyal to management according to business relationship theory (Desender et al., 2016). Foreign institutional investors would, therefore, voice their interests quicker and louder than domestic institutional investors as they are less likely to have business relations with the firm. This increased independence increases the ability of foreign institutional investors to monitor the firm. The monitoring of the firm becomes especially important in the case of shareholder activism via proposals, as these proposals are non-binding (Thomas & Cotter, 2007). Although proposals may receive strong support from shareholders, management may decide to not follow-up on them. However, with greater FIO, the probability that management will follow-up on proposals is likely to increase due to the stronger, more independent, monitoring capacity of foreign institutional investors. In turn, this can have a positive effect on performance.

governance-26

related proposals), firms with higher FIO can achieve higher revenues due to a limitation of agency problems within that firm resulting from increased monitoring and control activities.

Hence, I hypothesize:

Hypothesis 4a: Higher foreign institutional ownership of firms will positively moderate the positive effects of both categories of filing governance-related shareholder proposals on a firm’s financial performance

For the second dimension of shareholder proposals, that is social-related proposals, a higher FIO of a firm would mean that firms have to increasingly deal with social activists that voice their concerns quicker and louder (Chappie & Moon, 2005; Oh, Chang, & Martynov, 2011). In turn, this would mean that other stakeholders can easier become aware of shortcomings in a firm’s CSP and undertake their own action. Managers would then be alerted to a greater

extent to CSP problems, thus resulting in a stronger perceptiveness to these problems and an increased need to take corrective actions that would improve CSP. Hence, I hypothesize:

Hypothesis 4b: Higher foreign institutional ownership of firms will positively moderate the positive effects of filing environmental and social-related shareholder proposals on a firm’s social performance

2.5 Conceptual model

27

firm financial performance (H1-H2) as well as the proposed positive relationship between the independent variable of social-related shareholder proposals on the dependent variable firm social performance (H3). Additionally, the positive moderating effect of the variable FIO on the proposed shareholder activism performance relationship is incorporated in the model.

Figure 1: Conceptual model

3. Research methodology

3.1 Data collection

28

effects of FIO on this relationship will thereby be studied by a quantitative approach. The secondary data that is used for this approach comes from several databases. The data for the independent variables, the three categories of shareholder proposals, is obtained via the website ProxyMonitor.org, a website managed by the Manhattan Institute, a public, nonpartisan, independent, free-think tank consisting of scholars. Their database records all proxy proposals submitted by shareholders of the 250 largest American public companies for annual shareholder meetings. For the dependent variable firm financial performance as well as for the moderating variable FIO and all the control variables, data is obtained via a database called Orbis, which is provided by Bureau van Dijk. This database covers financial, statistical, market, and corporate governance-related information from publicly traded companies. Lastly, for the dependent variable firm social performance data is obtained via a database provided by Thomson Reuters EIKON: ASSET4 ESG.

3.2 Sample

29

data, 240 companies are included to form the final sample2. This is sufficient for a good sample size given the rule of thumb that a sample should include at least 10 observations per variable (VanVoorhis & Morgan, 2007).

3.3 Measurement of variables

3.3.1. Independent variables

The literature review reveals that the three different categories of shareholder proposals each have their own effect on firm performance. Following Karpoff et al. (1996), this study composes the different categories by grouping individual shareholder proposals into these categories. First, with regard to the category of executive compensation-related shareholder proposals, Say-on-Pay (SoP) and Say-on-Frequency (SoF) proposals are used to measure activism. Second, for the category of corporate governance-related proposals, chairman independence, director qualification, confidential and cumulative voting, declassify the board, rescinding poison pill and removal of golden parachutes proposals are applied. Lastly, human and animal rights, diversity, environmental, gender equality, lobbying and political spending are utilized for the third category of social/environmental-related proposals. For a more detailed overview of the different shareholder proposals see Appendix A.

How the exact measurement of shareholder activism should be performed is, however, debated in literature. The survey study of McCahery, Sautner & Starks (2016) finds that shareholders most frequently use ‘voice’ in their activist engagement, rather than ‘exit’.

Voicing concerns can be done utilizing a variety of tactics, ranging from public activism, by for example shareholder proposals, media campaigns, and publicized letters (Song &

30

Szewczyk, 2003; Reid & Toffel, 2009), to private activism, by for example private negotiations, phone calls, and letters (Becht et al., 2009; Logsdon & Van Buren, 2009). Especially shareholder proposals present shareholders with a cheap and relatively simple way to present matters to the company's owners. Although shareholder proposals are non-binding, they can play a critical role in facilitating the implementation of activists’ demands by sending powerful signals to management (Hillman et al., 2011). Hence, following previous research (e.g., Karpoff et al., 1996) I operationalize shareholder activism as the filing of each of the three identified shareholder proposals categories in the year 2015. In order to be able to conduct a regression analysis (they can only be run with ratio measured variables and dummy variables), I dummy code whether or not a proposal is filed to a specific company. Receiving a proposal receives a one, while not receiving a proposal receives a zero in the data. Important to note is that while shareholder activism for both the categories corporate governance and social/environmental–related proposals are measured using the filing of proposals to a target firm (via a dummy), the category executive compensation-related proposals is measured using the percentage of votes in favor that filed proposals received from the shareholders in a shareholders meeting (also done in the study of Thomas & Cotter (2007)). This is done because almost all the firms in the sample received an executive compensation-related proposal. Dummy coding would therefore not be a good measure.

3.3.2. Dependent variables

market-31

based measures (e.g., stock returns around announcement dates of shareholder proposals) or accounting-based measures (e.g., ROA). Market-based measures, however, bring about substantial event date uncertainty and other problems, which can undermine the ability of an event study to detect the impact of proposals (Del Guercio & Hawkins, 1999), making an accounting-based measure much more effective (Del Guercio, Seery, & Woidtke, 2008). Hence, I operationalize a firm’s financial performance as ROA, which can be calculated by dividing a firm’s net income by the total value of the firm’s assets at the end of the period (see

formula 1). This measure indicates how efficient a company is in using its assets to generate earnings. Important to note here is that the effects of shareholder activism are likely to show up only gradually over time (Karpoff et al., 1996; Del Guercio et al., 2008), making the adoption of a change in ROA over multiple years necessary (Gillan & Starks, 2007). In order to fully measure the impact of shareholder activism on ROA, I thus use a measure that takes into account both prior operating performance and post-activism operating performance. Therefore, following Karpoff et al. (1996), I calculate ROA one year before the firm receives a proposal (i.e. 2014) and three years after the received proposal (i.e. 2018) and then measure the percentage difference between these two. As the outcome of this is a percentage that can take any value, both positive and negative, this makes it a continuous variable.

Formula (1): ROA = Net income

End of period assets

32

companies with a controversies overlay (an ESG performance score based on negative media stories) as this increases the accurateness of the overall ESG score. This measure is also commonly used by previous studies (Velte, 2016). Following previous research, a lag of one year is chosen in order to fully take into account the effects of shareholder activism on firm social performance (David et al., 2007). The combined score is a number that can range between 0-100, which makes it a continuous variable.

3.3.3. Moderating variable

As (foreign) institutional investors3 need to be registered in the U.S. with a separate filing (13-F), they can be identified accordingly. As such, previous studies have used 13-F filings to identify the involvement of (foreign) institutional investors in firms (Elyasiani & Jia, 2017). This study uses a similar approach. Following Ferreira & Matos (2008), the moderator will be measured as the proportional number of shares owned by foreign (other than U.S.) institutional investors (see formula 2) in the year 2015. As the outcome of this measure is a percentage which can take any value, both positive and negative, this makes it a continuous variable.

Formula (2): Foreign institutional ownership =Shares owned by foreign institutional investors

Total outstanding firm shares

3

33 3.3.4. Control variables

34

three4 categories are dummy coded into either a value of one if the firm operates in the corresponding industry or zero if this is not the case.

3.4 Empirical data analysis

In order to test the suggested hypotheses, IBM SPSS version 24 is used. As the study examines multiple independent variables to explain two continuous dependent variables, two separate multiple linear regressions analyses are appropriate to use. Since there are moderation effects in this study, moderated multiple regressions are used. This requires the independent variables and the continuous moderator FIO to be mean-centered for the analysis, as well as the computing of interaction variables (Hayes, 2013).

In case there is missing data for one of the variables, the firm is still included in the analysis as this method (pairwise deletion) maintains valuable data (Enders, 2010).

3.5 Robustness tests

In order to test the sensitivity of the data and to make sure that the results have sufficient validity, the regression analyses are first repeated by using data of shareholder proposals filed in 2014 (combined with the corresponding data for all the other variables as specified in methodology). Additionally, another period is chosen for ROA change: the difference between ROA one year before the firm receives a proposal, to one year after the firm received the proposal. This is done for both the 2014 and 2015 datasets. Also, another accounting-based measure for the dependent variable financial performance is used for both years’ datasets: return on sales (ROS) (using the same change period as the original ROA measure).

35

ROS is also used in other shareholder activism-performance studies as a robustness test (e.g., Del Guercio & Hawkins, 1999), and can be calculated by dividing the MNC’s net income by net sales (see formula 3). The outcome is a percentage which can take any value, both positive and negative, making it a continuous variable.

Formula (3): ROS = Net income

Net sales

Finally, to check whether the identity of an activist shareholder has any influence in the entire process, this study makes a broad classification of the different types of submitters of shareholder proposals. This is done separately for each of the different categories of shareholder proposals, except for the category of executive compensation-related shareholder proposals since almost all the submitters for this category are the same in the obtained data (i.e. “other”). For corporate governance-related shareholder proposals, proposals filed by

36

member of the abovementioned two groups, that is, individuals and social institutions. A firm that receives a specific type of proposal from the abovementioned classified groups receives a one (e.g., an individual filing a corporate governance-related shareholder proposal), while a firm that receives a proposal from another type of activist shareholder in that shareholder proposal category (e.g., a proposal filed by a labor union in the corporate governance-related shareholder proposal category), as well as firms that do not receive a proposal at all, receive a zero in the data. This robustness test is performed for both year’s datasets, 2015 and 2014.

4. Results

4.1 Test of basic assumptions

37

meeting the assumption of linearity and homoscedasticity (Hayes, 2013). Additionally, several outliers occurred during the data analysis. As these outliers might influence the dataset, all variables are winsorized at 1% (except the dummies). For details see Appendix C.

4.2 Descriptive statistics

The descriptive statistics are presented in Table 1. Shareholder proposals that are submitted at the 240 sampled firms are mostly executive compensation-related (231), then corporate governance-related (63), and least of all social-related (58) (for the specific type of proposal frequency per category see Appendix D, Table D.1-D.3). Regarding industry type, 28 firms operate in natural resource-based industries, 78 in manufacturing industries, 129 in service industries, and four in other industries (see Appendix D, Table D.4).

The data for the 2015 sample shows that the average change in ROA for a firm is .41% (SD=5.05), which is quite alike to the robustness measure of the average change in ROS (.58%; SD=7.25), while the average CSP score is 54.98 (SD=16.31). From the independent variable executive compensation shareholder proposals, it becomes clear that shareholders highly approve this proposal category, as on average 91.52% of the votes are in favor, meaning that shareholders would like to see management implementing this proposal category.

38 Table 1: Descriptive statistics 2015

Variable N Min Max Mean SD

Dependent variables ROA change (%) 231 -12.31 13.49 .41 5.05 ROS change (%) 189 -19.51 20.08 .58 7.25 CSP 234 21.27 91.88 54.98 16.31 Independent variables Ex. compensation SP (%) 231 70.71 99.63 91.52 7.70 Corp. governance SP (dummy) 240 0 1 .26 .441 Social SP (dummy) 240 0 1 .24 .429 Moderating variable FIO (%) 235 0 33.20 14.43 6.27 Control variables Firm size 240 8.40 12.37 10.53 .61 Financial leverage (%) 197 -107.39 411.94 111.03 112.86 Firm age 240 0 165 44.11 36.91

Note: variables are presented in original form and not mean-centered. This table reports descriptive statistics for all the variables used in this study. All variables, except the dummies are winsorized at the 1% level in both tails. Presented are the number of observations (N), the mean, minimum (Min), maximum (Max) and standard deviation (SD) of all the variables for the full sample.

4.3 Correlations

Table 2: Correlation matrix 2015 ROA change CSP Ex. Compensation SP Corp.

governance SP Social SP FIO Firm size

Financial

leverage Firm age

4.4 Regression results

The results of the main regression analysis can be found in Tables 3 and 4. As can be seen in Table 3, six different models are used for the financial performance regression analysis. The first model analyzes the effects of the control variables on firm financial performance. It can be seen that none of these variables can explain the dependent variable at a significant level.

In model 2 the independent variable executive compensation-related shareholder proposals is added in a separate manner to the control variables in order to test hypothesis 1. The results show that this variable has a small positive (B=.003; p=.966), but insignificant effect, also when looking at the general regression model 4 (B=.005; p=.950). As both models do not support hypothesis 1, it can be rejected.

After this, the effect of the independent variable corporate governance-related shareholder proposals on firm financial performance (hypothesis 2) is tested in model 3. From Table 3 it becomes clear that the filing of corporate governance-related proposals also has a positive but insignificant effect (B=.059; p=.433), which is confirmed in model 4 (B=.060; p=.444). Therefore, both models do not support hypothesis 2, which can thus be rejected.

Table 3: Regression analysis 2015

Financial performance (ROA)

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Independent variables Executive compensation SP .003 (.049) .005 (.049) .011 (.049) .009 (.049) Corp. governance SP .059 (.887) .060 (.890) .055 (.882) .046 (.887) Moderating variable FIO -.155 (.059)** -.173 (.065)** Interactions

Executive compensation SP X FIO -.073 (.009)

Corp. governance SP X FIO .073 (.178)

Control variables

Firm size -.070 (.620) -.070 (.624) -.087 (.645) -.086 (.648) -.078 (.642) -.056 (.656)

Financial leverage .093 (.003) .093 (.003) .091 (.003) .091 (.003) .087 (.003) .080 (003)

Firm age -.038 (.010) -.038 (.010) -.041 (.010) -.041 (.010) -.037 (.010 -.028 (.010)

Natural resource-based industries .028 (2.793) .027 (2.810) .016 (2.807) .015 (2.824) .007 (.2.797) -.021(2.837) Manufacturing industries .041 (2.668) .040 (2.693) .020 (2.686) .019 (2.712) .008 (2.687) -.032 (2.720)

Service industries .066 (2.625) .064 (2.644) .050 (2.636) .048 (2.655) ..031 (2.631) -.012 (2.662)

R square .017 .017 .020 .020 .044 .054

Adjusted R square -.016 -.021 -.018 -.023 -.004 -.004

F-statistic .522 .445 .530 .462 .917 .924

42 Table 4: Regression analysis 2015

Social performance (CSP)

Model 1 Model 2 Model 3 Model 4

Independent variables Social SP -.022 (2.918) -.021 (2.944) -.026 (2.954) Moderating variable FIO .008 (.188) .039 (.210) Interactions Social SP X FIO -.072 (.510) Control variables Firm size -.192 (1.946)*** -.185 (2.045)** -.186 (2.055)** -.198 (2.088)** Financial leverage .070 (.011) .069 (.011) .069 (.011) .069 (.011) Firm age .107 (.033) .106 (.033) .106 (.033) .102 (.033)

Natural resource-based industries .-.194 (8.764) .-.187 (8.886) -.187 (8.910) -.188 (8.916)

Manufacturing industries -.216 (8.370) -.209 (8.430) -.209 (8.453) -.201 (8.464)

Service industries .-.227 (8.236) .-.222 (8.279) -.221 (8.303) -.223 (8.309)

R square .050 .051 .051 .055

Adjusted R square .020 .015 .010 .008

F-statistic 1.642 1.412 1.231 1.176

interact with FIO, which is in line with hypothesis 4A. However, since the effects are not significant, FIO does not seem to significantly positively moderate the shareholder activism effects on financial performance and hypothesis 4A needs to be rejected.

Looking at Table 4, four different models are used to perform a CSP regression analysis. Here, model 1 explores the effects of the control variables. It can be seen that firm size is the only variable that can explain the dependent variable on a significant level (p≤0.01). Model 2 then tests the effect of the independent variable

social/environmental-related shareholder proposals. The results show that social/environmental-social/environmental-related shareholder proposals have a negative insignificant effect (B= -.022; p=.771), thereby rejecting hypothesis 3. Next, in model 3 the direct effects of the moderating variable FIO on CSP are tested. The evidence shows that this variable has a positive insignificant effect (B=.008; p=.909).

When adding the interaction term in model 4, it becomes clear that there is a negative (B= -.072) relationship for the moderating effect of FIO on the shareholder activism-CSP relationship. However, given the fact that the B-coefficient is negative and insignificant, FIO does not seem to positively moderate the relationship between social/environmental activism and CSP. Therefore, hypothesis 4b is rejected.

4.5 Robustness tests results

44

The other independent variable, corporate governance-related shareholder proposals, also shows a negative effect (B= -.022) on financial performance, although at an insignificant level. Looking at the interaction effects in model 6, these still both remain insignificant in the 2014 analysis, thus rejecting hypothesis 4a. Next, the robustness CSP regression of 2014 also changes in quite some important ways to the main CSP regression of 2015. Although social/environmental proposals remain to have a negative insignificant effect in both models 2 (B= -.119) and 3 (B= -.113), the effect of these proposals on CSP does become negative significant in model 4 (B= -.400; p=.025). This means the exact opposite of the prediction that hypothesis 3 made. The interaction variable of social/environmental proposals with FIO in model 4 (B=.318; p=.077), on the other hand, shows a positive significant effect, which would be in line with hypothesis 4b. However, given the negative relation between social/environmental proposals and CSP in model 4, hypothesis 4b is still rejected.

The second and third robustness tests, using a different period for ROA, can be found in Appendix H (2015 data) and Appendix I (2014 data). Both regressions are quite similar to each other but somewhat different from the original ROA period used as the main regression analysis above. In both the robustness tests, executive compensation-related shareholder proposals have a negative insignificant effect on the different period of ROA. This, therefore, rejects hypothesis 1. Also, hypothesis 2 is rejected as the effect of corporate governance-related shareholder proposals on the different period of ROA is positive, albeit insignificant, in both years’ datasets. The interaction effect of FIO and executive compensation-related shareholder proposals in both robustness datasets is negative and insignificant, while for the other interaction, corporate governance-related proposals, and FIO, it is positive insignificant. This rejects hypothesis 4a.

45

previous tests, although the coefficients somewhat change. Both the independent variables in both years’ ROS regressions show an insignificant relationship with ROS, thereby rejecting hypotheses 1 and 2. Also, the interactions for the independent variables with FIO indicate an insignificant effect on ROS, thus rejecting hypothesis 4a.

The sixth and seventh robustness tests, using proposals filed by only specific shareholder groups, can be found in Appendix L (2015 data) and in Appendix M (2014 data). The results of both years’ ROA regressions show rather insignificant results, which is also largely the case for the CSP regressions. Yet, it can be noted that in the 2014 CSP regression social/environmental-related proposals filed by social institutions seem to have a negative significant effect on CSP (B=-.118; p≤0.1) in model 2. However, since this result only occurs in one model in one of both tested years, it should not be given too much attention.

All in all, these robustness tests show interesting results that are somewhat different from the main regression analysis. Although they thereby offer an important empirical basis for this study with regard to data sensitivity and validity, their results remain mostly insignificant. Therefore, it would be interesting to refine the most important measure in the analysis, that is, the measure for shareholder activism.

4.6. Regression results with a refined shareholder activism measure

46

is likely to urge shareholders backing the proposal in a more intense campaign for its implementation and attracts greater press coverage, which increases the political costs of ignoring the vote (Ertimur et al., 2010). Lower levels of voting support, on the other hand, are consistent with the view that shareholders do not believe the issues raised in the proposal as important to the firm.

The refined measure is operationalized as the percentage of votes in favor of each of the three identified shareholder proposals categories in the year 2015 (and also in the year 2014 for a robustness test)5. In the case that a company receives multiple identical proposals types in a year (e.g., multiple SoP proposals) the average amount of votes in favor of this type of proposal is taken.

Several regressions are performed using this measure, first for both the years 2015 and 2014 using ROA as the dependent variable, and second for both the years 2015 and 2014 using ROS as the dependent variable. All these conducted regressions thereby show that the refined measure of shareholder activism makes quite an impact in terms of the significance of the coefficients. Important to note, however, is that these regressions have a smaller sample (N≈50) compared to the main regression (N=240) and the robustness tests (N≈240), due to the use of listwise deletion of missing data. The disadvantage of this smaller sample can be that the results are somewhat less precise and generalizable, but when comparing the smaller sample with the larger sample, the descriptive statistics do not differ that much (see Appendix E). This indicates that the smaller sample is quite comparable to the larger sample and thus also its results.

First, the results of the regressions that use ROA as a dependent variable can be found for 2015 in Appendix N and for 2014 in Appendix O. Just as for the very first robustness test, these analyses produce quite different results compared to the main regression analysis

5

47

presented above, especially regarding the effect of activism on financial performance. Models 2 (2015: B= -.253; p=.055 (2014: B= -.338; p=.016)) and 4 (2015: B= -.267; p=.038 (2014: B= -.354; p=.014)) show negative significant effects of executive compensation-related proposals, thereby proving the exact opposite of hypothesis 1. This is also the case for the negative significant effect of corporate governance-related proposals, which is supported in models 3 (B= -.249; p=.070) and 4 (B= -.264; p=.048) of the 2015 dataset, thereby also proving the opposite of hypothesis 2. The interactions from executive compensation-related proposals and corporate governance-related proposals with FIO in these two analyses remain the same in both analyses, namely all insignificant, thus rejecting hypothesis 4a. An exception is the corporate governance shareholder proposal-related interaction with FIO in the regression analysis of 2014, as this one is negative significant (B= -.407; p≤0.01). Contrary to the abovementioned opposite effects of activism on financial performance when using a different measure for activism, the effects of activism on CSP remain almost unchanged when using the voting percentage robustness measure for both the years 2015 and 2014 and comparing these with the main analysis, namely insignificant. As this is also the case for the interaction of social/environmental proposals with FIO, hypotheses 3 and 4b can be rejected.

48

Overall, the robustness tests results, as well as the results from the tests using a refined measure of shareholder activism, are quite different from the main regression analysis, although not in the sense of their impact on the hypotheses. Both these differences and similarities between the robustness tests and the main regression are evaluated in the subsequent discussion part.

5. Discussion

The main objective of this thesis is to contribute to the debate on shareholder activism effects, with a special focus on the effect that activism has on firm performance. By examining the direct effects of three categories of shareholder proposals on both firm financial and social performance, this thesis expands the knowledge of shareholder activism research. The results of this study thereby indicate that almost all of the 250 largest U.S. listed firms do encounter activism, at least in the form of proposals. In turn, the filing of, as well as the voting on, these proposals generally lead to no or even negative effects on firm financial and social performance. Additionally, although only sporadically tested in existing activism research, interactions for foreign institutional ownership with the different proposal types are also researched within this study. However, overall these do not show a lot of explanatory power.

49

2010). Yet another view is that corporations are not sufficiently accountable to the societies in which they operate for their environmental and social impacts (O’Rourke, 2003). Each of these alternative perspectives has thereby studied individual performance implications of the activism phenomenon, without examining the mechanisms underlying its different aspects. Taking a multidimensional construct of shareholder activism, however, can help clarify some of the complexity of the activism-performance relationship. The multidimensional construct that this thesis has taken, consisting of executive compensation-related, corporate governance-related, and social/environmental-related types of proposals, as well as the two-way distinction of firm performance in financial and social performance, thereby brings together mechanisms from agency theory, institutional theory, managerial entrenchment theory, and business relationship theory.

Focusing on the first category of shareholder proposals, executive compensation-related proposals, the missing and even negative significant explanatory effects found in this thesis for this type of proposal are in stark contrast with this study’s predictions, but in line with the results of some other studies (e.g., Cai & Walkling, 2011; Larcker, Ormazabal, & Taylor, 2011). Where this study namely theorized a positive relationship between shareholder activism via executive compensation-related proposals on firm financial performance, the opposite is found. An explanation for these results might be that the filing and voting on executive compensation-related shareholder proposals do not lead to the effective monitoring of compensation, but that these proposals rather are an intrusive measure which undermine the board’s authority. The consequence of undermining the board’s freedom to decide may, in