Moderating effects of an underdeveloped judicial context on
firm level governance mechanisms:
An empirical study of India and the United States
Author: M. Meyer
maarten.meyer@yahoo.com
RUG: 1925482 - NCL: 110165073 Based at the Rijksuniversiteit Groningen
Faculty of Economic & Business:
Department of International Business & Management
Supervisor NL: Prof. A.A.J. van Hoorn Supervisor UK: Prof. C. Carter
5
thDecember 2011
Purpose – Examining the moderating effect of an underdeveloped judicial efficient context on the ability to discount cost of equity with board independence and monitoring abilities
Design/methodology/approach - It is proposed that governance mechanisms have an effect on COE. With dissimilar institutional contexts, firm-level governance is expected to have different values to create the most optimal governance mechanism in discounting or attracting equity capital. The independent variables are regressed upon dividend yield and price-to-book value that proxy the COE to examine the effect of board independence and monitoring abilities on COE in dissimilar judicial contexts.
Findings – Board independence from management discounts COE in four out of seven dimensions. For a board’s monitoring abilities there is no empirical evidence that it has a significant effect in discounting COE. Concerning the size of a board a significant moderating effect has been documented for the context of India. This board characteristic has a weakening effect on COE for a firm in India.
Originality/value – The paper contributes to current academic literature by providing a better understanding of what role the board of directors has for firms in institutional contexts with low judicial efficiencies (i.e. India) on discounting COE compared to firms in institutional contexts that have greater judicial efficiencies. From a practical point of view, this paper provides an examination on the magnitude of defence that the board of directors provide to shareholders in an environment with lacking shareholder protection.
Keywords – Cost of equity, judicial efficiency, shareholder protection, board of directors, firm-level governance. Paper type – Empirical Research
1
TABLE OF CONTENT
1. INTRODUCTION ... 2 2. LITERATURE REVIEW ... 7 2.1 Cost of Capital ... 7 2.1.1 Equity Capital ... 8 2.1.2 Agency Conflicts ... 92.1.3 Cost of Equity Capital ... 10
2.1.4 Corporate governance mechanisms ... 11
2.2 Low judicial efficiency and the supervisory board ... 15
2.2.1 The need for firm-level governance reforms ... 15
2.2.2 Complementary firm-level governance mechanisms ... 16
2.3 Board independence and monitoring abilities ... 18
2.3.1 (in)Dependence ... 18
2.3.2 Board independence from management ... 19
2.3.3 Board independence and cost of equity... 23
2.3.4 Supervisory board monitoring abilities ... 24
3. DATA & METHOD ... 27
3.1 Sample ... 27
3.2 Data operationalization ... 28
3.3 Descriptive statistics ... 28
3.4 Method ... 31
3.4.1 Model Specification Hypotheses A ... 31
3.4.2 Model Specification Hypotheses B ... 32
4. RESULTS ... 35 5. Discussion ... 51 6 References ... 57 6.1 Journal Articles. ... 57 6.2 Books ... 64 6.3 Websites ... 64
Appendix A: Data operationalization ... 65
APPENDIX B: Formal Shareholder Protection in India and the US ... 73
APPENDIX C: Graphical representation moderating effect India - US ... 76
2
1. INTRODUCTION
With the increasing economic- roles and growth of the four leading emerging markets, and notwithstanding the financial crisis of 2008 and the current Euro-crisis, world stock markets concentrate more and more on Brazil, Russia, India and China (BRIC). These markets compose a disproportionately large amount of global equity trading (Demirgui-Kunt & Levine, 1996). It is evident that globalization has gained increased attention over recent decades (Kose et. al 2003) and due to increasing trade and financial intergration of the world’s economy not merely developed market economies are the centre of attention for both academics and practitioners (Stulz, 1999). Moreover, BRIC markets “play increasingly important roles in global economic development and for the global monetary and financial systems” (Jensen & Larsen, 2004, p. 39). Due to their high economic growth, the gap between prosperity with more mature markets such as the US has diluted over the past decades and is expected to narrow even more in the future. Therefore, the role that BRIC countries have in the world economy will continue to grow larger (Jensen & Larsen, 2004).
However each BRIC country liberalise its markets gradually for (foreign) investors, evolution of such liberalisation policies are rather distinct1. “The distinctive characteristic of equity flows into India, has led to a relatively high level of portfolio equity financing” (Lane & Schmukler. 2007, p.2). This statement is furthermore strengthened by Samal (1997) and the IMF (1995), who both claim that in the period 1991-1996 India accounted for the largest number of equity issues of all emerging markets.
India’s specific institutional policies foster and encourage foreign equity providers to invest equity in India. Subsequent to the financial crisis in 1990s, the Indian government has implemented a broad series of financial institutional reforms to foster economic growth and foreign investments (Samal, 1997; Lane & Schmukler, 2007). Especially policy reforms concerning foreign equity inflows have been liberalized in India to greater degree than any other emerging market (Lane & Schmukler, 2007). Consequently, this evolution has led to improved governance practices in India and encouraged investments by foreign shareholders to provide equity capital in India.
1 For instance, China’s gradualist approach to financial liberalization has led to a surge of foreign direct investment (FDI) by investors and the
3
However formal governance mechanisms in India have improved significantly over the past decades, lots of scepticism still exists. The Indian institutional context, concerning shareholder protection law and regulation, is “de jure” the most developed out of the four emerging markets, but “de facto” significantly behind markets such as the US (Chakrabarti, 2008). Therefore, though formal law and regulations are present, enforcement of these laws is lacking extensively which dilutes the value of having such an institutional context in the first place.
Such formal shareholder protections laws, regulation and its degree of enforcement have effect on expropriations of shareholders and their incentive to monitor executive behaviour (Burkart & Panunzi, 2005) because they limit the ability of self-dealing by the executive body (La Porta et al, 2000). Such risk of expropriation has been evidential in the past and large corporate scandals (e.g. Enron; WorldCom) have not gone unnoticed (Cunningham & Harris, 2006; Hoffman & Rowe, 2007). To a certain extent it can be said that the institutional context has failed significantly to counter perverse behaviour by Machiavellian executives (Adler, 2007). But more important is failure of “the ultimate governing body of any corporation: the board of directors” (Hoffman & Rowe, 2007, p.553).
“The board of directors is the shareholder’s first line of protection” (Ramly & Rashid, 2010, pp.2199) and for this reason characteristics of the supervisory board are influential on the magnitude of protection that is provided to the shareholder on firm-level (Milstein & MacAvoy, 1998; Ramly & Rashid, 2010). Such firm-level governance has a positive effect on operating performance (Klapper & Love, 2003), firm valuation (Gompers et al., 2002) and accordingly the availability and cost of equity (COE) capital for a firm (La Porta et al. 1998).
4
Problem Statement
The current subject of debate is not whether there is potential in BRIC countries, but more how this potential is recognized (Jensen & Larsen, 2004). And due to distinct evolution of economic development each BRIC country has different potentials (Jensen & Larsen, 2004; Lane & Schmukler, 2007). Another subject of debate that has gained importance over the past decade is corporate governance (e.g. Ramly & Rashid, 2010). As aforementioned, various corporate scandals created the urge to improve firm- and country-level governance to counter unethical behaviour by corporate insiders (Adler, 2007). Particularly the focus on firm-level governance mechanisms has gained attention, since in case of firm failure the supervisory board has most influence of directing and disciplining the executive body (Hoffman & Rowe, 2007; Ramly & Rashid, 2010). “The primary reason for corporate governance is the separation of ownership, control and the agency problems it engenders” (John & Senbet,1998, p.372). Schleifer and Visney (1997, p.737) define corporate governance as “the ways in which the suppliers of finance to corporations assure themselves of getting a return on their investment”. Especially if formal country-level governance mechanisms (or their enforcement) are lacking, the role of the supervisory board becomes more important (Black, 2001). Despite the presence or lack of formal governance mechanisms or economic development, institutional contexts furthermore influence
firm-level governance specific2. The evolution of India’s economic environment and institutional
context created a rather liberal and open equity market for domestic and foreign investors (Lane & Schmukler, 2007). Even though formal shareholder protection is to a certain extent present, enforcement of these laws is lacking severely (Chakrabarti, 2008). This increases the likelihood and ability of the executive body to pursue own personal objectives and not the objective(s) of the shareholder. Such alignment disruptions are well explained with the agency theory or agency problem (Berle & Means, 1932; Jensen & Meckling, 1976; Eisenhardt, 1989).
With the separation of ownership and control, the supplier of capital has no direct control over its finance. The shareholder’s (principal) limited ability to influence the decision making process of management (agent) makes the cost and availability of capital for a firm challenging (Ramly & Rashid, 2010). Jensen and Meckling (1976, pp.388) define the agency relationship as: “a contract under which one or more principals engage another person to perform a service on
2
5
their behalf which involves delegating decision making authority to the agent”. Alignment disruptions occur for the reason that the agent doesn’t behave as he were to maximize the principals’ welfare (Jensen & Meckling, 1976). Thus, the principal’s empowerment in this relationship is a subject of interest to understand and counter the degree of alignment disruptions within the firm.
Such disruptions can with great extent be tackled with external, formal governance mechanisms which are less present in India than in the United States (Reese & Weisbach, 2002; Hail & Leuz, 2005). A developed institutional context not only acts in favour of the principal: by limiting expropriation, “the law raises the price that securities fetch in the market place” (La Porta et al., 1999, pp. 3), which in turn discount COE for a firm. Consequently, shareholder protection rights and the degree of enforcement are inherent with the risk a shareholder bears with its investment and therefore affects COE. Therefore, the link between law and finance is evident (Gompers, et al, 2002; Klapper & Love, 2003). Especially if formal law is lacking, various firm-level governance mechanisms have greater impact on stock price performance and firm valuation (Mitton, 2002). John and Senbet (1998) also argue that there is a significant interaction between external governance mechanisms and internal control mechanisms. They found that if external mechanisms are present internal governance mechanisms are to a lesser extent necessary and present. Therefore, interactions between external and internal governance mechanisms are a substitution for one another (John & Senbet, 1998) although firms can never completely substitute a weak legal system internally (Klapper & Love 2004).
6
However in India, which has no highly developed judicial context, the role of the supervisory board and its characteristics become increasingly important for discounting COE capital (Klapper & Love, 2003). This is furthermore proposed by Black (2001, p.20) who studied 21 firms’ corporate governance characteristics in relation with a firm’s valuation in Russia. He suggests that “firm-level governance behaviour has a huge positive effect on market value”.
Consequently, this postulates that there is variation in firm-level governance characteristics and the value of these characteristics among any market economy to moderate the weaknesses of the institutional context. Although such an interaction has been found on market level (John & Senbet, 1998) and countries (Joh, 2003), this relationship is still largely unexplored. For instance, Klapper and Love (2003, pp.704) argue that “the relationship between the country-level legal infrastructure and firm-level governance mechanisms is far from obvious”. Ramly & Rashid (2010, pp.2200) furthermore argue that “the value of the board is still a subject of debate” especially in emerging markets.
This paper aims to determine the value of specific board characteristics, which designate the principal’s empowerment to discipline and control the agent’s behaviour, in discounting COE in distinct judicial efficient contexts. Stated differently, what board characteristics create the most empowerment for the principal in India and the US and to what degree has this effect upon COE. Consequently the following research question is proposed:
Do similar supervisory board characteristics discount the cost of equity capital for firms in two distinct institutional contexts differently, and if so, how?
To provide an answer to this research question, first an extensive literature review is presented to answer the following sub questions:
1. How does firm- and country-level corporate governance affect the COE for a firm?
2. What board characteristics are regarded to have the most value in affecting firm-level governance if judicial efficiency is lacking?
3. How do these board characteristics lower COE for a firm in general?
Subsequent to the literature review the proposed relationships are tested empirically:
7
2. LITERATURE REVIEW
The literature review follows the order of the sub-questions presented above. First, it aims determine the theoretical relationship of sound governance mechanisms and its effect on discounting COE. This section services as an introduction to the second sub-question. Second, the literature review identifies specific board characteristics that have additional value in underdeveloped judicial efficient contexts. Third, when the most valued board characteristics are identified for countries with a weak legal infrastructure, these characteristics are examined in detail. The interest is on how these characteristics interact in different institutional contexts (i.e. India and the US), what their value is and whether the value of such a variable is dissimilar in both contexts. Furthermore this section provides hypotheses that are tested empirically. The fourth, and final, sub-question is answered through the use of an empirical research and testing the proposed hypotheses will lead to an answer on the main research question.
2.1 Cost of Capital
8
2.1.1 Equity Capital
The following section will examine the third component of a firm’s WACC, external financing by means of equity capital. Equity capital is raised with the use of shareholders and these are granted residual rights of disciplining a firm’s executive body. Though effective, control over their capital is transferred indirectly to the shareholder with the use of a board of directors. In general a shareholder, or any supplier of capital, is risk averse (Hakansson, 1969; Stulz, 1999). Consequently, the cost that is incurred with raising equity capital is dependent upon the risk a provider of equity has to bear and tolerates (Ramly, 2009).
9
2.1.2 Agency Conflicts
The agency theory comprises, in general, three parties; executive(s) or agent, a shareholder(s) or principal and creditors. All of these parties have different attitudes towards risk and therefore have different goals and division of labour (Jensen & Meckling, 1976; Eisenhardt, 1989). The theory uses a contract as a metaphor to distinguish the conflicts that may present themselves in an organisation (Jensen & Meckling, 1976). The agency theory portrays three contractual conflicts, agent/principal; controlling/minority; and creditor/principal (e.g. Heeremans, 2007; La Porta et al., 1999; Jensen & Meckling, 1976; Easterbrook, 2011) and the degree of conflict is distinct among organisations, industries and countries3. This paper will focus on the type 1 agency conflict: alignment disruption between the principal and agent (Shleifer & Vishny, 1997).
Alignment disruption concerns two problems (Eisenhardt, 1989). First, the agent and principal have different objectives, and due to information asymmetry the principal is limited in its ability to control and value the agents’ behaviour. Second, different attitudes towards risk create alignment disruptions between the agent and principal. The main component of the theory is the contractual efficiency that governs the relationship between principal and agent (Eisenhardt, 1989). Two main streams4 are developed within the theory and both focus on this contractual relationship (Jensen, 1983). This paper takes on a positivists’ view on the agency theory by examining governance mechanisms (firm- and country level) in relation with COE.
On one hand, the principal designates assets to a firm and in return receives residual rights and some effective control. On the other hand, the agent has control over a firm’s assets, but no ownership. The “pursuit of self-interest at the individual level and goal conflict at the organizational level” (Eisenhardt, 1989, p.63) is the main issue in contractual conflicts because the agent’s behaviour is not merely focused on the principal’s wealth maximization (Jensen & Meckling, 1976). In general, the agent’s effort will never meet the value of the agent’s contribution to the firm and its owners: executives earn less than the full return of their effort
3
For instance, the creditor/principal conflict is more pronounced in financial industries or in countries such as Germany or Japan (Shleifer & Vishny, 1996) than in the US which has a dissimilar institutional context.
4 A positivist view concerns alignment disruption between agent and principal and focuses on governance mechanisms to counter unsound
10
(Heeremans, 2007). Consequently, the agent’s incentive to act in self-interest dilutes the firm’s ultimate goal, shareholder’s wealth maximization (Shleifer & Vishny, 1996). Organizations may have incentives (e.g. stock option plans) for the agent to promote the pursuit of shareholder’s wealth maximization. But such incentives are rather a reflection of the principal-agent’s power relationship than the solution to the problem. Such incentives are a manifest of the imbalance of power between the agent and the principal (Bebchuk & Fried, 2003).
2.1.3 Cost of Equity Capital
The aforementioned paragraph discussed agency conflicts that are present in most publicly traded firms with focus on the imbalance of power and alignment disruption between the principal and agent. As a result, the principal is less inclined to provide equity if those conflicts are more pronounced. Increased probabilities of such conflicts create more risk in getting a full return on investment. And for the reason that the principal, or any supplier of capital, is risk averse (Hakansson, 1969) this risk must be compensated with higher returns on investment (Stulz, 1999). Such a rational requirement, to compensate risk with higher rates of return, consequently increases COE (Ramly, 2009). If the likelihood of such agency conflicts is limited, the principal will reduce its expected return and a firm induces lower cost for the use of a principal’s assets. An elaborate discussion on how COE must be perceived and measured is provided in the first section of Appendix A.
11
2.1.4 Corporate governance mechanisms
In its broadest sense, corporate governance refers to a complementary set of legal, economic, and social institutions that protect the interests of a corporation’s owners (Ahmadjian, 2000, p.5). The quality, reach and enforcement of this set of attributes are influential upon firm performance and valuation, for the reason that corporate governance minimizes agency conflicts and costs (Eisenhardt, 1989). As aforementioned, Schleifer and Visney (1997, p.737) define corporate governance as “the ways in which the suppliers of finance to corporations assure themselves of getting a return on their investment”. In general, the principal, who only cares about its expected rate of return and the volatility in these returns (Stulz, 1999), is more likely to provide capital and expects lower rates of return if the corporation is governed in a professional manner (Milstein & MacAvoy, 1998). Klapper and Love (2003) argue that improved corporate governance is highly correlated with better operating performance, market valuation and reduced firm-level specific risks. La Porta et al. (1999) found that greater investor protection increases the availability of equity capital for a firm and this furthermore would be reflected in a discounted COE. Such protection on country- (La Porta et al., 1999) and/or on firm-level (Black, 2001) reduces agency costs/conflict, COE (Klapper & Love, 2003) and improve firm performance significantly (Gompers et al., 2002).
Thus, governance mechanisms can be on country-level, for example with proper shareholder protection rights and regulation, or on firm-level, for example as with a board of directors. Nonetheless, corporate governance secures that the objectives of the principals and a firm’s executive body are more aligned than when such governance is lacking (Baysinger & Butler, 1985).
Country-level governance: Judicial efficiency
The focal issue in this paper concerns whether formal shareholder protection laws and regulations affect the value of specific firm-level governance mechanisms. For this reason
certain governance mechanisms5 are neglected. For an overview of specific governance
mechanisms that reduce agency conflicts/costs, please refer to Shleifer and Visney (1997) for a general overview or to Bushman and Smith (2001) for the topic of governance in relation with financial accounting.
12
Judicial efficiency is a broad term that comprises in general two aspects: (1) legal efficacy; and (2) law enforcement (Klapper & Love, 2008). Thus additional to the content, enforcement of these laws is of importance. Sarkar (2007) has examined the content of law in several least developed countries (LDCs) and concludes that “the rule of law is at a very miserable stage because of corruption and other imperfections”. Accordingly, the level of law regulation and enforcement has not been developed as abundant as it is in developed market economies such as the US. Frumin (2009, p.21) furthermore examines the lack of law enforcement in India and other developing markets and argues that “the enforcement of these (shareholder protection) laws is severely lacking and rather challenging”. This is furthermore argued by Frumin (2009), is it not necessary the lack of shareholder protection itself in developing economies that places a premium on the COE for a firm, but rather the lack of enforcement of these protection laws.
Next, the influence of judicial efficiency upon firm valuation and performance is examined and additionally its influence upon firm-level governance.
Judicial efficiency and firm valuation
Judicial efficiency or economy concerns formal protection and regulation and the effectiveness of enforcement. It “refers to efficiency in the operation of the courts and the judicial system. It is the efficient management of litigation so as to minimize duplication of effort” (USLegal, 2010). Usage of the term is diverse in various scholars, but in this paper the term concerns formal shareholder protection rights, regulation and enforcement. More specifically, the
methodology6 of Djankov et al. (2008) is used to examine the judicial efficiency of shareholder
protection in the Indian’s and US’s contexts. Please refer to Appendix B for a concise description Djankov et al.’s (2008) methodology in the context of India and the US.
In essence, four indices are created to indicate potential expropriation and alignment disruptions within a certain institution context (Djankov et al., 2008). The presence of judicial efficiency minimizes agency conflicts and consequently reduces agency costs (Hail & Leuz, 2005). In doing this, “the law raises the price that securities fetch in the market place” (La Porta et al., 1999, pp. 3), this in turn discounts the COE for a firm. Thus, shareholder protection rights, regulation and the degree of enforcement are inherent with the risk a shareholder bears with its
13
investment and therefore COE a firm bears for its external financing (La Porta, 1999). Consequently, a positive relationship between law and finance is evident (Gompers, et al, 2002; Klapper & Love, 2003). Though this relationship is evident, the value of each index component in discounting COE is distinct in each institutional context. For example, the extent of disclosure as a country level governance mechanism has more positive effect on firm valuation in countries such as Russia (Black, 2001) where managerial ownership as governance mechanism prevails
(Klapper & Love, 2004).
Judicial efficiency and firm-level governance
14
Firm-level governance: the supervisory board
The board’s function is to minimize transaction, or agency, costs that are accompanied with separation of ownership and control. “The board of directors, which has the power to hire, fire, and compensate senior management teams, serves to resolve conflicts of interest among decision makers and residual risk bearers” (Baysinger & Butler, 1985).
Ramly and Rashid (2010) even argue that the board of directors is the first line of defence for the principal and has the most direct effect in reducing COE due to improved corporate governance (Jensen, 1993; Gande, 2009). In the context of corporate governance literature, the primary related issues that have gained interest amid academics are board size and board composition’s effect on firm valuation and performance (e.g. John & Senbet, 1998; Ramly & Rashid, 2010). However board size is apparent in its meaning (i.e. number of directors), the definition of board composition needs more consideration.
15
2.2 Low judicial efficiency and the supervisory board
As aforementioned, firms can moderate the effect of a weak institutional context on COE by improving firm-level governance mechanisms. Although low judicial efficiency cannot completely be substituted by internal governance mechanisms (Klapper & Love, 2004) certain mechanisms have greater value in moderating a low judicial efficiency context. Due to merely rational and economical efficient reasons, firms in such contexts don’t have the flexibility to substitute or moderate their internal governance mechanisms with alternative mechanisms to counter balance the lacking institutional environment (Klapper & Lover, 2004). Notwithstanding, certain more universal/conventional internal governance mechanisms have more effect in discounting COE.
The following section will identify and describe the internal governance mechanisms that have greater value in institutional contexts with low judicial efficiencies in discounting COE for a firm. First, the distinct evolution of shareholder protection in India after the liberalization of its financial market is examined briefly. Second, academic literature is used to identify which firm-level governance mechanisms are ought to discount COE for a firm the most in a weak judicial efficient context.
2.2.1 The need for firm-level governance reforms
The 1990’s financial crisis in Asia and liberalization of various Asian financial markets have led to several reforms in corporate governance. The legal reforms and India’s economy are developing in rapid pace, but the magnitude of judicial efficiency is still lacking behind India’s economic development (Levine, 1998). Subsequent reforms in 2004 (i.e. Clause 49) focused specifically on firm-level governance. However, due to a lacking degree of enforcement, the judicial efficiency remained rather similar as prior to the implementation of Clause 49. These
reforms7 indicated the need for additional firm-level governance mechanisms that are ought to
have a profound effect in improving overall corporate governance. Chakrabarti (2008, p.64)
report argues that “thecomposition and proper functioning of the board of directors was one of
the key areas of focus”. Especially the focus of a director’s individual profile concerning its
7 The following five key mandatory reforms are stated: (1) composition of the board of directors; (2) composition and functioning of audit
16
independence, frequency of board meetings, remuneration and dictates certain disclosures for supervisory board members was amended in the clause (Chakrabarti, 2008).
Whether these reforms had a positive impact on firm-level governance mechanisms and reduced agency costs is not evident. Although the reforms aim to improve governance, scepticism is present due to the lack of enforcement of these legal requirements (Chakrabarti, 2008).
2.2.2 Complementary firm-level governance mechanisms
As aforementioned, firms in weak judicial efficient contexts don't have the flexibility to counterbalance their external environment with alternative internal governance mechanisms (Klapper & Love, 2004). Klapper and Love (2004, p.712) argue that when “enforcement of contracts is weak; firms are unable to “overwrite” their country’s legal system and therefore the lack the flexibility to improve governance”. But this does not imply the “de jure” of a legal context. On the contrary, if “de facto” is lacking, several observations are made in the existing, and rather limited, body of academic literature that prevent unlawful practices by corporate insiders in the first place, with the use of conventional governance mechanisms used in more developed market economies such as the US.
For instance, Mitton (2004) states that Russian firms are more commonly electing independent directors although the legal context does not imply, define or enforce board independence. In addition, Black (2001) furthermore examined 21 Russian firms and found that greater board independence is positively associated with firm performance, whereas in the US this relationship has not been recognized (e.g. Bhagat & Black, 1999) to the extent as Black (2001) shows in Russia. These dissimilar results are due to the circumstances that agency costs are already preceded and accounted for by the more developed judicial efficiency in the US than in Russia.
17
and managerial power”. Extending these results implies that, and in accordance with various other studies such as Dechow et al. (1996), board independence generates higher qualitative financial information/disclosure, discounts agency costs and reduces COE for a firm. This is also found by Chen et al. (2003). These authors argue that both disclosure governance mechanisms and non-disclosure governance mechanism discount COE for a firm in markets with low, or moderate, judicial efficiencies. Moreover, they found that non-disclosure firm-level governance mechanisms have the most effect in discounting COE in such an institutional context. “Reducing the expropriation risk by strengthening the overall corporate governance mechanisms appears to be more important than adopting a more forthright disclosure policy to reduce the cost of equity capital” (Chen et al., 2003, p.28). Kaymak and Bektas (2008) also test this proposition; however this effect is not that strong. Their specific focus on banks has implications for their results because the opaque nature of the banking industry places a burden on the monitoring ability and judgment of (unaffiliated) outside directors because they lack specific knowledge.
18
directors increases firm valuation. “Firms that have more exacerbated agency conflicts tend to incorporate professional directors to the boards, in an effort to improve corporate governance and ameliorate the agency problem” (Lefort & Urzua, 2008, p.621).
Although the existing body of (empirical) research concerning the subject of this paper is rather limited, an apparent disposition towards board independence and a board’s monitoring abilities has been documented.
2.3 Board independence and monitoring abilities
The following section will describe how the specific board characteristics defined in the prior paragraph are hypothetically constructed to have dissimilar values in discounting COE for a firm in distinct institutional contexts. First, a brief introduction of what (in)dependence actually means in the context of this paper and what types of independence are inherent with agency conflicts. Second, monitoring abilities by the board are described, similar as to which the proxy independence is examined. Third, hypotheses are constructed with the aim to empirically test the main research question in this paper.
2.3.1 (in)Dependence
First, it’s important to critically assess the usage and the definition of independence. The definition of the word, and its usage by stakeholders, postulates a certain degree of deception prescribed by its inherent agency conflicts. A deception that directors serve to control and monitor management in interest of shareholders, creditor or company (Ramly & Rashid, 2010) but not necessarily behaves in this manner suggests the inherent dependent nature of a director.
19
It is not in the scope of this paper to argue whether the term itself fulfils its purpose, but
the notion that independence has no fixed meaning8 in any contexts is important to understand.
Thus, according to the question -independence from whom or what- suggests that there are different types of director independence.
In general, there are three types of agency conflicts and in accordance with these conflicts there are three types of independence (Heeremans, 2007): independence of management (type I); controlling shareholders (type II); and financially independent (type III). This paper concerns the empowerment of the principal in reducing firm-specific risks and vice versa reducing COE for a firm. For this reason the following section will describe independence as a director that is independent of management. Such a director is capable to perform independent judgement of executive decision making and performance without being predisposed by the executive body. Such independency has in general a positive effect on firm-level governance (Kang et al., 2007; Ramly, 2009).
2.3.2 Board independence from management
As aforementioned, a board’s independence is not as straightforward as the definition postulates. For this reason various board characteristics that indicate the whole board’s independence from management are examined next. It follows various academic studies (e.g. Rosenstein, 1990; Ramly & Rashid, 2010) which point out the most common characteristics that affect a board’s independence from management. The following paragraph describes the following board characteristics that comprise board independence from management and derived from an extensive literature review: (1) board structure; (2) board size; (3) inside-outside directorships; (4) chairperson specifics; (5) director turnover; and (6) remuneration committee specifics. Subsequent, board independence and its relation with COE is asserted to provide the hypotheses that will be tested empirically.
8
20
Board structure
Gillette et al. (2008) found that a dual-tier board design is associated with more shareholder focussed decision making, thus a higher degree of type 1 independence. In this design, the board provides a better supervisory and control function over managerial behaviour. In turn, a single-tier board structure is associated with more opportunistic and managerial focussed behaviour. A single-tier board is more likely to be influenced by management and therefore provides a less independent audit upon management’s actions.
Board size
In a small board decisions can be more easily coordinated by the board and is therefore more difficult to control by management. A large board is easier to control by management; it creates a more conceivable chance that directors are influenced by management (Belkhir, 2009). A large board needs more time to come to a decision and therefore negatively affects the efficiency of a board’s duty. Larger board size increases the agent’s power and will ultimately
lead to underperformance. Ceteris paribus, smaller boards have higher firm valuation9 (Yermeck,
1995). Moreover various studies (e.g. Kini et al., 1995) show that “reducing board size has become a main concern for investors” (Yermack, 1996, pp.186). Consequently, an augmented board size negatively affects firm-level governance and a reduction in board size positively affects firm-level governance.
Inside-outside directorships
Two types of directors are profound to describe the magnitude of board independence and sound judgement. Inside, or executive, directors serve an internal function within the firm. These directors are considered prejudiced and not able to provide unbiased judgement, due to the fiduciary nature of their position (Weisbach, 1988). Outside directors are perceived to judge and control the CEO more rationally than inside directors would (Weisbach, 1988). Choi et al. (2007, pp.941) argue that “inside directors may not feel compelled to contradict the CEO, outside directors are in a better position to monitor managerial activities”. Choi et al.’s (2007) also confirm that in a country with weak shareholder protection, a low ratio of inside/outside directors is beneficial for firm valuation. Xie et al. (2003) argue that outside directors are important to
9
21
handle agency issues and inside directors may best use their knowledge and expertise for firm-specific issues.
Supervisory board chairperson specifics
The chairperson of the board is regarded as the most influential director of the board (Brickley et al., 1997). The chairperson provides leadership to executives and officers and ensures that the shareholder’s objectives are achieved. Coles and Hesterly (2000) argue that such a leadership structure, where the functions of board chairman and CEO are combined, has severe implications on a board’s independence, judgment of executive’s behaviour and shareholder value. In general, the chairperson acts as an intermediary between upper echelon management and the board. Therefore, characteristics of the chairperson have profound influence upon the board’s state of affairs (Chitayat, 1985; Pi & Timme, 1993). For more information regarding the role of a chairperson, the CEO and its power relations within the supervisory board please refer to Chitayat (1985).
A chairperson that furthermore holds the position of chief executive officer (CEO) has a fiduciary duty. “CEO duality exists when a firm's CEO also serves as the chairperson in the board of directors. Holding the highly symbolic position of board chair would provide the CEO with a wider power base and locus of control" (Boyd, 1994, p.338). Its judgments may be biased and the “executive body has de facto control” (Brickley et al., 1997, p.190). Hence, this will negatively affect COE for a firm (Pi & Timme, 1993).
Director turnover
22
Additionally, director turnover is related to poor prior firm performance (Hermalin & Weisbach, 1988; Gilson, 1990). This indicates that firms are willing to change their board composition to increase firm performance. And if performance is satisfactory, board composition will remain similar. Accordingly, (excessive) director turnover indicates improvement of firm-level governance.
Remuneration committee specifics
23
2.3.3 Board independence and cost of equity
In essence, board independence from management decreases the likelihood of shareholder expropriation and opportunistic behaviour by management (Ashbaugh et al., 2004). The role of the supervisory board is to “provide independent oversight of management and hold management accountable to shareholders for its actions” (Asbaugh et al., 2004, p.12). Ashbaugh et al. (2004) also found that the degree of board independence is negatively correlated with COE for a firm. Thus, board independence discounts equity costs for firms. Ramly (2009) argues that a high proportion of independent directors should be present on a board to experience a positive effect upon firm-level governance. This is furthermore confirmed by Kang (2007) who studied Australian firms, Ashbaugh et al. (2004) who studied changes in governance and consequently the changes in equity costs, or for instance Black (2001) who studied governance mechanism in Russia10.
In judicial developed market economies, firm-level variations in governance mechanisms have small effects on firm valuation (Black, 2001). In contrast, for Russian firms this has major effects on firm valuation. This postulates that firms in low judicial efficient market economies can “greatly improve their own share values, and thus reduce the cost of raising equity capital, through a determined effort to improve their corporate governance practices” (Black, 2001, p.20).
Consequently the following hypotheses are developed:
Hypothesis A: A higher degree of board independence consequently discounts the cost of equity capital for a firm in any judicial context.
Hypothesis B: In an underdeveloped judicial context board independence has a greater effect in discounting the cost of equity capital for a firm than it has in developed judicial efficient contexts.
10 Some authors find no significant relation (e.g. Baysinger & Butler, 1985) or even an inverse relationship (Agrawal & Knoeber, 1996). Their
24
2.3.4 Supervisory board monitoring abilities
Monitoring by the board of directors serves an internal governance mechanism that is ought to discount agency costs. Additionally, specialist audit firms monitor the firm at least annually and serve as an external monitoring device. Derived from the following review, the quality of monitoring is affected by: (1) a firm’s board structure; (2) the amount of corporate insiders on a board; (3) board size; and (4) quality of external auditing. First, monitoring as internal governance mechanism is argued upon and second monitoring as external governance mechanism is stated.
Internal monitoring
The board significantly contributes to the integrity of financial statements (Peasnell et al., 2005). It is widely believed that the supervisory board plays an important role in governance practices with particular focus on monitoring the executive body (Fama & Frenh, 1983). From an agency perspective, the supervisory board is used as a device to monitor management on behalf of principals (Fama & French, 1983; Eisenhardt, 1989). When boards provide richer information, incurred agency costs derived from the separation of ownership and control, decrease (Eisenhardt, 1989). Although there are many mechanisms to improve monitoring
abilities, such as incentive based compensation contracts (Eisenhardt, 1989), most literature11
considers the role of outside/inside directors in improving monitoring abilities (Bryd & Hickman, 1992). “Inside director provide valuable information concerning the firm’s activities, while outside directors contribute both expertise and objectivity in evaluating the management’s decisions” (Bryd & Hickman, 1992, p.196).
Overall, enhanced board monitoring creates more rich information for the principal and likewise reduces agency costs and COE for a firm. Although the relation with firm performance on COE is not directly apparent, by taking the perspective of incurred agency costs, enhanced monitoring abilities will significantly increase the board’s function in providing proper corporate governance on behalf of the principal.
Moreover, and contrary to what is argued in the prior paragraph, various author (e.g. Anderson et al., 2004; Adam & Mehran, 2003) argue that a larger board is able to provide more sound judgement and therefore creates better governance. For example, Anderson et al. (2004)
25
state that a larger board is more effective in monitoring the executive’s behaviour and workload can be spread more effectively among the directors. Cheng et al. (2007) argue that a large board is better able to judge executive behaviour due to the increased expertise on the board. Furthermore, increased board diversity is moreover argued to have a positive effect on corporate governance and consequently on COE. This argument is also known as the resource dependency theory. The focal issue here is that, a larger board creates a larger stock of human and social capital. Directors that complement each other with expertise should therefore be more able to provide sound judgement. Thus, a larger board could imply better monitoring abilities and therefore a lower COE, however theoretically this would reduce board independence.
External monitoring
In addition to internal monitoring mechanisms, external monitoring furthermore has a positive effect on reducing agency costs (Fan & Wong, 2002). If firms in emerging markets hire one of the big five auditors, (i.e. from 2002 four auditors: (1) Price Waterhouse Coopers; (2) Deloitte and Touche Tohmatsu; (3) Ernst & Young; and (4) KPMG) this will have a negative mitigate implied agency costs and COE.
Because of their international reputation these auditors are perceived to be more independent and provide better quality assurance than local auditors this will decrease the principal’s agency costs (Fan & Wong, 2002) and consequently COE incurred for a firm. Furthermore it is likely that the costs involved with the audit of such a big four auditor are greater than with for instance a local auditor. This indicates the willingness of a firm to sustain or to improve corporate governance. Chow and Wong-Boren (1987) also found that firms who project higher agency costs have a greater prospect to hire an external auditor.
26
preferred for internal control purposes if they are considered less likely to collude with the manager's subordinates” (Chow, 1982, p.227). Thus, accordingly this states the significance of independent judgement and monitoring abilities of a supervisory board in reducing agency conflicts, costs and COE.
Consequently the following hypotheses are developed and are similar to the prior stated hypotheses:
Hypothesis A: A higher degree of monitoring abilities discounts the cost of equity capital for a firm due to less implied agency costs.
27
3. DATA & METHOD
This chapter discusses the methods applied that test the hypotheses quantitatively. First, the sample is described and how the data is obtained. Second, operationalization of the variables is provided by means of table 3.1. Third, the descriptive statistics of the data are reported. Fourth, the model’s assumptions are discussed briefly to validate the use of the model. Finally, the method of research and specification of the regression models are discussed. Moreover, explanation of the model and consequently a brief discussion follows regarding how the results in the following chapter are interpreted.
3.1 Sample
This research will be performed as an empirical study. Two populations were chosen to compare and to examine the hypothesized interaction effects of board independence and monitoring abilities on COE in two distinct judicial efficient contexts. The two chosen sample populations, India and the US, must be consistent over time and both listed at similar exchanges (Fama & French, 1997). Therefore the BSE 100 (from 2011) and the S&P100 (from 2011) are
used respectively. By eliminating financials12 the sample constituted of 78 non-financials for the
BSE 100 and 87 non-financials for the S&P 100. Furthermore, it is preferred that the sample firms serve, to certain extent, similar markets (Aachen & Snidal, 1989). Moreover, the sample must be as complete as possible. This led to the exclusion of 79 additional companies. Consequently, the final sample constituted 42 BSE firms and 44 S&P firms. The timeline which this research covers is five years and prior to the 2008’s financial crisis to counter biased (volatility) results. Such a time period is also suggested by other studies (e.g. Fama & French, 1997) concerning COE to provide sound estimations. Hence, the chosen timeline is 2002-2007. Ultimately, the sample counted 430 firm-year observations. No available database covers Indian board profiles and the data that is available for US firms is not complete to the extent that is desirable for this research. Therefore, annual reports of Indian and US publicly traded firms are examined manually to create a comprehensive dataset. Furthermore, additional data was available in the DataStream and Orbis databases. Thus, secondary data is used for this research.
12
28
The date of publishing of each annual report13 and its financial year must be equal to the data that
is extracted from DataStream or Orbis. Moreover, the change in directorships in the year 2002 is examined with the use of annual reports preceding the sample period, thus 2001. All currency variables are stated in US dollar, expressions in thousands are altered into millions to ease
interpretation. Additionally, all data has been checked for defects and shortcomings14.
3.2 Data operationalization
Table 3.1 shows an overview in which the data is operationalized. Each variable is reported and its abbreviation that is used in the statistical processing software (PASW 18.0.3). Next, the source in which the data is found is presented. Furthermore, the variable category is shown and how the variable is computed or operationalized. In addition, the variables’ expected relationship with a particular dependent variable is articulated in the table. For an extensive discussion of how the variables are operationalized, please refer to Appendix A. The descriptive statistics of the data are provided subsequently in Table 3.2. Additionally, Table 3.3 and 3.4 show what effect the variables have as a proxy for board independence or monitoring abilities. It furthermore provides an indication which author(s) described these variables as a proxy and its theoretical effect on COE.
3.3 Descriptive statistics
The descriptive statistics of the data, separated according to their function in the regression model, are reported in Table 3.2. Subsequent, additional variables are computed for the regression that function as bilinear moderating variables to test the interaction effect that board independence/monitoring abilities have in a low judicial efficient context in discounting COE for firm. The table provides an overview of the data within the sample obtained from DataStream, Orbis and the annual reports of 42 Indian firms and 44 US firms over a period of 5 years. In addition, Appendix D. reports the Pearson correlation matrix to find dependence among the independent variables.
13 Especially in the context of India there is no fixed date of ending a financial year. For this reason the date of the annual general meeting (AGM)
is used to establish the proper year in DataStream and/or Orbis
14 The following variables have eliminated from this research because they were similar or not available for both sample populations:
29
Table 3.1 Data Operationalization
Variable Abr. Source Category Operationalization Expected Sign Dep. Var.
Dependent variables that proxy COE
Dividend Yield DY DataStream Interval Dividends / Current Share Price negative COE
Market to Book Value PB DataStream Interval (Book value assets - book value equity + market value equity - deferred taxes) / (book value assets)
negative COE
Independent variables that proxy independence
Board structure Tier AR Nominal 1 = Single-tier negative DYit
Board size BS AR Interval Number of board directors negative DYit
Non-executive directors on board
NED AR Interval (∑ NED) / BS positive DYit
Board chairperson BC AR Binary 1 = Executive Director negative DYit
ED Turnover TED AR Interval TED / Total turnover positive DYit
Chairperson of the Remuneration Committee
CRemm AR Nominal 1 = Non-executive director positive DYit
Presence of a
remuneration committee
PRemm AR Nominal 1 = Yes positive DYit
Independent variables that proxy monitoring abilities
Board structure Tier AR Nominal 1 = Single-tier positive PBit
Inside directors ED AR Interval (∑ED) / BS positive PBit
Board size BS AR Interval Number of board directors positive PBit
External Audit B4 AR Nominal 1 = Yes positive PBit
Independent control variables
Asset Composition AC BvD: Orbis Interval Fixed Capital / Total Sales
Growth Opportunity GO BvD: Orbis Interval (Sales t=2 / Sales t=1) * 100
Firm Size SizeLN BvD: Orbis Interval Natural logarithm of total sales
30
3.2 Descriptive Statistics
Variable N Minimum Maximum Mean Std. Deviation
Dependent Variables
Dividend yield 506 ,00 12,97 1,7783 1,55423
Price to book value 492 -60,76 86,49 4,7344 6,73812
Independent Variables
Board Structure 430 0 1 ,94 ,234
Board Size 428 4,00 21,00 11,5257 2,61804
Executive Directors on Board in % 428 ,00 60,00 20,5379 13,02168
Non-Executive Directors on Board in % 428 40,00 100,00 79,4823 13,03349
Executive Director Turnover 422 -8,00 11,00 ,2180 1,20563
Non-Executive Director Turnover 423 -9,00 10,00 ,6123 1,51955
Chairman is CEO 428 0 1 ,68 ,466
NED Chair of the Remuneration Committee 421 0 1,00 ,8599 ,34755
No Remuneration Committee Present 428 0 1,00 ,0935 ,29141
Big Four Audit Firm 430 ,0 1,00 ,6651 ,47250
India 516 ,0 1,00 ,4884 ,50035
Control Variables
Cross-listed in Another Country 430 ,00 1,00 ,3116 ,46370
Asset Composition 376 ,01 4,80 ,4336 ,53425
Sales Growth per Annum in % 305 -85,61 394,05 22,1499 39,69219
Firm Size as LN(turnover in Mil) 390 3,82 12,76 8,7179 1,91017
Table 3.3 Proxy effects on supervisory board independence
1. 1-tier boards show less board independence than dual-tier boards Gillette et al. (2008) 2. Higher executive board turnover increases board independence Erkens et al. (2009) 3. NED as chairperson of the remuneration comm. increases board independence Erkens et al. (2009) 4. A higher ratio NED directors on the board increases independence Choi et al., 2007
5. A non-executive as chairman of the board increases board independence Millstein & MacAvoy (1998)
6. Presence of a remuneration committee increases board independence Conyon&Peck (1998) 7. A lower board size will increase a board’s independence Erkens et al. (2009)
Table 3.3 Proxy effects on board monitoring abilities
31
3.4 Method
Following the considerations that derived from the literature review, two hypotheses are constructed, which are subsequently are divided into board independence and monitoring abilities. The following section will describe the multiple regression models that are used to quantitatively test these hypotheses.
3.4.1 Model Specification Hypotheses A
Next, the regression model for hypotheses 1.A and 2.A are specified. The two models that are presented concern an ordinary least squares (OLS) multivariate regression analysis, where the dependent variable is regressed upon the particular independent variables. The following model concerns the absolute sample, totalling 430 firm-year observations.
The following model is specified to test both the proxies ‘board independence’ and ‘board monitoring abilities’ interchangeably:
COEit = β0 + β1(Proxyit) + β2 (Cross-Listedit) + β3(Asset Compositionit) + β4(Growthit) +
β5(Firm Sizeit) + β6 (India )+ εit
In this equation, COEit refers to either ‘dividend yield’ or ‘price to book value’ of
company i in period t. Next, β0 refers to the intercept and is not fixed in this equation. It denotes
the constant in the regression model and has an interpretive value. Next, β1 refers to the
coefficient of the independent variable that is used to proxy15 either ‘board independence’ or ‘monitoring abilities’. Furthermore, β2 to 5 are firm-specific control variables and β6 denotes the
coefficient of the dummy variable ‘India’. In accordance with the dependent variable, ‘t’ refers to the period in which the observation is made, and ‘i’ concerns the company that is observed. Lastly, ε denotes the error term or noise. It refers to all other factors that have effect on the dependent variable other than the regressing independent variables.
This equation is used to quantitatively test the presence of a statistical significant relationship between the independent proxy independence/monitoring and the dependent proxy dividend yield/price to book value. This regression model is constructed to aid the next
32
regression model to test the presence of a board’s moderating effect in discounting COE in an underdeveloped judicial efficient context
3.4.2 Model Specification Hypotheses B
The following regression specification furthermore incorporates a bilinear moderating, or interaction, effect. That is, this model will test whether independence and monitoring abilities moderate the negative effect of a weak judicial efficient context. This moderating variable is a product of the interaction of an independent variable (Xit) with the regarding judicial efficient
context (Mit) in explaining the dependent variable (Yit). In its most simple for the equation is
denoted as:
Y = β0 + β1X1 + β2X2 + β3X1X2
Please note that additional bilinear moderating variables are computed for each particular
independent variable. This additional variable denotes the moderating effect of X2 on X1 and is
calculated by multiplying the particular independent variable by the moderating variable
‘institutional context’. The new parameter, or coefficient, that is created will be denoted as Mit.
Furthermore, multicollinearity must be accounted for, if necessary, the particular variable(s) will be altered16 accordingly to fit the regression.
The following model is specified to test both the moderating effect of proxies ‘board independence’ and ‘board monitoring abilities’ interchangeably:
COEit = β0 + β1(Proxyit) + β2(India) + M1(Proxyit * Indiait) + β3(Cross-Listedit)+) + β4(Asset
Compositionit) + β5(Growthit) + β6(Firm Sizeit) + εit
Similar to the prior stated regression specification, COEit refers to either ‘dividend yield’
or ‘price to book value’ of company i in period t. Next, β0 refers to the intercept and is not fixed
in this equation. Next, β1 refers to the coefficient of the independent variable that is used to proxy
either ‘board independence’ or ‘monitoring abilities’. Furthermore, β2 refers to the dummy
variable India. M1 denotes the bilinear moderating variable that is computed additionally as a
product term of the independent variable and the dummy variable that denotes the judicial
33
efficient context of firm i. Furthermore, β3 to 6 are firm-specific control variables. In accordance
with the dependent variable, ‘i’ refers to the period in which the observation is made, and ‘t’ concerns the company that is observed. Lastly, ε denotes the error term or noise.
3.5 Model Validation
For the results to be interpreted in a valid manner, the data that is used must meet several assumptions that are inherent with the OLS-regression model that is used for this study. However, due to the rather large sample size, these assumptions can be relaxed to certain extent. First, the relationships (between Xit and Yit) are all linear in the sample. Second, the data is checked for
normality. The normality of standard errors is examined to understand the significance (T-test) of a variable in a sound manner. However for most, this is not the case for all variables. Though, the central limit theorem argues that these assumptions can be relaxed with substantial sample size. Third, the error of variance is tested. A simple scatterplot where the standardize residuals are regressed with the standardized predicted values showed that the assumption is met. Finally, issues of independence are considered. However, due to the observations over a five year period and within two different groups of the same variables this assumption is of less significance. A correlation matrix is provided in Appendix
3.6 Interpretation of the model
The regression model is interpreted among multiple dimensions. First, the sign of the coefficients are examined whether they are similar to what is proposed in the literature review. Furthermore, the significance of the coefficients is assessed to indicate whether there is a likelihood that the coefficient has occurred incidentally. The significance levels are, respectively,
1%, 5% and 10%. Next the R2 and the regression model’s significance are examined at similar
P-values. The R2 indicates the explanatory power of the model. Furthermore, a negative coefficient
(-) indicates an inverse relationship (Xit negatively effects Yit) and a positive coefficient (+)
indicates a positive relationship (Xit positively effects Yit). The degrees of freedom (df) are of
less importance, however very useful examine whether the results can be generalized. A high df in the regression model is ought to make fair assumptions that the sample represents the
examined population. To test whether there is a moderating effect present, the R2 of
34
the data fits the model better in hypotheses B. Second, the R2 increment is assessed. This is
accomplished by simply subtracting the R2 of the A-models from the B-model. If this change is
regarded statistically significant, a moderating effect is present.
35
4. RESULTS
Next, the results of the regression analyses are reported. First, regression analyses are provided that regard to hypotheses A. Hence, they concern the un-moderated regression analyses. This functions as an introduction to the second part of this chapter. Thus, these results are of
interest to test whether a relationship, between Yit and Xit, actually exists or not. Each variable is
regressed in four manners. First, the variable is regressed independently of other variables. This gives an indication of the presumed relationship. Second and third, a dummy variable for India and subsequent firm-specific control variables are added to the regression. For the reason that dividend yield, or for that matter COE, is not merely a function of one variable, the regression analyses that includes all control variables (the third regression model) are of most importance to test the significance of the hypothesized relationship. Finally, the variables are regressed along all other independent variables that indicate board independence. This articulates the explanatory power of that particular variable within the whole model. More particular, if the results of the regressions show a statistical significant effect on the dependent variables these independent variables will be tested for their moderating effect in the Indian context.
4.1 Board Independence and its effect on dividend yield
Table 4.1 shows in a concise manner whether the proxies that indicate board independence are accepted or not. If a statistical significant relationship is present on usual significance levels, the alternate hypothesis is accepted. Please not that for one variable a significant correlating effect was documented with a contrary coefficient sign. This contradicting relationship will be elaborated upon in the discussion of this paper.
Table 4.1 Board independence and dividend yield
Board independence proxy Hypothesis
1. Board Structure Alternate hypothesis is accepted
2. Board Size Alternate hypothesis is accepted
3. Non-executive directors on board Null-hypothesis cannot be rejected
4. Turnover executive directors Alternate hypothesis is accepted
36
Table 4.2 Regression Analyses Dividend Yield (1/2)
37
Table 4.2 Regression Analyses Dividend Yield (2/2)
