Thesis BSc Economics and Business Economics
The influence of CEO duality on firm performance:
Evidence from Europe
Name: Noah Beukeboom
Student number: 11764368
Programme: BSc Economics and Business Economics Specialisation: Finance
Date: 29th of June, 2020
STATEMENT OF ORIGINALITY
This document is written by student Noah Beukeboom who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used
in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
A
bstract
When a Chief Executive Officer (CEO) of a firm simultaneously serves as the chairman of its board, directors have opposing objectives. This paper investigates the effect of CEO duality on firm performance relying on financial ratios (ROA, ROE and PM), to decide to what extent firm performance is being influenced by CEO duality. Whereas mainly US firm samples have been used in prior work, this paper performs research on 123 European publicly listed firms based in eight different European countries. The data is analysed by using robust OLS regressions controlled by several corporate governance variables. The results in this study reveal a significant and negative effect of CEO duality on firm performance. This finding is in line with the agency theory, which claims that the increased power of the CEO will hinder the independence between management and the board. However, the stewardship theory asserts that CEO duality has a positive influence on firm performance due to the unity of command it presents. The existing empirical evidence is largely inconclusive. Therefore, this study tries to contribute to the literature on CEO duality and firm performance in the context of European firms.
Table of content
Abstract……….………..…………...3
I. Introduction……….……….…5
II. Theory and hypotheses..………...……7
2.1 Agency theory………....……….……….…..7
2.2 Stewardship theory………..…..………..………….…………...8
2.3 Interaction of board size and CEO duality……….……….………....10
2.4 Interaction of board independence and CEO duality……….……..…..11
2.5 Conceptual framework………..…..…12
2.6 Empirical evidence on previous papers………..……12
2.7 Different board structures across countries……….….14
III. Methodology………..………..……...15
3.1 Data sample……….……….15
3.2 Variables……….………..….16
3.3 Empirical model specification……….………19
IV. Empirical findings……….20
4.1 Descriptive statistics ……….………..20
4.2 Regression output and hypotheses testing..………..….22
V. Discussion and concluding remarks……….…..29
I.
Introduction
Corporate governance covers a broad range of aspects in the process of managing a company. For a company to run efficiently it is essential to have the appropriate distribution of powers. The board of directors has the possibility of exercising a substantial amount of power in the decision-making process of a firm. One of the main purposes of the board is to function as a monitoring mechanism to align the interests of the Chief Executive Officer (CEO) with the interests of the shareholders of the firm.
Differences in leadership structure have contributed to several debates in the field of corporate governance. One of them is the aspect of CEO duality. CEO duality is defined as the situation where a CEO of the company simultaneously serves as the chairman of its board of directors. The main arguments of advocates and opponents of CEO duality are based on two approaches, respectively the stewardship theory and the agency theory. The stewardship theory asserts that the firm will benefit more from unified leadership, while the agency theory claims that the increased power of the CEO will hinder the independence between management and the board. Hence, for companies it is an important question whether CEO duality influences, either positively or negatively, firm performance.
There have been several studies on this subject matter with surprisingly different outcomes (Duru et al., 2016). In a study on companies in the Fortune 500, Rechner and Dalton (1991) reported that CEO duality was negatively associated with firm performance. Nevertheless, a meta-analysis on several other studies encountered an insignificant relationship between duality and firm performance (Berg & Smith, 1978; Chaganti et al.,1985; Daily & Dalton, 1992). These studies have been mainly conducted on publicly listed firms in the United States, which have a different board leadership structure compared to firms in Europe. Therefore, to underline the relevance of this paper, there has been chosen to conduct research on European publicly listed firms. The aim of this paper is to shed light on a new perspective next to the already existing literature regarding CEO duality. The research question that will be examined in this paper is:
To what extent is firm performance being influenced by CEO duality on European publicly listed firms?
This paper examines the relationship between CEO duality and firm performance. A sample of 123 European firms will be analysed. The corporate governance variables of the companies have all been manually collected in order to find the most recent and accurate data. In this study, a robust Ordinary Least Squares (OLS) regression will be conducted. The models are controlled by leverage, firm size and country effects. Next to this, two moderating variables, board size and board independence, will be included in the model to reflect the influence of CEO duality on firm performance as efficient as possible. The results of this study provide evidence that CEO duality has a statistically significant negative influence on firm performance that is positively moderated by board size. This implies that a larger board limits managerial entrenchment of a CEO and exploits the benefits of a dual leadership structure. This paper contributes to the existing literature by focusing on the European board leadership structure and by applying the moderating effect of corporate governance characteristics on the effect of CEO duality on firm performance.
Once again, nearly all studies have been conducted on firms in the United States. The most significant difference regarding board leadership structure between the United States and Europe is the distinction between a management board and a supervisory board. Whereas the business environment in the United States is known for its one-tier board structure, in Europe the two-tier board structure is being used in part of the countries. The two-tier board consists out of a supervisory board equipped with the task of supervising the management board. This setting entails that the chairman of the supervisory board cannot be the chairman of the management board, implying that CEO duality is not present in firms who hold a two-tier board structure. This will be further explained in one of the next sections.
In this study, there has been chosen to apply moderating variables on the relationship between CEO duality and firm performance. Moderating variables are variables which change the strength or direction of an effect between the independent (duality) and dependent variable (firm performance). Based on prior studies, there has been chosen to use board size and board independence as moderating variables (Kholeif, 2008; Ramdani & Witteloostuijn, 2010; Duru et al., 2016). The conceptual framework illustrates the model, including these variables, that will be analysed in this study.
The remainder of this paper is organized along the following lines. In the next section, the background of the different theories will be defined while the hypotheses of the research are being introduced. Thereafter, the methodology and the description of the data is being outlined. In the fourth section, the results of the data analyses are interpreted. Subsequently, this paper
II.
Theory and hypotheses
Agency Theory
One of the leading theories regarding the relationship between CEO duality and firm performance is the agency theory. Agency theorists say that the board of directors fulfils a function of checks and balances in a company. The control function of the board is present to monitor the behaviour of the CEO, limit managerial entrenchment and align the shareholders’ interests with the interests of top management of a company. Achieving these objectives will become more difficult when CEO duality is present in the governance structure of a company. Therefore, the agency theory pleads for a separation of CEO and chairman of the board.
Jensen and Meckling (1976) argue that duality increases the power of the CEO which results in a rise of potential managerial entrenchment. According to this theory, this opportunistic behaviour can lead to a reduction in shareholder wealth. In addition, Fama and Jensen (1983) find that the effective control mechanism of the board of directors is being jeopardized when a CEO gets assigned a significant role on the board. A potential solution is to provide the CEO with incentives to act more in line with the interests of shareholders. This can be accomplished by offering the CEO stock ownership in the company.
The ideal corporate governance structure according to Rechner (1989) should consist out of a board with a majority of outside directors and the chairman also being an outside director. She specified that firms with a dual leadership structure have a weaker form of corporate governance compared to firms who do not have a CEO as chairman of the board. An outside director is also more incentivized to intervene more quickly than an inside director when firm performance is decreasing because insiders may be beholden to a CEO for their careers (Fama & Jensen, 1983).
Finkelstein and D’Aveni (1994) affirm that one of the main arguments to avoid CEO duality is the fact that the CEO is both in charge of the agenda and the content of board meetings. Furthermore, CEO duality makes it challenging for the board to fire poorly performing executives (Goyal & Park, 2002). Chairmen are most of the time in control of nominating directors, meaning that this person can choose whoever is loyal to him. When this process is continuing for a long time, a CEO can compose a board with directors who are reluctant to stand up against him (Finkelstein & D’Aveni, 1994). Therefore, a board is positioned in a difficult situation when the chairman is not independent. Basically, when the
CEO is responsible for the management of a company while he also evaluates the efficacy of that particular policy a conflict of interest arises. Wan and Ong (2005) say that CEO duality is detrimental to firm performance, comparing it to someone who is marking his own examination paper. In addition to this, the phrase ‘who monitors the monitor?’ written down by Alchian and Demsetz (1992, p.7) advocates for a separation of CEO and chairman of the board. The before mentioned arguments result in the following hypothesis:
H10: CEO duality has no influence on the firm performance of European publicly listed
firms.
H1A: CEO duality has a negative influence on the firm performance of European publicly
listed firms.
Stewardship Theory
In contrast to the agency theory, the stewardship theory is approaching the relationship between CEO duality and firm performance from another perspective. CEO duality shows a unity of command and ensures that a firm is run effectively by unambiguous and clear authority (Finkelstein and D’Aveni, 1994). This theory underlines the managerial behaviour of an executive by focusing more on non-financial factors such as altruism, respect and reputation rather than focusing on the pursuing of personal gain by managers.
Combining the role of CEO and chairman of the board will create a more efficient flow of information. Normally, the information concerning management decisions should convey to the chairman and the CEO and vice versa. In case of a dual leadership structure, the costly, imperfect and time-consuming flow of information is minimalised to one person. Subsequently, this will lead to a decline in information costs (Jensen & Meckling, 1976).
Brickley et al. (1997) state that separating the positions of chairman and CEO can induce costs in monitoring the CEO. Besides, it involves incentive costs by creating a process in which a CEO gets promised the chairman title of the company. This might result in a CEO acting in his self-interest rather than maximizing shareholder value. They conclude that these costs might offset the advantage which arise when both positions are separated from each other. Their empirical evidence suggests that non-duality firms do not necessarily underperform firms who maintain a duality structure.
When essential information resides in one central management position it might enhance the individual’s ability to carry out the responsibilities of management. The firm will benefit more from unified leadership according to the stewardship theory (Donaldson & Davis, 1991). Likewise, Donaldson and Davis claim that clear leadership leads to remarkable returns for the shareholders of the company compared to firms which uphold a separate leadership structure.
The joint leadership structure provides consolidated command and discards any external or internal ambiguity regarding decision-making and its consequences (Donaldson, 1990; Finkelstein & D’Aveni, 1994). Consistent with this view, Baliga et al. (1996) clarify that the dual leadership structure empowers the position of the CEO and makes him feel more ambitious and inventive because no one above him summons him what to do.
Another theoretical suggestion implying the functionality of CEO duality is based on the strategy formation literature according to Finkelstein and D’Aveni (1994). Proponents of this perspective argue that strong leaders should issue commands to their lower level servants, set out strategic objectives and finally pass on instructions throughout the whole business so that all orders will be implemented as instructed (Andrews, 1987). The success of an organisation thrives on the strong leadership of a CEO. For instance, Mintzberg and Waters (1982) reveal in their study that firms adapting to environmental demands are positively correlated with powerful CEO leadership. In line with this, research has shown that businesses without unified and powerful leadership are unsuccessful, because Miller and Friesen (1977) claim that determined courses of action are more difficult to make in case of a split positions. Hence, the segregation of the positions of chairman and CEO can weaken the powerful leadership and lead to a diffusion of power and inconclusiveness at the top level of a firm.
These arguments focus on the centrality of CEO leadership and support the rationale behind the stewardship theory, which yields the following hypothesis:
H20: CEO duality has no influence on the firm performance of European publicly listed
firms.
H2A: CEO duality has a positive influence on the firm performance of European publicly
Interaction of board size and CEO duality
According to the agency theory, communication problems arise when the size of the board of directors becomes larger. Chaganti et al. (1985) assert that agency problems may have greater impact on larger boards which can result in the deterioration of the effective monitoring mechanism, a core task of a board. Besides, due to a larger amount of board members an incentive to freeride on other persons’ efforts might arise. All of these points in combination with CEO duality might decrease firm performance (Hermalin and Weisbach, 1998).
However, Zahra and Pearce (1989) proclaim that larger boards bring people together from diverse industrial and educational backgrounds. These board members share their different views on managerial activities which improve the quality of decision-making by a firm. Furthermore, the power of a CEO, especially in case of CEO duality, becomes less when board size increases which strengthens the corporate governance of a firm (Jackling & Johl, 2009). This implies that board size might moderate the impact of CEO duality on firm performance. Additionally, a larger board provides more monitoring resources, which might improve firm performance.
The study hypothesizes that board size moderates the association between CEO duality and firm performance. As explained previously, CEO duality can have a positive influence on firm performance due to the advantages of strong leadership or may impact firm performance negatively because of the weaker controlling mechanism according to the agency theory. In this study, there will be expected that CEO duality has a negative influence on firm performance, but due to the moderating effect of a larger board size, which may imply a decrease in CEO power, the following hypothesis has been formulated:
H30: The board size and CEO duality do not interact to influence firm performance.
H3A: The interaction of CEO duality and board size has a positive influence on firm
Interaction of board independence and CEO duality
Regarding this corporate governance feature, the agency and stewardship theory both display their opposing perspectives. Agency theorists argue that a larger proportion of independent directors on a board will positively affect firm performance. One of the reasons is that boards dominated by independent directors are expected to be more independent of management and so more likely to protect other stakeholders’ interests. Besides, the agency theory assumes that CEOs are opportunistic and selfish. Ramdani and Witteloostuijn (2010) argue that effective monitoring by an independent board is a crucial element of good corporate governance in order to ensure that a CEO pursues shareholder value rather than personal gain. Ergo, boards with a larger proportion of independent directors are able to effectively monitor managerial tasks. Empirical evidence revealed that boards with the right proportion of outside directors contributed more to firm performance than firms where insiders predominated the board (Jensen & Meckling, 1976; Fama & Jensen, 1983; Eisenhardt, 1989).
On the other hand, advocates of the stewardship theory claim that boards where a large proportion of inside directors dominate the board are preferred because CEOs are being considered trustworthy, altruistic and unitedly oriented. Further, Davis et al. (1997) assert that inside directors are better informed than their outside counterparts, which facilitates the support of effective decision-making. Consequently, stewardship theorists claim that boards with a predominance of inside directors increase firm performance (Donaldson, 1990).
Lastly, Duru et al. (2016) consider the importance of board independence as a crucial attribute of adequate monitoring. An interesting occurrence in line with this thinking, is that the U.S. Congress enacted the Sarbanes-Oxley Act in 2002, requiring a larger portion of independent directors on the board. This justifies the focus on board independence in this research. Supposedly, CEO duality will have a negative influence on firm performance, but because of the moderating effect of board independence, which may imply a decrease in CEO power, the following hypothesis has been formulated:
H40: The board independence and CEO duality do not interact to influence firm
performance.
H4A: The interaction of CEO duality and board independence has a positive influence on
Conceptual Framework
The conceptual framework of this research in summarized in the table below. The impact of the independent variable CEO duality on the dependent variable firm performance will be examined. Besides, the independent variable will be moderated by board size and board
independence. Next to this, firm performance will be controlled by firm size and leverage. In
the methodology section of this study, the details will be further underlined.
Conceptual framework
Empirical evidence of CEO duality on previous studies
In this section the results from previously examined papers will be discussed. As mentioned before, several academia have examined the relationship between firm performance and CEO duality before, which has resulted in a broad spectrum of diverse outcomes. While part of the studies supports the stewardship theory (Donaldson & Davis, 1991; Lin, 2005; Peng et al., 2010), others side with the agency theory (Rechner & Dalton, 1991; Pi & Timme, 1993; Duru et al., 2016). Even in other studies, there has been found a non-significant relationship between firm performance and CEO duality (Chaganti et al., 1985; Daily & Dalton, 1992; Baliga et al., 1996).
A possible explanation for the different outcomes of the studies is that the scholars applied different measures of firm performance. The study of Rechner and Dalton (1991), using a sample of 141 corporations over a six-year timespan, employed accounting based measures and found empirical evidence that firms with non-duality structures outperform firms with CEO duality. In line with this, Pi and Timme (1993) found evidence to support the separation
CEO duality Firm performance
Control Variables Firm size Leverage Moderating Variables Board size Board independence
of the CEO and chairman positions, also using accounting-based measures. However, Chaganti et al. (1985) came to the conclusion, by employing bankruptcy measures as performance benchmark, that they did not find any correlation between CEO duality and firm performance. Another difference between this study and the study of Pi and Timme (1993) is that the latter did not control for industry effects and limited its sample to banking firms, while Chaganti et al. (1985) concentrated on the retailing industry.
Another clarification for the mixed results is the effect of endogeneity in the regression models of the studies (Hermalin & Weisbach, 1998; Raheja, 2005; Faleye, 2007). Endogeneity occurs when a variable that is not included in the regression model is related to a variable incorporated in the model. In this case, the choice of board leadership structure (non-duality or duality) might be endogenous. Therefore, it is difficult to determine whether there is a relationship between CEO duality and firm performance (Harrison, Torres & Kukalis, 1988; Adams, Almeida, & Ferreira, 2005). Additionally, a research paper on 7,000 US based firms regarding board structure found that firm performance did not drive CEO duality (Linck et al., 2008). In line with this, Iyengar and Zampelli (2009) concluded in their research that studies did not suffer from selection bias in case CEO duality is treated as an exogenous variable. In this study CEO duality is treated as an exogenous variable.
A study conducted by Baliga et al. (1996) on a sample of Fortune 500 companies examined the announcement effects and accounting measures of operating performance for firms that changed their duality structure. They discovered that changes in duality structure did not significantly result in changes in operation performance and found hardly any evidence that duality affects long-term firm performance.
Both studies (Donaldson & Davis, 1991; Brickley et al., 1997) based their research on American firms. Whereas Brickley et al. (1997) found a higher return on capital for duality firms, did Donaldson and Davis (1991) discover a higher return on equity for firms relying on a joint leadership structure compared to firms maintaining a non-duality regime. Both studies thus found evidence for rejecting the agency theory.
Contrastingly, Duru et al. (2016) worked with a US based sample consisting out of 950 firms to estimate a dynamic model of the relationship between CEO duality, including board characteristics, and firm performance. Their results yielded compelling evidence that CEO duality has a negative impact on firm performance. Also, they found that this outcome is positively moderated by board independence. The conclusion of their study is in line with the agency theory and supports the fact that duality structures might reduce firm performance through managerial entrenchment.
Different board structures across countries
Compared to previously discussed studies which focused mainly on US firms, this paper is examining the relationship between CEO duality and firm performance on publicly listed firms in Europe. This section is deemed necessary to underline the differences in corporate governance systems between the United States and Europe, the Anglo-Saxon model1 and the Continental European model respectively. In practice, this means that the latter countries enforce a dual board structure while the United States implements a unitary board structure. The main focus in this section lies on the board structure in the Netherlands and Germany, because both countries in this research sample operate with a two-tier board system2. The two-tier board system consists out of a supervisory board which is responsible for monitoring policies and managerial behaviour, and accountable for remuneration and the appointment of members for the management board. Meanwhile, the management board is responsible for running the firm on a day-to-day basis. This system embraces a formal separation of executive and supervisory roles resulting in a less continuous stream of information. On the other hand, the one-tier structure might show a more smoother information flow because the supervisory and managerial functions are intertwined (Krivogorsky, 2006).
The two-tier board structure, embodied by the Netherlands and Germany, radically contradicts the Anglo-American model. Whereas the main goal of the Anglo-American model is the maximisation of returns to the shareholders of the firm and in such a way mitigates agency problems, the Continental European model tries to take into account the considerations of other stakeholders while making strategic firm decisions (Shleifer & Vishny, 1997; Fauver & Fuerst, 2006). However, the duties of the US board of directors demonstrate large similarities with the supervisory boards in the Dutch and German two-tier board structure. A fundamental distinction between the two board structures is that an individual can never serve on both the management and the supervisory board simultaneously in firms with a two-tier board structure (Bermig, 2012). When a CEO of the company chairs the management board, CEO duality in the Netherlands and Germany will never occur. While CEO duality in the US one-tier board structure often can be the case.
III. Methodology
Data sample
The sample consists out of 123 publicly listed firms based in eight countries across Europe. The countries with the biggest economies based on their GDP per capita have been chosen. Hence, the selected firms are located in Belgium, France, Germany, the Netherlands, Spain, Sweden, Switzerland and Great Britain. All firms are publicly listed on the largest stock exchange in the specific country. Firstly, the sample search consisted out of firms with at least 70 employees listed on Euronext Brussels, Euronext Paris, Börse Berlin, Euronext Amsterdam, Bolsa de Madrid, Nasdaq Stockholm, the Swiss exchange and the London stock exchange. In table 1, the process of the sample selection is summarized.
Firms based in the regulated financial industry were removed out of the sample, in line with Duru et al. (2016). Then, the firms with missing financial profitability ratios and other financial data variables were removed out of the sample. One of the last steps, was the elimination of firms who were not active during the whole sample period. This was relatively the lowest number of deleted firms. The final step was a particularly time-consuming task. An available database to collect information digitally about the corporate governance features (CEO duality or non-duality, board size and board composition) was not present. So, all the governance variables from the firms in the sample had to be collected manually. This was done by using annual reports, ‘investor relations’ sections on websites and corporate governance code manuals of each firm. A total of 98 firm did not have sufficient or available data on these governance variables and thus were removed out of the sample. This resulted in a final research sample of 123 firms. The data observations were collected for the period between 2015 and 2019 inclusive, a five-year timespan. This led to 615 firm-year observations. The necessary financial information on the firms has been collected from the ORBIS database.
Table 1: Details on sample selection (2015-2019)
Variables
Dependent variable
Firm performance
Several studies have demonstrated that there does not exist one universal performance measure. In the academic world researchers use accounting-based measures and market-based interchangeably to determine firm performance in the corporate governance literature. In this research, firm performance will be measured by return on assets (ROA), return on equity (ROE) and profit margin (PM). Accounting-based measures have been chosen, because these measures show accurate financial information about firms (Joskow et al., 1993) and prior studies have widely used these measures, which facilitates comparison between them (Dalton et al., 1998). One group of researchers used return on assets and profit margin as performance measures (Rechner & Dalton, 1991; Pi & Timme, 1993; Abdullah, 2004; Kholeif, 2008), another group used return on equity (Berg & Smith, 1978; Donaldson & Davis, 1991; Baliga et al., 1996).
This study assesses the performance of a firm by using the return on assets ratio (ROA), which shows the amount of earnings generated from invested capital (assets). As a matter of fact, managers (CEOs) are responsible for the utilization of the firm’s assets. So, the return on
Details Observations
Firm observations from selected stock exchanges
in Orbis database 1,043
Less: Firm observations from financial industry
(banks, insurance companies, etc.) 283
Less: Firm observations with insufficient financial
data in Orbis database 443
Less: Firm observations for firms who were not
active during the whole sample period 96
Less: Firm observations with unavailable
governance data in annual rapports 98
Firm observations in final sample
assets offers us an insight in how well the corporate governance mechanism is stimulating effective management of the business (Fooladi, 2011).
Maximising shareholder value is one of the main goals of publicly listed firms. Return on equity (ROE) is a measure which determines the amount of profit that has been generated from the invested capital in the firm (Duru et al., 2016).
Profit margin, equivalent to return on sales (ROS), is another accounting-based performance measure. A common method used to determine firm performance because it shows how much a firm is earning in relation to its sales (Rechner & Dalton, 1991; Duru et al., 2016).
Table 2: Measurements of firm performance
Variable Description/Measurement
Return on Assets (ROA) 100% X (net income divided by total assets) Return on Equity (ROE) 100% X (net income divided by total equity) Profit Margin (PM) 100% X (net income divided by total revenue)
Independent variable
CEO duality
CEO duality is determined by verifying whether a CEO serves simultaneously as chairman of the board of directors. The data are manually collected via annual reports of the companies in the research sample. CEO duality is being identified as a binary variable. So, in case CEO duality is present in a firm the value one (1) gets assigned to it, otherwise the value zero (0). The variable CEODUAL is used in the regression model.
Moderating variables
Board independence
The number of outside directors in a board when expressed as a fraction of the total amount of directors in a board is an adequate measure of board independence (Rechner & Dalton, 1991). Outside directors, both executive and non-executive, are considered as independent directors in this research. The variable BOARDIND is used in this study as variable for board independence.
Board size
Board size can be identified as the total number of directors in the board of each firm in the sample. This will include inside directors and outside directors, both executive and non-executive (Hermalin & Weisbach, 1991; Bhagat & Black, 2002). The natural logarithm of board size is used in this study (Duru et al., 2016). The variable log(BOARDSIZE) is used in the regression model.
Control variables
Firm size
The size of the firm can also affect firm performance because larger firms have greater capabilities which could positively influence firm performance (Majumdar and Chhibber, 1999). Nevertheless, larger firms may likewise have coordination problems which can lead to lower firm performance. Firm size is measured by the natural logarithm of net sales; the logarithmic form of the variable is used to mitigate the effect of heteroscedasticity (Dang et al., 2018). The variable log(FSIZE) is used in the regression model.
Leverage
Leverage is calculated as the ratio of total debt to total assets. This variable is incorporated as a control variable, because the amount of leverage can influence firm performance (Duru et al., 2019). The book value of leverage is used in this case, because it does not reflect any recent changes in the valuation of the firm. The variable LEV is used in the regression model.
Country dummy variable
The sample in this study consists out of multiple countries. Without controlling for the countries, it is possible that the relationship between CEO duality and firm performance is being influenced by differences in country settings. Therefore, there has been chosen to include country dummy variables in the regression model. The sample consists out of eight countries. So, in order to avoid the dummy variable trap one country, Spain, did not get a variable assigned. The following variables are incorporated into the regression model: BE (Belgium),
CH (Switzerland), DE (Germany), FR (France), GB (Great Britain), NL (the Netherlands) and SE (Sweden).
Empirical model specification
In this study an ordinary least square (OLS) multiple regression model is used with robust standard errors. This study employs a conditional mean regression. The firm observations over five years have been averaged and all assumptions of the OLS regression model has been satisfied. Model 1 and model 2 show the basic regressions and model 3 and model 4 are controlled for country effect. Based on the above-mentioned explanation on the variables, the following four regression models have been formulated:
Model 1 𝑌𝑖= 𝛽0+ 𝐶𝐸𝑂𝐷𝑈𝐴𝐿 ∗ 𝛽𝑖1+ log(𝐵𝑂𝐴𝑅𝐷𝑆𝐼𝑍𝐸) ∗ 𝛽𝑖2 + 𝐵𝑂𝐴𝑅𝐷𝐼𝑁𝐷 ∗ 𝛽𝑖3+ log(𝐹𝑆𝐼𝑍𝐸) ∗ 𝛽𝑖4+ 𝐿𝐸𝑉 ∗ 𝛽𝑖5 + 𝜀𝑖 Model 3 𝑌𝑖= 𝛽0+ 𝐶𝐸𝑂𝐷𝑈𝐴𝐿 ∗ 𝛽𝑖1+ log(𝐵𝑂𝐴𝑅𝐷𝑆𝐼𝑍𝐸) ∗ 𝛽𝑖2 + 𝐵𝑂𝐴𝑅𝐷𝐼𝑁𝐷 ∗ 𝛽𝑖3+ log(𝐹𝑆𝐼𝑍𝐸) ∗ 𝛽𝑖4+ 𝐿𝐸𝑉 ∗ 𝛽𝑖5 + 𝐵𝐸 ∗ 𝛽𝑖6 + 𝐶𝐻 ∗ 𝛽𝑖7 + 𝐷𝐸 ∗ 𝛽𝑖8 + 𝐹𝑅 ∗ 𝛽𝑖9 + 𝐺𝐵 ∗ 𝛽𝑖10 + 𝑁𝐿 ∗ 𝛽𝑖11 + 𝑆𝐸 ∗ 𝛽𝑖12 + 𝜀𝑖
Where Yi is alternatively profit margin, return on equity and return on assets.
Before being able to do the regression analysis in STATA, the data which was manually collected on the corporate governance features and the financial information from the Orbis
database were merged. Then, the necessary variables were created in order to finalize the data
analysis of this study. In the regression models below, the interaction variable of CEO duality with board size and the interaction variable of CEO duality with board independence are included. As mentioned above, model 4 is controlled for country effects.
Model 2
𝑌𝑖= 𝛽0+ 𝐶𝐸𝑂 𝐷𝑢𝑎𝑙𝑖𝑡𝑦 ∗ 𝛽𝑖1+ log(𝐵𝑂𝐴𝑅𝐷𝑆𝐼𝑍𝐸) ∗ 𝛽𝑖2 + 𝐵𝑂𝐴𝑅𝐷𝐼𝑁𝐷 ∗ 𝛽𝑖3+ log(𝐹𝑆𝐼𝑍𝐸) ∗
𝛽𝑖4+ 𝐿𝐸𝑉 ∗ 𝛽𝑖5 + Duality ∗ BOARDSIZE∗ 𝛽𝑖6 + Duality ∗ BOARDIND ∗ 𝛽𝑖7 + 𝜀𝑖
Model 4
𝑌𝑖= 𝛽0+ 𝐶𝐸𝑂 𝐷𝑢𝑎𝑙𝑖𝑡𝑦 ∗ 𝛽𝑖1+ log(𝐵𝑂𝐴𝑅𝐷𝑆𝐼𝑍𝐸) ∗ 𝛽𝑖2 + 𝐵𝑂𝐴𝑅𝐷𝐼𝑁𝐷 ∗ 𝛽𝑖3+ log(𝐹𝑆𝐼𝑍𝐸) ∗
𝛽𝑖4+ 𝐿𝐸𝑉 ∗ 𝛽𝑖5 + Duality ∗ BOARDSIZE∗ 𝛽𝑖6 + Duality ∗ BOARDIND ∗ 𝛽𝑖7 + 𝐵𝐸 ∗ 𝛽𝑖8 +
IV. Empirical findings
This section will show the descriptive statistics table, the correlation matrix and 3 tables with regression outputs. These tables consist each out of 4 regression. Two of them are not controlled for country effects, while the other two regression are controlled. Furthermore, the section will test the hypotheses formulated in this paper using t-tests and will interpret the coefficients in order to relate the data with an economic insight.
Descriptive Statistics
Table 2 provides the descriptive statistics about the European publicly listed companies used in the research sample. For each variable the mean, standard deviation, the minima and maxima are reported. The profit margin has a mean of 8.91% (SD = 9.41%) and is ranked between the return on equity (M = 15.66%, SD = 18.40%) and return on assets (M = 5.16%, SD = 5.46%). The table reveals that CEO duality, the independent variable in this study, is present in 21.1% of the corporate governance structures of the companies. One can argue whether this is high or low. Multiple studies show a range of various proportions. Kajola (2008) and Hewa and Locke (2011) reported relatively low percentages of CEO duality in their sample, whereas other studies documented the presence of CEO duality in more than half of the firms in their sample (Finkelstein & D’Aveni, 1994; Goyal & Park, 2002). A potential explanation for this discrepancy is the fact that latter studies used US based firm samples.
Furthermore, the average board size in the sample is 9 (SD = 9.41), which is similar to the results of Duru et al. (2016). Also, coherent with the suggestion of Lipton and Lorsch (1992), who state that the preferred board size is eight or nine members in order to manage a board effectively. The size of the boards fluctuates between a minimum of three board members to a maximum of twenty. The variable board independence discloses a mean percentage of 61.3%. This implies that more than half of the board members are outside directors. This finding is in line with the view of the agency theorists, which encourage independent management.
Table 3: Descriptive statistics table
Variable N Mean Std. Dev. Min Max
Profit margin (in %) 123 8.911 9.41 -17.923 50.909
ROE (in %) 123 15.66 18.402 -55.456 86.505 ROA (in %) 123 5.155 5.462 -19.496 20.272 CEO Duality 123 .211 .41 0 1 Board Independence 123 .613 .268 0 1 Board Size 123 8.74 3.885 3 20 Firm size (ln) 123 13.804 2.435 9.544 19.425 Leverage (in %) 123 12.968 10.883 0 68.300 Correlation matrix
Multicollinearity is the occurrence of high intercorrelation among the independent variables in a multiple regression model. The variables must be eliminated, when the degree of correlation is high. Multicollinearity can lead to skewedness of the results, which makes the data less reliable. In table 4, a correlation matrix on all variables is presented. Some of the variables show a relative high correlation, but none of them exceed 0.75. This benchmark tells whether there is multicollinearity in your data sample (Gujarati, 1999). This means that all variables can be treated in the analysis.
One of the main conclusions that can be drawn from this table is the significant negative correlation between CEO duality and the three performance measures, respectively profit margin, r(121) = -.19, p < .05, return on equity, r(121) = -.30, p < .05 and return on assets, r(121) = -.28, p < .05. Based on this, one might argue cautiously that there is evidence in support of hypothesis 1A implying that CEO duality negatively influences firm performance. But one has to keep in mind that correlation is not the same as causality. To determine a causal relationship, the hypotheses need to be tested, which will be done in the next paragraph.
Table 4: Correlation matrix
N=123 Profit
margin
ROE ROA CEO
Duality
Board Ind.
Board Size
Firm size Leverage
Profit margin 1.000 ROE 0.5138* 1.000 ROA 0.7214* 0.7419* 1.000 CEO Duality -0.1918* -0.2972* -0.2787* 1.000 Board Ind. 0.165 0.1872* 0.091 -0.2836* 1.000 Board Size -0.029 -0.036 -0.2050* 0.161 -0.038 1.000 Firm size -0.047 0.103 -0.161 0.126 0.2420* 0.6824* 1.000 Leverage -0.152 0.162 0.149 0.013 -0.142 -0.2781* -0.2584* 1.000 * p < 0.05.
Regression output and hypotheses testing
Table 5 presents the results of the OLS regression executed in this study on the dependant variable profit margin with robust standard errors. The t-statistics are stated between parentheses to facilitate the hypothesis tests. Model 3 and model 4 are controlled for country effects, whereas in model 2 and model 4 the interaction variables are included. As expected, the R2 increases when more variables are added.
In model 1, the table reports a statistically significant and negative relationship between CEO duality and profit margin, t(122) = -1.88, p < .001. Therefore, there is substantial evidence to reject the H10 and find support for H1A. Subsequently, H2A will be refuted. The coefficient of the variable CEO duality implies that profit margin decreases by -3.38% in case of CEO duality. This finding is in line with the agency theory, advocating for a separation between the positions of CEO and chairman of the board because CEO duality lowers firm performance.
Subsequently, the table discloses that leverage also has a significant negative effect on profit margin, t(122) = -1.81, p < .001. This shows that there is proof suggesting that an increasing amount of leverage in a firm will decrease profit margin. Next to these two
significant values in model 1, the variables board composition and board size show a positive and negative influence, respectively. However, both variables are not significant and thus do not influence profit margin. Board size is a logarithm expressed as the number of people
positioned in the board of directors. This indicates that, in model 1, a 10% increase in board size will cause a -2.25 X log(1.1) = 0.21% decrease in profit margin.
In model 2, the interaction variables regarding hypothesis 3 and hypothesis 4 are included. Here, it shows that CEO duality is still statistically significant and negatively influences profit margin, t(122) = -3.04, p < .001. Therefore, strong evidence is provided to reject H10. This is in line with the agency theory. Reasons for a negative influence of CEO duality on profit margin might be the occurrence of managerial entrenchment. The coefficient of CEO duality is in this model higher than in model 1, suggesting an even stronger (-16.77 vs -3.38) percentual influence on the dependent variable.
Additionally, the interaction variable of CEO duality and board size does also show a significant positive influence on firm performance, t(122) = 3.18, p < .001. Accordingly, the evidence suggest that H30 can be rejected. In practice, this means that a larger board in combination with CEO duality might positively influence firm performance. A reason for this is that the power of a CEO declines once more members are present in the board of directors.
Lastly, the inclusion of model 3 and model 4 in this regression analysis makes sure that the influence of CEO duality on firm performance is controlled by country effect. If the first two models are compared to the last two, one can see that the coefficients decrease once the country dummy variables are included. In the most complete model 4, CEO duality stays significant and the same applies to the interaction variable of board size and duality.
The interaction variable Duality*Board Independenceshows a positive effect on profit
margin, while CEO duality on its own has a negative impact. This entails that, in case board
independence is equal to zero (no independent directors on the board), CEO duality has a negative influence on firm performance, but this effect will be positively moderated when the number of independent directors on the board rises. However, the interaction variable is non-significant.
Table 5: OLS regression (robust) on dependent variable profit margin
(1) (2) (3) (4)
Variables Profit margin Profit margin Profit margin Profit margin
CEO Duality -3.384*** -16.77*** -2.046 -15.01*** (-1.881) (-3.036) (-0.771) (-2.668) Board Independence 2.415 1.461 1.848 0.254 (0.801) (0.452) (0.486) (0.0646) Board size(ln) -2.250 -4.267 -1.384 -3.258 (-0.779) (-1.391) (-0.455) (-1.018) Firm size(ln) -0.0784 -0.248 -0.121 -0.335 (-0.182) (-0.574) (-0.251) (-0.672) Leverage -0.158*** -0.164*** -0.136 -0.138 (-1.810) (-1.948) (-1.515) (-1.587) Belgium 4.953 4.561 (0.811) (0.753) Switzerland 3.495 4.459 (0.888) (1.133) Germany 3.203 4.052 (0.947) (1.184) France -0.408 -0.224 (-0.117) (-0.0643) Great Britain 0.369 0.580 (0.108) (0.166) Netherlands -0.102 1.406 (-0.0262) (0.364) Sweden 16.68*** 14.92*** (4.795) (3.935) Duality*Board Independence 2.130 0.533 (0.333) (0.0719) Duality*Board Size 1.288*** 1.335*** (3.176) (3.107) Constant 15.93*** 23.06*** 13.31*** 20.63*** (2.359) (3.041) (1.676) (2.319)
Country dummy NO NO YES YES
Observations 123 123 123 123
R-squared 0.075 0.123 0.120 0.168
t-statistics in parentheses *p<.05; **p<.01; ***p<.001
Table 6 presents the results of the OLS regression on the dependant variable return on
equity with robust standard errors. Also, in this table the t-statistics are stated in parentheses.
In line with the previous regression model, where profit margin functioned as the dependant variable, CEO duality has a significant influence on return on equity as well, t(122) = -3.41, p < .001. The value of -12.16 is considerably higher than in model 1 of table 5. An even higher value of -28.00 in model 2 is statistically significant. This is serious evidence in order to reject the H10. Due to the addition of the controlling country variables, there can be seen that the influence diminishes in model 3 and 4, but a rather strong effect is still observed.
Besides, in model 1 the two control variables firm size and leverage are both significant, t(122) = 2.19, p < .001 and t(122) = 2.55, p < .001 respectively. Whereas both variables had negative signs in the previous model, they now both have a positive influence on return on
equity. Additionally, both control variables are significant in all 4 models. Because firm size is
a logarithmic variable its effect on the dependent variable with a 10% increase, is a 2.08 X log(1.1) = 0.20% increase in return on equity in model 4.
The interaction variable Duality*Board size is included in model 2 and 4. Both models suggest that this interaction term asserts a positive and significant influence on return on equity. This is consistent with the results from table 5. This implies that there is evidence to reject H30 on a 0.1% significance level in both models.
From model 1 to model 4 the R2 increases by exactly 12%, revealing the explanatory power of the country dummy variables. A remarkable point is that the variable CEO duality does not show a significant value in the last model.
Regarding the control of country effect, one can see that almost all countries show a negative effect on the dependent variable except for Great Britain and Germany. This effect is relative to Spain, because that country functions as the reference country in the regression model. Only Sweden shows a significant and strong positive effect relative to Spain on return
on equity. A possible explanation for this high value is the relatively low number of firms and
Table 6: OLS regression (robust) on dependent variable return on equity
(1) (2) (3) (4)
Variables Return on equity Return on equity Return on equity Return on equity
CEO Duality -12.16*** -28.00*** -7.555*** -16.80 (-3.413) (-1.864) (-1.762) (-1.155) Board Independence 2.382 2.476 5.661 6.128 (0.455) (0.447) (1.001) (1.087) Board size(ln) -2.688 -5.680 -1.904 -4.232 (-0.637) (-1.423) (-0.419) (-1.015) Firm size(ln) 1.803*** 1.594*** 2.194*** 2.076*** (2.190) (1.890) (2.200) (1.969) Leverage 0.399*** 0.396*** 0.374*** 0.379*** (2.550) (2.618) (2.213) (2.305) Belgium -2.383 -1.973 (-0.503) (-0.427) Switzerland -2.352 -0.860 (-0.620) (-0.235) Germany 4.775 5.570 (1.093) (1.326) France -2.392 -2.768 (-0.486) (-0.574) Great Britain 5.034 5.819 (1.050) (1.255) Netherlands -1.025 -0.118 (-0.187) (-0.0206) Sweden 56.22*** 55.74*** (12.33) (12.07) Duality*Board Independence -9.519 -18.13 (-0.674) (-1.205) Duality*Board Size 2.125*** 1.936*** (2.603) (2.217) Constant -7.725 1.219 -19.33 -14.07 Country dummy (-0.563) NO (0.0844) NO (-1.364) YES (-0.990) YES Observations 123 123 123 123 R-squared 0.149 0.181 0.239 0.269 t-statistics in parentheses *p<.05; **p<.01; ***p<.001
In table 7, the last OLS regressions are executed on the dependent variable return on
assets with robust standard errors. Also, in this table the t-statistics are stated in parentheses.
In model 1, 2 and 4 CEO duality has a significant influence with a coefficient of -3.15, t(122) = -2.46, p < .001 in the first model. The values are in this regression table relatively lower than compared to the regressions on profit margin and return on equity. However, there is still strong support to reject H10.
There is no evidence that the moderating influence of board independence has an impact on firm performance. However, the influence of board size is significant on a 0.1% level. Hence, there has been found enough evidence to reject the H30 based on model 2 and 4, t(122) = 2.78, p < .001 and t(122) = 2.57, p < .001 respectively. As expected, these findings are in line with the regression models on the other two dependent variables, however the coefficients in this table show a less strong influence on return on assets than on profit margin and return
on equity.
Further, when comparing model 1 to model 3 and model 2 to model 4, the coefficient of CEO duality has declined by around 2% in each case. The reason for this percentual decrease is the inclusion of the controlling effect of the countries in the sample.
Consistent with the prevailing literature, the negative influence of board size on firm performance indicates that larger boards might work less effectively to achieve their monitoring tasks compared to smaller boards. As expected, the effect of board independence on firm performance suggests that a larger proportion of independent directors positively impacts firm performance. However, these data provide no evidence that the moderating variables influence the relationship between CEO duality and firm performance. Nevertheless, the interaction term of board size with CEO duality does show a significant influence.
Table 7: OLS regression (robust) on dependent variable return on assets
(1) (2) (3) (4)
Variables Return on assets Return on assets Return on assets Return on assets
CEO Duality -3.147*** -10.81*** -1.856 -8.412*** (-2.456) (-2.304) (-1.141) (-1.862) Board Independence 0.731 0.295 0.868 0.238 (0.431) (0.161) (0.446) (0.119) Board size(ln) -0.604 -1.812 -0.154 -1.210 (-0.449) (-1.426) (-0.109) (-0.918) Firm size(ln) 0.101 0.00374 0.161 0.0571 (0.347) (0.0127) (0.503) (0.174) Leverage 0.0797*** 0.0767 0.0805 0.0802 (1.662) (1.643) (1.592) (1.613) Belgium 0.487 0.364 (0.292) (0.230) Switzerland 1.141 1.716 (0.714) (1.100) Germany 1.585 2.035 (0.991) (1.316) France -0.402 -0.365 (-0.200) (-0.181) Great Britain 1.405 1.581 (0.813) (0.956) Netherlands 1.163 1.907 (0.537) (0.872) Sweden 16.05*** 15.25*** (9.495) (9.165) Duality*Board Independence 0.144 -1.759 (0.0316) (-0.360) Duality*Board Size 0.791*** 0.783*** (2.775) (2.569) Constant 4.190 8.315*** 1.081 4.787 Country dummy (0.964) NO (1.905) NO (0.240) YES (1.132) YES Observations 123 123 123 123 R-squared 0.091 0.142 0.160 0.207 t-statistics in parentheses *p<.05; **p<.01; ***p<.001
V. Discussion and concluding remarks
This paper examined the influence of CEO duality on firm performance on European publicly listed firms. This study differentiates from prior work by focusing on European firms rather than US firms. The distinction of board structures between the two business environments sheds a new light on the current corporate governance discussion regarding CEO duality. This study contributes to this debate by providing novel evidence on reasons for firms why or why not to select a certain board leadership structure. This can be either a duality or a non-duality structure.
To determine the influence of CEO duality on firm performance, a research sample of 123 European publicly listed firms, spread over eight different countries, has been used. Several corporate governance variables, such as board size and board independence, and control variables have been used in an OLS regression to find the results of the influence of CEO duality on firm performance. Three market-based performance measures were selected to assess firm performance (ROE, ROA, PM).
The empirical results in this paper show an overall significant and negative relationship between CEO duality and firm performance. Based on the fact that all three performance measures were negatively influenced by a duality structure, there has been found sufficient evidence to reject the H10 formulated in this paper. Furthermore, the significant effect of the interaction variable between CEO duality and board size indicates that larger boards weaken the power of a CEO (concerning a duality structure) and thus have a positive influence on firm performance. This is supportive evidence in order to refute the H30. These findings are in line with the agency theory, which states that CEO duality might result in managerial entrenchment leading to a decrease in firm performance and also a reduction in shareholder wealth. The dual leadership structure diminishes the ability of the board to exercise its governance function. Hence, firm performance is being negatively influenced by CEO duality on European publicly listed firms. This provides the answer to the research question of this study.
Several other studies found also evidence endorsing the agency theory (Rechner & Dalton, 1991; Pi & Timme, 1993; Duru et al., 2016). They all came to the conclusion that firm performance got negatively influenced by CEO duality. Whereas those studies based its findings on US firms, this paper focused on firms in Europe. However, the outcome of the research resulted in a similar conclusion.
On the other hand, several other researchers have found evidence in favour of the stewardship theory (Donaldson & Davis, 1991; Lin, 2005; Peng et al., 2010). The difference
regarding methodology lies in the fact that this study only uses accounting-based performance measures. It might be possible that the use of market-based performance measures revealed a different relationship between CEO duality and firm performance. Besides, most of the studies did not have to control for country effects because their sample was based in just one country. The inclusion of the country dummy variables in this study showed that CEO duality and firm performance vary across countries.
One of the main theoretical implications of this study is that CEO duality is not necessarily a bad thing. Some firms benefit form a duality structure, while others do not (Boyd, 1995; Brickley et al., 1997; Elsayed, 2007). There does not exist one universal leadership structure. Both leadership structures have costs and benefits. Whereas Brickley et al. (1997) claimed that the costs of separation of the two positions would outweigh the advantages of a non-duality structure and thereby supporting the stewardship theory, Rechner and Dalton (1991) came to the conclusion that non-duality structured firms performed better than firms with CEO duality.
As every other study, this research also features some limitations. The effect of CEO duality on firm performance can also be influenced by the different industries the companies operate in. In this research, there has not been controlled for industry effects. Additionally, a larger sample would have probably led to different results, which could have provided a better perspective on CEO duality and firm performances in Europe. Lastly, a restricted number of variables related to corporate governance were employed in this research, which confines the generalizability of the results. Furthermore, the utilization of the accounting-based performance measures in this study might have been not the best indicator for firm performance.
So, a suggestion for further research is a combination of accounting- and market-based performance measures because the latter reflect risk-adjusted firm performance. This study examined also the moderating effect of board size and board independence on the relationship between CEO duality and firm performance. Future research will enhance from a similar examination of such interdependencies between other governance mechanisms and the environment of the market (for example: bankruptcy laws, firm competition).
The outcome of this study is important to investors, regulators and governance experts. The results underline the difficulty of evaluating the reasons for CEO duality. Despite the fact that much needs to be explored on the relationship between CEO duality and firm performance, the results of this study imply that CEO non-duality is favourable for firm performance in Europe.
VI. References
1. Abdullah, S. N. (2004). Board composition, CEO duality and performance among Malaysian listed companies. Corporate Governance: The international journal of
business in society.
2. Adams, R. B., Almeida, H., & Ferreira, D. (2005). Powerful CEOs and their impact on corporate performance. The Review of Financial Studies, 18(4), 1403-1432.
3. Andrews, K. R., & David, D. K. (1987). The concept of corporate strategy (Vol. 101). Homewood, IL: Irwin.
4. Baliga, B. R., Moyer, R. C., & Rao, R. S. (1996). CEO duality and firm performance: What's the fuss?. Strategic management journal, 17(1), 41-53.
5. Berg, S.V. and S.K. Smith, 1978, CEO and Board Chairman: A Quantitative Study of Dual v. Unitary Board Leadership, Directors and Boards, 3, 34–49.
6. Bermig, A. (2012). German Supervisory Board Size and Composition: A Detailed Evaluation of Their Effects on Performance, Earnings Management and Cash Holdings (Doctoral dissertation).
7. Boyd, B. K. (1995). CEO duality and firm performance: A contingency model. Strategic management journal, 16(4), 301-312.
8. Brickley, J.A., Coles, J.L. & Jarrell, G. (1997). Leadership structure: separating the CEO and chairman of the board. Journal of Corporate Finance, 3, pp.189–220. 9. Chaganti, R. S., Mahajan, V., & Sharma, S. (1985). Corporate board size,
composition and corporate failures in retailing industry. Journal of management
studies, 22(4), 400-417.
10. Dalton, D. R., Daily, C. M., Ellstrand, A. E., & Johnson, J. L. (1998). Meta‐analytic reviews of board composition, leadership structure, and financial
performance. Strategic management journal, 19(3), 269-290.
11. Dang, C., Li, Z. F., & Yang, C. (2018). Measuring firm size in empirical corporate finance. Journal of Banking & Finance, 86, 159-176.
12. Davis, J. H., F. D. Schoorman and L. Donaldson (1997), Towards a Stewardship Theory of Management, Academy of Management Review, 22 (1): 20-47.
13. Donaldson, L. and J. H. Davis (1991), Stewardship Theory or Agency Theory: CEO Governance and Shareholder Returns, Australian Journal of Management, 16(1): 49-69.
14. Duru, A., Iyengar, R. J., & Zampelli, E. M. (2016). The dynamic relationship between CEO duality and firm performance: The moderating role of board
independence. Journal of Business Research, 69(10), 4269-4277.
15. Eisenhardt, K. M. (1989). Agency theory: An assessment and review. Academy of
management review, 14(1), 57-74.
16. Elsayed, K. (2007). Does CEO duality really affect corporate
performance?. Corporate governance: an international review, 15(6), 1203-1214. 17. Fama, E. F., & Jensen, M. C. (1983). Separation of ownership and control. The
journal of law and Economics, 26(2), 301-325.
18. Fauver, Larry & E. Fuerst, Michael (2006). Does good corporate governance include employee representation? Evidence from German corporate boards, Journal of
Financial Economics, 82, 673-710.
19. Finkelstein, S. & D'Aveni, R. A. (1994). CEO Duality as a Double-Edged Sword: How Boards of Directors Balance Entrenchment Avoidance and Unity of
Command. Academy of Management Journal, 37, 1079 -1109.
20. Fooladi, M., & Nikzad Chaleshtori, G. (2011). Corporate governance and firm performance. In International Conference on Sociality and Economics Development (ICSED 2011), 17 -19.
21. Goyal, V. K., & Park, C. W. (2002). Board leadership structure and CEO turnover. Journal of Corporate finance, 8(1), 49-66
22. Gujarati, D. N., & Porter, D. C. (1999). Essentials of econometrics (Vol. 2). Singapore: Irwin/McGraw-Hill.
23. Harrison, J. R., Torres, D. L., & Kukalis, S. (1988). The changing of the guard: Turnover and structural change in the top-management positions. Administrative
Science Quarterly, 211-232.
24. Hermalin, B. E., & Weisbach, M. S. (1998). Endogenously chosen boards of directors and their monitoring of the CEO. American Economic Review, 96-118.
25. Hewa Wellalage, N., & Locke, S. (2011). Does CEO duality really matter? Evidence
from an emerging market. Evidence from an Emerging Market (May 29, 2011).
Corporate Ownership & Control, 8(4).
26. Iyengar, R. J., & Zampelli, E. M. (2009). Self‐selection, endogeneity, and the relationship between CEO duality and firm performance. Strategic Management
27. Jackling, B., & Johl, S. (2009). Board structure and firm performance: Evidence from India's top companies. Corporate Governance: An International Review, 17(4), 492-509.
28. Jensen, M. C. and W. H. Meckling (1976). ‘Theory of the firm: Managerial behavior, agency costs, and ownership structure’, Journal of Financial Economics, 3, 305–350. 29. Joskow, P., Rose, N., Shepard, A., Meyer, J. R., & Peltzman, S. (1993). Regulatory
constraints on CEO compensation. Brookings Papers on Economic Activity.
Microeconomics, 1993(1), 1-72.
30. Kajola, S. O. (2008). Corporate governance and firm performance: The case of Nigerian listed firms. European journal of economics, finance and administrative
sciences, 14(14), 16-28.
31. Kholeif, A. (2008). CEO duality and accounting-based performance in Egyptian listed companies: A re-examination of agency theory predictions. Research in Accounting in
Emerging Economies, 8, 65-98.
32. Krivogorsky, V. (2006). Ownership, board structure, and performance in continental Europe. The International Journal of Accounting, 41(2), 176-197.
33. Lin, Y. F. (2005). Corporate governance, leadership structure and CEO compensation: Evidence from Taiwan. Corporate Governance: An International Review, 13(6), 824-835.
34. Linck, J. S., Netter, J. M., & Yang, T. (2008). The determinants of board structure. Journal of financial economics, 87(2), 308-328.
35. Lipton, M., & Lorsch, J. W. (1992). A modest proposal for improved corporate governance. The business lawyer, 59-77.
36. Majumdar, S. K., & Chhibber, P. (1999). Capital structure and performance: Evidence from a transition economy on an aspect of corporate governance. Public choice,
98(3-4), 287-305
37. Miller, D., & Friesen, P. H. (1977). Strategy‐making in context: ten empirical archetypes. Journal of Management Studies, 14(3), 253-280.
38. Mintzberg, H., & Waters, J. A. (1982). Tracking strategy in an entrepreneurial firm. Academy of management journal, 25(3), 465-499.
39. Peng, M. W., Zhang, S., & Li, X. (2007). CEO duality and firm performance during China's institutional transitions. Management and Organization Review, 3(2), 205-225.
40. Pi, L. and S. G. Timme (1993). 'Corporate control and bank efficiency', Journal of
Banking and Finance, 17, 515-530.
41. Raheja, C. G. (2005). Determinants of board size and composition: A theory of corporate boards. Journal of financial and quantitative analysis, 40(2), 283-306. 42. Ramdani, D., & Witteloostuijn, A. V. (2010). The impact of board independence and
CEO duality on firm performance: A quantile regression analysis for Indonesia, Malaysia, South Korea and Thailand. British Journal of Management, 21(3), 607-627. 43. Rechner, P. K., & Dalton, D. R. (1991). CEO duality and organizational performance:
A longitudinal analysis. Strategic Management Journal, 12(2), 155-160.
44. Shleifer, A., & Vishny, R. W. (1997). A survey of corporate governance. The journal
of finance, 52(2), 737-783.
45. Wan, D., & Ong, C. H. (2005). Board Structure, Process and Performance: evidence from public‐listed companies in Singapore. Corporate Governance: An International
Review, 13(2), 277-290.
46. Zahra, S. A., & Pearce, J. A. (1989). Boards of directors and corporate financial performance: A review and integrative model. Journal of management, 15(2), 291-334.
