Does the Leadership of the Board of Directors Affect Corporate Performance? Based on the Empirical Research of China’s SMEs

ABSTRACT

This study takes the chairman as the starting point to study whether the chairman-CEO duality can affect the performance, with a sample of 3568 the board of directors of the listed company leadership characteristics and performance during 2008–2018 in NEEQ market. The results found that the chairman as CEO is conducive to improving the business performance. Through the sample study of enterprise size division, it is found that in small enterprises, the chairman as CEO has a more significant positive impact on business performance than in large enterprises.

KEYWORDS:

• CEO Duality
• business performance
• modern stewardship theory
• resource dependence theory

1. Introduction

In recent years, The NEEQ market in China has witnessed rapid development. By the end of September 2020, the number of SMEs listed on the NEEQ market reached 8,402. The total share capital reached 541.367 billion shares, and the turnover reached 353.83663 million yuan. The turnover of the NEEQ market in 2020 has exceeded 100 billion yuan, reaching 101.155 billion yuan, a significant increase from 62.496 billion yuan in the same period of 2019 (up to October 18 of the same year). The NEEQ market, as an important part of China’s multi-layer capital market, is a typical representative of China’s small and medium-sized enterprises. Studying the NEEQ market is helpful to understand the profitability and growth of China’s SMEs and the vitality and potential of China’s economy. On the other hand, good corporate governance is of great significance for SMEs to maintain profitability and growth. Most of the traditional research literature takes large companies as the research object, and many studies point out that the duality of CEO weakens the supervision and control function of the board of directors based on agency theory. Does this conclusion apply to small and medium-sized enterprises?

CEO duality, defined as the practice that the same person holds the position of chairman and CEO at the same time, is an important leadership characteristic of the board of directors (Finkelstein and D’Aveni 1994). The impact of CEO duality on corporate performance is widely debated in academia and business (Doan 2020). The results of 31 studies by Dalton, Daily, Ellstrand, and Johnson were inconclusive, ranging from positive to negative to statistically insignificant (Dalton et al. 2007). CEO duality has different effects on the company in different market environments (Daily and Dalton 1994). Dalton and Dalton argue that there is almost no consistency between CEO duality and financial performance in existing studies (Dalton et al. 2011). Shufang Hsua and colleagues found that the performance of dual CEOs was not significantly higher or lower than that of non-dual CEOs (Shufang Hsua et al., 2019). The key roles of CEO and chairman of the board of directors are different in a company. CEO is responsible for the management and operation of the company, while the chairman is responsible for leading the board of directors (Doan 2020). Although the relationship between CEO duality and corporate performance has been extensively studied, there is little substantial and systematic empirical evidence for this relationship (Dalton et al. 1998; Dalton and Dalton 2011; Dalton et al. 2007; Faleye 2007; Heracleous 2008).

Therefore, empirical studies in recent years seek new contexts to help them reveal the effect of the strategic importance of CEO duality from various aspects (Dalton and Dalton 2011; Duru, Iyengar, and Zampelli 2016; Krause 2017; Krause, Semadeni, and Cannella 2014; Yang and Zhao 2014). Krause, Semadeni, and Cannella (2014) discuss mitigation attributes that might change the strength or direction of relationships in their study. Duru et al. (2016) found that the extent of board independence reduces the negative impact of CEO duality on corporate performance, and Hsu et al. (2019) found that when considering the impact of information cost on corporate performance, enterprises with high information cost of CEO duality are more likely to experience a decline in corporate performance than those with non-dual leadership. Tang (2017) found that when the CEO has the dominant power and the board of directors is controlled by outside directors, the duality of the CEO harms the corporate performance. However, Yang and Zhao (2014) showed that when the competitive environment changes, CEO dualistic enterprises perform better than non-dualistic enterprises. Elsayed (2010) emphasized that the optimal leadership structure varies with the company’s operating environment (He and Wang 2009; Krause 2017; Krause and Semadeni 2013; Ramdani and van Witteloostuijn 2010)

Large enterprises in Contemporary China are mainly state-owned enterprises. Due to the state-owned nature of enterprises, they are often monopolized in the industry and obtain high profits by under such monopolization (Chang et al. 2018). For small and medium-sized enterprises in China, their situation is quite different from that of large enterprises, such as insufficient government support and fierce industry competition. Based on the argument above, this paper, aiming at the specific situation of small and medium-sized enterprises in China, explores the influence of chairman-general manager duality on business performance. Based on the Ordered Probit model, this paper empirically examines the relationship between chairman and general manager and corporate performance of 3568 listed companies on the NEEQ market from 2008 to 2018. The results show that the chairman-general manager duality is beneficial to the company’s operating performance.

The marginal contribution of this paper lies in the following aspects: first, most of the existing works of literature take the whole board of directors or independent directors as the main research object to explore the relationship between their characteristics and corporate performance. This paper studies whether the chairman is the CEO or not. The research of this paper helps investors to identify excellent companies based on the characteristics of board leadership. Secondly, most of the existing researches focus on companies in the foreign market and domestic mainboard market. This paper focuses on the NEEQ market, which is densely populated by small and medium-sized enterprises and studies the relationship between board leadership characteristics and enterprise performance. The NEEQ market has developed rapidly in recent years, but due to the late start system and other reasons, there are still many problems to be solved. If these problems cannot be properly solved, they will seriously restrict the development potential of NEEQ enterprises and even drag down the development of the whole market. Small-sized and medium-sized enterprises are chosen because they account for the vast majority of the total number of Chinese enterprises and are the main force of micro subjects of China’s national economy. To explore how to improve the operation performance of small-sized and medium-sized enterprises, the development of small and medium-sized enterprises, to ensure the healthy operation of the national economy, increase employment opportunities, promote stable economic growth has an important significance. Third, most existing kinds of literature take a single indicator as a measure of operating performance, such as earnings per share (EPS), Tobin’s Q (ROA), Return on equity (ROE), net market value (MBV), and EBIT (before interest and tax), etc. (Duru et al. 2016). Earnings per share (EPS) does not take into account the time value and risk of currency, and cannot avoid the short-term behavior of enterprises. Tobin’s Q is a market-based measure of corporate performance, reflecting the present value of future cash flow based on current and future information (Singh et al. 2018), which is restricted by the development of China’s capital market.Net market value (MBV) is susceptible to the impact of stock market effectiveness and prosperity. To ensure the robustness of the empirical results, this article refer to the way of the Boston matrix, a method of cutting plane according to the different dimensions of classification and can be from two aspects of profitability and growth of company business performance comprehensive measure, while studying common profitability and development ability, will be explained variables is divided into four interval, avoided due to the inaccurate of single variables affect the empirical result, makes the measurement more comprehensive science.

Further study found that the ownership nature of the enterprise has no regulating effect on the relationship between the chairman-general manager duality and business performance; through the sample study of enterprise size division, it is found that in large-sized enterprises, the chairman-general manager duality has a more significant positive impact on business performance than in small-sized enterprises. After changing the model and variables, the conclusions of this paper are still robust.

The following arrangement of this paper is as follows: The second part is the theoretical basis and research hypothesis; the third part is the research design, which describes the samples, variables, and models of this paper, the fourth part is the empirical results and analysis; the fifth part is the robustness test; the sixth part is the conclusion and enlightenment.

2. Theoretical Basis and Research Hypothesis

The duality of CEO gives the CEO greater decision-making power, meanwhile, reduces the supervisory capacity of the board of directors (Dalton and Dalton 2011).On the one hand, it can bring stronger and more decisive leadership(Dalton and Kesner 1987). Under this circumstance, CEOs may make use of their position power to adopt self-interested behaviors that are not in line with the best interests of the company (Adams et al. 2005; Fama and Jensen 1983; Mizruchi 1983). Therefore, Elsayed (2010) discovers that the overall effect of CEO duality varies in different circumstances, as this effect is a trade-off between costs and benefits, which may depend on the actual situation. When a unified command is considered importantly, the benefits of CEO duality may outweigh the costs (Boyd 1995; Mueller and Barker 1997; Peng, Zhang, and Li 2007). In contrast, when THE AGENCY problem of the CEO was considered to be serious, stagger the board of directors may exacerbate the agency problem and lead to CEO pester (Amihud and Stoyanov 2017). When the monitoring of the board of directors became weak and external monitoring became insignificant, improper investment allocation and inefficiency would be caused (Aktas et al. 2019).In other words, any factor that affects the cost or benefit of CEO duality may mitigate the impact of CEO duality on business performance.

Any corporate governance arrangement has costs (defects). CEO duality of the difference between benefits (i.e., unity of command) and costs (i.e., agency problems) is mainly manifested in the agency theory and the modern housekeeper theory. CEO duality, by enhancing CEO power relative to the board, compromises the board’s monitoring and disciplining functions. By compromising the board’s monitoring and disciplining functions, CEO duality may also compromise the quality of the firm’s strategic decisions. Under the duality structure the CEO is more apt to make decisions aligned with his or her self-interest but deviant from the shareholders’ interest (i.e., employment contracts) . In a binary structure, the CEO is more likely to make bad or extreme decisions because he or she is removed from the board’s oversight function. Whereas agency theory highlights the costs of CEO duality, a stewardship theory perspective suggests that CEO duality has potential benefits, especially enhancing the unity of command at the top by clearing the confusion about who is the CEO or chair is in charge (Mutlu et al. 2018). The unity of command helps establish clear lines of authority and responsibility within the firm and thus has substantive and symbolic values. It enhances the power of the CEO and thereby enables fast, decisive decisions. This is very consistent with the characteristics of small and medium-sized enterprises to be studied in this paper. As we all know, small and medium-sized enterprises need strong, quick decisions.

The separation of ownership and control is a characteristic of most large companies in the modern economy, in which the CEO does not actually own the company and may therefore pursue his own interests at the expense of shareholders. According to the principal-agent theory, with the separation of ownership and control rights in modern companies, managers are given decision-making rights and have no control over shareholder capital, because managers and shareholders’ preferences cannot always be consistent (Berle and Means 1932; Dalton et al. 2007). In addition to the information asymmetry between managers and owners, managers, as the agents of shareholders, cannot always take actions for the maximization of shareholders’ value, but take opportunistic behaviors to grab too many benefits for themselves at the expense of shareholders’ interests. As an important part of corporate governance structure, the board of directors exercises its function of supervising managers by hiring, dismissing, and motivating the senior management team, etc., which is the highest point of modern corporate decision control system(Fama and Jensen 1983). It alleviates agency problems caused by the separation of ownership and control (Fama and Jensen 1983). There is no doubt that boards have become the key governance arrangement that limits the principal-agent problem for chief executives. The effectiveness of the board of directors in performing supervision and monitoring functions depends on the CEO’s power relative to the board of directors (Finkelstein, Hambrick, and Cannella 2009). In essence, the supervisory and disciplinary functions of the board are achieved by limiting the power of the CEO relative to the board. Therefore, the principal-agent theory is an important theoretical basis for the arrangement that the CEO and the chairman are held by different persons.

The duality of CEO, by strengthening the power of CEO relative to the board of directors, weakens the supervisory function of the board of directors and has a negative impact on corporate performance (Dalton et al. 1998; Finkelstein 1992; Krause, Semadeni, and Cannella 2014). When the CEO has excessive power under a dual leadership structure, it provides legitimacy beyond the control of the board for the CEO to engage in management activities that deviate from the interests of shareholders, and creates a productive environment. In the absence of a clear separation of the two leadership roles, the role of the board of directors in supervising and managing opportunism is weakened (Zona 2012).In addition, under the dual structure, the CEO’s decisions are unlikely to be subject to strict scrutiny by the board of directors; bad or extreme decisions are therefore more likely to go unchallenged (Tang, Crossan, and Rowe 2011). When board supervision becomes weak and external supervision becomes insignificant, investment mismatch and inefficiency become potential channels for CEO duality to adversely affect the corporate value (Aktas et al. 2019). Studies have shown that CEO duality in leadership structure is associated with high volatility in corporate performance and poor corporate performance on average (Bebchuk, Cremers, and Peyer 2011; Tang, Crossan, and Rowe 2011). Glaser, Lopez-De-Silanes, and Sautner (2013) also believe that when decision management and decision control are entrusted to the same person, the supervision of the board of directors will be weakened, and external supervision, which seems to restrain the actions of the CEO, will be negligible, because the internal capital market provides a mean for CEO to avoid supervision from the external financial market.

The duality of the CEO weakens the effectiveness of the board in terms of control and monitoring functions; this increases information asymmetry and agency costs (Boyd, Gove, and Hitt 2005; Finkelstein and D’Aveni 1994). At the same time, companies with dual CEO leadership structure are more likely to have a lower level of voluntary disclosure, which leads to a lack of transparency and a high degree of information asymmetry in turn (Allegrini and Greco 2013; Gul and Leung 2004). Byard, Li, and Weintrop (2006) study the relationship between corporate governance and the quality of information available to financial analysts and found that when the CEO was also the chairman of the board, the accuracy of analysts’ predictions would decline. The results also suggest that CEO duality may lead to higher levels of information asymmetry for shareholders. In addition, Bhattacharya, Desai, and Venkataraman (2013) believe that the impact of earnings quality on information asymmetry is affected by the corporate information environment and is more significant for companies operating in a relatively poor information disclosure environment, which means that earnings management will increase the uncertainty of investors. Such uncertainty will exacerbate information asymmetry between managers and shareholders, which is a necessary condition for earnings management (Dye 1988; Richardson 2000; Trueman and Titman 1988). Saleh, Iskandar, and Rahmat (2005) proved this point of view and found that the dual leadership structure of CEO improves the possibility of earnings management, indicating that the dual nature of CEO increases the level of information asymmetry.

Therefore, from this perspective, CEO duality is undesirable because it gives too much power to individual executives, weakens the board’s monitoring, and leads to management barriers; moreover, it may lead to the intensification of internal and external information asymmetry, which is likely to lead to a large increase in agency problems, thus harming the company’s performance. Only by separating the functions of managers and boards of directors can agency costs be reduced and corporate performance be improved (Fama and Jensen 1983). This view is widely supported by practitioners and a growing number of scholars who advocate the separation of two jobs, who believe that the combination of two jobs weakens corporate governance, while the separation of the two roles enables the CEO to focus on the operation of the business and the chairman to focus on the operation of the board (Dalton et al. 1998; Iannelli 2013; Krause 2017; Lublin 2009). Through empirical research, it is found that when two jobs are combined, CEO pay is higher and CEO turnover is less sensitive to corporate performance (Core, Holthausen, and Larcker 1999; Goyal and Park 2002). The data of 141 Fortune 500 companies with stable board leadership structure from 1978 to 1983 shows that there is a positive correlation between corporate performance and independent leadership (Rechner and Dalton 1991). The data of 112 American Banks from 1987 to 1990 shows that Banks with independent property rights have a higher return on assets (Pi and Timme 1993). Data from 22 non-US countries indicate that board independence is significantly positively correlated with company performance, especially in countries with low investor protection (Dahya, Dimitrov, and McConnell 2008). This conclusion is further confirmed (Aggarwal et al. 2009; Bruno and Claessens 2010).

Based on this theory, this paper proposes the following hypothesis:

H1a: In NEEQ enterprises, when other factors remain unchanged, the concurrent role of chairman and CEO is not conducive to improving the enterprise’s business performance.

From the perspective of management theory, researchers have long recognized the benefits of CEO duality, especially the enhancement of unified command at the top (Dalton et al. 2007; Davis, Schoorman, and Donaldson 1997; Finkelstein and D’Aveni 1994) .The dual nature of CEO combined with the authority of these two positions can provide clear leadership to better coordinate the proposition and implementation of corporate strategy, make decisions more effective, and thus improve corporate performance, especially in a relatively volatile business environment or high economic policy uncertainty(Anderson and Anthony 1986; DeBoskey, Luo, and Zhou 2019; Donaldson and Davis 1991; Peng, Zhang, and Li 2007; Pham, Oh, and Pech 2015; Stoeberl and Sherony). For example, in complex and less generous environments, in transition economies, and in turnaround situations, the benefits of CEO duality (especially unified command) may be particularly prominent. In such cases, the benefits of CEO duality may outweigh the inherent agency costs, so CEO duality may have a positive impact on corporate performance (Boyd 1995; Mueller and Barker 1997; Peng, Zhang, and Li 2007).

In addition, the modern housekeeper theory emphasizes the managers as a social person and hold the opinion that the principal-agent theory to the inherent self-interest opportunism behavior of managers and lazy behavior assumption is not appropriate, the manager has strong demand, self-actualization needs, have a strong internal motivation, have a high sense of responsibility at the same time. This allows them to work hard, to be able to consider the question from long-term interests and corporate interests, the enterprise managers and employers are not fully driven by their personal goals, but in the enterprise plays a seek cooperation, the pursuit of the enterprise target steward role (Davis, Schoorman, and Donaldson 1997). Modern butler theory has more advantages in ethics and is a better choice for organizations to pursue long-term interests. The governance model, based on modern butler theory, is more reasonable than that based on principal-agent theory because what enterprises need more is long-term interests. Although the research on modern butler theory has only been done for around 20 years, and many research issues are not in-depth enough, modern butler theory provides a new approach for corporate governance (Caldwell and Karri 2005). In this view, the agent may have motives except for money, such as achievement, recognition, and reputation. They complete challenging work to realize the inner satisfaction, with the right to exercise the rights and responsibility for peer reputation and boss recognition, and controls the company through the establishment of the company system development direction, strengthen their sense of accomplishment, personal self-esteem and corporate reputation is closely linked, the CEO incentive to pursue the best interest of shareholders, become a trusted butler (Donaldson and Davis 1991). Therefore, the board of directors and the management should develop a relationship of mutual cooperation and full trust, and the shareholders should fully trust the CEO, provide active support for the CEO’s work, and help develop the CEO’s leadership (Keay 2016). Keay (2016) found that in the board governance structure of China’s private listed companies, the chairman and general manager are usually held by one person. Bouillon M L, et al’s empirical studies have demonstrated that when managers accept organizational strategy and organizational goals, it can produce a strong attachment to the organization, earn higher psychological, stewards on behavior is behavior, and a housekeeper behavior of managers can help enterprises get higher long-term organizational performance (Bouillon et al. 2006). Corbetta and Salvato (2010) also think, in the housekeeper role in the enterprise means that managers and employees should not be a personal economic benefit maximization of the opportunist, on the contrary, he should be generous, loyal, dedication, and the spirit of the humanities concern, as a member of the enterprise oneself, make oneself of the behavior consistent with the goal of enterprise, for the enterprise benefit maximization, the pursuit of the welfare of the entire enterprise.

Based on this, scholars have found that the two-job structure will bring some beneficial effects, such as effective decision-making without disclosing strategically important information, rapid implementation, and flexible adjustment of decisions under highly uncertain conditions. CEO of the company unparalleled expertise and reflect the benefits of strong management ability, because the CEO “usually is perhaps the most understanding of strategic challenges and opportunities faced by companies”, so the two jobs one CEO should be able to coordinate the board of directors to act quickly and implement strategy, so that the company has a competitive advantage, especially in the difficult business environment(Jensen and Meckling 1995). The integration of the two jobs can also make the leadership and direction of the company clearer and promote effective communication with external parties (Dalton et al. 1998). It can have a positive impact on the company’s performance, because it can ensure a cohesive leadership relationship, and indicate stable development of the company and inspire confidence in the management of the company(Donaldson and Davis 1991). At the same time, the title of the chairman is a part of the CEO incentive contract. If a company does not award the title of chairman to CEO, its CEO’s work motivation may be reduced, and it may not even consider leaving the company (Brickley, Coles, and Jarrell 1997).

Resource dependence theory: The company is regarded as an open system of the company’s ability to survive depends on its ability to obtain key resources from the external environment (Pfeffer and Salancik 1978), that the board of directors and executives succession help the company to get external resources, and reduce reliance on resources enterprises, to provide resources to the CEO is an important function of the board of directors, Director business expertise and reputation for CEO provide information and resources companies doing business need (Hillman and Dalziel 2003; Kor and Sundaramurthy 2009). Companies often follow the process of “crossing”, which are required to keep the title of retired CEO and chairman, so that the valuable information to the new CEO and the board at the same time have a probation period to supervise the behavior of the new CEO. During this period, the new CEO often also has other titles, such as President and COO. If he successfully passed the test, the new CEO will receive an additional chairman title, the former chairman will resign from the board of directors. The two-job integration process simplifies the transition from active service to retirement for the elderly CEO, reduces the cost of succession planning, and reduces the possibility for the CEO to stay in this position (Vancil 1987). In this study, it is further found that the chairman who arranges the separation of two jobs will form his own agency problem in the form of “who will supervise the supervisor” (Brickley, Coles, and Jarrell 1997). In addition, the two-role leadership structure enables the chairman to have a higher power, more incentives for board members to provide resources, and at the same time reduces the possibility of conflicts caused by the division of roles of chairman and CEO (Boyd, Haynes, and Zona 2011). Ballinger and Marcel (2010) report that the succession of interim CEOs is associated with a low performance during the tenure of interim CEOs, and the duality of CEOs moderates the impact of this type of succession on business performance. Similarly, Krause and Semadeni (2013) found that when the current performance is poor (high), the separation of CEO and chairman will have a positive (negative) impact on the company’s future performance, among which the demotion has the most significant impact.

Therefore, compared with the leadership structure with separation of two jobs, the leadership structure with the integration of two jobs is conducive to enhancing the board’s ability to influence the strategic decisions of the enterprise, which may further restrict or promote the process of the board’s deep and broad capital influencing the internationalization strategy of the enterprise(Park et al. 2018). Due to the high cost of acquisition and transfer the company’s specific information, the chairman and general manager of the joining together of two post eliminated between rights to a certain extent and distribution of conflict, reduce the information asymmetry, reduce the internal transaction cost, the board of directors and business management to reduce low efficiency, and improve corporate performance, Pfeffer and Salancik point out that the leadership of the CEO duality structure provided by the greater discretion enhances the CEO in the dynamic business environment’s ability to respond more quickly and while ensuring the resources that are critical to the company’s success, the combination of the two jobs is better suited to the rapidly changing market environment(Brickley, Coles, and Jarrell 1997; Pfeffer and Salancik 1978). Recently, Yang and Zhao (2014), relying on the exogenous impact of industry competition, provide evidence that when the competitive environment of enterprises changes, companies with dual CEO leadership structure perform better than those with non-dual CEO structure, and the performance difference between companies with dual CEO structure and those with better corporate governance becomes larger. Peng, Zhang, and Li (2007) found that CEO duality may be particularly valuable in situations of resource scarcity and dynamic environment.

In practice, compared with large enterprises, small-sized and medium-sized enterprises are often difficult to develop large-sized production and sales, have strong research and development capabilities, quality, technology, reputation, and marketing are generally lower than large enterprises, and it is difficult to quickly develop business advantages such as cost leadership strategy and product differentiation strategy. SMEs face greater competitive pressure and a stronger sense of crisis, so they need more institutional innovation and technological innovation. They have a lean organizational structure, fewer management level, and a more flexible operation mechanisms. In the face of the rapidly changing market, they are more adaptable, tolerant and responsive, and have more flexibility in operation. The integration of the two jobs means rapid decision-making, which can improve the decision-making execution ability of small-sized and medium-sized enterprises, reduce the cost of information transmission, and improve the anti-risk ability and survival ability of small-sized and medium-sized enterprises. Based on the above theories, this paper proposes the following hypotheses:

H1b: While other factors remain unchanged, the concurrent role of chairman and CEO is conducive to improving the business performance of SMEs.

3. Research Design

In sample collection, in order to ensure the validity of data and minimize the impact of extreme abnormal samples, we follow the following criteria in sample selection:(1) according to the guidance on industry classification of listed companies issued by China securities regulatory commission; (2) Excluding financial and insurance industry data; (3) Exclude companies with incomplete annual report disclosure data; (4) Eliminate individual companies with extremely abnormal index values. This paper makes a detailed analysis of the leadership characteristics of the board of directors of NEEQ enterprises from 2008 to 2018, and a total of 28,095 observed values of 3568 NEEQ enterprises are obtained. The data were obtained from the CSMAR database.

In this paper, the business performance of enterprises is taken as the explained variable (Y). Most of the existing studies use a single variable to describe business performance, such as return on assets (ROA), return on equity (ROE), net market value (MBV), and EBIT. Among them, EBIT is independent of the company’s financial leverage and can effectively avoid the impact of non-current account, but like ROA and ROE, it is prone to be manipulated by the management.Net market value (MBV) is based on market value and is less likely to be manipulated by management, but is vulnerable to the impact of stock market effectiveness and prosperity. To ensure the robustness of the empirical results, this paper through using Boston matrix, the new three board enterprise profitability (total return on assets) and growth (main business revenue growth) in accordance with the average of threshold, is divided into four areas (as shown in Figure 1), the value of 1 ~ 4, to measure the operating performance of enterprise be explained variables, the operating performance of companies to be explained variable, the larger the value of the proof of enterprise operating performance, the better. When the revenue growth rate of the main business and the return on total assets of the enterprise are both at a high level, this paper believes that the business performance is the best at this time, taking the value of 4. Secondly, as most domestic studies take profitability as the criterion of business performance, this paper values the operating performance with high return on total assets and low growth rate of main business income as 3. Thirdly, the operating performance with high growth rate of main business income and low return on total assets is taken as 2. Finally, the operating performance of both the growth rate of main business income and the return on total assets at a low level is 1. The business performance was divided by Boston matrix, and the company’s business performance was comprehensively measured from two aspects of profitability and growth, so as to avoid the impact of the inaccurate single variable on the empirical results.

Figure 1. Business performance matrix of NEEQ enterprises.

In this paper, the leadership characteristics of the board of directors (chairman and general Manager concurrently) are taken as explanatory variables (X) to control the influence of company size (GSGM), main income growth rate (DEVE), company equity (DGGB), DEBT ratio (INDU), industry (INDU) and other variables on enterprise operation performance. The explanations of specific variables studied in this paper are shown in Table 1.

Table 1. Explanation of variables

Variable types	The variable name	Variable symbol	Variable definitions
Explained variable	Business performance	Y	Matrix rating, 1 to 4
Explanatory variables	Leadership characteristics of the Board	X	The concurrent is equal to 1, and vice versa is equal to 0
Control variables	The company size	GSGM	The natural log of the total assets of a business
	Growth rate of main revenue	DEVE	Enterprise main income growth rate
	Company’s equity	GSGB	Shares issued by an enterprise
	Asset-liability ratio	DEBT	Total liabilities/total assets of the business
	Industry variable	INDU	According to the China Securities Regulatory Commission’s 2012 First-level industry Directory, the values are 1–19 respectively

This table introduces all the variables used in the above models and their definition descriptions.

In order to verify the research hypothesis, the main regression model is established as follows:

(1) $Y_k\left(level\right)={\alpha }_{it}+{\beta }_1{X}_{it}++{\beta }_{2}GSG{M}_{it}+{\beta }_{3}DEV{E}_{it}+{\beta }_{4}GSG{B}_{it}+{\beta }_{5}DEB{T}_{it}+{\beta }_{6}IND{U}_{it}+{\epsilon }_{it}$(1)

4. Empirical Results and Analysis

4.1. Descriptive Statistics

Table 2 is the descriptive statistical results of the variables studied in this paper. Through statistical results can be seen, be explained variable average value is 2.812, the business performance of the new three board on enterprise performance level in the medium level, the standard deviation of 0.626, in a smaller range, between enterprise performance level fluctuation is small, performance is stable, the level of business performance difference between enterprises; the average value of explanatory variable X is 0.252, indicating that the proportion of chairman of the board holding the concurrent post of general manager is relatively low in Enterprises on the New Third Board. Only 25.2% of enterprises adopt the leadership structure of chairman of the board and general manager. Control variable size of company equity, asset-liability ratio, the poor were similar, average and median value is close to, you can see the new three board enterprise in company scale, equity and debt levels are relatively stable, less volatile, little differences between businesses, the above several control variables of the standard deviation in the smaller level, also determine the characteristics, the main business revenue growth, the differences between enterprises, from its maximum-minimum value, standard deviation value in the higher index can be seen in this feature.

Table 2. Descriptive statistics

	X	Y	INDU	GSGM	GSGB	DEVE	DEBT
Mean	0.252	2.812	4.806	21.989	20.005	4.701	0.469
Median	0.000	3.000	3.000	21.828	19.904	0.120	0.436
Maximum	1.000	4.000	19.000	28.509	25.933	59411.550	96.959
Minimum	0.000	1.000	1.000	14.108	17.371	2266.483	0.195
Std. Dev.	0.434	0.626	3.581	1.346	1.043	390.328	0.913
Skewness	1.143	2.323	2.118	0.648	0.771	139.036	66.790
Kurtosis	2.306	7.377	7.247	4.258	4.345	20601.020	5945.612
Jarque – Bera	6278.147	44838.86	39590.49	3592.98	4606.717	4.67 e + 11	3.89 e + 10
Observations	26410	26410	26410	26410	26410	26410	26410

This table reports the result of descriptive statistics; all variables are provided in Table 2.

4.2. Pearson and Spearman Correlation Test

Table 3 shows the test results of Pearson correlation between explained variables, explanatory variables, and control variables in this study. Can be seen from the results of Pearson correlation coefficient, and performance characteristics of CEO and chairman of X Y correlation coefficient is 0.0174, the correlation coefficient is positive and significant coefficient is 0.0037, shows that the correlation to 1% significance level, it can be seen that, without considering other factors, chairman, CEO, can significantly improve the business performance of enterprises. In this paper, the assumption of H1b preliminary support. In the control variable, in addition to the main business revenue growth, other variables and business performance in a certain level significantly, and that without considering the influence of other factors, in this paper, the control variables and interpreted all have certain relevance, indicates that a reasonable selection in this paper, the control variables, the possible factors affecting the results of the experiment have played a very good control effect.

Table 3. Pearson correlation test

	X	Y	INDU	GSGM	GSGB	DEVE	DEBT
X	1.0000
Y	0.0174***	1.0000
INDU	0.0458***	0.0403***	1.0000
GSGM	0.0842***	−0.1713***	0.0461***	1.0000
GSGB	−0.0161***	−0.1835***	0.0529***	0.8271***	1.0000
DEVE	0.0208***	−0.0011	−0.0021	0.0071	0.0086	1.0000
DEBT	−0.0644***	−0.0174***	0.0302***	−0.0672***	0.0285***	0.0005	1.0000

*, ** and *** respectively indicate that the regression coefficient is significant at the level of 10%, 5% and 1%.

Table 4 illustrates the interpreted variables in this study, explains the Spearman correlation between variables and control variables of the inspection results. We can find the performance characteristics of CEO and chairman of X Y correlation coefficient is 0.0147, the correlation coefficient is positive and significant coefficient is 0.0143, this shows that the correlation to 5% significance level, although with Pearson correlation test results there exist certain differences, but also can support this article preliminary assumption H1b, and in the Spearman correlation test results, all the control variables have a certain correlation with the explained variables, and it is verified that the control variables selected in this paper are reasonable and play a good role in controlling the non-experimental factors that may affect the experimental results.

Table 4. Spearman correlation test

	X	Y	INDU	GSGM	GSGB	DEVE	DEBT
X	1.0000
Y	0.0147**	1.0000
INDU	0.0701***	−0.0524***	1.0000
GSGM	0.0770***	−0.1835***	0.1031***	1.0000
GSGB	−0.0155***	−0.1831***	0.0825***	0.8191***	1.0000
DEVE	0.1394***	0.0276***	0.1017***	−0.0277***	−0.0580***	1.0000
DEBT	−0.1554***	−0.1622***	0.1333***	0.4337***	0.3047***	−0.0161***	1.0000

*, ** and *** respectively indicate that the regression coefficient is significant at the level of 10%, 5% and 1%.

4.3. Basic Regression Analysis

Table 5 reports the regression results of Ordered Probit models that did not use tool variables. According to the regression results, it can be seen that the explanatory variables are significantly positively correlated with the explained variables at the significance level of 1%, and the model regression coefficient can be obtained through the conversion of the probability density function output by the software. Compared with the separation of the two roles of chairman and general manager, the probability of improving the business performance of NEEQ enterprises increases by 2.16%.The hypothesis H1b in this paper is supported.

Table 5. Basic regression results

Variable	Coefficient	Std. Error	Z – Statistic	Prob.
X	0.0589 * * *	0.0221	2.6612	0.0078
INDU	0.0289 * * *	0.0027	10.7615	0.0000
GSGM	0.2879 * * *	0.0124	23.2544	0.0000
GSGB	0.3069 * * *	0.0163	18.8220	0.0000
DEVE	0.0027 * * *	0.0002	12.1339	0.0000
DEBT	0.1178 * * *	0.0116	10.1621	0.0000
Observations	26410	26410	26410	26410
The Pseudo R – squared	0.0440	0.0440	0.0440	0.0440

*, ** and *** respectively indicate that the regression coefficient is significant at the level of 10%, 5% and 1%.

4.4. Robustness Test

4.4.1. Instrumental Variable Method to Solve Endogenous Problems

To eliminate endogeneity problems that may arise from mutual causation or omission of variables, the two-stage instrumental variable method (2SLS) was used in this paper to conduct experiments again, and instrumental variables were selected to conduct regression analysis on the business performance of enterprises. In this paper, the average of chairmen and general managers in corresponding regions of NEEQ enterprises is taken as the instrumental variable X1. Besides, due to the practice, the chairman of the compensation in standing at the top of the board of directors, general manager’s salary is the leading member of the executive position, based on this reality, this paper collected samples, chairman of the compensation and compensation, general manager of comparative analysis of data for the chairman and general manager compensation levels of the same enterprise, thinks that the enterprise adopts the chairman and general manager leadership structure of the joining together of two jobs, chairman and general manager for different enterprises pay levels, argues that the enterprise adopts the chairman and general manager leadership structure of separation of two jobs, defines the content as instrumental variable X2.

In the first stage, OLS regression was conducted between the instrumental variable and the original explanatory variable X, and the model was established as follows:

(2) $X=\alpha +{\beta }_1{X}_1+{\beta }_{2}GSGM+{\beta }_{3}DEVE+{\beta }_{4}GSGB+{\beta }_{5}DEBT+{\beta }_{6}INDU+\epsilon$(2)

(3) $X=\alpha +{\beta }_1{X}_2+{\beta }_{2}GSGM+{\beta }_{3}DEVE+{\beta }_{4}GSGB+{\beta }_{5}DEBT+{\beta }_{6}INDU+\epsilon$(3)

In the second stage, Ordered Probit regression was conducted again by replacing endogenous variables and business performance in the original model with instrumental variables, and the model was established as follows:

(4) $Y_k\left(level\right)=\alpha +{\beta }_1{X}_1+{\beta }_{2}GSGM+{\beta }_{3}DEVE+{\beta }_{4}GSGB+{\beta }_{5}DEBT+{\beta }_{6}INDU+\epsilon$(4)

(5) $Y_k\left(level\right)=\alpha +{\beta }_1{X}_2+{\beta }_{2}GSGM+{\beta }_{3}DEVE+{\beta }_{4}GSGB+{\beta }_{5}DEBT+{\beta }_{6}INDU+\epsilon$(5)

Table 6 shows the regression results of the two-stage instrumental variable method on the average of regional chairmen and general managers. In stage 1 × 1 of X can be seen in the OLS regression results, instrumental variable X1 and the original is interpreted variable X to 1% significance level is positive, explained the selection of instrumental variables to meet the requirements of the instrumental variable and endogenous variables are highly correlated, to extract the original part explain the exogenous variable, instrumental variable selection in this paper is more reasonable. In the second stage, the regression results of X1 on Y show that the regression coefficient of the average level represented by the instrumental variable is significantly positive at 1% level, which is significantly higher than the regression coefficient of the unused instrumental variable. The model regression coefficient can be obtained by converting the probability density function output by the software. Compared with the separation of the two jobs of chairman and general manager, the probability of improving the business performance of NEEQ enterprises increases by 37.06%.

Table 6. Regression results of the two-stage instrumental variable method

	Variable	Coefficient	Std. Error	T – Statistic	Prob.
Step1	C	1.2997 * * *	0.0517	25.1464	0.0000
	The X1	0.9128 * * *	0.0366	24.9413	0.0000
	INDU	0.0044 * * *	0.0007	6.0579	0.0000
	GSGM	0.0273 * * *	0.0035	7.7971	0.0000
	GSGB	0.0330 * * *	0.0045	7.2892	0.0000
	DEVE	0.0000 * * *	0.0000	0.1527	0.8787
	Observations	26410	26410	26410	26410
	The Pseudo R – squared	0.0523	0.0523	0.0523	0.0523
Step2	The X1	1.0136 * * *	0.1311	7.7324	0.0000
	INDU	0.0277 * * *	0.0027	10.4482	0.0000
	GSGM	0.2913 * * *	0.0122	23.8681	0.0000
	GSGB	0.3110 * * *	0.0161	19.2953	0.0000
	DEVE	0.0028 * * *	0.0002	12.6334	0.0000
	DEBT	0.0997 * * *	0.0107	9.3144	0.0000
	Observations	26410	26410	26410	26410
	The Pseudo R – squared	0.0465	0.0465	0.0465	0.0465

*, ** and *** respectively indicate that the regression coefficient is significant at the level of 10%, 5% and 1%.

Table 7 shows the regression results of the two-stage instrumental variable method on the compensation levels of chairman and general manager. The X in the first stage X2 OLS regression results, we can find that instrumental variable X2 is with the original explained variable X to 1% significance level is positive, explained the selection of instrumental variables to meet the requirements of the instrumental variable and endogenous variables are highly correlated, to extract the exogenous part of the original repressors, instrumental variable selection in this paper is more reasonable. The regression results of X2 on Y in the second stage show that the regression coefficient of the average level represented by the instrumental variable is significantly positive at the 1% level, and the model regression coefficient can be obtained through the conversion of the probability density function output by the software. Compared with the separation of the two jobs of chairman and general manager, the probability of improving the business performance of NEEQ enterprises increases by 1.73%.

Table 7. Regression results of the two-stage instrumental variable method

	Variable	Coefficient	Std. Error	T – Statistic	Prob.
Step1	C	1.3973 * * *	0.0491	28.4421	0.0000
	X2	0.2239 * * *	0.0051	43.7091	0.0000
	INDU	0.0041 * * *	0.0007	5.8270	0.0000
	GSGM	0.0223 * * *	0.0034	6.5242	0.0000
	GSGB	0.0377 * * *	0.0044	8.5380	0.0000
	DEVE	0.0000	0.0000	0.1737	0.8621
	Observations	26410	26410	26410	26410
	The Pseudo R – squared	0.0955	0.0955	0.0955	0.0955
Step2	X2	0.0471 * *	0.0188	2.5070	0.0122
	INDU	0.0289 * * *	0.0026	10.9170	0.0000
	GSGM	0.2927 * * *	0.0122	23.9726	0.0000
	GSGB	0.3161 * * *	0.0161	19.6383	0.0000
	DEVE	0.0028 * * *	0.0002	12.3926	0.0000
	DEBT	0.1029 * * *	0.0107	9.6108	0.0000
	Observations	26410	26410	26410	26410
	The Pseudo R – squared	0.0442	0.0442	0.0442	0.0442

*, ** and *** respectively indicate that the regression coefficient is significant at the level of 10%, 5% and 1%.

4.4.2. Re-Experiment Based on OLS Regression Model

Given that the Ordered Probit model is classified regression, it is an explained variable comprehensively generated based on the two dimensions of profitability and development capacity of NEEQ enterprises. To test the adaptability of the research results to different models, we re-conducted the experiment based on OLS regression. In this paper, the OLS model is used to conduct regression for explanatory variables. Table 8 shows that: the leadership characteristics (X) of the board of directors are significantly positive at the significance level of 1%, and the conclusion of this paper is still valid.

Table 8. OLS regression models

Variable	Coefficient	Std. Error	T – Statistic	Prob.
C	2.7075 * * *	0.0741	36.5299	0.0000
X	0.0265 * * *	0.0088	2.9951	0.0027
INDU	0.0085 * * *	0.0011	8.0417	0.0000
GSGM	0.1339 * * *	0.0051	26.2893	0.0000
GSGB	0.1430 * * *	0.0066	21.7242	0.0000
DEVE	3.36 e-05 * * *	9.69 e-06	3.4644	0.0005
DEBT	0.0584 * * *	0.0042	14.0579	0.0000
Observations	26410	26410	26410	26410
The Pseudo R – squared	0.0369	0.0369	0.0369	0.0369

*, ** and *** respectively indicate that the regression coefficient is significant at the level of 10%, 5% and 1%.

4.4.3. Re-Measurement of Explained Variables

The experiment, the new three board enterprise profitability (total return on assets) and growth (main business revenue growth) in accordance with the average threshold, is divided into four regions, to measure the operating performance of enterprises be explained variable (Y), considering the robustness of variable metric aspect factor, to reclassify the research variables, we empirically test again. We classified the ROA quartile of NEEQ enterprises into four regions, and the ROA quartile of enterprises less than 25% was denoted as 1, 25%-50% as 2, 50%-75% as 3, and 75%-1 as 4.We carried out Ordered Probit regression again. According to the regression results (as shown in Table 9), it can be seen that explanatory variables and explained variables present a significant positive correlation at the significance level of 1%, and the model regression coefficient can be obtained by converting the probability density function output by the software. Compared with the separation of the two roles of chairman and general manager, the probability of improving the business performance of NEEQ enterprises increases by 2.59%. The experimental conclusion is consistent with the foregoing, so the conclusion of this paper is still valid.

Table 9. Regression results of ordered probit model remeasured

Variable	Coefficient	Std. Error	Z – Statistic	Prob.
X	0.0862 * * *	0.0252	3.4147	0.0006
INDU	0.0171 * * *	0.0031	5.4935	0.0000
GSGM	0.3436 * * *	0.0140	24.6047	0.0000
GSGB	0.3740 * * *	0.0185	20.2719	0.0000
DEVE	0.0002	0.0003	0.6824	0.495
DEBT	0.1223 * * *	0.0081	15.0795	0.0000
Observations	26410	26410	26410	26410
The Pseudo R – squared	0.0479	0.0479	0.0479	0.0479

*, ** and *** respectively indicate that the regression coefficient is significant at the level of 10%, 5% and 1%.

4.5. Further Analysis

On the basis of the basic regression empirical results to verify the establishment of hypothesis H1b, to further study the impact of chairman and CEO concurrently on business performance, this paper takes the nature of equity ownership as the regulating variable and the size of enterprises as the sample reclassification standard, and further studies the hypotheses respectively.

4.5.1. The Moderating Effect of Equity Nature

In China, State-Owned enterprises are mainly regulated by national policies, in which the chairman and CEO play limited roles. In private enterprises, the chairman and CEO play a more important role as the core decision-making group (Shen et al. 2020, 2021). Therefore, this paper takes the ownership nature of an enterprise as a regulating variable, and studies whether, under the regulatory effect of the ownership nature of an enterprise, it will promote the influence of the concurrent chairman and CEO on the business performance of the enterprise.

Table 10 reports the regression results of the Ordered Probit model with the addition of the Ordered Probit variable. According to the regression results, the product term coefficient of the moderator variable and the independent variable is positive, but not significant, indicating that the nature of an enterprise’s equity has no regulating effect on the relationship between the chairman and the CEO and the enterprise’s business performance. In other words, whether the enterprise is state-owned or private, the chairman and the CEO are conducive to improving the enterprise’s business performance.

Table 10. Moderating effect test of ownership nature

Variable	Coefficient	Std. Error	Z – Statistic	Prob.
X	0.0363	0.0315	1.1518	0.2494
INDU	0.0360 * * *	0.0031	11.5526	0.0000
GSGM	0.3015 * * *	0.0147	20.5129	0.0000
GSGB	0.2856 * * *	0.0194	14.7458	0.0000
DEVE	0.0025 * * *	0.0002	11.1359	0.0000
DEBT	0.0888 * * *	0.0113	7.8677	0.0000
EN	0.2507 * * *	0.0260	9.6257	0.0000
X * EN	0.0365	0.0626	0.5829	0.5600
Observations	24657	24657	24657	24657
The Pseudo R – squared	0.0450	0.0450	0.0450	0.0450

*, ** and *** respectively indicate that the regression coefficient is significant at the level of 10%, 5% and 1%.

4.5.2. Classification Analysis of Enterprise Size

This paper takes into account that in enterprises with different enterprise sizes, the difficulty of enterprise management varies greatly, the heterogeneity of management team is obvious, and the larger the enterprise size is, the more difficult it is to manage. In this case, the chairman and CEO may not be able to effectively manage the enterprise and give full play to the advantages of concurrent positions. Therefore, according to the number of employees, the sample enterprises are divided into large enterprises and small enterprises for regression respectively, as shown in Table 11.

Table 11. Grouping regression results

variable	A large company		A small company
Variable	Coefficient	Prob.	Coefficient	Prob.
X	0.0700 *	0.0614	0.1039 * * *	0.0002
INDU	0.0558 * * *	0.0000	0.0125 * * *	0.0007
GSGM	0.4162 * * *	0.0000	0.3480 * * *	0.0000
GSGB	0.3852 * * *	0.0000	0.4041 * * *	0.0000
DEVE	0.0046 * * *	0.0000	0.0023 * * *	0.0000
DEBT	1.8126 * * *	0.0000	0.0755 * * *	0.0000
N	1645	1645	1923	1923
The Pseudo R – squared	0.0783	0.0783	0.0499	0.0499

*, ** and *** respectively indicate that the regression coefficient is significant at the level of 10%, 5% and 1%.

Table 11 shows that for larger enterprises, the concurrent role of chairman of the board as CEO can not only fail to improve the company’s operating performance but also significantly reduce the enterprise’s operating performance. For large companies, both the size of the enterprise and the difficulty of decision-making require a strong supervision function of the board of directors, so as to limit the agency problem of management. CEO duality, by enhancing CEO power relative to the board, compromises the board’s monitoring and disciplining functions. The chairman and CEO of the board of directors control the agenda and information flow of the board of directors, which makes the independent directors and other directors fear the powerful CEO and weakens the function of the board of directors in supervising and managing the company’s management. And the smaller the enterprise, the chairman and CEO will significantly improve the company’s business performance. The companies in NEEQ are small and medium-sized enterprises that face a more complex financing environment and a tougher market environment than larger companies, which requires quick and strong decisions.I. n the case of a chairman as CEO, a unified command helps establish clear authority and responsibility within the company, and thus has both substantive and symbolic value. It gives the President more power, so he can make quick and decisive decisions.

5. Conclusions and Implications

In this paper, we use 3568 listed companies from 2008 to 2018 as samples to explore the influence of chairman-general manager duality on business performance in the practice of SMEs in China. Based on the test of the Ordered Probit model, it is found that compared with the separation of two jobs, the integration of the two jobs of chairman and general manager increases the probability of business performance improvement of NEEQ enterprises by 37.06%, that is, the simultaneous holding of chairman and general manager is conducive to improving business performance. Further study finds that the ownership nature of the enterprise has no regulating effect on the relationship between the chairman and the general manager and business performance. Through the sample study of enterprise size division, it is found that in small enterprises, the chairman and general manager have a more significant positive impact on business performance than in large enterprises. After the robustness test of the OLS model and explained variables, it is found that the conclusion of this paper is still valid. The innovation of this paper is as follows: First, most of the existing literature takes the whole board of directors or independent directors as the main research object to explore the relationship between their characteristics and corporate performance. This paper takes the chairman as the starting point to study whether the chairman-general manager duality can affect the performance. Second, most of the existing studies focus on companies in foreign markets and domestic main board markets. This paper focuses on the NEEQ market with SMEs, and studies the relationship between board leadership characteristics and corporate performance. Third, the existing literature on a single index as a business performance measurement, this paper reference the Boston matrix, can be from two aspects of profitability and growth of company business performance comprehensive measure, while studying common profitability and development ability, will be explained variables is divided into four interval, avoided due to the inaccurate of single variables affect the empirical result, makes the measurement more comprehensive science.

The research results of this paper can provide empirical evidence for corporate board structure optimization and business performance improvement. The following research direction will take the listed companies on the main board as samples, and further explore the deep relationship between the leadership characteristics of the board of directors and corporate performance through the comparative study between large companies and small and medium-sized enterprises represented by the NEEQ market.