Chair–CEO age dissimilarity and firm value – evidence from China Open Access

This study investigates whether age dissimilarity between the Chairman (Chair) and Chief Executive Officer (CEO) affects firm value. Further, we examine whether board monitoring intensity varies depending on the age dissimilarity between the Chair and CEO. Finally, we investigate whether the board monitoring intensity moderates the association between Chair and CEO age dissimilarity and firm value. The Upper Echelon Theory (UET) proposes that the personal characteristics of top managers, including their age, can significantly impact organisational outcomes. Empirical research based on UET evidence that top management team (TMT) attributes affect financial reporting quality (Habib and Hossain, 2013), strategic decision-making (Amason and Sapienza, 1997) and firm performance (Wang et al., 2016). The age of the Chair and CEO indicates their wealth of experience and cognitive abilities, which may directly or indirectly impact individual work attitudes, behaviours, and values (Li et al., 2008). Therefore, it is increasingly important to understand the impact of age dissimilarity on firms (Rabl and Triana, 2014). Given the critical role of the Chair and CEO individuals in shaping the firm’s strategy, their age dissimilarity may be a crucial element to consider.

The impact of the Chair-CEO age dissimilarity on firm value is a topic that has received limited attention in existing academic research. A harmonious relationship between the Chair and CEO increases the alignment of interests, promoting the maximisation of firm value and personal benefits (Li and Roberts, 2017). Goergen et al. (2015) assert that the Chair-CEO age dissimilarity can lead to cognitive conflict, resulting in more effective board oversight and higher firm value in Germany. At the same time, Zhou et al. (2019) reiterate that the Chair-CEO age dissimilarity can reduce excessive risk-taking between firms. Moreover, the Chair-CEO age dissimilarity could lead to conflict in the decision-making process, such as task conflict and relationship conflict (Amason, 1996). Task conflict positively impacts team performance, while relationship conflict negatively impacts team performance (Jehn, 1997; Simons and Peterson, 2000). Both types of conflict are destructive to firm performance (De Dreu and Weingart, 2003). Task conflict has a far less positive impact on enterprises than relationship conflict’s negative impact (Korsgaard et al., 2008), even if conflict management methods are used (Choi and Cho, 2011).

However, previous studies on the Chair-CEO age dissimilarity (Goergen et al., 2015; Zhou et al., 2019) are conducted based on western developed economies. The cultural background of Chinese listed firms significantly differs from that of firms in developed economies, and such differences in cultural backgrounds may lead to conflicts in cognitive styles among individuals (Chow, 2023). For individual culture, the Confucian values emphasise respect for elders and authority figures, which may lead to a more hierarchical power structure within Chinese listed firms, called “paternalistic leadership” (Farh et al., 2014). It combines strong discipline and authority with benevolence and moral integrity, which are couched in a “personalistic” atmosphere and are common in Chinese listed firms (Farh and Cheng, 2000). This may make the relationship between the Chair and CEO age dissimilarity and firm value in Chinese listed firms more complex than in Western firms. On the other hand, emerging market business cultures are relatively risk-averse because these markets used to be closed economies with less competition between firms. Therefore, firms rarely engage in risk-taking investments (Gilpin, 2000). Thus, the applicability of conclusions based on Western advanced economies has yet to be explored to a large extent in emerging markets. Additionally, the way conflicts are addressed may vary across different cultural backgrounds. Westerners tend to confront conflicts directly, while Asians tend to use avoidance to handle conflicts (Tse et al., 1994; Weldon and Jehn, 2018), especially in Chinese firms (Chen et al., 2005). Many cross-cultural management studies show these differences because many Asians are collectivists (Triandis, 1990), and their values may lead to an emphasis on harmony and conflict resolution to maintain relationships (Morris et al., 1998).

Further, we analyse how firm monitoring influences the relationship between Chair-CEO age dissimilarities and firm value. We argue that board meetings are one of the critical strategies to monitor top management. When there is high trust among the TMT, task conflict is unlikely to lead to relationship conflict between the TMT (Simons and Peterson, 2000). Based on the “similarity attraction” phenomenon, Chairs and CEOs with low age dissimilarity are more likely to share information frequently and have stronger trust, leading to more frequent board meetings. Due to cultural differences, Chinese managers are more inclined to avoid conflict than their Western counterparts (Fu et al., 2008; Morris et al., 1998). Therefore, the Chair and CEO will avoid conflict escalation by reducing social connections, which will decrease the frequency of board meetings and firm performance.

Amidst competing arguments, this study asserts that the Chair-CEO age dissimilarity can impact firm value. The Chair-CEO age dissimilarity may result in power struggles and communication difficulties that impede decision-making and strategy execution (Camelo-Ordaz et al., 2014; De Dreu and Beersma, 2005). This could stem from the divergence in perspectives and priorities that arise from their age-related experiences, which is heterogeneity. Heterogeneity can give rise to viewpoint conflicts and affect firm value. In addition, “similarity attraction” suggests that humans tend to interact and communicate with those who share similar attributes (McPherson et al., 2001); such interactions would yield more positive feedback and limit the emergence of divergent and conflicting opinions. In corporate governance, the Chair-CEO age dissimilarity can decrease effective communication and increase differences in opinions, thereby hindering decision-making efficiency and creating a lack of coordination and cooperation that delays firm operations (Lee et al., 2014).

Using a sample of 37,200 firm-year observations in the Chinese Stock Exchange for 2001–2021, this study finds a negative association between Chair-CEO age dissimilarity and firm value. Chair and CEO age dissimilarity could increase relationship conflict and reduce the frequency of board meetings and the efficiency of business operations (Lee et al., 2014). Regarding economic significance, a 1% increase in the Chair-CEO age dissimilarity reduces the firm’s value by 0.5%. Also, we find a negative association between the Chair-CEO age dissimilarity and board monitoring intensity. Results show that board monitoring intensity negatively moderates the effect of Chair and CEO age dissimilarity on firm value. Afterwards, based on a cross-sectional analysis, firms with higher free cash flow face more severe agency conflicts (Gugler and Yurtoglu, 2003), which exacerbates the negative impact of Chair and CEO age dissimilarity on firm value. In a low industry concentration business environment, firms have higher market competitiveness but lower profitability (Pahi and Yadav, 2022), which worsens the negative impact of Chair and CEO age dissimilarity on firm value. We find that Chair and CEO age dissimilarity has a more significant negative effect on firm value in firms with dispersed ownership. The negative impact of Chair and CEO age dissimilarity on firm value in SOEs is insignificant compared to non-SOEs. The results still hold after controlling the impact of management committee meetings. Regarding the endogeneity concern, our results regarding Heckman’s two-stage least squares and propensity score matching are robust.

The contribution of this paper is multi-fold. Firstly, the paper enriches the emerging research on the influence of Chair and CEO age dissimilarity on firm value and board monitoring intensity in the Chinese market. In addition, this study contributes to the very scarce literature on the effects of cooperation and conflict between the Chair and CEO. Thus far, extant research has primarily relied on the UET analysis, whereas this study incorporates “similarity attraction” theory, heterogeneity, age effect, cohort effect, social identity theory, social psychology theories, and socioemotional selectivity theory. Furthermore, this study contributes to future SOE-related research by including a comparison analysis between State-Owned Enterprises (SOE) and non-SOE in the cross-sectional analysis. An important implication of this study is that firms should consider aligning age diversity within top management to mitigate conflicts and enhance communication, thereby improving decision-making and firm value. For regulators and professional organizations, the study implies that governance frameworks and guidelines should account for the potential impact of leadership age diversity on firm performance. Regulators may recommend or encourage age alignment within top management to foster better communication and decision-making. Additionally, professional bodies could guide best practices for managing leadership dynamics, particularly in firms with significant age dissimilarity, to ensure more effective governance and oversight. These recommendations could help enhance corporate governance standards and improve overall firm performance.

This introduction is followed by a literature review and the development of hypotheses in Section 2. Section 3 discusses the methodology and sample selection process. Section 4 discusses empirical results and robustness tests. Finally, section 5 summarises and concludes this study.

The evolving TMT workforce, with an increasing presence of young members, has raised attention to age-related dynamics. Demographic factors like age, tenure, experience, education, nationality, and gender significantly influence behaviour and decision-making (Hambrick and Mason, 1984; Keck, 1997; Patzelt et al., 2008; Heenipellage et al., 2022; Kamran et al., 2023; Nielsen and Nielsen, 2013; Tonoyan and Olson-Buchanan, 2023). Age is a crucial basis for group classification, impacting team dynamics. Age dissimilarity in TMTs can manifest as Age Effects and Cohort Effects, influencing decision-making and strategic direction (Taylor, 1975). Older chairpersons, with more management experience, make accurate judgements and perform supervisory duties effectively (Hambrick et al., 2015). Age dissimilarity may impact the power dynamic between the CEO and Chair, leading to interpersonal conflicts, especially if the Chair is older than the CEO (Rahim, 2002).

Effective internal coordination and cooperation are critical for team performance, impacting firm outcomes. In the strategic decision-making process, conflicts within senior management teams are inevitable due to complexity and uncertainty (Simons and Peterson, 2000). Internal conflict is characterised by cognitive versus affective conflicts and task versus relationship conflicts (Amason, 1996; Jehn, 1997; Parayitam and Dooley, 2009). Task conflict, fostering diverse opinions, is theoretically positive for team performance (Mooney et al., 2007). However, empirical evidence varies, with some studies showing negative or insignificant correlations (De Dreu and Weingart, 2003; De Wit et al., 2012). Relationship conflict, fuelled by dissimilarity, harms team performance (Parayitam and Dooley, 2009). Age dissimilarity, particularly between Chairs and CEOs, may hinder personal connection and information sharing, leading to biased decision-making and ineffective performance (Gruenfeld et al., 1996). Both conflict types negatively impact firm performance, with relationship conflict potentially overshadowing the positive effects of task conflict (De Dreu and Weingart, 2003). Conflict management interventions addressing the root causes of relationship conflict may indirectly manage task conflict (Choi and Cho, 2011).

Nevertheless, both types of conflict can negatively impact firm performance (De Dreu and Weingart, 2003). Korsgaard et al. (2008) note that task conflict has a weak or non-existent impact beyond relationship conflict when examining both task conflict and relationship conflict in firms simultaneously. Moreover, the negative effects of relationship conflict on business performance may overshadow the positive effects of task conflict (De Dreu and Weingart, 2003). Therefore, conflict management interventions that focus on addressing the root causes of relationship conflict may indirectly lead to the management of task conflict (Choi and Cho, 2011).

Resource dependence theory demonstrates that dissimilarity can affect firm performance (Pfeffer and Salancik, 2015). Tanikawa et al. (2017) utilised the socioemotional selectivity theory ^[1] to investigate the impact of TMT age dissimilarity on a firm’s financial performance. Goergen et al. (2015) extended this investigation to explore the relationship between the Chair-CEO age dissimilarity and firm value. However, research on TMT age dissimilarity and firm value have produced conflicting theoretical arguments and mixed findings, with studies reporting a positive relationship (Kilduff et al., 2000), negative relationship (Ozer, 2010), and non-significant relationship (Nielsen and Nielsen, 2013).

Ning et al. (2022) have demonstrated a positive correlation between diverse boards and comparability of financial reports, which suggests that the age dissimilarity of the board can promote the achievement of corporate objectives and effectively enhance the quality of financial reporting (Hsieh et al., 2018; Thomsen and Conyon, 2012). Zhou et al. (2019) find that Chair and CEO age dissimilarity can positively affect bank loan portfolio risk. Similarly, Goergen et al. (2015) believe that Chair and CEO age dissimilarity can cause conflict, leading to improved decision-making quality and enterprise value. Some studies classify conflict resulting from age dissimilarity as task conflict, suggesting that age dissimilarity is a resource of information, knowledge, and professional experience (Williams and O'Reilly, 1998). From an information and decision-making standpoint, the cognitive conflict within the board caused by age dissimilarity could improve a firm’s moral development (Labelle et al., 2010), earnings quality (Hoang et al., 2017), and decision-making quality (Adams and Ferreira, 2007). Besides, lower Chair-CEO age dissimilarity is associated with weak corporate governance and lower firm value in US firms (Lee et al., 2014) and large French firms (Nguyen, 2012). Agency theory (Jensen, 1986) suggests that the Chair may not effectively monitor the CEO, as the trust and loyalty inherent in friendship can undermine the objectivity and willingness if the board is too friendly toward the CEO (Khedmati et al., 2020). This concept is referred to as a “friendly board”, which argues that a “friendly board” increases agency conflicts and information asymmetry, ultimately reducing firm value (Aebi et al., 2012). However, as noted by Milliken and Martins (1996), team heterogeneity with some aspects being beneficial and others detrimental. Some studies find a negative association between age diversity, corporate social responsibility disclosure (Khan et al., 2019), and earning management (Komal et al., 2023). Scholars also confirm the heterogeneity of executives’ views on risk, caution, and wealth, which makes knowledge sharing more difficult (Stahl et al., 2010), and is more likely to cause internal conflicts within the group during the decision-making process, resulting in a lack of attraction or even rejection which hinders the effectiveness of board meetings and weakens a firm’s profitability (Abt and Knyphausen-Aufseß, 2017).

Based on the “similarity-attraction” phenomena, McPherson et al. (2001) state the negative association between TMT members’ demographic diversity and firm performance. Individuals like to associate with similar people since it assures greater input while reducing the possibility of discrepancies and various perspectives. As a result, the Chair and CEO’s resemblance can help them to receive more favourable feedback and reduce disputes during communication (Lee et al., 2014). Directors of similar ages may have similar experiences, attitudes, opinions, and beliefs, which improves the operational efficiency and effectiveness of the organisation (age effect) (Wagner et al., 1984). In contrast, dissimilarity ages could heighten cognitive conflict and lead to relational conflict. According to social identity theory, group identification (cohort effect) is key to influencing commitment and motivation (Tajfel and Turner, 1979), while group differences could lead to social fragmentation (Guillaume et al., 2012). Diversity can impede collaboration among senior management (Schmid and Mitterreiter, 2020). Homogeneity may increase communication, foster successful working relationships, and promote team cohesiveness, which is favourably connected with team performance. TMTs’ internal consistency improves in avoiding internal setbacks and enables them to make good strategic decisions rapidly (Hambrick and Mason, 1984). Thus, TMTs with high age similarity among members will have higher collaboration efficiency and frequency, resulting in timely business decisions and increased firm value. Further, in executing corporate strategy, TMTs’ commitment to decision consistency is crucial for corporate performance (Pitcher and Smith, 2001).

Social categorisation (Knapp et al., 2011) and the similarity-attraction paradigm (Byrne, 1997) suggest that the Chair-CEO age dissimilarity negatively affects firm value. The categorisation of subgroups based on similarities and differences in work groups results in forming in-groups and out-groups, and people are more willing to cooperate with the former. Cooperation frequency and efficiency are stronger among TMT members with high age similarity than those with low similarity based on age effects, cohort effects, and UET. Furthermore, the greater the Chair-CEO age dissimilarity, the slower the decision-making process, leading to untimely business decisions and reduced firm value (Van Knippenberg and Schippers, 2007). Therefore, the Chair-CEO age dissimilarity may result in reduced communication, lower frequency of information sharing, and more communication errors, leading to poor business decision-making and a decrease in firm value (Williams and O'Reilly, 1998). Based on these studies, we developed the following hypothesis:

Board monitoring intensity is crucial to corporate governance, as it ensures that the board of directors can effectively oversee top management’s actions and protect shareholders’ interests. Age dissimilarity is one factor that has been shown to impact board functioning and monitoring intensity, particularly in the context of Chair and CEO relationships. Age dissimilarity can bring various perspectives and experiences to the boardroom, improving decision-making and reducing groupthink. On the other hand, age dissimilarity can also lead to communication difficulties and conflicts due to generational differences and varying experience levels. Based on the “similarity-attraction” phenomena, interactions and frequency of interpersonal relationships are higher among people with similarities (McPherson et al., 2001). Age dissimilarity between the Chair and CEO could negatively affect board monitoring intensity, leading to a power struggle and undermining the board’s effectiveness, specifically when the Chair and CEO have a significant age dissimilarity and career horizon. Also, the Chair may be less (or more) willing or able to challenge the CEO and hold the other accountable for their actions. This can lead to a lack of oversight and a lower monitoring intensity. Therefore, the following hypothesis is proposed based on the above discussion:

Board monitoring intensity significantly impacts firm operations and performance (Jaffar and Abdul-Shukor, 2016). Empirical Research indicates that increased monitoring is associated with improved performance, helping identify and rectify issues promptly and deterring misconduct (Vafeas, 1999). However, excessive monitoring may hinder innovation and risk-taking, which are crucial for long-term growth (Adams and Mehran, 2005). A highly effective board can navigate challenges of age diversity, leveraging its benefits like diverse opinions and in-depth discussions. Effective boards ensure age diversity contributes to decision-making and strategy creation, enhancing firm performance and value. In contrast, less effective boards overseeing senior management may amplify age diversity drawbacks, leading to conflicts and poor decision-making, negatively affecting firm value. Enhancing board monitoring efficiency could impact the relationship between age diversity and firm value. Increased board meeting frequency may improve communication and collaboration, particularly for chairs and CEOs with significant age dissimilarity, leading to better decision-making and improved firm performance (Ntim and Osei, 2011). Yet, it could also escalate conflicts, especially if age dissimilarity is large, potentially harming firm performance and value (Adams and Ferreira, 2007). Therefore, we develop the following hypothesis:

We employ two widely used firm value measures, namely Tobin’s Q (TOBINQ) and Market-to-Book ratio (MKTBK) (Hilliard and Zhang, 2015; Sucuahi and Cambarihan, 2016). TOBINQ is defined as the market capitalisation plus the difference between the book value of total assets and the book value of equity, all divided by the book value of total assets. If Tobin’s Q is greater than 1, it suggests that the firm is perceived to have higher growth potential and firm value because the asset’s market value is more than its replacement cost and vice versa. In comparison, the market-to-book ratio is another widely used measure of firm value. A higher MKTBK suggests greater perceived firm value and growth potential, as the market value of the firm’s equity is more than the book value suggests. This study measures MKTBK, the share price divided by the net book value per share.

Empirical research does not show any consensus on the measure of age dissimilarity. Therefore, we consider several measures for Chair and CEO age dissimilarity consistent with the existing literature (Goergen et al., 2015; Zhou et al., 2019; Zhu et al., 2021). This study uses two variables to measure the dissimilarity between the age of the Chair and the CEO. GAPS is the age dissimilarity between the Chair and CEO, calculated as the Chair’s age minus the CEO’s age. The purpose is to test whether the Chair is older or younger than the CEO. GAPU is the absolute value of GAPS. This variable captures the age difference between the Chair and CEO, regardless of who is older. In addition, GAPSQU is the squared GAPS used to test whether the regression model has a non-linear relationship (Younsi and Bechtini, 2020).

We use two proxies to measure board monitoring intensity. Studies have shown that the number of board meetings indicates board monitoring intensity (Brick and Chidambaran, 2010). Therefore, we follow prior studies (Goergen et al., 2015) and use board meetings (BMEET) to measure board monitoring intensity, defined as the sum of board meetings and the number of conferences held by the board during the fiscal year (Adams and Mehran, 2005). Following Brick and Chidambaran (2010), we use another proxy (LNBMEET), the natural logarithm of the number of board meetings per year, to measure board monitoring intensity.

Following Brick and Chidambaran (2010), we control for several Chair characteristics, CEO characteristics, supervisory board characteristics, and firm characteristics that may affect FV and BMI.

Chair and CEO characteristics include the Chair and CEO’s different gender (DIFGEN), different educational backgrounds (DIFEDU), cultural differences (DIFNAT), and length of time working together (JONTENU). Following Andres et al. (2014), we measure their corresponding power through the replacement (CHANGE1/CHANGE2) and the tenure (TENURE1/TENURE2) of the Chair and CEO (Kaplan et al., 2012). All of these are used to measure the possibility of tacit understanding or conflict between the Chair and CEO. Supervisory board characteristics include Board size (BODSIZE), which refers to the number of directors on the board. Board age (BODAGE) refers to the average age of board members. CV Board Age (CVBA) controls for age diversity on the board committee and supervisory board. These variables have also impacted firm value (Brick and Chidambaran, 2010; Yermack, 1996).

Firm characteristics include firm size (FSIZE), leverage (LEVER), beta coefficient (BETA), stock volatility (STOCKVOL) ^[2], sales growth (SALEGR), R&D expenses (R&D), the ratio of capital expenditure (CapEx/TA), firm listing age (FIRMAGE), and the family firm (FFIRM) (Audretsch et al., 2013; Bouslah et al., 2018; Burney et al., 2021; Custódio and Metzger, 2014; Dittmann et al., 2010; Le et al., 2006; Peng and Jiang, 2010; Popov and Ongena, 2011). Following Goergen et al. (2015), this study also includes a set of management committee characteristics that may affect the dependent variable as an additional control for models. As an essential part of the TMT, the board of supervisors plays a supervisory role on the board of directors and the CEO. Management committee characteristics include the management board committee size (MBSIZE), the average tenure of management board committee members (MBTENU), the age of the management board committee members (CVMBA), and the supervisory board size (SBSIZE) (Vafeas, 1999). We refer to  Appendix for further details on the variable definitions and the sources.

In H1, we examine the association of Chair and CEO age dissimilarity and firm value. The following Ordinary Least Square (OLS) regression tests the H1.

To test H1, the primary variable of interest is $β1GAPi,t$⁠. GAP_i,t is the proxy for the age dissimilarity between the Chair and CEO. As shown in Section 3.2, we consider two different measures of age dissimilarity: GAPS and GAPU. Firm value is the dependent variable in H1. This study uses two primary measures of firm value (FV), such as Tobin’s Q (TOBINQ) and Market-to-Book ratio (MKTBK). In equation (1), a positive (negative) coefficient of $β1GAPi,t$ will indicate that age dissimilarity increases (decreases) firm value.

In H2, we posit that the age dissimilarity between the Chair and CEO will likely affect the firm’s monitoring needs. Therefore, to test H2, we develop the following OLS regression model is developed:

In H2, the primary variable of interest is $γ1GAPi,t$⁠. A positive association on $γ1GAPi,t$ will indicate that age dissimilarity increases board monitoring intensity and vice versa. As shown in Section 3.3, we use BMEET and LNBMEET to measure the intensity of the board monitoring. We conduct both OLS and Poisson regression equations to test H2. Consistent with Goergen et al. (2015), the Poisson regression equation is an appropriate regression model for a variable measured on count data, such as BMEET (Hayat and Higgins, 2014).

In H3, we hypothesise that board monitoring intensity moderates the association between Chair and CEO age dissimilarity and firm value. We develop the following OLS regression model to examine the H3:

The primary variable of interest is $α3GAPi,t*BMIi,t$⁠. When α₁ is a positive coefficient, a positive association on α₃ would suggest that an increase in the board monitoring intensity could promote the positive impact of the Chair and CEO’s age dissimilarity on firm value (positive moderator). In contrast, a negative association on α₃ could indicate that an increase in the board monitoring intensity would negatively moderate the positive effect of the Chair and CEO’s age dissimilarity on firm value (negative moderator). Conversely, when α₁ is a positive coefficient, a positive association of α₃ could suggest that an increase in the intensity of board monitoring would reduce the negative impact of the Chair and CEO age dissimilarity on firm value (negative moderator; weakening the main effect), while a negative association on α₃ would suggest that an increase in the intensity of board monitoring would enhance the negative impact of the Chair and CEO age dissimilarity on firm value (negative moderator). In addition, this positive association on α₂BMI_i,t would suggest that firm value could increase with higher board monitoring intensity (Wang et al., 2018).

This study extracts the relevant financial and firm governance data of 5,049 Chinese listed firms from 1998 to 2021 from the China Stock Market & Accounting Research Database (CSMAR) (Du and Boateng, 2015), we drop the missing firm-year observations. Table 1 illustrates the sample selection process. After calculating the data for some firm-specific financial variables, we exclude 3,134 observations from 1998 to 2000, which are not part of the study period. In addition, we exclude missing values and outliers in the remaining sample data. Following Fang et al. (2021), we excluded financial institutes (SIC codes between J66 to J69), because of their unique regulatory environments. The final sample of this study is 37,200 firm-year observations from 3,870 unique firms. All the continuous variables are winsorised, top and bottom, at a 1% level to rule out any potential biases from outliers. Table 2 reports a distribution of samples based on industry classification. We follow the SIC code of CSMAR to classify the firms consistent with the prior study (Xie et al., 2021). The sample is well distributed. Sample observations of this study include a total of 17 industries. As an economic pillar industry in the Chinese market, industrial products account for the majority of the sample in this study.

Table 3 reports the descriptive statistics of all the variables used in this study. The mean (median) of TOBINQ, MKTBK, NAVPS, and MKTVAL are 2.45(1.87), 3.12(3.11), 4.40(3.75) and 22.65(22.50). Sample firms hold an average of 11.84 board meetings (BMEET) annually. Most CEOs are younger than Chairs. The GAPS (GAPU) mean is 3.59 (5.94) years, ranging from 0 to 28. A total of 6% of the sample firm-year has a 20-year age gap between Chair and CEO (GAP20), and the mean of GAP15 (GAP10) is 12% (24%) of the sample firm. The mean age of the Chair and CEO is 52.15 (48.56) years. Regarding control variables, 84% of the Chairs and CEOs have different educational backgrounds. A total of 8% of the sample firm-year observations have gender diversity. Finally, 43% of the sample firms are SOE firms, and 20% have a majority shareholder with over 50% ownership.

Table 4 reports the results of the difference in mean for all the variables for the lower age dissimilarity (GAPU ≤ median) and higher age dissimilarity (GAPU > median) categories. Firm value is lower, and the board meets less frequently when the Chair and CEO’s age dissimilarity is higher. The findings are consistent across all the measures of firm value. Also, differences in Chair and CEO education (DIFEDU), nationality (DIFNAT), and gender diversity (DIFGEN) are greater within the group of firms with higher Chair and CEO age dissimilarity (GAPU > median). The reported variables are statistically significant at the 10% level. This result tentatively supports our hypothesis.

Table 5 presents the Pearson correlations between the dependent and independent variables included in the baseline regression. The test variables GAPU and TOBINQ, MKTBK, NAVPS, and MKTVAL are all significantly and negatively correlated (correlation coefficients −0.04, −0.03, −0.04, −0.03, p < 0.01, respectively). In addition, the test variables GAPU and BMEET, and LNBMEET are also significantly and negatively correlated (correlation coefficients of −0.04, −0.02, p < 0.01). This provides univariate support for the negative correlation between the Chair and CEO age dissimilarity and firm value and monitoring intensity, respectively. Similarly, GAPS is also significantly negatively correlated with all dependent variables. Firms with a longer co-work time between Chair and CEO (JONTENU), firms with high R&D, and firms with high capital expenditures (CapEx/TA) have better performance and weaker board monitoring intensity (Full table in  Appendix). Firms with high sales growth rates (SALEGR) correlate positively with firm value and monitoring intensity.

This section uses the empirical model presented in section 3.1 to test the hypotheses, using sample data for OLS linear regression and Poisson regression analysis.

Table 6 reports the results of the OLS regression for H1. In the H1, we examine the association between Chair and CEO age dissimilarity and firm value. Column (1) only includes the control variables. In columns (2) and (3), we examine the impact of age dissimilarity (GAPS, GAPU) on the firm value (TOBINQ). Columns (5) and (6) report the impact of age dissimilarity (GAPS, GAPU) on the alternative measure of firm value (MKTBK). There is a sensitivity test in columns (4) and (7), both GAPU and GAPSQU are taken in the regression (Haans et al., 2016). We examine whether the association between Chair and CEO age dissimilarity and firm value has a non-linear relationship in columns (4) and (7). Finally, columns (8) and (9) illustrate whether the results are still available in a complex working environment.

Consistent with the expectation, the association between GAPS and GAPU on TOBINQ shows a negative coefficient (−0.003, −0.005; t-value = −3.854, −4.202; p < 0.01), which indicates that firm value reduces due to the Chair and CEO age dissimilarity. The economic significance of the findings suggests that a 1% increase in the Chair and CEO age dissimilarity causes a corresponding 0.5% reduction in the firm’s value. Also, we examine the impact of Chair and CEO age dissimilarity on the MKTBK and find that the association of both GAPS and GAPU are negative (coefficient = −0.0000337 & −0.0000616, t-value = −3.517 & −4.341) and the findings are statistically significant at a 1% level. There is a non-linear effect of Chair-CEO age dissimilarity on firm value. Column (4) shows the coefficient for GAPU is significantly negative, while the coefficient for GAPSQU is not statistically significant (−0.007, t-value = −2.480, p < 0.05; 0.0001001, t-value = 0.805, p > 0.10). Column (7) also shows similar results. The original inference remains unchanged when models include the four proxies (CVMBA, MBSIZE, MBTENU, SBSIZE) that reflect management board committee characteristics as additional control variables in columns (8) and (9). In other words, the significant negative associated relationship between Chair and CEO age dissimilarity and firm value are not impacted by the involvement of a firm’s management board committee.

Regarding the control variables, we find that a firm with a Chair and CEO from different nationalities (DIFNAT) and a firm with higher average age of board members (BODAGE) positively affects firm value. Firms with higher capital investment (CapEx/TA), higher R&D investment (R&D), and higher sales growth (SALEGR) increase firm value. Also, we find that family firms (FFIRM) have higher firm value (Bebchuk et al., 2009; Custódio and Metzger, 2014; Dittmann et al., 2010). Overall, we find that a negative coefficient between Chair and CEO age dissimilarity and firm value suggests an age dissimilarity between the Chair and CEO has a negative consequence on firm value. Thus supports H1.

Table 7 reports the results of H2. Column (1) shows the control variables on the board monitoring intensity. In columns (2), (3), (4), and (5), we examine the impact of age dissimilarity (GAPS, GAPU) on the board monitoring intensity (BMEET and LNBMEET). Finally, columns (6) and (7) examine the influence of Chair and CEO age dissimilarity on board monitoring intensity, including several additional attributes of management board committee members (CVMBA, MBSIZE, MBTENU, SBSIZE).

In contrast with prior research (Goergen et al., 2015), the results indicate that the association between GAPS and GAPU on BMEET presents a significantly negative association (coefficient = −0.001, −0.001; t-value = −4.301, −4.412; p < 0.01). Similarly, we examine the impact of Chair and CEO age dissimilarity on the LNBMEET and find the association of both GAPS and GAPU a negative coefficient (−0.001, −0.001; t-value = −5.406 & −4.837) and the findings are statistically significant at a 1% level. Specifically, columns (1) and (2) show that for every 1% increase in the age dissimilarity between the Chair and CEO, the strength of internal monitoring decreases at 0.1% accordingly. After adding additional characteristics of management board committee members in columns (6) and (7), Chair and CEO age dissimilarity still has a significant negative impact on board monitoring intensity (BMEET and LNBMEET). The significant negative relationship associated with the Chair and CEO’s age dissimilarity and the board monitoring intensity is not affected regardless of whether the firm’s management board committee members are involved. Thus, H2 is supported.

Table 8 shows the results for H3. We test whether the effect of the association between Chair and CEO age dissimilarity and firm value is related to board monitoring intensity. Therefore, we use TOBINQ and MKTBK as proxies for FV; we infirm BMEET_median or LNBMEET_median as proxies for BMI; we use GAPS as proxies for GAP. BMEET_median (LNBMEET_median) refers to a dummy variable of value 1 assigned if the value of BMEET (LNBMEET) is higher than the median and 0 otherwise. This variable’s 1(0) values mean that BMI is relatively stronger (weaker). The primary interest variable is the interaction term between GAP and BMI (i.e. GAPS*BMEET_median, GAPS*LNBMEET_median). Columns (1) and (2) of Table 8 show the results of a moderation test using TOBINQ to measure firm value. Consistent with the results in 4.4.1, the values of coefficient α1 in columns (1) and (2) are both −0.004, which is significant at the 1% level (t-value = −3.881 & −4.228). Also, the association of both BMEET_median and LNBMEET_median has a positive coefficient (0.072, 0.079; t-value = 3.877 & 4.604), and the findings are statistically significant at a 1% level. Importantly, in these two columns, the coefficients for the variables (i.e. GAPS*BMEET_median and GAPS*LNBMEET_median) are positive in columns (1) and (2), respectively and are statistically significant at the 10% and 5% levels. Similarly, we performed substitution tests on columns (1), and (2), respectively, and replaced the variable TOBINQ, which measures the value of the firm, with MKTBK (columns (3), (4)), keeping the moderator and independent variables unchanged. Similar to columns (1), and (2), GAPS*BMEET_median and GAPS*LNBMEET_median still maintain a significant positive correlation with MKTBK. As a result, the moderating effect of board monitoring intensity as a moderating variable again proves to be significant. Therefore, Chair and CEO age dissimilarity has a greater effect on firm value in firms with lower internal monitoring intensity, and firm performance is reduced due to the generation problem and cognitive washout between Chair and CEO. Thus, the results support H3.

We use Net Asset Value Per Share (NAVPS) and Market Value (MKTVAL) as alternative measures for TOBINQ and MKTBK to check the robustness of the findings (Andrikopoulos et al., 2022; Huang and Lu, 2022). The higher the NAVPS of a firm, the more assets the shareholders own (Li and Xia, 2013). MKTVAL can directly reflect the level of firm value (Black and Kim, 2012). We re-estimated equation (1). Table 9 shows the measurement results using alternative measures of firm value. Columns (1) and (2) examine the associations of NAVPS, MKTVAL, and GAPS, respectively. Columns (3) and (4) examine the associations of NAVPS, MKTVAL, and GAPU, respectively. Column (5) shows a non-linear analysis between Chair and CEO age dissimilarity and firm value involving MKTBK, GAPU, and GAPSQU. Overall, all findings support H1.

Following prior research, we performed additional tests using the inter-generational age gap. In line with the sociology literature, Strauss and Howe (1997), Goergen et al. (2015), and Zhou et al. (2019) used Gap20 as the primary research variable to study firm performance in European business settings. Among listed firms in the Chinese market, SOEs have strict restrictions on the length of employment and promotion of the Chair and CEO. It is rare for the Chair and CEO’s age dissimilarity to exceed 20 years old. Zhu et al. (2021) used Gap10 as the primary research variable to study the Chinese market. We combined these two classification methods and broadened them. We shortened the inter-generational age dissimilarity to 5 years as the dividing line and used the four different age dissimilarity dummy variables of Chair and CEO Gap20/15/10/5 (GAP20/GAP15/GAP10/GAP5) as the alternative measures of Chair and CEO age dissimilarity. These four dummy variables are used to more comprehensively and accurately test the influence trends of different age dissimilarity between the Chair and CEO on firm value and board monitoring intensity. We re-estimated equation (1) and equation (2).

Table 10 Panel A shows the relationship between the Chair and CEO inter-generational age dissimilarity (GAP20/15/10/5) with the primary firm value proxies TOBINQ (columns (1) to (4)) and MKTBK (columns (5) to (8)). The GAP20, GAP15, GAP10, and GAP5 coefficients consistently maintained significant negative associations in all the above models (p-values level at the 1% level). As with the previous findings, all findings support H1.