Chair–CEO age dissimilarity and investment efficiency: evidence from China

This research aims to empirically examine the impact of age dissimilarity between the chairman (hereafter Chair) and chief executive officer (hereafter CEO) on firm investment efficiency (IE). A stream of research has indicated that the diversity of top management team (hereafter TMT) encompasses various dimensions, including age (Rabl and Triana, 2014), tenure (Keck, 1997), work experience (Patzelt et al., 2008), gender (Tonoyan and Olson-Buchanan, 2023), nationality (Nielsen and Nielsen, 2013) and educational background (Heenipellage et al., 2022) affects firm operational outcome, such as firm performance, information disclosure, financial reporting quality, corporate social responsibility and risk-taking propensity (Ning et al., 2022). Empirical research argues that diversity facilitates many benefits, including heightened innovation, a more robust decision-making process and enhanced organizational adaptability (Wang et al., 2021). Building on these insights, Goergen et al. (2015) suggest that an age dissimilarity between Chair and CEO increases with firm value, and Trinh et al. (2022) propose a potential enhancement of bank performance. However, not all outcomes of age dissimilarity are uniformly positive. Zhou et al. (2019) observe a decline in bank risk-taking behaviour, which, while potentially beneficial from a regulatory standpoint, may also reflect reduced strategic boldness. Moreover, Zhu et al. (2021) highlight increased pay disparity with greater age dissimilarity, raising concerns about equity and internal cohesion. Despite these mixed findings, the relationship between Chair–CEO age dissimilarity and firm IE remains underexplored, highlighting the need for further investigation and motivating our contribution to this evolving area of corporate governance research.

The relationship between the Chair and CEO is pivotal in organizational dynamics, with each holding distinct roles. The Chair is responsible for long-term strategic direction and board leadership, while the CEO manages daily operations. Despite their differing roles, both rely on frequent interaction, particularly during critical organizational decisions, necessitating frequent interaction. A positive working relationship between them enhances board effectiveness and governance. However, according to agency theory, power disparities can lead to contentious communication and interpersonal conflicts. This dynamic relationship influences strategic decisions, resource mobilization and capability building within the firm. Additionally, Carmeli and Schaubroeck (2006) suggest that this relationship’s quality can impact broader organizational dynamics. Integrating this perspective with the Chair–CEO age gap and IE, a harmonious relationship may positively influence strategic decision-making and IE. At the same time, conflicts or communication breakdowns could lead to inefficiencies and hinder effective resource allocation and capability development. Therefore, understanding and managing the dynamics between the Chair and CEO, considering their respective ages and characteristics, are crucial for optimizing organizational IE.

Investment is core to firm’s operation and shareholder value creation. Efficient investment activities can significantly impact a company’s operating conditions, profitability, value creation and long-term development. In a well-functioning capital market, effective investment typically refers to firms undertaking projects with a positive net present value (NPV) (Jensen, 2019). However, agency conflicts and information asymmetries may cause firms to be risk-averse and focus on short-term performance due to sub-optimal investment behaviour. Empirical research has linked such inefficiencies, manifesting as both over-investment (investing in negative NPV projects) and under-investment (failing to invest in positive NPV projects) to broader issues of managerial decision-making and governance (Brooks et al., 2023).

Corporate governance literature posits that the responsibility for making investment decisions primarily resides with the board of directors (Estélyi and Nisar, 2016). As a pivotal decision-making body within the organization, the Chair assumes the critical role of formulating the long-term strategic trajectory of the enterprise and providing leadership to the board of directors (Parker, 1990). Concurrently, the CEO assumes the imperative task of translating the strategic vision into tangible business operations (Roberts and Stiles, 1999). Srour et al. (2022) emphasize that these two roles are inherently interdependent and complementary in nature. Consequently, in navigating the intricacies of decision-making processes, a symbiotic relationship necessitating frequent and meaningful interaction between the Chair and the CEO becomes indispensable for organizational efficacy and sound governance. The positive and harmonious relationship between the Chair and the CEO can shape supportive organizational communication (Srour et al., 2022) and promote effective governance. The relationship between the Chair and the CEO can significantly influence organizational dynamics, fostering positive or negative ramifications throughout the organizational structure (Roberts and Stiles, 1999). Grounded in agency theory, seminal work by Fama and Jensen (1983) underscores how the inherent disparities in roles and hierarchical positions between the Chair and the CEO can engender divergent power dynamics, precipitating adversarial communication patterns and interpersonal discord. This may result in adversarial communication patterns and interpersonal discord, a phenomenon that our ongoing research on the relationship between the Chair–CEO age gap and IE seeks to clarify further.

This study develops the argument following the “similarity-attraction hypothesis” proposed by Byrne (1971) and elaborated upon by McPherson et al. (2001), which posits that individuals are naturally drawn to others who share similar attributes, leading to increased interaction and communication. Thus, age dissimilarity within the Chair–CEO relationship can impact task and relationship conflicts within the TMT. Task conflict typically arises from differences in opinions, perspectives and approaches to work-related tasks and decisions. Age dissimilarity can contribute to task conflict within the TMT if it leads to divergent viewpoints on strategic direction, operational methods or organizational priorities. Similarly, age dissimilarity can also influence relationship conflict within the TMT, which arises from personal or interpersonal disputes between team members. Age differences may lead to varying communication styles, values and attitudes, exacerbating interpersonal tensions and contributing to conflicts within the TMT. Therefore, organizations must recognize and address the potential implications of age dissimilarity within their leadership teams, fostering open communication, mutual respect and conflict resolution strategies to mitigate the adverse effects and promote optimal investment outcomes.

From the lens of task conflict, empirical research suggests that younger CEOs often prioritize innovation and technological advancement, contrasting with the inclination of older chairs towards tradition and stability. These divergent perspectives can precipitate conflicts surrounding strategic priorities and decision-making processes, impeding task performance and hindering organizational effectiveness. Similarly, considering relationship conflict, empirical evidence indicates that generational disparities in work ethic, leadership preferences and collaboration approaches may engender friction between the Chair and CEO and among other members of the TMT. From a risk perspective, we argue that younger CEOs (Chair), especially those with less experience, may prioritize short-term gains, potentially compromising efficiency. Conversely, experienced younger CEOs (Chair) tend to focus on long-term profitability. Older CEOs (Chair) may take riskier strategies driven by motives like empire-building. Such conflicting approaches can erode trust, diminish cohesion and undermine collaborative efforts within the team, ultimately compromising overall team performance and impeding organizational success. While extant research has predominantly focused on the age of the CEO, there remains a dearth of empirical evidence regarding the role of the Chair and the impact of age diversity between the CEO and Chair on firm investment decisions. This research seeks to address this gap in the literature by shedding light on the overlooked aspect of age disparity within the leadership structure and its implications for firm IE.

This study delves into listed companies in China, acknowledging the region’s distinct cultural and business environment. Empirical evidence underscores that China’s unique cultural context significantly influences organizational behaviour, firm policies and outcomes, suggesting that the role of Chair–CEO age dissimilarity may differ compared to other countries. Specifically, research indicates that China’s cultural norms shape conflict resolution strategies within TMTs, shedding light on how age dissimilarity between CEOs and Chairs may impact investment decisions. In contrast to Western norms of direct conflict confrontation, Chinese cultural norms often favour avoidance as a primary strategy for conflict resolution (Weldon and Jehn, 2018). Despite Western research suggesting that conflict avoidance may stifle innovation, many Chinese TMTs exhibit a proclivity for this approach (Chen et al., 2005). This preference for conflict avoidance in Chinese firms can be attributed to the collectivistic values prevalent in Asian cultures, including China, prioritizing harmony and interpersonal relationships (Morris et al., 1998). Consequently, Chinese managers tend to eschew conflicts more frequently than their Western counterparts (Fu et al., 2008) . Moreover, empirical evidence suggests that the decision-making efficiency of Chinese TMTs can be enhanced by circumventing potential relationship conflicts, aligning with China’s collectivistic ethos (Chen et al., 2005). Chinese TMTs strive to foster cohesion and collaboration within the team by prioritizing harmony and interpersonal relationships, thereby facilitating more effective decision-making processes. Given this context, the age dissimilarity between CEOs and Chairs within Chinese TMTs emerges as a critical factor shaping conflict resolution strategies and investment decisions. The potential discord stemming from age disparities may influence the conflict resolution approaches adopted by TMTs, subsequently impacting their cohesion, decision-making effectiveness and firm IE. Understanding the nuanced interplay between age dissimilarity and China’s corporate governance context is essential for comprehensively assessing its implications for investment decisions and organizational performance.

Using data from the China Stock Market and Accounting Research Database (CSMAR) from 2001 to 2022, this study investigates the relationship between Chair–CEO age dissimilarity and firm IE. The findings reveal a significant negative association between Chair–CEO age dissimilarity and IE, demonstrating economic implications. Specifically, for each 1% increase in the age dissimilarity between the Chair and CEO, IE diminishes by 0.5%. Furthermore, the study conducts cross-sectional analyses to delineate the contextual nuances influencing the observed associations. The negative association is particularly pronounced in firms operating within highly competitive industries, with lower ownership concentration, and those not affiliated with state-owned enterprises (SOEs). This research suggests that, in industries characterized by higher competition, the importance of swift and effective decision-making for maintaining market position accentuates any disruptions or inefficiencies caused by age differences between the Chair and CEO. Likewise, in firms with lower ownership concentration, fragmented decision-making authority may exacerbate challenges in aligning strategic visions and reaching consensus on investment decisions. Additionally, in firms not affiliated with SOEs, where decision-making structures and long-term strategic goals may be less stable, the impact of age disparity between the Chair and CEO on IE is amplified.

This research has several contributions. While previous studies primarily focus on investigating the influence of board diversity, encompassing directors and CEOs, on firm IE, this study extends this inquiry by emphasizing the pivotal role of the Chair and the dynamics between the Chair and CEO in shaping firm IE. Recognized as the board’s leader, the Chair is central to shareholder engagement and developing positive relationships among board members. This research evidences that firm IE is reduced when the age gap between CEOs and Chairs is higher. Furthermore, the study unveils intriguing insights into the moderating effects of industry competitiveness, ownership structure and organizational ownership status on this relationship. Specifically, the association between the Chair–CEO age gap and IE is more pronounced in highly competitive industries, where strategic agility and decision-making adaptability are paramount. Moreover, firms with lower ownership concentration exhibit a heightened sensitivity to the Chair–CEO age gap, implying that governance structures characterized by dispersed ownership may necessitate greater alignment between executive leadership roles. Additionally, the observed effects are accentuated in firms not SOEs, underscoring the importance of corporate governance mechanisms in privately owned entities. These findings underscore the multifaceted nature of Chair–CEO age gap dynamics and their implications for firm performance across diverse organizational contexts. Thus, the contribution of this research lies in its examination of the age diversity between the CEO and Chair and its impact on IE, thereby enhancing the understanding of this crucial aspect within corporate governance dynamics.

The study is organized as follows. This section is the introduction portion. After that, the literature review and hypotheses are shown in Section 2. Section 3 presents the sample selection process, methodology and variable definition, followed by the empirical results and robustness tests in Section 4. Section 5 concludes the paper.

The theoretical framework of this study draws on agency theory and upper echelons theory. Agency theory focuses on the relationship between principals (shareholders) and agents (managers), highlighting potential conflicts of interest and information asymmetries. Agency theory explains how weak corporate governance can exacerbate agency conflicts, leading to sub-optimal investment decisions and reduced efficiency. Managers may over-invest to fulfil personal interests or under-invest to avoid risks, which diminishes IE. Conversely, robust governance mechanisms can mitigate these conflicts, enhancing IE. Upper echelons theory emphasizes how top demographic attributes and personal characteristics of managers influence decision-making, including investments (Bamber et al., 2010). Factors like age, education and professional background shape managerial decisions, with age particularly impacting decision-making preferences as managers evolve. The theory underscores the importance of understanding managers’ evolving behaviours and decision preferences as they age, which subsequently affect strategic decision-making and IE. Additionally, it emphasizes the significant influence of TMT dynamics, particularly the relationship between the Chair and CEO, on corporate governance strength and IE. The harmonious relationship between the Chair and CEO affects governance strength, impacting efficiency. Integrating both theories offers a comprehensive understanding of how agency dynamics and managerial attributes collectively influence organizational IE.

Drawing on upper echelons theory (Hambrick and Mason, 1984), this study recognizes the pivotal role of top executives’ demographic characteristics in shaping corporate decision-making and strategic outcomes. As Bamber et al. (2010) emphasize, managerial personal attributes profoundly influence corporate accounting behaviour, reinforcing that observable traits can serve as proxies for deeper cognitive frameworks and social values. Demographic attributes such as age, education, ethnicity and gender have been widely used to infer the cognitive bases and social processes within TMTs (Kor, 2006). The composition of TMTs, particularly in terms of age, professional background, education and tenure, has been empirically linked to a broad array of organizational outcomes, including investment decisions (Herrmann and Datta, 2005; Raj, 2025), financial reporting quality (Habib and Hossain, 2013), strategic decision-making (Iaquinto and Fredrickson, 1997), tax planning strategies (Plečnik and Wang, 2021) and overall firm performance (Wang et al., 2016). Among these demographic factors, age emerges as a particularly salient determinant, encapsulating a dynamic array of personal attributes reflective of individuals’ life experiences (Herrmann and Datta, 2005). Tanikawa et al. (2017) further assert that age-related changes in personality traits engender consequential shifts in behavioural patterns, values, cognitive orientations and strategic decision-making preferences among individuals over time.

Corporate governance research underscores that strategic decision-making within TMTs is often marked by internal conflict, particularly under complexity and uncertainty (Mooney et al., 2007; Simons and Peterson, 2000). Conflicts within TMTs have essential implications for internal coordination and, ultimately, for a firm’s IE, whether constructive or disruptive. Among the various factors contributing to TMT conflict, demographic diversity, especially age differences, plays a critical role. Upper echelons theory posits that age differences among executives can lead to divergent values, cognitive styles and decision-making preferences. Younger CEOs may exhibit a higher risk appetite and a forward-looking strategic orientation, whereas older executives are often more risk-averse and conservative in their approach (Chindasombatcharoen et al., 2022), which can affect their preferences for different firm-level policies, including investment decisions, particularly curbing over-investment in the case of Chairman-driven caution or, conversely, exacerbating under-investment due to excessive conservatism (Jiang et al., 2009). Thus, perceptual misalignment between the Chair and CEO, stemming from age dissimilarity, can reduce internal coherence and obstruct optimal capital allocation. Research on TMT diversity further suggests that age heterogeneity often triggers task and relationship conflicts (Jehn, 1997). Task conflicts arising from differences in opinions and approaches to work-related issues can be beneficial if they foster comprehensive analysis and improved decision quality (Mooney et al., 2007; Goergen et al., 2015). In contrast, relationship conflict driven by interpersonal friction, lack of trust or miscommunication typically undermines team dynamics, reduces information sharing and impairs performance (Simons and Peterson, 2000).

Age dissimilarity among team members can exacerbate relationship conflicts, as individuals of similar age are more likely to share attitudes, opinions and beliefs, facilitating information exchange (Wagner et al., 1984). Conversely, Chairs and CEOs of varying ages may struggle to establish personal connections, leading to friction during information sharing and exacerbating mutual distrust and hostility, intensifying relationship conflicts (Al-Dubai, 2025). The failure to share information can result in biased decision-making (Gruenfeld et al., 1996), ineffective firm performance (Aldag and Fuller, 1993) and adversely impact IE. Moreover, research indicates that, when relationship and task conflicts coexist within an organization, the impact of task conflicts may be weaker or non-existent compared to relationship conflicts (Korsgaard et al., 2008). Additionally, the adverse effects of relationship conflicts may overshadow the potential benefits of task conflicts (De Dreu and Weingart, 2003). Cognitive conflicts between the board and the TMT can also affect how the board makes recommendations (Veltrop et al., 2025). Hence, effective conflict management interventions can mitigate relationship conflicts and indirectly address task conflicts (Choi and Cho, 2011).

IE is a pivotal concept in corporate finance, as outlined by Modigliani and Miller (1958), who advocate for firms to undertake projects with positive NPV at the risk-adjusted cost of capital. However, deviations from this ideal are common, leading to agency problems such as over-investment and under-investment. These inefficiencies manifest in various forms, including prioritizing short-term over long-term projects and pursuing unfavourable investment deals, ultimately eroding shareholder value. Empirical evidence indicates that not all investments are efficient and contribute to firm value. Biddle et al. (2009) posit that IE involves firms’ capability to allocate resources to projects with positive NPV while avoiding those with negative NPV. Efficient investment, as highlighted by studies such as Amberger et al. (2021) and Chay et al. (2023), leads to the generation of positive NPV and adds value to the firm. Conversely, Verdi (2006) noted that inefficient investment entails deviations from optimal investment levels, resulting in either over-investment or under-investment. Over-investment occurs when firms exceed their resources or invest in negative NPV projects, leading to financial losses and hindering future growth (Myers, 1977). Conversely, under-investment arises when firms lack sufficient financing for profitable projects, leading to missed opportunities for value creation and impeding future development (Myers, 1977). Such inefficient investment practices not only result in significant financial losses but also undermine investor interests and impede the healthy development of firms. Therefore, investment decisions are pivotal in shaping firms’ IE and overall value creation.

Corporate capital structure is another critical determinant of investment behaviour. Financially slack firms tend to over-invest, leveraging their access to excess cash (Opler et al., 1999). Conversely, firms facing financial constraints may under-invest, especially when subject to credit rationing due to high leverage (Myers, 1977). Monitoring and governance mechanisms, such as bank relations and accounting practices, also play vital roles in influencing IE. Close ties with banks mitigate agency problems, thus enhancing IE (Hoshi et al., 1991). Additionally, conservative accounting practices and comprehensive information disclosure reduce informational asymmetries, improving IE (Biddle et al., 2009).

Research on IE has traditionally focused on firm characteristics and investment attributes, but recent studies highlight the growing importance of managerial traits and institutional factors. Ownership structure is a key determinant, as separating ownership and control often leads to agency problems and inefficient investment decisions (Jensen and Meckling, 1976). Despite their access to low-cost capital, listed firms face coordination challenges due to dispersed ownership (Asker et al., 2015). Family firms show mixed outcomes, while aligned ownership and control may promote efficiency, and conflicting family interests can distort investment (Anderson et al., 2012).

According to agency theory, agency problems (such as information asymmetry and goal conflicts) will erode corporate governance, lead to sub-optimal investment decisions and reduce IE (Jensen and Meckling, 1976). Agency conflicts may lead to investment inefficiencies under weak corporate governance (Stein, 2003). Richardson (2006) shows that self-interested managers (such as CEOs) will use free cash flow (FCF) to improve performance and consume additional expenses, resulting in over-investment and inefficient investment. In addition, Bertrand and Mullainathan (2003) believe that managers could refrain from investing in profitable NPV projects to avoid risks, resulting in under-investment and low IE. On the contrary, high-intensity corporate governance can reduce agency conflicts and improve IE (Biddle et al., 2009; Chen et al., 2011a, 2011b). Furthermore, incentivization schemes significantly impact investment behaviour. Managers require risk-taking incentives to pursue advantageous yet risky projects (Gormley et al., 2013). Compensation plans prioritizing long-term stock returns over short-term gains incentivize efficient investment behaviour (Bizjak et al., 1993).

Managerial attributes, such as professional background, optimism and overconfidence, play pivotal roles in investment decisions. Overconfident CEOs tend to over-invest, particularly in financially constrained firms (Malmendier and Tate, 2005), and often pursue aggressive investment strategies that can destroy shareholder value, particularly in losses in M&A deals (Ben-David et al., 2013). Khedmati et al. (2020) provide evidence suggesting that firm CEOs with strong ties to independent board members are linked to inefficient investment practices. According to upper echelon theory, a company’s financial decisions largely depend on the personal perceptions of board members, and their risk preferences and strategic choices will inevitably incorporate information obtained from the companies they serve (Masulis and Zhang, 2019). Ullah et al. (2021) noted that female CEOs are associated with higher IE because they focus more on curbing over-investment rather than under-investment when making investment decisions. Female CEOs are associated with less managerial opportunism, and agency risk suggests that less managerial opportunism and fewer agency problems will improve IE. Similarly, some research results show that cooperation between CEOs and Chairs with homogeneous backgrounds (gender) can improve IE, especially between female CEOs and female Chairs (Ismail and Abdul Wahab, 2024).

In addition, the careers of TMTs will also affect IE. Ullah et al. (2022) showed that CEOs with military experience reduce agency conflicts, are less likely to be self-interested and are more inclined to pursue value-added projects, thereby increasing IE. This type of CEO is better at making decisions under high risk and pressure because they have previously had systematic professional training and combat experience in the military (Franke, 2001). In addition, Chair tenure negatively correlates with R&D investment (Azzam and Alhababsah, 2025). A stream of research indicates that board tenure brings a range of perspectives, experience and expertise to the decision-making process, which fosters comprehensive decision-making through the evaluation of the investment process. Furthermore, a diverse range of board tenure can help mitigate group thinking and enhance critical analysis, leading to more informed, innovative investment decisions that result in higher IE (Phuong et al., 2022). Chowdhury et al. (2023) find that highly powerful CEOs often lead to decreased efficiency through increased over-investment, particularly in contexts with high information asymmetry and agency problems. They also reveal that the presence of co-opted independent directors or externally recruited powerful CEOs worsens the situation. Overall, the discussed literature underscores the complex interplay between demographic attributes of TMT, investment decisions and contextual factors influencing their effectiveness. While empirical research predominantly focuses on the demographic characteristics of directors, CEOs and Cheif Financial Officers, there is a lack of empirical evidence on the impact of diversity within the TMT, particularly regarding the age gap between the CEO and Chair, on IE.

Based on the upper echelons theory, TMTs with diverse background characteristics exhibit varied behavioural tendencies that influence corporate strategic decisions (Hambrick and Mason, 1984). Diversity, defined by differences in demographic attributes like age and gender (Harrison and Klein, 2007), results in board members with distinct policy and decision-making preferences. Past empirical research indicates that diversity can impact firm performance positively or negatively, or show no significant relationship with team performance (Webber and Donahue, 2001). Chairs and CEOs with the same personality traits positively affect bank profitability (Le et al., 2025). Hence, Chair–CEO age dissimilarity may significantly shape corporate governance and investment decision-making preferences.

Diversity in teams, particularly in leadership roles such as the CEO and Chair, enhances investment efficiency by promoting thorough decision-making, mitigating risks, fostering creativity and innovation, deepening market understanding and improving governance and oversight (Xie et al., 2020; Wu et al., 2023). The age similarity between the CEO and the board of directors may impact management of earnings through the board’s monitoring and advisory functions in mergers and acquisitions (Nguyen et al., 2025). Furthermore, age similarity between the audit committee chairperson and the engagement partner can improve the quality of accruals and reduce the likelihood of restatements (Alhababsah and Alhaj-Ismail, 2025). By bringing together individuals with different perspectives, experiences and cognitive styles, diverse teams are better equipped to identify risks, explore innovative investment opportunities and make well-rounded decisions tailored to market needs and long-term strategic goals (Bernile et al., 2018). Such an inclusive approach reduces the likelihood of biased or risky investments and ensures alignment with shareholder interests, ultimately leading to more effective and efficient investment strategies. Empirical research shows that diversity within the board, such as gender, age, education and race, can positively impact group performance (Harjoto et al., 2018; Usman et al., 2022). Kurtulus (2011) argues that diversity between CEOs and TMT members fosters information sharing, enhancing integrated learning. Additionally, Goergen et al. (2015) highlight that cognitive diversity arises from disparities in beliefs, attitudes, experiences and values between Chairs and CEOs of varying ages. This cognitive diversity may generate task conflict, enriching executives’ information pool (Kurtulus, 2011). Similarly, according to organizational information processing theory, age diversity within TMTs provides varied perspectives and information (Wei and Wu, 2013), facilitating comprehensive evaluations and better strategies. This enhances decision quality and benefits company performance (Shin and Zhou, 2007). Likewise, Ullah et al. (2020a, 2020b) find that board diversity is associated with high IE.

Age dissimilarity can be detrimental to firm outcomes. Age dissimilarity within teams may exacerbate internal divisions and foster discriminatory treatment, ultimately reducing team cohesion (Simons and Peterson, 2000). This can result in increased coordination difficulties (Schmid and Mitterreiter, 2020), leading to delays in decision-making and potentially hindering investment processes (Ebert et al., 2017). Additionally, research by Talavera et al. (2018) suggests that a significant age gap may trigger internal conflicts, impeding constructive information sharing. Furthermore, age dissimilarity can give rise to relationship conflicts stemming from personality preferences and cognitive biases (Zenger and Lawrence, 1989), offsetting any potential benefits from task conflict caused by Chair–CEO age dissimilarity (De Dreu and Weingart, 2003). This discordance may lead to sub-optimal investment decisions and diminish IE (Simons et al., 1999). Moreover, studies by Katsiampa et al. (2024) have indicated a negative association between age dissimilarity and various firm performance metrics, including return-on-assets, sales-on-assets, cash flow proficiency and market-to-book value.

Based on the discussion above, age dissimilarity between the CEO and Chair can profoundly affect firm investment and IE, as evidenced by insights from agency theory and upper echelons theory. From an agency theory perspective, the relationship between the CEO and Chair, akin to a principal-agent dynamic, can be influenced by divergent risk preferences and time horizons stemming from their disparate ages (Hambrick and Mason, 1984). Younger CEOs may favour riskier, long-term investment strategies to maximize growth (McGuinness, 2021), while older Chairs, nearing retirement, might prioritize stability and short-term returns to safeguard investments (Carmeli and Halevi, 2009). This misalignment in risk preferences and time horizons can lead to conflicts over investment decisions, potentially impeding the firm’s ability to pursue value-maximizing opportunities (Jehn, 1995; Simons and Peterson, 2000). Additionally, from the upper echelons theory standpoint, age differences between the CEO and Chair can shape decision-making processes through cognitive and demographic factors (Zenger and Lawrence, 1989). A wide age gap may result in cognitive diversity and generational perspectives (Wei and Wu, 2013), with younger CEOs advocating for innovative investment approaches aligned with emerging trends (Gan, 2019), while older Chairs may prefer traditional strategies rooted in industry experience. However, it can also exacerbate demographic homophily within the TMT, limiting the range of perspectives considered and potentially hindering innovation. Consequently, empirical research must carefully navigate age dynamics among top executives, ensuring alignment in strategic priorities and fostering diversity of thought to enhance firm performance. Thus, we hypothesize as follows:

The study starts with an initial sample consisting of 5,049 A-share Chinese firms listed on the Shanghai and Shenzhen Stock Exchanges from 1998 to 2022. We obtained data from the China Stock Market and Accounting Research (CSMAR) database (Du and Boateng, 2015). During the phase of data collection, a total of 56,478 firm-year observations are obtained. The sample selection process is illustrated in Table 1. We selected the sample using the following criteria (Majeed et al., 2018). Because some variables in the research model need the data samples collected from t − 1 to t − 3 years, such as the standard deviation of sales (SDSALE), the research period is from 2001 to 2022. We removed 3,134 observations produced from 1998 to 2000. Subsequently, we dropped missing observations on control variables, including Chair and CEO age, gender, tenure, change, Z-score, sales, state-owned shares percentage, the standard deviation of sales, management fee and K-structure. Third, we deleted firm-year observations with missing industry codes. We also eliminated finance industry firms (e.g. banks, insurance companies and investment trusts) (SIC codes between J66 and J69) because of their different financial characteristics (Fang et al., 2021). Ultimately, the sample consisted of 3,517 non-financial listed firms, comprising 22,792 firm-year observations between 2001 and 2022. To avoid potential biases from outliers, we winsorize all the continuous variables from top to bottom at a 1% level.

We report the sample distribution based on industry classification in Table 2. Following Xie et al. (2021), we classify firms by industry according to CSMAR’s SIC codes (17 industries in total). The sample distribution is in good condition. The manufacturing industry is the economically dominant industry of the Chinese market and accounts for the majority of the research samples. The top five industries in the sample are metal and non-metal (42.09%), chemical manufacturing (15.55%), IT (6.77%), machine manufacturing (5.33%) and retailing (5.03%).

This study measures IE using a widely accepted approach from prior literature (Richardson, 2006; Zhang et al., 2019). Richardson’s (2006) model estimates the expected (or normal) level of investment based on firm-specific characteristics, and the residuals from this regression are interpreted as deviations from optimal investment. A positive residual suggests over-investment, while a negative residual indicates under-investment, with both representing inefficiencies in capital allocation. Thus, residuals closer to zero imply higher IE. Two operational measures are used to capture and interpret this construct more effectively. First, the continuous variable IE is calculated as the negative absolute value of the residual (|residual| × –1), following Dinh et al. (2023) and Zhang and Michael (2023), where larger (i.e. less negative) values denote greater efficiency. Second, a binary variable investment efficiency dummy (IED) is constructed based on the annual distribution of residuals, consistent with McNichols and Stubben (2008). Firm-year observations falling within the interquartile range (25th to 75th percentile) of residuals are classified as having normal investment levels and assigned a value of 1, indicating higher efficiency. At the same time, those in the top or bottom quartiles (representing over- and under-investment, respectively) are coded as 0, reflecting lower IE. These complementary measures allow for a nuanced assessment of firm investment behaviour across both continuous and categorical dimensions.

We follow Goergen et al. (2015) to measure Chair–CEO age dissimilarity using two variables: GAPS (age gap signed) and GAPU (age gap unsigned). GAPS is defined as the Chair’s age minus the CEO’s age, producing either positive or negative values depending on whether the Chair is older or younger than the CEO. This signed measure captures potential cognitive and generational differences that may influence decision-making dynamics. Meanwhile, GAPU represents the absolute value of GAPS and reflects the magnitude of the age difference between the Chair and CEO without considering who is older (Zhou et al., 2019).

Previous studies suggest that Chair–CEO relationships affect IE and firm value (Duchin and Sosyura, 2013). We control demographic similarities and differences, including educational background (DIFEDU), cultural origin (DIFNAT), gender (DIFGEN) and shared tenure (JONTENU), following Khedmati et al. (2020) and Ullah et al. (2020a, 2020b). DIFEDU, DIFNAT and DIFGEN are dummy variables equal to 1 if the Chair and CEO differ in education, nationality or gender, respectively. JONTENU measures the number of years they have served together. Given that younger or newly appointed CEOs invest more efficiently (Xie, 2015; Guizani, 2024), we control CEO and Chair age (CEOAGE, CHAIRAGE), position changes (CHANGE1, CHANGE2) and tenure (TENURE1, TENURE2) (Baran and Forst, 2015; Zhang and Michael, 2023). Board attributes also affect investment decisions. Board diversity improves efficiency, while larger boards may increase under-investment risk (Nor et al., 2017). BODSIZE measures board size (Javeed and Azeem, 2014).

Firm-level characteristics controls include the book-to-market ratio (BKTMK) (Majeed et al., 2018), financial leverage (LEVER) measured as the firm’s total liabilities over total assets (Ling and Wu, 2022), FCF (Richardson, 2006; Biddle et al., 2009), firm size (FSIZE) expressed as the natural logarithm of the firm’s total assets (Biddle et al., 2009; Majeed et al., 2018), sales volatility (SDSALE) measured as sales divided by the average total assets from year t − 3 to t (Jung et al., 2014; Khedmati et al., 2020; Biddle et al., 2009), and sales growth (SALEGR) measured by the changes in sales between year t and year t − 1 (Majeed et al., 2018). To account for ownership effects, we include the percentage of state-owned equity (SOE%) (Tran, 2020; Ullah et al., 2020a, 2020b; Majeed et al., 2018). Financial performance controls include LOSS (Biddle et al., 2009), Altman’s Z-score for financial distress (Tran, 2020) and asset tangibility (TANGI) measure TANGI as the property, plant and equipment ratio to total assets (Tran, 2020; Biddle et al., 2009). Market structure (K-structure) follows Biddle et al. (2009). We include management expense (MFE), defined as management cost over total assets, and firm age (FIRMAGE), which is the log of years since establishment. We refer to  Appendix for further details on the variable definition.

To empirically examine the relationship between Chair–CEO age dissimilarity and firm IE (H₁), we develop the regression model presented in equation (1). To measure IE (INVEFFI), we use two dependent variables: IE, a continuous measure, and IED, a binary variable indicating optimal investment levels. When the dependent variable is IE, we use ordinary least squares (OLS) regression, and for IED, we apply logistic regression. The key independent variable is the Chair–CEO age gap (GAP). We refer to Section 3.2 for the details on variable measurements. Equation (1) is as follows:

To test H₁, the primary variable of interest is $β1GAPi,t$⁠, $GAPi,t$ is the proxy for the age difference between Chair and CEO. As shown in Section 3.2, we consider two different measures of the age gap: GAPS and GAPU. In equation (1), a positive (negative) coefficient of $β1GAPi,t$ will indicate that age dissimilarity increases (decreases) IE.

Table 3 reports the descriptive statistics for the key variables used in this study. The average IE is −0.10, with a median of −0.08 and a standard deviation of 0.08, suggesting that, on average, firms in the sample tend to under-invest. Additionally, 48% of the sample firms exhibit higher IE, as indicated by the IED. The average age gap between the CEO and Chair (GAPS) is 3.76 years, with the absolute age difference between CEO–Chair (GAPU) averaging 5.93 years. The mean age of the Chair and CEO is 52.44 (48.68) years. The sample shows a significant educational diversity between CEOs and Chairs, with 87% of the pair having different educational backgrounds, but there is limited national diversity, as only 2% of the CEOs and Chairs come from different countries. The average co-work tenure between the Chairs and CEOs is 3.41 years, consistent with Zhu et al. (2021). The mean board size is 8.75. The mean of the firm size is 22.05. The sample firm’s leverage (LEVER) is 0.45, which indicates that 45% of the sample firm’s assets are tied to liabilities. The mean cash holding (CASH) is 0.17, suggesting that 17% of the total assets are cash. While the sample firm has 14% observations and reports loss, the sales growth mean is 0.22. The sample’s average SOE holding (SOE%) is relatively small at 8%. Overall, the descriptive statistics are consistent with existing literature, especially Chinese market studies (Zhai and Wang, 2016; Zhou et al., 2019; Zhu et al., 2021).

Table 4 presents the results of the mean difference test comparing key variables between firms with Chair–CEO age dissimilarity (GAPD = 1) and those without (GAPD = 0). Firms that have Chair–CEO age dissimilarity exhibit significantly lower average IE (IE and IED), suggesting a potential negative relationship between the age gap and investment performance. These firms also allocate a smaller proportion of resources to investment activities. In contrast, differences in education (DIFEDU), nationality (DIFNAT) and gender diversity (DIFGEN) between Chair and CEOs are more prevalent in firms without age dissimilarity (GAPD = 0), indicating greater demographic diversity in that group. Interestingly, joint tenure of Chair and CEOs (JONTENU) is longer in firms with an age gap, suggesting a potentially more established working relationship despite generational differences. Firms with existing Chair–CEO age dissimilarity groups are relatively older (FIRMAGE) and larger in size (FSIZE). However, companies with non-existent age differences have a higher state-owned percentage (SOE%), larger board size (BODSIZE) and higher FCF. All reported differences are statistically significant at the 5% level, supporting the robustness of the findings. These findings provide initial empirical support for the hypothesis that Chair–CEO age dissimilarity may influence firm investment efficiency through organizational and governance dynamics.

Table 5 presents the pairwise correlations among key variables. Chair–CEO dissimilarity, both GAPS and GAPU, shows a significant negative correlation with IE (IED), supporting the hypothesis that greater age gaps are associated with lower IE. Specifically, IED is negatively correlated with GAPS (−0.017, p < 0.01) and GAPU (−0.012, p < 0.05). GAPS and GAPU significantly correlate with several control variables, indicating their broader relevance within the corporate governance structure. Specifically, firms with higher leverage, size, sales volatility, sales growth and SOE ownership tend to have lower IE. In contrast, higher book-to-market ratio, FCF, Z-scores and asset tangibility are linked to better IE. The results are consistent with prior studies that used the Chinese market as a research sample (Ullah et al., 2020a, 2020b).

This section reports the findings of H₁, which examines the relationship between Chair–CEO age dissimilarity and firm IE. Table 6 reports the results of both OLS and Logit regression using two dependent variables: IE and IED. Chair–CEO age dissimilarity is measured using two proxies: GAPS and GAPU. Columns (1) and (2) report GAPS and GAPU on IE, respectively. Column (1) shows a significant negative association between GAPS and IE (coefficient = −0.0013***, t-statistics = −2.86), statistically significant at a 1% level. In Column (2), the coefficient on GAPU is negative (coefficient = −0.0002*, t-statistics = −1.76). Columns (3) and (4) report the findings of GAPS and GAPU on IED, respectively. Column (3) shows the coefficients are negative for GAPS (coefficient = −0.055***, t-statistics = −2.825) and GAPU (coefficient = −0.010***, t-statistics = −3.165), statistically significant at the 1% level. Our findings indicate that IE reduces when the age gap between Chair and CEOs increases, supporting the H₁. In terms of the economic significance of our findings, a one-standard-deviation increase in GAPS is associated with [(−0.001 * 8.24/−0.10) = 0.0824, i.e. 8.24%] an 8.24% decrease in IE to the mean for the sample firms indicates the Chair–CEO age dissimilarity effect towards firm IE is economically significant. Economic significance using the GAPU suggests that a one-standard-deviation increase in GAPU is associated with [(−0.0002 * 6.95/−0.10) = 0.0139, i.e. 1.39%] a 1.39% decrease in IE to the mean for the sample firms, which indicates the Chair–CEO age dissimilarity effect towards firm IE is economically significant.

In terms of control variables, we find that a firm which has Chair and CEO have different educational backgrounds (DIFEDU), older Chair (CHAIRAGE), longer Chair tenure (TENURE1) or CEO tenure (TENURE2) has a positive association with IE. Also, we find that the firms with higher FCF, less performing firms (LOSS), high distress risk (Z-score) and more tangible assets (TANGI) increase IE. In contrast, a firm with a longer joint tenure between Chair and CEO (JONTENU) and an older CEO (CEOAGE) has lower IE. Also, a firm with higher sales growth (SALEGR) and more volatile sales (SDSALE) reduces IE. The results are consistent with the previous findings (Biddle et al., 2009; Ullah et al., 2020a, 2020b; Khedmati et al., 2020). To rule out the concern of multicollinearity within our analysis, we perform variance inflation factor (VIF) diagnostic tests in the regression model. The results demonstrate that all VIFs are below 10 (Marquaridt, 1970), indicating that multicollinearity does not impact the results. The adjusted R² (pseudo-R²) ranges between 14 and 26.

We follow Richardson (2006) in the primary analysis to measure IE. Empirical research also suggests alternative measures of IE and, therefore, to examine the robustness of findings (Shahzad et al., 2018), we use alternative measures of IE following Biddle et al. (2009) and Chen et al. (2011a, 2011b). Consistent with the primary measure of IE (IE and IED), we develop four additional measures for dependent variables. Following Biddle et al. (2009), we developed IE_Biddle and IED_Biddle. Similarly, with the measure of Chen et al. (2011a, 2011b), we develop IE_Chen and IED_Chen. Findings are reported in Table 7. We re-perform the equation (1). Columns (1) and (2) examine the association between IE_Biddle and GAPS and GAPU, respectively. We find a negative association between Chair–CEO age dissimilarity and IE (coefficients = −0.00005, −0.0002**; t-statistics = −0.15, 2.305). Columns (3) and (4) are the association between IED_Biddle, GAPS and GAPU. Consistently, we find a negative association between Chair–CEO age dissimilarity and IE (coefficients = −0.010, −0.02014***; t-statistics = −0.296, 3.366). Columns (5) and (6) test the association between IE_Chen and GAPS and GAPU, respectively. Columns (7) and (8) correspond to the association relationship between IED_Chen and GAPS and GAPU.

To assess the robustness of our results, we construct four alternative binary measures of Chair–CEO age dissimilarity: GAP20, GAP15, GAP15 and GAP5, which equal 1 if the age gap between the Chair and CEO is greater than or equal to 20, 15, 10 or 5 years, respectively, and 0 otherwise. Our approach to measuring GAP20, GAP15, GAP15 and GAP5 is consistent with Goergen et al. (2015) and Zhou et al. (2019). Also, Zhu et al. (2021) measure 10-year intervals for the Chair–CEO age dissimilarity. We re-estimate equation (1) by incorporating these four alternative measures. Table 8 reports the findings. In Column (1), we find a negative association with GAP20 (coefficient −0.0056**, t-statistics = −2.242), indicating that firm IE reduces when the Chair–CEO age gap (GAP20) is higher. In Column (2), we report the findings on GAP15, which is consistent (coefficient −0.0049***, t-statistics = −2.709) with the earlier findings. The findings consistently show a negative association in Columns (3) and (4). We also consider the alternative measure of IE (IED) and report the findings in Columns (5) to (8). Overall, these findings reinforce the conclusion that greater Chair–CEO age dissimilarity is associated with lower IE, supporting H₁.

We conduct cross-sectional analyses based on market competition, ownership concentration and state ownership to further explore the mechanisms underlying the observed negative relationship between Chair–CEO age dissimilarity and IE. These tests serve two purposes:

1. to clarify the conditions under which the age gap impacts IE and

2. to enhance the robustness of our findings by ruling out confounding economic factors.

Market competition is an essential mechanism of corporate governance that can influence managerial behaviour and strategic decisions. Prior research evidence shows that firms operating in more competitive environments exhibit higher IE due to stronger external pressure (Ullah et al., 2022). In contrast, monopolistic firms may be more insulated from internal conflicts, making managerial misalignment less consequential. To test whether market competition moderates the relationship between Chair–CEO age dissimilarity and IE, we split the sample based on industry concentration using the Herfindahl–Hirschman Index (HHI). Firms with an HHI below or equal to the mean are classified as operating in competitive industries, while those above the mean are classified as monopolistic.

Table 9 presents the regression results. In competitive industries [Columns (3), (4), (7) and (8)], we find a statistically significant negative association between Chair–CEO age dissimilarity and IE, supporting the notion that interpersonal conflicts have greater adverse effects when firms are under competitive pressure. In contrast, for monopolistic firms [Columns (1), (2), (5) and (6)]. The relationship is statistically insignificant, indicating that the impact of Chair–CEO age dissimilarity on investment outcomes is attenuated in less competitive environments. Our findings suggest that the detrimental effect of age dissimilarity on IE is more pronounced in competitive markets, where coordination and unified strategic direction are crucial. In a monopolistic setting, the same misalignment appears less disruptive due to reduced external pressures and greater strategic flexibility.

Ownership concentration plays a critical role in shaping corporate governance and influencing managerial decisions, including investment behaviour (Buckley et al., 2018). Aggarwal and Samwick (2006) report that, when controlling shareholders have absolute decision-making power, they tend to influence managers’ investment decisions. Thus, examining whether ownership concentration affects the association between Chair–CEO age dissimilarity and IE is necessary. In China’s corporate landscape, many listed firms have concentrated ownership, where a dominant shareholder holds significant equity. This setup poses governance challenges, diminishing oversight from smaller shareholders. Concentrated ownership can mitigate agency costs, but dispersed ownership may lead to internal power struggles and affect investment decisions (He et al., 2022). Additionally, age disparity between the Chair and CEO can exacerbate conflicts, impacting IE and highlighting the complex relationship between ownership structure, managerial dynamics and corporate performance in China.

To test whether ownership concentration moderates the relationship between the Chair–CEO age gap and IE, we divide the sample based on block-holder dominance. A dummy variable is generated, assigning a value of 1 to instances where the largest ownership owns 50% or more of the firms and 0 otherwise. Table 10 presents the regression results. Columns (1), (2), (5) and (6) represent firms characterized by significant ownership concentration (Blockholder 50% = 1). Meanwhile, Columns (3), (4), (7) and (8) report firms without significant ownership concentration (Blockholder 50% = 0). In firms with lower ownership concentration (Blockholder <50%), a significant negative association emerges between Chair–CEO age dissimilarity and IE [Columns (3), (4), (7) and (8)]. This suggests that, in the absence of dominant ownership, internal dynamics such as generational or cognitive conflicts between the Chair and CEO can lead to fragmented decision-making and inefficient investment outcomes. Conversely, in firms with high ownership concentration [Columns (1), (2), (5) and (6)], the relationship is statistically insignificant, implying that strong shareholder control may neutralize or override executive misalignment. Findings indicate that the negative effect of Chair–CEO age dissimilarity on IE is more pronounced in firms with dispersed ownership, where decision-making is less centralized and more susceptible to interpersonal executive conflict.

SOEs in China present unique institutional characteristics that may influence the relationship between the Chair–CEO age gap and IE. Due to governmental involvement in ownership and executive appointments, SOEs often operate under dual mandates, pursuing both economic and socio-political objectives (Faccio et al., 2006). These entities may benefit from easier access to capital and preferential treatment, which could reduce the negative impact of executive misalignment on investment decisions. Furthermore, collectivist cultural values prevalent in SOEs may promote internal cooperation and mitigate interpersonal conflicts stemming from the age difference between the Chair and CEO. To assess the moderating role of SOEs, we divide the sample into SOEs (SOE = 1) and non-SOEs (SOE = 0), with SOEs accounting for 36.83% of the total sample.

Table 11 presents the results. In non-SOEs [Columns (3), (4), (7) and (8)], Chair–CEO age dissimilarity is significantly negatively associated with IE, with all coefficients statistically significant at the 1% level. In contrast, the relationship is insignificant among SOEs [Columns (1), (2), (5) and (6)]. These findings suggest that, in non-SOEs, where firms are typically more profit driven and exposed to market discipline, leadership misalignment reflected by an age gap can hinder effective decision-making and resource allocation. In SOEs, however, institutional buffers and emphasis on hierarchical harmony may dilute the adverse effects of Chair–CEO age disparity. Overall, the findings highlight that the negative impact of age dissimilarity on IE is more pronounced in market-oriented, non-SOEs.

We apply Heckman’s two-stage procedure to address potential endogeneity arising from sample selection bias, particularly since Chair and CEO appointments are non-random. Sample selection bias may occur if unobservable factors influencing Chair–CEO age dissimilarity also affect IE, thus potentially biases our main results (Tucker, 2010). To address this, we use the Heckman (1979) two-stage model for mitigating sample selection bias. Following Zhu et al. (2021), we use a dummy variable, CRG (cultural revolution generation), as an instrumental variable for Heckman’s two-stage test. CRG is equal to 1 if the Chair or CEO was at least 16 during the Cultural Revolution (1966–1976) and 0 otherwise ^[1]. This variable captures the generational cognitive conflicts due to differing values and thought processes developed during this tumultuous period, which affect firm governance. The CRG variable, correlated with age but not with firm IE, prevents multicollinearity between explanatory variables and the inverse Mills ratio (IMR). As shown in Table 3, the mean values of ChairNCN and CEONCN are both 0.98, indicating that in 98% of observations, both the Chair and CEO are Chinese nationals. Therefore, this instrumental variable is appropriate for the study sample.

We incorporated the dummy variable GAPS_mean or GAPU_mean as the dependent variable in the first stage of the model ^[2]. The IMR is computed in this first-stage probability model and included in the second-stage regression model ^[3] as an additional independent variable to assess sample selection bias. The first-stage regression results are presented in Columns (1) and (4) of Table 12, where a Probit regression was performed using GAPS_mean and GAPU_mean, respectively. The association coefficient between GAPS_mean and CRG is positive (coefficient = 0.781, t-statistics = 8.880) and significant at the 1% level. Similarly, the coefficient between GAPU_mean and CRG is positive (coefficient = 1.645, t-statistics = 53.815) and significant at the 1% level. The second-stage regression results, shown in Columns (2), (3), (5) and (6), indicate that the coefficients of GAPS/GAPU on IE [Columns (2) and (5), OLS regressions] and on IED [Columns (3) and (6), Logit regressions] are negative and significant. The association between IMR and IE/IED is negative and mostly insignificant [Columns (5) and (6)]. Thus, this study has no sample selection bias, and all findings support our hypothesis.