Do board-CEO ties in outsider CEO succession raise companies' returns?

The selection of a new chief executive officer (CEO) is one of the most strategic decisions for a corporation’s board of directors (Boivie et al., 2021). In the past, boards could rely almost exclusively on internal successions as the required expertise was company-specific and acquired through increasingly senior roles within a company (Kotter, 1982). However, over recent decades, several forces have pushed up the proportion of outsider CEO appointments to around a third of all CEO successions (Balsmeier et al., 2013). In particular, globalization and outsourcing have transformed the industrial structures of high-income economies toward specialized services in a world-wide competitive environment (Palmisano, 2006). Consequently, the CEO role has become more focused on strategy and decision-making capabilities requiring generalist rather than company-specific knowledge (Frydman, 2007; Shapiro, 2000).

The greater emphasis on hiring outsider CEOs has been accompanied by the rising challenge of determining their true abilities, as prospective external hires are relatively unknown to the new employers. Unlike internally groomed CEOs, whose suitability for promotion was determined by trusted historical performance data, outside appointments are selected in the context of lesser information about performance (Karaevli, 2007), with lower reliability than internally-generated data (Cappelli and Hamori, 2014).

Companies have therefore had to develop new means to assess the suitability of prospective CEOs. These include directors referring known prospective candidates and the contracting of executive search companies to mediate exchanges between the employer and outsider CEO candidates (Khurana, 2002). Several studies have explored aspects of the information sharing and influencing role that directors play in recommending and appointing other corporate officers, including CEOs (Westphal and Shani, 2016; Westphal and Stern, 2007). However, the outcomes of these referrals, particularly when sourced from professional director ties has not been examined. This knowledge gap is important because CEOs’ characteristics shape the development and execution of strategy which underpins organizational performance (Finkelstein et al., 2009; Hambrick and Mason, 1984; Meeks, 2015; Withers and Fitza, 2016). The lack of insight into the effects of professional director–CEO ties in the CEO labor market serves as the motivation for this study.

Professional director–CEO relationships likely cover performance information that is of value to the employing company (Ekinci, 2016), especially in CEO labor markets characterized by lower transparency and market disclosure requirements about the identify of CEOs and directors and their affiliations (Bushman et al., 2004; Djankov et al., 2008). While connectivity is expected to reduce the informational disadvantage of the employing company about the track record of a CEO candidate, it also creates the possibility of collusive behaviors: is this the case? Do director-CEO past professional connections matter in the selection of a new outsider CEO and, if so, are they viewed positively by the investors of the employing company?

We address these questions by examining the role of board–CEO ties in outsider CEO successions and their effects on company performance across 22 countries in high- and middle-/low-income countries following different institutional environments and approaches to corporate governance. As a result, we contribute to a better understanding of outsider CEO successions and its relationship with models of corporate governance. We expect that these relationships will lead to a discernible positive reaction by investors in securities markets because they provide boards with an additional tool to screen otherwise unknown and unobservable features of outsider CEO candidates. The theory of information economics informs us that when there is greater information asymmetry between investors and boards, the role of board–CEO ties becomes more important as a screening mechanism to reduce that information asymmetry (Akerlof, 1970; Bushman et al., 2004; Stiglitz, 1975). By focusing on professional relationships, we can better identify the source of informative news about the past performance of a prospective CEO, and anticipate them to positively influence the employing company’s market return outcomes (Jensen and Meckling, 1976). By carrying out the analysis across several markets differing in national jurisdictional environments (Aggarwal and Goodell, 2015), we can explore whether any benefit of director–CEO relationships will vary according to the degree of transparency in specific CEO labor markets (Bushman et al., 2004).

In sum: (1) we interpret the connectivity between outsider CEOs and one or more members of the board of the hiring company as reducing the uncertainty about the suitability of the outsider CEO; (2) we view differences across jurisdictions as indicators of the degree to which investors hold boards to account; and (3) we expect such varying degrees of governance transparency to alleviate the informational disadvantage of boards about the new CEO’s suitability. The insight of this relationship extends the CEO succession and management literature’s understanding of the role of asymmetric information between investors, boards and CEOs in outsider CEO successions (Bergh et al., 2019).

The empirical analysis draws on a sample of 1,136 outside CEO succession events and their public companies’ performance. This sample spans all outsider CEO successions that occurred between 1992 and 2018 across 22 countries that include English common law-founded markets such as the United States (139 Fortune 1000 companies), the United Kingdom, Australia and Canada as well as markets in Europe, Asia, South Africa and Latin America. The global coverage of the sample enables us to examine how varying levels of governance transparency place boundary conditions on the information asymmetry that affects boards and CEOs.

As a base-line, we use a market-based measure of company performance – the cumulative market-adjusted returns (CARs). This variable captures investors’ viewpoint about the expected effect of connectivity on overcoming the employing company’s informational disadvantage in appointing an external CEO. Such a measure, utilizing companies’ stock returns minus an industry benchmark, is the most commonly used market-based measure in the CEO performance literature (Cook and Zhang, 2023). A market-based measure also captures a more comprehensive measure of corporate value than accounting-based measures because it incorporates investors’ expectations, market trends and potential future growth. CEO performance studies that have used market-based measures, when exploring aspects of board–CEO relationships, have found a range of results. For example, investors can react favorably to a new CEO appointment when there is greater board and CEO human and social capital or adversely when perceived political interference may impede effective CEO selection (Fan et al., 2007; Tian et al., 2011).

We complement this analysis with three accounting-based measures of financial performance – return on assets (ROA), return on invested capital (ROIC) and return on sales (ROS) – to reflect the institutional viewpoint, as set by accounting standards, of the same event (CEO succession). Accounting-based measures of financial performance have the advantage of more closely reflecting companies’ actions, over which CEOs can exert direct control. ROA, in particular, has been used extensively as a benchmark in the CEO performance literature, including in studies that have shown the benefits of board–CEO ties in CEO succession events (Wang, 2021; Zhu and Shen, 2016). By drawing on both market- and accounting-based measures of CEO performance and including a range of jurisdictions which exhibit differences in corporate governance, capital market regulation and accounting standards, our study aims to provide a comprehensive understanding of CEO performance.

To quantify the effect of connectivity over time and overcome the potential masking effect of confounders, separate analyses are carried out for the 12-month period before and after the appointment, with windows expanding up to the subsequent two years.

The results suggest that board–CEO ties improve companies’ performance, serving both investors’ and management’s interests. We also find that such outcomes vary across jurisdictions in line with different degrees of market transparency and corporate governance regulation. In the United States, United Kingdom, Australia and Canada – characterized by the prevalence of company disclosures concerning the backgrounds, identity and relationships of managers and directors – board–CEO ties have little empirical effect on market- and accounting-based measures. This is likely because greater governance transparency means any relevant information about the identity and affiliations of new CEOs has already been disclosed to the market and is already widely and publicly available (Lubatkin et al., 2016).

By contrast, in markets where average governance transparency indicators are lower, as in continental Europe, opposite results emerge: connected CEOs positively affect CARs and both ROA and ROIC. These results are consistent with the argument that where a more severe information asymmetry is likely, board–CEO ties can reduce the hiring company’s disadvantage toward the new CEO (Westphal, 1999).

In countries with markets listing large family-owned companies but where governance transparency is relevant because of their openness to foreign institutional investors, such as Hong Kong and Japan, the appointment of a non-connected CEO positively affects CARs. However, the appointment of a connected CEO – which somewhat maintains the company under broad family control – has no statistical effect (Froot, O’Connell and Seasholes, 2001). Finally, in emerging markets that are open to fewer international investors, like Brazil, India and South Africa, prior professional director–CEO relationships have no detectable effect nor do non-connected CEOs.

Together, these results represent three contributions to the existing CEO succession literature. First, they show that governance transparency may underpin an informational advantage for boards when appointing a new outsider CEO. This advantage moderates the role of prior board-CEO ties in easing outsider CEO successions. Second, the results show that varying governance transparency may play a role in CEO succession events globally. Finally, they show that CEO-led company performance varies according to whether one uses indicators reflecting investors’ reactions or accounting reporting. Overall, the results paint a far more nuanced picture of the role of directors in outside CEO successions than currently explored by the CEO succession literature. Building on Khurana (2002) that the act of directors referring CEO candidates in CEO successions reduce information asymmetry and Wang (2021) that professional director–CEO relationships result in a positive ROA effect, we show that the role of directors, in referring outside CEO candidates, carries out a far more complex function than has previously been explored.

The CEO labor market has transformed in recent decades as low-skilled manufacturing industries have been outsourced to lower-cost, developing countries and high-income economies’ industrial bases have become more concentrated in higher value, specialized, technological and professional services, in a more competitive and globalized economy (Bryson et al., 2004; Saxenian, 2006). As a result, high-income economies now consist of a larger proportion of knowledge-intensive industries made up of companies that seek to exploit greater intellectual capital and informational labor that is capable of handling more complex information flows. These capabilities are not necessarily dependent on CEOs having previously worked for the company (Bertrand, 2009; Groysberg et al., 2006; Murphy and Zabojnik, 2007; Pyöriä, 2005). These industrial and sociological changes have driven greater demand for CEOs with proven strategic leadership and decision-making abilities that can be obtained in multiple company environments; and they have reduced the value of internally developed CEOs whose strengths lie with more operational and company-specific knowledge (Gathmann and Schonberg, 2010; Gibbons and Waldman, 2004; Kambourov and Manovskii, 2009; Pfeffer, 1994, 1998; Rizzi et al., 2020).

Over the same period, deregulation and the liberalization of labor markets in the United States and other western economies have increased labor mobility by loosening ties between employers and employees and leading to greater employee turnover (Finlay and Coverdill, 2000; Peck, 1996). Consequently, companies’ internal labor markets have weakened, and the need to meet succession planning requirements has made the global search for talent a greater priority for managers in multinational corporations (Faulconbridge et al., 2009). Outside CEO appointments now account for over 20% of CEO successions, up from less than 10% fifty years ago (Khurana, 2002; Murphy et al., 2007; Nickisch, 2016) although there is some recent evidence that this growth is plateauing (Cappelli et al., 2024; Cziraki and Jenter, 2021). These political, economic, sociological and legal changes have also seen the emergence of a multi-billion dollar global executive search industry that supports boards in identifying new CEOs and top managers in succession events (Beaverstock et al., 2010).

As the proportion of outsider CEOs has grown, companies have had to address the increased asymmetric information problem of attracting new leaders of which they have limited direct experience and knowledge. Any selection of a new CEO involves considerable complexity and uncertainty. Public company CEOs are responsible for the overarching performance of companies, including the development of and implementation of strategy, leadership, decision-making and the effective management of a range of constituents and external stakeholders (Andrews, 1971; Dewar et al., 2019; Mintzberg, 1973). There is also considerable uncertainty surrounding emergent challenges that a new CEO will face. These could include external environmental shocks, increased competitive intensity and the threat of new market entrants, technological disruption and changing regulatory environments (Aguilar, 1967).

Adding to the difficulty in predicting CEO effectiveness, new CEOs are expected to possess a range of potentially relevant attributes, only some of which might be observable to boards (Quigley et al., 2020; Zajac, 1990; Zhang, 2008). This is particularly problematic in the case of outsider CEOs, where even less is known about a given CEO’s character, leadership and decision-making capabilities as they relate to the specific CEO appointment.

A large literature addressing the topic of asymmetric information highlights that when agents are better informed than principals about their true skills and capabilities, the risk of selecting agents with inferior talents is higher (Akerlof, 1970; Shen and Cannella, 2002). In the context of outsider CEO successions, this potential problem is worsened by the incentives that low-quality agents have to misrepresent abilities that boards cannot completely verify. This can result in negative surprises such as company underperformance (Quigley et al., 2019) or overcompensation.

To address this problem, companies have had to develop new means, such as the greater use of labor market intermediaries (including executive search companies) that reduce information asymmetry by coordinating and mediating exchanges between directors and candidate pools and which legitimize the search process by signaling to constituents that a professional process has been undertaken. Specific CEO search activities undertaken by the executive search consultants include: liaising with directors within their professional networks to generate prospective CEO candidates, conducting structured interviews, checking 360 degree references, carrying out multi-faceted assessments and verifying prospective CEOs’ credentials through referral networks, as well as triangulating negotiations (Khurana, 2002). In some cases, directors and executive search consultants target executives at stronger competing companies as a proxy for ability. Despite the adoption of these search and assessment practices, there are still substantial acknowledged limitations (Boivie et al., 2025; Hamori, 2010; Sackett and Lievens, 2008; Schmidt and Hunter, 1998).

The lack of goal alignment between managers and owners is a well-known problem to managerialist theories of the corporation (Fama, 1980). Managerialism’s principal-agent problem has grown in recent decades, particularly in the United States and other common law countries such as the United Kingdom and those with Commonwealth heritage, as arms’ length institutional investors play a more prominent role as distributed owners of corporations (Dalton et al., 2007; Khurana, 2002). To counter the problem of management acting in its own interests, a board that is independent of management has been proposed to ensure management’s actions align with investors’ goals (Fama and Jensen, 1983a; Jensen and Ruback, 1983). Any director–CEO ties would necessarily reduce this independence and therefore could be collusive, enabling management to act in its own interests.

In an international context, approaches to corporate governance that affect the relationship between owners and managers vary. In the common law setting, the large proportion of dispersed, arms’ length institutional investors that make up company share registers, necessitates independent boards to effectively monitor management (Boyd et al., 2011). However, in Europe, where many countries have civil-law origins, and Asia, where there are typically a smaller number of large private or family group investors that are also actively involved in company management, board independence matters less because there is not the same distinction between owners and management (Lubatkin et al., 2016; Yusof, 2016). In these “insider” systems, a more collaborative, stewardship-focused approach to corporate governance exists where management and investors work together in the pursuit of shared goals (Claessens et al., 2000; Davis et al., 1997; Fan and Wong, 2005). Consequently, it is possible that any possible collusive effects associated with board–CEO ties are more pronounced in common law countries, where board independence is particularly important.

Management scholars, focused largely on the United States and United Kingdom-focused common law contexts, have explored the impact of director–CEO ties on company performance by focusing on professional, political, social and friendship ties (Boyd et al., 2011; Finkelstein et al., 2009; Schepker et al., 2017; Westphal and Park, 2020). Researchers have used a combination of empirical, field and survey-based methods, with instrumental variables spanning professional director–CEO connections, joint political party or leisure club memberships or friendship ties to examine the impact of board–CEO collaboration on company performance (Chidambaran et al., 2010; Fan et al., 2007; Fracassi and Tate, 2012). Other studies have focused on board capabilities, diversity, the proportion of inside directors, CEO duality or whether the former CEO has stayed on the board as indicators of how board–CEO collaboration impacts company performance (Boivie et al., 2025; Quigley and Hambrick, 2012; Zhu et al., 2016). Controls have included the succession type, the extent of managerial discretion available to CEOs and prior CEO experience, including as a director on other boards (Boyd et al., 2011; Tuggle et al., 2010), and large data samples.

These studies have shown that while board–CEO collaborations can enhance strategic decision-making (Tian et al., 2011; Westphal, 1999) and board–CEO ties are associated with lower levels of fraud (Chidambaran et al., 2010), other director-CEO political, social or friendship ties are associated with reduced company performance (Hwang and Kim, 2009; Khanna et al., 2015). Therefore, it is possible that professional relationships are similarly collusive, particularly in common-law countries. In other words, the appointment of known CEOs, in a “boys’ club” type scenario may reduce company performance.

However, in the context of external CEO succession, the reasoning that directors use narrow, professional relationships to screen and identity more appropriate prospective hires is more likely. This is because, in practice, directors are known to refer target outsider CEO candidates to board nominating committees, in many instances through executive search companies that limit the scope for collusion (Khurana, 2002). Moreover, hiring through referrals in general is known to reduce information asymmetry, resulting in positive hiring outcomes (Burks et al., 2015). Consequently, if these board–CEO ties are considered a form of proxy referral – given the opportunity they provide for employing companies to fill information gaps from trusted sources, while simultaneously representing a lower level of risk associated with personal affiliation that comes from friendship, social or other forms of relationships – then it may be expected that they will prove advantageous to the company. Thus, the following hypothesis can be established.

While it is hypothesized that board–CEO ties enable boards to reduce information asymmetry in the appointment of new outsider CEOs, these relationships exist within contextualized institutional settings that affect the information asymmetry between investors and boards and therefore that of the CEO labor market. This second information asymmetry that affects outsider CEO successions focuses on how corporate governance and institutional transparency impacts issues of moral hazard and the way in which companies act in their own interests rather than their investors’ interests (Fama and Jensen, 1983b; Jensen et al., 1976).

In an international context, investors face the challenge of navigating different institutional structures that affect corporate governance. National institutional structures vary around the world because of their basis in national behavioral, ethical, social, political and economic norms and circumstances (Aggarwal and Goodell, 2014). They are important because they impact the costliness of information needed for the measurement and enforcement of exchanges (North, 1990). In doing so, they can lower or raise transaction costs in market exchange (including that for CEOs) through their ability to inform participants about the nature of contracts and their enforceability, as well as the temperament and motivations of other market participants (Aggarwal and Goodell, 2009b; Hart, 1995, 2001).

Specifically, differences between German civil law and English common law have led to divergent approaches to the development of capital markets, where banks dominate financial intermediation in continental Europe, with its German civil law underpinnings, while equity investors play a more prominent role in Commonwealth countries, founded on English common law (Modigliani and Perotti, 2000; Tirole, 2006). In civil law countries, a more literal interpretation of contract language by the courts increases the risk of an unfair verdict and makes the writing of one-time bilateral contracts problematic (Ergungor, 2004). To reduce this risk, banks have emerged in these countries as institutions that can resolve conflicts and enforce contracts without court intervention through their ability to collect fixed-claim returns from financing activity (Aggarwal and Goodell, 2009a). Banks, by nature of their focus on fixed claims as collateral for participation in financial intermediation, prefer less disclosure and transparency to protect their claims. This focus on fixed claims as collateral also means banks are less able to distinguish between low- and high-quality companies, which, in turn, incentivizes banks to reduce transparency to protect their relationships with weaker companies (von Thadden and Perotti, 2001). In a reinforcing cycle, this reduced transparency further shields weaker companies because competitors may misjudge expectations of other competitors’ output and act more aggressively than they otherwise would. Under strategic competition, greater transparency and disclosure requirements augments price transparency and higher variability of profits and output. Conversely, a lack of transparency can reduce strategic competition, leading to lower variability in profits and output as well as lower expected profits and volatility. Reduced volatility increases the return to holders of fixed claims on the company (Perotti and von Thadden, 2003). This reasoning implies that there is a natural preference by lenders for less information dissemination as they do not gain from higher profits but suffer higher risk (Jensen et al., 1976). In less transparent institutional environments where banks control superior information on clients’ credit quality, they can also charge higher interest rates, or information rents (Rajan, 1992). Therefore, national institutional environments such as in France, Germany, Italy and Spain as well as other European countries that are based on civil law – which in turn favors bank-financing and possess a unique communitarian business system, emphasizing the role of government in economic and social affairs – have strong institutional incentives for less governance transparency (Millar et al., 2005).

By contrast, the flexibility and discretion that English common law allows judges in interpreting the law, enable common law courts to cast a wider net around the law to penalize opportunistic behavior more effectively. Consequently, common law jurisprudence facilitates the enforcement of incomplete contracts with greater nuance which, in turn, enables the development of both stock markets and banks to act as financial intermediaries because of the greater legal protections for equity investors (Ergungor, 2004; Modigliani et al., 2000).

The greater proportion of equity investors in financial markets-led financial systems encourages greater institutional governance transparency as a means to lower the transaction costs of obtaining reliable information about corporations. These information costs are shared with the suppliers of equity, adding to its expense where, by definition, it is already riskier than bank lending because equity suppliers hold residual and not fixed claims on corporations. Consequently, equity investors have a stronger incentive to secure detailed, reliable information on companies to participate in financial intermediation. In this context, greater transparency of transaction costs lowers the cost of equity-focused market financing by lowering the costs of resolving asymmetric information (Aggarwal et al., 2009a; Aggarwal and Goodell, 2010; Ergungor, 2004; Kwok and Tadesse, 2006). As a result, the wider participation of financial markets in financial intermediation in common law countries encourages greater institutional governance transparency to reduce asymmetric information and transaction costs (Bertaut, 1998; Hail and Leuz, 2006). This is born out in objective measures of governance transparency where the major common law countries, the United States, United Kingdom, Canada and Australia report higher overall averages of governance transparency than the major civil law countries in Europe including France, Germany, Spain, Italy and Portugal across a range of measures of ownership transparency and anti-self-dealing (Aggarwal et al., 2015; Bushman et al., 2004; Djankov et al., 2008).

For the CEO labor market specifically, greater governance transparency means more market disclosures about the identity of CEOs and directors, as well as their affiliations. Where there is greater governance transparency, CEO labor markets participants, such as directors and other labor market intermediaries, have access to more information on prospective outsider CEOs. However, in institutional environments where there are fewer disclosures about the identity of CEOs and directors, the residual level of information asymmetry in the CEO labor market is greater because of the lack of corporate leadership and governance market disclosure. In this context, directors may be able use their relationships with prospective CEOs to reduce this information asymmetry. Therefore, in less transparent institutional environments, board–CEO ties could lead to an observable effect where outsider connected CEOs outperform non-connected CEOs.

By contrast, outsider connected CEOs may underperform non-connected CEOs in the opposite case, where there is less residual information asymmetry. For those countries associated with indicators of greater governance transparency, the characteristics and track records of incoming CEOs have already been disclosed to the labor market such that any potential value of information transmitted through board–CEO ties is negated. Consequently, in institutional environments with greater average levels of governance transparency, it is expected that no observable effect can be determined, and outsider connected CEOs will not be observed to outperform outsider non-connected CEOs.

We test these hypotheses using CARs, a market-based measure that reflects investors’ reactions to the actions taken by CEOs. We expect investors to react favorably to newly appointed outsider connected CEOs because of the informational advantage that they provide. We also expect that the extent of these reactions will be moderated by the degree of informational advantage that they provide, as reflected by the level of institutional governance transparency in unique national institutional environments.

For robustness and to further test the effect of connected CEOs on company financial performance, we also include three accounting-based measures of financial performance, ROA, ROIC and ROS. These directly reflect CEOs’ actions but also reflect an institutional viewpoint to CEO performance rather than those of investors. As such, we do not anticipate that connected CEOs will deliver uniformly greater company profitability than non-connected CEOs by these measures and that overall results may, in fact, be negative. This expectation is driven by several factors that uniquely underpin the institutional viewpoint. First, outsider CEOs are typically appointed to enact change (Lorsch and MacIver, 1989; Vancil, 1987) and the negative effects of corporate restructuring on net income and assets, which bear on accounting-based ratios such as ROA, ROIC and ROS, can drive lower overall profitability. Second, connected CEOs may be emboldened by their board relationships to pursue change more assertively placing further additional downward pressure on these accounting measures of profitability. Further, if the confounding effects of change are expected to result in lower accounting-based measures of profitability then the moderating effect of institutional governance transparency is less likely to be visible through these measures. Consequently, a third hypothesis is:

Data are sourced from several financial databases including Bloomberg, Compustat Global (accessed through the Wharton Research Data Services platform), Datastream and BoardEx. The Bloomberg equity screen function is used to identify 1,855 outsider CEOs appointed between 1992 and 2018 across 22 counties: Australia, Belgium, Brazil, Canada, China (including Hong Kong), Czech Republic, Denmark, Finland, France, Germany, India, Italy, Japan, the Netherlands, Norway, Portugal, South Africa, Spain, Sweden, Switzerland, the United Kingdom and the United States. Bloomberg searches focused on publicly-owned corporations listed on recognized securities exchanges. These are the Australian Securities Exchange, New York Stock Exchange, NASDAQ exchange, Toronto Stock Exchange, London Stock Exchange, Euronext NV, Frankfurt Stock Exchange, Helsinki Stock Exchange, Stockholm Stock Exchange, Borsa Italiana, National Stock Exchange of India, Shanghai Stock Exchange, Stock Exchange of Hong Kong, Tokyo Stock Exchange, Johannesburg Stock Exchange and the Brasil Bolsa Balcao S.A. stock exchange.

Information regarding company size and industry sector is collated, and based on market capitalization (US dollars) at the time of data collection along with the relevant Morgan Stanley Capital International (MSCI), Global Industry Classification Standard (GICS) sector to which a company belongs. The MSCI GICS framework incorporates 11 industry sectors spanning Communication Services, Consumer Discretionary, Consumer Staples, Energy, Financials, Health Care, Industrials, Information Technology, Materials, Real Estate and Utilities.

Biographical and résumé data for CEOs and directors are drawn from BoardEx. Using the BoardEx Connections and Matching functions, overlapping tenure at prior employers between the newly appointed, outsider CEOs and company directors is identified. Where a prior professional connection exists between the new CEO and an individual board director, the CEO is marked as a connected CEO. Only prior professional, executive and non-executive connections, where a CEO and director overlapped at the same company, are considered relevant. Overlapping, non-executive directorships at other corporations, a common feature within elite labor markets (Westphal, 1999) are included because they are professional in nature but director–CEO social relationships and other non-work-related connections, for example, studying at the same university during at least some of the same years, are not considered a relevant professional connection and so are not considered in the definition of a professional relationship.

Data on the financial performance of companies in the sample along with industry performance benchmarks is also collected from Bloomberg. Compustat Global and Datastream are used as ancillary databases to source company financial performance data for smaller companies where Bloomberg recorded missing data. CARS, ROA, ROIC and ROS data are gathered for each company over a 30-year period, spanning 1990 to 2020. Monthly cumulative GICS sector total returns are also collected for each of the 11 nominated GICS sectors across each country or region represented in the sample. In developed markets (including Australia, Canada, Europe and the United States) monthly total returns data are available for each of the 11 GICS sectors. In developing markets including Brazil, China (including Hong Kong), India and South Africa, MSCI provides singular individual country indices rather than an in-country range of the 11 GICS sector indices articulated above. In these cases, the singular country GICS indices are used to collect benchmark monthly total returns data. MSCI GICS also conflates individual European countries into a Europe-wide series of industry benchmarks and these data were utilized for the performance analysis of all European companies in the sample. Focal company, monthly CARs are then calculated as the individual focal company’s cumulative monthly total return, including the reinvestment of all dividends, minus the relevant country and GICS sector benchmark total return for the same monthly time period.

Based on the availability of relevant pre- and post-succession financial performance data, the final sample settles at 1,136 companies. All CEO successions occur between 1992 and 2018. Of these 1,136 companies, 500 are led by connected CEOs (44%) and 636 by non-connected CEOs (56%). 606 are headquartered in the United States, 146 in the United Kingdom, 95 in Australia, 83 in India, 47 in France, 44 in China (including Hong Kong), 38 in Canada, 23 in both Italy and South Africa, 18 in Brazil and Switzerland, 14 in Sweden, 9 in Finland and Spain, 8 in Germany, 7 in Denmark, 6 in Japan, 5 in Norway, 4 in the Netherlands, 3 in Belgium and 2 in the Czech Republic and Portugal.

The empirical model developed to evaluate the company financial performance delivered by connected or non-connected CEOs for each of the continuous CARs, ROA, ROIC and ROS dependent variables is presented below. This model includes a categorical connected CEO independent variable and a series of continuous and categorical control variables including Industry Conditions, Company Health, Company Size and Industry as well as Country (C) and Year (Y) fixed effects.

Y_it represents the dependent variable of interest for company i in year t. For the accounting-based dependent variables, ROA, ROIC and ROS, Y_it includes the subtraction of the relevant Industry Conditions variable from the reported accounting measure to improve the empirical model’s precision. The coefficient of interest is β₁ connected CEO.

The key independent variable is whether the incoming, externally appointed, CEO has previously worked with at least one member of the board at another company earlier in their career. The independent variable is a dichotomous variable, connected CEO, equal to zero if the CEO is non-connected, and one otherwise.

The market-based measure, CARs, captures value creation for investors through capital gains and dividends (Fan et al., 2007). CARs are calculated as the focal company’s cumulative monthly total shareholder return, including the reinvestment of all dividends, minus the relevant country and GICS sector benchmark total shareholder returns for the same monthly time period. For those developing country GICS indices that do not include specific GICS sector indices, the country GICS index benchmark is used.

ROIC is calculated as the focal company’s net income minus dividend, divided by the sum of total debt and equity. ROA is calculated as the focal company’s 12-month trailing net income divided by average total assets. ROS is calculated as the focal company’s earnings before interest and tax divided by net annual sales revenue. These accounting-based measure are commonly used because they focus on operational performance (Hambrick and Finkelstein, 1995; Shen et al., 2002; Yoo et al., 2009).

Drawing on the established CEO in context (Hambrick and Quigley, 2014) approach, control variables include categorical Year (capturing macroeconomic effects), Company Size, Country and Industry variables. Continuous measures are Industry Conditions and Company Health.

Company Size is based on market capitalization at the time of data collection. Categories of Company Size are nano-cap (<USD50m), micro-cap (USD50m–USD300m), small-cap (USD300m–USD2b), mid-cap (USD2b–USD10b), large-cap (USD10b–USD200b) and mega-cap (>USD200b). Country reflects the country where the focal company is incorporated.

Industry is a vector that includes a categorical variable identifying the particular MSCI GICS sector to which the focal company belongs. The vector industry also includes a continuous control measure that is included to isolate any variation in company performance that may be as a result of the economic conditions present in a particular industry during the measurement period (Industry Conditions). For example, financial companies’ profitability suffered during and after the 2008 financial crisis, which is included in the overall 1990–2020 span of measurement years. Industry Conditions is calculated as the median CARs, ROA, ROIC and ROS for all companies in the sample that are in the same GICS sector as the focal company. This measure represents an annual return percentage value and is calculated for each of the first three years of the new CEO’s tenure. Given the large sample, the focal company’s outcome is not excluded from the calculation of this industry control variable.

Company Health is included to further isolate the CEO’s impact through capturing any variation in company performance that is attributable to the general state of affairs present in the companies to which the new outsider CEOs are appointed; that is, how successful, or otherwise, were the companies immediately prior to the new CEO being appointed. Company Health is calculated as the focal company’s market-to-book (MTB) value divided by the relevant GICS sector median MTB at the close of the fiscal year prior to each CEO’s start. For example, if a company’s pre-succession MTB is 4.0 and the relevant GICS sector median MTB is 2.0, the Company Health indicator for the focal company would be 2.0 (i.e. 4.0/2.0).

The CEO in Context specification of the variance partitioning analysis method is adopted (Bertrand and Schoar, 2003; Hambrick et al., 2014; Lieberson and O'Connor, 1972; Mackey, 2008; Quigley et al., 2012). Data are organized in a panel with monthly values spanning January 1990 to December 2020. The CEO performance measurement period consists of years one, two and three from the first full month from the newly appointed CEO’s start date, as up to three years post-succession is considered an appropriate measure of short-term CEO performance (Karaevli, 2007; Zhang, 2008). A Variance Inflation Factor test revealed no evidence of multicollinearity between variables. Hausman and Breusch-Pagan Lagrange Multiplier tests indicated that a random effects regression model is appropriate rather than fixed effects or ordinary least squared, using the xtreg function in Stata. Descriptive statistics and correlations are provided in Table 1.

CARs results (refer Tables 2 and 3) for the Pooled sample are inconclusive and are not necessarily consistent with H1 and the central argument that board–CEO ties confer a screening advantage. Over an initial 12-month period connected CEOs outperform by delivering greater CARs (top left cell in Table 2: ß = 0.75, p = 0.00) than non-connected CEOs (ß = 0.73, p = 0.00) but this result is marginal and may not reflect a significant difference. Moreover, over a 36-month period, non-connected CEOs deliver greater CARs (ß = 0.89, p = 0.00) than connected CEOs (top left cell in Table 3: ß = 0.80, p = 0.00), a result that indicates board–CEO ties may only confer a marginal short-term advantage, if at all.

When considering the degree of transparency in the CEO labor market, a range of different results emerge. In the United States and Commonwealth countries (Australia, Canada and the United Kingdom), connected CEOs do not outperform non-connected CEOs as measured by CARs over 12- (ß = 0.79, p = 0.00 for connected vs ß = 0.96, p = 0.00 for non-connected CEOs) or 36-month (ß = 0.87, p = 0.00 for connected vs ß = 1.06, p = 0.00 for non-connected CEOs) periods, a result that is consistent with H2a.

For those countries in continental Europe, where there are typically fewer market disclosure about CEOs and directors, connected CEOs outperform non-connected CEOs by CARs over both 12- (ß = 1.01, p = 0.00 for connected CEOs vs ß = −0.26, p = 0.54 for non-connected CEOs) and 36-month (ß = 0.67, p = 0.05 for connected CEOs vs ß = 0.38, p = 0.30 for non-connected CEOs) periods, a result that is consistent with both H1 and H2a.

Similarly, in Asia, where large family-controlled conglomerates are common and where there are also fewer market disclosures about CEOs and directors, connected CEOs create greater value for shareholders through larger CARs over the longer term, 36-month measurement period (ß = 1.00, p = 0.10 for connected CEOs vs ß = 0.61, p = 0.36 for Non-connected CEOs), a result that is consistent with H1 and H2a. In the short-term, non-connected CEOs perform better over 12-months (ß = 0.45, p = 0.54 for connected CEOs vs ß = 1.54, p = 0.10 for non-connected CEOs). In Brazil, India and South Africa (Rest of the World), emerging markets countries that are open to international investors, neither connected CEOs or non-connected CEOs deliver meaningfully different market-based outcomes over initial 12- (ß = 0.22, p = 0.71 for connected CEOs vs ß = −0.21, p = 0.65 for non-connected CEOs) or 36-month (ß = 0.20, p = 0.66 for connected CEOs vs ß = 0.13, p = 0.74 for non-connected CEOs) periods. These results do not support H1 or H2a.

For the accounting measures, ROA, ROIC and ROS, connected CEOs underperform non-connected CEOs in the overall pooled sample across both 12- (ß = −0.23, p = 0.13; ß = −0.61, p = 0.00; ß = −2.02, p = 0.01 for connected CEOs, ROA, ROIC and ROS respectively vs ß = −0.23, p = 0.15; ß = 0.12, p = 0.52; ß = −0.78, p = 0.29 for non-connected CEOs) and 36-month (ß = −0.94, p = 0.00; ß = −1.01, p = 0.00; ß = −5.25, p = 0.00 for connected CEOs, ROA, ROIC and ROS respectively vs ß = −0.30, p = 0.00; ß = −0.60, p = 0.00; ß = −3.88, p = 0.00 for non-connected CEOs) periods. These results are consistent with H3.

In the United States and Commonwealth countries, connected CEOs also underperform non-connected CEOs (ß = −0.43, p = 0.03; ß = −1.02, p = 0.00; ß = −3.41, p = 0.00 for connected CEOs, ROA, ROIC and ROS respectively vs ß = 0.26, p = 0.15; ß = 0.37, p = 0.10; ß = 0.56, p = 0.55 for non-connected CEOs), an outcome that is consistent with H3. Across continental Europe, results show that connected CEOs deliver positive and superior results over a 12-month period (ß = 0.30, p = 0.05; ß = 0.70, p = 0.00; ß = −0.29, p = 0.42 for connected CEOs, ROA, ROIC and ROS respectively vs ß = −0.72, p = 0.00; ß = −1.12, p = 0.00; ß = −2.25, p = 0.00 for non-connected CEOs). These results are consistent with the CARs outcomes for Europe but not H3.

In Asia, results for accounting-based measures over both the initial 12-month (ß = 0.13, p = 0.76; ß = 1.08, p = 0.07; ß = 6.61, p = 0.10 for connected CEOs ROA, ROIC and ROS respectively vs ß = 0.46, p = 0.38; ß = −4.69, p = 0.00; ß = −57.10, p = 0.00 for non-connected CEOs) and longer term 36-month (ß = −1.28, p = 0.00; ß = −1.29, p = 0.01; ß = 4.01, p = 0.20 for connected CEOs, ROS, ROIC and ROA respectively vs ß = −47.50, p = 00; ß = −4.18, p = 0.00; ß = 1.43, p = 0.01 for non-connected CEOs) period are negative and consistent with H3.

For Brazil, India and South Africa (Rest of the World) connected CEOs do underperform initially over 12-months (ß = −0.95, p = 0.00; ß = −0.58, p = 0.05; ß = −1.45, p = 0.05 for connected CEOs ROA, ROIC and ROS respectively vs ß = −0.65, p = 0.00; ß = −0.76, p = 0.00; ß = −1.27, p = 0.04 for non-connected CEOs) as indicated by the accounting-based measures, a result that is consistent with H3. Over 36-months (ß = 0.77, p = 0.00; ß = 0.01, p = 0.97; ß = −1.68, p = 0.09 for connected CEOs, ROA, ROIC and ROS respectively vs ß = −0.14, p = 0.48; ß = −0.07, p = 0.76; ß = −1.81, p = 0.02 for non-connected CEOs), there are very few observable differences between connected CEOs and non-connected CEOs, an outcome that is also consistent with H3.

The aim of this study was to explore what effect, if any, prior board–CEO ties have on the selection of a new outsider CEO as assessed by associated company performance. It was thought that these relationships might confer a screening advantage because of their professional nature and because in practice directors refer known candidates to executive search companies and board nominating committees in external CEO successions. It was also thought that there may be international differences to the results because of varying rates of market disclosures by corporations about the identity of CEOs and directors. We believed that these differences in the levels of governance transparency would moderate the effects of board–CEO ties and, in combination with different approaches to corporate governance in unique national jurisdictional environments, would also create boundary conditions that affect the role of board–CEO ties.

The study’s results are broadly consistent with these aims and expectations, although the outcomes in the United States, United Kingdom, Canada and Australia, where board–CEO ties are shown to have less value, was somewhat surprising. In achieving these results, this study makes three important contributions. First, we show that board–CEO ties matter in outsider CEO successions. Second, we show that a country’s legal origins and institutional settings, which affect CEO labor market transparency, affect how professional board–CEO ties function. Finally, we highlight the importance of considering both the market-based and accounting measures of company performance.

A substantive, United States/United Kingdom, board–CEO ties literature exists that spans a range of theoretical perspectives from agency, resource dependency, upper echelons, stewardship and institutional to social network theory (Boyd et al., 2011 provides a comprehensive review). This literature includes evidence where a collaborative board approach can enhance strategic decision-making (Adams and Ferrerira, 2007; Balsmeier et al., 2013; Chidambaran et al., 2010; Hoitash and Mkrtchyan, 2018, 2022; Tian et al., 2011; Westphal, 1999; Wiersema et al., 2018; Zhu et al., 2016) and, conversely, where an independent board is important to monitor management appropriately (Fan et al., 2007; Fan et al., 2019; Fracassi et al., 2012; Hwang et al., 2009; Khanna et al., 2015). It has not conclusively explored the role of professional board–CEO ties in outsider CEO successions (Cziraki and Jenter, 2021; Khurana, 2002; Wang, 2021).

Our study contributes to this literature by examining company performance outcomes associated with board–CEO ties in outsider CEO successions. Our overall results, also based on a sample that is heavily weighted to the common law-founded model of corporate governance, indicate that relationships ties are associated with a slight increase in value creation to investors (greater CARs) while also being associated with more pronounced lowered profitability as determined by accounting-based measures of financial performance. While prima facie these accounting-based results are not consistent with our screening advantage argument, they are consistent with our expectation that accounting-based measures would mask any effects of board–CEO ties and not necessarily reflect profitable returns to investors because of the effects of CEO-led change on the constitution of these financial ratios (Aguirre et al., 2016; Hambrick et al., 1984; Nickisch, 2016; Schepker et al., 2017).

As such, we extend Wang’s (2021) United States’ results, through our significantly more nuanced international findings, that board–CEO ties reduce information asymmetry in outsider CEO successions and are associated with positive company financial outcomes. Our results are also consistent with the broader board–CEO ties literature that suggests a collaborative board approach can enhance strategic decision-making. In particular, we extend Khurana’s (2002) pioneering findings that directors play an integral role in CEO successions through referring target CEO candidates to executive search companies, by showing the effect that a narrower strand of these types of referrals (past professional relationships) has on a range of different company financial outcomes specifically in outsider CEO successions.

A second contribution that our results make is to the literature addressing boundary conditions to information asymmetry that are imposed by national, cultural and institutional differences across a range of international jurisdictions. While there is an extensive literature that explores the effects of institutional and governance transparency in moderating the impact of information asymmetry on transaction costs in market exchange (Aggarwal et al., 2015; North, 1990; Williamson, 1975), there has been very little attention paid to its effect on transaction costs and exchange in the CEO labor market (Cziraki et al., 2021) or on the broader aspect of board–CEO ties in an international context outside of the United States/United Kingdom environments (Boyd et al., 2011). Our study is the first comparative analysis that we are aware of that takes into account the effect of board–CEO ties in a global setting where there are significant differences in CEO labor market transparency and approaches to corporate governance.

Where lower levels of governance transparency add to the opacity of the CEO labor market, such as in continental Europe, board–CEO ties are associated with superior market-based measures of financial performance. Unexpectedly, European connected CEOs are also associated with superior accounting-based measures of financial performance. These results are consistent with arguments that board–CEO ties, in outsider CEO successions, support cooperation rather than monitoring. It has been conjectured that a cooperative relationship between board and management is more likely to be observed outside of the more oppositional common law-founded model of corporate governance (Doidge et al., 2007; Eisenhardt, 1989; Forbes and Hodgkinson, 2015; Lubatkin et al., 2016; Young et al., 2008). Our European and Asian results are consistent with this line of reasoning.

Our United States and Commonwealth CARs results are consistent with our argument that CEO labor market transparency moderates the effects of board–CEO ties. In the United States and those Commonwealth national institutional environments that are associated with greater CEO labor market transparency, board–CEO ties, flip to being collusive. That is, they are associated with less shareholder value creation than their non-connected peers. This may be in part a reflection of both the greater amount of information on CEOs that is publicly available as well as the more pronounced adversarial common law-founded model of corporate governance. By contrast, accounting-based measures show that connected CEOs are associated with lower levels of profitability, a result that could be attributed to a range of factors such as more assertive corporate restructuring.

Interestingly, our regional results in the emerging markets suggest that principal-agent issues are also a factor, in particular in those Asian countries that are more open to international investors (i.e. Hong Kong and Japan) possibly attributable to the globalization of common law-founded governance (Dignam and Galanis, 2009). Whereas in those jurisdictions where family-controlled conglomerates are common but there is less international investment (i.e. Brazil, India and South Africa), board–CEO ties do not lead to observable collusive effects. This outcome is consistent with the reasoning that principal-principal rather than principal-agent problems are a more common feature in family-dominated emerging markets economic environments (Young et al., 2008).

Our results also have implications for the literature that suggests hiring through referrals can reduce information asymmetry in labor markets, such as that for CEOs, that are prone to adverse selection (Aurelie et al., 2021; Cziraki et al., 2021; Ekinci, 2016). Specifically, we demonstrate that any realized value of hiring new outsider CEOs through board referrals is conditional on the latent degree of information asymmetry that exists in the CEO labor market. As such, these results also extend the economic and institutional transparency literature’s understanding of the role of macro-institutional settings in affecting the functioning of the CEO labor market.

Our study includes several limitations. First, we do not assess the strength of board–CEO ties or discriminate board interlocks from prior executive relationships. Thus, we are unable to contribute to the literature on strong or weak connections (Granovetter, 1973). Second, we do not incorporate qualitative information on CEOs and so cannot add to the upper echelons discussion (Hambrick, 2007; Jalbert et al., 2010). Third, by focusing purely on financial performance, we examine only a very narrow strand of CEO performance. While public company CEOs are primarily incentivized to grow earnings, they are also assessed against leadership, strategy, people management, operating metrics and relationships with external constituencies (Kaufman, 2008) which we do not examine. Fourth, it is likely that a new CEO is appointed at particularly critical times and the choice of a connected CEO is appointed at particularly crucial times. Consequently, the choice of a connected or non-connected CEO may be related to unobserved determinants of company performance. To address this issue, an instrumental variable would be necessary but based on the available data we have been unable to find a sufficiently exogenous indicator to apply a two-stages least squares approach. Notwithstanding this limitation, it is likely that connectivity is positively related to such unobserved determinants (for example, the possible disinterest of non-connected prospective CEOs given the dire financial prospects of the company), which attenuate the coefficient of interest. Hence, non-zero coefficients of connectivity are likely lower bounds of the true estimates.

Future research may include separating professional connections into past executive and non-executive, interlocking board relationships. Knowledge about a CEO’s past executive experience (rather than board experience) may provide a clearer reflection of their likely performance as CEO. Future research may also continue to focus on international data where there remains limited available evidence.

Our results, that shed light on the outcomes of professional board–CEO ties in outsider CEO successions, may be of interest to investors and a range of other stakeholders. Directors have long been known to refer prospective CEO candidates to executive search companies and board nominating committees (Cziraki and Jenter, 2021; Khurana, 2002). We show that these referrals are material and that investors react favorably. We also show that the institutional, rather than market, viewpoint as represented by accounting-based return measures provides a less clear picture on these ties’ effects. Furthermore, we show that the context provided by the degree of CEO labor market transparency is important. Where there is greater CEO labor market transparency, these relationships add little to no additional shareholder value. By contrast, in less-transparent CEO labor markets, board–CEO ties provide a clear advantage to companies and their investors.