The dark side of intangibles? Organizational capital and corporate investment efficiency Mohamed Shaker Ahmed Cairo University, Giza, Egypt, and Timothy King School of Accounting and Finance, University of Vaasa, Vaasa, Finland Abstract Purpose –Organizational capital (OK) represents an important intangible productive firm asset, yet one subject to agency problems. This paper provides a first examination of how OK impacts corporate investment inefficiency using an unbalanced panel of listed US companies from 2009 to 2020. Design/methodology/approach –We utilize fixed-effect regressions to explore the relationship between OK and investment efficiency. Additionally, we implement a battery of robustness tests of the main study findings based on a variety of panel data techniques, including firm fixed effects, alternative measures of investment efficiency, estimators, including Newey–West as well as additional steps to address endogeneity concerns using system-GMM, two-stage least-squares and entropy balancing analyses. Findings –OK is associated with reduced investment efficiency (underinvestment and overinvestment). A one- standard deviation increase in OK to total assets is associated with a 4.42% decrease in investment efficiency. Based on average firm investment, this represents a $390.88 million decrease in investment. CEO gender and career concerns as well as R&D intensity, positively moderate this relationship, while CEOs’ age, power, tenure and connections as well as corporate governance and disclosure quality, negatively moderate it. The findings can be understood from the perspective of agency theory, whereby informational asymmetries surrounding OK make it challenging for firm outsiders to monitor and evaluate managerial investment choices. Originality/value – We lack understanding as to how it impacts the efficiency of corporate investment. We contribute the first evidence in this regard by demonstrating that OK is associated with lower investment efficiency. Keywords Investment efficiency, Organizational capital, Capital allocation, Intangible assets, CEO characteristics, Agency theory Paper type Research paper 1. Introduction Capital investment decisions are among the most important decisions made within firms that have a direct influence on firm value. The efficient market hypothesis presumes that investment and financing decisions are made independently and that firms pursue projects that maximize shareholder value (Modigliani and Miller, 1958). However, in the presence of capital market frictions, such as adverse selection, moral hazard, and information asymmetries (Biddle and Hilary, 2006; Anagnostopoulou et al., 2023), firms can deviate significantly from optimal investment levels due to conflicts of interest between corporate managers and outside shareholders as well as financial constraints, as outlined by agency theory (Jensen and Meckling, 1976). The importance of corporate investment for corporate success has inspired a growing literature that seeks to understand the firm-level factors that influence corporate investment efficiency. This literature finds that factors including the quality of accounting information and JAL 47,5 444 JEL Classification — E22, G30 © Mohamed Shaker Ahmed and Timothy King. Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non- commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode The current issue and full text archive of this journal is available on Emerald Insight at: https://www.emerald.com/insight/0737-4607.htm Received 11 June 2024 Revised 23 January 2025 Accepted 19 April 2025 Journal of Accounting Literature Vol. 47 No. 5, 2025 pp. 444-489 Emerald Publishing Limited e-ISSN: 2452-1469 p-ISSN: 0737-4607 DOI 10.1108/JAL-06-2024-0120 http://creativecommons.org/licences/by/4.0/legalcode https://doi.org/10.1108/JAL-06-2024-0120 reporting (Biddle and Hilary, 2006; Biddle et al., 2009; Chen et al., 2011; Gomariz and Ballesta, 2014; Barth et al., 2017), level of employee education (Jin et al., 2023), accounting conservatism (Lara et al., 2016; Laux and Ray, 2020), choice of accounting standards (McClure and Zakolyukina, 2024), earnings management (McNichols and Stubben, 2008), managerial quality (Chemmanur et al., 2009; Garc�ıa-S�anchez and Garc�ıa-Meca, 2018), cash holdings (Richardson, 2006), firm ownership (Richardson, 2006; Chen et al., 2017), can influence corporate investment efficiency. Although existing studies are informative, we lack understanding as to how organizational capital (OK), a key intangible asset [1], influences corporate investment efficiency. OK involves the accumulation of technological knowledge, experience, processes, designs, and business practices, which facilitates unique matching of labor and physical production resources, resulting in more efficient use of firm resources (Li et al., 2018; Eisfeldt and Papanikolaou, 2013). Given that OK stems directly from the input of senior managers (Carlin et al., 2012; Eisfeldt and Papanikolaou, 2013, 2014; Li et al., 2018; Attig and El Ghoul, 2018; Boguth et al., 2022), it has formerly been defined as “a production factor that is embodied in the firm’s key talent and has an efficiency that is firm-specific, [and one where] both shareholders and key talent have a claim to its cash flows” (Eisfeldt and Papanikolaou, 2013, p. 1). Managers both choose to make investments in OK as well as influence its value to the firm (e.g. Carlin et al., 2012; Peters and Taylor, 2017; Eisfeldt and Papanikolaou, 2013, 2014; Boguth et al., 2022). Therefore, OK can potentially represent a source of sustainable competitive advantage for firms (Lev et al., 2009). Yet how OK should influence corporate investment efficiency is theoretically ambiguous. On one hand, investments in OK could be associated with reduced investment efficiency, consistent with an agency-based interpretation (Myers and Majluf, 1984). Firms with high OK are understood to be risky (Boguth et al., 2022) and intangible assets are thought to exacerbate information asymmetries, which calls for enhanced disclosures of relevant and accurate information by firm executives to mitigate such information problems (Biondi and Reb�erioux, 2012). Thus, heightened information asymmetries and uncertainty surrounding managerial investment choices, mean that shareholders fail to recognize or fully value the future benefits of intangible investments (Eisfeldt and Papanikolaou, 2013, 2014). Such informational asymmetries are compounded due to issues with disclosure, reporting, and general measurement difficulties associated with intangible assets in general (Barth et al., 2001) and OK in particular (Eisfeldt and Papanikolaou, 2013, 2014). For example, intangible assets are not normally recognized on firms’ balance sheets because of the difficulties in determining their value and the uncertain nature of future payoffs related to their use. Thus, the information asymmetries associated with OK faced by corporate boards, shareholders, and other firm outsiders could result in, for example, managerial performance being evaluated more extensively based on the outcomes of their project choices in firms characterized by high OK. Consequently, managers may invest suboptimally when OK is high from the perspective of shareholders because of exacerbated career concerns (Goel et al., 2004; Kim et al., 2021). On the other hand, OK could be associated with better corporate investment efficiency. Notably, investments in OK have been shown to facilitate greater quality and extent of informational disclosures by managers. In turn, this reduces information asymmetries surrounding managerial investment choices and could result in more efficient investment as senior managers face, ceteris paribus, lower career concerns (Goel et al., 2004; Chemmanur et al., 2009; Attig and El Ghoul, 2018; Garc�ıa-S�anchez and Garc�ıa-Meca, 2018) consistent with an agency theory perspective. Moreover, managers in high-OK firms may also invest more efficiently because of both better matching of managerial capital with OK and because of superior management ability; notably, investments in OK can improve the quality of matching between labor and physical production resources (Lev et al., 2009). In this way, OK can constitute a valuable and unique resource that can provide a source of sustainable competitive advantage for firms (Carlin et al., 2012; Hasan and Cheung, 2018). Consistent with these arguments Chemmanur et al. (2009) show that firms with better managers tend to choose Journal of Accounting Literature 445 superior investment projects at any scale, resulting in both increased capital expenditures and investment in research and development (R&D), while Boguth et al. (2022, p. 857), who consider OK as “a firm-specific production factor provided by key employees,” find that while firms with high OK levels are risky, they are associated with higher future stock returns. In this paper, we draw on agency theory to question whether OK can impact the efficiency of corporate investments. The underlying notion is that since OK can influence the extent of informational asymmetries in firms between insiders and outsiders, this could also impact managerial investment choices. We set our study in the context of the US, which, as we explain in detail in Section 2, has experienced significant growth in the use of intangible assets within firms and changes in investment efficiency over time driven. For example, the Sarbanes-Oxley Act of 2002, especially Section 404, introduced stricter requirements for financial reporting and internal controls, requiring management and external auditors to assess their effectiveness. These measures enhanced corporate governance and financial transparency, indirectly fostering greater investment efficiency by reducing managerial opportunism (Aubert and Grudnitski, 2014). We construct an unbalanced panel of 486 listed US firms, totaling 5,819 firm-year observations for the period 2009 to 2020, to provide, to the best of our knowledge, the first empirical examination as to how OK impacts corporate investment efficiency. We employ panel regression models that utilize fixed effects at the firm and year levels, where OK is measured using the perpetual inventory method of Peters and Taylor (2017) and corporate investment is modeled based on a firm’s growth opportunities (Biddle et al., 2009). To mitigate potential concerns associated with our OK measure, we employ two measures of OK throughout our analyses (the ratio of OK to total assets and the ratio of OK to total capital). By way of preview, our main results document a positive (negative) and significant impact of OK on investment inefficiency (efficiency). That is, increases in OK investments are more likely to increase investment distortions and, in turn, decrease investment efficiency. In terms of economic significance, a one standard deviation increase in the ratio of organizational capital to total assets (the ratio of organizational capital to total capital) results in a 4.42% (1.85%) decrease in investment efficiency. We also conduct analyses whereby we separate investment distortions into overinvestment and underinvestment and find that high OK is associated with both investment distortions. Thus, consistent with an agency-based interpretation, our main findings support the idea that increases in firms’ OK are more likely to result in information asymmetries (Biondi and Reb�erioux, 2012) and, in turn, increase corporate investment inefficiency (Myers and Majluf, 1984). In further analyses we examine several key CEO-specific factors that could moderate the relationship between OK and corporate investment. For example, we find that OK investments by female CEOs are associated with greater investment efficiency; this is consistent with the idea that female CEOs are more likely to engage in quality management practices, which reduce agency conflicts and information asymmetries (Adams and Ferreira, 2009; Jurkus et al., 2011). We also observe a negative moderating effect of CEO power. We interpret this result from the perspective that a potentially harmful effect of CEO power on investment efficiency may be mitigated in high-OK firms since this reduces the opportunity for such CEOs to overinvest (Chowdhury et al., 2023). Furthermore, we show that CEOs who are older, longer-tenured, and have more external board seats are also associated with more efficient investments when OK is higher. Moreover, having shown that OK is associated with both under- and overinvestment, we consider, in corporate governance, disclosure quality and CEO career concerns, several viable economic channels through which the negative impact of OK on investment efficiency may be conveyed. We find that the quality of corporate governance, including board size, independence and gender diversity negatively moderates the relationship between OK and investment inefficiency. In other words, high-OK firms with good governance are associated with more efficient investment. Similarly, we find that improvements in a firm’s information environment matter; firms with higher OK and with higher disclosure quality tend to make more efficient (less inefficient) corporate investments compared to those with lower disclosure quality. We also examine the impact of CEO career JAL 47,5 446 concerns by exploiting the passage of the inevitable disclosure doctrine as a proxy for CEO mobility and find that CEOs in high-OK firms who face greater career concerns make more inefficient investments. Finally, we examine whether the relationship between OK and investment efficiency varies according to whether a firm belongs to an R&D intensive industry, and therefore may face greater information asymmetries, and find that firms with high-OK who operate in such industries are associated with lower investment efficiency. We run a battery of robustness tests to validate our main findings. Importantly, we take several further steps to address potential endogeneity concerns, including using alternative estimators, including the system-GMM estimator and the 2SLS method, as well as implementing entropy balancing. We also run additional tests based on two subsamples based on the nature of products firms produce ((e.g. a subsample of industries that produce more tangible assets: agriculture (01–09), mining (10–14), construction (15–17), manufacturing (20–39)), and those that produce more intangible assets ((transportation (40–49), wholesale trade (50–51), retail trade (52–59), finance, insurance, and real estate (60–67), and services (70–89)), and find that the relationship between OK and investment efficiency holds even when splitting the sample based on the (tangible and intangible) nature of products firms produce. To add credence to the findings we report, we also check the sensitivity of our results over time. We do so, by splitting the main sample into two equal time periods: 2009 to 2014 and 2015 to 2020. Moreover, given that intangible assets have implications for firm risk, as firms with higher intangibles are generally associated with greater information asymmetry (e.g. Biondi and Reb�erioux, 2012; Wu and Lai, 2020), we test the stability of our findings with respect to levels of firm risk, by creating two equal subsamples based on median company beta. Finally, we rerun our main regressions using the Newey-West estimator as an alternative method to adjust standard errors. Our results remain fully robust for these tests. We make several contributions. To the best of our knowledge, our paper is the first to directly investigate the impact of OK on corporate investment inefficiency. To do so, we combine two main research streams. First, we expand the accounting literature on corporate investment efficiency, which has mostly focused on the role of accounting disclosure and quality in influencing corporate investment (Biddle and Hilary, 2006; Biddle et al., 2009; Chen et al., 2011; Gomariz and Ballesta, 2014; Laux and Ray, 2020), to consider the impact of OK on corporate investment efficiency. Second, we complement a recent interdisciplinary OK literature that demonstrates its relevance for various firm strategic choices and outcomes (Lev et al., 2009; Carlin et al., 2012; Hasan and Cheung, 2018; Attig and Cleary, 2014; Boguth et al., 2022; Provaty et al., 2024). For example, Attig and Cleary (2014), proxying for OK with management quality practices, find that firms with higher quality management practices exhibit lower investment sensitivity to internal cash flows. Eisfeldt and Papanikolaou (2013) show that OK is riskier and faces higher costs of equity financing. Boguth et al. (2022) emphasize the “fragile” nature of OK and show how the risk of OK reducing firm performance correlates with the outside employment options of top managers. We add to this literature by advancing understanding regarding the implications of OK for firms’ strategic choices by showing that OK is associated with lower investment efficiency, ceteris paribus. Aside from presenting new findings regarding the impact of OK on investment efficiency, we further contribute to existing knowledge by showing a role for several key CEO characteristics (external directorships, power, age, tenure, gender, and career concerns) as moderators of the relationship between OK and investment efficiency. In this way, we also extend a recent strand of the literature on the efficiency of capital allocation, which has explored the effect of such CEO-specific factors on investment efficiency, including social ties (Khedmati et al., 2020), power (Aktas et al., 2019), age and tenure (Li et al., 2017), and gender (Faccio et al., 2016). For example, we extend the work of Faccio et al. (2016) on gender, who show that although female-run firms have lower leverage and less volatile earnings, the preference of female CEOs for lower risk-taking is also associated with suboptimal investment. Importantly, our findings speak to a more nuanced effect of female CEOs on Journal of Accounting Literature 447 investment efficiency since we demonstrate that when female CEOs lead firms with high OK, they exhibit superior investment efficiency compared to male-led high-OK firms. Finally, our results also have several valuable practical implications for practitioners and investors that help explain the actual impacts of OK investments on investment decisions. From a practical perspective, our findings add to nascent understanding regarding the implications of OK for firm policies (e.g. Attig and Cleary, 2014; Attig and El Ghoul, 2018; Hasan et al., 2021). First, they infer that firms should be willing to improve information environments, including reporting and disclosure practices associated with intangible assets, to help investors overcome informational asymmetries associated with OK and, thus, better understand the value proposition of OK, allowing managers greater freedom from, for example, career concerns, to invest more efficiently. In this respect, since we show that female- led firms and firms with more gender diverse boards can help mitigate the negative impact informational asymmetries on investment efficiency, we lend support to international efforts to increase female leadership in large corporations, which is an important policy challenge in all countries (Huang and Kisgen, 2013). Second, in demonstrating an important negative association between OK and the efficiency of firm investment, our findings underscore the importance of robust corporate governance practices to address agency conflicts and information asymmetries linked to intangible assets such as OK. This is especially relevant for industries with high levels of OK and other intangible assets. Thus, strengthening governance structures can play a critical role in mitigating the adverse effects of OK on investment efficiency. In this vein, the importance of corporate governance mechanisms may be especially pronounced in specific industries, such as those that are R&D-intensive and require more input from human capital. The remainder of this paper is structured as follows: Section 2 discusses intangible assets and investment efficiency reforms in the US. Section 3 explains the theoretical framework. Section 4 reviews the relevant literature on OK and corporate investment efficiency and develops the study hypotheses. Section 5 describes the research design. Section 6 presents the empirical analyses and discussion of the findings. Finally, Section 7 concludes. 2. Intangible assets and investment efficiency reforms in the US Intangible assets play an important role in US industry. Much of the observable decline in physical investment in the US since the early 2000s, yet stable or increasing firm valuations (Gomme et al., 2011), has been attributed to the growing significance of intangible assets for economic rents (Crouzet and Eberly, 2023). For example, Lev and Gu (2016) emphasize that annual investments in intangible assets grew by 60% over the period 1977–2014 while investments in tangible assets decreased by 35%. Similarly, Corrado et al. (2022) illustrate that US investments in intangible assets had surged substantially from 1985 to 2021 and reached around 16% of the total GDP as of 2021, whereas annual investments in tangible assets had dropped by 4% during the same period, from 12.5% to 8.5% of the total GDP. Moreover, an increasing number of studies demonstrate that intangible assets, and OK in particular, are a key determinant of value creation and competitive advantage within the US economy (e.g. Corrado et al., 2005; Lim et al., 2020). Corrado et al. (2005) find that intangible assets contributed 75% of US economic growth in the early 2000s and valued US intangible assets at $3.4 trillion in the same time period, while Atkeson and Kehoe (2005) show that OK accounts for 40% of the total earnings produced by intangible assets in the US. International regulations and standards for intangible assets have evolved markedly over the past 30 years to address specific challenges surrounding the treatment of intangible assets [2]. From 1994 to 2004, early reforms included the voluntary implementation of the International Accounting Standards (IAS). This resulted in many regulatory bodies around the world rewriting and revising their standards to grant companies more flexibility to choose the appropriate accounting guidelines. In 2001, these guidelines were retitled the International Financial Reporting Standards (IFRS). Like many countries, intangibles in the US accounting JAL 47,5 448 framework are governed by detailed standards emphasizing recognition, measurement, and disclosure. Their treatment under US accounting standards reflects their growing importance and the challenges of accurately valuing these assets. Initially, APB Opinion No. 17 (1970) required amortization of goodwill and other intangibles over up to 40 years, emphasizing simplicity over economic realism. This shifted with SFAS 142 (2001), which replaced amortization with annual impairment testing for goodwill, aiming to better reflect an asset’s true value. Codified under ASC 350, this change aligned with the increasing importance of fair value measurements. In 2007, the IFRS and the US-preferred Generally Accepted Accounting Principles (US GAAP) began to harmonize their standards. However, despite this harmonization process and the introduction of new standards, accounting practices remain subject to ongoing debates, and there are significant differences between the treatment of intangibles in the US and internationally. For example, under US GAAP, intangibles, such as internally developed R&D costs, must be expensed immediately (ASC 730) (Canace et al., 2022), while IFRS (IAS 38) allows capitalization when specific criteria, like technical feasibility, are met. Goodwill is another key point of divergence; while US GAAP mandates annual impairment testing (ASC 350), IFRS allows amortization alongside impairment testing, reducing complexity but potentially obscuring economic reality. Aside from the treatment of intangibles under prevailing accounting frameworks, there have been changes in firms’ investment efficiency within the US, which have followed the passage of key regulatory changes. For example, through Section 404, the Sarbanes-Oxley Act of 2002 (SOX) added stricter requirements for financial reporting and controls, including that management and external auditors assess the effectiveness of internal controls over financial reporting. From this perspective, SOX enhanced corporate governance and financial transparency, indirectly fostering greater investment efficiency by reducing managerial opportunism (Aubert and Grudnitski, 2014). Furthermore, Section 201 of SOX, which imposed stricter rules on auditing, including prohibiting the outsourcing of internal auditing, also impacted investment efficiency through its influence on auditor conservatism (Lu and Sapra, 2009). Lu and Sapra (2009) find that Section 201 had a negative impact on investment efficiency and audit quality through its mandatory restriction of nonaudit services. More recently, the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 also had an influence on firm investment. For example, as shown by Du and Heo (2022), the introduction of the Dodd-Frank whistleblowing provision, which enhanced the protection afforded to firm whistleblowers, resulted in a negative relationship between political corruption and firm investment becoming insignificant. The above discussion infers that while investment efficiency and intangibles are important topics within the accounting discipline, they are also subject to much ongoing debate and are characterized by significant differences between the US compared to internationally. This divergence underscores ongoing debates about the relevance, reliability, and comparability of financial reporting for intangibles (Hussinki et al., 2024). Taken together, we consider that these differences may mean that the relationship between OK and investment efficiency may differ in the US compared to other country settings. We therefore argue that these differences make the US setting a highly relevant and interesting setting to study the effect of OK on corporate investment efficiency. 3. Theory Investment opportunities are a key determinant of corporate investment under the framework of Modigliani and Miller (1958). From the perspective of shareholders, managers should select positive net present value (NPV) projects and forgo those with negative NPVs. However, in corporations the separation of firm ownership and control can result in agency problems, attributable to the misalignment of managerial and shareholder incentives (Jensen and Meckling, 1976). Agency theory (Jensen and Meckling, 1976) posits that a firm is just a Journal of Accounting Literature 449 spectrum of contracts among shareholders and managers (principals and agents). Jensen and Meckling (1976) argue that firms are mere legal fictions and agency relationships can be reflected in contracts. Bonding, monitoring vehicles, and incentives pursued to diminish agency costs are the key ingredients of the contracts. Based on these contracts, principals (shareholders) delegate the running of the business to agents (managers), who they expect will run the business to maximize shareholder value. Here, agency theory presumes that corporate managers will engage in managerial opportunistic behavior (e.g. empire building and expense preference behavior) when there is a conflict of interests between managers and shareholders. These are important observations, given that the prior investment literature considers agency conflict and asymmetric information as the most important frictions that can result in corporate investment deviating from an optimal level (Biddle and Hilary, 2006; McNichols and Stubben, 2008; Biddle et al., 2009; Anagnostopoulou et al., 2023). Myers and Majluf (1984) and Myers (1984) demonstrate that asymmetric information between shareholders and corporate managers does not only affect the cost of capital but also the investment choice. Incentive misalignment and asymmetric information between managers and firm outsiders can result in under- or overinvestment (Jensen and Meckling, 1976; Lambert et al., 2007; Biddle et al., 2009). The presence of high informational asymmetries between managers and firm outsiders also gives rise to moral hazard and adverse selection issues, which can impact the efficiency of investment (Biddle et al., 2009). An important reason is that managers are more likely to pursue self-interest at the expense of shareholders when agency costs and informational asymmetries are high. For example, as argued by Jensen (1986), managers may overinvest in certain projects, such as pursuing empire building through acquisitions, when there is an abundance of resources to invest (Chen et al., 2017). Similarly, when corporate managers know that the company’s equities are overvalued, they tend to issue new equities and obtain more funds. However, corporate shareholders predicting this behavior are more likely to discount new equities if there are high information asymmetries (Lara et al., 2016; Wu et al., 2022). In line with these arguments, Barth and Kasznik (1999) predict and find that in firms with high intangible assets, which they associate with more substantial informational asymmetries, the stock market reacts more positively to share repurchase announcements. The reason being that such repurchases provide important information regarding managers’ beliefs regarding firm value, which helps mitigate information asymmetries present in high intangible asset firms. Conversely, managers may underinvest for several reasons, including career concerns (Goel et al., 2004; Kim et al., 2021), which are generally understood to be higher when information asymmetries between firm insiders and outsiders are more pronounced (Goel et al., 2004; Chemmanur et al., 2009; Attig and El Ghoul, 2018; Garc�ıa-S�anchez and Garc�ıa-Meca, 2018), such as in firms with significant intangible investments (Biondi and Reb�erioux, 2012). Furthermore, suppliers of capital, recognizing the potential for managers to deviate from shareholder objectives in firms with high information asymmetries, may restrict funding to such firms, which could exacerbate the tendency for self-interested managers to invest suboptimally. For example, underinvestment may be more likely as managers face financial constraints and, in turn, higher costs of firm capital. Consistent with prior studies, we consider that intangibles, including OK, impact firm information asymmetries and agency problems and that this matters from the perspective of investment efficiency (e.g. Barth and Kasznik, 1999; Barth et al., 2001; Park, 2019; Choi et al., 2013). Based on the consensus in the investment literature that firms with higher (lower) informational asymmetries exhibit lower (better) investment inefficiency (Barth and Kasznik, 1999; Barth et al., 2001; Biddle and Hilary, 2006; Biddle et al., 2009; Chemmanur et al., 2009; Chen et al., 2011, 2023; Benlemlih and Bitar, 2018), our main conjecture, as we hypothesize in subsection 4.3, is therefore that firms with high OK make less efficient investments, ceteris paribus. This is based on the notion that agency problems and informational asymmetries between firm insiders and outsiders are increasing with intangible assets (Barth and Kasznik, 1999; Barth et al., 2001; Biondi and Reb�erioux, 2012), which exacerbates agency problems and increases the scope for managers to act according to self-interest. There are several reasons JAL 47,5 450 why intangible assets should be associated with higher informational asymmetries and agency problems. For example, there is substantial economic uncertainty surrounding intangible assets, which helps explain why they are not recognized in financial statements given measurement difficulties (Barth and Kasznik, 1999; Barth et al., 2001; Hussinki et al., 2024). Moreover, since intangibles are not typically unrecognized and fair values undisclosed, ceteris paribus, firms with high intangibles have less informative prices (e.g. Barth et al., 2001). 4. Empirical literature review and hypotheses development 4.1 Corporate investment efficiency Corporate investment efficiency relates to how firms choose to invest, with deviations from optimal investment representing under- or overinvestment. The importance of understanding firm investment choices has led to a rich literature on corporate investment efficiency emerging over the past two decades [3]. The majority of this literature focuses on the implications of accounting quality and conservatism for investment efficiency. A seminal paper by Biddle and Hilary (2006) focuses on the relationship between accounting quality and capital investment. They show that accurate and transparent financial reporting enhances the effectiveness of investment choices by bridging the knowledge gap between company management and external investors. Similarly, Biddle et al. (2009) present evidence consistent with the notion that financial reporting quality reduces frictions such as moral hazards and adverse selection, which can hinder effective investment. While these studies focus on large, publicly listed US firms, Chen et al. (2011) and Gomariz and Ballesta (2014) focus on other country contexts. Chen et al. (2011) show similarity to Biddle et al. (2009) in the case of private firms in emerging markets, while Gomariz and Ballesta (2014) examine the impact of financial reporting quality on investment efficiency using a sample of Spanish-listed companies covering the period 1998–2008. They demonstrate that higher reporting quality can enhance investment efficiency and reduce investment distortions. In the same vein, Lara et al. (2016) investigate the impact of accounting conservatism on investment efficiency. They find that accounting conservatism enhances investment efficiency, mitigates equity-debt conflicts and underinvestment, and facilitates access to debt finance. Finally, Laux and Ray (2020) also study the relationship between accounting conservatism and investment efficiency in a theoretical setting and argue that the relationship depends on prevailing managerial incentives. A separate strand of the literature focuses on other firm-level factors that can influence investment efficiency, including earnings management (McNichols and Stubben, 2008), managerial quality (Chemmanur et al., 2009; Garc�ıa-S�anchez and Garc�ıa-Meca, 2018), cash holdings (Richardson, 2006), firm ownership (Richardson, 2006; Chen et al., 2017), and a firm’s choice of accounting standards (McClure and Zakolyukina, 2024). For example, McNichols and Stubben (2008) present evidence that managers who engage in firm earnings manipulation also overinvest during these periods, while McClure and Zakolyukina (2024) find that firms’ reporting standards matter for how managers invest. Specifically, they consider whether a firm’s choice to report earnings under GAAP or non-GAAP accounting standards has relevance for managerial investment. The authors show that when investors exclusively rely on a firm’s GAAP earnings, this encourages managers to distort from optimum levels (e.g. underinvestment). However, when non-GAAP earnings are used, this can mitigate underinvestment yet also result in overinvestment as it allows managers to more easily hide inefficient investment. 4.2 Organizational capital (OK) OK, still quite a new concept within the accounting and finance literature, involves the accumulation of technological knowledge, experience, processes, designs, and business practices that facilitate unique matching of labor and physical production resources, which combine to facilitate the efficient use of firm resources (Eisfeldt and Papanikolaou, 2013, Journal of Accounting Literature 451 2014; Li et al., 2018). Existing studies suggest that OK is partially embedded within senior managers and other key employees (Attig and El Ghoul, 2018; Eisfeldt and Papanikolaou, 2013, 2014). Carlin et al. (2012) theorize that the value of OK for firm performance depends on the extent to which managers are willing and able to coordinate within the firm to share OK, while a body of literature supports a positive relationship between managerial quality and the value of OK (Peters and Taylor, 2017; Eisfeldt and Papanikolaou, 2013, 2014). Boguth et al. (2022) argue that the value of OK for firms also depends on the “fragility” of OK and that this fragility is determined by the marginal contribution of managers to OK. Specifically, their findings demonstrate that high-OK firms suffer more significant declines in firm market performance from the departure of key managers when they have smaller top management teams compared to firms with larger teams. Extant literature presents mixed findings as to whether OK should convey a positive or negative impact on firms’ strategic choices and outcomes. From one perspective, OK can represent a source of unique and sustainable competitive advantage for firms (Lev et al., 2009; Carlin et al., 2012; Hasan and Cheung, 2018). In this way, OK can help mitigate agency conflicts between managers and shareholders by improving the information environment. Consistent with this, Attig and El Ghoul (2018) present evidence that OK can reduce information asymmetries between managers and firm outsiders. They show that firms’ investments in OK are rewarded with superior management practices; specifically, such firms disclose more and better-quality information to investors, which reduces information asymmetries and reduces the implied cost of firm equity. Provaty et al. (2024, p. 2) make comparable arguments, using a signaling theory perspective, to propose that firms with high OK may undertake initiatives to reduce greenhouse gas (GHG) emissions to “communicate and signal their outstanding capabilities”. In support of such arguments, there are a number of studies that highlight several benefits to firms associated with OK, including, a lower cost of equity financing (Attig and El Ghoul, 2018), higher firm values and stock returns (Orhangazi, 2019; Lev et al., 2009; Eisfeldt and Papanikolaou, 2013; Gao et al., 2021), and higher firm productivity (Gao et al., 2021). For example, Gao et al. (2021) show that investment in OK is associated with enhanced firm productivity and shareholder value, while Eisfeldt and Papanikolaou (2013) estimate that firms with more OK realize average returns 4.6% higher compared to low OK firms. A few studies focus on various other dimensions of firm performance. The influence of OK on merger and acquisition (M&A) performance is the focus of Li et al. (2018). Using data on US acquisitions between 1984 and 2014, they show that higher OK is associated with better deal performance – both in terms of short-term market returns to M&A announcements and post-merger operating and stock performance. Hasan et al. (2021) find that higher OK is associated with increased future cash flows and that tax avoidance strategies, including investing in tax haven subsidiaries, are a mechanism through which this can occur. Finally, Provaty et al. (2024) demonstrate that high-OK firms are associated with lower direct and indirect GHG emissions. Conversely, OK may have a negative impact on firms’ strategic choices and outcomes. Investments in OK are often considered risky and uncertain (Eisfeldt and Papanikolaou, 2013, 2014; Attig and El Ghoul, 2018; Boguth et al., 2022), and information asymmetries and uncertainty surrounding managerial investment choices mean that firm outsiders fail to recognize the value of OK investments (Eisfeldt and Papanikolaou, 2013, 2014). Consistent with this, Eisfeldt and Papanikolaou (2013) show that firms with higher OK are more exposed to firm-specific and aggregate shocks, which raises the cost of equity financing. Furthermore, issues with disclosure, reporting, and general measurement of OK could lead to managerial performance being evaluated more extensively based on the outcomes of their project choices in firms characterized by high OK. Consequently, managers may invest suboptimally in high- OK firms because of exacerbated career concerns (Goel et al., 2004; Kim et al., 2021). This could also imply that managers may hold more cash to offset such risks. Consistent with this notion, Marwick et al. (2020) focus on the relationship between OK and corporate cash holdings and show that firms that invest more in OK hold more cash and that this effect is JAL 47,5 452 strongest amongst financially constrained firms and those with riskier cash flows. To the extent that high cash holdings may reflect an agency concern from the perspective of shareholders, shareholders may be wary of allowing managers to cash in firms with high OK. For example, Hasan and Uddin (2022) examined the relationship between OK and payout policies. Using a sample of US firms from 1980–2017, they find that firms with higher OK tend to have higher levels of cash dividends and share repurchases. The authors interpret their results from the perspective of agency theory, whereby managers in high-OK firms may invest suboptimally to maximize private benefits and dividend payouts, which therefore serves as a means to discipline managers. 4.3 Linking organizational capital (OK) with corporate investment efficiency Since we lack prior studies that examine the implications of OK for corporate investment efficiency, in this section we integrate the existing literature on OK and the literature on corporate investment efficiency to outline two hypotheses regarding the effect of OK on investment efficiency. The first hypothesis we consider relates to the fundamental relationship between OK and investment efficiency. As we discussed earlier in the introduction section, the effect of OK on investment is theoretically ambiguous. Yet as we explain, we consider that the balance of arguments supports a negative relationship between OK and investment efficiency. From one perspective, OK could be positively related to investment efficiency. For example, OK could potentially enhance corporate investment efficiency by improving the quality and extent of managerial disclosures, thereby reducing information asymmetries and lowering managerial career concerns (Goel et al., 2004; Chemmanur et al., 2009; Attig and El Ghoul, 2018; Garc�ıa-S�anchez and Garc�ıa-Meca, 2018). Additionally, high-OK firms may be more likely to possess superior management capabilities, including better alignment of managerial capital with OK, resulting in more efficient resource allocation (Carlin et al., 2012; Hasan and Cheung, 2018). Consistent with this, Chemmanur et al. (2009) demonstrate that firms with strong managers make better investment choices, increasing capital expenditures and R&D investment. Similarly, Boguth et al. (2022) find that while firms with high OK are riskier, they also yield higher future stock returns, highlighting OK as a source of sustainable competitive advantage. However, we postulate that, ceteris paribus, OK should be negatively associated with investment efficiency for several reasons. Investments in OK in particular and intangible assets in general can exacerbate agency conflicts between shareholders and managers. One important reason is that the high information complexity of intangible assets increases information problems and moral hazards, rendering it more difficult for outsiders to evaluate firm operations (Barth et al., 2001; Gu and Wang, 2005; Lev et al., 2009). For example, it is more difficult for analysts to understand and predict the future performance of intangible-intensive firms, which can result in higher forecast errors (Gu and Wang, 2005). The difficulties firm outsiders face with assessing the value of firms’ OK are further compounded by the fact that OK experiences measurement challenges under current accounting principles (Atkeson and Kehoe, 2005; Crouzet and Eberly, 2023). For instance, OK is always expensed on the income statement, does not appear as an asset on the balance sheet (Danielova et al., 2023), and is typically not capitalized (Crouzet and Eberly, 2023). Furthermore, current accounting principles largely lack detailed requirements related to the reporting and disclosure of intangible assets, and the use of book value to value OK can result in an under- or over-valuation problem (Peters and Taylor, 2017; Chan et al., 2022; Crouzet and Eberly, 2023). Thus, a fair value for OK is unlikely to exist in financial statements. Such issues may lead investors to not fully recognize the future benefits of OK investment, which increases information asymmetries between investors and managers (Eisfeldt and Papanikolaou, 2013, 2014). For these reasons, external shareholders may resist increases in OK investments, which eventually deepens the agency conflict between shareholders and top managers (Boguth et al., Journal of Accounting Literature 453 2022) and makes it more difficult for shareholders to monitor (Richardson, 2006; Morellec and Sch€urhoff, 2011). Moreover, firms with higher OK tend to have less informative prices (Barth et al., 2001) and are more prone to negative financial shocks, which increases the riskiness of their cash flows (He, 2018). Consequently, managers of high-OK firms may invest suboptimally because of exacerbated career concerns (Goel et al., 2004; Kim et al., 2021). One reason is that evaluations of their performance by shareholders are more closely linked to the outcome of their investment decisions in firms with higher information asymmetries (Goel et al., 2004). Consistent with this, Carlin et al. (2012) present a model in which managers in firms with stronger employment protection regulations, and hence who face lower managerial career concerns, are more likely to invest in OK. In other words, managerial career concerns in firms with high OK are likely to be more salient and are expected to impact corporate investment efficiency. Thus, based on these arguments, we hypothesize: H1. Firms with high levels of organizational capital have lower investment efficiency, ceteris paribus. However, while intangible assets are typically associated with more pronounced agency problems and informational asymmetries (e.g. Biondi and Reb�erioux, 2012), there are several important factors that could moderate the expected impact of OK on investment efficiency. Notably, firms with higher OK have been associated with more efficient organizational structures that facilitate better quality information flows, management practices, product quality, reputations, and greater stability of firm operations and production (Attig and Cleary, 2014; Hasan et al., 2022; Danielova et al., 2023). Such findings support the notion that OK investments can contribute to valuable and sustainable advantages for firms (Carlin et al., 2012; Hasan and Cheung, 2018; Danielova et al., 2023). Furthermore, aside from corporate governance, investments in OK have been shown to facilitate greater quality and number of informational disclosures by managers (Goel et al., 2004; Chemmanur et al., 2009; Attig and El Ghoul, 2018; Garc�ıa-S�anchez and Garc�ıa-Meca, 2018). Based on such observations, better corporate governance arrangements and information disclosures by managers could mitigate information asymmetries in high-OK firms. Therefore, we further posit that stronger corporate governance arrangements, better disclosure quality, and the extent of labor market frictions (e.g. Biddle and Hilary, 2006; Gomariz and Ballestra, 2014; Chen et al., 2021, 2023; Danielova et al., 2023; Jiang et al., 2024) [4] could positively moderate the expected negative impact of OK on corporate investment efficiency; that is, they may represent key economic channels through which OK investments could potentially lead to an increase in investment efficiency. We test for these possibilities in Section 6.1, Economic channels. If true, this would help explain recent findings of a positive relationship between OK for and the stock price of firms (Lev et al., 2009; Eisfeldt and Papanikolaou, 2013) and the finding of Danielova et al. (2023) that OK is negatively related to the cost of bank loans. Based on this discussion, while H1 predicts a negative relationship between OK and corporate investment efficiency, we further conjecture that the impact of OK on corporate investment may be positive in the context of firms with lower information asymmetries. That is, we posit that specific firm-level conditions associated with lower agency costs could serve to moderate the negative impact of OK on investment efficiency: H2. Organizational capital is associated with better investment efficiency in firms with lower agency costs. 5. Research design 5.1 Sample and data Our unbalanced panel dataset is based on the constituents of the S&P 500 index, which comprises the leading 500 companies in the US economy with a contribution of more than 80% of the US market capitalization (S&P 500, 2023). We source firm financials and other relevant JAL 47,5 454 data from LSEG’s DataStream platform (formerly Thomson Refinitiv) for all companies present in the S&P 500 over the period 2009–2020 to mitigate survivorship bias [5]. This yields a final sample of 486 firms with 5,819 firm-year observations [6]. All data are winsorized at 1 and 99% to mitigate the effect of potential outliers. Since S&P 500 firms encounter the lowest degree of information asymmetry between management and shareholders [7], due to the continuing flow of private and public information, we select only S&P 500 constituents rather than, for example, the wider S&P 1500 in order to isolate the effect of OK on corporate investment efficiency. The S&P 500 is widely considered as the best representation of the U.S. stock market (Jiang and Fang, 2015; Cornaggia et al., 2017; Krauss et al., 2017). Furthermore, our dataset is not extended to cover the post-COVID-19 pandemic period 2021–2023 in order to mitigate the potential influence of the COVID-19 pandemic on corporate investment policies [8]. In this vein, Howe et al. (2021) and Buchheim et al. (2022) illustrate that the COVID-19 pandemic has left a deep effect on corporate behavior and has forced firms to change all aspects of corporate strategy (investment, finance, marketing, operations, and HR). 5.2 Investment inefficiency In a theoretical world absent of market imperfections or “frictions,” investment efficiency infers the efficient use of firm resources in projects that return positive net present value (hereinafter, NPV) (Modigliani and Miller, 1958). However, in reality, frictions exist. These frictions render it challenging for investors to observe managerial investments and firm cash flows (Stulz, 1990). Accordingly, this can give rise to overinvestment, whereby managers pursue projects with zero or negative NPVs, or underinvestment, where managers forego positive NPV investments (Biddle et al., 2009). Following the corporate investment literature, we construct our main measure of investment efficiency by first modeling the expected levels of total investment as a function of investment opportunities, internal finance competencies (McNichols and Stubben, 2008), and growth opportunities (Biddle et al., 2009). Investment inefficiency is then estimated as the deviations or residuals from the optimal investment model as follows (e.g. Richardson, 2006; Biddle et al., 2009; Gomariz and Ballesta, 2014; Barth et al., 2017). Positive (or negative) residuals from equation (1) infer overinvestment (underinvestment) relative to the expected level of investment. Investmenti;tþ1 ¼ β0 þ β1SalesGrowthi ;t þ β2TobinsQi;t þ β3CFOi;t þ ei;t (1) where Investmenti;tþ1 is the total investment of a firm, computed as the net increase in tangible and intangible assets scaled by lagged total assets; SalesGrowthi ;t is the percentage change in sales from t�1 to t; TobinsQi;t is the market-to-book ratio; CFOi;t is the operating cash flows; and ei;t is the error term, which is also the key residual from the optimal investment model and our measurement of investment inefficiency, our main dependent variable. 5.3 Organizational capital (OK) measures To estimate a firm’s OK we join Hasan et al. (2021) in adopting the perpetual inventory method proposed by Peters and Taylor (2017), which utilizes a firm’s selling, general, and administrative (SG&A) expenses and includes a firm’s operating expenses that are not directly related to the production of goods or services, including costs relating to IT outlays, employee training costs, brand enhancement activities, payments to consultants, and costs associated with establishing and maintaining online supply and distribution channels (Lev et al., 2009). Specifically, we follow Peters and Taylor (2017) in computing the level of OK for each firm year as follows: OKi;t ¼ ð1� δOCÞOCi;t−1 þ ðSG&Ai;t 3 λOCÞ (2) And where the initial value of OK is calculated as: Journal of Accounting Literature 455 OKi;t ¼ � SG&Ai;t 3 λOC gþ δOC � (3) whereOKi;t represents the OK of firm i at time t, δOC indicates the rate at which OK depreciates, and is set equal to 20% (0.2) (Eisfeldt and Papanikolaou, 2013; Peters and Taylor, 2017; Hasan et al., 2021). SG&Ai;t expenses of firm i in year t, are denoted by SG&Ai;t: The percentage of SG&A expenditure allocated to OK investment is shown by λOC and is set equal to 30% (0.3) (Eisfeldt and Papanikolaou, 2013; Peters and Taylor, 2017; Hasan et al., 2021). Lastly, the average growth rate in firm-level SG&A expenses is represented by g. 5.4 Model specification and control variables To examine the effect of OK on corporate investment efficiency, we follow the latest developments in investment efficiency and estimate panel regressions with firm- and year- fixed effects and with standard errors clustered at the firm level (following Hasan and Cheung, 2018; Hasan et al., 2021; Chan et al., 2022; Xu et al., 2024). Equation (4) presents the baseline fixed effects model [9]: Investment inefficiencyi;t ¼ β0 þ β1OKi;t þ Σ β2CONTROLSi;t þ β3FirmiFE þ β4YeartFE þ εi;t (4) where Investment inefficiencyi;t denotes the residuals from the optimal investment model and is the dependent variable. OKi;t, is the main variable of interest, which is represented by the ratio of OK to total assets (OKTA) or the ratio of OK to total capital (OKTC) (Hasan et al., 2021). Right-hand side variables are contemporaneous in the model, consistent with best practices in the literature (Hasan and Cheung, 2018; Chan et al., 2022; Hasan et al., 2021). We include a vector of control variables motivated by prior literature. SIZE is the log of annual market capitalization. Prior literature infers that firm size can convey a strong impact on firm investment decisions (Wu et al., 2022). Moreover, larger firms are less prone to agency problems and default (Rajkovic, 2020). LEVERAGE is computed as the ratio of total debt to total capital. The effect of leverage on investment efficiency can vary and result in both over- and underinvestment (Biddle et al., 2009). Based on agency theory, Biddle et al. (2009) explain that firms with low (high) leverage and those that have low (high) cash holdings are more likely to overinvest (underinvest). Agrawal and Knoeber (1996) argue that leverage can reduce manager-shareholder agency problems through the additional monitoring provided by debtholders. However, Aivazian et al. (2005) find a negative and significant impact of leverage on corporate investment and confirm that this relationship is stronger among firms with lower growth opportunities. ROA represents return on investment and is computed as the ratio of earnings before interest and taxes to total assets. We control for whether a firm has negative earnings with LOSS, a dummy variable that takes a value of 1 if a firm makes a loss and 0 if it does not (e.g. Chen et al., 2017). Both variables, ROA and LOSS, are employed to control firm performance. Following Biddle and Hilary (2006), we control for a firm’s investment opportunities using TOBIN’S Q, proxied by the market-to-book equity ratio. Biddle and Hilary (2006) find a negative relationship between TOBIN’S Q and investment efficiency. AGE is firm age, computed as the number of years since the company was listed on the stock market. Firm age is typically found to have a positive impact on investment efficiency, as younger firms tend to be less visible and less known by investors and debtholders (Xu et al., 2024). Also, more established firms often have easier access to external finance resources compared to younger firms. Consequently, older firms tend to have more capital investments and more efficient investments (Sufi, 2007). SLACK is measured using the ratio of cash holdings to total assets and is used to control for the impact of cash on investment efficiency (Biddle and Hilary, 2006; Biddle et al., 2009; Gomariz and Ballesta, 2014), since JAL 47,5 456 this could result in either underinvestment or overinvestment (Biddle et al., 2009). To better isolate any effect of OK on corporate investment efficiency, we also control for the degree of tangibility of a firm’s assets with TANGIBILITY, measured as the ratio of fixed assets to total assets. Ferris et al. (2017) demonstrate that the inclusion of tangibility and leverage can control the likelihood of firm financial distress. Finally, we include firm (Chen et al., 2017) and year- fixed effects (Lara et al., 2016; Chen et al., 2017) to account for unobserved firm- and year- specific factors that could influence investment (Chen et al., 2017). 6. Empirical results and discussion 6.1 Baseline results 6.1.1 Summary statistics. Table 1 presents the descriptive statistics for the dependent and independent variables based on the full sample of 486 firms and 5,819 firm-year observations. The average organizational capital to total assets, OKTA, is 0.193 with a standard deviation of 0.206, while the average organizational capital to total capital,OKTC, is 0.284 with a standard deviation of 0.325. The minimum values ofOKTA andOKTC are 0.000 for both, while the 25th (75th) percentile value is 0.054 (0.260) for OKTA and 0.080 (0.364) for OKTC. Finally, the maximum values are 1.110 and 1.798, respectively. These values are very similar to those reported in Marwick et al. (2020), who use a comparable sample of US firms. The average sample investment inefficiency is �0.001 with a range from �0.018 to 0.030, which is also similar to prior studies (e.g. Benlemlih and Bitar, 2018). These values suggest that both overinvestment (positive residuals) and underinvestment (negative residuals) are present in our sample. The instrumental variable (IV), industry-level growth uncertainty, has a sample mean of 0.127 and a standard deviation of 0.121. The minimum value of the IV is 0.001, while the 25th (75th) percentile value is 0.055 (0.149) and the maximum value is 0.704. These estimates are comparable to prior studies (e.g. Li et al., 2018). Table 1. Descriptive statistics for organizational capital and investment efficiency Variable Mean Std. dev. Min Max 25th 75th Observations OKTA 0.193 0.206 0.000 1.110 0.054 0.260 5,742 OKTC 0.284 0.325 0.000 1.798 0.080 0.364 5,706 Investment inefficiency �0.001 0.005 �0.018 0.030 �0.002 0.001 5,161 Size (log) 16.513 1.253 12.735 19.521 15.961 17.354 5,742 ROA 0.010 0.011 �0.110 0.161 0.041 0.101 5,685 Slack 0.144 0.163 0.000 1.000 0.025 0.189 5,712 Tangibility 0.625 0.220 0.067 1.000 0.468 0.835 5,742 Firm age (log) 23.864 12.280 0.000 50.000 14.000 35.000 5,479 Leverage 0.663 0.512 0.106 1.000 0.253 0.564 5,783 Tobin’ Q 0.002 0.002 0.001 0.025 0.008 0.016 5,676 Loss 0.054 0.226 0.000 1.000 0.000 0.000 5,819 Industry-level growth uncertainty 0.127 0.121 0.001 0.704 0.055 0.149 5,797 CEO age 55.763 7.139 21 86 52 61 5,769 CEO gender 0.035 0.182 0.000 1.000 0.000 0.000 5,762 CEO duality (power) 0.472 0.499 0.000 1.000 0.000 0.000 5,783 CEO tenure 6.420 6.347 0.000 49.000 2.000 10.000 5,745 CEO external board seats 2.382 2.450 0.000 29.000 1.000 3.000 5,273 IDD 0.365 0.481 0.000 1.000 0.000 0.000 5,819 Note(s):This table presents the descriptive statistics, including mean, standard deviation, minimum, maximum, 25th and 75th percentiles and number of observations of the empirical data for all the study variables, including the instrumental variable: Industry-level growth uncertainty. All variables are defined in Appendix 1 Source(s): Authors’ own work Journal of Accounting Literature 457 Table 2 shows correlations between our main variables. OK measures (OKTA and OKTC) are negatively correlated with investment, in line with the descriptive statistics presented in Table 2. All correlation coefficients range from �0.600 to 0.554, suggesting they are within commonly acceptable levels. Hence, we do not expect to have a multicollinearity problem. 6.1.2 Baseline regressions. Table 3 exhibits the results of the baseline multivariate regressions that examine the impact of OK on investment efficiency. The first two columns present the results of the first measure of organizational capital (OKTA), while the last two columns present the results of the second measure of OK (OKTC). We found that OK is associated with a positive impact on investment inefficiency, with regression coefficients of 0.021 (t 5 4.82) and 0.009 (t 5 5.35) for OKTA and OKTC, respectively, statistically significant at a 1% significance level. The positive sign on the OK coefficients indicates that a higher OK is associated with increased investment inefficiency [10]. The results are also economically meaningful. For example, a one-standard deviation increase in the ratio ofOKTA (OKTC) is associated with a 4.42% (1.85%) decrease in investment efficiency. To calculate the dollar economic value of investment inefficiency, we multiply the percentage of deviation by the absolute mean of investments. This gives an average of $390.88 million (8843.382 3 4.42%) and $163.60 (8843.382 3 1.85%). This means that a one standard deviation in OKTA (OKTC) is associated with an investment distortion of $390.88 (163.60) million. We therefore find support for our first hypothesis and for an agency-based interpretation of OK investments. That is, OK investments can exacerbate agency conflicts between managers and stakeholders (e.g. Biondi and Reb�erioux, 2012), resulting in deviations from optimal investment (Goel et al., 2004; Kim et al., 2021). In this vein, managers may suboptimally invest because of heightened informational asymmetries surrounding OK that make it more difficult and costly for shareholders to monitor and evaluate managerial investment choices (Richardson, 2006; Morellec and Sch€urhoff, 2011). Furthermore, managers may also invest suboptimally because of heightened career concerns attributable to the fact that investors choose to rely more extensively on the outcome of managers’ project choices as indicators of managerial ability. This could lead to overinvestment in short-term projects and underinvestment in longer-term projects (Kim et al., 2021). With respect to context, our results may be partially explained by the US setting. For example, they tentatively infer that ASC 730, which requires that internally developed R&D costs must be expensed immediately (ASC 73), may be ineffective in reducing information asymmetries stemming from intangible assets. An important reason is that ASC 730 may encourage financial officers to capitalize R&D expenditures to manipulate earnings and investment levels (Canace et al., 2023). 6.1.3 Overinvestment and underinvestment. Having established that OK investment is associated with reduced investment efficiency, in this section we examine the direction of deviations from investment efficiency in terms of under- and overinvestment (Biddle et al., 2009; Richardson, 2006; Kim et al., 2021). Although investment efficiency is generally expected to decrease in organizational environments characterized by high informational asymmetries because managers have more opportunity to prioritize their own interests over those of shareholders, as predicted by agency theory, this could result in both underinvestment and overinvestment. If informational asymmetries surrounding OK investments render it difficult for firm outsiders to correctly value OK, then managers operating in high-OK firms could deviate, and underinvest, from optimal levels. For example, higher informational asymmetries associated with OK could lead investors to rely more extensively on the outcome of managers’ project choices as indicators of managerial ability (Goel et al., 2004). In turn, this would be expected to mitigate the likelihood that managers overinvest and could result in underinvestment due to exacerbated managerial career concerns (Kim et al., 2021). More specifically, this could result in overinvestment in short-term projects and underinvestment in longer-term projects (Kim et al., 2021). Another important reason for potential underinvestment relates to managers’ concerns that the market will apply a discount to new equity or debt securities issued because JAL 47,5 458 Table 2. Correlation matrix Variable OKTA OKTC Investment inefficiency Size ROA Slack Tangibility Firm age Loss Lev TQ OKTA 1.000 OKTC 0.929 1.000 Investment inefficiency �0.079 �0.050 1.000 Size �0.054 �0.046 �0.244 1.000 ROA 0.374 0.329 �0.323 0.135 1.000 Slack 0.118 0.126 �0.040 0.038 0.171 1.000 Tangibility �0.262 �0.352 0.031 0.023 �0.224 �0.600 1.000 Firm age 0.032 0.037 �0.118 0.199 0.033 �0.102 0.045 1.000 Loss �0.006 �0.014 0.072 �0.101 �0.445 0.079 �0.001 �0.060 1.000 Lev �0.004 0.046 �0.039 0.092 �0.008 �0.148 0.106 �0.012 0.035 1.000 Tobin’s Q 0.361 0.319 �0.353 0.095 0.554 0.400 �0.297 �0.064 0.016 �0.096 1.000 Note(s): This table presents Pearson correlations for the main study variables. All variables are as defined in Appendix 1 Source(s): Authors’ own work Journalof A ccounting Literature 459 investors anticipate that managers will choose to issue new securities when managers hold private information that they are overvalued (Chen et al., 2017). High-OK firms could also be associated with overinvestment. The weakened information environment in high-OK firms could provide opportunities for managers to overinvest by selecting projects that fail to increase, or could erode, shareholder value. This is consistent with existing evidence that managers may overinvest when information asymmetries between firm insiders and outsiders are more pronounced, and therefore monitoring managerial investment is more costly and difficult (e.g. Richardson, 2006; Morellec and Sch€urhoff, 2011). Overinvestment could also occur in high-OK firms because managers accelerate investment to signal the strength of their human capital and management practices and, ultimately, that they are a “good firm” (Wang and Yu, 2023). Table 4 presents the results. Overinvestment is represented by positive residuals from the optimal investment model, and second, underinvestment is captured by negative residuals. Concerning overinvestment, we observe a positive impact of OK on overinvestment for both Table 3. Baseline regression results of OK on corporate investment inefficiency Variables Investment inefficiency (1) Investment inefficiency (2) Investment inefficiency (3) Investment inefficiency (4) Intercept �0.004*** 0.013 �0.003*** 0.018** (0.001) (0.008) (0.000) (0.008) OKTA 0.017*** 0.021*** (0.003) (0.004) OKTC 0.007*** 0.009*** (0.001) (0.002) Size �0.001** �0.001** (0.000) (0.000) ROA �0.000 �0.000 (0.000) (0.000) Slack 0.010*** 0.010*** (0.003) (0.003) Tangibility 0.008*** 0.008*** (0.002) (0.002) Firm age �0.001** �0.002** (0.001) (0.001) Leverage 0.000 �0.000 (0.000) (0.000) Tobin’s Q �0.000*** �0.000*** (0.000) (0.000) Loss �0.001 �0.001 (0.001) (0.001) Firm FE Yes Yes Yes Yes Year FE Yes Yes Yes Yes R2 (%) 42.02 51.76 41.31 51.26 F-statistics (Wald) 24.07 25.65 25.07 27.86 p-value (0.000) (0.000) (0.000) (0.000) Observations 5,160 4,909 5,150 4,902 Note(s): The table presents the results of baseline regressions that examine the impact of OK on investment inefficiency. The dependent variable is investment inefficiency, which captures deviation from optimal investment based on the residuals from the investment model. The main independent variable is OK, measured by either OKTA, the ratio of OK to total assets, or OKTC, which is the ratio of OK to total capital and is our independent variable. Control variables include size, ROA, slack, tangibility, firm age, leverage, Tobin’s Q, and Loss. All variables are as defined in Appendix 1. Figures in parentheses are standard errors that are clustered at the firm level for heteroscedasticity. (***) (**) (*) significance at the (1%) (5%) (10%) levels Source(s): Authors’ own work JAL 47,5 460 measures of OK, with regression coefficients of 0.021 (t 5 3.92) and 0.007 (t 5 3.51) for OKTA and OKTC, respectively. With respect to underinvestment, we multiply the variable of underinvestment by�1 to facilitate its interpretation (following Chen et al., 2011). In contrast, the impact of OK on underinvestment is negative and significant, with regression coefficients of�0.007 (t 5�3.80) and�0.003 (t 5�3.62) for OKTA and OKTC. Thus, OK investments tend to increase deviations from optimal investment in terms of overinvestment and decrease the deviation for underinvestment. 6.2 CEO characteristics as moderating variables In this section, we consider the role of several CEO traits as potential moderators (CEO gender, age, power, tenure, and CEO external directorships) of the relationship between OK and firm investment efficiency. The results are displayed in Table 5. 6.2.1 CEO gender. We first explore a potential moderating effect of CEO gender on the relationship between OK and investment efficiency. Prior literature documents that female CEOs are less likely to experience behavioral biases (such as overconfidence), which can distort investment decisions (Huang and Kisgen, 2013). Female CEOs are also more likely to engage in quality management practices, such as corporate governance, monitoring, and financial reporting (Adams and Ferreira, 2009; Jurkus et al., 2011), which improve the information environment and reduce agency conflicts, thus making it easier for the market to evaluate managerial investment choices. Given this, we predict that firms with higher OK and whose CEOs are female will make more efficient investments compared to high OK firms with male CEOs. To examine the moderating impact of CEO gender, we constructed a dummy variable that takes one if the CEO is female and zero otherwise. The first column of Table 5 reports the results of the moderating role of CEO gender. For reasons of brevity, we restrict the analysis to our main measure of OK: OKTA [11]. The key interaction term, OKTA 3 CEO gender, is Table 4. Overinvestment and underinvestment Variables Overinvestment (1) Underinvestment (2) Overinvestment (3) Underinvestment (4) Intercept 0.017** �0.018*** 0.020** �0.016*** (0.008) (0.005) (0.008) (0.005) OKTA 0.021*** �0.007*** (0.005) (0.002) OKTC 0.007*** �0.003*** (0.002) (0.001) Control variables Yes Yes Yes Yes Firm FE Yes Yes Yes Yes Year FE Yes Yes Yes Yes R2 (%) 72.55 78.40 71.62 78.32 F-statistics (Wald) 5.18 46.24 5.04 46.50 p-value (0.000) (0.000) (0.000) (0.000) Observations 1,914 2,896 1,914 2,890 Note(s): The table presents the results of regressions that examine how OK impacts the direction of deviations from efficient investment (investment inefficiency). The sample period is divided into two subsamples, Overinvestment and Underinvestment, based on positive (overinvestment) or negative residuals (underinvestment) from the investment model. OKTA is the ratio of OK to total assets. OKTC is the ratio of OK to total capital. The investment inefficiency is the deviation from the optimal investment and is computed as the residuals from the investment model. Control variables include size, ROA, slack, tangibility, firm age, leverage,Tobin’sQ, andLoss. All variables are as defined in the Appendix 1. Figures in parentheses are standard errors that are clustered at the firm level for heteroscedasticity. (***) (**) (*) significance at the (1%) (5%) (10%) levels Source(s): Authors’ own work Journal of Accounting Literature 461 negative and significant at a 10% significance level (�0.009 with a t-value of �1.70). This finding supports our a priori prediction that CEO gender should act as a negative moderator of the relationship between organizational capital and investment inefficiency. In other words, firms with higher OK whose CEOs are female are associated with greater investment efficiency compared to those whose CEOs are male. Thus, our finding of greater investment Table 5. The moderating role of CEO gender, age, power, tenure, and external board seats Variables CEO gender (1) CEO age (2) CEO power (3) CEO tenure (4) CEO external directorships (5) Intercept 0.062* 0.022*** 0.019*** 0.024*** 0.023*** (0.035) (0.008) (0.001) (0.008) (0.008) OKTA 0.045*** 0.008*** 0.007*** 0.006*** 0.008*** (0.017) (0.002) (0.000) (0.001) (0.002) CEO gender �0.003* (0.002) CEO age 0.001*** (0.000) CEO power 0.001*** (0.001) CEO tenure 0.001*** (0.000) CEO external directorships 0.001*** (0.000) OKTA * CEO gender �0.009*** (0.005) OKTA * CEO age �0.006*** (0.001) OKTA * CEO power �0.005*** (0.000) OKTA * CEO tenure �0.004*** (0.000) OKTA * CEO external directorships �0.006*** (0.002) Control variables Yes Yes Yes Yes Yes Firm FE Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes R2 (%) 34.87 50.48 50.41 49.92 49.90 F-statistics (Wald) 6.32 21.63 23.51 25.63 22.70 p-value (0.000) (0.000) (0.000) (0.000) (0.000) Observations 4,867 4,909 4,840 4,909 4,909 Note(s): This table presents the regression coefficients of the panel regression model. The dependent variable, Investment inefficiency, is the deviation from optimal investment and computed as the residuals from the investment model. OKTA is the ratio of organizational capital to total assets and is the independent variable of interest. CEO gender is the first moderator and is constructed as dummy variable that takes one if the CEO is female and zero otherwise. CEO age is the second moderator and is constructed as dummy variable that takes one if the CEO’s age is above the sample median and zero otherwise. CEO power is the third moderator and is constructed as a dummy variable that takes the value of one if the CEO is also the chairman of the board of directors and zero otherwise. CEO tenure is the fourth moderator. It takes the form of a dummy variable equal to one if a CEO’s tenure is above the sample median and zero otherwise. CEO external directorships is the fifth moderator, which is constructed as a dummy variable that takes a value of one if the number of external board membership seats of a CEO is above the sample median and zero otherwise. Control variables include size,ROA, slack, tangibility, firm age, leverage, Tobin’s Q, and Loss. All variables are as defined in Appendix 1. Figures in parentheses are standard errors that are clustered at the firm level for heteroscedasticity. (***) (**) (*) significance at the (1%) (5%) (10%) levels Source(s): Authors’ own work JAL 47,5 462 efficiency in firms that have higher OK and are run by female leaders adds support to international efforts to increase female representation in leadership positions in private firms — an important policy challenge in all countries. 6.2.2 CEO age. Next, we consider a potential moderating role for CEO age. Prior literature documents several reasons why younger CEOs could be associated with lower investment efficiency compared to older peers. First, given that a CEO’s current performance determines her future employment opportunities, younger CEOs with longer career paths tend to have greater career concerns and may invest less and more efficiently (Yim, 2013). Yet, if younger CEOs choose to increase investments in OK to improve their outside career options, this would also be expected to increase information asymmetries associated with intangible asset investments. Accordingly, younger CEOs leading firms with higher OK may be more likely to underinvest because of exacerbated career concerns. Similarly, younger CEOs in high-OK firms could also overinvest relative to older CEOs to signal their talents and capabilities to the market and other external stakeholders (Xie, 2015). Both reasons lead us to predict that CEO age should negatively moderate the relationship between OK and investment efficiency. In other words, we conjecture that firms with higher OK whose CEOs are older tend to have more efficient investments compared to high OK firms with younger CEOs. To examine the moderating effect of CEO age, we construct a dummy variable that takes one if CEO age is above the sample median and zero otherwise. The second column of Table 5 reports the results. We find that the interaction term OKTA 3 CEO age is negative and significant at a 1% significance level (�0.006 with a t-value of�4.17). This finding supports our prediction that CEO age serves as a negative moderator of the relationship between OK and investment inefficiency. 6.2.3 CEO power. Here we explore the potential moderating effect of CEO power on the relationship between OK and investment efficiency. Existing literature suggests that powerful CEOs are more prone to behavioral biases, especially overconfidence, and that this can distort investment efficiency (Aktas et al., 2019; Chowdhury et al., 2023). For example, Chowdhury et al. (2023) find that powerful CEOs are associated with reduced investment efficiency and that this is attributable to a tendency to overinvest. Similarly, Aktas et al. (2019) find that powerful CEOs (those who hold duality of the CEO and chairperson positions in the firm) are more likely to invest inefficiently by overinvesting in low-growth opportunity business segments at the expense of investing in market segments with high growth opportunities. Therefore, since powerful CEOs are prone to overestimating their ability as well as the probability of success associated with their strategic choices, we expect powerful CEOs to make less efficient investments (e.g. Han et al., 2016). To test this prediction, we construct a dummy variable that takes one if the CEO is also the chairperson of the board of directors and zero otherwise (Aktas et al., 2019). The third column of Table 5 reports the result. Interestingly, we observe that the interaction term OKTA 3 CEO power is negative and significant at the 1% significance level (�0.005 with a t-value of�4.69). Therefore, this finding does not support our prediction that CEO power tends to be a positive moderator of the relationship between OK and investment inefficiency. Instead, it infers that firms with higher OK and powerful CEOs make more efficient (i.e. less inefficient) investment decisions. This result may be attributed to the fact that power may improve a CEO’s ability to leverage higher OK to ensure the firm works effectively towards organizational objectives. For example, holding dual CEO and chairperson roles may enable them to take prompt decisions in times of uncertainty and thus enhance the responsiveness of the firm to external environmental changes. Moreover, another reason stems from the fact that powerful CEOs tend to have fewer career concerns due to their influence and control over the board, thereby mitigating CEOs’ natural risk aversion– especially given the presence of heightened informational asymmetries associated with OK investments (Eisfeldt and Papanikolaou, 2013). 6.2.4 CEO tenure. Next, we consider a moderating effect of CEO tenure, a proxy for career experience (Adams, 2005), on the relationship between OK and investment efficiency. Journal of Accounting Literature 463 Consistent with agency theory, existing literature generally supports the idea that long- tenured CEOs make less efficient investments and are more prone to self-maximizing decisions that erode shareholder value (Hope and Thomas, 2008), with recent evidence finding that long tenured CEOs are associated with lower firm values and profitability (Hope and Thomas, 2008; Brochet et al., 2021). Such findings are consistent with the idea that long tenure reflects higher agency costs. One reason being that over time CEOs gain more structural power and influence over the board of directors and firm resources, which helps explain why long- tenured CEOs are less likely to be dismissed for poor performance (Dikolli et al., 2014). Long- tenured CEOs have also been shown to prefer the status quo and to be associated with less risky investment policies (Weng and Lin, 2014). For example, they are less likely to initiate strategic change (Weng and Lin, 2014) and to respond aggressively to external threats to the firm’s environment (Hambrick and Mason, 1984). Thus, they may underinvest in positive NPV projects from the perspective of shareholders. Moreover, they could overinvest in certain types of investments, such as acquisitions. Based on this discussion, we expect tenure to be positively associated with investment inefficiency. Yet we posit that longer tenure in higher OK firms to be associated with improve investment efficiency. One important reason relates to a reduction in information asymmetries and agency costs in higher OK firms, attributable to more knowledge regarding managerial ability. Specifically, OK can promote greater informational disclosures by managers, which help mitigate information asymmetries and reduce agency costs, thereby reducing the likelihood that managerial investment choices deviate from shareholder objectives (Goel et al., 2004; Chemmanur et al., 2009; Attig and El Ghoul, 2018; Garc�ıa-S�anchez and Garc�ıa-Meca, 2018). Moreover, in high-OK firms, longer tenure may allow CEOs to better leverage the value of OK to produce more efficient investments. One reason being that with tenure, managers gain more opportunities to learn and gain valuable firm-specific knowledge, which could improve decision-making quality and increase the value of human capital (Leng and Pan, 2023). Possibly reflecting these arguments,Cain and McKeon (2016) find that long-tenured CEOs are associated with more completed merger and acquisition deals yet lower firm risk. Consequently, we conjecture that tenure interacts with OK to reduce investment inefficiency. To examine the moderating impact of CEO tenure on the relationship between OK and investment efficiency, we include a dummy variable that takes one if the CEO tenure is above the sample median and zero otherwise. Column 4 of Table 5 reports the results of the moderating role of CEO tenure. As predicted, we first find that the singular effect of CEO tenure is positive and highly significant, which indicates that tenure is associated with reduced investment efficiency. Importantly, the interaction term of interest OKTA 3 CEO tenure is negative and significant at the 1% significance level (�0.004 with a t-value of�4.33), which supports our conjecture that firms with higher OK whose CEOs are long-tenured make less inefficient (more efficient) investments. Taken together, since we find that tenure is individually positively associated with investment inefficiency yet the interaction with OK serves to reduce investment inefficiency, our test in this section provides support for the information asymmetry channel. An important takeaway is that OK investments can be valuable from the perspective of shareholders when long-tenured managers are given the scope to leverage OK and wider firm resources in firm environments with higher information disclosures to outsiders, which serve to offset potential agency costs, including a tendency for managers to make self-maximizing decisions (Hope and Thomas, 2008). 6.2.5 CEO connections through external directorships. This section explores the potential moderating effect of CEOs outside directorships on the relationship between OK and investment efficiency. Outside directorships provide CEOs with important social and professional connections, which could allow them access to valuable and unique resources that help improve their job performance (Geletkanycz and Boyd, 2011). Outside directorships may also increase the visibility of CEOs in the executive labor market and, thus, increase their outside career opportunities, thereby mitigating career concerns (Fahlenbrach et al., 2010). This could mitigate the tendency of CEOs to underinvest. Yet, from an agency cost JAL 47,5 464 perspective, external directorships could be associated with reduced investment efficiency if they increase the busyness of the CEO and divert their attention from running their own firm efficiently (Mutlu et al., 2021). Given the tension in the existing literature, we are agnostic with regards to which direction the moderating effect of CEO directorships on the relationship between OK and investment efficiency will take. To examine a moderating effect for CEOs’ external directorships, we construct a dummy variable that equals one if the number of external board membership seats a CEO holds is above the sample median and zero otherwise. From column 5 of Table 5, it can be observed that the interaction term OKTA 3 CEO external directorships is negative and significant at the 1% significance level (�0.006 with a t-value of�3.16). Thus, we find support for the notion that firms with higher OK and whose CEOs have more external board seats tend to make less inefficient (more efficient) investments. This result is consistent with the arguments that external board seats facilitate access to unique resources and can mitigate career concerns that could be associated with suboptimal investment choices. 6.3 Economic channels So far, we have shown that OK is associated with both under- and overinvestment. In this section, we consider, in corporate governance, information environment and CEO career concerns, several viable economic channels through which the negative impact of OK on investment efficiency may be conveyed. Doing so allows us to better identify whether the agency theory perspective or information asymmetry argument drives our main findings. We discuss each in turn and present the results in Table 6. To conserve space, we restrict the tabulated output to our main OK measure: OKTA. 6.3.1 Corporate governance. The fields of accounting and finance have long established that the issues of asymmetric information and agency costs can result in significant distortions away from efficient corporate investment (Jensen, 1986; Agrawal and Knoeber, 1996; Chen et al., 2021) and that corporate governance serves as an important mechanism to mitigate their effects by improving the alignment of manager and shareholder interests (e.g. Chen et al., 2021). Therefore, in this section we consider a potential moderating effect for firm-level corporate governance, which allows us to further explore whether the agency theory perspective drives our main findings. Empirical studies are informative as to how corporate governance arrangements influence corporate investment efficiency. Chen and Chen (2017) demonstrate that firms with stronger internal and external governance mechanisms (i.e. those with more independent and less busy boards, more performance-tied executive compensation, better audit quality, greater institutional ownership, and stronger shareholder rights) make more efficient investments. Similarly, Chen et al. (2021) find that good firm governance is a necessary condition for increased corporate disclosure of project-specific investments to have a positive impact on future investment efficiency, while Rajkovic (2020) finds that boards with greater power in the hands of independent directors and larger boards are associated with more efficient investment. Relatedly, Regier (2023) presents indirect evidence consistent with the notion that stronger overall firm corporate governance, which implies higher levels of monitoring, is associated with more efficient investment. Specifically, they show that an exogenous shock to the duration of executive compensation contracts (towards a long-term focus) can act as a partial substitute to weak corporate governance arrangements, leading to an increase in investment efficiency. Two further studies do not consider governance in isolation but instead combined ESG performance. Bilyay-Erdogan et al. (2024) examine the impact of European firms combined ESG scores on investment efficiency and show that higher ESG performance mitigates overinvestment; furthermore, they find it reduces underinvestment but only in firms with higher information asymmetries. Based on this discussion, we anticipate that firms with both high OK as well as strong corporate governance arrangements will make more efficient investments because effective Journal of Accounting Literature 465 Table 6. Economic channels Panel A: Corporate governance and board attributes Variables Corporate governance (1) Board size (2) Board independence (3) Board gender diversity (4) Intercept 0.018*** 0.012** �0.028** 0.025** (0.018) (0.016) (0.033) (0.050) OKTA 0.009*** 0.014*** 0.007*** 0.005*** (0.001) (0.002) (0.011) (0.026) Corporate governance 0.023*** (0.048) Board size �0.044** (0.058) Board independence �0.008*** (0.037) Board gender diversity �0.024** (0.087) OKTA * CG �0.031*** (0.000) OKTA * Board size �0.265*** (0.003) OKTA * Board independence �0.709*** (0.001) OKTA * Board gender diversity �0.087** (0.031) Control variables Yes Yes Yes Yes Firm FE Yes Yes Yes Yes Year FE Yes Yes Yes Yes R2 (%) 43.29 41.05 29.95 37.51 F-statistics (Wald) 17.58 24.85 26.97 29.49 p-value (0.000) (0.000) (0.000) (0.000) Observations 4,846 4,764 4,808 4,819 Panel B: Other economic channels Variables Information environment (1) CEO career concerns (2) R&D intensity (3) Intercept 0.059** 0.026*** 0.005** (0.038) (0.014) OKTA 0.065*** 0.015*** 0.027** (0.019) (0.006) (0.052) Information environment 0.173*** (0.032) CEO career concerns 0.039*** (0.015) R&D intensity �0.094** (0.082) OKTA * Information �0.011*** (0.018) OKTA * CEO career concerns 0.013*** (0.028) OKTA * R&D 0.044*** (0.071) (continued ) JAL 47,5 466 corporate governance mechanisms should typically result in closer alignment of manager and shareholder interests. In Table 6 Panel A we test this conjecture and examine the moderating effect of corporate governance on the relationship between OK and investment efficiency. We do so by following recent studies (e.g. Regier, 2023) in first focusing on a firm’s overall corporate governance pillar score by collecting the “G” pillar of ESG from the LSEG (formerly Refinitiv) database. This pillar contains three subcategories, namely, shareholders, management, and corporate social responsibility strategy, and measures the efficacy of corporate governance arrangements in terms of board structure, board functions, shareholder rights, compensation policies, vision, and strategy. We then create a dummy variable for corporate governance, CG that equals 1 if the “G” score for a firm is above the sample median and zero otherwise. In Column 1 of Table 6 we then interact this variable with OK (OKTA 3 CG) and observe a negative and statistically significant coefficient on the interaction term at a 1% level (coefficient 5�0.031). Therefore, this result supports the prediction that firms with higher OK and strong corporate governance tend to have less inefficient (more efficient) corporate investments compared to those with weaker corporate governance. This finding adds credence to the idea that good governance is a necessary condition for substantial OK investments to lead to increases in investment efficiency. Table 6. Continued Panel B: Other economic channels Variables Information environment (1) CEO career concerns (2) R&D intensity (3) Control variables Yes Yes Yes Firm FE Yes Yes Yes Year FE Yes Yes Yes R2 (%) 39.40 44.25 47.98 F-statistics (Wald) 9.41 24.94 22.07 p-value (0.000) (0.000) (0.000) Observations 4,865 4,914 4,902 Note(s): This table presents economic channels analyses through which the effect of OK on investment inefficiency are conveyed. The dependent variable, Investment inefficiency is the deviation from optimal investment and is computed as the residuals from the investment model. OKTA is the ratio of organizational capital to total assets and is our main independent variable. Panel A shows results for a corporate governance channel that relates to overall firm governance quality (Corporate governance), which is captured by a dummy variable equal to one if the “G” pillar score (based on the ESG data from the LSEG database) for a firm is above the sample median and zero otherwise, as well as several board attributes: Board size is a dummy variable that equals one if a firm has an above median board size and zero otherwise; Board independence is a dummy variable equal to one if a firm has above the median proportion of independent directors and zero otherwise. Board genderdiversity is a dummy variable equal to one if a has above the median proportion of female directors on the board and zero otherwise. In Panel B, Information environment is a dummy variable equal to one if the quality of corporate disclosure is above the sample median and zero otherwise (disclosure quality is based on stock price synchrony as detailed in Section 6.3.2). CEO career concerns is constructed as a dummy variable equal to one if a firm’s headquarters is in a state adopting the Inevitable Disclosure Doctrine (IDD) and zero otherwise.R&D intensity is a dummy variable that equals one if a firm is a member of an intensive R&D industry and zero otherwise (intensive R&D industries are telecommunications, cable, electronics, pharmaceuticals, chips, computer/peripherals, software, wireless/internet, satellite, tech retail, tech blue chips, and networking, which are identified based on 3-digit SIC codes). Control variables in both panels include size, ROA, slack, tangibility, firm age, leverage, Tobin’s Q, and Loss. All remaining variables are as defined in Appendix 1. Figures in parentheses are standard errors that are clustered at the firm level for heteroscedasticity. (***) (**) (*) significance at the (1%) (5%) (10%) levels Source(s): Authors’ own work Journal of Accounting Literature 467 Second, in order to provide a more comprehensive analysis of the impact of corporate governance on the relationship between OK and investment efficiency, In Panel A of Table 6 we also examine the impact of specific board characteristics on the abovementioned relationship. Specifically, in addition to “CEO power” (or CEO duality) discussed in section 6.2.3, we examine the moderating influence of three board characteristics, namely, board size, board independence, and board gender diversity on the relationship between OK and investment efficiency. We do so, using similar interactive terms between each of the three board attributes with OK. We first observe thatOKTA3Board size is negative and statistically significant at the 1% level, which we argue is consistent with prior findings that larger boards are more effective monitors and are associated with improved investment efficiency (e.g. Rajkovic, 2020). Similarly, in columns 3 and 4, we find that the coefficients for OK 3 Board independence and OKTA 3 Board gender diversity are both similarly negative and highly significant. We consider that the finding for OKTA 3 Board independence provides support for prior literature (e.g. Rutherford and Buchholtz, 2007; Musteen et al., 2010; Chang et al., 2017), which suggests that independent directors play an important positive role in alleviating agency conflicts by improving the monitoring quality. Put differently, less independent boards are less willing to challenge CEOs and other top executives, which increases the scope for managers to pursue behaviors and make decisions not compatible with shareholder objectives, resulting in lower investment efficiency. Further, the finding for OK 3 Board gender diversity is consistent with earlier findings in the literature (e.g. Upadhyay and Zeng, 2014) that firms with more gender diversity tend to enjoy better information environments due to better monitoring, which we show results in higher investment efficiency. 6.3.2 Information environment. Next, we develop a proxy for the information environment within the firm. Prior literature finds that improved disclosure quality enhances the overall information environment of the firm (Byard and Shaw, 2003) by reducing information asymmetries between management and shareholders, and, in turn, agency costs (Brown and Hillegeist, 2007). Consistent with this, several studies find that increases in disclosure quality are associated with a lower cost of firm capital (Lambert et al., 2007), increased analyst coverage (Sundgren et al., 2018), improved market liquidity (Heflin et al., 2005), and ultimately more efficient corporate investments (Biddle and Hilary, 2006; Gomariz and Ballesta, 2014; Chen et al., 2021, 2023). Conversely, improved disclosure may not necessarily translate into improved investment efficiency. For example, Dutta and Nezlobin (2017) find that while improved disclosure of a firm’s future capital stock can reduce managerial incentives to underinvest, improved disclosures of a firm’s future cash flows can increase the incentives for managers to under- or overinvest. Although we acknowledge that increases in the quality of information disclosure may not always be associated with more efficient corporate investment (Dutta and Nezlobin, 2017), the general consensus in the literature is that informationally transparent firms make more efficient investments. Therefore, we conjecture that firms with high-OK and with high levels of disclosure quality will make more efficient (less inefficient) corporate investments. We define the information environment as the total quantity of idiosyncratic information about a firm made available to investors (Biddle et al., 2009; Myers and Majluf, 1984; Zhang, 2006). Firm-specific information may be provided by the firm through mandated or voluntary disclosures, market intermediaries such as analysts and the media, and investors. It is spread through public and private channels, as well as price discovery. Although price-based measurements do not identify the source, route, or content of such information, they capture the whole of the information in the environment (Piotroski et al., 2015). The quality of information disclosure has been a key focus of post-Enron and post-GFC regulatory endeavors to enhance financial market stability, quality, and efficiency. In this vein, as discussed earlier within Section 2, the Dodd–Frank Act of 2010 and the Sarbanes–Oxley Act of 2002 aim to determine the main ingredients of quality corporate disclosure to safeguard investors by enhancing the reliability and accuracy of corporate disclosures with respe