Three essays on corporate governance and narrative disclosure

• Autores: Cristina Grande Herrera
• Directores de la Tesis: Beatriz García Osma (dir. tes.), Encarna Guillamón Saorín (codir. tes.)
• Lectura: En la Universidad Carlos III de Madrid ( España ) en 2019
• Idioma: español
• Tribunal Calificador de la Tesis: Ane Tamayo (presid.), Juan Manuel García Lara (secret.), Andrew J. Leone (voc.)
• Programa de doctorado: Programa de Doctorado en Empresa y Finanzas / Business and Finance por la Universidad Carlos III de Madrid
• Materias:
  • Ciencias económicas
    • Contabilidad económica
      • Contabilidad financiera
• Enlaces
  • Tesis en acceso abierto en: e-Archivo
• Resumen

  • This Thesis contains three studies related to corporate governance and narrative disclosure. In Chapter 1, we study whether new CEOs engage in qualitative strategies to extend their tenure. In Chapter 2, we analyze a friction in the managerial labor market. In particular, we study the relationship between a decrease in the pool of replacement CEOs, the incumbent CEO level of entrenchment, firms’ financial reporting quality and narrative disclosure. Finally, in Chapter 3, we study the effect of hostile takeover susceptibility on narrative disclosure.

    The concept of Corporate Governance can be defined as the set of rules applied to control and lead a company. Corporate boards have been extensively studied in the literature. Every company has a board of directors and many studies want to discern whether differences across boards can explain firms’ behaviour and performance (Hermalin and Weisbach, 2010). The main goals of corporate boards are to advise and monitor management. One of the duties of the board is to appoint the new CEO and decide about dismissing the incumbent one in case, for instance, of poor performance.

    On the other hand, managers can decide about the narratives they use in their 10-K reports. Narrative disclosure is an efficient way to disclose relevant firms’ information (Merkley, 2014). There is an increasing stream of literature showing that not only quantitative but also qualitative information is relevant for investors and has relevant economic effects (Frazier et al., 1984; Gibbins et al., 1990; Tetlock, 2007; Tetlock et al., 2008; Feldman et al., 2010; Huang et al., 2014b).

    When a new CEO is appointed, the board applies its monitoring role to assess his or her ability using all the available information (both quantitative and qualitative). This ability assessment is an important component of corporate governance (Hermalin and Weisbach, 2017).

    Prior literature confirms that managers worry about how firm performance influences the labor market assessment of their managerial skills and, also, that they are willing to take real and reporting decisions to improve this perception and construct good reputations (e.g., Nagar, 1999; Ali and Zhang, 2015). Linking with this idea, in Chapter 1, Optimistic Disclosure Tone and CEO Career Concerns, we argue that CEOs use narrative disclosure tone to assuage career concerns. In particular, we predict that more talented CEOs use a more optimistic tone at the beginning of their tenures to build a reputation of strong performance ability. Ali and Zhang (2015) show that newly appointed CEOs are likely to engage in earnings management activities (i.e., quantitative strategies) given their career concerns. Our argument is that new CEOs are likely to engage not only in quantitative but also qualitative strategies to extend their tenure. We argue that highly able CEOs anticipate the importance of narrative tone and use it in response to career concerns, to improve the beliefs of third parties concerned with CEO assessment. High ability CEOs differently use tone to signal their ability and superior future performance for at least two reasons. First, the use of optimism in narrative disclosures increases litigation risk (Rogers et al., 2011), acting as a deterrent to engage in cheap talk. Second, optimism in disclosure unravels, i.e., it is soon revealed to be either informative or opportunistic. This is because managerial disclosures trigger additional information searches by analysts and other market participants (Barron et al., 2002). CEOs who fail to meet their voluntary disclosures may damage their reputation and the firm’s image (Ferreira and Rezende, 2007), leading to a loss of investors’ confidence (Graham et al., 2005). These CEOs would self-reveal as the low ability ones. Given these costs, low ability CEOs are unlikely to imitate high ability CEOs’ narrative practices. In line with this view, Nagar (1999) shows that managerial uncertainty about performance evaluation can deter managerial disclosure, and that it is the ‘not very talented’ managers that are particularly prone to nondisclosure.

    Thus, while high ability CEOs could use optimistic tone to assuage career concerns, the opposite may hold for low ability CEOs. Without considering CEO ability or tenure, Huang et al. (2014a) show that optimistic narrative disclosures relate, on average, with managerial attempts to mislead market’s perceptions of future firm results and hide poor future realizations. Assuming that optimistic disclosures attract market attention and increase litigation risk (Rogers et al., 2011), new CEOs may use lower optimistic tone to avoid excessive market attention, particularly, given the evidence that these CEOs report inflated earnings via the use of income-increasing accruals (Ali and Zhang, 2015). Also, new CEOs may prefer to use big bath strategies, and blame their predecessors for any initial poor firm performance. This would lead to negative disclosure tone in CEO early years of tenure. Given the above discussion, we predict that high (low) ability CEOs use greater (lower) optimistic tone early in their tenures. As CEO tenure progresses, managers develop a reputation and, absent any shocks, monitoring decreases (Dikolli et al., 2014; Pan et al. 2015), as they are expected to maintain or improve their knowledge and skills (Wu et al., 2005). Thus, optimistic tone is less useful after the early years. The findings in Ali and Zhang (2015) that CEOs do not engage in earnings management in the middle years of their careers is consistent with this attenuation in career concerns. Regarding the final year of tenure, Brickley et al. (1999) show that strongly-performing departing CEOs are more likely to serve in their own board or join other boards. Monetary post-retirement benefits may also hinge on late career performance (Kalyta, 2009). Thus, CEOs may face career concerns at this late stage, if they stay active in the job market. However, CEOs may be entrenched by this point, making such strategies redundant. In line with this later argument, Holmstrom (1999) analytically demonstrates that managers work harder in the first years of tenure than in the last one. In addition, it is unclear whether CEOs can perfectly foresee which will be their last year in the job and thus, disclosure tone in the last year of tenure is an empirical question of interest.

    We analyze the links between CEO ability, CEO career concerns, and tone in 10-K reports using a large sample of US firms for the period 1993 to 2013. To conduct our tests, we follow the approach of Ali and Zhang (2015), who study the links between CEO career concerns and earnings management. We follow Huang et al. (2014a) to calculate optimistic disclosure tone, and condition our analyses on managerial ability as measured by Demerjian et al. (2012). Our results show that highly skilled CEOs use more optimistic tone, particularly, as a response to career concerns. For our sample of CEOs, optimistic tone is linked to higher future access to debt, greater future capital investment and dividend payments. Overall, this suggests that CEOs use optimistic tone to convey their true managerial ability. We validate that our CEO ability proxy measures management skills by showing that CEO ability is negatively associated with the probability of forced turnover. Forced turnover is also less probable for CEOs with more optimistic disclosure tone. Overall, our evidence is consistent with CEOs influencing firm narratives, and with CEOs narrative disclosures being driven, at least partly, by their career concerns. Thus, we show that skilled CEOs engage in qualitative strategies (i.e., narrative disclosure) at the beginning of their tenure to ensure long tenures.

    When the board needs to hire a new CEO, the options are choosing an internally or externally appointed CEO. Boards need to look at the pool of potential new CEOs which includes every insider and outsider who has the required abilities to be the next CEO of the firm. Donatiello et al. (2018) run a survey to directors of the largest US companies and find that 73% of the interviewed directors agree on that less than 5 people (including both insiders and outsiders) have the required abilities to be the next CEO. Then, it is fair to assume that any shock that decreases the pool of replacement CEOs is likely to affect firms. In this line, in Chapter 2, CEO Labor Market Incentives and Accounting Quality: The Unintended Consequences of Trade Secret Regulation, we examine the impact of a decrease in the pool of replacement CEOs on the incumbent CEO level of entrenchment and firms’ financial reporting quality. By integrating the staggered enactment of the Uniform Trade Secrets Act with the pre-existing pool of talent, we develop a novel firm and time specific measure of changes in the pool of replacement CEOs.

    UTSA aims to protect firms’ competitive advantage, by means of protecting their proprietary information, i.e. trade secrets, from rivals (Barney, 1991; Grant, 1996; Kogut and Zander, 1992). UTSA lowers uncertainty on the legal protection afforded to trade secrets, limiting information misappropriation (Samuels and Johnson, 1990). Trade secrets -commonly referred to as the jewel crown- (Jorda et al., 2007; Castellaneta et al., 2017), are an important source of firm risk and, if disclosed, can lead to significant impairments in competitive advantage and economic losses (Klasa et al., 2018), which have been estimated to be as high as $50 billion annually (PriceWaterhouseCooper, 2002). Given their economic relevance, it is not surprising that trade secrets litigation is on the rise both in state and federal courts (Almeling et al., 2010, 2011).

    Although the main goal of UTSA is to protect firms’ competitive advantage, we argue that it impacts an important labor market institution: the pool of replacement CEOs. This is because after the enactment of UTSA firms can more easily litigate against top management team members that disclose firm trade secrets, as well as against any firm that hires these departing executives. Thus, greater trade secrets protection reduces the mobility of incumbent CEOs and of other top executives (supply side), as well as lowers the probability that they receive offers from other firms (demand side). Overall, this means that by increasing litigation risk for managers in possession of trade secrets and proprietary information, UTSA may reduce both the availability and attractiveness of labor-market opportunities (Castellaneta et al., 2017). To test our predictions and shed light on these contrasting views, we create a firm-specific, time-varying measure of the annual decrease in the pool of replacement CEOs by integrating the staggered enactment of UTSA with the existing within-industry pool of potential new CEOs (Pool Decrease). In particular, our proxy captures, for each firm, the annual percentage of the within-industry pool of talent that is impaired, i.e. of firms belonging to the same industry that are incorporated in states with UTSA. This percentage is then multiplied by the quartile of one over the total pool of top management team members available in the industry. This measure follows the quasi-experimental shift share research designs used in previous literature (Borusyak et al., 2018). The intuition underlying Pool Decrease is that a firm experiences a stronger decrease in its potential pool of replacement CEOs as more firms in the same industry are incorporated in states where UTSA is enacted, particularly, when the number of individuals forming the available pool of talent is small. The use of the quasi-natural experiment provided by the staggered adoption of UTSA in 48 U.S. states and the District of Columbia, reduces endogeneity concerns as (1) firms cannot control the state of incorporation of other firms in their industry; (2) UTSA was enacted by policy-makers without considering the specific economic and political situation in each state (Png, 2017); and (3) the focus of the study is on the unintended consequences of UTSA over managerial labor markets.

    We implement a difference-in-difference research design and use a large sample of U.S. firms from the period 1980 to 2016. We find that decreases in this pool lead to longer tenure, lower forced turnover, and higher compensation for incumbent CEOs, as well as lower financial reporting quality and narrative disclosure quality. Decreases in the pool of replacement CEOs are also associated with lower CEO-firm match, lower firm efficiency, lower performance and higher over-investment. The results are robust to alternative measures of decreases in the pool of talent and to controlling for additional trade secrets protection measures. Our collective findings indicate that labor market institutions are important drivers of firm outcomes and accounting quality.

    Corporate boards have also a role in takeover situations. During the early 1980s there were several waves of hostile takeovers affecting even to the largest US companies (Gompers et al., 2003). As a result, firms and states implement antitakeover provisions and laws, respectively. In Chapter 3, Takeover Protection through Narrative Disclosure, we assess the effect of hostile takeover susceptibility on narrative disclosure. We predict that firms use narrative disclosures as a takeover defense mechanism. In particular, we expect that managers use more pessimistic tone in 10-K reports to lower firm visibility and drive away bidders’ attention, protecting themselves from takeovers. To assess whether pessimistic language acts as a defense mechanism, we study the use of negative tone, as well as analyze abnormal negative tone, which we denote pessimism. Pessimism therefore means use of negative tone beyond what would be expected given the firm’s fundamentals (such as performance, risk or complexity). This focus on negative and pessimistic tone allows us to contribute to prior work, as the effects of negative disclosure are scarce (e.g., Huang et al., 2014b), as prior literature usually focuses on the strategic use of positive rather than negative disclosure (e.g., Huang et al., 2014a; Bochkay et al., 2018).

    US companies experienced several waves of hostile takeovers in the early 1980s. At the time of this heightened takeover environment, many rules were enacted at the state and firm level. The main goal of these rules was to protect managers from unexpected takeovers. Antitakeover provisions have been widely studied by previous literature. Previous studies analyze how antitakeover provisions affect managerial preferences and corporate governance (Bertrand and Mullainathan, 1999; 2003), firm value (Gompers et al., 2003) or shareholder wealth (DeAngelo and Rice, 1983). However, to the best of our knowledge, there is no previous study analyzing whether managers use more negative tone in their disclosures when confront higher probability of experiencing a hostile takeover. This is, whether managers use narrative disclosure to protect their companies from unwelcome bids. Takeover protection effectiveness depends not only on the type of protection adopted but also on the investors’ view of firms’ managers (Coates, 2000).

    Understanding the consequences of higher probability of hostile takeovers and, thus, more antitakeover provisions (i.e., higher takeover protection) on narrative disclosure is important, as narratives are an efficient tool to disclose relevant information (Merkley, 2014) and have an impact on investors’ decisions (Tetlock, 2007; Tetlock et al., 2008). Indeed, a growing literature shows that market participants consider not only firms’ quantitative information but also its qualitative disclosures, and provides mounting evidence that narratives have economic consequences (Frazier et al., 1984; Gibbins et al., 1990; Tetlock, 2007; Tetlock et al., 2008; Feldman et al., 2010; Huang et al., 2014b). Companies disclosures are useful to stakeholders with different interests in the firm such as investors who want to discern their investment opportunities or financial analysts who issue their buy or sell recommendations. There is a debate in the literature about whether firms should issue accurate information to attract resource providers or whether issuing valuable information may attract rivals (Darrough and Stoughton, 1990; Verrechia, 1983) or potential acquirers. Antitakeover provisions have the main goal of making the firm unattractive to potential unwelcome bidders. Some provisions such as poison pills or pension parachutes make the target less attractive to the acquirer. Other antitakeover provisions such as fair price or silver parachutes increase the acquisition price. Director duties, unequal voting, supermajority, written consent, special meeting, black check and staggered boards complicate that the bidder can acquire the control over the target company. Then, our main prediction is that a higher probability of confronting a hostile takeover is likely to increase the negative tone and pessimism used by managers in their 10-K disclosures. More negative tone and pessimism in firms’ narratives is likely to dissuade potential acquirers as the firm looks less attractive.

    To test our predictions, we use the Cain et al. (2017)’s takeover index which is referred as Hostile Takeover in our study. The authors construct the takeover index using first the state-level variation of takeover activity which is plausibly exogenous to discretional firm decisions. After analyzing which are the relevant takeover laws and court cases, they construct a firm-level index of hostile takeover susceptibility. To measure negative tone, we use the Huang et al. (2014a) proxy of raw disclosure tone to construct our Negative Tone variable. The residual from the Huang et al. (2014a)’s model is our abnormal pessimistic disclosure variable (Pessimism). Our sample ranges from 1994 to 2013.

    Our results show that managers of firms with higher likelihood of receiving an unwelcome bid have more negative and pessimistic tone in their 10-K reports. We also find that our main results are stronger for firms more attractive for potential acquirers and that more pessimistic firms in hostile environments are less related with new acquisition announcements. Our results are robust to a shock that represents lower need of firm-initiated antitakeover provisions.