Kibria, Ishrar
(2023)
CEO overconfidence and corporate policy.
PhD thesis, University of Nottingham.
Abstract
The PhD thesis consists of three individual but related pieces of research. We extend the literature of CEO overconfidence influence in corporate strategies and provide latest findings on R&D investment, corporate governance structure and HR policy. This PhD thesis adopts empirical method of research, and the results are validated by several robustness checks.
Strategic decisions, according to Hambrick and Mason (1984), reflect senior managers' attitudes and intellectual prejudices. They present the "upper-echelons" theory, which concludes that managerial characteristics, such as those of CEOs, can impact business decisions and organizational performance outcomes (Hambrick and Mason, 1984; Hambrick, 2007). This theory hypothesise that CEOs are boundedly rational. Their personality and ideals are fairly stable, and their choices are inextricably influenced by their personal characteristics (Finkelstein, 1992). Several studies based on upper echelons theory find evidence that CEOs' background characteristics, personality traits, and leadership styles have a significant impact on corporate strategic dynamism (Crossland et al., 2014), R&D spending (Barker and Mueller, 2002), mergers and acquisitions (Yim, 2013), corporate risk-taking propensity (Serfling, 2014), and firm profitability (Wang et al., 2016). A growing field of research demonstrates that a substantial number of senior business executives suffer from overconfidence in their judgements. The willingness of executives to keep their personal wealth undiversified by holding stock options even if they are substantially in-the-money, has been the key measure of executives’ overconfidence. The presence of CEO overconfidence, that is, the belief that the firm stock price should be greater than it is, seems to matter for a number of corporate strategies. As documented in the literature, because of the significant impact of CEO and executives’ attributes over firm policies, this thesis particularly considers the impact of C-suit Executive Overconfidence on corporate policies. Overconfident CEOs are inclined to inflate their knowledge and control, overestimate the predicted returns, and undervalue the risk of an investment. Overconfident CEOs are described as those who underestimate the volatility of earnings (Hackbarth, 2008) and firm assets (Hackbarth, 2009); or overvalue predicted investment proceeds (Malmendier, Tate and Yan, 2011) and the accuracy of their knowledge (Gervais, Heaton and Odean, 2011). Different aspects of overconfidence are studied in empirical research (Oskamp, 1965; Weinstein, 1980; and Puri and Robinson, 2007). Moore and Healy (2008) find that literatures are based on three discrete explanations of overconfidence: overestimation (64 percent), over-precision (31 percent) and over-placement (5 percent). Since overconfidence increases alongside the corporate ladder (Camerer and Lovallo, 1999), CEOs and Board members are arguably most susceptible to this bias. We use the overexposure of executives to firms' idiosyncratic risk to construct measures of overconfidence. Executives have substantial stock and option grants embedded in their compensation plan. The stock options are vested, and the firm only allows trading after the period is over. These restrictions make the executives portfolio highly undiversified and risky. As soon as the vesting period is over, a risk-averse executive should exercise their options if the stock price is sufficiently high (Lambert, Larcker and Verrecchia, 1991; Meulbroek, 2001; Hall and Murphy, 2000, 2002). However, some executives persistently choose to hold their stock options because they believe the stock price will increase more in the future. We interpret this behaviour to measure overconfidence following Glasso and Simmcoe (2011) and Malmendier and Tate (2005a, 2008). An overconfident executive takes the value 1 if s/he postpones the exercise of 67 percent in-the-money stock options, and 0 otherwise. Using this technique, if the executive displays overconfidence at least twice, the individual is considered as overconfident for the rest of their tenure. We apply Campbell et al. (2011) for calculating the average moneyness of the executive's option portfolio. For each CEO-year, the average realizable value per option is calculated by dividing total realizable value of the options by the number of options held by him. The strike price is calculated as the fiscal year-end stock price minus the average realizable value. The average moneyness of the options is then calculated by dividing average realisable value by strike price. Majority of the past research on executive’s overconfidence, for instance, Malmendier and Tate (2005a, 2008), Glasso and Simmcoe (2011) and Hirshleifer et al. (2012), followed the same approach.
There is a substantial amount of literature that examines the connection between CEO overconfidence and various corporate strategies. Literature document the impact of CEO overconfidence on dividend policy (Deshmukh et al., 2013), investments strategies (Cronqvist et al., 2012; Hackbarth, 2009; Huang and Kisgen, 2013), mergers and acquisitions (Malmendier and Tate, 2008; Yim, 2013), financial risk (Malmendier and Nagel, 2011) and corporate governance (Kim and Lu, 2011, Van den Steen, 2004).
Our first paper extends the work of Francis, Hasan and Sharma (2011) (finds increased innovation with CEO overconfidence), Galasso and Simcoe (2011) (finds 25%-35% more patent citation with CEO overconfidence) and Hirshleifer et al. (2012) (finds increased innovation and patent citations with CEO overconfidence); however, their findings are limited to the year 2003. Also the oscillations of this associations during economic fluctuations have not been studied yet. Our paper fills this gap and document the shifting relation between CEO overconfidence and R&D expenditure during the 2000-01 dot com crash and 2007-08 Global Financial Crisis (GFC). Financial crises limit firms' access to capital, reducing their ability to borrow, causing to cut back on their investment in innovative ventures. We look further to uncover the underlying mechanism and conclude imposed corporate governance as the cause of this shift. Sierra-Morán et al. (2021) and Hoskisson et al. (2002) find negative association between directors who hold more shares and innovation. We complement these findings and conclude that unlike overconfident CEOs, an overconfident board invests less in R&D, more so after the 2000-01 dot com crash and furthermore after the 2007-08 GFC. Due to regulatory enforcements in the post-crash periods (Sarbanes–Oxley Act of 2002, Emergency Economic Stabilization Act of 2008, Wall Street Reform and Consumer Protection Act of 2010 (Dodd-Frank)), the overconfident board intervenes more in overseeing firms’ strategic decisions, which diminishes the positive link between CEO overconfidence and R&D. With empirical support, we take out financial constraints as a potential explanation for reduced R&D spending in post crisis period. We document the novel paradigm shift in corporate governance from CEO to board directors when deciding for R&D investment.
The second paper is developed upon the findings from our first work and looks into the structural formation of corporate governance for firms with overconfident CEOs. We take the discussion further and segment our research question into internal and external components of board structure. For internal element of corporate governance, we look into Board Size, Board Independence (absence of CEO duality), Board Age Demographic and Presence of Co-opted Directors in the Board. For external corporate governance, we investigate the presence of Blockholders and nature of auditor in the firm. We conclude that Board structure of Overconfident CEOs is smaller, less independent and have more younger directors. They also have higher co-opted directors (directors who are hired after the CEO took office) than firms lead by non-overconfident CEOs. Smaller boards are likely to be intimidated by the CEO, lack of board independence gives CEOs more power, younger directors are typically more risk taker and co-opted directors are likely to assign their allegiance to the CEO. These characteristics of the board implies that overconfident CEOs prefers to lead the board and have control over firms’ decisions; and hence likes to keep less crowded structure. For external mechanism of corporate governance, we find lower shareholding and number of affiliated and outside Blockholders in firms lead by overconfident CEOs. Blockholders are large investors (holding more than 5% share of the firm) who have the resource and motivation to implement corporate scrutiny; and having lower concentration of them certainly leaves more power for the CEO. Lastly, firms with overconfident CEOs have lower probability to hire Big 4 or Big 5 accounting firms as their financial auditor. Large audit firms perform quality audits and mitigates managements propensity to interfere with the audit process. By not choosing them, overconfident CEOs retain control over firm financials. Our findings point out the preference of overconfident CEOs for a less forceful board structure and highlights the difference of governance compared to non-overconfident CEOs. There is not enough definite research on the impact of CEO overconfidence on governance processes (Parker, 2017). Through the findings, we are able to clear the ambiguity surrounding corporate governance composition and clarify the unorthodox corporate policies of overconfident CEOs (Malmendier et al., 2011).
Our third paper directly answers to the request for additional research into the role of internal corporate governance, particularly CEOs, in defining HR systems and their impact on an organization's people (Mullins, 2018; Wood and Budhwar, 2021); by looking into the hiring-firing patterns of overconfident CEOs. We apply different measures of net hiring, such as the Employment Growth (which takes the value 1 if net hiring is greater than 1 percent, 5 percent, 10 percent, 15 percent, and 20 percent), and the Employee Reduction (which takes the value 1 if net hiring is less than 0 percent, -1 percent, -5 percent, -10 percent, -15 percent, and -20 percent). We use a number of control variables, year and industry fixed effects in our testing to control for time-invariant industry level effects. We conclude that overconfident CEOs have higher job growth and lesser downsizing in their companies. To assess the hiring pattern during industry expansion-contraction, we use four methods of expansion and contraction, along with their CEO overconfidence interaction as a control to determine the impact.
First, we use industry expansion, which takes the value 1 if the industry's sales median over the last three years is greater than zero. We find that the interaction coefficient for overconfident CEOs is significantly positive for net hiring and employment growth, and significantly negative for Employee Reduction. Second, we employ rolling cash flow volatility, which calculates the standard deviation of industry cashflow over the previous five years. With the industry's high cash volatility, we find that overconfident CEOs hire more and fire less than their counterparts. Third, we divide the sample into three equal sub samples using the firm Kaplan Zingales index. We ran separate analyses and argue that when a company's finances are tight, CEO overconfidence has a greater influence and is a substantial positive for employment growth. Lastly, we split the sample into three groups based on the total asset value and discover that overconfident CEOs are more likely to have higher employment growth in smaller enterprises. They do, however, have more recruitment in three sub-samples, but have higher influence in smaller businesses. Our findings contribute to the ongoing discussion on the value implications of over-hiring and under-firing decisions, alongside CEO risk preferences. According to certain studies, when CEOs choose the quiet life (Bertrand and Mullainathan, 2003), engage in empire building (Pagano and Volpin, 2005), or seek private rewards associated with creating relationships with people, sub-optimal investments in human resources are destructive (Atanassov and Kim, 2009; Landier et al., 2009). However, the actions of overconfident CEOs may add value, not only because retaining workers during a bad economy saves firms money in employment adjustment costs, retains firm-specific human capital, and boosts morale; but it also allows firms to be prepared for future growth opportunities, which is consistent with the concept of maintaining optimal investment in employees (Falato and Liang, 2016; Ellul et al., 2018).
Our findings reveal novel relationship between CEO overconfidence and various corporate and governance policies. We contribute to the literature an empirical framework to help integrate and evaluate the growing research on CEO overconfidence. Our first paper adds to the literature on CEO overconfidence by presenting the most recent research on the shifting relation between CEO overconfidence and innovation. Ours is one of the first efforts to look at both the 2000-01 dot com bubble crash and 2007-08 GFC and discover increased board dominance in post crisis era. This strengthen governance is the evidence of successful incorporation of Sarbanes–Oxley Act of 2002, Emergency Economic Stabilization Act of 2008, Wall Street Reform and Consumer Protection Act of 2010 (Dodd-Frank) and gives assurance to investors about prevention of another crisis due to poor monitoring. Investors rights are now stronger due to board intervention and close scrutiny. On another view, the vibrant and positive relationship between CEO overconfidence and innovation between crashes relevant to investors who are interested in investing in innovative and growth firms. Investors are now better-off and well-guarded with stronger governance in firms headed by overconfident CEOs.
Taking the study further, our second paper investigates the internal and external determinants of corporate governance for overconfident CEOs, particularly focusing on board structure. Boards have substantial impact on shaping an organizational culture and governance (Lightle et al., 2009). Overconfident CEOs have unique board formation which could help in explaining their non-conventional corporate strategies. Our findings suggest that strong internal control may limit CEO decision-making ability, as since firms decision making process is highly standardized. However, if a company's internal control is inadequate, it may result in a rise in agency costs, impeding innovation's efficiency and productivity. Our findings add to the understand of the firm's central agency problem, specifically the separation of ownership and control. Equity based compensation might help to minimize typical CEO incentive misalignment but fails to reduce agency cost. These findings aid investors with a probable reasoning on why firms with overconfident CEOs might incur reduced agency costs due to CEO power dominance, however there prevails a probability of more risky investments and skewed earnings manipulation.
Our third paper contributes to the expanding literature about company HR policy, and we investigate how CEO overconfident impacts firms hiring-firing decisions. We show that overconfident CEOs hire more employees and have a positive employment growth over time when using multiple hiring and firing criteria. We also find the increased influence of CEO overconfidence for smaller and financially constrained firms. Retaining workers in the bad time saves employment adjustment costs, retains firm-specific human capital and boosts morale. Investors of such firms could expect to be better prepared for future growth opportunities. In this context, CEO’s preferences could become more closely aligned with those of the employees and of the shareholders.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
![[thumbnail of Ishrar_Kibria_ID_14342495_PhDThesis.pdf]](https://eprints.nottingham.ac.uk/style/images/fileicons/application_pdf.png "Ishrar_Kibria_ID_14342495_PhDThesis.pdf")