Abstract
A significant amount of research has examined firms’ decisions to adopt poison pills; however, firms today are increasingly repealing or allowing poison pills to expire. Based on agency theory, the authors examine competing perspectives of governance mechanisms as having complementary or substitutive effects within the context of poison pill repeal. They test whether firms repeal poison pills when governance is strong (complementary effects) or allow for other governance mechanisms to compensate for potential agency costs associated with poison pill renewal (substitutive effects). Using a sample of 288 firms who made decisions to terminate or renew poison pills, the authors find that firms with CEO duality, fewer directors nominated by the CEO, and higher levels of outside director ownership and pressure-resistant institutional shareholdings are more likely to repeal poison pills. A curvilinear relationship between managerial ownership and poison pill repeal is also found. The results provide greater support for the notion that firms use governance mechanisms as complements rather than substitutes.
Following a significant wave of mergers and acquisitions in the 1980s, firms began to adopt a number of antitakeover provisions; however, in recent years firms have begun allowing such provisions to expire. Early research on antitakeover adoptions attempted to explain why firms would choose to adopt such provisions and what were the consequences (Bebchuk, Cohen, & Ferrell, 2009; Brickley, Coles, & Terry, 1994; Mallette & Fowler, 1992; Sundaramurthy, Mahoney, & Mahoney, 1997). The recent wave of repeal of one such provision, the poison pill, has led to a gap in the literature to examine why firms are now choosing to repeal a prior decision. This article addresses this gap by examining governance characteristics of firms who must decide whether to repeal or renew such provisions, and whether such firms see governance mechanisms designed to limit managerial opportunism as serving as complements to or substitutes for other mechanisms.
The poison pill, contingent securities that are triggered in the event of the firm becoming a takeover target that lead to unwanted obligations on the acquirer’s end (Mallette & Fowler, 1992), has been identified as one of the most effective antitakeover mechanisms. Poison pills lead to the dilution of shareholder equity, which significantly reduces the likelihood of acquisition. Advocates against poison pills (e.g., Mahoney & Mahoney, 1993; Malatesta & Walking, 1988) argue that they are harmful to shareholders, as they can insulate top managers from competition and reduce the effectiveness of the market for corporate control (Pound, 1987), which serves as a mechanism to punish managers whose interests are not aligned with shareholders (Fama, 1980). However, others (e.g., Heron & Lie, 2006; Jennings & Mazzeo, 1986) argue that poison pills enhance shareholder value by increasing premiums paid to shareholders in the case of acquisition or by fending off corporate raiders looking to buy an undervalued firm.
Recent changes in the institutional environment have led to firms repealing or not renewing poison pill provisions at a much higher rate (Laide, 2008). Two perspectives exist as to why poison pills may be renewed or repealed by boards. One perspective, that governance mechanisms can be complementary in effects, argues that firms should bundle as many governance mechanisms as possible to minimize agency problems. Agency theory (e.g., Jensen & Meckling, 1976) argues that agents (e.g., managers) will pursue their own interests rather than those of the principals (e.g., shareholders), unless they are properly monitored or their incentives are aligned with the principals’. The perspective of governance mechanisms as complements thus argues that firms should be less likely to retain poison pills for two reasons. First, prior research shows that poison pills harm shareholder value, primarily through a negative stock market reaction (e.g., Bebchuk et al., 2009; Jarrell & Poulsen, 1987), resulting in agency costs to shareholders. Second, poison pills limit the effectiveness of the market for corporate control (Walsh & Seward, 1990), resulting in less monitoring of managerial behavior. Research on poison pill adoption has shown that organizations with strong monitoring and incentive alignment are less likely to adopt poison pill provisions (e.g., Mallette & Fowler, 1992).
An alternative perspective also exists, arguing that firms use governance mechanisms not in combination, but as substitutes, in order to reduce agency problems (Rediker & Seth, 1995; Zajac & Westphal, 1994). This perspective notes that the use of governance mechanisms is costly and that use of multiple mechanisms may be unnecessary. Consistent with this argument, prior research shows that stock market reactions to antitakeover provision adoption are more favorable when evidence exists that monitoring of managerial behavior is effective (e.g., Brickley et al., 1994). From this perspective, adoption of a poison pill does not increase agency costs to organizations when monitoring and incentive alignment are strong. Instead, poison pills may enhance shareholder value through increased premiums received if the firm is acquired.
Based on these competing perspectives, the questions this study asks are the following:
Do firms with strong monitoring and incentive alignment repeal poison pills as part of a pattern of strong corporate governance (complementary perspective), or do firms with strong monitoring and incentive alignment prefer to keep poison pills in place as monitoring and incentive alignment provides protection for shareholders (substitutive perspective)?
To date, research has not yet examined the reasons for boards choosing to repeal poison pills. This decision is important, as poison pills represent one of the most effective antitakeover provisions a firm can adopt (Mallette & Fowler, 1992; Walsh & Seward, 1990). It also allows for modeling how firms, which vary on a variety of continuous governance characteristics, engage in a dichotomous decision. This study examines the governance characteristics of 288 firms whose poison pills were set to expire to predict whether firms will renew poison pills.
This study makes several interesting and important contributions to the existing literature on antitakeover defenses. First, this study develops hypotheses based on the competing perspectives of governance mechanisms as substitutes or complements in order to test whether firms renew or repeal poison pill provisions when other governance mechanisms are in place at the time the decision is made. We find support primarily for the notion that governance mechanisms have complementary effects with less support for the notion of substitutive effects. Second, this study examines whether corporate governance characteristics that lead to poison pill repeal are the reciprocal of those that lead to pill adoption. Findings inconsistent with previous research on pill adoption may illustrate a separate pattern of governance. If an inconsistent pattern exists, future research can attempt to explain why shifts have occurred in corporate governance and board decision making from the time that poison pills were adopted to the time of repeal. Third, this study utilizes a unique data set of firms who face a decision as to whether to renew or terminate the poison pill provision, which has not yet been examined in the literature. Our data set provides a unique view into a firm’s decision, as it specifically examines the point at which firms make decisions on whether to renew a poison pill.
Theory and Hypotheses
Theories on Poison Pills and Their Effects on Shareholder Value
Among potential antitakeover defenses that may be adopted, the poison pill has proven to be one of the most effective. Two competing perspectives exist as to whether poison pill provisions are beneficial to shareholder value. The management entrenchment hypothesis (e.g., Walsh & Seward, 1990) argues poison pills are detrimental, as they insulate managers from the market for corporate control (e.g., Bebchuk et al., 2009). A poison pill reduces the chances of an unwanted takeover and thus the probability that managers will be removed for incompetence, shirking, or self-serving behaviors (Walsh & Seward, 1990). Such insulation allows for managers to make decisions that are not in the best interests of principals. For instance, Borokhovich, Brunarski, and Parrino (1997) find that CEOs of firms that adopt antitakeover provisions receive higher salaries and options than those at similar firms without such provisions.
These arguments in particular argue that poison pills harm both minority shareholders and shareholders with significant investments in organizations. On the one hand, minority shareholders are harmed as the existence of a poison pill reduces the value of equity held and the potential for acquisition of the organization. These investors also lack the ability to directly monitor managerial behavior. Transaction costs exist that also reduce the potential value from liquidating such shareholdings. On the other hand, significant investors, such as large institutions, have less ability to transfer investments from one organization to another, as selling a large volume of holdings can significantly influence stock prices. Instead, these shareholders see a significant reduction in value of holdings when poison pills are adopted. However, these shareholders may be able to influence managerial behavior due to the size of their holdings.
An alternative perspective, the shareholder interest hypothesis (e.g., Heron & Lie, 2006), argues that poison pills might be beneficial for shareholder value. For example, Danielson and Karpoff (2006), using 302 firms adopting poison pills from 1984 to 1992, find that long-term operating performance generally improves after poison pill adoption. In these situations, poison pills can be an effective tool in warding off corporate raiders that look for short-term gains at the expense of long-term value. Defensive measures against takeover might also enable managers to negotiate greater benefits for stockholders (Grossman & Hart, 1980; Jennings & Mazzeo, 1986). Potential buyers should provide additional economic value to relieve poison pill protection, leading to higher bidding prices (Heron & Lie, 2006), while not necessarily reducing the probability that a firm is acquired (Comment & Schwert, 1995). Additionally, it has been argued that the threat of a takeover causes managers to pay attention too closely to short-term earnings even at the expense of long-term investment (e.g., Mahoney, Sundaramurthy, & Mahoney, 1997).
Agency Theory, Corporate Governance Mechanisms, and the Repeal of Poison Pills
While different theories exist on whether poison pills are beneficial to organizations, competing viewpoints using agency theory can help provide insight as to why firms may choose to repeal poison pill provisions. These viewpoints argue that governance mechanisms designed to limit managerial opportunism can be either complementary or substitutive in effect. Agency theory (Eisenhardt, 1989; Jensen & Meckling, 1976) argues that boards should act as agents for shareholders, monitoring managerial behavior and limiting opportunism. Problems, however, arise when interests of the principals, the shareholders, are not aligned with those of the agents, the firm’s managers. Without proper oversight, research has shown managers will demonstrate risk aversion through diversification (Amihud & Lev, 1981), accept golden parachutes (Singh & Harianto, 1989), and reprice underwater stock options (Pollock, Fischer, & Wade, 2002).
In order to limit opportunism by agents, several mechanisms can be implemented. First, monitoring devices may limit the behavior of executives. Independent directors can prevent managerial opportunism by imposing greater oversight over managerial actions and by serving as an approval function for organizational decisions. The external market for corporate control also can serve as a monitoring mechanism should internal governance and monitoring fail (Jensen, 1986; Walsh & Seward, 1990). The external market for corporate control can allow firms to acquire underperforming firms and replace management. Finally, principals may be represented by active or vigilant owners with significant stakes in the organization.
Second, incentive systems may be implemented in order to align management’s interests with those of the shareholders (e.g., Jensen & Meckling, 1976). Managerial incentive alignment allows for managerial gains to also result in shareholder gains. Incentive systems can include stock options and long-term incentive plans and can be aligned in outcome-oriented or behavior-oriented contracts (Eisenhardt, 1989).
Governance mechanisms as complements
Two potential perspectives exist regarding agency theory and the employment of governance mechanisms designed to limit opportunism by agents. The first argument suggests that governance mechanisms have complementary effects (Hoskisson, Castleton, & Withers, 2009; Schmidt & Spindler, 2004), whereby effective monitoring internally (e.g., board and shareholder oversight) and externally (e.g., the market for corporate control) and incentive alignment between principals and agents, reduce the potential for managerial opportunism. Principals will seek to limit agency problems by implementing as many governance mechanisms as possible, inducing synergistic effects among activities, in order to reduce the potential for agent opportunism.
Research on poison pill adoption finds results consistent with this perspective, as seen in Table 1. For instance, firms with CEO duality (Field & Karpoff, 2002; Mallette & Fowler, 1992), low insider ownership (Davis, 1991; Field & Karpoff, 2002; Heron & Lie, 2006; Mallette & Fowler, 1992; Sundaramurthy, 1996), dispersed ownership (Davis, 1991), and outside director loyalty to managers (Sundaramurthy, 1996) are all more likely to adopt poison pill provisions. Additionally, evidence exists that shareholders react more negatively to the adoption of a poison pill when governance is considered strong (Sundaramurthy et al., 1997).
Summary of Prior Research on Poison Pill Adoption
Note: (+)/(–) indicates more/less likelihood of poison pill adoption/existence.
These arguments are consistent with the managerial entrenchment hypothesis and the notion that governance mechanisms have complementary effects, as firms with strong monitoring by boards and incentive alignment between agents and principals are less likely to adopt a poison pill. Taken together, such arguments suggest a continuum of weak to strong governance, whereby stronger governance is represented by more cumulative monitoring and incentive alignment. As seen in Figure 1, from the perspective that governance mechanisms have complementary effects, poison pill repeal should therefore be more likely when monitoring by multiple parties is strong and incentive alignment between principals and agents exists. Stronger governance should lead to poison pill repeal in order to reduce the likelihood that managers become entrenched from the maintenance of the poison pill.
A Model Predicting Poison Pill Repeal with Governance Mechanisms as Complements
Governance mechanisms as substitutes
An alternative perspective, whereby governance mechanisms have substitutive effects, argues that opportunism does not need to be reduced through as many governance mechanisms as possible. This perspective argues that these mechanisms do not operate independently and have differing effects that are combinative (Rediker & Seth, 1995). For instance, Zajac and Westphal (1994) find that the firm’s use of incentive alignment is negatively associated with its use of monitoring mechanisms. That is, firms with weak incentive alignment have stronger monitoring mechanisms, or vice versa. Consistent with this perspective, research on poison pill adoption has shown that shareholders respond favorably when poison pills are adopted by firms with more independent directors (Brickley et al., 1994). Additionally, Danielson and Karpoff (1998) find that firms with poison pills had a greater proportion of outside directors on the board, while Mallette and Fowler (1992) find a positive relationship between institutional shareholdings and poison pill adoption. This last result potentially indicates that institutional shareholders are willing to accept poison pill adoption, as they can provide strong oversight of managerial behavior.
This perspective aligns closely with the shareholder interest hypothesis, as alternative governance mechanisms can prevent problems associated with poison pills (e.g., managerial entrenchment). The existence of a poison pill then may increase shareholder value by allowing shareholders to gain an additional premium if the firm were to be acquired and enables managers to focus on long-term competitiveness without the fear of acquisition. Figure 2 presents the theoretical predictions for the likelihood of poison pill repeal based on the assumption that governance mechanisms are substitutive in nature. From this perspective, when monitoring of managerial behavior is strong or incentives are aligned with shareholders, boards should be more likely to renew a poison pill. In these situations, shareholders are protected by alternative governance mechanisms, reducing the risk of managerial entrenchment. The remaining theoretical sections of this article outline competing hypotheses that will be used to test the likelihood of poison pill repeal based on the notion that governance mechanisms are either complementary or substitutive in effect.
A Model Predicting Poison Pill Repeal with Governance Mechanisms as Substitutes
Board Characteristics and Poison Pill Repeal
Proportion of independent directors
Agency theory argues that effective board composition serves to align the interests of managers with those of shareholders through higher quality monitoring (Walsh & Seward, 1990). However, in order to properly govern management’s decisions, the board must maintain independence from firm managers. As the proportion of managers (e.g., insiders) who serve on the board increases, there is a higher probability that the board will be dominated by management, which is more likely to align the board’s interests with those of the agents (Boeker, 1992; Fama & Jensen, 1983). Outside directors are presumed to be more independent than inside directors, given that outsiders can challenge the CEO with less fear of losing their positions (Pitcher, Chreim, & Kisfalvi, 2000). Consistent with these arguments, prior research has found that CEOs with power unchecked by outside directors are more likely to take actions that decrease shareholder wealth (Combs, Ketchen, Perryman, & Donahue, 2007; Dunn, 2004; Frankforter, Berman, & Jones, 2000). On the contrary, boards with more independent directors have been found to be more likely to dismiss the CEO when warranted (Boeker, 1992) and accept takeover bids (Buchholtz & Ribbens, 1994).
An important distinction must be made, however, in that not all outside directors are equal. Directors with business relationships with the firm are less independent, as they may be dependent upon such relationships for future income. These “affiliated” directors may also have interests that are more closely aligned with those of the agents, rather than principals. Thus, we specifically focus on outside directors (e.g., nonmanagerial directors) that do not have relationships with the firm (e.g., independent directors).
With regard to poison pill repeal, the perspective that governance mechanisms have complementary effects argues that more independent boards should be more likely to repeal poison pills, as such provisions reduce shareholder value and lead to managerial entrenchment (Borokhovich et al., 1997). Consistent with this logic, Mallette and Fowler (1992) argue that firms with a higher proportion of independent directors are less likely to adopt poison pills, although results did not find such a relationship. Guo, Kruse, and Nohel (2008) find that firms choosing to destagger boards of directors had a greater proportion of independent directors. In other words, research suggests that a board with more independent members is more likely to implement multiple governance mechanisms in order to limit managerial opportunism.
An alternative perspective using agency theory argues that governance mechanisms have substitutive effects. Consistent with agency theory, a more independent board of directors serves as a stronger monitor of managers’ behavior (e.g., Combs et al., 2007), reducing the likelihood for managerial opportunism. From this perspective, strong monitoring by independent directors can reduce the harmful effects of entrenchment associated with a poison pill. The reduced market for corporate control is counterbalanced by strong monitoring by the board and also allows for greater shareholder value through potential acquisition premiums. Consistent with these arguments, some research has found that firms are more likely to have a poison pill when the proportion of outsiders on a board is higher (Danielson & Karpoff, 1998).
Hypothesis 1a (complementary effects): The higher the proportion of independent directors on a board (e.g., strong monitoring), the higher the probability that the board will repeal or terminate a poison pill provision.
Hypothesis 1b (substitutive effects): The higher the proportion of independent directors on a board (e.g., strong monitoring), the lower the probability that the board will repeal or terminate a poison pill provision.
Proportion of outside directors appointed by CEO
While independent directors may better serve shareholder interests, not all outside directors are equally independent. Daily and Dalton (1994) coined the term “interdependent directors” to refer to directors nominated by the current CEO. Such directors are expected to act more favorably to the CEO, as they feel some level of gratitude for their current position, and they are also likely to be similar to the current CEO (Westphal & Zajac, 1995). Some directors may not overtly criticize the CEO in order to not jeopardize their position (Hirsch & Friedman, 1986), and are more likely to be individuals with whom the current CEO is comfortable and with whom will feel some loyalty to the CEO (Villiers, Naiker, & Staden, 2011; Wade, O’Reilly, & Chandratat, 1990).
Following the notion of governance mechanisms as having complementary effects, boards with a greater proportion of interdependent directors should be more likely to renew existing poison pill provisions. Boards loyal to the CEO have reduced effectiveness in monitoring CEO behavior and are thus more likely to make decisions in accordance with the CEO’s wishes. Consistent with these arguments, Sundaramurthy (1996) finds that outsiders with greater loyalty to the CEO have a strong effect on antitakeover provision adoption. From this perspective, ineffective governance in the form of weak board oversight is likely to lead to ineffective governance in the form of renewing a poison pill provision.
From the alternative perspective of governance mechanisms as having substitutive effects, firms with weak governance via board monitoring may be more likely to repeal a poison pill. In situations where internal monitoring of managerial behavior is weak, boards may need to allow for more effective external monitoring of managerial behavior. Thus, as the percentage of board outsiders appointed by the CEO increases, the likelihood of poison pill repeal may decrease.
Hypothesis 2a (complementary effects): The higher the proportion of outside directors appointed by the current CEO on a board, the lower the probability that the board will repeal or terminate a poison pill provision.
Hypothesis 2b (substitutive effects): The higher the proportion of outside directors appointed by the current CEO on a board, the higher the probability that the board will repeal or terminate a poison pill provision.
CEO duality
Agency theory argues that as managers gain greater power over the board, the board’s effectiveness in limiting opportunism is reduced. One means to assess managerial power is whether the CEO also holds the dual title of board chair (Hambrick, 1981; Harrison, Torres, & Kukalis, 1988; Mizruchi, 1983; Ocasio, 1994). CEOs with duality are more aligned with management and are likely to act to protect their own job (Firstenberg & Malkiel, 1994; Herman, 1981; Mace, 1971). Overall, CEO duality harms board independence and enables managerial entrenchment (Cannella & Lubatkin, 1993; Crystal, 1991; Daily & Dalton, 1997).
From the perspective that governance mechanisms have complementary effects, separating the roles of chairperson and CEO should enhance board monitoring and independence, leading to the repeal of poison pill provisions. Consistent with these arguments, prior studies on poison pill adoption find that CEO duality is significantly related to the adoption of a poison pill (Field & Karpoff, 2002; Mallette & Fowler, 1992). Thus, a perspective of governance mechanisms as complements predicts that firms without duality are more likely to repeal a poison pill in order to enhance the full effectiveness of monitoring managerial behavior.
The alternative perspective that governance mechanisms have substitutive effects provides a different prediction. Duality may lead to CEO domination of the board; however, the board can improve monitoring by allowing for an efficient market for corporate control. In firms with duality, the board may compensate for managerial power on the board with weaker protection from the market for corporate control. Consistent with this logic, Heron and Lie (2006) find that the adoption of poison pills was less likely when duality existed.
Hypothesis 3a (complementary effects): Existence of CEO/chairperson duality will lead to a lower probability that the board will repeal or terminate a poison pill provision.
Hypothesis 3b (substitutive effects): Existence of CEO/chairperson duality will lead to a higher probability that the board will repeal or terminate a poison pill provision.
Firm Ownership and Poison Pill Repeal
Top management team ownership
According to agency theory, agents should act in the interests of principals when incentives are aligned through the issuance of equity to managers. Managers with greater ownership are less likely to make value destroying decisions (Kosnik, 1987), given that they stand to benefit if equity is sold at a premium. For instance, CEOs with greater ownership are less likely to accept golden parachutes (Wade et al., 1990).
If governance mechanisms have complementary effects, firms with greater levels of incentive alignment through managerial shareholdings should also repeal poison pills. Past research has shown that pills are more likely to be adopted under conditions of lower insider ownership (Danielson & Karpoff, 1998; Davis, 1991; Field & Karpoff, 2002; Heron & Lie, 2006; Malatesta & Walking, 1988). From this perspective, given evidence that poison pills harm shareholder value, managers with higher levels of stockholdings may be more likely to repeal poison pill provisions. This better alignment of managerial and principal interests should lead to decisions that are in the best interests of shareholders. Additionally, the board may seek to limit opportunism through both incentive alignment and an effective market for corporate control.
On the contrary, the perspective that governance mechanisms have substitutive effects offers an alternative prediction. Managerial incentive alignment is designed to ensure that managers act in the interests of the firm’s principals. If mechanisms have substitutive effects, an effective market for corporate control is not as necessary when incentive alignment exists, as managerial interests are already aligned to promote decisions that increase shareholder value and managerial entrenchment is less likely. In these situations, firms may reduce managerial employment risk and promote investment in long-term strategic options by renewing poison pills.
Hypothesis 4a (complementary effects): The higher the percentage of shareholdings owned by the firm’s managers, the higher the probability that the board will repeal or terminate a poison pill provision.
Hypothesis 4b (substitutive effects): The higher the percentage of shareholdings owned by the firm’s managers, the lower the probability that the board will repeal or terminate a poison pill provision.
Prior studies have examined not only linear relationships between managerial shareholdings and incentive alignment, but curvilinear relationships as well. For instance, Morck, Shleifer, and Vishny (1988) noted a curvilinear relationship between incentive alignment and Tobin’s q, such that firm performance was highest at low and high levels of ownership, with moderate ownership leading to managerial entrenchment. Prior studies on antitakeover provision adoption find a nonlinear relationship between insider ownership and antitakeover provision adoption; however, these results are in the opposite direction suggesting that high ownership is likely to lead to adoption (Sundaramurthy, 1996).
We believe, however, that the findings of Morck and colleagues (1988) provide greater insight with regard to poison pill repeal, as greater ownership provides protection from the market for corporate control. At low levels of managerial ownership, managers lack the power to influence board decision making, leading to a greater likelihood of poison pill repeal. However, as managerial ownership rises, managers will seek to enhance power and remain entrenched by encouraging the board to renew a poison pill. Removing the poison pill protection at moderate levels of ownership will expose managers to significantly greater employment risk. At these levels of ownership, managers may be less willing to trade the increase in unemployment risk by removing the poison pill protection for gains that would result if the firm is acquired. Finally, as managers increase their shareholdings to higher levels, the probability of acquisition and the assumption of related unemployment risk will decrease. At higher levels of ownership, managerial incentives will be more closely aligned with shareholder interests (Fama & Jensen, 1983; Morck et al., 1988; Zahra, Neubaum, & Huse, 2000), as gains from acquisition will most likely outweigh increases in employment risk. Additionally, at the highest levels of ownership, managers may have the power to prevent or block potential acquisitions.
Hypothesis 4c: The relationship between managerial shareholdings and likelihood of poison pill repeal is curvilinear such that poison pills are less likely to be repealed as managerial shareholdings increase to a point, at which point poison pills are more likely to be repealed as managerial shareholdings increase (i.e., a convex relationship).
Outside director ownership
Stock ownership not only can be used to affect managerial decision making, but also can impact board decision making. Ownership can increase the personal stakes of directors and aligns interests less with management and more with shareholders, leading to decisions that increase the value of the firm (Dalton, Daily, Certo, & Roengpitya, 2003; Jensen & Warner, 1988).
From the perspective that governance mechanisms have complementary effects, greater levels of outside director ownership should increase the likelihood of poison pill repeal. For instance, Guo and colleagues (2008) find that firms with a higher percentage of director ownership were more likely to destagger the board of directors. Additionally, Mallette and Fowler (1992) find that inside director ownership decreases the likelihood of poison pill adoption. These arguments suggest that when the interests of directors are aligned with those of principals, directors are more likely to utilize multiple governance mechanisms to limit opportunism. Thus, directors with greater ownership should be more likely to repeal existing poison pill provisions.
An alternative approach argues that if director incentives are aligned with shareholders, an effective market for corporate control is less necessary in order to properly monitor managerial behavior and limit opportunism. From this perspective, the existence of effective monitoring due to incentive alignment is enough to overcome the potential disadvantages of retaining a poison pill provision.
Hypothesis 5a (complementary effects): The higher the percentage of shareholdings owned by outside directors, the higher the probability that the board will repeal or terminate a poison pill provision.
Hypothesis 5b (substitutive effects): The higher the percentage of shareholdings owned by outside directors, the lower the probability that the board will repeal or terminate a poison pill provision.
Pressure-resistant and pressure-sensitive institutional ownership
In recent years, a greater concentration of ownership has been held by institutional investors, especially investors with more than 5% of a firm’s stock. Oversight from institutions is expected to lead to decisions made in shareholders’ best interests, due to a stronger incentive and ability to monitor managerial actions (Agrawal & Mandelker, 1990, 1992; Kochhar & David, 1996). However, while such investors have greater incentive to monitor management, the incentives of all such institutions are not necessarily the same (e.g., Hoskisson, Hitt, Johnson, & Grossman, 2002). In particular, research has paid attention to the differences in propensity to engage in monitoring between pressure-resistant and pressure-sensitive institutional investors (Brickley, Lease, & Smith, 1988; Dalton et al., 2003; David, Kochhar, & Levitas, 1998; Kochhar & David, 1996).
Pressure-resistant institutional investors, such as pension funds, mutual funds, and investment companies, do not have a direct business relationship with firms in which they own shares. As a result, they are more likely to engage in monitoring of managers’ decisions and are often labeled “activist institutional investors” (Brickley et al., 1988). Since they do not have business relationships with the firm, these investors’ intervention in corporate governance will not create conflict with investment relationships. Prior research finds a negative association between ownership by pressure-resistant investors and the adoption of antitakeover amendments (e.g., Evans, Pyles, & Choo, 2009). For example, Brickley et al. (1988) find that pressure-resistant institutions are more likely to oppose antitakeover amendments than pressure-sensitive institutions. These arguments suggest that pressure-resistant institutional investors are strongly against poison pill provisions. Thus, if governance mechanisms have complementary effects, higher levels of pressure-resistant institutional ownership will lead to poison pill repeal.
The alternative perspective of governance mechanisms as having substitutive effects suggests that pressure-resistant institutional ownership will lead to poison pill renewal. In situations where activist institutional investors exist, managerial opportunism may be limited by the propensity of such investors to monitor managerial decision making. Renewing a poison pill may allow for a firm to reap the potential benefits without the associated problems. Consistent with these arguments, research on antitakeover provision adoption has shown that firms were more likely to adopt such provisions when institutional shareholdings were greater (Danielson & Karpoff, 1998; Davis, 1991; Mallette & Fowler, 1992). Furthermore, Evans and colleagues (2009) find that the stock market reacts positively to antitakeover provision adoption when there is a high level of pressure-resistant investors, indicating the potential positive effect of renewing poison pills when effective monitoring by pressure-resistant investors exists.
Hypothesis 6a (complementary effects): The higher the percentage of pressure-resistant institutional ownership, the higher the probability that the board will repeal or terminate a poison pill provision.
Hypothesis 6b (substitutive effects): The higher the percentage of pressure-resistant institutional ownership, the lower the probability that the board will repeal or terminate a poison pill provision.
Unlike pressure-resistant institutional investors, pressure-sensitive institutional investors, such as banks, insurance companies, and nonbank trusts, are less likely to be vigilant, as they have a business relationship with the firm (David et al., 1998; Kochhar & David, 1996). Specifically, Borokhovich, Brunarski, Harman, and Parrino (2006), measuring stock returns around the announcement of antitakeover amendment proposals, note that blockholders with potential business ties to a firm are less effective monitors.
From the perspective of governance mechanisms as having complementary effects, higher levels of pressure-sensitive institutional shareholdings should lead to poison pill renewal. Pressure-sensitive investors may be more likely to side with managers due to concerns that managers may terminate business relationships if they oppose managerial preferences (Heard & Sherman, 1987; Zahra et al., 2000). Additionally, renewing poison pill provisions allows pressure-sensitive institutions to maintain existing business relationships, which may be lost if the firm is acquired. Consistent with this logic, Evans et al. (2009) find that the likelihood of poison pill adoption is positively associated with shareholdings by pressure-sensitive investors.
The alternative perspective, governance mechanisms as having substitutive effects, suggests that firms with higher levels of pressure-sensitive investor ownership may be more likely to repeal a poison pill. First, the substantial amount of ownership by such investors can be a substitute for antitakeover provisions. Second, in situations where monitoring by shareholders is weak, the board may need to use external monitoring mechanisms to prevent opportunism.
Hypothesis 7a (complementary effects): The higher the percentage of pressure-sensitive institutional ownership, the lower the probability that the board will repeal or terminate a poison pill provision.
Hypothesis 7b (substitutive effects): The higher the percentage of pressure-sensitive institutional ownership, the higher the probability that the board will repeal or terminate a poison pill provision.
Method
Sample and Data
Our sample consists of 288 U.S.-based, publicly traded firms that made a poison pill renewal or termination decision over the period 2005–2007 and with full data available from COMPUSTAT, Corporate Library, and FactSet SharkRepellent. 1 During these three years, 244 firms made renewal or repeal decisions at the defined poison pill expiration date and 44 firms terminated their pill earlier than its scheduled expiration. Data on poison pill expiration and associated firm decisions to renew or repeal the pill were obtained from FactSet SharkRepellent, a product offering by TrueCourse, Inc., a research company specializing in such provisions.
According to FactSet SharkRepellent, renewal rates on expiring poison pills hit a high of 85% in 2001, then decreased rapidly to a low of 32% over the period 2005–2007 (Laide, 2008), the lowest renewal rate since the institute began tracking poison pill renewal. In our sample, 74 firms extended or renewed their poison pill while 214 firms repealed or terminated their poison pill provision.
Dependent Variable
Upon their passage, poison pills have an effective period in which they are in place with a defined expiration date. As such, firms usually make the decision whether to renew the poison pill on or near that date. Firms can also eliminate poison pill provisions early by (a) a redemption payment method (i.e., payment of a defined amount of money to shareholders) or (b) accelerating the pill’s expiration date. Given this defined period of provision existence, firms do not typically reexamine poison pill provisions annually. Instead, the decision to renew a poison pill provision is best modeled by examining the time at which the provision is actually set to expire.
Our dependent variable examines whether firms renew a poison pill provision at the pill’s expiration or whether the firm eliminated (e.g., repealed or allowed it to expire) a poison pill. Given that we attempt to predict the repeal or dropping of a poison pill provision, the dependent variable is a dichotomous variable coded as 1 where the firm terminates or repeals a poison pill provision and 0 if the firm chooses to renew a poison pill.
Independent Variables
Most of our independent variables, board and ownership characteristics of the firm, are drawn from the Corporate Library database, recognized as a reliable independent rating agency offering corporate governance data for most publicly traded U.S. corporations and has been used by corporate governance researchers (e.g., Deckop, Merriman, & Gupta, 2006). We also used firm proxy statements (i.e., SEC Form DEF 14A) for ownership information for pressure-resistant investors and pressure-sensitive investors.
Prior researchers find that board composition and characteristics affect the degree of managerial decisions in line with shareholder interests (e.g., Daily & Dalton, 1994). Proportion of independent directors is calculated by dividing the number of unaffiliated outside directors on a board by the total number of board members (Mallette & Fowler, 1992). Proportion of directors nominated by CEO, or interdependent directors (e.g., Daily & Dalton, 1994), is the proportion of the firm’s directors who were nominated by the current CEO. This variable was obtained by comparing the tenures of the current CEO and each director. We measured CEO duality by creating a dummy variable coded as 1 if the CEO also served as the chairman of the board, 0 otherwise. Data on board characteristics were gathered at the time the poison pill provision was set to expire or as of the date of early termination. 2
Since major shareholders also have been shown to impact the decision to adopt antitakeover proposals (Evans et al., 2009; Sundaramurthy, 2000), we include managerial ownership, outside director ownership, pressure-resistant institutional ownership, and pressure-sensitive institutional ownership. Managerial ownership represents the percentage of shares held by top managers including the CEO. Outside director ownership is measured as the percentage of a firm’s outstanding stock held by independent directors. Following prior research (Brickley et al., 1988; Evans et al., 2009), we calculate pressure-resistant institutional ownership as the percentage of shares owned by public pension funds, mutual funds, foundations, endowments, and investment companies. Pressure-sensitive institutional ownership is the percentage of shares owned by insurance companies, banks, nonbank trusts, and employee stock ownership plans (ESOPs). All information on firm ownership is lagged by one year.
Control Variables
We include firm size, market-to-book ratio, debt ratio, poison pill age, board size, number of board meetings, and number of other takeover provisions in force, as well as year and industry dummies as control variables. All control variables, except board size, are lagged by one year. Firm size has been shown to be related to the decision to adopt a poison pill (e.g., Heron & Lie, 2006). Thus, we use the number of employees as a measure of firm size. Since the distribution of this variable is skewed, we used a log transformation of the number of employees. Firms with poor market performance compared to the firm’s book value are assumed to be at a greater risk of takeover. Following previous findings (e.g., Sundaramurthy et al., 1997), market-to-book ratio, calculated as total market value divided by the firm’s book value and adjusted by a firm’s average industry level (using two-digit SIC), is also controlled. Because firms with high leverage are less attractive to potential buyers, it is reasonable to assume that they would not need poison pills. Therefore, we control for a firm’s debt ratio, as measured by long-term debt divided by total assets. Board size is the total number of active directors on the firm’s board, and board meetings is the number of board of directors’ meetings held throughout the entire year. We also include poison pill age by measuring the time period in years since the initial enactment. Last, the number of other antitakeover provisions is included in the analysis. The number of takeover provisions a firm has adopted has also been shown to be significantly related to poison pill adoption (Davis, 1991). Other antitakeover provisions may affect the decision of whether to renew the poison pill. This variable, other provisions, is measured by the sum of five other antitakeover protections at the time of poison pill expiration: (a) classified board provision, (b) dual class stock, (c) supermajority merger approval provision, (d) bylaw amendment vote, and (e) corporate charter amendment vote. 3 Thus, this variable ranges from 0 to 5. To control for time and industry effects given a multi-year, multi-industry sample, we include year dummies (i.e., years 2005 and 2006) and industry dummies. We use six different industry categories including agricultural, fishing, mining, and construction (SIC < 20), manufacturing (20 ≤ SIC < 40), transportation, communication, and utility (40 ≤ SIC < 50), wholesale and retail (50 ≤ SIC < 60), finance, insurance, and real estate (60 ≤ SIC < 70), and service (SIC ≥ 70). 4
Analysis
Given the nature of our context, our sample does not represent a random sample of firms. Every firm in our sample already had a poison pill in existence in 2005. Given that prior research shows that the decision to adopt a poison pill is related to a variety of factors, the inclusion in our sample likely represents an endogenous choice by each firm in our analysis. In order to address the endogeneity associated with poison pill existence, we use a Heckman selection model (see Heckman, 1979), a two-staged procedure that corrects for sample selection bias. This model was used in order to correct for any sample selection error that may exist regarding firm size and governance characteristics that are correlated with poison pill existence. As a first step, we estimate the likelihood of the presence of a poison pill in 2005 using a probit regression model to address potential sampling bias at firms with or without poison pills. This sample was inclusive of all firms in the Corporate Library database for which all data could be gathered (N = 1,393). Multiple variables were used to predict the presence of a poison pill, and all explanatory variables were lagged by one year. The model took the following form:
After the first-stage procedure, we tested our hypotheses using probit regression. The selection model generated the inverse Mills ratio, which is included as an additional explanatory variable in our second-stage probit regression predicting poison pill repeal. The results of the first stage probit regression are reported in the appendix. The inclusion of the inverse Mills ratio controls for any sample selection bias that might be due to nonrandom sampling. We used mean-centered variables for all continuous variables and reported variance inflation factors (VIFs) based on the full model. Since the general rule is that VIF values should be below 10 (Belsley, Kuh, & Welsch, 1980), we assume that our sample does not have serious multicollinearity issues.
Results
Table 2 shows the descriptive statistics and correlations among variables except year and industry dummies. The decision to repeal a poison pill is positively and significantly correlated with firm size, board size, board meetings, proportion of independent directors, and ownership by pressure-resistant institutions. In our sample, 74.3% of firms (N = 214) eliminated poison pills, which indicates a significant decline in the use of poison pills during this time frame.
Means, Standard Deviations, and Correlations
Note: Two-tailed coefficient test (N = 288).
p < .05. **p < .01. ***p < .001.
Our hypotheses are tested using probit regression as reported in Table 3. Model 1 includes only control variables, including the inverse Mills ratio for possible selection bias. Model 2 tests the effects of our hypothesized variables, including board and ownership characteristics, on the poison pill termination decision. We include the square term of managerial ownership to test for a curvilinear relationship in Model 3. In our sample, the Breusch–Pagan/Cook–Weisberg test revealed the presence of heteroscedasticity (p < .001). To correct for heteroscedasticity, robust standard errors are reported and used for hypothesis testing (Huber, 1967; White, 1982).
Probit Analysis of Poison Pill Renewal or Repeal Decision with Heckman Correction
Note: Dependent variable is whether the firm repealed the poison pill (n = 214) or not (n = 74). VIF (variance inflation factor) values are based on Model 3. Two-tailed coefficient tests.
p < .1. *p < .05. **p < .01. ***p < .001.
In examining control variables, results show that larger firms are more likely to repeal poison pills, consistent with past research (e.g., Davis, 1991; Mallette & Fowler, 1992) on pill adoption. Larger firms have more resources concerning takeover provisions and are more difficult to acquire due to high capital requirements. Results also find that firms with frequent board meetings, a proxy for board vigilance, are more likely to repeal a poison pill.
In Models 2 and 3, we examine the effects of board and ownership characteristics on the decision to repeal or renew a poison pill. Hypothesis 1 predicts the positive (1a) or negative (1b) effect of the proportion of independent directors on a board on the likelihood of poison pill repeal. Results, however, do not indicate any significant effect of the proportion of independent directors on repeal. Thus, Hypotheses 1a and 1b are not supported. Hypothesis 2 examines a complementary (2a) or substitutive (2b) effect of directors appointed by the CEO on the decision to repeal or terminate a poison pill. Results indicate that as the CEO appoints more directors to a firm’s board, the board is more likely to renew poison pill provisions (p ≤ .05). Thus, Hypothesis 2a is supported. Consistent with agency theory, this finding suggests that the board acts in the interests of the CEO when there are more directors appointed by the current CEO. Hypothesis 3 predicts that firms with CEO duality are less (3a) or more (3b) likely to repeal or terminate poison pill provisions. Consistent with the view of governance mechanisms as substitutes, duality has a positive and significant association with the repeal of poison pills in both Model 2 and Model 3 (p ≤ .05). These results indicate that when boards allow managers greater power with duality, executives may face a trade-off with a more efficient market for corporate control leading to greater employment risk.
In terms of managerial ownership, results do not suggest any linear relationship, providing no support for Hypothesis 4a or Hypothesis 4b. Hypothesis 4c argues that a U-shaped pattern between managerial ownership and poison pill repeal exists, as a poison pill will be repealed at high and low levels of ownership. In Model 3, the squared term of managerial ownership is statistically significant (p ≤ .01). These results provide support for Hypothesis 4c, and this convex relationship can be examined in greater detail in Figure 3. At low levels of ownership by top managers, the likelihood of repeal decreases. Managers may be less likely to trade employment risk for the associated gain from a potential acquisition. However, as managers gain more shareholdings, the repeal of a poison pill is more likely. At this point, managers may be more willing to trade employment risk for the benefits associated with acquisition premiums for shareholdings. This may also be the point where managerial interests become more closely aligned with shareholder interests. These results are consistent of those of Morck et al. (1988) and suggest that incentive alignment by ownership may not be effective until ownership reaches higher levels of stockholdings. At the highest levels of ownership in our sample, poison pills are unlikely to be needed due to managers already having obtained significant control.
Curvilinear Relationship between Managerial Ownership and Probability of Poison Pill Repeal
Hypothesis 5 argues that the effects of independent director ownership is positively (5a) or negatively (5b) associated with poison pill repeal. Consistent with the perspective that governance mechanisms have complementary effects, we find that higher levels of ownership by outside directors leads to the repeal of poison pill provisions in Model 3, but this relationship is significant at the .10 level. This result provides some support for the notion that when outside directors’ incentives are aligned with shareholders, directors make decisions that enhance shareholder value and allow for a greater reduction in managerial opportunism.
Ownership by institutional blockholders also plays a strong role in the decision to repeal poison pill provisions. We argue the positive (H6a) or negative (H6b) effects of shareholdings by pressure-resistant institutional investors on the repeal of poison pills. Our findings indicate that there is a greater likelihood of poison pill repeal when the stock ownership by pressure-resistant investors is high (p ≤ .001), supporting Hypothesis 6a. This significant effect of pressure-resistant investors reflects recent shareholder activism against poison pills (e.g., Lindstrom, 2005) and indicates that pressure-resistant institutional blockholders are pushing for the elimination of poison pills. We do not find any significant results for pressure-sensitive institutional investors, thus neither Hypothesis 7a nor Hypothesis 7b is supported.
Table 4 presents a summary of our findings related to governance mechanisms as having complementary or substitutive effects. Our results provide significant support for the perspective that firms view governance mechanisms as complements rather than substitutes. Firms with stronger monitoring by more independent board members (e.g., those not appointed by the CEO) and activist institutional investors are significantly more likely to repeal poison pills. Additionally, aligning the incentives of managers and directors with higher levels of equity also increases the likelihood of poison pill repeal. Our results provide limited support for the notion that governance mechanisms have substitutive effects, with only results indicating that firms with duality are more likely to repeal poison pill provisions providing support for this perspective. In sum, these results suggest that firms bundle multiple governance mechanisms in order to attempt to limit managerial opportunism.
Comparison of Results in Support of Governance Mechanisms as Complementary versus Substitutive Effect
ns refers to non significant results for hypotheses tested.
The effect of outside director ownership on poison pill repeal is significant at the p = .10 level.
Supplemental Analyses
We conducted a number of supplemental analyses to check the robustness of our results. First, since the early termination of a poison pill and the expiration of a poison pill on schedule are separate decisions, we ran a probit regression to compare firms who terminate poison pills prior to the scheduled expiration date (n = 44) and those who terminate provisions on the scheduled date (n = 170). As reported in Table 5, the proportion of directors nominated by the CEO is positively related to the likelihood of early termination and pressure-sensitive investor ownership is negatively associated with early termination. However, these relationships are only marginally significant, indicating that governance attributes are not significantly different at the conventional level. Furthermore, we conducted an additional test of our hypotheses (n = 244) without including the 44 firms that chose to terminate poison pill provisions before the expiration date. Results for these models were nearly identical to the models reported in Table 3, which provide some support for our discussion that the drivers of repeal are similar whether firms allowed poison pills to expire or chose to terminate such pills early.
Sensitivity Analysis: Probit Regression of Early Termination versus Expiration on Schedule
Note: Dependent variable is whether the firm repealed the poison pill before the scheduled expiration date (n = 44) or allowed it to expire on schedule (n = 170). VIF (variance inflation factor) values are based on Model 3. Industry dummies are not reported for parsimony, and no firms repealed the poison pill before the scheduled date in the agricultural, fishing, mining, and construction (SIC < 20) industries. Two-tailed coefficient tests.
p < 0.1. *p < .05. **p < .01. ***p < .001.
We also utilized a number of alternative approaches for checking the robustness of our findings. First, we tested models using each antitakeover provision as its own dummy variable, rather than the summed approach included in our tables. This approach did not change the pattern of results reported. Furthermore, the results of our probit regression model are almost the same without the inclusion of the inverse Mills ratio. In addition, our result shows same the pattern without controlling for industry. Last, we measured the effects of pressure-resistant and pressure-sensitive institutional investors by using the number of investors, instead of percentage of ownership. We confirmed the same significant effects. 5
Discussion
Previous research on poison pills has examined governance characteristics of firms that choose to adopt such provisions, arguing that they are harmful to shareholder value and lead to management entrenchment. In recent years, however, firms have begun to repeal poison pills at a much higher rate. Despite this trend, research has not examined why firms choose to repeal or terminate poison pills. Repealing a poison pill provision is an important decision, as removal of the provision exposes the firm to a more efficient market for corporate control and increases the likelihood of the firm being acquired. While potential acquisition is likely to benefit shareholders through payment of a premium, an acquisition also increases the likelihood of managerial unemployment (e.g., employment risk increases). Thus, repeal of a poison pill may be beneficial to shareholders, but risky for managers. Our article makes several meaningful contributions to the literature on both corporate governance and antitakeover provisions.
First, we use agency theory arguments on governance mechanisms as having complementary or substitutive effects to provide competing hypotheses as to whether firms will repeal poison pills. The perspective that governance mechanisms have complementary effects suggests poison pills are more likely to be repealed when other governance mechanisms have been implemented by an organization. In these cases, the organization is more likely to subject itself to an efficient market for corporate control when monitoring by multiple parties and incentive alignment between managers and shareholders exists, as managers may have an incentive to renew a poison pill provision to limit employment risk. Alternatively, the notion that governance mechanisms have substitutive effects suggests that firms may not repeal poison pills when other governance mechanisms are in place, as these mechanisms can serve to limit opportunism, while the poison pill’s protection may encourage other value creating behaviors by management. Overall, our results find much greater support for the notion that existing governance mechanisms have a complementary effect on the decision to repeal a poison pill, whereby greater levels of governance mechanisms in existence at the decision time were more likely to lead to poison pill repeal. Specifically, we find support for the complements perspective for three hypothesized variables (interdependent directors, pressure-resistant institutional ownership, and outside director ownership), while results support the substitute perspective for only one relationship (CEO/chair duality). Our results relating to managerial ownership are more difficult to classify as they provide support for both perspectives, whereby both nominal and significant ownership by managers encourages poison pill repeal (ownership has a complementary effect), while a moderate amount of ownership reduces the likelihood of repeal (ownership has a substitutive effect). This relationship is discussed in greater detail below.
Based on our results, we conclude that firms prefer to bundle multiple governance mechanisms in order to limit problems associated with opportunism. As such, firms that exhibit characteristics of stronger monitoring and incentive alignment are more likely to repeal a poison pill, a provision that has been shown to negatively impact shareholder value.
Specifically, we find that firms with fewer directors appointed by the CEO and greater ownership by both pressure-resistant institutional blockholders and outside directors are more likely to repeal poison pill provisions. These results indicate that when monitoring is stronger and there is a greater level of incentive alignment, firms are more likely to repeal poison pill provisions, indicating that firms identify these mechanisms as working together as complements to reduce managerial opportunism. Consequently, our results support the perspective that governance mechanisms have substitutive effects in only one instance, CEO duality. These results are inconsistent with prior results on poison pill adoption and duality (e.g., Field & Karpoff, 2002; Mallette & Fowler, 1992), and suggest that boards may provide greater power to managers in the form of duality, but managers must be willing to accept the trade-off of increased employment risk. Furthermore, managers who have duality may be willing to accept greater employment risk as the increased power provided by duality may allow the CEO to block takeover threats. This result is consistent with studies regarding duality that indicate that it may reduce the likelihood of poison pill adoption (Heron & Lie, 2006) or increase the level of corporate performance (e.g., Donaldson & Davis, 1991; Lin, 2005). In addition, others argue that duality allows for greater unity of command and strategic information (e.g., Boyd, 1995; Finkelstein & D’Aveni, 1994) and fosters better communication between management and the board (Anderson & Anthony, 1986), indicating that duality may be beneficial in certain circumstances.
Second, this study examines the decision of whether to renew an expiring poison pill provision and finds that governance characteristics of firms that repeal poison pills are not necessarily the reciprocal of those that chose to adopt poison pills. Work on adoption of poison pills primarily supports the position of governance mechanisms as having complementary effects (e.g., Mallette & Fowler, 1992), such that poison pills are adopted when firms have poorer governance. Our results are similar in this vein, in that we note significant support for the perspective that governance mechanisms are complementary in nature. However, our results note two major factors that drive poison pill repeal which deviate from those of studies on adoption. Specifically, results show that firms with CEO duality and greater shareholdings by pressure-resistant institutional investors are more likely to repeal poison pill provisions. While we find these factors increase the likelihood of repeal, prior research (Danielson & Karpoff, 1998; Field & Karpoff, 2002; Mallette & Fowler, 1992) finds that these same factors increase the chances of poison pill adoption. This inconsistent pattern of results indicates a separate decision-making pattern may exist from what drove poison pill adoption. More fine-grained studies of board processes may be able to help understand this ambiguity.
While our results are different from prior research on poison pill adoption in key ways, some findings are consistent between poison pill repeal and adoption that support the perspective of governance mechanisms having complementary effects. Firms with a greater proportion of directors appointed by a CEO and with greater inside ownership are more likely to repeal poison pills. Additionally, while the general pattern between managerial ownership and repeal is consistent with studies on pill adoption (e.g., Danielson & Karpoff, 1998; Davis, 1991; Mallette & Fowler, 1992), the curvilinear relationship we find is opposite that found by Sundaramurthy (1996). Consistent with Mallette and Fowler (1992), we find no effect of independent directors with regard to decisions on poison pills. In sum, while our results and studies on poison pill adoption both provide support for the notion that governance mechanisms have complementary effects in general, our results have major differences from prior studies which indicate that the factors that drive pill repeal are not the same factors that lead to pill adoption. Interestingly, in several cases the factors that drove pill adoption also drive repeal (e.g., duality, institutional ownership). Future research should examine whether changes in the corporate governance environment affected these decisions or whether the decision to repeal a poison pill represents a different decision-making process than the decision to adopt a poison pill.
Third, this study is the first to examine the decision boards must undertake of whether to renew or terminate a poison pill. Given the significant amount of focus placed on the adoption of antitakeover provisions (e.g., Danielson & Karpoff, 1998; Davis, 1991; Mallette & Fowler, 1992; Sundaramurthy, 1996), we believe there is significant interest in why firms abandon such provisions. Prior research on the repeal of antitakeover provisions has identified similar findings, including that firms choosing to destagger boards of directors have better governance and greater monitoring of management by boards and shareholders (Guo et al., 2008). Future research should continue to investigate the reasons that firms are abandoning antitakeover provisions and the consequences of these decisions.
Fourth, we find a curvilinear relationship between managerial ownership in the firm and the likelihood of poison pill repeal. Our findings are similar to those of Morck et al. (1988) and indicate that the likelihood of repeal decreases as managerial shareholdings go from low to moderate levels. As managerial ownership increases beyond a certain point, as shown in Figure 3, the likelihood of repeal begins to increase significantly. These findings are significant for several reasons. First, at extremely low levels of ownership, our results may indicate that managers lack power to influence the decision to renew a poison pill. However, as levels of ownership increase in our sample, our results suggest that ownership functions as an entrenching mechanism (rather than an aligning mechanism) and managers are unwilling to trade increased employment risk for the potential gains related to ownership. This result may indicate that low to moderate levels of ownership entrench managers or that moderate levels of ownership serve as a substitute for an efficient market for corporate control. However, once ownership reaches a certain level, managers may be more willing to trade employment risk for gains if the firm is acquired. At higher levels of ownership, managers may also be able to block potential unwanted acquisitions. The findings at low and high levels of ownership are consistent with the notion that incentive alignment is achieved, providing support for the perspective that governance mechanisms have complementary effects. Taken together, these findings also may indicate either that managers are risk averse to the potential loss of an executive position at moderate levels of ownership or that incentive alignment does not exist until managers gain a larger share of ownership. In order to prevent this problem, managers may need to be provided significant ownership stakes if the firm expects managers to repeal provisions that affect their employment security. Alternatively, due to the significant ownership needed for incentive alignment, it may be less costly for boards to employ other governance mechanisms to reduce opportunism.
Practically, our results are also important. Our results indicate that keeping CEOs in check by restricting directors they appoint can limit managerial opportunism. Directors appointed by a CEO are more likely to act in the CEO’s interest in order to exhibit reciprocity for the CEO’s decision to nominate the director. Furthermore, our study suggests that providing greater equity in the firm can also truly align the interests of managers, directors, and shareholders. However, providing only small or token equity amounts may actually promote managerial entrenchment. Firms should be careful when providing equity to managers to ensure that incentive alignment is reached. Additionally, our results provide evidence that pressure-resistant institutional investors can significantly impact firm decision making with regard to corporate governance. It is important for managers and shareholders to understand how such investors can alter board decision making.
Future Research
While we believe that board and ownership characteristics play a significant role in the decision to renew a poison pill, future research should continue to explore other characteristics that may play a strong role in the process. For instance, building on institutional theory (e.g., DiMaggio & Powell, 1983), research could examine whether firms choose to repeal a poison pill when other firms that are connected to the firm’s board make a similar decision. Additionally, the effects of pressure-resistant institutional investors may indicate that firms with significant investments from such institutions may make a similar decision at the same time. Thus, this process may be driven by a few institutions with shareholdings across a variety of firms.
Future studies can examine whether other characteristics of a firm may impact the decision to repeal. For instance, consistent with research on pill adoption, our findings show that large firms are more likely to repeal poison pill provisions. Large firms require greater capital to acquire and thus are considered to have a lower likelihood of being a takeover target. Future research should examine whether governance characteristics matter less when large firms face decisions to renew or repeal poison pills. Additionally, prior research has argued that poison pills provide incentives for managers to invest in long-term investments (Mahoney et al., 1997). Given that long-term investments are considered evidence of strong governance, future research can benefit by examining whether firms that invest more in long-term strategic activities (e.g., research and development) are less likely to repeal poison pills and whether interactions exist between other governance mechanisms (e.g., monitoring, incentive alignment) and long-term investment to predict poison pill renewal. Firms may be unwilling to repeal poison pills in order to continue to encourage investments in long-run strategic options.
The decision of whether to renew a poison pill may also be driven by existing conditions in the market for mergers and acquisitions. For instance, in an inactive market for mergers and acquisitions, a poison pill may not be necessary in order to reduce employment risk, as there is a lesser risk of acquisition. However, managers may prefer a poison pill when there is a strong market for acquisitions, especially when ownership is low. Managers with low levels of ownership may lack power to prevent loss of employment in the case of acquisition, and thus may push the board to renew a poison pill provision. Lower levels of ownership yield lesser premiums in the case of the firm being acquired. Since active markets for acquisitions tend to yield greater premiums, managers with significant shareholdings may be more willing to expose themselves to greater levels of employment risk due to the potential for gains in premiums for shares. Future research should examine how market conditions affect a board’s decision to renew or repeal a poison pill provision.
Finally, industry conditions may also strongly affect the use of governance mechanisms as complements or substitutes. In regulated industries such as banking, the market for corporate control may be less active due to government oversight (Adams & Mehran, 2003). As such, stronger oversight from the board and investors may be necessary to reduce managerial opportunism. In these industries, the determinants of poison pill repeal as well as the effects of poison pills might be structurally different.
Limitations
Despite our results, this study is not without its limitations. First, although our data cover poison pill decisions from 2005 to 2007 and we use lagged explanatory variables, our study is cross-sectional. Stronger causal inferences may be formed in future studies by examining changes in governance characteristics from the date of pill adoption until decisions are made regarding renewal. A longitudinal study of firms from adoption to repeal or renewal may provide greater insights into changes in corporate governance. Insights from this research could provide an indication as to whether an endogenous change in firm governance occurred or whether changes were driven by exogenous forces. Second, our study uses only proxies for corporate governance. For example, we find no support that a greater proportion of independent directors lead to poison pill repeal, indicating better proxies may be needed to measure board monitoring. Future research should examine the black box of characteristics that go into helping firms better monitor managerial behavior and should dive deeper into corporate governance to understand why firms undertake decisions such as whether to terminate antitakeover provisions.
Third, our study adopts an agency theory approach in order to explain poison pill repeal or nonrenewal; however, other theories may further explain this phenomenon. For instance, theories on the spread of managerial fads and decision making through social networks (e.g., Davis, 1991; Haunschild, 1993) may explain additional variance in how interlocking directorates and other networks encourage firms to repeal poison pills. While future research should examine the explanatory power of such theories, our results provide strong evidence that agency theory can predict the decision of whether or not firms will repeal expiring poison pill provisions. We encourage future authors to examine the role of board interlocks and institutional mimicking of firm decision making with regard to poison pill repeal.
In conclusion, past research primarily examines the causes and effects of poison pill adoption. However, we examine a different firm decision, whether to renew or terminate a poison pill provision, and attempt to model whether governance mechanisms designed to limit managerial opportunism can be either complementary or substitutive in effect. Our results indicate that firms with fewer directors appointed by the CEO, CEO duality, and greater pressure-resistant institutional stockholdings are more likely to repeal poison pill provisions. Taken together, these results provide significant evidence that multiple governance mechanisms are used by boards and owners as complements in order to limit managerial opportunism; however, firms may also use other potential governance mechanisms as trade-offs in order to limit managerial power and entrenchment in the organization.
Footnotes
Appendix
Sample Selection Models: The Presence of a Poison Pill
| Coeff. | SE | VIF | |
|---|---|---|---|
| Intercept | 0.09 | (0.06) | |
| Firm size | −0.10*** | (0.02) | 1.06 |
| Debt ratio | −0.17 | (0.19) | 1.01 |
| Market-to-book ratio | 0.00 | (0.00) | 1.01 |
| Other provisions | 0.12*** | (0.03) | 1.01 |
| Independent director proportion | 1.19*** | (0.24) | 1.16 |
| CEO duality | 0.03 | (0.07) | 1.01 |
| Institutional holdings | 0.79*** | (0.19) | 1.15 |
| Block ownership | −0.24 | (0.25) | 1.13 |
| Insider ownership | −2.14*** | (0.28) | 1.20 |
| −2 log likelihood | 868.63 | ||
| Chi-square | 153.76 (df = 9) | ||
| Pseudo R2 | .09 | ||
Note: Dependent variable is whether the firm had a poison pill (1) or not (0) before 2005. Two-tailed coefficient tests (N = 1,393). VIF = variance inflation factor.
p < .05. **p < .01. ***p < .001.
Acknowledgements
The authors would like to thank Vince Barker and Laura Poppo for their assistance throughout the process of completing this article. Additionally, we would like to thank the action editor from the Journal of Management, as well as the anonymous reviews of two very helpful and insightful colleagues from the Journal of Management and three helpful colleagues from the Academy of Management conference. We also gratefully acknowledge data provided by FactSet SharkRepellent from FactSet Research Systems, Inc.