<%BANNER%>

The Impact of the Changes in the Takeover Market on Managerial Entrenchment


PAGE 1

THE IMPACT OF THE CHANGES IN THE TAKEOVER MARKET ON MANAGERIAL ENTRENCHMENT By RALPH M. SCHOLTEN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

Dedicated to Kristi

PAGE 3

iii ACKNOWLEDGMENTS I wish to thank all the faculty members in the Department of Finance at the University of Florida for a great five years. In particular, I wish to express my gratitude to Dave Brown, Mahendrarajah Nimalendran, and Chris James for their helpful discussions. I am also fortunate to have Mark Rush fr om the Department of Economics serve on my committee. Finally, I am especially grateful to Mike Ryngaert for his continuous support and encouragement throughout my years at Florida.

PAGE 4

iv TABLE OF CONTENTS ACKNOWLEDGMENTS..................................................................................................iii LIST OF TABLES............................................................................................................. .vi LIST OF FIGURES...........................................................................................................v ii ABSTRACT................................................................................................................... ..viii CHAPTER 1 INTRODUCTION........................................................................................................1 2 EVOLUTION OF CORPORAT E GOVERNANCE MECHANISMS........................5 2.1 The Corporate Takeover Market........................................................................5 2.2 External Pressure from Institutions and Shareholders.....................................10 2.2.1 Role of Institutional Investors.................................................................10 2.2.2 Shareholder Activism..............................................................................12 2.3 Improved Internal Controls..............................................................................14 2.3.1 Board Size and Composition..................................................................14 2.3.2 Director Incentive Compensation...........................................................16 2.4 Interaction between Internal and External Control Markets............................18 3 INVESTMENT DECISIONS AND MANA GERIAL DISCIPLINE: EVIDENCE FROM THE TAKEOVER MARKET........................................................................25 3.1 Motives and Wealth Effects of Takeovers.......................................................28 3.2 Data and Methodology.....................................................................................29 3.2.1 Sample of Acquisitions...........................................................................29 3.2.2 Disciplinary Actions by External Control Market..................................31 3.2.3 Disciplinary Action by the Board...........................................................34 3.2.4 Measuring the Wealth Effects of the Acquisition...................................36 3.3 Results..............................................................................................................37 3.3.1 Univariate Results...................................................................................37 3.3.2 Logistic Regression of Managerial Turnover.........................................42 3.4 A closer Look at the Determ inants of Forced Turnover..................................44 3.5 Concluding Remarks........................................................................................48

PAGE 5

v 4 IMPLICATIONS OF DEFEATING TAKEOVER BIDS: EVIDENCE PRE AND POST TIME-WARNER.............................................................................................63 4.1 Previous Literature and Hypothesis Development..........................................66 4.1.1 Managerial Entrenchment, Takeover Bids, and Agency Theory............66 4.1.2 Hostile Takeovers and Target Defenses.................................................68 4.1.3 Hypotheses Regarding Defeated Takeover Bids in 1980s and 1990s....70 4.2 Data Description and Methodology.................................................................71 4.2.1 Sample of Defeated Takeover Bids........................................................71 4.2.2 Target Characteristics.............................................................................76 4.2.3 Bidder Characteristics.............................................................................78 4.3 Takeover Defenses and Leverage....................................................................80 4.3.1 Takeover Defenses..................................................................................80 4.3.2 Leverage as a Takeover Defense............................................................83 4.4 Target Performance..........................................................................................84 4.4.1 Event Study Analysis..............................................................................84 4.4.2 Post-Offer Valuation...............................................................................86 4.4.3 Multivariate Analysis..............................................................................89 4.5 Post-Offer consequences to Firms and Management.......................................94 4.5.1 Target Independence...............................................................................94 4.5.2 Managerial Turnover..............................................................................95 4.6 Concluding Remarks........................................................................................97 5 CONCLUSION.........................................................................................................112 LIST OF REFERENCES.................................................................................................114 BIOGRAPHICAL SKETCH...........................................................................................120

PAGE 6

vi LIST OF TABLES Table Page 3.1 Description of the sample.........................................................................................50 3.2 Frequency and type of disciplinary events...............................................................51 3.3 Abnormal stock market performance of acquiring firms..........................................52 3.4 Differences of abnormal re turns of acquiring firms.................................................54 3.5 Frequency and type of control-redu cing event for good versus bad acquirers.........56 3.6 Pooled logit regression on CEO turnover and forced CEO turnover........................58 3.7 Ownership and board charac teristics of the 1992-1998 sample...............................59 3.8 Pooled logit regression on CEO tur nover and forced CEO turnover for the 19921998 sample..............................................................................................................60 4.1 Sample selection.......................................................................................................99 4.2 Sample of defeated hostile deals.............................................................................100 4.3 Target characteristics..............................................................................................101 4.4 Bidder characteristics..............................................................................................102 4.5 Takeover defenses...................................................................................................103 4.6 Target leverage changes..........................................................................................104 4.7 Event returns...........................................................................................................10 5 4.8 Post-acquisition valuation.......................................................................................106 4.9 OLS regression on Reval1 and Reval2...................................................................107 4.10 OLS regressions on Discount1 and Discount2.......................................................108 4.11 Distribution of delisted targets................................................................................109 4.12 CEO turnover statistics and l ogit regression on CEO turnover..............................110

PAGE 7

vii LIST OF FIGURES Figure Page 2.1 Volume of takeover activity of U.S. firms scaled by GDP......................................20 2.2 Proportion of takeover o ffers that are hostile...........................................................21 2.3 Proportion of tender offers rejected by targets.........................................................22 2.4 Dollar volume of junk bonds issued each year........................................................23 2.5 Proportion of U.S. equity he ld by institutional investors.........................................24 3.1 Stock price reaction to acquisition announcements for firms that subsequently become a hostile or friendly target...........................................................................61 3.2 Stock price reaction to acquisition announcements for firms that subsequently experience CEO turnover.........................................................................................62 4.1 Target leverage changes surrounding unsuccessful hostile takeover attempts......111

PAGE 8

viii Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE IMPACT OF THE CHANGES IN THE TAKEOVER MARKET ON MANAGERIAL ENTRENCHMENT By Ralph M. Scholten August 2004 Chair: Michael Ryngaert Major Department: Finance, In surance, and Real Estate It is often argued that the takeover market is the only effective corporate control mechanism that penalizes underperforming ma nagers. Over the past two decades, the takeover market has experienced substantial changes and m oved from an active hostile environment in the late 1980s to a more friendly and less contentious environment in the 1990s. Some researchers have argued that mana gers have become more entrenched as a result of these changes. The goal of this disse rtation is to examine the implications of the changes in the takeover market and the lega l environment on managerial entrenchment. Chapter Three focuses on the relative im portance of board monitoring and the hostile takeover market in disciplining mana gers that make poor acquisition decisions. The results show that there is little eviden ce that takeovers that are poorly received by capital markets result in a firm becoming the target of a hostile takeover. However, a forced turnover of a top executive is negativ ely related to the abnormal return associated with an acquisition announcement. The relationship between forced turnover and

PAGE 9

ix negative acquisition returns is stronger duri ng periods of less intense hostile takeover activity. Hence, it appears that being disciplined for making a poor acquisition is more a function of internal discipline than the workings of the takeover market. Chapter Four examines the claim that the Delaware Supreme Court decision of the so called just say no takeover defens e in the Time-Warner case in 1989 greatly undercut the ability of hostile takeovers to di scipline entrenched management teams. I examine this claim by analyzing firms that de feated hostile takeovers in the pre and post Time-Warner periods. I find that the availability of the just say no takeover defense in the 1990s significantly lessened the use of gr eenmail, standstill agreements, share repurchases, special dividends and stock placem ents with friendly third parties. Despite the changes in defensive tactics, there are litt le differences in target revaluations between the 1980s and 1990s. The market price of targets after defeating an offer relative to the offer price is virtually identical for both pe riods as are target abnormal returns over the takeover contest period. Firms are also less li kely to remain independent, and managers more likely forced out in the two year s after defeating a bid in the 1990s. Collectively, the results do not support the argument that the changes in the takeover market and the legal environment ha ve exacerbated the pr oblem of managerial entrenchment. Instead, the results are largely consistent with the idea that alternative control mechanisms such as monitoring by institutions and shareholders and improvements in internal corporate governance in terms of smaller, more independent boards, and director incentive compensation are at least partially replacing the need for an active hostile takeover market.

PAGE 10

1 CHAPTER 1 INTRODUCTION The problem of so called managerial entrenchment in corporations is characterized by two key conditi ons. The first condition is th e presence of a management team that is unwilling and/or unable to put in place the most value maximizing strategy for shareholders. The second condition is the in ability to remove that management team from office. One popular argument is that the problem of entrenched management has become more severe in recent times as a result of changes in takeover rules and regulations in the late 1980s. Jensen (1993) argues that the corporate takeover market can initiate changes in mismanaged firms within a reasonably short peri od of time, but that boards are frequently too slow to effect a major change of flawed corporate policies. Jensen maintains that there are no effective timely alternatives to the takeover market in disciplining managers and their firms, and argues that few boards will initiate actions, such as restructuring and replacing manage ment, without explicit pressure from the takeover market. Jensens argument would suggest that managerial entrenchment has increased during the 1990s as corporate takeover defenses became stronger, tougher state takeover rules predominantly favored targets, and junk bond financing for hostile takeovers diminished. In fact, the leve l of hostile takeover activity dropped significantly in the 1990s relative to the 1980s. The decline in hostile takeover activity, however, has not occurred in isolati on. Alternative external and particular ly internal governance

PAGE 11

2 mechanisms have also seen dramatic change s over the past two d ecades. U.S. equity ownership of institutional holders has increased steadily, laws and regulations regarding shareholder meetings have been relaxed, boards have become smaller and more independent, and outside board members increa singly receive much of their remuneration in the form of incentive compensation. Hence, while the hostile ta keover market has the potential to limit managerial entrenchment, it is plausible that these developments have offset the need for an act ive hostile takeover market. The primary goal of this dissertation is to investigate the impact of the changes in the takeover and legal environment on mana gerial entrenchment. I examine two key managerial actions that are often associated with entrenchment. The first action is the decision to make a large acquisition. The motiv es for acquisitions are often questionable, and managers are often accused of empire-building as opposed to value-maximizing when major acquisitions are announced. Specifi cally, I focus on the effectiveness of the takeover market and the board of directors in disciplini ng managers that make poor acquisition decisions. I compare the disciplinar y events that acquirers face in the 1980s relative to the 1990s. The results indicate that there is no relationshi p between the wealth effects of acquisitions and the probability of becoming a take over target in either period. However, I find that managers are more lik ely to be replaced as a result of valuedestroying acquisitions in the 1990s. The evidence is consistent with the idea that boards have become more responsive in replaci ng poorly performing managers despite the decline in hostile takeover activity. The second action I examine is the def eat of an unsolicited takeover offer. Takeover bids are highly public and shares of unsuccessful takeove r targets generally

PAGE 12

3 trade at prices below the offer price. Mana gement is often heavily criticized for not accepting takeover offers and therefore fa iling to maximize shareholder wealth. Arguably, it has become easier for managers to mount passive just say no takeover defenses with the widely accepted poison pill defense in the 1990s. Consistent with this belief, I find fewer alternative defensive tactics such as share repurchases, recapitalizations, and greenmail ar e used in the 1990s relative to the 1980s. On the other hand, the revaluation of the targets and the pric e gap between what is offered in the deal and what the target trades for after the contest are very similar between the 1980s and 1990s. Furthermore, firms are slightly more likely to subsequent ly get acquired and replace managers after defea ting a hostile bid in the latter period. Hence, the results do not support the theory that management teams th at defeat hostile take over bids can stray further from value maximization in the 1990s relative to the 1980s. Collectively, the results are consistent with the idea that boards can function without the explicit pressure from the host ile takeover markets. More importantly, it appears that managers have generally beco me less entrenched over time. The results suggest that alternative governance mechanis ms have largely substituted for declining hostile (disciplinary) takeover markets. This has important implications for takeover defenses and the nature of ta keover activity in the future. It is possible that hostile takeover activity as seen in the late 1980s will not return as alternative control mechanisms become more important in disciplining firms that deviate from valuemaximization. If this is the cas e, then modern takeover defens es, in particul ar the poison pill defense, will have less far reachi ng consequences than once thought.

PAGE 13

4 This dissertation is outlined as follows. The second chapter describes in detail the major changes in the takeover market and corporate governance mechanisms over the past two decades. Section 2.1 describes the changes in the takeove r environment between 1980 and 2000 and offers some explanations why hostile takeover activity declined post 1989. Section 2.2 discusses the role of institutional holders a nd large shareholders in the monitoring process. Changes in the size a nd composition of boards and their impact on valuation are discussed in s ection 2.3.1. Section 2.3.2 discusses the impact of increased reliance on director equity based compensa tion on managerial entrenchment. Chapter 2 concludes with a discussion on the interacti on of the internal and external control mechanisms in Section 2.4. Chapter 3 and 4 empi rically test the impact of the changes in the corporate governance mechanisms on mana gerial entrenchment. Chapter 3 examines the managerial decision to acquire anot her firm and focuses on the intertemporal relationship between the wealth effects of acquisitions and managerial turnover. Chapter 4 compares the consequences of targets that successfully defeat unwanted acquisition attempts in the 1980s and 1990s. Concluding re marks and implications are discussed in Chapter 5.

PAGE 14

5 CHAPTER 2 EVOLUTION OF CORPORATE GOVERNANCE MECHANISMS 2.1 The Corporate Takeover Market It has long been recognized that the separation of owners hip and control in modern firms can create a situation in whic h management may have incentives to stray from value-maximization. There are several in ternal and external control mechanisms designed to prevent, or at least reduce, these so called agency costs. One of these mechanisms is the external takeover market. Manne (1965) is the first to recognize the ability of the takeover market to act as a disciplinary mechanism. He argues that large observable deviations between actual and pot ential stock prices could induce outside parties to acquire the firm and operate it more efficiently. In other words, underperforming firms will likely face pressure from the takeover market. While the takeover market certainly has the potential to discipline underperforming firms, it was relatively weak until the 1980s. Comment and Schwert (1995) document that fewer than 0.5 percent of public firms received a takeover bid in month in 1975 compared to 1.5 percent of firms in 1987 and 1988. In addition to the increase in the number of takeover events in the 1980s, there was also an emergence of so called hostile takeover activity. Hostile or uns olicited takeover offers were often made by private firms and individuals such as Ca rl Icahn and T. Boone Pickens, with the objective to replace management and/or re structure firm operations. Targets of

PAGE 15

6 unsolicited takeover attempts ofte n rejected those offers and tried to fend off bidders by paying greenmail or by initiating a massive firm restructuring of their own. The literature offers several explana tions for the emergence of the hostile takeover market. One explanation offered by Je nsen (1993), is that the changes in the corporate control markets were long overdue. He asserts that the hostility in the takeover markets was in response to the failure of the internal control mechanisms (or the board of directors) to put in place value-enhancin g strategies. Changes in technology and regulation had led to a large amount of ex cess capacity in many firms in the 1980s. However, managers were often reluctant to restructure their operations and pay out free cash flow to shareholders in the form of dividends. Instead, the practice of empirebuilding continued despite the growing di ssatisfaction of shar eholders. Leverage increases through leveraged acquisitions, buyouts, and stock buybacks successfully eliminated free cash flow, by forcing firms to become more efficient in order to service debt payments (Jensen, 1986). It is not clear that excess capacity is the only factor driv ing the takeover market in the 1980s as Jensen (1993) suggests. If firm s with excess capacity spend too much on capital expenditures, we should expect to find that these companies spen d less after being acquired. The empirical evidence is mixed. Ka plan (1989) reports that management buyout firms made large cuts in capital expenditures. Serv aes (1994), however, does not find evidence that targets are overinvesting prior to being acqui red. Moreover, Healy, Palepu, and Ruback (1992) report that the pre and post acquisition ratio of capital expenditures to sales do not significan tly differ for targets in the 1980s.

PAGE 16

7 Another possible explanation for the increasingly hostile takeover environment in the 1980s was the dissatisfaction of the dive rsifying mergers that firms made just a decade earlier. Shleifer and Vishny (1988) ar gue that the takeover wave of the 1980s was to a large extent a response to the di sappointment with conglomerates. Companies divested unrelated busin ess to increase their corporate fo cus. Their view supports the idea that conglomeration, at least in hindsight, was a mistake, and managers were too slow or unwilling to realize this until the ta keover market pressured them. Regardless of the motives of the hostile takeover market, most early empirical evidence shows that targets of hostile take overs were underperforming their benchmarks. Morck, Shleifer, and Vishny (1989) find that hostile targets are characterized by lower growth and lower Tobins Q compared to friendly targets. Mitchell and Lehn (1990) document that hostile targets make acquisiti ons that, on average, reduce shareholder value. Martin and McConnell (1991) provid e evidence on the disciplinary motives of takeovers by reporting th at targets of successful tender o ffers experience abnormally high managerial turnover after being acquired. Agrawal and Walkling (1994) argue that takeover bids generate information about the firm that is used by the labor markets to discipline managers. In part icular, they document a negati ve relationship between the targets performance prior to the takeover a nd subsequent managerial turnover. The last two papers interpret their results as ev idence for the disciplinary motive of many takeovers. More recent evidence, however, suggests that hostile and friendly targets may not be so different after all. Agrawal and Jaffe (2004) extensively examine the stock price and operating performance for targets of hos tile takeover activity and fail to find

PAGE 17

8 evidence that targets underperform their benchmarks. They document that firms that receive hostile takeover attempts have simila r operating performance as firms that receive friendly bids. Similarly, Schwert (2000) finds that hostile and frie ndly targets cannot be distinguished from each other in terms of their stock price and accounting performance. He argues that hostile offers reflect a different bargaining strategy, where hostile offers are disclosed earlier in the deal process. Despite the conflicting evidence on the pre-takeover performance of hostile targets, many academics agree that the hostile takeover market of the 1980s had positive effects on firms and their valuation. Targets of hostile offers experienced on average positive revaluations even if the attempt fa iled (Bradley, Desai, and Kim (1983), Jandik and Makhija (2003)). Mitc hell and Mulherin (1996) report that nearly half of all major corporations received a takeover offer in the 1980s. As evidence for the effectiveness of hostile takeovers, firms that were not succe ssfully acquired frequently responded to hostile pressure with a co rporate restructuring. After the active takeover period of th e 1980s, the number of takeovers dropped significantly in the early 1990s. Figure 2.1 illust rates the total acquisition volume scaled by GDP over the 1980-2000 period. The figure s hows that acquisition volume was over 4 percent of GDP during the peak of the hos tile takeover market in 1988, and declined significantly in 1990 to only 1.5 percent. After the 1990-1992 Recession, however, the takeover market recovered and acquisition volume rebounded and soon exceeded levels of the 1980s. In 2000, during the height of the internet bubble, acquisition volume scaled by GDP reached an all-time high of over 15 percent.

PAGE 18

9 While the takeover markets rebounded str ongly after 1992, the nature of the acquisitions changed drastically as the fre quency of hostile takeover attempts declined substantially. Figure 2.2 shows the proportion of all acquisitions of publicly traded firms that are classified hostile or unsolicite d from 1980 to 2001. Figure 2.3 illustrates the proportion of tender offers which were opposed by management and hence perceived as hostile. Both figures clearly illustrate the dramatic drop in the fraction of hostile attempts after 1989. Consistent with the figures, Andrade, Mitchell and Stafford (2002) report that the percentage of CRSP firms that receive d a hostile offer dropped from 14.3% in 19801989 to 4.0% in 1990-1998. The decline in hostile takeover activity coin cided with two major events. First, the size of the junk bond market de clined significantly in 1990. Figure 2.4 shows the junk bond market volume over 1980-2000. The pattern is similar to the volume of acquisitions in Figure 2.1. After a sharp increase in the 1980s, the junk bond market dried up in 1990 largely as a result of the insider trading scandals involving Michael Milken (See e.g. Livingston and Williams (2003)). Without access to the junk bond market, many wouldbe hostile acquirers, such as private firms and investors, were unable to finance acquisitions (Comment and Schwert (1995)). Second, and perhaps more importantly, the regulatory environment surrounding hostile takeovers changed. One court case in pa rticular, the Time-Warner decision in July 1989, made it more difficult for acquirers to obt ain control of a target firm without the approval of management. The court favored Time, the target of a hostile offer by Paramount Communications, and allowed Time to keep its poison pill in place. This court decision set a precedent in that targets could protect themse lves from unwanted acquirers

PAGE 19

10 by adopting poison pills and cons equently making hostile offers prohibitively expensive. Figure 2.4 illustrates that, the junk bond market has recovered since the early 1990s, and the volume of junk bond issuance has risen well above levels seen in the 1980s. At the same time, however, the level of hostile takeover activity di d not return s uggesting that the anti-takeover legislation has had a nontrivial effect on the takeover markets. While hostile takeover activity declined, other corporate governance mechanisms began to play a larger role in the 1990s. In particular shareholders and institutional investors became more active in monitoring mana gement. Also, internal controls such as board composition, size, and incentive compensa tion have received more attention. It is plausible that a combination of these alte rnative monitoring and incentive mechanisms partially substitute for the absence of hostile takeovers in resolving agency problems. The next sections will describe so me of the most important changes in corporate governance over the past two decades. 2.2 External Pressure from Institutions and Shareholders 2.2.1 Role of Institutional Investors Institutional equity ownership of U.S. firms has increased dramatically in recent years. Gompers and Metrick (2000) report that large institutional shareholders almost doubled their share of U.S. stock mark et ownership between 1980 and 1996. Figure 2.5 illustrates the increase in the proportion of U.S. corporate equity owned by insurance companies, mutual funds, and private pensi on and government retirement funds between 1980 and 2002. Consistent with Gompers and Metrick (2000) the fi gure clearly shows that aggregate ownership by these institutions increased from approximately 30 percent

PAGE 20

11 in 1980 to over 50 percent in 2001. The bi ggest increase comes from mutual fund holdings; while their holdings were small in 1980, their share of the U.S. equity market rose to over 20 percent in 2001. Due to the nature of the business, institutional investors are primarily concerned with stock price performance. Of course, in place of monitoring, institutional investors could opt to sell their shares of firms with below acceptable performance. In fact, Parrino, Sias, and Starks (2003) document that the number of institutions and aggregate institutional ownership decline in the year prior to forced CEO turnover, suggesting that institutional selling occurs as a result of di ssatisfaction. However, they also report that many institutional investors refrain from selli ng, which may be partially explained by the fact that many institutions have selling restri ctions, in particular when they track one of the broader indices. Institutional investors gene rally hold relatively large e quity stakes and hence have greater incentives to take an active role in monitoring management than do small atomistic shareholders, since the benefits they receive from monitoring are greater than the costs they bear (Grossman and Hart ( 1980), Shleifer and Vi shny (1986)). Also, because of their large stakes, institutions can exert more influence on corporate decisions. Consistent with this argument, Denis and Serrano (1996) find that large unaffiliated blockholders, which are often institutional inve stors, play a large role in a boards decision to replace management. They report that almost 60 percent of the CEO turnovers in firms that successfully defeat a tender o ffer are preceded with the addition of outside blockholders. In contrast, only a fourth of th e firms that do not experience CEO turnovers

PAGE 21

12 add a large unaffiliated blockholder in the tw o years subsequent to the unsuccessful takeover contest. It is natural to think that institutional investors are involved in improving other corporate governance mechanisms of the firm as well. In fact, besides directly monitoring the firm and its management, proposals to impr ove the internal control systems can have tremendous value. Hartzell and Starks (2003) document that institutional ownership concentration is positively related to pay-for-performance sensitivity of executive compensation. Moreover, they find an inverse relationship between in stitutional holdings and the level of compensation, even afte r controlling for firm size, investment opportunities and performance. 2.2.2 Shareholder Activism In addition to the emergence of instit utional holders, shareholders have also become more active in the 1990s as disclosu re requirements increased and regulations concerning shareholder meetings and comm unications were relaxed. Starting in 1992, firms are required to include a graph of their firms stock price performance relative to industry peers in their annual proxy statemen t. While the timeline and peer firms to which the firm is compared are at the di scretion of management, the law has arguably resulted in an increased focus on stock price performance. At the same time in 1992, the Securities and Exchange Commission (SEC) also lowered the costs of mounting proxy figh t contests that challenged boards and management teams. In the past, shareholde rs were required to file detailed proxy statements with the SEC before communicatin g with more than ten other shareholders.

PAGE 22

13 After the new rule change, shareholders can now freely communicate with each other at any given time, provided that a copy of the discussion is sent to the SEC afterwards (Holmstrom and Kaplan (2000)). This rule ch ange has effectively lowered the cost of coordinating shareholder actions and likely contributes to incr eased shareholder activism. Evidence on increased shareholder activism is provided by Gillan and Starks (2000). They document that public funds and shareholder groups initiated 616 shareholder proposals related to corporat e governance during 1989-1994 compared to only 60 proposals during 1983-1988. Karpoff (2001) summarizes the empirical evidence on the effectiveness of shareholder activis m on target firms. While there is some controversy about the impact of shareholder targeting, he concludes that firms with high institutional ownership and poor stock and ear nings performance are more likely to be targeted. He cautions, however, that share holder activism does gene rally not lead to subsequent changes in earnings or the likeliho od of a change in control. The impact of shareholder proposals targeting a firms gove rnance policies on managerial turnover is also mixed. While Huson (1997) reports an increased rate of CE O and board turnover during the three years following shareholder targeting, Opler and Sokobin (1997) report a decrease in the rate of CEO turnover among their samp le of firms that were the target of shareholder activism. In conclusion, it appe ars that shareholder proposals and private negotiations have made at best minor change s in firms governance rules, and are more effective when combined with other governance mechanisms.

PAGE 23

14 2.3 Improved Internal Controls 2.3.1 Board Size and Composition The goal of the board of directors is to reduce the agency costs between shareholders and management. However, their fiduciary duties of monitoring and offering advice to management are often a point of controversy. Even though shareholders ultimately elect board members using proxy voting once a year, management and the incumbent board nominat e the slate of directors. This process potentially causes serious govern ance issues when entrenched managers elect insiders and others close to management that are unlikely to be effective monitors.1 Several studies have examined the link between board characteristics and firm valuation and decisions regard ing managerial turnover. So me researchers have argued that board size plays an important role in the decision making process. Jensen (1993), for example, argues that boards are less likely to function effectively once they consist of more than 7 or 8 directors. In addition, he claims that CEOs more easily control larger boards since each board member is less likely to question leadership. There is some empirical evidence that larger boards are less efficient. Ye rmack (1996) finds an inverse relationship between board size and Tobins Q. He also documents that smaller boards are more likely to dismiss CEOs for poor perf ormance, and that the threat of dismissal declines as board size increases. Kini, Kracaw, and Mian (1995) show that board size is reduced after successful tender offers for unde rperforming firms, which suggests that the takeover market helps improve in ternal control mechanisms. 1 The nomination process for directors is in the process to undergo major changes. The Securities and Exchange Commission has proposed a set of rules that would allow long-term shareholders with a minimum stake of one percent to nominate on e director to the slate of directors.

PAGE 24

15 The majority of research in this area has focused on the composition of the board of directors. It is widely be lieved that outside directors, us ually defined as directors with no immediate ties to management or the firm, are better monitors than inside directors. The empirical evidence is largely consistent with this argument. Since the hiring and firing of CEOs is the one of the primary fu nctions of the board of directors, several papers examine the impact of board compos ition on the relationship between managerial turnover and performance to evaluate th e quality of the board. Weisbach (1988) documents that the link between managerial turnover and performance is stronger for firms with outsider-dominated boards. Bo rokhovich, Parrino, and Trapani (1996) find that outsider-dominated boards are more lik ely to consider a broader range of CEO replacements. In addition, they find that the stock market reaction is significantly higher for outsider appointments compared to insi der appointments and conclude that their evidence suggests that outsider-dominated board s are more beneficial to shareholders when changes in firm policies are important. Other papers focus on the relations hip between board composition and firm valuation. Evidence by Byrd and Hickman (199 2) on bidders and Cotter, Shivdasani, and Zenner (1997) on targets suggests that inde pendent boards increase value for their shareholders during an acquisition attempt. Brickley, Coles, and Terry (1994) find that markets only react positively to poison pill adoptions for outsider-dominated boards. Their finding implies that investors anticip ate independent boards to select takeover defenses with shareholder valuation in mind. Finally, markets generally perceive independent directors to be va luable, as the addition of inde pendent directors to the board is associated with a positive stock pri ce reaction (Rosenstein and Wyatt (1990)).

PAGE 25

16 The results of these papers generally s uggest that smaller and more independent boards function as better monitors. The impli cation is that firms (and regulators) have made significant steps to reduce board size a nd include more independent directors over the last few decades (Hermalin and Weisb ach (2003), Borokhovich, Parrino, and Trapani (1996)). Hence, if smaller and more independent boards are better monitors, and boards have become smaller and more independent ov er time, then there is reason to believe that, on average, boards have become more ef fective monitors inde pendent of pressure from the takeover market. 2.3.2 Director Incentive Compensation The primary function of the board of di rectors is to represent shareholder interests. Legal and structural mechanisms are designed to ensure that outside board members act independently of management. In the past, however, the judicial system has been slow to act to discipline directors for violating their fi duciary responsibilities towards shareholders. Perhaps a greater ince ntive for directors to act in shareholder interest is the labor market for directors. Fama and Jensen (1983) argue that outside directors have incentives to signal to the la bor market that they add value through their insights and monitoring. Ther e are several papers that document the relationships between firm performance and number of di rectorships. Kaplan and Reishaus (1990) document that dividend decreases have implicat ions in the director labor market. They find that executives of firms that cut dividends are less likely to receive additional directorships than are other executives. Kini, Kracaw, and Mian (1995) show that boards of the combined firm tend to be constructed in a more balanced manner than the target

PAGE 26

17 board following the acquisition (e.g. insider-d ominated boards are replaced by boards with more outside seats). The primary source of director compensa tion is in the form of a cash-retainer. Brickley, Coles, and Linck (1999) document for a sample of large firms, that directorships can be very lu crative with the average di rector earning almost $45,000 in annual compensation. In addition, they find that directors often obtain profitable consulting opportunities with the firm th at pay anywhere from $20,000 to $83,333 per month. The problem with cash-retainers is that they can distort the motives for directors, especially in the event of a takeover offer. Ha rford (2003) uses a samp le of firms that are the target of successful tender o ffers to show that board positi ons are difficult to replace, and concludes that there is a large financia l cost for failing to monitor management. To provide better incentives, independe nt boards have increasingly received a larger portion of their rem uneration in the form of equ ity-based compensation. Perry (1999) documents that the fraction of compensa tion for directors that is incentive-based increased from 25% in 1992 to 39% in 1995. Ryan and Wiggins ( 2003) report over a 58% increase of equity based compensation for a sample of 600 firms between 1995 and 1997. They also find that firms with inde pendent boards and weak er CEOs are more likely to pay directors incentive compensation. Yermack (2003) finds that total outside director remuneration is very sensitive to firm performance with a change in wealth of about $285,000 for a one standard deviation cha nge in firm performance for the largest publicly traded firms. Incentive compensation for independent directors has a positive impact on board monitoring and as a result on firm performa nce. Perry (1999) documents an increased

PAGE 27

18 probability of CEO turnover following poor pe rformance for firms where independent boards receive incentive compensation. Becher Campbell, and Frye (2004) show that equity-based compensation for bank directors has increased drastically after the deregulation in the early 1990s. In additi on, they find a positive relationship between bank performance and the level of equity-b ased compensation for directors. They conclude that incentive compensation incr eases the effectiven ess of monitoring. 2.4 Interaction between Internal and External Control Markets Because of the decline in hostile takeover activity in the early 1990s, there has arguably been a shift to rely on other co ntrol mechanisms to step up monitoring. However, Jensen (1993) argues th at alternatives to the mark et for corporate control are ineffective. In particular, he argues that boards have not adequately fulfilled their responsibilities of hiring, fi ring, and properly compensate CEOs. In his view, few boards have initiated necessary cha nges concerning top management in the absence of external crises. The empirical evidence on the impact of the takeover market on the effectiveness of board monitoring is mixed. Hadlock and Lumer (1997) and Mi kkelson and Partch (1997) provide support for Jensens assertion, an d find that takeover activity is necessary for the board of directors to apply effectiv e disciplinary pressure. Hadlock and Lumer document lower than normal turnover rates during the 1933-1941 period that was also characterized by few corporate control transa ctions. Using more recent data, Mikkelson and Partch report that as the takeover mark et became less active in the early 1990s, the link between managerial turnover and poor performance weakened. They find that

PAGE 28

19 management turnover in unacquired firms is substantially highe r during the active takeover period of 1984-1988 compared to th e inactive period of 1989-1993. Moreover, they argue that the change in the takeover regime, from hostile a nd contested to more friendly takeovers, may result in less manage rial discipline among poor performing firms in the future. These studies suggest that when the threat of a hostile takeover diminishes, boards face less external pressure to replace underperforming managers. In contrast, Denis and Kruse (2000) a nd Huson, Parrino, and Starks (2001) find evidence that boards of directors and the takeover market are substitute mechanisms. In their view, the effectiveness of the board of directors is independent of the takeover market. Denis and Kruse (2000) report that while the number of disciplinary events underperforming firms face during the inactive takeover period declines, the frequency of internally precipitated CEO turnovers is ac tually higher. Huson et al. (2001) document that the number of forced resignations a nd outside successions increases between 1971 and 1994. More importantly, they find no evid ence that the sensitivity of forced CEO turnover to performance varies with the intensity of the takeover market over that period. Hence, there is no consensus on the effect of the takeover market on board monitoring and managerial entrenchment. It is important to realize th at the studies above use data from the early 1990s. Since substant ial changes have occurred during the latter part of the 1990s, a reexamination of the imp act of the changes in the takeover market on managerial entrenchment is valuable. In part icular, the changes in the legal environment may have permanently altered the nature of the takeover markets. The next two chapters examine whether the changes in the takeover ma rket are at least partially offset by the changes in the other corporate governance mechanisms.

PAGE 29

20 0 2 4 6 8 10 12 14 16 18 19801982198419861988199019921994199619982000 Year Percentage Figure 2.1 Volume of takeover activity of U.S. firms scaled by GDP Sources: Data on takeover activity are from Mergerstat (various issues). GDP data are from the U.S. Department of Commerce: Bur eau of Economic Analysis.

PAGE 30

21 0 5 10 15 20 25 30 19801982198419861988199019921994199619982000 Year Percentage Figure 2.2 Proportion of takeovers that are hostile Source: Thomson Financial SDC database. The selectio n criteria are that both acquirer and target are publicly traded, the deal value exceeds $50 million, an d the deal attitude is either friendly, unsolicited, or hostile. Unsolicited deals are also considered hostile.

PAGE 31

22 0 5 10 15 20 25 30 35 40 45 19801982198419861988199019921994199619982000 Year Percentage Figure 2.3 Proportion of tender offers rejected by targets Source: Mergerstat (various issues)

PAGE 32

23 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 19801982198419861988199019921994199619982000 Year Volume in millio n Figure 2.4 Dollar volume of junk bonds issued each year Source: Table 2 from Livingston, Miles, and Glenn Williams, 2003, Drexel Burnham Lambert, underwriting fees, and market power, Working paper, University of Florida.

PAGE 33

24 0 10 20 30 40 50 60 19801982198419861988199019921994199619982000 Year PercentageInstitutional Investors Private Pension and Government Retirement Insurance Firms Mutual Funds Figure 2.5 Proportion of U.S. equity held by institutional investors Source: Federal Reserve Board, Flow of Funds Accounts of the United States, http://www.federalreserv e.gov/releases/Z1.

PAGE 34

25 CHAPTER 3 INVESTMENT DECISIONS AND M ANAGERIAL DISCIPLINE: EVIDENCE FROM TH E TAKEOVER MARKET Considerable research attention has been given to examining the effectiveness of internal and external governance mechanisms in mitigating potential agency costs that arise from the separation of ownership and control. Two of the most controversial mechanisms that have received attention in both the popular press and academia are hostile takeovers and the board of directors. While it is generally believed that hostile takeovers are effective in di sciplining at least some poorly performing firms and their managers, the effectiveness of the board of directors is less clear. Jensen (1993) comments, “It appears that internal control sy stems have two faults. They react too late, and they take too long to effect a major change .” In addition, he also argues that boards rarely act without explic it disciplinary pressure from capital markets. The empirical evidence on the impact of the takeover market on the effectiveness of board monitoring is largely an unresol ved issue. Hadlock and Lumer (1997) and Mikkelson and Partch (1997) pr ovide support for Jensen’s as sertion. They find that managerial turnover related to poor perfor mance is higher duri ng periods of more takeover activity. Both studies suggest that when the threat of a hostile takeover diminishes, boards face less ex ternal pressure to replace underperforming managers. In contrast, Denis and Kruse (2000) and Huson, Parrino, and Starks (2001) argue that the takeover market and the board of directors are substitute mechanisms and that the effectiveness of board monitoring is independent of takeover activity.

PAGE 35

26 The goal of this chapter is to provide additional insights on the relative importance of the takeover market and board of directors in disciplining poorly performing managers during periods of high and low hostile takeover activity. More specifically, this chapter a ddresses the question of whet her an active hostile takeover market is necessary for a well-functioning boar d. It is plausible th at an active takeover environment, in which underperforming fi rms are likely targets of hostile offers, stimulates boards to act in the interest of shareholders by genera ting public information about the target and its management or realigning incentives of outside directors.1 It is also plausible, however, that boards can di scipline management independent of these forces. Recent developments in corporate governance, such as shareholder activism, greater focus on stock price performance due to increased institutional holdings, and board incentive compensation, may create an environment where boards feel compelled to discipline management abse nt a hostile takeover threat. To investigate this question, I focus on a sample of 285 firms that announce and complete major acquisitions during the activ e hostile takeover period from 1980-1987 and the inactive hostile takeover period from 1992-1998. I document the number of disciplinary events each acquirer faces from th e external control market in the form of receiving a hostile or friendly takeover offer, and the internal control market in the form of managerial turnover, for up to three y ears after the acquisiti on announcement. Similar to Mitchell and Lehn (1990), I use the wealth effects of the acquisitions, measured by the announcement returns, as a measure of managerial performance. Unlike other 1 Harford (2003) reports that outside directors of target firms are rarely retained and that their financial impact is typically negative following a completed offe r. Hence, there is a cost to outside directors for failing to actively monitor management, forcing the external control market to act for them. This cost is arguably lower when the threat of a takeover diminishes.

PAGE 36

27 performance measures such as long-term stock price performance or operating performance that are influenced by macro economic factors and industry-dynamics, acquisition decisions are publicly observabl e discretionary investments initiated by management, and the success or failure of acquis itions is often attributed directly to the CEO of the acquiring firm. Consistent with Figures 2.2 and 2.3, I document that hostil e takeover activity sharply diminishes in the 1990s. I find no re lationship between the wealth effects of acquisitions and the likelihood of becoming a ( hostile) target. However, deviating from value-maximization has implications for mana gement. I find a strong inverse relationship between the abnormal returns around the acquisition announcement and forced management turnover. Logit regressions indica te that this relationship only holds during the period of less hostile takeover activity in the 1990s.2 In addition, there is no overwhelming evidence that monitoring by active outside investors measurably facilitates internal control efforts. Collect ively, the results indicate th at the internal mechanism does to some degree discipline managers that make poor acquisition decisions, the hostile takeover market appears to be less effective, and there is no enhancement effect associated with the internal mechanism working better in an active hostile takeover environment. The remainder of the chapter is organi zed as follows. Section 3.1 discusses the motives and wealth effects of takeovers. Section 3.2 describes the data and methodology. 2 In a contemporaneous paper, Zhao (2002) tests the relationship between wealth effects of acquisitions and CEO turnover for firms that remain independent over a period of five years. She examines firms that make acquisitions between 1990 and 1997, and also finds a negative relationship between announcement returns and CEO turnover. In this chapter, I investigate the relationship between the wealth effects of acquisitions and managerial discipline from both the external corporate control mark et and the board of directors. In addition, unlike Zhao, my research design allows for making inferences about the effectiveness of board monitoring during a changing takeover environment.

PAGE 37

28 The relationship between the wealth effects of acquisitions and disciplinary actions by the takeover market and board are presented in section 3.3. Section 3.4 contains a more detailed examination of the forced turnove r cases. Conclusions and implications are summarized in section 3.5. 3.1 Motives and Wealth Effects of Takeovers A vast literature has examined the wealth effects of takeovers. Studies generally find that the excess returns to target shareholders are positive, while the excess returns to shareholders of acquiring firms are either ne gative or not statistically different from zero (see e.g. Bradley et. al. (1988) Schwert (2000), and Mulherin and Boone (2000)). Since the bidding firm determines the timing and ini tial offer conditions, much of the research in this area has focused on explaining why acquirers fail to gain in acquisitions. There are several competing theories that attempt to explain the dismal returns most bidders experience around the merger announcement. The first is based on asymmetric information between manageme nt and the capital market. Following the Myers and Majluf (1984) argument, firms tend to issue equity when management believes the current stock pri ce is overvalued, signaling this private information to the market. Empirical evidence suppor ts the asymmetric informa tion theory, and shows that abnormal returns around the announcement date are negative for stock offers and generally insignificant for cash offers (Travl os (1987) and Martin (1996)). Loughran and Vijh (1997) show that this relationship persists during the post-acquisition period. The method of payment, however, does not fully explain the variation in returns for stock acquirers, or why in some instances the market reacts unfavorably to cash

PAGE 38

29 offers. Other theories argue that disappoin ting reactions to acquisition announcements are not the result of the form of financing, but rather caused by non-value-maximizing behavior by management in the form of overpaying for the acquisition. One explanation is based on an empire bui lding/rational agency cost theory that assumes it is impossible, or at least very cos tly, to write contracts that perfectly align the incentives of managers with shareholders. Jensen (1986) and Shleifer and Vishny (1988) argue that managers overpay for targets, not because managers make valuation errors, but to reap personal benefits from acquisitions that are non-value maximizing to the acquiring shareholders. An alternative explan ation for overpayment in acquisitions is managerial optimism. Roll (1986), and more re cently Heaton (2002), argue that managers suffer from ‘hubris’ and are too optimistic a bout takeover prospects. In their view, managers try to maximize firm value, but overestimate the value of targets and as a result overpay. Regardless of whether overpayment is the result of self-dealing or hubris, the interesting question is whether and which control systems step in and discipline management for deviating from value-maximization. 3.2 Data and Methodology 3.2.1 Sample of Acquisitions The initial sample consists of acquis itions that are announced between 1980 to 1987 and 1992 to 1998. The sample periods are constructed to examine the effects of changes in the intensity of the hostile takeover market on internal monitoring. Data on takeovers are obtained from Th omson Financial Securities Data Company and satisfy the following criteria. (1) For each transaction, both bidder and target trade on the NYSE,

PAGE 39

30 AMEX or NASDAQ, with data available on CRSP around the merger announcement. (2) Minimum transaction value of the deal is $100 million for the 1980-1987 period and $300 million for the 1992-1998 period.3 (3) Transactions in whic h either party operates in a regulated industry are exclude d (banks, financials, railroad s, and utilities). (4) The acquirer gains complete control of the target. (5) The ratio of market values of target and bidder is between 10 and 70 percent. The mark et values of equity are computed four weeks prior to the initial announcement, reduc ing the impact of information leakage on the stock price. The selection range for the target-to-bidder market value ratio guarantees that acquisitions will determine a significant part of the acquirer’s future cash flows, but yet precludes a large number of ‘mergers of equals’. Mergers of equals are excluded because they generally experience signifi cant changes in management and board structure, and make CEO succession more difficult to interpret.4 This selection process potentially excludes other ‘major’ acquisitions that the firm completed during the same time period, which may have had a greater im pact on the firm. To address this issue, acquisitions with relative size greater than 70 percent are included only in cases where the bidder makes multiple acquisitions in a three year period. Applying these criteria results in a sample of 334 acquisitions. Nine mergers that exhibit merger of equals char acteristics (equal board repres entation), and five acquirers that experience CEO turnover in the m onth surrounding the takeover announcement are excluded. The final sample consists of 319 acquisitions; 137 during the 1980-1987 period and 182 acquisitions during the 1992-1998 peri od. Since some firms made multiple 3 This chapter focuses on the largest acquisitions during the two subperiods. The rationale for choosing a three-to-one ratio of the deal values is to reflect the increasing firm size over time. The value-weighted CRSP return is 317% over the period between the midpoints of the two samples. The return of the S&P 500 is 233% over the same period. 4 See Wulf (2003) for examples of governance changes in mergers of equals.

PAGE 40

31 acquisitions in a three year period, there are 124 and 161 firms in the 1980-1987 and 1992-1998 periods, respectively.5 Table 3.1 reports the number of acquisiti ons and unique acquirers for each year. The acquisitions are evenly distributed for the 1980-1987 sample. On the other hand, the 1992-1998 sample experiences a surge in acquisitions between 1996 and 1998, with 70 percent of the acquisitions occu rring during those three years. The table also reports the median market values of the bidder and targ ets, the transaction value and the ratio of market values of target to bidder. The ma rket values of the bidder and target are calculated 20 trading days before the acquisitio n. By construction, bidder, target and deal values are significantly larg er for the latter period.6 3.2.2 Disciplinary Actions by External Control Market Since hostile takeovers are more likely to be disciplinary (Morck et al. (1989)), I distinguish between hostile and friendly takeover attempts.7 Similar to Mitchell and Lehn (1990), I classify acquirers as hostile target s when they are targ ets of successful or unsuccessful hostile tender offers, unsolicite d and rejected offers, and proxy contests. Friendly targets are firms that receive frie ndly merger or tender offer proposals, and successful and unsuccessful ma nagement buyouts, and are not the target of any hostile activity. 5 There are 28 firms that make multiple acquisitions: 23 firms make 2 acquisitions, 3 firms make 3 acquisitions, and 2 firms make 4 acquisitions within a th ree year period. One firm experiences a change of CEOs between acquisitions, and both acquisitions are included as separate observations. 6 For the multivariate tests, firm size is adjusted by the CRSP value-weighted index. 7 Schwert (2000) argues that hostile deals cannot be distinguished from friendly deals based on economic terms. He argues that the offers perceived as hostile are at the beginning stages of the bargaining process and are publicized earlier. To take this possibility into acco unt, I also examine th e disciplinary role of friendly takeover offers.

PAGE 41

32 Jensen (1993) argues that changes motiv ated by the corporate takeover markets are frequently accomplished within one and a ha lf to three years. In line with Jensen’s argument, I focus on the number of hostile an d friendly takeover attempts that bidders receive for up to three years following the acq uisition. Table 3.2 reports that there are 44 acquirers (15%) that become either a hostil e or friendly target. Across the two periods, only 19 firms (7%) are subject to hostile or disciplinary takeover offers, and 25 (9%) become friendly targets. The remaining 241 firms do not receive any type of takeover offer. More interesting are the differences in the number of hostile and friendly takeover attempts between the two periods. In the fi rst period, 11% of the acquirers received a hostile takeover attempt within three years after the acquisiti on. In contrast, only 3% of the firms in the 1990s received a hostile offe r. The decline in the fraction of hostile takeovers is significant at the one percent level, and is cons istent with figure 2.1 and the findings of Andrade et al. ( 2002). To support the disciplina ry motive of hostile takeover offers, Denis and Kruse (2000) fi nd a significant decline in disciplinary takeovers that result in target managers losing thei r position from 1985-1988 to 1989-1992. Conversely, there is a significant increase in the number of friendly attempts from the first to the latter period, with 13 percent of the acquirers becoming a friendly target in 1992-1998, compared to 3 percent in 1980-1987. To put the frequency of takeover activity for the 1980-1987 sample in perspective, I compare the numbe r of takeover attempts to th ose reported by Mitchell and Lehn (1990). One key difference between the two methodologies is the length of time considered after the acquisi tion announcement. Mitchell and Lehn classify firms as

PAGE 42

33 targets if they receive a takeover offer for up to seven years after the acquisition. They report that 28 percent of acquirers received an offer for control, where 21 and 7 percent received a hostile and friendly offer, respectively compared to 11 and 3 percent for my sample.8 When I extend the postacquisition period to five years, the frequency of takeover offers in my sample is very simila r. Within five years, 31 (25%) acquirers are targets of takeover attempts, of which 24 ( 19%) are hostile, and 7 (6%) are friendly. While there is not a “right” cut-off poi nt, extending the post-acquisition period beyond three years arguably reduces the “effec tiveness” of the corporate control market. Goodyear’s acquisition of Celeron, which Mitc hell and Lehn (1990) use as an example, illustrates this point. At the announcement of the acquisition, Goodyear’s stock price dropped by about 10 percent, and subseque ntly declined even further. Goodyear, however, did not receive a hostile offer for over three and half years after they made this ‘bad’ acquisition, questioning the effectiveness of the takeover market as a disciplinary mechanism. Moreover, while Goodyear did st art a massive restructuring program after the hostile attempt, the CEO was never repl aced as a result of his poor investment decision.9 8 In Mitchell and Lehn (1990) a firm is classified as a takeover target when it receives a takeover offer between 1982 and 1988. Their sample consists of acquisitions announced between 1982 and 1986. Hence, the length of time between the acquisition and subse quent takeover offers ranges from two years for acquisitions announced at the end of 1986 and up to seven years for acquisitions announced at the beginning of 1982. 9 In my sample, 11 of the 19 hostile takeover targets remain independent. Only 3 firms that are the target of an unsuccessful hostile attempt subsequently experience forced turnover. Hence, th e majority of the CEOs retain their position after an unsuccessful hostile attempt. Thus, even with an explicit takeover threat, it is not clear that boards are much more likely to replace the CEO.

PAGE 43

34 3.2.3 Disciplinary Actions by the Board To examine the effectiveness of the board, I concentrate on board actions that lead to CEO turnover. While boards have alternative mechanisms available to discipline management, such as varying compensa tion and bonus structures, a change in management is a less ambiguous measure of monitoring. For instance, Denis and Denis (1995) find that forced resigna tions signal a turnaround and ar e associated with positive market reactions and large improvements in performance. There are two broad definitions of CEO turnover in the literature. The first definition, used by Weisbach (1988) a nd Mikkelson and Partch (1997), does not distinguish between disciplinary and routine tu rnover. These research ers claim that there is no reliable way to classify the motive for managerial turnove r. They argue that it is especially difficult to detect forced tur novers, since both the board and the CEO have incentives to downplay the succession as retirement. An alternative definition of disciplinary CEO succession is proposed by Huson et al. (2001). If the news reports i ndicate that the CEO is fire d, forced from his position, or leaves for unspecified reasons, then the turnover is classified as forced. In addition, succession is classified as forced when the CEO is under age 60 and the WSJ articles about the succession (1) do not report the reas on for departure due to death, poor health, or acceptance of another positi on, or (2) report that the CEO is retiring, but do not announce the retirement at least 6 months in advance. The analysis relies on both definitions of turnover to ensure that the findings are robust to different definitions of turnover. In this chapter, CEO turnover refers to the succession of the top manager, excluding successions that are the result of health reasons

PAGE 44

35 and death. The method proposed by Huson et al. (2001) is used to classify succession as ‘routine’ (retirement, health reasons, other employment), and ‘forced’ (forced by board, no reasons listed, performance related). Reas ons for CEO turnover ar e obtained from the Wall Street Journal proxy statements, and 10K filings with the SEC. Firms that do not experience CEO turnover are referred to as nonturnover firms. The nonturnover group also contains firms that are acquired before replacing their CEO. In other words, being acquired is not classified as eith er forced or routine turnover. Table 3.2 also reports the frequency of CEO turnover and forced turnover. The rate of CEO turnover is subs tantially higher than the take over rate. Of the 285 firms in the sample, 88 experience CEO turnover in th e three years after the acquisition. The annualized turnover rate is approximately 10%, which is similar to turnover rates reported in previous studies (see, e.g. Hu son et. al. (2001), Denis and Denis (1995)).10 Of these 88 firms, 43 experience a forced change in top officer. About 20 percent experience a disciplinary event, defined as experienci ng non-routine turnover or receiving a hostile offer. Despite the decline in hostile takeover activity from the first to the latter period, the fraction of firms that were subject to disc iplinary turnover increas ed. Of the 43 forced turnovers, only 9 occurred during the 19801987 period. The remaining 34 cases took place in the latter period. The increase in th e fraction of disciplinary turnovers is significant at the one percent level, and indica tes that boards were more likely to act and replace management during the latter period. 10 The annualized turnover rate is calculated as follows: 88 turnovers occur for 285 firms within three years. Hence, the annual turnover rate is equal to (88/285)/3 = 10%.

PAGE 45

36 3.2.4 Measuring the Wealth Effects of the Acquisition Similar to Mitchell and Lehn (1990), I use the market reaction to a firm’s acquisition announcement as an indicator of managerial performance. In estimating the wealth effects of acquisitions, I implicitly a ssume that the capital markets react efficiently to the information of the acquisition announcem ent. In other words, capital markets are able to distinguish between ‘good’ and ‘bad’ acqui rers, and ‘bad’ acqu irers experience more negative stock price reactions at the announcement.11 CRSP daily returns are used to comput e market-adjusted abnormal returns for each acquiring firm. The wealth effects of acqui sitions are measured by the change in the acquirer’s equity value around the announcemen t date. Abnormal returns are estimated over several windows, including the standa rd [-1,1] and [-5,5] estimation windows.12 Since this chapter focuses on large acquisitions where information about the deal is likely to leak out before the announcement, the longer [-5,5] window is arguably a better measure for the wealth e ffects of the acquisition. Market-adjusted returns ar e calculated by differencing the compounded return of the firm and the CRSP value-weighted market index return over the event windows. The mean cumulative abnormal return is obtai ned by taking the average of the individual abnormal returns. Market-adjusted returns are used instead of the market model because acquirers, on average, outperform the market prior to the acquisition. This implies that the intercept term for the market model is li kely to be biased upward and over-estimates 11 Empirical evidence shows that the initial market reaction to acquisition announcements contains information about the long-term consequences of a takeover to the bidding firm and its management. Kaplan and Weisbach (1992), for ex ample, find that acquirer returns and combined (acquirer and target) returns at the acquisition announcement are significantly lower for divestitures that are performance related, than for successful divestitures and acquisiti ons not divested. They conclude that stock market prices correctly react to fundamenta ls during the announcement period, and that managers that ignore the unfavorable signals from the capital market s are motivated by managerial hubris. 12 Qualitatively similar results are obtained for [-5,1] and [-5,40] estimation windows.

PAGE 46

37 the negative impact of the acquisition announcement.13 Moreover, for narrow event windows, abnormal returns do not differ subs tantially between the market and means models (Brown and Warner (1985)). For firm s that make multiple acquisitions, wealth effects are measured by the abnormal return averaged across all its acquisitions.14 3.3 Results 3.1.1 Univariate Results Table 3.3 reports abnormal returns for the full sample and the two sub-periods. Abnormal returns are estimated over the [1,1] and [-5,5] windows a nd are reported in panels A and B, respectively. The a nnouncement returns corresponding to the acquisitions made by the full sample are -2.3% and significantly different from zero at the one percent level for both estimation windows. Of the 319 acquisitions, 103 are financed using cash only, 154 are pure stock acqui sitions, and 36 consist of a mix of cash, stock, and debt. Consistent with previous findings (e.g. Travlos (1987)), announcement returns are more negative whenever acquisition s are financed entirely or partially using stock. The average [-1,1] abnormal return on st ock acquisitions is -3.9%, compared to 0.4% for cash acquisitions. The difference be tween the announcement returns of the cash and stock acquisitions is statistically signi ficant at the one percent level for both 13 The parameters of the market model can be estimated using the equation Rit = i + i Mt + it where Rit is the return of firm i, Mt is the return on the market portfolio, and i and i are the market model parameters. The residuals it can be used as an alternative measure of abnormal performance. Since acquirers, on average, outperform the market, is biased upward, over-estimating the negative impact of the merger announcement (see Schwert ( 2000)). Evidence of this is provided by the market-adjusted Buy-and-Hold abnormal returns (BHAR) one year pr ior to the acquisition. The mean BHAR is 10.05% with a p-value of 0.001. 14 The results reported in the chapter are based on th e equally-weighted return of the acquisitions that the acquirer makes. The results are robust to differen t specifications; using all acquisitions as separate transactions or only the first or last acquisition th at a firm makes does not af fect the significance of the results reported in the chapter.

PAGE 47

38 estimation windows.15 The returns are comparable to H ealy et al. (1992), for example, who find an abnormal bidder return of -2.2% for the largest 50 tran sactions during 19791984. Table 3.3 also reports the average and me dian stock price reactions associated with announcements of acquisitions made by nontargets, targets, hostile targets, and friendly targets, as well as turnover, forced turnover, and nonturnover firms. The results show that nontargets perform slightly better than targets. There is no evidence that acquirers that subsequently become a hostile target perform significantly worse; hostile targets, on average, experien ce a return of -2.7%, and -1.3% for the [-1,1] and [-5,5] windows, respectively. Figure 3.1 graphs the stoc k price reactions for the targets, hostile targets, friendly targets, a nd nontargets from five days prior to 40 days after the acquisition announcement. The figure illustrates that while friendly targets perform worse the few days directly following the announc ement, the difference between the stock market performance of hostile, friendl y, and nontargets disappears over time. The [-1,1] and [-5,5] announcement retu rns by firms in the CEO turnover category are -3.1% and -4.1%, re spectively. In contrast, firm s in the nonturnover category have announcement returns of –1.9% and –1. 6%. Many turnovers are simply normal retirements, or precipitated by factors othe r than poor performance, such as planned succession or other employment opportunities. Therefore, if poor acquisition decisions lead to removal of CEOs, the lowest return s are expected for the forced turnover group. Consistent with this assertion, abnormal re turns are substantially lower for this group relative to any other category. The average re turns for the forced turnover group are 15 The proportions of turnovers for the cash and (partial) stock samples are 11/103 and 32/216 respectively. A chi-square test indicates that th ere is no significant difference be tween the frequencies of turnovers across the two samples, and should therefore not bias the qualitative results.

PAGE 48

39 4.8% and -6.7% for the two estimation windows. The last two rows re port the returns for firms that receive disciplinary action from the hostile takeover market or experience forced turnover. With returns of -4.3% a nd -5.4% for the [-1,1] and [-5,5] windows, the returns for this group are considerably lowe r than for firms that do not experience a disciplinary event. Figure 3.2 depicts the market reaction of the turnover, forced turnover, and nonturnover categories for the [-5,40] event window. Whil e all categories underperform subsequent to the takeover annou ncement, it is clear from the graph that firms that experience forced turnover do s ubstantially worse. The difference in the cumulative abnormal returns between the fo rced turnover and nonturnover samples is approximately 6% after five days. More importantly, unlike the difference between targets and nontargets, the return difference between forced turnove rs and nonturnovers stays practically constant ove r the entire [-5,40] window. Panel A and B of Table 3.4 reports the diffe rence in returns for nontargets versus (1) targets, (2) hostile target s, and (3) friendly targets, fo r the [-1,1] and [-5,5] windows, respectively. The difference in announcement returns betwee n target and nontarget firms is only significant at the ten percent level for the [-1,1] window for the full sample. For any other estimation windows or sub-sample the difference is insignificant. Moreover, there is little evidence that hostile targets do worse than nontarget firms. In fact, hostile targets appear to outperform nontargets during acquisitions during the 1980-1987 period. This result is inconsistent with the findings of Mitchell and Lehn (1990), who find that takeover targets, especially hostile targ ets, have significantly more negative announcement returns surrounding acquisi tion announcements than nontargets.16 16 The difference in results between my study and Mitchell and Lehn (1990) during the 1980s period is likely the result of differences in the sample selecti on. This study focuses on acquirers of public targets

PAGE 49

40 The evidence on managerial turnover is more compelling. The mean and median return differences between nonturnover vers us turnover and forced turnover firms are significant for the full and the 1992-1998 sample In panel B, for example, the average difference in returns between turnover and nont urnover firms is -2.5% for the full sample. That difference is -5.1% when comparing firms that experience forced turnover with firms that experience no turnover. Interestingly, this differen ce is entirely driven by the 1992-1998 sample, which suggests that board scrutiny has increased, and CEOs are dismissed quicker after costly investment mi stakes. Hence, it appe ars that boards have become more effective over time without relia nce on external pressu re from the hostile takeover market. To gain a better understandi ng of the relationship between a “bad” acquisition and subsequent disciplinary actions from the internal and external control markets, the sample is partitioned into quartiles based on the announcement returns. The type and frequency of disciplinary actions are documented for firms in the top and bottom quartiles. For the full sample, the top quartile includes firms that made acquisitions with an average return greater than 1.6% and 2.7% for the [-1,1] and [-5,5] windows, respectively. The average announcement return for firms in the bottom qua rtile is less than -6.1% and -7.4% for the with a minimum deal value of $100 million. Their sample consists of a sample of firms that are covered by Valueline and make acquisitions where the ratio of deal value and bidder market value exceeds 5%. While they do not report the mean or median bidder values, their sample selection allows for many smaller bidders. Shivdasani (1993) and Mitchell and Lehn (1990) show that the probability of becoming a takeover target is inversely related to bidder size. Hence, large firms in the 1980s were more isolated from hostile takeover pressure. The results that follow indicate that the internal governance mechanisms also fail to penalize managers of large firms in the 1980s.

PAGE 50

41 same windows.17 The Chi-squared statistic tests the hypothesis that the proportion of firms experiencing each event is equal between the two samples. Table 3.5 reports the fraction of firms that fa ll in the target, hostile target, friendly target, turnover, forced turnover, and disc iplined categories for both the top and bottom quartiles. These events are not mutually exclusiv e since in a few instances CEO turnover precedes a takeover attempt, and vice versa, an unsuccessful takeover attempt precedes CEO turnover. The results show that there is no difference in the number of takeover or hostile offers “bad” and “good” acquirers receive. Focusing on Panel B, 15 percent of the acquirers in the top quartile become targets, while 13 percent of the acquirers in the bottom quartile become targets. Contrary to what we would expect, 6 percent of the top performers receive a hostile offer compared to 3 percent of the bottom performers. The proportion of forced turnovers is signi ficantly higher for firms in the lowest quartile with 20 of 72 firms (28%) experienci ng forced turnover, compared to 6 of 72 (8%) for firms in the top quartile. Another interesting result is the difference in the proportion of firms that face a disciplin ary action after making a poor acquisition between the two time periods. Of the 31 acquirers with returns in the bottom quartile only 10% or 3 firms face disciplinary actions in the three years following the acquisition. In contrast, of the 41 “bad” acqui rers in the 1992-1998 sample, 44% or 18 firms are either the target of a hostile takeover or experien ce forced turnover. The difference in the fraction of firms that face disciplinary actions is significant at the one percent level. Even more interesting is that all of the discipline in the later period comes from boards. None of the acquirers in the bottom quartile become the target of a hostile offer. The results 17 The [-1,1] and [-5,5] announcement returns for top and bottom quartiles for the 1980-1987 sample are 0.4% and 2.7%, and -5.5% and -7.3%. The [-1,1] and [-5,5] announcement returns for top and bottom quartiles for the 1992-1998 sample are 1.8% and 2.4%, and -6.8 and -7.4%.

PAGE 51

42 indicate that “bad” acquisitions result in fo rced turnover more frequently only during the 1992-1998 period. 3.3.2 Logistic Regressions of Managerial Turnover The results in the previous section indi cate an inverse relationship between the wealth effects of acquisition and subsequent disciplinary events from the board of directors. To better understand the time vari ation in managerial tu rnover and its relation to acquisition performance, I estimate log it regressions relating the probability of managerial turnover and the abnormal return around the acquisition announcement, bidder size, CEO age, pre-acquisition performa nce, a dummy variable equaling one if the merger occurred during 1980-1987, as well as an interaction term measuring the difference in the sensitivity of turnover to the wealth effects of the acquisition between the two takeover regimes.18 The abnormal return of the acquisition a nnouncement is used as the measure of the wealth effects of the acquisition. The Buy-and-hold market-adjusted return (BHAR), calculated over the [-250,-10] window, is in cluded as a measure of the firm’s preacquisition performance. To control for firm size in the regression, I include the logarithm of the firm’s market value calcu lated 20 days before the first announcement scaled by the value-weighted CRSP index. A dummy variable (D80-87) is equal to one when the acquisition took place during the 1980 to 1987 period, and zero otherwise. An 18There are 19 firms that are successfully acquired befo re any CEO turnover occurs. Including these firms in the analysis is ambiguous for two reasons. First, since the firm exits the sample before its three-year post-acquisition period, it is unclear whether the intern al control systems would have acted if no takeover had taken place. Second, information about the managers and their positions within the new firm are often difficult to obtain. Moreover, it is unclear how to classify managers that are assigned a similar position with the new firm. For these reasons, I exclude these 19 firm s, and perform the regression analysis on a sample of 266 firms.

PAGE 52

43 interaction term (D80-87*AR) is included to allow for changing sensitivity of disciplinary events to acquisition performan ce that may be induced due to the general decline in hostile takeover activity in the 1990s relative to the 1980s. The results of the logit regression for CEO turnove r and forced turnover are reported in Table 3.6. The negative coefficien t estimates on the announcement returns indicate a strong inverse relationship between the wealth effects of the acquisition and managerial turnover. In other words, valu e-reducing behavior by management increases the probability of forced turnover. The co efficient estimates on the period dummy are negative and significant when forced tur nover is the dependent variable. Hence, controlling for age and the wealth effects of the acquisition, firms are more likely to experience forced CEO succession in the latter period. This is consistent with Huson et al. (2001), who find that the frequency of forced turnover has increased over time. The most interesting result is the coefficients on the interaction variable. The coefficients on the interaction dummy are insi gnificant for the [-1,1] window, but significantly positive for the [-5,5] window. Hence, there is no eviden ce that the sensitivity of forced turnover to acquisition wealth effects is lower as a result of the decline in hostile takeover activity. In fact, for the [-5,5] window the results are co nsistent with internal control mechanisms becoming more effective in quickly repl acing management who make poor acquisition decisions, despite the decline in hostile takeover activity. I also estimate the implied probabilities of forced turnover for firms at the 10th, 25th, 50th, 75th, and 90th percentiles of the announcemen t returns for the sample of acquirers. The implied probabilities are calculat ed using the logistic equations predicting the likelihood of forced turnover duri ng the three years af ter the acquisition

PAGE 53

44 announcement. The probabilities are estimate d holding the independent variables at the sample mean. The probability of forced tu rnover for the [-5,5] announcement return at the 10th percentile (-12.4%) is 20.3%, compared to the probability of forced turnover of only 7.8% when the [-5,5] announcement return is at the 90th percentile (8.9%). Thus, the probability of forced turnover almost tr iples and increases by over 12 percent as announcement returns fall from the 90th to the 10th percentile.19 To gain additional insights about the ne gative relation between turnover and announcement returns across the tw o sample periods, I estimate logit regressions for the 1980-1987 and 1992-1998 periods separately. The resu lts (not reported) indicate that the coefficient on the announcement return is ne gative and highly significant for the 19921998 period. In contrast, the coefficient is not significant duri ng the 1980-1987 period. Since the hostile takeover market also fails to discipline managers, it appears that managers are less likely to be penalized fo r making “bad” acquisiti ons during the earlier period. 3.4 A Closer Look at the Determinants of Forced Turnover So far, I have argued that the board of directors is the pr imary disciplinary mechanism in forcing out top executives th at make poor acquisition decisions. While the threat of a hostile takeover sharply diminish ed during the 1990s, other forms of external pressure may have played a more important ro le in influencing board decisions regarding managerial turnover. Large shareholders or bl ockholders, in particul ar, have incentives to exert considerable pressure to replace underperforming managers. 19 The implied probabilities for the 25th, 50th, and 75th,percentiles are 16.5%, 13.3%, and 10.6% respectively. The implied probabilities at the 10th, 25th, 50th, 75th, and 90th percentiles of the [-1,1] announcement return wi ndow are 20.2%, 16.6%, 13.0%, 11.3%, and 9.3% respectively.

PAGE 54

45 In order to investigate the role blockholde rs play in pressuring the board to force management out, I restrict the analysis to the 1992-1998 sample. Excluding 14 firms that are acquired before replacing the top executive, a nd one firms for which proxy material or 10K reports before or within two m onths after the acqui sition announcement is missing results in a sample of 146 firms. Ta ble 3.7 reports the means and medians of the ownership and board characteristics for thes e 146 firms. CEO characteristics show that the average top executive is 53 years old, owns 3.58 percent of the voting shares, and holds the additional position of chairman of the board 87 percent of the time. The table also reports boar d characteristics. On averag e, the board of directors consists of 10 board members, with 70 percent of the directors classified as outsiders. A director is considered an outsider when the individual is not currently, or has never been employed by the firm, has no business relationships with the firm, and has no family ties to management. A board is defined to be inde pendent or outsider-dominated when at least 60 percent of the board member s are outsiders. Using this de finition, almost 73 percent the firms have outsider-dominated boards. This number is substantially higher than in Weisbach (1988), who documents that 35 perc ent of the boards are independent for a sample of NYSE firms in 1980, and is consis tent with improved regulations that move towards more independent boards. The averag e shareholdings by offi cers and directors, including the CEO, are 10 percent. Unaffiliated blockholders are defined as shareholders that own at least five percent of the outstanding voting stock, and do not have any type of agreement with the firm. ESOPs and standstill agreements are cons idered affiliated block holdings, and are therefore not included. The major ity (76 percent) of the firms has at least one blockholder

PAGE 55

46 at the time of the acquisition. The median numb er of blockholders is two, and they own on average approximately 15% of the voting shares. Comparing the governance variables for the forced turnove r and non-forced turnover samples reveals few differences. Inte restingly, managers of firms that do not experience forced turnover are, on average, a bout three years older. The reason is that the control sample includes both firms where the managers retain thei r position, and firms that experience planned succession through re tirement. Otherwise, only mean block holdings are significantly higher at the ten pe rcent level for the firms in the control sample. Jensen (1993) warned that CEOs that are also the chairman of the board entrench themselves. The fraction of firms that have one individual who serves in a dual role as CEO and chairman is 0.79 for the forced tur nover sample and 0.89 for the control sample, but is not significant at conventional levels. To provide evidence on the role of blockhol ders in the decision to replace the top executive, I re-estimate the logit model from the previous secti on. If blockholders are active monitors of the firm, we would expect greater sensitivity of turnover to acquisition returns for firms with blockholders. To allow for this possibility, I include an interaction term (BLOCKHOLDER*AR), where BLOCKHOL DER is a dummy variable equal to one when a blockholder is present at the time of the acquisition. Previous research has found that board char acteristics also play an important role in the decision to replace management. Weisb ach (1988) reports that the sensitivity of turnover to performance incr eases with the independence of boards. To examine the impact of the composition of the board on the relationship between turnover and acquisition performance, I include a second interaction term (OUTSIDER*AR).

PAGE 56

47 OUTSIDER is a dummy variable equal to on e when at least 60 percent of the board members are outsiders. To control for poten tial entrenchment effects of management through shareholdings, the vari able CEOHOLD is included, wh ich is the proportion of stock owned by the top executive imme diately preceding the acquisition. The results are reported in Table 3.8. Ev en after controlling for the various ownership and board characteristics, the nega tive relationship between the wealth effect of acquisitions and subsequent forced turnove r is still highly significant. Ownership by management is only significant when CEO turnover is the dependent variable. The coefficient estimates on the interaction terms ar e not statistically signi ficant. Hence, there is no evidence that either unaffiliated block holdings or outsider-dominated boards affect the sensitivity of forced turnover to the wealth effects of the acquisition. One potential downside of the regression results in Table 3.8 is that the governance variables are measured before and around the time of the acquisition. However, CEOs may not be replaced until two or three years later. Denis and Serrano (1996) argue that blockholders often app ear and disappear be tween two successive proxies. Searching news accounts around the date of management turnover, they find that monitoring by active outside inve stors facilitates internal co ntrol efforts. To account for this possibility, I search Factiva during the year before and th ree months after the forced CEO succession. Out of 34 firms that force out th eir CEO, I am only able to identify five cases (15%) in which a large shareholder or activist group exer ts pressure that subsequently leads to replacement of the CEO. Hence, there is little direct evidence that active blockholders are the main initiators of CEO replacements.

PAGE 57

48 There are two interpretations for these fi ndings. First, much of the external pressure from institutional holders and large shareholders is communicated verbally and not reported. Consequently, it is difficult to measure the extent and effects of shareholder activity on board decisions rega rding the issue of replacing management. A second, and perhaps more satisfying interpretation, is th at boards have become more responsive to fulfilling their fiduciary obligations in acting on behalf of shareholders. Incentives in the form of equity-based compensation and the move towards smaller and more independent boards are potentially contributing factors in the increased sensitivity of turnover to performance. 3.5 Concluding Remarks In the current capital mark et, where valuations are high and effective antitakeover devices are deterring potential raiders from making control-reducing bids, there is an emphasized need to increase the role of the board of directors in monitoring management. This chapter investigates the m onitoring function of the board of directors and its effectiveness during pe riods of high and low external pressure from the hostile takeover market, by focusing on a sample of firms that make large acquisitions. Contrary to Mitchell and Le hn (1990), there is little evidence for the disciplinary role of hostile takeove rs. Firms that become targets of a hostile takeover attempt in the three years after their acquisi tion announcement do not appear to make acquisitions that are significantly worse than those of nontarget firms. In addition, only a small fraction of the firms (approximately 7%) receives a hostile offer within three years after making an acquisition. Moreover, hostile offers are rare ly successful and while they may induce

PAGE 58

49 firms to make value-improvements, they seldom result in forced resignations of the top executive. Therefore, given the small number of hostile attempts combined with the fact that hostile targets do equally well dur ing acquisitions seriously questions the “effectiveness” of the external control ma rket in disciplining management for making value-reducing acquisitions in a timely manner. Internal control mechanisms, however, are e ffective in disciplining what appear to be poor acquisition decisions. The results i ndicate a strong negative relationship between the abnormal returns around the acquisition announcement and subsequent forced turnover of the top executive. Contrary to the implicatio ns of Mikkelson and Partch (1997), the effectiveness of the internal control syst em headed by the board of directors is stronger during periods of less intense hos tile takeover activity. Improvements in corporate governance in terms of increased shareholder involvement, greater focus on stock price performance, and proper board incentives appear to provide viable alternatives to the corporate control market Collectively, the evidence in this chapter indicates that being disciplined for maki ng a poor acquisition is more a function of internal discipline than the workings of the takeover market.

PAGE 59

50 Table 3.1 Description of the sample The table reports the number of acquisitions, number of unique bidders, median market values of the bidder, target, deal, and relative size of target to bidder for each year. The market values are computed four weeks prior to the initial announcement. The sample consists of acquisitions that satisfy the following criteria: (1) For each transaction, bo th bidder and target trade on the NYSE, AMEX or NASDAQ, with data available on CRSP. (2) Minimum transaction value is $100 million for acquisitions in the 1980-1987 period and $300 million for acquisitions in the 1992-1998 period. (3) Transactions in which either party operates in a regulated industry are excluded – this excludes all banks, financials, railroads, and utilities. (4) The acquirer gains complete control of the target. (5) The ratio of market values of targ et and bidder is between 10 and 70 percent. Acquisitions with relative size greater than 70 percent are incl uded in cases where the bidder makes additional acquisitions in the three years be fore or after the acquisition that satisfies criteria 1 through 4. Year Number of acquisitions Number of firms Value Bidder ($ mil) Value Target ($ mil) Transaction Value ($ mil) MV Target/ MV Bidder 1980 5 5 1652 442 483 0.20 1981 21 20 1523 249 331 0.19 1982 11 8 1194 388 555 0.35 1983 18 17 992 333 386 0.25 1984 16 16 860 255 332 0.41 1985 24 21 1599 389 568 0.37 1986 26 23 957 316 364 0.24 1987 16 14 634 176 251 0.37 1992 5 5 1953 328 529 0.29 1993 9 9 2371 631 842 0.21 1994 18 17 1905 651 849 0.25 1995 23 23 2128 883 1459 0.31 1996 34 28 2801 509 1039 0.24 1997 49 43 1892 440 776 0.29 1998 44 36 3630 950 1623 0.30 1980-1987 137 124 999 270 361 0.28 1992-1998 182 161 2254 555 921 0.26

PAGE 60

51 Table 3.2 Frequency and type of disciplinary event The table reports the frequency of disciplinary control ev ents acquiring firms experience in the three years following the acquisition. Firms are classified as inde pendent (nontarget), hostile target, friendly target, experiencing CEO turnover, or forced turnover. The total number of disciplinary events (disciplined), defined as experiencing forced turnover or becoming a hostile target is also reported. A hostile target is a firm that is the target of successful or unsuccessful hostile tender offers, unsolicited and rejected offers, and proxy contests. A firm is a friendly target if the fi rm is the target of friendly merger or tender offer proposals, or successful and unsu ccessful management buyouts, and doe s not receive any hostile offers. A firm is classified as a nontarget if it does not receive any hostile or fri endly offers. A firm experiences CEO turnover whenever a succession of the top officer occurs, excluding th ose that are the result of health reasons and death. Forced turnover excludes top executive turnover labeled as normal succession using the classification scheme of Huson et al. (2002).The sample periods are 1980-1987 and 1992-1998. The pvalue of the Chi-squared statistic that tests the hypot hesis that the proportion of firms experiencing each event is equal in the two sub-periods is reported in the last column. Period Full Sample 1980-1987 1992-1998 Number of Firms Percent of Firms Number of Firms Percent of Firms Number of Firms Percent of Firms p-value Full Sample 285 100 124 100 161 100 Nontargets 241 85 106 85 135 84 All Targets 44 15 18 15 26 16 0.51 Friendly 25 9 4 3 21 13 0.01 Hostile 19 7 14 11 5 3 0.00 CEO Turnover 88 31 34 27 54 34 0.16 Forced Turnover 43 15 9 7 34 21 0.00 Disciplined 59 21 21 17 38 24 0.19

PAGE 61

52 Table 3.3 Abnormal stock market performance of acquiring firms The table reports the acquisition announcement return s for firms that remain independent (nontarget), become a hostile target, become a friendly target, experience no turnover (n onturnover), CEO turnover, forced turnover, or are disciplined defined as forced tu rnover or target of a hostile takeover. A hostile target is a firm that is the target of successful or unsuccessf ul hostile tender offers, unsolicited and rejected offers, and proxy contests. A firm is a friendly target if the firm is the target of friendly merger or tender offer proposals, or successful and unsu ccessful management buyouts, and doe s not receive any hostile offers. A firm is classified as a nontarget if it does not receive any hostile or fri endly offers. A firm experiences CEO turnover whenever a succession of the top officer occurs, excluding th ose that are the result of health reasons and death. Forced turnover excludes top executive turnover labeled as normal succession using the classification scheme of Huson et al. (2001).The sample periods are 1980-1987 and 1992-1998. The abnormal return is the market-adjusted return measured over the [-1,1] and [-5,5] windows. For firms with multiple acquisitions that meet the criteria discussed in table 3.1, the abnormal return is the arithmetic average of the announcement returns for all its acquisitions. The p-values of the t-test and signed rank test are reported in parentheses below the mean and median, respectively. Panel A: [-1,1] window Period Full Sample 1980-1987 1992-1998 Mean Median Mean Median Mean Median -0.023 -0.023 -0.022 -0.021 -0.024 -0.025 Full Sample (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) -0.020 -0.020 -0.020 -0.018 -0.020 -0.023 Nontargets (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) -0.039 -0.035 -0.031 -0.030 -0.044 -0.041 All Targets (0.00) (0.00) (0.00) (0.00) (0.03) (0.02) -0.027 -0.033 -0.031 -0.030 -0.016 -0.039 Hostile Target (0.01) (0.01) (0.00) (0.01) (0.46) (0.63) -0.047 -0.036 -0.033 -0.030 -0.050 -0.045 Friendly Target (0.02) (0.00) (0.00) (0.13) (0.02) (0.02) -0.019 -0.021 -0.021 -0.022 -0.018 -0.021 Nonturnover (0.00) (0.00) (0.00) (0.00) (0.01) (0.00) -0.031 -0.027 -0.024 -0.016 -0.036 -0.021 CEO Turnover (0.00) (0.00) (0.01) (0.00) (0.00) (0.00) -0.048 -0.044 -0.028 -0.028 -0.054 -0.056 Forced Turnover (0.00) (0.00) (0.06) (0.13) (0.00) (0.00) -0.043 -0.040 -0.032 -0.033 -0.049 -0.051 Disciplined (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

PAGE 62

53 Table 3.3 continued. Panel B: [-5,5] window Period Full Sample 1980-1987 1992-1998 Mean Median Mean Median Mean Median -0.023 -0.027 -0.021 -0.022 -0.026 -0.031 Full sample (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) -0.022 -0.024 -0.021 -0.018 -0.022 -0.031 Nontargets (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) -0.035 -0.038 -0.019 -0.045 -0.046 -0.033 All Targets (0.02) (0.01) (0.28) (0.14) (0.04) (0.04) -0.013 -0.038 -0.008 -0.030 -0.028 -0.038 Hostile Target (0.41) (0.20) (0.70) (0.43) (0.19) (0.13) -0.051 -0.038 -0.055 -0.059 -0.051 -0.029 Friendly Target (0.00) (0.02) (0.00) (0.13) (0.07) (0.08) -0.016 -0.014 -0.021 -0.017 -0.012 -0.014 Nonturnover (0.01) (0.00) (0.02) (0.00) (0.37) (0.63) -0.041 -0.041 -0.021 -0.030 -0.054 -0.051 CEO Turnover (0.00) (0.00) (0.09) (0.05) (0.00) (0.00) -0.067 -0.061 -0.022 -0.042 -0.079 -0.077 Forced Turnover (0.00) (0.00) (0.48) (0.43) (0.00) (0.00) -0.054 -0.054 -0.021 -0.041 -0.072 -0.066 Disciplined (0.00) (0.00) (0.20) (0.13) (0.00) (0.00)

PAGE 63

54 Table 3.4 Differences in abnormal returns of acquiring firms The table reports the differences in means and me dians of the acquisition announcement returns for acquiring firms that are subsequently classified as target, hostile target, friendly target, experiencing CEO turnover, forced turnover, and disciplined (forced turnover or hostile target). A hostile target is a firm that is the target of successful or unsu ccessful hostile tender o ffers, unsolicited and rejected offers, and proxy contests. A firm is a friendly target if the firm is th e target of friendly merger or tender offer proposals, or successful and unsuccessful manageme nt buyouts, and does not receive any hostile offers. A firm is classified as a nontarget if it does not receive any hostile or friendly offers. A firm experiences CEO turnover whenever a succession of the top officer occurs, excluding th ose that are the result of health reasons and death. Forced turnover excludes top executive turnover labeled as normal succession using the classification scheme of Huson et al. (2001).The sample periods are 1980-1987 and 1992-1998. The abnormal return is the market-adjusted return measured over the [-1,1] and [-5,5] windows. For firms with multiple acquisitions that meet the criteria discussed in table 3.1, the abnormal return is the arithmetic average of the announcement returns for all its acqui sitions. The p-values of the t-test and Wilcoxon zstatistic for the tests of differences in means an d medians, respectively, are in parentheses. Panel A: [-1,1] window Period Full Sample 1980-1987 1992-1998 Mean Median Mean Median Mean Median -0.018* -0.015 -0.011 -0.012 -0.023 -0.018 Targets vs. Nontargets (0.08) (0.15) (0.41) (0.34) (0.13) (0.26) -0.007 -0.013 -0.011 -0.012 0.004 -0.016 Hostile vs. Nontargets (0.64) (0.73) (0.48) (0.45) (0.90) (0.86) -0.027* -0.016 -0.013 -0.012 -0.030* -0.022 Friendly vs. Nontargets (0.05) (0.12) (0.63) (0.50) (0.08) (0.18) -0.012 -0.006 -0.004 0.006 -0.018 0.000 Turnover vs. Nonturnover (0.14) (0.32) (0.74) (0.90) (0.14) (0.21) -0.029*** -0.023*** -0.003 -0.006 -0.036** -0.035*** Forced turnover vs. Nonturnover (0.01) (0.01) (0.69) (0.92) (0.02) (0.01) -0.022** -0.023** -0.013 -0.016 -0.029** -0.034*** Disciplined vs. Nondisciplined (0.02) (0.01) (0.32) (0.36) (0.02) (0.01) ***, **, denote significance at the one, five, and ten percent level, respectively.

PAGE 64

55 Table 3.4 continued. Panel B: [-5,5] window Period Full Sample 1980-1987 1992-1998 Mean Median Mean Median Mean Median -0.013 -0.014 0.002 -0.027 -0.024 -0.002 Targets vs. Nontargets (0.35) (0.63) (0.91) (0.97) (0.21) (0.65) 0.008 -0.014 0.013 -0.012 -0.006 -0.007 Hostile vs. Nontargets (0.67) (0.73) (0.58) (0.69) (0.88) (0.96) -0.030 -0.014 -0.034 -0.031 -0.028 0.002 Friendly vs. Nontargets (0.11) (0.37) (0.42) (0.40) (0.19) (0.62) -0.025** -0.029** -0.001 -0.013 -0.042*** -0.037*** Turnover vs. Nonturnover (0.02) (0.02) (0.96) (0.57) (0.00) (0.00) -0.051*** -0.049*** -0.001 -0.025 -0.067*** -0.063*** Forced turnover vs. Nonturnover (0.00) (0.00) (0.97) (0.88) (0.00) (0.00) -0.028** -0.039** 0.001 -0.020 -0.046*** -0.053*** Disciplined vs. Nondisciplined (0.03) (0.01) (0.96) (0.96) (0.00) (0.00) ***, **, denote significance at the one, five, and ten percent level, respectively.

PAGE 65

56Table 3.5 Frequency and type of control-reducing event for good acquirers versus bad acquirers The table reports the proportion of acquiring firms that experi ence a disciplinary control even t in the three years following t he acquisition. The sample is partitioned in quartiles based on the [-1,1] and [-5,5] abnormal returns. Firms are cl assified as target, hostile target, frien dly target, CEO turnover, forced turnover, or discip lined (forced turnover or ho stile target). A firm is a target when it receives any takeover offer within thr ee years after the announcement. A hostile target is a firm that is the target of successful or uns uccessful hostile tender offers, unsolicited an d rejected offers, and proxy contests. A firm is a friendly target if the firm is the targ et of friendly merger or tender offer proposals, or successful an d unsuccessful management buyouts, and does not receive any hostile offers. A firm experi ences CEO turnover whenever a su ccession of the top officer occu rs, excluding those that are the result of health reasons and death. Forced turnover ex cludes top executive turnover labe led as normal succession using the classification scheme of Huson et al. (2001).The sample periods are 1980-1987 and 1992-1998. The p-values of the Chi-squared statistic that tests the hypothesis that the proportion of firms experiencing each event is equal between the two periods are also reported. Panel A: [-1,1] window Target Hostile Target Friendly Target CEO Turnover Forced Turnover Disciplined N Fraction p-value Fraction p-value Fraction p-value Fraction p-value Fraction p-value Fraction p-value Full sample Top quartile 72 0.11 0.46 0.04 0.65 0.07 0.26 0.26 0.28 0.10 0.03 0.13 0.04 Bottom quartile 72 0.15 0.03 0.13 0.35 0.24 0.26 1980-1987 Top quartile 31 0.10 0.45 0.10 0.69 0.00 0.31 0.26 1.00 0.10 0.64 0.16 0.74 Bottom quartile 31 0.16 0.13 0.03 0.26 0.06 0.19 1992-1998 Top quartile 41 0.15 0.39 0.02 0.31 0.12 0.24 0.27 0.34 0.12 0.06 0.15 0.11 Bottom quartile 41 0.22 0.00 0.22 0.37 0.29 0.29

PAGE 66

57Table 3.5 continued. Panel B: [-5,5] window Target Hostile Target Friendly Target CEO Turnover Forced Turnover Disciplined N Fraction p-value Fraction p-value Fraction p-value Fraction p-value Fraction p-value Fraction p-value Full sample Top quartile 72 0.15 0.63 0.06 0.40 0.08 0.77 0.22 0.05 0.06 0.00 0.10 0.00 Bottom quartile 72 0.13 0.03 0.10 0.38 0.28 0.29 1980-1987 Top quartile 31 0.10 0.30 0.10 0.30 0.00 1.00 0.26 0.35 0.06 1.00 0.13 0.69 Bottom quartile 31 0.03 0.03 0.00 0.16 0.06 0.10 1992-1998 Top quartile 41 0.17 0.78 0.02 1.00 0.15 0.76 0.20 0.00 0.05 0.00 0.07 0.00 Bottom quartile 41 0.20 0.02 0.17 0.54 0.44 0.44

PAGE 67

58 ***, **, denote significance at the one, five, and ten percent level, respectively. Table 3.6 Pooled Logit Regressions on CEO tu rnover and forced CEO turnover Logistic regressions estimating the relation between the probability of turnover and the abnormal return around the acquisition announcement, bidder size, CE O age, a pre-acquisition performance variable, a dummy variable equaling one if the merger occurred during 1980 to 1987, as well as an interaction term measuring the difference in the sens itivity of merger performance to turnover across the two periods A firm experiences CEO turnover whenever a succession of the top officer occurs excluding those that are the result of health reasons and deat h. Forced turnover excludes top executive turnover labeled as normal succession using the classification scheme of Huson et al. (2001). The abnormal return is the marketadjusted return measured over the [-1,1] and [-5,5] windows. For firms with multiple acquisitions that meet the criteria discussed in table 3.1, the abnormal return is the arithmetic average of the announ cement returns for all its acquisitions. Bidder size is measured by the lo garithm of the market value of the bidder 20 days prior to the first announcement scal ed by the value-weighted CRSP re turn index (1980=1). The CEO age dummy is equal to one if the CEO is 60 years or older at the time of the announcement, and zero otherwise. The Buy-and-Hold abnormal return is the market-adjusted return calcu lated over the [-250,-10] window. The sample consists of 266 firms that remained independent before CEO turnover or for the three years after the acquisition. CEO turnover occurred in 88 firms, with 43 firms experiencing forced turnover. pvalues are in parentheses. Firm experiences turnover (N=266) Firm experiences forced turnover (N=266) [-1,1] [-5,5] [-1,1] [-5,5] Intercept -0.658 -0.981 -0.486 -0.984 (0.67) (0.53) (0.80) (0.62) Abnormal Return -6.581** -6.778*** -9.050*** -9.008*** (0.03) (0.00) (0.01) (0.00) Bidder Size -0.030 -0.009 -0.081 -0.050 (0.81) (0.94) (0.59) (0.75) CEO Age>60 1.393 1.401 (0.00) (0.00) D80-87 -0.501 -0.430 -1.017** -0.877** (0.12) (0.17) (0.03) (0.05) D80-87*AR 4.956 6.201* 6.078 8.561* (0.34) (0.08) (0.43) (0.09) BHAR -0.141 -0.088 -0.025 0.049 (0.61) (0.75) (0.93) (0.87) Pseudo R2 0.082 0.067 0.092 0.117

PAGE 68

59 Table 3.7 Ownership and board characteristics of the 1992-1998 sample The table reports the board and ownership characteristics for 146 acquirers of the 1992-1998 sample. Firms that are acquired during the three-year post-acquisition period or have missing proxy statements are excluded. CEO characteristics are CEO age at the time of the acquisition, proportion of shares held by the CEO, and whether the same individual holds the CEO and chairman position. Board characteristics are the size of the board, proportion of board comprised of outsiders, and proportion of shares held by officers and directors. Directors are classified as outsiders when the individual is not currently, or has never been employed by the firm, has no business relationships with the firm, and has no family ties to management. Unaff iliated blockholders are defined as shareholders that own five percent of the stock, and do not have any type of agreement with the firm. The proportion of firms with at least one blockholder, the average number of blockholders, and blockholder holdings are reported. The sample is partitioned into firms that experience forced turnover and of those that do not. Forced turnover is defined as any succession of the top officer, excluding thos e that are the result of he alth reasons and death, or are labeled as normal succession using the classification scheme of Huson et al. (2001). Full Sample (N=146 ) No Forced Turnover (N=112) Forced Turnover (N=34) Mean Median Mean Median Mean Median CEO characteristics CEO age at time of the acquisition 53 53 54 54 50** 52** Proportion of shares held by CEO 0.04 0.01 0.04 0.01 0.03 0.01 Proportion firms CEO is Chairman 0.87 1 0.89 1 0.79 1 Board Characteristics Number of directors 10.34 10 10.43 10 10.03 10 Proportion of outside directors 0.70 0.73 0.70 0.75 0.69 0.71 Proportion of shares O&D 0.10 0.04 0.11 0.04 0.08 0.05 Unaffiliated Blockholdings Proportion at least one blockholder 0.76 1 0.78 1 0.71 1 Number of unaffiliated blockholders 1.73 2 1.76 2 1.62 2 Proportion of shares blockholders 0.15 0.14 0.16 0.15 0.11* 0.12 ***, **, denote significance at the one, five, and ten percent level, respectively.

PAGE 69

60 Table 3.8 Pooled Logit Regressions on CEO turnover and forced CEO turnover for the 1992-1998 sample Logistic regressions estimating the relation between the pr obability of turnover and the abnormal return around the acquisition announcement, bidder size, CEO age, pre-acquisition performance variable, CEO ownership, and two interaction terms. The first interaction term measures th e difference in the sensitivity of acquisition performance to turnover for outsider-dominated boards. The second me asures the difference in the sensitivity of acquisition performance to turnover for firms with blockholders. A firm experiences CEO turnover whenever a succession of the top officer occurs, excluding those that are the result of health reasons and death. Forced turnover excludes top executive turnover labeled as normal succession using the classification scheme of Huson et al. (2001) The abnormal return is the market-adjusted return measured over the [-1,1] and [-5,5] windows. For firms with multiple acquisitions that meet the criteria discussed in table 3.1, the abnormal return is the arithmetic average of the announcement returns for all its acquisitions. Bidder size is measured by the logarithm of the market value of the bidder 20 days prior to the first an nouncement scaled by the value-weighted CRSP return index (1980=1). The CEO age dummy is equal to one if the CEO is 60 years or olde r at the time of the announcem ent, and zero otherwise. The Buy-and-Hold abnormal return is the market-adjusted return calculated over the [-250,-10] window. The sample consists of 146 firms that are not acquired before CEO turnover or the three years after the acquisition announcement and for wh ich proxy information on governance variables are available. Firm experiences turnover (N=146) Firm experiences forced turnover (N=146) [-1,1] [-5,5] [-1,1] [-5,5] Intercept 0.442 -0.195 -0.592 -1.696 (0.83) (0.93) (0.81) (0.49) Abnormal Return -13.7 -12.1** -21.4** -18.7*** (0.11) (0.05) (0.04) (0.00) Bidder Size -0.092 -0.047 -0.065 0.015 (0.58) (0.78) (0.73) (0.94) CEO Age>60 1.096** 1.105*** (0.01) (0.01) CEOHOLD -0.081* -0.086* -0.040 -0.043 (0.06) (0.06) (0.38) (0.36) Outsider*AR 4.331 4.192 10.538 8.692 (0.48) (0.44) (0.12) (0.15) Blockholder*AR 5.112 3.022 5.303 3.903 (0.50) (0.53) (0.55) (0.45) BHAR -0.177 -0.165 -0.232 -0.229 (0.59) (0.64) (0.50) (0.56) Pseudo R2 0.078 0.101 0.081 0.113 ***, **, denote significance at the one, five, and ten percent level, respectively.

PAGE 70

61 -0.1 -0.08 -0.06 -0.04 -0.02 0 0.02 -50510152025303540Cumulative Abnormal Retur n Days relative to announcement Friendly Target Hostile Target Nontarget Figure 3.1 Stock price reaction to acquisition announcements for firms that subsequently become a hostile or friendly target

PAGE 71

62 -0.1 -0.08 -0.06 -0.04 -0.02 0 0.02 -50510152025303540Cumulative Abnormal Retur n Days relative to announcement Forced Turnover CEO Turnover Non-Turnover Figure 3.2 Stock price reaction to acquisition announcements for firms that subsequently experience CEO turnover

PAGE 72

63 CHAPTER 4 IMPLICATIONS OF DEFEATING TAKEOVER BIDS: EVIDENCE PRE AND POST TIME-WARNER A popular viewpoint is that the problem of entrenched management has been exacerbated by new takeover defenses that have become available and legally permissible in the past twenty years. Many commentat ors point to the Dela ware Supreme Court’s decision in Paramount Communi cations, Inc v Time, Inc (“Ti me-Warner”) in July 1989 that upheld the use of the so called poison pill takeover defense even in the case of a fully financed cash takeover offer for all shares of a corporation’s stock.1 The Time-Warner decision helped give birth to what the popular press termed the “just say no defense” in which management can simply refuse to bargain with a potentia l buyer on the grounds that the offer is inadequate. At the same time, the poison pill defense makes it nearly impossible for the bidder to acquire the shares needed to replace management. This chapter addresses the issue of whethe r the advent of the “just say no takeover defense” has had a significant impact on managerial entrenchment by examining a specific example of alleged entr enchment that has long draw n the attention of academics and the media: the defeat of an unsolicited takeover bid. Unsolicited takeover bids are generally at prices that exceed current mark et prices and often entail replacement of management. Given that unsolicited bidders are unsuccessful in acquiring control of the target firm, the question becomes, how much, if at all, do shareholders suffer as a result 1 See, for instance, Hilder (1990).

PAGE 73

64 of defeating a bid. As the name of the “j ust say no defense” implies, managers in the post Time Warner era are argua bly better equipped to simp ly go about their business and make fewer efforts to move toward valu e maximization when fending off unsolicited bids. Hence, the entrenchment hypothesis predicts that post Time Warner we should observe fewer takeover defenses that are often associated with moving toward value maximization such as debt financed recapit alizations. Post Time-Warner, we should observe smaller positive revaluations of targ ets resulting from failed bids and the posttakeover bid share prices of these targets shoul d trade at larger disc ounts relative to the takeover price offered by the failed bidders. The results show that the “just say no defense” has had a dramatic impact in terms of how managers defend against unsolic ited takeover bids. I document significant declines in the use of share buybacks, special dividends, gree nmail, placement of shares with friendly third parties, and standstill ag reements during the post Time-Warner period. Managers were also far less likely to incr ease leverage in the 1990s as opposed to the 1980s when defending against takeovers. Instead, the poison pill (or its threatened use) is the primary defense in the 1990s. The evidence is consistent with managers viewing the “just say no” takeover defense as a “lower cost” defense option. Despite the ability to “just say no,” the analysis shows that the average valuation implications of successfully defending take over offers pre and post Time-Warner are very similar. I calculate the valuation gap between what is offered in a takeover bid and the post offer share price, and find that it is virtually identical between the two time periods. The average market-adjusted targ et revaluation, calcula ted over the takeover contest, is also nearly identical between th e two periods. One po ssible explanation of

PAGE 74

65 this finding is that the nature of the bi ds differ between the two time periods. For instance, two-tier bids and bi ds by non-public en tities are more common in the pre-Time Warner period. Even after controlling for these (and other) factors, however, the valuation consequences of a failed bid stil l are quite similar between the two time periods. With respect to the nature of the defense, simila r to Safieddinie and Titman, I find that firms that increase leverage in th e pre Time Warner period increase their revaluations relative to those that do not, but no such pattern appear s post Time Warner. Also, firms that pay greenmail as a defensiv e tactic appear to have more negative revaluations associated with the failed takeover episode. Finally, I followed the sample of firms in the period after defeating a hostile offer. I find that Post Time-Warner, targets are no le ss likely to be acquire d and CEOs are more likely to lose their executive positions after defeating a bid. Collectively, these results suggest that managers that successfully defeat takeovers in the 1990s are no more entrenched than those in the 1980s. This is consistent with othe r governance devices placing limitations on the deviation from w ealth maximization caused by “just say no” takeover defenses for firms that defeat a hostile bid. In interpreting the results, we should bear in mind that we are focusing on a set of firms where the management team (or at leas t the existing board of directors) remains intact for some period of time after defeati ng a takeover bid. The significance of this group is to determine just how far a management team might be able to stray from value maximization, while maintaining independenc e in the face of an unsolicited bid in alternative legal environments Obviously, the number of ex ternal challenges and the frequency with which they succeed also need to be factored in when assessing the degree

PAGE 75

66 to which the average manager may be able to deviate from value maximization in different legal environments. This is but one piece of the entrenchment puzzle. This chapter is organized as follows. Th e next section reviews the literature on managerial entrenchment, hostile takeovers, and contains the hypotheses development. Section 4.2 describes the data Section 4.3 documents the time variation of the takeover defenses. Section 4.4 focuses on intertemporal variation of managerial entrenchment. The post-contest consequences of defeating a bid on to the firm and management are discussed in Section 4.5. Concludi ng remarks follow in section 4.6. 4.1 Previous Literature and Hypotheses Development 4.1.1 Managerial Entrenchment, Takeover Bids, and Agency Theory One of the most cited examples of ag ency conflicts between management and shareholders is management de feating a takeover with a higher stated value than the price of the shares at the time the takeover is de feated. While management teams certainly have the potential motivation to fight for th eir managerial independe nce at the potential cost of forgoing a takeover premium, such actions are rather curious in an agency theoretic context. Generally, agency probl ems are viewed as obs ervability problems and defeating a takeover is a fa irly public spectacle. There are various reasons why manageme nt may choose to defeat a takeover offer. Management can make claims that their shares are underv alued based on their access to confidential information. In additi on, targets may deny a bidder due diligence

PAGE 76

67 lest competitors learn about competitive adva ntages that can be exploited by rivals.2 Often, targets argue that the compensation be ing proffered by a bidder is not worth what it is claimed to be worth. This might be partic ularly true of offers that include contingent securities that have no market or that may have lower valuat ions if we had more complete knowledge of the bidder’s business or the potenti al synergies from a merger. Taking this argument a step further, management often ma kes claims that the bidding party is not serious about actually ac quiring the target firm. A bidder may simply be trying to profit on an existing investment position in a firm by mounting a takeover bid and imposing costs on management so as to extract so ca lled greenmail payments, or by creating a buzz about potential takeovers that temporarily infl ates the price of the firm’s securities and allows the “bidder” to sell their position at a more attractive price. Nevertheless, to the extent that takeover bids are “serious,” th e decision to resist a bid is quite public. Successfully defeating an offer, with a resulting share price often well below the perceived value of the bid, might i nvoke disciplinary mechanisms to come into play. Competing bidders may step in, or the original bidder might return with enhanced credibility that management was in the wrong in defeating their prior bid. In many cases, management is pressured to execute some, if no t all, of the plans that a potential bidder had for the firm such as asset divestitures to refocus the firm or increasing financial leverage to capture tax benefits. However, management may be able to deviate from value maximization to a larger extent, the greater the entrenching devices that are 2 Some bidders also request compensation for expenses as a precondition into entering into some preliminary agreement that includes due diligence rights. This too could prove costly to target shareholders.

PAGE 77

68 available to them. The sophistic ation and availability of such devices, in particular the poison pill defense, has evolved with the ta keover market in the past two decades. 4.1.2 Hostile Takeovers and Target Defenses The 1980s experienced an unprecedented wave of so called hostile takeover activity in which companies received and initially rejected unsolicited takeover proposals. Andrade, Mitchell, and Stafford (2002) find that 14 percent of CRSP firms receive a hostile offer between 1980 and 1989. Larger firms were even more likely to receive a hostile offer. Mitc hell and Mulherin (1996) find th at 23% of Value Line firms are the target of hostile takeover activity during the 1980s. The wave of hostile takeover activity reflected a number of fact ors, such as better access to capital markets for investor groups and smaller corporations in the “junk bond” market, undoing th e diversification of corporations incurred during previous merger waves, and changes in antitrust policy. Firms employed numerous defensive tactics such as greenmail, the placement of securities with friendly third parties, sell ing off so called crown jewel assets, increasing leverage via share repurchases and instituting asset restru cturings. While many of the defenses were viewed as “entrenching devi ces,” it has been argued, that many of the defenses were simply enacting certain elemen ts of what outside acquirers would have done themselves if they had acquired the company. In other words, many defenses were merely actions to increase share price by moving toward value maximization for shareholders. For instance, Safieddine and Titman (1999) find for a sample of defeated takeover bids from 1981-1991, that firms that increase their leverage are much more

PAGE 78

69 likely to have positive revalu ations. Leverage increases accompanying defensive share buybacks and special dividends are common ta keover defenses utilized in the 1980s. The most dramatic development in combating hostile takeover defenses was the so called poison pill takeover defense (see Ryngaert (1988)). The defense first became prominent in 1985. It was effective in delayi ng acquirers and rais ing the legal costs associated with takeovers. Neve rtheless, the pill de fense did little to mitigate the wave of takeover activity in part because the courts placed limits on the use of the defense. For instance, it was generally taken as conventiona l wisdom that for Delaware incorporated firms, the pill could be used to combat two-tier takeovers (see Unocal Corp v. Mesa Petroleum, 1988), but that the pill could not be deployed indefinitely to defeat any or all takeovers (see Grand Metroplitan, PLC v Pillsbury, 1988). This changed with the Delaware Supreme Court’s decision in Pa ramount Communications, Inc v Time, Inc in July 1989. The decision suggested that managers could reject a bid if they felt the offer was too low and suggested that management was under no obligation to remove a poison pill takeover defense even in the face of an all-cash, fully financed tender offer for their firm. The decision basically gave managers th e right to “just say no” to takeover bids. The only recourse to a potential acquirer facing such a defe nse was to replace the board of directors in a proxy fight, a much more pr otracted and difficult undertaking, especially when board terms were staggered. Subsequent popular press articles suggest ed that this defense would severely damage shareholders by crippling the hos tile takeover mechanism for disciplining management. Consistent with the chilling effect of the Time-Warner decision, the incidence of hostile takeover activity pl unged in the early 1990s and never fully

PAGE 79

70 recovered. The Time-Warner decision is not th e exclusive explanation for this decline, however. Jensen (1993), for example, points to the closing down of the junk bond market to would be hostile acquirers and the cumulative impact of state takeover laws as contributing factors. 4.1.3 Hypotheses regarding defeated takeover bids in 1980s and 1990s If the ability to “just say no” has allowed firms to stray further from value maximization, a natural corollary would be that firms are under far less pressure to value maximize when they do receive a hostile takeover offer post Time-Warner. This also implies that, all else equal, firms should have to take fewer alternative defensive actions and would be less likely to lose their executive positions due to some subsequent actions taken against them. More importantly, to the extent that the “just say no” legal environment leads to entrenchment, the share prices of firms that defeat takeover bids should have less positive revaluation in the wake of an unsuccessful takeover bid and should also trade at a larger discount relative to the price of fered in the unsolicited bid. The alternative hypothesis is that substitute mechanisms have arisen to force management teams to “come close” to the value offered by an external bidder or risk dismissal. This would suggest that we might see few diffe rences in the outcomes associated with defeated hostile takeover bids in the 1980s and 1990s. The above arguments, of course, assume that the set of takeover targets and bidders is similar between the 1980s and 1990s. For instance, differences in post takeover defeat executive turnover rates and/or share price revaluat ion resulting from unsuccessful takeover bids may be a function of how poorly target firms performed in the

PAGE 80

71 pre and post Time-Warner samples. Similarly, there may be differences in the nature of the bidding parties, the nature of the bi ds themselves and the corporate governance structure of the firms that receive the bids in the two sample periods. I control for these differences in the analysis. 4.2 Data Description and Methodology 4.2.1 Sample of Defeated Takeover Bids The goal is to construct a sample of firms where management opposed a takeover bid and the firm remained independent for a period of time. The definition of “remaining independent” in the face of a hostile bid must meet the following th ree criteria: (1) the initial bid is unsolicited, (2) the target fi rm management rejects the bid, management pursues litigation against the bi dder or the bidder cites manage rial resistance as a reason for dropping its bid, and (3) the bidder subsequently “drops th e bid” with the exception of cases where the target had agreed to be acquired, but the bidder subsequently backed out of the agreement. While the criteria sound simple, the ch allenge is determining when a bid is dropped by the bidder and when the firm has “s uccessfully” fended off an unsolicited bid. Frequently acquirers drop a bid, only to come back shortly thereafter with another offer or some other means to achieve control such as a proxy contest or acquisitions of shares in the open market. In other cases, bidders dr op a bid because they realize there are other parties bidding “behind the scenes.” In fact this behind the scenes activity may be the very reason target management is resisting an existing bid in the first place. Finally, in many cases, the bidder simply “goes away” without comment when its bid is rejected,

PAGE 81

72 and the bid is never publicly dropped at all. In all these different types of contests, I seek to not only classify whether the firm has “s uccessfully” remained independent but to also identify a date of resolution for purposes of measuring wealth impacts of the takeover contest on the target. Using the Factiva database of news st ories and taking into account the above considerations, one of the following criteria must be met for a bid to be considered defeated, and a resolution date for each type of contest is determined as follows. (1) The bidder withdraws the bid or sells shares in the target and ceases taking actions to influence the firm’s policies or purchase the firm’s assets for three months, and there are no other takeover bids for the firm over the ne xt three months after the bid withdrawal. In this case the bid withdrawal date is c onsidered the resolution date. (2) The bidder drops the initial bid but clearly st ates that it will actively seek to control the firm or takes subsequent actions to control the firm. In these cases, the bid is deemed dropped when four consecutive months go by without the bidd er taking any actions to influence firm policy or acquire firm assets and with no ot her takeover bids being made. The resolution date is 40 days after the rejection or the last bidder action. (3) In the event that a bid is rejected and the bidder never formally drop s its bid, then the bid is deemed defeated when the bidder ceases to take actions to influence control of the target for four consecutive months and with no other takeover bids being made. Th e resolution date is 40 days after the rejection or the last bidder acti on. (4) Management de clares that it is putting the firm (or pieces of it) up for sale as a means to gain peace with a hostile bidder or as a means of winning a proxy contest. If no takeover related announcements occur for four months, then the resolution date is 40 days af ter the date of the for sale

PAGE 82

73 announcement. If there are stories updating the progress of this “sale process,” the resolution date is whenever target management declares that the search process is over or has turned up no buyers. The primary data source is the Thomson Financial SDC database listing of hostile and/or unsolicited takeover bids. Additiona l observations were hand collected from annual issues of W. T. Gri mm Mergerstat Review. Using the SDC database, there are 1477 instances between the years 1980-1999 wher e a firm received a takeover bid that was deemed “unsolicited” or “hostile” by SDC. From this list I identify the resolution dates of these contests and split the firms into pre and post Time-Warner groupings. There are 676 instances pre Time Warner an d 801 instances post Time Warner. First, I eliminate all instances that are related to banks, REITs, limited partnerships, and closed end mutual funds. These firms have very di fferent capital struct ures, tax issues, and regulatory issues relative to firms in othe r industries and they al so appear in large numbers. I also eliminate offers for privat e firms and divisions of firms since there are no stock quotes for these assets. I further elim inate offers for firms in bankruptcy or for takeover bids conditional on debt workouts. Negotiations in these cases must take place with a broader set of parties than bidder a nd target management teams. Offers by unions and government entities are also eliminated du e to their unusual na ture and the potential that these offers are parts of broader bargaini ng strategies. Offers that are not for control are also excluded. These include so called “clean up” offers where the bidder already controls 40% of the stock of the firm and offe rs to invest in the company or acquire only a foothold in the company less than 40%. Three false offers, where the press release announcing the bid turned out to be fraudul ent are also excluded. Offers by the

PAGE 83

74 management team of the firm (executives and/ or chairman of the board) are excluded. So called Pac Man offers made by firms that ar e the subject of a hos tile bid are excluded since these are more defensive maneuvers than unsolicited bids. Other eliminations include offers that were really “friendly” agreements between the target and the bidder, offers where no terms were disclosed, offers for less than $1.50 per share, and offers where the rejection of the bid became public knowledge well after the event itself. After these various screen s displayed in Table 4.1, th e remaining number of firms is 456 pre Time-Warner and 337 post Time-Warne r. After screening out firms where the target had multiple entries for the same take over contest, the number of pre Time-Warner cases drops to 389 and the number of post Ti me Warner Cases drops to 294. Next, I exclude cases where the firm receiving the unsolicited bid got acquired, announced it would liquidate or had the bidd ing entity gain working contro l of the board of directors during the takeover contest. There are 250 such “change of control” events in the pre Time-Warner period and 141 in the post Time Warner period. Given that the pre TimeWarner period encompasses a slightly shorter pe riod of time, this is consistent with prior empirical work that demonstrates greater le vels of hostile takeover activity in the 1980s versus the 1990s. After these eliminations, there are 139 and 153 firms respectively that are classified as remaining independent in the pre and post Time-Warner periods. I eliminate cases where management agreed to be acquire d, but the bidder backed out latter on its own accord, since target management cannot be blamed for such occurrences. Similarly, I also eliminate cases where the bidder back ed out without any blaming of management and without a rejection of the bid by the target firm. The fi nal concern was to eliminate

PAGE 84

75 offers where the true value of the offer was difficult to decipher. I eliminate one offer where different prices were offered to different classes of stock. I also eliminate offers where securities of questionable value were to make up a significant portion of the offer price. These offers include cases where more than 45% of the compensation offered by a private firm was in securities or where a pr ivate firm was offering more than 20% of the offer’s alleged value in common stock of a new firm. In the case of public firms, I keep any offer where at least half of the compensa tion offered is in cash. Firms are eliminated, however, if less than half the compensation is cash and the offer to the target exceeded the market value of the bidding firm. The va lue of securities in su ch cases will largely depend on how successful the merger would be and in the case of preferred securities there might be considerable disagr eement over their true value. Finally, any predominantly stock offer wher e the number of shares to be paid to the target is fixed and the target firm shareh olders would own more than one-third of the firm is excluded. The value of such near “merger of equals” offers can only be determined as the product of th e bidders share price times the number of shares offered to the target shareholders. Unfo rtunately, in such large mergers the value of the bidder’s securities would be highly re lated to the perceived desirability of any takeover. Hence, it is difficult to value the offer after it is defeated. In cases, where there are price guarantees that allow the number of shares to be issued to float with the bidder’s share price, this is not an issue. After these various screens, we are le ft with 122 pre Time-Warner observations and 109 post Time-Warner observations from the original SDC sample. Clearly, the number of defeated offers relative to changes of control is lower in the pre TimeWarner

PAGE 85

76 era. This might be interpreted to mean that managers had greater ability to stay independent in the event of a hostile takeover bid in the post time Warner era. The purpose of this study is not to dispute this interpretation. Rather, the focus is on what happens in the event of defea ting a hostile bid. To obtain th e largest possible sample of such defeats, an additional 22 observations in the pre-Time Warner period and an additional 17 in the post Time Warner obs ervations are added from the Mergerstat Review database of defeated bids from 1981-1999. These defeats meet all the same criteria for inclusi on as the SDC sample. The final sample of defeated offers consists of 144 pre Time-Warner and 126 post Time-Warner observations. The summary statisti cs for the sample of defeated offers are presented in Table 4.2. The Table shows the num ber of defeated offers, competing offers, multiple offers by the same bidder, and the median duration of the takeover contest per year. I am able to find 195 clear resolution da tes. When a precise w ithdrawal date could not be identified, the resolution date was 40 da ys after the last rejection date or bidder activity date as previously described in th is chapter. For purposes of calculating the median duration of the contest, I added 10 days to the last of fer/rejection date for events without clear withdrawal dates. The number of contests with multiple offers or competing bidders is similar between the two periods. 4.2.2 Target Characteristics It is often argued that hostile acquirers target underperforming firms. Early research finds results consistent with this idea. Morck et al. (1989) report that hostile takeover targets have lower Q ratios than friendly targets. Mitchell and Lehn (1990)

PAGE 86

77 show that firms that make value-reduci ng acquisitions are likely to become hostile targets. Both studies argue that hostile take overs serve as a disciplinary mechanism to reverse prior investment mistakes. More recent studies have found no underperformance of targets prior to the takeover offer. Agrawal and Jaffe (2001) use a large sample of target firms and examine stock and accounting performance of friendl y and hostile targets over a period of ten years prior to the offer. Contrary to popular belief, they find no significant evidence that either friendly or hostile targets underper form. Similarly, Schwert (2002) shows that hostile targets cannot be disti nguished from friendly targets in terms of accounting and/or stock performance prior to a takeover bid. Table 4.3 reports the firm characteristics for the targets in the sample. Panel A reports the stock and accounting performance In terms of differences in the two sample periods, the post Time-Warner targets tend to have poorer pre-announcement stock market performance and higher market to book ratios. EBIT and FCF to asset ratios are similar across periods. Hence, the evidence on ex-ante performance is not clear. The market capitalization of targets is la rger during the pre Time-Warner sample. Panel B reports the industry breakdown for the targets based on the SIC codes reported by Compustat. The majority of the ta rgets have industry classifications of 30003999. There are no clear differences in industr ies between the two periods. Proportionally there are more firms in the 30003999 SIC code range in the 1990s.

PAGE 87

78 4.2.3 Bidder Characteristics Previous studies have largely ignored bidder characteristics in the takeover process. The majority of deals are conditi onal on financing, due diligence, and other conditions such as the removal of the pois on pill defense. In some cases, however, bidder’s demands are highly conditional and ar guably unreasonable, which leaves target management no options but to reject the offer. In my sample design, I try to exclude the most preposterous takeover attempts. Neverthe less, it is plausible that for some of the offers in the sample, the bidder’s reputation and prior dealings in the takeover market are a major factor in target rejections. Acquirers may be motivated to make an offer to exert disciplinary pressure on management to implem ent value-enhancing strategies or to put the firm up for sale and sell its equity stake for profit later. Hostile acquirers are often accused of trying to profit at the expense of other shareholders. But even when acquirers own little or no equity stake in the target, it is not always clear that bidders are able to secure financing. It is plausibl e that targets reject some of fers conditional on financing for good reasons. Table 4.4 reports bidder characterist ics for the preand post Time-Warner samples. It is clear that bidde r characteristics have changed over time. Post Time-Warner, a larger fraction of the bidders are publicly traded firms. In addition, it appears that bidders have become more “serious” over time. Financial buyers, such as private investment firms and firms that have been paid greenmail in the five years prior to the offer, are more common in the 1980s. One r eason for fewer financial buyers during the latter period is the collapse of the junk bond market in 1989 which has made financing

PAGE 88

79 deals more difficult. This might suggest that the pre Time-Warner period has more dubious bids, and I control for this in s ubsequent analysis in a couple of ways. First, I classify eight bidders as ha ving a bad reputation (BADREP) when the principal of the bidding firm management is under criminal indictment, has run a formerly bankrupt firm as noted in contem poraneous press accounts, or has previously backed out of a merger agr eement in the past two years due to financing problems. Second, I create a variable called SYNERGY, which is a dummy equal to one if the bidder has some potential vertical or horizonta l strategy for use of the target’s assets in conjunction with the acquirer’s own field of expertise. In cases where the bidder is publicly traded I counted the offer as synerg istic except in obvious cases where business lines were clearly unrelated or in cases where the bidding firm was a shell company acquiring other firms to allow it to utilized tax loss carry forwards. Acquisition bids by private firms were generally deemed to not be synergistic unless ne ws accounts discussed how the bidding entity saw some synergy be tween existing operating divisions of the private firm and the target firm or a busine ss description of the bidding firm suggested synergies with the public target. It appears that there are greater syne rgistic motives in the 1990s as two thirds of the deals are defined as having a synergy versus fewer than half in the earlier period. The average premium, calculated as the ra tio of offer price to target price 20 days before the initial offer was made public, is 34. 6 percent. Using the date that the firm is first put in play, the premium increases to over 40 percent. The premiums are relatively constant over time and comparable to premiums for successful hostile and friendly public

PAGE 89

80 targets.3 Since the focus of this chapter is on ho stile offers, it is not surprising that the vast majority of the deals are all-cash offers. Grossman and Hart (1980) argued that ta rget shareholders will not tender their shares unless the price reflects all value impr ovements and synergies that will be created by the merger. To circumvent this free ride r problem, bidders can engage in so called two-tier offers where the first offer for cont rol is generally priced different than the backend. Two-tier offers are much more comm on in the earlier period. In fact, in the sample only four two-tier offers occurred af ter 1990. This is consiste nt with the courts justifying more stringent ta keover defenses (beginning in the middle of the 1980s) if the bid was two-tiered. Finally, almost half of th e bidders hold at leas t five percent of the target before the initial offer, and the av erage foothold is approx imately 6.5 percent. Consistent with the idea that the majority of financial bidders ar e more likely to make offers for reasons other than control, foot holds are more common and larger during the pre Time-Warner period. Post foothold refers to the equity stak e held by the bidder following the contest, and is si milar during the two periods. 4.3 Takeover Defenses and Leverage 4.3.1 Takeover Defenses Changes in the takeover market during the past two decades have also changed the way firms are able to defend themselves from unwanted acquirers. While targets have an arsenal of takeover defenses available, in this section I focus on share repurchases, defensive dividends, greenmail, standstill ag reements, majority blocking positions, the 3 Andrade et al. (2002) report a medi an premium of 37.7 percent and 34.5 percent for public targets during the 1980-89 and 1990-1998 periods, respectively.

PAGE 90

81 sale of shares to White Squi res or ESOPs, staggered boards and poison pills. The next section describes the effect of buybacks, greenmail, and dividends, on the capital structure of the firm. BUYBACK is defined as the repurchase of 10 percent or more of stock either through a tender offer or open market purcha ses. DIVIDEND is defined as a special dividend payout in excess of 10 percent of fi rm value and includes dividends initiated by spin-offs. GREENMAIL is defined as either a repurchase of shares from the hostile acquirer at a premium or a payment of cash or in kind to secure a standstill agreement. STANDSTILL refers to cases where the bidder is subject to a standstill agreement with the target. RECAPITALIZATION is defined as any comb ination of share buyback, greenmail, and dividend payout that exceeds 10 percent of equity value. In order to be included as a defensive measure, recapitalizat ions must be announced between one month before the first offer and one month after th e withdrawal date and executed within one year. Information on recapitalizations was collected from searching the Factiva database. A firm has a BLOCK when the combined stake of directors, affiliated blockholders, and ESOPs exceeds 25 percent during the contest. Some of these blocks are created during the contest in large part to fight off the takeover offer. STAGGERED implies that the board of directors consis ts of multiple classes. PRE-CONTEST VETO and POST-CONTEST VETO refer to shareholder blocks that are in excess of 40 percent before and after the contest, respectively, an d have the ability to effectively block a takeover. Information on shareholder blocks, standstill agreements, and staggered boards were collected from detailed news accounts and proxy statements.

PAGE 91

82 In response to growing hostility in takeove rs, poison pill plans were introduced in the mid 1980s. Poison pills allow target sharehol ders to purchase shares of the target or the surviving entity in case of a successful me rger at deep discounts, and therefore have the potential to make takeovers ‘prohibi tively expensive’. Data on poison pills are obtained from three sources. First, in many cases news accounts surrounding the takeover attempt document the existence of poison pi lls. If no news accoun ts mentioned poison pills, I checked the dataset used in Comment and Schwert (1995).4 Finally, I searched multiple editions of the “Corporate Anti-Take over Defenses: The Poison Pill Device” for evidence of the use of poison pills. POIS ON PILL is a dummy equal to one when the firm has a poison pill defense before the reso lution date of the contest identified by any of these three sources. Table 4.5 reports the fraction of targets that use the va rious defensive strategies available to the target during the takeover contes t. It is clear from the table that there has been a shift in the use of takeover defens es over time. Greenmail, dividend payouts, general share repurchases, a nd standstill agreements were more common in the pre TimeWarner period. The post Time-Warner peri od is characterized by few management initiated defenses. The fraction of firms that ha ve poison pills is highe r, but there are still many firms without them in the 1990s. It is important to remember however, that a firm can adopt a poison pill simply with board a pproval. In other words, it is quite possible that merely the threat of adopting a poison pi ll can fend off unsolicited bids. In fact, a number of firms adopted poison pills subsequent to the failed bids, presumably for future contests. Regardless, it appears that target s can more cheaply thwart unwanted raiders after the Time-Warner decision. Firms also we re more likely to have staggered boards of 4 The dataset can be found at Bill Schwert’s webs ite at http://schwert.ssb.rochester.edu/data.htm.

PAGE 92

83 directors in the post Time-Warner period, wh ich makes a good complement to the “just say no” takeover defense. 4.3.2 Leverage as a Takeover Defense There is an ongoing debate about the effect s of leverage in targets of unsuccessful takeover attempts. Berger et al. (1997) find th at firms increase leverage after being the target of an unsuccessful tender offer. They re port that the mean increase in total debt to assets is almost 12%. They argue that entren ched managers seek to avoid debt, but are forced to increase their firm’s leverage when their firm is targeted by an unfriendly suitor. Safieddine and Titman (1999) find a positive relationship between levering up and post-acquisition performance and conclude that leverage forces target management to make value improvements. In contrast, Ja ndik and Makhija (2003) argue that debt entrenches managers and that targets that remain independent subsequently underperform. With the structural changes in the legal environment as described in section 4.1.2 and the ease with which targets are able to deter unwanted acquirers, targets may no longer need to increase firm leverage. In f act, the evidence in Table 4.5 suggests that firms make fewer decisions that have effects on their capital structure. To investigate the impact of hostile offers on target leverage, Table 4.6 reports the mean leverage ratios for the target firms from two years prior to the offer until two years after the offer.5 Targets in both periods have very similar debt to asse ts ratios the two years prior to the offer. In the year of the offer and the two subsequent years, however, leverage ratios drastically 5 The pattern is very similar when median debt ratios are used and are therefore omitted from the table.

PAGE 93

84 increase for the pre Time-Warner targets, but stay virtually the same for the post TimeWarner targets. Figure 4.1 illustrates the changes in leve rage for firms with available Compustat data during the two years prior until two years afte r the offer. It is apparent that leverage ratios increase substantially only for the preTime-Warner targets. This evidence suggests that firms are able to defeat takeover bids without levering up through either defensive buyback (e.g., greenmail) strategies or issu ing debt in the post Time Warner legal environment. 4.4 Target Performance 4.4.1 Event Study Analysis Contested takeover battles are often very complicated and evolve over time. In this section I examine the market reaction to various events during the contest. Table 4.7 reports the abnormal returns for the targets around several event dates. Abnormal returns are net-of-market and calculated from one day before the event through one day after. I focus on five events: the offer date, rejection da te, last offer date, last rejection date, and the resolution date. The offer date is the first date that the bidder made a formal offer. The average abnormal return around the offer date is 17.6 percent. By construction, all target s reject and subsequently resist the takeover attempt.6 Many offers are rejected within one day; hence some of the returns on the rejection dates are difficult to interpret. In order to unders tand the market reaction of the rejection, I only report abnormal returns for rejection dates that are at least two days after the initial offer date. This results in 168 clean observations. The market reaction around the rejection date 6 One rejection date could not be obtained.

PAGE 94

85 is negative suggesting that the market does not view the rejection as a bargaining tool, but perhaps rather as an entrenchment move. Interestingly, the mean and median rejection returns are more negative during the pos t Time-Warner period, which may be an indication that the market believes that target s are more likely to successfully defeat the offer. The sample contains 101 deals where the original acquirer ma kes multiple offers. I define a multiple offer as an offer that is different from the first in terms of price or structure of the deal (e.g. from stock offer to all cash offer). The offer price was increased in 74 deals, remained the same in 17 deals, and was decreased for 10 deals. The average increase between the first and last offer is 9.8 percent, but the mean abnormal returns around the last offer date is only slightly over 4 percent. The aver age abnormal return on the last rejection date is -2.9 percent, and appears to be driven mostly by the latter sample. Finally, I report the abnormal returns on the resolution date. On the resolution date the bidder clearly indicates that its intentions are to withdraw from the takeover contest. Withdrawals includes cases where the bidder looks for altern atives such as the acquisition of similar targets, sells stock back to the target (whether it is for a premium or not), signs a standstill agreement, or lets the tender offer expire wi thout an indication of renewing. By searching Factiva I ensured that the resolution date represents the last real effort of the acquirer. I found 176 resolution dates using these crite ria. Resolution dates were easier to determine during the earlier period, where I identified 113/144 versus 83/126 in the latter period. One of the reasons for this differe nce is that the reliance on greenmail, which clearly suggests withdrawal declined significantly during the latter

PAGE 95

86 period. On average, the market reaction is sharply negative with an average abnormal return of -7.3 percent. This further indicat es the dismay of the market for targets’ management actions in successfully defeating the offer. 4.4.2 Post-Offer Valuation While markets generally react negatively to news of target rejections and bidder withdrawals, the positive abnorma l returns around the offer date appear not to be entirely offset. In fact, a quick calcula tion suggests that the average gain for typical targets is approximately 8 percent.7 To more carefully document the impact of the hostile offer on target valuation, I calculate two revaluation measures. The first revaluation measure (REVAL1) is computed as the difference be tween the target’s return and the CRSP value-weighted return calculated over th e period from 20 da ys prior to the leak date to 10 days after the withdrawal date. The second revaluation measure (REVAL2) is calculated from 20 days prior to the offer date through 10 days after the withdrawal date.8 Both measures are calculated with the assumpti on that all dividends paid out during the estimation windows are reinvested in the CRSP value-weighted market index. For thirteen cases, I adjusted the post-offer target prices for major self-tender offers for which CRSP did not make appropriate adjustments of returns to tendering sh areholders or cases where large special dividends are paid.9 7 This calculation assumes only one offer, and uses the average results in Table 7. The combined abnormal return from the offer, rejection, and withdrawal is equal to 17.6 – 2.3 – 7.3 = 8.0 percent. 8 The leak date is the first date for which there were rumors about the firm becoming a potential target. For the majority of the targets this is also the offer date. 9 The returns for firms with defensive self-tenders are calculated using post-offer prices computed as follows: [(%Bought in Tender)(Tender Price)] + [(1-%bought in Tender)(Presolve(CRSP))]. The post-offer return for firms with special divide nds is calculated assuming that speci al dividends are reinvested at the 30-day T-bill rate.

PAGE 96

87 Table 4.8 reports the mean and median revaluations for the pre and post TimeWarner samples. Approximately two thir ds of the targets experience a positive revaluation during the takeover contest, and th e mean target revaluation is 10.6 percent for REVAL1 and 6.3 percent for REVAL2. Cont rary to the management entrenchment hypothesis, the revaluations are higher dur ing the post Time-Warner period. Neither REVAL1 nor REVAL2, however, are statistica lly different between the two periods at any conventional significance level. Before further analysis of the impact of the takeover offer on target valuation, a few comments are worth making about the positive revaluation of the targets. The literature offers three potential explanations for the positive revaluation of targets of hostile takeover attempts. The first argues that the offer discloses positive information about the target not known to the market prior to the bid. This is sometimes coined the sitting on a gold mine hypothesis. The sec ond explanation, proposed by Bradley, Desai and Kim (1983) argues that targ ets with the greatest revaluation get acquired within a short period of time and this revaluation reflects future synergies from future acquisitions. I examine Bradley, Desai and Kim’s result by examining the revaluations for firms that are acquired within one, two, or three years. The results (not reported) indicate that targets that are acquired shor tly after defeating an offer generally have higher revaluations, but the differences are economically small and statistically insignificant due to the small sample size. The final explanation for positive target revaluation is that firms must restructure and recapitalize with more financial leverage on their own to avoid being acquired and this tr iggers a positive revaluation (see Safieddine and Titman (1999)). The effect of leverage on va luation is explored in the next section.

PAGE 97

88 All three explanations s uggest that the average post offer performance of the targets will merely match market returns if markets correctly anticipate the average tendencies of the targets. A bnormal returns for 250 and 500 da ys after defeating a hostile bid are reported in Table 4.8 confirm that pos t bid returns are not si gnificantly different from zero. It is intere sting to note though that post bid performance is actually higher for the post Time-Warner sample. This confir ms that the pre-Time-Warner group does not outperform the post Time-Warner sample in wa ys that were not anticipated by capital markets at the time of the bid’s defeat. While it is clear that targets experience a positive revaluation during the takeover contest, rejecting a takeover offer can also be viewed as an opportunity cost to shareholders when the price of the target tr ades at a discount relative to the offer price after the contest. I define two variables DISCOUNT1 and DI SCOUNT2 that measure this opportunity cost. DISCOUNT1 (DISCOUNT2) is the percentage difference between the target’s post-offer price and the initial (last) offer price. The post-offer price is adjusted for stock splits and for the thirteen special cases described in footnote 9. (4.1) DISCOUNT1 = (4.2) DISCOUNT2 = Table 4.8 also reports the mean a nd median numbers for DISCOUNT1 and DISCOUNT2. Less than one fifth of the targets trade at or above the final offer price ten days after the resolution date. On average, targets trade at a 14 to 20 percent discount offer resolve offerP P P lastoffer resolve lastofferP P P

PAGE 98

89 relative to the offer price subsequent to de feating the offer. Interestingly, while the takeover environment and the defenses to fend off unwanted acquirers have changed drastically, there is no significant time vari ation in the magnitude of the levels of DISCOUNT1 and DISCOUNT2. 4.4.3 Multivariate Analysis The previous section showed that there is little difference in post-offer valuations of takeover targets between the pre and po st Time-Warner periods. One interesting question is what determines these post-offer valuations? This section attempts to answer this question by examining how target charac teristics, bidder characteristics, and the takeover regime affect the reva luation of targets and the disc ount targets trade at relative to the offer price. It is natural to think that a target’s g overnance structure plays an important role in the success of takeover offers. Entrenched mana gers with large holdings or the backing of large affiliated shareholders have the ability to effectively resist takeover offers since takeover bids must be approved by the major ity of shareholders. In addition to large shareholdings, staggered board terms can be effective defense mechanism since it would take the bidder more than one proxy to ga in enough board representation necessary to either remove management or a poison pill. Hence, the variables BLOCK and STAGGERED are expected to be negatively rela ted to target revaluations and positively related to target discounts. In this analysis, I also try to control fo r the “seriousness” of the offer since many offers in the sample, especially tender offers are only for a controll ing interest in the

PAGE 99

90 target firm. In some cases, the bidder discusse s what will be offered to shareholders in a second step merger if the offer succeeds. However, most bidders are under no legal obligation to complete a second step offer. Even back-end offers that are discussed are often for unspecified securities and/or cash. It is likely th at the premiums suggested by the offering price in these bids are worth le ss if one considers the blended premium that targets would receive (shares bought on the fron t end averaged with shares bought on the back end). PERCENT SOUGHT captures this variation and is calculated as the percentage of shares ou tstanding net of foothold, that th e bidder is offering to acquire in the front-end. For non two-tier o ffers this variable is set equal to 100 percent. Other variables that proxy for the se riousness of the bid are BADREP, Foothold, SYNERGY, and CASH. Foothold is the percentage change in bidder’s stake from prior to the initial leak date until after the resolution date. We would expect positive signs on the coefficients for PERCENT SOUGHT, SYNERGY, Foothold, and CASH for the regressions on target revalu ation, and the opposite signs for regressions on target discount. The emphasis of this study is on the change s in the legal structure that allowed the use of the poison pill as a defense against unwanted acquirers. Since adoption of a poison pill only requires board appr oval and not a shareholder vo te, I include a dummy TW, which equals one for the post Time-Warner period and zero before to allow for the changes in post-offer target valuations between the two periods. Other important differences between the two pe riods are takeover defenses a nd the use of debt. Previous research has shown that leverage has a positive effect on returns of unsuccessful targets (Berger et al. (1997)). I incl ude change of leverage ( Leverage) and an interaction

PAGE 100

91 variable of change of leve rage with the TW dummy ( Leverage*TW) to examine the effect of leverage changes on target valuation during the two periods. A significant coefficient on the interaction dummy would impl y that leverage has a different impact on target valuation pre and post Time-Warner. Finally, I also include the more precise breakdown of the leverage cha nges and include dummies for use of buybacks, greenmail, and special dividends. The results of the regressions on RE VAL1 and REVAL2 are presented in Table 4.9. Firms with large blocking positions (25% or greater) affiliated with management have lower revaluations suggesting that large blockholders either have desires to keep the firm independent with non-value maximizing (for shareholder) intent or that they are apt to overvalue the plans of the existing management team (which often includes themselves). The presence of staggered boards is not a determinant in the revaluation. Surprisingly, revaluations are no higher wh en more than one bidder was involved in the takeover contest. Bidder characteristics appear to have larg e effects on target revaluations. Offers where the bidder has a bad reputation (BADR EP) have lower revaluations which may indicate that the market as well as mana gement finds such bids less compelling. The coefficient on Percent Sought is significantly positive at the one percent level, and shows that markets factor in the uncertainty of the value of the second step and consequently the seriousness of the offer. Targets in syne rgistic deals (SYNERGY) experience higher revaluations as well. This may reflect that these are bids with a real rational that have to be taken more seriously. CASH and Foothold do not appear to affect the revaluation measures.

PAGE 101

92 Equations (1) and (4) in Table 4.9, s how that leverage changes are positively related to REVAL1 and REVAL2. However, the in teraction variable indicates that this effect disappears and actually negatively impacts revaluations during the post TimeWarner period. In other words, the positive effect of leverage on firm value reported by Safieddine and Titman (1999) is unique to the pre Time-Warner sample. Equations (2) and (5) show that the form of levering up has a great effect as well. Large special dividend payouts have a positive effect on ta rget revaluations. Greenmail (payments made to the bidder for stock or standstill ag reements), on the other hand, reduces target revaluations. Share buybacks appear to posit ively impact the revalu ations, although this effect is not significant. Agai n, these results are probably more germane to the Pre-TimeWarner period since such transactions are rela tively rare in the post Time-Warner era. I conjecture that the different response to le verage pre and post Time-Warner may have more to do with the manner in which the le verage is increased. The pre-Time-Warner leverage increases are quite intentional whereas it is possible that the post Time-Warner increases are more of a reflection of pe rformance-driven leve rage increases. Columns (3) and (6) show revaluation regressions that omit defensive tactics such as leverage changes, greenmail, and buybacks. The adoption of such defenses may be driven by the pre and post Time-Warner le gal environments. However, there are no intertemporal differences in target revaluati ons as measured by the TW dummy even after eliminating differences in target defensiv e tactics that may be induced by the TimeWarner decision itself. Hence, even though defenses are quite different, valuation consequences related to changes in the take over environment do not appear in the data.

PAGE 102

93 Table 4.10 reports the OLS regressi ons of DISCOUNT1 and DISCOUNT2 on bidder, target, and takeover environment char acteristics. The only additional regressor included is the CRSP value-weighted market return (Market Return) which is calculated from the first (last) offer date to the resolution date for DISCOUNT1 and DISCOUNT2, respectively. Not surprisingly, Market Return is negatively related to the target discount measures, which means that the post-contest target price trades at a lower discount relative to the (last) offer price when ma rket returns are higher during the contest. Otherwise, the results are very similar to those presented in Table 4.9. Again, the market value of the post-contest target is higher rela tive to the offer price when firms leverage themselves up only in the pre-Time-Warner period. Also, the value of the post offer firm (relative to the final offer) is higher when th e firm buys back stock from all shareholders and when it pays a special dividend. Targets trade at larger discounts when management has a potential ownership blocking position owned by it or aligned with it. The nature of the bidder also comes into play. For some of the regressions, th e post-offer value is smaller the larger the change in foothold of the bidder during the contest. This suggests that there might be some value associated w ith the bidder co ntinuing to monitor the target firm. Also bidder interest an d seriousness are important dete rminants for the relative post-offer price. The post offer market price is lower relative to the final offer price when the bid is two-tiered and the bidder has a bad reputation. The combined evidence suggests that the credibility of a bidding firm has a major role in determining the post offer valuation of the target.

PAGE 103

94 4.5 Post-Offer Consequences to Firms and Management 4.5.1 Target Independence There are several reasons why management may not agree to sell to a hostile bidder. One reason explored in the previous s ections is that the offer and/or financing may not be serious. Another reason is in anticip ation of higher bids from either the same or competing bidders. Comment and Schwert (1 995), for example, argue that targets use the poison pill defense not to entrench themselves, but to bargain for a higher offer. By rejecting and resisting the initial unsolicited offer, management may be speculating that the same or other bidders return to the target with an alternative offer. The majority of targets in the sample, however, remain i ndependent for a prolonged period of time. Table 4.11 reports the number of targets that leave the CRSP tapes within one, two, and three years after the su ccessfully defeating the hostile offer. The panels separate the pre and post Time-Warner targets. During the full sample there are seven firms that delist due to bankruptcy within three years af ter the resolution date. Five firms undergo major restructuring and are no longer in CRSP as their original firms.10 Contrary to the idea that takeover defenses have made manage rs more entrenched, more targets in the 1990s are acquired subsequent to the unsucces sful takeover contest. There are 29 firms that are subsequently acquired in the 1980s versus 32 in the 1990s. Hence, targets that defeated offers post Time-Warner are a littl e more likely to subsequently be acquired. The table also reports the median value gain s/losses that are calculated using the final 10 Compared to Safieddine and Titman (1999), a larger fraction of our sample firms remain independent. To put the numbers into perspective, 40 percent of their firms leave the sample within three years because they are subsequently acquired compared to only 20 percent for our sample. This distinction is in large part due to differences in sample design. Wh ereas they include all withdrawn offers regardless of competing bids, we only consider failed hostile bids when no alternative offers are on the table.

PAGE 104

95 offer price for the targets. It appears that firms in the 1990s did not do worse when they finally did decide to sell. 4.5.2 Managerial Turnover So far I have established that targ et managers do not always maximize shareholder value by defeating takeover offers with prices well above current market prices. While the price differential between what has been offered and post market valuation hints at managerial entrenchment, it does not appear that these decisions are viewed differently between the pre and post Time-Warner periods despite the dramatic changes in the takeover environment a nd available defensive strategies. To further investigate the impact of the changes in the legal environment on managerial entrenchment, I examine the rela tionship of firm performance and forced departures of top executives. Entrenched ma nagers are difficult to force out, especially because of the firm governance structure they developed during their tenure. If the legal environment has contributed to entrenching management, we should expect to find fewer forced resignations and lower sensitivity of forced turnover to pe rformance for the post Time-Warner sample. I follow the top executive (CEO in mo st cases, president otherwise) for each target during the contest and for up two year s after the withdrawal date. If a turnover occurs during this time, I document whether the turnover was voluntary (retirement, health reasons, other employment) or forced.11 Panel A in Table 4.12 shows the fraction 11 Similar to Huson et al. (2001), we define forced turnover to be cases in which the CEOs were ousted, forced, or left without mention of other employment or reasons before the age of 60. Turnovers for CEOs under age 60 and the Wall Street Journal (i) does not report the reason for departure as involving death, poor health, or the acceptance of another position, or (ii) reports that the CEO is retiring, but does not

PAGE 105

96 of targets that replace their top executive in the two years af ter defeating a hostile offer. About one third of the firms experience CE O turnover, which compares to Denis and Serrano (1996) who find a turnover rate of 34 percent within two years following a failed takeover contest. Forced turnover occurs in 18 percent of the firms, with the majority of those initiated by the board. C onsistent with Huson et al. (2001), the number of forced resignations has substantially increased over time.12 Panel B reports the logist ic regressions of CEO tur nover and forced turnover. Since one of the goals is to establish the relationship between def eating an offer above current market prices and ma nagerial turnover, I exclude nine cases where the CEO is replaced during the takeover contest.13 The results indicate that performance prior and during the offer year are major determinants of CEO turnover. Th e industry-adjusted ROA in the year of the offer is significantl y negatively related to forced turnover. The variable DISCOUNT2, which measures the pr ice differential between the post-contest target price and the offer price, is negativel y related to turnover, albeit only significant for the full sample. This should not come as a su rprise as targets arguably trade at higher discounts when management teams have blocking positions and are hence likely to be more entrenched. After controlling for firm performance and ot her determinants of turnover, there is little evidence that CEOs that defeat hostile bids are more en trenched in the latter period. announce the retirement at least six months before succe ssion, are also classified as forced. Turnovers are reclassified as voluntary when in formation after the CEO turnover in dicates that the departure was not forced. 12 One obvious difficulty with assessing the frequencies of post-contest turnover is the overlap of the postcontest periods. For several firms that defeat an of fer prior to the Time-Warner decision, the CEO turnover will not occur until after the Time-Warner decision. When the years immediately around the Time-Warner decision are excluded, the differences in turnover rates are more significant. 13 Including these nine cases where th e CEO is replaced prior to the withdrawal date does not affect the results.

PAGE 106

97 The sensitivity of turnover to performance, measured by the interaction vari able has also not changed significantly over time. These results are consistent with my claim that the advances in corporate governance have made it more difficult for managers to entrench themselves. 4.6 Concluding Remarks This chapter examines the impact of ch anges in the legal e nvironment on target defense strategies, managerial entrenchment, and post-offer valuation. I divide the sample in two periods based on the Time-Warner c ourt decision in July 1989 that recognized the poison pill as a legal defense measure. Not surprisingly, I find substantial differences in managerial defensive actions in response to unsolicited offers between the two periods. Pre Time-Warner, targets use defensive stra tegies such as buybacks, dividend payouts, and greenmail to fend off unwanted acquirers Consistent with previous studies (e.g. Berger et al. (1997) and Sa fieddine and Titman (1999)), I find significant leverage increases and that higher leve rage leads to higher post-acq uisition valuations for firms that remain independent. Targets respond very differently in th e post Time-Warner period. More firms either already have or adopt a poison pill defense in response to a hostile takeover. The ‘just say no’ defense appears to work as a substitute defense mechanism as defensive recapitalizations virtually disappear. Contra ry to the pre Time-Warner period, hostile targets no longer lever up, and when they do, they underperform. The legal environment appear s to have little effect on the average “entrenchment” of management for this sample of firms. The post-offer valuations of targets that

PAGE 107

98 successfully defeat hostile bids are similar across the two pe riods. In addition, despite the decline in hostile takeover activity and a dvances in takeover defenses, managers are forced out from their position more freque ntly post Time-Warner. In conclusion, the evidence presented in this chapter is c onsistent with Kahan and Rock (2002) and contradicts the theory that the Time-Warner decision led to increased managerial entrenchment in the 1990s.

PAGE 108

99 Table 4.1 Sample Selection The table reports the selection criteria for the sample of hostile or solicited offers from the SDC database and Mergerstat. The final sample of defeated offe rs is reported in the last row of the table. Pre Time Warner Post Time-Warner Initial Sample of hostile or unsolicited SDC targets 676 801 Banks and S&Ls 56 75 REITS, Funds, Partnerships, Trusts 18 46 Offer for assets only 9 48 Offer for non-traded firm 39 124 Offer for bankrupt firm 9 40 Debt workout required 1 6 Offers for stakes (not control) or expressions of interest 32 35 Government or union led offers 3 1 Cleanup Offers 10 24 Offer price less than $1.50 2 8 Fraudulent Offers 2 1 Pac Man offers 5 1 Management led bid 17 8 No bidder identified or insufficient information on dates 12 29 Friendly bid 3 4 No Terms given on offer 2 13 News of offer/rejection comes out well after the fact 0 1 Remaining Firms 456 337 Redundant Entries (same contest) 67 43 Remaining Firms 389 294 Target acquired (unsolicited bid follows friendly bid) 39 42 Target shares sought acquired by bidder or other bidder 205 95 Firm announces intent to liquidate 1 0 Bidder acquires board control 5 4 Target firm changes of control for SDC data 250 141 Remaining Firms 139 153 Bidder backs out without target opposition 5 12 Bidder backs out after an agreement from target to a deal 4 9 Different prices offered to different classes of stock 0 1 Security offers of questionable value 7 8 "Mergers of equals" 1 14 Sample from SDC data sources 122 109 Additional observations from Mergerstat Review 22 17 Total sample observations 144 126

PAGE 109

100 Table 4.2 Sample of Defeated Hostile Deals The table reports the number of defeated offers in each year for a sample of 2 70 targets that successfully defeat a hostile offer between 1980 and 1999. Columns 2-5 report the number of contests with multiple offers, competing offers, and the median duration of th e takeover contests. Multiple Offers indicates that the target received more than one offe r from the original bidder. Competi ng Offers indicates that there were more than one bidder during the takeover contest. A contest is classified as pre Time-Warner (pre TW) when its withdrawal date is before July 17, 1989. C ontest Duration refers to the number of days between the offer date and resolution date. Defeated Offer Multiple Offers Competing Offers Contest Duration 1980 3 0 0 12 1981 5 0 1 21 1982 7 2 0 29 1983 5 0 0 21 1984 13 5 3 24 1985 18 4 2 23 1986 19 11 0 20 1987 30 11 1 25 1988 26 7 2 53 1989 29 15 2 58 1990 18 8 2 94 1991 10 3 0 95 1992 4 3 0 35 1993 6 1 0 39 1994 9 6 1 24 1995 12 5 0 28 1996 12 4 0 20 1997 14 5 0 32 1998 11 5 3 24 1999 16 6 1 21 2000 3 0 1 62 Pre TW 144 49 11 25 Post TW 126 52 8 32

PAGE 110

101 Table 4.3 Target Characteristics The table reports firm characteristics for the sample of targets that successfully defeat a hostile offer between 1980 and 1999. Panel A reports the stock and accounting performance and Panel B reports the industry breakdown of the targets. The abnormal returns are net-of -market returns calculated over 500 and 250 days prior to 21 days before the first leak date. A ccounting data is calculated the fiscal year prior to the offer. ROA is the ratio of EBIT and beginning period assets. FCF = (Operating Income before Depreciation)-(Income Tax Change Deferred Tax) Interest expense – Preferred dividends common dividends. Market/book is measured as the ratio of the year-end market value of common stock to the book value of equity for the prior fiscal year. Target Size (in millions) is the CPI-adjusted (1980=1) equity value of the target is calculated 20 days prior to the initia l takeover announcement. A contest is classified as pre Time-Warner when its resoultion date is before July 17, 1989. Panel A: Stock and Accounting Performance Full Sample Pre TimeWarner Post TimeWarner p-value difference in means Abnormal Return (-520,-21) -11.50** -7.75 -15.80 0.44 Abnormal Return (-270,-21) -9.04*** -4.63 -14.04*** 0.07 ROA 9.14 9.36 8.88 0.72 FCF/Assets 11.49 11.87 11.06 0.53 Market/Book 1.51 1.28 1.77 0.01 Target Size 403 508 286 0.05 Panel B: Target Industries SIC codes Pre Time Warner Post Time Warner 1000-1999 13 4 2000-2999 23 17 3000-3999 39 53 4000-4999 22 18 5000-5999 19 8 6000-6999 11 8 7000-7999 11 11 8000-8999 5 6 9000-9999 1 1

PAGE 111

102 Table 4.4 Bidder Characteristics The table reports bidder characteristics at the time of the offer for the full sample and the preand post Time-Warner periods. A contest is classified as pre Ti me-Warner when its withdrawal date is before July 17, 1989. PUBLIC bidders are exchange-listed domestic or foreign firms. FINANCIAL BUYER is a private investment firms or a firm that has been paid greenmail in the five years prior to the bid. BADREP refers to eight firms in which the principal is under indictment, has run a firm into bankruptcy, or backed out of a negotiated deal in the past two years due to financing problems. SYNERGY is a dummy equal to one if the bidder has some potential vertical or horizontal strategy for use of the target’s assets in conjunction with the acquirer’s own field of expertise. Premium is defined as the ratio of offer price to price of target 20 days prior to offer and is in percentages. CASH is defined as an offer that consists of cash only. TWO-TIER refers to offers that are front-en d loaded. FOOTHOLD refers to bidders that own more than five percent of the target’s stock prior to the initial offer. Percent Foothold is the average stake (in percent) that the bidder owns prior to the first initia l bid. Percent Post-Foothold is the average stake of the bidder following the contest. Bidder Size (in millions) is the CPI-adjusted (1980=1) equity value of the target calculated 20 days prior to the initial takeover announcement. Full sample Pre Time-Warner Post Time-Warner p-value difference PUBLIC 0.56 0.53 0.59 0.29 Bidder Size 832 778 883 0.78 FINANCIAL BUYER 0.49 0.58 0.37 0.00 BADREP 0.03 0.04 0.02 0.21 SYNERGY 0.57 0.48 0.67 0.00 Premium 34.6 33.4 35.9 0.39 CASH 0.83 0.84 0.82 0.62 TWO-TIER 0.09 0.15 0.03 0.00 FOOTHOLD 0.49 0.55 0.42 0.04 Percent Foothold 6.47 7.06 5.79 0.16 Percent Post-Foothold 4.73 4.66 4.82 0.85

PAGE 112

103 Table 4.5 Takeover Defenses The table reports the fraction of targets that used any of the following takeover defenses to fend off hostile acquirers. BUYBACK is the repurchase of 10 percent or more of stock either through a tender offer or open market purchases. DIVIDEND is defined as a dividend greater than 10 percent or more of firm market value and includes dividends initiated by spin-offs. GR EENMAIL is defined as either a repurchase of shares from the hostile acquirer at a premium or the payout of a premium to secure a standstill. STANDSTILL refers to cases where the bidder is subj ect to a standstill agreement with the target. ESOP refers to the creation of an ESOP during the takeover contest. WHITE SQUIRE refers to the placement of a firm’s securities to a friendly party. RECAPITALI ZATION is defined as an y combination of share buyback, greenmail, and dividend payout that exceeds 10 percent. In order to be included as a defensive measure, recapitalizations must be announced between one month before the first offer and one month after the withdrawal date and executed within one year STAGGERED implies that the board of directors consists of multiple classes. A firm has a BLOCK when the combined equity stake of directors, affiliated blockholders, and ESOPs exceeds 25 percent. PR E-CONTEST VETO and POST-CONTEST VETO refer to a combined equity stake of dir ectors, affiliated blockholders, and ESOPs exceeds 40 percent before and after the contest, respectively. POISON PILL refers to firms that have a poison pill in place before or adopt one during the takeover contest. A takeover contest is classified as pre Time-Wa rner when its withdrawal date is before July 17, 1989. The p-value of the differences in fractions between the preand post TimeWarner samples is based on the t-test (equal variances). Full Sample Pre TimeWarner Post TimeWarner p-value difference BUYBACK 0.13 0.19 0.07 0.01 DIVIDEND 0.07 0.11 0.03 0.01 GREENMAIL 0.15 0.23 0.06 0.00 STANDSTILL 0.16 0.22 0.09 0.00 ESOP 0.06 0.08 0.04 0.20 WHITE SQUIRE 0.08 0.10 0.05 0.08 RECAPITALIZATION 0.25 0.36 0.12 0.00 STAGGERED 0.51 0.45 0.58 0.04 BLOCK 0.36 0.35 0.37 0.75 PRE-CONTEST VETO 0.13 0.13 0.14 0.67 POST-CONTEST VETO 0.20 0.21 0.20 0.84 POISON PILL 0.48 0.40 0.58 0.00

PAGE 113

104Table 4.6 Target Leverage Changes The table reports mean and standard devia tions of leverage ratios and changes for ta rgets of hostile takeover attempts. The lev erage ratios are calculated from two years prior to two years after the initial offer date. The initial o ffer occurs between year t-1 an d t. Leverage is defined as the ratio of total debt to total assets. Leverage ratios are set equal to one whenever total debt exceed s total assets. The number of observations is different for each year due to missing Compustat data and firm delistings. A takeover contest is cla ssified as pre Time-Warner when its withdr awal date is before July 17, 1989. The p-value of differences in leverage is based on the t-test (equal variances). Full Sample Pre Time-Warner Post Time-Warner Mean difference Mean Standard Dev. Mean Standard Dev. Mean Standard Dev. p-value Leverage (t-2) 26.6 17.5 27.0 16.1 26.1 19.2 0.68 Leverage (t-1) 26.2 17.8 26.5 16.7 25.8 19.0 0.74 Leverage (t) 31.0 22.0 34.1 23.1 27.4 20.2 0.01 Leverage (t+1) 31.6 22.8 35.5 23.6 27.0 21.1 0.00 Leverage (t+2) 32.4 23.4 37.3 23.6 26.9 22.1 0.00 Leverage (t-1,t+1) 6.21*** 19.4 9.8*** 21.4 2.04 15.7 0.00 Leverage (t-1,t+2) 7.12*** 20.5 11.0*** 21.1 2.71 19.0 0.00 **** indicates significance at the one percent level

PAGE 114

105Table 4.7 Event Returns The table reports the market-adjusted returns around event dates du ring the takeover contest. The market adjusted returns are c alculated over the (-1,1) window. Offer is the initial offer date. Rejection is the date of the fi rst rejection. For contests with multiple offers, Last Offer is the date of the final offer, and Last Rejection is the rejection date of the last offer. Returns aro und the Rejection and Last Rejec tion dates are calculated only wh en the rejections occur at least two days after the Offer and Last Offer dates re spectively. Resolution is the official withdr awal date if available. A takeover con test is classified as pre Time-Warner when its withdrawal date is before July 17, 1989. Full Sample Pre Time-Warner Post Time-Warner N Mean Std Dev. N Mean Std Dev. N Mean Std Dev. Offer 270 17.7*** 15.6 144 16.1*** 13.2 126 19.5*** 17.9 Rejection 168 -2.2*** 7.5 104 -1.8*** 6.8 64 -2.5*** 8.5 Last Offer 101 4.3*** 7.6 49 3.7*** 7.4 52 4.5*** 7.9 Last Rejection 73 -2.9** 9.4 33 -1.5 11.7 40 -3.9*** 6.9 Resolution 195 -7.3*** 8.8 113 -7.5*** 8.9 83 -7.1*** 8.7 ***, ** indicates significance at th e one and five percent respectively.

PAGE 115

106 Table 4.8 Post-Acquisition Valuation The table reports the mean values and standard deviations of REVAL1, REVAL2, DISCOUNT1, DISCOUNT2, and abnormal returns. REVAL1 is defined as the market-adjusted target return from 20 days prior to the leak date to 10 days after the resolution date. REVAL2 is defined as market-adjusted target return from 20 days prior to the offer date to 10 days after the resolution date. Both measures assume that dividends paid out during the estimation windows are reinvested in the CRSP value-weighted market index. DISCOUNT1 is defined as the ratio of the target price 10 days after the resolution date and the initial offer price. DISCOUNT2 is defined as the ratio of the target price 10 days after the resolution date and the last offer price. The abnormal returns are net-of-market returns calculated over 500 and 250 from 10 days after the resolution date. A takeover contest is classified as pre Time-Warner when its withdrawal date is before July 17, 1989. P-values of the differences in means are insignificant and are excluded. Full Sample Pre Time-Warner Post Time-Warner Mean Std Dev. Mean St d Dev. Mean Std Dev. REVAL1 10.6*** 32.7 9.8*** 27.4 11.6*** 38.0 REVAL2 6.3*** 30.8 4.2*** 25.2 8.7*** 36.1 DISCOUNT1 15.7*** 20.9 17.2*** 18.4 14.1*** 23.4 DISCOUNT2 18.5*** 19.6 19.3*** 16.8 17.6*** 22.4 Abnormal Return (1,250) 2.2 56.0 -3.6 44.2 8.9 66.7 Abnormal Return (1,500) 6.5 109.6 -6.7 70.8 21.3 140.0 **** indicates significance at the one percent level

PAGE 116

107 Table 4.9 OLS Regressions on REVAL1 and REVAL2 The table reports the OLS regression coefficients of the post-acquis ition revaluation measures (REVAL1 and REVAL2) on target and bidder characteristic s and legal environment characteristics. REVAL1 (REVAL2) is defined as the market-adjusted target return from 20 days prior to the leak (offer) date to 10 days after the resolution date. Both measures assume that dividends paid out during the estimation windows are reinvested in the CRSP value-weighted market index. DISCOUNT1 (DISCOUNT2) is defined as the ratio of the target price 10 days after the resolution date and the initial (last) offer price. TW is a dummy equal to one if the takeover contest is resolved before July 17, 1989. in Leverage is the difference in total debt / total assets ratio from one year prior to the offer to one year after withdrawal. Percent Sought is defined as (frontend sought-foothold)/(1-(foothold/100)). Foothold is the percentage change in bidder’s stake from prior to the initial leak date until after the resolution date. COMPETITION is a dummy variable equal to one for contests with multiple bidders. All other variables are reported in Tables 4.2, 4.4, and 4.5. P-values for the two-tailed test that the coefficients equal zero are reported below the coefficients. REVAL1 REVAL2 (1) (2) (3) (4) (5) (6) Intercept -37.81* -19.30 -35.00* -36.11** -16.62 -32.27* (0.06) (0.27) (0.06) (0.04) (0.34) (0.07) TW 1.58 2.16 0.93 2.98 2.58 2.62 (0.70) (0.60) (0.82) (0.40) (0.51) (0.48) Leverage 29.75** 29.69** (0.02) (0.02) ( Leverage)*TW -40.17** -55.70*** (0.05) (0.00) BUYBACK 1.49 2.13 (0.78) (0.70) GREENMAIL -4.13 -9.85** (0.37) (0.03) DIVIDEND 29.23*** 20.56*** (0.00) (0.01) Percent Sought 0.28*** 0.27** 0.31** 0.29** 0.24** 0.29** (0.03) (0.02) (0.01) (0.02) (0.03) (0.01) BADREP -23.25* -22.11* -23.98** -29.16* -24.95** -25.99** (0.07) (0.05) (0.03) (0.07) (0.02) (0.01) SYNERGY 9.86** 9.58** 10.86*** 9.89** 9.62** 11.21*** (0.02) (0.02) (0.01) (0.01) (0.02) (0.01) Foothold 0.60 1.02* 1.23** 0.36 0.51 0.77* (0.15) (0.06) (0.01) (0.23) (0.31) (0.09) CASH 5.40 1.44 2.03 5.27 1.27 1.56 (0.29) (0.81) (0.73) (0.28) (0.83) (0.80) BLOCK -5.93 -6.52 -6.76 -10.21*** -11.56*** -11.82*** (0.15) (0.14) (0.13) (0.01) (0.01) (0.01) STAGGERED -3.17 -7.24* -6.39 -1.08 -6.69* -6.02 (0.41) (0.07) (0.12) (0.76) (0.08) (0.13) COMPETITION 6.36 14.75 17.84 -5.80 6.50 8.92 (0.50) (0.34) (0.25) (0.40) (0.67) (0.56) Size 0.88 -0.19 0.92 0.48 -0.13 0.65 (0.37) (0.86) (0.39) (0.57) (0.89) (0.51) Adjusted R2 5.4 12.5 8.2 10.4 11.8 8.7 ***, **, indicates significan ce at the one, five, and ten percent level respectively.

PAGE 117

108 Table 4.10 OLS Regressions on DISCOUNT1 and DISCOUNT2 The table reports the OLS regression coefficients of DISOUNT1 and DISCOUNT2 on target and bidder characteristics and legal environment characteristics. REVAL1 (REVAL2) is defined as the marketadjusted target return from 20 days prior to the leak (offer) date to 10 days after the resolution date. Both measures assume that dividends paid out during the estimation windows are reinvested in the CRSP valueweighted market index. DISCOUNT1 (DISCOUNT2) is defined as the ratio of the target price 10 days after the resolution date and the initial (last) offer pri ce. TW is a dummy equal to one if the takeover contest is resolved before July 17, 1989. Market Return is the CRSP value-weighted return from the first (last) offer date to the resolve date for Discount1 and Discount2, respectively. in Leverage is the difference in total debt / total assets ratio from one year prior to th e offer to one year after w ithdrawal. Percent Sought is defined as (front-end sought-foothold)/(1-(foothold/100)). Foothold is the percentage change in bidder’s stake from prior to the initial leak date until after the resolution date. COMPETITION is a dummy variable equal to one for contests with multiple bidders. All other variables are reported in Tables 4.2, 4.4, and 4.5. P-values for the two-tailed test that the coefficients equal zero are reported below the coefficients. DISCOUNT1 DISCOUNT2 (1) (2) (3) (4) (5) (6) Intercept 92.59*** 87.99*** 96.49*** 98.27*** 92.70*** 100.79*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) TW -2.61 -2.09 -1.60 0.38 -0.04 -0.14 (0.30) (0.40) (0.51) (0.87) (0.99) (0.95) Market Return -0.41*** -0.46*** -0.44*** -0.50*** -0.55*** -0.53*** (0.01) (0.00) (0.00) (0.00) (0.00) (0.00) Leverage -21.34*** -15.74** (0.01) (0.01) ( Leverage)*TW 41.35*** 32.89** (0.00) (0.02) BUYBACK -5.14 -3.56 (0.16) (0.26) GREENMAIL 4.49 5.56* (0.24) (0.06) DIVIDEND -14.84*** -13.40*** (0.00) 0.00 Percent Sought -0.25** -0.24*** -0.26*** -0.21** -0.20*** -0.23*** (0.01) (0.01) (0.00) (0.01) (0.01) (0.00) BADREP 22.02** 18.91** 20.39** 21.18** 18.76** 19.81** 0.02 0.03 (0.01) (0.02) (0.02) (0.01) SYNERGY -3.83 -3.3 8 -4.07 -3.34 -2.97 -3.83 (0.13) (0.20) (0.13) (0.17) (0.23) (0.12) Foothold -0.22 -0.27 -0.43 -0.09 -0.15 -0.31 (0.31) (0.34) (0.10) (0.67) (0.60) (0.22) CASH -0.86 1.53 0.90 2.85 3.17 2.73 (0.83) (0.70) (0.82) (0.45) (0.42) (0.48) BLOCK 6.35** 6.52** 7.17*** 4.94** 5.23** 5.39** (0.01) (0.01) (0.01) (0.04) (0.04) (0.03) STAGGERED 4.50* 7.05*** 6.72*** 3.58 6.21*** 5.87** (0.06) (0.00) (0.01) (0.12) (0.01) (0.01) COMPETITION 2.58 -0.15 -2.14 1.34 -1.70 -3.41 (0.54) (0.98) (0.78) (0.73) (0.83) (0.65) Size -1.00 -0.41 -1.13* -0.87 -0.26 -0.85 (0.12) (0.53) (0.08) (0.14) (0.66) (0.15) Adjusted R2 15.3 17.8 14.3 14.6 17.6 14.3 ***, **, indicates significan ce at the one, five, and ten percent level respectively.

PAGE 118

109 Table 4.11 Distribution of Delisted Targets The table reports the number of targ ets that delist within 12, 24, and 36 months after successfully defeating a hostile takeover attempt. Panel A and B report the delistings for pre and post Time Warner targets respectively. Delistings are the result of bankruptcy restructuring, and acquisitions. Bankruptcies are identified by CRSP. Restructurings are delistings where the firm may still exist as a stand alone firm after substantially changing its character istics as a result of large acquisiti ons, sell offs, and recapitalizations. Acquisitions are identified by CRSP. The last two column s report the median percentage difference in price at which targets are acquired and th e last offer price by the hostile acquirer. Column 5 reports the raw differences between the last offer pr ice and acquisition price. The last column reports the ratio where the last offer price is market adjusted to the date of delisting. Panel A: pre Time Warner Reason for delisting Value gain/loss Months after withdrawal Bankruptcy Restructuring Acquired /merger Pacquired / Poffer Market-adj Pacquired / Poffer 1-12 0 2 10 16.7 2.2 13-24 2 1 14 19.8 9.1 25-36 2 0 5 1.67 -35.2 Panel B: post Time Warner Reason for delisting Value gain/loss Months after withdrawal Bankruptcy Restructuring Acquired /merger Pacquired / Poffer Market-adj Pacquired / Poffer 1-12 0 0 10 46.7 33.3 13-24 2 2 8 21.5 -4.5 25-36 1 0 14 23.6 -9.7

PAGE 119

110 Table 4.12 CEO Turnover Statistics and Logit Regression on CEO Turnover Panel A reports the fraction of targets that experience turnover of the top execu tive during the two years after successfully defeating a hostile offer for the full sample and the pr eand post Time-Warner periods. A contest is classified as pre Time-Warner when its withdrawal date is before July 17, 1989. A firm experiences CEO turnover whenever a succession of th e top officer occurs, excluding those that are the result of health reasons an d death. Forced turnover excludes top executive turnover labeled as normal succession using the classification scheme of Huson et al. (2002). Forced by boar d indicates that the CEO is forced out by its own board of directors. Forced by Other Acquirer means that the CEO is ousted after its firm is acquired. Panel B reports the logistic regressions of CEO Turnover and Forced Turnover. Marketadjusted return is the difference between the return on the target and the market during the two years prior to the offer date. ROA is the ratio of EBIT to beginning of period assets less the median ROA for all firms in the same two-digit SIC code in withdrawal year. Age is CEO age at the time of the withdrawal. Unaffiliated is the equity stake of unaffiliated blockholders. The remaining variables are described in Table 4.11. P-values for the two-tailed test that the co efficients equal zero are below the coefficients. Panel A: Fraction of Targets Experienci ng CEO Turnover and Forced Turnover Full Sample Pre Time-Warner Post Time-Warner p-value difference CEO Turnover 0.34 0.29 0.40 0.05 Forced 0.18 0.15 0.22 0.10 by Board 0.13 0.10 0.15 0.25 by Other Acquirer 0.06 0.04 0.07 0.29 Panel B: Logistic Regressions of CEO Turnover and Forced Turnover Full Sample (N=242) Independent Firms (N=213) CEO Turnover Forced Turnover CEO Turnover Forced Turnover Intercept -2.28 -0.42 -3.82 -0.86 (0.14) (0.79) (0.03) (0.64) TW 0.22 0.25 0.36 0.24 (0.47) (0.55) (0.31) (0.63) Age 0.04* 0.06** (0.07) (0.01) Market-adjusted Return -0.06 -0.35 -0.14 -0.90* (0.74) (0.32) (0.55) (0.07) Industry-adjusted ROA -6.00** -7.41** -5.91* -8.98** (0.04) (0.04) (0.06) (0.03) ROA*TW 2.54 1.91 3.23 3.76 (0.44) 0.65 0.36 (0.43) DISCOUNT2 -0.02* -0.02 -0.01 -0.01 (0.07) (0.20) (0.17) (0.27) BLOCK -0.5 2 -0.57 -0.75* -0.73 (0.13) (0.21) (0.06) (0.19) Unaffiliated 0.00 -0.01 0.01 -0.00 (0.86) (0.70) (0.72) (0.89) Size -0.02 -0.10 -0.03 -0.09 (0.83) (0.44) (0.79) (0.52) Pseudo R2 5.4 7.8 7.2 13.0 ***, **, indicates significan ce at the one, five, and ten percent level respectively.

PAGE 120

111 Target Leverage pre Time Warner Offer date Target Leverage post Time-Warner Industry Leverage pre Time-Warner Industry Leverage post Time-Warner0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 -2-1012Date relative to offer dateTotal debt/Total assets Figure 4.1 Target Leverage Changes Surrounding Unsu ccessful Hostile Takeover Attempts The figure shows the mean leverage ra tios for target firms that successful ly defeat a hostile takeover offer. Leverage is defined as the ratio of total debt to total assets. Leverage ratios are set equal to one whenever total debt exceeds total assets. Only firms with complete data on leverage two years prior to trough two years after the offer date are included. A takeover contest is classified as pre Time-Warner when its withdrawal date is prior to July 17, 1989. Industry Leve rage ratio is the mean leverage ratio for a portfolio of firms in the same industry as target firms based on the two-digit SIC code.

PAGE 121

112 CHAPTER 5 CONCLUSION The primary goal of this dissertation is to examine the impact of the changes of the takeover and legal envi ronment over the past two decades on managerial entrenchment. Some researchers have argued that managerial entrenchment has been exacerbated as the pressures of the takeove r markets subsided as advanced takeover defenses became legally permissible and fina ncing deals became more difficult. To test this conjecture, I examine two key managerial actions that are often associated with managerial entrenchment, namely the deci sion to make a large acquisition and the decision to resist a takeover offer that re sults in a firm remaining independent. Chapter Three examines the effectivene ss of the hostile takeover market and the board of directors in disciplining manage rs that make value-destroying acquisition decisions. Unlike Mitchell and Lehn (1990), I find no evidence that “bad acquirers” are more likely to become hostile takeover targets. Instead, there is an inverse relationship between the wealth effects of acquisitions and managerial turnover. This relationship is stronger during the 1990s period when board s faced less pressure from the takeover markets. This suggests that boards and takeove r markets act as subs titute mechanisms. Chapter Four examines a sample of targ ets that successfully defeat an unsolicited takeover offer. I document that takeover defe nses have significantly changed over time. Managers no longer have to resort to “expensive” buybacks, standstill agreements, and greenmail payments. Instead, they rely on an insider block and/or the poison pill defense

PAGE 122

113 in the 1990s. Nevertheless, the post-acquisition target valuation char acteristics are very similar between the pre and post Time-Warne r periods. Also, post Time-Warner targets are more likely to be acquired and their management more lik ely to be replaced in the years immediately after the takeover contest. Hence, the results in Chapter Four are not consistent with the theory that firms that maintain their independence can stray further from value maximization. This research has several implications for the takeover market and regulators. Previous research has suggested that the ta keover market, and in particular the hostile takeover market was vital to correct the actions of non-value maximizing management teams (e.g. Jensen (1993), Mikkelson and Pa rtch (1997)). Hence, the introduction and wide acceptance of the poison pill defense led many to think that managers would become more entrenched as the principal control mechanism was weakened. The empirical evidence in Chapters Three and Four, however, does not support this conjecture. Consistent with Kahan and Rock (2002) the evidence presented here suggests that alternative control mechanisms are eff ectively substituting for the decline in hostile takeover activity, and that the Time-War ner decision had far fewer negative consequences than originally thought.

PAGE 123

114 LIST OF REFERENCES Agrawal, Anup, and J. Jaffe, 2003, Do ta keover targets under-perform? Evidence form operating and stock returns, Journal of Financial and Quantitative Analysis 38, 721-746. Agrawal, Anup, and Ralh A. Walkling, 1994, Executive careers and compensation surrounding takeover bids, Journal of Finance 49, 985-1014. Andrade, Gregor, Mark Mitchell, and Erik Stafford, 2002, New evidence and perspectives on mergers, Journal of Economic Perspectives 15, 103-120. Becher, David A., Terry L. Campbell, an d Melissa B. Frye, In Press, Incentive compensation for bank directors: The impact of deregulation, Journal of Business. Berger, Philip G., Ofek, Eli, and David L. Yermack, 1997, Managerial entrenchment and capital structure decisions, Journal of Finance 52, 1411-1438. Borokhovich, Kenneth A., Robert Parrino, and Teresa Trapan i, 1996, Outside directors and CEO selection, Journal of Financial and Quantitative Analysis 31, 337-355. Bradley, Michael, Arnand Desai, and E. Ha n Kim, 1983, The rationale behind interim tender offers: information or synergy? Journal of Financial Economics 11, 183-206. Bradley, Michael, Arnand Desai, and E. Han Kim, 1988, Synergistic gains from corporate acquisitions and thei r division between the stockholde rs of target and acquiring firms, Journal of Financial Economics 21, 3-40. Brickley, James, Jeffrey Coles, and James Linck, 1999, What happens to CEOs after they retire? New evidence on career concerns horizon problems, and CEO incentives, Journal of Financial Economics 52, 341-377. Brickley, James, Jeffrey Coles, and Rory Terry, 1994, The board of directors and the enactment of poison pills, Journal of Financial Economics 35, 371-390. Brown, Steven J., and Jerold B. Warner, 1985, Using daily stock returns: The case of event studies, Journal of Financial Economics 14, 3-32. Byrd, John, and Kent Hickman, 1992, Do outside directors monitor management? Evidence from tender offer bids, Journal of Financial Economics 32, 195-221.

PAGE 124

115 Comment, Robert, and G. William Schwert, 1995, Poison or placebo? Evidence on the deterrents and wealth effects of modern antitakeover measures, Journal of Financial Economics 39, 3-43. Cotter, James, Anil Shivdasani, and Marc Zenner, 1997, Do independent directors enhance target shareholder wealth during tender offers? Journal of Financial Economics 43, 195-218. Denis, David J., and Diane K. Denis, 1995, Performance changes following top management dismissals, Journal of Finance 50, 1029-1057. Denis, David J., and Timothy A. Kruse 2000 Managerial discipline and corporate restructuring following performance declines, Journal of Financial Economics 53, 391424. Denis, David J., and Jan M. Serrano, 1996, Active investors and management turnover following unsuccessful control contests, Journal of Financial Economics 40, 239-266. Fama, Eugene, and Michael Jensen, 1983, Se paration of ownership and control, Journal of Law and Economics 26, 301-325. Federal Reserve Board, Flow of Funds Accounts of the United States, http://www.federalreserve.gov/releases/Z1 accessed March 2004. Gillan, Stuart C., and Laura T. Starks 2000, Corporate governance proposals and shareholder activism: The role of institutional investors, Journal of Financial Economics 57, 275-305. Gompers, Paul, and Andrew Metrick, 2000, In stitutional investors and equity prices, Quarterly Journal of Economics 116, 229-259. Grossman, Sandy, and Oliver Hart, 1980. Takeove r bids, the free-rider problem, and the theory of the firm. Bell Journal of Economics Spring, 42-64. Hadlock, Charles J., and Gerald Lume r, 1997, Compensation, turnover, and top management incentives: Historical evidence, Journal of Business 70, 153-187. Harford, Jarrad, 2003, Takeover bids and target di rectors incentives: The impact of a bid on directors wealth and board seats, Journal of Financial Economics 69, 51-83. Hartzell, Jay C., and Laura L. and Starks, 2003, Institutional investors and executive compensation, Journal of Finance, Forthcoming. Healy, Paul M., Krishna G. Palepu, and Ri chard S. Ruback, 1992, Does corporate performance improve after mergers? Journal of Financial Economics 31, 135-175.

PAGE 125

116 Heaton, J. B., 2002, Managerial optimism and corporate finance, Financial Management 31, 33-45. Hermalin, Benjamin E., and Michael S. We isbach, 2003, Boards of directors as an endogenously determined institution: a survey of the economic literature, FRBNY Economic Policy Review April 2003. Hilder, David B., 1990, Ruling by court on Time Inc.s me rger affirms the power of corporate boards, Wall Street Journal February 28, p. A3. Holmstrom, Bengt, and Steven N. Kaplan, 2000, Corporate governance and merger activity on the U.S.: Making sense of the 1980s and 1990s, Working Paper, MIT, Cambridge, MA. Huson, Mark R., 1997, Does governance matter? Evidence from CalPERS interventions, Working paper, University of Alberta. Huson, Mark R., Robert Parrino, and Laur a L. Starks, 2001, Internal monitoring mechanisms and CEO turnover: A long-term perspective, Journal of Finance 56, 22652297. Jandik, Tomas, and Anil K. Makhija, 2003, Debt, debt structure, and corporate performance after unsuccessful takeovers: Ev idence from target firms that remain independent, Working Paper, Ohio State University, Columbus. Jensen, Michael C., 1986, Agency costs of free cash flow, corporate finance, and takeovers, American Economic Review 76, 323-329. Jensen, Michael C., 1993, The modern industria l revolution, exit, and the failure of internal control systems, Journal of Finance 48,831-880. Kahan, Marcel, and Edward B. Rock, 2002, How I learned to stop worrying and love the pill: Adaptive responses to takeover law, University of Chicago Law Review 69, 871-915. Kaplan, Steven, 1989, The effects of mana gement buyouts on operations and value, Journal of Financial Economics 24, 217-254. Kaplan, Steven, and David Reishaus, 1990, Outside directorships and corporate performance, Journal of Financial Economics 27, 389-410. Kaplan, Steven N., and Michael S. We isbach, 1992, The success of acquisitions: Evidence from divestitures, Journal of Finance 47, 107-138. Karpoff, Jonathan M., 2001, The impact of sh areholder activism on target companies: A survey of empirical findings, Working paper, University of Washington.

PAGE 126

117 Kini, Omesh, William Kracaw, and Shehzad Mi an (1995) Corporate takeovers, firm performance, and board composition, Journal of Corporate Finance 1, 383-412. Loughran, Tim, and Anand M. Vijh, 1997, Do long-term shareholders benefit from corporate acquisitions? Journal of Finance 52, 1765-1790. Livingston, Miles, and Glenn Williams, 2003, Drexel Burnham Lambert, underwriting fees, and market power, Working paper, University of Florida, Gainesville. Manne, Henry G., 1965, Mergers and th e market for corporate control, Journal of Political Economy 73, 110-120. Martin, Kenneth J., 1996, The method of paymen t in corporate acquisitions, investment opportunities, and management ownership, Journal of Finance 51, 1227-1246. Martin, Kenneth J., and John J. McConnell, 1991, Corporate performance, corporate takeovers, and management turnover, Journal of Finance 46, 671-687. Mikkelson, Wayne H., and Megan Partch 1997, The decline of takeovers and disciplinary managerial turnover, Journal of Financial Economics 44, 205-228. Mitchell, Mark L., and Kenneth Lehn, 1990, Do bad bidders become good targets? Journal of Political Economy 98, 372-398. Mitchell, Mark L., and Harold Mulherin, 1996, The impact of industry shocks on takeover and restru cturing activity, Journal of Financial Economics 41, 193-229. Morck, Randall, Andrei Shleif er, and Robert W. Vishny, 1989 Characteristics of targets of hostile and friendly takeovers, In Alan J. Auerbach, editor, Corporate Takeovers: Causes and Consequences Chicago: University of Chicago Press, 1988. Mulherin, Harold, and Audra L. Boone, 2000, Comparing acquisitions and divestitures, Journal of Corporate Finance 6, 117-139. Myers, Stewart C., and Nicholas S. Majluf, 1984, Corporate financing and investment decisions when firms have information that investors do not have, Journal of Financial Economics 13, 187-222. Opler, Tim, and Jonathan Sokobin, 1997, Does coordinated institut ional activism work? An analysis of the activities of the council of institutional investors, Working paper, Ohio State University. Parrino, Robert, Richard W. Sias, and Laur a T. Starks, 2003, Votin g with their feet: Institutional ownership change s around forced CEO turnover, Journal of Financial Economics 68, 3-46.

PAGE 127

118 Perry, Tod, 1999, Incentive compensation for outside directors and CEO turnover, Working paper, Arizona State University, Tempe. Roll, Richard, 1986, The hubris hypot hesis of corporate takeovers, Journal of Business 59, 197-216. Rosenstein, Stuart, and Jeffrey Wyatt, Ou tside directors, board independence, and shareholder wealth, Journal of Financial Economics 26, 175-192. Ryan, Harvey, and Roy A. Wiggins, 2003, Who is in whose pocket? Director compensation, board independence, and barriers to effective monitoring, Journal of Financial Economics, Forthcoming. Ryngaert, Michael D., 1988, The effect of pois on pill securities on shareholder wealth, Journal of Financial Economics 20, 476-418. Safieddine, Asem, and Sheridan Titman, 1999, Leverage and corporate performance: Evidence from unsuccessful takeovers, Journal of Finance 54, 547-580. Schwert, William G., 2000, Hostility in ta keovers: In the eyes of the beholder? Journal of Finance 55, 2599-2640. Servaes, Henry, 1994, Do takeover targets overinvest? Review of Financial Studies 7, 253-278. Shivdasani, Anil, 1993, Board composition, owne rship structure, and hostile takeovers, Journal of Accounting and Economics 16, 167-198. Shleifer, Andrei, and Robert W. Vishny, 1986, Large shareholde rs and corporate control, Journal of Political Economy 94, 461-488. Shleifer, Andrei, and Robert W. Vishny, 1988, Value maximization and the acquisition process, Journal of Economic Perspectives 2, 7-20. Travlos, Nickolaos G., 1987, Co rporate takeover bids, methods of payment, and bidding firms' stock returns, Journal of Finance 42, 943-963. Weisbach, Michael S., 1988, Outside directors and CEO turnover, Journal of Financial Economics 20, 431-460. Wulf, Julie, 2003, Do CEOs trade power for premium? Evidence from mergers of equals, Working paper, University of Pennsylvania, Philadelphia. Yermack, David, 1996, Higher market valuation of companies with a small board of directors, Journal of Financial Economics 40, 185-211.

PAGE 128

119 Yermack, David, 2003, Remuneration, retention, and reputation incentives for outside directors, Journal of Financial Economics, Forthcoming. Zhao, Mengxin, 2002, Acquisition decisions and CEO turnove r: Do bad bidders get fired? Working paper, University of P ittsburgh, Pittsburgh.

PAGE 129

120 BIOGRAPHICAL SKETCH Ralph M. Scholten received a B.S. in mathematics in December 1996 and M.S. in economics in December 1998 from the University of Texas at El Paso He arrived at the University of Florida to pursue his Ph.D. in finance in August 1999. His research focuses on corporate finance, with particular inte rests in mergers and acquisitions, corporate governance, and valuation. He currently works for Cornerstone Research in Washington, D.C.


Permanent Link: http://ufdc.ufl.edu/UFE0004962/00001

Material Information

Title: The Impact of the Changes in the Takeover Market on Managerial Entrenchment
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0004962:00001

Permanent Link: http://ufdc.ufl.edu/UFE0004962/00001

Material Information

Title: The Impact of the Changes in the Takeover Market on Managerial Entrenchment
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0004962:00001


This item has the following downloads:


Full Text











THE IMPACT OF THE CHANGES INT THE TAKEOVER MARKET
ON MANAGERIAL ENTRENCHMENT














By

RALPH M. SCHOLTEN


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2004





































Dedicated to Kristi















ACKNOWLEDGMENTS

I wish to thank all the faculty members in the Department of Finance at the

University of Florida for a great five years. In particular, I wish to express my gratitude to

Dave Brown, Mahendraraj ah Nimalendran, and Chris James for their helpful discussions.

I am also fortunate to have Mark Rush from the Department of Economics serve on my

committee. Finally, I am especially grateful to Mike Ryngaert for his continuous support

and encouragement throughout my years at Florida.



















TABLE OF CONTENTS


ACKNOWLEDGMENT S .............. .................... iii


LIST OF TABLES......................... .......vi


LIST OF FIGURES ................. ..............vii..............


AB STRACT ....._.. ................ ......._._ ..........vi



CHAPTER


1 INTRODUCTION ................. ...............1.......... ......


2 EVOLUTION OF CORPORATE GOVERNANCE MECHANISMS ........................5


2. 1 The Corporate Takeover Market. ................. ............. ...............5.....
2.2 Extemnal Pressure from Institutions and Shareholders ................. .................10O
2.2. 1 Role of Institutional Investors................... ...............................10
2.2.2 Shareholder Activism ................. ...............12........... ....
2.3 Improved Intemnal Controls............... ...............14
2.3.1 Board Size and Composition .............. ...............14....
2.3.2 Director Incentive Compensation .................. ............... ................ ..16
2.4 Interaction between Internal and Extemnal Control Markets ................... .........18



3 INVESTMENT DECISIONS AND MANAGERIAL DISCIPLINE: EVIDENCE
FROM THE TAKEOVER MARKET .....__.....___ ..........._ ............2


3.1 Motives and Wealth Effects of Takeovers ......____ ........_ ................28
3.2 Data and Methodology ............ ............ ...............29..
3.2. 1 Sample of Acquisitions ............... ... ......... .............._ ...........2
3.2.2 Disciplinary Actions by External Control Market ............... .................31
3.2.3 Disciplinary Action by the Board .............. .. ...............34...
3.2.4 Measuring the Wealth Effects of the Acquisition ................. ................3 6
3.3 R esults.................. .. .............3
3.3.1 Univariate Results ................. .. .......... ...............37....
3.3.2 Logistic Regression of Managerial Tumnover ................. ................ ..42
3.4 A closer Look at the Determinants of Forced Tumnover ................. ...............44
3.5 Concluding Remarks............... ...............48












4 IMPLICATIONS OF DEFEATING TAKEOVER BIDS: EVIDENCE PRE AND
POST TIME-WARNER ................. ...............63.................


4.1 Previous Literature and Hypothesis Development .............. ...... .................6
4.1.1 Managerial Entrenchment, Takeover Bids, and Agency Theory............66
4. 1.2 Hostile Takeovers and Target Defenses .............. ......... ................6
4. 1.3 Hypotheses Regarding Defeated Takeover Bids in 1980s and 1990s ....70
4.2 Data Description and Methodology ............ ...... .__ ... .._............7
4.2.1 Sample of Defeated Takeover Bids .............. ...............71....
4.2.2 Target Characteristics .............. ...............76....
4.2.3 Bidder Characteristics ......_......._.__... .....__ ............7
4.3 Takeover Defenses and Leverage ......_......._.__..... ..._._..........8
4.3.1 Takeover Defenses ........._... ...... ...............80.
4.3.2 Leverage as a Takeover Defense .............. ...............83....
4.4 Target Performance ........._.._..... .__ ...............84...
4.4.1 Event Study Analysis............... ...............84
4.4.2 Post-Offer Valuation ......_......._.__... .....__ ............8
4.4.3 M ultivariate Analysis................. .... .... .. .......8
4.5 Post-Offer consequences to Firms and Management............... ...............9
4.5.1 Target Independence ................. ...............94...............
4.5.2 M anagerial Turnover .............. ...............95....
4.6 Concluding Remarks............... ...............97


5 CONCLUSION ................. ...............112................


LIST OF REFERENCES ................. ...............114................


BIOGRAPHICAL SKETCH ................. ...............120......... ......


















LIST OF TABLES


Table Page


3.1 Description of the sample ................. ...............50...._.__ ....

3.2 Frequency and type of disciplinary events............... ...............5..1

3.3 Abnormal stock market performance of acquiring firms............_._ .........._._ ....52

3.4 Differences of abnormal returns of acquiring firms .............. .....................5

3.5 Frequency and type of control-reducing event for good versus bad acquirers .........56

3.6 Pooled logit regression on CEO turnover and forced CEO turnover..............._._.....58

3.7 Ownership and board characteristics of the 1992-1998 sample .............. ................59

3.8 Pooled logit regression on CEO turnover and forced CEO turnover for the 1992-
1998 sam ple .............. ...............60....

4.1 Sample selection .............. ...............99....

4.2 Sample of defeated hostile deals ...._. ......_._._ .......__. ...........0

4.3 Target characteristics .............. ...............101....

4.4 Bidder characteristics ........._._ ...... .... ...............102.

4.5 Takeover defenses............... ...............10

4.6 Target leverage changes............... ...............104

4.7 Event returns ........._._ ...... .__ ...............105...

4.8 Post-acqui siti on valuati on ...._.. ................. ...............106 ...

4.9 OLS regression on Revall and Reval2 ........._.._.. ......._ ......._.........10

4.10 OLS regressions on Discount1 and Discount2 .............. ...............108....

4. 11 Distribution of delisted targets ........._._. ...._... .. ...............109.

4. 12 CEO turnover statistics and logit regression on CEO turnover .........._.................110

















LIST OF FIGURES


Figure Page

2.1 Volume of takeover activity of U. S. firms scaled by GDP .............. ................. 20

2.2 Proportion of takeover offers that are hostile ................. .............................21

2.3 Proportion of tender offers rej ected by targets ................. ............................22

2.4 Dollar volume of junk bonds issued each year .............. ...............23....

2.5 Proportion of U. S. equity held by institutional investors ................. ................ ...24

3.1 Stock price reaction to acquisition announcements for firms that subsequently
become a hostile or friendly target ................. ...............61...............

3.2 Stock price reaction to acquisition announcements for firms that subsequently
experience CEO turnover ................. ...............62................

4.1 Target leverage changes surrounding unsuccessful hostile takeover attempts ......111
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

THE IMPACT OF THE CHANGES INT THE TAKEOVER MARKET
ON MANAGERIAL ENTRENCHMENT

By

Ralph M. Scholten

August 2004

Chair: Michael Ryngaert
Maj or Department: Finance, Insurance, and Real Estate

It is often argued that the takeover market is the only effective corporate control

mechanism that penalizes underperforming managers. Over the past two decades, the

takeover market has experienced substantial changes and moved from an active hostile

environment in the late 1980s to a more friendly and less contentious environment in the

1990s. Some researchers have argued that managers have become more entrenched as a

result of these changes. The goal of this dissertation is to examine the implications of the

changes in the takeover market and the legal environment on managerial entrenchment.

Chapter Three focuses on the relative importance of board monitoring and the

hostile takeover market in disciplining managers that make poor acquisition decisions.

The results show that there is little evidence that takeovers that are poorly received by

capital markets result in a firm becoming the target of a hostile takeover. However, a

forced turnover of a top executive is negatively related to the abnormal return associated

with an acquisition announcement. The relationship between forced turnover and










negative acquisition returns is stronger during periods of less intense hostile takeover

activity. Hence, it appears that being disciplined for making a poor acquisition is more a

function of internal discipline than the workings of the takeover market.

Chapter Four examines the claim that the Delaware Supreme Court decision of

the so called "just say no" takeover defense in the Time-Warner case in 1989 greatly

undercut the ability of hostile takeovers to discipline entrenched management teams. I

examine this claim by analyzing firms that defeated hostile takeovers in the pre and post

Time-Warner periods. I find that the availability of the just say no takeover defense in the

1990s significantly lessened the use of greenmail, standstill agreements, share

repurchases, special dividends and stock placements with friendly third parties. Despite

the changes in defensive tactics, there are little differences in target revaluations between

the 1980s and 1990s. The market price of targets after defeating an offer relative to the

offer price is virtually identical for both periods as are target abnormal returns over the

takeover contest period. Firms are also less likely to remain independent, and managers

more likely forced out in the two years after defeating a bid in the 1990s.

Collectively, the results do not support the argument that the changes in the

takeover market and the legal environment have exacerbated the problem of managerial

entrenchment. Instead, the results are largely consistent with the idea that alternative

control mechanisms such as monitoring by institutions and shareholders and

improvements in internal corporate governance in terms of smaller, more independent

boards, and director incentive compensation are at least partially replacing the need for an

active hostile takeover market.
















CHAPTER 1
INTTRODUCTION

The problem of so called managerial entrenchment in corporations is

characterized by two key conditions. The first condition is the presence of a management

team that is unwilling and/or unable to put in place the most value maximizing strategy

for shareholders. The second condition is the inability to remove that management team

from office. One popular argument is that the problem of entrenched management has

become more severe in recent times as a result of changes in takeover rules and

regulations in the late 1980s. Jensen (1993) argues that the corporate takeover market can

initiate changes in mismanaged firms within a reasonably short period of time, but that

boards are frequently too slow to effect a major change of flawed corporate policies.

Jensen maintains that there are no effective timely alternatives to the takeover market in

disciplining managers and their firms, and argues that few boards will initiate actions,

such as restructuring and replacing management, without explicit pressure from the

takeover market.

Jensen's argument would suggest that managerial entrenchment has increased

during the 1990s as corporate takeover defenses became stronger, tougher state takeover

rules predominantly favored targets, and junk bond Einancing for hostile takeovers

diminished. In fact, the level of hostile takeover activity dropped significantly in the

1990s relative to the 1980s. The decline in hostile takeover activity, however, has not

occurred in isolation. Alternative external and particularly internal governance










mechanisms have also seen dramatic changes over the past two decades. U.S. equity

ownership of institutional holders has increased steadily, laws and regulations regarding

shareholder meetings have been relaxed, boards have become smaller and more

independent, and outside board members increasingly receive much of their remuneration

in the form of incentive compensation. Hence, while the hostile takeover market has the

potential to limit managerial entrenchment, it is plausible that these developments have

offset the need for an active hostile takeover market.

The primary goal of this dissertation is to investigate the impact of the changes in

the takeover and legal environment on managerial entrenchment. I examine two key

managerial actions that are often associated with entrenchment. The first action is the

decision to make a large acquisition. The motives for acquisitions are often questionable,

and managers are often accused of "empire-building" as opposed to value-maximizing

when major acquisitions are announced. Specifically, I focus on the effectiveness of the

takeover market and the board of directors in disciplining managers that make poor

acquisition decisions. I compare the disciplinary events that acquirers face in the 1980s

relative to the 1990s. The results indicate that there is no relationship between the wealth

effects of acquisitions and the probability of becoming a takeover target in either period.

However, I find that managers are more likely to be replaced as a result of value-

destroying acquisitions in the 1990s. The evidence is consistent with the idea that boards

have become more responsive in replacing poorly performing managers despite the

decline in hostile takeover activity.

The second action I examine is the defeat of an unsolicited takeover offer.

Takeover bids are highly public and shares of unsuccessful takeover targets generally









trade at prices below the offer price. Management is often heavily criticized for not

accepting takeover offers and therefore failing to maximize shareholder wealth.

Arguably, it has become easier for managers to mount passive "just say no" takeover

defenses with the widely accepted poison pill defense in the 1990s. Consistent with this

belief, I find fewer alternative defensive tactics such as share repurchases,

recapitalizations, and greenmail are used in the 1990s relative to the 1980s. On the other

hand, the revaluation of the targets and the price gap between what is offered in the deal

and what the target trades for after the contest are very similar between the 1980s and

1990s. Furthermore, firms are slightly more likely to subsequently get acquired and

replace managers after defeating a hostile bid in the latter period. Hence, the results do

not support the theory that management teams that defeat hostile takeover bids can stray

further from value maximization in the 1990s relative to the 1980s.

Collectively, the results are consistent with the idea that boards can function

without the explicit pressure from the hostile takeover markets. More importantly, it

appears that managers have generally become less entrenched over time. The results

suggest that alternative governance mechanisms have largely substituted for declining

hostile (disciplinary) takeover markets. This has important implications for takeover

defenses and the nature of takeover activity in the future. It is possible that hostile

takeover activity as seen in the late 1980s will not return as alternative control

mechanisms become more important in disciplining firms that deviate from value-

maximization. If this is the case, then modern takeover defenses, in particular the poison

pill defense, will have less far reaching consequences than once thought.










This dissertation is outlined as follows. The second chapter describes in detail the

major changes in the takeover market and corporate governance mechanisms over the

past two decades. Section 2.1 describes the changes in the takeover environment between

1980 and 2000 and offers some explanations why hostile takeover activity declined post

1989. Section 2.2 discusses the role of institutional holders and large shareholders in the

monitoring process. Changes in the size and composition of boards and their impact on

valuation are discussed in section 2.3.1. Section 2.3.2 discusses the impact of increased

reliance on director equity based compensation on managerial entrenchment. Chapter 2

concludes with a discussion on the interaction of the internal and external control

mechanisms in Section 2.4. Chapter 3 and 4 empirically test the impact of the changes in

the corporate governance mechanisms on managerial entrenchment. Chapter 3 examines

the managerial decision to acquire another firm and focuses on the intertemporal

relationship between the wealth effects of acquisitions and managerial turnover. Chapter

4 compares the consequences of targets that successfully defeat unwanted acquisition

attempts in the 1980s and 1990s. Concluding remarks and implications are discussed in

Chapter 5.
















CHAPTER 2
EVOLUTION OF CORPORATE GOVERNANCE MECHANISMS

2.1 The Corporate Takeover Market

It has long been recognized that the separation of ownership and control in

modern firms can create a situation in which management may have incentives to stray

from value-maximization. There are several internal and external control mechanisms

designed to prevent, or at least reduce, these so called agency costs. One of these

mechanisms is the external takeover market. Manne (1965) is the first to recognize the

ability of the takeover market to act as a disciplinary mechanism. He argues that large

observable deviations between actual and potential stock prices could induce outside

parties to acquire the firm and operate it more efficiently. In other words,

underperforming firms will likely face pressure from the takeover market.

While the takeover market certainly has the potential to discipline

underperforming firms, it was relatively weak until the 1980s. Comment and Schwert

(1995) document that fewer than 0.5 percent of public firms received a takeover bid in

month in 1975 compared to 1.5 percent of firms in 1987 and 1988. In addition to the

increase in the number of takeover events in the 1980s, there was also an emergence of so

called "hostile" takeover activity. Hostile or unsolicited takeover offers were often made

by private firms and individuals such as Carl Icahn and T. Boone Pickens, with the

objective to replace management and/or restructure firm operations. Targets of









unsolicited takeover attempts often rejected those offers and tried to fend off bidders by

paying greenmail or by initiating a massive firm restructuring of their own.

The literature offers several explanations for the emergence of the hostile

takeover market. One explanation offered by Jensen (1993), is that the changes in the

corporate control markets were long overdue. He asserts that the hostility in the takeover

markets was in response to the failure of the internal control mechanisms (or the board of

directors) to put in place value-enhancing strategies. Changes in technology and

regulation had led to a large amount of excess capacity in many firms in the 1980s.

However, managers were often reluctant to restructure their operations and pay out free

cash flow to shareholders in the form of dividends. Instead, the practice of "empire-

building" continued despite the growing dissatisfaction of shareholders. Leverage

increases through leveraged acquisitions, buyouts, and stock buybacks successfully

eliminated free cash flow, by forcing firms to become more efficient in order to service

debt payments (Jensen, 1986).

It is not clear that excess capacity is the only factor driving the takeover market in

the 1980s as Jensen (1993) suggests. If firms with excess capacity spend too much on

capital expenditures, we should expect to find that these companies spend less after being

acquired. The empirical evidence is mixed. Kaplan (1989) reports that management

buyout firms made large cuts in capital expenditures. Servaes (1994), however, does not

find evidence that targets are overinvesting prior to being acquired. Moreover, Healy,

Palepu, and Ruback (1992) report that the pre and post acquisition ratio of capital

expenditures to sales do not significantly differ for targets in the 1980s.









Another possible explanation for the increasingly hostile takeover environment in

the 1980s was the dissatisfaction of the diversifying mergers that firms made just a

decade earlier. Shleifer and Vishny (1988) argue that "the takeover wave of the 1980s

was to a large extent a response to the disappointment with conglomerates." Companies

divested unrelated business to increase their corporate focus. Their view supports the idea

that conglomeration, at least in hindsight, was a mistake, and managers were too slow or

unwilling to realize this until the takeover market pressured them.

Regardless of the motives of the hostile takeover market, most early empirical

evidence shows that targets of hostile takeovers were underperforming their benchmarks.

Morck, Shleifer, and Vishny (1989) find that hostile targets are characterized by lower

growth and lower Tobin's Q compared to friendly targets. Mitchell and Lehn (1990)

document that hostile targets make acquisitions that, on average, reduce shareholder

value. Martin and McConnell (1991) provide evidence on the disciplinary motives of

takeovers by reporting that targets of successful tender offers experience abnormally high

managerial turnover after being acquired. Agrawal and Walkling (1994) argue that

takeover bids generate information about the firm that is used by the labor markets to

discipline managers. In particular, they document a negative relationship between the

target's performance prior to the takeover and subsequent managerial turnover. The last

two papers interpret their results as evidence for the disciplinary motive of many

takeovers .

More recent evidence, however, suggests that hostile and friendly targets may not

be so different after all. Agrawal and Jaffe (2004) extensively examine the stock price

and operating performance for targets of hostile takeover activity and fail to find









evidence that targets underperform their benchmarks. They document that firms that

receive hostile takeover attempts have similar operating performance as firms that receive

friendly bids. Similarly, Schwert (2000) finds that hostile and friendly targets cannot be

distinguished from each other in terms of their stock price and accounting performance.

He argues that hostile offers reflect a different bargaining strategy, where hostile offers

are disclosed earlier in the deal process.

Despite the conflicting evidence on the pre-takeover performance of hostile

targets, many academics agree that the hostile takeover market of the 1980s had positive

effects on firms and their valuation. Targets of hostile offers experienced on average

positive revaluations even if the attempt failed (Bradley, Desai, and Kim (1983), Jandik

and Makhija (2003)). Mitchell and Mulherin (1996) report that nearly half of all major

corporations received a takeover offer in the 1980s. As evidence for the effectiveness of

hostile takeovers, firms that were not successfully acquired frequently responded to

hostile pressure with a corporate restructuring.

After the active takeover period of the 1980s, the number of takeovers dropped

significantly in the early 1990s. Figure 2.1 illustrates the total acquisition volume scaled

by GDP over the 1980-2000 period. The figure shows that acquisition volume was over 4

percent of GDP during the peak of the hostile takeover market in 1988, and declined

significantly in 1990 to only 1.5 percent. After the 1990-1992 Recession, however, the

takeover market recovered and acquisition volume rebounded and soon exceeded levels

of the 1980s. In 2000, during the height of the internet bubble, acquisition volume scaled

by GDP reached an all-time high of over 15 percent.









While the takeover markets rebounded strongly after 1992, the nature of the

acquisitions changed drastically as the frequency of hostile takeover attempts declined

substantially. Figure 2.2 shows the proportion of all acquisitions of publicly traded firms

that are classified hostile or unsolicited from 1980 to 2001. Figure 2.3 illustrates the

proportion of tender offers which were opposed by management and hence perceived as

hostile. Both Eigures clearly illustrate the dramatic drop in the fraction of hostile attempts

after 1989. Consistent with the Eigures, Andrade, Mitchell and Stafford (2002) report that

the percentage of CRSP firms that received a hostile offer dropped from 14.3% in 1980-

1989 to 4.0% in 1990-1998.

The decline in hostile takeover activity coincided with two maj or events. First, the

size of the junk bond market declined significantly in 1990. Figure 2.4 shows the junk

bond market volume over 1980-2000. The pattern is similar to the volume of acquisitions

in Figure 2.1. After a sharp increase in the 1980s, the junk bond market dried up in 1990

largely as a result of the insider trading scandals involving Michael Milken (See e.g.

Livingston and Williams (2003)). Without access to the junk bond market, many would-

be hostile acquirers, such as private firms and investors, were unable to finance

acquisitions (Comment and Schwert (1995)).

Second, and perhaps more importantly, the regulatory environment surrounding

hostile takeovers changed. One court case in particular, the Time-Warner decision in July

1989, made it more difficult for acquirers to obtain control of a target firm without the

approval of management. The court favored Time, the target of a hostile offer by

Paramount Communications, and allowed Time to keep its poison pill in place. This court

decision set a precedent in that targets could protect themselves from unwanted acquirers










by adopting poison pills and consequently making hostile offers prohibitively expensive.

Figure 2.4 illustrates that, the junk bond market has recovered since the early 1990s, and

the volume of junk bond issuance has risen well above levels seen in the 1980s. At the

same time, however, the level of hostile takeover activity did not return suggesting that

the anti-takeover legislation has had a non-trivial effect on the takeover markets.

While hostile takeover activity declined, other corporate governance mechanisms

began to play a larger role in the 1990s. In particular, shareholders and institutional

investors became more active in monitoring management. Also, internal controls such as

board composition, size, and incentive compensation have received more attention. It is

plausible that a combination of these alternative monitoring and incentive mechanisms

partially substitute for the absence of hostile takeovers in resolving agency problems. The

next sections will describe some of the most important changes in corporate governance

over the past two decades.



2.2 External Pressure from Institutions and Shareholders

2.2.1 Role of Institutional Investors

Institutional equity ownership of U.S. firms has increased dramatically in recent

years. Gompers and Metrick (2000) report that large institutional shareholders almost

doubled their share of U.S. stock market ownership between 1980 and 1996. Figure 2.5

illustrates the increase in the proportion of U.S. corporate equity owned by insurance

companies, mutual funds, and private pension and government retirement funds between

1980 and 2002. Consistent with Gompers and Metrick (2000) the figure clearly shows

that aggregate ownership by these institutions increased from approximately 30 percent









in 1980 to over 50 percent in 2001. The biggest increase comes from mutual fund

holdings; while their holdings were small in 1980, their share of the U.S. equity market

rose to over 20 percent in 2001.

Due to the nature of the business, institutional investors are primarily concerned

with stock price performance. Of course, in place of monitoring, institutional investors

could opt to sell their shares of firms with below acceptable performance. In fact, Parrino,

Sias, and Starks (2003) document that the number of institutions and aggregate

institutional ownership decline in the year prior to forced CEO turnover, suggesting that

institutional selling occurs as a result of dissatisfaction. However, they also report that

many institutional investors refrain from selling, which may be partially explained by the

fact that many institutions have selling restrictions, in particular when they track one of

the broader indices.

Institutional investors generally hold relatively large equity stakes and hence have

greater incentives to take an active role in monitoring management than do small

atomistic shareholders, since the benefits they receive from monitoring are greater than

the costs they bear (Grossman and Hart (1980), Shleifer and Vishny (1986)). Also,

because of their large stakes, institutions can exert more influence on corporate decisions.

Consistent with this argument, Denis and Serrano (1996) find that large unaffiliated

blockholders, which are often institutional investors, play a large role in a board's

decision to replace management. They report that almost 60 percent of the CEO turnovers

in firms that successfully defeat a tender offer are preceded with the addition of outside

blockholders. In contrast, only a fourth of the firms that do not experience CEO turnovers










add a large unaffiliated blockholder in the two years subsequent to the unsuccessful

takeover contest.

It is natural to think that institutional investors are involved in improving other

corporate governance mechanisms of the firm as well. In fact, besides directly monitoring

the firm and its management, proposals to improve the internal control systems can have

tremendous value. Hartzell and Starks (2003) document that institutional ownership

concentration is positively related to pay-for-performance sensitivity of executive

compensation. Moreover, they find an inverse relationship between institutional holdings

and the level of compensation, even after controlling for firm size, investment

opportunities and performance.



2.2.2 Shareholder Activism

In addition to the emergence of institutional holders, shareholders have also

become more active in the 1990s as disclosure requirements increased and regulations

concerning shareholder meetings and communications were relaxed. Starting in 1992,

firms are required to include a graph of their firm's stock price performance relative to

industry peers in their annual proxy statement. While the timeline and peer firms to

which the firm is compared are at the discretion of management, the law has arguably

resulted in an increased focus on stock price performance.

At the same time in 1992, the Securities and Exchange Commission (SEC) also

lowered the costs of mounting proxy fight contests that challenged boards and

management teams. In the past, shareholders were required to file detailed proxy

statements with the SEC before communicating with more than ten other shareholders.









After the new rule change, shareholders can now freely communicate with each other at

any given time, provided that a copy of the discussion is sent to the SEC afterwards

(Holmstrom and Kaplan (2000)). This rule change has effectively lowered the cost of

coordinating shareholder actions and likely contributes to increased shareholder activism.

Evidence on increased shareholder activism is provided by Gillan and Starks

(2000). They document that public funds and shareholder groups initiated 616

shareholder proposals related to corporate governance during 1989-1994 compared to

only 60 proposals during 1983-1988. Karpoff (2001) summarizes the empirical evidence

on the effectiveness of shareholder activism on target firms. While there is some

controversy about the impact of shareholder targeting, he concludes that firms with high

institutional ownership and poor stock and earnings performance are more likely to be

targeted. He cautions, however, that shareholder activism does generally not lead to

subsequent changes in earnings or the likelihood of a change in control. The impact of

shareholder proposals targeting a firm's governance policies on managerial turnover is

also mixed. While Huson (1997) reports an increased rate of CEO and board turnover

during the three years following shareholder targeting, Opler and Sokobin (1997) report a

decrease in the rate of CEO turnover among their sample of firms that were the target of

shareholder activism. In conclusion, it appears that shareholder proposals and private

negotiations have made at best minor changes in firms' governance rules, and are more

effective when combined with other governance mechanisms.










2.3 Improved Internal Controls

2.3.1 Board Size and Composition

The goal of the board of directors is to reduce the agency costs between

shareholders and management. However, their fiduciary duties of monitoring and

offering advice to management are often a point of controversy. Even though

shareholders ultimately elect board members using proxy voting once a year,

management and the incumbent board nominate the slate of directors. This process

potentially causes serious governance issues when entrenched managers elect insiders

and others close to management that are unlikely to be effective monitors.l

Several studies have examined the link between board characteristics and firm

valuation and decisions regarding managerial turnover. Some researchers have argued

that board size plays an important role in the decision making process. Jensen (1993), for

example, argues that boards are less likely to function effectively once they consist of

more than 7 or 8 directors. In addition, he claims that CEOs more easily control larger

boards since each board member is less likely to question leadership. There is some

empirical evidence that larger boards are less efficient. Yermack (1996) finds an inverse

relationship between board size and Tobin's Q. He also documents that smaller boards

are more likely to dismiss CEOs for poor performance, and that the threat of dismissal

declines as board size increases. Kini, Kracaw, and Mian (1995) show that board size is

reduced after successful tender offers for underperforming firms, which suggests that the

takeover market helps improve internal control mechanisms.




SThe nomination process for directors is in the process to undergo major changes. The Securities and
Exchange Commission has proposed a set of rules that would allow long-term shareholders with a
minimum stake of one percent to nominate one director to the slate of directors.









The maj ority of research in this area has focused on the composition of the board

of directors. It is widely believed that outside directors, usually defined as directors with

no immediate ties to management or the firm, are better monitors than inside directors.

The empirical evidence is largely consistent with this argument. Since the hiring and

firing of CEOs is the one of the primary functions of the board of directors, several

papers examine the impact of board composition on the relationship between managerial

turnover and performance to evaluate the quality of the board. Weisbach (1988)

documents that the link between managerial turnover and performance is stronger for

firms with outsider-dominated boards. Borokhovich, Parrino, and Trapani (1996) find

that outsider-dominated boards are more likely to consider a broader range of CEO

replacements. In addition, they find that the stock market reaction is significantly higher

for outsider appointments compared to insider appointments and conclude that their

evidence suggests that outsider-dominated boards are more beneficial to shareholders

when changes in firm policies are important.

Other papers focus on the relationship between board composition and firm

valuation. Evidence by Byrd and Hickman (1992) on bidders and Cotter, Shivdasani, and

Zenner (1997) on targets suggests that independent boards increase value for their

shareholders during an acquisition attempt. Brickley, Coles, and Terry (1994) find that

markets only react positively to poison pill adoptions for outsider-dominated boards.

Their finding implies that investors anticipate independent boards to select takeover

defenses with shareholder valuation in mind. Finally, markets generally perceive

independent directors to be valuable, as the addition of independent directors to the board

is associated with a positive stock price reaction (Rosenstein and Wyatt (1990)).









The results of these papers generally suggest that smaller and more independent

boards function as better monitors. The implication is that firms (and regulators) have

made significant steps to reduce board size and include more independent directors over

the last few decades (Hermalin and Weisbach (2003), Borokhovich, Parrino, and Trapani

(1996)). Hence, if smaller and more independent boards are better monitors, and boards

have become smaller and more independent over time, then there is reason to believe

that, on average, boards have become more effective monitors independent of pressure

from the takeover market.



2.3.2 Director Incentive Compensation

The primary function of the board of directors is to represent shareholder

interests. Legal and structural mechanisms are designed to ensure that outside board

members act independently of management. In the past, however, the judicial system has

been slow to act to discipline directors for violating their fiduciary responsibilities

towards shareholders. Perhaps a greater incentive for directors to act in shareholder

interest is the labor market for directors. Fama and Jensen (1983) argue that outside

directors have incentives to signal to the labor market that they add value through their

insights and monitoring. There are several papers that document the relationships

between firm performance and number of directorships. Kaplan and Reishaus (1990)

document that dividend decreases have implications in the director labor market. They

find that executives of firms that cut dividends are less likely to receive additional

directorships than are other executives. Kini, Kracaw, and Mian (1995) show that boards

of the combined firm tend to be constructed in a more balanced manner than the target









board following the acquisition (e.g. insider-dominated boards are replaced by boards

with more outside seats).

The primary source of director compensation is in the form of a cash-retainer.

Brickley, Coles, and Linck (1999) document for a sample of large firms, that

directorships can be very lucrative with the average director earning almost $45,000 in

annual compensation. In addition, they find that directors often obtain profitable

consulting opportunities with the firm that pay anywhere from $20,000 to $83,333 per

month. The problem with cash-retainers is that they can distort the motives for directors,

especially in the event of a takeover offer. Harford (2003) uses a sample of firms that are

the target of successful tender offers to show that board positions are difficult to replace,

and concludes that there is a large Einancial cost for failing to monitor management.

To provide better incentives, independent boards have increasingly received a

larger portion of their remuneration in the form of equity-based compensation. Perry

(1999) documents that the fraction of compensation for directors that is incentive-based

increased from 25% in 1992 to 39% in 1995. Ryan and Wiggins (2003) report over a

58% increase of equity based compensation for a sample of 600 firms between 1995 and

1997. They also find that firms with independent boards and "weaker" CEOs are more

likely to pay directors incentive compensation. Yermack (2003) finds that total outside

director remuneration is very sensitive to firm performance with a change in wealth of

about $285,000 for a one standard deviation change in firm performance for the largest

publicly traded firms.

Incentive compensation for independent directors has a positive impact on board

monitoring and as a result on firm performance. Perry (1999) documents an increased










probability of CEO turnover following poor performance for firms where independent

boards receive incentive compensation. Becher, Campbell, and Frye (2004) show that

equity-based compensation for bank directors has increased drastically after the

deregulation in the early 1990s. In addition, they find a positive relationship between

bank performance and the level of equity-based compensation for directors. They

conclude that incentive compensation increases the effectiveness of monitoring.



2.4 Interaction between Internal and External Control Markets

Because of the decline in hostile takeover activity in the early 1990s, there has

arguably been a shift to rely on other control mechanisms to step up monitoring.

However, Jensen (1993) argues that alternatives to the market for corporate control are

ineffective. In particular, he argues that boards have not adequately fulfilled their

responsibilities of hiring, firing, and properly compensate CEOs. In his view, few boards

have initiated necessary changes concerning top management in the absence of 'external

crises'.

The empirical evidence on the impact of the takeover market on the effectiveness

of board monitoring is mixed. Hadlock and Lumer (1997) and Mikkelson and Partch

(1997) provide support for Jensen's assertion, and find that takeover activity is necessary

for the board of directors to apply effective disciplinary pressure. Hadlock and Lumer

document lower than normal turnover rates during the 1933-1941 period that was also

characterized by few corporate control transactions. Using more recent data, Mikkelson

and Partch report that as the takeover market became less active in the early 1990s, the

link between managerial turnover and poor performance weakened. They find that










management turnover in unacquired firms is substantially higher during the active

takeover period of 1984-1988 compared to the inactive period of 1989-1993. Moreover,

they argue that the change in the takeover regime, from hostile and contested to more

friendly takeovers, may result in less managerial discipline among poor performing firms

in the future. These studies suggest that when the threat of a hostile takeover diminishes,

boards face less external pressure to replace underperforming managers.

In contrast, Denis and Kruse (2000) and Huson, Parrino, and Starks (2001) find

evidence that boards of directors and the takeover market are substitute mechanisms. In

their view, the effectiveness of the board of directors is independent of the takeover

market. Denis and Kruse (2000) report that while the number of disciplinary events

underperforming firms face during the inactive takeover period declines, the frequency of

internally precipitated CEO turnovers is actually higher. Huson et al. (2001) document

that the number of forced resignations and outside successions increases between 1971

and 1994. More importantly, they find no evidence that the sensitivity of forced CEO

turnover to performance varies with the intensity of the takeover market over that period.

Hence, there is no consensus on the effect of the takeover market on board

monitoring and managerial entrenchment. It is important to realize that the studies above

use data from the early 1990s. Since substantial changes have occurred during the latter

part of the 1990s, a reexamination of the impact of the changes in the takeover market on

managerial entrenchment is valuable. In particular, the changes in the legal environment

may have permanently altered the nature of the takeover markets. The next two chapters

examine whether the changes in the takeover market are at least partially offset by the

changes in the other corporate governance mechanisms.














18 -


16 -


14 -


12 -


10 -






6-


4-


2




1980


1982 1984 1986 1988 1990 1992 1994


1996 1998 2000


Year



Figure 2.1
Volume of takeover activity of U.S. firms scaled by GDP


Sources: Data on takeover activity are from Mergerstat (various issues). GDP data are from the U.S.
Department of Conunerce: Bureau of Economic Analysis.














30



25



20






~10












1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
Year



Figure 2.2
Proportion of takeovers that are hostile

Source: Thomson Financial SDC database. The selection criteria are that both acquirer and target are
publicly traded, the deal value exceeds $50 million, and the deal attitude is either friendly, unsolicited, or
hostile. Unsolicited deals are also considered hostile.














45


40








20





15


10






1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
Year



Figure 2.3
Proportion of tender offers rejected by targets

Source: Mergerstat (various issues)




















































Source: Table 2 from Livingston, Miles, and Glenn Williams, 2003, Drexel Burnham Lambert,
underwriting fees, and market power, Working paper, University of Florida.


45000


40000


35000


30000


25000


20000


15000


10000


5000


0
1


980


1982 1984 1986 1988 1990 1992 1994 1996 1998


2000


Figure 2.4
Dollar volume of junk bonds issued each year















60




50
Institutional Investors


40




~30
Private Pension and Government Retirement





Mutual Funds





10



1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000

Year



Figure 2.5
Proportion of U.S. equity held by institutional investors

Source: Federal Reserve Board, Flow of Funds Accounts of the United States,
htll tp u sal.federalreserve~gov/releases/Z1.
















CHAPTER 3
INVESTMENT DECISIONS AND MANAGERIAL DISCIPLINE:
EVIDENCE FROM THE TAKEOVER MARKET

Considerable research attention has been given to examining the effectiveness of

internal and external governance mechanisms in mitigating potential agency costs that

arise from the separation of ownership and control. Two of the most controversial

mechanisms that have received attention in both the popular press and academia are

hostile takeovers and the board of directors. While it is generally believed that hostile

takeovers are effective in disciplining at least some poorly performing firms and their

managers, the effectiveness of the board of directors is less clear. Jensen (1993)

comments, "It appears that internal control systems have two faults. They react too late,

and they take too long to effect a major change." In addition, he also argues that boards

rarely act without explicit disciplinary pressure from capital markets.

The empirical evidence on the impact of the takeover market on the effectiveness

of board monitoring is largely an unresolved issue. Hadlock and Lumer (1997) and

Mikkelson and Partch (1997) provide support for Jensen's assertion. They find that

managerial turnover related to poor performance is higher during periods of more

takeover activity. Both studies suggest that when the threat of a hostile takeover

diminishes, boards face less external pressure to replace underperforming managers. In

contrast, Denis and Kruse (2000) and Huson, Parrino, and Starks (2001) argue that the

takeover market and the board of directors are substitute mechanisms and that the

effectiveness of board monitoring is independent of takeover activity.










The goal of this chapter is to provide additional insights on the relative

importance of the takeover market and board of directors in disciplining poorly

performing managers during periods of high and low hostile takeover activity. More

specifically, this chapter addresses the question of whether an active hostile takeover

market is necessary for a well-functioning board. It is plausible that an active takeover

environment, in which underperforming firms are likely targets of hostile offers,

stimulates boards to act in the interest of shareholders by generating public information

about the target and its management or realigning incentives of outside directors.l It is

also plausible, however, that boards can discipline management independent of these

forces. Recent developments in corporate governance, such as shareholder activism,

greater focus on stock price performance due to increased institutional holdings, and

board incentive compensation, may create an environment where boards feel compelled

to discipline management absent a hostile takeover threat.

To investigate this question, I focus on a sample of 285 firms that announce and

complete major acquisitions during the active hostile takeover period from 1980-1987

and the inactive hostile takeover period from 1992-1998. I document the number of

disciplinary events each acquirer faces from the external control market in the form of

receiving a hostile or friendly takeover offer, and the internal control market in the form

of managerial turnover, for up to three years after the acquisition announcement. Similar

to Mitchell and Lehn (1990), I use the wealth effects of the acquisitions, measured by the

announcement returns, as a measure of managerial performance. Unlike other



SHarford (2003) reports that outside directors of target firms are rarely retained and that their financial
impact is typically negative following a completed offer. Hence, there is a cost to outside directors for
failing to actively monitor management, forcing the external control market to act for them. This cost is
arguably lower when the threat of a takeover diminishes.










performance measures such as long-term stock price performance or operating

performance that are influenced by macroeconomic factors and industry-dynamics,

acquisition decisions are publicly observable discretionary investments initiated by

management, and the success or failure of acquisitions is often attributed directly to the

CEO of the acquiring firm.

Consistent with Figures 2.2 and 2.3, I document that hostile takeover activity

sharply diminishes in the 1990s. I find no relationship between the wealth effects of

acquisitions and the likelihood of becoming a (hostile) target. However, deviating from

value-maximization has implications for management. I find a strong inverse relationship

between the abnormal returns around the acquisition announcement and forced

management turnover. Logit regressions indicate that this relationship only holds during

the period of less hostile takeover activity in the 1990s.2 In addition, there is no

overwhelming evidence that monitoring by active outside investors measurably facilitates

internal control efforts. Collectively, the results indicate that the internal mechanism does

to some degree discipline managers that make poor acquisition decisions, the hostile

takeover market appears to be less effective, and there is no enhancement effect

associated with the internal mechanism working better in an active hostile takeover

environment.

The remainder of the chapter is organized as follows. Section 3.1 discusses the

motives and wealth effects of takeovers. Section 3.2 describes the data and methodology.


2 In a contemporaneous paper, Zhao (2002) tests the relationship between wealth effects of acquisitions and
CEO turnover for firms that remain independent over a period of five years. She examines firms that make
acquisitions between 1990 and 1997, and also finds a negative relationship between announcement returns
and CEO turnover. In this chapter, I investigate the relationship between the wealth effects of acquisitions
and managerial discipline from both the external corporate control market and the board of directors. In
addition, unlike Zhao, my research design allows for making inferences about the effectiveness of board
monitoring during a changing takeover environment.










The relationship between the wealth effects of acquisitions and disciplinary actions by the

takeover market and board are presented in section 3.3. Section 3.4 contains a more

detailed examination of the forced turnover cases. Conclusions and implications are

summarized in section 3.5.



3.1 Motives and Wealth Effects of Takeovers

A vast literature has examined the wealth effects of takeovers. Studies generally

find that the excess returns to target shareholders are positive, while the excess returns to

shareholders of acquiring firms are either negative or not statistically different from zero

(see e.g. Bradley et. al. (1988), Schwert (2000), and Mulherin and Boone (2000)). Since

the bidding firm determines the timing and initial offer conditions, much of the research

in this area has focused on explaining why acquirers fail to gain in acquisitions.

There are several competing theories that attempt to explain the dismal returns

most bidders experience around the merger announcement. The first is based on

asymmetric information between management and the capital market. Following the

Myers and Majluf (1984) argument, firms tend to issue equity when management

believes the current stock price is overvalued, signaling this private information to the

market. Empirical evidence supports the asymmetric information theory, and shows that

abnormal returns around the announcement date are negative for stock offers and

generally insignificant for cash offers (Travlos (1987) and Martin (1996)). Loughran and

Vijh (1997) show that this relationship persists during the post-acquisition period.

The method of payment, however, does not fully explain the variation in returns

for stock acquirers, or why in some instances the market reacts unfavorably to cash









offers. Other theories argue that disappointing reactions to acquisition announcements are

not the result of the form of financing, but rather caused by non-value-maximizing

behavior by management in the form of overpaying for the acquisition.

One explanation is based on an empire building/rational agency cost theory that

assumes it is impossible, or at least very costly, to write contracts that perfectly align the

incentives of managers with shareholders. Jensen (1986) and Shleifer and Vishny (1988)

argue that managers overpay for targets, not because managers make valuation errors, but

to reap personal benefits from acquisitions that are non-value maximizing to the

acquiring shareholders. An alternative explanation for overpayment in acquisitions is

managerial optimism. Roll (1986), and more recently Heaton (2002), argue that managers

suffer from 'hubris' and are too optimistic about takeover prospects. In their view,

managers try to maximize firm value, but overestimate the value of targets and as a result

overpay. Regardless of whether overpayment is the result of self-dealing or hubris, the

interesting question is whether and which control systems step in and discipline

management for deviating from value-maximization.



3.2 Data and Methodology

3.2.1 Sample of Acquisitions

The initial sample consists of acquisitions that are announced between 1980 to

1987 and 1992 to 1998. The sample periods are constructed to examine the effects of

changes in the intensity of the hostile takeover market on internal monitoring. Data on

takeovers are obtained from Thomson Financial Securities Data Company and satisfy the

following criteria. (1) For each transaction, both bidder and target trade on the NYSE,










AMEX or NASDAQ, with data available on CRSP around the merger announcement. (2)

Minimum transaction value of the deal is $100 million for the 1980-1987 period and

$300 million for the 1992-1998 period.3 (3) Transactions in which either party operates in

a regulated industry are excluded (banks, financial, railroads, and utilities). (4) The

acquirer gains complete control of the target. (5) The ratio of market values of target and

bidder is between 10 and 70 percent. The market values of equity are computed four

weeks prior to the initial announcement, reducing the impact of information leakage on

the stock price. The selection range for the target-to-bidder market value ratio guarantees

that acquisitions will determine a significant part of the acquirer' s future cash flows, but

yet precludes a large number of 'mergers of equals'. Mergers of equals are excluded

because they generally experience significant changes in management and board

str-ucture, and make CEO succession more difficult to interpret.4 This selection process

potentially excludes other 'major' acquisitions that the firm completed during the same

time period, which may have had a greater impact on the firm. To address this issue,

acquisitions with relative size greater than 70 percent are included only in cases where

the bidder makes multiple acquisitions in a three year period.

Applying these criteria results in a sample of 334 acquisitions. Nine mergers that

exhibit merger of equals characteristics (equal board representation), and five acquirers

that experience CEO turnover in the month surrounding the takeover announcement are

excluded. The final sample consists of 319 acquisitions; 137 during the 1980-1987 period

and 182 acquisitions during the 1992-1998 period. Since some firms made multiple


3 This chapter focuses on the largest acquisitions during the two subperiods. The rationale for choosing a
three-to-one ratio of the deal values is to reflect the increasing firm size over time. The value-weighted
CRSP return is 317% over the period between the midpoints of the two samples. The return of the S&P
500 is 233% over the same period.
4 See Wulf (2003) for examples of governance changes in mergers of equals.










acquisitions in a three year period, there are 124 and 161 firms in the 1980-1987 and

1992-1998 periods, respectively.'

Table 3.1 reports the number of acquisitions and unique acquirers for each year.

The acquisitions are evenly distributed for the 1980-1987 sample. On the other hand, the

1992-1998 sample experiences a surge in acquisitions between 1996 and 1998, with 70

percent of the acquisitions occurring during those three years. The table also reports the

median market values of the bidder and targets, the transaction value and the ratio of

market values of target to bidder. The market values of the bidder and target are

calculated 20 trading days before the acquisition. By construction, bidder, target and deal

values are significantly larger for the latter period.6



3.2.2 Disciplinary Actions by External Control Market

Since hostile takeovers are more likely to be disciplinary (Morck et al. (1989)), I

distinguish between hostile and friendly takeover attempts.' Similar to Mitchell and Lehn

(1990), I classify acquirers as hostile targets when they are targets of successful or

unsuccessful hostile tender offers, unsolicited and rejected offers, and proxy contests.

Friendly targets are firms that receive friendly merger or tender offer proposals, and

successful and unsuccessful management buyouts, and are not the target of any hostile

activity.




5 There are 28 firms that make multiple acquisitions: 23 firms make 2 acquisitions, 3 firms make 3
acquisitions, and 2 firms make 4 acquisitions within a three year period. One firm experiences a change of
CEOs between acquisitions, and both acquisitions are included as separate observations.
6 For the multivariate tests, firm size is adjusted by the CRSP value-weighted index.
SSchwert (2000) argues that hostile deals cannot be distinguished from friendly deals based on economic
terms. He argues that the offers perceived as hostile are at the beginning stages of the bargaining process
and are publicized earlier. To take this possibility into account, I also examine the disciplinary role of
friendly takeover offers.









Jensen (1993) argues that changes motivated by the corporate takeover markets

are frequently accomplished within one and a half to three years. In line with Jensen's

argument, I focus on the number of hostile and friendly takeover attempts that bidders

receive for up to three years following the acquisition. Table 3.2 reports that there are 44

acquirers (15%) that become either a hostile or friendly target. Across the two periods,

only 19 firms (7%) are subject to hostile or disciplinary takeover offers, and 25 (9%)

become friendly targets. The remaining 241 firms do not receive any type of takeover

offer.

More interesting are the differences in the number of hostile and friendly takeover

attempts between the two periods. In the first period, 11% of the acquirers received a

hostile takeover attempt within three years after the acquisition. In contrast, only 3% of

the firms in the 1990s received a hostile offer. The decline in the fraction of hostile

takeovers is significant at the one percent level, and is consistent with Eigure 2.1 and the

Endings of Andrade et al. (2002). To support the disciplinary motive of hostile takeover

offers, Denis and Kruse (2000) Eind a significant decline in disciplinary takeovers that

result in target managers losing their position from 1985-1988 to 1989-1992. Conversely,

there is a significant increase in the number of friendly attempts from the first to the latter

period, with 13 percent of the acquirers becoming a friendly target in 1992-1998,

compared to 3 percent in 1980-1987.

To put the frequency of takeover activity for the 1980-1987 sample in

perspective, I compare the number of takeover attempts to those reported by Mitchell and

Lehn (1990). One key difference between the two methodologies is the length of time

considered after the acquisition announcement. Mitchell and Lehn classify firms as










targets if they receive a takeover offer for up to seven years after the acquisition. They

report that 28 percent of acquirers received an offer for control, where 21 and 7 percent

received a hostile and friendly offer, respectively compared to 11 and 3 percent for my

sample.8 When I extend the post-acquisition period to five years, the frequency of

takeover offers in my sample is very similar. Within five years, 31 (25%) acquirers are

targets of takeover attempts, of which 24 (19%) are hostile, and 7 (6%) are friendly.

While there is not a "right" cut-off point, extending the post-acquisition period

beyond three years arguably reduces the "effectiveness" of the corporate control market.

Goodyear's acquisition of Celeron, which Mitchell and Lehn (1990) use as an example,

illustrates this point. At the announcement of the acquisition, Goodyear's stock price

dropped by about 10 percent, and subsequently declined even further. Goodyear,

however, did not receive a hostile offer for over three and half years after they made this

'bad' acquisition, questioning the effectiveness of the takeover market as a disciplinary

mechanism. Moreover, while Goodyear did start a massive restructuring program after

the hostile attempt, the CEO was never replaced as a result of his poor investment

deci sion.9










SIn Mitchell and Lehn (1990) a firm is classified as a takeover target when it receives a takeover offer
between 1982 and 1988. Their sample consists of acquisitions announced between 1982 and 1986. Hence,
the length of time between the acquisition and subsequent takeover offers ranges from two years for
acquisitions announced at the end of 1986 and up to seven years for acquisitions announced at the
beginning of 1982.
9 IH In Sample, 11 of the 19 hostile takeover targets remain independent. Only 3 firms that are the target of
an unsuccessful hostile attempt subsequently experience forced turnover. Hence, the majority of the CEOs
retain their position after an unsuccessful hostile attempt. Thus, even with an explicit takeover threat, it is
not clear that boards are much more likely to replace the CEO.










3.2.3 Disciplinary Actions by the Board

To examine the effectiveness of the board, I concentrate on board actions that lead

to CEO turnover. While boards have alternative mechanisms available to discipline

management, such as varying compensation and bonus structures, a change in

management is a less ambiguous measure of monitoring. For instance, Denis and Denis

(1995) find that forced resignations signal a turnaround and are associated with positive

market reactions and large improvements in performance.

There are two broad definitions of CEO turnover in the literature. The first

definition, used by Weisbach (1988) and Mikkelson and Partch (1997), does not

distinguish between disciplinary and routine turnover. These researchers claim that there

is no reliable way to classify the motive for managerial turnover. They argue that it is

especially difficult to detect forced turnovers, since both the board and the CEO have

incentives to downplay the succession as retirement.

An alternative definition of disciplinary CEO succession is proposed by Huson et

al. (2001). If the news reports indicate that the CEO is fired, forced from his position, or

leaves for unspecified reasons, then the turnover is classified as forced. In addition,

succession is classified as forced when the CEO is under age 60 and the WSJ articles

about the succession (1) do not report the reason for departure due to death, poor health,

or acceptance of another position, or (2) report that the CEO is retiring, but do not

announce the retirement at least 6 months in advance.

The analysis relies on both definitions of turnover to ensure that the findings are

robust to different definitions of turnover. In this chapter, CEO turnover refers to the

succession of the top manager, excluding successions that are the result of health reasons










and death. The method proposed by Huson et al. (2001) is used to classify succession as

'routine' (retirement, health reasons, other employment), and 'forced' (forced by board,

no reasons listed, performance related). Reasons for CEO turnover are obtained from the

Wall Street Journal, proxy statements, and 10K filings with the SEC. Firms that do not

experience CEO turnover are referred to as nonturnover firms. The nonturnover group

also contains firms that are acquired before replacing their CEO. In other words, being

acquired is not classified as either forced or routine turnover.

Table 3.2 also reports the frequency of CEO turnover and forced turnover. The

rate of CEO turnover is substantially higher than the takeover rate. Of the 285 firms in

the sample, 88 experience CEO turnover in the three years after the acquisition. The

annualized turnover rate is approximately 10%, which is similar to turnover rates

reported in previous studies (see, e.g. Huson et. al. (2001), Denis and Denis (1995)).10 Of

these 88 firms, 43 experience a forced change in top officer. About 20 percent experience

a disciplinary event, defined as experiencing non-routine turnover or receiving a hostile

offer. Despite the decline in hostile takeover activity from the first to the latter period, the

fraction of firms that were subject to disciplinary turnover increased. Of the 43 forced

turnovers, only 9 occurred during the 1980-1987 period. The remaining 34 cases took

place in the latter period. The increase in the fraction of disciplinary turnovers is

significant at the one percent level, and indicates that boards were more likely to act and

replace management during the latter period.







"' The annualized turnover rate is calculated as follows: 88 turnovers occur for 285 firms within three years.
Hence, the annual turnover rate is equal to (88/285)/3 = 10%.










3.2.4 Measuring the Wealth Effects of the Acquisition

Similar to Mitchell and Lehn (1990), I use the market reaction to a firm's

acquisition announcement as an indicator of managerial performance. In estimating the

wealth effects of acquisitions, I implicitly assume that the capital markets react efficiently

to the information of the acquisition announcement. In other words, capital markets are

able to distinguish between 'good' and 'bad' acquirers, and 'bad' acquirers experience

more negative stock price reactions at the announcement."

CRSP daily returns are used to compute market-adjusted abnormal returns for

each acquiring firm. The wealth effects of acquisitions are measured by the change in the

acquirer's equity value around the announcement date. Abnormal returns are estimated

over several windows, including the standard [-1,1] and [-5,5] estimation windows.12

Since this chapter focuses on large acquisitions where information about the deal is likely

to leak out before the announcement, the longer [-5,5] window is arguably a better

measure for the wealth effects of the acquisition.

Market-adjusted returns are calculated by differencing the compounded return of

the firm and the CRSP value-weighted market index return over the event windows. The

mean cumulative abnormal return is obtained by taking the average of the individual

abnormal returns. Market-adjusted returns are used instead of the market model because

acquirers, on average, outperform the market prior to the acquisition. This implies that

the intercept term for the market model is likely to be biased upward and over-estimates

11 Empirical evidence shows that the initial market reaction to acquisition announcements contains
information about the long-term consequences of a takeover to the bidding firm and its management.
Kaplan and Weisbach (1992), for example, find that acquirer returns and combined (acquirer and target)
returns at the acquisition announcement are significantly lower for divestitures that are performance
related, than for successful divestitures and acquisitions not divested. They conclude that stock market
prices correctly react to fundamentals during the announcement period, and that managers that ignore the
unfavorable signals from the capital markets are motivated by managerial hubris.
12 Qualitatively similar results are obtained for [-5,1] and [-5,40] estimation windows.










the negative impact of the acquisition announcement.13 Moreover, for narrow event

windows, abnormal returns do not differ substantially between the market and means

models (Brown and Wamner (1985)). For firms that make multiple acquisitions, wealth

effects are measured by the abnormal return averaged across all its acquisitions.14




3.3 Results

3.1.1 Univariate Results

Table 3.3 reports abnormal returns for the full sample and the two sub-periods.

Abnormal returns are estimated over the [-1,1] and [-5,5] windows and are reported in

panels A and B, respectively. The announcement returns corresponding to the

acquisitions made by the full sample are -2.3% and significantly different from zero at

the one percent level for both estimation windows. Of the 319 acquisitions, 103 are

financed using cash only, 154 are pure stock acquisitions, and 36 consist of a mix of cash,

stock, and debt. Consistent with previous findings (e.g. Travlos (1987)), announcement

returns are more negative whenever acquisitions are financed entirely or partially using

stock. The average [-1,1] abnormal return on stock acquisitions is -3.9%, compared to -

0.4% for cash acquisitions. The difference between the announcement returns of the cash

and stock acquisitions is statistically significant at the one percent level for both



13 The parameters of the market model can be estimated using the equation Rt = a, + P, Mt + Et where Rt is
the return of firm i, Mt is the return on the market portfolio, and a, and P, are the market model parameters.
The residuals ext can be used as an alternative measure of abnormal performance. Since acquirers, on
average, outperform the market, a is biased upward, over-estimating the negative impact of the merger
announcement (see Schwert (2000)). Evidence of this is provided by the market-adjusted Buy-and-Hold
abnormal returns (BHAR) one year prior to the acquisition. The mean BHAR is 10.05% with a p-value of
0.001.
14 The results reported in the chapter are based on the equally-weighted return of the acquisitions that the
acquirer makes. The results are robust to different specifications; using all acquisitions as separate
transactions or only the first or last acquisition that a firm makes does not affect the significance of the
results reported in the chapter.










estimation windows." The returns are comparable to Healy et al. (1992), for example,

who Eind an abnormal bidder return of -2.2% for the largest 50 transactions during 1979-

1984.

Table 3.3 also reports the average and median stock price reactions associated

with announcements of acquisitions made by nontargets, targets, hostile targets, and

friendly targets, as well as turnover, forced turnover, and nonturnover firms. The results

show that nontargets perform slightly better than targets. There is no evidence that

acquirers that subsequently become a hostile target perform significantly worse; hostile

targets, on average, experience a return of -2.7%, and -1.3% for the [-1,1] and [-5,5]

windows, respectively. Figure 3.1 graphs the stock price reactions for the targets, hostile

targets, friendly targets, and nontargets from fiye days prior to 40 days after the

acquisition announcement. The Eigure illustrates that while friendly targets perform worse

the few days directly following the announcement, the difference between the stock

market performance of hostile, friendly, and nontargets disappears over time.

The [-1,1] and [-5,5] announcement returns by firms in the CEO turnover

category are -3.1% and -4.1%, respectively. In contrast, firms in the nontumnover category

have announcement returns of -1.9% and -1.6%. Many turnovers are simply normal

retirements, or precipitated by factors other than poor performance, such as planned

succession or other employment opportunities. Therefore, if poor acquisition decisions

lead to removal of CEOs, the lowest returns are expected for the forced turnover group.

Consistent with this assertion, abnormal returns are substantially lower for this group

relative to any other category. The average returns for the forced turnover group are -

15 The proportions of turnovers for the cash and (partial) stock samples are 11/103 and 32/216 respectively.
A chi-square test indicates that there is no significant difference between the frequencies of turnovers
across the two samples, and should therefore not bias the qualitative results.










4.8% and -6.7% for the two estimation windows. The last two rows report the returns for

firms that receive disciplinary action from the hostile takeover market or experience

forced turnover. With returns of -4.3% and -5.4% for the [-1,1] and [-5,5] windows, the

returns for this group are considerably lower than for firms that do not experience a

disciplinary event. Figure 3.2 depicts the market reaction of the turnover, forced turnover,

and nonturnover categories for the [-5,40] event window. While all categories

underperform subsequent to the takeover announcement, it is clear from the graph that

firms that experience forced turnover do substantially worse. The difference in the

cumulative abnormal returns between the forced turnover and nontumnover samples is

approximately 6% after five days. More importantly, unlike the difference between

targets and nontargets, the return difference between forced turnovers and nonturnovers

stays practically constant over the entire [-5,40] window.

Panel A and B of Table 3.4 reports the difference in returns for nontargets versus

(1) targets, (2) hostile targets, and (3) friendly targets, for the [-1,1] and [-5,5] windows,

respectively. The difference in announcement returns between target and nontarget firms

is only significant at the ten percent level for the [-1,1] window for the full sample. For

any other estimation windows or sub-sample the difference is insignificant. Moreover,

there is little evidence that hostile targets do worse than nontarget firms. In fact, hostile

targets appear to outperform nontargets during acquisitions during the 1980-1987 period.

This result is inconsistent with the findings of Mitchell and Lehn (1990), who find that

takeover targets, especially hostile targets, have significantly more negative

announcement returns surrounding acquisition announcements than nontargets.16


16 The difference in results between my study and Mitchell and Lehn (1990) during the 1980s period is
likely the result of differences in the sample selection. This study focuses on acquirers of public targets










The evidence on managerial turnover is more compelling. The mean and median

return differences between nonturnover versus turnover and forced turnover firms are

significant for the full and the 1992-1998 sample. In panel B, for example, the average

difference in returns between turnover and nonturnover firms is -2.5% for the full sample.

That difference is -5.1% when comparing firms that experience forced turnover with

firms that experience no turnover. Interestingly, this difference is entirely driven by the

1992-1998 sample, which suggests that board scrutiny has increased, and CEOs are

dismissed quicker after costly investment mistakes. Hence, it appears that boards have

become more effective over time without reliance on external pressure from the hostile

takeover market.

To gain a better understanding of the relationship between a "bad" acquisition and

subsequent disciplinary actions from the internal and external control markets, the sample

is partitioned into quartiles based on the announcement returns. The type and frequency

of disciplinary actions are documented for firms in the top and bottom quartiles. For the

full sample, the top quartile includes firms that made acquisitions with an average return

greater than 1.6% and 2.7% for the [-1,1] and [-5,5] windows, respectively. The average

announcement return for firms in the bottom quartile is less than -6.1% and -7.4% for the









with a minimum deal value of $100 million. Their sample consists of a sample of firms that are covered by
Valueline and make acquisitions where the ratio of deal value and bidder market value exceeds 5%. While
they do not report the mean or median bidder values, their sample selection allows for many smaller
bidders. Shivdasani (1993) and Mitchell and Lehn (1990) show that the probability of becoming a takeover
target is inversely related to bidder size. Hence, large firms in the 1980s were more isolated from hostile
takeover pressure. The results that follow indicate that the internal governance mechanisms also fail to
penalize managers of large firms in the 1980s.










same windows." The Chi-squared statistic tests the hypothesis that the proportion of

firms experiencing each event is equal between the two samples.

Table 3.5 reports the fraction of firms that fall in the target, hostile target, friendly

target, turnover, forced turnover, and disciplined categories for both the top and bottom

quartiles. These events are not mutually exclusive since in a few instances CEO turnover

precedes a takeover attempt, and vice versa, an unsuccessful takeover attempt precedes

CEO turnover. The results show that there is no difference in the number of takeover or

hostile offers "bad" and "good" acquirers receive. Focusing on Panel B, 15 percent of the

acquirers in the top quartile become targets, while 13 percent of the acquirers in the

bottom quartile become targets. Contrary to what we would expect, 6 percent of the top

performers receive a hostile offer compared to 3 percent of the bottom performers.

The proportion of forced turnovers is significantly higher for firms in the lowest

quartile with 20 of 72 firms (28%) experiencing forced turnover, compared to 6 of 72

(8%) for firms in the top quartile. Another interesting result is the difference in the

proportion of firms that face a disciplinary action after making a poor acquisition

between the two time periods. Of the 3 1 acquirers with returns in the bottom quartile only

10% or 3 firms face disciplinary actions in the three years following the acquisition. In

contrast, of the 41 "bad" acquirers in the 1992-1998 sample, 44% or 18 firms are either

the target of a hostile takeover or experience forced turnover. The difference in the

fraction of firms that face disciplinary actions is significant at the one percent level. Even

more interesting is that all of the discipline in the later period comes from boards. None

of the acquirers in the bottom quartile become the target of a hostile offer. The results

17 The [-1,1] and [-5,5] announcement returns for top and bottom quartiles for the 1980-1987 sample are
0.4% and 2.7%, and -5.5% and -7.3%. The [-1,1] and [-5,5] announcement returns for top and bottom
quartiles for the 1992-1998 sample are 1.8% and 2.4%, and -6.8 and -7.4%.










indicate that "bad" acquisitions result in forced turnover more frequently only during the

1992-1998 period.



3.3.2 Logistic Regressions of Managerial Turnover

The results in the previous section indicate an inverse relationship between the

wealth effects of acquisition and subsequent disciplinary events from the board of

directors. To better understand the time variation in managerial turnover and its relation

to acquisition performance, I estimate logit regressions relating the probability of

managerial turnover and the abnormal return around the acquisition announcement,

bidder size, CEO age, pre-acquisition performance, a dummy variable equaling one if the

merger occurred during 1980-1987, as well as an interaction term measuring the

difference in the sensitivity of turnover to the wealth effects of the acquisition between

the two takeover regimes.ls

The abnormal return of the acquisition announcement is used as the measure of

the wealth effects of the acquisition. The Buy-and-hold market-adjusted return (BHAR),

calculated over the [-250,-10] window, is included as a measure of the firm's pre-

acquisition performance. To control for firm size in the regression, I include the

logarithm of the firm's market value calculated 20 days before the first announcement

scaled by the value-weighted CRSP index. A dummy variable (D80-87) is equal to one

when the acquisition took place during the 1980 to 1987 period, and zero otherwise. An


1sThere are 19 firms that are successfully acquired before any CEO turnover occurs. Including these firms
in the analysis is ambiguous for two reasons. First, since the firm exits the sample before its three-year
post-acquisition period, it is unclear whether the internal control systems would have acted if no takeover
had taken place. Second, information about the managers and their positions within the new firm are often
difficult to obtain. Moreover, it is unclear how to classify managers that are assigned a similar position with
the new firm. For these reasons, I exclude these 19 firms, and perform the regression analysis on a sample
of 266 firms.









interaction term (D80-87*AR) is included to allow for changing sensitivity of

disciplinary events to acquisition performance that may be induced due to the general

decline in hostile takeover activity in the 1990s relative to the 1980s.

The results of the logit regression for CEO turnover and forced turnover are

reported in Table 3.6. The negative coefficient estimates on the announcement returns

indicate a strong inverse relationship between the wealth effects of the acquisition and

managerial turnover. In other words, value-reducing behavior by management increases

the probability of forced turnover. The coefficient estimates on the period dummy are

negative and significant when forced turnover is the dependent variable. Hence,

controlling for age and the wealth effects of the acquisition, firms are more likely to

experience forced CEO succession in the latter period. This is consistent with Huson et

al. (2001), who find that the frequency of forced turnover has increased over time. The

most interesting result is the coefficients on the interaction variable. The coefficients on

the interaction dummy are insignificant for the [-1,1] window, but significantly positive

for the [-5,5] window. Hence, there is no evidence that the sensitivity of forced turnover

to acquisition wealth effects is lower as a result of the decline in hostile takeover activity.

In fact, for the [-5,5] window the results are consistent with internal control mechanisms

becoming more effective in quickly replacing management who make poor acquisition

decisions, despite the decline in hostile takeover activity.

I also estimate the implied probabilities of forced turnover for firms at the 10th,

25th, 50th, 75th, and 90th percentiles of the announcement returns for the sample of

acquirers. The implied probabilities are calculated using the logistic equations predicting

the likelihood of forced turnover during the three years after the acquisition









announcement. The probabilities are estimated holding the independent variables at the

sample mean. The probability of forced turnover for the [-5,5] announcement return at

the 10th percentile (-12.4%) is 20.3%, compared to the probability of forced turnover of

only 7.8% when the [-5,5] announcement return is at the 90th percentile (8.9%). Thus, the

probability of forced turnover almost triples and increases by over 12 percent as

announcement returns fall from the 90th to the 10th percentile.19

To gain additional insights about the negative relation between turnover and

announcement returns across the two sample periods, I estimate logit regressions for the

1980-1987 and 1992-1998 periods separately. The results (not reported) indicate that the

coefficient on the announcement return is negative and highly significant for the 1992-

1998 period. In contrast, the coefficient is not significant during the 1980-1987 period.

Since the hostile takeover market also fails to discipline managers, it appears that

managers are less likely to be penalized for making "bad" acquisitions during the earlier

period.



3.4 A Closer Look at the Determinants of Forced Turnover

So far, I have argued that the board of directors is the primary disciplinary

mechanism in forcing out top executives that make poor acquisition decisions. While the

threat of a hostile takeover sharply diminished during the 1990s, other forms of external

pressure may have played a more important role in influencing board decisions regarding

managerial turnover. Large shareholders or blockholders, in particular, have incentives to

exert considerable pressure to replace underperforming managers.

19 The implied probabilities for the 25 50 and 75 h,percentiles are 16.5%, 13.3%, and 10.6%
respectively. The implied probabilities at the 10 25 50 75 and 90" percentiles of the [-1,1]
announcement return window are 20.2%, 16.6%, 13.0%, 11.3%, and 9.3% respectively.









In order to investigate the role blockholders play in pressuring the board to force

management out, I restrict the analysis to the 1992-1998 sample. Excluding 14 firms that

are acquired before replacing the top executive, and one firms for which proxy material

or 10K reports before or within two months after the acquisition announcement is

missing results in a sample of 146 firms. Table 3.7 reports the means and medians of the

ownership and board characteristics for these 146 firms. CEO characteristics show that

the average top executive is 53 years old, owns 3.58 percent of the voting shares, and

holds the additional position of chairman of the board 87 percent of the time.

The table also reports board characteristics. On average, the board of directors

consists of 10 board members, with 70 percent of the directors classified as outsiders. A

director is considered an outsider when the individual is not currently, or has never been

employed by the firm, has no business relationships with the firm, and has no family ties

to management. A board is defined to be independent or outsider-dominated when at least

60 percent of the board members are outsiders. Using this definition, almost 73 percent

the firms have outsider-dominated boards. This number is substantially higher than in

Weisbach (1988), who documents that 35 percent of the boards are independent for a

sample of NYSE firms in 1980, and is consistent with improved regulations that move

towards more independent boards. The average shareholdings by officers and directors,

including the CEO, are 10 percent.

Unaffiliated blockholders are defined as shareholders that own at least five

percent of the outstanding voting stock, and do not have any type of agreement with the

firm. ESOPs and standstill agreements are considered affiliated block holdings, and are

therefore not included. The maj ority (76 percent) of the firms has at least one blockholder









at the time of the acquisition. The median number of blockholders is two, and they own

on average approximately 15% of the voting shares.

Comparing the governance variables for the forced turnover and non-forced

turnover samples reveals few differences. Interestingly, managers of firms that do not

experience forced turnover are, on average, about three years older. The reason is that the

control sample includes both firms where the managers retain their position, and firms

that experience planned succession through retirement. Otherwise, only mean block

holdings are significantly higher at the ten percent level for the firms in the control

sample. Jensen (1993) warned that CEOs that are also the chairman of the board entrench

themselves. The fraction of firms that have one individual who serves in a dual role as

CEO and chairman is 0.79 for the forced turnover sample and 0.89 for the control sample,

but is not significant at conventional levels.

To provide evidence on the role of blockholders in the decision to replace the top

executive, I re-estimate the logit model from the previous section. If blockholders are

active monitors of the firm, we would expect greater sensitivity of turnover to acquisition

returns for firms with blockholders. To allow for this possibility, I include an interaction

term (BLOCKHOLDER*AR), where BLOCKHOLDER is a dummy variable equal to

one when a blockholder is present at the time of the acquisition.

Previous research has found that board characteristics also play an important role

in the decision to replace management. Weisbach (1988) reports that the sensitivity of

turnover to performance increases with the independence of boards. To examine the

impact of the composition of the board on the relationship between turnover and

acquisition performance, I include a second interaction term (OUTSIDER*AR).









OUTSIDER is a dummy variable equal to one when at least 60 percent of the board

members are outsiders. To control for potential entrenchment effects of management

through shareholdings, the variable CEOHOLD is included, which is the proportion of

stock owned by the top executive immediately preceding the acquisition.

The results are reported in Table 3.8. Even after controlling for the various

ownership and board characteristics, the negative relationship between the wealth effect

of acquisitions and subsequent forced turnover is still highly significant. Ownership by

management is only significant when CEO turnover is the dependent variable. The

coefficient estimates on the interaction terms are not statistically significant. Hence, there

is no evidence that either unaff61iated block holdings or outsider-dominated boards affect

the sensitivity of forced turnover to the wealth effects of the acquisition.

One potential downside of the regression results in Table 3.8 is that the

governance variables are measured before and around the time of the acquisition.

However, CEOs may not be replaced until two or three years later. Denis and Serrano

(1996) argue that blockholders often appear and disappear between two successive

proxies. Searching news accounts around the date of management turnover, they find that

monitoring by active outside investors facilitates internal control efforts. To account for

this possibility, I search Factiva during the year before and three months after the forced

CEO succession. Out of 34 firms that force out their CEO, I am only able to identify Hyve

cases (15%) in which a large shareholder or activist group exerts pressure that

subsequently leads to replacement of the CEO. Hence, there is little direct evidence that

active blockholders are the main initiators of CEO replacements.









There are two interpretations for these Eindings. First, much of the external

pressure from institutional holders and large shareholders is communicated verbally and

not reported. Consequently, it is difficult to measure the extent and effects of shareholder

activity on board decisions regarding the issue of replacing management. A second, and

perhaps more satisfying interpretation, is that boards have become more responsive to

fulfiling their Eiduciary obligations in acting on behalf of shareholders. Incentives in the

form of equity-based compensation and the move towards smaller and more independent

boards are potentially contributing factors in the increased sensitivity of turnover to

performance.



3.5 Concluding Remarks

In the current capital market, where valuations are high and effective anti-

takeover devices are deterring potential raiders from making control-reducing bids, there

is an emphasized need to increase the role of the board of directors in monitoring

management. This chapter investigates the monitoring function of the board of directors

and its effectiveness during periods of high and low external pressure from the hostile

takeover market, by focusing on a sample of firms that make large acquisitions.

Contrary to Mitchell and Lehn (1990), there is little evidence for the disciplinary

role of hostile takeovers. Firms that become targets of a hostile takeover attempt in the

three years after their acquisition announcement do not appear to make acquisitions that

are significantly worse than those of nontarget firms. In addition, only a small fraction of

the firms (approximately 7%) receives a hostile offer within three years after making an

acquisition. Moreover, hostile offers are rarely successful and while they may induce









firms to make value-improvements, they seldom result in forced resignations of the top

executive. Therefore, given the small number of hostile attempts combined with the fact

that hostile targets do equally well during acquisitions seriously questions the

"effectiveness" of the external control market in disciplining management for making

value-reducing acquisitions in a timely manner.

Internal control mechanisms, however, are effective in disciplining what appear to

be poor acquisition decisions. The results indicate a strong negative relationship between

the abnormal returns around the acquisition announcement and subsequent forced

turnover of the top executive. Contrary to the implications of Mikkelson and Partch

(1997), the effectiveness of the internal control system headed by the board of directors is

stronger during periods of less intense hostile takeover activity. Improvements in

corporate governance in terms of increased shareholder involvement, greater focus on

stock price performance, and proper board incentives appear to provide viable

alternatives to the corporate control market. Collectively, the evidence in this chapter

indicates that being disciplined for making a poor acquisition is more a function of

internal discipline than the workings of the takeover market.







50


Table 3.1
Description of the sample

The table reports the number of acquisitions, number of unique bidders, median market values of the
bidder, target, deal, and relative size of target to bidder for each year. The market values are computed four
weeks prior to the initial announcement. The sample consists of acquisitions that satisfy the following
criteria: (1) For each transaction, both bidder and target trade on the NYSE, AMEX or NASDAQ, with data
available on CRSP. (2) Minimum transaction value is $100 million for acquisitions in the 1980-1987 period
and $300 million for acquisitions in the 1992-1998 period. (3) Transactions in which either party operates
in a regulated industry are excluded this excludes all banks, financial, railroads, and utilities. (4) The
acquirer gains complete control of the target. (5) The ratio of market values of target and bidder is between
10 and 70 percent. Acquisitions with relative size greater than 70 percent are included in cases where the
bidder makes additional acquisitions in the three years before or after the acquisition that satisfies criteria 1
through 4.

Value Transaction
Number of Number of Bidder Value Target Value MV Target/
Year acquisitions firms ($ mil) ($ mil) ($ mil) MV Bidder

1980 5 5 1652 442 483 0.20
1981 21 20 1523 249 331 0.19
1982 11 8 1194 388 555 0.35
1983 18 17 992 333 386 0.25
1984 16 16 860 255 332 0.41
1985 24 21 1599 389 568 0.37
1986 26 23 957 316 364 0.24
1987 16 14 634 176 251 0.37

1992 5 5 1953 328 529 0.29
1993 9 9 2371 631 842 0.:21
1994 18 17 1905 651 849 0.25
1995 23 23 2128 883 1459 0.31
1996 34 28 2801 509 1039 0.24
1997 49 43 1892 440 776 0.29
1998 44 36 3630 950 1623 0.30

1980-1987 137 124 999 270 361 0.28


1992-1998 182


161 2254 555


921 0.26







51



Table 3.2
Frequency and type of disciplinary event

The table reports the frequency of disciplinary control events acquiring firms experience in the three years
following the acquisition. Firms are classified as independent nontargett), hostile target, friendly target,
experiencing CEO turnover, or forced turnover. The total number of disciplinary events (disciplined),
defined as experiencing forced turnover or becoming a hostile target is also reported. A hostile target is a
firm that is the target of successful or unsuccessful hostile tender offers, unsolicited and rejected offers, and
proxy contests. A firm is a friendly target if the firm is the target of friendly merger or tender offer
proposals, or successful and unsuccessful management buyouts, and does not receive any hostile offers. A
firm is classified as a nontarget if it does not receive any hostile or friendly offers. A firm experiences CEO
turnover whenever a succession of the top officer occurs, excluding those that are the result of health
reasons and death. Forced turnover excludes top executive turnover labeled as normal succession using the
classification scheme of Huson et al. (2002).The sample periods are 1980-1987 and 1992-1998. The p-
value of the Chi-squared statistic that tests the hypothesis that the proportion of firms experiencing each
event is equal in the two sub-periods is reported in the last column.

Period
Full Sample 1980-1987 1992-1998
Number Percent Number Percent Number Percent
of Firms of Firms of Firms of Firms of Firms of Firms p-value

Full Sample 285 100 124 100 161 100

Nontargets 241 85 106 85 135 84

All Targets 44 15 18 15 26 16 0.51

Friendly 25 9 4 3 21 13 0.01

Hostile 19 7 14 11 5 3 0.00

CEO Tumnover 88 31 34 27 54 34 0.16

Forced Tumnover 43 15 9 7 34 21 0.00


Disciplined 59


21 21 17


38 24 0.19










Table 3.3
Abnormal stock market performance of acquiring firms

The table reports the acquisition announcement returns for firms that remain independent nontargett),
become a hostile target, become a friendly target, experience no turnover (nontumnover), CEO turnover,
forced turnover, or are disciplined defined as forced turnover or target of a hostile takeover. A hostile target
is a firm that is the target of successful or unsuccessful hostile tender offers, unsolicited and rejected offers,
and proxy contests. A firm is a friendly target if the firm is the target of friendly merger or tender offer
proposals, or successful and unsuccessful management buyouts, and does not receive any hostile offers. A
firm is classified as a nontarget if it does not receive any hostile or friendly offers. A firm experiences CEO
turnover whenever a succession of the top officer occurs, excluding those that are the result of health
reasons and death. Forced turnover excludes top executive turnover labeled as normal succession using the
classification scheme of Huson et al. (2001).The sample periods are 1980-1987 and 1992-1998. The
abnormal return is the market-adjusted return measured over the [-1,1] and [-5,5] windows. For firms with
multiple acquisitions that meet the criteria discussed in table 3.1, the abnormal return is the arithmetic
average of the announcement returns for all its acquisitions. The p-values of the t-test and signed rank test
are reported in parentheses below the mean and median, respectively.

Panel A: [-1,1] window Period
Full Sample 1980-1987 1992-1998
Mean Median Mean Median Mean Median

Full Sample -0.023 -0.023 -0.022 -0.021 -0.024 -0.025
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Nontargets -0.020 -0.020 -0.020 -0.018 -0.020 -0.023
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
All Targets -0.039 -0.035 -0.031 -0.030 -0.044 -0.041
(0.00) (0.00) (0.00) (0.00) (0.03) (0.02)
Hostile Target -0.027 -0.033 -0.031 -0.030 -0.016 -0.039
(0.01) (0.01) (0.00) (0.01) (0.46) (0.63)
Friendly Target -0.047 -0.036 -0.033 -0.030 -0.050 -0.045
(0.02) (0.00) (0.00) (0.13) (0.02) (0.02)
Nontumnover -0.019 -0.021 -0.021 -0.022 -0.018 -0.021
(0.00) (0.00) (0.00) (0.00) (0.01) (0.00)
CEO Tumnover -0.031 -0.027 -0.024 -0.016 -0.036 -0.021
(0.00) (0.00) (0.01) (0.00) (0.00) (0.00)
Forced Tumnover -0.048 -0.044 -0.028 -0.028 -0.054 -0.056
(0.00) (0.00) (0.06) (0.13) (0.00) (0.00)
Disciplined -0.043 -0.040 -0.032 -0.033 -0.049 -0.051
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00)











Table 3.3 continued.

Panel B: [-5,5] window


Period
1980-1987


Full Sample


1992-1998
Mean Median


Mean

-0.023
(0.00)
-0.022
(0.00)
-0.035
(0.02)
-0.013
(0.41)
-0.051
(0.00)
-0.016
(0.01)
-0.041
(0.00)
-0.067
(0.00)
-0.054
(0.00)


Median

-0.027
(0.00)
-0.024
(0.00)
-0.038
(0.01)
-0.038
(0.20)
-0.038
(0.02)
-0.014
(0.00)
-0.041
(0.00)
-0.061
(0.00)
-0.054
(0.00)


Mean

-0.021
(0.00)
-0.021
(0.00)
-0.019
(0.28)
-0.008
(0.70)
-0.055
(0.00)
-0.021
(0.02)
-0.021
(0.09)
-0.022
(0.48)
-0.021
(0.20)


Median

-0.022
(0.00)
-0.018
(0.00)
-0.045
(0.14)
-0.030
(0.43)
-0.059
(0.13)
-0.017
(0.00)
-0.030
(0.05)
-0.042
(0.43)
-0.041
(0.13)


Full sample

Nontargets

All Targets

Hostile Target

Friendly Target

Nonturnover

CEO Turnover

Forced Turnover

Disciplined


-0.026
(0.00)
-0.022
(0.00)
-0.046
(0.04)
-0.028
(0.19)
-0.051
(0.07)
-0.012
(0.37)
-0.054
(0.00)
-0.079
(0.00)
-0.072
(0.00)


-0.031
(0.00)
-0.031
(0.00)
-0.033
(0.04)
-0.038
(0.13)
-0.029
(0.08)
-0.014
(0.63)
-0.051
(0.00)
-0.077
(0.00)
-0.066
(0.00)










Table 3.4
Differences in abnormal returns of acquiring firms

The table reports the differences in means and medians of the acquisition announcement returns for
acquiring firms that are subsequently classified as target, hostile target, friendly target, experiencing CEO
turnover, forced turnover, and disciplined (forced turnover or hostile target). A hostile target is a firm that
is the target of successful or unsuccessful hostile tender offers, unsolicited and rejected offers, and proxy
contests. A firm is a friendly target if the firm is the target of friendly merger or tender offer proposals, or
successful and unsuccessful management buyouts, and does not receive any hostile offers. A firm is
classified as a nontarget if it does not receive any hostile or friendly offers. A firm experiences CEO
turnover whenever a succession of the top officer occurs, excluding those that are the result of health
reasons and death. Forced turnover excludes top executive turnover labeled as normal succession using the
classification scheme of Huson et al. (2001).The sample periods are 1980-1987 and 1992-1998. The
abnormal return is the market-adjusted return measured over the [-1,1] and [-5,5] windows. For firms with
multiple acquisitions that meet the criteria discussed in table 3.1, the abnormal return is the arithmetic
average of the announcement returns for all its acquisitions. The p-values of the t-test and Wilcoxon z-
statistic for the tests of differences in means and medians, respectively, are in parentheses.

Panel A: [-1,1] window Period
Full Sample 1980-1987 1992-1998
Mean Median Mean Median Mean Median

Targets vs. -0.018* -0.015 -0.011 -0.012 -0.023 -0.018
Nontargets (0.08) (0.15) (0.41) (0.34) (0.13) (0.26)
Hostile vs. -0.007 -0.013 -0.011 -0.012 0.004 -0.016
Nontargets (0.64) (0.73) (0.48) (0.45) (0.90) (0.86)
Friendly vs. -0.027* -0.016 -0.013 -0.012 -0.030* -0.022
Nontargets (0.05) (0.12) (0.63) (0.50) (0.08) (0.18)
Turnover vs. -0.012 -0.006 -0.004 0.006 -0.018 0.000
Nontumnover (0. 14) (0.32) (0.74) (0.90) (0.14) (0.21)
Forced turnover -0.029*** -0.023*** -0.003 -0.006 -0.036** -0.035***
vs. Nontumnover (0.01) (0.01) (0.69) (0.92) (0.02) (0.01)
Disciplined vs. -0.022** -0.023** -0.013 -0.016 -0.029** -0.034***
Nondisciplined (0.02) (0.01) (0.32) (0.36) (0.02) (0.01)


***, **, denote significance at the one, five, and ten percent level, respectively.











Table 3.4 continued.

Panel B: [-5,5] window


Period
1980-1987
Mean Median


Full Sample
Mean Median


1992-1998
Mean Median


-0.013
(0.35)
0.008
(0.67)
-0.030
(0.11)
-0.025**
(0.02)
-0.051***
(0.00)
-0.028**
(0.03)


-0.014
(0.63)
-0.014
(0.73)
-0.014
(0.37)
-0.029**
(0.02)
-0.049***
(0.00)
-0.039**
(0.01)


0.002
(0.91)
0.013
(0.58)
-0.034
(0.42)
-0.001
(0.96)
-0.001
(0.97)
0.001
(0.96)


-0.027
(0.97)
-0.012
(0.69)
-0.031
(0.40)
-0.013
(0.57)
-0.025
(0.88)
-0.020
(0.96)


-0.024
(0.21)
-0.006
(0.88)
-0.028
(0.19)
-0.042***
(0.00)
-0.067***
(0.00)
-0.046***
(0.00)


-0.002
(0.65)
-0.007
(0.96)
0.002
(0.62)
-0.037***
(0.00)
-0.063***
(0.00)
-0.053***
(0.00)


Targets vs.
Nontargets
Hostile vs.
Nontargets
Friendly vs.
Nontargets
Turnover vs.
Nontumnover
Forced turnover
vs. Nontumnover
Disciplined vs.
Nondisciplined


***, **, denote significance at the one, five, and ten percent level, respectively











Table 3.5
Frequency and type of control-reducing event for good acquirers versus bad acquirers

The table reports the proportion of acquiring firms that experience a disciplinary control event in the three years following the acquisition. The sample is
partitioned in quartiles based on the [-1,1] and [-5,5] abnormal returns. Firms are classified as target, hostile target, friendly target, CEO turnover, forced
turnover, or disciplined (forced turnover or hostile target). A firm is a target when it receives any takeover offer within three years after the
announcement. A hostile target is a firm that is the target of successful or unsuccessful hostile tender offers, unsolicited and rejected offers, and proxy
contests. A firm is a friendly target if the firm is the target of friendly merger or tender offer proposals, or successful and unsuccessful management
buyouts, and does not receive any hostile offers. A firm experiences CEO turnover whenever a succession of the top officer occurs, excluding those that
are the result of health reasons and death. Forced turnover excludes top executive turnover labeled as normal succession using the classification scheme
of Huson et al. (2001).The sample periods are 1980-1987 and 1992-1998. The p-values of the Chi-squared statistic that tests the hypothesis that the
proportion of firms experiencing each event is equal between the two periods are also reported.

Panel A: [-1,1] window
Target Hostile Target Friendly Target CEO Tumnover Forced Tumnover Disciplined
N Fraction p-value Fraction p-value Fraction p-value Fraction p-value Fraction p-value Fraction p-value


Full sample
Top quartile 72
Bottom quartile 72


0.46 0.04
0.03


0.45 0.10
0.13


0.39 0.02
0.00


0.65 0.07
0.13


0.69 0.00
0.03


0.31 0.12
0.22


0.26 0.26
0.35


0.31 0.26
0.26


0.24 0.27
0.37


0.28 0.10
0.24


1.00 0.10
0.06


0.34 0.12
0.29


0.03 0.13
0.26


0.64 0.16
0.19


0.06 0.15
0.29


0.04


1980-1987
Top quartile
Bottom quartile

1992-1998
Top quartile
Bottom quartile


0.74




0.11











Table 3.5 continued.


Panel B: [-5,5] window
Target Hostile Target Friendly Target CEO Turnover Forced Turnover Disciplined
N Fraction p-value Fraction p-value Fraction p-value Fraction p-value Fraction p-value Fraction p-value

Full sample
Top quartile 72 0.15 0.63 0.06 0.40 0.08 0.77 0.22 0.05 0.06 0.00 0.10 0.00
Bottom quartile 72 0.13 0.03 0.10 0.38 0.28 0.29

1980-1987
Top quartile 31 0.10 0.30 0.10 0.30 0.00 1.00 0.26 0.35 0.06 1.00 0.13 0.69
Bottom quartile 31 0.03 0.03 0.00 0.16 0.06 0.10

1992-1998
Top quartile 41 0.17 0.78 0.02 1.00 0.15 0.76 0.20 0.00 0.05 0.00 0.07 0.00
Bottom quartile 41 0.20 0.02 0.17 0.54 0.44 0.44











Table 3.6
Pooled Logit Regressions on CEO turnover and forced CEO turnover

Logistic regressions estimating the relation between the probability of turnover and the abnormal return
around the acquisition announcement, bidder size, CEO age, a pre-acquisition performance variable, a
dummy variable equaling one if the merger occurred during 1980 to 1987, as well as an interaction term
measuring the difference in the sensitivity of merger performance to turnover across the two periods A
firm experiences CEO turnover whenever a succession of the top officer occurs, excluding those that are
the result of health reasons and death. Forced turnover excludes top executive turnover labeled as normal
succession using the classification scheme of Huson et al. (2001). The abnormal return is the market-
adjusted return measured over the [-1,1] and [-5,5] windows. For firms with multiple acquisitions that meet
the criteria discussed in table 3.1, the abnormal return is the arithmetic average of the announcement returns
for all its acquisitions. Bidder size is measured by the logarithm of the market value of the bidder 20 days
prior to the first announcement scaled by the value-weighted CRSP return index (1980=1). The CEO age
dummy is equal to one if the CEO is 60 years or older at the time of the announcement, and zero otherwise.
The Buy-and-Hold abnormal return is the market-adjusted return calculated over the [-250,-10] window.
The sample consists of 266 firms that remained independent before CEO turnover or for the three years
after the acquisition. CEO turnover occurred in 88 firms, with 43 firms experiencing forced turnover. p-
values are in parentheses.

Firm experiences turnover Firm experiences forced turnover
(N=266) (N=266)
[-1,1] [-5,5] [-1,1] [-5,5]

Intercept -0.658 -0.981 -0.486 -0.984
(0.67) (0.53) (0.80) (0.62)
Abnormal Retumn -6.581** -6.778*** -9.050*** -9.008***
(0.03) (0.00) (0.01) (0.00)
Bidder Size -0.030 -0.009 -0.081 -0.050
(0.81) (0.94) (0.59) (0.75)
CEO Age>60 1.393 1.401
(0.00) (0.00)
D80-87 -0.501 -0.430 -1.017** -0.877**
(0.12) (0.17) (0.03) (0.05)
D80-87*AR 4.956 6.201* 6.078 8.561*
(0.34) (0.08) (0.43) (0.09)
BHAR -0.141 -0.088 -0.025 0.049
(0.61) (0.75) (0.93) (0.87)
Pseudo R2 0.082 0.067 0.092 0.117


***, **, denote significance at the one, five, and ten percent level, respectively.











Table 3.7
Ownership and board characteristics of the 1992-1998 sample

The table reports the board and ownership characteristics for 146 acquirers of the 1992-1998 sample. Firms that are
acquired during the three-year post-acquisition period or have missing proxy statements are excluded. CEO
characteristics are CEO age at the time of the acquisition, proportion of shares held by the CEO, and whether the
same individual holds the CEO and chairman position. Board characteristics are the size of the board, proportion of
board comprised of outsiders, and proportion of shares held by officers and directors. Directors are classified as
outsiders when the individual is not currently, or has never been employed by the firm, has no business relationships
with the firm, and has no family ties to management. Unaffiliated blockholders are defined as shareholders that own
five percent of the stock, and do not have any type of agreement with the firm. The proportion of firms with at least
one blockholder, the average number of blockholders, and blockholder holdings are reported. The sample is
partitioned into firms that experience forced turnover and of those that do not. Forced turnover is defined as any
succession of the top officer, excluding those that are the result of health reasons and death, or are labeled as normal
succession using the classification scheme of Huson et al. (2001).

Full Sample No Forced Tumnover Forced Tumnover
(N=146) (N=112) (N=34)
Mean Median Mean Median Mean Median

CEO characteristics
CEO age at time of the acquisition 53 53 54 54 50** 52**
Proportion of shares held by CEO 0.04 0.01 0.04 0.01 0.03 0.01
Proportion firms CEO is Chairman 0.87 1 0.89 1 0.79 1

Board Characteristics
Number of directors 10.34 10 10.43 10 10.03 10
Proportion of outside directors 0.70 0.73 0.70 0.75 0.69 0.71
Proportion of shares O&D 0.10 0.04 0.11 0.04 0.08 0.05

Unaffiliated Blockholdings
Proportion at least one blockholder 0.76 1 0.78 1 0.71 1
Number of unaffiliated blockholders 1.73 2 1.76 2 1.62 2
Proportion of shares blockholders 0.15 0.14 0.16 0.15 0.11* 0.12

***, **, denote significance at the one, five, and ten percent level, respectively.











Table 3.8
Pooled Logit Regressions on CEO turnover and forced CEO turnover for the 1992-1998 sample

Logistic regressions estimating the relation between the probability of turnover and the abnormal return around the
acquisition announcement, bidder size, CEO age, pre-acquisition performance variable, CEO ownership, and two
interaction terms. The first interaction term measures the difference in the sensitivity of acquisition performance to
turnover for outsider-dominated boards. The second measures the difference in the sensitivity of acquisition
performance to turnover for firms with blockholders. A firm experiences CEO turnover whenever a succession of
the top officer occurs, excluding those that are the result of health reasons and death. Forced turnover excludes top
executive turnover labeled as normal succession using the classification scheme of Huson et al. (2001) The
abnormal return is the market-adjusted return measured over the [-1,1] and [-5,5] windows. For firms with multiple
acquisitions that meet the criteria discussed in table 3.1, the abnormal return is the arithmetic average of the
announcement returns for all its acquisitions. Bidder size is measured by the logarithm of the market value of the
bidder 20 days prior to the first announcement scaled by the value-weighted CRSP return index (1980=1). The CEO
age dummy is equal to one if the CEO is 60 years or older at the time of the announcement, and zero otherwise. The
Buy-and-Hold abnormal return is the market-adjusted return calculated over the [-250,-10] window. The sample
consists of 146 firms that are not acquired before CEO turnover or the three years after the acquisition
announcement and for which proxy information on governance variables are available.

Firm experiences turnover Firm experiences forced turnover
(N= 146) (N= 146)
[-1,1] [-5,5] [-1,1] [-5,5]

Intercept 0.442 -0.195 -0.592 -1.696
(0.83) (0.93) (0.81) (0.49)
Abnormal Retumn -13.7 -12.1** -21.4** -18.7***
(0.11) (0.05) (0.04) (0.00)
Bidder Size -0.092 -0.047 -0.065 0.015
(0.58) (0.78) (0.73) (0.94)
CEO Age>60 1.096** 1.105***
(0.01) (0.01)
CEOHOLD -0.081* -0.086* -0.040 -0.043
(0.06) (0.06) (0.38) (0.36)
Outsider*AR 4.331 4.192 10.538 8.692
(0.48) (0.44) (0.12) (0.15)
Blockholder*AR 5.112 3.022 5.303 3.903
(0.50) (0.53) (0.55) (0.45)
BHAR -0. 177 -0.165 -0.232 -0.229
(0.59) (0.64) (0.50) (0.56)
Pseudo R2 0.078 0.101 0.081 0.113

***, **, denote significance at the one, five, and ten percent level, respectively.












0.02



Days relative to announcement

0 1\5/ V10V 15 2~1 V 25 30 35 A
Hostile Target

9 -0.02
Nontarget



-0.04


I V \ Friendly Target

-0.06




-0.08




-0.1




Figure 3.1
Stock price reaction to acquisition announcements for firms that subsequently become a hostile or friendly target








62




0.02




0 Days relative to announcement

-5 5 10 15 20 25 30 35 4


-00 -z
Non-Turnover




-0.04

I \ CEO Turnover


-0.06


Forced Turnover

-0.08




-0.1




Figure 3.2
Stock price reaction to acquisition announcements for firms that subsequently experience CEO turnover
















CHAPTER 4
IMPLICATIONS OF DEFEATING TAKEOVER BIDS:
EVIDENCE PRE AND POST TIME-WARNER


A popular viewpoint is that the problem of entrenched management has been

exacerbated by new takeover defenses that have become available and legally permissible

in the past twenty years. Many commentators point to the Delaware Supreme Court's

decision in Paramount Communications, Inc v Time, Inc ("Time-Warner") in July 1989

that upheld the use of the so called poison pill takeover defense even in the case of a fully

financed cash takeover offer for all shares of a corporation's stock.l The Time-Warner

decision helped give birth to what the popular press termed the "just say no defense" in

which management can simply refuse to bargain with a potential buyer on the grounds

that the offer is inadequate. At the same time, the poison pill defense makes it nearly

impossible for the bidder to acquire the shares needed to replace management.

This chapter addresses the issue of whether the advent of the "just say no takeover

defense" has had a significant impact on managerial entrenchment by examining a

specific example of alleged entrenchment that has long drawn the attention of academics

and the media: the defeat of an unsolicited takeover bid. Unsolicited takeover bids are

generally at prices that exceed current market prices and often entail replacement of

management. Given that unsolicited bidders are unsuccessful in acquiring control of the

target firm, the question becomes, how much, if at all, do shareholders suffer as a result



SSee, for instance, Hilder (1990).









of defeating a bid. As the name of the "just say no defense" implies, managers in the

post Time Warner era are arguably better equipped to simply go about their business and

make fewer efforts to move toward value maximization when fending off unsolicited

bids. Hence, the entrenchment hypothesis predicts that post Time Warner we should

observe fewer takeover defenses that are often associated with moving toward value

maximization such as debt financed recapitalizations. Post Time-Warner, we should

observe smaller positive revaluations of targets resulting from failed bids and the post-

takeover bid share prices of these targets should trade at larger discounts relative to the

takeover price offered by the failed bidders.

The results show that the "just say no defense" has had a dramatic impact in terms

of how managers defend against unsolicited takeover bids. I document significant

declines in the use of share buybacks, special dividends, greenmail, placement of shares

with friendly third parties, and standstill agreements during the post Time-Warner period.

Managers were also far less likely to increase leverage in the 1990s as opposed to the

1980s when defending against takeovers. Instead, the poison pill (or its threatened use)

is the primary defense in the 1990s. The evidence is consistent with managers viewing

the "just say no" takeover defense as a "lower cost" defense option.

Despite the ability to "just say no," the analysis shows that the average valuation

implications of successfully defending takeover offers pre and post Time-Warner are

very similar. I calculate the valuation gap between what is offered in a takeover bid and

the post offer share price, and find that it is virtually identical between the two time

periods. The average market-adjusted target revaluation, calculated over the takeover

contest, is also nearly identical between the two periods. One possible explanation of









this finding is that the nature of the bids differ between the two time periods. For

instance, two-tier bids and bids by non-public entities are more common in the pre-Time

Warner period. Even after controlling for these (and other) factors, however, the

valuation consequences of a failed bid still are quite similar between the two time

periods. With respect to the nature of the defense, similar to Safleddinie and Titman, I

find that firms that increase leverage in the pre Time Wamner period increase their

revaluations relative to those that do not, but no such pattern appears post Time Warner.

Also, firms that pay greenmail as a defensive tactic appear to have more negative

revaluations associated with the failed takeover episode.

Finally, I followed the sample of firms in the period after defeating a hostile offer.

I find that Post Time-Warner, targets are no less likely to be acquired and CEOs are more

likely to lose their executive positions after defeating a bid. Collectively, these results

suggest that managers that successfully defeat takeovers in the 1990s are no more

entrenched than those in the 1980s. This is consistent with other governance devices

placing limitations on the deviation from wealth maximization caused by "just say no"

takeover defenses for firms that defeat a hostile bid.

In interpreting the results, we should bear in mind that we are focusing on a set of

firms where the management team (or at least the existing board of directors) remains

intact for some period of time after defeating a takeover bid. The significance of this

group is to determine just how far a management team might be able to stray from value

maximization, while maintaining independence in the face of an unsolicited bid in

alternative legal environments. Obviously, the number of external challenges and the

frequency with which they succeed also need to be factored in when assessing the degree










to which the average manager may be able to deviate from value maximization in

different legal environments. This is but one piece of the entrenchment puzzle.

This chapter is organized as follows. The next section reviews the literature on

managerial entrenchment, hostile takeovers, and contains the hypotheses development.

Section 4.2 describes the data. Section 4.3 documents the time variation of the takeover

defenses. Section 4.4 focuses on intertemporal variation of managerial entrenchment. The

post-contest consequences of defeating a bid on to the firm and management are

discussed in Section 4.5. Concluding remarks follow in section 4.6.



4.1 Previous Literature and Hypotheses Development

4.1.1 Managerial Entrenchment, Takeover Bids, and Agency Theory

One of the most cited examples of agency conflicts between management and

shareholders is management defeating a takeover with a higher stated value than the price

of the shares at the time the takeover is defeated. While management teams certainly

have the potential motivation to Eight for their managerial independence at the potential

cost of forgoing a takeover premium, such actions are rather curious in an agency

theoretic context. Generally, agency problems are viewed as observability problems and

defeating a takeover is a fairly public spectacle.

There are various reasons why management may choose to defeat a takeover

offer. Management can make claims that their shares are undervalued based on their

access to confidential information. In addition, targets may deny a bidder due diligence










lest competitors learn about competitive advantages that can be exploited by rivals.2

Often, targets argue that the compensation being proffered by a bidder is not worth what

it is claimed to be worth. This might be particularly true of offers that include contingent

securities that have no market or that may have lower valuations if we had more complete

knowledge of the bidder' s business or the potential synergies from a merger. Taking this

argument a step further, management often makes claims that the bidding party is not

serious about actually acquiring the target firm. A bidder may simply be trying to profit

on an existing investment position in a firm by mounting a takeover bid and imposing

costs on management so as to extract so called greenmail payments, or by creating a buzz

about potential takeovers that temporarily inflates the price of the firm's securities and

allows the "bidder" to sell their position at a more attractive price.

Nevertheless, to the extent that takeover bids are "serious," the decision to resist a

bid is quite public. Successfully defeating an offer, with a resulting share price often well

below the perceived value of the bid, might invoke disciplinary mechanisms to come into

play. Competing bidders may step in, or the original bidder might return with enhanced

credibility that management was in the wrong in defeating their prior bid. In many cases,

management is pressured to execute some, if not all, of the plans that a potential bidder

had for the firm such as asset divestitures to refocus the firm or increasing financial

leverage to capture tax benefits. However, management may be able to deviate from

value maximization to a larger extent, the greater the entrenching devices that are





SSome bidders also request compensation for expenses as a precondition into entering into some
preliminary agreement that includes due diligence rights. This too could prove costly to target
shareholders.









available to them. The sophistication and availability of such devices, in particular the

poison pill defense, has evolved with the takeover market in the past two decades.



4.1.2 Hostile Takeovers and Target Defenses

The 1980s experienced an unprecedented wave of so called hostile takeover

activity in which companies received and initially rejected unsolicited takeover

proposals. Andrade, Mitchell, and Stafford (2002) Eind that 14 percent of CRSP firms

receive a hostile offer between 1980 and 1989. Larger firms were even more likely to

receive a hostile offer. Mitchell and Mulherin (1996) Eind that 23% of Value Line firms

are the target of hostile takeover activity during the 1980s. The wave of hostile takeover

activity reflected a number of factors, such as better access to capital markets for investor

groups and smaller corporations in the "junk bond" market, undoing the diversification of

corporations incurred during previous merger waves, and changes in antitrust policy.

Firms employed numerous defensive tactics such as greenmail, the placement of

securities with friendly third parties, selling off so called crown jewel assets, increasing

leverage via share repurchases and instituting asset restructurings. While many of the

defenses were viewed as "entrenching devices," it has been argued, that many of the

defenses were simply enacting certain elements of what outside acquirers would have

done themselves if they had acquired the company. In other words, many defenses were

merely actions to increase share price by moving toward value maximization for

shareholders. For instance, Safieddine and Titman (1999) Eind for a sample of defeated

takeover bids from 1981-1991, that firms that increase their leverage are much more









likely to have positive revaluations. Leverage increases accompanying defensive share

buybacks and special dividends are common takeover defenses utilized in the 1980s.

The most dramatic development in combating hostile takeover defenses was the

so called poison pill takeover defense (see Ryngaert (1988)). The defense first became

prominent in 1985. It was effective in delaying acquirers and raising the legal costs

associated with takeovers. Nevertheless, the pill defense did little to mitigate the wave of

takeover activity in part because the courts placed limits on the use of the defense. For

instance, it was generally taken as conventional wisdom that for Delaware incorporated

firms, the pill could be used to combat two-tier takeovers (see Unocal Corp v. Mesa

Petroleum, 1988), but that the pill could not be deployed indefinitely to defeat any or all

takeovers (see Grand Metroplitan, PLC v Pillsbury, 1988). This changed with the

Delaware Supreme Court's decision in Paramount Communications, Inc v Time, Inc in

July 1989. The decision suggested that managers could reject a bid if they felt the offer

was too low and suggested that management was under no obligation to remove a poison

pill takeover defense even in the face of an all-cash, fully financed tender offer for their

firm. The decision basically gave managers the right to "just say no" to takeover bids.

The only recourse to a potential acquirer facing such a defense was to replace the board

of directors in a proxy fight, a much more protracted and difficult undertaking, especially

when board terms were staggered.

Subsequent popular press articles suggested that this defense would severely

damage shareholders by crippling the hostile takeover mechanism for disciplining

management. Consistent with the chilling effect of the Time-Warner decision, the

incidence of hostile takeover activity plunged in the early 1990s and never fully









recovered. The Time-Wamner decision is not the exclusive explanation for this decline,

however. Jensen (1993), for example, points to the closing down of the junk bond market

to would be hostile acquirers and the cumulative impact of state takeover laws as

contributing factors.



4.1.3 Hypotheses regarding defeated takeover bids in 1980s and 1990s

If the ability to "just say no" has allowed firms to stray further from value

maximization, a natural corollary would be that firms are under far less pressure to value

maximize when they do receive a hostile takeover offer post Time-Wamner. This also

implies that, all else equal, firms should have to take fewer alternative defensive actions

and would be less likely to lose their executive positions due to some subsequent actions

taken against them. More importantly, to the extent that the "just say no" legal

environment leads to entrenchment, the share prices of firms that defeat takeover bids

should have less positive revaluation in the wake of an unsuccessful takeover bid and

should also trade at a larger discount relative to the price offered in the unsolicited bid.

The alternative hypothesis is that substitute mechanisms have arisen to force management

teams to "come close" to the value offered by an external bidder or risk dismissal. This

would suggest that we might see few differences in the outcomes associated with

defeated hostile takeover bids in the 1980s and 1990s.

The above arguments, of course, assume that the set of takeover targets and

bidders is similar between the 1980s and 1990s. For instance, differences in post

takeover defeat executive turnover rates and/or share price revaluation resulting from

unsuccessful takeover bids may be a function of how poorly target firms performed in the










pre and post Time-Warner samples. Similarly, there may be differences in the nature of

the bidding parties, the nature of the bids themselves and the corporate governance

structure of the firms that receive the bids in the two sample periods. I control for these

differences in the analysis.



4.2 Data Description and Methodology

4.2.1 Sample of Defeated Takeover Bids

The goal is to construct a sample of firms where management opposed a takeover

bid and the firm remained independent for a period of time. The definition of "remaining

independent" in the face of a hostile bid must meet the following three criteria: (1) the

initial bid is unsolicited, (2) the target firm management rejects the bid, management

pursues litigation against the bidder or the bidder cites managerial resistance as a reason

for dropping its bid, and (3) the bidder subsequently "drops the bid" with the exception of

cases where the target had agreed to be acquired, but the bidder subsequently backed out

of the agreement.

While the criteria sound simple, the challenge is determining when a bid is

dropped by the bidder and when the firm has "successfully" fended off an unsolicited bid.

Frequently acquirers drop a bid, only to come back shortly thereafter with another offer

or some other means to achieve control such as a proxy contest or acquisitions of shares

in the open market. In other cases, bidders drop a bid because they realize there are other

parties bidding "behind the scenes." In fact, this behind the scenes activity may be the

very reason target management is resisting an existing bid in the first place. Finally, in

many cases, the bidder simply "goes away" without comment when its bid is rejected,









and the bid is never publicly dropped at all. In all these different types of contests, I seek

to not only classify whether the firm has "successfully" remained independent but to also

identify a date of resolution for purposes of measuring wealth impacts of the takeover

contest on the target.

Using the Factiva database of news stories and taking into account the above

considerations, one of the following criteria must be met for a bid to be considered

defeated, and a resolution date for each type of contest is determined as follows. (1) The

bidder withdraws the bid or sells shares in the target and ceases taking actions to

influence the firm's policies or purchase the firm's assets for three months, and there are

no other takeover bids for the firm over the next three months after the bid withdrawal.

In this case the bid withdrawal date is considered the resolution date. (2) The bidder

drops the initial bid but clearly states that it will actively seek to control the firm or takes

subsequent actions to control the firm. In these cases, the bid is deemed dropped when

four consecutive months go by without the bidder taking any actions to influence firm

policy or acquire firm assets and with no other takeover bids being made. The resolution

date is 40 days after the rejection or the last bidder action. (3) In the event that a bid is

rejected and the bidder never formally drops its bid, then the bid is deemed defeated

when the bidder ceases to take actions to influence control of the target for four

consecutive months and with no other takeover bids being made. The resolution date is

40 days after the rejection or the last bidder action. (4) Management declares that it is

putting the firm (or pieces of it) up for sale as a means to gain peace with a hostile bidder

or as a means of winning a proxy contest. If no takeover related announcements occur

for four months, then the resolution date is 40 days after the date of the for sale










announcement. If there are stories updating the progress of this "sale process," the

resolution date is whenever target management declares that the search process is over or

has turned up no buyers.

The primary data source is the Thomson Financial SDC database listing of hostile

and/or unsolicited takeover bids. Additional observations were hand collected from

annual issues of W. T. Grimm Mergerstat Review. Using the SDC database, there are

1477 instances between the years 1980-1999 where a firm received a takeover bid that

was deemed "unsolicited" or "hostile" by SDC. From this list I identify the resolution

dates of these contests and split the firms into pre and post Time-Warner groupings.

There are 676 instances pre Time Warner and 801 instances post Time Warner. First, I

eliminate all instances that are related to banks, REITs, limited partnerships, and closed

end mutual funds. These firms have very different capital structures, tax issues, and

regulatory issues relative to firms in other industries and they also appear in large

numbers. I also eliminate offers for private firms and divisions of firms since there are

no stock quotes for these assets. I further eliminate offers for firms in bankruptcy or for

takeover bids conditional on debt workouts. Negotiations in these cases must take place

with a broader set of parties than bidder and target management teams. Offers by unions

and government entities are also eliminated due to their unusual nature and the potential

that these offers are parts of broader bargaining strategies. Offers that are not for control

are also excluded. These include so called "clean up" offers where the bidder already

controls 40% of the stock of the firm and offers to invest in the company or acquire only

a foothold in the company less than 40%. Three false offers, where the press release

announcing the bid turned out to be fraudulent are also excluded. Offers by the










management team of the firm (executives and/or chairman of the board) are excluded. So

called Pac Man offers made by firms that are the subject of a hostile bid are excluded

since these are more defensive maneuvers than unsolicited bids. Other eliminations

include offers that were really "friendly" agreements between the target and the bidder,

offers where no terms were disclosed, offers for less than $1.50 per share, and offers

where the rej section of the bid became public knowledge well after the event itself.

After these various screens displayed in Table 4. 1, the remaining number of firms

is 456 pre Time-Wamner and 337 post Time-Warner. After screening out firms where the

target had multiple entries for the same takeover contest, the number of pre Time-Wamner

cases drops to 389 and the number of post Time Warner Cases drops to 294. Next, I

exclude cases where the firm receiving the unsolicited bid got acquired, announced it

would liquidate or had the bidding entity gain working control of the board of directors

during the takeover contest. There are 250 such "change of control" events in the pre

Time-Warner period and 141 in the post Time Warner period. Given that the pre Time-

Warner period encompasses a slightly shorter period of time, this is consistent with prior

empirical work that demonstrates greater levels of hostile takeover activity in the 1980s

versus the 1990s.

After these eliminations, there are 139 and 153 firms respectively that are

classified as remaining independent in the pre and post Time-Warner periods. I eliminate

cases where management agreed to be acquired, but the bidder backed out latter on its

own accord, since target management cannot be blamed for such occurrences. Similarly, I

also eliminate cases where the bidder backed out without any blaming of management

and without a rejection of the bid by the target firm. The Einal concern was to eliminate









offers where the true value of the offer was difficult to decipher. I eliminate one offer

where different prices were offered to different classes of stock. I also eliminate offers

where securities of questionable value were to make up a significant portion of the offer

price. These offers include cases where more than 45% of the compensation offered by a

private firm was in securities or where a private firm was offering more than 20% of the

offer' s alleged value in common stock of a new firm. In the case of public firms, I keep

any offer where at least half of the compensation offered is in cash. Firms are eliminated,

however, if less than half the compensation is cash and the offer to the target exceeded

the market value of the bidding firm. The value of securities in such cases will largely

depend on how successful the merger would be and in the case of preferred securities

there might be considerable disagreement over their true value.

Finally, any predominantly stock offer where the number of shares to be paid to

the target is fixed and the target firm shareholders would own more than one-third of the

firm is excluded. The value of such near "merger of equals" offers can only be

determined as the product of the bidders share price times the number of shares offered to

the target shareholders. Unfortunately, in such large mergers the value of the bidder's

securities would be highly related to the perceived desirability of any takeover. Hence, it

is difficult to value the offer after it is defeated. In cases, where there are price

guarantees that allow the number of shares to be issued to float with the bidder's share

price, this is not an issue.

After these various screens, we are left with 122 pre Time-Warner observations

and 109 post Time-Warner observations from the original SDC sample. Clearly, the

number of defeated offers relative to changes of control is lower in the pre Time- Warner









era. This might be interpreted to mean that managers had greater ability to stay

independent in the event of a hostile takeover bid in the post time Warner era. The

purpose of this study is not to dispute this interpretation. Rather, the focus is on what

happens in the event of defeating a hostile bid. To obtain the largest possible sample of

such defeats, an additional 22 observations in the pre-Time Warner period and an

additional 17 in the post Time Warner observations are added from the Mergerstat

Review database of defeated bids from 1981-1999. These defeats meet all the same

criteria for inclusion as the SDC sample.

The final sample of defeated offers consists of 144 pre Time-Warner and 126 post

Time-Warner observations. The summary statistics for the sample of defeated offers are

presented in Table 4.2. The Table shows the number of defeated offers, competing offers,

multiple offers by the same bidder, and the median duration of the takeover contest per

year. I am able to find 195 clear resolution dates. When a precise withdrawal date could

not be identified, the resolution date was 40 days after the last rejection date or bidder

activity date as previously described in this chapter. For purposes of calculating the

median duration of the contest, I added 10 days to the last offer/rejection date for events

without clear withdrawal dates. The number of contests with multiple offers or competing

bidders is similar between the two periods.



4.2.2 Target Characteristics

It is often argued that hostile acquirers target underperforming firms. Early

research finds results consistent with this idea. Morck et al. (1989) report that hostile

takeover targets have lower Q ratios than friendly targets. Mitchell and Lehn (1990)









show that firms that make value-reducing acquisitions are likely to become hostile

targets. Both studies argue that hostile takeovers serve as a disciplinary mechanism to

reverse prior investment mistakes.

More recent studies have found no underperformance of targets prior to the

takeover offer. Agrawal and Jaffe (2001) use a large sample of target firms and examine

stock and accounting performance of friendly and hostile targets over a period of ten

years prior to the offer. Contrary to popular belief, they find no significant evidence that

either friendly or hostile targets underperform. Similarly, Schwert (2002) shows that

hostile targets cannot be distinguished from friendly targets in terms of accounting and/or

stock performance prior to a takeover bid.

Table 4.3 reports the firm characteristics for the targets in the sample. Panel A

reports the stock and accounting performance In terms of differences in the two sample

periods, the post Time-Warner targets tend to have poorer pre-announcement stock

market performance and higher market to book ratios. EBIT and FCF to asset ratios are

similar across periods. Hence, the evidence on ex-ante performance is not clear. The

market capitalization of targets is larger during the pre Time-Warner sample.

Panel B reports the industry breakdown for the targets based on the SIC codes

reported by Compustat. The majority of the targets have industry classifications of 3000-

3999. There are no clear differences in industries between the two periods. Proportionally

there are more firms in the 3000-3999 SIC code range in the 1990s.









4.2.3 Bidder Characteristics

Previous studies have largely ignored bidder characteristics in the takeover

process. The majority of deals are conditional on financing, due diligence, and other

conditions such as the removal of the poison pill defense. In some cases, however,

bidder's demands are highly conditional and arguably unreasonable, which leaves target

management no options but to reject the offer. In my sample design, I try to exclude the

most preposterous takeover attempts. Nevertheless, it is plausible that for some of the

offers in the sample, the bidder's reputation and prior dealings in the takeover market are

a major factor in target rejections. Acquirers may be motivated to make an offer to exert

disciplinary pressure on management to implement value-enhancing strategies or to put

the firm up for sale and sell its equity stake for profit later. Hostile acquirers are often

accused of trying to profit at the expense of other shareholders. But even when acquirers

own little or no equity stake in the target, it is not always clear that bidders are able to

secure Einancing. It is plausible that targets rej ect some offers conditional on Einancing for

good reasons.

Table 4.4 reports bidder characteristics for the pre- and post Time-Warner

samples. It is clear that bidder characteristics have changed over time. Post Time-Warner,

a larger fraction of the bidders are publicly traded firms. In addition, it appears that

bidders have become more "serious" over time. Financial buyers, such as private

investment firms and firms that have been paid greenmail in the five years prior to the

offer, are more common in the 1980s. One reason for fewer financial buyers during the

latter period is the collapse of the junk bond market in 1989 which has made financing









deals more difficult. This might suggest that the pre Time-Warner period has more

dubious bids, and I control for this in subsequent analysis in a couple of ways.

First, I classify eight bidders as having a bad reputation (BADREP) when the

principal of the bidding firm management is under criminal indictment, has run a

formerly bankrupt firm as noted in contemporaneous press accounts, or has previously

backed out of a merger agreement in the past two years due to financing problems.

Second, I create a variable called SYNERGY, which is a dummy equal to one if the

bidder has some potential vertical or horizontal strategy for use of the target's assets in

conjunction with the acquirer's own field of expertise. In cases where the bidder is

publicly traded I counted the offer as synergistic except in obvious cases where business

lines were clearly unrelated or in cases where the bidding firm was a shell company

acquiring other firms to allow it to utilized tax loss carry forwards. Acquisition bids by

private firms were generally deemed to not be synergistic unless news accounts discussed

how the bidding entity saw some synergy between existing operating divisions of the

private firm and the target firm or a business description of the bidding firm suggested

synergies with the public target. It appears that there are greater synergistic motives in the

1990s as two thirds of the deals are defined as having a synergy versus fewer than half in

the earlier period.

The average premium, calculated as the ratio of offer price to target price 20 days

before the initial offer was made public, is 34.6 percent. Using the date that the firm is

first put in play, the premium increases to over 40 percent. The premiums are relatively

constant over time and comparable to premiums for successful hostile and friendly public










targets.3 Since the focus of this chapter is on hostile offers, it is not surprising that the

vast maj ority of the deals are all-cash offers.

Grossman and Hart (1980) argued that target shareholders will not tender their

shares unless the price reflects all value improvements and synergies that will be created

by the merger. To circumvent this free rider problem, bidders can engage in so called

two-tier offers where the first offer for control is generally priced different than the

backed. Two-tier offers are much more common in the earlier period. In fact, in the

sample only four two-tier offers occurred after 1990. This is consistent with the courts

justifying more stringent takeover defenses (beginning in the middle of the 1980s) if the

bid was two-tiered. Finally, almost half of the bidders hold at least five percent of the

target before the initial offer, and the average foothold is approximately 6.5 percent.

Consistent with the idea that the majority of financial bidders are more likely to make

offers for reasons other than control, footholds are more common and larger during the

pre Time-Warner period. Post foothold refers to the equity stake held by the bidder

following the contest, and is similar during the two periods.



4.3 Takeover Defenses and Leverage

4.3.1 Takeover Defenses

Changes in the takeover market during the past two decades have also changed

the way firms are able to defend themselves from unwanted acquirers. While targets have

an arsenal of takeover defenses available, in this section I focus on share repurchases,

defensive dividends, greenmail, standstill agreements, majority blocking positions, the


3 Andrade et al. (2002) report a median premium of 37.7 percent and 34.5 percent for public targets during
the 1980-89 and 1990-1998 periods, respectively.









sale of shares to White Squires or ESOPs, staggered boards, and poison pills. The next

section describes the effect of buybacks, greenmail, and dividends, on the capital

structure of the firm.

BUYBACK is defined as the repurchase of 10 percent or more of stock either

through a tender offer or open market purchases. DIVIDEND is defined as a special

dividend payout in excess of 10 percent of frm value and includes dividends initiated by

spin-offs. GREENMAIL is defined as either a repurchase of shares from the hostile

acquirer at a premium or a payment of cash or in kind to secure a standstill agreement.

STANDSTILL refers to cases where the bidder is subject to a standstill agreement with

the target. RECAPITALIZATION is defined as any combination of share buyback,

greenmail, and dividend payout that exceeds 10 percent of equity value. In order to be

included as a defensive measure, recapitalizations must be announced between one month

before the first offer and one month after the withdrawal date and executed within one

year. Information on recapitalizations was collected from searching the Factiva database.

A firm has a BLOCK when the combined stake of directors, affiliated

blockholders, and ESOPs exceeds 25 percent during the contest. Some of these blocks are

created during the contest in large part to fight off the takeover offer. STAGGERED

implies that the board of directors consists of multiple classes. PRE-CONTEST VETO

and POST-CONTEST VETO refer to shareholder blocks that are in excess of 40 percent

before and after the contest, respectively, and have the ability to effectively block a

takeover. Information on shareholder blocks, standstill agreements, and staggered boards

were collected from detailed news accounts and proxy statements.









In response to growing hostility in takeovers, poison pill plans were introduced in

the mid 1980s. Poison pills allow target shareholders to purchase shares of the target or

the surviving entity in case of a successful merger at deep discounts, and therefore have

the potential to make takeovers 'prohibitively expensive'. Data on poison pills are

obtained from three sources. First, in many cases news accounts surrounding the takeover

attempt document the existence of poison pills. If no news accounts mentioned poison

pills, I checked the dataset used in Comment and Schwert (1995).4 Finally, I searched

multiple editions of the "Corporate Anti-Takeover Defenses: The Poison Pill Device" for

evidence of the use of poison pills. POISON PILL is a dummy equal to one when the

firm has a poison pill defense before the resolution date of the contest identified by any of

these three sources.

Table 4.5 reports the fraction of targets that use the various defensive strategies

available to the target during the takeover contest. It is clear from the table that there has

been a shift in the use of takeover defenses over time. Greenmail, dividend payouts,

general share repurchases, and standstill agreements were more common in the pre Time-

Warner period. The post Time-Warner period is characterized by few management

initiated defenses. The fraction of firms that have poison pills is higher, but there are still

many firms without them in the 1990s. It is important to remember however, that a firm

can adopt a poison pill simply with board approval. In other words, it is quite possible

that merely the threat of adopting a poison pill can fend off unsolicited bids. In fact, a

number of firms adopted poison pills subsequent to the failed bids, presumably for future

contests. Regardless, it appears that targets can more cheaply thwart unwanted raiders

after the Time-Warner decision. Firms also were more likely to have staggered boards of

4 The dataset can be found at Bill S chwert' s web site at http://schwert. ssb.rochester. edu/data. htm.









directors in the post Time-Warner period, which makes a good complement to the "just

say no" takeover defense.



4.3.2 Leverage as a Takeover Defense

There is an ongoing debate about the effects of leverage in targets of unsuccessful

takeover attempts. Berger et al. (1997) find that firms increase leverage after being the

target of an unsuccessful tender offer. They report that the mean increase in total debt to

assets is almost 12%. They argue that entrenched managers seek to avoid debt, but are

forced to increase their firm's leverage when their firm is targeted by an unfriendly

suitor. Safleddine and Titman (1999) find a positive relationship between levering up and

post-acquisition performance and conclude that leverage forces target management to

make value improvements. In contrast, Jandik and Makhija (2003) argue that debt

entrenches managers and that targets that remain independent subsequently

underperform .

With the structural changes in the legal environment as described in section 4. 1.2

and the ease with which targets are able to deter unwanted acquirers, targets may no

longer need to increase firm leverage. In fact, the evidence in Table 4.5 suggests that

firms make fewer decisions that have effects on their capital structure. To investigate the

impact of hostile offers on target leverage, Table 4.6 reports the mean leverage ratios for

the target firms from two years prior to the offer until two years after the offer.5 Targets

in both periods have very similar debt to assets ratios the two years prior to the offer. In

the year of the offer and the two subsequent years, however, leverage ratios drastically


5 The pattern is very similar when median debt ratios are used and are therefore omitted from the table.










increase for the pre Time-Warner targets, but stay virtually the same for the post Time-

Warner targets.

Figure 4.1 illustrates the changes in leverage for firms with available Compustat

data during the two years prior until two years after the offer. It is apparent that leverage

ratios increase substantially only for the pre-Time-Wamner targets. This evidence suggests

that firms are able to defeat takeover bids without levering up through either defensive

buyback (e.g., greenmail) strategies or issuing debt in the post Time Wamner legal

environment.



4.4 Target Performance

4.4.1 Event Study Analysis

Contested takeover battles are often very complicated and evolve over time. In

this section I examine the market reaction to various events during the contest. Table 4.7

reports the abnormal returns for the targets around several event dates. Abnormal returns

are net-of-market and calculated from one day before the event through one day after. I

focus on five events: the offer date, rejection date, last offer date, last rejection date, and

the resolution date. The offer date is the first date that the bidder made a formal offer.

The average abnormal return around the offer date is 17.6 percent.

By construction, all targets reject and subsequently resist the takeover attempt.6

Many offers are rejected within one day; hence some of the returns on the rejection dates

are difficult to interpret. In order to understand the market reaction of the rej section, I only

report abnormal returns for rejection dates that are at least two days after the initial offer

date. This results in 168 clean observations. The market reaction around the rejection date

6 One rejection date could not be obtained.










is negative suggesting that the market does not view the rej section as a bargaining tool, but

perhaps rather as an entrenchment move. Interestingly, the mean and median rejection

returns are more negative during the post Time-Wamner period, which may be an

indication that the market believes that targets are more likely to successfully defeat the

offer.

The sample contains 101 deals where the original acquirer makes multiple offers.

I define a multiple offer as an offer that is different from the first in terms of price or

structure of the deal (e.g. from stock offer to all cash offer). The offer price was increased

in 74 deals, remained the same in 17 deals, and was decreased for 10 deals. The average

increase between the first and last offer is 9.8 percent, but the mean abnormal returns

around the last offer date is only slightly over 4 percent. The average abnormal return on

the last rejection date is -2.9 percent, and appears to be driven mostly by the latter

sample.

Finally, I report the abnormal returns on the resolution date. On the resolution

date the bidder clearly indicates that its intentions are to withdraw from the takeover

contest. Withdrawals includes cases where the bidder looks for alternatives such as the

acquisition of similar targets, sells stock back to the target (whether it is for a premium or

not), signs a standstill agreement, or lets the tender offer expire without an indication of

renewing. By searching Factiva, I ensured that the resolution date represents the last real

effort of the acquirer. I found 176 resolution dates using these criteria. Resolution dates

were easier to determine during the earlier period, where I identified 113/144 versus

83/126 in the latter period. One of the reasons for this difference is that the reliance on

greenmail, which clearly suggests withdrawal, declined significantly during the latter










period. On average, the market reaction is sharply negative with an average abnormal

return of -7.3 percent. This further indicates the dismay of the market for targets'

management actions in successfully defeating the offer.



4.4.2 Post-Offer Valuation

While markets generally react negatively to news of target rejections and bidder

withdrawals, the positive abnormal returns around the offer date appear not to be entirely

offset. In fact, a quick calculation suggests that the average gain for typical targets is

approximately 8 percent.' To more carefully document the impact of the hostile offer on

target valuation, I calculate two revaluation measures. The first revaluation measure

(REVAL1) is computed as the difference between the target's return and the CRSP

value-weighted return calculated over the period from 20 days prior to the leak date to 10

days after the withdrawal date. The second revaluation measure (REVAL2) is calculated

from 20 days prior to the offer date through 10 days after the withdrawal date.8 Both

measures are calculated with the assumption that all dividends paid out during the

estimation windows are reinvested in the CRSP value-weighted market index. For

thirteen cases, I adjusted the post-offer target prices for maj or self-tender offers for which

CRSP did not make appropriate adjustments of returns to tendering shareholders or cases

where large special dividends are paid.9



SThis calculation assumes only one offer, and uses the average results in Table 7. The combined abnormal
return from the offer, rejection, and withdrawal is equal to 17.6 2.3 7.3 = 8.0 percent.
SThe leak date is the first date for which there were rumors about the firm becoming a potential target. For
the majority of the targets this is also the offer date.
9 The returns for firms with defensive self-tenders are calculated using post-offer prices computed as
follows: [(%/Bought in Tender) x(Tender Price)] + [(1-%/bought in Tender) x(Presole(CRSP,))]. The post-offer
return for firms with special dividends is calculated assuming that special dividends are reinvested at the
30-day T-bill rate.









Table 4.8 reports the mean and median revaluations for the pre and post Time-

Warner samples. Approximately two thirds of the targets experience a positive

revaluation during the takeover contest, and the mean target revaluation is 10.6 percent

for REVAL1 and 6.3 percent for REVAL2. Contrary to the management entrenchment

hypothesis, the revaluations are higher during the post Time-Warner period. Neither

REVAL1 nor REVAL2, however, are statistically different between the two periods at

any conventional significance level.

Before further analysis of the impact of the takeover offer on target valuation, a

few comments are worth making about the positive revaluation of the targets. The

literature offers three potential explanations for the positive revaluation of targets of

hostile takeover attempts. The first argues that the offer discloses positive information

about the target not known to the market prior to the bid. This is sometimes coined the

sitting on a gold mine hypothesis. The second explanation, proposed by Bradley, Desai

and Kim (1983) argues that targets with the greatest revaluation get acquired within a

short period of time and this revaluation reflects future synergies from future

acquisitions. I examine Bradley, Desai and Kim's result by examining the revaluations

for firms that are acquired within one, two, or three years. The results (not reported)

indicate that targets that are acquired shortly after defeating an offer generally have

higher revaluations, but the differences are economically small and statistically

insignificant due to the small sample size. The final explanation for positive target

revaluation is that firms must restructure and recapitalize with more financial leverage on

their own to avoid being acquired and this triggers a positive revaluation (see Safleddine

and Titman (1999)). The effect of leverage on valuation is explored in the next section.









All three explanations suggest that the average post offer performance of the

targets will merely match market returns if markets correctly anticipate the average

tendencies of the targets. Abnormal returns for 250 and 500 days after defeating a hostile

bid are reported in Table 4.8 confirm that post bid returns are not significantly different

from zero. It is interesting to note though that post bid performance is actually higher for

the post Time-Wamner sample. This confirms that the pre-Time-Wamner group does not

outperform the post Time-Wamner sample in ways that were not anticipated by capital

markets at the time of the bid' s defeat.

While it is clear that targets experience a positive revaluation during the takeover

contest, rejecting a takeover offer can also be viewed as an opportunity cost to

shareholders when the price of the target trades at a discount relative to the offer price

after the contest. I define two variables DISCOUNT1 and DISCOUNT2 that measure this

opportunity cost. DISCOUNT1 (DISCOUNT2) is the percentage difference between the

target's post-offer price and the initial (last) offer price. The post-offer price is adjusted

for stock splits and for the thirteen special cases described in footnote 9.



(4.1) DISCOUNT1 = e o
offer


(4.2) DISCOUTNT2 = [Paster ""Iv



Table 4.8 also reports the mean and median numbers for DISCOUNT1 and

DISCOUNT2. Less than one fifth of the targets trade at or above the final offer price ten

days after the resolution date. On average, targets trade at a 14 to 20 percent discount









relative to the offer price subsequent to defeating the offer. Interestingly, while the

takeover environment and the defenses to fend off unwanted acquirers have changed

drastically, there is no significant time variation in the magnitude of the levels of

DISCOUNT and DISCOUNT.



4.4.3 Multivariate Analysis

The previous section showed that there is little difference in post-offer valuations

of takeover targets between the pre and post Time-Warner periods. One interesting

question is what determines these post-offer valuations? This section attempts to answer

this question by examining how target characteristics, bidder characteristics, and the

takeover regime affect the revaluation of targets and the discount targets trade at relative

to the offer price.

It is natural to think that a target's governance structure plays an important role in

the success of takeover offers. Entrenched managers with large holdings or the backing of

large affiliated shareholders have the ability to effectively resist takeover offers since

takeover bids must be approved by the majority of shareholders. In addition to large

shareholdings, staggered board terms can be effective defense mechanism since it would

take the bidder more than one proxy to gain enough board representation necessary to

either remove management or a poison pill. Hence, the variables BLOCK and

STAGGERED are expected to be negatively related to target revaluations and positively

related to target discounts.

In this analysis, I also try to control for the "seriousness" of the offer since many

offers in the sample, especially tender offers, are only for a controlling interest in the










target firm. In some cases, the bidder discusses what will be offered to shareholders in a

second step merger if the offer succeeds. However, most bidders are under no legal

obligation to complete a second step offer. Even back-end offers that are discussed are

often for unspecified securities and/or cash. It is likely that the premiums suggested by

the offering price in these bids are worth less if one considers the blended premium that

targets would receive (shares bought on the front end averaged with shares bought on the

back end). PERCENT SOUGHT captures this variation and is calculated as the

percentage of shares outstanding net of foothold, that the bidder is offering to acquire in

the front-end. For non two-tier offers this variable is set equal to 100 percent. Other

variables that proxy for the seriousness of the bid are BADREP, a Foothold, SYNERGY,

and CASH. a Foothold is the percentage change in bidder's stake from prior to the initial

leak date until after the resolution date. We would expect positive signs on the

coefficients for PERCENT SOUGHT, SYNERGY, a Foothold, and CASH for the

regressions on target revaluation, and the opposite signs for regressions on target

discount.

The emphasis of this study is on the changes in the legal structure that allowed the

use of the poison pill as a defense against unwanted acquirers. Since adoption of a poison

pill only requires board approval and not a shareholder vote, I include a dummy TW,

which equals one for the post Time-Warner period and zero before to allow for the

changes in post-offer target valuations between the two periods. Other important

differences between the two periods are takeover defenses and the use of debt. Previous

research has shown that leverage has a positive effect on returns of unsuccessful targets

(Berger et al. (1997)). I include change of leverage (A Leverage) and an interaction










variable of change of leverage with the TW dummy (A Leverage*TW) to examine the

effect of leverage changes on target valuation during the two periods. A significant

coefficient on the interaction dummy would imply that leverage has a different impact on

target valuation pre and post Time-Warner. Finally, I also include the more precise

breakdown of the leverage changes and include dummies for use of buybacks, greenmail,

and special dividends.

The results of the regressions on REVAL1 and REVAL2 are presented in Table

4.9. Firms with large blocking positions (25% or greater) aff61iated with management

have lower revaluations suggesting that large blockholders either have desires to keep the

firm independent with non-value maximizing (for shareholder) intent or that they are apt

to overvalue the plans of the existing management team (which often includes

themselves). The presence of staggered boards is not a determinant in the revaluation.

Surprisingly, revaluations are no higher when more than one bidder was involved in the

takeover contest.

Bidder characteristics appear to have large effects on target revaluations. Offers

where the bidder has a bad reputation (BADREP) have lower revaluations which may

indicate that the market as well as management Einds such bids less compelling. The

coefficient on Percent Sought is significantly positive at the one percent level, and shows

that markets factor in the uncertainty of the value of the second step and consequently the

seriousness of the offer. Targets in synergistic deals (SYNERGY) experience higher

revaluations as well. This may reflect that these are bids with a real rational that have to

be taken more seriously. CASH and a Foothold do not appear to affect the revaluation

measures.