STRUCTURING INITIAL TAKEOVER OFFERS: MANAGEMENT RESISTANCE, COMPETITION AND BID SUCCESS
Daniel Chersky
A thesis submitted in partial fulfillment of the requirements for the degree of Masters of Commerce (Honours) to the University of New South Wales
2004 CERTIFICATE OF ORIGINALITY
I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, nor material which to a substantial extent has been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis.
I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged
Daniel Chersky
11 ABSTRACT
A current and comprehensive data set of 304 Australian takeover transactions, completed in the eight year period between January 1, 1996 and December 31, 2003, is employed in this thesis to address the question of how initial bidders can structure their offer so as to maximise, or at least increase, the likelihood of a profitable acquisition. An initial bidder will go a long way to achieving this aim if two transaction-specific hurdles are successfully negotiated: the securing of target management support and the deterrence of potential rival bidders from entering the takeover contest.
Theoretical and empirical literature suggests that a bidder can increase his chances of achieving these two goals, as well as the ultimate goal of a successful acquisition, via the deliberate choice of bid structure. Using Logit and Bivariate Probit models, this thesis provides a comprehensive empirical test of this claim, focusing on the key bid structure elements of bid premium, medium of exchange, and bidder toehold.
We find that higher-premium bids are associated with a lower likelihood of target management resistance and an increased probability of bid success, however, they do not appear to deter competing bidders. Neither the use of cash in the offer, nor the cost of acquiring information about a target, are correlated with the frequency of competition, resistance or bid success. However, ownership of large toeholds by initial bidders serves to deter competing bidders, reduce the likelihood of resistance and increase the probability of bid success.
We also confirm the importance of target management support in takeovers by finding that in transactions where target management resist the initial bid, the likelihood of a competing offer emerging is significantly higher, while the probability of initial bid success is reduced significantly.
iii ACKNOWLEDGMENTS
Completion of this thesis would not have been conceivable without the enormous contribution of my supervisor, Dr. Sian Owen, whose enthusiasm and 'aim high' attitude motivated me to produce my best work. I am extremely thankful for her commitment to this thesis and the responsiveness and generosity with which she gave up large amounts of her time. I learnt a great deal from her expertise, whilst her accessibility and light-hearted nature made enjoyable what can often be stressful times. Needless to say, all errors are mine.
I am also grateful to the helpful people at Bloomberg LP and Rubicon Partners for providing data used in the empirical analysis in this thesis. Additionally, I would like to thank the relevant people at JP Morgan, whose understanding was essential to the completion of this research.
Finally, since this thesis is a culmination of nearly six years at UNSW, I would like to acknowledge those people who have provided me with tremendous support during this time. I am forever grateful to my parents for their ever-present love and constant support, both emotional and financial, without which any personal achievements would be unattainable. To Natalie, I rely on you greatly and you have never let me down over many years. I appreciate your unwavering support enormously and, needless to say, any successes are partly yours. To the rest of my family, my second family, the Fletchers, and friends too numerous to name, thanks for always being there and I look forward to celebrating bigger and better things in our lives.
IV TABLE OF CONTENTS
Certificate of Originality ...... ii Abstract ...... iii Acknowledgments ...... iv Table of Contents ...... v List of Tables ...... ix
CHAPTER 1 INTRODUCTION ...... 1
1.1 INTRODUCTION ...... 1
1.2 MOTIVATION AND CONTRIBUTION ...... 3
1.3 THESIS OVERVIEW ...... 5
CHAPTER 2 LITERATURE REVIEW ...... 6
2.1 INTRODUCTION TOTAKEOVER LITERATURE ...... 6
2.2 CYCLICAL NATURE OFTAKEOVER ACTIVITY ...... 7
2.2.1 Merger Waves ...... 7
2.3 MOTIVATION BEHIND TAKEOVER ACTIVITY ...... •...... 11
2.3.1 Synergy ...... 12
2.3.2 Target Undervaluation ...... 13
2.3.3 Disciplinary Theory ...... 14
2.3.4 Management Hubris Theory ...... 15
2.3.5 Free Cashflow Theory ...... 16
2.4 PERFORMANCE AND RETURNS ISSUES ...... •...... 17
2.4.1 Pre-Takeover Performance ...... 18
2.4.1.1 Characteristics of Target Firms ...... 18
2.4.1.2 Characteristics of Acquirer Firms ...... 19
2.4.2 Announcement Period Returns ...... 19
2.4.2.1 Target Firm Shareholders ...... 19
2.4.2.2 Bidder Firm Shareholders...... 20
2.4.2.3 Do Takeovers Create or Redistribute Wealth? ...... 21
2.4.3 Post·Takeover Performance ...... 22
2.4.3.1 Share Price Performance ...... 22
2.4.3.2 Operating Performance ...... 23
V 2.5 BID STRUCTURE AND TRANSACTION OUTCOMES ...... 24
2.5.1 Elements of Bid Structure ...... 25
2.5.1.1 Bid Premium Decision ...... 25
2.5.1.2 Payment Method Decision ...... 26
2.5.1.3 Toehold Acquisition Decision ...... 28
2.5.2 Transaction Outcomes ...... 29
2.5.2.1 Target Management Response ...... 29
2.5.2.2 Bid Competition ...... 32
2.5.2.3 Bid Success ...... 33
2.5.3 Impact of Bid Structure on Transaction Outcomes ...... 34
2.5.3.1 Relationship between Bid Structure Elements and Bid Competition ...... 35
2.5.3.1.1 Bid Premium and Competition ...... 35
2.5.3.1.2 Payment Method and Competition ...... 37
2.5.3.1.3 Toehold and Competition ...... 39
2.5.3.2 Relationship between Bid Structure Elements and Target Management Resistance ...... 40
2.5.3.2.1 Bid Premium and Target Management Resistance ...... 40
2.5.3.2.2 Payment Method and Target Management Resistance ...... 41
2.5.3.2.3 Toehold and Target Management Resistance ...... 42
2.5.3.3 Relationship between Bid Structure Elements and Bid Success ...... 43
2.5.3.3.1 Bid Premium and Bid Success ...... 43
2.5.3.3.2 Payment Method and Bid Success ...... 45
2.5.3.3.3 Toehold and Bid Success ...... 45
2.6 LITERATURE REVIEW CONCLUSION ...... 46
CHAPTER 3 THE AUSTRALIAN REGULA TORY ENVIRONMENT ...... 48
3.1 ACQUISITION OF A TOEHOLD STAKE ...... 48
3.2 MAKING A MERGER OR ACQUISITION PROPOSAL...... 49
3.3 DIRECTORS' RECOMMENDATIONS TO TARGET SHAREHOLDERS ...... 49
CHAPTER 4 HYPOTHESES DEVELOPMENT ...... 52
4.1 COMPETITION HYPOTHESES ...... 52
4.2 MANAGEMENT RESISTANCE HYPOTHESES ...... 56
4.3 BID SUCCESS HYPOTHESES ...... 58
vi CHAPTER 5 DATA ...... 61
5.1 DATA SOURCES ...... 61
5.2 SAMPLE CRITERIA ...... 62
5.3 DATA CHARACTERISTICS ...... 64
5.3.1 Unprocessed Data ...... 64
5.3.1.1 Takeover Mode ...... 64
5.3.1.2 Bid Outcome ...... 64
5.3.1.3 Directors' Recommendations to Shareholders and Target Management Resistance ...... 65
5.3.1.4 Competition Status ...... 67
5.3.1.5 Offer Type and Terms ...... 67
5.3.1.6 Bidder Toehold ...... 68
5.3.2 Processed Variable Definitions ...... 68
5.3.2.1 Bid Premium ...... 69
5.3.2.1.1 Target's pre-bid value ...... 69
5.3.2.1.2 Offer Value ...... 71
5.3.2.1.3 Bid Premium Calculation ...... 72
5.3.2.2 Medium of Exchange ...... 73
5.3.2.3 Target Size ...... 73
5.3.3 Descriptive Statistics ...... 74
CHAPTER 6 METHODOLOGY ...... 76
6.1 LOGISTIC REGRESSION ...... •...... 76
6.2 PR OBIT REGRESSION ...... 80
6.3 SIMULTANEOUS EQUATION ANALYSIS - BIVARIATE PROBIT...... 81
CHAPTER 7 EMPIRICAL TESTS AND RESULTS ...... 83
7 .1 COMPETING BIDS ...... 83
7.2 TARGET MANAGEMENT RESISTANCE ...... 89
7.3 BID SUCCESS ...... 95
7.4 SIMULTANEOUS DECISION ANALYSIS ...... 98
7.4.1 Simultaneous Determination of the Resistance and Competition Decisions ...... 98
7.4.2 Simultaneous Determination of the Resistance Decision and Initial Bid Outcome ...... 101
Vll CHAP'I'ER 8 CONCLUSION ...... 103
8.1 SUMMARY OF RESULTS ...••....•...•..•...... •...... •..•...•...... •...... •...... •.•...... •... 103
8.2 LIMITATIONS AND SUGGESTIONS FOR FURTHER RESEARCH •...... 106
RE FEREN CES ...... 109
APPENDIX A - TABLE A.1 ...... 117
APPENDIX B ROBUSTNESS TESTS ...... 123
Bl INTRODUCTION ••••.•..••••.••••••••••.•.••.••...... •...... ••..•.••.•.•..•.•.•.••••••..••.•.••••••.•..••••••. 123
B2 COMPETING BIDS ·••••••·••·•·•·•·•••············•··•··••·········· .•...... •.•.••••••••••.••••.•.•.••.•.•...... ••...... •....•...... 123
B3 TARGET MANAGEMENT RESISTANCE ...... •••.••...... •..•.••••.•••.••••••.••.•.••••.•.•••.•.•••••.••.•.••.• 125
B4 BID SUCCESS ...... •...... •.•.•.•...... •...•.••.•..••.••••••.•.••••••.•••••...•.••.••.•.•...... 127
APPENDIX C ANALYSIS OF NON-INITIAL BIDS ...... 129
Cl INTRODUCTION ••...... ••.••....•..•..••..•...••.••..•.•.•.•••..•.••••.•.••..•...... •.•••.•.••.•...••.••.•.•..•.•..... 129
C2 DATA •..••••••••.•.••••••.•.••••••.•.•.••.•.••.••••••.•••••••••.•.•.•••••••...... •...... •...... •.....•.•.••••.••..•...... ••...... 129
C2.1 Competition Status ...... 130
C2.2 Bid Outcome ...... 130
C2.3 Bidder Toehold ...... 131
C3 METHODOLOGY ...... •..•.•.....••.•..•.•...... •...... •...... •..•.....••.•••.••••••. 133
C3.1 Bid Premium Calculation ...... 133
C3.1.1 Target's pre-bid value ...... 133
C3.l.2 Offer Value ...... 134
C4 RESULTS ...... •...... •.•..•..•...•.•..•...•...... ••.....•..•...... •...••...... •...... 136
C4.1 Competing Bids...... 136
C4.2 Target Management Resistance ...... 139
C4.3 Bid Success ...... 140
CS CONCLUSION ...... •....•.•..•.••••.•....•.•...•..•...... •...... ••.•.•...... ••.• 142
Vlll LIST OF TABLES
Table 5.1: Descriptive Statistics of 'Initial' Bids Data Set ...... 75 Table 7.1: Binary Logistic Regression - Likelihood of Bid Competition ...... 84 Table 7.2: Trinomial Logistic Regression - Likelihood of Target Resistance ...... 90 Table 7.3: Binary Logistic Regression - Likelihood of Bid Success ...... 96 Table 7 .4: Bivariate Probit Regression - Likelihood of Target Resistance and Bid Competition ...... 99 Table 7.5: Bivariate Probit Regression - Likelihood of Target Resistance and Bid Success ...... 102 Table A.1: Identity of Target and Initial Bidder Companies ...... 117 Table B.1: Separate Binary Logistic Regression Analysis - Likelihood of Bid Competition ...... 124 Table B.2: Separate Trinomial Logistic Regression Analysis - Likelihood of Target Resistance ...... 125 Table B.3: Separate Binary Logistic Regression Analysis - Likelihood of Bid Success ...... 127 Table C.1: Descriptive Statistics of 'Expanded' Data Set - Initial and Non-Initial Bids ...... 132 Table C.2: Binary Logistic Regression on 'Expanded' Data - Likelihood of Bid Competition ...... 137 Table C.3: Trinomial Logistic Regression on 'Expanded' Data - Likelihood of Target Resistance ...... 139 Table C.4: Binary Logistic Regression on 'Expanded Data' - Likelihood of Bid Success ...... 141
lX Chapter 1 Introduction
1.1 Introduction
Global Merger and Acquisition activity has increased dramatically in the last decade, both in volume and size of transactions. For example, the value of completed M&A transactions in the Australian market in 1993 totalled A$4.1 billion1. In 2003, the value of Australian M&A deals soared to A$58.6 billion, representing a fourteen fold increase in just the last decade2• In the U.S. market, M&A activity peaked in
1999 with a massive US$1,500 billion worth of transactions completed, an increase of over 600% on the 1994 total (McDonald, 2003).
Likewise, the amount of attention devoted to these transactions by the financial and popular press has increased significantly. The elevated profile enjoyed by the M&A market is partly due to the steep rise in the level of stock ownership amongst individuals during this period. This is evident in Australia, where approximately 5.5 million individuals, or 37% of the adult population, now own shares directly. This rises to approximately 50% when share ownership through a mutual fund is included3• Hence, M&A transactions no longer simply represent the transfer of corporate assets between institutions, but rather affect the wealth of millions of individual shareholders.
1 Mergers & Acquisitions Index, Ernst & Young Corporate Finance (2002) 2 Thomson Financial Research, 6 January 2004 3 ASX Share Ownership Survey, 5 February 2003. 1 In terms of organisation, only a minority of merger and acquisition transactions are pre-negotiated between both sets of senior management4• Hence, the vast majority of these approaches are largely unsolicited and require bidder management to construct an offer with which to approach target management and shareholders. This being the case, the following question arises: How can a potential acquirer structure his initial bid so as to maximise, or at least increase, the likelihood of completing a profitable acquisition?
The evidence from the market suggests that for an initial bidder, target management resistance and the presence of competing bidders significantly reduces the chances of successful acquisition (Walkling (1985) and O'Sullivan and Wong (1998)). Even if the bidder is able to overcome management resistance and acquire the target, this will usually take place at an increased price via a revised bid. Likewise, an initial bidder may acquire the target in the face of competition from rival bidders, but this will certainly be at an increased price, potentially inflated by multiple rounds of counter-bidding. Hence, the most desirable outcome sought by a firm launching an initial takeover offer is initial bid success. As the above comments suggest, the securing of target management support and the deterrence of potential rival bidders are key to achieving this aim.
Extant literature suggests that a bidder's strategic choice of offer structure can increase his chances of achieving these three outcomes, including the ultimate goal of a successful acquisition. This thesis tests this claim via a comprehensive empirical
4 Only 20% of the transactions examined in this thesis were organised via a Scheme of Arrangement. Pompilio (2000), examining 249 Australian transactions between 1992-1998, reported that only 11 % of these transactions were organised via Schemes. 2 analysis. The key bid structure elements of bid premium, medium of exchange, and bidder toehold are focused on to determine how changes in the size of these variables affect the likelihood of observing target management resistance, encountering competition from rival bidders, and ultimately enjoying bid success.
1.2 Motivation and Contribution
The next chapter illustrates that a vast body of literature has been devoted to the study of past levels of takeover activity, potential reasons speculated to motivate takeover activity, the wealth effects of takeovers, and the impact of takeovers on corporate performance. Whilst extremely relevant and interesting, these areas of research tend to examine takeovers ex-post and thus are of limited use to corporate managers and M&A practitioners. Of more interest to these groups, once the decision to pursue acquisitions is made, is the formulation of an effective acquisition strategy. A key part of this strategy is the bid structure with which one approaches a takeover target or potential merger partner. However, it is interesting, and somewhat surprising, that the amount of literature devoted to studying the impact of bid structure on the takeover process and outcome is dwarfed by that related to each of the areas mentioned above. This is not just the trend in Australia but is also the case globally. Hence, this thesis adds to, and extends, this largely neglected and very important area of corporate finance research. Additionally, it is hoped that this study will be of interest to M&A practitioners and corporate managers, thereby strengthening the link between academia and the market.
This thesis utilises a comprehensive data set of 304 Australian takeover transactions taking place over an eight year period. This study will add significantly to the
3 available Australian evidence, which to date is provided by a total of three empirical studies. Furthermore, these studies fail to address the impact of bid structure on offer outcome, an issue examined empirically in this thesis. Such an analysis was beyond the scope of the three Australian studies since each of them only considered the impact of bid structure on one of the three areas of interest mentioned above5•
Finally, the data set utilised in this study is the most current of all the available evidence, Australian and international, with the eight year period examined ending in
December 2003.
Methodologically, this thesis introduces an important innovation. Following Henry
(2004), this study uses Directors' Recommendation to Shareholders as a direct proxy for target management's response. Whilst most recommendations are in the form of
'Reject the offer' or 'Accept the offer', on rare occasions, other less explicit responses such as 'Take No Action at this Time' or 'No Recommendation' are made.
These are normally interpreted by the market as an admission that the takeover bid is attractive to target shareholders, but perhaps not to managers (Henry, 2004 ).
Therefore, in an innovation of this study, all such recommendations will be grouped into a third category, in which the response of target management is considered
Neutral. It is expected that this three-way classification of target management response, facilitated by a trinomial logistic regression, will provide a more precise proxy for the presence, or absence, of Target Management Resistance than has been used in past studies.
5 Eddey and Casey (1989) and Henry (2004) considered the impact of bid structure on the likelihood of observing target management resistance, while Pompilio (2000) examined the impact of bid structure on the likelihood of encountering bid competition.
4 1.3 Thesis Overview
The remainder of this thesis proceeds as follows. The next chapter reviews the extant literature related to takeovers and, more specifically, the empirical analysis in this thesis. Chapter Three introduces some specific aspects of the Australian regulatory environment as they relate to takeovers, knowledge of which is necessary to understand the analysis in this thesis. Chapter Four develops the hypotheses to be tested. Chapter Five introduces the data upon which the empirical testing in this thesis is carried out and the processes followed to arrive at the final data set. Chapter
Six describes the econometric techniques used to test whether our data supports our hypotheses, while Chapter Seven provides a comprehensive discussion of the results.
The thesis is concluded with Chapter Eight.
5 Chapter 2 Literature Review
2.1 Introduction to Takeover Literature
Academic literature on corporate mergers and acquisitions can be traced back to the early 1900s. However, as the size, sophistication, and public profile of the market for corporate control has increased over the decades, so too has the amount of literature devoted to its examination. The aim of this chapter is to introduce the main themes in the takeover literature.
Beginning with the next section, takeover activity is placed in a historical context by examining past periods of intense takeover activity, known as merger waves. Section
2.3 discusses the main motivations or rationales put forward to justify and explain the existence of acquisitive activity, whilst the vast amount of literature examining performance and returns issues related to acquisitions will be reviewed in section
2.4. Some of the issues covered in this section will be: the characteristics of firms involved in takeover activity, the winners and losers from takeovers, and the impact of these transactions on operating efficiency and share price performance of the combined firm. The literature reviewed in section 2.5 considers some of the decisions facing bidders in structuring their offer and the transaction-specific hurdles that must be negotiated before they can enjoy bid success. In particular, sub-section
2.5.3 contains a review of literature directly related to the empirical analysis in this thesis.
6 2.2 Cyclical Nature of Takeover Activity
2.2.1 Merger Waves
Corporate takeover activity, like many other economic phenomena, is not stable over time but rather occurs in waves. Periods of high takeover activity are most commonly referred to as Merger Waves and represent an area of considerable interest for corporate finance researchers. The vast majority of literature on this subject examines merger waves in the context of the American takeover market.
Consequently, the following discussion is primarily focused on this market.
It is generally accepted that there have been five takeover waves in the 20th century.
Following is a sequential discussion of these waves, the circumstances in which they occurred and the theories posited to explain their occurrence.
1900's: The Monopoly Wave
The first U.S. takeover wave began at the close of the 19th century and ended in
1904. This wave was primarily a product of the paradoxical Sherman Antitrust Act, which allowed takeovers to result in the formation of near monopolies, and was further fostered by a rising stockmarket. The wave came to an end in 1904 when the courts vastly strengthened the Sherman Antitrust Act, leading to significant limitations on horizontal mergers (Shleifer and Vishny, 1991).
7 1920's: The Oligopoly Wave
The second takeover wave took place in the 1920s and resembled the first wave in many aspects. Although takeover regulations were tightened to prohibit the creation of monopolies, the courts did allow oligopolies. Acquirers were again assisted by a rising stockmarket, allowing them to issue equity on favorable terms to finance their takeovers. The second wave was halted by the beginning of the Great Depression, and the resultant collapse of the stockmarket, which saw takeover activity ground to an almost complete halt by the early 1930s (Shleifer and Vishny, 1991).
1960's: The Conglomerate Wave
The level of takeover activity in the 1940s and 1950s moved up and down in a manner typical of many economic phenomena, however, the next takeover wave did not emerge until the second half of the 1960s. In the U.S. it coincided with a period of sustained economic prosperity and stockmarket strength. Unsurprisingly, cash rich U.S. firms in this environment turned to takeovers as a means of growth.
However, due to the Celler-Kefauver Act of 1950, which clamped down vigorously on takeovers designed to increase market power, mergers between firms in the same industry became the exception rather than the rule in the 1960s, leading to the formation of numerous conglomerates (Lev, 1992).
Despite the various reasons put forward to explain conglomerates, perhaps the most obvious reason is also the most valid. That is, takeover regulations effectively prohibited U.S. firms, which prospered in the 1960s, from merging with other firms
8 in their industry. Cash-rich and facing a rising stockmarket, managers of these firms were intent on pursuing acquisitive growth and did so in the only manner available to them; takeovers and mergers with firms in unrelated industries (Shleifer and
Vishny, 1991). This scenario seems to explain a significant proportion of takeover activity in all countries since legislators began to take takeover regulation more seriously.
1980's: The Leveraged Hostile Wave
The fourth takeover wave took place in the 1980s. This wave was not only large in the volume of transactions but was unprecedented in terms of the size of deals. Firms that were once considered untouchable, due to their large size, were regularly becoming the targets of hostile takeovers (Jarrell, Brickley, and Netter, 1988). The
1980s takeover wave has attracted the largest amount of comment from academic researchers, resulting in numerous reasons being put forward to explain both the intensity of takeover activity and size of takeover transactions in the 1980s.
Periods of takeover activity at least partly reflect the takeover legislation prevailing at that time. For example, in the 1980s the U.S. government relaxed restrictions on takeovers imposed by antitrust laws, which undoubtedly contributed to the 1980s takeover wave (Jensen, 1992).
Shleifer and Vishny (1991) proposed that the 1980s takeover wave represented a return to specialization after the failed experimentation with conglomerates in the
1960s. Indeed, the prominence in the 1980s of horizontal mergers and 'bust-ups',
9 which saw a conglomerate taken over and its unrelated divisions sold off to specialist buyers, is consistent with this argument.
Adapting a theory first posited by Gort (1969), others suggests that the takeover wave of the 1980s partly reflected an adaptation of industry to broad-based industry shocks such as deregulation, changes in input costs, such as oil, and technological innovation. It is proposed that takeovers often provide the least-cost means for an industry to restructure in response to such shocks and that at least part of the 1980s takeover activity facilitated such restructuring (Mitchell and Mulherin, 1996)6.
A more traditional motivation put forward to explain the 1980s takeover wave is that of management inefficiency. Jensen (1992) suggested that ineffective internal governance mechanisms and weak management incentives lead to corporate mismanagement in the 1970s. Proponents of this explanation suggest that, consistent with the disciplinary theory of takeovers, the market for corporate control took over in the 1980s, armed with a combination of favourable regulatory changes, effective takeover methods and financing innovations.
Although hostile tender offers and the use of leverage to finance takeover transactions was not an innovation of the 1980s, the frequency and aggression with which these takeover methods were combined in the 1980s was unprecedented. For example, it is estimated that almost 50% of all major U.S. companies received
'hostile' (unsolicited) takeover bids in the 1980s (Holstrom and Kaplan, 2001). The
6 U.S. industries which experienced significant shocks in the 1980s include financial services, oil and gas, and broadcasting. 10 original issuance of high-yield non-investment-grade bonds, also known as 'junk bonds', to finance takeover transactions only added to this leveraged buyout (LBO) phenomenon. This innovative source of financing, which effectively removed size as a barrier in takeovers, opened the way for bidders to acquire much larger targets, partly explaining the unprecedented size of transactions in the 1980s (Jensen, 1988).
1990's: The Mega-Merger Wave
The most recent takeover wave took place in the 1990s and was particularly intense in the second half of that decade. This wave is the largest to date, eclipsing previous waves both in the number and value of deals. Although academic literature on this takeover wave is as yet not plentiful, several things are known about the U.S. experience. Firstly, hostile takeovers virtually disappeared, with only 4% of U.S. takeovers in the 1990s classified as hostile. Secondly, acquirers turned overwhelmingly to stock as a method of payment. The use of stock was approximately 50% more prevalent than in the 1980s. Finally, the 1990s, consistent with all previous merger waves with the exception of the 1960s, were dominated by same-industry mergers (Andrade, Mitchell, and Stafford, 2001).
2.3 Motivation behind takeover activity
It is important to understand a potential acquirer's motivation in initiating the takeover as this may influence the nature and structure of the bid. Furthermore, the motivation as perceived by target management will influence their response to the offer. Combined, these two factors will go a long way to determining the success, or
11 otherwise, of the proposal. Theories of takeover motivation fall into two categories: value maximising theories, where bidder management seek to maximise shareholder value via takeovers, and non-value maximising theories, where reasons other than shareholder value maximisation motivate takeovers. The most prominent of these theories are discussed below and each is relevant in explaining at least some of the observed takeover activity.
2.3.1 Synergy
Perhaps the most popular rationale put forward to justify mergers and acquisitions, by both practitioners and academics, is synergy. It is proposed that real economic gains can be created through the combination of target and acquirer firms, particularly when these firms operate in related markets or industries (Jensen and
Ruback, 1983). Synergistic gains are often categorised as being financial or operating in nature. Examples of financial synergies are tax benefits and increased debt capacity, while operating synergies can accrue via economies of scale, leading to cost reductions, or a consolidation of market power (Lev, 1992).
Despite its popular use by corporate managers as a justification for merger, empirical research on the subject is not as conclusive. Bradley, Desai, and Kim (1983, 1988) found support for synergy as a rationale for takeovers by examining the returns to target and acquirer firms involved in both successful and unsuccessful tender offers.
However, perhaps the largest barrier to the wider acceptance of synergy as a motive and valid rationale for takeovers lies in the seeming inability of firms to take advantage of supposed synergies post-takeover. As section 2.4.3 discusses, empirical research has failed to conclusively show that combined firms experience increases in
12 efficiency or profitability post-merger, which somewhat undermines the validity of synergy as a takeover rationale.
2.3.2 Target Undervaluation
Target undervaluation as a motive for takeover is based on the simple notion that the target represents a bargain at its current market value. However, this theory can be sub-divided according to the nature of the undervaluation. Firstly, some propose that such bargains exist due to inefficiencies in stockmarket pricing, which contradicts the central notion of market efficiency (Scherer, 1988). An example of this is if large differences exist between firms' market values and the replacement value of their assets (Lev, 1992).
The second source of target undervaluation put forward as a motive for takeovers relates to 'specialist information'7• In this instance it is claimed that the managers of bidder firms have specialist information that the market does not have, which allows them to identify that the target is undervalued. This may involve the knowledge of more efficient operating strategies unknown to incumbent management, implementation of which would lead to an increase in firm value (Halpern, 1983). To the extent that bidders possess such information, the target undervaluation theory represents a legitimate motive for acquisition, consistent with the principle of shareholder value maximisation. In a U.K. study, Limmack (1994) documented results consistent with the target undervaluation motive by finding that the gains to
7 Specialist information in this context relates to information which may become apparent to industry specialists due to their expertise and industry experience. It is not to be confused with 'Inside Information', acting on which is illegal in the Australian market, as well as in developed markets globally.
13 targets m abandoned bids did not disappear even when further bids were not forthcoming. He interprets this as an indication that some takeover bids in his sample revealed target undervaluation, which was subsequently corrected due to the information revealed by these abandoned bids.
2.3.3 Disciplinary Theory
The Disciplinary Theory proposes that takeovers are motivated by the replacement of inefficient target-firm management via transfers in the market for corporate control (Jensen, 1988). The concept of a market for corporate control was first introduced by Manne (1965), who suggested that corporate control, like other assets, is likely to attract buyers in the marketplace if its price falls below its fair value. If a company's share price is sufficiently depressed due to inefficient management, other management teams may seek to gain control of that company via a takeover bid.
From the bidder's point of view, the rationale is simple: gains can be achieved by purchasing a company which is underperforming due to management inefficiency and replacing the incumbent management team with their own, more efficient, managers. To the extent that such action can generate real economic gains, the disciplinary theory presents a value-maximising motive for takeovers.
Supporters of the disciplinary theory argue that takeovers provide the ultimate incentive for management to perform efficiently and deliver added protection for minority shareholders (Grossman and Hart (1980), and Rappaport (1990)).
It is to be expected that if takeovers serve a disciplinary function for inefficient management then targets of disciplinary (hostile) takeovers should exhibit inferior 14 pre-takeover performance and be followed by a higher degree of management turnover when compared to targets of non-disciplinary (friendly) takeovers and non target firms. However, the empirical results are mixed. While it is generally accepted that targets of disciplinary (hostile) takeovers experience higher levels of post takeover management turnover than targets of non-disciplinary (friendly) takeovers
(Martin and McConnell (1991) and Agrawal and Walkling (1994)), the findings in relation to pre-takeover performance of targets are less consistent with the disciplinary theory. Franks and Mayer (1996) found that 'targets of hostile takeovers are performing neither poorly nor worse than the targets of friendly bids' (p.177), while Agrawal and Jaffe (2003) similarly fail to find evidence of pre-acquisition under-performance by targets in their comprehensive sample, whether measured by operating or stock returns. These findings seem to reject the basis of the disciplinary theory.
2.3.4 Management Hubris Theory
The influential paper by Roll (1986) proposed a hubris theory of corporate takeovers under which bidding firm managers systematically overpay for targets. Put simply, the rationale for takeover from the point of view of these managers is based on their belief that they are getting a bargain, whereas in fact they are just making overvaluation errors due to their excessive self-confidence.
Whereas Roll's theory proposes that managers try to maximise shareholder value but make valuation errors unconsciously, a more cynical view exists, in which takeovers are driven by the self-interest of acquiring firm managers (Malatesta (1983)). This self-interest sees managers engage in empire-building, where they seek to expand the
15 scope of their control and increase their status, remuneration and importance to the firm. This version of the management hubris theory suggests that managers do not unconsciously overpay for targets but rather do so willingly in the pursuit of these personal benefits (Shleifer and Vishny, 1988). Thus, while Roll's original theory is somewhat consistent with value-maximising takeover motives, this version of the management hubris theory is not. Whilst empirical work testing the hubris theory has been limited, Morck, Shleifer, and Vishny (1990) did find evidence consistent with managerial self-interest driving poor acquisitions.
2.3.5 Free Cashflow Theory
Jensen's Free Cashflow Theory of takeovers is based on the agency relationship inherent in public corporations, which may cause managers to act against the best interests of shareholders. Jensen (1986) proposes that the agency issues with respect to one particular management decision, the dividend payout decision, motivates some of the observed merger and takeover activity. More specifically, the reluctance of management to pay out free cashflow8 to shareholders, in the form of dividends, leads two distinct types of corporate managers to undertake mergers and acquisitions, with two distinct motivations.
The first is a cash-hoarding manager operating in a firm with ineffectual internal control processes. This manager refuses to pay out free cashflow for self-interest reasons, perhaps motivated by empire-building, increased status or remuneration, instead using the free cashflow to fund negative net present-value projects, including takeovers. Jensen's theory predicts that such managers will engage in value-
8 Jensen (1988) defines free cashflow as 'cashflow in excess of that required to fund all of a firm's projects that have positive net present values when discounted at the relevant cost of capital' (p.28). 16 decreasing takeovers, typical of diversifying takeovers. The second type of manager seeks to identify and combine with cash-hoarding firms of the type described above.
The bidding manager in this case attempts to maximise efficiency and shareholder value by investing the target's free cash flow in positive net present-value projects of the combined entity and returning any remaining free cashflow to shareholders.
Jensen predicts that such takeovers will be value-increasing and are likely to be hostile in nature. Hence, the free cash flow theory suggests that the conflicts between managers and shareholders over the payout of free cashflow gives rise to two distinct takeover motives, only one of which is consistent with the principle of shareholder value maximisation. Harford (1999), in a comprehensive U.S. study, provides some support for the free cashflow theory by finding that cash-rich firms are more likely to initiate value-decreasing transactions, as measured by the announcement-period returns and long-run operating performance. This is consistent with one of Jensen's central predictions.
2.4 Performance and Returns Issues
Sub-section 2.4.1 deals with some of the characteristics typical of target and bidders firms, demonstrating the difficulty faced by targets in fighting off approaches from larger and better-performing firms. Sub-section 2.4.2 shows that target shareholders are the big winners in takeovers. The abnormal returns they experience reflect the bid premium, the size of which will drive the response of target management, the decision of other bidders to enter the contest, and ultimately bid success. Returns to bidders are far less spectacular and are shown to partially reflect the payment method decision and the signal it conveys to the market. The literature presented in
17 subsection 2.4.3 shows that, overall, performance of combined firms, both stock and operating, does not improve significantly post-takoever, providing some justification for the high levels of target management resistance observed.
2.4.1 Pre-Takeover Performance
2.4.1.1 Characteristics of Target Firms
A large body of literature has been devoted to the study of takeover targets and their characteristics. At issue is whether takeover targets are fundamentally different to other firms. A majority of recent evidence suggests that, consistent with the disciplinary and target-undervaluation theories, takeover targets are less profitable and have lower market-to-book ratios than other firms (Nuttal (1999) and Trimbath
(2001)). However, the evidence with respect to share price performance contradicts these findings, with the most recent studies, such as those by Franks and Mayer
(1996) and Agrawal and Jaffe (2003), unable to conclude that takeover targets under perform compared to matched non-target firms 9• Perhaps the most distinguishable feature of takeover targets is their size, with the vast majority of studies finding that takeover targets tend to be smaller than other firms (Morck, Shleifer, and Vishny
(1989) and Billett (1996)). The findings with respect to other characteristics such as capital structure, asset structure and age have been inconclusive as to how they affect a firm's likelihood of becoming a takeover target.
9 In both these studies, target firms were matched with non-target firms according to size, as measured by market value of equity, and industry of operation. 18 2.4.1.2 Characteristics of Acquirer Firms
Compared to target firms, the characteristics of acquirer firms have been less well explored. It is generally agreed that, at the time of a takeover bid, acquiring firms are larger in size and demonstrate superior long-run abnormal stock returns when compared to other firms (Lev, 1992). Larger firms obviously have greater access to external capital markets, as well as a larger internal pool of resources, and as such are more able to fund a takeover. Additionally, a higher share price effectively reduces the cost of the takeover in the case of scrip bids, since less of the acquirer's shares need to be offered than would otherwise be the case.
2.4.2 Announcement Period Returns
2.4.2.1 Target Firm Shareholders
Since takeover offers usually incorporate substantial premiums to the current value of the target's shares, it is unsurprising that target-firm shareholders are the big winners in takeover transactions. This fact is confirmed by the empirical literature, which abounds with evidence of target shareholders earning substantial positive abnormal returns, ranging from 13% to 34% in the months surrounding the takeover
(Jensen and Ruback (1983), Servaes (1991) and Andrade, Mitchell, and Stafford
(2001)). Australian studies present evidence consistent with the above international results, with cumulative abnormal returns (CAR) for successful takeover targets reported to be between 17% and 23% (Bugeja and Walter (1995), and da Silva Rosa et al. (2000)). However, one might expect these gains to dissipate in the case of unsuccessful transactions, where the target remains independent. Yet the empirical
19 evidence shows that target shareholders are sometimes able to retain a significant proportion of the abnormal returns even in unsuccessful transactions (Bradley (1980) and Bishop et al. (1987)).
2.4.2.2 Bidder Firm Shareholders
Whereas target-firm shareholders realise almost immediate gains at the announcement of a takeover offer, largely representative of the premium offered by the bidder, any gains to bidding-firm shareholders are far less certain. Following the announcement of a takeover bid, the adjustment in the bidding firm's share price will reflect the market's view of the proposed takeover. Hence, it is not uncommon for bidder-firm shareholders to earn significantly negative abnormal returns, since far from all takeovers represent sound business decisions likely to result in increased shareholder wealth. The empirical evidence shows that, on average, bidder firms earn either normal rates of return or modest abnormal returns in the months surrounding the takeover. For example, Jarrell and Poulsen (1987) reported abnormal returns of -1.10% for successful bidders by tender offer, while Andrade, Mitchell, and Stafford (2001) found abnormal returns of -3.9% over the period beginning 20 days before the announcement date to the deal closing date. Similar results, showing modest yet significantly negative abnormal returns, are presented for Australia by
Bellamy and Lewin (1992) and da Silva Rosa et al. (2000).
Travlos (1987), in an influential study, showed that method of payment is also significant in explaining returns to bidder firms, finding that stock bidders in his sample earned an average abnormal return of -1.6%, whilst cash bidders earned
20 normal (zero) rates of return10• Results consistent with this relationship between abnormal returns and method of payment were also found by Servaes (1991), as well as Bellamy and Lewin (1992) and da Silva Rosa et al. (2000) in the Australian market. These results are consistent with the Myers and Majluf (1984) hypothesis that stock offers elicit a signal that bidders believe their shares are overvalued.
In summation, the empirical evidence shows that, on average, bidder firms earn either normal rates of return or modest abnormal returns, which can be either positive or negative depending on the nature of the transaction and the method of payment. Hence, it is clear that for bidder-firm shareholders takeovers do not deliver the windfall experienced by target shareholders.
2.4.2.3 Do Takeovers Create or Redistribute Wealth?
The evidence reviewed in the previous two sections prompts the question whether takeovers create overall net gains, as measured by the aggregate dollar wealth change of target and bidder shareholders, or simply redistribute wealth between these two groups. Empirical evidence suggests that the former is true. For example,
Bradley, Desai, and Kim (1983) reported that the total dollar value of the combined target and bidder firms increased by an average 10.5% in the two months surrounding the takeover announcement. Stulz, Walkling, and Song (1990) report findings consistent with this. The above evidence has led authors like Jensen and
Ruback (1983) and Andrade, Mitchell, and Stafford (2001) to conclude that
10 The cumulative abnormal returns were measured from 10 days prior to the takeover announcement date to 10 days after the takeover announcement date. 21 shareholder-gains in takeovers represent a valid response by the market to what it considers an economically beneficial reorganisation of productive assets.
2.4.3 Post-Takeover Performance
2.4.3.1 Share Price Performance
In justifying takeovers, many acquirer-firm managers claim that the proposed transaction will create value for their shareholders. On the contrary, early studies by
Langetieg (1978) and Jensen and Ruback (1983) reported negative abnormal stock returns for combined firms in the several years following takeover completion, indicating that takeovers destroy shareholder value in the long-term 11 • More recent studies, implementing the methodological suggestions of authors like Lyon, Barber, and Tsai (1999), find no evidence of abnormal returns in the three to five-year period post-takeover (Franks, Harris, and Titman (1991), Agrawal, Jaffe, and Mandelker
(1992), and Higson and Elliott (1998)). However, perhaps the most comprehensive study on this topic is that provided by Loughran and Vijh (1997), who studied 947
U.S. acquisitions between 1970 and 1989, using firms matched by book-to-market ratio, firm-size and industry as a benchmark. They find that the post-acquisition returns to combined firms is significantly related to the mode of acquisition and the method of payment. Over the five-year period following acquisition, mergers, on average, were found to earn an abnormal return of -16%, compared to + 43% for acquisitions via tender offer. Stock acquirers earned abnormal returns of -24%, while cash acquirers earned +19%.
11 It must be noted that these early studies suffered from the poor benchmark problem (also known as the joint hypothesis problem), typically measuring combined firms' returns against benchmarks such as the market model or a CAPM-based measure. 22 2.4.3.2 Operating Performance
Supporters of corporate takeover activity, including firm-managers, most often validate their arguments by suggesting that the market for corporate control facilitates the efficient redistribution of assets to their highest value use. This issue has since been examined empirically, with researchers trying to determine whether operating performance of combined firms improves following takeover, compared to pre-takeover levels. Healy, Palepu, and Ruback (1992) and Linn and Switzer (2001) provide evidence that combined firms experience significant post-acquisition improvements, reporting annual increases in post-merger cashflow returns of 2.8% and 1.8% respectively 12• Additionally, Linn and Switzer (2001) found that cash acquirers dominated stock acquirers, exhibiting annual improvements of 3.1 % and
0.8% respectively. However, Ghosh (2001) proposes that the use of industry-median firms as a benchmark by the two afore-mentioned studies led them to overstate the level of operating performance improvements. Ghosh (2001), using non-merging firms matched on pre-takeover performance and firm-size as a benchmark, found that, on the whole, there is no evidence that operating performance improves following acquisitions. However, when acquisitions were split according to method of payment, results consistent with Linn and Switzer (2001) were found, with cash acquirers experiencing superior improvements to stock acquirers.
12 The benchmark used by both Healy, Palepu, and Ruback (1992) and Linn and Switzer (2001) to determine whether the detected improvements in operating performance were abnormally good was the performance of firms in the same industry as the merged firms. Specifically, the reported improvements represent excess or abnormal improvements after adjusting for the improvement of the industry-median firm. Annual performance was compared over 5 pre-merger years (using pro-forma aggregated numbers of the target and the acquirer) and 5 post-merger years.
23 In conclusion, the evidence suggests that, on the whole, firms completing
acquisitions experience slight improvements in operating performance. However, the more certain result seems to be that firms completing cash-acquisitions experience performance improvements far in excess of stock-acquirers, who appear to show very little improvement. These results are consistent with the evidence in Loughran and Vijh (1997) and are supportive of the notion that announcement-period stock revaluations reflect anticipated operating performance changes in merged firms
(Linn and Switzer, 2001).
2.5 Bid Structure and Transaction Outcomes
In this section, literature directly related to the research questions in this thesis is reviewed. Beginning with sub-section 2.5.1, the key bid structure decisions facing bidders, as well as their motivation in choosing a certain bid structure, are reviewed.
Sub-section 2.5.2 presents literature on the transaction-specific hurdles that must be negotiated before a bidder can enjoy success. These include target management resistance and bid competition. A bidder's ability to successfully negotiate these hurdles will depend on the adopted bid structure. Sub-section 2.5.3 considers the evidence on these relationships in detail, thus reviewing literature directly related to the empirical analysis in this thesis.
24 2.5.1 Elements of Bid Structure
2.5.1.1 Bid Premium Decision
The Bid Premium in a takeover transaction refers to the excess of the offer value over the pre-offer price of the target. Since target shareholders can sell their shares in the market, it goes without saying that they will not accept a takeover offer unless its value at least exceeds the current market value of their shares. In other words, the bid premium decision for the bidder relates entirely to the size of the premium that he chooses to offer, since not offering a premium is not a viable option. However, in practice, takeover offers usually incorporate substantial premiums to the pre-bid value of the target's shares. For example, Jarrell and Poulsen (1987) estimate the bid premiums paid in over 600 transactions between 1962 and 1985. They found that bid premiums averaged 19% in the 1960s, 35% in the 1970s, and 30% in the first-half of the 1980s. Bradley (1980) also reported that for a sample of successful tender offers, the average price paid by the bidder was 49% higher than the target's share price two months before the offer.
Several empirical studies have attempted to explain the cross-sectional variation in bid premium size. Slusky and Caves (1991) find evidence that financial synergy is related to bid premium size. Specifically, bid premium was found to be positively related to the acquirer's financial slack, when acquiring a cash-poor target. Officer
(2003) examined multiple determinants of bid premium size, finding that bid premium is positively related to operational synergy, hostility of the offer, the
25 presence of a termination fee clause 13 and the bidder's market to book ratio, whilst the existence of a toehold was found to be negatively related to bid premium. Israel
(1992), via a theoretical model, proposes that bid premium size is related to the target firm's capital structure and ownership structure. More specifically, bid premium is positively related to a firm's gearing and the proportion of the firm owned by management.
2.5.1.2 Payment Method Decision
Payment method, or medium of exchange, in takeovers refers to the type of consideration offered to target shareholders in exchange for their holdings. The most commonly used forms of consideration are cash and scrip, the latter offering target shareholders direct ownership in the combined firm via an exchange of bidder firm stock for target firm stock. Less common forms of consideration include debt and derivative securities, such as options. In practice, mixed takeover offers, which involve the offer of more than one of the above forms of consideration, are also commonplace.
The payment method decision of bidders has been studied both theoretically and empirically. Perhaps the most influential contribution is that of Myers and Majluf
(1984), whose market-timing hypothesis suggests that if managers are better informed about their firm's prospects than the market, they will acquire with stock when it is overpriced, and use cash otherwise. Shleifer and Vishny (2003) present a
13 A termination fee clause refers to a contractual agreement entered into by a target and a bidder under which one party must pay the other a fixed cash fee if it does not consummate the proposed merger. Target termination fee clauses are most common, where the target pays the bidder if the target fails to complete the proposed merger. Nevertheless, bidder termination fee clauses also exist. 26 stockmarket-based framework of takeovers consistent with this hypothesis. Under their model, cash bids are only made for severely undervalued targets and stock bids are a defensive strategy used by overvalued acquirers to bid for relatively less overvalued targets.
Eckbo, Giammarino, and Heinke! (1990) develop a medium of exchange model specific to mixed offers of cash and scrip. Their model proposes that a bidder's true value is revealed to the target via the proportionate use of cash and scrip in the offer and, consistent with Myers and Majluf (1984), predicts that the bidder's revealed value is monotonically increasing with the proportion of the total offer that consists of cash. Hansen (1987) presents a model with similar implications.
In terms of empirical evidence, a U.S. study by Martin (1996) found that the use of stock-financing is positively related to the acquiring-firm's investment opportunities and level of institutional shareholdings, and negatively related to the acquiring firm's level of management ownership and cash balances. Faccio and Masulis (2004) present evidence on the determinants of payment method in European transactions and show that constraints on acquirer's cash availability encourage stock-financing, whilst potential corporate control threats resulting from the use of stock-financing encourage the use of cash. Additionally, consistent with the predictions of Myers and
Majluf (1984), the pre-bid stock price run-up of acquirers was positively related to the use of stock in the bid.
27 2.5.1.3 Toehold Acquisition Decision
A toehold refers to the bidder's pre-bid ownership stake in the target. The method used to acquire these stakes, as well as their disclosure, is subject to regulatory provisions, a summary of which is provided in Chapter 3. A bidder might acquire a toehold stake for several reasons. The theoretical literature is quick to point out that bidders inevitably profit from the acquisition of toehold stakes. Firstly, by acquiring a toehold stake in the open market prior to the official takeover offer, a successful bidder is able to purchase these shares without paying the premium inherent in takeover offers (Kyle and Villa (1991) and Shleifer and Vishny (1986, a)). Secondly, in the event of the target being acquired by a competing bidder, the losing bidder will at least profit from the sale of its toehold stake (Grossman and Hart (1980), Shleifer and Vishny (1986, a), and Bulow, Huang, and Klemperer (1999)). Lastly, a toehold stake provides its owner with a strategic advantage in the presence of competing bidders, increasing their likelihood of success. This feature of toeholds will be discussed further in section 2.5.3.
These proposed benefits of toeholds are difficult to reconcile with empirical research, which shows that a large proportion of bidders do not acquire toehold stakes in the target prior to launching takeover bids. For example, Bradley, Desai, and Kim (1988) found that less than half of the firms in their sample acquired toeholds. Similar evidence is provided by Betton and Eckbo (2000).
Several theoretical studies have attempted to reconcile this contradiction between theory and empirical research. Chowdhry and Jegadeesh (1994) propose a model in which the size of a bidder's toehold conveys private information about the extent of
28 potential synergies. Specifically, the size of the toehold is positively related to the value of these gains. Hence, under this model, a smaller toehold provides a signal which allows the bidder to bid less than he would have to otherwise in the presence of the free-rider problem (Grossman and Hart, 1980)14. Bris (2002) similarly models the toehold acquisition decision in light of the costly private information it reveals to the market. Ravid and Spiegel (1999) examine the toehold acquisition decision in the context of rival bidders. Amongst other things, their model proposes that if no rival bidders are expected, no toehold should be purchased.
2.5.2 Transaction Outcomes
2.5.2.1 Target Management Response
Target firm management, particularly in cases of hostile and unsolicited offers, may choose to resist the takeover or merger proposal. Most commonly, resistance is waged verbally, whereby the target-firm management recommend rejection of the offer to their shareholders and vigorously denounce the offer in the press. However, resistance can also extend beyond words to specific actions, such as amendment of company charters, the solicitation of white knights, and poison pills15 • All of these actions are aimed at making the transaction less attractive and more difficult for the unwelcome bidder to execute.
14 Grossman and Hart (1980) present a model where target shareholders are atomistic and assume their response will not affect the outcome of a takeover offer. Consequently, these shareholders do not tender their shares unless the bid premium at least matches the expected value of takeover-induced gains, otherwise preferring to free-ride on the improvements made by the bidder. This may effectively p.revent the bidder from profiting from, or even completing, the takeover. . 5 Shleifer and Vishny (1986, b), Dann and DeAngelo (1988), Ambrose and Megginson (1992), Bhagat and Jefferis (1994), Datta and Iskandar-Datta (1996), and Bae and Simet (1998) examine the commonly used anti-takeover defences. 29 Interestingly, management resistance seems to be significantly more prevalent in
Australian takeovers than other developed markets. For example, Eddey and Casey
(1989) report that 41 % of the 400 Australian bids in their sample, made between
1972 and 1985, encountered resistance in the form of reject recommendations from directors. Henry (2004), examining 440 Australian transactions between 1991 and
2000, also reported a rejection rate of 49% for initial bids. In comparison, U.K. studies by Wier (1997) and O'Sullivan and Wong (1998) showed U.K. managers responding in a hostile fashion in only 24% and 26% of transactions, respectively. In the U.S., North (2001) reports that a mere 9% of takeovers were contested by target management between 1990 and 199716• Henry (2004) proposes that the significant limitations on the use of anti-takeover mechanisms in the Australian market can partially explain the elevated levels of hostility observed. More specifically, the effective use of these mechanisms in other markets discourages bidders from launching takeover bids without the support of target management, since they are likely to be thwarted. Conversely, bidders in the Australian market are more willing to attempt hostile transactions due to the reduced potency of takeover defence mechanisms at the disposal of target management.
The statistics in the above literature, which show that target management resist a significant proportion of takeover offers, seem difficult to reconcile with the large amount of evidence showing that target shareholders profit substantially from successful takeovers, as discussed in section 2.4.2.1. Consequently, the potential motivations behind target management resistance warrant further discussion. Baron
16 In this study, 38% of the 304 initial bids received a Reject recommendation from target directors, with a further 4% receiving neutral recommendations such as "Take No Action at this Time" or "No Recommendation". 30 (1983) proposes that target-firm management choosing to resist a takeover offer can be divided into three groups, according to their motivation. In the first group, target management believe the offer is inadequate compared to the true value of their firm and therefore resist. In the second group, target management believe the offer is adequate in this regard but resist as a means of soliciting a higher offer, either from the same or a new bidder. Lastly, target management may resist in order to retain control of the target, regardless of the offer's value proposition.
First and foremost, company directors have a fiduciary duty to act in the best interests of shareholders. Beginning with the first group, resistance by these directors is clearly consistent with this principle. The primary concern of directors in this group is to ensure that the amount and form of consideration offered does not undervalue the target-firm. Evidence on this potential motivation of resistance is discussed in more detail in section 2.5.3.2 and forms part of the empirical analysis in this thesis.
The second group's motive for resistance is also consistent with the principle of shareholder wealth maximisation. Not surprisingly, empirical evidence confirms that competition amongst bidding firms increases target shareholder wealth (Bradley,
Desai, and Kim, 1988). However, following target firm resistance, an upwards revision by an existing bidder or the emergence of a competing bidder is far from certain. In such instances, where management defeat a bid and remain independent, target firm shareholders are likely to suffer wealth losses (Jarrell, 1985). Hence, the trade-off between increased final consideration, in the event of deal success, and decreased probability of success will determine the impact on target shareholders
31 wealth. Cotter and Zenner (1994) propose that this trade-off does not benefit target shareholders, finding that, on average, managerial resistance does not increase shareholder wealth. On the other hand, Jennings and Mazzeo (1993) find that target management resistance is not harmful to their shareholders' wealth.
Finally, resistance by the third group of directors is clearly inconsistent with the best interests of shareholders and illustrates the agency problem inherent in public companies (Jensen and Meckling, 1976). These directors will resist any takeover offer, regardless of the value it represents to shareholders, in order to retain their job and related perquisites. This has proven to be an effective strategy, with management resistance found to be highly positively correlated with bid failure (Walkling, 1985).
In reality, even managers driven by self-interest are likely to accept an offer if it results in a sufficiently large increase in their personal wealth. This notion is supported by Cotter and Zenner (1994), Buchholtz and Ribbens (1994), and Holl and
Kyriazis (1997), who find that the greater their personal wealth gains from an offer, the less likely managers and directors are to resist that offer.
2.5.2.2 Bid Competition
Following the announcement of an initial takeover offer, it is not uncommon to see a competing bid entered by a second and even third bidder for the same target firm 17 .
A competing bidder may choose to enter the contest upon the discovery of new information revealed by the initial bid or in order to preserve or improve their competitive position, particularly if the other bidder/s is a direct competitor.
17 Betton and Eckbo (2000) reported the entry of a competing bidder in 16% of the 1,353 transactions examined. Similarly, competing bidders were present in 14% of the 304 transactions examined empirically in this thesis. 32 Consistent with evidence from other asset markets, competitive bidding situations in the market for corporate control result in higher sale prices, often after multiple rounds of counter-bidding. Empirical evidence is consistent with this intuitive outcome, confirming that transactions involving competing bidders lead to greater wealth gains for target-firm shareholders than single-bidder transactions. Predictably, these gains are made at the expense of acquirer-firm shareholders, who experience reduced returns as a result of having to pay a higher price to purchase the target
(Bradley, Desai, and Kim, 1988)18.
In summation, the presence of competing bidders generally leads to positive outcomes for target-firm shareholders by increasing the price obtained for the firm.
On the contrary, the presence of other bidders reduces the likelihood of success for each individual bidder, whilst also increasing the purchase price for the winning bidder.
2.5.2.3 Bid Success
Bid outcome refers to whether a particular takeover bid is ultimately accepted by shareholders, leading to a successful transfer of control. Bid success must not be confused with deal completion, since a completed deal can also involve one or more unsuccessful bids. Hence, in a transaction involving multiple bids, only the final bid by the winning bidder can be considered successful, with all other bids considered unsuccessful. Deal completion in the Australian market was found to be very high, with a completion rate of 95% over the six year period ending in 2001 (Rubicon,
18 Returns are calculated as the cumulative abnormal return to the acquiring firm measured from five trading days before the announcement of the first offer made by this firm through to five trading days after the announcement of the ultimately successful bid. 33 2001). Statistics on bid outcome are sparse, however, Betton and Eckbo (2000) found 78% of single-bid deals were successful.
Several empirical studies have examined determinants of bid success. Clearly, the amount and form of consideration offered, together with other bid structure elements, will significantly influence bid outcome. The past empirical evidence with regard to these elements will be reviewed in section 2.5.3.3. Additional evidence will be provided by the empirical analysis in this thesis. Intermediate transaction outcomes such as the presence of target management resistance and/or a competing bid reduce the likelihood of bid success (Walkling, 1985). Bates and Lemmon (2003) find that target-firm ownership structure is also related to the likelihood of bid success.
2.5.3 Impact of Bid Structure on Transaction Outcomes
This section will review the theoretical and empirical literature dealing with the hypothesised and discovered relationships between the bid structure elements of bid premium, payment method, and toehold, and the likelihood of observing transaction outcomes such as target management resistance, bid competition and ultimate bid success.
34 2.5.3.1 Relationship between Bid Structure Elements and Bid Competition
2.5.3.1.1 Bid Premium and Competition
Numerous theoretical papers have proposed that the likelihood of observing a competing bid for a takeover target is related to the structure of the initial bidder's offer. For example, Giammarino and Heinkel (1986), Fishman (1988), and
Hirshleifer and Png (1989) all present models where a high-premium initial bid reduces the likelihood of a competing bidder emerging. In these models, a high premium bid signals that the initial bidder has a high valuation of the target. This signal causes potential competing bidders to revise downwards their expectation of winning the contest at a profitable price, thereby deterring them from committing funds to investigating the target and bidding.
In Giammarino and Heinkel (1986) and Hirshleifer and Png (1989) each bidder is limited to making just one bid. Therefore, the motive for the initial high-premium bid is to deter the competing bidder by making the acquisition cost he faces (the value of the initial bid) unprofitably high. In the model of Fishman (1988), where bidders can make as many bids as they wish, the rationale is slightly different.
Bidders in this environment submit initial high-premium bids to signal their high valuation, also in the hope of deterring a potential competing bidder. However, if competing bidders are not deterred, the initial bidder can raise the acquisition cost via subsequent bids. Baron (1983) also presents a model where each bidder is limited to one bid, however, unlike Giammarino and Heinkel (1986) and Hirshleifer and Png
(1989), a second bidder can only bid if the initial bidder's offer is rejected by target management. Hence, in this situation, the initial bidder's motive for a high-premium 35 bid is to induce target management to accept the offer, which ends the contest in the initial bidder's favour.
All of the above studies model an environment where bidding is costly, reflecting the significant informational and other costs of mounting a takeover bid. In this setting, a high-premium initial bid is an effective deterrent since the costs of making a bid are non-trivial. Khanna (1997) presents a model where information, and therefore the act of bidding, is costless. Consequently, in this model a second bidder always enters the contest, whereas further bidders only arrive if the second bidder can top the initial offer. Under these assumptions, the motive for a high initial offer is to reduce the probability of later competing bidders being given the opportunity to enter.
Despite some differences in the assumptions and set-up of their models, the empirical implication with respect to the relationship between bid premium and takeover competition presented by the above studies is identical. These studies unanimously predict a negative relationship between the size of the bid premium in initial takeover offers and the likelihood of competing bids emerging. The available empirical evidence is generally consistent with this theoretical prediction. Jennings and Mazzeo (1993), in a U.S. study of 647 initial acquisition proposals made between 1979 and 1987, find evidence that high-premium bids do deter competing bidders. Betton and Eckbo (2000) report the same relationship with respect to the
1,353 initial tender offers in their sample. Consistent Australian evidence comes courtesy of Pompilio (2000). In a study of 249 initial takeover and merger proposals made for ASX-listed targets between 1992 and 1998, Pompilio reports a
36 significantly negative relationship between bid premium size and the likelihood of a competing bid emerging.
2.5.3.1.2 Payment Method and Competition
A second element of bid structure proposed to influence bid competition is the medium of exchange, or payment method. In the model presented by Fishman
(1989), the method of payment provides a signal of a bidder's valuation of the target in a setting of asymmetric information. In Fishman's model, it is assumed that each bidder, but not the target, has private information on the proposed acquisition's profitability. This leads to the assumption that bidders know the value of their securities offers, since the value of the offered securities is dependent on the future profitability of the acquisition, but the target does not. However, the value of cash offers is unconditional, effectively eliminating the informational disadvantage to the target. Only those bidders who expect a highly profitable acquisition, and thus hold a high valuation of the target, can afford to reveal these beliefs to the target by offering cash. Additionally, these high-valuation bidders use cash in order to retain ownership of the combined firm's securities, which they believe will have a high future value.
In contrast, those bidders with lower expectations for the acquisition's profitability will choose to hide this belief by offering stock. They are also willing to share the combined firm's securities with target-firm shareholders since the expected value of these shares is lower. Hence, the use of cash by a bidder signals a high valuation, which can serve to deter or competing bidders. The empirical prediction made by
Fishman is that the probability of competing bids emerging is higher for initial
37 securities offers as compared to initial offers of cash19• Berkovitch and Narayanan
(1990) also stress the preemptive qualities of cash, suggesting that bidders will structure their offer to include a higher proportion of cash in situations where the potential for competition is high.
The empirical evidence again comes courtesy of Jennings and Mazzeo (1993),
Pompilio (2000), and Betton and Eckbo (2000). Contrary to predictions from theory,
Jennings and Mazzeo (1993) find that greater use of cash in the offer does not seem to deter competition. In fact, proportion of cash is found to be somewhat positively associated with the likelihood of competition. Pompilio (2000) similarly finds a significant positive relation between proportion of cash in the offer and the likelihood of competition. However, in an innovative extension that is replicated in this thesis, Pompilio re-estimates his results after removing Schemes of Arrangement from the sample. Removal of these inherently friendly transactions results in the relation between proportion of cash in the offer and the likelihood of competition becoming insignificant. This is consistent with the results of Betton and Eckbo
(2000). Pompilio interprets this result as showing that medium of exchange is often a proxy for the type of transaction, which is in fact the underlying influence on competition, not the payment method. For example, it is the friendly nature of
Schemes of Arrangement that deter competing bidders rather than the fact that these transactions are often consummated with stock-offers. In summation, empirical evidence fails to support the proposition that cash as a payment method has a
19 Hansen (1987) and Eckbo, Giammarino, and Heinkel (1990) likewise examine the signaling effect of the payment method. In these models, bidders offer cash if their equity is relatively undervalued. However, these models focus on single-bidder environments and thus make no predictions relating to the likelihood of competition. 38 deterring effect on competing bidders, which is contrary to the hypothesis of
Fishman (1989) and related theoretical literature.
2.5.3.1.3 Toehold and Competition
As alluded to in section 2.5.1.3, the price offered by a bidder with a toehold stake simultaneously represents a bid price for the outstanding shares and an ask price for their toehold stake, obliging any competing bidders to pay at least this ask price to win the contest. Thus, toeholds give rise to an incentive for both a low acquisition price and a high acquisition price. Placed in this position, toeholders have an obvious incentive to overbid, a notion formalised by the models of Burkart (1995), Singh
(1998), and Bulow et al. (1999). If the initial bidder owns a toehold, his propensity to overbid is likely to reduce the chances of competing bids emerging, particularly from bidders without toeholds, who bid rationally. This intuitive result is proposed by
Bulow et al. (1999), who suggest that "a toehold may make it much less likely that an outside bidder will enter a takeover battle" (p. 430). Dewatripont (1993) and
Ravid and Spiegel (1999) also propose that a toehold can be effective in deterring rival bidders.
The empirical evidence is consistent with this hypothesis, with both Betton and
Eckbo (2000) and Pompilio (2000) finding that ownership of larger toeholds by initial bidders reduces the probability of rival bidder entry.
39 2.5.3.2 Relationship between Bid Structure Elements and Target Management Resistance
2.5.3.2.1 Bid Premium and Target Management Resistance
Intuitively speaking, the size of the bid premium m an initial offer should be negatively related to the likelihood of encountering target management resistance.
Firstly, by endorsing an offer with a higher bid premium target management are better able to discharge their fiduciary duty to shareholders, as well as satisfy their self-interest as shareholders. The negative relationship between management wealth changes (resulting from the offer) and the likelihood of target management resistance is confirmed in the empirical studies of Buchholtz and Ribbens (1994) and Cotter and Zenner (1994). Additionally, Giammarino and Heinkel (1986) and Hirshleifer and Png (1989), whose models examine the resistance and competition decisions simultaneously, propose that target management may resist an initial offer in order to encourage alternative bidders. As discussed in section 2.5.3.1.1, the primary implication of these models is that higher-premium bids deter competing bidders.
Hence, higher-premium bids make resistance for the purpose of soliciting competing bids a higher-risk strategy, which should, all else being equal, reduce the likelihood of resistance.
Early empirical evidence came from Walkling and Long (1984), who found that bid premium size was the least important variable, of those examined in their model, in explaining target management resistance. This counter-intuitive result could partially be explained by the modest data-set of 95 tender offers. However, Jennings and
Mazzeo (1993), examining both mergers and tender offers, reported a significantly
40 negative relation between the size of initial bid premium and the likelihood of encountering target management resistance. Australian evidence is likewise mixed.
Eddey and Casey (1989) found higher bid premiums to be associated with a lower likelihood of resistance, whilst Henry (2004) found no significant relationship in his study. St-Pierre, Gagnon and Saint-Pierre (1996) provide evidence from the
Canadian market, finding a negative relationship between bid premium size and resistance in a sample of 112 initial cash offers.
Notably, the two Australian studies use directors' recommendations to shareholders to proxy for target management response20• This is a more direct and superior proxy to that used by other studies, which usually resort to searching through media sources for statements or evidence of certain actions by target management potentially indicating dissatisfaction with the offer.
2.5.3.2.2 Payment Method and Target Management Resistance
Fishman (1989), in the model described above, proposes that target management is less likely to resist a cash offer as opposed to a securities offer. Under this asymmetric information model, a securities offer signals a low-valuation bidder and a less profitable acquisition than that proposed by a cash offer. Hence, upon receiving this signal, target management are more likely to reject securities offers as compared to cash offers. Ghosh and Ruland (1998) present a contrasting hypothesis.
Their theory focuses on transactions where target management have high ownership stakes in the firm. They propose that target management with large ownership stakes generally wish to retain their position or at least their voting power in the combined
20 This thesis also uses directors' recommendations in the same way. 41 firm. Consequently, they will prefer to receive stock in exchange for their ownership interest, as opposed to cash. This theory implies that larger amounts of cash in the offer will be associated with a higher likelihood of resistance, but only in situations of high target management ownership.
Jennings and Mazzeo (1993) find that larger amounts of cash are associated with a lower frequency of target resistance. However, this result is only mildly supportive of Fishman (1989) since the observed relationship is weak in the merger sub-sample and not robust to the definitions of resistance used in the study. Furthermore, Henry
(2004) reports a significant positive relation between cash offers and target management resistance, as proxied by reject recommendations from directors. This surprising result is somewhat consistent with Pompilio (2000), who found that cash offers are positively related to bid competition.
2.5.3.2.3 Toehold and Target Management Resistance
There is an absence of theoretical papers devoted directly to exploring the relationship between initial bidder's toehold and the likelihood of observing target management resistance. However, as discussed in section 2.5.3.1.3, toeholds can be an effective deterrent for competing bidders. Thus, one can sensibly argue that target management is less likely to resist if the initial bidder also owns a toehold, since the likelihood of further suitors emerging is reduced. For example, Bulow et al. (1999) suggest that their model, which primarily examines bidder competition in the presence of toeholds, is consistent with results that find "toeholds lower the probability of management resistance" (p. 430).
42 Walkling and Long (1984) were the first to examine this relationship empirically.
They found that bidders who did not encounter resistance possessed an average toehold of 27%, whilst bidders in contested transactions had an average pre-bid stake of only 11 %, evidence consistent with a negative relationship between toehold size and the probability of encountering management resistance. Jennings and Mazzeo
(1993) and Cotter and Zenner (1994), using binomial logistic estimation, likewise found that larger toeholds reduce the probability of target management resistance, whilst Betton and Eckbo (2000) confirm this relationship using multinomial probability estimates. Australian evidence is consistent with the above U.S. results, with both Eddey and Casey (1989) and Henry (2004) reporting a significant negative relationship between toehold size and directors' reject recommendations. Hence, the empirical evidence strongly supports the notion that greater bidder toeholds reduce the probability of encountering target management resistance.
2.5.3.3 Relationship between Bid Structure Elements and Bid Success
2.5.3.3.1 Bid Premium and Bid Success
Giammarino and Heinke! (1986), Fishman (1988), and Hirshleifer and Png (1989) all present multiple-bidder models featuring a positive relationship between the bid premium and the likelihood of offer success for the initial bidder. These models, as described in section 2.5.3.1.1, are analysed within an asymmetric information setting, where an initial bid provides a signal to potential competing bidders. The positive relation between bid premium and the likelihood of offer success in these papers is to a large degree a function of the competition-deterring effect of high premium bids.
43 Importantly, these papers assume that offers are made to management, rather than directly to shareholders, effectively assuming away the free-rider problem. On the other hand, the model presented by Hirshleifer and Titman (1990) accounts for the free-rider problem among target shareholders. Under this model, takeover bids perfectly reveal the bidder's private information as to the value of takeover gains.
High-value bidders reveal themselves by making high-premium bids, thereby increasing their probability of success. Hence, the theoretical literature reinforces the intuitive notion that higher-premium bids are more likely to succeed than lower premium bids.
Early empirical evidence comes courtesy of Walkling (1985), who found higher bid premiums indeed increased the probability of success in his study of 158 initial cash tender offers made in the U.S. between 1972 and 1977. Importantly, Walkling also showed that previous studies, such as Hoffmeister and Dyl (1981), which found that bid premium size did not significantly affect offer outcome, suffered from severe methodological problems which affected their results. At the forefront of these problems were the misspecification of the bid premium and the use of linear models, the assumptions of which were not met by the data. Consistent with Walkling
(1985), and employing multinomial logistic estimation, Betton and Eckbo (2000) also found that the probability of initial bid success is positive with respect to the bid premium. Officer (2003), using both initial and competing bids in the same analysis, likewise reports that higher-premium bids are more likely to succeed. Hence, the intuitively appealing positive association between bid premium and offer success seems to be well supported empirically.
44 2.5.3.3.2 Payment Method and Bid Success
Several theoretical studies have emphasized the signaling effect of the payment method. Fishm,m (1989) showed this in the context of competing bidders, whilst
Hansen (1987) and Eckbo, Giammarino, and Heinkel (1990) focused on the payment method decision in a single-bidder environment. Unfortunately, these studies did not directly examine the impact of payment method on the likelihood of bid success, providing no empirical predictions about this relationship. Empirical evidence on this relationship has likewise been less than plentiful. Betton and Eckbo (2000) and
Officer (2003) found that payment method was insignificant in explaining the offer outcome, while Bates and Lemmon (2003) actually found deal success to be positively related to the use of equity consideration.
2.5.3.3.3 Toehold and Bid Success
As noted previously, theoretical works by Burkart (1995), Singh (1998) and Bulow et al. (1999) find that bidders with toeholds always overbid, whilst bidders without toeholds do not pay above their valuation. These models consider multiple-bidder situations and show that the toeholder is able to increase the bid price enough to beat zero-toehold bidders. The first of these models considers bidding for 'common value' targets, for whom all bidders have similar post-acquisition strategies, whilst the other two models consider 'private-value' transactions, where each bidder has a different use in mind for the target assets. However, these studies are consistent in their empirical prediction that ownership of a toehold stake increases a bidder's chances of winning a takeover contest, be they the initial or following bidders.
45 Whereas the studies above ignore the free-rider problem, the model of Chowdhry and Jegadeesh (1994) takes it in to account. Thus, their model does not assume that target shareholders automatically sell their shares to the highest bidder. However, their signaling model does suggest that toeholders will bid above those bidders without toeholds, consequently improving their chances of bid success. Empirical evidence is supportive of this theoretical prediction, with Walkling (1985), Betton and Eckbo (2000), and Officer (2003) all finding that larger toeholds increase the probability of bid success.
2.6 Literature Review Conclusion
The literature review presented in the above sections was aimed at highlighting some of the main issues and findings in the extant takeover literature. Section 2.2 showed that the intensity of takeover activity deviates over time, with several periods of particularly high takeover intensity, known as merger waves, evident in the twentieth century. In section 2.3, the main reasons thought to motivate takeover activity were reviewed. It was shown that bidder management initiate takeovers with one or more of the following goals in mind: leveraging synergies between the target and acquirer, acquiring an undervalued target, replacing inefficient target management with new and better management, satisfying their personal wealth and career objectives, and dispensing with free cashflows. The next section looked at the performance and returns literature related to takeovers. Section 2.4.1 shows that, predictably, acquirers tend to be larger and better-performed, on average, than target firms. Empirical evidence reviewed in section 2.4.2 confirmed that target-firm shareholders are the big winners from takeovers and that these transactions, on average, are wealth creating, rather than simply redistributing wealth from acquirer to target
46 shareholders. Section 2.4.3 considered post-takeover performance, suggesting that takeovers, on the whole, do not increase shareholder wealth in the long-term, while operating performance of combined firms improves slightly. Section 2.5 reviewed literature directly related to the empirical analysis in this thesis. Specifically, evidence on the impact of a bid's structure on the likelihood of observing target management resistance, bid competition and ultimate bid success was explored thoroughly.
47 Chapter3 The Australian Regulatory Environment
Takeovers of publicly-listed companies in Australia are primarily regulated under the provisions embodied in Chapter 6 of the Corporations Act (2001) and enforced by the Australian Securities and Investments Commission (ASIC). This section focuses on those aspects of the Australian regulatory framework that are of particular relevance to the analysis in this thesis. Furthermore, familiarity with the terminology, processes, and legislation introduced below is necessary for a full understanding of this analysis.
3.1 Acquisition of a Toehold Stake
In the Australian market, individuals or corporations holding a stake of 5% or more in a listed firm must declare this holding to the Australian Stock Exchange (ASX).
The identity of these parties and the size of their stake then becomes public information. Additionally, Section 606 of the Corporations Act (2001) prohibits individuals or firms from increasing their stake in a listed company from 20%, or below 20%, to more than 20%, or from a starting point that is between 20% and
90%, via ordinary stockmarket purchases. Rather, they must do this via a publicly announced takeover offer21 .
21 There are certain circumstances under which the contravention of section 606 is allowed. These are listed in Section 611 of the Corporations Act 2001.
48 3.2 Making a Merger or Acquisition Proposal
Bidders in the Australian market can choose between three methods, or modes of takeover, in implementing their proposal. The first, and by the far the most common, method is an off-market bid. This method is also referred to as a Part A offer and enables the acquiring firm to make an off-market offer of cash or shares to target shareholders, whilst stipulating certain conditions which must be satisfied before acceptances are considered binding. The second method is a market bid, or Part C offer, which is implemented on-market, must offer cash-only consideration and must be free of any conditions22• The third method involves merging two or more companies using a Scheme of Arrangement. Under the Corporations Act (2001), mergers executed via this method require minimum target shareholder approval of
50% in number of shareholders and 75% in value of shares held. Additionally, court approval is necessary for the transaction to succeed.
3.3 Directors' Recommendations to Target Shareholders
Under the Corporations Law, companies making a takeover proposal via a Part A or
Part C offer have 21 days after announcing the takeover to formalise the offer by issuing a Bidder Statement to the target company, outlining the offer in detail. This document must then be sent to all target shareholders no later than 28 days after it is sent to the target company itself. In response, target management are required to issue a Target Statement (or Response Statement), which must be prepared and
22 Whereas Part A bidders can choose to make an offer for only a specified proportion of the securities in the bid class, an offer for securities under a market bid must be an offer to buy all outstanding securities in the bid class. 49 distributed to target shareholders, the bidding company and the regulator within 15 days of the Initial Bidder Statement being received by target shareholders.
Target statements must include a section in which the target's directors provide a
Recommendation to Shareholders on how to deal with the proposed takeover offer, giving detailed reasons to support their recommendation. The vast majority of recommendations made are in the form of "Reject the offer" or "Accept the offer".
However, on rare occasions, other less explicit responses such as "Take No Action at this Time" or "No Recommendation" are made. There are also instances, usually in multiple-bid transactions, where a Target Statement from directors is not received.
This can occur in the following situations:
(i) Target directors are only obliged to provide a Target Statement in
response to initial bids, not competing or revised bids;
(ii) A bidder can revise their bid numerous times in the 21 days before
actually issuing a Bidder Statement, before which a target is not required to
respond with a Target Statement;
(iii) A bid may be withdrawn before a Target Statement can be issued.
However, in these cases, directors usually send a letter directly to target shareholders, or at least make a media release, through which they advise their shareholders on how to respond to an offer. These Recommendations are usually of the identical form as those made by directors in a Target Statement and are used in the same manner in this study to proxy for target management response.
Following Henry (2004), this study uses the Directors' Recommendation to
Shareholders as a direct proxy for the response of target management. This
50 particular feature of the Australian M&A market, which sees target directors make direct and explicit recommendations to their shareholders with respect to takeover proposals, enables us to use a more precise proxy for the response of target management than has been used in the majority of past studies. The use of Directors'
Recommendations in this study is discussed further in Chapter 5: Data.
51 Chapter4 Hypotheses Development
The purpose of this chapter is to clearly introduce the hypotheses that will be empirically tested. In doing so, this chapter flows on from section 2.5 of the literature review, where the extant literature motivating these hypotheses is reviewed in detail. The reasoning behind the hypothesised relationships will be re-introduced with reference to the guiding literature, yet this will be done with brevity in order to avoid excessive repetition.
4.1 Competition Hypotheses
H-1: The likelihood of a rival bidder entering the takeover contest is:
(H-la) Negatively related to the size of the initial bid premium;
(H-lb) Negatively related to the proportion of cash in the initial offer;
(H-lc) Negatively related to the size of the initial bidder's toehold stake;
(H-ld) Negatively related to the cost of acquiring information about the
target firm;
(H-le) Positively related to the presence of target management resistance
towards the initial offer;
Prior to launching a takeover bid firms need to formulate an acquisition strategy.
One of the critical decisions relates to the structuring of the bid. In fact, authors like
Fishman (1988) suggest that bid structure can be used by initial bidders to deter
52 competing bidders from entering the contest, thereby improving their chances of acquiring the target at a profitable price. The deliberate manipulation of initial bid structure for this purpose is referred to in the literature as 'pre-emptive bidding'.
Specifically, several elements of bid structure have been proposed as having the ability to reduce the likelihood of rival bidder entry. For example, it is sensible to expect that other potential bidders are less likely to compete with a high initial offer.
This intuitive notion is formalised in the theoretical work of Giammarino and
Heinkel (1986), Fishman (1988), and Hirshleifer and Png (1989), all of whom also stress the signaling effect of high-premium initial bids. Their common prediction, that competing bidders are less likely to emerge when initial bid premium is high, is confirmed empirically by Jennings and Mazzeo (1993), Pompilio (2000), and Betton and Eckbo (2000). Fishman (1989) extended his work of the previous year to consider the payment method used by bidders. The signaling effect of bid structure is again highlighted, with Fishman suggesting the use of cash by a bidder signals a high valuation, which can serve to deter or preempt competing bidders. The predicted implication is that bids offering a higher proportion of cash are less likely to attract competition.
Bidders may also choose to acquire an ownership interest in the target firm, commonly referred to as a toehold stake, prior to launching a formal takeover bid.
Theoretical work by the likes of Burkart (1995), Singh (1998) and Bulow et al.
(1999) shows that toehold-owning bidders commit to an aggressive bidding strategy that causes them to overbid. As explained in section 2.5 previously, this is because toeholders have some incentive for offering both a low acquisition price and a high acquisition price. Such over-aggressive bidding by an initial bidder can serve as a
53 deterrent to competing bidders, a notion confirmed empirically by Betton and Eckbo
(2000) and Pompilio (2000), who found that ownership of larger toeholds by initial bidders reduces the probability of rival bidder entry.
In summation, existent literature suggests that the likelihood of a rival bidder entering the takeover contest is lower when: the initial bidder offers a high bid premium; a high proportion of the initial consideration is in the form of cash; the initial bidder owns a large toehold stake. These expected relationships are expressed in hypotheses H-la to H-lc above.
The very next hypothesis examines a target-specific variable. In addition to their predictions regarding bid premium and payment method, Fishman (1988, 1989), and
Hirshleifer and Png (1989) propose that the likelihood of rival bidder entry may also be dependent on the cost of becoming informed about the target. Consider that bidders must commit resources to learn the target's value prior to launching a bid.
Assuming that costs involved in this process are fixed, such as investment banking advice and commitment of managerial time, the bidder must consider the payoff from bidding in light of this up-front cost. Hence, Fishman (1988, 1989) and
Hirshleifer and Png (1989) propose that the likelihood of competing bids emerging will be inversely related to this information cost. Consistent with Jennings and
Mazzeo (1993) and Pompilio (2000), this study uses the size of the target firm to proxy for the cost of becoming informed. This is an adequate proxy since firm size is expected to be highly positively correlated with the number of analysts following the firm as well as the amount of historical market information available. Hence, a positive relationship is expected between target-firm size and the likelihood of
54 competition as bidders will find it easier and less costly to become informed about larger targets, leading to a higher frequency of rival-bidder entry for such targets.
This expected relationship is expressed in hypotheses H-1 d.
As suggested by hypotheses H-la to H-lc, bidder management may understandably structure their offer so as to reduce the likelihood of encountering competition from rival bidders. On the other hand, target management may take actions which will encourage the entry of rival bidders, motivated by the knowledge that competition amongst bidders can drive the acquisition price substantially higher. Theoretical work by the likes of Baron (1983), Giammarino and Heinke! (1986), Hirshleifer and
Png (1989), and Berkovitch and Khanna (1990) suggests that one such course of action available to target-firm management is to resist the initial offer. These models examine the resistance and competition decisions simultaneously, arguing that target management may resist an initial offer as a means of soliciting competing bids. If this proposition were to be true, one would expect that the likelihood of a competing offer emerging would be higher in transactions where the initial offer is resisted by target management. The last hypothesis in this section, hypothesis H-le, expresses this expected relationship. This hypothesis is tested using simultaneous equation analysis, employing the bivariate probit model, results of which are reported in
Section 7.4.1: Simultaneous Determination of the Resistance and Competition
Decisions.
55 4.2 Management Resistance Hypotheses
H-2: The likelihood of observing target management resistance towards an
initial offer is:
(H-2a) Negatively related to the size of the initial bid premium;
(H-2b) Negatively related to the proportion of cash in the initial offer;
(H-2c) Negatively related to the size of the initial bidder's toehold stake;
(H-2d) Negatively related to the cost of acquiring information about the
target firm;
In launching a takeover bid, perhaps the single largest influence on the outcome of the offer is the response of target-firm management. Target management can thwart takeover attempts by utilising various takeover-defence mechanisms or, at least, significantly influence the decision of shareholders to tender their shares via the management's recommendations. Conversely, an initial offer endorsed by target management is much more likely to succeed without competition from rival bidders or opposition from shareholders. Hence, in preparing their offer, bidders would understandably seek to avoid target management resistance. In section 4.1, above, it was proposed that bid structure could be used to pre-empt competing bidders. In a similar manner, extant literature suggests that bid structure can be used by initial bidders to avoid target-management resistance, likewise improving their chances of acquiring the target at a profitable price.
Considering initial bid premium first, it seems sensible to suggest that higher premium offers are less likely to be resisted. Examining the resistance decision from
56 a self-interest perspective, Buchholtz and Ribbens (1994) and Cotter and Zenner
(1994) provide empirical evidence consistent with this proposition. Giammarino and
Heinke! (1986) and Hirshleifer and Png (1989), whose models examine the resistance and competition decisions simultaneously, also imply a negative relation between bid premium and likelihood of resistance, although this is primarily a function of the competition-deterring effect of high premiums.
The motivation for hypotheses H-2b and H-2d comes from Fishman (1989), who's primary proposition is that target management is less likely to resist a cash offer as opposed to a securities offer. Under the asymmetric information model in Fishman's paper, a securities offer signals a low-valuation bidder and a less profitable acquisition than that proposed by cash offers. This signalling model suggests that initial bids offering a higher proportion of cash are less likely to be resisted. Fishman also proposes that the likelihood of resistance should be higher when it is easier to become informed about the target, partly because such targets are more likely to attract competition. This expected relationship is expressed in hypothesis H-2d.
A plethora of empirical studies have found evidence consistent with an inverse relationship between the likelihood of an initial offer being resisted and the size of the initial bidder's toehold (Walkling and Long (1984), Eddey and Casey (1989),
Jennings and Mazzeo (1993), and Henry (2004)). The model of Bulow et al. (1999) likewise suggests that ownership of larger toeholds by initial bidders lower the probability of target-management resistance. This expected relationship is expressed in hypothesis H-2c.
57 In summation, hypotheses H-2a to H-2d propose that the likelihood of observing target management resistance towards an initial offer is lower when the initial bidder offers a high bid premium, a high proportion of the initial consideration is in the form of cash, the initial bidder owns a large toehold stake and, the cost of acquiring information about the target is high.
4.3 Bid Success Hypotheses
H-3: The likelihood of initial bid-success, where the transaction is
consummated under the terms of the initial offer, is23 :
(H-3a) Positively related to the size of the initial bid premium;
(H-3b) Positively related to the proportion of cash in the initial offer;
(H-3c) Positively related to the size of the initial bidder's toehold stake;
(H-3d) Negatively related to the presence of target management resistance
towards the initial offer;
Whereas pre-empting potential competing bidders and gaining target management approval are highly desirable outcomes for a bidder, these are simply means to an end, with the ultimate goal for any bidder being the enjoyment of bid success, leading to the transfer of control of the target's assets. However, although not
23 The cost of acquiring information about the target firm was not considered as a determinant of the likelihood of bid success. The reason for this is straightforward and intuitive. This investigation cost was already borne by actual bidders, since they entered a bid, and it will have no further bearing on the likelihood of their bid succeeding. Rather, this cost is only relevant to potential bidders considering entering the takeover contest, and as such may explain the likelihood of observing competing bidders. Less directly, this cost may also impact on the likelihood of observing target management resistance, since management may resist as a means of soliciting competing bids.
58 necessarily essential, or even sufficient, to ensure bid success, deterrence of competition and target management approval are highly positively correlated with bid success. Hence, in the above hypotheses, the impact of the bid structure variables on the likelihood of bid success in part reflects the ability to use these elements to deter competing bidders and secure target management support.
Beginning with hypothesis H-3a, ample theoretical motivation is provided by the likes of Giammarino and Heinkel (1986), Fishman (1988), and Hirshleifer and Png
(1989). These multiple-bidder models are set in an asymmetric information framework, where a high-premium initial bid signals a high-valuation bidder, a signal that deters potential competing bidders. These models feature a positive relationship between bid premium and the likelihood of offer success for the initial bidder. However, this is to a large degree a function of the competition-deterring effect of high premium bids.
Several theoretical studies have emphasized the signaling effect of the payment method. Fishman (1989) proposed that cash as a payment method can deter competing bidders, whilst Hansen (1987) and Eckbo, Giammarino, and Heinkel
(1990) showed that bidders offer cash if their equity is relatively undervalued. The signalling effects proposed by these authors suggest that the use of cash should be positively related to bid success. The relationship between payment method and bid success is dealt with in hypothesis H-3b, which is motivated by the absence of dedicated theoretical work, as well as the modest amount of empirical evidence, on this relationship. Interestingly, the extant empirical evidence is inconsistent with related theoretical work, with both Betton and Eckbo (2000) and Officer (2003)
59 finding the payment method to be insignificant in explaining offer outcome, while
Bates and Lemmon (2003) actually found offer success to be weakly positively related to the use of equity consideration.
Theoretical pieces by Burkart (1995), Singh (1998), and Bulow et al. (1999) find that bidders with toeholds always overbid, whilst bidders without toeholds do not pay above their valuation. These models consider multiple-bidder situations and show that bidders with larger toeholds commit to more aggressive bidding strategies, which improves their chances of winning a takeover contest, be they the initial or following bidders. This expected relationship is expressed in hypothesis H-3c.
Again, this hypothesised relationship is largely reflective of the competition deterring effect of high-toeholds.
Finally, as mentioned in section 4.2, perhaps the single largest influence on the outcome of a takeover offer is the response of target-firm management. This notion is supported by numerous empirical studies who find that resisted offers are significantly less likely to succeed, as compared to offers endorsed by target management (Walkling (1985), Eddey and Casey (1989), Cotter and Zenner (1994),
Holl and Kyriazis (1997), and O'Sullivan and Wong (1998)). Hypothesis H-3d expresses this expected relationship and is tested using simultaneous equation analysis, employing the bivariate probit model. The results of this analysis are reported in Section 7.4.2: Simultaneous Determination of the Resistance Decision and Initial Bid Outcome.
60 Chapter 5 Data
This study examines Merger and Acquisition transactions in the Australian market during the period January 1996 to December 2003. To be considered for the sample, transactions had to involve targets listed on the Australian Stock Exchange (ASX).
However, bidders in the sample can originate from any country and can be either exchange-listed or privately owned.
5.1 Data Sources
Two primary sources were used to obtain data on Mergers and Acquisitions in the
Australian market. These were the Connect 4 Takeovers Database and the
Bloomberg LP M&A Database. Taken alone, neither of these sources provided a truly complete record of transactions, however, use of these sources together resulted in a larger and more accurate data set than would otherwise have been possible.
The Connect 4 Database contains data on Australian transactions, where Australian transactions are defined as those involving targets listed on the Australian Stock
Exchange (ASX). The Bloomberg Database contains data on transactions from all over the globe, involving both exchange-listed and privately owned targets. In order to obtain relevant data for this study, a single criterion had to be imposed in searching this database: takeover targets had to be listed on the ASX. Connect 4 contains data on transactions announced on or after January 1, 1997, while
Bloomberg only has complete data on transactions announced after January 1, 1999.
Data on transactions initiated in 1996 was kindly provided by Rubicon Partners,
61 Sydney-based managers of the Rubicon M&A Fund24. Initial stages of the data collation process involved combining the information provided by these three data sources and, where details of a particular transaction were available from more than one of the sources, cross-checking the data for accuracy25 .
5.2 Sample Criteria
To be included in the final sample the proposal must be an offer to obtain a majority interest in an ASX-listed target by a business entity that does not currently hold a majority interest in that target. That is, the bid must be for at least 50.1 percent of the target and the bidder cannot hold a pre-bid stake in the target exceeding 50 percent.
The reason for imposing these two criteria is that our interest lies in analysing bona fide bids for control of a firm, rather than 'mop-up' bids by companies already exercising control or minority-stake investments26. These two criteria reduced the original data set of 524 transactions by 169, leaving 355 transactions.
A further 22 transactions had to be excluded due to an inability to calculate vital bid structure elements such as bid premium and proportion of cash in the offer. Four of these transactions were lost due to unclear bid details or an inability to identify the bidder with certainty, 10 due to the offer including share options or unlisted stock
24 Thanks are due to the principals of Rubicon Partners for granting us access to their data. 25 In the few cases where discrepancies existed between the two primary data sources, a third source was consulted in an attempt to identify the correct data. This was usually the Signal G database of ASX Company Announcements, provided by SIRCA, or a search of media publications via the Factiva database. 26 Although individuals and firms are prohibited from increasing their stake in a listed company from below 20% to more than 20%, or from a starting point that is between 20% and 90%, via ordinary stockmarket purchases, situations do occur where bidders own toeholds greater than 20%. Where such toeholds are less than 50%, we feel they still represent bona fide bids for control of a firm and therefore retain them in the analysis. 62 and eight due to the target's shares being inactive at the time of the bid. Another 14 transactions involved bidders and targets that were significantly related, beyond the holding of a toehold stake, and therefore had to be excluded from the sample. The majority of these transactions involved the 'rebadging' of property trusts, which represented a rearrangement and renaming of assets within the same business. Once again, such transactions cannot be considered as bona-fide control contests.
In some cases, transactions were not consummated for reasons other than bid inadequacy. That is, transaction failure was caused by circumstances unrelated to the offer value and form of payment. Since the focus of this study is on the structure of takeover bids and the ability of bidders to structure their offer so as to pre-empt competing bids, receive target management support and ultimately enjoy bid success, bids that were withdrawn due to reasons other than the inadequacy of the bid had to be excluded from the sample. This resulted in a further 15 transactions being excluded. These transactions involved the bidder withdrawing their bid for a variety of reasons, including the commencement of unrelated legal proceedings against the bidder or target, lack of necessary approval from regulatory bodies such as the
Australian Competition and Consumer Commission (ACCC) and Foreign
Investment Review Board (FIRB), the target being forced into administration, or the post-bid discovery of accounting or other irregularities with the target.
Following the necessary exclusions, the final sample consists of 304 initial takeover bids announced between January 1, 1996 and December 31, 2003.
63 5.3 Data Characteristics
5.3.1 Unprocessed Data
For each transaction, certain data was readily available from the sources described and was included in the data-set without further manipulation. This information included: Takeover Announcement Date, Identities of the Target and Bidder,
Takeover Mode, Percentage of Target shares Sought, Bid Outcome and Final Result,
Directors' Recommendation to Shareholders, Competition Status, Offer Type and
Terms, and Bidder Toehold. These transaction characteristics are defined in the following sections.
5.3.1.1 Takeover Mode
Takeover Mode refers to the legal form used to implement the takeover proposal.
This study includes bids made via all of the three available takeover modes: Part A and Part C offers, and Schemes of Arrangement. The takeover modes, and their distinct features, have been discussed in Chapter 3 above.
5.3.1.2 Bid Outcome
Bid Outcome distinguishes between Successful and Unsuccessful Initial bids, while
Final Result refers to the bidder's stake in the target following completion of the transaction. For the purpose of this study, an initial bid is considered Successful when the Final Result is equal to or greater than the stake initially sought by the bidder, as outlined in the offer documentation. It must also be stressed that our interest lies in the ability to structure successful initial bids. Hence, for the purpose
64 of this study, only initial bids that lead to the acquisition of the target without further revised bids (by the initial bidder) are considered successful. That is not to say that a transaction cannot be successfully completed via future upward-revised bids.
5.3.1.3 Directors' Recommendations to Shareholders and Target Management Resistance
As discussed in Chapter 3, target directors in Australia make direct and explicit recommendations to their shareholders with respect to takeover proposals, usually accompanied by equally unambiguous views about the adequacy of the offer. This particular feature of the Australian M&A market provides an ideal tool for identifying and classifying the response of the target's top management and directors to the takeover proposal. This compares favourably to previous American studies, such as Jennings and Mazzeo (1993), which had to use far less accurate proxies to gauge the response of target management. Most such studies resorted to searching through media sources for statements or evidence of certain actions by target management potentially indicating satisfaction or dissatisfaction with the offer.
Therefore, following Henry (2004), this study uses the Directors' Recommendation to Shareholders as a direct proxy for the response of target management. Reject recommendations will be considered to indicate the presence of target management resistance, while Accept recommendations will be regarded as an indicator of target management support, or the absence of resistance. As mentioned previously, on rare occasions, other less explicit responses such as "Take No Action at this Time" or
"No Recommendation" are made. These are normally interpreted by the market as an admission that the takeover bid is attractive to target shareholders, but perhaps not to managers (Henry, 2004 ). Therefore, in an innovation of this study, all such
65 recommendations will be grouped into a third category, in which the response of target management is considered Neutral. Such recommendations arguably represent an agency conflict where the target directors stop short of explicitly recommending the offer due to self-interest. It is expected that this three-way classification of target management response, using directors' recommendations, provides a more precise proxy for the response of target management than has been used in the majority of past studies. Betton and Eckbo (2000) also performed a three-way categorisation of target management response (supportive, neutral, and opposed), but had to use a far less precise proxy than is used here.
It should also be noted that in the vast majority of cases directors are unanimous in their recommendations to shareholders. However, in the very few instances where directors disagree and make divergent recommendations, for the purpose of this study the majority opinion will be used to represent the response of the target's board.
Merger proposals made via a Scheme of Arrangement are somewhat different to Part
A and Part C offers in that they are inherently friendly transactions that are the result of direct cooperation between bidder and target management. Therefore, in this study, offers made via a Scheme of Arrangement will be classified as having received 'Accept' recommendations from target directors, indicating the absence of resistance.
66 5.3.1.4 Competition Status
Competition Status refers to classifying each bid according to whether or not another bidder entered a new bid, for the same target, after the announcement date of the initial bid in question. An initial bid is only considered to have received a competing bid if the first offer remained open to shareholders on the date the competing bid was announced.
5.3.1.5 Offer Type and Terms
Offer Type refers to the consideration offered by the bidder. Bids in this study fall into three distinct categories: Cash Offers, Scrip Offers, and Mixed Offers, the latter offering target shareholders a combination of cash and listed shares in exchange for their holdings. Offer Terms are used to calculate the value of a takeover offer and include the cash amount and the scrip exchange ratio. Where target shareholders are given more than one payment alternative, the alternative with the highest proportion of cash in the offer is used to calculate bid structure elements in this study. This is often a pure cash alternative. This approach reflects the market perception that cash consideration is favoured by target shareholders.
Additionally, there were seven transactions in which the offer consideration included foreign-listed stock or cash denominated in a foreign currency. In these cases, the relevant exchange rate ( closing price) prevailing on the trading day immediately prior to offer announcement was used for conversion to Australian Dollar values.
67 5.3.1.6 Bidder Toehold
As previously defined, Bidder Toehold refers to the size of the stake (percentage) in the target held by the bidder immediately prior to the announcement of the first offer by that bidder. The impact of toehold-ownership by the initial bidder on the likelihood of observing competing bids, target-management resistance, and bid success is empirically tested in Chapter 7. In line with previous studies, a large proportion of initial bidders in the sample did not own a toehold stake27• Following
Jennings and Mazzeo (1993), we include an intercept effect for toehold, TOEDUM, in addition to the slope effect. Hence, while the TOEHOLD variable represents exact toehold size (percentage), TOED UM is a binary variable taking a value of 1 if the initial bidder acquired prior ownership in the target. It is felt that this added variable will improve the information content of the model
5.3.2 Processed Variable Definitions
For each transaction, certain variables were not readily available from the sources described or, if available, were not calculated in the manner required by this study.
These variables are Bid Premium, Medium of Exchange (Proportion of Cash) and
Target Size. The manner in which these variables were calculated in this study is defined in the following sections.
27 118 out of the 304 initial bidders in the sample used here reported not owning a toehold stake. 68 5.3.2.1 Bid Premium
5.3.2.1.1 Target's pre-bid value
The Bid Premium in a takeover transaction refers to the excess of the offer value over the pre-offer price of the target. More specifically of interest is the premium over the stand-alone value of the target as this is the value that the target's share price is likely to return to in the event that the transaction is not consummated, as reported by Bradley, Desai, and Kim (1983). However, in practice, the value of the target immediately prior to the announcement of an initial takeover offer rarely represents this stand-alone value. Information about an upcoming takeover proposal almost always leaks into the market, causing the price of the target to appreciate as investors purchase the stock, either with the view to capturing the bid premium or to become shareholders in the future merged entity. Jabbour, Jalilvand and Switzer
(2000) identified two sources of the leakage effect. Firstly, investors' anticipation of an impending takeover bid, based on analysis of information in the media, corporate reporting disclosures, or activity by arbitrageurs, can cause pre-bid price run-ups.
Secondly, illegal insider trading activity was found to be related to abnormal returns in the target firm's shares as early as 45 days before the initial acquisition announcement. Therefore, the value of the target immediately prior to the announcement of an initial takeover offer will usually incorporate the likelihood that a successful takeover proposal is made, rather than reflecting the pure stand-alone value of the target company. The extent to which such leakage, and pre-bid run-up in target share price, occurs differs with every transaction, yet its existence is rarely disputed and has been confirmed by several empirical studies.
69 Jennings and Mazzeo (1993) use the target's share price 10 days prior to the initial offer announcement as the 'no leakage' price. Schwert (1996) found that the pre-bid run-up in the share price of the target begins on average 42 trading days prior to the announcement of the initial offer. Betton and Eckbo (2000) use a similar period to account for the leakage effect, using the target's share price prevailing 60 calendar days prior to initial offer announcement to calculate bid premium. Closely consistent with this are Pompilio (2000) and Eckbo and Langohr (1989), who use the target's share price 40 trading days and 56 calendar days, respectively, as the 'no leakage price'. In this study, the target's pre-bid value used to calculate bid premium is its share price 40 trading days prior to the announcement of the Initial offer, pT(T-40/ 8•
However, this price is adjusted for any dividends and capitalisation changes in the target between day (T-40) and the announcement date29• More specifically, where an
Ex-Dividend date occurs in this period, the target's (T-40) price is adjusted downwards by the per-share value of the dividend. The Ex-Dividend date is of significance here as it is on this date that a company's share price adjusts for the value of the dividend, causing the price to fall. Similarly, for capitalisation changes effected between day (T-40) and the announcement date, such as stock splits or bonus issues, the target's (T-40) price is adjusted by replicating the share price effects of the capitalisation change. These adjustments are necessary to ensure that the effects of dividends and capitalisation changes on the target's share price
28 A 'no leakage' period of 40 trading days was chosen following the Australian study of Pompilio (2000) and is closely consistent with that used in most other recent studies. 29 Details of dividends and capitalisation changes were sourced from Bloomberg. 70 between day (T-40) and announcement date do not distort the bid premmm calculation3°.
5.3.2.1.2 Offer Value
The second component required for the calculation of the bid premium in takeover transactions is the value of the offer. For cash offers this is simply the amount of cash offered. However, for scrip offers and mixed offers, this requires knowledge of the bidder's pre-bid share price. Consistent with Jennings and Mazzeo (1993) and numerous other studies, we use the bidder's price on the trading day immediately prior to the initial announcement (T-1) to calculate bid premium for initial bids. This method seems to ignore any potential leakage effect on the bidder's share price but this is justified since the bidder's stand-alone share price is of little relevance to target shareholders and directors when they consider the takeover offer. Instead, their interest lies in the share price of a potential combined firm. Consequently, it can be argued that, from their point of view, of most relevance in assessing the adequacy of an initial scrip bid is the bidder's most recent share price, as this provides the best available estimate of the share price of the potential combined firm. Hence, in this study, the offer value used to compute bid premium is calculated as in equation (5.1) below.
30 Consider a situation where an offer of $0.60 per share is made for a company. At announcement date, 100 million target shares were trading at $0.50 (Bid Premium is 20% ). At T -40, this same company only had 50 million shares outstanding, each trading at $1.00. The difference in shares outstanding is due to a 2:1 stock split occurring at T-2O. Calculating bid premium using the target's share price at T-4O, and without properly accounting for the stock split, would give a bid premium result of -40%. This is clearly incorrect given that Offer Value = $6OOM, while MVE(T-40) = $5OOM. This simplified example illustrates the need to adjust for capitalisation changes taking place between day T-4O and the announcement date. 71 Offer Value (per share) =[Cash per share+ (Exchange RatioxPB(r-1) )] (5.1)
where;
Exchange Ratio = Number of bidder shares offered for each target share
P\r-1) = Bidder's share price on the trading day immediately
prior to the initial announcement
5.3.2.1.3 Bid Premium Calculation
As mentioned previously, the bid premium in a takeover transaction refers to the excess of the offer value over the pre-offer price of the target. It is often expressed in percentage terms and can be written as in equation (5.2) below.
BPREM (%) = [( Offer Value )-l] x lOO (5.2) Target's Pre-bid Value
Using the definitions developed in the previous two sections, bid premium, as calculated in this study, can be expressed as follows:
01. [( [cash per share+ (Exchange Ratio x PBrr-1> )]) 1 BPREM (-;o) = T - 1] x 00 (5.3) p (T-40)
72 5.3.2.2 Medium of Exchange
In this study, the Offer Type and Terms are also used to calculate the Proportion of
Cash in each offer. 87% of offers in the sample are purely cash or scrip offers, made up of 100% and 0% cash respectively. In the case of mixed offers, the proportion of cash is calculated as follows, with the value of bidder scrip being calculated in the manner outlined in section 5.3.2.1.2 above:
PCASH = Cash per share (5.4) [cash per share+ (Exchange Ratio x P 8(r-1) }]
5.3.2.3 Target Size
Following Jennings and Mazzeo (1993), target size is used to proxy for the Cost of
Acquiring Information about a Target. Target size will be measured by the Market
Value of Equity (MVE) 40 trading days prior to the announcement of the initial offer
(T-40) and calculated by multiplying the price prevailing on that date by the number of ordinary shares outstanding on that date. Once again, T-40 is used as the reference date to ensure that target size data is not distorted by the leakage effect31 •
(5.5)
31 Due to unavailability of data, the number of shares outstanding at announcement date, (T-0), was used in the calculation of target size in some transactions, instead of the number outstanding 40 trading days prior to the announcement date. However, checks for capitalization changes and dividends between date T-0 and T-40 were carried out to ensure that these figures were approximately equal.
73 where;
MVET(T-40) = Target firm's Market Value ofEquity 40 trading days prior
to the announcement of the initial offer
pT(T-40 >=Target's share price 40 trading days prior to the
announcement of the initial offer
ST(T-40) = Number of Ordinary Shares outstanding 40 trading days
prior to the announcement of the initial offer
It must also be noted that in the empirical analysis, the log of target-firm size will be used rather than the raw variable, again following Jennings and Mazzeo (1993). This was done because the distribution of the target-size variable was too extreme to use in its raw form 32 •
5.3.3 Descriptive Statistics
Table 5.1 below provides some descriptive statistics on the 304 initial bids used in the empirical analysis in Chapter 7. The vast majority of transactions were initiated via a takeover bid, with only 20% organised via a Scheme of Arrangement. Bid competition was rare, with only 14% of initial bidders having to deal with the entry of a competing bidder. Directors recommended acceptance of the bid in 58% of cases and rejection in 38% of cases. Successful single-bid transactions accounted for
54% of all transactions. However, many more transactions, 84% of the total, were eventually consummated, leading to the transfer of control in the target. That is, some transactions required more than one bid, and perhaps more than one bidder, to
32 All variables were examined for the presence of significant outliers. Using Windsorisation, no such outliers were found and no action was taken. 74 be completed. Cash was the dominant payment method, accounting for 60% of all
bids. Acquirer stock was used in 29% of bids, with the remainder utilising a mixture
of cash and scrip. Bidders had a toehold in 62% of cases. Finally, no significant time
clustering of events is evident in the sample.
TableS.l Descriptive (percentage) statistics summarising a sample of 304 Initial takeover bids between January l, 1996 and December 31, 2003.
Takeover Mode Scheme of Arrangement 61 20% Takeovers 243 80% 304
Medium ofExchange Cash Bids 181 59% Scrip Bids 88 29% Mixed Bids 35 12% 304
Bid Competition Competing Bid Received 42 14% No Competing Bid Received 262 86% 304
Initial Bid Success Successful Bids 165 54% Unsuccessful Bids 139 46% 304
Transaction Outcome Not Consummated 49 16% Consummated 255 84% 304
Director's Recommendation Accept 175 58% Reject 116 38% Other 13 4% 304
Freguency_ Distribution o[Bid Premium Freguency_ Distribution of Calendar Year o[Initial Toehold ProJ2Q_sal BP:50.10 29% TOE=0 38% 0.10 < BP :5 0.20 22% 0.00 < BP :5 0.05 7% 1996 7% 0.20 < BP :5 0.30 16% 0.05 < BP :5 0.10 11% 1997 11% 0.30 ~ BP :5 0.40 10% 0.10 ~ BP :5 0.15 9% 1998 13% 0.40 < BP :5 0.50 10% 0.15 < BP :5 0.20 21% 1999 13% 0.50 < BP :5 0.60 3% 0.20 < BP :5 0.25 4% 2000 17% 0.60 < BP :5 0.70 3% 0.25 < BP :5 0.30 2% 2001 18% 0. 70 < BP :5 0.80 2% 0.30 < BP :5 0.35 3% 2002 11% 0.80 < BP :5 1.0 2% 0.35 < BP :5 0.40 2% 2003 10% BP> 1.0 3% 0.40 < BP :5 0.50 3% (Median BP= 0.20) (Median TOE = 0.086)
75 Chapter 6 Methodology
Due to the nature of the research question, the empirical analysis in this thesis relies entirely on discrete regression models. The most common of these, the logistic regression (LOGIT), is utilised heavily, while the largely similar Probit Regression is also used to perform simultaneous equation analysis.
6.1 Logistic Regression
The Logistic regression model (LOGIT) is one of the most popular of the discrete regression models. As the name of the family of models suggests, the dependent variable in such models can only take discrete values. These models are commonly used for research topics in which the dependent variable is restricted and are particularly useful in analysing whether some event is likely to occur or not. In this thesis we are concerned whether events such as the arrival of a competing bid or the successful completion of a takeover occurred. Examples of other fields where discrete regression models are often used include studies of voting patterns, mortality rates, and business failure.
The dependent variable ( output) from a binary LOGIT model can assume one of two possible values, which are usually denoted O and 1, where the latter represents the event occurring, while the former signifies the non-occurrence of some event. The dependent variable reflects a probability. These probabilities can be denoted as follows:
76 (6.1.1)
The probabilities of these outcomes will be affected by multiple discrete and continuous variables, which need to be included in the model. The linear regression model represents the basis from which a binary choice model is created (Hastie and
Tibshirani, 1995). The linear regression model can be written as:
p 11 = ao + }: a ixii (6.1.2) j-1 where;
17 = the alteration in the probability ;r resulting from a change in xii
X;i = a vector of variables
a i = a vector of coefficients.
The first adjustment that needs to be made to this model is that the coefficients need to be restricted in some way so that the output, reflecting a probability, will always lie in the 0 to 1 range. This is achieved by using a transformation, g(1t), called the link function, that maps the output obtained using the linear regression model (OLS) onto the unit interval without imposing any restrictions on the coefficients (Maddala,
1983, 1992). The logistic link function is one of the most well-known and popular link functions and takes the following form:
77 g(rr) = log{_!!_} 1- rr (6.1.3)
The logistic link function is combined with the classical linear regression model to give the logistic regression (LOGIT) model. The logistic regression model is simply a non-linear transformation of the linear regression and can be derived from equation
(6.1.4) below.
(6.1.4)
The left hand side of equation (6.1.4) is known as the log odds ratio, that is the natural logarithm of the probability of y being equal to one divided by the probability of y being equal to zero. Rewriting the expression in terms of the exponential operator, and solving for n:, we obtain:
(6.1.5) (a0 + .f 1 a J·X··) l] e ]=
1l' = + (a 0 ·-1fa ]·X··) I] l+ e J-
The derived result is the binary LOGIT model, illustrating the relationship between the probability of the outcome denoted 1 and the covariates. Hence:
78 (6.1.6)
The discussion of LOGIT models above has focused on the binary case. However, the number of possible outputs that discrete choice models can consider is by no means limited to two. In fact, extending these models to a multinomial framework takes little effort. Multinomial discrete choice models, such as LOGIT, can consider any number of possible outcomes given that these outcomes are unambiguously and clearly defined.
For example, considering a multinomial LOGIT model in which the output can take any of J possible values, the model simply becomes:
(6.1.7)
In this thesis, we employ both binary LOGIT, in which the dependent variable can assume one of two possible values, and trinomial LOGIT, where the output can take one of three possible values.
79 6.2 Probit Regression
Probit regression is introduced here because a variant of this model is utilised in
LIMDEP to perform simultaneous equation analysis, rather than the previously introduced Logistic regression (LOGIT). This change in method is not expected to affect the substance of the results or comparability across the results since, in practice, Probit models deliver essentially the same results as Logit models in many cases. Instead of using the logistic link function, Probit models utilise the Inverse
Normal link function. This link function is combined with the classic linear regression model to give the Probit regression model, as shown in equation (6.2.1) below:
p
u 1 -12
The difference between the Logit and Probit models lies in their assumptions.
Logistic regression is based on the assumption that the discrete dependent variable reflects an underlying qualitative variable and therefore uses the binomial distribution. On the other hand, Probit regression assumes that the discrete dependent variable reflects an underlying quantitative variable and therefore uses the cumulative normal distribution. However, in the majority of cases, the Probit and
Logit models will yield results that are very close to one another, especially in large samples (Maddala, 1983).
80 6.3 Simultaneous Equation Analysis - Bivariate Probit
In this thesis, simultaneous equation analysis is utilised to examine the possibility that some decisions in the takeover process are determined simultaneously. The simultaneous equation function in the LIMDEP statistical package allows us to carry out this analysis using the Bivariate Probit modet33•
The Bivariate Probit model consists of two simultaneous equations. For example, in
Section 7.4.2., the first equation models the binary outcome, 'bid success', of each bid, y *Ii , while the second relates to the binary decision of whether to resist a bid.
Under this model y *1; and y *2 ; can be defined as follows:
(6.3.1)
(6.3.2)
where;
x ji = 1 x k i vectors of explanatory variables, j = 1,2
pi = a vector of coefficients
* = subscript used to indicate an unobserved variable.
33 Recently, the bivariate probit model has been used in studies of financial economics (Boyes, Hoffman, and Low (1989), Roszbach (2003), and Jacobson and Roszbach (2003)), social economics (Evans and Schwab (1995), Ettner (1996), and Evans, Farrelly and Montgomery 1999)), and labour economics (Christofides, Stengos, and Swidinsky 1997).
81 Additionally, the disturbances are assumed to be bivariate normally distributed, with
E[t:J= E[t: 2 ]= 0, Var[t:J= Var[s 2 ]= 1, and Cov[t:i,t: 2 ]= p.
The results of the Bivariate Probit models are reported in Tables 7.4 and 7 .5 and include the estimated coefficient of the p statistic. If p = 0 , then the model estimated simply consists of independent and unrelated probit equations. If this is the case, the equations should be estimated separately and the bivariate probit is not applicable
(Greene, 1983). The null hypothesis being tested is that p = 0 and the associated p- value is reported in brackets beside the estimated p statistic.
The logistic regression model is used in this thesis to determine the impact of bid structure on the likelihood of competition, resistance, and bid success, separately.
However, the value of using simultaneous equation analysis lies in the ability to analyse any interaction between these outcomes. More specifically, the possibility that target management resistance influences the likelihood of observing competing bids and initial bid success can be tested.
82 Chapter 7 Empirical Tests and Results
This chapter will present the empirical results and relate these results to the hypotheses outlined in chapter four. The empirical testing in this chapter is performed on two sets of data. Firstly, the entire sample, which includes transactions of all type, is tested. Secondly, transactions undertaken via a Scheme of
Arrangement are excluded from this full sample to produce a 'Takeovers only' sub sample which is more representative of hostile or unsolicited offers, since it excludes the inherently friendly transactions organised via Schemes of Arrangement. The empirical tests were structured in this manner for two reasons. Primarily, by comparing the two sets of results, we can test the proposition that some elements of bid structure are merely proxies for transaction-type (Jennings and Mazzeo (1993),
Pompilio (2000)). Secondly, while some of the existent theoretical models are equally applicable to transactions of all type, others focus their analysis, and empirical predictions, on either mergers or takeovers. Hence, structuring the testing in this manner provides empirical results that allow for a better comparison with the available predictions from theory.
7 .1 Competing Bids
Table 7.1 presents results of a logistic regression analysis of the likelihood of encountering bid competition.
83 Table 7.1 Results of a Logistic regression examining the likelihood of a competing bid arising using the bid premium, the proportion of cash in the offer, size of the toehold, existence of a toehold, and target size as explanatory variables from a sample of 304 initial bids between January 1996 and December 2003 1•2•
Sample Sample All Bids 304 -2.6153 0.0012 0.1486 -0.0718*** 1.0491 ** 0.0364 (0.1627} (0.7601} (0.6983} (0.0060} (0.0303} (0.7147} LLR =9.7102 (0.0839}* HL =2.6618 (0.9538} Takeovers 243 -3.7226* 0.0054 -0.1142 -0.0733*** 1.1064** 0.1063 (0.0684} (0.3273} (0.7858} (0.0074) (0.0383} (0.3340} LLR = 11.0732 (0.0499}** HL =2.6599 (0.9539} 1 The equation estimated is: COMP= ao+ a1(BPREM) + a2(PCASH)+ a3(TOEHOLD)+ il4(TOEDUM)+ a5(LNSIZE) + E1 (7.1) where COMP = binary variable taking a value of 1 if a competing bid arises BPREM = percentage bid premium PCASH = fraction of the offer in cash TOEHOLD = fraction of the target firm owned by the initial bidder before making the offer TOEDUM = binary variable taking a value of 1 if the initial bidder acquired prior ownership in the target LNSIZE = log of the target firm's market value of equity 2 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. ***, **, * denotes significance at the 1 %, 5% and 10% level respectively. In addition to the ordinary t statistics, and their corresponding P values, the Likelihood Ratio34 and Hosmer-Lemeshow35 test statistics are reported for the regressions estimated in this thesis. The first competition hypothesis (H-la) suggests that an inverse relationship should exist between the likelihood of a competing bid emerging and the size of the initial bid premium. However, the results provide no support for this proposition as the sign of the estimated bid premium coefficient (a1) is positive, although insignificant. While contrary to hypothesis (H-la), this result is 34 The Likelihood Ratio Test (LLR) is a joint-test to determine whether all the coefficients are equal to zero. We wish to reject this hypothesis and are able to do so for all the models estimated in the main body of this thesis. 35 The Hosmer-Lemeshow Test examines whether the model is a good fit for the underlying data. We wish to accept this hypothesis and are able to do so for all the models estimated in the main body of this thesis. 84 consistent with the undervaluation motive for takeovers. Proponents of this theory claim that in some instances bidder-firm managers have access to 'special information' that the market does not have, perhaps via their industry experience and knowledge, which allows them to identify undervalued targets (Halpern, 1983). Initial bidders with special information may choose a high-premium bid as a means of pre-empting potential competition and securing a significantly undervalued target. In so doing, the initial bidder may in fact reveal the undervaluation to the market, leading previously uninformed rivals to investigate the situation and lodge a competing bid for the target. This sequence of actions, beginning with a pre-emptive bid for a significantly undervalued target, could potentially explain the result in Table 7.1, which suggests that high-premium bids do not deter competing bidders. Motivated by the theoretical work of Fishman (1989), hypothesis H-lb suggests the likelihood of encountering competition should be negatively related to the proportion of the initial offer consisting of cash. However, the results in Table 7.1 lend no support to the stated hypothesis. This finding is entirely consistent with existing empirical evidence, which finds that greater use of cash in the offer does not serve to deter competing bidders (Jennings and Mazzeo (1993), Pompilio (2000), and Betton and Eckbo (2000)). Hence, in finding against hypothesis H-lb, this thesis adds to the empirical evidence that refutes the payment method prediction of Fishman (1989). Theoretical models developed by the likes of Burkart (1995), Singh (1998) and Bulow et al. (1999) all show that bidders who own toeholds always overbid in takeover contests. Furthermore, the incentive to overbid, and the observed overbidding, increases with the size of their toehold. A natural implication of this 85 aggressive bidding strategy is that an inverse relationship will exist between the likelihood of bid competition and the size of the initial bidder's toehold stake. The results in Table 7.1 show that TOEHOLD is indeed negatively and significantly related to the likelihood of competition, at the 1 % level. This result lends strong support to hypothesis H-lc: that ownership of large toeholds by initial bidders serves to pre-empt or deter competing bidders, a notion confirmed empirically by Betton and Eckbo (2000) and Pompilio (2000). However, the result with respect to TOEDUM, the dummy term, is somewhat contradictory to this notion. TOEDUM is found to be positively and significantly related to the likelihood of competition, at the 5% level, which suggests that ownership of a toehold stake is positively associated with the frequency of observing competition. Hence, the empirical evidence on the pre-emptive qualities of initial toehold stakes implies a joint result that is supportive of hypothesis H-1 c. Taken together, the above results suggest that initial bidders with small toeholds are more likely to encounter competition than bidders with zero-toeholds, however, as the size of the initial bidder's toehold increases, the likelihood of competition declines. We propose a simple signalling hypothesis that provides a potential explanation for this joint result. Presumably, toehold-owning bidders have an informational advantage with respect to the operations and prospects of the firm in which they hold a stake, compared to non-shareholders. Thus, when such a shareholder makes a takeover bid, a positive signal is released to the market regarding the value of the target firm. Relating this hypothesis to the likelihood of observing competition, one could argue that initial bidders who own toeholds are more likely to encounter competition that zero-toehold bidders due to the 86 undervaluation signal their bid conveys. Assuming that the level of superior information is increasing in the size of the bidder's toehold, the strength of the undervaluation signal should be positively related to toehold size and thus toehold size should be positively related to the likelihood of bid competition. However, in reality, this positive association would be truncated at a certain toehold level due to the competition-deterring effect of large toeholds (Burkart (1995), Singh (1998) and Bulow et al. (1999)). In other words, under the proposed hypothesis, an offer by a bidder with a larger toehold might convey a stronger signal that the target is undervalued, attracting potential competing bidders. However, the likelihood of observing a competing bid would be significantly off-set by the difficulty of competing with a bidder who already owns a large stake in the target. In summation, this simple signalling hypothesis provides an explanation for the toehold results in Table 7.1, suggesting that initial bidders with small toeholds are more likely to encounter competition than bidders with zero-toeholds, however, as the size of the initial bidder's toehold increases, the likelihood of competition declines. The final coefficient (a5) in Table 7.1 shows that the relation between target firm size and the likelihood of bid competition, although positive as predicted, is statistically insignificant. Assuming that target firm size is an adequate proxy for the cost (faced by a potential bidder) of acquiring information about the target, this result is not supportive of hypothesis H-ld. However, it is consistent with Australian evidence provided by Pompilio (2000) and we suggest that such a relationship cannot be found in the Australian market due to its small size, particularly with respect to listed companies. Quite simply, as target-size increases, the number of potential acquirers falls dramatically. This is the case because size is still the major barrier to takeover, 87 despite recent innovations in takeover financing. On the other hand, in substantially larger markets such as the U.S., where there are numerous listed firms in most industries, the number of potential suitors will be larger across all target-size levels. Unsurprisingly, the theoretical literature motivating hypothesis H-ld, provided by Fishman (1988, 1989) and Hirshleifer and Png (1989), and the sole empirical evidence confirming it (Jennings and Mazzeo, 1993) originates in the U.S. market36. We now turn our attention to the results obtained from the 'Takeovers' sub-sample in Table 7.1. These results do not appear to be materially different from those of the 'All Bids' sample discussed above, with the exception of the coefficient on PCASH. The sign of this coefficient is different between the two samples examined, although statistically insignificant in both cases. These results appear to contradict the proposition that payment method is in fact a semi-proxy for transaction type (Pompilio, 2000). Pompilio proposes that cash offers are more often associated with hostile transactions, whereas stock offers are more likely to represent friendly transactions arranged cooperatively by target and bidder management. This implies that the association between proportion of cash in the offer and the likelihood of competition should be more positive when examining hostile transactions as compared to friendly transactions. However, the results in this study are contrary to this proposition. The 'Takeovers' sub-sample, which contains a higher proportion of hostile or unsolicited transactions, actually showed a somewhat negative relationship between proportion of cash and competition. 36 The use of alternate proxies for the cost of acquiring information about a target, such as the number of equity analysts following the target firm (Jennings and Mazzeo (1993)), are unlikely to change our results since these proxies will be highly correlated with the utilised proxy of target-firm size. 88 In summation, results from a binomial logistic regression analysis show that the likelihood of a competing bid arising is higher for initial bidders with small toeholds, as compared to bidders with zero-toeholds, however, as the size of the initial bidder's toehold increases, the likelihood of competition declines. However, initial bid premium size, proportion of cash in the initial offer, and the cost of acquiring information about the target, as proxied by target-firm size, are not found to be significantly related to the likelihood of observing bid competition37• For further confirmation of the robustness of these results, following Jennings and Mazzeo (1993), we also estimate these effects separately via three logistic regressions. The results obtained serve to confirm the relationships found in the combined analysis in Table 7.1. The results for these three models, as well as the corresponding models for the resistance and bid outcome analyses, can be found in Appendix B. 7 .2 Target Management Resistance In this thesis, a three-way classification is used to record target management's response to a takeover offer. Consequently, the resistance decision is modelled using a trinomial logistic regression, as reported in Table 7.2 below. Considering Panel A first, we find bid premium size to be negatively related to the likelihood of encountering target-management resistance. This result is statistically significant at the 5% level and is supportive of hypothesis H-2a. Additionally, this result is consistent with existing empirical evidence from the U.S. (Jennings and Mazzeo 37 Hypothesis H-le, which suggests that the likelihood of competition is positively related to the presence of target management resistance towards the initial offer, is dealt with in section 7.4 below. 89 (1993)) and Australia (Eddey and Casey (1989)). Interestingly, despite seeming somewhat obvious, the significant negative relationship between bid premium size and resistance has only been found in approximately half of the empirical studies examining it. Table 7.2 Results of a Logistic regression analysis exammmg the likelihood of target management resistance using the bid premium, the proportion of cash in the offer, size of the toehold, existence of a toehold, and target size as explanatory variables from a sample of 304 initial bids between January 1996 and December 2003 1•2• Sample Sample Uz (l5 Grou Size Panel A: Characteristics of Probability [Y=l] All Bids 304 -0.6845 -0.0097** 0.0385 -0.0509*** 1.8084*** -0.0074 (0.6242) (0.0139) (0.8925) (0.0018) (0.0000) (0.9217) Takeovers 243 -2.2829 -0.0123*** -0.2162 -0.0515*** 1.3949*** 0.1309 (0.1373) (0.0077) (0.4972) (0.0037) (0.0012) (0.1194) LLRALLBIDS = 39.3825 (0.0000)*** Panel B: Characteristics of Probability [Y=2] All Bids 304 4.0021 -0.0055 1.2689 0.0023 0.6974 -0.4493** (0.2428) (0.5060) (0.1302) (0.9473) (0.4772) (0.0206) Takeovers 243 2.7607 -0.0063 1.0359 0.0045 0.2829 -0.3361 * (0.4226) (0.4843) (0.2156) (0.8956) (0.7750) (0.0882) LLRrAKEOVERS = 29.3186 (0.0011)*** 1 The equation estimated is: REST= ao+ a 1(BPREM) + a2(PCASH)+ a3(TOEHOLD)+ a4(TOEDUM)+ a5(LNSIZE) + e:5 (7.2) where REST = trinomial variable taking a value of O if no resistance, 1 if target management resisted, and 2 if a neutral response was given BPREM = percentage bid premium PCASH = fraction of the offer in cash TOEHOLD = fraction of the target firm owned by the initial bidder before making the offer TOEDUM = binary variable taking a value of 1 if the initial bidder acquired prior ownership in the target LNSIZE = log of the target firm's market value of equity 2 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. ***, **, * denotes significance at the 1 %, 5% and 10% level respectively. The next variable examined is the fraction of the offer denominated in cash, which displays a positive coefficient, but is statistically insignificant. Although clearly unsupportive of hypothesis H-2b, this result is consistent with extant empirical 90 evidence. The Australian study of Henry (2004) reported a significant positive relation between cash offers and target management resistance, while Jennings and Mazzeo (1993) found the relationship between payment method and resistance in their sample to be weak and not robust to the examination of mergers. Consistent with hypothesis H-2c, the ownership of larger toeholds by initial bidders is shown to significantly reduce the likelihood of target resistance. This negative relationship is statistically significant at the 1 % level and is consistent with all available empirical evidence (Eddey and Casey (1989), Jennings and Mazzeo (1993), Betton and Eckbo (2000), and Henry (2004)). However, as in the competition analysis, the TOEDUM variable is positive and significant, indicating that ownership of a toehold by initial bidders is positively associated with the frequency of observing target-management resistance. Taken together, estimates of the two toehold coefficients suggest that initial bidders with small toeholds are more likely to be resisted than bidders with zero-toeholds, however, as the size of the initial bidder's toehold increases, the likelihood of resistance declines. Again, this joint result is largely supportive of hypothesis H-2c and is fully consistent with Jennings and Mazzeo (1993), who also used two toehold variables. Jennings and Mazzeo propose a partial explanation for the joint-result, claiming that small toeholds are associated with hostile bids, while bidders with larger toeholds, whether friendly or hostile, are difficult to resist. Presumably these small-toehold bidders would only acquire a small pre-bid stake so as not to arouse suspicion in the market concerning their intended hostile bid. Our data set is, however, unsupportive of this proposition, with bidders in hostile or unsolicited transactions acquiring larger toeholds than 91 bidders in friendly transactions38. An additional implication of the argument put forward by Jennings and Mazzeo (1993) is that zero-toehold bidders are more frequently associated with friendly bids than small-toehold bidders, presumably reflecting the fact that, unlike in hostile transactions, securing a toehold is not vital to success in a friendly transaction. Our data set is supportive of this proposition39• An alternate explanation for these results lies in the signalling-hypothesis outlined in section 7.1. That is, initial bidders with small-toeholds are more likely to encounter target-management resistance than zero-toehold bidders due to the undervaluation signal their bid conveys. In other words, the new information revealed by a bid from these informed investors may increase the likelihood of target-management resistance, if for no other reason than to solicit competing bids. Finally, contrary to hypothesis H-2d, the relation between target-firm size and the likelihood of target resistance is found to be negative, although statistically insignificant. Thus, the cost of acquiring information about a target does not seem to be related to the likelihood of resistance. Panel B of Table 7.2 essentially considers bids that received a neutral response (REST = 2) and compares them, on aggregate, to those bids that were not resisted 38 The median toehold acquired by bidders in Takeovers was 10.1 %, compared to 0% in the friendly transactions organised via a Scheme of Arrangement (Mean is 11.8% for Takeovers and 5.4% for Schemes). We can also use directors' response to proxy for transaction hostility, where an accept recommendation signifies a friendly transaction, while all other recommendations indicate hostility. The results using this proxy are similar. The median toehold acquired by bidders in hostile transactions was again 10.1 %, compared to 0.7% in the friendly (Accept) transactions (Mean is 11.0% for hostile transactions and 10.2% where an Accept recommendation was made). 39 73% of bidders in the friendly Scheme transactions were zero-toehold bidders, compared to 30% of bidders in Takeovers. The result is similar, although less pronounced, using the directors' response proxy. 47% of bidders in the friendly (Accept) transactions were zero-toehold bidders, compared to 26% of bidders in the other transactions. 92 (REST =0). Interestingly, LNSIZE is returned as the only statistically significant variable. This indicates that accepted bids, on aggregate, are not structured differently to bids that receive a neutral response from target-management, in terms of bid premium, payment method, or toehold size. The estimated coefficient on the LNSIZE variable is negative, indicating a positive relationship between likelihood of resistance and cost of acquiring information about a target. This is again contradictory to Hypothesis H-2d and suggests that this information cost tends, on aggregate, to be higher in transactions where target management proffer neutral responses, as compared to non-resisted bids. It must however be noted that the full 'All bids' sample included only 13 neutral target-management responses (REST = 2), a very small proportion of the whole sample 4°. We now tum our attention to the results obtained from the 'Takeovers' sub-sample in Table 7.2. These results do not appear to be materially different from those of the 'All Bids' sample discussed above, with two exceptions. Firstly, similarly to the competition analysis, the sign of the estimated coefficient on PCASH is different between the two samples examined, although statistically insignificant in both cases. These results again contradict the proposition that payment method is in fact a semi proxy for transaction type (Pompilio, 2000). Pompilio proposes that cash offers are more often associated with hostile transactions, whereas stock offers are more likely to represent friendly transactions arranged cooperatively by target and bidder management. Our data set provides some support for this summary proposition41 • 40 The full sample of 304 initial bids consisted of 116 resisted bids, 175 bids that were accepted (not resisted), and only 13 bids receiving neutral responses from target-firm management. 41 The average proportion of cash used by bidders in takeovers was 67%, compared to 43% in the friendly transactions organised via a Scheme of Arrangement (Median is 100% for takeovers and 18% for Schemes). 93 However, a further implication of Pompilio's proposition is that the association between proportion of cash in the offer and the likelihood of target resistance should be more positive when examining hostile transactions as compared to friendly transactions. However, the results in this study are not supportive of this proposition. On the contrary, the sign of the estimated coefficient on PCASH in the 'Takeovers' sub-sample, which contains a higher proportion of hostile or unsolicited transactions, is negative, while the coefficient estimated in the 'All Bids' sample, which includes the inherently friendly Scheme transactions, showed a positive sign. However, this is a weak result since the estimated coefficients are statistically insignificant in both cases. The second material difference between the results from the two samples relates to the LNSIZE variable. Whereas a negative relation with likelihood of resistance is found in the 'All Bids' sample, this changes to a positive relationship when the 'Takeovers' sub-sample is considered. This result, which is almost significant at traditional levels, is consistent with hypothesis H-2d. This result seems to suggest that the cost of acquiring information about the target only affects target management's decision to resist in hostile transactions. This makes intuitive sense since target management will not resist friendly offers, particularly those organised cooperatively via Schemes of Arrangement, regardless of the information-cost to other potential bidders. Hence, it seems likely that the hypothesised negative relationship between likelihood of resistance and information cost, as proxied by a positive relation with LNSIZE, is not present in the 'All Bids' sample due to the influence of the friendly Scheme transactions. It can be seen that when these 94 transactions are excluded to form the 'Takeovers' sub-sample, the hypothesised relationship is found. In summation, results of the trinomial logistic regression show that the likelihood of encountering target management resistance is higher for initial bidders with small toeholds, as compared to bidders with zero-toeholds, however, as the size of the initial bidder's toehold increases, the likelihood of resistance declines. Additionally, initial bid premium size is found to be significantly negatively related to the likelihood of resistance, while the cost of acquiring information about a target-firm is also somewhat negatively related to resistance, but only when largely unsolicited transactions are considered. 7 .3 Bid Success Table 7.3 presents results of a logistic regression examining the likelihood of bid success using bid structure elements as explanatory variables. Consistent with hypothesis H-3a, bid premium size is found to be positively and significantly related to the likelihood of bid success, at the 5% level. In other words, consistent with extant empirical evidence (Walkling (1985), Betton and Eckbo (2000), and Officer (2003)), higher-premium initial bids are found to have a greater likelihood of success. The proportion of cash in the offer is found to be negatively related to the likelihood of bid success, although not significantly so. Although contrary to hypothesis H-3b, and the related theoretical work motivating it, this result is entirely consistent with the extant empirical evidence. Bates and Lemmon (2003) similarly found deal success to be somewhat positively related to the use of equity consideration, whilst Betton and Eckbo (2000) and Officer (2003) found payment 95 method to be insignificant in explaining offer outcome. Additionally, this result is consistent with evidence from the previous two sections, which show that bids offering a higher proportion of cash are somewhat more likely to encounter bid competition and target-management resistance. Table 7.3 Results of a Logistic regression examining the likelihood of bid success using the bid premium, the proportion of cash in the offer, size of the toehold, and existence of a toehold as exrlanatory variables from a sample of304 initial bids between January 1996 and December 20031•• Sample Sample Clz Group Size All Bids 304 0.4059* 0.0094** -0.2453 0.0605*** -1.5069*** (0.0823) (0.0126) (0.3684) (0.0002) (0.0001) LLR = 25.7700 (0.0000)*** HL = 3.7715 (0.8771) Takeovers 243 -0.4282 0.0118*** 0.0708 0.0628*** -1.1735*** (0.1437) (0.0082) (0.8200) (0.0005) (0.0062) LLR = 20.7578 (0.0004)*** HL = 6.9222 (0.5451) 1 The equation estimated is: OUT= !lo+ a 1(BPREM) + a2(PCASH) + a3(TOEHOLD) + 14(TOEDUM) + E6 (7.3) where OUT = binary variable taking a value of 1 if the initial bid is successful BPREM = percentage bid premium PCASH = fraction of the offer in cash TOEHOLD = fraction of the target firm owned by the initial bidder before making the offer TOEDUM = binary variable taking a value of 1 if the initial bidder acquired prior ownership in the target 2 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. * * *, * *, * denotes significance at the 1 %, 5% and 10% level respectively. TOEHOLD is positive and again significant at the 1 % level, indicating that initial bidders with larger toeholds are more likely to enjoy bid success. This result is therefore supportive of hypothesis H-3c and is also consistent with prior empirical evidence (Walkling (1985), Betton and Eckbo (2000), and Officer (2003)). However, mirroring the results from the competition and resistance analyses, the estimated coefficient on TOEDUM is again of the opposite sign to TOEHOLD and is significant at the 1% level. This result suggests that the existence of a toehold is 96 negatively associated with the likelihood of initial bid-success. Taken together, estimates of the two toehold coefficients suggest that initial bidders with small toeholds are less likely to enjoy bid success than bidders with zero-toeholds, however, as the size of the initial bidder's toehold increases, the likelihood of success increases. This result may reflect the potential impact of toehold size on target management resistance and bid competition, as proposed in the previous two sections. Considering the 'Takeovers' sub-sample in Table 7.3, we find results strikingly similar to the resistance and competition analyses. Again, the coefficient on PCASH stands out as the only result found to materially differ when compared to the 'All Bids' sample discussed above. When going from the full-sample to the 'Takeovers' sub-sample, the sign of the coefficient on PCASH is again reversed, in this instance from negative to positive, although remaining statistically insignificant in both cases. In summation, the above results show that the likelihood of enjoying bid success is higher for initial bidders with zero toeholds, as compared to bidders with small toeholds, however, as the size of the initial bidder's toehold increases, the likelihood of bid success increases. Additionally, initial bid premium size is found to be significantly positively related to the likelihood of bid success. Noticeably, the results in Table 7.3 seem to reflect those of the competition and target-management resistance analyses. That is, bid structure variables that are found to reduce the likelihood of bid competition and target-management resistance are 97 also shown to increase the likelihood of bid success, and vice versa 42• Hence, it is possible that the relation between these bid structure variables and the likelihood of bid success is driven by the effect these variables have on the likelihood of observing bid competition and/or target-management resistance. For example, the relationship between the two toehold variables (TOEHOLD and TOEDUM) and the likelihood of bid success might simply reflect the signalling effect of small-toehold bidders on the likelihood of encountering bid competition, a theory developed in section 7.1. The following section contains analysis which, amongst other things, allows for this possibility. 7 .4 Simultaneous Decision Analysis 7 .4.1 Simultaneous Determination of the Resistance and Competition Decisions As mentioned in earlier chapters of this thesis, the bid premiums paid in corporate acquisitions tend to be significantly higher in competitive transactions. Being aware of this outcome, it would be reasonable to expect that, at least in some transactions, target-firm management may take actions aimed at soliciting offers from competing bidders. It has been proposed that one such action available to target management is to resist the initial offer, presumably indicating to other potential acquirers that they are open to alternate bids. Based on this theory, several authors have examined the resistance and competition decisions simultaneously (Baron (1983), Giammarino and Heinkel (1986), Hirshleifer and Png (1989), and Berkovitch and Khanna (1990)). 42 The one exception to this inverse association is the observed positive relation between bid premium size and likelihood of bid competition. 98 Essentially, these authors propose that target management may resist an initial takeover offer as a means of encouraging higher alternate bids. We can test this proposition by estimating the simultaneous equation system reported in Table 7.4 below using a Bivariate Probit model. This equation system combines the previous Table 7.4 Results of a Bivariate Probit regression analysis of the likelihood of a competing bid and the likelihood of target management resistance using the bid premium, the proportion of cash in the offer, size of the toehold, existence of a toehold, and target size as exogenous variables from a sample of304 initial bids between January 1996 and December 20031•2• Sample Sample 1fo ac1 ac2 ac3 ac4 CJ.Cs Grou Size All Bids 304 -1.7685 0.0033 0.1544 -0.0227** 1.4726*** 0.0058 (0.1068) (0.2444) (0.5230) (0.0243) (0.0074) (0.9194) Takeovers 243 -0.9813 0.0075*** 0.1132 -0.0089 2.1904*** -0.0573 (0.2580) (0.0094) (0.5718) (0.2989) (0.0000) (0.2480) CJ.Ro a\ a\ a\ a\ a\ All Bids 304 -0.7901 -0.0053** -0.0394 -0.0322*** 1.0903*** 0.0160 (0.3354) (0.0411) (0.8203) (0.0012) (0.0000) (0.7162) Takeovers 243 -2.0675 -0.0059** -0.1742 -0.0274*** 0.6294*** 0.1197** {0.0201} {0.0331} {0.3722} {0.0042} {0.0014} {0.0152} p All Bids = -0.4263 (0.2661) Prakeovers = -0.9995 (0.0000) 1 The equations estimated are: COMP= ac0 + ac1(BPREM) + acz(PCASH)+ ac3(TOEHOLD)+ ac4(BIREST) (7.4.1) + ac s(LNSIZE) + £1 BIREST =a\+ a\(BPREM) + a\(PCASH) + a\(TOEHOLD) + a\(TOEDUM) (7.4.2) + aR 5(LNSIZE) + Es where COMP = binary variable taking a value of 1 if a competing bid arises BIREST = binomial variable taking a value of 1 if target management resists, 0 otherwise BPREM = percentage bid premium PCASH = fraction of the offer in cash TOEHOLD = fraction of the target firm owned by the initial bidder before making the offer TOED UM= binary variable taking a value of 1 if the initial bidder acquired prior ownership in the target LNSIZE = log of the target firm's market value of equity 2 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. ***, **, * denotes significance at the 1 %, 5% and 10% level respectively. 99 resistance and competition analyses, utilising a linking variable between the COMP and BIREST variables. The binomial variable BIREST replaces the trinomial REST variable utilised in the separate analysis of resistance 43 . The vast majority of the relationships discovered in the separate analyses of COMP and REST are maintained, with the exception of the estimated coefficients on PCASH and LNSIZE in equation (7.4.2). The signs of these coefficients are different to those noted in the separate analyses. However, the relations were found to be statistically insignificant in both analyses. It must also be noted that, whilst included as an explanatory variable in the separate analysis of COMP performed earlier, TOEDUM is excluded from equation (7.4.1). The reason for this is primarily econometric in nature. More specifically, we found that the TOED UM and BIREST variables are both significantly positively related to COMP, the likelihood of observing competition. As a result, including these two variables in the same equation lead to misspecification. Although not ideal, removing TOEDUM from sub-equation (7.4.1) was considered the preferred solution to this problem. Importantly, this issue needed to be addressed if the primary purpose of estimating the simultaneous equation analysis, which was to determine the relationship between COMP and BIREST, was to be achieved. Following the necessary exclusion of TOEDUM, the estimated coefficient on BIREST (ac4) shows a positive relationship with COMP that is significant at the 1 % level. Hence, we find that the likelihood of a 43 Those management recommendations classified as 'Neutral' in the separate Resistance analysis (REST=2) are merged with the 'No Resistance' recommendations. Together, these observations are classified as BIREST=O ('No Resistance') in the simultaneous equation analysis. The rationale for this classification is that 'Neutral' bids are more representative of 'No Resistance' bids (as compared to 'Resisted bids') since both are adequately structured but the former are often not accepted by management for self-interest reasons. This is done because when estimating simultaneous equation systems using the LIMDEP software, all dependent variables must be binomial. 100 competing offer emerging is significantly higher m transactions where target management resist the initial bid, a result which is consistent with the afore mentioned propositions of Giammarino and Heinkel (1986) and Hirshleifer and Png (1989). 7.4.2 Simultaneous Determination of the Resistance Decision and Initial Bid Outcome Numerous empirical studies suggest that perhaps the largest single influence on the outcome of a takeover offer is the response of target-firm management (Walkling (1985), Eddey and Casey (1989), and O'Sullivan and Wong (1998)). These studies find that initial offers that encounter target management resistance are significantly less likely to succeed than those offers endorsed by target management. Again, simultaneous equation analysis allows us to determine whether our data supports this conclusion. The equation system estimated, and the results, are reported in Table 7.5 below. Similarly to sub-equation (7.4.1) in the previous section, the combination of the TOEDUM and BIREST variables resulted in misspecification. Again, removing TOEDUM from sub-equation (7.5.1) was considered the preferred solution to this problem. We can see in Table 7.5 below that the estimated coefficient on BIREST (ac4) shows a negative relation with OUT that is significant at the 1% level. Consistent with the afore-mentioned empirical evidence, we find that the likelihood of an initial bid being successful is reduced significantly in transactions where target management resist the initial bid. This does not necessarily mean that the initial bidder will not successfully acquire the target via a future revised bid, however, this will almost 101 certainly take place at a higher acquisition cost and may involve competing bidders. Hence, this result highlights the importance of gaining target-management support for an initial bid, since doing so increases the likelihood of an expeditious single-bid acquisition significantly. Table 7.5 Results of a Bivariate Probit regression analysis of the likelihood of bid success and the likelihood of target management resistance using the bid premium, the proportion of cash in the offer, size of the toehold, existence of a toehold, and target size as exogenous variables from a sample of304 initial bids between January 1996 and December 2003 1•2• a.c Sample Sample a.co a.c, a.c2 3 a.c4 Grou Size All Bids 304 1.0798*** 0.0016 -0.2472 0.0146* -2.6924*** (0.0000) (0.5967) (0.2652) (0.0527) (0.0000) Takeovers 243 0.7519 0.0024 -0.1250 0.0163* -2.3113*** (0.1255) (0.5311) (0.6232) (0.0721) (0.0006) UR a.\ a.\ a.\ a.\ a.\ All Bids 304 -0.4139 -0.0056** -0.0334 -0.0327*** 1.0910*** -0.0043 (0.5892) (0.0317) (0.8486) (0.0011) (0.0000) (0.9159) Takeovers 243 -1.5444* -0.0073** -0.1872 -0.0326*** 0.8650*** 0.0867* (0.0786} (0.0130} (0.3487} (0.0019} (0.0009} (0.0726} p All Bids = 0.5489 (0.2478) Prakeovers = 0.34 70 (0.5357) 1 The equations estimated are: OUT = a.c0 + a.ci(BPREM) + a.c2(PCASH)+ a.c3(TOEHOLD)+ a.c4(BIREST) + £9 (7.5.1) BIREST = a.\+ a.\(BPREM) + a.R2(PCASH)+ a.\(TOEHOLD)+ a.R4(TOEDUM) (7.5.2) + a.\(LNSIZE) + £10 where OUT = binary variable taking a value of 1 if the initial bid is successful BIREST = binomial variable taking a value of 1 if target management resists, 0 otherwise BPREM = percentage bid premium PCASH = fraction of the offer in cash TOEHOLD = fraction of the target firm owned by the initial bidder before making the offer TOED UM= binary variable taking a value of 1 if the initial bidder acquired prior ownership in the target LNSIZE = log of the target firm's market value of equity 2 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. ***, **, * denotes significance at the 1 %, 5% and 10% level respectively. 102 Chapter 8 Conclusion 8.1 Summary of Results The most desirable outcome sought by a firm launching a takeover offer is initial bid success. An initial bidder will increase his chances of achieving this outcome by successfully clearing two transaction-specific hurdles: the securing of target management support and the deterrence of potential rival bidders from entering the takeover contest. Put simply, the initial offer will likely represent an acquisition cost that the bidder considers favourable. However, if the initial offer needs to be revised upwards, due to the presence of target resistance or competition from rival bidders, or both, the expected profitability of the transaction naturally declines. Extant theoretical and empirical literature suggests that initial bidders can increase their chances of successfully negotiating the two hurdles mentioned above, as well as achieving the ultimate goal of initial bid success, via the deliberate choice of bid structure. Focusing on the key bid structure elements of bid premium, medium of exchange, and bidder toehold, this thesis provides a comprehensive empirical test of this claim. The impact that each of these elements has on the likelihood of observing target management resistance, encountering competition from rival bidders, and ultimately enjoying bid success is examined thoroughly. Beginning with initial bid premium, the results indicate, as hypothesised, that high premium bids are associated with a lower likelihood of target management resistance 103 and an increased probability of bid success. However, a higher bid premium does not appear to deter competing bidders. Although contrary to the advanced hypothesis, this result is consistent with the undervaluation motive for takeovers. In terms of the payment method, Fishman (1989) proposes that bids offering a higher proportion of cash are less likely to attract competition or encounter target management resistance. However, the results indicate that greater use of cash in the offer does not serve to deter competing bidders, nor does it lower the likelihood of target management resistance or increase the probability of bid success. These results, although contrary to the advanced hypotheses, are supported by extant empirical evidence and cast further doubt on the predictions of Fishman (1989).These results are also in contrast to the long-held adage that 'cash is king' in the market for corporate control, suggesting that target directors and shareholders consider the overall value in responding to a takeover offer, rather than placing undue importance on the form of consideration offered. Overall, the toehold results suggest that initial bidders with small toeholds are more likely to encounter bid competition and target resistance than bidders with zero toeholds, however, as the size of the initial bidder's toehold increases, the likelihood of competition and resistance declines. A simple signalling hypothesis that provides a potential explanation for this joint result is also advanced in this thesis. Unsurprisingly, given these results, the likelihood of enjoying bid success is higher for initial bidders with zero toeholds, as compared to bidders with small toeholds, however, as the size of the initial bidder's toehold increases, the likelihood of bid success increases. These results confirm that, consistent with the advanced 104 hypotheses, ownership of large toeholds by initial bidders serves to deter competing bidders, lowers the likelihood of resistance and increases the probability of bid success. The impact of a fourth variable, one unrelated to bid structure, on the response of target management and the potential entry of competing bidders was explored. This variable is the cost (to a potential bidder) of acquiring information about the target and is proxied by target-firm size. The results suggest that this cost does not appear to be significantly related to the likelihood of observing bid competition or target management resistance. By exammmg the impact of bid structure on the likelihood of observing bid competition, target resistance and offer success in the one study, this thesis allows for a further examination of how these three factors interact. Namely, the impact of the presence of target resistance on the likelihood of observing bid competition and offer success is examined. Firstly, we find that the likelihood of a competing offer emerging is significantly higher in transactions where target management resist the initial bid. This result is as hypothesised and is consistent with the theoretical propositions of Giammarino and Heinke! (1986) and Hirshleifer and Png (1989). With respect to bid success, we find that the likelihood of an initial bid being successful is reduced significantly in transactions where target management resist the initial bid. This result is also as hypothesised and is consistent with extant empirical evidence (Walkling (1985), Eddey and Casey (1989), and O'Sullivan and Wong (1998)). 105 Overall, the results suggest that, for an initial bidder, the likelihood of a successful and expeditious acquisition is increased significantly if he owns a large toehold, offers a high bid premium, and is able to deter competing bidders and secure the support of target management. However, it seems that structuring the offer to include a greater amount of cash has no such effect. 8.2 Limitations and Suggestions for Further Research The limitations of this study are methodological in nature but only relate to a small proportion of the empirical analysis. Firstly, it is possible that target-firm size is not an appropriate proxy for the 'Cost of Acquiring Information about the Target' in this study. This is because in the relatively small Australian market, as target-size increases, the number of potential acquirers falls dramatically. The use of alternate proxies such as the number of equity analysts following the target firm (Jennings and Mazzeo, 1993) is unlikely to provide a solution since such proxies are usually highly correlated with target-firm size. The discovery of a more robust proxy for the 'Cost of Acquiring Information about the Target' represents a valid avenue for future research. Such a proxy would allow us to determine clearly the significance of this variable. The second limitation relates to the simultaneous decision analysis performed in section 7.4. TOEDUM is excluded as an explanatory variable from this analysis due to its high correlation with BIREST, the binary variable indicating the presence of target resistance. Although not ideal, removing TOEDUM from equations (7.4.1) and (7.5.1) was considered the preferred solution to this problem, particularly since 106 determining the relationship between BIREST and the dependent variables in these equations (COMP and OUT) was the primary purpose of estimating the simultaneous equation analysis. Several potentially fruitful avenues for future research emerge from this study. Perhaps the most natural step forward is research dedicated to the prediction of the outcomes examined in this thesis. That is, whilst this study examines how individual bid structure elements affect the likelihood of observing target resistance, bid competition and success, future work could use these bid structure elements and other variables, not necessarily related to bid structure, to develop models capable of predicting these outcomes ex-ante. Models capable of predicting target management's response, whether a competing bid will be entered or ultimate bid success would be of enormous interest to corporate managers launching a takeover bid, M&A practitioners in formulating their advice, and, not least of all, M&A arbitrageurs. As mentioned previously, part of the toehold result obtained suggests that initial bidders with small toeholds are more likely to encounter bid competition and target resistance than bidders with zero-toeholds. Jennings and Mazzeo (1993) also obtained this result with respect to target resistance. A simple signalling-hypothesis is proposed in this thesis that provides a potential explanation for this toehold result. When toehold-owning bidders, who have an informational advantage as to the prospects of the firm in which they hold a stake, make a takeover bid for the said firm, a positive signal is released to the market regarding the value of the target firm. Thus, initial bidders who own small toeholds are more likely to encounter 107 competition that zero-toehold bidders due to the undervaluation signal their bid conveys. However, this positive association between bid premium and competition would be truncated at a certain toehold level due to the competition-deterring effect of large toeholds (Burkart (1995), Singh (1998) and Bulow et al. (1999)), explaining the second part of the toehold result. Likewise, the new information revealed by a bid from these informed investors may increase the likelihood of target-management resistance, if for no other reason than to solicit competing bids. Future research could test this theory empirically and attempt to quantify the toehold levels at which these effects are present, if at all. This task would involve multiple steps. Firstly, examine the relationship between toehold size and competition on different sets of data. Secondly, if the results found are consistent with those presented in this thesis, examine bids by small-toehold bidders for evidence of target-undervaluation. Finally, examine different toehold levels in order to quantify these effects. 108 References Agrawal, A. and Jaffe, J. (2003), "Do Takeover Targets Under-perform? Evidence from Operating and Stock Returns", Journal of Financial and Quantitative Analysis 38 (5), pp. 721-746. Agrawal, A., Jaffe, J., and Mandelker, G. (1992), "The Post-Merger Performance of Acquiring Firms: A Re-Examination of an Anomaly", Journal of Finance 47 (4), pp. 1605-1621. Agrawal, A. and Walkling, R. (1994), "Executive Careers and Compensation Surrounding Takeover Bids", Journal ofFinance 49 (3), pp. 985-1014. Ambrose, B. and Megginson, W. (1992), "The Role of Asset Structure, Ownership Structure, and Takeover Defenses in Determining Acquisition Likelihood", Journal ofFinancial and Quantitative Analysis 27 (4), pp. 575-589. Andrade, G., Mitchell, M., and Stafford, E. (2001), "New Evidence and Perspectives on Mergers", Journal ofEconomic Perspectives 15 (2), pp. 103-120. Australian Stock Exchange. (2003), "Share Ownership Survey", Unpublished Manuscript, ASX. Bae, S. and Simet, D. (1998), "A Comparative Analysis of Leveraged Recapitalization Versus Leveraged Buyout as a Takeover Defense", Review ofFinancial Economics 7 (2), pp. 157-172. Baron, D. (1983), "Tender Offers and Management Resistance", Journal of Finance 38 (2), pp. 331-343 Bates, T. and Lemmon, M. (2003), "Breaking Up Is Hard To Do? An Analysis of Termination Fee Provisions and Merger Outcomes", Journal of Financial Economics 69 (3), pp. 469-504. Bellamy, D. and Lewin, W. (1992), "Corporate Takeovers, Methods of Payment, and Bidding Firms' Shareholder Returns: Australian Evidence", Asia Pacific Journal of Management 9 (2), pp. 137-149. Berkovitch, E. and Khanna, N. (1990), "How Target Shareholders Benefit from Value Reducing Defensive Strategies in Takeovers", Journal of Finance 45 (1), pp. 137- 156. Berkovitch, E. and Narayanan, M. (1990), "Competition and the Medium of Exchange in Takeovers", Review ofFinancial Studies 3 (2), pp. 153-174. Betton, S. and Eckbo, B. (2000), "Toeholds, Bid-jumps and Expected Payoffs m Takeovers", Review of Financial Studies 13 (4), pp. 841-882. 109 Bhagat, S. and Jefferis, R. (1994), "The Causes and Consequences of Takeover Defense: Evidence from Greenmail",Journal of Corporate Finance 1 (2), pp. 201-231. Billett, M. (1996), "Targeting Capital Structure: The Relationship Between Risky Debt and the Firm's Likelihood of Being Acquired", Journal ofBusiness 69 (2), pp. 173-192. Bishop, S., Dodd, P., and Officer, R. (1987), "Australian Takeovers: The Evidence 1972- 1985", Centre for Independent Studies Policy Monograph No. 12. Boyes, W., Hoffman, D., and Low, S. (1989), "An Econometric Analysis of the Bank Credit Scoring Problem", Journal ofEconometrics 40 (1 ), pp. 3-14. Bradley, M. (1980), "Interfirm Tender Offers and the Market for Corporate Control", Journal ofBusiness 53 (4), pp. 345-376. Bradley, M., Desai, A., and Kim, E. (1983), "The Rationale Behind Interfirm Tender Offers: Information or Synergy?", Journal ofFinancial Economics 11 (2), pp. 183-206. Bradley, M., Desai, A., and Kim, E. (1988), "Synergistic Gains from Corporate Acquisitions and their Division Between the Stockholders of Target and Acquiring Firms", Journal of Financial Economics 21 (1), pp. 3-40. Bris, A. (2002), "Toeholds, Takeover Premium, and the Probability of Being Acquired", Journal of Corporate Finance 8 (3), pp. 227-253. Buchholtz, A. and Ribbens, B. (1994), "Role of Chief Executive Officers in Takeover Resistance: Effects of CEO Incentives and Individual Characteristics", Academy of Management Journal 37 (3), pp. 554-579. Bugeja, M. and Walter, T. (1995), "An Empirical Analysis of Some Determinants of the Target Shareholder Premium in Takeovers", Accounting and Finance 35 (1), pp. 33- 60. Bulow, J., Huang, M., and Klemperer, P. (1999), "Toeholds and Takeovers", Journal of Political Economy 107 (3), pp. 427-454. Burkart, M. (1995), "Initial Shareholdings and Overbidding in Takeover Contests", Journal of Finance 50 (5), pp. 1491-1515. Chowdry, B. and Jegadeesh, N. (1994), "Pre-tender Offer Share Acquisition Strategy in Takeovers", Journal of Financial and Quantitative Analysis 29 (1), pp. 117-129. Christofides, L., Stengos, T., and Swidinsky, R. (1997), "Welfare Participation and Labour Market Behaviour in Canada", Canadian Journal ofEconomics 30 (3), pp. 595-621. Cotter, J. and Zenner, M. (1994), "How Managerial Wealth Affects the Tender Offer Process", Journal ofFinancial Economics 35 (1), pp. 63-97. 110 da Silva Rosa, R., Izan, H., Steinbeck, A., and Walter, T. (2000), "The Method of Payment Decision in Australian Takeovers: An Investigation of Causes and Effects", Australian Journal ofManagement 25 (1), pp. 67-94. Dann, L. and DeAngelo, H. (1988), "Corporate Finance Policy and Corporate Control: A Study of Defensive Adjustments in Asset and Ownership Structure", Journal of Financial Economics 20 (1), pp. 87-128. Datta, S. and Iskandar-Datta, M. (1996), "Takeover Defenses and Wealth Effects on Securityholders: The Case of Poison Pill Adoptions", Journal of Banking and Finance 20 (7), pp. 1231-1250. Dewatripont, M. (1993), "The Leading Shareholder Strategy, Takeover Contests and Stock Price Dynamics", European Economics Review 37 (5), pp. 983-1004. Eckbo, B., Giammarino, R., and Heinkel, R. (1990), "Asymmetric Information and the Medium of Exchange in Takeovers: Theory and Tests", Review of Financial Studies 3 (4), pp. 651-675. Eckbo, B. and Langhor, H. (1989), "Information Disclosure, Method of Payment, and Takeover Premiums: Public and Private Tender Offers in France", Journal of Financial Economics 24 (2), pp. 363-403. Eddey, P. and Casey, R. (1989), "Directors Recommendations in Response to Takeover Bids: Do They Act in Their Own Interests", Australian Journal of Management 14 (1), pp. 1-28. Ernst and Young Corporate Finance. (2002), "Mergers and Acquisitions Index". Ettner, S. (1996), "The Opportunity Costs of Elder Care", Journal of Human Resources 31 (1), pp. 189-205. Evans, W., Farrelly, M., and Montgomery, E. (1999), "Do Workplace Smoking Bans Reduce Smoking?",American Economic Review 89 (4), pp. 728-747. Evans, W. and Schwab, R. (1995), "Finishing High School and Starting College: Do Catholic Schools Make a Difference?", Quarterly Journal of Economics 110 (4), pp. 941-974. Faccio, M. and Masulis, R. (2004), "The Choice of Payment Method in European Mergers and Acquisitions", Unpublished Working Paper, Vanderbilt University. Fishman, M. (1988), "A Theory of Preemptive Takeover Bidding", RAND Journal of Economics 19 (1), pp. 88-101. Fishman, M. (1989), "Preemptive Bidding and the Role of the Medium of Exchange in Acquisitions", Journal of Finance 44 (1), pp. 41-57. Franks, J., Harris, R., and Titman, S. (1991), "The Postmerger Shareprice Performance of Acquiring Firms", Journal of Financial Economics 29 (1 ), pp. 81-96. 111 Franks, J. and Mayer, C. (1996), "Hostile Takeovers in the UK and the Correction of Management Failure", Journal ofFinancial Economics 40 (1), pp. 163-181. Ghosh, A. (2001), "Does Operating Performance Really Improve Following Corporate Acquisitions?", Journal of Corporate Finance 7 (2), pp. 151-178. Ghosh, A. and Ruland, W. (1998), "Managerial Ownership, the Method of Payment for Acquisitions, and Executive Job Retention", Journal ofFinance 53 (2), pp. 785-798. Giammarino, R. and Heinkel, R. (1986), "A Model of Dynamic Takeover Behavior", Journal of Finance 41 (2), pp. 465-480. Gort, M. (1969), "An Economic Disturbance Theory of Mergers", Quarterly Journal of Economics 83 ( 4 ), pp. 624-642. Greene, W. (1993), "Econometric Analysis", (2nd ed.), Macmillan, New York. Grossman, S. and Hart, 0. (1980), "Takeover Bids, the Free Rider Problem, and the Theory of the Corporation", Bell Journal ofEconomics 11 (1), pp. 42-64. Halpern, P. (1983), "Corporate Acquisitions: A Theory of Special Cases? A Review of Event Studies Applied to Acquisitions", Journal of Finance 38 (2), pp. 297-317. Hansen, R. (1987), "A Theory for the Choice of Exchange Medium in Mergers and Acquisitions", Journal ofBusiness 60 (1), pp. 75-95. Harford, J. (1999), "Corporate Cash Reserves and Acquisitions", Journal of Finance 54 (6), pp. 1969-1997. Hastie, T. and Tibshirani, R. (1995), "Generalized Additive Models", Chapman and Hall, London. Healy, P., Palepu, K., and Ruback, R. (1992), "Does Corporate Performance Improve After Mergers?", Journal ofFinancial Economics 31 (2), pp. 135-175. Henry, D. (2004), "Directors Recommendations in Takeovers: An Agency and Governance Analysis", Journal ofBusiness Finance and Accounting, forthcoming. Higson, C. and Elliott, J. (1998), "Post-takeover Returns: The UK Evidence", Journal of Empirical Finance 5 (1), pp. 27-46. Hirshleifer, D. and Png, I. (1989), "Facilitation of Competing Bids and the Price of a Takeover Target", Review of Financial Studies 2 (4), pp. 587-606. Hirshleifer, D. and Titman, S. (1990), "Share Tendering Strategies and the Success of Hostile Takeover Bids", Journal ofPolitical Economy 98 (2), pp. 295-324. Hoffmeister, J. and Dyl, E. (1981), "Predicting Outcomes of Cash Tender Offers", Financial Management 10 (1), pp. 50-58. 112 Holl, P. and Kyriazis, D. (1997), "Wealth Creation and Bid Resistance in U.K. Takeover Bids", Strategic Management Journal 18 (6), pp. 483-498. Holstrom, B. and Kaplan, S. (2001), "Corporate Governance and Merger Activity in the U.S.: Making Sense of the 1980s and 1990s", Unpublished Working Paper, National Bureau of Economic Research. Israel, R. (1992), "Capital and Ownership Structures, and the Market for Corporate Control", Review ofFinancial Studies 5 (2), pp. 181-198. Jabbour, A., Jalilvand A., and Switzer, J. (2000), "Pre-bid Price Run-ups and Insider Trading Activity Evidence from Canadian Acquisitions", International Review of Financial Analysis 9 (1), pp. 21--43. Jacobson, T. and Roszbach, K. (2003), "Bank Lending Policy, Credit Scoring and Value-at Risk", Journal ofBanking and Finance 27 (4), pp. 615-633. Jarrell, G. (1985), "The Wealth Effects of Litigation by Targets: Do Interests Diverge in a Merge?", Journal ofLaw and Economics 28 (1), pp. 151-177. Jarrell, G., Brickley, J., and Netter, J. (1988), "The Market for Corporate Control: The Empirical Evidence since 1980", Journal ofEconomic Perspectives 2 (1), pp. 49-68. Jarrell, G. and Poulsen, A. (1987), "Bidder Returns", Unpublished Working Paper, University of Georgia. Jennings, R. and Mazzeo, M. (1993), "Competing Bids, Target Management Resistance, and the Structure of Takeover Bids", Review of Financial Studies 6 ( 4 ), pp. 883-909. Jensen, M. (1986), "Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers", American Economic Review 76 (3), pp. 323-329. Jensen, M. (1988), "Takeovers: Their Causes and Consequences", Journal of Economic Perspectives 2 (1 ), pp. 21-48. Jensen, M. (1992), "The Takeover Controversy: Analysis and Evidence", in Stern, J. and Chew, D.: The Revolution in Corporate, (2nd ed.), Blackwell Finance, U.S. Jensen, M. and Meckling, W. (1976) "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure", Journal of Financial Economics 3 (4), pp. 305-360. Jensen, M. and Ruback, R. (1983), "The Market for Corporate Control: The Scientific Evidence", Journal ofFinancial Economics 11 (1), pp. 5-50. Khanna, N. (1997), "Optimal Bidding for Tender Offers", Journal of Financial Research 20 (3), pp. 323-342. 113 Kyle, A. and Villa, J. (1991), "Noise Trading and Takeovers", RAND Journal of Economics 22 (1), pp. 54-71. Langetieg, T. (1978), "An Application of a Three-Factor Performance Index to Measure Stockholders Gains from Mergers", Journal of Financial Economics 6 (4), pp. 365- 384. Lev, B. (1992), "Observations on the Merger Phenomenon and a Review of the Evidence", in Stern, J. and Chew, D.: The Revolution in Corporate, (2nd ed.), Blackwell Finance, U.S. Limmack, R. (1994), "Synergy or New Information as a Source of Wealth Change in Acquisitions: The Case of Abandoned Bids", Accounting and Business Research 24, pp. 255-265. Linn, S. and Switzer, J. (2001), "Are Cash Acquisitions Associated with Better Postcombination Operating Performance than Stock Acquisitions?", Journal of Banking and Finance 25 (6), pp. 1113-1138. Loughran, T. and Vijh, A. (1997), "Do Long-term Shareholders Benefit from Corporate Acquisitions?", Journal of Finance 52 (5), pp. 1765-1790. Lyon, J., Barber, B., and Tsai, C. (1999), "Improved Methods for Tests of Long-Run Abnormal Stock Returns", Journal ofFinance 54 (1), pp. 165-201. Maddala, G. (1983), "Limited-Dependent and Qualitative Variables in Econometrics", University of Cambridge Press, U.S. Maddala, G. (1992), "Introduction to Econometrics", (2nd ed.), Macmillan, New York. Malatesta, P. (1983), "The Wealth Effect of Merger Activity and the Objective Functions of Emerging Firms", Journal ofFinancial Economics 11 (2), pp. 155-181. Manne, H. (1965), "Mergers and the Market for Corporate Control", Journal of Political Economy 73 (2), pp. 110-120. Martin, K. (1996), "The Method of Payment in Corporate Acquisitions, Investment Opportunities, and Management Ownership", Journal of Finance 51 (4), pp. 1227- 1246. Martin, K. and McConnell, J. (1991), "Corporate Performance, Corporate Takeovers, and Management Turnover", Journal ofFinance 46 (2), pp. 671-688. McDonald, B. (2003) "The U.S. M&A Market: Overview and Outlook", Unpublished Manuscript, Houlihan Lokey Howard and Zukin. Mitchell, M. and Mulherin, J. (1996), "The Impact of Industry Shocks on Takeover and Restructuring Activity", Journal ofFinancial Economics 41 (2), pp. 193-229. 114 Morck, R., Shleifer, A., and Vishny, R. (1989), "Alternative Mechanisms for Corporate Control", American Economic Review 79 (4), 842-852. Morck, R., Shleifer, A., and Vishny, R. (1990), "Do Managerial Objectives Drive Bad Acquisitions?", Journal of Finance 45 (1), pp. 31-48. Myers, S. and Majluf, N. (1984), "Corporate Financing and Investment Decisions When Firms Have Information that Investors Do Not Have", Journal of Financial Economics 13 (2), pp. 187-221. North, D. (2001), "The Role of Managerial Incentives in Corporate Acquisitions: The 1990s Evidence", Journal of Corporate Finance 7 (2), pp. 125-149. Nuttal, R. (1999) "Takeover Likelihood Models for UK Quoted Companies", Unpublished Working Paper, University of Oxford. O'Sullivan, N. and Wong, P. (1998), "The Impact of Board Composition and Ownership on the Nature and Outcome of UK Takeovers", Corporate Governance: An International Review 6, pp. 92-100. Officer, M. (2003), "Termination Fees in Mergers and Acquisitions", Journal of Financial Economics 69 (3), pp. 431-467. Pompilio, D. (2000), "An Examination of Preemptive Takeover Bidding Strategies", Unpublished Working Paper, University of Melbourne. Rappaport, A. (1990), "The Staying Power of the Public Corporation", Harvard Business Review 1 (1 ), pp. 96-104. Ravid, S. and Spiegel, M. (1999), "On Toeholds and Bidding Contests", Journal of Banking and Finance 23 (8), pp. 1219-1242. Roll, R. (1986), "The Hubris Hypothesis of Corporate Takeovers", Journal of Business 59 (2), pp. 197-216. Roszbach, K. (2003), "Bank Lending Policy, Credit Scoring and the Survival of Loans", Unpublished Working Paper, Sveriges Riksbank Working Paper Series. Rubicon Partners. (2001), "Rubicon M&A Fund: Profit from Mergers and Acquisitions", Prospectus. Scherer, F. (1988), "Corporate Takeovers: The Efficiency Arguments", Journal ofEconomic Perspectives 2 (1), pp. 69-82. Schwert, G. (1996), "Markup Pricing in Mergers and Acquisition", Journal of Financial Economics 41 (2), pp. 153-192. Servaes, H. (1991), "Tobin's Q and the Gains from Takeovers", Journal of Finance 46 (1), pp. 409-419. 115 Shleifer, A. and Vishny, R. (1986, a), "Large Shareholders and Corporate Control", Journal ofPolitical Economy 94 (3), pp. 461-488. Shleifer, A. and Vishny, R. (1986, b ), "Greenmail, White Knights, and Shareholders' Interest", RAND Journal ofEconomics 17 (3), pp. 293-309. Shleifer, A. and Vishny, R. (1988), "Value Maximization and the Acquisition Process", Journal ofEconomic Perspectives 2 (1), pp. 7-20. Shleifer, A. and Vishny, R. (1991), "The Takeover Wave of the 1980s", Journal of Applied Corporate Finance 4 (3), pp. 49-56. Shleifer, A. and Vishny, R. (2003), "Stock Market Driven Acquisitions", Journal of Financial Economics 70 (3), pp. 295-311. Singh, R. (1998), "Takeover Bidding with Toeholds: The Case of the Owner's Curse", Review ofFinancial Studies 11 (4 ), pp. 679-704. Slusky, A. and Caves, R. (1991), "Synergy, Agency, and the Determinants of Premia Paid in Mergers", Journal of Industrial Economics 39 (3), pp. 277-296. St-Pierre, J., Gagnon, J., and Saint-Pierre, J. (1996), "Concentration of Voting Rights and Board Resistance to Takeover Bids",Journal of Corporate Finance 3 (1), pp. 45-73. Stulz, R., Walkling, R., and Song, M. (1990), "The Distribution of Target Ownership and the Division of Gains in Successful Takeovers", Journal of Finance 45 (3), pp. 817- 833. Thomson Financial. (2004), "Australian M&A Research", Unpublished Manuscript, Thomson Financial. Travlos, N. (1987), "Corporate Takeover Bids, Methods of Payment, and Bidding Firms' Stock Returns", Journal ofFinance 42 (4), pp. 943-963. Trimbath, S. (2001), "High Yield Financing and Efficiency-Enhancing Takeovers", Unpublished Working Paper, Milken Institute. Walkling, R. (1985), "Predicting Tender Offer Success: A Logistic Analysis", Journal of Financial and Quantitative Analysis 20 (4), pp. 461-478. Walkling, R. and Long, M. (1984), "Agency Theory, Managerial Welfare, and Takeover Resistance", RAND Journal ofEconomics 15 (1), pp. 54-68. Weir, C. (1997), "Corporate Governance, Performance and Takeovers: An Empirical Analysis of UK Mergers", Applied Economics 29, pp. 1465-1475. 116 Appendix A - Table A.1 Identity of Target and Initial Bidder Companies Comprising the Data Set Announcement Date Target Company Initial Bidder Company (Initial Bid) 17/01/1996 Gasgoyne Gold Mines Sons Of Gwalia NL 17/01/1996 Burmine Ltd Sons Of Gwalia NL 29/01/1996 Associated Gold Fields NL Resolute Samantha Group Ltd 14/02/1996 Ampolex Ltd Mobil Exploration & Producing 16/02/1996 Austoft Holdings Case Corp Pty Ltd(Case Com) 29/02/1996 Solomon Pacific Resources NL Acacia Resources Ltd 18/03/1996 Shomega Inc Pacific Magazines & Printing 28/03/1996 QIWLtd Davids Ltd(Davids Holdings) 9/04/1996 Rothbury Wines Ltd Mildara Blass Ltd(Foster's) 11/04/1996 Golden Shamrock Mines Ltd Ashanti Goldfields Co Ltd 15/04/1996 Sunbeam Victa Holdings Ltd GUO Holdings Ltd 26/04/1996 Clyde Industries Ltd Evans Deakin Industries Ltd 1/05/1996 Azon Ltd(Boral Ltd) Pacific BBA Ltd 3/05/1996 Elders Australia Ltd Futuris Corp Ltd 16/05/1996 Crusader Ltd Clyde Petroleum(Australia) Pty Ltd 17/06/1996 Pacific Mutual Australia Ltd Mercantile Mutual Life Insurance 8/08/1996 Orion Resources NL Sons Of Gwalia NL 23/09/1996 Discovery Petroleum(Aztec) Premier Oil Australia Pty Ltd 2/10/1996 TNT Ltd KPN (Australia)Ltd 14/10/1996 Advance Bank Of Australia St George Bank Ltd 12/11/1996 McIntosh Securities Ltd Merrill Lynch & Co Inc 30/12/1996 Century Drilling Downer Group Ltd 10/01/1997 Techway Ltd Nova Pacific Capital Ltd 22/01/1997 Mount Edon Gold Mines (Aust) Ltd Reachwest Pty Ltd 22/01/1997 Mt Edon Gold Mines(Australia) Reachwest (Camelot) Pty Ltd 11/02/1997 Loscam Ltd GE Capital International Holdings Corp. 18/02/1997 Leutene1n1:er Ltd Hancock & Gore Ltd 25/02/1997 Australian Innovation Ltd Tailgate Pty Ltd 27/02/1997 Spatial Systems Ltd Spacetec IMC Corporation 05/03/1997 Defiance Mills Ltd Bunge Foods Queensland Pty Ltd 13/03/1997 Mt Leyson Gold Mines Ltd St Barbara Mines Ltd 27/03/1997 Arcadia Minerals NL Sipa Resources International NL 3/04/1997 Bank Of Melbourne Limited Westpac Banking Corp Ltd 09/04/1997 Stanley Mining Services Ltd Layne Christensen Australia Pty Ltd 23/04/1997 Australian Vintage Ltd Simeon Wines Ltd 09/05/1997 Mining World Ltd Maior Drilling Australia Pty Ltd 13/05/1997 Bush Capital Ltd Zennyar Pty Ltd 21/05/1997 Resource & Industry Ltd Axiom Properities Ltd 29/05/1997 Takoradi Gold NL West African Gold Corporation 23/06/1997 Hawker Richardson Ltd Austrim Ltd 14/08/1997 Lanes Ltd Barlow Australia Pty Ltd 21/08/1997 Eagle Mining Corporation NL Great Central Holdings Pty Ltd 117 21/08/1997 Wiluna Mines Ltd Great Central Investments Pty Ltd 25/08/1997 Datacraft Ltd Dimension Data Australia Ptv Ltd 27/08/1997 AHCLtd Strategic Capital Investment Ltd 28/08/1997 Ghana Gold Mines NL West African Gold Corporation 29/08/1997 Gibson Chemical Industries Ltd Ecolab Australia Ptv Ltd 10/09/1997 Weston Investments Ltd Consolidated African Mines Jersey Ltd 18/09/1997 Memtec Ltd USFC Acquisition Inc 24/09/1997 Champion Compressors Ltd Austrim Ltd 29/09/1997 Mitec Ltd Filtronic Comtek (Australia) Pty Ltd 15/10/1997 Country Road Ltd Woolworths International (Australia) Pty Ltd 05/12/1997 AAPCLtd Cobefin SA 11/12/1997 Australian Chemical Holdings Ltd Nuplex Industries Ltd 14/01/1998 National Consolidated Ltd Austrim Ltd 15/01/1998 Clements Marshall Consolidated Ltd Webster Ltd 19/01/1998 Panorama Resources NL Tanganyika Gold NL 29/01/1998 Newsnet ITN Ltd Telegroup Network Services Ptv Ltd 09/02/1998 Peter Kurts Properties Ltd Forrester Parker Group Ltd 11/02/1998 Davids Ltd Metro Cash & Carrv Ltd 13/02/1998 Allgas Energy Ltd TU Australia (Queensland) Pty Ltd 02/03/1998 Agen Ltd Biotech International Ltd 20/03/1998 Dioro Exploration NL Diamond Rose NL 20/03/1998 Striker Resources NL Diamond Rose NL 31/03/1998 Motors Holdings Ltd Doncaster Motors Pty Ltd 17/04/1998 Primac Holdings Ltd Milltoc Pty Ltd 28/04/1998 Aberfoyle Ltd Western Metals Ltd 29/04/1998 Pauls Ltd National Foods Milk Ltd 30/04/1998 Kilpatrick Green Holdings Ltd United Group Ltd 08/05/1998 Coms21 Ltd Atlantic International Entertainment Australia Pty Ltd 29/05/1998 Howard Financial Holdings Ltd Challenger Life Ltd 22/06/1998 Gearhart Australia Ltd Scientific Services Ltd 26/06/1998 Australasian Technology Corporation Leong Yew Cheong Boon Kee And Hemlock Ltd Capital Pte Ltd 16/07/1998 Frontier Petroleum NL West Oil NL 20/07/1998 JNA Telecommunications Ltd Lucent Technologies Australia Pty Ltd 21/07/1998 Tandou Ltd Colly Cotton Ltd 31/07/1998 Unitel Corporation Ltd Mcgregor Marketing Pty Ltd 07/08/1998 Tiger Investment Company Ltd Metals Exploration Ltd 14/08/1998 Consolidated Transport Industries Darowa Corporation Ltd Ltd 25/08/1998 GIO Australia Holdings Ltd AMP Insurance Investment Holdings Pty Ltd 26/08/1998 KLZLtd Avatar Industries Ltd 08/09/1998 Reef Mining NL Hudson Resources Ltd 10/09/1998 SGIO Insurance Ltd Wesfarmers Insurance Investments Ptv Ltd 17/09/1998 Sipa Resources International NL Lynas Gold NL 18/09/1998 Great Southern Oil NL Amity Oil NL 23/09/1998 FAI Insurances Ltd HIH Investment Holdings Ltd 23/09/1998 Metal Manufactures Ltd Marsh Electrical Pty Ltd 05/10/1998 Abednego Nickel Ltd Murrin Murrin Investments Ptv Ltd 06/10/1998 Australian National Industries Ltd Evans Deakin Investments Pty Ltd 20/10/1998 Savage Resources Ltd Pasminco Investments Pty Ltd 12/11/1998 Berela Ltd Citie Centre 4 Pty Ltd 07/12/1998 Geographe Resources Ltd Golden Prospect Mining Company Pie 14/12/1998 Ozemail Ltd Uunet Holdings Australia Pty Ltd 18/12/1998 Perseverance Corporation Ltd Melbourne Square Pty Ltd 21/12/1998 Diversified Mineral Resources NL Hargraves Resources NL 03/01/1999 Taipan Resources NL Paulsens Gold Ptv Ltd 11/01/1999 Great Central Mines Ltd Yandal Gold Pty Ltd 118 15/02/1999 Australian Premium Wines Ltd Cranswick Estate Wines Ltd 16/02/1999 Tyndall Australia Ltd Royal & Sun Alliance Insurance Group 03/03/1999 FAI Life Ltd Tower Group (Australia) Pty Ltd 24/03/1999 Hoyts Cinemas Group Consolidated Press Holdings 29/03/1999 CIM Resources Ltd RIB Australia Pty Ltd 01/04/1999 Wesfi Ltd Blend Investments Pty Ltd 15/04/1999 Cultus Petroleum NL OMV Australia Pty Ltd 16/04/1999 Star City Holdings Ltd Tabcorp Holdings Limited 28/04/1999 Capcount Property Trust Goodman Hardie Management Australia Ltd 28/04/1999 Reef Mining NL Hudson Resources Ltd 07/06/1999 Australian Commercial Property Stockland Trust Group Trust 17/06/1999 Coca Cola Beverages Pk Hellenic Bottling Company 24/06/1999 Cuppa Cup Vineyards Ltd Southcoro Ltd 29/06/1999 BT Hotel Group GPT Nominees Pty Ltd 12/07/1999 Hargraves Resources NL ORD Australasia Aps 15/07/1999 Hudson Conway Ltd Publishing & Broadcasting 29/07/1999 Holyman Ltd Adsteam Enterprises Pty Ltd 09/08/1999 Omega Oil NL Cue Enern:v Resources NL 02/09/1999 Acacia Resources Ltd Delta Gold NL 08/09/1999 Australian Gold Resources Ltd Centaur Mining & Exploration Ltd 08/09/1999 Heine Management Ltd Mercantile Mutual Holdings Ltd 08/09/1999 Joe White Maltings Ltd GPG (No3) Pty Ltd 13/09/1999 Armstrong Jones Office Fund Prime Credit Property Trust 15/09/1999 AAPTLtd TCNZ Australia Investments Pty Ltd 23/09/1999 AWA Ltd Jupiters Ltd 11/10/1999 Niugini Mining Ltd Lihir Gold Ltd 12/10/1999 OAMPSLtd Ultramar Investments Pty Ltd 21/10/1999 Australian Wine & Horticulture Fund AIDCLtd 25/10/1999 Walker Corporation Ltd Australand Holdings Limited 26/10/1999 Parbury Ltd Atkins Carlyle Ltd 08/11/1999 Gilt-Edged Mining NL Croesus Mining NL 08/11/1999 Metal Manufactures Ltd Marsh Electrical Pty Ltd 18/11/1999 Minproc Ltd Gold And Resource Development NL 19/11/1999 Metakoro Ltd Smorgon Steel (Acquisitions) Pty Ltd 22/11/1999 Australian Properties & Resorts Ltd Wee Boo Kuan 29/11/1999 Pioneer International Ltd Hanson Pk 06/12/1999 Biotech International Ltd Peptech Ltd 23/12/1999 Siddons Ramset Ltd ITW Holdings Pty Ltd 12/01/2000 Dome Resources NL Durban Roodepoort Deep Ltd 12/01/2000 Hotcopper Australia Ltd Bourse Data Ltd 07/02/2000 Heron Resources NL Centaur Mining & Exploration 09/02/2000 Eastern Aluminium Ltd Elval Australia Pty Ltd 10/02/2000 Allstate Eexplorations NL Beaconsfield Gold NL 23/02/2000 Tyndall Property Trust Meridian Investment Trust 01/03/2000 Ross Mining NL Delta Gold Ltd 07/03/2000 Amway Asia Pacific Ltd New Aap Ltd 09/03/2000 Colonial Ltd Commonwealth Bank Of Austral 31/03/2000 Darling Park Trust Amp Asset Management Australia Ltd 06/04/2000 Brickworks Ltd Guinness Peat Group Pk/Old 10/04/2000 Infratil Australia Ltd Australian Infrastructure Fund Ltd 17/04/2000 J Boag & Son Ltd San Miguel Australia Holdings Pty Ltd 19/04/2000 Email Ltd Smorgon Steel Group Ltd 03/05/2000 Greenchip Emerging Growth Ltd Vital Capital Ltd 03/05/2000 Flinders Industrial Property Trust Macquarie Goodman Industrial 05/05/2000 Crevet Ltd lplex Pipelines Australia Pty Ltd 18/05/2000 Challenger Property Income Trust Challenger Property Nominees Pty Ltd 05/06/2000 First Australian Building Society Ltd Bendigo Bank Ltd 119 05/06/2000 St Lukes Group Ltd Westfield Trust 19/06/2000 Carillon Development Ltd Aimtree Pty Ltd 20/06/2000 Eisa Ltd Austar United Communications 23/06/2000 North Ltd Rio Tinto Pie 27/06/2000 Finemore Holdings Ltd Toll Holdings Ltd 05/07/2000 Market Faxts Ltd Teleo Australia Ltd 13/07/2000 Paladin Commercial Trust Commercial Investment Trust 13/07/2000 Paladin Industrial Trust Industrial Investment Trust 18/07/2000 Libertyone Ltd Cybersentry Inc 18/07/2000 Pinnacle VRB Ltd Federation Resources NL 25/07/2000 Bounty Investments Ltd Argo Investments Limited 25/07/2000 Wakefield Investments (Australia) Argo Investments Limited Ltd 31/07/2000 Ashton Mining Ltd De Beers Australia Holdings Pty Ltd 08/08/2000 Channel E Ltd My Money Group Ltd 08/08/2000 Advance Property Fund Mirvac Funds Ltd 15/08/2000 Macquarie Industrial Trust Macquarie Goodman Industrial 28/08/2000 QCT Resources Ltd Metcoal Holdings (Qld) Ptv Ltd. 05/09/2000 Bemax Resources NL Iluka Resources Limited 13/09/2000 Petroz NL Novus Petroleum Ltd 15/09/2000 Bligh Ventures Ltd Panaseer Ltd 19/09/2000 Taipan Resources NL Troy Resources NL 28/09/2000 Email Ltd Smorgon Distribution Ltd 10/10/2000 Hazelton Airlines Ltd AnsettHoldings Ltd 16/10/2000 Petroz NL Novus Petroleum Ltd 30/10/2000 Realestate.Com.Au Ltd Rea Holdings fNo.11 & fNo.21 Pty Ltd 02/11/2000 Finemore Holdings Ltd Toll Holdings Limited 15/11/2000 Evans Deakin Industries Ltd Downer EDI Limited 16/11/2000 Spicers Paper Ltd Paperlinx Limited 20/11/2000 Armstrong Jones Retail Fund ING Groep NV 21/11/2000 Australian Hospital Care Ltd Mayne Group Ltd 22/11/2000 IAMALtd Wesfarmers Limited 06/12/2000 Arthur Yates & Company Ltd Yates Limited 19/12/2000 New Hampton Goldfields Ltd Harmony Gold Mining Co Ltd 20/12/2000 Palmer Corporation Ltd Colorado Group Limited 02/01/2001 Vincorp Wineries Ltd Simon Gilbert Wines Ltd 18/01/2001 Wesfi Ltd Arnatek Holdings Ltd 13/02/2001 Franked Income Fund Wesfarmers Ltd 21/02/2001 Vivanet Ltd Datafast Telecommunications 06/03/2001 Media Entertainment Group Ltd Television & Media Services 15/03/2001 Namakwa Diamond Company NL Maiestic Resources NL 26/03/2001 Cable & Wireless Optus Ltd Singapore Telecommunications 27/03/2001 Wine Planet Holdings Ltd Foster's Group Ltd 30/03/2001 Northern Gold NL Pilatus Gold Corp Pty Ltd 06/04/2001 Rebel Sport Ltd Harvey Norman Holdings Ltd 09/04/2001 Natural Gas Australia Ltd Santos Ltd 09/04/2001 Alpha Healthcare Limited Ramsay Health Care Limited 27/04/2001 Australian Liquor Group Ltd Coles Myer Ltd 30/04/2001 Delfin Ltd Lend Lease Corp Limited 01/05/2001 Telecasters Australia Ltd Southern Cross Broadcasting 07/05/2001 Mtm Media Entertainment Trust Babcock & Brown Pty Ltd 09/05/2001 Mobile Communications Holdings Vodafone Group Pie Ltd 09/05/2001 Kusp Ltd Senetas Corporation Limited 28/05/2001 Consolidated Rutile Ltd Iluka Resources Limited 31/05/2001 F H Faulding & Co Ltd Mayne Group Ltd 13/06/2001 Howard Smith Ltd Wesfarmers Limited 27/06/2001 Just Jeans Group Ltd Prudential Pie 120 05/07/2001 Montana Group (Nz) Ltd Millstream Equities Ltd 10/07/2001 Homestake Mining Company Barrick Gold Corporation 23/07/2001 TOG Logistics Ltd Patrick Corp Ltd 30/07/2001 Bigshop.Com.Au Ltd Fast Scout Ltd 02/08/2001 Wealthpoint Ltd St George Bank Limited 07/08/2001 Baycorp Holdings Ltd Data Advantage Ltd 23/08/2001 Pacmin Mining Corporation Ltd Sons Of Gwalia Limited 24/08/2001 Aurora Gold Ltd Silvara Pty Ltd 27/08/2001 Nautronix Ltd OHi Asa 05/09/2001 Normandy Mining Ltd Anglogold Ltd 13/09/2001 Combined Communications Network Cabcharge Australia Ltd Ltd 14/09/2001 lpoh Ltd Recosia Pte Ltd 17/09/2001 Delta Gold Ltd Auriongold Ltd 19/09/2001 Homemaker Retail Group General Property Trust 21/09/2001 Central Pacific Minerals NL Southern Pacific Petroleum NL 27/09/2001 Banksia Wines Ltd Lion Nathan Limited 01/10/2001 Petaluma Ltd Lion Nathan Limited 08/10/2001 Scientific Services Ltd SGS Societe General De Surveillance Holding SA 08/10/2001 Contact Energy Ltd Mission Enern:v Five Star Holdings 11/10/2001 Otter Gold Mines Ltd Normandy Nfm Ltd 12/10/2001 Central Norseman Corporation Ltd Croesus Mining NL 18/10/2001 Pipers Brook Vineyard Ltd G & C Kreglinger NV 23/10/2001 Cambooya Investments Ltd Milton Corporation Limited 24/10/2001 Frucor Beverages Group Ltd Danone Asia Pte Ltd 02/11/2001 Asia Pacific Speciality Chemicals Symex Holdings Limited Ltd 08/11/2001 2 Park Street Trust Macquarie Office Trust 22/11/2001 Efinancial Capital Ltd Challenger International Ltd 26/11/2001 King Island Company Ltd fThe l National Foods Limited 29/11/2001 Brisbane Broncos Ltd Magic Millions League Pty Ltd 29/11/2001 Westgold Resources NL Redsummer Pty Ltd 10/12/2001 Hill 50 Gold NL Harmony Gold Mining Co Ltd 12/12/2001 Admiralty Resources NL Equity-1 Resources Ltd 11/01/2002 ITG Ltd Flight Centre Limited 21/01/2002 Orogen Minerals Ltd Oil Search Ltd 05/02/2002 Spencer Gulf Telecasters Ltd Southern Cross Broadcasting 12/02/2002 Clarity International Ltd Powerlan Limited 20/02/2002 Simeon Wines Ltd Brian Mcguigan Wines Ltd 21/02/2002 Pos.lt.Ive Technologies Ltd Tele2000 Ltd 28/02/2002 SME Growth Ltd H-G Ventures Ltd 05/03/2002 Arc Energy NL Tap Oil Limited 15/3/2002 Australian Oil & Gas Corporation Ensign Resource Service Grp Ltd 03/04/2002 Asia Pacific Speciality Chemicals Symex Holdings Limited Ltd 10/04/2002 Software Communication Group Ltd Data & Commerce Ltd 18/04/2002 Ranger Minerals Ltd Revesco Group Ltd 03/05/2002 Nautronix Ltd First Technology Pie 10/05/2002 Foundation Healthcare Ltd Independent Practitioner Net 27/05/2002 Aurion Gold Ltd Placer Dome Inc 29/05/2002 Basin Minerals Ltd Iluka Resources Limited 31/05/2002 Online Advantage Ltd Mcwilliam Nominees Pty Ltd 05/06/2002 Snack Foods Ltd Campbell Soup Co 13/06/2002 Utility Services Corporation Ltd DVT Holdings Ltd 14/06/2002 Cranswick Premium Wines Ltd Evans & Tate Limited 18/06/2002 Ausdoc Group Ltd ABN Amro Holding NV 121 26/06/2002 Sonacom Ltd Zylotech Limited 16/07/2002 Recruitment Solutions Ltd Ego Pty Ltd 30/07/2002 Colonial First State Property Trust Commonwealth Property Office Fund Group 14/08/2002 Permanent Trustee Company Ltd Trust Company of Australia Ltd 20/08/2002 Winepros Ltd Starmore Investments Pty Ltd 30/09/2002 Balmoral Corporation Ltd Formrace Pty Ltd 07/10/2002 Breakwater Island Trust Jupiters Limited 21/10/2002 Guinness Peat Group Plc Brunel Holdings Plc 04/11/2002 Joe White Maltings Ltd Ausbulk Ltd 06/12/2002 Jetset Travelworld Ltd Sintack Pty Ltd 13/12/2002 Goodman Fielder Ltd Burns, Philp & Company Ltd 16/12/2002 Magnetic Minerals Ltd Ticor Ltd 13/01/2003 Tandou Ltd Volcot Holding Ag 15/01/2003 Sanford Ltd IWLLtd 17/01/2003 BRL Hardy Ltd Constellation Brands Inc 20/01/2003 Challenger International Ltd CPH Investment Corp 21/01/2003 Anaconda Nickel Limited Matlinpatterson Global Oooor 11/02/2003 Sirtex Medical Ltd Ceohalon Inc 18/02/2003 International Goldfields Ltd International Goldfields Ltd 26/02/2003 Abelle Ltd Harmony Gold Mining Co Ltd 04/03/2003 Bristile Ltd Brickworks Limited 18/03/2003 AMP Shoooing Centre Trust Centro Properties Group 04/04/2003 Hamilton Island Ltd General Property Trust 07/04/2003 Mim Holdings Ltd Xstrata Plc 10/04/2003 Futureone Ltd ABC Leaming Centres Ltd 28/04/2003 Neverfail Springwater Ltd Coca-Cola Amatil Limited 30/04/2003 OPSM Group Ltd Luxottica Group Spa 01/05/2003 Rcr Tomlinson Ltd MTQ Corp Ltd 08/05/2003 Sabre Group Ltd AMP Limited 22/05/2003 Milnes Holdings Ltd Crane Group Limited 26/05/2003 Principal Office Fund Investa Property Group 28/05/2003 AMP Diversified Property Trust Stockland Trust Management Ltd 28/05/2003 Prudential Investment Company Of Fexco Australia 09/07/2003 Securenet Ltd Bank One Corp 10/07/2003 AMP Industrial Trust Macquarie Goodman Funds Management Ltd 22/08/2003 Australian Growth Properties Ltd Trans Tasman Properties (AGP) Pty Ltd 29/08/2003 Peter Lehmann Wines Ltd Hess Group Australia Pty Ltd 1/09/2003 Hamilton Island Ltd 21st Century Resorts Holdings Pty Ltd Environmental Recovery Services 30/09/2003 Ltd Transpacific Industries Pty Ltd 22/10/2003 Hamilton Island Ltd Voyages Hotels & Resorts Pty Ltd 24/10/2003 Hamilton Island Ltd 21st Century Resorts Holdings Pty Ltd 122 Appendix B Robustness Tests Bl Introduction As mentioned in Chapter 7, we wish to confirm the robustness of the main results obtained in the combined analyses and presented in sections 7.1 to 7.3. Unlike linear regression models, relatively few effective tests for goodness of fit of a logistic regression exist. Following Jennings and Mazzeo (1993), in this Appendix, we confirm the robustness of the main results by estimating the bid structure effects separately via three logistic regressions. The results from this analysis are reported below and serve to confirm the relationships found in the combined analyses in Chapter 7. B2 Competing Bids Beginning with the likelihood of competition, the 'separate', or disaggregated, analysis is presented in Table B.1 below. The results show that, consistent with the combined analysis, the coefficient estimate on BPREM is positive, the coefficient estimate on PCASH is positive in the 'All Bids' sample but negative in the 'Takeovers' sub-sample, although statistically insignificant in both cases, the coefficient estimate on TOEHOLD is significantly negative, the coefficient estimate on TOEDUM is significantly positive, while the coefficient estimate associated with LNSIZE is positive in all 3 separate equations estimated. These results serve to confirm the relationships found in the combined analysis in Table 7.1. 123 Table 8.1 Results of a Logistic regression analysis examining the likelihood of a competing bid arising using three separate equations on a sample of 304 initial bids between January 1996 and December 20034• Sample Sample ao a, ll2 ll3 ll4 ll5 Group Size Pane!A1 All Bids 304 -3.0078* 0.0020 0.0621 (0.0939} (0.6079} (0.5231) LLR =0.5955 (0.7425} HL = 14.2181 (0.0763)* Takeovers 243 -4.1907** 0.0057 0.1292 (0.0333) (0.2791} (0.2274) LLR =2.4071 (0.3001} HL =6.9711 (0.5398} Pane!B2 All Bids 304 -2.9770* 0.1359 0.0584 (0.0981} (0.7083} (0.5474} LLR =0.4896 (0.7829} HL = 13.5774 (0.0935)* Takeovers 243 -3.8033** -0.1101 0.1200 (0.0489} (0.7801} (0.2555} LLR = 1.3495 (0.5093} HL =5.0859 (0.7484) Panel c3 All Bids 304 -2.4748 -0.0715*** 1.0968** 0.0339 (0.1785) (0.0058} (0.0206) (0.7318) LLR =9.4644 (0.0237)** HL = 4.9977 (0.7578) Takeovers 243 -3.5013 -0.0742*** 1.1296** 0.0972 (0.0810)* (0.0065} (0.0309} (0.3721} LLR = 10.0914 (0.0178)** HL = 2.0744 (0.9786) 1 The equation estimated is COMP= ao + a 1(BPREM) + a5(LNSIZE) + E1 (B.1.1) 2 The equation estimated is COMP = ao + a2(PCASH) + a5(LNSIZE) + E2 (B.1.2) 3 The equation estimated is COMP = 4 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. ***, **, * denotes significance at the 1%, 5% and 10% level respectively. 124 B3 Target Management Resistance Results from the 'separate' analysis of the likelihood of target management resistance are presented in Table B.2. Table B.2 Results of a Logistic regression analysis examining the likelihood of target management resistance using three separate equations on a sample of 304 initial bids between January 1996 and December 20034• Sample Sample PanelB2 BJ: Characteristics of Probability {Y=l] All Bids 304 -0.4144 0.2427 -0.0081 (0.7493) (0.3478) (0.9082) Takeovers 243 -2.4376* -0.1237 0.1405* (0.0935} (0.6721} (0.0814} LLRALL BIDS =10.8163 (0.0287}* * B2: Characteristics of Probability {Y=2] All Bids 304 4.5160 1.4508* -0.4666** (0.1792) (0.0749) (0.0162) Takeovers 243 2.8769 1.1132 -0.3407* (0.4003} (0.1719} (0.0856} LLRrAKEOVERS = 10.2979 (0.3570} Panelc3 CJ: Characteristics of Probability [Y=l J All Bids 304 -1.0192 -0.0470*** 1.6521 *** 0.0036 (0.4578) (0.0035) (0.0000) (0.9613) Takeovers 243 -2.6525* -0.0475*** 1.1785*** 0.1331 (0.0778} (0.0065} (0.0042} (0.1060} LLRALLBIDS =29.0709 (0.0001}*** C2: Characteristics of Probability [Y=2] All Bids 304 4.2099 0.0019 0.9396 -0.4250** (0.2068) (0.9535) (0.3163) (0.0262) Takeovers 243 2.8235 0.0042 0.4819 -0.3117 (0.4028} (0.8992} (0.6123} (0.1101} LLRrAKEOVERS =18.3527 (0.0054}*** 125 1 The equation estimated is REST= Cio + a 1(BPREM) + a5(LNSIZE) + E4 (B.2.1) 2 The equation estimated is REST= Cio + a2(PCASH) + a5(LNSIZE) + E5 (B.2.2) 3 The equation estimated is REST= ao + aJ(TOEHOLD) + ~(TOEDUM) + a5(LNSIZE) + E6 (B.2.3) where REST = trinomial variable taking a value of O if no resistance, 1 if target management resisted, and 2 if a neutral response was given BPREM = percentage bid premium PCASH = fraction of the offer in cash TOEHOLD = fraction of the target firm owned by the initial bidder before making the offer TOEDUM = binary variable taking a value of 1 if the initial bidder acquired prior ownership in the target LNSIZE = log of the target firm's market value of equity 4 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. ***, **, * denotes significance at the 1 %, 5% and 10% level respectively. Panels Al, B1, and Cl, above, consider the characteristics of resisted bids compared to those bids that were not resisted. We focus on these results since these bids make up the vast proportion of the data set. The results show that, consistent with the combined analysis, the coefficient estimate on BPREM is significantly negative, the coefficient estimate on PCASH is insignificant, the coefficient estimate on TOEHOLD is significantly negative, the coefficient estimate on TOEDUM 1s significantly positive, while the coefficient estimate associated with LNSIZE is insignificant in all 3 separate equations estimated. These results serve to confirm the relationships found in the combined analysis in Table 7.2. Additionally, the results with respect to Neutral bids (Panels A2, B2, and C2), as well as those for the 'Takeovers' sub-sample in all the panels, are also consistent with the results of the combined analysis reported in Table 7.2. 126 B4 Bid Success Table B.3 presents results from the 'separate' analysis of the likelihood of bid success. Table B.3 Results of a Logistic regression analysis examining the likelihood of bid success using three separate equations on a sample of 304 initial bids between January 1996 and December 20034• Sample Sample a2 Group Size PanelA1 All Bids 304 0.0172 0.0067* (0.9015) (0.0517) LLR = 4.4105 (0.0357)** HL = 4.4774 (0.8117) Takeovers 243 -0.4228** 0.1033** (0.0106) (0.0144) LLR =6.5032 (0.0108) ** HL = 4.8531 (0.7732) PanelB2 All Bids 304 0.3663* -0.3099 (0.0609) (0.2141) LLR = 1.5521 (0.2128) HL = 15.7002 (0.0469)** Takeovers 243 -0.2364 0.0938 (0.3072) (0.7424) LLR = 0.1081 (0.7423) HL = 9.4486 (0.3059) Panelc3 All Bids 304 0.4482** 0.0568*** -1.4126*** (0.0176) (0.0005) (0.0001) LLR = 17.7145 (0.0001)*** HL = 7.4471 (0.4893) Takeovers 243 -0.1624 0.0592*** -1.0176** (0.4864) (0.0009) (0.0135) LLR = 12.8470 (0.0016)*** HL = 5.7070 (0.6800) 1 The equation estimated is OUT= no+ a1(BPREM) + E7 (B.3.1) 2 The equation estimated is OUT = 4 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. ** *, * *, * denotes significance at the 1 %, 5% and 10% level respectively. 127 The results presented are again consistent with the combined analysis. The coefficient estimate on BPREM is significantly positive, the coefficient estimate on PCASH is insignificant, the coefficient estimate on TOEHOLD is significantly positive, and the coefficient estimate on TOEDUM is significantly negative. These results serve to confirm the relationships found in the combined analysis in Table 7.3. 128 Appendix C Analysis of Non-Initial Bids Cl Introduction The empirical analysis in the main body of this thesis considers only initial takeover bids. However, a significant proportion of transactions involve more than one bid. Subsequent bids can come from competing bidders or may represent revised bids by the initial bidder as part of the negotiation process typical of takeover transactions. These non-initial bids represent takeover offers of equal validity to the initial offers included in the main data set, as outlined in Chapter 5. As a result, in this appendix, these bids will be added to the 'Initial' data set. Empirical testing of the same nature as that carried out in Chapter 7 will be performed on this expanded data set. The results from this analysis will then be used to demonstrate why, contrary to first appearance, this expanded data set cannot be used, in aggregated form, to determine the validity of the hypotheses presented in Chapter 4. C2 Data Using the same sample selection criteria as the 'Initial' sample, the expanded sample contains 439 takeover bids announced between January 1, 1996 and December 31, 2003. This consisted of 274 Initial bids, 127 Revised bids and 38 Competing bids. A definition of these terms is provided here: 129 (I) Initial Bids - this refers to the first bid/offer made for a target in a takeover transaction (i.e. no other offer for the target exists at the time). This is obviously made by the first or initial bidder. (11) Competing Bids - this refers to the first bid/offer made for a target by a bidder other than the first bidder. In some competitive transactions there may be more than one competing bid. (111) Revised Bids - this refers to all bids/offers in a transaction other than the first bid made by each bidder. C2.1 Competition Status Competition Status refers to classifying each bid depending on whether or not another bidder entered a new offer, for the same target, after the announcement date of the bid in question, be it initial or otherwise. In the case of multiple-bidder transactions, where a bidder revises upwards their bid in response to an opposing bidder's latest bid, the opposing bidder's bid is deemed to have received competition. However, where a bidder revises their bid upwards in a single-bidder transaction, this does not represent competition. C2.2 Bid Outcome For the purpose of this analysis, in a successful transaction that involves multiple bids, only the final bid by the winning bidder is considered successful, with all other bids by that bidder and all other bidders considered unsuccessful. 130 C2.3 Bidder Toehold As previously defined, Bidder Toehold refers to the size of the stake (percentage) in the target held by the bidder immediately prior to the announcement of the first offer by that bidder. Hence, in multiple-bidder transactions, the recorded toehold size is the same for all of a particular bidder's bids Table C.1 below provides some descriptive statistics on the 439 takeover bids used in the empirical analysis in Section C4 that follows. 131 Table C.l Descriptive (percentage) statistics summarising a sample of 439 Initial, Competing and Revised Bids made by 311 bidders in 273 takeover transactions between January 1, 1997 and December 31, 2003. Transaction Comu.etition Bid Distribution Competing Transactions 37 14% Initial Bids 273 62% Non-Competing Transactions 236 86% Revised Bids 128 29% 273 Competing Bids 38 9% Takeover Mode 439 Scheme of Arrangement 52 19% Bid Success Takeovers 221 81% Successful Bids 218 49.7% 273 Unsuccessful Bids 221 50.3% 439 Transaction Outcome Not Consummated 55 20% Director's Recommendation Consummated 218 80% Accept 243 55% 273 Reject 171 39% Other 25 6% 439 Bid Comu.etition Competing Bid Received 100 23% No Competing Bid Received 339 77% 439 Medium ofExchange Cash Bids 288 66% Scrip Bids 96 22% Mixed Bids 55 12% 439 Frequency Distribution of Bid Premium Frequency Distribution of Toehold BP:'.S0.10 25% TOE=0 38% 0.10 < BP :'.S 0.20 17% 0.00 < BP :'.S 0.05 9% 0.20 < BP :'.S 0.30 12% 0.05 < BP :'.S 0.10 10% 0.30 ~ BP :'.S 0.40 12% 0.10 ~ BP :'.S 0.15 8% 0.40 < BP :'.S 0.50 11% 0.15 < BP :'.S 0.20 21% 0.50 < BP :'.S 0.60 5% 0.20 < BP :'.S 0.25 5% 0.60 < BP :'.S 0.70 5% 0.25 < BP :'.S 0.30 2% 0.70 < BP :'.S 0.80 4% 0.30 < BP :'.S 0.35 2% 0.80 < BP:'.S 1.0 4% 0.35 < BP :'.S 0.40 2% BP> 1.0 5% 0.40 < BP :'.S 0.50 3% Median BP = 0.27 Median TOE = 0.072 Calendar Year oflnitial Proposal 1997 11% 1998 15% 1999 15% 2000 19% 2001 20% 2002 12% 2003 8% 132 C3 Methodology C3.1 Bid Premium Calculation Whereas the calculation of bid premium for initial bids is reasonably straightforward, some adjustments must be made when performing this calculation for subsequent bids. Failure to estimate these values correctly can result in vastly distorted measures of bid premium. C3.1.1 Target's pre-bid value As mentioned previously, the bid premium in a takeover transaction refers to the excess of the offer value over the pre-offer price of the target. In the empirical analysis in the main body of this thesis, the target's share price 40 trading days prior to the announcement of the Initial offer is used to calculate bid premium. It is expected that taking the target's price on this date will effectively account for the leakage effect. With respect to the analysis in this appendix, it is important to note that this price at time T-40 is used to calculate bid premium for all bids made in a transaction: Initial, Competing, and Revised Bids. The reason for this again relates to information leakage and the desire to determine the premium over the target's stand alone value, the real bid premium. When calculating bid premium for revised or competing bids, an alternative that may seem appealing at first is to adjust this reference date, and the target's share price, for each new bid. However, closer consideration reveals that such a method would result in gross miscalculation of the real bid premium. The most obvious example of such a miscalculation arises when a competing or revised bid is made more than 40 trading days after the initial bid. The 133 price of the target 40 trading days prior to such a new bid would obviously incorporate explicit information of the initial bid and therefore the bid premium calculated using this price would clearly not be a measure of premium over stand alone value, or the true premium. Therefore, consistent with Betton and Eckbo (2000), bid premium for all bids in this study are calculated relative to the target's price 40 trading days prior to the announcement of the initial offer (T-40). Again, this price was adjusted for any dividends and capitalisation changes in the target between day (T-40) and the announcement date. C3.1.2 Offer Value Consistent with Jennings and Mazzeo (1993) and numerous other studies, the bidder's price on the trading day immediately prior to the initial announcement (T-1) was used to calculate offer value, and therefore bid premium, for initial bids that included stock. However, for competing bids, the competing bidder's share price on the trading day immediately prior to the announcement of the competing bid (C-1) is used to calculate bid premium, not T-1. Similarly, for subsequent revised bids by the initial or competing bidders, the bidder's price on the trading day immediately prior to the announcement of that revised bid is used (R-1). It must be noted that using this method can lead to a situation where an upwards revision of a scrip offer sees a reduction in the bid premium from the initial ( or competing) bid to the revised bid. The explanation for this lies in the abnormal returns literature related to takeovers. Previous studies have shown that the stock 134 price of bidders making scrip and mixed offers tend to fall, often significantly, shortly after announcement of the initial offer. Although such a situation seems somewhat counter-intuitive, it is entirely consistent with the reporting of bid premium, and the experience of bidders, in the takeover market. When scrip consideration is used in the offer, the bid value is exposed to the whims of the market. The bid value fluctuates with the bidder's share price, which often experiences high post-announcement volatility as the market expresses its view of the takeover proposal. Cases where the bidder's shares are sold down heavily following the initial bid, significantly affecting the bid value and premium offered to target shareholders, are not uncommon in the marketplace. In such cases bidders are often forced to significantly increase their bid in order to consummate the transaction, at times doing so by including cash in the offer. A potential alternative to the above method is to use the bidder's share price prior to their involvement in the transaction in calculating bid premium for revised bids (e.g. use T-1 for all bids by an initial bidder). However, doing so would not be representative of the situation facing target shareholders and directors, as well as bidder management, in the takeover market. As mentioned previously, from the point of view of target shareholders and directors, of most relevance in assessing the adequacy of a scrip bid is the bidder's most recent share price, as this provides the best available estimate of future share price (of the combined firm). The bidder's share price prior to its involvement in the transaction is irrelevant since there is no reason to believe the combined firm's share price will reflect the bidder's pre-transaction price in the event of the transaction being consummated. On the contrary, studies have found that, post-takeover, share prices of successful acquirers paying with scrip are prone to long-run underperformance. For example, Loughran and Vijh (1997) found that 135 stock-returns of acquirers in U.S. scrip mergers underperformed a control sample of non-merging firms in each of the first three years post-takeover. Similarly, the sample of all scrip-paying acquirers underperformed by 24.2% over a five-year period. Hence, the advantage of the method chosen to calculate bid premium is that it more closely reflects the situation faced by target directors when deciding whether they should recommend the bid, as well as that of target shareholders in their decision to accept or reject the bid. This consequently impacts on the subsequent actions of the bidder. C4 Results C4.1 Competing Bids Table C.2 below presents results of a logistic regression examining the likelihood of observing bid competition. The regression run is identical to Equation (7 .1) in Chapter 7 but uses the expanded sample of 439 initial, competing and revised bids. Compared to the analysis of initial bids only, the above results differ noticeably. Firstly, bid premium is shown to be positively and significantly related to the likelihood of competition. This result is completely opposite to that predicted by competition hypothesis (H-la) and suggests that a higher bid premium actually increases the likelihood of observing competition. Secondly, the toehold variables, TOEHOLD and TOEDUM, lose all significance in their relation to competition. At first glance, these results seem nonsensical, particularly when compared to the results in Table 7.1. However, a simple methodology-related explanation exists. These results are due to the addition of competing and revised bids to the analysis and reflect the competitive bidding situation captured by their inclusion. Firstly, the 136 Table C.2 Results of a Logistic regression examining the likelihood of a competing bid arising using the bid premium, the proportion of cash in the offer, size of the toehold, existence of a toehold, and target size as explanatory variables from a sample of 439 initial, competing and revised bids between January 1997 and December 2003 1•2• Sample Sample Expanded 439 -1.8616 0.0079** 0.3333 -0.0127 0.4020 0.0265 (0.1406) (0.0023) (0.2412) (0.4078) (0.2421) (0.6882) LLR = 15.9049 (0.0071)*** HL = 15.9830 (0.0426)** 1 The equation estimated is COMP= where COMP = binary variable taking a value of 1 if a competing bid arises BPREM = percentage bid premium PCASH = fraction of the offer in cash TOEHOLD = fraction of the target firm owned by the initial bidder before making the offer TOEDUM = binary variable taking a value of 1 if the initial bidder acquired prior ownership in the target LNSIZE = log of the target firm's market value of equity 2 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. ***, **, * denotes significance at the 1 %, 5% and 10% level respectively. mere presence of these bids signals a competitive bidding situation where at least two bidders are seeking to acquire the same target. In such situations, the likelihood of observing competition in the form of upward revisions by the opposing bidders, and to a lesser extent the entry of additional bidders, is higher compared to the analysis of initial bids only. This is because the majority of transactions are non competitive in the sense that initial bidders do not encounter any competition from other bidders. The second implication of including non-initial bids in the analysis is that these bids will necessarily have higher bid premiums when compared to the initial bids sample. It is obvious that competing and revised bids must be of a higher value to target shareholders, than the previous offer by the opposing bidder, in order to have any chance of success. Combining these two characteristics of non-initial bids, the reason behind the above results becomes apparent. Taking the bid premium result first, the 135 non-initial bids included in the analysis provided observations 137 that, on average, had higher bid premiums and a higher likelihood of attracting competition. These characteristics combined to produce the positive and significant relationship between BPREM and the likelihood of observing competition, giving the impression that higher bid premiums encouraged, rather than deterred, competing bidders. However, these results conceal the true relationship and rather represent the auction-like situation that develops in multiple-bidder transactions. The toehold results can likewise be explained by the simultaneous inclusion of initial and non-initial bids in the analysis. In this instance, non-initial bids again introduce the competitive bidding situation, where actual bidders, having invested their time, funds and reputation into a public takeover bid are more likely to continue fighting against an opposing bidder, by making upward-revised bids, regardless of the size of their opponent's toehold. In other words, the likelihood of these actual bidders making a revised bid is higher than the likelihood of a potential bidder making a competing bid (which is the likelihood being measured in the analysis of initial bids in Table 7.1). Unlike bid premium, each bidder's recorded toehold does not increase as they make subsequent revised bids, rather remaining at the same level. However, the addition of non-initial bids to the analysis still distorts the relationships observed when only initial bids were considered, resulting in the TOEHOLD and TOEDUM variables losing their significance, when compared to the results in Table 7.1. 138 C4.2 Target Management Resistance Table C.3 presents results of a logistic regression analysis examining the likelihood of encountering target-management resistance. The regression run is identical to Equation (7.2) in Chapter 7 but uses the expanded sample of 439 initial, competing and revised bids. Unlike the competing bids analysis in the previous section, the inclusion of non-initial bids does not significantly alter the results obtained, or the conclusions drawn, when compared to the analysis of initial bids in the main body of this thesis. Table C.3 Results of a Logistic regression analysis exammmg the likelihood of target management resistance using the bid premium, the proportion of cash in the offer, size of the toehold, existence of a toehold, and target size as explanatory variables from a sample of 439 initial, revised and competing bids between January 1997 and December 20031•2• Sample Sample Clo U1 Uz U3 U4 U5 Grou Size Pane/A: Characteristics of Probability [Y=l] All Bids 439 0.0447 -0.0068** -0.0673 -0.0412*** 1.5555*** -0.0390 (0.9678) (0.0130) (0.7857) (0.0018) (0.0000) (0.5056) Panel B: Characteristics of Probability [Y=2] All Bids 439 3.3092 0.0032 1.5476** 0.0118 0.0358 -0.4077*** (0.1949} (0.3781} (0.0371} (0.6121} (0.9571} (0.0040} LLR = 52.6586 (0.0000}*** 1 The equation estimated is REST= Cio + a 1(BPREM) + a2(PCASH) + a 3(TOEHOLD) + ~(TOED UM)+ as(LNSIZE) + E2 where REST = trinomial variable taking a value of O if no resistance, 1 if target management resisted, and 2 if a neutral response was given BPREM = percentage bid premium PCASH = fraction of the offer in cash TOEHOLD = fraction of the target firm owned by the initial bidder before making the offer TOEDUM = binary variable taking a value of 1 if the initial bidder acquired prior ownership in the target LNSIZE = log of the target firm's market value of equity 2 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. * * *, * *, * denotes significance at the 1 %, 5% and 10% level respectively. 139 As per the 'Initial' results in Table 7.2, BPREM and TOEHOLD are found to be negatively and significantly related to the likelihood of resistance, while TOEDUM is positively related at the 0% significance level. When comparing bids that received a neutral response (REST = 2) with those bids that were not resisted (REST =0), Panel B again shows PCASH (positive) and LNSIZE (negative) to be the only statistically significant variables, mirroring the 'Initial' results. This consistency in results serves to confirm the conclusions drawn in the main body of the thesis with respect to the likelihood of encountering target-management resistance. C4.3 Bid Success Table C.4 presents results of a logistic regression examining the likelihood of a takeover bid being successful. The regression run is identical to Equation (7.3) in Chapter 7 but again uses the expanded sample of 439 initial, competing and revised bids. These results are consistent with those obtained in the 'Initial' bids analysis in Table 7.3, with one exception. This exception relates to the bid premium variable. Whereas BPREM was found to be positively and significantly related to the likelihood of bid success in the 'Initial' bids analysis, this variable loses all significance when non-initial bids are included in the analysis. The reason for this result, much like those of the competition analysis in section C4.1 of this appendix, lies in the addition of non-competing bids to the analysis and reflects the competitive bidding situation captured by their inclusion. Competing and revised bids carry higher bid premiums, on average, than initial bids due to the fact that they are either part of a bidding war with a competing bidder/s, are encountering target management resistance, or both. However, for the same reasons, they are more likely 140 Table C.4 Results of a Logistic regression examining the likelihood of bid success using the bid premium, the proportion of cash in the offer, size of the toehold, and existence of a toehold as explanatory variables from a sample of 439 initial, competing and revised bids between January 1997 and December 20031'2• Sample Sample Clz Grou Size All Bids 304 0.1846 0.0019 -0.3304 0.0350*** -0.6312** (0.3562) (0.4054) (0.1523) (0.0042) (0.0298) LLR = 11.3378 (0.0230)** HL = 12.1622 (0.1441) 1 The equation estimated is OUT= Clo + a 1(BPREM) + a 2(PCASH) + aJ(TOEHOLD) + a 4(TOEDUM) + &3 where OUT = binary variable taking a value of 1 if the initial bid is successful BPREM = percentage bid premium PCASH = fraction of the offer in cash TOEHOLD = fraction of the target firm owned by the initial bidder before making the offer TOEDUM = binary variable taking a value of 1 if the initial bidder acquired prior ownership in the target 2 The P-levels of the estimated coefficients from a two-tailed test of significance are in parentheses. ***, **, * denotes significance at the 1 %, 5% and 10% level respectively. to be unsuccessful. For example, consider a transaction involving two competing bidders who make three bids each, with each bid offering a higher amount than the previous bid. This bidding produces five non-initial bids with inflated bid premiums, due to the competitive-bidding situation, however only one of these bids, if any, will be successful. Hence, in this setting, it is clear that a strong positive relation between bid premium and bid success cannot emerge. On the other hand, when analysing initial bids only, we are dealing primarily with single-bidder transactions. In this setting, a strong positive relation between bid premium and bid success does hold, as confirmed by the results in Table 7.3. Thus, the combination of bids from these two settings has an offsetting effect, blurring the actual relationship between bid premium and the likelihood of bid success and leading to the insignificant result shown in Table C.4 above. 141 CS Conclusion The empirical analysis undertaken in this appendix, whilst valid, highlighted some significant issues in the simultaneous analysis of initial takeover bids and non-initial takeover bids. These issues largely reflect the different bid-settings: the largely non competitive setting of initial bids and the highly competitive setting in which non initial bids are made. This leads to results that are distorted and less straightforward to interpret than those obtained from separate analyses of initial and non-initial bids. This was evident in the competition and bid success analyses and most severe with respect to the bid premium variable. The issues highlighted in this appendix perhaps explain why analysis of bid competition and bid outcome in takeovers has, in the existing literature, typically been restricted to the analysis of initial bids, as was the case in the main body of this thesis. 142