1......
. . .1...
. ,1 . . :
ﬁrm: .

.7...

.5 .

.

3;.
. .. _
3 U...

 

a! . I?! .
It: .1

Ian—Jhrkﬁ .3: . :5
a: 3 a”... .
an. max... a

.t.
,.

.05.... . O. m“
. 2...... . , _ . 2 . V 5.... we 5 5......“
. , . V . . .2 “Sun. .3.» o in.»
:3»). . V. . . . In. «3:1,... . svl| I:
_.. 5...: . ant}. . bra
o r a. , . . . 71.7.: a...» I: a.
y. 2;. , 2.. . . . . Sad... 6.1.2.32... .
v.33 . .55: .2 an
9; ,

 

 

, .35..
to... hurl
pt

. 28.... . . ﬂag.
an? ._ amahwﬁmqmgugmﬂuﬂ
. .. 2

 

.1

I
..:\l .1 3.: 83-9.. ’ Fix...
1... . a Shitanh‘ 19.: 5N
.a .max. ark-pr. .

an". nun:

7....
2.3. .3:
. 5.591;!

 

..... . . 1 9 £1.“an i

. . . , a 2 - .

. x. . . , 1:1. 5.... . at ..
i . . . . a ,

in... b... _. . . . 2 . . Juan”? 5.2. . 23.1%....

gasps... . . . . . .

9‘ At. 1 KIA-u- at
g
u ,
1......r.m..m§.. . ».
«Eb. am»...

.3. 3.
int...

1. 1‘.
.3 ?9..§
.4. . It}. t:)le..€§u
3.1:}. €1.90“? 59'.
39.523!!! 3.5!! “no... ‘
IIAIrltJ .. 3. cl...
:2! . , . It“...
~ 01"
. .3513!!! I.
.4121: sis... (0:535:21:
ii 3:..vlltxhrrap13: :35)!
.3 ..iz..§v)sa..1...f.. ’55.... 7 ,
1...! 43H..As.u¢l
7: . .. :.. A €9.23.
0...)!!!

 

 

2 «L.
1....973.
5.1.1.5.?
33.4. . I!

.0

 

 

t: . 1...! .

1:. :22 :ii!
at... $33.4... «‘23:.
, 3.79....3

i

91.... 1.1L:
.(‘xﬁlii
4.?

 

 

 

 

 

1):] Ir 1.... .nli’:.)l.lvtll
0113......(4! ow. :?,ltl4v. :0... 2...

 

 

          
 

       

IV '11:! . .u

   

     

           

   
  

        

      
     

        

  
 
 

 

h f A. I» 3., 2.15.4315: 3 , .. .3 f: z. .1 . .. A -1. ._ .-,...v... . . ._ .. x 1 tum—Ly . . nwx.:.\.1v.mm..u..wblf.¢
gag maﬁa". . .ﬁagﬁgﬁ gaﬁ.ﬁg¥%§-§§ Egg..- .....1
_. a... may? ,.!. V. .. ._.._ .. . ..:;... .H. ._ ..........s .3. - .. : , . .. e . . . .

  

l’l

 

 

‘1'

 

LIBRARY

Michigan State
University

 

 

 

This is to certify that the

dissertation entitled

The Role of Managerial Incentives in Corporate
Acquisitions and Corporate Mergers and Top
Management Turnover: The 19903 Evidence

presented by

David Scott North

has been accepted towards fulﬁllment
of the requirements for

 

 

Ph . D . degree in Finance

awe [jg/{ck

 

Major pr‘fessor

Datejjzojoo

MS U is an Afﬁrmative Action/Equa! Opportunity Institution 0- 12771

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mm mm.“

 

THE ROLE OF MANAGERIAL INCENTIVES IN CORPORATE ACQUISITIONS
AND CORPORATE MERGERS AND TOP MANAGEMENT TURNOVER: THE
19908 EVIDENCE.

By

David Scott North

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the degree requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Finance

2000

ABSTRACT

THE ROLE OF MANAGERIAL INCENTIVES IN CORPORATE ACQUISITIONS
AND CORPORATE MERGERS AND TOP MANAGEMENT TURNOVER: THE
19908 EVIDENCE.

By

David Scott North

This dissertation investigates the 19905 market for corporate control. The ﬁrst
essay examines the relationship between the likelihood a ﬁrm is acquired and the
governance and ﬁnancial characteristics of the ﬁrm. Given many of the developments in
the corporate control market in the late 19805, I suspect that the process governing
takeover likelihood may have changed in the 19905. I examine a sample of 342
NYSE/AMEX ﬁrms that were acquired during the 1990-1997 period and compare them
to a matched sample of non-acquired ﬁrms. 1 ﬁnd that ﬁrms that were acquired over this
period can be characterized as having lower growth, lower managerial ownership, higher
ownership by outside directors, and greater ownership by unafﬁliated blockholders with
board representation. The fact that managerial ownership is negatively related to
takeover likelihood is consistent with studies using data from the 19705 and 19805. This
suggests that managerial ownership helps managers maintain control, or alternatively that
ownership proxies for how much managers care about control.

The second essay examines the different roles a merger plays is disciplining or
rewarding the performance of top managers. One role is to discipline the top managers of
poorly performing target ﬁrms, and another, to reward the management practices of well

performing top managers. Given many of the developments in the corporate control

market in the late 19805, I suspect that the role of disciplinary takeovers may have
changed in the 19905. I examine a sample of 236 NYSE/AMEX ﬁrms that were acquired
during the 1990-1997 period and document that the turnover rate for the top manager of
target ﬁrms signiﬁcantly increases following completion of the merger. Target ﬁrms
with no post-merger change in the top executive were signiﬁcantly out-performing other
ﬁrms in their industry, performing at least as well as others in the market, and out-
performing targets with a change in the top executive prior to the takeover. Target ﬁrms
with a post-merger change in the top executive were signiﬁcantly under-performing other

targets within their industry as compared with ﬁrms with no post-merger change.

ACKNOWLEDGEMENTS

I would like to express my appreciation to the members of my dissertation
committee - Dr. Edward Fee, Dr. Richard Simonds, Dr. Michael Mazzeo, and Dr.
Charles Hadlock, for their insightful guidance and encouragement. I would especially

like to thank Dr. Charles Hadlock for all of his help patience, and commitment of time.

iv

TABLE OF CONTENTS

 

LIST OF TABLES vii
ESSAY 1. The Role of Managerial Incentives in Corporate Acquisitions: The 19905
Evidence
1 .1 Introduction 1
1.1.1 Historical Perspective of Managerial Ownership and the Market for
Corporate Control 1
1.1.2 Changes in the Market for Corporate Control 2
1.1.3 Summary of Essay 7
1.2 Hypothesized Effects of Governance Structures and Financial Characteristics 9
1.2.1 Governance Based Hypotheses 10
1.2.2 Financial Hypotheses 12
1.3 Sample Selection and Empirical Design 13
1.3.1 Population Selection 13
1.3.2 Sample Selection 15
1.3.3 Variable Construction 17
1.3.2.1 Ofﬁcers and Directors 18
1.3.2.2 Block Ownership 19
1.3.2.3 The Top Executive 20
1.3.2.4 Financial Information 21
1.4 Empirical Results 22
1.4.1 Univariate Results 22
1.4.1.1 Financial Characteristics 22
1.4.1.2 Board of Directors Characteristics 24
1.4.1.3 Ownership Characteristics 24
1.4.1.4 CEO Characteristics 25
1.4.2 Multivariate Results 25
1.4.3 Hypotheses Tested 28
1.4.3.1 Governance-based Hypothesis 29
1.4.3.2 Financial Hypotheses 30
1.4.4 Robustness 30
1.4.4.1 Voting Rights 31
1.4.4.2 Additional Control Variables 31
1 .5 Conclusion 33

ESSAY 2.

2.1

2.2

2.3

2.4

2.5

2.6

2.7

Introduction

Background and Hypothesis Development

2.2.1 Previous Literature on Ownership and Takeovers
2.2.2 The Changing Corporate Control Environment
2.2.3 Hypothesis Development

2.2.4

Data .

2.3.1 Population Selection

2.3.2 Sample Selection

Management Turnover
2.4.1 Turnover Frequency
2.4.2 Top Manager Turnover Pre- vs. Post-Merger

Measuring Corporate Performance

2.5.1 Industry-Adjusted Returns

2.5.2 Market Model Procedure

2.5.3 Performance Results
2.5.3.1 Disciplinary vs. Nondisciplinary Takeovers and Pre-Takeover

Performance

2.5.3.2 Disciplinary vs. Nondisciplinary: Replacement Classiﬁcation
2.5.3.3 Disciplinary vs. Nondisciplinary: Reason Classiﬁcation
2.5.3.4 Pre-Takeover Performance: Sub-Periods

Additional Investigations
2.6.1 Hostile versus Friendly Takeovers
2.6.2 Additional Characteristics

Conclusion

APPENDIX 1: TABLES OF ESSAY 1

APPENDIX 2 TABLES OF ESSAY 2

APPENDIX 3 DATA DEFINITIONS

BIBLIOGRAPHY

vi

Corporate Mergers and Top Management Turnover: The 19905 Evidence

37

4O
40
41
43

45
45
46

47
47
52

53
54
54
55
55
57
58
59
61
61
62
67
69
96
107

111

Table 1.1

Table 1.2

Table 1.3

Table 1.4

Table 1.5

Table 1.6

Table 1.7

Table 1.8

Table 1.9

Table 1.10

Table 2.1

Table 2.2

Table 2.3

Table 2.4

Table 2.5

Table 2.6

Table 2.7

Table 2.8

LIST OF TABLES

Merger Classiﬁcation: Hostile vs. Non-Hostile
Hypotheses Tested

Population Summary

Univariate Results

Univariate Results, Continued

Multivariate Logit Results (% of stock owned)
Hypotheses Results

Multivariate Logit Results (% of votes owned)
Additional Multivariate Results

Hypotheses Results Comparison

Merger Classiﬁcation: Hostile vs. Non-Hostile

Population Summary, 1990 to 1997

Top Manager Turnover for Target Firms: Turnover Frequency,

1990 -— 1997

Top Manager Turnover for Target Firms: Reasons for Turnover,

1990 - 1997

Top Manager Turnover for Target Firms: Replacement Manager

Origin, 1990 — 1997

Turnover Characteristics: Pre-merger vs. Post-merger

Pre-Takeover Cumulative Abnormal Returns for Disciplinary and

Nondisciplinary Takeover Targets

Cumulative Industry-Adjusted Returns and Cumulative Average
Prediction Errors for Various Time Periods Relative to the

Merger Announcement

vii

70

71

73

74

78

82

86

88

92

94

97

98

99

100

101

102

104

LIST OF TABLES (CONTINUED)

Table 2.9 Post-merger: Top Manager Remains or Departs 105

Table 2.10 Post-merger: Top Manager Remains or Departs, Continued 106

viii

ESSAY l: The Role of Managrial Incentives in Corporate Acquisitions: The 1990s

Evidence

1.1 Introduction
1.1.1 Historical Perspective of Managerial Ownership and the Market for Corporate
Control

There has been a great deal of interest in managerial ownership in the ﬁnance
literature over the past 30 years. In particular, literature dating back to Jensen and
Meckling (1976) hypothesizes that increased managerial ownership encourages
managerial diligence and effort because managers bear a higher proportion of the cost of
any inappropriate decisions. Nowhere is the potential conﬂict of interest between
managers and shareholders larger than in the arena of corporate takeovers. While
agreeing to be acquired is often quite proﬁtable for shareholders, it often results in
managers losing their jobs.

Jensen and Meckling (1976) argue that as managerial ownership of the ﬁrm increases
the interest of managers and outside shareholders become more closely aligned. As the
interests of the two parties converge, fewer acquisitions will be made for motives other
than shareholder wealth maximization. Walkling and Long (1984) and Morck, Shleifer,
and Vishny (1988a) ﬁnd evidence that the more votes managers control, in a target ﬁrm,
the more likely the takeover succeeds. This is consistent with the view that the cost to
bidders of negotiating with a small number of large shareholders is less than the cost of
negotiating with managers with diverse ownership. Demsetz and Lehn (1985) argue that

there exists a positive relationship between takeovers and managers' voting control. They

ﬁnds that managers are less likely to maximize ﬁrm value when they have majority
voting control. This reduced ﬁrm value with majority ownership by managers can attract
takeover offers. A bidding ﬁrm is attracted to the possibility of taking over the ﬁrm,
replacing management, maximizing ﬁrm value, and in turn creating value. Stulz (1988)
describes a model in which high levels of managerial stock ownership are harmful to
shareholders because managers become insulated ﬁom some corporate governance
mechanisms, such as the market for corporate control.

The empirical evidence on the impact of managerial ownership on the likelihood of a
ﬁrm being acquired has been mixed. Song and Walkling (1993) ﬁnd evidence of
management valuing control during the 1977 — 1986 market for corporate control. They
ﬁnd that targets of acquisition have lower managerial ownership than either industry
matched non-targets or randomly selected non-targets. Shivdasani (1993), in a study of
hostile takeovers that took place during 1980-1988, ﬁnds that hostile targets have
signiﬁcantly lower managerial ownership than matched non-targets. Mikkelson and
Partch (1989) ﬁnd that the likelihood of successful acquisitions of ﬁrms is unrelated to
managerial ownership in their sample from the 1972 - 1987 market for corporate control.
Taken together, the empirical evidence on the impact of managerial ownership in the

market for corporate control appears inconclusive.

1.1.2 Changes in the Market for Corporate Control
There has been a great deal of variation in the volume and type of control activity
over time. The late 19705 were characterized by two major trends: takeovers of large

public corporations and a return to corporate focus, with the divestiture of unrelated and

unproﬁtable operations. During the 19705 the number of merger deals declined, while
the size of the deals increased.1 The 1980's became the decade of the mega-deals with
the acceptance of hostile takeovers, leveraged buyouts and strategic mega-mergers. The
creation of the junk bond market in the 19805 made deal ﬁnancing easier to obtain and
allowed banks and brokerage ﬁrms to ﬁnance leveraged buyouts. Beginning in the late
19805 and continuing into the very early 19905, the junk bond market weakened and this
type of ﬁnancing began more difﬁcult to obtain. The total value and number of all
corporate deals hit a low in 1991 due to the aforementioned junk bond market weakening
and political and economic uncertainty in the US. Starting in 1992, and continuing until
1997, the number and value of deals increased annually as the US. economy experienced
economic recovery.2

The mergers of the 1990’s appeared to differ from those of the 19705 and 19805.
The latest round of takeover activity seemed to be ﬁiendlier in nature and focused more
toward expanding markets. This is compared with the control contests, breakups, and
downsizing of the previous periods. Following Walkling and Long (1984), I classiﬁed all
mergers in my sample based on the degree of resistance by management.3 Table 1.1
provides a summary of this classiﬁcation and describes a corporate control market that is
not very hostile. For the 374 mergers from the population considered in this essay, only
8.8% of the mergers can be classiﬁed as hostile, with most of the hostile classiﬁcation

being only verbal opposition by management or initially contested, but target

 

1 See Mergerstat Review (1999)

2 See Mergerstat Review (1999)

3 Walkling and Long (1984) classiﬁed tender offers in their sample into one of six
categories: 1.) No resistance by management, 2.) Verbal opposition by management, 3.)
White knight sought, 4.) Court action ﬁled, 5.) White knight and court action ﬁled, 6.)
Initially contested, but target management later agrees to a higher premium.

3

management later agrees to a higher premium. The number of successful hostile
takeovers remained fairly constant throughout the 1990's with a low of 7.1% in 1996 and
1997, and a high of 13.8% in 1995 of all completed mergers being classiﬁed as hostile.
This evidence is compared with Shivdasani (1993), where he ﬁnds approximately 50% of
all corporate control activity was classiﬁed as hostile from 1980 to 1988.4

Much has been made of this distinction of takeovers as either hostile or friendly.
Recently, Schwert (2000) in a study of 2,346 takeover contests from 1975 to 1996
investigates the issue of hostile vs. friendly classiﬁcation. He ﬁnds that the highest rates
of hostile offers take place between 1982 to 1989. Additionally, he ﬁnds that after 1991
there was a statistically signiﬁcant shift away from hostility comparable in magnitude to
the shift towards hostility seen in the early 19805. My classiﬁcation results are the same
as that of Schwert’s classiﬁcation of the period 1990 to 1997. His annual proportion of
hostile deals during this period ranges from 2.0% to 12.0%, and mine from 7.1% to
13.8%. However, Schwert’s empirical tests show that most deals described as hostile in
the press are not distinguishable from friendly deals in economic terms, except that
hostile transactions involve publicity as part of the bargaining process. Therefore, I do
not make much about this distinction except to describe the corporate control
environment as friendlier in the 19905.

There are some underlying factors as to why the corporate control environment
became 'friendlier'. The hostility of the 19705 and 19805 was slowed by the actions taken
by both corporations and state governments. To slow the activity in the corporate control

market corporations adopted many forms of anti-takeover amendments and adopted

 

4 Shivdasani (1993) uses all ﬁrms in Value Line lnvestrnent Survey as his population of
ﬁrms.

shareholder rights plans. State governments passed many forms of anti-takeover
legislation to stop hostile transactions.

Corporations adopted corporate amendments such as 1.) Super-majority
provisions - the requirement that a super-majority of shareholders, typically 66% to 80%,
approve a merger, 2.) Fair price provisions - require the bidder to meet either a speciﬁed
fair price or to receive super-majority approval of the voting shareholders, 3.) Staggered
terms of directors on the board, 4.) Blank check preferred provision - provide directors
with broad discretion to issue new stock and establish voting rights, 5.) Restricting the
ability of shareholders to call special meetings, and 6.) Elimination of cumulative voting
rights.5 All of these amendments increase the cost of successfully completing a hostile
takeover of a corporation.

Corporations additionally passed shareholder rights plans, or what are otherwise
known as poison pills. Shareholder rights plans create impossibly burdensome
obligations for anyone who buys a controlling block of shares without management
approval. These plans can take many forms, with the most common form being ﬂip-over
rights. Flip-over rights allow the target ﬁrm shareholders to purchase shares of the
acquiring ﬁrm at a substantial discount to the prevailing market price. These shareholder
rights plans obviously increase the cost to acquire a ﬁrm who does not want to be taken
over.

Additionally, states passed many types of anti—takeover legislation such as control
share laws and business combination laws intending to slow the corporate control market.
Control share laws restricted the voting rights of a controlling shareholder. These laws

generally call for a shareholder vote to decide if the law's restrictions on voting rights are

to apply in a given instance. These laws are triggered by a disclosure of intent to seek
control. Business combination laws delay any type of business combination. Business
combination laws are also referred to as merger moratorium laws or freeze out laws. The
main difference between control share laws and business combination laws is that
business combination laws are triggered by the actual purchase of shares rather than the
disclosed intent to purchase for control reasons. In 1988 the state of Delaware passed its
version of business combination laws covering approximately ﬁfty percent of all NYSE
and AMEX listed ﬁrms. Both types of laws apply to the buyer of a controlling block of
shares who buys without management approval and can delay takeover by up to ﬁve
years.

The market for corporate control in the 19905 was governed by a markedly
different set of rules than that of the 19705 and 19805. By 1991, 87% of NYSE and
AMEX listed ﬁrms were covered by at least one form of an anti-takeover measure; 35%
were covered by self-imposed poison pills, 80% by business combination laws, and 24%
by control share laws. This is compared to 1985, where less than 7% of NYSE and
AMEX listed ﬁrms were covered by at least one form of an anti-takeover measure; 1%
were covered by self-imposed poison pills, 5% by business combination laws, and 5% by
control share laws.6

With the introduction of poison pills, new corporate anti-takeover amendments,
control share laws, and business combination laws, absolute deterrence by management
Opposed to a takeover was a possibility. These anti-takeover measures would seem to

protect managers' interest, and in particular, entrench them. The fact that hostile

 

5 See Borokhovich, Brunarski, & Parrino (1997)
6 See Comment and Schwert (1995) for their methodology in calculating these statistics.

6

takeovers are harder to complete today would suggest that management would not need
higher ownership to block an unwanted deal, all they would need is some form of anti-

takeover amendment or protection under state law.

1.1.3 Summary of Essay

The majority of empirical research on the subject of corporate control uses
samples of mergers that took place during the 19705 and 19805. Since the environment
for corporate control in the 19905 has been markedly different than the 19705 and 19805,
a re-examination of the previous acquisition literature would seem insightful. This essay
examines if the previous empirical results hold in a different time period, within a
different corporate control environment.

To investigate these issues, this essay examines the impact of management
incentives on the likelihood of a ﬁrm being acquired in the corporate control market of
1990 to 1997. I examine a sample of 342 ﬁrms that were successfully acquired, and
compare them to a matched sample of 342 ﬁrms that were not acquired. I examine
whether ownership characteristics, ﬁnancial characteristics, and management
characteristics of the frrrn are systematically different between acquired ﬁrms and
matched ﬁrms that were not acquired. I classify managers and directors into the
categories of executives, directors who are employed by the ﬁrm, and directors who are
not employed by the ﬁrm. I also classify blockholders as being afﬁliated with
management or not and as having representation on the board or not.

In my empirical analysis, I ﬁnd that certain ﬁnancial characteristics of a ﬁrm are

signiﬁcantly related to the probability that a ﬁrm is acquired, in particular the ﬁrm's

liquidity and sales growth. I ﬁnd that ﬁrms with lower liquidity (lower relative excess
liquidity, an indicator of sound management) characterize ﬁrms that were acquired. This
is evidence against a discipline hypothesis, where ﬁrms that are mismanaged are more
likely to be acquired. Financial characteristics that I ﬁnd not signiﬁcantly related to the
probability that a ﬁrm is acquired include return on equity, leverage, market-to-book
ratio, and price-to-earnings ratio.

I also ﬁnd that ﬁrms with a higher percent of outside directors and higher
ownership by blockholders who are represented on the board by outside directors are
relatively more likely to be acquired. This provides support to the theory that outside
directors and blockholders represented on the board by outside directors act as monitors
of management and are representing all shareholders not afﬁliated with management.
This empirical evidence provides support for what I later term the monitoring hypothesis.
1 ﬁnd no evidence that blockholders who are not represented on the board either facilitate
or deter control changes.

I ﬁnd that ﬁrms with higher managerial ownership and a higher percentage of
insiders on the board are relatively less likely to be acquired. Also, the CEOs of ﬁrms
that were acquired had on average 2 years less experience than matched counterparts.
Greater managerial control reduces the likelihood of a change in control. This empirical
evidence provides strong support for what I later term the entrenchment hypothesis.

Overall, I ﬁnd that target ﬁrms of the 19905 can be characterized as ﬁrms with
lower growth, lower managerial ownership, higher ownership by outside directors, and
higher ownership by blockholders afﬁliated with an outside director then their non-target

counterparts. The fact that non-target ﬁrms have higher managerial ownership, higher

percent of inside directors, and longer tenure of the CEO is an interesting result. Given

that the 19905 were governed by the adoption of poison pills, new corporate anti-takeover
amendments, control share laws, and business cbmbination laws, it would seem that
managers would not need ownership to deter an unwanted deal. However, the results
show that even with these outside methods of deterrent, non-target ﬁrms’ managers still
ﬁnd it necessary to have larger relative ownership to help in deterring unwanted changes
of control.

Another interesting result is that target ﬁrms have a signiﬁcantly larger percent of
blockholders unafﬁliated with management. However, all of signiﬁcance is in the
blockholders that are represented on the board. These blockholders represented on the
board could be classiﬁed as active or monitoring blockholders. These blockholders are
going to monitor and act in their best interest and those of other shareholders. The
evidence suggests that large blockholders, unafﬁliated with management and represented
on the board, play an important third-party role in facilitating takeovers.7

The rest of the essay is organized as follows. In section 1.2 I brieﬂy describe the
theories of the role of managerial ownership and of certain ﬁnancial characteristics on the
probability of being acquired. In section 1.3 I describe the construction of my data set.
In section 1.4 I present my empirical ﬁndings. Section 1.5 discusses the results and

concludes.

1.2. Hypothesized Effects of Governance Structures and Financial Characteristics

 

7 Shivdasani (1993) ﬁnds similar results in his study of hostile targets vs. non-targets
9

 

 

I examine whether variables related to ownership structure, management
characteristics, and ﬁnancial characteristics are systematically different between acquired
ﬁrms and ﬁrms that are not acquired. This leads to the testing of several governance

based and ﬁnancial based hypotheses listed in Table 1.2.

1.2.1 Governance Based Hypotheses

Increased managerial ownership gives additional control to managers, which may
be used to deter acquisitions. Managers must weight the beneﬁts from the wealth gain
from acquisition with the lost beneﬁts of control. There are several hypotheses that might
predict the sign of the coefﬁcients on the ownership and governance variables that I
examine. I will outline these hypotheses and label them using the terminology of
Hadlock, Houston, and Ryngaert (1999).

One possibility is that managers who have signiﬁcant ownership positions gain
ﬁnancially by accepting an attractive takeover premium and thus these are the ﬁrms that
are most inclined to receive and accept acquisition offers. This is the situation where the
beneﬁts from the wealth gain from acquisition outweigh the beneﬁts of control. I will
call this possibility the ﬁnancial incentive hypothesis.

A second hypothesis possibility is that managers do not like to lose control of
their ﬁrms despite the potentially large ﬁnancial gains form doing so. If managers with
signiﬁcant ownership positions are the ones with the highest preference for control or the
greatest ability to maintain control by resisting the demands of other shareholders to sell

out to an acquirer, then we would expect their ﬁrms to be the least likely to be acquired.

10

This is the situation where the beneﬁts from the wealth gain from acquisition do not
outweigh the beneﬁts of control. I will call this possibility the entrenchment hypothesis.

A third possibility is that ﬁrms with signiﬁcant levels of managerial ownership,
outside dominated boards, and/or large block ownership unafﬁliated with management,
should be particularly efﬁcient because of their strong incentives to maximize value. If a
motive of corporate mergers is to replace poor management teams, then ﬁrms with higher
levels of management ownership, outside dominated boards, and/or higher levels of
unafﬁliated block ownership should be relatively less likely to be acquired. I call this
possibility the discipline hypothesis.

The discipline hypothesis argues that managerial ownership is related to
managerial efﬁciency and that inefficient management teams are likely to be targets.
Firms with signiﬁcant levels of management ownership or outside dominated boards
should be particularly efﬁcient because of their strong incentive to maximize value. This
hypothesis also suggests that other measures of managerial performance, for example
return on equity, should also increase the probability of being acquired. Inefficient
management of these ﬁrms, measured by any appropriate ﬁnancial metric, should create
opportunities for bidders to acquire these ﬁrms and increase their value.

Previous research suggests that outside directors play an important role in
monitoring top managers (e.g. Weisbach (1988)). Following Shivdasani (1993), I
examine the role that the composition of the board and the ownership interest of outside
board members have in predicting takeover likelihood. A hypothesis, which I termed the
monitoring hypothesis, asserts that ﬁrms with a larger number of outside directors or

ﬁrms where outside directors hold a substantial fraction of the ﬁrm's equity will be the

11

ﬁrms that are the most likely to make value maximizing decisions in corporate control
transactions. Higher ownership by outside directors provides them with the incentive to
look after shareholder interests and the ability to threaten removal of inside directors via a
proxy contest. A larger percentage of the board seats provides outside directors with a
greater ability to look after the interests of shareholders. I also examine the role of large
blockholders who are unafﬁliated with management. Higher ownership by unaffiliated
blockholders gives unafﬁliated blockholders more incentive to monitor management
performance closely. Under the monitoring hypothesis I expect that higher ownership by

unafﬁliated blockholders will increase the likelihood of acquisition.

1. 2. 2 Financial Hypotheses

Following Palepu (1986), included in my model are ﬁnancial characteristics based
on several hypotheses suggested from previous research, which are likely indicators of
acquisition likelihood. The following ﬁnancially based hypotheses are tested: the
growth-resource mismatch hypothesis, the size hypothesis, the book-to-market
hypothesis, and the price-earnings hypothesis.

Firms with a mismatch between their growth and the ﬁnancial resources at their
disposal may be particularly attractive targets. A low-growth, resource rich ﬁrm may be
a likely target since it has ﬁnancial resources at its disposal, yet it doesn't have many
opportunities available to it to utilize the ﬁnancial resources. A bidding ﬁrm would ﬁnd
this attractive, since it might be able to better utilize the ﬁnancial resources of target. A
high-growth, resource poor ﬁrm may also be a likely target due to the attractiveness of its

growth opportunities. This type of ﬁrm would seem to be in need of ﬁnancial resources

12

to ﬁrlly utilize its growth potential. A ﬁrm with excess resources available would ﬁnd
this type of ﬁrm an attractive target. A dummy variable based on growth-resource
mismatch following the treatment of Palepu (1986) is deﬁned in the Appendix 3. The
growth-resource mismatch dummy variable utilizes three individual variables: sales
growth, liquidity and leverage. These variables have also been used individually in
previous papers as indicators of a target ﬁrm.

Another pOpular ﬁnancial based hypothesis is the size hypothesis. This
hypothesis asserts that the likelihood of acquisition decreases with the size of the ﬁrm.
This hypothesis is based on the possibility that transaction costs increase with size of the
target ﬁrm. A larger ﬁrm size may reduce takeover likelihood because ﬁnancing costs to
acquire a large ﬁrm are prohibitive and the cost of integrating a ﬁrm increases with target
size. Net book value of assets is included in the model to test the size hypothesis. Net
book value of assets is included instead of any form of ﬁrm value that utilizes market
value of common equity, since market value of common equity is one of the criteria for
choosing the matched ﬁrms in the sample.

The market-to-book hypothesis asserts that ﬁrms whose market values are low
compared to their book values are likely acquisition targets. This hypothesis is based on
the notion that ﬁrms with low market-to-book value ratios are relatively good values.
However, market-to-book may also proxy for managerial quality. This hypothesis is
included to allow comparisons to previous research. The market-to-book value ratio of
common equity is included in the model to test this hypothesis.

A ﬁnal ﬁnancial hypothesis, the price-earnings hypothesis, states that ﬁrms with

low P/E ratios are likely acquisition targets. This hypothesis is based on the theory that

13

bidders with high P/E ratios seek to acquire low P/E ﬁrms to realize an instantaneous
capital gain because of the belief that the stock market values the earnings of the
combination at the higher P/E ratio of the acquirer. The rationale behind this hypothesis
is difﬁcult to support; however it has been used in previous literature and therefore is
included for completeness.

A ﬁnal hypothesis is that mergers occur for reasons completely independent of
management and ﬁnancial characteristics. Mergers take place for reasons such as to
capture synergistic gains or the desire of a bidder to enter a new line of business. This
hypothesis would imply that all of the managerial characteristics examined should be

insigniﬁcant. I will call this possibility the irrelevance hypothesis.

1.3 Sample Selection and Empirical Design
1.3.1 Population Selection

My study focuses on mergers that took place from 1990 - 1997. Following studies
that concentrated on mergers of the 1970's and 1980's, I made decisions on the population
on which my sample was derived. (1) To start with, ﬁrms had to have data available in
both CRSP and COMPUSTAT databases. (2) Firms had to be listed on either the New
York Stock Exchange or American Stock Exchanges. Certain ﬁrms were then excluded
from the population. (3) Firms not listing the United States as its home country were
excluded from the population. (4) American Depository Receipts (ADRs) and (5)

Limited partnerships were also excluded. (6) Due to the regulatory and speciﬁc nature of

 

8 My restriction to NYSE and AMEX is similar to the treatment in Palepu (1985),
Mikkelson and Partch (1989) and Billett (1996). Song & Walkling (1993) use any ﬁrm
available in Value Line, which is mostly NYSE and AMEX ﬁrms, due to the fact that
Value Line tracks the largest ﬁrms.

l4

particular industries, some entire industries were excluded. In particular, ﬁnancial,
banking, insurance, real estate and utilities ﬁrms were excluded. 9 Included in this
deletion were industries classiﬁed in 8105 6000 to 6999 (ﬁnance, insurance, real estate)
and SIC's 4900 - 4999 (public utilities).

Table 1.3 outlines the total number of ﬁrms listed by year that were left after the
above characteristics were controlled for. From 1990 to 1997 there was an average of
1,814 ﬁrms listed on the NYSE and AMEX each year. A list of merged ﬁrms by year was
prepared from the delistments from the stock exchanges due to mergers and acquisitions,
obtained from the CRSP and COMPUSTAT databases, and manually conﬁrmed by Dow
Jones Interactive searching The Wall Street Journal, Dow Jones News Service, Dow
Jones Newswires, and Press Release Wires. During the period, an average of 47 ﬁrms
per year were delisted from the NYSE and AMEX due to merger. This represents 2.58%
of the listed ﬁrms annually being delisted due to merger. Over the period, there appears
to be an increasing number of mergers taking place starting in 1991 and increasing each

year up until 1997.

1.3.2 Sample Selection

From the above population of ﬁrms a sample of 342 ﬁrms that were acquired
during the period 1990-1997 and a matched sample of 342 ﬁrms that were not acquired
are used for the estimation of an acquisition model. For a merged ﬁrm to be included in

the sample I required that an announcement date and news story for the merger

 

9 Similar Palepu (1985) who only includes manufacturing and mining ﬁrms, Mikkelson
and Partch (1989) who exclude banking, insurance, ﬁnancial, utilities, transportation,
Song and Walkling (1993) who excluded ﬁnancial, banking, and utilities, Billett (1996)
who excludes regulated and ﬁnancial industries.

15

transaction be identiﬁed. Dow Jones Interactive (searching The Wall Street Journal, Dow
Jones News Service, Dow Jones Newswires, and Press Release Wires) was used to
identify the speciﬁc announcement and rumor date of the merger. Additionally, a proxy
statement for the ﬁrm must be available previous to the announcement date or rumor date
of the merger. Table 1.3 identiﬁes 374 ﬁrms delisted due to merger during the period
1990 - 1997. Incomplete or not ﬁled proxy statements reduce the number of ﬁrms in the
sample by 32 to a total of 342 ﬁrms.

The 342 matched ﬁrms were chosen in the following manner”). The matching ﬁrm
was chosen based on three criteria; the target's industry, degree of focus, and market
value of equity preceding the armouncement date of the merger. To be considered as a
matching ﬁrm, a ﬁrm could not have been considered an acquisition target for ﬁve
calendar years centered on the announcement date. Canceled acquisitions were identiﬁed
from the Mergerstat Review Cancellation Rosters for 1988 - 1998.

For the industry criterion, I started with the target ﬁrms four-digit SIC code. If
there were at least ten non-target ﬁrms in the same four-digit SIC, the four-digit SIC
industry was used to identify the matching ﬁrm. If less then ten non-target ﬁrms were
available in the four-digit SIC, then the three—digit SIC industry code was used if at least
ten ﬁrms were available in the three-digit SIC. If less then ten non-target ﬁrms were
available in the three-digit SIC classiﬁcation, then the two-digit SIC is used no matter

how many ﬁrms available to identify the matching ﬁrm.

 

'0 Song & Walkling (1993) select a matched ﬁrm based on Value Line industry
classiﬁcation & another matched ﬁrm based on random sampling. Palepu (1985) selects
a matched ﬁrm by random sampling. Hadlock, Houston, Ryngaert (1999) selects a
matched bank based on size and location. Mikkelson & Partch (1989) and Billett (1996)
do not select a matched ﬁrm. They both follow a sample of ﬁrms through time and ﬁnd
which ones were targets.

16

Once the industry classiﬁcation was determined, ﬁrm focus was used to screen for
matches. FASB No. 14 and SEC Regulation S-K require ﬁrms to report segment
information for ﬁscal years ending after December 15, 1977. Firms must report audited
footnote information for segments whose sales, assets, or proﬁts exceed 10% of
consolidated totals. The COMPUSTAT database reports the number of segments
reported for each ﬁrm. A ﬁrm with only one segment is commonly referred to as a
focused ﬁrm or single segment ﬁrm. A ﬁrm with more than one segment is referred to as
a diversiﬁed ﬁrm or multi segment ﬁrm. Berger and Ofek (1995) calculate a value 1055
from diversiﬁcation of 13% to 15% on average during 1986-1991. They ﬁnd that over-
investrnent and cross-subsidization contribute to the value 1055 for diversiﬁed ﬁrms.
Berger and Ofek (1996) ﬁnd that ﬁrms with greater value losses, due to diversiﬁcation,
are more likely to be taken over. This so called 'diversiﬁcation discount' and the possible
undesirability of multiple segments to acquiring ﬁrms were deciding factors in using ﬁrm
focus as a matching criterion. To be a matched ﬁrm, a ﬁrm had to have the same
classiﬁcation of multi-segment or single segment as the target ﬁrm.

The market value of common equity was calculated for the end of the calendar year
proceeding the announcement date of the merger transaction. The closest match to the
market value of common equity of the target ﬁrm, in the correct SIC classiﬁcation and
with the same focus, was chosen as the matched ﬁrm. For each target I refer to the

alternative target as a match for the target.

1.3. 3 Variable Construction

17

Previous studies concerning managerial ownership in relation to the market for
corporate control utilize summary data for managerial ownership.ll For instance, these
studies utilize a summary number that includes ownership for all directors and executive
ofﬁcers of the ﬁrm. These studies do not break out ownership of directors from executive
ofﬁcers and also do not break out the different types of directors. This summary data was
used in previous studies due to the time involved in collecting more speciﬁc data. I
follow the more time consuming procedures outlined by Shivdasani (1993) and Hadlock,
Houston and Ryngaert (1999) to classify directors (described in 3.2.1).

For the set of all target and matched ﬁrms, I searched the Disclosure and Lexus-
Nexis databases for the proxy statement for the year ending prior to the date the merger
was announced. If a proxy was not found in the above database searches the Securities
and Exchange Commissions Q-ﬁle Data Corporation database on microﬁche was used. I
collected detailed data from the proxy statement concerning management characteristics,
management compensation, ownership structure, and board structure of the ﬁrm. From
the proxy statements I attempt to deﬁne relevant variables in a similar manner to previous

studies.

1.3.2.1 Ofﬁcers and Directors

For each ﬁrm I classify all directors as either inside, outside or gray directors. Inside
directors are individuals who were full time employees of the ﬁrm or one of its
subsidiaries along with their immediate family members. Gray directors are individuals

who were retired employees of the ﬁrm, non-immediate family members of employees of

 

" Palepu, Song & Walking, and Mikkelson & Partch all utilize summarized managerial
ownership.

18

the ﬁrm and individuals where the proxy statement indicated an ongoing business
relationship between the individual and the ﬁrm. Most people classiﬁed as gray outsiders
were outside legal counsel or retired employees of the ﬁrm. All other directors are
deﬁned to be outside directors. For each individual I record their ownership of common
stock and options exercisable within 60 days. The sum of common stock and options
exercisable within 60 days is commonly referred to as beneﬁcial ownership. In addition
to the directors' beneﬁcial ownership, proxy statements list a summary number for all
executive ofﬁcers and directors as a group. I used this number to calculate the beneﬁcial
ownership of all executive ofﬁcers. Beneﬁcial ownership of all executive ofﬁcers is
calculated by subtracting the holdings of all directors from the summary number of all
ofﬁcers and directors holdings. I then created a variable called inside directors and
executives that is the beneﬁcial ownership of inside directors and executive ofﬁcers

together as a group.

1.3.2.2 Block Ownership

In addition to the officers and directors holdings, I deﬁne variables related to
blockholdings. Beginning in 1978, ﬁrms were required to disclose in proxy statements
the security ownership of holders of 5.0% or more of a class of voting securities.
Following Shivdasani (1993) and Hadlock, Houston, Ryngaert (1999), I classify
blockholders into two groups: aﬂiliated blockholders and unafﬁliated blockholders. The
aﬁiliated blockholder variable is the percentage of stock holdings by: individuals related
to inside directors, ﬁrrn’s with an ongoing business relationship with the ﬁrm, charitable

institutions and trusts for which an insider director of the ﬁrm served as an officer or

19

trustee, ESOPs, proﬁt sharing plans, and ﬁrm pension plans. These are blockholders
whom would be expected to vote in concert with the inside director they are afﬁliated
with. In some cases, they are blocks of shares inside directors actually have the authority
to vote. The block holdings of all other blockholders are included in the variable
unaﬂiliated blockholders. Unafﬁliated blockholders are further classiﬁed as to whether
outside directors represent them on the board or not. Two additional variables,
unafﬁliated blockholders represented on the board of directors and unaﬁ‘iliated
blockholders not represented on the board of directors are identiﬁed. Unaﬂiliated
blockholders represented on the board of directors is the sum of unafﬁliated
blockholdings that has an ongoing relationship with an outside director, and unafﬁliated
blockholders not represented on the board of director is the sum of unafﬁliated

blockholdings that don't.

1.3.2.3 The Top Executive

Following Hadlock, Houston, Ryngaert (1999), for each ﬁrm I identiﬁed an
individual as the top executive for the ﬁrm as of the proxy date. The top executive is
deﬁned to be the ﬁrrn's CEO if the ﬁrm assigns that title to an individual on the board of
directors. If the ﬁrm does not have a designated CEO then I deﬁne the top executive to
be the individual with the highest pay amongst the set of individuals on the board who
had the title of Chairman or President during the previous ﬁscal year. If one of these
individuals had retired by the proxy date, I deﬁne the top executive to be the individual
with the highest ownership amongst the set of individuals on the board holding the title

Chairman or President as of the proxy date.

20

For the top executive I collect data on his/her beneﬁcial ownership in the ﬁrm, his
age, his tenure as a director, and his cash compensation timing the year proceeding the
proxy date. I adjust all cash compensation to 1996 dollars for comparison purposes. I
also create an indicator variable for a ﬁrm equal to one if only one individual holds the
titles CEO, President, and Chairman of the Board. This variable attempts to identify
powerful CEOs that could have a special ability to resist takeover bids, thereby making

them less attractive targets. ‘2

1.3.2.4 Financial Information

Financial information was collected from the COMPUSTAT database. I deﬁne
relevant ﬁnancial variables in a similar manner to previous studies. In particular, I follow
the procedures outlined by Palepu (1985), Mikkelson and Partch (1990), Billett (1996)
and Song and Walkling (1993). All ﬁnancial variables were collected for the ﬁscal year
ending prior to the announcement date of the merger for each target. The matched ﬁrrn's
data was also collected for the ﬁscal year ending prior to the announcement date of the
target for which it was a match.

Financial variables such as sales growth, liquidity, leverage, and return on equity
are averaged for three ﬁscal years prior to the announcement date of the target's merger.
If three years of data is not available, then two years are averaged. If two years of data
are not available, then one year is utilized. Other ﬁnancial variables, such as total assets,
market-to-book ratio, price-to-earnings ratio, net plant, property and equipment to total

assets, and free cash ﬂow to total assets are calculated for the ﬁscal year ending prior to

 

12 Shivdasani (1993) ﬁnds this to be signiﬁcant in comparing hostile targets to matched
non-targets.

21

the announcement date of the target's merger. Market value of common equity is
calculated at the end of the calendar year preceding the announcement date. All ﬁnancial

variables utilized in this essay are described in detail in the Appendix 3.

1.4. Empirical Results

Section 1.4.1 reports descriptive univariate statistics on ownership, ﬁnancial
characteristics and other variables. The multivariate analysis of acquisition likelihood is
discussed in section 1.4.2. In section 1.4.3 the results of the hypotheses tests are
discussed. To check the robustness of the results, several speciﬁcation checks are

conducted in 'section 1.4.4.

I . 4. I Univariate Results

Table 1.4 presents summary statistics for the sample of 342 target ﬁrms and the
matched sample of 342 non—target ﬁrms. For each variable I test for a difference in the
means by a t-test with unequal means. Additionally, a Wilcoxon-Z test is computed for
the difference in medians. The univariate results are divided into four sections: ﬁrm
ﬁnancial characteristics, board of director's characteristics, ownership characteristics and

CEO characteristics.

1 .4. l .1 Financial Characteristics
The targets and matched ﬁrms are very similar in size. This is due to the fact that the
matching ﬁrm was chosen based on the criterion of market value of equity. Out of total

assets, market value of equity and total ﬁrm value, only the difference in medians for

22

total assets is statistically signiﬁcant. Song and Walkling (1993) compare their target
ﬁrms’ average market values with those of Mikkelson and Partch (1989). They estimate
that Mikkelson and Partch's targets average market value of equity is $787 million (in
1983 dollars). Song and Walkling's targets average market value of equity is $731
million (in 1983 dollars). In 1997 dollars the average size of my target ﬁrms is $927
million. Adjusting this number to 1983 dollars yields an average size of market value of
equity for my targets of $590 million. Therefore, it appears that the targets in my sample
are slightly smaller in mean size than the targets of Song and Walkling and Mikkelson
and Partch. The median size of targets is also comparatively smaller in my sample then
those of the previous studies.

As for the other ﬁnancial variables, only sales growth and liquidity are statistically
different. Three-year average sales growth for matched ﬁrms is 27%, while only 15% for
target ﬁrms. Three-year average liquidity is higher for matched ﬁrms than for target
ﬁrms. Three-year average return on equity is signiﬁcantly different in the means (p-value
= .0926). Target ﬁrms averaged 5.7% return on equity over a three-year period, while
non-target average a negative 6.5%. However, there is no statistical difference in the
medians. Leverage, the growth-resource dummy, market-to-book ratio, price-to-earnings
ratio, net plant property and equipment to total assets and free cash ﬂow to total assets are
all not signiﬁcantly different.

Another indicator variable, the existence of two or more classes of voting stock
outstanding, was statistically different at the 1% level. Only 9% of target ﬁrms had two
or more classes of voting stock, while 18% of matched ﬁrms had two or more classes of

voting stock outstanding. Twice as many matched ﬁrms had multiple voting classes as

23

that of the targets. The existence of two classes of voting stock usually concentrates the
control of the ﬁrm with a smaller group of shareholders. This concentration of voting
power if held by managers could make it easier for managers to maintain control of a

ﬁrm.

1.4.1.2 Board of Directors Characteristics

Although both targets and non-targets average approximately eight and a half board
members, the composition of the board of directors for targets and non-targets are slightly
different. Targets have on average 29% insiders / 66% outsiders / 5% gray outsiders,
while non-targets have 35% insiders / 62% outsiders / 3% gray outsiders on average. All
three types’ means and medians are statistically different at the 5% level. Mikkelson and
Partch (1989) include a dummy variable for staggered voting for board of directors. Both
targets and non-targets have almost identical numbers of ﬁrms with staggered voting.
Staggered voting does not appear to deter outsiders' interest in takeover activity, nor does

it identify managers who are likely to resist takeover offers.

1 .4.1.3 Ownership Characteristics

Table 1.5 displays the univariate results for ownership characteristics. All variables
related to ownership by 'insiders', such as holdings by inside directors, executives, the top
executive, the tOp three executives or the top ﬁve salaried employees, are higher in non-
target ﬁrms as compared with targets ﬁrms. All of these insider ownership variables are
statistically different for both the means and medians. Inside directors and executives as

a group have beneﬁcial ownership, on average, of 22% in matched ﬁrms and only 16% in

24

target ﬁrms. Afﬁliated blockholders for non-targets also have higher ownership than
targets, but this difference is not statistically signiﬁcant. In aggregate, all inside directors,
executives and afﬁliated blockholders own 21% of non-targets and only 15% of targets,
on average. This aggregate variable is statistically different in the means and medians.
Ownership by outside directors, gray outside directors and unafﬁliated
blockholders is higher in target ﬁrms than in non-target ﬁrms. Gray outsiders and
unafﬁliated blockholders are statistically different, with outside directors not statistically
different. Aggregated, outside directors, gray outside directors and unaﬂrliated
blockholders own 22% of non-targets and 31% of targets, on average. This aggregate

variable is statistically different in the means and medians.

1.4.1.4 CEO Characteristics

The age of the top executive is not statistically different between targets and non-
targets. However, the tenure of the top executive, as measured by the years served on the
board of directors does differ signiﬁcantly. The mean difference is more than two years.
Non-target CEOs have an average tenure of 13 years while targets CEOs have an average
tenure of 11 years. This evidence suggests entrenchment of long tenure CEOs. The
indicator for the one-person management team and the cash compensation for the top

executive are not statistically different between the target fums and the matching sample.

1. 4.2 Multivariate Results
Table 1.6 reports the results from logit regressions where the dependent variable is set

to unity if the ﬁrm is a target and is zero otherwise. Five different logit models are

25

presented in Table 1.6. Models 1 and 5 include eight ﬁnancial independent variables in
addition to a constant term. The eight variables correspond to the hypotheses discussed
in section 1.2. Models 2 through 5 include independent variables that correspond to
managerial and block ownership of equity to test the various govemance-based
hypotheses. The parameter estimates of the logit acquisition models and the associated p-
values are presented in Table 1.6.

Model 1 is a replication of Palepu's (1986) acquisition model.13 The growth-resource
dummy variable was positive and signiﬁcant in his sample. In my sample, the growth-
resource dummy is positive, but not signiﬁcant. Sales growth and liquidity are both
negative and signiﬁcant, while leverage is negative and insigniﬁcant. The signs indicate
that low growth and low liquidity characterize target ﬁrms in my sample. These
coefﬁcients all have the same signs as as those reported by Palepu (1986). However,
Palepu (1986) ﬁnds that only sales growth and leverage are signiﬁcant in his sample.
Low growth and low leverage characterize Palepu's (1986) sample of target ﬁrms. Song
and Walkling (1993) ﬁnd leverage, liquidity and sales grth all are insigniﬁcant.
Consistent with Palepu’s (1986) and Song and Walkling’s (1993) ﬁndings, return on
equity, market-to-book value and price-to-earnings ratio are all statistically insigniﬁcant
in my model. Palepu and Song and Walkling both found size to be a signiﬁcant factor,
with target ﬁrms being smaller in size. Size is not signiﬁcant in my acquisition model.
However, this is not surprising due to the fact that market value of equity was used as a

matching criterion. Therefore, size is mostly controlled for.

 

'3 Palepu also included an industry dummy variable to control for industry speciﬁc
takeover activity. This is not used in this model due to the fact that non-target ﬁrms
where matched on the basis of industry.

26

In models 2 through 4 I include only ownership variables based on the methodology
of Shivdasani (1993) and Hadlock, Houston, Ryngaert (1999). Both break out ownership
based on classiﬁcation of the board members into inside, outside or gray, and
classiﬁcation of blockholders as afﬁliated or unafﬁliated.14 In model 2 inside directors
and executives, outside directors, gray directors, afﬁliated block holders and unafﬁliated
blockholders share ownership is included. The coefﬁcient on ownership by inside
directors and executives is negative and highly signiﬁcant (p-value = .012). The
coefﬁcients on ownership by outside directors and gray outside directors are both
positive, with the coefﬁcient on gray outside directors being signiﬁcant at the 10% level.
The coefﬁcient on holdings by unafﬁliated blockholders is positive and highly
signiﬁcant, while the corresponding coefﬁcient for affiliated blockholders is negative but
insigniﬁcant.

In model 3 I found it appropriate to sum both types of outside directors, both true
outside directors and gray outside directors. Most directors classiﬁed as gray in my
sample were former employees or lawyers who provided legal counsel to the ﬁrm. In
model 2 gray and outside directors have the same coefﬁcient sign, indicating that they
both act as monitors to management. Therefore, in model 3 the two holdings were
aggregated into one all outside directors variable. When aggregated, the all outside
directors variable is positive, but not quite signiﬁcant at conventional levels (p-value of

.131).

 

‘4 Mikkelson & Partch (1989) use total votes controlled by all executive ofﬁcers and
directors and also votes controlled by holders of the top three ofﬁces. Song & Walkling
(1993) use total ownership by executive officers and directors as a group. Neither of
these papers classiﬁes directors as inside/outside. Mikkelson & Partch also use a slightly
different classiﬁcation of blockholders.

27

Inside directors and executives and affiliated blockholders both have negative
coefficients. If the interests of afﬁliated blockholders are aligned with inside directors
and executives, the two groups could provide a united front to defend against unwanted
acquisition attempts. Thus, a case could be made for a speciﬁcation with an aggregate
inside ownership variable that sums afﬁliated blockholdings with inside directors and
executives. A similar case could be made for the outside ownership variables.
Unafﬁliated blockholders and all outside directors both have positive coefﬁcients. In
model 4, these two aggregate measures of inside and outside ownership are used in a logit
model. Both variables are highly signiﬁcant, with the coefﬁcients on the inside
ownership variables supporting the entrenchment hypothesis and the coefﬁcients on the
outside ownership variables providing evidence of outside monitoring.

In the ﬁnal model, model 5, both the ownership variables from model 3 and the
ﬁnancial variables from model 1 are included. Ownership by inside directors and
executives remains negative and highly signiﬁcant at the 5% level, while ownership by
all outside directors is positive and becomes signiﬁcant at the 10% level. The coefﬁcient
on ownership by unafﬁliated blockholders remains positive and highly signiﬁcant at the
1% level, while the corresponding coefﬁcient for afﬁliated blockholders remains negative
and insigniﬁcant. As for the ﬁnancial variables, sales growth and liquidity remain
negative and signiﬁcant, while most other variables remain insigniﬁcant. Return on

equity becomes signiﬁcant at the 10% level.

1.4.3 Hypotheses Tested

28

The following hypothesis tests are summarized in Table 1.7 and utilize all univariate

results and multivariate results from model 5, Table 1.6.

1.4.3.1 Governance-based Hypotheses
In my empirical analysis I ﬁnd that ﬁrms with higher management ownership are
relatively less likely to be acquired. This result appears inconsistent with the irrelevance
hypothesis and the ﬁnancial incentive hypothesis, but is highly consistent with both the
discipline hypothesis and the entrenchment hypothesis. To test between these two
hypotheses note that the results suggest that ﬁrms with high numbers of outside directors,
high outside ownership, and high unafﬁliated blockholdings are relatively more likely to
be acquired. Since these ﬁrms are the ones we expect to be heavily monitored and thus
least in need of discipline, these results cast doubt on the discipline hypothesis.
Additionally, ﬁnancial indicators of management quality do not suggest that target ﬁrms
are poor performers. Target ﬁrms have lower liquidity, indicating that these ﬁrms have
sound management. The return on equity and leverage variables indicates that targets are
not managed worse than matched ﬁrms. Taken together, these additional ﬁndings cast
doubt on the discipline hypothesis.
My results appear to offer the strongest support for the entrenchment hypothesis.
This suggests that managers value the beneﬁts of control more than the beneﬁts of
increased wealth due to acquisition. Supporting evidence includes the fact that the CEOs
of non-target ﬁrms average two years more tenure on the board of directors than CEOs of
target ﬁrms. Additionally, target ﬁrms have a larger percent of inside board members

than matched ﬁrms.

29

There also appears to be empirical evidence supporting the monitoring
hypothesis. Target ﬁrms have a signiﬁcantly larger number of outside directors sitting on
their board of directors then non-targets. Additionally, in the logit model, ownership by
all outside directors as a group is positively related to the probability that a ﬁrm is a
target, and is signiﬁcant. Large unafﬁliated blockholdings is also positively related to the
probability that a ﬁrm is a target, and is highly signiﬁcant. Therefore, it appears that
outside monitoring is taking place and value maximizing decision are being made in
corporate control transactions involving higher outsider ownership and increased outside

presence.

1.4.3.2 Financial Hypotheses

The growth—resource mismatch hypothesis is not supported empirically by either the
univariate analysis or the logit model results from my sample of mergers from the 1990's.
The growth-resource coefﬁcient is positive in the logit model, consistent with the
hypothesis, but is not signiﬁcant. Palepu (1986) found empirical support for the growth-
resource mismatch hypothesis in his sample of 1970's and 1980's mergers.

Size is not signiﬁcant in either the univariate analysis or the logit model. This is
not surprising due to the fact the matched ﬁrms were chosen on the criterion of market
value of equity in my sample. The market-to-book and price-earnings hypotheses have
no empirical support from my model, Palepu's (1986) model or Song and Walkling's

(1993) model.

1.4.4 Robustness

30

This section discusses several different speciﬁcation checks. Votes controlled are
considered rather than shares controlled, additional variables from previous literature are

considered, and model robustness is checked.

1.4.4.1 Voting Rights

There are two ways to measure ownership; using beneﬁcial ownership or using
percentage of votes controlled. In my univariate analysis and logit acquisition model I
used beneﬁcial ownership where all outstanding classes of stock are treated equally. This
may be appropriate for arguments based on ﬁnancial considerations, however it may not
be completely satisfactory when analyzing control considerations. In previous research
one or both methods are utilized. Therefore, as a robustness check, I replace beneﬁcial
ownership with percentage of votes controlled in my logit acquisition model. The results
are reported in Table 1.8. The statistical signiﬁcance of each variable remains
unchanged. The only substantial difference is that the magnitude of the coefﬁcients on

all the ownership variables is slightly lower.

1.4.4.2 Additional Control Variables

Shivdasani (1993) uses management characteristics in his logit model of hostile
targets vs. matched non-targets. He considers the age of the CEO, the CEOs tenure on
the board of directors and an indicator variable that identiﬁes strong CEOs who hold the
three titles CEO, President and Chairman of the Board. Shivdasani also tabulates the
average number of additional outside directorships held by each type of director, inside

and outside. I also tabulated the average number of outside directorships listed in the

31

proxy statement excluding charitable organizations and obvious non—proﬁt

5. Table 1.9, model 1 is my logit acquisition model with ﬁve additional

organizationsl
corporate governance variable included. The age of the top executive, the number of
years the top executive served as a director, the one person management team indicator,
and the average number of additional directorships held by outside directors are all
statistically insigniﬁcant. The average number of additional directorships held by inside
directors is negative, and signiﬁcant.

Mikkelson and Partch (1990) try to identify blockholders who have representation
on the board, and those who do not. I do not replicate their methodology. Instead, I
continue to use Shivdasani's method of classifying unafﬁliated blockholders. However, I
further classify unafﬁliated blockholders into two groups: those with board
representation, and those without board representation. By deﬁnition, the blockholders
with board representation are represented by outside directors. All affiliated blockholders
are by deﬁnition represented by inside directors. The results of the logit regression are
reported in model 2 of Table 1.8.

Unafﬁliated blockholdings represented on the board is positive and higth
signiﬁcant. The higher the ownership by unafﬁliated blockholders represented on the
board, the more likely the ﬁrm is to be a target. This is evidence that outside
blockholders and their corresponding outside directors together play a key role in
monitoring management and will make value-maximizing decision in corporate control

transactions. However, unafﬁliated blockholdings that are not represented on the board

do not appear to play a role in monitoring management.

 

'5 Shivdasani only includes directorships of companies included in the Fortune 1000 or
Fortune Service 500 lists.

32

Billett (1996) ﬁnds empirical evidence on the relationship between a ﬁrm's
takeover likelihood and the coinsurance potential of its debt in a sample of ﬁrms ﬁom
1979-1990“. Billett studies the risk of the ﬁrm's debt, measured by its credit rating, and
its leverage to measure the coinsurance potential of a ﬁrrn's debt. He ﬁnds that takeover
likelihood decreases as the coinsurance potential of its debt increases. I collected public
bond ratings from Standard and Poor's COMPUSTAT and Moody's credit ratings. I
followed the methodology outlined by Billett to create a coinsurance variable”. Using
his methodology, I was unable to ﬁnd a signiﬁcant relationship between coinsurance
potential and takeover likelihood. In particular, the interaction variable of Billett's non-
investrnent grade public debt dummy and leverage was in all cases insigniﬁcant. Several
differences in sample selection and methodology could explain these differences. Billett
(1996) uses all ﬁrms on the NYSE or AMEX that have sufﬁcient CRSP and
COMPUSTAT data available, where I only use targets and matched non-target ﬁrms.
Another possible explanation is the difference in time periods. The Billett (1996) sample

is from 1977-1986, where the sample I use is from 1990-1997.

1.5 Conclusion
The corporate control market in the US. has changed signiﬁcantly over the past 25
years. Most of the literature to date concerning these control markets has focused on the

merger activity of the 1970's and the 1980's. However, anecdotal evidence suggests that

 

16 Billett deﬁnes coinsurance as the wealth transfers from the bidder and target equity
holders to the target debt holders.

17 Billett creates a dummy variable for the existence of non-investment grade public
bonds. He interacts this dummy variable with the ﬁrrn's leverage. In his logit model he
ﬁnds this interaction variable to be negative and highly signiﬁcant. This implies that

33

the current market differs in important ways from this earlier period. In particular,
today's takeovers appear to be ﬁiendlier in nature and focused more on expanding
markets compared to the hostile transactions and downsizing that characterized the
1980's. Corporations and state governments took actions to slow the market for corporate
control of the 1980's. Corporations adopted many forms of anti-takeover amendments
and state governments passed forms of legislation aimed to protect ﬁrms form hostile
type takeovers. This essay attempts to ﬁrrther our knowledge about managerial,
ownership, and ﬁnancial characteristics that affect the likelihood of acquisition in the
'new' market for corporate control of the 1990’s.

I used a sample of 342 ﬁrms acquired from 1990 to 1997 and 342 non-acquired
ﬁrms matched by industry, focus, and size to examine these issues. Following the
methodology of Shivdasani (1993) and Hadlock, Houston, and Ryngaert (1999), I
classify managers and directors into the categories of executives, directors who are
employed by the ﬁrm, and directors who are not employed by the ﬁrm. I also classify
blockholders as being afﬁliated with management or not and as having representation on
the board or not.

With regards to ﬁnancial characteristics it appears that the mergers of the 19905
are different in some ways from those of the 19705 or 19805. Targets in the 19905 can be
characterized as having low grth and low liquidity, both of which are statistically
signiﬁcant in determining the likelihood of acquisition. The discipline hypothesis would
suggest that ﬁrms with excess liquidity, excess leverage, and/or relatively low return on

equity would be likely targets since these variables are likely to proxy for poor

 

ﬁrms with non-investment grade public debt are less likely to be targets, due to the
coinsurance effects.

34

management. However, it appears that targets are well managed without excess liquidity,
casting doubt on the discipline hypothesis.

I also examine the effect of blockholders and outside ownership on acquisition
likelihood. I classify large blockholders into afﬁliated blockholders and unafﬁliated
blockholders. I further classify unafﬁliated blockholders into unafﬁliated blockholders
who are represented on the board by an outsider and unafﬁliated blockholders who are
not. I ﬁnd that ﬁrms with higher levels of unafﬁliated blockholders represented on the
board are more likely to be acquired. However, I ﬁnd that unafﬁliated blockholders who
are not represented on the board do not have any effect on control changes. I also ﬁnd
evidence that ﬁrms with higher levels of outside director ownership are more likely to be
acquired. Additionally, target ﬁrms have a larger percentage of outside directors on their
board. This evidence all supports an outside monitoring hypothesis, where ﬁrms with
substantial outside monitoring are most likely to make value-maximizing decisions in
corporate control transactions.

My results also indicate that ﬁrms with higher levels of management ownership
are less likely to be acquired. This supports an entrenchment hypothesis where higher
ownership managers value control over any potential wealth gains from being acquired. I
ﬁnd additional evidence supporting the entrenchment hypothesis in other variables as
well. The tenure of the top executive of non-target ﬁrms is on average two years longer
than on target ﬁrms. Additionally a larger percentage of the boards of directors of non-
target ﬁrms are inside directors.

Table 1.10 compares my results with that of other studies. In particular, the

studies of Mikkelson & Partch (1990) and Song & Walkling (1993) are most similar to

35

this essay. The main difference being the time period considered. Both the Mikkelson &
Partch and Song & Walkling papers are based on samples from the 19705 and 19805,
whereas this essay deals with mergers of the 19905. I ﬁnd similar results to that of Song
& Walkling (1993) in that we both ﬁnd support for the entrenchment hypothesis.
Shivdasani (1993), in his study of hostile targets and non-targets, and Hadlock, Houston,
Ryngaert (1999), in there study of bank targets and non-targets, both ﬁnd supporting
evidence for the entrenchment hypothesis.

My evidence supporting the monitoring hypothesis is collaborated by
Shivdasani’s (1993) study. However, no other paper has created the variables to tests this
hypothesis. As for the growth-resource mismatch hypothesis, 1 ﬁnd no evidence while
Palepu (1986) ﬁnds evidence. Most studies that don’t cOntrol for size by matching their
sample on size ﬁnd evidence of the size hypothesis. All studies ﬁnd little to no evidence
of the market-to-book and price-to-earnings hypotheses.

These ﬁndings raise the question as to the role beneﬁcial stock ownership has in
aligning the managers with the interests of the shareholders at least with regards to being
acquired. Taken as a whole my results are consistent with Morck, Shleifer, and Vishny
(1988b) who suggest that the entrenching properties of high ownership may actually

destroy corporate value.

36

ESSAY 2: Corporate Mergers and Top Management Turnover: The 19905

 

Evidence

2.1 Introduction

Financial economists have identiﬁed many reasons for corporate mergers. Some
reasons postulated for mergers include; disciplinary, synergistic, market expansion,
growth-resources mismatches”, and sometimes even for “empire building”. In this
essay, I examine the role a merger plays in disciplining or validating the performance of
top managers. An acquirer can create value by replacing non-value maximizing top
managers with value-maximizing top managers. Martin and McConnell (1991) have
documented this disciplining mechanism in their sample of tender offer takeovers from
1958 through 1984. Alternatively, value-creating management practices may be an
attractive asset for ﬁrms in the acquisition market. A well-managed ﬁrm could be
attractive for synergistic, growth-resource mismatches, and market expansion type
acquisitions.

Prior studies are largely based on pre-19905 data, a period with a signiﬁcantly
different control environment from what we observe today. With the introduction of
poison pills, new corporate anti-takeover amendments, control share laws, and business
combination laws, absolute deterrence by management opposed to a takeover is a real
possibility today. As discussed by Schwert (2000), in the early 19905 there was a
statistically signiﬁcant shift towards more friendly deals. To the extent that this shift

reﬂects a structural change in the corporate control market, perhaps arising from the

 

'8 Palepu (1986) reports that ﬁrms with a mismatch between their growth opportunities
and ﬁnancial resources at their disposal may be particularly attractive targets.

37

development of modern takeover-defenses and legal rulings in the late 19805, our
understanding of takeovers and their use as a disciplinary or rewarding device would
appear to be quite limited. One might suspect that modern takeover defenses increased
the ability of poorly performing managers to fend off unwanted acquisition attempts and
therefore, disciplinary takeovers.

In this essay I attempt to update the evidence on top management evaluation
within a merger event with a sample of 236 acquisitions from 1990-1997. In particular, I
test for evidence supporting a discipline hypothesis and evidence supporting a
performance evaluation hypothesis. For my sample, targets are classiﬁed as either
disciplinary or nondisciplinary. I classify a merger as disciplinary if within two years

19 A target is

following a merger, the pre-merger target CEO departs the new ﬁrm.
classiﬁed as nondisciplinary if the pre-merger target CEO remains with the ﬁrm for the
two years following a merger. To measure managerial performance, industry-adj usted
and market model returns are calculated for the four years prior to the merger
announcement. To test the performance evaluation hypothesis, for the nondisciplinary
sample, the cumulative industry-adjusted returns and cumulative market model prediction
errors are tested to examine they are signiﬁcantly greater than zero. To test the
disciplinary hypothesis, for the disciplinary sample, the cumulative industry-adjusted
returns and cumulative market model prediction errors are tested on whether they are
signiﬁcantly less than zero. Additionally, I examine whether the cumulative industry-

adjusted returns and cumulative market model prediction errors are signiﬁcantly different

between the disciplinary and nondisciplinary sample.

 

'9 The period evaluated is from the announcement of the merger until 2 years following
the completion of the merger. On average this time period is 28 months.

38

I ﬁrst identify and characterize top management turnover both pre- and post-merger. I
ﬁnd a substantial increase in top management turnover following a merger. Out of the
236 target ﬁrms in my sample, 72.5% (171) experience top manager turnover within the
announcement date of the merger through the two years following completion of a
merger. The pre-merger turnover rate for this same time span would be a substantially
different 26.6%.20 I also document a substantial difference in the characteristics of
turnover events pre-merger vs. post-merger.

Next I measure corporate performance by calculating cumulative industry-adjusted
returns and cumulative market model prediction errors for four years prior to the merger
announcement. The results suggest that, on average, the target of a merger classiﬁed as
nondisciplinary, out-performed other ﬁrms in its industry and performed at least as well
as the market during the pre-merger period. On average, targets in mergers classiﬁed as
disciplinary, performed almost as well as other ﬁrms in their industry and at best, equal to
the market during the pre-merger period. The nondisciplinary targets signiﬁcantly out-
performed the disciplinary targets as measured by industry-adjusted returns and
performed at least as well as the disciplinary on a risk adjusted market model basis, if not
better.

I ﬁnd evidence supporting the discipline hypothesis in line with the results of Martin
and McConnell (1991). Martin and McConnell (1991) document stronger evidence of
the discipline hypotheses, in that ﬁrms in their disciplinary sample are signiﬁcantly
under-performing within their industries, whereas in my sample the disciplinary sample

under-performance is not as signiﬁcant. However, I ﬁnd stronger evidence of what I call

 

20 26.6% is calculated by using the 11.4% average annual turnover rate, pre-merger, times
the 2.33 years (on average) considered following a merger.

39

the performance evaluation hypothesis, where managers who are signiﬁcantly out-
performing within their industry remain on with the new ﬁrm. Sirnilary to Martin and
McConnell (1991), I document that ﬁrms in the nondisciplinary sample signiﬁcantly out-
performed ﬁrms in the disciplinary sample.

Additionally, the characterization of a merger as being hostile or friendly is
investigated. Next, some collaborating evidence for the discipline and performance
evaluation hypotheses is presented. A statistical summary of additional characteristics
concerning whether a manager remains or departs post-merger is presented. The
characteristics include data collected on the top manager, the ﬁrm, and the takeover event
itself.

The rest of this essay is organized as follows. In 2.2 I brieﬂy discuss some of the
relevant literature and the various hypothesis that will be tested. In section 2.3 I describe
the population and sample selection methodology. The data on management turnover is
summarized in section 2.4. Section 2.5 describes industry-adjusted returns and market
model methodologies. The results of the pre-merger performance measures are also
reported in section 2.5. Section 2.6 considers some additional investigations into issues
such as hostile vs. friendly mergers and characteristics of whether a tOp manager remains

or departs post-merger. Section 2.7 summarizes the essay.

2.2 Background and hypothesis development
2. 2. 1 Previous Literature on Ownership and Takeovers
The benchmark for this essay is the work completed by Kenneth J. Martin and John J.

McConnell in 1991, “Corporate Performance, Corporate Takeovers, and Management

40

Turnover.” Their sample includes 720 inter-corporate tender offers over the period 1958
through 1984. Their ﬁnal sample contains 253 takeovers after ﬁrms are deleted
insufﬁcient data availability and the presence of large toeholds. They document that
corporate takeovers play an important role in disciplining the tOp manager of under-
performing targets. They document that the turnover rate for the top manager of target
ﬁrms in tender offer-takeovers signiﬁcantly increases following completion of the
takeover and that prior to the takeover these ﬁrms were signiﬁcantly under-performing
other ﬁrms in their industry as well as other target ﬁrms which had no post-takeover
change in the top manager.

Morck, Shleifer and Vishny (1988) provided additional motivation for this essay. In
a study of large ﬁrm takeovers during the period 1981 to 1985 they document the role of
hostile takeovers as a disciplinary tool in the market for corporate control. They ﬁnd that
targets of hostile takeovers are poor performers, while the targets of ﬁiendly takeovers
are average performers. In summary, most of the previous research suggests that

takeovers are a mechanism for disciplining poorly performing managers.

2. 2.2 The Changing Corporate Control Environment

While the results in this earlier literature have expanded our knowledge of the role
of takeovers in disciplining poor management, it is not clear that these results provide an
accurate description of the role a takeover plays today. In particular, these results rely
exclusively on deals pre-l990. However, the corporate control environment, while quite
active today, appears to have changed in substantial ways since the late 19805. This

change is certainly apparent in my sample. While Shivdasani (1993) reports that

41

approximately 50% of the takeovers in his Value Line sample from 1980-1988 were
hostile, in my sample from 1990-1997 only 8.8% of the deals were hostile (see Table
2.1). Schwert (2000) discusses these trends in detail in his study of 2,346 takeover
contests from 1975 to 1996. He ﬁnds that the highest rates of hostile offers occurred
between 1982 and 1989. Additionally, he ﬁnds that after 1991 there was a statistically
signiﬁcant shift away from hostility comparable in magnitude to the shift towards
hostility seen in the early 19805. Schwert’s (2000) empirical tests show that most deals
described as hostile in the press are not distinguishable from friendly deals in economic
terms, except that hostile transactions involve publicity as part of the bargaining process.

There are some underlying factors as to why the corporate control environment
has become 'ﬁiendlier' in the 19905. In particular, in the late 19805 many corporations
adopted anti-takeover provisions including: a) super-majority provisions, b) fair price
provisions, c) staggered elections for directors, (1) blank check preferred provisions, e)
restrictions on special meetings, and f) elimination of cumulative voting (see
Borokhovich, Brunarski, and Parrino (1997)). In addition, many corporations adopted
shareholder rights plans (i.e. poison pills).

Finally, many states passed many types of anti-takeover legislation such as
control share laws and business combination laws intending to slow hostile acquisition
activity. As Comment and Schwert (1995) report, these anti-takeover provisions and
legal developments certainly changed the rules for control contests in the 19905. By
1991, 87% of NYSE and AMEX listed ﬁrms were covered by at least one form of an
anti-takeover measure; 35% were covered by self-imposed poison pills, 80% by business

combination laws, and 24% by control share laws. These ﬁgures are in stark contrast to

42

1985, where less than 7% of NYSE and AMEX listed ﬁrms were covered by at least one
form of an anti-takeover measure; 1% were covered by self-imposed poison pills, 5% by
business combination laws, and 5% by control share laws.

With the introduction of poison pills, new corporate anti-takeover amendments,
control share laws, and business combination laws, absolute deterrence by management
opposed to a takeover became a real possibility. These anti-takeover measures would
seem to help entrench any management team that was opposed to an acquisition. The
fact that hostile takeovers are harder to complete today would suggest management could
be protected from any unwanted disciplinary takeovers, because all that they would need
is some form of anti-takeover amendment or protection under state law. This raises the
possibility that the takeovers play a more limited role in the disciplining of poor

managers in the post 19805 era.

2. 2.3 Hypothesis Development

Following a merger, a target’s top manager can either remain with the new ﬁrm, or
depart. Prior to a merger, a target’s ﬁnancial performance can be classiﬁed as under-
performing, average performing, or out-performing, vs. some chosen benchmark. This
leads to six possible states, post-merger: 1.) top manager departs and pre-merger under-
perforrnance, 2.) top manager departs and pre-merger average-performance, 3.) top
manager departs and pre-merger out-perforrnance, 4.) top manager remains and pre-
merger under-performance, 5.) top manager remains and top manager average-

perforrnance, and 6.) top manager remains and pre-merger out-performance.

43

The state where the top manager departs and there is pre-merger under-performance
is the outcome of a mechanism that I will refer to as the discipline hypothesis. The
discipline hypothesis is the situation where target management is using non-value
maximizing operating strategies. An acquiring ﬁrm takes control of the target, replaces
management, and creates value by changing the non-value maximizing operating
strategies of management. I will test this hypothesis by analyzing post-merger turnover
events and pre-merger ﬁnancial performance of managers using industry-adjusted
returns, market model returns, and other characteristics.

The state in which the targets top manager remains and there is evidence of pre-
merger out-performance, is an outcome that reﬂects what I will call the performance
evaluation hypothesis. This is the case where target management is doing a better than
average job, and is acquired for just that reason. The target is singled out because of its
superior performance, and the manager is retained because of his/her managerial abilities.
Similarly, I will test this hypothesis by analyzing post-merger turnover events and pre-
merger ﬁnancial performance of managers using industry-adjusted returns, market model
returns, and other characteristics.

The two cases of average performance say nothing about the target’s top manager,
except for the fact that they are average. There is no direct link to be made with whether
a manager remains or departs because of average performance. Another state is where
the top manager departs and pre-merger out-performance is experienced. In this case the
acquirer would probably like to retain the manager, yet the manager has other things he

would like to do. The ﬁnal case is that of where the manager is retained and pre-merger

44

under-performance is experienced. In this case, it would appear that the acquirer does not

discipline the targets top manager for under-performance.

2.3 Data
2.3.] Population Selection

My study focuses on mergers that took place from 1990 - 1997. Following studies
that examine mergers in the 19705 and 19805, I made decisions on the population on
which my sample was derived. (1) To start with, ﬁrms had to have data available in both
CRSP and COMPUSTAT databases. (2) F inns had to be listed on either the New York
Stock Exchange or American Stock Exchange. Certain ﬁrms were then excluded from
the population. (3) Firms not listing the United States as its home country were excluded
from the population. (4) American Depository Receipts (ADRs) and (5) Limited
partnerships were also excluded. (6) Due to the regulatory and speciﬁc nature of
particular industries, some entire industries were excluded. In particular, ﬁnancial,
banking, insurance, real estate and utilities ﬁrms were excluded. Included in this deletion
were industries classiﬁed in 8105 6000 to 6999 (ﬁnance, insurance, real estate) and SIC's
4900 - 4999 (public utilities).

Table 2.2 outlines the total number of ﬁrms listed by year that were left after the
above characteristics were controlled for. From 1990 to 1997 there were an average of
1,814 ﬁrms listed on the NYSE and AMEX in any given year. A list of merged ﬁrms by
year was prepared from the delistrnents from the stock exchanges due to mergers and
acquisitions, obtained from the CRSP and C 0MPUSTAT databases, and manually

conﬁrmed by Dow Jones Interactive searching The Wall Street Journal, Dow Jones News

45

Service, Dow Jones Newswires, and Press Release Wires. During the period, an average
of 47 ﬁrms per year were delisted from the NYSE and AMEX due to merger. This ﬁgure
represents 2.58% of all listed ﬁrms in any given year. Over the period there appears to

have been an increasing number of mergers, with a steady increase starting in 1991.

2. 3.2 Sample Selection

From the above population of ﬁrms a sample of 236 ﬁrms that were acquired
during the period 1990-1997 are used. For a merged ﬁrm to be included in the sample I
required that an announcement date and news story for the merger transaction be
identiﬁed. Dow Jones Interactive (searching The Wall Street Journal, Dow Jones News
Service, Dow Jones Newswires, and Press Release Wires) was used to identify the
speciﬁc announcement and rumor date of the merger. Five years of monthly stock
returns in the CRSP monthly returns ﬁle prior to the merger announcement must be
available in order to calculate industry-adj usted returns and market model prediction
errors prior to the merger.21 Table 2.2 identiﬁes 374 ﬁrms delisted due to merger during
the period 1990 - 1997. Incomplete data availability reduces the number of ﬁrms in the
sample by 113. Following Martin and McConnell (1991) and others, I deﬁne a transfer
of control as occurring when a bidding ﬁrm owns less than 50% of the target’s shares
before the tender offer. Deleting those targets whose acquirer had a toehold of 50% or
greater reduces the number of ﬁrms in the sample by 25, to a total sample size of 236

ﬁrms.

46

2.4 Management Turnover
2. 4.1 Turnover Frequency

A takeover is classiﬁed as disciplinary if there is turnover of the top manager of the
target ﬁrm following the takeover. The top manager is deﬁned as the person holding the
position of CEO, 'or if no CEO exists, the president. Following Martin and McConnell
(1991), turnover data is collected for ﬁve years prior to, to two years after completion of
the merger. The starting point is the ofﬁcial announcement date of the merger. The 12
calendar months prior to the announcement date are considered to be year -1. The next
12 calendar months prior to that are year -2, and so forth. Year +1 begins with the
announcement date and continues for 12 months following completion of the takeover.
On average, year +1 is 16 months in length. Year +2 begins with the ﬁrst month
following year +1 and ends 12 months later.

To document management turnover, the individual occupying the position of CEO or
president is identiﬁed in the Standard and Poor ’5 Register of Corporations, Directors,
and Executives (S&P Register) for each of the ﬁve years prior to the merger
announcement and for the two years following completion of the merger. In addition to
using the S&P Register, each target ﬁrrn’s top manager immediately prior to merger
announcement, is fully researched by name on Dow Jones Interactive, Disclosure, and

Lexus-Nexis for the full two years after the completion of the merger.22 In completing

 

2' I require that at least 12 consecutive monthly returns be available over the period form
month -108 to -49 prior to the merger announcement in order to compute the market
model outlined by Brown and Warner (1980).

22 The Wall Street Journal, Dow Jones News Service, Dow Jones Newswires, and Press
Release Wires are used in the search on Dow Jones Interactive. Disclosure and Lexus-
Nexis are used to search proxy statements and annual reports.

47

these name searches 100% of top managers were accounted for, for the full two years. It
should be noted that in conducting the name searches in the online news services there
were notable differences in what was reported in the S&P Register and the news services.
Not always, but frequently the S&P Register had a signiﬁcant lag that could be measured
in years. The S&P Register seemed to keep previous year data in their publication even
after a turnover transaction.

For clariﬁcation, if the top manager of a target ﬁrm remains as the head of the same
operation within the new ﬁrm, no turnover event is counted. This is true even if his/her
title is not president or CEO. If the target’s top manager departs the new ﬁrm and a new
manager replaces him/her as head of the same Operation, then that replacement is used as
the replacement manager, even if his/her title is not president or CEO. If the top manager
of the target departs the new ﬁrm and is not replaced or the old operation is not stand
alone within the new ﬁrm, then the acquirers CEO is considered the replacement
manager. This methodology is slightly different from Martin and McConnell’s (1991)
who report that CEOs are not replaced in approximately 25% of post-merger turnover
events. In this study, the acquirer CEO replaces the ‘not replaced’ in their study.

Table 2.3 is a summary of top manager changes for years -5 to +2 for all target ﬁrms.
For years -5 through -1 , the annual rate of change in the top manager ranges from 8.1% to
14.0% with an average of 11.4%. Turnover in year +1 leaps to 69.5%. In year +2 the
rate of change in the top manager drops back down to the normal range and a value of

8.1%.23 For comparison, Martin and McConnell (1991) report a range of 7.1% to 11.1%

 

23 Included in year +1 and +2 are all reported changes in the top manager. This includes
the top manager in ofﬁce at the time of the merger and any subsequent changes in the top
manager. This is different from some other tables in this essay where only the top

48

with an average of 9.9% for years -5 through -1. Their tumover in year +1 leaps to
41.9% and in year +2, 19.0%.

Out of the 236 target ﬁrms, 171 (72.5%) experience a change in the top manager
(whom was in ofﬁce immediately preceding the merger announcement) within year +1
through year +2. These ﬁrms that experience a change in the top manager are placed in
the disciplinary sample. The other 65 (27.5%) target ﬁrms that did not experience a
change in the top manager are placed in the non-disciplinary sample. Out of the 180
changes in the top manager within year +1 and year +2 listed in Table 2.3, 9 are
subsequent changes of the top manager in ﬁrms that had already experienced turnover in
the top manager. This is a higher turnover rate in the top manager after a control action
then reported in Martin and McConnell’s (1991). They report top manager turnover in
141 (55.7%) of their 253 tender offers over 1958 - 1984 and no top manager turnover in
112 (44.3%) of their target ﬁrms.

The classiﬁcation methodology that categorized takeovers as disciplinary if the top
manager is replaced, and nondisciplinary if the top manager is not replaced, is not
necessarily perfect. Sometimes management change following a merger is normal, and
not necessarily disciplinary. Martin and McConnell (1991) considered two additional
classiﬁcation schemes for disciplinary and non-disciplinary takeovers that try to
compensate for this.

The ﬁrst method tries to eliminate normal turnover events by examining the stated
reasons for the top manager’s departure. When the top manager of a ﬁrm departs, there is

usually some form of ofﬁcial announcement by the ﬁrm along with a stated reason.

 

manager in office at the time of the merger is included, and subsequent changes are not
included.

49

These stated reasons can then be classiﬁed as either disciplinary or nondisciplinary and
therefore, also the turnover event. Dow Jones Interactive was used to collect the ofﬁcial
stated reason of every turnover event for target ﬁrms from year -5 to year +2. The
reasons indicated for the departing tOp manager are categorized in Table 2.4. The reasons
are separated into non-disciplinary and disciplinary reasons. The non-disciplinary
reasons being; normal retirement, accepted higher level position in ﬁrm, accepted lower
level position in ﬁrm, died, and accepted directorship in ﬁrm. The disciplinary reasons
being; change in control, ﬁred for poor performance, policy differences, early retirement,
other personal or business interests, took similar position with another ﬁrm, and no
reason given. Based on stated reasons, pre—merger top management turnover is
disciplinary 42.5% of the time vs. 75.4% for post-merger turnover. The difference is
statistically signiﬁcant at the 1% level.

For the 171 turnover events post-merger, 129 are classiﬁed as disciplinary, and 42 are
classiﬁed as non-disciplinary. The most common reason cited post-merger is “change of
control”, which accounted for 62 out of the 129 disciplinary events. The next most
common reason was “took similar position with another ﬁrm” with 44 out of the 129
disciplinary events. These results indicate that takeovers are an important device for
changing top management, however, it does not say anything as to why the management
change was needed.

The second method for eliminating instances of nondisciplinary management
turnover is to classify targets according to the origins of the replacement manager.
Furtado and Rozeff (1987), Reinganum (1985), and Vancil (1987) document that the

majority of replacement managers come from inside the ﬁrm in their studies of non-

50

takeover related turnover. Therefore, a replacement manager that comes ﬁ'om outside the
target ﬁrm following a merger is more likely to indicate a disciplinary action than if the
replacement manager comes from inside the target ﬁrm.

For every target 1 document the origin of the replacement manager for all top
manager turnover events ﬁom year -5 through +2. Each event is investigated and
classiﬁed using Dow Jones Interactive.“ Table 2.5 summarizes top management
replacements by pre-merger (time-5 through -1) and post-merger (time +1 through +2).
Pre-merger replacements are outsiders 37.3% of the time. This is compared with
Reinganum (1985) who documents that 13% of replacement managers are outsiders and
Vancil (1987) who documents that 25% of replacement managers are outsiders, in their
respective samples of non-takeover related top manager turnover.

Post-merger replacements are outsiders in 158 out of the 171 top manager turnover
events of target ﬁrms, or 92.4% of the time. This is a difference of 55.1% from pre-
merger turnover of the target ﬁrms, and the difference is signiﬁcantly different from zero
at the 1% level. ~This 92.4% outside replacement rate is dramatically different from the
37.3% pre-merger, from Reinganum’s (1985) 13%, and from Vancil’s (1987) 25%.
Additionally, Martin and McConnell (1991) report that 55.5% of post-merger
replacements are from outside the target ﬁrm in their sample of tender offer takeovers
from 1958 to 1984. However, if you include the ‘not replaced’ events in Martin and
McConnell (1991) with ‘outsider’, as I have done, you get a rate of approximately 70.0%.

This is closer to the 92.4% in my sample, but still substantially different.

 

2" The Wall Street Journal, Dow Jones News Service, Dow Jones Newswires, and Press
Release Wires are used in the search on Dow Jones Interactive.

51

2. 4.2 T op Manager Turnover Pre- vs. Post-Merger

Additional statistics on pre-merger vs. post-merger top management turnover are
displayed in Table 2.6. The age of the departing manager is, on average, 2.43 years less
in a post-merger turnover event. The age of the replacement manager is, on average, 2.29
higher in a post-merger turnover event. The age differential, as measured by the age of
the departing top manager minus the age of the replacement manager, is substantially
different pre- vs. post-merger. The age differential for pre-merger turnover is 7.56 years
vs. 2.73 years for post-merger turnover. The difference is 4.84 years and is statistically
signiﬁcant at the 1% level. This is evidence that pre-merger turnover is more of a natural
type turnover, whereas post—merger turnover seems to be something different. Pre-
merger turnover seems more ‘natural’ in that managers seem to work, get to an age that
they move on to something else, then are replaced by someone 7.56 years younger, on
average. In post-merger turnover events, younger managers replace top managers, but
only younger by 2.73 years, on average. This 5 year difference between the two age
differentials suggests that something different is taking place in post-merger turnover
events. The tenure of the departing top manager is not signiﬁcantly different pre- vs.
post-merger.

A few additional statistics are presented in Table 2.6. The ﬁrst is a dummy variable
that indicates whether the top manager remains with the frrrn in some capacity after
he/she losses the title CEO. The top manager could remain with the ﬁrm in a lower
position, as a director, or possibly chairman of the board. Pre-merger CEOs that lose the
CEO title remain on with the target ﬁrm 32.09% of the time vs. 2.90% of the time post-

merger. The difference is statistically signiﬁcant at the 1% level.

52

The next variable considers whether a regime change took place or not. I consider a
regime change an event where both the titles CEO and chairman are simultaneously
replaced by new manager(s). Regime changes take place in 41.0% of pre-merger
turnover events vs. 89.5% in post-merger turnover events. The difference is statistically
signiﬁcant at the 1% level. A ﬁnal statistic is a dummy variable for whether a CEO holds
both the titles of CEO and chairman at the time of the turnover event. There is no
difference pre- vs. post-merger in this statistic, with both event groups having a CEO
with the dual titles approximately 68% of the time.

All together, the facts point to a difference in the top manager turnover events for
ﬁrms pre-merger vs. post-merger. The substantial increase in turnover following a
merger, the difference in turnover characteristics pre- vs. post-merger, and that a large
percentage of replacements post-merger are from outside the target, all imply that
takeovers are a method of replacing managers of target ﬁrms. However, none of these
facts alone, nor taken together, indicate whether this top management turnover is related
to poor performance. To include corporate performance in this analysis, industry-

adjusted returns and market model returns are calculated below.

2.5 Measuring Corporate Performance

Two methodologies are used to evaluate pre-merger performance of target ﬁrms.
Following Martin & McConnell (1991), I calculate industry-adjusted returns motivated
by Morck, Shleifer, and Vishny(l988)25 and use the market model procedure described

in Brown and Warner (1980).

 

25 Morck, Shleifer, and Vishny (1988) ﬁnd that there is an industry effect in the pre-
takeover performance of the targets of hostile takeovers.

53

2. 5. I Industry-Adjusted Returns

Martin and McConnell’s (1991) methodology of calculating industry adjusted returns
(lARs) is as follows. First, for each month from 1981 to 1997, all ﬁrms on the CRSP ﬁle,
excluding the 374 target ﬁrms, are grouped into 4-digit SIC code portfolios and equally
weighted monthly'retums calculated.26 Next, the S&P Register for the year in which the
target is acquired, is used to identify all of the 4-digit SIC codes of each target ﬁrm.
Next, for each target ﬁrm, the industry portfolios for each of its 4—digit SIC codes are
combined, on an equally weighted basis, to form an industry index for that ﬁrm. If the
CRSP ﬁle contains no ﬁrms in any of the 4—digit SIC categories of a target ﬁrm, the ﬁrst
three-digits of that code are used. If no three-digit match is possible, the ﬁrst two digits
are used, and if no two—digit match is possible, a single digit is used.

IARs are computed by subtracting the return of the industry index from the return of
the target ﬁrm during the same calendar month. Cumulative industry-adjusted returns
(ClARs) are calculated by summing the IARs and are calculated over various pre-merger
time periods from 48 months prior to the merger announcement to the merger
announcement. Statistical tests are performed to determine the signiﬁcance of the

CIARs.
2. 5. 2 Market Model Procedure

The market model procedure described in Brown and Warner (1980) is used to

evaluate risk adjusted corporate performance. Market model parameters are estimated

54

over the period beginning 108 months prior to the merger announcement and ending 49
months prior to the merger announcement. If stock returns are not available for all 60
months, then the available data is used down to a minimum requirement of 12 months.
This requirement is the crucial data availability constraint. Monthly stock returns must
be available for a full ﬁve years prior to the announcement of the merger. Therefore, any
conclusions that are stated in this essay only apply to ﬁrms with ﬁve years worth of
monthly stock returns available. The equally weighed index of all stocks in the CRSP
monthly returns ﬁle is used as the market index. Average monthly prediction errors
(PEs) and cumulative average prediction errors (CPEs) are calculated over various pre-
merger time periods from 48 months prior to the merger announcement to the merger

announcement. Statistical tests are performed to determine the signiﬁcance of the CPEs.

2. 5. 3 Performance Results
2.5.3.1 Disciplinary vs. Nondisciplinary Takeovers and Pre—Takeover Performance
CIARs and CPES are calculated for all 236 target ﬁrms for the time period 48 months
prior until 3 months prior to the merger armouncement. Month 0 is excluded since it is
the month of the merger itself and should not be considered pre-takeover performance.
Studies by Dodd and Ruback (1978), Kummer and Hoffrneister (1978), and Jarrell and
Bradley (1980) all report that information is leaked during the few months prior to a
merger announcement that affects stock prices. Therefore, the two months prior to the

merger announcement are also excluded from the pre-takeover performance period.

 

26 The 374 target ﬁrms include all ﬁrms in the population that were delisted due to
merger. The ﬁnal sample is 236 ﬁrms, after 113 are dropped due to insufﬁcient data
availability and 25 dropped for having a toehold of 50% or greater.

55

Table 2.7 summarizes CIARs and CPEs over the pre-takeover performance period, -
48 to -3 months prior to the merger announcement for the 236 target ﬁrms in the sample.
For the full sample of takeover targets, the pre-merger CIAR is 2.84% and the pre-merger
CPE -3.90%. Neither is signiﬁcant. Additionally, for the full sample, the median CIAR
is -1.55%, and the median CPE is -7.35%. These results might suggest that on average,
the full sample of targets is performing worse than the market, yet they are performing
around average for their industry.

Next in Table 2.7, the sample of 236 target firms is subdivided into two groups based
on if turnover of the top manager takes place or not. The target ﬁrm is classiﬁed as
disciplinary if the pre-merger top manager departs within the year +1 through +2 time
frame and nondisciplinary if the top manager remains on through the end of year +2. The
disciplinary sub-sample includes 171 target ﬁrms and the nondisciplinary sub-sample
includes 65 target ﬁrms. The differences in the two samples are substantial. The pre—
merger CIAR for the disciplinary sample is -3.92% (p-value = .50), and for the
nondisciplinary sample, 20.67% (p-value = .04). The difference between the CIARs of
the two samples is 24.59% (p-value = .03). Additionally, the difference in the medians is
highly signiﬁcant.27 These results indicate that during the pre-merger period, ﬁrms in the
disciplinary sample performed worse or about as well as their industry average, but ﬁrms
in the nondisciplinary sample performed signiﬁcantly better than other ﬁrms in their
industry did. Also, ﬁrms in the nondisciplinary sample performed signiﬁcantly better
than ﬁrms in the disciplinary sample.

The CPEs present a similar story, but with diminished signiﬁcance. The pre-merger

CPE for the disciplinary sample is -11.30% (p-value = .25), and for the nondisciplinary

56

sample, 15.64% (p-value = .29). The difference between the CPEs of the two samples is
26.94% (p-value = .14), and almost signiﬁcant.

The results suggest that, on average, the target of a merger classiﬁed as
nondisciplinary, out-performed other ﬁrms in its industry and performed at least as well
as the market during the pre-merger period. On average, targets in mergers classiﬁed as
disciplinary, performed almost as well as other ﬁrms in their industry and at best, equal to
the market during the pre-merger period. The nondisciplinary targets signiﬁcantly out-
performed the disciplinary targets as measured by industry-adjusted returns and
performed at least as well as the disciplinary on a risk adjusted market model basis, if not
better.

These results suggest that there is support for both the discipline hypothesis and the
performance evaluation hypothesis during the 19905 market for corporate control. The
support for the discipline hypothesis is particularly interesting due to the fact that a vast
array of anti-takeover mechanisms was in place. These anti-takeover mechanisms
allowed managers’ absolute deterrence if they wanted to fend off any unwanted, possibly

disciplinary takeovers.

2.5.3.2 Disciplinary vs. Nondisciplinary: Replacement Classiﬁcation

Additional classiﬁcations of disciplinary and non-disciplinary are considered to check
the robustness of the results. First, the 236 target ﬁrms are subdivided into two groups
based on the origins of the replacement manager. The takeover is classiﬁed as
disciplinary if an outsider replaces the pre-merger top manager within the year +1

through +2 time frame. The takeover is classiﬁed as nondisciplinary if there is no

 

27 Wilcoxon-Z test was used to test for a dif§e7rence in the medians. (P-value = .03)

turnover in the top manager or if an insider replaces the pre-merger top manager within
the year +1 through +2 time frame. The disciplinary sub-sample includes 157 target ﬁrms
and the nondisciplinary sub-sample includes 79 target ﬁrms.

The results are almost identical to the original classiﬁcation of disciplinary and
nondisciplinary, both in magnitude and statistical signiﬁcance. The differences in the
two samples are substantial. The pre-merger CIAR for the disciplinary sample is -5.00%
(p-value = .42), and for the nondisciplinary sample, 18.43% (p-value = .04). The
difference between the CIARs of the two samples is 23.43% (p-value = .03).
Additionally, the difference in the medians is highly signiﬁcant.28 The pre-merger CPE
for the disciplinary sample is -10.l3% (p-value = .33), and for the nondisciplinary
sample, 8.34% (p-value = .52). The difference between the CPEs of the two samples is
18.67% (p-value = .29). Therefore, by reclassifying targets based on the replacement

manager’s origin, the basic results remain the same.

2.5.3.3 Disciplinary vs. Nondisciplinary: Reason Classiﬁcation

Finally, the 236 target fums are subdivided into two groups based on the stated
reason for the top manager’s departure. The takeover is classiﬁed as disciplinary if the
reason indicated for the top manager turnover was change in control, ﬁred for poor
performance, policy differences, early retirement, other personal or business interests,
took similar position with another ﬁrm, or no reason given. The takeover is classiﬁed as
nondisciplinary if there is no turnover in the tOp manager or if the reason indicated for the
top manager turnover was normal retirement, accepted higher level position in the ﬁrm,

accepted lower level position in ﬁrm, died, or accepted directorship in the ﬁrm. The

58

 

 

 

disciplinary sub-sample includes 128 target ﬁrms and the nondisciplinary sub-sample
includes 108 target ﬁrms.

The results are similar to the original classiﬁcation of disciplinary and
nondisciplinary, but not as signiﬁcant. The pre-merger CIAR for the disciplinary sample
is -3.73% (p-value = .60), and for the nondisciplinary sample, 10.64% (p-value = .14).
The difference between the CIARs of the two samples is 14.37% (p-value = .16).
Additionally, the difference in the medians is signiﬁcant at the 11% level. The pre-
merger CPE for the disciplinary sample is -9.00% (p-value = .44), and for the
nondisciplinary sample, 2.19% (p-value = .85). The difference between the CPEs of the
two samples is 11.18% (p-value = .50).

By reclassifying targets based on the stated reasons for the managers departure,
resulted in weakening the signiﬁcance of the results. However, this classiﬁcation scheme
is the one with the most uncertainty. The ﬁrst scheme is based on whether there was a
turnover event or not. The second classiﬁcation scheme was based on the origin of the
replacement manager. Both of these are know with certainty. The scheme based on
reason classiﬁcation is not so certain. There could be misrepresentation in the press
release for various reasons. Also, the reason classiﬁcation itself is some what arbitrary

and perhaps ﬂawed for the purpose of my investigation.

2.5.3.4 Pre-Takeover Performance: Sub-Periods
The CIARs and CPEs for the Turnover Classiﬁcation are broken down into sub-
periods in Table 2.8. The full -48 to -3 months period, prior to the merger announcement,

is listed then broken down into sub-periods of —48 to -25, -24 to ~3, and ~12 to -3 months

 

28 Wilcoxon-Z test was used to test for a difgegrence in the medians. (P-value = .03)

prior to the merger announcement. Considering the CIARs, most of the under-
perforrnance of the disciplinary sample takes place in the -24 to -3 sub-period, with a
value of -9.11%, and statistically signiﬁcant. As for the nondisciplinary sample, the out-
performance seems to take place evenly throughout both the -48 to -25 and -24 to -3 sub-
periods. The most substantial and signiﬁcant difference between the disciplinary and
nondisciplinary samples takes place in the -24 to -3 sub-period. The disciplinary sample
under-performs within its industry by -9.11% and the non-disciplinary sample out-
performs other ﬁrms in its indusz by 9.75%, with a difference of 18.87%. The
difference is statistically signiﬁcant at the 1% level.

The same results play out when considering the CPE sub-periods. Over the entire
time period, the nondisciplinary sample out-performs the disciplinary sample, however
the difference in only signiﬁcant at the 14% level. The disciplinary sample is under-
performing the market very signiﬁcantly throughout the -24 to -3 sub-period. Whereas, it
seems the nondisciplinary targets are performing as well as the market over the same
period.

Overall, in considering the sub-periods of pre-merger performance, it seems most of
the action is in the -24 to -3 months prior to the merger announcement. The disciplinary
sample is signiﬁcantly under-performing within its indusz, signiﬁcantly under-
performing the market, signiﬁcant under-performing the nondisciplinary within the
respective industries, and most likely under-performing the nondisciplinary sample on a
risk-adjusted basis. The nondisciplinary sample is out-performing within its industry,
performing at least as well as the market, signiﬁcantly out-performing the disciplinary

sample within the respective industries, and probably out-performing the disciplinary

60

 

 

sample on a risk-adjusted basis. The results provide more deﬁnition as to when the pre-
perforrnance of targets is signiﬁcant. The results also add supporting evidence to the

existence of the discipline hypothesis and the performance evaluation hypothesis.

2.6 Additional Investigations
2. 6. 1 Hostile versus Friendly Takeovers

Much has been made about the hostile vs. ﬁiendly characterization of takeovers in
economic literature. As discussed earlier, the 19905 could be characterized as a ‘friendly’
corporate control environment. With the advent of poison pills, new corporate anti-
takeover amendments, control share laws, and business combination laws, absolute
deterrence by management opposed to a takeover became a real possibility. Therefore,
within this environment not many hostile takeovers were initiated. In the initial
population of all ﬁrms delisted due to merger from 1990 to 1997, only 8.8% of the 374
mergers were classiﬁed as being hostile. In my ﬁnal sample of 236 targets, only 9.3%, or
22 were classiﬁed as hostile.

Morck, Shleifer, and Vishy (1988), in a sample of large ﬁrm targets in the early 19805
ﬁnd that the targets of hostile takeovers are poor performers, and therefore disciplinary
control events. In my sample of targets, 171 are classiﬁed as disciplinary and 65 as non-
disciplinary. Of those classiﬁed as disciplinary, 10.5% (18) are classiﬁed as hostile and
of those classiﬁed as nondisciplinary, 6.2% (4) are classiﬁed as hostile. The difference
between the two is not signiﬁcant (p-value = .30).

To test the conjecture of Morck, Shleifer, and Vishy (1988) that the motive for a

takeover determines its character, Martin and McConnel (1991) sub-divide their sample

61

 

 

based on the hostile and friendly classiﬁcation of the takeover. They classify 50% of the
takeovers in their sample as hostile. They ﬁnd no evidence that the hostile classiﬁcation
distinguishes between disciplinary and nondisciplinary takeovers and no evidence that the
pre-takeover performance of targets of disciplinary takeovers differ from the pre-takeover
performance of nondisciplinary takeover targets. I ﬁnd the same results as Martin and
McConnell (1991). In calculating pre-merger CIARs and CPEs over the 48 to 3 month
time period prior to the merger announcement, I ﬁnd no signiﬁcant statistics and no
signiﬁcant differences based on the hostile (22 targets) and friendly (214 targets)

classiﬁcation of targets as disciplinary (hostile) and nondisciplinary (friendly).29

2. 6. 2 Additional Characteristics

Some additional data was collected on the top manager, the ﬁrm, and the takeover
event itself. For the set of all target ﬁrms, I searched the Disclosure and Lexus-Nexis
databases for the proxy statement for the year ending prior to the date the merger was
announced. If a proxy was not found in the above database searches the Securities and
Exchange Commissions Q-ﬁle Data Corporation database on microﬁche was used. I
collected detailed data from the proxy statement concerning the top manager’s
characteristics, management compensation, ﬁrm characteristics, and board structure of
the ﬁrm.

Tables 2.9 and 2.10 report the values for the variables based on the disciplinary /
nondisciplinary split, the split on whether the top manager remains with the ﬁrm through

year +2 or departs. The ﬁrst group of variables concerns the top manager’s

 

29 CIARS: hostile 1.8% (p-value = .88), ﬁiendly 2.9% (p-value = .59), difference 1.1% (p-
value = .95). CPEs: hostile -22.5% (p-value = .32), friendly -2.0% (p-value = .82),

62

characteristics. The age of the manager does not seem to be an important factor with
both the remaining and departing top managers being approximately 55 years old. The
average tenure of the remaining top manager is greater by almost 2 years, but is not quite
statistically signiﬁcant. The remaining manager owns on average 8.37% of the target
ﬁrm, while the departing top manager owns on average only 5.87%, however this

0

difference is not quite statistically signiﬁcant. 3 Cash compensation is statistically

different between the remaining and departing managers. The remaining managers make

3’ The difference is in

signiﬁcantly more cash compensation than the leaving managers.
the annual cash salary of the top managers, not the annual cash bonuses. This higher cash
compensation for the remaining top managers probably has to do with the fact that
remaining top managers are out-performing other ﬁrms within their industry as reported
earlier in the CIAR analysis.

Additionally, certain ﬁrm ﬁnancial characteristics were collected from the
COMPUST AT database. Firm value, total assets and free cash ﬂow to assets were
collected for the ﬁscal year ending prior to the announcement date of the merger for each
target. Return on equity is averaged for three ﬁscal years prior to the announcement date

of the target's merger. All ﬁnancial variables utilized in this essay are described in detail

in the Appendix 3.

 

difference 20.6% (p-value = .47).

30 For each top manager I record their ownership of common stock and options
exercisable within 60 days. The sum of common stock and options exercisable within 60
days is commonly referred to as beneﬁcial ownership. This data is collected from proxy
statements immediately prior to the announcement of the merger.

3 1 For each tOp manager I record annual cash salary and annual cash bonuses. This data
is collected ﬁom proxy statements immediately prior to the announcement of the merger.
All cash compensation statistics are stated in a 1996 dollars.

63

 

 

Size does not seem to be a factor in determining whether a manager remains or
departs after a merger. Both total ﬁrm value and total assets are not signiﬁcantly
different. Free cash ﬂow to assets and return on equity are however signiﬁcantly
different. Both are signiﬁcantly higher in the ﬁrms in which the top manager remains.
These results collaborate with previous CIARs and CPEs results reported earlier. This
evidence reinforces the fact that the ﬁrms in which the top manager remains are
performing signiﬁcantly better then in ﬁrms where the top manager departs.

I also collected public bond ratings from Standard and Poor’s C 0MPUST AT and
Moody's for all target ﬁrms. 1 created a dummy variable for non-investment grade public
debt. 32 This statistic tells a similar story about performance; the ﬁrms in which the top
manager departs have signiﬁcantly more non-investment grade bond ratings than ﬁrms
do where the top manager remains.

Additional ﬁrm characteristics were collected from the target ﬁrms’ proxy statement
immediately preceding the announcement of the merger. These characteristics include
staggered voting for board members, the existence of two or more classes of voting stock,
and board member composition.

Staggered voting for board members is signiﬁcantly different between the two sub-
sarnples. Whereas 63.1% of the targets where the manager remains have staggered
voting, only 48.5% of the targets where the manager departs. This is possibly evidence
that staggered voting is an effective anti-takeover measure. The fact that what I have
called disciplinary targets have signiﬁcantly less instances of staggered voting suggests
that top managers, who under-perforrn are less protected from unwanted takeovers then

those ﬁrms who out-performed and were classiﬁed as nondisciplinary. However, a more

64

appropriate assessment would be to compare under-performing, non-targets with the
under-performing disciplinary targets. The dummy variable for the existence of two or
more classes of voting shares, another anti-takeover measure, is not signiﬁcantly different
between the two sub-samples.

Morck , Shleifer, and Vishny (1989) ﬁnd evidence that corporate boards do not
always do an adequate job of removing unresponsive managers. They document that
takeovers come to play a role in replacing managers who the board is unable or unwilling
to discipline. For this reason I investigate the composition of the board. It could be that
certain types of board members are at fault for not disciplining non-value maximizing
managers.

For each ﬁrm I classify all directors as either inside, outside or gray directors.33
Inside directors are individuals who were full time employees of the fum or one of its
subsidiaries along with their immediate family members. Gray directors are individuals
who were retired employees of the ﬁrm, non-immediate family members of employees of
the ﬁrm and individuals where the proxy statement indicated an ongoing business
relationship between the individual and the ﬁrm. Most people classiﬁed as gray outsiders
were retired employees of the ﬁrm or family members. All other directors are deﬁned to
be outside directors.

If the board was not performing its monitoring job on an ongoing basis, a reason
could be differences in the board composition between the target ﬁrms whose manager
departs and the target ﬁrms whose manager remains. In particular, a board dominated by

insiders might be less inclined to remove the top manager when the manager is

 

32 I used the methodology outlined by Billett (1996).
33 I follow the methodology outlined by Hadlock, Houston, and Ryngaert (1999).

65

performing poorly. Alternatively, a board dominated by outsiders might not be doing an
adequate job of removing poor performing top managers. Therefore, there might be
some difference in the composition of the board based on post-merger disciplinary
turnover activity.

The percent of board members classiﬁed as insiders is not signiﬁcantly different.
Therefore is seems that insiders do not have any undo inﬂuence in turnover events.
However, there is some difference in those board members classiﬁed as outsiders and
gray. There are on average about 4% more outsiders on boards in ﬁrms where the
manager departs post-merger. This would suggest that outside board members have not
been doing an adequate job of monitoring poor-performing managers pre-merger. This
result is not very strong with signiﬁcance only at the 13% level. The percent of gray
board members is signiﬁcantly different. Targets where the manager remains have 7.4%
gray directors vs. 4.2% in targets where the manager departs. This difference could
suggest that gray directors do a good job of replacing poorly performing top managers at
when needed. Additionally, since the targets in which the top manager remains are better
performing ﬁrms, maybe the inﬂuence of gray directors has something to do with
performance. However, given an average of 10 board members, this difference is
roughly 1/3 of a board member so not much can really be made of this result.

The ﬁnal statistics in Table 2.10 have to do with characteristics of the takeover itself.
The takeover was classiﬁed as being hostile or not, whether multiple bidders existed,
whether the offer was for all cash, whether the acquisition was by a foreign ﬁrm, and

whether the acquirer had a toehold of less than 50% prior to the merger.34 Only one of

 

34 Firms with toehold of equal to or greater than 50% were dropped from the entire
analysis in this essay.

66

these variables was signiﬁcantly different between the two sub-samples, the toehold
variable. 12.3% of the targets in which the manager remained had a toehold in the target

pre-merger vs. only 2.9% of the targets in which the manager departed.

2.7 Conclusion

Most of the existing literature concerning corporate control transactions has focused
on the merger activity pre-19905. However, substantial evidence suggests that the current
corporate control market differs in important ways from this earlier period. In particular,
today's takeovers appear to be ﬁiendlier in nature and focused more on expanding
markets compared to the hostile transactions and downsizing that characterized the
19805. The relative “friendliness” of these more modern transactions is likely driven by
the adoption of anti-takeover provisions and legal developments in the late 19805. This
essay attempts to explore the role a merger plays in disciplining or rewarding the
performance of tOp managers in this new environment.

I study a sample of 236 industrial NYSE/AMEX ﬁrms acquired from 1990 to 1997.
Following the methodology of Martin and McConnell (1991), I document evidence
supporting the discipline hypothesis and the performance evaluation hypothesis. Initially,
I document that turnover in the top manager increases signiﬁcantly following a merger.
The characteristics of top manager turnover following a merger are also signiﬁcantly
different then top manager turnover pre-merger. In particular, post-merger turnover is
characterized as being more disciplinary than pre-merger turnover. For example, 75.4%

of post-merger top manager turnover is classiﬁed as disciplinary, vs. 42.5% pre-merger,

67

based on the stated reasons for departure. Additionally, the top manager is replaced with
an outsider 92.4% of the time post-merger, vs. 37.3% pre-merger.

Next I document the relationship between pre-merger performance and top manager
turnover. The results suggest that, on average, the target of a merger classiﬁed as
nondisciplinary, out-performed other ﬁrms in its industry and performed at least as well
as the market during the pre-merger period. On average, targets in mergers classiﬁed as
disciplinary, performed almost as well as other ﬁrms in their industry and at best, equal to
the market during the pre-merger period. The nondisciplinary targets signiﬁcantly out-
performed the disciplinary targets as measured by industry-adjusted returns and
performed at least as well as the disciplinary on a risk adjusted market model basis, if not
better.

I also provide additional evidence on which top managers remain and which depart.
Variables such as free cash ﬂow to assets, return on equity and a non-investment grade
bond rating all point to manager who remain out-perform managers who departed.

Taken as a whole, my ﬁndings suggest that there is support for both the discipline
hypothesis and the performance evaluation hypothesis. These are similar results to
Martin and McConnell’s (1991) results based on data from 1954 to 1984. The support
for the discipline hypothesis is particularly interesting due to the fact that a vast array of
anti-takeover mechanisms were in place during my sample period. These anti-takeover
mechanisms allowed managers’ absolute deterrence if they wanted to fend off any
unwanted, possibly disciplinary takeovers. Yet even with these in place, it appears
takeovers are still needed to perform these disciplinary actions to maximize shareholder

value.

68

APPENDIX 1

TABLES OF ESSAY 1

69

Table 1.1
Merger Classiﬁcation: Hostile vs. Non-Hostile

 

Total successfully merged ﬁrms from a population of ﬁrms available from 1990 to 1997.
Population includes 1.) ﬁrms with CRSP and C 0MPUSTAT data available, 2.) NYSE and
AMEX listed ﬁrms only, 3.) United States ﬁrms only, 4.) ADR’s deleted, 5.) limited
partnerships deleted, 6.) SIC’s 4900 - 4999 (utilities) and SIC ’s 6000 - 67999 (ﬁnancials)
deleted. Hostile classiﬁcation is based on Morek, Shleifer, and Vishny (1988)

methodology.

 

Successfully Merged Firms

 

 

 

Non- Non-

Total Hostile hostile Hostile hostile
1990 43 4 39 9.3% 90.7%
1991 16 2 14 12.5% 87.5%
1992 23 2 21 8.7% 91.3%
1993 26 2 24 7.7% 92.3%
1994 40 3 37 7.5% 92.5%
1995 58 8 50 13.8% 86.2%
1996 70 5 65 7.1% 92.9%
1997 98 7 91 7.1% 92.9%
Total 374 33 341 8.8% 91.2%

 

70

Table 1.2

Hypotheses Tested

 

Hypothesized sign of logit regression coefﬁcients. Dependent variable = 1 if the ﬁrm
was acquired and 0 if it was not acquired.

 

 

 

 

 

 

 

% of BOD Outside Unaff.
Mgrl. Tenure that are director Blkhldg.
Ownership of CEO Outsiders Ownership Ownership Liquidity
Fin. Incentive positive
Entrenchment negative negative
Discipline negative negative negative negative positive
Monitoring negative positive positive positive
Irrelevance no relation no relation no relation no relation no relation no relation

 

iG-Resource

 

Size

 

Mkt-to-Book

 

Price-Earnings

 

 

 

 

 

 

 

 

71

 

Table 1.2, Continued
Hypotheses Tested

 

Hypothesized sign of logit regression coefﬁcients. Dependent variable = 1 if the ﬁrm
was acquired and 0 if it was not acquired. '

 

 

 

 

 

 

 

 

 

 

 

Growth- Book Value Market/ Price/
Resource Total Book Earnings
Leverage ROE Indicator Assets Ratio Ratio
Fin. Incentive
Entrenchment
Discipline negative negative
Monitoring
Irrelevance no relation no relation no relation no relation no relation no relation
G-Resource positive
Size negative
Mkt-to-Book negative
Price-Earnings positive

 

 

 

 

 

 

 

 

72

 

Table 1.3

Population Summary, 1990 to 1997

 

Population of ﬁrms available from 1990 to 1997. Population includes 1.) ﬁrms with
CRSP and COMPUSTATdata available, 2.) NYSE and AMEX listed ﬁrms only, 3.)
United States ﬁrms only, 4.) ADR’s deleted, 5.) limited partnerships deleted, 6.) SIC’s
4900 - 4999 (utilities) and SIC’s 6000 - 67999 (ﬁnancials) deleted.

 

 

Firms delisted

Mergers as a %

 

 

 

 

Total listed ﬁrms due to merger of total listed ﬁrms
NYSE AMEX Tot. NYSE AMEX Total NYSE AMEX Tot.
1990 1244 400 1644 28 15 43 2.3% 3.8% 2.6%
1991 1283 400 1683 11 5 16 0.9% 1.3% 1.0%
1992 1345 417 1762 12 11 23 0.9% 2.6% 1.3%
1993 1396 431 1827 19 7 26 1.4% 1.6% 1 .4%
1994 1438 428 1866 25 15 40 1.7% 3.5% 2.1%
1995 1518 445 1963 44 14 58 2.9% 3.1% 3.0%
1996 1495 437 ' 1932 51 19 70 3.4% 4.3% 3.6%
1997 1422 414 1836 78 20 98 5.5% 4.8% 5.3%
Avg. 1393 422 1814 34 13 47 2.4% 3.1% 2.6%
268 106 374 Total Mergers
32 Incomplete data avail.
342 Final Sample

 

 

73

Table 1.4
Univariate Results

 

Summary statistics comparing a sample of 342 ﬁrms that were acquired between 1990 —
1997 with a match sample of 342 ﬁrms that were not acquired. (p-values in parenthesis)

 

Matched Target
Characteristic Firms Firms Diff. t-test

(Mean) (Mean) (M51)

 

 

Firm Financial Characteristics

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

total assets 1,010 1,078 -68 -0.364
(716)

market value of equity

at e.o.c.y. 835 927 ~93 -0.535
(593)

total ﬁrm value 1,343 1,477 -134 -0.513
(608)

sales growth .27 .15 .12 2.203 **
(028)

liquidity -.12 -. 15 .04 2.261 * *
(.024)

leverage .94 -.79 1 .73 0.803
(423)

growth-resource dummy .23 .26 -.04 -1.065
(287)

% return on equity -.07 .06 -.12 -1.685
(.093)

market to book ratio 2.97 6.13 -3.16 -0.919
(359)

price to earnings ratio 8.95 13.02 -4.08 -0.699
(485)

nppe to total assets .36 .35 .01 0.488
(.626)

fcf to total assets .05 .07 -.01 -1.085
(278)

2 classes of voting stock exist .18 .09 .09 3.484 ***

- (.001)

Board of Directors Characteristics

% of board that are insiders .35 .29 .06 4.508 ***
(.000)

% of board that are outsiders .62 .66 -.04 -3.237 ***
(.001)

% of board that are grey .03 .05 -.01 -2.073 **
(029)

 

74

Table 1.4, Continued
Univariate Results

 

Summary statistics comparing a sample of 342 ﬁrms that were acquired between 1990 -
1997 with a match sample of 342 ﬁrms that were not acquired. (p-values in parenthesis)

 

Matched Target
Characteristic Firms Firms Diff. t-test
(Mean) (Mean) (Mean)

 

 

# of other directorships

 

 

 

held by inside directors .78 .78 -.01 -0.063
(.950)

# of other directorships

held by outside directors 1.55 1.70 -.15 -1.499
(134)

total number of directors 8.35 8.51 -.16 -0.769
(.442)

staggered voting for BOD .48 .47 .01 0.229
(.819)

 

*** denotes signiﬁcance at the 1.0% level
** denotes signiﬁcance at the 5.0% level
* denotes signiﬁcance at the 10.0% level

75

Table 1.4, Continued
Univariate Results

 

Summary statistics comparing a sample of 342 ﬁrms that were acquired between 1990 —
1997 with a match sample of 342 ﬁrms that were not acquired. (p—values in parenthesis)

 

Matched Target

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Characteristic Firms Firms m t-_te_st
(Median) (Median) (Median)

Firm Financial Characteristics

total assets 238 344 -106 -2.094 **
(.036)

market value of equity

at e.o.c.y. 203 227 -24 -0.748
(455)

total ﬁrm value 326 440 -115 -l .239
(215)

sales growth .11 .08 .03 2.380 **
(.017)

liquidity -. 14 -.16 .03 2.321 **
(.020)

leverage .48 .45 .02 -0515
(607)

growth-resource dummy .00 .00 .00 -1.065
(287)

% return on equity .09 .08 -.01 0.099
(.921)

market to book ratio 1.92 1.78 .14 1.740
(.083)

price to earnings ratio 13.26 12.99 .27 0.612
(541)

nppe to total assets .29 .29 .00 0.505
(.613)

fcf to total assets .08 .08 .00 0.170
(865)

2 classes of voting stock exist .00 .00 .00 3.456 ***
(.001)

Board of Directors

Characteristics

% of board that are insiders .33 .25 .08 4.477 ***
(.000)

% of board that are outsiders .63 .67 -.04 -3.277 ***
(.001)

% of board that are grey .00 .00 .00 -2.517 **
(.012)

 

76

Table 1.4, Continued
Univariate Results

 

Summary statistics comparing a sample of 342 ﬁrms that were acquired between 1990 —
1997 with a match sample of 342 ﬁrms that were not acquired. (p-values in parenthesis)

 

Matched Target
Characteristic Firms Firms Diff. t-test
(Median) (Median) (Median)

 

 

 

 

 

# of other directorships

held by inside directors .40 .33 .07 0.550
(.583)

# of other directorships

held by outside directors 1.29 1.5 -.22 -1 .250
(211)

total number of directors 8.00 8.00 .00 -l.317
(188)

staggered voting for BOD .00 .00 .00 0.229
(.819)

 

*** denotes signiﬁcance at the 1.0% level
** denotes signiﬁcance at the 5.0% level
* denotes signiﬁcance at the 10.0% level

77

Table 1.5
Univariate Results, Continued

 

Summary statistics comparing a sample of 342 ﬁrms that were acquired between 1990 -
1997 with a match sample of 342 ﬁrms that were not acquired. (p-values in parenthesis)

 

Matched Target
Characteristic Firms Firms Diff. t-test

(Mean) (Mean) (Mean)
Ownership Characteristics (% of stock owned)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

inside directors .16 .11 .05 3.773 ***
(.000)

outside directors .03 .03 .00 -0.265
(.791)

grey directors .00 .01 .00 -1.546
(118)

all directors .19 .15 .05 3 .149 * *
(.002)

executives .02 .02 .01 2.290 **
(.022)

inside directors / executives .18 .12 .06 4.183
(.000) ***

outside and grey directors .03 .04 -.01 -0.867
(386)

directors / executives .22 .16 .05 3.516 ***
(.001)

top executive .10 .07 .03 2.940 ***
(.003)

CEO/C/P .13 .09 .05 3.618 ***
(.000)

top 5 salaried executives .14 .10 .04 3.412 ***
(.001)

unaff. blockholders .19 .27 -.09 -5.064 ***
(.000)

aff. blockholders .03 .02 .01 0.851

. (.395)

unaff. blockholders not

represented on BOD .13 .14 .00 -0.393
(.695)

unaff. blockholders

represented on BOD .05 .13 -.08 -5.158 ***
(.000)

all outside directors / unaff.

blockholders .22 .31 -.09 -5.326 ***
(.000)

 

78

Table 1.5, Continued
Univariate Results, Continued

 

Summary statistics comparing a sample of 342 frrrns that were acquired between 1990 -
1997 with a match sample of 342 ﬁrms that were net acquired. (p—values in parenthesis)

 

Matched Target

 

 

 

 

 

 

 

 

Characteristic Firms Firms DE, t_-t_e_s_t
(Mean) (Mean) (Mean)

inside directors / aff. blockholders .19 .13 .06 3 .905 * * *
(.000)

inside directors / executives / aff.

blockholders .21 .15 .06 4.296 * * *
(.000)

CEO Characteristics

age of top executive 55.67 54.85 .82 1.221
(.222)

# of years the top exec. served

as a director 13.01 1 1.08 1.93 2.494 **
(.01 3)

Indicator: one man management

team .52 .56 .00 -1.070
(.284)

cash compensation top executive

(000) 686 , 715 -29 -0.505
(.614)

 

*** denotes signiﬁcance at the 1.0% level
** denotes signiﬁcance at the 5.0% level
* denotes signiﬁcance at the 10.0% level

79

Table 1.5, Continued
Univariate Results, Continued

 

Summary statistics comparing a sample of 342 ﬁrms that were acquired between 1990 —
1997 with a match sample of 342 ﬁrms that were not acquired. (p—values in parenthesis)

 

Matched Target
Characteristic Firms Firms Diff. t-test
(Median) (Median) (Median)

 

 

Ownership Characteristics L% of stock owned)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

inside directors .07 .04 .04 4.583 ***
(.000)

outside directors .01 .00 .00 0.550
(582)

grey directors .00 .00 .00 -2.205 * *
(.027)

all directors .1 1 .06 .05 3.648 ***
(.000)

executives .01 .01 .00 1.617
(.106)

inside directors / executives .09 .05 .04 4.645 * * *
(.000)

outside and grey directors .01 .01 .00 -0.317
(.75 1)

directors / executives .13 .08 .05 3.845 ***
(.000)

tOp executive .03 .02 .01 3.422 ***
(.001)

CEO/C/P .05 .03 .03 3.301 ***
(.001)

top 5 salaried executives .06 .04 .03 4.150 ***
(.000)

unaff. blockholders .13 .21 -.08 -4.678 ***
(.000)

aff. blockholders .00 .00 .00 0.755
(450)

unaff. blockholders not

represented on BOD .00 .00 .00 -0.743
(458)

unaff. blockholders

represented on BOD .00 .00 .00 -5.059 ***
(000)

all outside directors / unaff.

blockholders .16 .25 -.09 -5.l93 ***
(.000)

 

80

Table 1.5, Continued
Univariate Results, Continued

 

Summary statistics comparing a sample of 342 ﬁrms that were acquired between 1990 —
1997 with a match sample of 342 ﬁrms that were not acquired. (p-values in parenthesis)

 

Matched Target

 

 

 

 

 

 

 

 

Characteristic Firms Firms m ties!
(Median) (Median) (Median)

inside directors / aff. blockholders .10 .05 .05 4.413 ***
(.000)

inside directors / executives / aff.

blockholders .12 .07 .05 4.547 ***
(000)

CEO Characteristics

age of top executive 55.00 55.00 .00 1.000
(318)

# of years the top exec. served

as a director 9.00 8.00 1.00 2.906 ***
(.004)

Indicator: one man management

team 1.00 1.00 .00 -1.070
(.284)

cash compensation top executive

(000) 491 518 -27 -0.778
(436)

 

*** denotes signiﬁcance at the 1.0% level
** denotes signiﬁcance at the 5.0% level
* denotes signiﬁcance at the 10.0% level

81

Table 1.6
Multivariate Logit Results (% of stock owned)

 

Logit results relating the likelihood a ﬁrm is acquired to a set of explanatory variables
including ownership structure and ﬁnancial characteristics of a ﬁrm. Dependent variable
= 1 if the ﬁrm was acquired and 0 if it was not acquired. (p-values in parenthesis)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(1) (2) (3)
% of stock owned by:
inside directors / executives -l .159 ** -1.152 **
(.013) (.013)
all outside directors 1.558
(131)
outside directors 0.990
(.366)
gray outside directors 5.318 *
(.099)
unafﬂiated blockholders 1.490 *** 1.495 ***
(.000) (.000)
affliated blockholders -0.249 -0.248
(.804) (.805)
inside directors / executives
/ afﬁliated blockholders
all outside directors
/ unafﬁlitate blockholers
% return on equity 0.125
(.188)
growth-resource dummy 0.182
(325)
sales growth -0.407 *
(.070)
liquidity -1.044 **
(.021)
leverage -0.026
(.440)
In of assets 0.092
(.079)
market to book ratio 0.004
(.643)

 

82

Table 1.6, Continued
Multivariate Logit Results (% of stock owned)

 

Logit results relating the likelihood a ﬁrm is acquired to a set of explanatory variables
including ownership structure and ﬁnancial characteristics of a ﬁrm. Dependent variable
= 1 if the ﬁrm was acquired and 0 if it was not acquired. (p-values in parenthesis)

 

 

 

 

(1) (2) (3)

price to earnings ratio 0.001

(520)
constant -0.639 * -0.212 -0.212

(.046) (.205) (.204)
Pseudo R2 0.026 0.038 0.036
Number of observations 653 684 684
Prob > chi2 (.003) *** (.000) *** (.000) ***

 

*** denotes signiﬁcance at the 1.0% level
** denotes signiﬁcance at the 5.0% level
* denotes signiﬁcance at the 10.0% level

83

Table 1.6, Continued
Multivariate Logit Results (% of stock owned)

 

Logit results relating the likelihood a ﬁrm is acquired to a set of explanatory variables
including ownership structure and ﬁnancial characteristics of a ﬁrm. Dependent variable
= 1 if the ﬁrm was acquired and 0 if it was not acquired. (p-values in parenthesis)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(4) (5)
°/o of stock owned by:
inside directors / executives -0.945 **
(.049)
all outside directors 2.093 *
(.062)
outside directors
gray outside directors
unafﬂiated blockholders 1.524 ***
(.000)
afﬂiated blockholders -0.688
(519)
inside directors / executives -1.029 **
/ afﬁliated blockholders (.018)
all outside directors 1.482 ***
/ unafﬁlitate blockholers (.000)
% return on equity 0.183 *
(.086)
growth-resource dummy 0.215
(.258)
sales growth -0.435 "‘
(068)
liquidity -1.215 ***
(.010)
leverage -0.023
(.509)
In of assets 0.070
(231)
market to book ratio 0.001
(.900)

 

84

Table 1.6, Continued
Multivariate Logit Results (% of stock owned)

 

Logit results relating the likelihood a ﬁrm is acquired to a set of explanatory variables
including ownership structure and ﬁnancial characteristics of a ﬁrm. Dependent variable
= 1 if the ﬁrm was acquired and 0 if it was not acquired. (p-values in parenthesis)

 

 

 

 

(4) (5)
price to earnings ratio 0.001
(408)
constant -0.204 0.801 *
(.220) (.061)
Pseudo R2 0.036 0.059
Number of observations 684 653
Prob > obi2 (.000) (.000)

 

*** denotes signiﬁcance at the 1.0% level
** denotes signiﬁcance at the 5.0% level
* denotes signiﬁcance at the 10.0% level

85

Table 1.7

Hypotheses Results

 

Hypothesized sign of logit regression coefﬁcients. Dependent variable = 1 if the ﬁrm
was acquired and 0 if it was not acquired. ‘

 

 

 

 

 

 

 

 

 

% of BOD Outside Unaff.
Mgrl. Tenure that are director Blckhldg

Ownership of CEO Outsiders Ownership Ownership Liquidity
Fin. Incentive positive
Entrenchment negative negative
Discipline negative negative negative negative positive
Monitoring negative positive positive positive
Irrelevance no relation no relation no relation no relation no relation no relation
lG-Resource
Size

 

Market-to-Book

 

Price-Earnings

 

 

 

 

 

 

 

 

 

Results of study

 

 

negative

 

negative

 

positive

 

positive

 

positive

 

negative

 

86

 

 

Table 1.7, Continued
Hypotheses Results

 

Hypothesized sign of logit regression coefﬁcients. Dependent variable = 1 if the ﬁrm
was acquired and 0 if it was not acquired.

 

 

 

 

 

 

 

 

 

 

 

Book
Growth— Value Market/ Price/
Resource Total Book Earnings
Leverage ROE Indicator Assets Ratio Ratio
Fin. Incentive
Entrenchment
Discipline negative negative
Monitoring
Irrelevance no relation no relation no relation no relation no relation no relation
G-Resource positive
Size negative
Market-to-Book negative

 

Price-Earnings

 

 

 

 

 

positive

 

 

 

Results of study

 

no relation

 

no relation

 

no relation

 

no relation

 

no relation

 

no relation

 

87

 

 

Table 1.8
Multivariate Logit Results (% of votes owned)

 

Logit results relating the likelihood a ﬁrm is acquired to a set of explanatory variables
including ownership structure and ﬁnancial characteristics of a ﬁrm. Dependent variable
= 1 if the ﬁrm was acquired and 0 if it was not acquired. (p-values in parenthesis)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

( 1) (2) (3)
% of votes owned by:
inside directors / executives -1.031 ** -1.022 **
(.012) ( .012)
all outside directors 1.350
(191)
outside directors 0.750
(497)
gray outside directors 5.105
(105)
unafﬂiated blockholders 1.396 *** 1.400 ***
(.000) (.000)
aﬁliated blockholders 0.011 0.018
(.989) (.983)
inside directors / executives
/ afﬁliated blockholders
all outside directors
/ unafﬁlitate blockholers
% return on equity 0.125
(188)
growth-resource dummy 0.182
(325)
sales growth -0.407 **
(.070)
liquidity —1.044 **
(.021)
leverage -0.026
(.440)
In of assets 0.092
(.079)
market to book ratio 0.004
(.643)
price to earnings ratio 0.001
(520)

 

88

Table 1.8, Continued
Multivariate Logit Results (% of votes owned)

 

Logit results relating the likelihood a frrm is acquired to a set of explanatory variables
including ownership structure and ﬁnancial characteristics of a ﬁrm. Dependent variable
= 1 if the ﬁrm was acquired and 0 if it was not acquired. (p-values in parenthesis)

(1) (2) (3)

 

 

constant -0.639 * -0.188 -0.189
- (046) (252) (249)
Pseudo R2 0.026 0.036 0.034
Number of observations 653 684 684
Prob > chi2 (.003) (.000) (.000)

 

*** denotes signiﬁcance at the 1.0% level
** denotes signiﬁcance at the 5.0% level
* denotes signiﬁcance at the 10.0% level

89

Table 1.8, Continued
Multivariate Logit Results (% of votes owned)

 

Logit results relating the likelihood a ﬁrm is acquired to a set of explanatory variables
including ownership structure and ﬁnancial characteristics of a ﬁrm. Dependent variable
= 1 if the ﬁrm was acquired and 0 if it was not acquired. (pcvalues in parenthesis)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(4) (5)
% of votes owned by:
inside directors / executives -0.897 **
(.029)
all outside directors 1.934 *
(086)
outside directors
gray outside directors
unafﬂiated blockholders 1.450 ***
(.000)
aﬂliated blockholders -0.3 52
(.690)
inside directors / executives -0.867 **
/ affiliated blockholders (.023)
all outside directors 1.380 ***
/ unafﬁlitate blockholers (.000)
% return on equity 0.178 *
(.094)
growth-resource dummy 0.196
(.301)
sales growth -0.433 **
(.067)
liquidity -1.240 ***
(.010)
leverage -0.024
(.491)
ln of assets 0.070
(.228)
market to book ratio 0.002
(.869)
price to earnings ratio 0.001
(.419)

 

90

Table 1.8, Continued
Multivariate Logit Results (% of votes owned)

 

Logit results relating the likelihood a fum is acquired to a set of explanatory variables
including ownership structure and ﬁnancial characteristics of a ﬁrm. Dependent variable
= 1 if the ﬁrm was acquired and 0 if it was not acquired. (p—values in parenthesis)

 

 

(4) (5)
constant -0.185 -0.768 *
(.259) (068)
Pseudo R2 0.032 0.058
Number of observations 684 653
Prob > chi2 (.000) (.000)

 

*** denotes signiﬁcance at the 1.0% level
** denotes signiﬁcance at the 5.0% level

* denotes signiﬁcance at the 10.0% level

91

Table 1.9

Additional Multivariate Logit Results (% of stock owned)

 

Logit results relating the likelihood a ﬁrm is acquired to a set of explanatory variables
including ownership structure and ﬁnancial characteristics of a ﬁrm. Dependent variable
= 1 if the ﬁrm was acquired and 0 if it was not acquired. (p-values in parenthesis)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

( 1) (2)
% of stock owned by:
inside directors / executives -0.903 * -l .007 **
(.070) (.035)
all outside directors 2.018 * 2.037 *
(.074) (.070)
unafﬂiated blockholders 1.520 ***
(.000)
aﬁliated blockholders -0.772 -0.663
(.477) (.533)
unafﬂiated blockholders 1.904 **
repr. on board (.000)
unafﬂiated blockholders 0.564
not repr. on board (.374)
% return on equity 0.183 * 0.180 *
(.081) (.098)
growth-resource dummy 0.212 0.202
(.269) (.290)
sales growth -0.485 * -0.453 *
(.053) (.059)
liquidity -1.118** -1.220***
(.019) (.010)
leverage -0.026 -0.022
(448) (.528)
In of assets 0.101 0.056
(.115) (.341)
market to book ratio 0.002 0.001
(.839) (.905)
age of top executive -0.006
(588)
# of years the tOp exec. -0.078
served as a director (.447)
Indicator: one man 0.089
top management team (.646)
# of other directorships -0.156 *
held by inside directors (.068)

 

92

Table 1.9, Continued
Additional Multivariate Logit Results (% of stock owned)

 

Logit results relating the likelihood a ﬁrm is acquired to a set of explanatory variables
including ownership structure and ﬁnancial characteristics of a ﬁrm. Dependent variable
= 1 if the ﬁrm was acquired and 0 if it was not acquired. (p-values in parenthesis)

 

 

 

 

(1) (2)
# of other directorships 0.043
held by outside directors (.552)
constant -0.486 -0.605
(.481) (.167)
Pseudo R2 0.066 0.064
Number of observations 651 653
Prob > chi2 (.000) (.000)

 

*** denotes signiﬁcance at the 1.0% level
** denotes signiﬁcance at the 5.0% level
* denotes signiﬁcance at the 10.0% level

93

Table 1.10

Hypotheses Results Compared

 

 

 

Mikkelson Song &
Essay 1 & Partch Walkling
(2000) (1990) (1993)
[Sample Years 1990 - 1997 1972 - 1983 1977 - 1986
No No No
Sample Notes fmancials ﬁnancials ﬁnancials

or utilities

or utilities

or utilities

 

Fin. Incentive

no evidence

no evidence

no evidence

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Entrenchment evidence no evidence evrdence
Discipline no evidence n.a. n.a.
Monitoring evidence n.a. n.a.
Irrelevance no evidence evidence no evidence
G-Resource no evidence n.a. n.a.
Size n.a. evidence ”We
Mkt-to-Book no evidence n.a. no evidence
P / E no evidence n.a. no evidence

 

94

 

 

Table 1.10, Continued
Hypotheses Results Compared

 

 

 

 

 

 

 

 

 

 

 

 

Hadlock
Palepu Shiv. et. all
(1986) (1993) (1999)
Sample Years 1971 - 1979 1980 - 1988 1982 - 1992
No hostile banks
Sample Notes ﬁnancials targets only
or utilities only
Fin. Incentive n.a. no evidence no evidence
Entrenchment n.a. evidence evidence
Discipline n.a. no evidence no evidence
iiMonitoring n.a. evidence * no evidence
Irrelevance no evidence no evidence no evidence
G-Resource evidence n.a. n.a.
Size evidence n.a. no evidence
Mkt-to-Book no evidence - evidence 12 n.a.
P / E no evidence n.a. n.a.

 

 

 

 

 

 

95

APPENDIX 2

TABLES OF ESSAY 2

96

Table 2.1
Merger Classiﬁcation: Hostile vs. Non-Hostile

 

Total successfully merged ﬁrms from a population of ﬁrms available from 1990 to 1997.
Population includes 1.) ﬁrms with CRSP and COMP USTAT data available, 2.) NYSE and
AMEX listed ﬁrms only, 3.) United States ﬁrms only, 4.) ADR’s deleted, 5.) limited
partnerships deleted, 6.) SIC’s 4900 - 4999 (utilities) and SIC’s 6000 - 67999 (ﬁnancials)

deleted. Hostile classiﬁcation is based on Morck, Shleifer, and Vishny (1988)

 

 

 

 

methodology.
Successfully Merged Firms

Non- Non-

Total Hostile hostile Hostile hostile

1990 43 4 39 9.3% 90.7%
1991 16 2 14 12.5% 87.5%
1992 23 2 21 8.7% 91.3%
1993 26 2 24 7.7% 92.3%
1994 40 3 37 7.5% 92.5%
1995 58 8 50 13.8% 86.2%
1996 70 5 65 7.1% 92.9%
1997 98 7 91 7.1% 92.9%
Total 374 33 341 8.8% 91 .2%

 

97

Table 2.2
Population Summary, 1990 to 1997

 

Population of ﬁrms available from 1990 to 1997. Population includes 1.) ﬁrms with
CRSP and C 0MPUSTAT data available, 2.) NYSE and AMEX listed firms only, 3.)
United States ﬁrms only, 4.) ADR’s deleted, 5.) limited partnerships deleted, 6.) SIC’s
4900 - 4999 (utilities) and SIC’s 6000 - 67999 (ﬁnancials) deleted.

 

 

Firms delisted Mergers as a %

Total listed ﬁrms due to merger of total listed ﬁrms

 

NYSE AMEX Tot. NYSE AMEX Total NYSE AMEX Tot.

 

 

 

1990 1244 400 1644 28 15 43 2.3% 3.8% 2.6%
1991 1283 400 1683 ll 5 16 0.9% 1.3% 1.0%
1992 1345 417 1762 12 11 23 0.9% 2.6% 1.3%
1993 1396 431 1827 19 7 26 1.4% 1.6% 1.4%
1994 1438 428 1866 25 15 40 1.7% 3.5% 2.1%
1995 1518 445 1963 44 14 58 2.9% 3.1% 3.0%
1996 1495 437 1932 51 19 70 3.4% 4.3% 3.6%
1997 1422 414 1836 78 20 98 5.5% 4.8% 5.3%
Avg. 1393 422 1814 34 13 47 2.4% 3.1% 2.6%
268 106 374 Total Mergers

113 Incomplete data avail.
25 >50% toehold

 

236 Final Sample

 

 

98

Table 2.3
Top Manager Turnover for Target Firms: Turnover Frequency, 1990 - 1997

 

Frequency distribution of changes in the top manager, either CEO or President, for the 5
years prior to the announcement of the takeover and for the 2 years following completion
of the takeover of 236 targets over the period 1990 - 1997.

 

 

 

Time Period Rate (Number) of
Relative to Change(s)
Takeover in the Top Manager
Year -5 8.1% (19)
Year -4 10.6% (25)
Year -3 14.0% (33)
Year -2 13.6% (32)
Year -1 10.6% (25)
Year +1 69.5% (161)
Year +2 8.1% (19)

 

Note: Years -5 through -1 are the years preceding the announcement of the merger. Year
+1 begins with the announcement of the merger and ends 12 months after the completion
of the merger. Year +2 begins 1 year after the completion of the merger and continues
for 12 months. The turnover rate is calculated by the number of top manager changes in
a year divided by the sample size, 236.

99

Table 2.4
Top Manager Turnover for Target Firms: Reasons for Turnover, 1990 - 1997

 

Frequency distribution of reasons for departures of the top managers in 236 successful
takeover targets, pre and post merger as indicated in Dow Jones Interactive. Pre-merger
includes years -5, -4, -3, -2, and -1. Post-merger includes years +1 and +2. Post-merger
only includes managers that were in place in the target at the time of the announcement of
merger.

 

Percent (Number) of Top Managers Departing

 

 

Pre-merger Post-merger
Reason Cited (5 years) (2 years)
Normal retirement 26.9% (36) 7.6% (13)
Accepted high-level position
in ﬁrm (mainly chairman) 22.4% (30) 7.0% (12)
Accepted lower-level position
in ﬁrm 3.7% (5) 4.7% (8)
Died 2.2% (3) 0.0% (0)
Accepted directorship in ﬁrm 2.2% (3) 5.3% (9)
Total Non-disciplinary 57.5% (77) 24.6% (42)
Change in control 3.0% (4) 39.2% (62)
Fired, poor performance cited 3.7% (5) 1.2% (2)
Policy difference 0.7% (l) 0.6% (1)
Early retirement 1.5% (2) 0.0% (0)
Other personal or business interests 22.4% (30) 9.4% (16)
Took similar position with
another ﬁrm 5.2% (7) 25.7% (44)
No reason given 6.0% (8) 2.3% (4)
Total Disciplinary 42.5% (57) 75.4% (129)
Total 100% (134) 100% (171)

 

100

 

Table 2.5
Top Manager Turnover for Target Firms: Replacement Manager Origin, 1990 -
1997

 

Frequency distribution of top managers arriving at 236 successful takeover target ﬁrms
classiﬁed according to previous afﬁliation. Pre-merger includes years -5, -4, -3, -2, and -
1. Post-merger includes years +1 and +2. Post-merger only includes managers that were
in place in the target at the time of the announcement of merger. (p-values in parenthesis
for difference)

 

 

 

 

 

 

Percent (N umber)
Total Number Of Top Managers
Of Changes In Replaced By:
Top Manager Outsider Insider
Pre-merger 1 34 3 7.3% 62.7%
(50) (84)
Post-merger l 71 92.4% 7.6%
(158) (13)
Difference 55.1% -55.1%
(.00).... (.00)...

 

 

***indicates signiﬁcantly different from zero at the 0.01 level.
Note: An outsider is an individual who was not employed by the target ﬁrm at the time he

assumed the top manager position. An insider is an individual who was employed by the
target ﬁrm at the time he assumed the top manager position.

101

Table 2.6
Turnover Characteristics: Pre-merger vs. Post-merger

 

Turnover characteristics of the top manager turnover at 236 successful takeover target
ﬁrms. Pre-merger includes years -5, -4, -3, -2, and -1. Post-merger includes years +1 and
+2. Post-merger only includes managers that were in place in the target at the time of the

announcement of merger. (p-values in parentheses)

 

Top Manager Turnover

 

 

 

 

 

 

 

 

Pre-merger Post-merger
(N = 134)8 (N = 171)“ Difference
Age of departing top manager 57.93 55.50 2.43”
(.01)
Age of replacement top manager 50.47 52.76 -2.29.
(.01)
Age differential (departing minus
replacement) 7.56 2.73 4.84
(00)." (.00)... (.00)...
Tenure of departing top manager 9.78 9.25 .53
(00).“ (00)” (.63)
Top manager remains with ﬁrm after
turnover event (includes different position
with ﬁrm, director, and chairman) .3209 .0290 .2930
(00)” (02)” (00)”
Regime change. Dummy variable = 1 if both
the titles of CEO and chairman turnover
concurrently, 0 otherwise. .4104 .8953 -.4849
(.00)... (.00)... (.00)...
Top manager held both titles CEO and
Chairman before turnover event. Dummy
variable = 1 if manager holds both titles, 0 .6716 .6802 -.0086
otherwise.
(00)“ (00)” (.87)

 

 

a Not every pre-merger statistic has 134 observations and not every post-merger statistic
has 171 observations due to missing data.

***indicates signiﬁcantly different from zero at the 0.01 level.

** indicates signiﬁcantly different ﬁom zero at the 0.05 level.
* indicates signiﬁcantly different from zero at the 0.10 level.

102

Table 2.7
Pre-Takeover Cumulative Abnormal Returns for Disciplinary and Nondisciplinary
Takeover Targets

 

Cumulative average industry-adjusted returns (CIAR) and cumulative average prediction
errors (CPE) over months ~48 through -3 relative to the announcement of the takeover of
236 successful takeover target ﬁrms classiﬁed according to turnover in the top manager
during the 2 years following completing of takeovers occurring between 1990 and 1997.
The top manager is deﬁned as the individual occupying the CEO position, or, if the ﬁrm

does not have a CEO position, the presidency. (p-values in parentheses)

 

 

 

 

 

 

 

 

Number
Of CIAR CPE
Sample Firms (%) (%)
Full sample of takeover targets 236 2.84 ~3.90
(.60) (.75)
Turnover Classiﬁcation
Takeover targets with turnover in the top manager 171 ~3.92 ~11.30
(.50) (.25)
Takeover targets without turnover in the top
manager 65 20.67. 1 5.64
(.04) (.29)
Difference 24.59 26.94
(03)” ( 14)
Replacement Classiﬁcation
Takeover target with top manager replaced by an
outsider 157 ~5.00 ~10.13
(.42) (.33)
Takeover target with no turnover or top manager
replaced by an insider 79 18.43 8.54
(04)" (.52)
Difference 23.43 18.67
(03)” (.29)
Reason Classiﬁcation
Takeover targets with reasons indicated as
disciplinary. 128 ~3.73 ~9.00
(.60) (.44)
Takeover targets with no turnover in top manager
or turnover with reasons indicated as nondiscp. 108 10.64 2.19
(.14) (.85)
Difference 14.37 1 1.18
(.16) (.50)

 

***indicates signiﬁcantly different from zero at the 0.01 level.
** indicates signiﬁcantly different from zero at the 0.05 level.
* indicates signiﬁcantly different from zero at the 0.10 level.

103

".‘Y

I."

Table 2.8
Cumulative Industry-Adjusted Returns and Cumulative Average Prediction Errors
for Various Time Periods Relative to the Merger Announcement.

 

(p-values in parentheses)

 

 

CIAR (in percent) CPE (in percent)
Non- Non-
Time Discp. Discp. Discp. Discp.
Period Sample Sample Sample Sample
(N=171) (N=65) Diff. (N=171) (N=65) Diff.
'1']
~48 to ~3 ~3.92 20.61 24.59 ~11.29 15.65 26.94
(.50) (.04) (.03) (.25) (.28) (.14) ..
~48 to ~25 6.03 10.92 4.61 0.57 12.05 11.48
(.22) (.11) (.62) (.94) (.21) (.37)
~24 to -3 ~91]. 9.75 18.81 ~11.119 4.11 15.20
(.03)’ (.13) (.01) (04) (.58) (12)
~12 to ~3 ~3.58 3.80 7.39 624. ~1.12 5.11
(.16) (.36) (.13) (.04) (.80) (.36)

 

***indicates signiﬁcantly different from zero at the 0.01 level.
** indicates signiﬁcantly different from zero at the 0.05 level.
* indicates signiﬁcantly different ﬁ'om zero at the 0.10 level.

104

Table 2.9
Post-merger: Top Manger Remains or Departs

 

Top manager characteristics and ﬁrm ﬁnancial characteristics at 236 successful takeover
target ﬁrms. Targets are sub-divided base on if the top manager (immediately preceding
the merger) remains on with the new ﬁrm or departs post-merger. Post-merger includes

years +1 and +2. (p-values in parentheses)

 

 

 

 

 

 

Target Target
Firrn’s Firm’s
Top Top
Manager Manager
Remains Departs
(N = 65) (N = 17]) Difference
Top Manager Characteristics
Age, pre-merger 55.66 55.12 .54
(.73)
Tenure, pre-merger 10.91 8.98 1.93
(.18)
% of targets stock owned 8.37 5.87 2.49
(.14)
Cash compensation (1996 dollars) 93 7,729.6 707,264.3 230,465.3
(.09)‘
Annual salary, pre-merger (1996 dollars) 608,245.6 448,975.9 159,269.7
(03)"
Annual bonus, pre-merger (1996 dollars) 324,415.0 244,694.2 79,720.7
(.42)
Target Firm Financial Characteristics
log of total ﬁrm value (1997 dollars) 5.47 5.32 .15
(.56)
log of total assets (1997 dollars) 5.81 5.73 .08
(.73)
free cash ﬂow to assets .0879 .0694 .0184
(.09)‘
return on equity (3 year average) .0898 ~.O252 .1150
(.09)‘
dummy variable = 1 if ﬁrm has a investment
grade bond, 0 otherwise .5380 .7250 ~.1870
(.01 )‘Ct

 

***indicates signiﬁcantly different from zero at the 0.01 level.
** indicates signiﬁcantly different from zero at the 0.05 level.
* indicates signiﬁcantly different from zero at the 0.10 level.

105

Table 2.10

Post-merger: Top Manger Remains or Departs, Continued

 

F irrn characteristics and takeover characteristics at 236 successful takeover target ﬁrms.
Targets are sub-divided based on if the top manager (immediately preceding the merger)
remains on with the new ﬁrm or departs post-merger. Post-merger includes years +1 and

+2. (p-values in parentheses)

 

 

 

 

 

 

Target ﬁrm’s Target frrrn’s
Top Top
Manager Manager
Remains Leaves
(N = 65) (N = 171) Diff.
Target Firm Other Characteristics
Staggered voting for board of directors
exists. Dummy variable = 1 if staggered
voting, 0 otherwise .6308 .4854 .1454.
(-04)
Two classes of voting shares exist. Dummy
variable = 1 if two classes exist, 0 otherwise .1385 .0994 .0390
(.39)
% of board members classiﬁed as insiders 29.2 28.6 .006
(.79)
% of board members classiﬁed as outsiders 63.4 67.2 ~3.8
(. 1 3)
% of board members classiﬁed as gray 7.4 4.2 3.2"
(.01)
Takeover Characteristics
Takeover classiﬁed as hostile. Dummy
variable =1 if hostile, 0 otherwise. .0615 .1053 ~.0437
(.30)
Takeover with multiple bidders. Dummy
variable = 1 if multiple bidders, 0 otherwise .1846 .2398 ~.0552
(.37)
Foreign acquisition in the US. Dummy
variable = 1 if acquirer is a foreign ﬁrm, 0
otherwise. .1230 .1400 ~.0170
(.73)
Toe holder. Dummy variable = 1 if
acquiring ﬁrm had toe hold prior to merger .1231 .0292 .0938
announcement
(00)”

 

***indicates signiﬁcantly different from zero at the 0.01 level.
** indicates signiﬁcantly different from zero at the 0.05 level.
* indicates signiﬁcantly different from zero at the 0.10 level.

106

APPENDIX 3

DATA DEFINITIONS

107

DATA DEFINITIONS

Stock ownership ~ is the sum of shares owned and options exercisable within 60 days for

all classes of stock, divided by total shares outstanding by the ﬁrm.

Voting ownership ~ is the sum of shares owned and options exercisable within 60 days

multiplied by votes per share held, divided by the total votes outstanding by the ﬁrm.

a.-
.‘l .
‘:

Firm value (M&P) = (current liabilities + long-term debt + book value of prefered stock

+ market value of common stock). This variable is adjusted to 1997 dollars.

Total Assets (Palepu) = In of (total assets). This variable is adjusted to 1997 dollars.

Leverage (M&P) = (current liabilities + long-term debt)/(total debt + book value of

preferred stock + market value of common stock)

Net plant, property and equipment to total assets (Billett) = (net plant, property and

equipment)/ (total assets)

Free cash flow to assets (Billett) = (operating income before depreciation ~ interest

expense ~ common stock dividends - preferred stock dividends - taxes + deferred

taxes)/(total assets)

108

Leverage (Palepu) = (long-terrn debt)/(book value of common equity + book value of
preferred stock); The leverage ratio is computed for three ﬁscal years prior to the
observation year, and the average is used as the leverage variable. If three years are not

available, two years is averaged. If two years is not available, one year is utilized.

Liquidity (Palepu) = (cash and cash equivalents ~ current liabilities)/(total assets); The
liquidin ratio is computed for three ﬁscal years prior to the observation year, and the
average is used as the liquidity variable. If three years are not available, two years is

averaged. If two years is not available, one year is utilized.

Sales growth (Palepu) = [(netsales,)/(netsales,-1) ~1]; Sales growth is computed for three
ﬁscal years prior to the observation year, and the average is used as the sales grth
variable. If three years are not available, two years is averaged. If two years is not

available, one year is utilized.

Growth resource dummy (Palepu) - The growth-resource dummy is a 0/1 variable
deﬁned on the basis of its three variables; growth, liquidity and leverage deﬁned above.
The dummy variable is assigned a value one if the ﬁrm has a combination of either low
growth-high liquidity-low leverage or high growth-low liquidity-high leverage. The
dummy is set to zero for all the other combinations. Each of the three variables growth,
liquidity and leverage is deﬁned as 'high' if its value for a ﬁrm is larger than the average

for all the C 0MPUS TAT ﬁrms, otherwise, it is deﬁned as 'low'.

109

Return on equity (Palepu) = (income before extraordinary adjustrnents)/(book value of
common stock + book value of preferred stock); The return on equity ratio is computed
for three ﬁscal years prior to the observation year, and the average is used as the return on
equity variable. If three years are not available, two years is averaged. If two years is not

available, one year is utilized.

Market value of common equity = (common stock price at end of calendar

year*common shares outstanding). This variable is adjusted to 1997 dollars.

Market-to-book value (Palepu) = (common stock price at end of ﬁscal year * common

shares outstanding)/(book value of common equity)

Price-to-earnings ratio (Palepu) = (common stock price at end of ﬁscal year)/[(net

income)/(common shares outstanding)]

110

 

BIBLIOGRAPHY

111

BIBLIOGRAPHY

Berger, P. and E. Ofek, 1995, "Diversiﬁcation's effect on ﬁrm value, " Journal of
Financial Economics 3 7, 1995.

Berger, P. and E. Ofek, 1996, "Bustup Takeovers of Value-Destroying Firms, " Journal of
Finance Vol. LI, No 4, 1996.

Billett, M., 1996, "Targeting Capital Structure: The Relationship between Risky Debt and
the F irm's Likelihood of Being Acquired," Journal of Business, vol. 69, no. 2, 173~
192.

Borokhovich, K.A., K. Brunarski, and R. Parrino, 1997, "CEO Contracting and
Antitakeover Amendments," Journal of Finance Vol. LII, No. 4, 1997, 1495 ~1517.

Brown, S. and J. Warner, 1980, “Measuring Security Price Performance,” Journal of
Financial Economics 8, 205 ~ 258.

Comment, R. and G.W. Schwert, 1995, "Poison or placebo? Evidence on the deterrence
and wealth effects of modern antitakeover measures," Journal of Financial
Economics 39, 1995.

Cotter, J. and M. Zenner, 1994, "How managerial wealth affects the tender offer process,"
Journal of Financial Economics 35, 63-97.

Demsetz, J ., and K. Lehn, 1985, "The Structure of Corporate Ownership: Cause and
Consequences," Journal of Political Economy 93, 1 155-1 177.

Dodd, P. and R. Ruback, 1977, “Tender offers and stockholder returns: An empirical
analysis,” Journal of Financial Economics 5, 351 ~ 373.

Furtado, E.P.H. and MS. Rozeff, 1987, “The wealth effect of company initiated
management changes,” Journal of Financial Economics 5, 351 ~ 373.

Hadlock, C., J. Houston, M. Ryngaert, 1999, “The Role of Managerial Incentives in Bank
Acquisitions,” Journal of Banking and Finance, 1999.

Jarrell, G. and M. Bradley, 1980, “The economic effects of federal and state regulation of
cash tender offers,” Journal of Law and Economics 23, 371 ~ 407.

Jensen, M.C., and W. Meckling, 1976, “Theory of the Firm: Managerial Behavior,
Agency Costs, and Ownership Structure," Journal of Financial Economics 3, 305-
360.

Kummer, D. and J. Hofﬁneister, 1978, “Valuation consequences of cash tender offers,”

112

Journal of Finance 33, 505 - 516.

Martin, K. J. and J. McConnell, 1991, “Corporate Performance, Corporate Takeovers,
and Management Turnover," The Journal of Finance, June 1991, 671 ~ 687.

Mikkelson, W. and M. Partch, 1989, "Managers' Voting Rights and Corporate Control,"
Journal of Financial Economics 25, 263-290.

Mikkelson, W. and M. Partch, 1997, “The decline of takeovers and disciplinary
managerial turnover”, Journal of Financial Economics 44, 205-228.

Morck, R., A. Shleifer, and R. Vishny, 1988a, "Characteristics of targets of hostile and l‘“
ﬁiendly takeovers," in: Alan J Auerbach, Corporate takeovers: Causes and
consequences (University of Chicago Press, Chicago, IL).

Morck, R., A. Shleifer, and R. Vishny, 1988b, "Management ownership and market
valuation: An empirical analysis,” Journal of Financial Economics 20, 293-315.

Morck, R., A. Shleifer, and R. Vishny, 1989, "Alternative Mechanisms for Corporate
Control," The American Economic Review, September 1989, 842 ~ 852.

Palepu, K., 1986, "Predicting takeover targets: a methodological and empirical analysis,"
Journal of Accounting and Economics 8, 3-35.

Reinganum, M., 1985, “The effect of executive succession on stockholder wealth,”
Administrative Science Quarterly 30, 45 ~ 60.

Schwert, G. William, 2000, "Hostility in Takeovers: In the Eyes of the Beholder?,"
Journal of Finance, forthcoming.

Shivdasani, A, 1993, "Board composition, ownership structure and hostile takeovers,"
Journal of Accounting and Economics 16, 167-198.

Song, M. and R. Walkling, 1993, "The impact of managerial ownership on acquisition
attempts and target shareholder weatlh," Journal of Financial and Quantitative
Analysis, 28-4, 439 ~ 457.

Standard & Poor 's Register of Corporations, Directors, and Executives, 1984 ~ 2000,
(Standard & Poor’s Corporation).

Stulz, R.M., "Managerial Control of Voting Rights, Shareholder Wealth and the Market
for Corporate Control," Journal of Financial Economics 20, 1988, 25-54.

Vancil, RF, 1987, Passing the Baton: Managing the Process of CEO Succession,”
(Harvard Business School Press, Cambridge, MA).

113

Walkling, R. and M. Long, 1984, "Agency theory, managerial wealth, and takeover bid
resistance," Rand Journal of Economics 15, 54-68.

Weisbach, M. 1988, "Outside directors and CEO turnover," Journal of Financial
Economics 20, 431-460. '

 

114

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

111)IIITIIIIIMIJII16111