City Research Online
City, University of London Institutional Repository
Citation: Salami, A.R. (1994). Determinants and financial consequences of the method of
payment in corporate acquisitions. (Unpublished Doctoral thesis, City University London)
This is the accepted version of the paper.
This version of the publication may differ from the final published
version.
Permanent repository link:
https://openaccess.city.ac.uk/id/eprint/7544/
Link to published version:
Copyright: City Research Online aims to make research outputs of City,
University of London available to a wider audience. Copyright and Moral
Rights remain with the author(s) and/or copyright holders. URLs from
City Research Online may be freely distributed and linked to.
Reuse: Copies of full items can be used for personal research or study,
educational, or not-for-profit purposes without prior permission or
charge. Provided that the authors, title and full bibliographic details are
credited, a hyperlink and/or URL is given for the original metadata page
and the content is not changed in any way.
City Research Online:
http://openaccess.city.ac.uk/
publications@city.ac.uk
DETERMINANTS AND FINANCIAL CONSEQUENCES OF
TILE METHOD OF PAYMENT IN CORPORATE
ACQUISITIONS
By
AYODELE. R. SALAMI
Submitted for.the degree of
Doctor of Philosophy
City University, London
The research was conducted at:
City University Business School
Department of Business Studies
December 1994
1
Table of Contents.
List of Tables.
I
List of Figures.
V
VI
Acknowl edgeinents.
Declaration
....................VIII
Abstract .......................IX
CHAPTER 1.
OBJECTIVES AND OUTLINE OF THE THES IS .........
1.1. Introduction
1.2. Choice of payment method in corporate
acquisitions
1.3. Impact of the method of payment on
shareholder wealth
1.4. Objectives of the thesis
1.5. Outline of the thesis ...........
1
1
2
13
17
18
CHAPTER 2.
DETERMINANTS OF THE METHOD OF PAYMENT IN MERGERS AND
ACQUISITIONS :- THEORY & EMPIRICAL EVIDENCE. . . . . 20
2.1. Introduction ...............20
2.2. Payment methods available in corporate
acquisitions...............21
2.3. Accounting policy choice and the method of
payment..................28
2.4. Information asymmetry and the method of
payment..................30
2.5. Taxation and the method of payment . . . 37
2.6. Share ownership structure and the method of
payment..................40
2.7. Market conditions and the method of
payment..................47
2.8. Capital structure and the method of
payment..................48
2.9. Growth opportunities in the merging firms
and the method of payment .........50
2.10. Cash resources and the method of payment 52
2.11. Target managerial resistance and the method
ofpayment ................52
2.12. Previous research on the choice of payment
method..................53
2.13. Conclusion ................58
ii
CHAPTER 3.
DETERMINANTS OF THE ACCOUNTING METHOD IN MERGERS AND
ACQUISITIONS :- THEORY & EMPIRICAL EVIDENCE. .
63
3.1. Introduction
63
3.2. UK rules governing the method of accounting
64
for corporate acquisitions
3.3. Implications of UK accounting rules for the
69
choice of accounting policy ........
3.4. Empirical evidence on the accounting policy
81
choice
market
manipulation
and
Earnings
3.5.
90
efficiency
3.6. Managerial considerations and accounting
92
policychoice ...............
93
3.7. Conclusion
CHAPTER 4
INTERACTION BETWEEN ACCOUNTING POLICY AND THE METHOD OF
98
PAYMENT: METHODOLOGY, DATA AND RELATED ISSUES.
98
4.1. Introduction
4 .2. Methodology ................
99
4.3. Control variables ............. 104
123
4.4. The complete simultaneous equations model
124
4.5. Sample selection
136
4.6. Conclusion
CHAPTER 5
INTERACTION BETWEEN ACCOUNTING POLICY AND THE METHOD OF
PAYMENT: RESULTS OF EMPIRICAL ANALYSIS ........140
5 .1. Introduction ...............140
5 .2. Univariate test ..............141
5.3. Determinants of the payment method . .
156
164
5.4. Determinants of the accounting policy .
5.5. Robustness checks .............174
5 .6. Conclusion ................178
CHAPTER 6
TARGET SHAREHOLDERS' CHOICE BETWEEN
CONSIDERATION.
6.1. Introduction
6.2. Theoretical background .
6.3. Definition of variables .
6.4. Sample
6 .5. Results ...........
6.5. Conclusion
iii
CASH AND EQUITY
181
181
185
198
206
209
221
CHAPTER 7
THE IMPACT OF THE METHOD OF PAYMENT ON SHAREHOLDER WEALTH: THEORY AND EMPIRICAL EVIDENCE ...........224
7.1 Introduction ................224
7.2. The impact of the method of payment on
shareholder wealth ............225
7.3. Information asymmetry and shareholder
wealth..................229
7.4. Taxation and shareholder wealth ......233
7.5. Debt co-insurance and shareholder wealth 241
7.6. Other studies on the method of payment and
shareholder wealth ............248
7.7. Conclusion
................250
CHAPTER 8
DETERMINANTS OF THE DIFFERENCE IN THE BID PREMIUM BETWEEN
CASH OFFERS AND EQUITY OFFERS
...........253
8 .1. Introduction
...............253
8 .2. Methodology ................254
8 .3. Data
...................255
8.4. Wealth gains surrounding the bid
announcement...............256
8.5. Role of underwriters ...........293
8.6. Impact of capital gains tax, information
asymmetry and debt
co-insurance on
shareholder wealth ............296
8 .6. Conclusion ................317
APPENDIX 8.1.
EVENT STUDY METHODOLOGY ...............319
CHAPTER 9.
334
SUMMARY, CONCLUSIONS AND IMPLICATIONS.
334
9.1. Introduction
335
9.1. Determinants of the method of payment
337
9.2. Determinants of the accounting policy
9.3. The choice of payment method by target
339
shareholders
9.4. Difference in bid premium between cash and
equityoffers ............... 344
9.5. Issues for further research ........ 344
BIBLIOGRAPHY .....................347
iv
List of Tables.
Number
Title
Page No
2.1.
Summary of factors influencing the method
of payment in corporate acquisitions
59
3.1.
Summary of studies testing the impact of
goodwill on the choice of accounting
method
87
3.2.
Summary of factors influencing the choice
of accounting policy in corporate
acquisitions
95
4.1.
Definition of control variables used in
the simultaneous equations model
124
4.2.
Sample distribution by payment method and
accounting method
128
4.3
Descriptive statistics for the control
variables used in the simultaneous
equations model
133
4.4.
Pearson correlation coefficients among
the explanatory/control variables
135
A4 . 1.
Distribution
of
the
sample
by
announcement year and method of payment
137
A4 .2.
Distribution of the sample by the
proportion of equity in each acquisition
138
A4.3.
Distribution of the sample by year and
accounting method
139
5.1.
Difference in means between the "equity
group" and the "cash group": Tinivariate
tests
143
5.2.
Difference in means between the "merger
group" and the "acquisition group":
Univariate tests
150
5.3.
Accounting method employed partitioned by
the ability of the bidder to use merger
accounting
153
5.4.
Difference in means between the "merger
group" and the "qualified acquisition
accounting group": Univariate tests
154
5.5.
Maximum likelihood estimates of Two Limit
Tobit models explaining the proportion of
equity in the method of payment
158
I
5.6.
Maximum likelihood estimates of 2 group
Logit regressions discriminating between
the "merger group" and the "acquisition
group"
165
5.7.
Classification matrix for the Logit
regressions presented in Table 5.6.
169
5.8.
Maximum likelihood estimates of 2 group
Logit regressions discriminating between
the "merger group" and the "qualified
acquisition accounting group"
172
5.9.
Classification matric for the Logit
regressions presented in Table 5.8.
173
5.10.
Maximum likelihood estimates of 2 group
Logit model of the choice of the payment
method based on the reduced sample of
observations either "all equity" or "all
cash"
175
5.11.
Maximum likelihood estimates of 2 group
Logit regressions discriminating between
the "merger group" and the "acquisition
group" based on the reduced sample of
observations either "all equity" or "all
cash"
176
6.1.
Choice of payment medium by target
shareholders: Definition of explanatory
variables
207
6.2.
Frequency distribution of the variable
PROEQUI
208
6.3.
Value difference between equity offers
and cash offers
210
6.4.
Impact of value difference between cash
and equity offers on target shareholder
choice of payment method
212
6.5.
Impact of the economic fundamentals of
the acquisition on target shareholder
choice of payment method
214
6.6.
Impact of bid dynamics on target
shareholders choice of payment method
215
6.7.
Maximum likelihood estimates of Logit
models explaining the proportion of
target accepting the equity offer
217
II
6.8.
Regression of the value difference
between the cash and equity offer on
explanatory variables
219
7.1.
Cumulative abnormal returns to target
shareholders around the acquisition
announcement date partitioned by the
method of payment
227
7.2.
Cumulative abnormal returns to bidder
shareholders around the acquisition
announcement date partitioned by the
method of payment
228
7.3.
The effect of variance changes on the
value of debt and equity
243
8.1.
Cumulative abnormal returns to targets
over different windows surrounding bid
announcements
257
8.2.
Cumulative abnormal returns to targets
over the period -40 to +40 days by the
method of payment
263
8.3.
Pairwise comparison of the target's
abnormal return over the period -40 to
+40 days for different methods of payment
265
8.4.
Cumulative abnormal returns to targets
over the period -40 to -1 days by the
method of payment
270
8.5.
Pairwise comparison of the target's
abnormal return over the period -40 to -1
days f or different methods of payment
271
8.6.
Cumulative abnormal returns to targets
over the period +1 to +40 days by the
method of payment
272
8.7.
Pairwise comparison of the target's
abnormal return over the period +1 to +40
days by the method of payment
273
8.8.
Cumulative abnormal returns to bidders
over different windows surrounding bid
announcements
274
8.9.
Cumulative abnormal returns to bidders
over the period -40 to ^40 days by the
method of payment
282
8 . 10
Pairwise comparison of the bidder's
abnormal return over the period -40 to
+40 days for different methods of payment
283
III
8.11
Cumulative abnormal returns to bidders
over the period -40 to -1 days by the
method of payment
289
8.12.
Pairwise comparison of the bidder's
abnormal return over the period -40 to -1
days for different methods of payment
290
8.13.
Cumulative abnormal returns to bidders
over the period +1 to ^40 days by the
method of payment
291
8.14.
Pairwise comparison of the bidder's
abnormal return over tie period -40 to -1
days for different methods of payment
292
8 . 15.
Cumulative abnormal returns to bidders
using "equity offers with a cash
alternative" as the method of payment
294
8 . 16.
Explanatory variables influencing the
wealth of shareholders in takeovers
303
8 . 17.
Impact of the method of payment on
shareholder wealth: Descriptive
statistics for the explanatory variables
304
8 . 18.
Difference in means test for the
variables explaining the impact of the
method of payment on shareholder wealth
305
8 . 19.
Regression of the cumulative abnormal
returns (CAR) to the target on
explanatory variables
309
8.20.
Regression of the cumulative abnormal
returns (CAR) to the bidder on
explanatory variables
314
A8.1.
Average betas using different procedures
to correct for thin trading
328
A8.2.
Number of event days between the
announcement date and the unconditional
date by method of payment
331
A8.3.
Pairwise comparison of the number of
event days between the announcement date
and the unconditional date for different
methods of payment.
333
Iv
List of Figures.
Number
Title
Page No
8.1.
Daily abnormal returns for targets
(Dimson Model)
259
8.2.
Daily abnormal returns for targets
(Market adjusted Model)
260
8.3.
Daily abnormal returns for targets (Mean
adjusted Model)
261
8.4.
Cumulative abnormal returns for targets
262
8.5.
CAR for targets by method of payment
(Dimson Model)
267
8.6.
CAR for targets by method of payment
(Market Adjusted Model)
268
8.7.
CAR for targets by method of payment
(Mean Adjusted Model)
269
8.8.
Daily abnormal returns for bidders
(Dimson Model)
277
8.9.
Daily abnormal returns for bidders
(Market Adjusted Model)
278
8 .10.
Daily abnormal returns for bidders (Mean
Adjusted Model)
279
8 . 11.
Cumulative abnormal returns for bidders
280
8 . 12.
CAR for bidders by method of payment
(Dimson Model)
285
8 . 13.
CAR for bidders by method of payment
(Market Adjusted Model)
286
8 .14.
CAR for bidders by method of payment
(Mean Adjusted Model)
287
V
Acknowledgements.
I am grateful to several people for their generous
assistance and help in the course of my research. In
particular my mentor, Dr Sudi Sudarsanam, who supervised
this work. His suggestions and criticisms of the method of
analysis have greatly improved the quality of this work.
Without his friendship and support, I doubt if I could have
successfully coped with the strains and difficulties
involved in completing this project.
I acknowledge the helpful cmments of colleagues and
members of staff at City University Business School,
especially Professor R.J. Taffler, Professor G. Gemmill,
Professor R. Batchelor, Professor M. Levis, Dr P. Holl, Mr
D. Citron, Mr D. Thomas, Dr M. Lasfer, Mr A. Mahate, Dr D.
Kyriazis.
I express my thanks to Extel Financial Limited (for
granting me access to their library), to City University
Business School and the Chartered Accountants Trust for
Education and Research (for providing financial assistance
towards this project)
I also wish to thank Christopher Burley and Demetra
Kalogerou for the encouragement and succour which their
companionship has provided me throughout all the stages of
my research.
Finally in recognition and appreciation of the immense and
encouraging support of my family and in particular my
parents, this thesis is dedicated to my father and the
memory of my mother.
VI
To my father and the memory of my mother
VII
Declaration.
I grant powers of discretion to the University
Librarian to allow this thesis to be copied in whole or in
part without further reference to me. This permission
covers only single copies made f or study purposes, subject
to normal conditions of acknowledgement.
VIII
Abstract.
Although wide variation in the type of consideration
offered in corporate acquisitions is observed in practice,
little is known about how bidders or targets choose the
method of payment in takeovers. Further, several empirical
studies report that shareholders of both targets and
bidders earn higher returns in cash offers than in equity
offers, but the reasons for this more favourable impact of
cash offers have not been empirically established.
This study attempts to fill these gaps in the
literature by addressing three research questions:1) What factors determine the method of payment used
by bidders in corporate acquisitions?
2) How do target shareholders choose between cash and
equity when the bidder has offered "equity with a cash
alternative" as the method of payment?
3) Why are bid premia higher in cash offers than in
equity offers?
In examining how bidders choose the method of payment
this study in contrast to all previous studies, emphasises
that there is a simultaneous and joint relationship between
the method of payment and the choice of accounting policy.
Accordingly, we adopt a simultaneous equations framework
with payment method and accounting policy choices as
endogenous variables. Our results show that payment method
has a significant impact on accounting policy choice
whereas the reciprocal effect is not significant. This
result reflects the fact that UK accounting rules have
eroded the distinction between merger and acquisition
accounting which is more clearly observed in the US.
We study how target shareholders choose the payment
currency by examining how a choice is made between cash and
equity when the bidder has offered "equity with a cash
alternative" as the method of payment. We find that
information about the opportunities for realising synergies
in the acquisition have no influence on the choice of
payment method by target shareholders. The choice is based
primarily on the difference in value between the cash and
equity offers. This is consistent with the theoretical
predictions based on the efficient market hypothesis, that
all publicly available information about a security can be
reduced into a single index, namely the share price.
We tested some of the popular explanations which have
been advanced in the literature to explain the higher
returns to cash offers. The capital gains tax compensation
and the wealth redistribution hypotheses are rejected.
Information asymmetry between managers and shareholders can
explain some of the higher returns observed in cash offers.
This is consistent with signalling models which predict
that the use of equity to finance investments signals a
belief by managers that their shares are overvalued.
Ix
CH.APTER 1.
OBJECTIVES AND OUTLINE OF THE THESIS.
1.1. Introduction
While there is still an unresolved controversy on
whether corporate acquisitions are a good or bad
phenomenon, it is unarguably true that takeovers have now
become an integral part of the Anglo-Saxon economic
environment. In recognition of the importance of takeovers
as part of the economic landscape, a significant amount of
academic research in financial economics, industrial
economics, organisational theory etc has been devoted to
the study of the causes and consequences of corporate
acquisitions.
Research over the last twenty years has made
substantial progress in enhancing our understanding of
corporate acquisitions. Academic research into takeovers
has established some well documented and fairly robust
results: 1) Shareholders in target firms gain significant
returns in the immediate period surrounding the
announcement of the takeover (Jensen & Ruback, 1983;
Jarrell, Brickley & Netter, 1988)
2) Returns to the target shareholders are higher in
hostile bids than in friendly bids (Franks & Harris, 1989)
3) Returns to the target shareholders are higher in
bids where there are multiple bidders than in bids with a
single bidder (Bradley, Desai & Kim, 1988; Franks & Harris,
1
1989)
4) Shareholder returns are higher in cash offers than
in equity offers f or both targets and bidders (Franks,
Harris & Mayer, 1988).
Despite the large amount of published research on
corporate acquisitions there are still gaps in our
understanding of the causes and consequences of corporate
takeovers. In particular the role of the method of payment
in corporate acquisitions and its impact on the wealth of
shareholders in participating firms are not very well
understood.
Despite the documented result that the returns to the
shareholders of both the bidder and the target are higher
in acquisitions where the method of payment is cash than in
takeovers where the payment currency is equity (see Table
7.1 and 7.2 in Chapter 7 below) the source of this higher
return to cash offers has not yet been identified. A number
of theoretical arguments based on capital gains tax,
information asymmetry, transfer of wealth from shareholders
to bondholders have been proposed to explain the higher
returns to cash offers, but empirical evidence on these
theories is minimal.
1.2. Choice of payment method in corporate acquisitions
Although there are wide variations in the types of
consideration offered in corporate acquisitions' very
1 Franks et al (1988) document in respect of US and UK acquisitions
details of the different forms of consideration offered, the importance
of each form and trends over the 30 year period 1955 - 1985.
2
little is known about how the choice of payment currency in
takeovers is made by bidders.
One of the principal benefits from studying how the
method of payment is chosen in takeovers is that it can
potentially contribute to the debate on whether firms have
any systematic preference for the means of financing
investments. In Modigliani & Miller's (1958), no-tax,
perfect market environment, firms should be indifferent to
the means by which investments are financed. With the
introduction of taxes and market imperfections (Modigliani
& Miller, 1963; Miller, 1977; Myers, 1977; Ross, 1977) the
relationship between financing decisions and the value of
the firm becomes ambiguous. An understanding of any
systematic financing preferences which managers have with
regard to their investment decisions can help us in
explaining the resultant capital structure of the firm.
Unfortunately it is difficult to test empirically, any
preferences which managers have for the means of financing
project type investments, since information on the
financing of individual projects is not publicly available.
However, a corporate acqi.iisition represents a unique
occasion, where the means of financing an investment is
publicly disclosed. Accordingly, in this thesis, we seek to
understand the factors which determine the method of
payment in corporate acquisitions.
Conventional finance theory provides that new
projects must provide a rate of return which is greater
than their risk adjusted cost of capital where the cost of
3
capital is a weighted average of the cost of debt and the
cost of equity (Copeland & Weston, 1983: Chapter 12) . This
implies that acceptance of new projects which influences
the growth rate of the firm is dependent on the proportions
of debt and equity used in financing the projects. In
seeking to explain the choice between equity and debt, a
great deal of the existing literature has focused on the
capital structure of the whole firm (Harris & Raviv, 1991).
Yet if given their individual characteristics, projects are
financed by a mix of debt and equity which is appropriate
for each project then the capital structure of the whole
firm would represent an average of the capital structure of
the firm's portfolio of projects. This implies that a
greater understanding of the financing decisions of firms
can be gained by studying the financing of individual
projects rather than the capital structure of the whole
firm which inherently masks a variety of different capital
structure decisions. We believe that the empirical evidence
which we provide can strengthen our understanding of those
factors which influence the way projects are financed by
firms thereby helping to improve the financial management
and capital budgeting decisions of managers.
There is some theoretical literature which has
examined acquisition financing, but this has been mainly
concerned with the role of taxes and information asymmetry
as determinants of the method of payment (Hansen, 1987;
Eckbo, Giammarino & Heinkel, 1990; Myers & Majiuf, 1984;
Niden, 1988; Berkovitch & Narayanan, 1990)
4
1.2.1. Capital Gains Tax (CGT) and acquisition financing
The principal effect of taxes in an acquisition is the
possible crystallisation of CGT liability. In a cash offer,
the target shareholders are liable for immediate payment of
CGT on any gains realised on the sale of their shares to
the bidder. This tax is levied on the difference between
the price paid by the bidder for the target's shares and
the shareholder's original cost. In an equity offer the
target shareholders can defer the realisation of any CGT
until the subsequent disposal of the shares received from
the bidder.
The ability in an equity offer to defer the
realisation of gains until a future date reduces the
present value of the CGT liability relative to that in a
cash offer. In an efficient market, the higher CGT
liability in a cash offer would result in target
shareholders demanding compensation from the bidder in the
form of a higher bid premium. Bidders wishing to avoid
paying this higher premium would be obliged to offer equity
as the method of payment. Niden (1988) and Higson (1990a)
found no evidence to support the argument that CGT affects
the method of financing acquisitions.
There are at least two reasons why CGT may not have a
significant impact on acquisition financing:1) A number of shareholders are exempt from CGT, in
particular institutional investors such as pension funds,
unit trusts, investment trusts and charitable trusts.
2) There are a number of legitimate ways for tax
S
paying investors to reduce their CGT liability, (i)
shareholders can claim indexation allowance on their gains,
(ii) there is an annual exemption limit of £6,0002 before
private investors become liable to CGT (iii) capital losses
on one investment can be set off against capital gains on
another investment (iv) shares can be held in a tax exempt
form e.g, Personal Equity Plans (PEP)
1.2.2. Information asymmetry and acquisition financing
Where the managers of either the target or the bidder
possess information about the value of the transaction that
is not available to the other party, then the method of
payment offers opportunities for the informed party to
exploit its information advantage.
The bidder's managers may have information about the
true value of their firm which is not reflected in the
current share price (e.g, the opportunities for future
profitable investment) . If the bidder's managers believe
that their shares are overvalued then they have an
incentive to offer equity as the method of payment (Myers
& Majluf, 1984) . Alternatively, if the bidder's managers
have private information about the potential gains
realisable from the acquisition, then they will offer cash
as the method of payment. The advantage of a cash offer is
that it prevents target shareholders from participating in
any post merger gains (Fishman, 1989)
Target managers may have information which is not
2 This
is the exemption limit for 1995/96 financial year.
6
available to the bidder (e.g, the physical condition of
assets, future contractual obligations etc) . Target
managers then have an incentive to deceive by accepting a
cash offer which is greater than the value of the target's
assets. An equity offer makes the returns to the target
shareholders conditional on the future profitability of the
combined firm, and hence protects the bidder from the
"adverse selection" problem 3 (Hansen, 1987; Fishman, 1989)
Hansen (1987) and Smith & Jennings (1993) using US
data tested and rejected the hypothesis that information
asymmetry affected the choice of payment currency. This
issue has so far not been examined in the UK.
1.2.3. Other factors affecting the method of payment
Inexplicably, the literature on acquisition financing
has concentrated on taxes and information asymmetry, to the
exclusion of other factors which are equally likely to
influence the choice of payment currency. In particular,
the relationship between the method of payment and the
choice of accounting policy has not received any attention
in the literature. This thesis argues that the payment
method and the choice of accounting policy are interrelated
and jointly determined.
The two main methods of accounting for a business
combination are acquisition and merger accounting. Although
the choice of accounting method has no impact on the cash
3Adverse selection refers to the risk that a bidder discovers ex
post that it has overvalued the target.
7
flows arising from an acquisition, there are still economic
reasons why managers may have a preference for either
merger or acquisition accounting. Profit related pay
clauses in managerial compensation contracts and debt
covenants in loan agreements are usually calculated by
reference to accounting numbers which can be directly
affected by the choice of accounting policy.
Accounting and statutory regulations in the UK allow
some latitude in how business combinations are accounted
for in the financial statements. As a result of this
flexibility, the method of payment chosen can subsequently
affect how the acquisition is presented in the financial
statements. The inter-relationship between the method of
payment and the accounting for the acquisition in the
financial statements must be considered at the planning
stage in order to take advantage of any benefits which the
latitude in accounting rules presently permit.
The joint and simultaneous determination of the
payment method and the accounting policy arises because at
least 90% of the total consideration offered by the bidder
must be in the form of equity in order to qualify for
merger accounting. The 90% rule would imply that the choice
of accounting method is partly determined by the method of
payment. Similarly the method of payment is partly
determined by the choice of accounting method.
Consequently, in contrast to previous studies, this study
investigates the relationship between the method of payment
and the accounting policy choice decision of the bidder
8
within the framework of a simultaneous equations model
which recognises their mutual dependence.
In modelling the simultaneous relationship between the
method of payment and the choice of accounting policy this
study examines the determinants of the accounting policy in
corporate acquisitions. Most of the previous empirical
evidence on the choice of accounting policy in corporate
acquisitions is US based (Gagnon, 1967; Copeland & Wojdak,
1969; Anderson & Louderback, 1975). However, the US
evidence cannot be directly applied to the UK since the UK
has distinctive accounting rules on business combinations
(eg, writing off goodwill against reserves and the
availability of merger relief) which differentiates the UK
from the US. The evidence presented in this study on the
choice of accounting policy in the UK provides an
opportunity to examine the robustness of the US evidence
(see Chapter 3) to changes in the institutional
environment.
In addition to the choice of accounting policy,
information asymmetry and capital gains tax, there are a
number of other factors which are likely to influence the
type of consideration offered. These include: the capital
structure of the bidder and the target, the effects of the
payment currency on the dilution of existing blockholdings,
the cash resources of the bidder and the target, the
ability of the bidder to raise new funds in the capital
market, the response of the target's management (hostile or
friendly), relative size of the bidder to the target etc.
9
There is limited evidence on the impact of these variables
on the choice of payment method in takeovers (see Table 2.1
below)
1.2.4. Empirical evidence on acquisition financing
A number of studies have examined the determinants of
the method of payment in corporate acquisitions. Hansen
(1987) examined the influence of debt and size on the
method of payment. Amihud, Lev & Travios (1990)
concentrated on the role of insider ownership and size.
Mayer & Walker (1992) examine the role of insider
ownership, size and debt on the method of payment. However
these studies have adopted a fragmented and piecemeal
approach to this research. Individual studies have
concentrated on the influence of one or two specific
factors without controlling for the effects of other
possibly relevant variables.
Higson (1990a) and Smith & Jennings (1993) are two
studies which use a relatively broad range of explanatory
variables to examine the determinants of the method of
payment. However these two studies omit any discussion of
the influence of accounting policy choice on the method of
payment.
Apart from suffering a possible omitted variables
problem, all previous studies adopt a single equation Logit
methodology in the estimation of their models. There are a
number of deficiencies in this approach:
(a) a single equation estimation procedure does not
10
allow for an examination of the joint relationship between
the method of payment and the choice of accounting policy;
(b) the Logit discriminant methodology requires a
binary dependent variable, whereas the payment method is a
continuous variable representing a mix of equity, cash and
debt. Previous studies use selective sampling procedures as
a method of obtaining a binary dependent variable. This
means that only those observations where the method of
payment is "pure cash" or "pure equity" are included in the
sample.
In this study we address some of these inadequacies of
the earlier literature:1) we explicitly model the relationship between the
choice of accounting policy and the method of payment as
jointly and simultaneously determined variables;
2) we use a comprehensive set of explanatory variables
in investigating the choice of payment method in corporate
acquisitions;
3) we use a statistical methodology (i.e, the Two
Limit Tobit Model) which obviates selective sampling and
permits observations with a mixture of cash and equity to
be included in the estimation of the model.
1.2.5. Target shareholders and acquisition financing
The role of target shareholders in determining the
method of payment has been largely ignored in the empirical
11
literature4 . This omission probably results from the fact
that researchers only have access to data on the final
method of payment offered by the bidder. Any concessions
which the target's managers/shareholders may have made to
the bidder in pre-bid negotiations on the method of payment
are not easily observed.
Studying how target shareholders make decisions on
whether to accept cash or equity can enhance our
understanding of the information asymmetry problem that
exists between targets and bidders. One problem facing
target shareholders in an equity offer is how to value the
bidder's share offer. This problem arises since it is
possible that the bidder may be offering overvalued equity
(see Section 1.2.2) . In "all equity" or "all cash" bids we
observe only one method of payment hence we cannot gain any
insight into how the target shareholders resolve this
valuation problem and decide that cash or equity is the
acceptable method of payment.
The institutional environment in the United Kingdom
offers an opportunity to study the choice of payment method
by target shareholders. In the UK, we observe a large
number of bids where the bidder makes a "cash or equity"
offer. In these bids the target shareholders are allowed
the opportunity of deciding which method of payment to
accept. By studying the method of payment accepted in "cash
or equity" offers, we can gain some insights into how
4 me CGT liability of target shareholders and its impact on the
method of payment has received some attention in the extant literature
(see Niden, 1988)
12
target shareholders resolve the problem of valuing the
bidder's equity. In this regard, this study is the first to
consider the choice of payment method from the target
shareholders' perspective and provide empirical evidence on
the determinants of that choice.
1.3. Impact of the method of payment on shareholder wealth
The empirical literature on corporate acquisitions has
established that bidders and targets earn higher returns in
cash offers than in equity offers (see Table 7.1 and 7.2
below) . The most common explanations for the difference in
the abnormal returns associated with cash and equity offers
include information asymmetry between the bidder's managers
and shareholders in both participating firms, compensation
for capital gains tax, and transfer of wealth from
shareholders to bondholders.
1.3.1. Capital Gains Tax (CGT) and shareholder wealth
If there is a CGT compensation effect in takeovers
(see 1.2.1. above), then we would expect that in a cash
offer the premium paid to target shareholders would be
positively related to the level of CGT payable. If the CGT
compensation premium to target shareholders is paid out of
the bidder's share of expected merger gains then we would
expect that in a cash offer the bidder's returns are
negatively correlated with the CGT payable. While the
empirical evidence that the returns to target shareholders
are higher in cash offers is consistent with the CGT
13
compensation hypothesis, the evidence that the bidder also
gains in a cash offer is inconsistent with this hypothesis.
The empirical evidence does not support the CGT
compensation hypothesis 5 (Niden, 1988; Franks et al, 1988;
Hayn, 1989)
1.3.2. Information asymmetry and shareholder wealth
Information asymmetry theory suggests that the method
of payment conveys information to the stock market (Myers
& Majluf, 1984; Fishman, 1988; Eckbo, Giammarino & Heinkel,
1990; Brown & Ryngaret, 1991) . Information asymmetry models
are based on the assumption that the true value of a
company's assets is not known to the market. Managers who
have better information about that value because they are
insiders, may attempt to exploit this information advantage
by issuing new equity when they believe that their shares
are overvalued (Myers & Majiuf, 1984) . However such a
strategy would reveal any overvaluation to the market and
lead to a downward revision in the company's share price.
The negative returns observed around the time of seasoned
equity offers has been interpreted as evidence in support
of the information asymmetry ary-ument (Asquith & Mullins,
1986; Masulis & Korwar, 1986; Mikkelson & Partch, 1986;
]Jierkens, 1991)
If the market believes that equity is used as a method
of payment when the bidder's managers deem their shares to
5 See 1.2.1. above for a discussion of reasons why CGT may not
affect the returns to shareholders.
14
be overvalued, then this should cause the returns to
bidders offering paper to be negative. Since the market can
adjust for the information advantage of the bidder's
managers, there should not be any effect on the returns to
the target (i.e, target shareholders can demand additional
shares from the bidder to compensate for any downward
revaluation in the bidder's share price) . We are not aware
of any studies which have tested the information asymmetry
hypothesis.
1.3.3. Wealth redistribution and shareholder wealth
Wealth transfer theory suggests that equity offers
result in the transfer of wealth from shareholders to
bondholders. This wealth redistribution from shareholders
to bondholders is the result of a fall in the variance of
the cash flows of the combined firm. Since the variance of
the combined firm's cash flows is a weighted average of the
variance and co-variance of the individual firms' cash
flows, then where the correlation between the cash flows of
the merging firms is low or negative, the variance of the
combined firm's cash flow will fall.
The fall in the variance of the combined firm's cash
flow reduces the default risk on the firms' outstanding
debt. In the absence of any increase in the firm's cash
flow resulting from synergy gains a reduction in the
default risk of debt increases the market value of the debt
at the expense of the shareholders (Higgins & Schall, 1975;
Galai & Masulis, 1976)
15
In order for the wealth transfer effect to be present,
the cash flows of the two firms must be combined and this
implies that significant resources must not leave the
group. The wealth transfer effect is stronger in business
combination effected via an exchange of equity (Eger, 1983;
Travlos, 1987) . In a cash offer it is possible that the
amount of assets (particularly liquid resources) leaving
the group are so large that the cash flow profile of the
combined firm is not a simple combination of the cash flows
of the two independent firms. Additionally, with resources
leaving the group in a cash offer, the increase in asset
backing for debt is reduced, hence eroding the scope for a
wealth redistribution from shareholders. The empirical
evidence based largely on US studies has not supported the
wealth redistribution hypothesis (Asquith & Kim, 1982;
Dennis & McConnell, 1986; Travios, 1987) . We are not aware
of any UK studies testing the wealth redistribution
hypothesis.
1.3.4. Empirical evidence on the method of payment and
shareholder wealth
The majority of studies that have examined the impact
of the method of payment on announcement period abnormal
returns have concentrated their analysis on addressing the
question of whether the medium of exchange has an impact on
the wealth experience of the parties to an acquisition.
There is a limited literature which tests the theoretical
explanations which have been advanced for the difference in
16
returns between cash offers and equity offers. The CGT
hypothesis has been examined and rejected by studies in the
UK and the US (Franks et al, 1988; Niden, 1988; Hayn,
1989) . The wealth redistribution hypothesis has not been
tested in the UK, although it has been rejected by some US
studies (Asquith & Kim, 1988; Dennis & McConnell, 1986;
Travios, 1987). The information asymmetry theory has not
been tested in the literature. While we know that the
method of payment affects shareholder wealth, we cannot yet
explain why the payment currency influences shareholder
returns.
1.4. Objectives of the thesis
Having laid out the broad issues to be examined by
this thesis, we now set out our specific objectives. This
thesis focuses on the method of payment in corporate
acquisitions and addresses three research questions:
1) What are the factors that influence the bidder's
choice of payment currency in corporate acquisitions?
2) What are the factors that influence the choice by
target shareholders of accepting cash or equity, when
the bidder offers such a choice?
3) Why do cash offers result in higher returns to
shareholders than equity offers?
17
As a subsidiary issue arising from examining the
simultaneous relationship between the method of payment and
the choice of accounting policy, we study the determinants
of the accounting policy choice in corporate acquisitions.
1.5. Outline of the thesis
In Chapter 2 we argue that the method of payment and
the choice of accounting policy are interrelated, and
should be viewed as jointly and simultaneously determined
decisions. The chapter reviews the existing literature on
the determinants of the method of payment in takeovers. Our
review identifies a number of variables that should
influence the method of payment, but whose role has been
largely ignored in the empirical literature on acquisition
financing (eg choice of accounting policy, external
blockholding, growth opportunities in the merging firms,
free cash flow)
Chapter 3 reviews the literature on the determinants
of the accounting policy in corporate acquisitions. The
evidence from the US has concentrated exclusively on the
impact of goodwill on the choice of accounting policy and
ignored a number of other variables (e.g, size of the firm,
gearing ratio) which have been found to significantly
affect the choice of accounting policy in non-takeover
contexts. The literature review reveals a lack of UK
evidence on the determinants of accounting policy.
Chapter 4 outlines the simultaneous equations model
which investigates the interaction between payment method
18
and accounting policy choices in corporate acquisitions.
Additionally the definition of variables and descriptive
statistics on the data for the study are provided in the
chapter.
Chapter 5 presents the results from estimating the
simultaneous equations model. The results are analysed with
a view to understanding how the method of payment and the
choice of accounting policy are related. Variables which
influence the method of payment and the choice of
accounting policy are identified and their economic meaning
is discussed.
Chapter 6 reports and discusses the results of
empirical tests on the factors affecting the choice of
accepting cash or equity by target shareholders, when the
bidder has offered "equity with a cash alternative" as the
method of payment.
Chapter 7 reviews the existing literature on the
effects of the payment method on shareholder wealth. The
literature survey shows that higher returns to cash offers
than equity offers is a robust result, across different
time periods and institutional environments. The chapter
discusses the theoretical arguments advanced to explain
this result. Our survey reveals a lack of empirical tests
of these theoretical explanations. Chapter 8 reports an
empirical analysis of these theoretical explanations.
Chapter 9 summarises the results of our research,
discusses implications for policy makers and suggests
directions for future research.
19
CHAPTER 2.
DETERMINANTS OF THE METHOD OF PAYMENT IN MERGERS AND
ACQUISITIONS :- THEORY & EMPIRICAL EVIDENCE.
2.1. Introduction
In paying for an acquisition, the bidder can choose
shares, cash, loan notes or some combination of all three.
The importance of each method of payment is likely to
fluctuate with market conditions and, in some instances,
with the prevailing fashion. The bidder's share price,
gearing structure and cash resources, will all influence
the method of payment. The tax position of the target
shareholders will also have an impact on the method of
payment. Target shareholders are more likely to accept an
offer, if the consideration received is tax efficient from
their viewpoint.
The impact of the method of payment on the percentage
holding of existing bidder shareholders may also have an
important impact on the final choice. Private information
held by managers may affect the method of payment.
Information asymmetry based models predict that the
bidder's managers will only offer equity when they believe
that its shares are overvalued and offer cash when the
shares are undervalued. One of the main tasks confronting
the merchant bankers advising the parties involved in an
acquisition is to devise an appropriate consideration
package which satisfies the financial and tax based
preferences of both the bidder and the target shareholders.
20
The relationship between the method of payment and the
choice of accounting policy may also have a bearing on how
an acquisition is structured. There are two methods of
accounting for a business combination which are permitted
in the UK: acquisition and merger accounting. The rules
governing the availability of merger accounting establish
a major link between the accounting method and the payment
currency. This inter-relationship must be considered at the
acquisition planning stage if bidders wish to report the
combination in a favourable light.
Accounting rules require, inter alia, that the bidder
must offer at least 9O of the fair value of the total
consideration in the form of equity in order to qualify to
use merger accounting. The 9O rule implies that the choice
of accounting method is partly determined by the payment
method. Reciprocally, the payment method is partly
determined by the choice of accounting method.
This chapter discusses the relative advantages of each
of the main methods of payment and reviews the theoretical
arguments and empirical evidence on the factors which are
likely to influence the bidder in choosing the payment
currency.
2.2. Payment methods available in corporate acquisitions
2.2.1. Shares
A bidder wishing to use equity as a method of payment
faces a number of institutional and statutory requirements
which have to be fulfilled in respect of the new shares
21
being issued. The primary institutional rule is the
requirement to publish "listing particulars".
If the new shares issued will increase the number of
a class already listed by more than 10%, then listing
particulars must be published. The precise contents of the
listing particulars are set out in Section 3 of the Stock
Exchange's
H
Admission of Securities to Listing" (the
Yellow Book, 1984 edition)'. If any new class of securities
is to be listed as a result of the bid, then a formal
notice must appear in a national daily newspaper, which
must specify the address where the full listing particulars
are available for inspection. The bidder must make
arrangements for details of listing particulars to be
circulated in the Extel Statistical Service. In an equity
offer, the requirement to publish listing particulars can
place a significant cost burden on the bidder.
A bidder wishing to use equity, must also consider the
statutory protection given to its existing shareholders by
Section 80 of the 1985 Companies Act. Sec.80 CA 1985
provides that no shares (or securities carrying conversion
rights into shares) may be issued by the directors without
the authority of that company given either in a general
meeting or by the Articles of Association. Such authority
must state the maximum amount of shares that may be issued
and the time period not to exceed five years, in which this
may be done. If shareholder approval is required under
Section 80, an extraordinary general meeting of the company
1 Since
1/12/93 called the "Stock Exchange Listing Rules"
22
must be convened, for which 21 days notice is required.
There are circumstances in which the bidder wishes to
issue equity, but the target shareholders would prefer to
receive cash. In such situations, this difference may be
reconciled through a vendor placing or vendor rights. Under
these schemes the bidder issues shares to the target's
shareholders, but arranges for its merchant bankers to buy
back these shares at a fixed price.
Any shares bought by the merchant bankers can either
be offered to the bidder's existing shareholders (vendor
rights) or placed in the market (vendor placing). One of
the problems with an equity offer is that the target's
shareholders are compelled to place their own value on the
bidder's shares. A cash underwritten alternative reduces
this problem by providing a secure value for the bidder's
shares. However this backing is provided at some cost in
terms of the fees payable to underwriters.
As the shares issued under a vendor placing or vendor
rights are not issued for cash, the statutory pre-emption
rights provided by Section 89 of the 1985 Companies Act 2 do
not apply. Consequently existing shareholders may suffer a
dilution in their percentage shareholding when these new
share are placed in the market.
In its memorandum on "shareholders' pre-emption rights
2 Section 89 Companies Act 1985 imposes statutory pre-emption
rights in favour of existing shareholders whenever there is a new issue
of shares. These rights provide that any new issue of shares must be
made proportionately to existing shareholders in terms of the nominal
value of existing shares. This rule makes a rights issue compulsory.
However the pre-emption rights do not apply where the issue of shares
is wholly or partly for a consideration other than cash.
23
and vendor placings" issued in February 1989, the Investor
Protection Committee (IPC) of the Association of British
Insurers (ABI) stated that where the shares issued by the
bidder exceeds 1O of its existing issued share capital,
then a claw back offer must be made available to the
existing shareholders 3 . Although the guidelines from the
IPC do not have the force of law, they are usually
persuasive given the power of institutional shareholders.
Under a claw back offer, the bidder's existing shareholders
are given the right of first refusal over the new shares.
The claw back offer must be available for at least 21 days.
2.2.2. Cash
Cash offers are quick, clean and avoid most of the
legal and institutional complexities associated with equity
offers. Unless internal cash resources are adequate to
finance the bid, the bidder would have to resort to medium
or long term borrowing. The ability of the bidder to borrow
to finance a cash offer will depend on its existing level
of borrowing, the type of security it can offer, the amount
of additional cash flows which will be generated from the
acquisition etc.
Borrowing funds in order to finance a cash offer might
result in some restrictions being placed on the bidder by
the lenders, since lending agreements tend to incorporate
restrictive covenants (Smith & Warner, 1979; Citron,
3 The
claw back is popularly referred to as a "vendor rights"
offer.
24
1992a). The terms on which the bidder is able to borrow, in
order to finance a cash offer will be important in choosing
the method of payment.
Alternatively the bidder could finance the cash offer
via a rights issue. In this case, new shares are offered to
its existing shareholders on a pro-rata basis at a discount
to the existing market price. A rights issue requires the
preparation of listing particulars (Section 3, Chapter 1,
Yellow Book, 1984 edition) . Additionally the issue is
normally underwritten in order to guarantee that the
necessary funds will be raised. Underwriting commissions
can be quite substantial. As a rights issue must be open
for at least 21 days this can represent a substantial time
delay in raising the funds required to finance the cash
offer.
In certain situations, the bidder may not have a
choice of which method of payment to offer. A cash offer
may be mandatory under the City Code on Takeovers and
Mergers (the City Code). Rule 9 of the City Code provides
that if a person acquires 30% or more of the voting rights
of the target, or when they already have more than 30%
acquire 1% of the voting rights within a 12 month period,
then a general offer in cash or with a cash alternative
must be made for the balance of the company's shares at the
highest price paid in the previous 12 months. Rule 11 of
the City Code provides that, where a person has purchased
4p rior
to 3/3/1993, the relevant percentage was 2
25
in the last 12 months, preceding a general offer 1O6 or
more of the voting rights, then a subsequent general offer
must be for cash or must include a cash alternative at the
highest price paid during the period.
2.2.3. Loans
From the bidder's point of view, loans have a number
of advantages as a method of payment. Interest payments on
loans are tax deductible, while dividends are not. Loan
financing could lead to an increase in earnings per share
(so long as profits from the acquisition cover the interest
payments) . Target shareholders accepting loan notes as
consideration can claim roll over relief against any
capital gains tax liability 6 . Loan capital can be issued in
a variety of ways, listed loan stock, convertible loan
stock and warrants.
Any company issuing listed loan stock has to draw up
listing particulars for the loans issued. The exemption
available for small equity offers (i.e, any issue less than
1O of a class already in issue) does not apply. The
contents of the listing particulars are detailed in
chapters 1 and 2 of section 3 of the "Yellow Book (1984
5 Prior to 25/6/1989, the threshold was l5.
6 1f the loan stock is classified as a 'qualifying corporate bond'
(ie, loan stock issued after 13 March 1984 by a company whose shares
are quoted on a recognised stock exchange) then target shareholders are
entitled to a 'hold over' rather than 'roll over' relief. The
distinction between the two is that under 'hold over' relief capital
gains accrued by target shareholders are frozen, carried forward and
become liable for CGT when the loan stock is eventually disposed off.
This treatment arises because qualifying corporate bonds are exempt
from CGT.
26
edition)
tI
The loan stock will usually be secured by a
trust deed entered into between the bidder and a trustee
appointed to represent the interest of the bondholders. The
minimum contents which must be included in the trust deed
are detailed in chapter 2 of section 9 of the "Yellow
Book".
Convertible loan stocks are loan stocks with equity
conversion rights. The conversion rights usually enable the
stockholders to convert from debt into equity at a future
date and fixed price. The stockholders get a guaranteed
income in the form of interest, security for their holding
(if the loans are secured) and the opportunity to convert
into equity, if the bidder prospers. Additionally any
increase in the price of the bidder's equity will be
reflected in the price of the convertible loan stock. The
advantages to the bidder of a convertible issue include:(1) a lower interest rate is paid in exchange for the
conversion rights in comparison to a straight loan issue;
(2) it is a form of borrowing which if everything goes
well never has to be repaid.
However the bidder would need to consider some
potential disadvantages of a convertible issue. The fixed
conversion price will usually represent a discount on the
existing market price (otherwise target shareholders have
no incentive to accept the convertible). This discount can
only be at the expense of the existing shareholders. Also
when the conversion takes place, the percentage
shareholding of existing shareholders will be diluted. A
27
convertible involves greater administrative costs than a
straight loan stock. Conversion notices have to be sent to
stockholders prior to every conversion period and the share
register has to be updated regularly with small parcels of
shares.
Warrants entitle the holders to subscribe for the
bidder's shares at some future date at a fixed price. The
main difference between a warrant and a convertible is that
warrants are separated from the underlying security with
which they were issued and are traded separately on the
stock exchange. At some stage in the future, warrants will
result in the inflow of cash to the bidder, although the
amount and timing of these inflows is uncertain. They are
usually issued in order to make the terms of a bid more
attractive to the target shareholders, but without an
immediate cost to the bidder. The negative effects which a
convertible has on existing shareholders also apply to
warrants.
2.3. Accounting policy choice and the method of payment
Accounting policy preferences of firms are influenced
by a variety of considerations. Watts and Zimmerman (1978)
argue that management's support or opposition to a proposed
accounting standard depends upon the size of the firm and
whether the proposed standard increases or decreases the
firm's reported earnings.
Large firms are more likely to support income reducing
standards due to tax, political and regulatory benefits. On
28
the surface a decision, to support an accounting standard
which reduces profits may appear to be counterproductive.
However, large firms have high public visibility and a high
profit fig-ure can result in (i) accusations of
profiteering 7 (ii) political pressure f or increased taxes
on these profits (iii) increased scrutiny by the antimonopoly agencies.
Large firms can seek toe, reduce their vulnerability to
these pressures by supporting and adopting accounting
policies which reduce their profits (Watts & Zimmerman,
1978; Hagerman & Zmijewski, 1979; Zmijewski & Hagerman,
1981). These benefits are traded off against the costs of
reporting lower earnings such as loss of earnings related
compensation, lower interest and dividend cover,
downgrading of the company's future prospects by the stock
market.
While the above considerations are important, the
choice of accounting method (i.e, acquisition or merger
accounting) is in practice restricted by the relevant
accounting standard (SSAP
23)8.
In order to qualify to use
merger accounting, bidders are required to offer equity f or
not less than 9O of the fair value of the total
consideration paid. Thus method of payment in acquisitions
7mere is a general tendency for the public to associate high
profits by large firms with monopoly rent (Watts & Zimmerman, 1978).
8 SSAP 23 which was the relevant accounting standard operating
throughout the period covered by the sample in this study (1/1/80 to
31/12/90) was withdrawn in September 1994 and replaced by Financial
Reporting Standard (FRS) 6. However FRS 6 requires that the majority of
the consideration paid to target shareholders must be equity in order
to qualify for merger accounting (see Section 3.2.2 below)
29
is one of the determinants of the accounting method choice.
Choice of payment method in acquisitions is, however,
not determined entirely by accounting considerations. A
number of studies in the finance literature have suggested,
and empirically identified, factors which bear upon the
payment method choice by bidders. These factors encompass
bidder's capital availability, financial and ownership
structure as well as information asymmetry between bidder
and target shareholders.
There have been few studies concerned with the impact
of the bidder's choice between merger and acquisition
accounting on payment method. Yet such a choice must be
reflected in the payment method decision because of the 9O
rule referred to above. As Wyatt (1967) notes:
the accounting for a combination is commonly
decided in advance of consummation of the transaction.
That is, the accounting treatment is one of the
variables that must be firmed up before the final
price is determined."
2.4. Information asymmetry and the method of payment
Information asymmetry theory is based on the
assumption that in any transaction, if one participant has
superior private information then an attempt will be made
to exploit this information advantage to the possible
detriment of the other participants to the transaction. In
takeovers either the bidder or the target can have superior
information about the value of its assets.
Information asymmetry affects cash offers and equity
offers differently due to the contingent pricing effects of
30
equity. With a cash offer the value of the consideration
paid to acquire the target is determined ex-ante (i.e, at
the date of the offer) since there is no private
information about the value of cash. The true value of an
equity offer is determined ex-post (i.e, after the date of
the offer) since it depends on the post merger
profitability of the enlarged group.
When the target has p4vate information about the
value of its assets it will only agree to a trade with the
bidder, if the value of its assets is less than the value
of the consideration offered by the bidder. The bidder
therefore faces a valuation risk or adverse selection
problem9 . To mitigate the negative impact of this valuation
risk the bidder can make an equity offer, whose value is
determined by the future profitability of the group. The
advantage here is that the target shareholders bear part
the valuation risk.
When the bidder has private information on the value
of its equity then the target shareholders face a similar
adverse selection problem that the equity offer may be over
valued. Indeed, the target shareholders' likely presumption
is that the bidder will attempt to exploit any over
valuation and convert its over valued equity into real
assets. Similarly target shareholders may presume that cash
offers are only made when the bidder believes that its
equity is undervalued. The conclusions of some of the main
9valuation risk refers to the possibility of the bidder
discovering ex-post that its valuation of the target's assets was
incorrect.
31
models in the literature which investigate the conditions
for equilibrium in the presence of information asymmetry
are discussed below.
2.4.1. Relative size of the bidder to the target and the
method of payment
Hansen (1987) examined the bargaining process in a
merger or acquisition. He deyeloped a model in which the
determination of the choice of exchange medium is the
result of a two agent bargaining game under imperfect
information. Hansen concludes that when the target is
assumed to know the true value of both its own assets and
the bidder's assets, but the bidder is only aware of the
true value of its own assets (i.e, the target is assumed to
have an information advantage) then:a) Equity offers can establish a trade'° between the
bidder and the target in circumstances where cash offers
cannot effect a trade. (i.e., equity dominates cash as a
medium of exchange);
b) The probability of an equity offer decreases as the
size of the bidder increases relative to the size of the
target;
c) The probability of an equity offer increases with
the level of financial gearing of the bidder and decreases
with the level of financial gearing of the target.
Hansen uses a significant amount of algebra to
10 Trade
refers to the acceptance of the bidder's offer by the
target.
32
establish his results, but the intuitive reasoning behind
his conclusions is as follows:Conclusion (a) reflects the fact that an equity offer
establishes the actual acquisition price ex-post.
Consequently the bidder is able to use an equity offer to
reduce the adverse selection problem associated with its
uncertainty of the true value of the target's assets. If
the bidder makes a cash of fe with a value of C, and the
true value of the target's assets is V1 the bidder faces
the risk of discovering after consummation of the deal that
V < C (i.e, with a cash offer the bidder runs the risk of
buying a "lemon")
With an equity offer some of the adverse selection
risk is passed on to the target. If the bidder makes an
equity offer with a value
discovered that
V < E,
E
and after the acquisition it is
the post transaction value of the
enlarged group will fall and so will the value of the
acquisition price paid to target shareholders. From the
target's point of view the dominance of an equity offer
still holds, because the target has no difficulty in
determining the value of an equity offer since it has an
information advantage. Therefore if a cash offer with a pre
transaction value of C is acceptable to the target, then an
equity offer
E
which is at least equal to C will also be
acceptable to the target (i.e, C puts a floor under the
consideration)
It must be pointed out that the dominance of an equity
offer established in Hansen's model is based on the
33
assumption that the bidder is always able to create value
out of the acquisition. Formally Hansen assumes that if
W(V) is the value of the target's assets to the bidder then
W(V) is an increasing function of V and W(V) > V for all
values of V. While it may be possible to argue that from
the bidder's point of view W(V) > V is a necessary
condition for making an acquisition there is no reason for
target shareholders always to believe that the bidder will
create value out of the acquisition.
If the possibility that target shareholders can face
a post-acquisition moral hazard problem (i.e, that the
bidder fails to create value out of the acquisition) is
introduced into Hansen's model, it is doubtful if, ceteris
paribus, the dominance of equity offers will still hold.
Conclusion (b) is based on the idea that the
contingent pricing advantage of equity offers depends on
the target's assets being a significant addition to the
bidder. As the bidder increases in size relative to the
target the valuation risk and contingent pricing advantage
of equity offers diminish. When the bidder is too large
relative to the target, the beneficial price contingent
effect of an equity offer is negligible.
Conclusion (c) reflects the fact that the contingent
pricing advantage of an equity offer is stronger the larger
the equity of the target is relative to the equity of the
bidder. If we imagine two bidders 1 and 2, with the same
level of total assets but with different levels of debt Dl
and D2, where Dl > D2, then the equity component El < E2.
34
If both bidders were to acquire a target with a given size
V1 bidder 1 will gain more from an equity offer than bidder
2, because the smaller size of its equity increases the
contingent pricing advantage of the equity offer. Therefore
as the bidder's gearing increases, the contingent pricing
advantage of an equity offer increases. A similar argument
can be developed to show that the contingent pricing
advantage of an equity offer to the bidder falls as the
target's gearing increases.
2.4.2. Overvaluation of the bidder and the method of
payment
Myers & Majiuf (MyM) examine the behaviour of managers
and investors, in a model which assumes that managers have
superior information about the value of a firm's existing
assets and future investment opportunities. The MyM model
is built on two central assumptions:
1)
Managers act in the interests of existing
shareholders.
2) Existing shareholders are passive and do not adjust
their portfolios in response to the firm's investment
decisions.
With these assumptions the MyM model leads to the
following conclusions:
1) Where a company has no financial slack" the firm
may reject positive NPV projects rather than issue equity
"Financial slack is defined as cash and marketable securities
held by the firm plus total risk free debt that can be issued by the
firm.
35
to finance the project, if managers believe that the firm's
equity is undervalued.
2) The issue of stock will always result in a fall in
the value of existing shareholders' wealth.
MyM establish that the issue of equity would only
result in the maximization of existing shareholder wealth,
if the proportion of existing firm value accruing to the
new shareholders is less than the increase in the value of
the firm resulting from using the proceeds of the equity
issue. If this condition is violated the firm should reject
the project irrespective of whether it might be a positive
NPV project. This reflects the fact that the under pricing
of the new issue may be so severe that the new shareholders
earn more than the NPV of the new project with a consequent
loss to existing shareholders. Presumably a rights issue
can avoid this problem, since existing shareholders are
given the first opportunity to acquire the whole of the NPV
from the new project. However an equity offer in a
takeover, creates a distinct body of new shareholders who
can acquire all the gains generated by the acquisition and
part of the wealth of existing shareholders, if the
underpricing in the bidder's equity is severe.
Conclusion 2 seems to contradict existing finance
literature. If a project has a positive NPV, why should the
issue of equity to finance the project lead to a fall in
the share price?. This pessimistic result arises directly
out of the existence of information asymmetry between the
managers and investors. In the MyM model a value maximising
36
manager will issue new equity and invest in a zero or
negative NPV project so long as the value of the new assets
accruing to the new shareholders is less than or equal to
the value of the equity issued (i.e, equity may be issued
at a premium). Therefore the issue of new equity does not
automatically signal the existence of a positive NPV
project. Since investors are aware of the possibility that
a new equity issue could simply result in the transfer of
wealth from themselves to existing shareholders, the market
is likely to interpret any new issue of equity as bad news.
2.5. Taxation and the method of payment
Although the taxation of mergers and acquisitions in
the UK is a complicated subject, the actual scope for
increasing value through tax planning in an acquisition is
extremely limited. While bad tax planning can be costly,
good tax planning is not likely to result in the
exploitation of tax opportunities to create value in an
acquisition (except in specific and unusual cases, e.g, to
utilise irrecoverable advance corporation tax)
In this regard the UK corporate tax environment is in
sharp contrast to the USA where prior to the Tax Reform Act
(1986), it was generally believed that tax planning could
independently add value to an acquisition. Niden (1988:
Chapter 2) provides a comprehensive discussion of the tax
planning opportunities available in the USA prior to
37
198612
While value creation by the bidder as a result of tax
benefits is limited in the UK, there can be a direct
taxation impact on the wealth of the target shareholders as
a result of the type of consideration received from the
bidder. Under UK tax laws a disposal of shares by a
chargeable person is a taxable event unless the
consideration received by the vendor is in the form of
shares or debentures in another company. (Section 85
Capital Gains Taxes Act 1979)
In an equity offer the target shareholders can claim
roll over relief and avoid the crystallisation of any
capital tax liability, until the vendor sells the new
shares received in the equity offer. If prior to the bid
the bidder held more than 5 of any class of the share or
loan capital in the target, then in order for roll over
relief to apply, it will be necessary to demonstrate that
the transaction was effected for bona fide commercial
reasons and not designed primarily for tax avoidance
purposes. Where doubt exists, then a procedure for
obtaining prior clearance from the Inland Revenue is
available under Section 88 CGTA 1979.
In an efficient market, the differences in the
taxation treatment of equity and cash offers will tend to
compel bidders to either finance with an equity offer or
12 The main sources of tax benefits in US takeovers were: (i) the
step up in value of the target's assets in calculating tax allowances
(see Section 7.4. below for a further discussion), (ii) the utilisation
of the target's losses and tax credits by the bidder.
38
else offer a higher premium in a cash offer to compensate
target shareholders for the capital gains tax arising on
the disposal of their shares. Carleton et al (1983),
Wansley et al (1983), Huang & Walkling (1987), Franks et al
(1988) provide evidence that the bid premium is
significantly higher in cash offers than in equity offers.
These results would be consistent with the tax compensation
hypothesis. However, it should be noted that alternative
arguments based on information asymmetry (Myers & Majiuf,
1984) would also be consistent with these results.
Franks et al (1988) cast serious doubt on the ability
of the tax compensation hypothesis to explain the higher
bid premium observed in cash offers. They observe that:1) Higher bid premium in cash offers were observed
prior to 1965 when capital gains tax was introduced in the
UK.
2) Target bid premia in "cash or equity" offers were
comparable with the bid premia in "all cash" offers. Since
"cash or equity" offers should reduce any adverse personal
tax consequences of the offer, it was expected that "cash
or equity offers" would be associated with a lower bid
premia than all cash offers.
Niden (1988: Chapter 4) is the only study that has
examined the role of taxation in acquisition financing. She
used the following variables as proxies for the impact of
the acquisition on the CGT position of the target
shareholders on the method of payment:- proportion of
target shares held by institutional investors, target's
39
dividend yield' 3 , variability of the target's market
adjusted return in the pre bid period' 4 , increase in the
target's share price over the six months period preceding
the bid. She performed Logit regressions to examine whether
the type of consideration offered was a function of the tax
status of the target's shareholders. Her Logit regressions
had very low explanatory power and most of the coefficients
of her tax variables were not significant nor did they have
the predicted sign. Niden concluded that there was no
relationship between the tax status of the target's
shareholders and the form of consideration offered in the
acquisition.
The result that CGT cannot explain either the higher
bid premium in cash offers or the choice of payment method
may be rationalised by the availability of legitimate tax
schemes which allow individual investors to reduce their
CGT liability'5.
2.6. Share ownership structure and the method of payment
There have been some recent attempts in the literature
on the theory of capital structure to explain the existence
of an optimal debt-equity ratio in terms of managerial
13Niden suggested that high income tax investors bought shares in
low yield firms, while low tax investors bought shares in high yield
firms.
14Without giving an adequate explanation, Niden suggested that the
average tax rate of a firm's shareholders is an increasing function of
the variability of that firm's common stock returns.
15 See Section 1.2.1. for a summary of the principal methods
available for reducing CGT liability.
40
control of voting rights. The related models marry the
literature on the market for corporate control with the
capital structure theories by exploiting the fact that
equity shares have voting rights while debt does not. The
main models in this area are Stulz (1988) and Harris &
Raviv (1988) . These two models develop a relationship
between the voting rights controlled by management and the
value of the firm.
2.6.1. Managerial control of voting rights and the value of
the firm
Stulz (1988) uses managerial control of voting rights
(MCVR) as a construct to examine the attractiveness of a
company as a takeover target. In Stulz's model the
proportion of voting rights controlled by managers affects
the firm's likelihood of being a takeover target and the
size of the bid premium received. An increase in MCVR will
increase the value of the firm because it increases the
size of the takeover premium. This result is derived from
the assumption that the passive shareholders who hold the
remaining shares not held by managers have heterogeneous
opportunity costs for selling their shares (ie, the premium
demanded by the owner of the
ith
share is not equal to the
premium demanded by the owner of the
jth
share).
Target shareholders can demand different premia due to
their differing capital gains tax position, loyalty to the
existing management etc. Therefore the bigger the level of
MCVR the greater the proportion of passive investors shares
41
that must be obtained by the bidder 16 and hence the larger
the premium the bidder must offer to target shareholders.
Above a certain level, an increase in MCVR will result
in a fall in the value of the firm. This result is based on
the assumption that as MCVR increases the probability of a
takeover declines. If managers control more than 5O of the
votes the probability of a hostile takeover bid is zero. As
the probability of a takeover declines, so does the
probability that the passive investors will realise the
control premium associated with a takeover. Stulz shows
that there is an optimal level of MCVR at which the value
of the firm is maximised.
Stulz then argues that given a personal wealth
constraint limit, managers can increase their control of
voting rights by increasing the firm's leverage'7.
Increasing leverage will have two effects on the value of
the firm: (i) increase the potential premium available from
a future bidder by increasing MCVR (ii) reduces the
probability that the bid premium will be realised, since an
increase in MCVR reduces the chances of a future bidder
succeeding. With these two opposing effects, Stulz derives
the result that there is an optimal level of leverage that
'6 1f managers control l0 of the votes, then the bidder would have
to acquire 55.55 (i.e. 50/90) of the votes held by passive investors
in order to gain control. With managers controlling 20 of the votes,
the required percentage rises to 62.5 (i.e, 50/80)
17 For example, let us assume that managers have a total wealth of
£100. They wish to invest in a project costing £1000. If they issue
1000 shares with a nominal value of £1, then they will control l0 of
the votes. Alternatively by issuing 500 shares and raising £500 by debt
instrument, they can increase their control over the firm's votes to
20w.
42
maximises the value of the firm.
Since equity offers dilute
MCVR
while cash offers do
not, Stulz's model suggests that where managers value the
benefits of control the probability of an equity offer will
be inversely related to both the bidder's
MCVR
and
leverage.
2.6.2. Managerial control of voting rights and the type of
takeover attempt
Harris and Raviv (1988) analyse the effects of insider
control and financial leverage on the type of takeover
method (i.e, tender offer or proxy contest) . Their model
postulates that the target's management can influence the
type of takeover attempt and its probability of success by
choosing the level of
MCVR.
Changes in
MCVR
are affected by
the level of debt. Issuing more debt increases the level of
MCVR.
Increasing leverage (by increasing MCVR) will reduce
the probability of the incumbent management being voted
out. This increases the probability of reaping the benefits
of control, but reduce the likelihood of obtaining capital
gains through a takeover. Additionally increasing leverage
could reduce the benefits of control by increasing the
probability of bankruptcy, and by increasing the monitoring
activities of creditors which reduce the ability of
management to mis-allocate free cash flow. By trading of f
these factors, the target's management can determine an
optimal level of MCVR (and an optimal capital structure)
43
This would simultaneously determine the probability of
various takeover methods.
If the target's management has sufficient control over
votes (by choosing an appropriately high level of debt) to
guarantee that a hostile bidder fails, then we will observe
an unsuccessful tender offer. Conversely a very low level
of debt (and MCVR) will result in a successful tender
offer. Intermediate levels of debt, which imply that
neither the bidder nor the incumbent management can be sure
of achieving control, will result in a proxy fight.
The main insight of the model is that the target's
management can influence the type of takeover method and
its outcome by manipulating its control of the firm's votes
through its policy on capital structure.
2.6.3. Dilution of external blocklioldings
While there is some literature which discusses the
attenuating effects of equity offers on managerial
shareholding (Harris & Raviv, 1988; Stulz, 1988), very
little attention has been paid to the dilution of shares
held by large external blockholders. There are legal rights
associated with different levels of shareholdings which the
external blockholder may not wish to see diluted:5%: the right to object to the courts against a proposed
re-registration of the company as a private one
(sec.54 CA 1985);
10%: the right to requisition an extraordinary general
meeting of the company (sec.368 CA 1985);
44
the right to petition the Department of Trade to
appoint inspectors to investigate the company's
affairs (sec.431 CA 1985) ;
l5: the right to object to a court against a proposed
variation of class rights (sec.l27 CA 1985)
the right to object to the courts against a proposed
change in the company's objects clause (sec.4 CA
1985) ;
26%: the right to block any actions of the company which by
virtue of the Companies Act 1985 can only be carried
out by a special resolution of the company (e.g,
voluntary liquidation, change of articles, change of
name, scheme of arrangement etc)
Apart from the dilution of the legal rights discussed
above, there are economic benefits associated with a large
block of shares which the blockholder may not wish to see
diluted. Principally a large blockholding provides a
platform or toehold from which the blockholder can launch
a future bid' 8 . The benefit of the toehold is that it can
help the bidder avoid the "free-rider" problem. This issue
is addressed in the models of diffuse shareholding
discussed by Grossman & Hart (1980) and Shleifer & Vishny
(1986)
The "free-rider" problem arises when individual
shareholders who hold small numbers of shares, refuse to
18Although blockholders such as institutional investors are
unlikely to bid for a company themselves, they can use their block of
shares to help facilitate a bid by another raider.
45
accept an offer from the bidder. Each shareholder reasons
that his decision with regard to the bidder's offer will
not affect the outcome of the bid, since his shareholding
is small. They therefore hold on, in the hope of
participating in any value increases resulting from the
takeover. The cumulative result of these individual
decisions, is that the bid will fail.
Both Grossman & Hart (1980) and Shleifer & Vishny
(1986) argue that as a result of the "free-rider" problem
a bid will not succeed unless the offer price exceeds the
value of the company under the bidder's management. However
the bidder is only willing to pay a maximum price equal to
the potential value of the target less the cost of prebid
monitoring and the cost of the bid. This implies that the
attempt by the small investors to "free ride" and
participate in any post merger gains could ensure that the
offer price is insufficient to guarantee victory for the
bidder. At the extreme the "free rider" problem would
suggest that takeovers would not occur at all, so long as
the incumbent shareholders demand the entire value of any
surplus in return for their shares.
A large pre-bid toehold can mitigate the "free rider"
problem. If a bidder has a sufficiently large initial
shareholding, then the capital gains profit accruing on
this stake when the bid is announced, could be enough to
compensate the bidder for his monitoring and bid costs,
hence providing an incentive for the bidder to launch a
bid.
46
In the Shleifer & Vishny model, as the proportion of
the firm held by the bidder rises so does its share of any
takeover gains. Additionally as the bidder's toehold rises,
a takeover becomes more likely and the target's share price
increases. When a takeover does occur the premium above the
prevailing stock price would be lower. The lower premium
results from two factors (i) the increase in the pre
takeover market price of the target (ii) the bidder is
willing to launch a takeover for a smaller increase in the
post takeover profits of the target.
An external blockholder will wish to avoid any
dilution of their shareholding resulting from an equity
offer, if the shares represent a strategic investment (ie,
a toehold) which forms the platform for launching a future
a takeover bid.
2.7. Market conditions and the method of payment
It may be reasonably deduced that a manager's decision
to raise funds through a seasoned equity offer will be
affected by the recent performance of the company's shares
and the returns on the market as whole. Smith (1977), Marsh
(1979), Levis (1993) found evidence of significant abnormal
returns on a company's stock in the period preceding a
seasoned equity issue. This suggests that firms time the
issue of new equity to coincide with periods when their
share prices are rising This association between share
price performance and the timing of seasoned equity offers
could be explained by one of the following factors :47
1) Managers as insiders believe that their shares are
over valued. This is a feasible scenario, if the market is
strong form inefficient, and managers attempt to take
advantage of this mispricin.g.
2) The recent financial performance of the company and
its managers has been excellent and this is presently
reflected in the share price. The managers have therefore
decided to raise new capital in a period when they are the
'darlings of the market'.
The recent return on the overall market could also
have an impact on the timing of a seasoned equity offer. In
periods of a rising market, investor confidence is high and
there will be a general willingness to invest money in the
market. In a falling market, investors are likely to be
more risk averse and generally unwilling to undertake
speculative investments. Taggart (1977) found evidence that
seasoned equity issues tend to follow periods of market
rises.
The empirical evidence from the seasoned issues
literature would lead to an expectation that the level of
the market index and the bidder's share price will be
positively correlated with the proportion of equity in the
method of payment.
2.8. Capital structure and the method of payment
Since the Modigliani & Miller (1958) seminal paper,
the question of whether an optimal capital structure exists
has dominated much of the literature in financial
48
economics. If capital structure does not affect the value
of the firm, then a priori we cannot expect that the impact
of an acquisition on the bidder's capital structure should
influence the choice of payment method. However if an
optimal capital structure does exist, then the impact of
the acquisition on capital structure can be expected to
affect the choice of exchange medium.
The balance of the empirical evidence at present would
suggest that there is some pre-set capital structure which
firms try to attain. Marsh (1982), Jalinvand & Harris
(1984), Lasfer (1991) provide evidence that firms attempt
to maintain target levels of gearing. Assuming the
existence of a target capital structure, we may expect that
managers will view acquisition financing as part of their
normal financing decisions. Firms above their pre-set
capital structure may seek to issue equity in an
acquisition in order to adjust down their gearing ratios.
Firms operating at the desired level of gearing will be
inclined to use internal funds to finance acquisitions
(i.e, cash offers). Firms below the desired level of
gearing will be inclined to finance the acquisition through
new debt instruments.
It is arguable, however, whether capital structure
should have any influence on the method of payment, even if
one believes that some optimal or target gearing ratio
exists. If a bidder uses equity to finance an acquisition,
thereby reducing the gearing ratio to a level below
optimal, this implies that debt is more likely to be used
49
to finance future projects. Conversely if the bidder
finances the acquisition with debt, then a rights issue
could be used in future to restore the equity base.
Ignoring the effect of an acquisition on capital structure
will depend on how comfortable the bidder is with temporary
deviations from the "normal" gearing ratio.
It is conceivable that bidders with a high level of
gearing would find it comparatively more difficult to raise
the funds necessary to finance a cash offer. Increases in
leverage, the reduction of liquid resources and a possible
loss of control to creditors associated with a cash offer
could be more expensive for a bidder already carrying a
high level of debt. Consequently bidders with a high level
of debt may choose to finance acquisitions with equity.
The capital structure of the target may also influence
the method of payment. Targets with a high level of gearing
would already be subject to a high level of external
monitoring by the creditors (Jensen, 1986). A bidder can
free ride on the monitoring activities of the creditors.
This reduces the valuation risks 19 faced by the bidder with
a consequent reduction in the price contingent advantage of
an equity offer. This implies that targets with high
gearing are more likely to be acquired via a cash offer.
2.9. Growth opportunities in the merging firms and the
method of payment
Myers' (1977) under-investment model describes the
19 See 2.4. above for an explanation of valuation risk.
50
firm as a combination of assets in place and call options
on future growth opportunities. The value of these call
options depends on the probability that managers will
exercise them. Myers shows that situations may arise where
managers of firms with outstanding risky debt may refuse to
exercise the option on projects with a positive net present
value because acceptance of the project by increasing the
asset backing for the outstanding debt (with a consequent
increase in the value of the debt) reduces the value of
shareholder's wealth.
Myers (1977) also shows that the higher the level of
debt in the firm's capital structure the greater the
probability that managers will be forced to take sub
optimal investment decisions 20 . If a high level of debt
will result in an under investment incentive problem, then
a possible solution is that, firms with growth options
should reduce the level of debt in their capital structure.
Myers predicts that the firm's leverage is inversely
related to the proportion of the firm's value which is
accounted for by growth options. Smith & Watts (1993) and
Gayer & Gayer (1993) show that firms with growth options
have low debt/equity ratios. This suggests that takeovers
in which either the bidder or the target have significant
amounts of growth options are more likely to be financed
20 Since debt holders have a senior claim on project cash flows,
the greater the amount of debt in the capital structure, the greater
the probability that the firm's cash flows are paid out to creditors
rather than shareholders. Positive NPV projects thus add to the
security of the creditors and provide less benefits to shareholders.
This increases the probability that managers acting in the interests of
shareholders would under-invest (i.e, reject positive NPV projects).
51
with equity.
2.10. Cash resources and the method of payment
The free cash flow hypothesis (Jensen, 1986) argues
that where managers have cash flow in excess of that
necessary to finance positive NPV projects, instead of
returning such excess cash flow to shareholders, managers
would choose to make acquisitions. Returning excess cash
flow to shareholders might signal that managers do not have
any more positive NPV projects with the result that their
shares could be downgraded by analysts. If free cash flow
motivates the acquisition then it is more likely to be
financed with cash. The liquidity of the bidder will also
be expected to affect the method of payment. Cash rich
bidders are in a better position to make a cash offer than
cash poor firms.
Additionally a bidder acquiring a cash rich target is
more likely to make a cash offer since the cash resources
of the target can be used to help finance part of the cost
of the acquisition. The utilisation of the target's cash
resources in this manner, is limited by Section 151, 1985
Companies Act, which makes it unlawful for a company to
give financial assistance, directly or indirectly, for the
acquisition of its own shares.
2.11. Target managerial resistance and the method of
payment
When a bid has not received the support of the
52
target's management, the target's shareholders would
probably prefer to receive a consideration that is easy to
value (i.e, cash). Resistance by the target's management
could therefore affect the method of payment.
Persuasion is an important and critical component of
defence strategies in the UK. The target management usually
try to persuade their shareholders that they will be worse
of f under the bidder. Defence documents will almost
certainly refer to the "poor" financial and share price
performance of the bidder and the general "incompetence" of
the bidder's management. (Sudarsanam, 1991). The
effectiveness of these defence tactics will be undermined
in a cash offer (Peterson & Peterson, 1991; Sudarsanam,
1994b) . With cash the bidder can reduce the scope of the
target's management defence strategy.
Cash may also have a pre-emptive role in a hostile
takeover. The use of cash as a method of payment can signal
that the bidder has a high valuation for the target and
hence forestall the emergence of a competing bid (Fishman,
1989)21.
2.12. Previous research on the choice of payment method
There is limited published empirical research
investigating the choice of payment method in corporate
acquisitions.
Carleton et al (1983) was the first paper to
21 A bidder with a high valuation for the target will offer a high
premium. This discourages rival bidders because of the increased cost
associated with mounting a bid.
53
specifically examine the factors that may influence the
choice of payment in corporate acquisitions. They estimated
binomial probit regressions comparing equity and cash
offers using a sample of 61 firms (30 cash offers and 31
equity offers) which were acquired during the years 1976
and 1977. They found that:
(1) the higher the dividend payout ratio of the
target, the higher the probability that the target would be
acquired via an equity offer. However they were unable to
offer any explanation for this result.
(2) the lower the market-to-book ratio of the target,
the higher the probability that the target is acquired in
a cash offer. Carleton et al suggested that where the
market-to-book ratio was positively correlated with the
amount of goodwill created in a cash offer, then this
result would be consistent with an attempt by bidders to
avoid the creation of goodwill and its resultant depressing
effect on earnings per share, when amortised.
Hansen (1987) focused on the role of debt and size on
the choice of payment method. His model predicts that the
probability of an equity offer is negatively related to the
relative size of the bidder to the target, the gearing of
the target and positively related to the gearing of the
bidder (see Section 2.4.1). He used a sample of 106
acquisitions over the period 1976-78. He estimated some
Logit regressions comparing equity offers and cash offers
using the bidder's and the target's gearing and the
54
relative size of the bidder to the target 22 as the
explanatory variables. In some of his models Hansen found
that the probability of an equity offer increased with the
size of the bidder's total liabilities (not deflated for
size), although in other models this result was not
sustained. Additionally none of the measures of gearing in
Hansen's study was significant. Hence the best conclusion
we can reach is that Hansen provides moderately supportive
evidence that the level of total liabilities of the bidder
might have an impact on the method of payment.
Amihud, Lev and Travios (1990) examined the role of
size and insider ownership as determinants of the method of
payment. Using a sample of 209 acquisitions over the period
1981-83, they ran binomial Probit regressions comparing
equity financed and cash financed acquisitions. Amihud et
al found a significant negative relationship between the
probability of an equity offer and the fraction of shares
held by the bidder's managers. This is consistent with the
hypothesis that managers use the method of financing to
increase their control over the firm. However they did not
find any evidence in support of Hansen's (1987) proposition
that the size of the target relative to the bidder affected
the method of payment.
Higson (1990a) is the only study which examines the
choice of the method of payment using UK data. Using a
sample of 373 completed takeovers over the period 1976 to
1987, he estimated Logit regressions comparing cash offers
22 Size
was measured as the book value of total assets
55
with equity offers. Higson found a significant positive
relationship between the probability of an equity offer,
the size of the goodwill arising on the acquisition and the
recent returns on the bidder's equity. Higson also found a
significant positive relationship between the probability
of a cash offer and the bidder's liquidity.
While Higson's study is important, the methodology
adopted can be improved. Higson used Logit models with a
dichotomous dependent variable equal to 1 for equity offers
and 0 for cash offers. In order to get a two group
classification Higson arbitrarily defined bids as equity
offers when over 5O of the consideration was equity and as
cash offers when over 5O of the consideration was cash.
This classification rule did not recognise that a whole
range of values is possible between 5O and lO0, if l0O&
is regarded as the theoretically pure definition of an all
cash or all equity offer. Higson also used some cash flow
variables as proxies for the free cash flow hypothesis.
However no attempt was made to distinguish between bidders
with high cash flow and those with free cash flow23 . This
may partly explain why the cash flow variable was not
significant in Higson's study.
Mayer & Walker (1992) used a sample of 181 bids over
the period 1979-1990. This is the first study that moves
away from the traditional 2 group Logit model and uses a
methodology in which bids with a mixture of cash and equity
23 Cash flow is only free if the firm has run out of positive NPV
projects (Jensen, 1986) . Hence a firm can have high cash flow which is
not free provided the firm has got profitable investment opportunities.
56
can be handled easily by the statistical model (i.e., the
Two Limit Tobit Model, see Section 4.2.1). They found a
significant positive correlation between the probability of
a cash offer, the fraction of shares held by managers in
the bidder and the bidder's liquidity.
Smith & Jennings (1993) examined whether information
asymmetry between the bidder and the target affected the
choice of payment method (see Section 2.4). They used the
following variables as proxies for the uncertainty
surrounding the value of the bidder's and the target's
assets:- coefficient of variation of analyst forecast of
the firm's earnings, number of analysts making an earnings
forecast about the firm, relative size of the bidder to the
target. Using a sample of 140 acquisitions occurring over
the period 1979 to 1987, they estimated Logit regressions
comparing equity offers and cash offers. Although the
variables used as proxies for information asymmetry had the
predicted sign none was statistically significant.
Additionally they found that the likelihood of a cash offer
was significantly positively related to the potential value
of stepping-up the target's assets, the proportion of the
bidder's shares held by management and competition from
other bidders. They concluded that the weakness of their
results regarding the information asymmetry variables could
be due to the substantial noise in the proxies for
unobservable private information.
There is an alternative explanation for the weak
results in Smith & Jennings (1993) . The discussion in
57
Section 2.4 suggests that information asymmetry on its own
does not affect the method of payment. It is the possible
overvaluation of a bidder resulting information asymmetry
that causes the acquiring managers to offer equity.
Similarly it is the desire of a bidder to avoid overpaying
for the target, which results in an equity offer. When
investors are not fully informed about the value of a firm
(ie, there is information asymmetry in the market) the firm
could be either overvalued or undervalued. If information
asymmetry causes a firm to be undervalued, we would expect
to observe a cash offer. In their proxies Smith & Jennings
measure the level of information asymmetry, they are not
measuring the over or under valuation in the bidder and the
target, which is what drives the use of equity as the
method of payment.
The most significant methodological shortcoming in the
existing literature is the use of the single equation Logit
model. The problems with this model have been discussed in
Section 1.2.4.
2.13. Conclusion
In this chapter we have reviewed the literature
concerned with factors likely to influence bidders in their
choice of payment currency. The main picture that emerges
is that the determinants of the method of payment in
corporate acquisitions are diverse and complex. No single
theoretical model can adequately encompasses all these
factors. A summary of the variables identified by the
58
0
C
U)
a,
0
C)
CD
C 0
C C)
a) C)
QD
C)
t.C
C)
V.-
a)
C
a) .Cl)
0
Cu
a)
0
>
4J
-H
Cu
•1
a)
U)
CO
0
I-
a.
E
a)
CO
0
>
a)
0
-a
-o
a)
C
a)
C
E
gI
E
CU
><
a)
CD
x
a)
>-.
U)
0
>
a)
ii
a
zgz
4-.
0
z
C U)C
CD
CO
a,
4-
C
CCC
0
a.
0
Z.CZ
4-
-DC)
a) '-
-d
C
2'
(I)
40
z
a)
C
'-0
CDU,
E
CO
>1
>1
-C
C) .2
a)
>U,
U)
0
>
CD C) >
.&i -, U'
a)
.0)
COlD
-a
a)D a)
C c
0CD
E
CO
>
E
a)
V
C
C
C')
0
>
CD
0
.9
>CD
-C
0 .2'
ZC/)
a)
- Z
CO
a
40
z
C/)
0
p4
U)
0
0
4-.
4-
a)
CO
-C
ci.
CO
0
U)
.lJ
a)
0
0
4J
a)
E
a)
4J
0
a)
U)
C,
4-.
CO
1-'
U)
CO
a)
Ui
p4
C')
C
a)
a)
E
CD
a4a
-a
0
-c
4-.
a)
E
a)
0
a)
H
U-4
CO
a
E
c
a.
CD a.
U)0
'-' 0
a.
a)
a)
C
. >
C.
o a0
C -a
CD
'-'O
OCD
-C
a)0
'4-.
a)'-'-a)
'4-'4- 0
CD
U
0 a-a)
U) -
CD
a) -
4-.
0
a)
a)>
U)0
a00
I—co
_c
CDa)
0
C
a)'0 -a .
a-a
ci) 0
1-a)
>
4-a)
c a,•I-
_C
CD
- — a)
4-I
a)
aJC
a)
-a
CO
HE
CO
>
a)
o
_a .
-C
4-.
'I0
90
a,
0
CD
a),-.
>
.2
4-
CD ,-.
a)
-a
CO (I)
Ia)
a)
-C
a)
-C
9a)
- 0
U) .' 0
CO
4-'CD
> C
a,
C
0
C
CD
C
>'
a-
-
COU)
4
4-
—
CDU)Q
CCO
a)
90
a)
oE
4-,
CU,
-
0a)
>
a 1
CO Cl)
o 0
>CO
o2
a)
U)
a).
0..
a)
aCD
'-a)
CD
-C-COU)
a)
4-.
_c '
4-, a)
>
C
IU)
0
a,
C
C
a)
-C 4-
a)
D
—0>
( Ti >
CO-5
E
-D U)
_5
0
U)
>CD
II
rO
>CD
-CO
0
0
0
U)
a)
a)
a,-0
U) -
CD
CD
Cl)
0
Cr1
4-I
EC
a)
I-
r1
0
CO
C
>CO
4-.
C
0
0
a)
>
U)
>
-C
-H
U)
1_
a,
ICO
a)
ICD
4-. 9-
a) 4-C4'0
a)
4-.
CO
4-.
a)0
U,
-C
0
COa)
4-. 1
U)CO
C
0
4CO
CD'-.
a)
-a
a,D
a,
CD
I--
>a)
i5-a
0.-a
C,
C
-C
0
-C
U)
1
CO
-C
C0
C
a,
C
0
-C
0
0
-05
-s-a
CO
4-
a
CD
0
a)
-c
4-.
C
U)
C
0
CD
a.
CD>
Oa
59
CDa)
4-a)
>1
LI_I
0CD
oE
-a
a,
>CO
C
Co
C
CO
'- i__U) -
N
QD
-
0)
C
a,
C']
0
-H
0
C
a,
I,
>
C
CO-C
2
>
a)
1.
-
E
CO
><
U]
-H
0
(Ii
a
E
a,
U)
a)
4J
('j
0
0
>
a)
I-.
ci
-
a)
4
t
C
Co
CO
a)
> .2
C
o
Z3&5
(f )
0
0
E
CO
X
a)
a
U)
'4-
0
oa)
.0 C
a)CO
C
U,
a)
E
>
ID
C c C/
Ia).Co
4-,
0
XC,)
a,U)
CO
.4J
a
a)
4-
E
0
ci)
-a
0
4-.
a)
E
a)
CC,)
-a
a)
0,99- a)
o.
O)a)
>u,
I-
C
0
4-.
Co
a
E
-
4-. 0)- CO
cC
oCO
Q)a)C
2
Co - C
CO5
4-, 9- '4-
e - 2
Xa)>.
0- 0
zz
(C)
a) a) -
U' ,, .2
4-' 4-'
a,CO
o ,
-
-C
U)
CO
-a
C
a, 0 _c .c
Cl)
0)
i__
C) - ,
a
-
CoC
cO
:
>.
acl)
- CO
4
OOU)
zz
-- cn
40
z
0)
C
0
U)a,
CO.C)Co
-VU)
COO
0a)
'4U) ._
; 0
=
_; 0
> 9-
--C
C/,
2co
9- 0
a)
Cl)cD
U)
(/)
Co
4-.
9- 0)
-C .-U)
Cl)
'4-
-
a,
a) 0
CO
4-
a)
4-.
o 9-'._
E C
co
a)
>
0
C
9U)
a)
U) a)
0Co
OCO
EE
a)-
C 0
C)C
4-.
E
0
0
U)
-
.c2c
0 1
o
4-.
>C
0
>
Ia
U)
1
a C)
-a
CO
1k
U) co
O -
a)
U)
a)
I-
a)
4-.
E
CO
X
a)
>
U)
U)
a
U)
CO
a)
C
U)
C
0
4-.
C
CO
0
U)
a)
0
'40
4-.
1
0)
4-.
rCO a)
C
2 E
9- IOu,
-U)
U) - -
a) C
a) -- Co
'9r C/)
0)
C-)
C
Co
C
a,Oa)
.5? > -D
- .0
0
-- 0
H
(I)3&)
o >
9- '4-
a)
.2
>
E
C)
a)
-H
0
Co
o
E" E
0
z
ECO
b)
CO
a
>
a,
1.a.
0)0
CU)
C
>CO
0
C
.2'- 0
C/)Z
CO
C 4-' 4-.
C 4
- C
CO C Co
0Co0
U)
I-
0
-
4-. 0) 0
-c
U,
-H
9-i
0
C
a)
0
>9-i Cl)
a)
C
a,
-H
CO
0
CO
Co C0)
U) 0
ci
)
-
C/,
c 0)
a) 0,_S
C
a,
0
0)
C)
Co
0
0)
0)
-
U) 4-
4-.
-- CO
-H
>a, .0
99- a)
00
C
CO
COU)
0 -.' i__
a)
iu
C
a)
-c
4--.
90
a,
-c
09-i
a,
a)
C-
ti-I
0
0
I-
OC
a,
-c
4-.
90
C/,
a)
0
U)
0
U)
Co
a.
Oaj
0
-i-1
4J-J
0C
rO
a)
rE
QE
(
.0
(0
Co
>
4-a)
-ao
CO -0
o6
4-.
90
CO
-
U)
a)
a
a)
oCo
-aa,
0a,E'
CO
'-I
4-' 9-
U)
a,
60
1
_c a,
U)
Co 0
a,
'40
U)
0
I0
U,
a,
I-
a)
U) C)
CoCo
o 4-
0
'4-C
U)
CO
0
9U)
Ia)
Ca,
a)
a)
> CO
.0 C
CO
4-.
II
a,
0
literature review is provided in Table 2.1.
A robust result which emerges from the US based
studies on the determinants of the method of payment is the
negative relationship between the use of equity and insider
control of voting rights (Amihud et al, 1990; Mayer &
Walker, 1992; Smith & Jennings, 1993). This issue has not
been examined by any UK studies.
The literature review has revealed a number of
variables which have been found to be related to the
financing decisions of the firms in other areas, but which
have not been studied within the context of corporate
acquisitions: (i) choice of accounting policy (ii) dilution
of shares held by external blockholders (iii) free cash
flow of the bidder (vi) growth opportunities in the bidder.
Our review reveals a disturbing lack of UK based
empirical literature on the determinants of the financing
method in acquisition. The only UK based study in this area
is Higson (1990a). Additionally some variable which have a
significant impact on acquisition financing in the US
literature have not been tested in the UK: (ii managerial
shareholding in the bidder (ii) capital structure of the
target (iii) liquid resources of the target. Later in this
thesis we attempt to fill this gap in the literature.
In this chapter we have argued that the method of
payment and the choice of accounting policy are jointly
determined (Section 2.3. above) . This interaction requires
the use of a simultaneous equations framework which
necessitates that we identify the determinants of the
61
accounting policy choice in corporate acquisitions. The
next chapter discusses these factors.
62
CHAPTER 3.
DETERMINANTS OF THE ACCOUNTING METHOD IN MERGERS AND
ACQUISITIONS :- THEORY & EMPIRICAL EVIDENCE.
3.1. Introduction
There are two methods of accounting for a business
combination i.e, merger or acquisition accounting. While
the cash flows under both methods may be the same and,
therefore, the choice of accounting method may have little
valuation impact on the firm's securities, managers may
still perceive relatively greater benefits from one
accounting method than from the other.
There are several possible economic motives for
managerial preference for a particular accounting method.
Among these are: the impact of accounting numbers on
management compensation, dividend restrictions from debt
covenants, restriction on borrowing capacity imposed by
such covenants and the political cost of reporting high
earnings. Since business combinations give rise to monopoly
concerns and tend to attract antitrust regulatory scrutiny,
the political cost of reporting higher earnings is
particularly important in the choice between merger and
acquisition accounting.
One of the key determinants of the accounting policy
is the method of payment. This relationship is a direct
result of the rule which requires that at least 9O of the
fair value of the total consideration offered by the bidder
must be equity in order to qualify to use merger accounting
63
(SSAP 23)'. Conversely, this 9O rule implies that the
method of payment is affected by the choice of accounting
policy.
In this chapter we discuss the relative merits of
acquisition and merger accounting and the factors which are
likely to influence the choice of accounting policy by the
bidder' s managers.
3.2. UK rules governing the method of accounting for
corporate acquisitions
Accounting for business combinations is a topic which
has caused considerable controversy in the UK. The two main
methods of accounting f or a business combination are
acquisition (purchase) or merger (pooling) accounting.
Traditionally the acquisition method was the main
technique of accounting for business combinations. However
in the 1960s, as a result of its popularity in the USA, the
merger method began to find favour in the UK. ED 3
'Accounting for acquisitions and mergers', issued in
January 1971 was the first attempt to specify the
situations in which each method could be used.
ED 3 was never converted into an accounting standard,
because there were doubts as to the legality of merger
accounting. As the merger method required that shares be
recorded at nominal value rather than their market values,
'SSAP 23 (issued in April 1985) which was the relevant standard
operating for most of the sample period (1/1/80 to 31/12/90) covered by
this thesis was withdrawn in September 1994 and replaced by Financial
Reporting Standard (FRS) 6.
64
it appeared to contravene Section 56 of the 1948 Companies
Act. The illegality of the merger method was finally
confirmed in the case of
Shearer vs Bercain Ltd 1980.
The 1981 Companies Act legalized merger accounting by
providing that a share premium account need not be set up
for any equity shares issued as consideration where one
company has acquired at least 9O of the equity in another
company. This provision permits but does not require the
use of merger accounting.
3.2.1. Statement of Standard Accounting Practice (SSAP) 23
Following the 1981 Companies Act, the Accounting
Standards Committee issued SSAP 23 in April 1985. SSAP 23
requires that all business combinations should be accounted
for using acquisition accounting unless all of the
following conditions have been complied with:(1) there must have been an offer to the shareholders
of the target for all the shares and all the voting shares
not held by the bidder at the date of the offer;
(2) the offer must result in the bidder securing
ownership of at least 9O of all equity shares (taking each
class of equity separately) and at least 9O of the votes
of the target;
(3) prior to the offer, the bidder must not hold more
than 2O of either the equity or the votes of the target;
(4) at least 9O of the fair value of the total
consideration given by the bidder to secure ownership of
the equity and the equity non-voting shares of the target
65
must be in the form of the bidder's equity.
The 1989 Companies Act now gives legal status to the
rules governing the use of merger accounting contained in
SSAP 23. However in one respect, the 1989 Act is more
demanding in defining the conditions under which merger
accounting may be used. For merger accounting to apply, the
cash element of the total consideration should not exceed
1O of the nominal value of the shares issued as opposed to
lO of the fair value of total consideration which is the
looser condition specified by SSAP 23.
These rules would imply that merger accounting is only
applicable where the method of payment offered by the
bidder is primarily equity. Cash consideration would be
consistent with acquisition accounting. However devices
exist for ensuring that the conditions for merger
accounting are satisfied even though the method of payment
received by target shareholders is primarily cash.
In the 1986 Hanson bid for Imperial Group or Turner &
Newall's bid for AE, nominee companies funded by the bidder
were sent into the market to acquire the target's shares
for cash. These nominee companies then accepted the
bidder's equity in a general offer and were subsequently
wound up. The bidder's shares held by these nominee
companies were placed in the market. Since the bidder had
made an equity rather than a cash offer, it could obtain
the benefits of merger accounting.
Vendor rights and vendor placing schemes (see Section
2.2.1) also allow companies to violate the spirit of the
66
SSAP 23 rules. These schemes allow bidders wishing to buy
a company for a cash consideration, but who also wish to
use merger accounting to have their cake and eat it. The
bidder would make a share for share exchange offer, while
its merchant bankers would arrange for the target
shareholders to convert the shares they receive into cash,
either by selling them to a third party (vendor placing) or
selling the shares to the bidder's own shareholders (vendor
rights) . The final result is that shares are purchased for
cash while technically merger accounting can still be used.
Since merger accounting is not available, if the
bidder holds more than 2O of the target's shares prior to
the offer, bidders who fall foul of this rule, can
temporarily 'warehouse' the offending shares with their
bankers or a friendly third party and buy them back in a
general offer.
3.2.2. Financial Reporting Standard (FRS) 6
In September 1994, the Accounting Standards Board
(ASE) issued FRS 6 and withdrew SSAP 23. FRS 6 addresses
some of the abuses of SSAP 23 discussed above and limits
the use of merger accounting. FRS 6 states that merger
accounting can only be used for a merger which is defined
as: "A business combination that results in the creation
of a new reporting entity formed from the combining
parties, in which the shareholders of the combining
entities come together in a partnership for the mutual
sharing of risks and benefits of the combined entity,
and in which no party to the combination in substance
obtains control over any other, or is otherwise seen
to be dominant, whether by virtue of the proportion of
67
its shareholders' rights in the combined entity, the
influence of its directors or otherwise."
The broad objective of FRS 6 is to limit the use of
merger accounting to only those business combinations where
the combining parties join together on an equal footing to
form a new enterprise for their mutual benefit. The FRS
contains five detailed criteria which must be met in order
to use merger accounting:1) no party to the combination is portrayed as either
acquirer or acquired;
2)
the board of directors of all parties to the
combination participate in the management of the combined
entity;
3) no party to the combination dominates the other
parties by virtue of its relative size;
4) the consideration received by equity shareholders
of each party to the combination comprises primarily equity
shares in the combined entity;
5) no equity shareholder of any of the combining
entities retain any material interest in the future
performance of only part of the combined entity.
Unlike SSAP 23, FRS 6 requires that if these
conditions are satisfied, then merger accounting is
mandatory.
3.2.3. Acquisition versus merger accounting
The main differences between acquisition and merger
accounting are summarised below:
68
ACQUISITION ACCOUNTING
MERGER ACCOUNTING
Investment in the target is
recorded in the bidder's
accounts at the fair value
of the consideration paid.
Investment in the target is
recorded in the bidder's
accounts at the nominal
value of the shares issued.
A share premium account
arises if the method of
payment used is equity
unless merger relief is
available.
No share premium account
arises.
Target's results are
consolidated from the date
of acquisition.
Target's results are
consolidated for the whole
of the year in which the
acquisition took place.
Target's net assets at the
date of acquisition have to
be restated at fair value
before consolidation in the
group's accounts.
Restatement of the target's
net assets is not necessary.
Goodwill should be
recognised as the difference
between the fair value of
the consideration given and
the fair value of the
net
separable
assets
acquired.
Goodwill is not recognised.
Any difference between the
nominal values of the shares
issued and the shares
acquired is treated as a
reserve
arising
on
consolidation.
Pre-acquisition reserves of
the target cannot be treated
as distributable profits by
the bidder.
All reserves are
distributable irrespective
of whether they are pre or
post acquisition reserves.
3.3. Implications of UK accounting rules for the choice of
accounting policy
The implications of the differences between merger and
acquisition accounting for the choice of accounting policy
by the bidder are discussed below.
3.3.1. Accounting treatment of goodwill
SSAP 23 provides for 2 methods of eliminating goodwill
in the consolidated accounts:
69
(1) Goodwill is written off immediately in the balance
sheet against reserves.
(2) Goodwill is capitalised and amortised through the
profit and loss account over its economic life.
The amount of goodwill to be eliminated can be subject
to a considerable amount of ingenuity on the part of
companies. The requirement to ascribe fair values to the
separable net assets of the target can result in bidders
making significant adjustments to the target's net assets
in the year of acquisition with the purpose of benefiting
earnings in future years. Provisions and write-downs can be
made against the assets of the target prior to
consolidation, thereby increasing the amount of goodwill2.
If the goodwill is then eliminated against reserves, it is
excluded from the bidder's profit and loss account.
Subsequent expenses are then written off to such provisions
in the balance sheet rather than to the profit & loss
account. Any over-provision being released to the P&L
account at a later date, if the original provisions prove
to be excessive.
The immediate write-off of goodwill has the advantage
that future earnings are not affected by the annual charge
2 Statement of Standard Accounting Practice No 22, Paragraph 14,
permits the setting up of provisions at the time of an acquisition in
respect of anticipated future losses or costs of reorganisation. Such
provisions reduce the net assets of the target and increase the value
of goodwill. Grinyer et al (1991) found that the use of provisions to
write down the cost of assets in the balance sheet was negatively
related to the bidder's gearing ratio. This suggests that bidders' with
a high level of gearing preferred to assign high values to their
targets' tangible assets in order to help reduce gearing ratios in the
post acquisition group balance sheet. The setting up of provisions for
future losses and reorganisation costs expected to be incurred as a
result of the acquisition is now prohibited by FRS 7 issued in
September 1994.
70
for goodwill amortisation. However the goodwill write-off
against reserves will deplete the balance sheet. The
acquisition of J Walter Thompson by WPP in 1986 all but
eliminated WPP's shareholders' funds. This can result in a
reduction in borrowing capacity and breaches of existing
loan covenants. In order to repair the damage done to the
balance sheet by the goodwill write-off, some companies
have resorted to valuing intangible assets like brands or
trade marks. In 1986 when Guinness acquired Distillers, it
had to write off £1.39bn by way of goodwill. In 1989, it
simply added back £1.38bn by way of brand valuation.
The amortisation of goodwill through the P&L account
has the clear disadvantage that future earnings are
adversely affected. Consequently most bidders may find this
method unacceptable. The UK is unique in allowing an
immediate write off against reserves. In the USA, it is
mandatory that goodwill is amortised through the P&L
account over a maximum period of 40 years, while in
Continental Europe the mandatory write off period is
between 10 to 20 years. Given the international dimension
of takeover activity, foreign companies have argued that UK
companies have an unfair advantage when bidding against
foreign competitors, because the immediate write off method
leaves their future earnings unaffected (Ivancevich, 1993).
Under acquisition accounting goodwill should be
recognised and somehow it has to be eliminated. Merger
accounting however avoids all the problems associated with
eliminating goodwill. Where an acquisition will result in
71
the recognition of a large element of goodwill, the bidder
has the incentive to structure the method of payment so as
to obtain the benefits of merger accounting and avoid the
recognition of goodwill.
3.3.2. Merger relief
The desire to use merger accounting as a route to
avoiding the problem of goodwill may not be the only reason
why bidders may choose to offer equity as a method of
payment. Prior to the 1981 Companies Act only the
consolidated revenue reserves could be used to write off
goodwill. The 1981 Companies Act introduced the concept of
"Merger Relief". Under the merger relief provisions where
a bidder issues shares to acquire a subsidiary and the
shares issued have a market value greater than the nominal
value, then the difference can be credited to a merger
reserve account rather than the usual share premium
account. Any goodwill arising on the acquisition can be
written off against the merger reserve.
The following conditions have to be satisfied in order
to obtain relief from the requirement to create a share
premium account:
(1) the bidder must secure ownership of at least 9O
of each class of equity in the target;
(2) at least part of the consideration offered by the
bidder must be in the form of equity shares.
Since there is no requirement that a minimum
percentage of the total consideration paid by the bidder
72
must be in the form of equity, the size of the merger
reserve against which any goodwill can be written off will
be directly proportional to the amount of equity in the
final consideration.
It is not immediately clear that the use of the merger
reserve account to write off goodwill has any advantage to
the bidder. If goodwill arises on consolidation, then
immediate write off against reserves has no effect on
distributable profits 3 . Consequently it is arguable whether
the elimination of goodwill against merger reserve as
opposed to the profit and loss reserve results in any
advantage to the bidder. However there appears to be a
belief amongst company directors that a large profit and
loss reserve in the group's consolidated balance sheet is
a sign of good corporate health and therefore this reserve
should not be depleted by the write off of goodwill. (See
Holgate, 1990: p.20)
The belief that writing off goodwill against the
profit and loss reserve gives an unfavourable impression to
investors has led to the practice in the UK of eliminating
goodwill via a dangling debit when a merger reserve account
is not available. Under this technique a new reserve is
started with a zero balance and goodwill is written off
3 me distributable profits of the group depend on the size of the
accumulated profit and loss reserve in the holding company's balance
sheet. This is not affect by writing off goodwill on consolidation
against the group's accumulated profit and loss reserve.
73
against it. The result is of course a debit balance 4 . This
method has the advantage that the user is able to see the
total of goodwill on all acquisitions since the policy was
first adopted.
3.3.3. Bidder's access to the target's pre acquisition
reserves
Pre acquisition reserves refer to the accumulated
revenue reserves of the target at the date of the
acquisition. Under acquisition accounting and prior to the
1981 Companies Act, the pre acquisition reserves of the
target have to be capitalised by the bidder and are not
available for distribution in the future as dividends by
the bidder. Any dividend received by the bidder out of the
target's pre acquisition reserves is used to write down the
cost of its investment in the target. The bidder cannot
treat such dividends as realized profits which could
subsequently be distributed to its own shareholders.
The 1981 Companies Act, now reiterated in Schedule 9
of the 1985 Companies Act, introduced changes to the
treatment of dividends paid out of pre-acquisition profits.
Where a dividend is paid out of pre-acquisition profits, it
does not necessarily have to be used to write down the cost
4 LIT Holdings Plc used acquisition accounting with a dangling
debit to account for its takeover of Johnson Fry Plc in 1988. In its
last balance sheet prior to the acquisition, LIT had total capital and
reserves of £43.7m including accumulated revenue reserves of £13.1.
Goodwill recognised from the takeover was £24.lm which would have
eliminated all of LIT's revenue reserve if goodwill had been written
off to the revenue reserve account. Presumably to avoid disclosing a
negative figure for accumulated revenue reserve LIT wrote off goodwill
via a dangling debit.
74
of the investment in the target, except to the extent that
it is necessary to provide for a permanent diminution in
the value of the investment in the target. If such a
provision is not necessary, dividends received out of preacquisition reserves can be treated as realised profits in
the hands of the bidder and can legally be distributed to
the bidder's shareholders.
This change in the law regarding the treatment of pre
acquisition reserves, legalised the merger accounting rules
which allow dividends paid out of pre acquisition reserves
to be treated as distributable profits by the bidder.
Where the bidder is likely to need the pre acquisition
reserves of the target for paying future dividends to its
own shareholders, the probability of using merger
accounting in the consolidated accounts will be an
increasing function of the size of the target's pre
acquisition reserves. Bidders with declining dividend cover
or a high ratio of current dividends to revenue reserve may
prefer merger to acquisition accounting where the target is
relatively liquid and has significant pre acquisition
reserves.
3.3.4. Depreciable fixed assets and post merger profits
Acquisition accounting requires that when
consolidating a subsidiary's activities for the first time
in the parent company's accounts, all the assets of the
subsidiary must be restated to their fair value at the date
of the acquisition (Companies Act 1985, Schedule 4A and
75
SSAP 23) . The requirement to restate assets to their fair
value can have an impact on the future reported profits of
the group. In order to analyse this potential impact we
must distinguish between two broad classes of depreciable
fixed assets.
1)
"Value Increasing Depreciable Fixed Assets"
(VIDFA) : These are depreciable fixed assets whose values
are likely to increase over time. These assets would have
a fair value which is likely to be greater than the
reported net book values eg, property which is not classed
as an investment property and brands.
2)
"Value Decreasing Depreciable Fixed Assets"
(VDDFA) : These are depreciable fixed assets whose values
are likely to decrease over time. These assets would have
a fair value which is likely to be lower than the reported
net book values eg, plant and machinery.
When VIDFA are restated at fair value this can have a
negative impact on the future reported profits as higher
depreciation charges are passed through the profit and loss
account. Restatement of VD]JFA will have a positive impact
on future profits, since future depreciation charges will
be lower. Nurnberg and Sweeney (1989) show that when the
fair value of assets is increasing, higher post combination
profits are reported under merger accounting, while
decreasing fair asset values result in acquisition
accounting showing higher post combination profits.
The probability of the bidder adopting merger
accounting is likely to be an increasing function of the
76
proportion of the target's net worth which is composed of
value increasing depreciable assets and a decreasing
function of the value decreasing depreciable fixed assets
included in the target's net worth.
3.3.5. Enhancing current period earnings
Acquisition accounting requires that the target's
activities can only be consolidated from the date of
acquisition on a time apportionment basis. This implies
that in the year of acquisition only that proportion of the
target's profits which has accrued from the date of
acquisition can be included in group profits. Merger
accounting, however, treats both the bidder and the target
as one single entity for the entire year in which the
acquisition took place (SSAP 23) . A bidder with small
profits may wish to utilise the current year's preacquisition profits of the target in boosting the reported
profits of the group. Merger accounting offers the
advantage of reporting higher profits, especially in
combinations involving a low profit bidder and a high
profit target.
By making an equity offer bidders may be able to "buy"
earnings and incorporate the target's full year's results
in the consolidated P&L account even if the acquisition
takes place at the end of the relevant accounting period.
Although the reported earnings may rise so does the number
of shareholders, hence the impact on earnings per share may
not be uniform across all bidders. The effect of the
77
acquisition on the EPS depends on the relative price
earnings ratio and the share exchange ratio (Sudarsanam,
l990a)
Where the level of the target's earnings in the year
of acquisition is a consideration for the bidder the
probability of the latter using merger accounting is an
increasing function of the profitability of the target
relative to the bidder.
3.3.6. High reported earnings and their political cost
Watts & Zimmerman (1978) argue that the "political
costs" of reporting large profit numbers increases with the
size of the firm. Large firms reporting maximum profit
figures are more likely to face increased regulatory
pressure in terms of challenges to their acquisition
programmes and pricing policies from the anti-monopoly
agencies. In order to obtain a quiet life large firms are
more likely to choose conservative accounting policies
which do not result in the reporting of maximum profits.
(Watts & Zimmerman, 1978; Hagerman & Zmijewski, 1979;
Bowen, Lacy & Noreen, 1981; Zmijewski & Hagerman, 1981;
Daley & Vigeland, 1983)
In the UK, it is not immediately obvious whether the
reporting advantage lies with merger or acquisition
accounting. Since goodwill can be eliminated against
reserves, the future accounting profits of the group are
not affected by the recognition of goodwill. In the year of
the combination, the consolidation of pre-acquisition
78
earnings favours merger accounting. Conversely acquisition
accounting offers the opportunity to create provisions in
the combination year which can then be subsequently fed
back to inflate earnings in future years 5 . Additionally
acquisition accounting allows the bidder to write down the
value of the target t s tangible fixed assets prior to
consolidation which reduces future depreciation charge with
a favourable impact on future profits.
3.3.7. Goodwill accounting and breach of debt covenants
In order to control the agency conflict between
shareholders and bondholders, debt agreements have
restrictive covenants which limit the financing, investment
and dividend policies of the firm (Smith & warner, 1979).
Citron (1992c) found that these restrictions usually
circumscribe the ability of managers to increase the
gearing of the company and dispose of assets.
Since generally accepted accounting principles are
used in the measurement of these accounting based
covenants, the closer a firm is to breaching a specific
covenant, the more likely it is that it will adopt
accounting policies to help avoid a breach of the
particular covenant.
A number of empirical studies have demonstrated that
a high debt/equity ratio is associated with the adoption of
income increasing accounting policies (Dhaliwal, 1980;
5 The creation of provisions for future losses or reorganisation
costs at the date of acquisition is no longer permitted under FRS 7.
79
Holthausen, 1981; Bowen et al, 1981; Lilien & Pastena,
1982; ]Jaley & Viegland, 1983; Christie, 1990) . Implicit in
all these studies is the assumption that a high debt/equity
ratio is a reasonable proxy for closeness to breach of
accounting based covenants.
Since the elimination of goodwill under acquisition
accounting will eventually result in the reduction of the
bidder's net worth and possible violation of gearing
covenant restrictions, bidders with high gearing may have
a preference for merger accounting. However Citron (1992b)
suggests that the impact of gearing covenant restrictions
on the choice of accounting policy in the UK may not be
significant. He finds that where breach of a covenant is
caused by goodwill write off, the relaxation of the
covenant is the most likely response of the lenders.
Additionally he finds that it is not uncommon for loan
agreements in the UK to provide for goodwill on
acquisitions to be included in net worth for the purpose of
calculating gearing.
3.3.8. Summary of implications of accounting rules for the
choice of accounting policy
The above discussion suggests that the choice of
merger accounting by the bidder
iS:
i) affected by the method of payment, due to the rule
that equity must form at least 90 of the total
consideration paid in order to use merger accounting;
ii)
positively related to the size of goodwill
80
realisable in the acquisition (Section 3.3.1), proportion
of the target's assets composed of value increasing
depreciable fixed assets (Section 3.3.4), size of target's
pre acquisition reserves (Section 3.3.3), relative
profitability of the bidder to the target (Section 3.3.5)
and bidder's gearing (Section 3.3.7);
iii) negatively related to the size of the bidder and
the target (Section 3.3.6) and proportion of the target's
assets composed of value decreasing depreciable fixed
assets (Section 3.3.4)
3.4. Empirical evidence on the accounting policy choice
The preponderance of empirical evidence, on the
determinants of accounting policy choice, has come mainly
from the US. Accounting for business combinations in the US
is governed by Accounting Principles Board Opinion 16 (APBO
16) . APBO 16 is similar to SSAP 23 in the UK in that it
requires that at least 90% of the consideration paid must
be equity in order to use merger accounting. However there
is a sharp difference in the prescribed treatment of
goodwill under acquisition accounting between the US and
the UK rules. APBO 16 requires that in the US any goodwill
recognised under acquisition accounting must be amortised
through the P&L account over a maximum period of 40 years,
while in the UK bidders have a choice of either amortising
goodwill through the P&L account or writing it off against
reserves.
In the US bidders face a choice between not
81
recognising goodwill (merger accounting) and amortising
goodwill against future profits (acquisition accounting).
Although in the UK goodwill is recognised under acquisition
accounting, its elimination does not necessarily depress
future profits. The difference between the US and the UK
treatment of goodwill could reduce the relevance of US
results for understanding accounting policy choice
decisions by UK bidders.
Because the US accounting rules require that goodwill
must be amortised through the P&L account, most of these
studies have concentrated on examining the impact of
goodwill on the merger/acquisition accounting choice.
Gagnon (1967) was the first study to investigate
whether under certain conditions a business combination
would be accounted for using merger or acquisition
accounting. Gagnon's study was driven by an income
maximising hypothesis. He argued that the choice between
merger and acquisition accounting could be used to
manipulate future income, because different asset values
are reported under the two methods. If the price paid for
the target (P) exceeds the book value of the target's
assets (BV) merger accounting would minimise the asset
value reported by the combined entity since goodwill is not
recognised under merger accounting. Where P is less than By
then acquisition accounting would minimise asset values.
Gagnon suggested that the method which minimises asset
values would maximise profit, since this reduces future
charges against income. Gagnon defined a predictor variable
82
K= P-By
E
where
P = price paid for the target
By = Book Value of the target's assets.
E = Expected earnings of the combined firm6.
Gagnon's hypothesised that 1idders would choose merger
accounting when K was positive and acquisition accounting
when K was negative. His sample consisted of 219 equity
offers from the years 1955 and 1956. Gagnon found that,
where P-By was positive, the probability of merger
accounting being used was positively related to the size of
K. However contrary to prior expectation his results also
showed a similar positive association when P-By was
negative. Gagnon concluded that
K
was a good predictor of
accounting choice only when P-By was positive. Although he
discusses other variables which could influence the choice
between merger and acquisition accounting (eg size of the
bidder) Gagnon did not present the results of any
multivariate regressions.
In discussion articles both Sapienza (1967) and Wyatt
(1967) criticise the Gagnon study for the choice of time
period used. Sapienza points out that, in 1955, the
understanding of merger accounting among practitioners and
6 Expected earnings is defined as the market value weighted average
of the bidder's and target's earnings over the ten years preceding the
bid multiplied by the accounting rate of return over the same 10 year
period.
83
academics was limited. Wyatt observed that over the period
to 1967, there had been a substantial erosion of the
stringent guidelines for using merger accounting. As a
consequence there was a trend towards merger accounting
with the possible result that combinations accounted for as
acquisitions in 1955-56 would probably be treated as
mergers in 1967, when Gagnon's study was published, thus
questioning the relevance of results based on 1950's data
f or researchers and practitioners in later periods.
Copeland & Wojdak (1969) addressed the time criticism
directed at Gagnon's study. Using a sample of 118 randomly
selected equity offers over the period July 1966 to July
1967, they found that there had indeed been a significant
trend towards increased use of merger accounting since the
sample period chosen by Gagnon. 51% of Gagnon's sample had
used merger accounting compared to 85% for Copeland and
Wojdak. Copeland and Wojdak also found that when P-B y was
positive 92.66% of bidders chose merger accounting, while
when P-By is negative, only 55.55% of bidders chose
acquisition accounting. The results indicate that when P-By
was positive there was a greater probability that the
bidder would chose merger accounting. When P-BV was
negative the results did not establish any clear preference
by bidders between merger and acquisition accounting.
Anderson & Louderback (1975) investigated the impact
of Accounting Principles Board Opinion 16 (APBO 16) on the
choice of accounting method. In October 1970, APBO 16
introduced the rule that the bidder had to acquire at least
84
9O of the target's shares via an exchange of voting shares
in order to qualify for merger accounting. They wanted to
examine whether the results of Gagnon and Copeland & Wojdak
were still applicable in the post
16 period. The
APBO
sample consisted of 114 takeovers in the pre
and 64 takeovers in the post
that when
P-B y
APBO
P-B y
16 period
16 period. They found
was positive about 87 of bidders chose
merger accounting in both thepre and post
When
APBO
APBO
16 period.
was negative 66 of bidders chose acquisition
accounting in both the pre and post
APBO
16 period. They
concluded that the stricter rules introduced by
APBO
16 did
not significantly influence the choice of accounting method
by the bidder.
Nathan (1988) investigated whether bidders were
willing to pay higher bid premia in order to obtain the
benefits of merger accounting. Nathan suggested that the
conditions for using merger accounting contained in
APBO 16
made it impossible to use merger accounting without the cooperation of the target's management 7 . If bidders prefer to
use merger accounting then the target may demand a higher
premium for its co-operation. Using a sample of
461
takeovers over the period 1963 to 1978, Nathan regressed
the bid premium on proxies for potential goodwill write off
7 unfortunately, Nathan did not clarify his argument, that the cooperation of target managers was required for the bidder to use pooling
accounting in the US. The use of pooling accounting is not barred as a
result of resistance by target managers under the rules laid out in
APBO 16. Additionally, Nathan's argument suggests that target bid
premia will be higher in friendly bids than in hostile bids. This is
inconsistent with the evidence from Huang & Walkling (1987); Franks,
Harris & Mayer (1988); Franks & Mayer (1993)
85
and several control variables 8 . Contrary to his hypothesis
that goodwill was positively related with bid premium9,
Nathan found a significant negative relationship between
goodwill write off and bid premium.
To explain this result Nathan suggested that the
goodwill write of f variable was a proxy for the target's q
ratio' 0 . If a low q ratio was a sign of inefficient
management, then low q targets may offer greater scope for
realising post merger gains and hence attract higher bid
premiums. Although bidders may be willing to pay higher bid
premium in order to use merger accounting, the relationship
between q ratio and bid premium overwhelmed any positive
relationship between goodwill and bid premium. Similar to
Gagnon (1967), Nathan finds that as the potential goodwill
to be written off rises, the proportion of bidders using
merger accounting also rises. This result was consistent in
both the pre APBO 16 and post APBO 16 period.
The results of studies that have examined the impact
of goodwill on the choice of accounting method are
summarised in Table 3.1. The results generally support the
proposition that when goodwill is positive bidders choose
merger accounting in order to avoid the reduction in future
8These regressions are only meaningful, if we accept Nathan's
argument that bidders will offer a higher bid premium in order to
obtain the benefits of merger accounting.
9Nathan's proposition is that a large goodwill figure will induce
the bidder to choose merger accounting. If the target demands a larger
premium in order to cooperate with the bidder, then a positive
relationship will exist between goodwill and bid premium.
10Nathan's proposition is that, a low q ratio is synonymous with
a small goodwill figure.
86
0) 0) W ).
•
E0___ ___ ___ ___ ___ ___ ___ ___
-H -H 0 -H
UUOU
-H
ci
cc
cc
c
c-'j
F-1W
U)
•-O
CDccc-.i
C)00 -H
U)
C)
C)
t
00 •
c 0
CC
CD
0)
CC
N
00 b1
4J
tU
-O\
-
-
- U)
___
___ ___ ___ ___ ___ ___
-
0 caiRo
N
(N
N
CD
_
_
__ C') __ __ __ __ __ __
-
(d4J--0
0
C:1
a)c)
cc
CI) (N
'-
'-
EE N
U)-
LC)
U)-
CC
i-
U)
CD
CD
CD
CD
EN
(j
>
(0
iQ
1)
ri-H
H
t
w
< N
U)
CD
CDCD
CD
C)
CD0
N
(N
Cs
U)
CD
0
CD
0
g
(1
.
cD
U
•:j•
_
(N
Ci
(N
CC
'-
0 ü)
WD4W .___ ___ ___ ___ ___ ___ ___ ___
0 czo
Epob
'-1
rd
E
CD
C)
0)
0
•-0
0U)CC
'
CD(N
U)W-H-H
-H
CD
_
(N
(N
N
rlQ
U)
0
g)
0
CD
CC CD
CD
C)
C)
a
0 rnWi
U
C)
-
-H
Ll1
0
flj
cz
E
-H
W
.4)
C)
•)J
'd4-l-H
rtjOJ E
-r-1-1-H
<
-H
C)
CD
:
o
0
5
0E
b___ __ __ __ __ ___ __ __
j
0
0
WUc
0
NO
NO
U
CD
'C) N
H
CDC)'0)
N'N
0)CC
W-H
C)
C')
J)
Zc
U-HUQ)Q
'-°
''
E
1.HQ
)
'C') 'Cl,4Jp _____ __ __ __ __ __
-
' d
U
4-1
0
•
Ci>,
-i
Wfli
rH
•
(
E-C')
ti)E
-
H-H(j
-
U) U)
R
O
W0.Hb)
> U
-HU)n1-H
UOJU)U
'0-H-H
w0E
ir-H0
CD
0
N N cc
OCD °
CC
C- a— C)
0
0
'
0
4
o
Q)(0
-o-J
w0(0
a)
0)
(0
oCq'- _c.
-C
c
Z
Z
C
<0 <0 (0
0
cc
0
-
(0
P.Eni ______ ___ ___ ___ ___ ___ ___
87
earnings caused by the amortisation of goodwill through the
P&L account. However the alternative proposition that when
goodwill is negative bidders choose acquisition accounting
is not supported. The validity of the results when goodwill
is negative is questionable since the samples sizes are
small.
Because the above studies all concentrate exclusively
on the impact of goodwill on the choice of accounting
policy, they are all affected by the omitted variables
problem which limits their usefulness in providing insights
into how bidders choose the method of accounting for
business combinations. Higson (l990b) uses a broader range
of explanatory variables hence reducing the impact of
omitted variables on his results.
Higson (1990b) is the only study investigating the
choice of accounting method in takeovers using a
multivariate regression methodology. In this regard Higson
improved on the previous US studies which had focused
exclusively on goodwill. He used a sample of 69 takeovers
where the bidder was qualified to use merger accounting
over the period 1976-87, and performed a binary Logit
regression with accounting method as the dependent
variable. Higson's model showed that the relative
profitability of the target had a significant and positive
impact on the use of merger accounting. With a single
equation Logit model, Higson was forced to use selective
sampling (i.e, pure equity bids) to control for the impact
of the method of payment on the choice of accounting
88
policy. Additionally Higson's study did not test for the
impact of political cost (see Section 3.3.6) and debt
covenants (see Section 3.3.7) on the choice of accounting
policy. In this regard Higson's study is still affected by
the omitted variables problem.
Robinson and Shane (1990) investigated the impact of
the accounting method chosen by the bidder on the premium
paid to the target shareholders. They suggested that where
a takeover was structured to obtain significant economic
benefits from the accounting method' 1 , then the benefits
derived from using that accounting method may be reflected
in the bid premia paid for target shares.
Using a sample of 95 pure equity offers over the
period 1972-82, they regressed the bid premium on the
method of accounting and several control variables. Both
the univariate and multivariate tests showed that the bid
premium to target shareholders was higher when bidders used
merger accounting relative to acquisition accounting. The
interpretation of this result is, however confounded by the
problem that the method of accounting and the bid premium
may be jointly determined.
A statistical association between accounting method
and bid premium may indicate that bidders derive economic
benefits from using merger accounting which is
11 See Robinson & Shane (1990: p.28) for a discussion of the
possible economic benefits resulting from using merger accounting.
These include relaxation of restrictions on dividend and financing
policies imposed by debt covenants, increases in managerial
compensation, resulting from the higher profits reported under merger
accounting.
89
correspondingly reflected in the bid premium. Alternatively
it may indicate that high value targets systematically
attract higher bid premium, consequently increasing the
probability of using merger accounting due to the resultant
increase in the size of goodwill write of f. The regressions
in Robinson and Shane do not offer any insights into
disentangling the reported association between method of
accounting and bid premium.
As a brief summary, the above review shows that:1) when goodwill is positive, bidders prefer to use
merger accounting, although negative goodwill does not
always lead to the use of acquisition accounting;
2) the majority of studies focus on the impact of
goodwill on the choice of accounting policy. There is very
little use of multivariate regression techniques to analyse
jointly the influence of other relevant variables on the
choice of accounting method;
3) the bid premium to target shareholders is higher
when bidders use merger accounting than when bidders use
acquisition accounting. However the interpretation of this
relationship is not clear.
3.5. Earnings manipulation and market efficiency
Given the empirical evidence that capital markets are
efficient (Foster, 1986: Chapters 9 and 11), should
managers have a preference between the choice of merger and
acquisition accounting?.
The discounted cash flow model suggests that the
90
market price of any security is the present value of its
expected future cash flows discounted at the appropriate
risk-adjusted rate. Since annual accounts report earnings
and not cash flows an important question is whether the
stock market values the effects of managerial decisions on
cash flows or on earnings. If the market efficiently prices
securities (ie, it only values cash flows), then investors
should be indifferent between,1 managerial choice of merger
or acquisition accounting, since the choice of accounting
method does not affect the cash flows resulting from the
takeover.
Hong, Kaplan & Mandelker (1978) examined the effects
of merger and acquisition accounting on the stock prices of
bidders. Using a sample of equity offers over the period
1954-64, they compared the abnormal returns for a sample of
122 bidders who used merger accounting with 37 bidders who
used acquisition accounting. Abnormal returns centred on
two event periods were examined (i) the month of the first
earnings announcement following the completion of the
merger (ii) the merger announcement month.
Bidders using merger accounting did not display any
statistically significant returns around either of the two
event months 12 . Bidders using acquisition accounting showed
statistically significant positive returns around the month
of merger announcement. These results do not suggest that
12unfortunately the paper only presents the graphs for abnormal
returns for the 12 months before and after the relevant event month.
The actual abnormal returns calculated for the event month are not
reported in the paper.
91
firms using merger accounting were able to benefit in terms
of increased share price at the time of either merger or
earnings announcement.
A number of other studies in a non takeover context
have examined whether the stock market is deceived by
changes in accounting policies, which are not accompanied
by real economic changes in cash flows (Kaplan & Roll,
1972; Archibald, 1972; Eall,, 1972; Sunder, 1973) . These
studies show that accounting policy changes which have no
impact on the firm's cash flows do not affect the firm's
share price (Foster, 1986: Chapter 11)
3.6. Managerial considerations and accounting policy choice
Despite the empirical evidence above that managers
cannot increase the value of their firm by manipulating the
accounting numbers (see Section 3.5) there are still
several possible economic motives for managers to prefer
merger to acquisition accounting.
Among these are: the restrictions on dividend payments
and borrowing capacity imposed by debt covenants (see
Section 3.10), the political costs of reporting high
accounting earnings ( see Section 3.9) and the impact of
accounting numbers on managerial remuneration.
The remuneration of managers is linked to accounting
numbers through the inclusion of profit related pay clauses
in their compensation contracts. The purpose of these
clauses is to align managerial interests with those of
shareholders (Watts & Zimmerman, 1978; Sloan, 1993; Forbes
92
& Watson, 1993)
Murphy (1985) found that profit related pay accounted
for over 25% of managerial compensation, while stock
options account for around l0. Jensen & Murphy (1990) and
Lambert & Larcker (1987) found that managerial compensation
was related to both changes in accounting earnings and
stock price performance. However, in both studies
accounting earnings had greater power in explaining cross
sectional variations in cash compensation to managers than
share prices.
Mangel & Singh (1993) found that the cash compensation
to the chief executive officer in large US corporations was
positively related to the firm's accounting return on
equity, while in a study of UK Building Societies, Ingham
& Thompson (1993) found that the salary received by the
highest paid director was positively related to the return
on assets and the growth in profitability. These studies
confirm the existence of a link between managerial pay and
the firm's accounting profit.
Where managerial compensation is dependent on the
reported accounting numbers, then irrespective of a
managerial belief in efficient markets, managers will not
regard the choice of accounting policy as an insignificant
detail in the firm's acquisition strategy.
3.7. Conclusion
In this chapter we reviewed the literature on the
determinants of the accounting policy in corporate
93
acquisitions. A summary of the determinants of accounting
policy in corporate acquisitions is provided in Table 3.2.
US based studies all adopt a univariate approach in
their analysis and focus exclusively on the importance of
goodwill and the desire by the bidder to choose merger
accounting as a means of avoiding the recognition of
goodwill. This evidence suggests that when goodwill is
positive bidders are more,, likely to choose merger
accounting. However when goodwill is negative there is no
clear preference by bidders for a particular accounting
method.
There is very little UK based evidence on the
determinants of the accounting policy in takeovers. Higson
(1990b) which is the only relevant UK study in this area
found that goodwill was not a significant determinant of
accounting policy. Higson found that the main determinant
of the accounting policy was the relative profitability of
the target to the bidder.
One of the main criticisms of the existing literature
is the omission of relevant variables from the analysis.
Some variables (e.g, size and gearing) which have been
found to be significant determinants of accounting policy
in other contexts have not been tested in relation to
corporate acquisitions. This is a gap in the literature
which our empirical work seeks to fill.
Additionally the existing studies use samples
restricted to pure equity offers in order to control for
the effect of the payment method on the choice of
94
(I)
0
-'-I
-i-i
-H
-H
0
('j
w
4J
(Ii
0
-H
R CO
Co
)
0) N
I-
a)
0
C
a)
-o
>
a)
Co
0
0
E
a)
0
0
U)
0)
C
0
a)
=
CO
U)
-
0
>
4-.
-
0
z
Ci
0
0
0
('j
0
4-I
0
C
0.
0)
C
0
Ci
0
Co
9-
0
a)
o
0
0
o r
oE
0) 0
4-' 9U)a)
Co
a) 4-.
CO
a)
-I-)
0
a)
4-.
C
0
4-.
0
0
0-CoCO
Eo
a) 4-.
a)
0)
I-
U)
0 00
zz
CO
9--,
CO
--C
0
a)
C--
C
C -' C
>
g
-
C
0
a)
U)0
C
Co
C.)
--0
OC
Co U) 0 C
0-
o '- -c Co j .2
Co
a) -
'-a)
E
-
0
0)
4- > 40 _
C
C
)
C)a
a) L
>
0 0) 0) C
.0 C.)
0
0a)COQ
a) .0
U) Co
a) 0
I- (1) .0 0 '.- C
c .2
0 0
CO
Co.0 C (0 0 '4
a)
0)
ci)
U)
CE
2 = •- 0
C0
0
0) 0
0 0) C
OCOa)
o
0 '- - C
= 0) CO
>0
-- S 0
=
0)
U;
a)dC
4.-,4--'
.0
9--,
Co 0
4- -
C/)4-,
0-
a)Co
-a) ,.0) U)
Ca)
C
0)
C0
ICO•'Q) 0a)0)
4a)
0) 0) '- E '0.
.0 Co 0. a) '0
> 0)
C9- '4
0 L
a)
0)'-
Co . C a) a)
0)
U) .
CO1JCOC
C
• 9 CO
C -0 )
0)
Co
.2 a) (0
.' )<.0 > o C Co 0
--
a) '-CO
a)
0--
a)
.0
0- C
:2 a 0 0) .0
a).0 0) CO .
%--4--,
U)
4-'
C
a)
a)
U)
a)
CO
0.
4-4
L. 0
0) a)
r>i
.22-o
.0 0
OJai
Co .0
U)
4-. U)
CO
0
0
>
(3
0)
'-
0CoO
I-Co
________
95
-e C
CO
0) Q) CO 0)C
C 0C 0.2 0
Dc oHD oO
a)
.0
CO
C.)
a)
a)'-
90
c'.i0
E-'U)
U)
a)
>
E
>-
4-I
0>
CoU)
U)
o
nj
C.)
Q9- CC4-
C
0)
0
0 C.) 0- - 0
4-000a)a)
0) (0 CO C
CO
,.a)a)4-
Ca)°
CO C 0)
-H
0
4.)
C
0. i5 a)
•
C
.0 (00
U) 0.
CO- C C -.0 = 0
-
00
I-0
a)
C
a)
CCo
L. L.
a)
0.
E
0.
O)
4- U)
0
CO
4-1
.2
>E
0 00
zz
0 2 c
•'
0 •'-'
a)
-c
H
.29> E
.2)
a)
0
0
-H
0
4-i-
Co U) C 0 .0
'4'0
4-
CDI
(1)0 C Co
Co
>-
0
o?j oj a) .2
(fl5 52Z
4-,
0-
-H
-H
0
...i a)
=0
-H
Q4
Co'
C
U)Co
>
>
a)
C
a)o
CU)
O-5
Co
><
a)
0
CU)
C
•E
Co
E
0
0
H
0
=
0
I-C/)
0)
C
a)
>1
U)
0
0
C
-
0 0)
U)
SC!) U)
.i--. CO
a)
I-a)
CO)<
-5
C))
-C
0
0)
0
-H
-H
a)
0
(I]
-H
a)
-c
0
C0)
C)) 'C —
EC
Co
-2)
>
a)
Co
0
I-
U)
C-
E
0
>1
C
a)
C
E
E
E
CO
CO
CO
0
C))
>Co
U)
-0
U)
4-. C
0
>
Z(/)
>
a)
o .2
C-
40
z
C))
-'-I
C))
E5
a)
0
0
-C
C))
C
Ia-
><
><
a)
a)
>-
>-
U)
U)
0
>
C))
0
>
C))
C4-'
0
z
40
z
a-
>1
0
-H
H
0
a)
ci)
Q)0
C)._
+
U)
C. C
Q)O)
_-;
-
p4
bi
E
3)
4-.
Co C
>_ -c >-= - C))
—C--C C
CO C 0 C
0 0
-c
:
C
C C))
-- r. 4- -- Co
0 4
C))
C
0
4-.
0
Co
C-
H
-U-I
E
'ri -H
a)
4CO U)
a)-_C -- C)) a)
C- 0 0
0 0 0
Co
CO
-
>
a)-
U)
cO0
a) 0 0
a)C)C)) Co°
0
C)) - 4-. C))
-- - Co Co i. a)
' -0
C))C/)2
- '4_C)
0t.
'a) > CO C)
00 C - C
o -C
a) -.j
C) C) C
C-0 o 40
0
0
C0-
U) -- -- 0
4-' 4-' 4a)
C
=
U)
0)
-- 4-'
L QU)
CO o
_a)0a)0
I c.
SC-)
Co
a)
0)
4- ci 0 L
= CO Co
0
Co
a)4
ErC -- .,-' Co E
C
4-CO
C)Co
CC))
C 4-' L C -.
- 0 ci) -0
C
-- 0
C)2
COO)
--
-
C
-
0
0L
Co
CoQCO
C))
0) 0OU)
'- —C
-
C
a) -.
0
8E2
C) CO c -
.
E '
= = CI)
0
a
LU
- > a.
-
C) C)
_c C -- C
-
U) 0 C -a) j5 0
0
0
Co
a)
-O
4C))
0)
-C
4-.
a)
-c
a)
-C
9-
9-
0
0
C))
N
a)
N
(I)
-ci)
I-
'4- ci)
u-I
>-
rO
0
—4-.
C))
a)
-
-Q
Co
Co
>
o
a)
CO
4-.
4-
E-'c/)
96
0
0
CoCO
4-
-o
-.
C C C
-.
C-- C 0 C 0
Q Co CO
CO
CO
4-.C))
0
r-
0 a)
0 c
- 4-' 0
Co
C4-.
-
C 0 -- 0 =
0
CO
CO
C - CO
4.0
4 -o CC
o-
C))
9- :-=
CO Co
c_)
'4- o
U)
L
Oni
HE
CCa)0
C4-.
CO 0
a)
C)
-H 0
-I-i-I-i
rn>1
0 . 0
>-U) M-'4-
- -O
0)
c
-
0)
C)
I-
0
-H
0
C
0o
Co
4-.
a)COC
C))
CoCO0
0 0 >
9-
.L)
C-r C
0 4-' CO
4-'
--
Co
E
0
-
-4-.
CO
C 4-'
-. .2:2
CoCa
4-.O0
0
ci)
0
0
U)
0 r
L
Co C) 0 '-
4-.
0
C-)
0
C.
ao
Q)U)
Co
CO
0
ci)
0
C0)
u-I
0
a)
0
-H
0
U)
C) C))
C
9- 0 C
0 C CO C))
>C-)
a)
'4-
>-.E
50
-H
4J
0
0
0
(ii
CO
C
-o
-4-.
U)
C))
U) I- -- U)
o C))
C))
o 0)
a)
C)
C
ICO
a)
C)
U)
a)
0
0
accounting policy. This use of selective sampling has
prevented an investigation of the interaction between the
choice of accounting policy and the method of payment. Our
use of a simultaneous equations approach, improves on this
methodological deficiency in the extant literature as
discussed in the next two chapters.
97
CHAPTER 4
INTERACTION BETWEEN ACCOUNTING POLICY AND THE METHOD OF
PAYMENT: METHODOLOGY, DATA AND RELATED ISSUES.
4.1. Introduction
The literature review in Chapter 2 has shown that
despite the wide cross sectional variations that exist in
the types of consideratior offered in mergers and
acquisitions very little is known about how bidders choose
the consideration offered or the factors which explain the
cross sectional variations which are observed in the types
of consideration offered by acquiring companies.
In particular there has been no investigation of the
association between the method of payment and choice of
accounting policy. In equilibrium the method of payment and
the choice of accounting method may be jointly determined
(see Section 2.3) . This joint relationship requires the use
of a simultaneous equations framework, if we wish to study
the interaction between the choice of accounting policy and
the method of payment.
In this chapter we outline such a simultaneous
equations model and discuss some of the related empirical
problems in estimating it.
To estimate the model and to avoid an omitted
variables problem, we use a number of exogenous control
variables. The chapter presents the definitions of these
control variables. We also describe the criteria used in
the selection of the sample and descriptive statistics on
98
the data.
4.2. Methodology
Earlier studies investigating the choice of accounting
method which were reviewed in the last chapter, control for
the effect of payment method by restricting their samples
primarily to equity only offers (Gagnon, 1967; Copeland &
Wojdak, 1969; Anderson & Lquderback, 1975; Robinson &
Shane, 1990; Higson, 1990b) . On the other hand, studies
concerned with the determinants of the payment method (see
Chapter 2) adopt a two group Logit discriminant
methodology. The Logit model requires a binary dependent
variable. Hence these studies restrict their sample to only
those observations in which the method of payment is either
"all cash" or "all equity" (see Hansen, 1987; Amihud et al,
1988; Higson, 1990a) . A single equation approach is flawed,
given the joint determination of the two choices (see
Section 1.2.4 for a discussion of the limitations of the
single equation approach).
Our simultaneous equations methodology obviates
selective sampling as a means of controlling for the effect
of either payment method or accounting method on the other.
By using the Two Limit Tobit Model for the choice of
payment method, we are able to overcome the restriction of
the two group Logit methodology with a specification in
which the offer of a mixture of cash and equity is
naturally reflected in the model. Our sampling is,
therefore, not limited to 100 cash or 100% equity cases.
99
Again, in modelling the determinants of the accounting
method, our sample is not restricted to 1OO equity offers.
4.2.1. Simultaneous equations model
The relationship between the method of payment and the
choice of accounting method is examined within the context
of the following simultaneous equations system
PM= f (A CCMET, I)
(la)
A CCMET = .f (PM, Y )
(lb)
where
PM is the method of payment
ACCMET is the method of accounting
X and Y represent vectors of exogenous control
variables.
Under the statistical specification of this equation
system, we define an unobserved variable PMIS which is an
index of the propensity to use equity financing in any
particular acquisition. The index PM1 may be interpreted as
reflecting the perceived differential benefits to the
bidder of equity and cash financing in the ith acquisition.
When the benefits of equity financing are greater than
those of cash financing the propensity to use equity
financing will be greater than zero, ie, PM > 0.
The propensity to use a particular method of financing
cannot be directly observed. The observed variable is PM1
which is the proportion of equity in the ith acquisition.
100
PM has an unrestricted range of values, while PM is
bounded by zero and one. When the propensity to use equity
is large and PM
1 then PM1
^
=
1. When 0
<
PM1
<
1, PN
=
PM1
the acquisition is financed by a mixture of cash and
equity. When the propensity to use equity is small and
PMIS ^
0 then PM1
=
0. The various combinations are
summarised below
If PMI* ^
If 0
<
If PM*
1 then PM1
PM
^
<
=
1
1 then PM
0 then
PM =
"all equity offer"
=
PM' "Mix of equity & cash"
0
"all cash offer"
Similarly ACCMET1 is an index of the perceived
differential propensity for the bidder to use merger or
acquisition accounting in the ith acquisition. The index is
not directly observed, what we observe is a dummy variable
ACCMET1 which is defined by
1 if
A CCMET1' >
0,
.ie, Merger Account.ing
ACCMET =
0 otherwise, .ie, A cquisition A ccounting
This equation system is estimated by a two stage
procedure. First we estimate the reduced form for PM by the
Two Limit Tobit method (See Maddala, 1983: p.160) and the
reduced form for ACCMET by the Logit method (See Maddala,
1983: p.22) . The structural equation for PM is then
estimated by the second stage Two Limit Tobit (2STM) and
the structural equation for ACCMET is estimated by second
stage Logit (2SLM).
101
4.2.2. Econometric problems in estimating the simultaneous
ecpiatious models
Because this is an equation system with a mixture of
doubly censored (payment method) and dichotomous
(accounting method) endogenous variables, we have not been
able to derive an efficient asymptotic covariance matrix
estimator for the model. This implies that the t-statistics
which we report for the estimated model parameters may not
be efficient (see Maddala, 1983: p.246)
As a partial, admittedly imperfect, solution to the
problem of inefficient t-statistics, we attempt an
alternative model formulation by recoding PM as a
dichotomous rather than a doubly censored variable. If both
endogenous variables are dichotomous we can use a two stage
Logit method to estimate the system (see Maddala, 1983:
p.246) with efficient t-statistics. Dichotomising the
payment method variable PM is achieved by restricting the
sample to only those observations in which the method of
payment is either "all equity" or "all cash". The results
from this two stage Logit method can then be compared with
the results for the complete sample to check how sensitive
our conclusions are to the problem of inefficient tstatistics.
We also examine whether a single equation model is
statistically adequate and satisfactory despite the
conceptual validity and methodological superiority of the
simultaneous equations model. This procedure allows our
results to be more easily compared to those of earlier
102
studies employing the single equation model. We estimate
the functional relationship f or PM and ACCMET as single
equation models and use the Wu-Hausman test (Hausman, 1978)
to check for the existence of any simultaneity bias which
may necessitate the simultaneous equations model approach.
4.2.3. Test for simultaneity bias
In the simultaneous equation model we have two
endogenous variables PM and ACCMET. For simplicity we
represent the vector of exogenous variables in the first
equation by X 1 and the vector of exogenous variables in the
second model by X2 . To test for simultaneity bias in the
single regression model we first obtain the predicted
values PM (by
Two
Limit Tobit method) and ACCMET (by Logit
method) from the reduced form equations for PM and ACCMET.
We then estimate the models
PM = a 1 + 1 1 ACCMET+A1 X1 +PACCMET +p.1
and
ACCMET= a2
+
I2
PM^A2 X2
+
PM
+
A1 and A2 represent vectors of the estimated coefficients
for the exogenous variable vectors X1 and X2 respectively.
If the test
I3
=
0 and
12 =
0 is rejected then the
single equation regression results are subject to
simultaneity bias. For a fuller description of this test,
see Hausman (1978)
103
4.3. Control variables
The finance and accounting literature reviewed in
Chapters 2 and 3 suggests several variables which may
affect either the method of payment or the choice of
accounting method. In order to avoid an omitted variables
problem and to facilitate the identification of the
simultaneous equations system we introduce these variables
into our model as control vaiables. These variables and
how they are defined in this study are discussed below.
4.3.1. Control variables for the method of payment
The following variables are used as control variables
in the method of payment equation:-
Potential goodwill arising on the acquisition (GWILL)
The use of merger relief provisions in conjunction
with acquisition accounting, implies that goodwill can have
a direct impact on the method of payment independent of the
use of merger accounting (see Section 3.3.2). When the
bidder chooses not to use merger accounting, the method of
payment could still be of concern if goodwill needs to be
written off to the merger reserve account. The size of the
merger reserve created in an acquisition is a direct
function of the proportion of equity used in the method of
payment.
104
Goodwill is proxied by'
V alue Of - Target s's Net
GW ILL = The Offer
A ssets
Bidder"s Net A ssets
The net assets of both the bidder and the target are
measured at the last balance sheet prior to acquisition
announcement. We recognise th rat balance sheet values are
not equivalent to fair values. However problems with data
availability restricted our ability to use a more refined
proxy. We initially attempted to collect the goodwill
figure as disclosed by the bidder in the annual report but
we encountered a number of difficulties in the process:1) Many bidders make multiple acquisitions in a year
and only the figure for total goodwill is disclosed.
2) Many bidders disclosed goodwill figures in their
accounts which were net of any available merger relief and
could not be disentangled.
3) Prior to the issue of SSAP 22 (revised) in July
1989 there was no requirement for bidders to provide
details of the fair value adjustments in their annual
reports hence data on fair values prior to this period
which covers almost the entire sample is very patchy.
4) Bidders using merger accounting do not have to
1 We have chosen to normalise the difference between the price paid
for the target's assets and the book value of those assets by the
bidder's net assets rather than the bidder's earnings (see Gagnon, 1967
in Section 3.4) because in the UK, very few companies amortise goodwill
through the P&L account. Any adverse impact of goodwill write off is
more likely to be felt in the balance sheet than in the P&L account.
105
disclose the goodwill that would have arisen if acquisition
accounting had been used.
In cases where bidders revalue the targets' assets
upward prior to consolidation then the use of book values
will overstate goodwill (see Section 3.3.1). Conversely if
bidders make provisions against the target's asset values
then the use of book values will understate goodwill. As we
have no prior reason to believe that bidders in our sample
will systematically choose to either revalue assets upward
or make provisions against asset values, there is no reason
to believe that the use of book values as a proxy for fair
values will introduce any systematic bias into our results.
Additionally the practice of frequently revaluing fixed
assets which is allowed and encouraged by the UK accounting
rules ensures that the difference between book values and
the fair values of assets may not be too large.
The value of the offer is equivalent to the market
value of the target at the unconditional date. In the case
of cash offers this should be equivalent to the offer price
per share multiplied by the number of outstanding target
shares. In equity offers changes in the bidder's share
price affects the value of the offer. The market value of
the target at the unconditional date is a reasonable
approximation of the purchase price paid by the bidder.
Because the offer value is determined at the
unconditional date we could be criticised for using an expost measure of goodwill. It could be argued that since the
bidder has to decide the method of payment ex-ante,
106
goodwill should correspondingly be calculated ex-ante ie,
the value of the offer should be determined at a date
before the unconditional date. We would argue that the
choice of method of payment and the accounting policy are
never fixed at any time during the bid process. The bidder
can always choose to revise the method of payment (and
correspondingly the chosen accounting method) if this will
prove to be advantageous. , If the bidder's ex-ante
calculation of goodwill turns out to be incorrect, then a
rational bidder will, if necessary, revise the method of
payment 2 and the accounting policy since it is the ex-post
value of goodwill that will be used in the consolidated
accounts.
Bidder and target gearing (BIDGEAR and TAGGEAR)
We expect the proportion of equity in the payment
currency to increase with the bidder's gearing as the cost
of financing a cash offer increases with the bidder's
gearing (see Section 2.8)
We expect the proportion of equity in the method of
payment to be negatively correlated with the target's
gearing, since the bidder can free ride on the monitoring
activities of external creditors (see Section 2.8)
Gearing for both the bidder and the target is measured
some cases, a cash offer may be mandated by rule 9 or 11 of
the City Code (see Section 2.2.2) . This will not necessarily constrain
the bidder's choice since a cash alternative underwritten by the
bidder's merchant bank (see Section 2.2.1) allows a bidder wishing to
make an equity offer to comply with the mandatory requirements of rules
9 and 11.
107
as
GEA R = Total Liabilities
Total A ssets
Relative size of the bidder to the target (RELSIZE)
Information asymmetry reflects unshared private
information held by parties to a transaction. Such an
asymmetry exposes the bidder to valuation risk, ie, ex-post
it turns out to have overvalued the target (see Section
2.4.1) . The cost to the bidder of such a valuation error
increases in the target size relative to the bidder and the
proportion of cash in consideration. With equity some of
this cost is borne by the target shareholders themselves
due to the contingent nature of equity valuation. Bidders
seeking to minimise this cost would offer a higher
proportion of consideration in equity with increasing
relative target size (Hansen, 1987).
The relative size of the bidder to the target is
measured by
Bidder1s Market V alue
RELSIZE = ___________________
Target's Market V alue
Market value is price per share 2 months before the
acquisition times the nuither of outstanding shares.
108
Dilution of existing block shareholdings (BIDDIR and
BIDLGE)
Amihud et al (1990) suggest that the method of payment
is a function of the ownership structure of the bidder.
Equity offers dilute the holdings of managers and large
shareholders in the bidder (see Section 2.6). The
shareholding structure of the bidder is proxied by
EI]JDIR = Proportion of , shares (beneficial and non-
beneficial) held by the directors of the bidder.
BIDLGE = Proportion of bidder shares held by large
investors. This is equivalent to the total of all
shareholdings greater than 5 in the bidder (excluding
shares held by the directors).
As large external blockholders will wish to avoid a
dilution of their holding (see Section 2.6.3), we expect a
negative relationship between BI]JLGE and the proportion of
equity in the method of payment.
We postulate a non-linear relationship between BIDIJIR
and the probability of an equity offer. This reflects the
intuitive idea that at low levels of shareholdings
directors do not control enough shares to be concerned
about the dilution effects of an equity offer. At very high
levels of shareholdings, the directors have such a solid
control over the firm that concern about the dilution
effects of an equity offer would be minimal.
We use a piecewise linear regression model to capture
this non-linearity. The variable BIDDIR is modified as
follows:
109
if BIDDIR < 0.1
BIDDIR 0
BIDDIR 0 .1
to 0.25
0.1
=
0.1
if
BIDDIR ^ 0 . 1
10 if BIDDIR K 0.1
= BIDDIR minus 0.1 if 0.1 ^ BIDDIR K 0.25
Lo.15 if BIDDIR
^
0.25
0 if BIDDIR K 0.25
BIDDIRovero25
=
IDDIR minus 0.25 if BIDDIR
^
0.25
The piecewise regression model is a technique for
estimating non-linear patterns in a dataset (Morck,
Shleifer & Vishny, 1988). It is a procedure which allows
for multiple changes in the slope of the regression line
describing the relationship between two variables.
At low levels of shareholding the directors do not
control enough shares to be particularly concerned about
dilution. Hence, we expect a positive relationship between
BIDDIR
01 and the probability of an equity offer. Beyond
the 10 level, any new issue of shares will have an impact
on directors' control. Any dilution of control between the
1O to 25% range could affect the ability of directors to
defeat any challenge to their authority. Hence we expect a
negative relationship between BIDDIR01025 and the proportion
of the acquisition which is equity financed. At high levels
110
of percentage holding the dilution effects of an equity
offer are insignificant, hence we expect a positive
relationship between BI]JDIR over025 and the probability of an
equity offer.
In the above piecewise definition we allow for two
changes in the slope coefficient on BIDDIR (ie, at lO and
25) . There is no theoretical justification for choosing
these particular points. To mke sure that our results are
robust, we estimate the model allowing for slope changes at
different points. The results are fairly insensitive to
variations in the points at which the slope changes are
allowed to occur3.
Conditions in the capital market (RETMKT and RETBID)
Bidders faced with favourable conditions in the stock
market will be inclined to use equity as a method of
payment (see Section 2.7)
We use two alternative proxies to capture the
condition of the market:
RETMKT = Cumulative return on the market index during
the 80 trading days beginning 120 days before the
announcement of the bid4.
RETBID = Cumulative unadjusted return on the bidder's
equity during the 80 trading days beginning 120 days before
3As a further robustness test to examine whether the relationship
is non-linear with two slope changes, we estimated the model using
BIDDIR in the linear, squared and cubic form (ie, BIDDIR, BIDDIR 2 and
BIDDIR3 ) without any changes in the results.
4Returns are calculated as log returns. See Appendix 8.1. for the
formulas.
111
the announcement of the bid.
The bidder's pre-bid returns have not been adjusted
for market movements because what we are trying to measure
is the general direction in which the bidder's shares are
moving and not its movement relative to the market e.g, a
bidder whose shares fall by 10% when the market falls by
15% will have an abnormal return of ^5% (assuming beta
=
1). A measure of pre-bid perormance which is relative to
the market may wrongly indicate that such a bidder was in
a favourable market.
Cash resources (BIDCASH and TAGCASH)
Cash rich bidders are in a better position to make a
cash offer than cash poor firms. Additionally the cash
resources of the target can be used to help finance part of
the cost of the acquisition 5 . We expect the liquidity of
the bidder and the target to be negatively related to the
proportion of equity in the method of payment (see Section
2.10) . Cash liquidity for both the bidder and the target is
measured by:
CASH
Cash In
Hand
=
Marketable
Securities At
Market Value
Net Assets
+
5 This is subject to the Companies Act 1985 regulation which
prohibits a company from giving financial assistance for the
acquisition of its own shares (see Section 2.10)
112
Growth opportunities (VRBID & VRTAG)
Since a high level of debt acts as a disincentive for
managers to undertake positive NPV projects (growth
options), firms with growth options should reduce their
leverage (Myers, 1977) . This suggests that an acquisition
in which either the bidder or the target have a high
proportion of their value represented by growth options is
more likely to be financed wiph equity (see Section 2.9).
Knowledge of a firm's investment opportunity set is
required to distinguish firms with high growth options from
firms with low growth options. Unfortunately a firm's
investment opportunity set is not observable, hence we need
a suitable proxy. Firms with a high Tobin's q ratio are
presumed to have growth opportunities and hence positive
NPV projects, while low q bidders are not likely to have
positive NPV projects. (Lang, Stulz & Walkling, 1991; Smith
& Watts, 1993; Ga yer & Gayer, 1993) . Tobin's q is defined
as:
Tobin's q
Market Value of Total Assets
Replacement Cost of Total Assets
Data is not available in the UK to calculate the
replacement cost of assets in place, hence we approximate
the Tobin's q ratio with the Valuation Ratio (VR) which is
defined as:
Market Value + Book Value of
= Of Equity
Total Debt
Book Value of Total Assets
113
VRBID and VRTAG represent the valuation ratios of the
bidder and the target respectively. The valuation ratio was
used as a measure of investment opportunities by Lewellen,
Loderer & Martin (1987), Collins & Kothari (1989), Chung &
Charoenwong (1991), Smith & Watts (1993), Ga yer & Gayer
(1993) . Smith and Watts (1993) argue that firms with a high
proportion of their total value made up of "assets in
place" have low growth options and that the book value of
assets was a suitable proxy for "assets in place". In a
similar vein, Collins & Kothari (1989) argue that the
difference between the market value and the book value of
assets is a reasonable approximation of the investment
opportunities facing the firm.
Free cash flow (FREECASE
Jensen's (1986) free cash flow hypothesis predicts
that bidders with free cash flow will invest in negative
NPV projects rather than return excess cash flow to the
shareholders. If free cash flow is the motivating factor
behind an acquisition then we would expect the probability
of an equity offer to be negatively correlated with the
level of free cash flow.
Free cash flow exists when a firm lacks profitable
investment opportunities. The valuation ratio (see yR
above) serves as a proxy for the firm's investment
opportunities. For bidders with a valuation ratio greater
than one we don't expect any association between cash flow
and the method of payment, while for bidders with a
114
valuation ratio less than 1, we expect a negative
association between cash flow and the probability of an
equity offer.
Free cash flow is measured as an interaction variable.
If the bidder has a valuation ratio greater than 1 then
free cash flow equals zero otherwise free cash flow equals
the firms cash flow 6 , ie:
Free Cash Flow
=
OIfVR.BID>
Cash flow if
1
VR.BID < 1
As there are many cash flow measures which have been
proposed in the accounting literature, we use several
proxies to measure cash flow in order to investigate
whether our results depend on the cash flow measure used.
The definitions below are similar to those used in Bowen,
Noreen & Lacy (1981) . The first proxy for cash flow is
NPDNA
=
Operating - Minority - Divi- -Taxes
Interest dend
Profit
Net Assets
+
Depreciation
The second cash flow measure is
WCONA
=
Increase In
- Associate
Profit
Long
Term
NPDNA ^ Minority
Interest
Provisions
Net Assets
6 This measure is superior to using just the cash flow measure as
a proxy for free cash flow because, high cash flow in a firm does not
necessarily equate to free cash flow if the firm has growth
opportunities. Cash flow is only free when there are no growth
opportunities. Lang, Stulz & Walkling (1991) used a similar approach in
their definition of free cash flow.
115
The final cash flow measure is
CFONA =
In - Increase
In+Increase
WCONA - Increase
Debtors
CurrentIn
Stocks
Liabilities
Net A ssets
Type of bid - hostile or friendly (HOSTILE)
Bidders use cash as method of overcoming bid
resistance by the target (see Section 2.11). We expect that
the use of equity will be negatively correlated with the
hostility of the target's management. Hostility of the
target's management to the bid is measured by the variable
HOSTILE, which is defined as a dummy variable equal to 1 if
the first offer from the bidder is rejected by the target's
management and 0 otherwise7.
Capital gains tax (CGAIN)
In a cash offer, the bidder may be required to
compensate
target
shareholder for the
immediate
crystallisation of their CGT liability (see Section 2.5)
The higher the CGT liability of the target's shareholders
the higher the premium demanded from the bidder. Assuming
that the bidder wishes to minimise the bid premium paid we
expect the proportion of equity in the method of payment
7 There is no agreed definition of a hostile bid in the literature.
Sudarsanam (1994a) argued that the initial rejection of a bid, followed
quickly by an acceptance of an increased offer, does not leave much of
a contest to talk about. He defined a hostile bid as one rejected by
the target's managers who then undertake one or more of a range of
defensive strategies.
116
will be positively related to the potential CGT liability
of target shareholders.
CGT is levied on the difference between the selling
price of a share and its original purchase price, subject
to any available indexation allowance. Data on the original
purchase price for most investors is unavailable, hence
estimating the potential CGT payable by target shareholders
is difficult. Some of the proxies which have been used in
the literature to measure the CGT payable by target
shareholders include: (i) the difference between the offer
value and the target's lowest share price over a specified
period (Niden, 1988; Hayn, 1989) ; (ii) the proportion of
target shares held by tax exempt investors (Niden, 1988;
Peterson & Peterson, 1991)
We use share price appreciation over a one year period
as a proxy for the CGT payable. The variable is defined as:
Pre-bid Mkt V alue Lowest Mkt V alue
Of The Target - Of The Target
CGA IN =
Lowest Mkt V alue Of The Target
The lowest market value of the target over the
preceding one year period is used as the base cost for
approximating the CGT purchase price of target
shareholders. The pre-bid market value of the target is
measured 41 days before the announcement of the bid. We
have chosen to use the pre-bid market value of the target
in calculating CGAIN rather than the offer value in order
to avoid the problem of capturing any tax compensating
117
premium which may be present in cash offers in the CGAIN
variable.
We also use the notion of tax clienteles as a proxy
for target shareholders' CGT tax rates. We expect that the
higher the proportion of tax paying shareholders, the
higher is the average tax rate of the target's
shareholders. The CGT profile of the target's shareholders
is proxied by
TAGLGE = Proportion of target shares held by large
investors. This is equivalent to the total of all
shareholdings greater than 5 in the target (excluding
shares held by the directors and the bidder).
The majority of institutional investors are exempt
from capital gains tax (le, pension funds, investment
trusts, unit trusts, charitable trusts etc) . If we assume
that institutional investors are also likely to be the
large investors in companies, we expect the average tax
rate of the target's shareholders to be inversely related
to TAGLGE. This proxy suffers from a number of problems:1) Not all shareholders with a percentage holding
greater than 5 will be tax exempt institutions.
2)
The vast majority of shareholders will have
percentage holdings less than 5. Consequently the tax
profile of the major shareholders may not be representative
of the tax profile of the whole body of shareholders.
Unfortunately in the UK only data on shareholders with
a percentage holding greater than 5 was publicly available
during the study period. Lack of data prevents us from
118
using proxies which are better defined8.
4.3.2. Control variables for the choice of accounting
method
The control variables in the choice of accounting method
equation are:-
Potential goodwill arising from the acquisition (GWILL)
Under acquisition accounting goodwill can either be
written off against net assets or amortised through the
profit and loss account while goodwill is not recognised
under merger accounting (see Section 3.3.1) . The impact of
goodwill on the choice of accounting method is mainly
concerned with the bidder's ability to write off goodwill
in the consolidated accounts. The bidder's ability to write
of f any goodwill arising on the acquisition depends on the
size of the goodwill relative to bidder's net worth. We
expect a positive association between GWILL (see Section
4.3.1 for definition) and the use of merger accounting as
the bidder tries to avoid the problem of recognising
goodwill.
Target's distributable reserves (DISRES & PAYOUT)
Under acquisition accounting the pre-acquisition
reserves of the target have to be capitalised whereas under
merger accounting they can be distributed as dividends (see
8A more suitable proxy would be the proportion of shares held by
tax exempt investors.
119
Section 3.3.3). Bidders with a. high dividend payout may
wish to access the target's pre-acquisition reserves and
the larger these reserves are the greater is the bidder's
preference for merger accounting. The target's preacquisition reserves are proxied by
DISRES = Target "s Accumula ted Di s tribu tab] e Reserves
Target's Net Assets
We expect DISRES to be positively related with the use
of merger accounting.
The bidder's pre-acquisition dividend payout is
proxied by
Bidder"s Dividend
Ridder's Net Income
PAYOUT = ________________
We expect PAYOUT to be positively related with the use
of merger accounting.
Target's depreciable fixed assets (DEPFA)
Acquisition accounting requires that the target's
fixed assets be restated to their fair values prior to
consolidation. Any revaluation of fixed assets would have
an impact on future earnings as depreciation charges are
passed through the P&L account (see Section 3.3.4). Where
fixed assets are revalued upward the bidder would like to
avoid the higher depreciation charges in the future. Where
fixed assets are revalued downward the bidder would welcome
120
the lower future depreciation charges. The importance of
the target's depreciable fixed assets is measured by
DEPFA Target's Depreciable Fixed Assets
Target's Net Assets
Ideally the depreciable fixed assets should be split
into Value Increasing Depreciable Fixed Assets (VIDFA) and
Value Decreasing Depreciable Fixed Assets (VDDFA) as
discussed in Section 3.3.4. Unfortunately the availability
of data from Datastream to enable us to carry out this
split is very patchy. For a significant number of
observations in the sample Datastream does not provide any
analysis on the composition of fixed assets. In order to
avoid a significant reduction in the sample size due to
missing data we have used DEPFA in our analysis. We do not
attach any a priori sign to the impact of DEPFA on the
probability of choosing merger accounting.
Relative profitability of the bidder to the target (PROFIT)
Under acquisition accounting only the post-acquisition
profits of the target can be included in group profits
whereas under merger accounting both pre and post
acquisition profits can be included in group profits (see
Section 3.3.5). Bidders with low profitability who acquire
highly profitable targets will value the opportunity
provided by merger accounting to boost the reported
earnings in the year of acquisition. The profitability of
the target relative to the bidder is measured by
121
PROFIT = Target's Return On Equity
Bidder s's Return On Equity
We expect a positive association between PROFIT and
the probability of using merger accounting.
Size of the merging firms (BIDSIZE & TAGSIZE)
Large firms reporting high profits following
acquisitions may attract antitrust scrutiny and possibly
invite antitrust sanctions (see Section 3.3.6) To avoid
these political costs, acquisitions involving large firms
are more likely to be accounted for using accounting
policies which do not result in the reporting of high
profits. The effect of size on the choice of accounting
method is proxied by:BIDSIZE = Market value of the bidder's equity 41
trading days before the announcement of the bid.
TAGSIZE = Market value of the target's equity 41
trading days before the announcement of the bid.
We expect BIDSIZE and TAGSIZE to be negatively related
to the use of merger accounting.
Bidder's gearing (BIDGEAR)
Since debt covenant constraints increase in severity
with gearing, bidders with high gearing are likely to
attempt to loosen these constraints by choosing income
increasing accounting policies (see Section 3.3.7) . We
expect that bidders with high pre-acquisition gearing will
122
choose merger accounting. If this preference is reflected
in the bidders' choice of equity, such a choice in itself
may attenuate the severity of debt covenants.
4.4. The complete simultaneous equations model
The simultaneous equations system incorporating both
endogenous and control variables is as follows:
PM =
f(ACCMET, GWILL, RELSIZE, BIDGEAR, TAGGEAR,
BIDDIR, BIDLGE, RETMKT, BIDCASH, TAGCASH, VR,
FREECASH, HOSTILE, CGAIN)
(2A)
ACCMET = f(PM, GWILL, DISRES, PAYOUT, DEPFA, PROFIT, SIZE,
BIDGEAR)
(2B)
In equation 2A, as stated earlier, we assume a
nonlinear relationship between payment method and
director's shareholding. BIDDIR has a piecewise
specification as defined above.
The accounting variable (ACCMET) is dichotomous equal
to 1 when merger accounting is used and 0 when acquisition
accounting is employed.
The method of payment variable (PM) is a continuous
variable representing the proportion of equity in the final
offer and ranges from 0 to 1.
Definitions of the control variables and their data
sources are summarised in Table 4.1. Both accounting and
share price data were drawn from the Datastream
123
Table 4.1.
Definition of control variables used in the simultaneous
equations model.
Variable
Definition
Method of_Accounting Variables'.
GWILL
(Offer value - target's net assets) / bidder's net assets
DISRES
Target's pre acquisition accumulated distributable reserves
assets.
PAYOUT
Bidder's pre-acquisition dividend / net income.
DEPFA
Target's depreciable fixed assets / net assets.
I net
Target's return on equity / bidder's return on equity. Return on
PROFIT
_____________ equity is profit for ordinary shares after tax / net assets.
BIDSIZE
Market value of bidder's equity at day -41 relative to bid
announcement day 0.
TAGSIZE
Market value of target's equity at day -41 relative to bid
announcement day 0.
BIDGEAR
Bidder's total liabilities / total assets.
Method of_Payment Variables2
TAGGEAR
Target's total liabilities / total assets.
RELSIZE
Market value of the bidder's equity / market value of the
_____________ target's equity. Market value is at day -41.
BIDDIR
Proportion (%) beneficial & non beneficial shares held by the
directors of the bidder at acquisition announcement.
BIDLGE
Proportion (%) of all shareholdings greater than 5% in bidder
_____________ (excluding directors' shares) at acquisition date.
RETMKT
Cumulative market return during days -1 20 to -41 before bid
announcement day 0.
RETBID
Cumulative raw returns to bidder's equity over days -1 20 to 41.
VRBID
(Market value of the bidder's equity at day -41 pIus book value
of debt) / book value of total assets.
124
Table 4.1. (Continued)
Definition of control variables used in the simultaneous
equations model.
Variable
Definition
VRTAG
(Market value of the target's equity at day -41 plus book value
of debt) / book value of total assets.
NPDNA
Bidder's profit after tax, dividends and minority interest plus
_____________ depreciation / net assets.
NPDFLOW
Dummy variable = NPDNA if the bidder's valuation ratio
____________ (VRBID) is less than 1, otherwise = 0.
WCONA
NPDNA plus minority iiterest share of profit plus increase in
long term provisions minus associate company profit / net
assets.
WCOFLOW
Dummy variable = WCONA if the bidder's valuation ratio
_____________ (VRBID) is less than 1, otherwise = 0.
CFONA
WCONA plus increase in current liabilities minus increase in
stocks minus increase in debtors / net assets.
Dummy variable = CFONA if the bidder's valuation ratio
CFO FLOW
_____________ (VRBID) is less than 1, otherwise = 0.
BIDCASH
Bidder's cash plus marketable securities / net assets.
TAGCASH
Target's cash plus marketable securities / net assets.
Dummy variable = 1 if the bidder's first offer is rejected by
HOSTILE
_____________ target_management, otherwise = 0.
CGAIN
(Target's day -41 market value - target's lowest market value
over the preceding one year) / the lowest market value.
TAGLGE
Proportion (%) of all shareholdings greater than 5% in target
(excluding directors' and bidder's holding) at acquisition date.
Notes:
1) For GWILL, DISRES, PAYOUT, DEPFA, PROFIT & BIDGEAR accounting
data are drawn from the last financial statement before the first bid
announcement.
2) For TAGGEAR, VRBID, VRTAG, NPDNA, WCONA, CFONA, BIDCASH &
TAGCASH accounting data are drawn from the last financial statement
before the first bid announcement. For BIDDIR, BIDLGE & TAGLGE
shareholding data at the announcement date are drawn from the Extel
financial news summaries.
125
International database.
4.5. Sample selection
The sample contains all takeover bids which meet the
following criteria
1) The bid was declared unconditional between 1st
January 1980 and 31st December 1990. The sample period was
chosen to reflect a full takeover cycle following the
previous peak of 1976-78.
2) Both the bidder and the target had to be quoted on
the London International Stock Exchange (LISE) at the time
of the bid. Companies not quoted on the LISE generally do
not have the option of offering equity as a method of
payment.
3) The bid must have been successful and the bidder
must have achieved control of the target. Unsuccessful bids
were excluded from the sample because the accounting policy
which would have been chosen by the bidder cannot be
observed.
4) Datastream must have one year of pre-bid accounting
data for both the bidder and the target.
5) ]Jatastream must have at least 120 trading days of
pre-bid daily share prices for both the bidder and the
target. This ensure that we can calculate the pre-bid
return on the bidder's equity (see definition of RETBID in
Section 4.3.1).
The announcement date for each bid is defined as the
earliest of the following three dates:126
i) First approach date: This is the date when the
stock exchange is initially informed that merger talks are
underway.
ii) First bid date: This is the date of the first
formal offer. In the case of offers where there are
multiple bidders, the event day for the target is the date
of the formal bid from the first bidder. In cases where the
subsequent bidder is successful, then the bidder and the
target need not necessarily have the same event date.
iii) Unconditional date: In a mandatory bid 9 , this is
the date when the bidder acquires more than 50 of the
target's voting rights. In a voluntary bid the bidder may
set a higher minimum acceptance level (eg, 90) for the
offer to become unconditional.
The final sample consists of 505 takeover bids.
Various parts of this thesis require different amounts of
data for the relevant analysis. These data requirements may
reduce the sample size. Reductions in the sample size due
to missing data are indicated at the appropriate places.
505 is the maximum sample size used in any part of this
study.
4.5.1 Sample characteristics
Table 4.2 gives the distribution of the sample
classified by both payment and accounting methods. As seen
9 under Rule 9 of the City Code on Takeovers and Mergers, where a
bidder has acquired 3O of the target's voting shares or if already
owing 3O or more, has increased its holding by 1% or more in the last
12 months, then a mandatory full offer must be made for the remaining
shares. A voluntary offer is any bid other than a mandatory offer.
127
0
o
LUQ
I-
•
I_
0
LU0
00
LUQ
0
I-
0
ZC
>-
0)
C
C
0
(00
W
II a)
0
-c
U.
a)o
0_a)
—
.Ca).t C)
coO
cON
cOo
U)
LUQ
a)
0)
a)
C
oa(
0)R
LUN
'I
_a)
80)
c'J
.:
0
>.)
C
a)
ci)
E
0
C
a)
a)
.0
a)
a)
C
-
iC
•C
>.0
C')
a)
>..i.D
E
>.
a)
U. UJ
-H
0
—
a)
0
0
LU
N
•0
•0
C
0.
.
U]
Ui
(I)
U]
ci)
' d
o.
,
.lJ
4)
ci)
E
.IJ
(I)
-i
('I
P4
>
V—
-0
a_
>.
cO V
cO
w
a)
C
a)
U.
a)
>..0
>. a)
D_a)
a)
—
II
0C 0
-I
<
w
a)
0
0)a)
a)
'I)
0
0
0
0
0
0
(1
U.-
N
'LU
wJO
-
I
c..t
C.)
ci
>(i
>1
.Q
o
-H
.J
-H
-1
.1-)
• U]
(N-H
c11
0—
w .0
-H
bi
C
0 <t,
C'J
00)
r-Z
LU
c,l
V
J
U.
0
.1)
t1.l
0
—
V
0
.0
a)
0
a)
LU
C.,
C
0
C.,
-c
I)
0•
C.,
b
UI
cUr-I
(I)
r 1P4 0
E-J
0)
C
—
r1
rtl
ci)
QE
(lnl
— a)
h_a)
00
.Ca)
0
-C
a)
II
—
U.
-C
cON
0
<8
C.)
a)
-i
cij
4
a)
0
z
128
in Panel A, equity offer with a cash alternative is the
most widely used method of payment. This method of payment
constitutes 44% of the entire sample. The cash alternative
can either be provided out of the bidder's own resources or
through an underwriter. Such a large proportion of offers
with a cash alternative attached to equity offers may be
due to the requirements of rules 9 and 11 of the City Code
(see Section 2.2.2).
Alternatively, the popularity of equity offers with a
cash alternative could be due to its tax efficiency. Target
shareholders who wish to avoid the crystallisation of CGT
can choose the equity offer, while shareholders who do not
want the bidder's equity in their portfolio can choose the
cash offer. The underwriter provides a source of cash for
an illiquid bidder. The underwriter can also act as a
mechanism for signalling the true value of the bidder's
assets to the market as in an underwritten rights issue
(Slovin et al, 1990) . Any shares left with the underwriters
after the bid is either offered to the existing
shareholders of the bidder in a claw back offer (vendor
rights) or placed in the open market (vendor placing).
The data in Panel A is classified by the method of
payment offered to the target shareholders. It does not,
however, tell us what proportion of the final consideration
is cash or equity. Panel B of Table 4.2 classifies the
sample by the proportion of equity in the final
consideration. This classification demonstrates that in the
UK it is possible to divorce the consideration offered by
129
the bidder from that received by the target shareholders.
In offers where PM1
=
lOO (all equity with underwriting)
the bidder would record a l00 equity offer in its
accounts, but target shareholders accepting the cash
alternative would have received cash not equity.
Analysis of the sample not reported in Table 4.2 shows
that the 1987 stock market crash had a dramatic impact on
equity offers 10 . As a proportion of the annual bids total,
all equity offers (PM I = l00) dropped from 58.24 in 1987
to 30.55 in 1988, while all cash offers (PM1 = 0) rose
from 6.59 in 1987 to 23.6l in 1988. The market crash also
diminished investment banks' enthusiasm for underwriting
the cash alternative to equity offers. As a proportion of
total annual bids, equity offers with an underwritten cash
alternative rose from 7.69 in 1981 to 43.88 in 1986
coinciding with the bull market rise of the 1980's. While
the stock market crash of October 1987 did not
significantly reduce takeover activity it had a devastating
impact on cash underwritten equity offers. Their proportion
of the annual total dropped to 9.72 in 1988 and remained
close to this level in 1989 and 1990.
Panel C of Table 4.2 gives the distribution of the
sample by accounting method used by the bidder. We find
that only l0.95 of UK bidders chose to use merger
accounting. This is similar to the figure of 9.l1 found in
Higson (1990b) . It is interesting to observe from Panel B
that while at least 201 firms (PM = 100) satisfy the SSAP
10 See
the more detailed tables in Appendix 4.
130
23 condition for using merger accounting (i.e, at least 90
of the total consideration must be equity) only 54 firms
actually used merger accounting. This suggests that bidders
probably don't attach a great deal of value to the ability
to use merger accounting.
The possibility of combining acquisition accounting
with merger relief may partly explain why only a small
number of bidders choose merger accounting. Panel C shows
that more than 5O of bidders choose to combine acquisition
accounting with merger relief. In fact Appendix 4.3 shows
that bidders adopted the use of merger relief quicker than
they did merger accounting after the 1981 Companies Act
legalised merger accounting and introduced merger relief.
Four bids with announcement dates in 1980 but whose
consolidated accounts were published in 1981 took immediate
advantage of the merger relief provisions of the 1981
Companies Act. The first use of merger accounting by any
bidder in our sample did not occur until 1983 by which time
the use of merger relief was already quite popular.
Acquisition accounting with a dangling debit refers to
the practice of creating a new reserve with a zero balance
and writing off goodwill to this reserve thus creating a
debit balance. The net effect of the two methods - goodwill
written off to reserves or as a dangling debit - is the
same, viz, reduction in shareholders' funds. These methods
only differ in balance sheet presentation.
Acquisition accounting with a capital reserve refers
to negative goodwill which is credited to a special reserve
131
account on the balance sheet.
4.5.2. Descriptive statistics
Table 4.3 reports the mean, median and standard
deviation on the control variables. The high standard
deviations of GWILL, BIDSIZE, TAGSIZE, RELSIZE and VRBID
indicate that the distributions of these variables are
highly skewed. To minimise the impact of extreme outliers
in these variables, except GWILL, we use the natural
logarithms of these variables in our regression analyses.
Since GWILL includes negative values we are unable to apply
a similar transformation.
Table 4.4. reports the correlation matrix among the
variables 11 . There isn't a serious problem of multicollinearity in the models 12 . Out of a total of 190
pairwise correlations among 20 control variables only 7
exceeded 0.30 and only 15 equalled or exceeded 0.20. The
largest correlations were: 0.59 between the bidder's size
(BIDSIZE) and relative size (RELSIZE), 0.55 between the pre
bid return on the market index (RETMKT) and pre bid return
on the bidder's shares (RETBI]J), -0.46 between the target's
' 1Due to constraint of space, only one of the proxies for growth
opportunities (VRBID), free cash flow (NPDFLOW) and size (BIDSIZE) are
included in Table 4.4. Growth opportunities in the target (LOG of
VRTAG) had a correlation coefficient of 0.54 with growth opportunities
in the bidder (LOG of VRBID) but was not correlated with any other
variables. Target size (LOG of TAGSIZE) had a correlation coefficient
of 0.56 with bidder size (LOG of BIDSIZE) but had low correlations with
other variables. Apart from a high correlation with NPDFLOW (which
results from the definition of these variables) the other two proxies
for free cash flow (WCOFLOW and CFOFLOW) had very low correlation
coefficients with the other control variables.
t2None of the variables with high correlation co-efficient occur
simultaneously in the same equations.
132
Table 4.3.
Descriptive statistics for the control variables used in
the simultaneous equations model.
The variables are defined in Table 4.1. The number of
observations is different for each variable because
complete data was not available for all variables.
VARIABLE
MEAN
MEDIAN
STD DEV
No OF OBS
GWILL
0.93
0.22
3.42
505
DISRES
0.30
0.35
0.69
505
PAYOUT
0.41
0.38
0.35
504
DEPFA
0.86
0.72
0.71
505
PROFIT
1.18
0.77
2.76
499
BIDSIZE
( EM )
424.43
504
90.91
1197.21
_____________ ___________ _____________ _____________
504
LOG OF
4.63
4.51
1.74
BIDSIZE___________ __________ ___________ ___________
TAGSIZE
( EM )
191.50
65.79
13.75
505
______________ ___________ _____________ _____________
LOG OF
2.91
2.62
1.49
505
TAGSIZE___________ _________ __________ __________
BIDGEAR
0.28
0.26
0.15
505
TAGGEAR
0.28
0.27
0.15
505
RELSIZE
29.67
4.97
164.80
504
LOG OF
1.73
1.60
1.53
504
RELSIZE___________ __________ ___________ ___________
BIDDIR
0.12
0.03
0.17
505
BIDLGE
0.12
0.06
0.16
505
RETMKT
0.06
0.07
0.10
505
RETBID
0.09
0.08
0.22
505
VRBID
1.76
1.14
7.36
504
LOG OF
VRB ID
0.16
0.14
0.84
504
133
Table 4.3. (Continued)
Descriptive statistics for the control variables used in
the simultaneous equations model.
The variables are defined in Table 4.1. The number of
observations is different for each variable because
complete data was not available for all variables.
VARIABLE
MEAN
MEDIAN
VRTAG
1.36
1.01
STD DEV No OF OBS
2.65
505
LOG OF
0.06
0.02
505
0.60
VRTAG___________ __________ ___________ __________
NPDNA
0.14
0.l4
0.24
505
NPDFLOW
0.05
0
0.09
504
WCONA
0.15
0.15
0.28
476
WCOFLOW
0.05
0
0.09
503
CFONA
0.15
0.15
0.48
476
CFOFLOW
0.06
0
0.27
503
BIDCASH
0.26
0.16
0.34
503
TAGCASH
0.18
0.07
0.32
503
HOSTILE
0.23
0
0.42
505
CGAIN
0.55
0.33
1.27
505
TAGLGE
0.17
0.12
0.17
502
distributable reserves (DISRES) and the target's
depreciable fixed assets (DEPFA) and 0.40 between the
bidder's size (BIDSIZE) and directors' shareholding in the
bidder (BIDDIR).
The negative relationship between firm size and
directors' shareholding is consistent with evidence from
Demsetz & Lehn (1985), Mikkelson & Partch (1989) and Song
& Walkling (1993) . Demsetz & Lehn (1985) argue that ceteris
paribus larger firms have larger capital requirements.
Hence any attempt to preserve a high level of ownership
concentration by a small group of investors implies a
134
C)
00—O,-0
0N1.QQ
>dd c?99ddd9dd dd9dc?9d9
U)
N
;d9
CD CD
CD C)
LC) CD c' CD N
000O00QC)
0 '-
9999c99 d909990>Qz
C')
, N
d99dcd9999d dd999d
>>
d
CD CD
•
Ho
a
iHrl
CD
C)
C) N
QQ0N
Ododod00000 000
N
(U N
N
C)
CD
CD -
0
w
N
-
0-
0o
CD).tLO0CDOONOC) 0'CD
_000
>dd69dd99ci ?9oo
II
II
CDC\J
W
;dd999dd q d c 9o9
- ('4 N
0 0 CD '- N ('4 N
O000O0oOO
H
CD
4J
LU
N
CD C')
90>
UHO
U)()
0 uj
IL)
o >dC?999dd9dd o
N
(d
>>>
"NC)
>>>
CD '- -
;d999d9dd9d
;d99d99dc99;
0
W
W m L)
C) (D
;o99odcoc?9;
0,,,
",
NO
N'0000,O)
- > 0 çqcD 000 >
U)
00W
>-go
<ml
NCDtN
N
0
N 0 0
II II
"ON
C)NCD
4-4bCD .'-OO •CD
N1-
-H >°0009>
r-
(')rcO
w
Odd
0_J
4)(U
H
"CD
>999>
>>>
E•
O
>90>
-
0>
II
H-H
(WW .
Hd >>
CD
-
135
CD.-
commitment of a larger proportion of their wealth to a
single enterprise. Normal risk aversion results in a
negative relationship between size and the concentration of
shares held by any small group of investors (e.g,
managers)
4.6. Conclusion
In this chapter we develop a conceptual framework in
which a bidder's choice of payment method and its choice of
accounting policy to account, post-acquisition, for the
business combination are jointly determined. We formulate
a simultaneous equations model to reflect this joint
determination (see Section 4.2.1) . A number of exogenous
factors affecting the choice of either payment currency or
accounting policy are included in our model as control
variables.
We also describe the criteria used in the selection of
the firms in our sample and we examine the sample
characteristics in order to get a feel for the data. The
Pearson correlation coefficients among the control
variables show that most of the variables have very low
correlations. This suggests that multicollinearity will not
be a problem in the simultaneous equations model presented
in the next chapter.
136
-
0
0
-
0
0
0
0
0
0
2o 2d
2
2
2
0
0
d
2
ä
2
o- o -0
-
-
>-
0
0
0
0
2
2
2
0
0
0
2
0
CD
C.lC0 ,-N NO.
-
(_) =
-
-
N
o8
CD
.•N
N
-
L
CD
-
0
0
0
Oo
LLQ
d
2CD2
2
N
0
O C0
CD
LC)
N
CD
-O C0
C) 'N
N
N
0)
0)
CD
0
0
Nh
2 2 ! !
CD
0)
o W
LC)
N
0
oO oQ -'-
--
0
0
2
C)
0
LD0 CC
0 03N
N
_
6
o
-c - o
CO
0
N- CO
0)
-o
C
o
6
co
R
6
6
-E
R
0 (DC
CD ,_CO
N
CO
0 03 c'J
0 NCO 0)03
N 'O
-
5 () C
-c
o '
-a
6
N
N
0
(0
CD
CD
6
CD
s-CD ,-
CO
6
N coN NO
o
LU
-c
U,
CO
0
0
6
.
CD
2
-
C) L50
- -
CO
6
0)
6
0
0
-
6
6
6
6
CD
R
f-'- oq o0 -'-:
-
-
In
CD
0
N
o
COQ
6
CO
6 6 6 6co'R 6 6
CD
0
-
oQ
0
-
oq
0
-
-.
0)
-
-
N°
03
-
R
N
-
U,
o
-o
2
o
o0_ sN
C
-
OQ
C)
s-
a)
oil>
I 0)
I-
-
-ico
-H G)
><
2
N
-
6
-
CO
o =
O
(0°-
6
N
0
R 6
N
s-'
6 6
CD- 03'
C'J
N
0
CD'
(O'
6 R 6 - 6 6
.,-N
R
'-'
6
.-
6
2
6
6
NC0
CO>
6
6
CO
(0
CO
N
-
N
6
R 6
6
6
6
6
NN 10'- ,-CD N0
N
cd
-IJ
-H
'd
-H
II
aJjJ
4 u)
IN
I
lCD
Ia)
I CO
I
CD
I N103
I103 lCD lCD 103 lCD 103 lCD lCD 103 lCD
10
U
-1
-H W
I C)
I'-
Ia>
I'
IC)
I
IC)
I.—
137
IC)
I.-
IC)
I-
ID)
i-
IC)
1s
IC)
I '—
0
0,
iC)
i.
.0
il-
R
0
0
0
a
o
0
0) -
o—
0
0
0
0
a
o
2o0
0
a
o0
0
0
0
0
o
0
o0
0
0
o
0
0
0
o
0
0
0
o
0
o
0
0
a
o0
0
O
0.,
o
R - 6 6 6
6 -
-H
0 cON
0
C4
6
0 c0
cO
3-
-H
U]
-H
o
o
-a.
u
w
C4
0
N-
2
-
co
.
co.
o
.
0
4J
—
-H
R 6 o — — 6
V
-
C
V
>1
N
N
N
-
tr
—
N0.
0
N
-
6
—
N
-.-
-
r—
'-
6
IL)
-
6 —c'L
6 R 6 6
NN cOL
'cj
N
—
.—
LL)
—
-.
O
N
co
-
6
—0
6
°
-
NIL,
-
-H
N
WW
U]
4-W '
°
v
—
—
csJ
N
O
-.
N
0)
c')0
-.
0)-
U)
<0C)
u,C'
V
ow
0
p4
0
0-H
V—
—
0
a)
.—
N
3V
IL'
N
6
0
—
-
o6.
2
— — —
N
6
6
N-
2—
-
6
—
—
4Jb—
>1W
0) V
WH OGH(
Q1jj
v
EQ
(I]
r-4
6
o°
— —
6
-- OQ N
N
6
2
N'
6
-Q
LD
0)
—
6
0
-
U)
6
N
—
wrj
44(T
0
N 0
• 0
cJl -H a)
4Jt3-)
0
0
6 6 ,CC
R.—IL)6
U
C
0
0)
(0
,—
U)
0)
—
••(0
N
.
6 R
0)°
'-
0)'1
6
N
-
'Ce)
—
0)
U0
LL)
('4
('1
—
(0 0)N
-
-.
___
-
0
—
—
N
0)
U)
(0
N
(0
(0
(0
0)
0
=
=
0)
—
a)4J U
Q.1U]
P.1-H
1q
a)
fl
=
=
=
=
=
=
=
138
=
=
=
=
°'i J J 4
0
I-
iJ J
C)
6 6 6
o 0 oQ 00.
2 2 9
0)0
'
0
.
LX
co
LC
R
6
oQ
2
2
0)
.
:
0)
0
0•
O
C=
C)
00C
0
0
o
0
0
0
0
c)
-C'
N
0
QC
0
C')
IC)
-'
C\J
-
6
6
(0
IC)'
LL)C'
N
0)
6
6
0
00
0
N
<CD
'ci
o
U)
N
:
-
N
(I]
b'Q)
w
N
o
cs
6
N
C')
-
0
0
'
6
0
d
00 Nh
oa
-HU)
<
4J,
OW
O
C')
R
CD
2
0
Li LU
(0
0)
.
(0
.
.
(0
C')
IC)
IC)
CO
6 6 6
(0C• N
C')
'0)
C)<
'O)
- 6 R 6 6
(00
'N
-C'
o
N
C')O
'
0
,-LC)
00
(j
nl
W-H
rn
0
•o
WH
Ic
6 6 6 R 6 6 6 6 6
c',°
oQ oO. c 'C
oQ
9
9 9
2
6
6 6
oO 00
2
<
0
J
<0
(iu
(I]
iH
W1
CC)
0CC)
4J
CD
C')
,,
L4
U
o
00.0
.
C'J
0
IC)
2
C')
C')
(0
(,)C'J
N
'-d
C')
C')
.-
(0
P.-H
N
N
C')
N
--
•
u-I
mo
• o ____
ct' -H 1)
4Jb)
Q) >w4_J 0
p4(i) -I
-_
..
-
2
(0
C)
(0
C)
T
U)
_____ -
139
C')
c')
IC)
CD
cC)
c0
IL)
CD
(0
0)
(0
0)
-
CD
CD
.,j-IC)
C0
N
(0
C)
N
ç.JN
C')
-
(0
(0
C)
0
0)N
NC,)
IC)
C)
(0
C)
0
0
(0
LC)
C)
CL)
(0
N
0
0)
C)
e
0
I-
CHAPTER 5
INTERACTION BETWEEN ACCOUNTING POLICY AND THE METHOD OF
PAYMENT: RESULTS OF EMPIRICAL ANALYSIS.
5.1. Introduction
The purpose of this chapter is to examine the
relationship between the method of payment and the
accounting policy choice decision of the bidder within the
context of a simultaneous equations model which recognises
the mutual dependence between the method of payment and the
choice of accounting method. In the chapter we use the Two
Limit Tobit model which overcomes some of the
methodological limitations of previous studies (see
Sections 1.2.4 and 4.2)
Some of the exogenous/control variables used in the
simultaneous equations model are new, and allow us to
investigate issues which hitherto have remained unexplored
in the literature. In our model of the determinants of the
choice of accounting policy, we test the debt covenant
restriction hypothesis (see Section 3.3.7) and the
political cost of earnings hypothesis (see Section 3.3.6)
which have not been previously tested in the context of
accounting policy choice in corporate acquisitions.
Some of the new issues examined in our study of the
determinants of the payment currency in corporate
acquisitions include the relationship between the method of
financing new investments and:
(i) the dilution of shares held by managers and
140
external blockholders;
(ii) the growth opportunities in the firm;
(iii) managerial use of free cash flow.
The major contribution of this chapter is to recognise
explicitly and model the mutual dependency between
accounting policy choice and payment method. We believe
that this approach provides an insight into the more
general issue of the interaction between financing
decisions and accounting policy choice.
5.2. Univariate test
As a preliminary step in our analysis and to get a
feel for the data, we carry out univariate tests 1 in which
we compare the statistical significance of the difference
in group means for the following subsamples:
1) All equity offers (equity group ie, PM 1 = lOO) with
all cash offers (cash group ie,PM = O)
2) Bids where the bidder has used merger accounting
(merger group ie, ACCMET = 1) with bids where the bidder
has used acquisition accounting (acquisition group ie,
ACCMET = 0)
3) Bids where the bidder has used merger accounting
(merger group) with bids where the bidder was qualified to
1 We apply in addition to parametric tests, non-parametric testing
procedures which make no assumptions about the underlying distribution
of the data. Where the non-parametric and the parametric tests give the
same results, then we are reassured that our conclusions are
insensitive to deviations from non-normality in the data.
141
use merger accounting 2 but instead used acquisition
accounting (qualified acquisition accounting group).
5.2.1. Cash offers and equity offers
Table 5.1 reports the results of difference-in-means
tests comparing the "equity group" with the "cash group".
Since the parametric t-test for the difference in group
means can be performed based on the assumption of either
equality of variance or inequality of variance between the
two groups 3 , we report both sets of results. The nonparametric test of the difference in group medians [MannWhitney-Wilcoxon test 4 (MWW)] is also reported, in order to
check whether any deviations from non-normality in the data
will significantly influence our conclusions.
Goodwill as a proportion of the bidder's net assets
(GWILL) is 10.91% in the "cash group" and 146.69% in the
"equity group". The average size of potential goodwill in
equity offers (146.69%-) is very large and different from
the overall sample average of 93% (see Table 4.3) . This
suggests the presence of some outliers in the equity group.
However, since the non-parametric MWW test is significant
at better than the 1% level, this result is robust to nonnormality in the data.
The larger size of GWILL in the equity group suggests
2Bids where the bidder fulfilled all the conditions for using
merger accounting stipulated in SSAP 23 (see Section 3.2.1)
3 See Wonnacott and Wonnacott (1977: chapter 8)
4See Wonnacott and Wonnacott (1977: Chapter 16)
142
Table 5.1.
Difference in means between the "equity group" and the
"cash group": Univariate tests.
The "equity group" represents bids where PM 1 = lOO and the
"cash group" represents bids where PM1 = O. PM1 is the
proportion of equity in the final method of payment. The
"equity group" has 184 observations and the "cash group"
has 89 observations. The means for the "cash group 1' and
"equity group", the t-Statistics testing the difference in
group means assuming both unequal and equal group variances
and the non-parametric Mann-Whitney-Wilcoxon test statistic
are reported. The variables are defined in Table 4.1.
Variable
Mean For
Cash
Group
Mear For
Equity
Group
GWILL
0.11
1.47
BIDGEAR
0.27
0.31
TAGGEAR
0.30
RELSIZE
t-Stat
Assuming
Unequal
Variance
t-Stat
Assuming
Equal
Variance
MannWhitneyWilcoxon
Test.
3.05"
7.01"
1.73"
1.82"
2.72'"
0.28
0.50
0.52
0.23
94.88
10.67
2.15"
3.08'"
6.16'"
LOG OF
RELSIZE
2.70
1.37
6.17"
7.05'"
6.16'"
BIDDIR
0.11
0.14
1.18
1.26
4.05"
BIDLGE
0.12
0.11
0.37
0.40
0.58
RETMKT
0.04
0.08
2.92'"
3.07'"
3.18'"
RETBID
0.05
0.13
3.02'"
2.63'"
2.60'"
VRBID
1.12
2.58
1.63'
1.14
5.22'"
LOG OF
VRBID
-0.06
0.29
3.28'"
2.61'"
5.22m
VRTAG
1.16
1.29
1.12
1.23
2.73m
LOG OF
V RTAG
-0.05
0.11
2.17"
2.26"
2.73'"
NPDNA
0.14
0.13
0.33
0.24
0.84
WCONA
0.15
0.12
0.70
0.52
0.25
CFONA
0.25
0.05
2.99'"
2.68"
2.48'"
4.25*
Note:
1) "' "' Significant at 1 %, 5%, 10% levels respectively, one tail test.
143
Table 5.1 (Continued).
Difference in means between the "equity group" and the
"cash group": Univariate tests.
The "equity group" represents bids where PM 1 = 100% and the
"cash group" represents bids where PM1 = 0%. PM1 is the
proportion of equity in the final method of payment. The
"equity group" has 184 observations and the "cash group"
has 89 observations. The means for the "cash group" and
"equity group", the t-Statistics testing the difference in
group means assuming both unequal and equal group variances
and the non-parametric Mann-Whitney-Wilcoxon test statistic
are reported. The variables are defined in Table 4.1.
Variable
Mean For
Cash
Group
Means For
Equity
Group
t-Stat
Assuming
Unequal
Variance
t-Stat
Assuming
Equal
Variance
MannWhitneyWilcoxon
Test.
NPDFLOW
0.05
0.03
2.42m
2.53w
3.15"
WCOFLOW
0.06
0.04
1.79"
1.86"
CFOFLOW
0.13
0.03
2.31"
3.17"
2.14"
2.46w
BIDCASH
0.35
0.21
2.67"
3.04"
4.07"
TAGCASH
0.19
0.20
0.13
0.13
1.11
HOSTILE
0.20
0.21
0.08
0.08
0.08
CGAIN
0.41
0.51
1.43
1.46
1.63
TAGLGE
0.16
0.16
0.24
0.25
0.92
Note:
1)
" Significant at 1 %,
5%, 10% levels respectively, one tail test.
that bidders who face a substantial goodwill write off are
more likely to use equity offers as a method of avoiding
the negative impact of writing off goodwill. Equity offers
can ameliorate the adverse consequences of goodwill in the
consolidated accounts through the use of either merger
accounting or the merger reserve. This choice is discussed
further below (see Tables 5.2 and 5.4)
Bidders using equity have an average gearing (BIDGEAR)
of 30.56% compared to 27.17% for bidders using cash. This
144
is consistent with our prior hypothesis that highly geared
bidders are more likely to use equity because the costs of
raising the necessary funds to finance a cash offer are
positively related to the firm's level of gearing.
In cash offers bidders are 94.88 times larger than the
target (RELSIZE) but only 10.67 times larger in equity
offers. This is consistent with the information asymmetry
hypothesis that large targets are acquired with equity in
order to reduce the valuation risk faced by the bidder
(Hansen, 1987) . The statistically significant difference in
relative size between cash and equity offers is still
maintained when we use the LOG OF RELSIZE and the MWW nonparametric test. This indicates that this result is robust
and is not being driven by outlier values.
Over a period of 81 trading days (4 calender months)
ending 41 days before the announcement of the bid, bidders
using equity experienced an average rise in their share
price of 13% (RETBID) compared to 5% for bidders using
cash. This difference is significant at the 1% level. Over
a similar period, the average rise in the stock market when
bidders used equity was 8% compared to an average rise of
4% when bidders used cash. This supports the hypothesis
that equity is more likely to be used in a favourable
market.
Growth opportunities in the bidder (VRBID) and the
target (VRTAG) are higher in bids financed with equity than
for bids financed with cash. This is consistent with Myers'
(1977) model that firms with growth options should finance
145
new investments with equity (see Section 2.9) . The presence
of outliers in the distributions for VRBID and VRTAG (see
Section 4.5.2) may explain why better results are obtained
with the natural logarithms of these variables and the nonparametric MWW tests.
Bidders using cash have higher cash flows (NPDNA,
WCONA and CFONA) than bidders using equity, irrespective of
the cash flow definition used. However only the CFONA
variable is statistically significant. The impact of cash
flow on the method of payment is based on the concept of
free cash flow. To investigate the free cash flow idea we
use variables which combine our cash flow measures with the
firm's growth opportunities proxied by the valuation ratio
(NPDFLOW, WCOFLOW and CFOFLOW). With all three cash flow
measures we find that cash paying bidders have
significantly higher free cash flow than bidders using
equity. This suggests that among bidders with no growth
opportunities, those with a higher level of cash flow tend
to use it to make cash acquisitions. This observation is
consistent with Jensen (1986)
Not surprisingly we find that bidders using cash are
more liquid than bidders using equity (BIDCASH) . It is
difficult to decide how much importance should be attached
to this result, since we may simply be observing the
effects of anticipatory financing by bidders planning to
use cash. However when we consider that bidders using cash
have both higher liquid resources and higher cash flow than
bidders using equity, anticipatory bid financing may not be
146
the factor driving this result 5 . We do not find any
evidence that the cash resources of the target have an
impact on the method of payment, suggesting that bidders do
not use the liquid resources of the target to help finance
the bid. This is probably due to the 1985 Companies Act
provisions which prohibit the target from providing
assistance for the acquisition of its own shares (see
Section 2.10)
There is no significant difference in hostility
between cash offers and equity offers. 23 of the entire
sample was made up of hostile bids (See Table 4.3) . The
percentages for cash and equity offers were 20 and 21%
respectively6 . This result is inconsistent with Sudarsanam
(1994b) who found that cash offers increased the
probability of success for the bidder in hostile bids. The
evidence here does not support the hypothesis that cash has
a pre-emptive role to play in a hostile takeover7.
Short run capital gains (CGAIN) are higher in equity
than in cash offers. This is consistent with the capital
gains tax compensation hypothesis. However in the
5 A preliminary analysis of rights issues in the six months
preceding our sample bids, showed that only a small number of bidders
(about 40) made such rights issues.
6The proportion of hostile bids for the overall sample is greater
than the proportion for cash and equity offers because 26% of mixed
offers were hostile bids.
7Differences in the sample selection criterion may explain this
inconsistency. Our sample is made up entirely of completed bids which
may be hostile or friendly. Sudarsanam (1994b) uses a sample of only
hostile bids, which includes both failed and completed bids.
Sudarsanam's results may be indicating that once a bid turns hostile,
then cash is more advantageous to the bidder without suggesting that
cash should be more prevalent in hostile bids.
147
multivariate regressions reported below, the effect of
CGAIN on the choice of payment method is insignificant.
Since the univariate results are based on a sub-sample of
"pure equity" and "pure cash" bids, this suggests that this
result is not uniform across the whole sample. Contrary to
prior expectation (see Section 4.3.1) targets acquired in
cash offers have higher levels of large shareholding
(TAGLGE) than targets acquired in equity offers. This could
reflect the fact that this proxy f or potential capital
gains is measured with some error.
While equity offers have higher levels of directors'
shareholdings (EID]JIR) than cash offers, the difference is
not statistically significant using the parametric t-test.
This is not inconsistent with our prior suggestion of a
non-linear relationship between directors' shareholding and
the proportion of the acquisition price financed by equity
(see Section 4.3.1) . At very high and very low levels of
directors' shareholding we expect to observe the use of
equity. When directors' shareholding is in a middle range
where the method of payment can dilute directors' control
over the firm then cash offers are used. Due to this nonlinear relationship we cannot predict a priori what the
average level of directors' shareholding would be in cash
offers relative to equity offers.
Consistent with our prior expectation (see Section
4.3.1) cash bidders have higher levels of large
shareholding (BIDLGE) than equity bidders (12 for cash
bidders compared to 11 for equity bidders). However the
148
difference is not statistically significant. In
multivariate regressions (discussed below) the impact of
the presence of large shareholders in the bidder on the
method of payment is consistent with our prior expectation
and statistically significant.
5.2.2. Merger accounting and acquisition accounting
Table 5.2. reports the results of difference-in-means
tests comparing the "merger group" with the "acquisition
group". Goodwill as a proportion of the bidder's net assets
(GWILL) is higher for bidders using merger accounting
compared to those using acquisition accounting. Although
the evidence here supports our earlier conclusion that
bidders structure the method of payment in order to avoid
the negative impact of goodwill write off, it does not
provide an answer to the question whether bidders have a
preference for using merger accounting as opposed to using
acquisition accounting with merger relief.
Table 5.2 is likely to overstate the importance of
goodwill write off as a determinant of accounting policy
choice because we have not controlled for the difference in
the method of payment between the "merger group" and the
"acquisition group". While the method of payment is uniform
in the "merger group" (i.e, primarily equity) the payment
method in the "acquisition group" includes cash. Since bid
premia is higher in cash offers than in equity offers
(Franks et al, 1988), this could cause goodwill to be
higher in the "acquisition group" than in the "merger
149
Table 5.2.
Difference in. means between the "merger group" and the
"acquisition group": Univariate tests.
The "merger group" represents bids where the bidder has
used merger accounting and the "acquisition group"
represents bids where the bidder used acquisition
accounting. The "merger group" has 54 observations and the
"acquisition group" has 439 observations. The means for the
"acquisition group" and "merger group", the t-Statistics
testing the difference in group means assuming both unequal
and equal group variances and the non-parametric MannWhitney-Wilcoxon test are reported. The variables are
defined in Table 4.1.
Variable
Mean For
Acquisition
group
Mean For
Merger
group
t-Stat
Assuming
Unequal
Variance
t-Stat
Assuming
Equal
Variance
MannWhitneyWilcoxon
Test
GWILL
0.74
2.51
1.94"
3.61"
5.26"
DISRES
0.29
0.35
1.03
0.51
0.03
PAYOUT
0.41
0.37
0.93
0.90
2.19"
DEPFA
0.87
0.77
1.47
0.97
0.77
PROFIT
1.08
1.28
0.89
0.77
1.44
BIDSIZE
( £M )
LOG OF
BIDSIZE
TAGSIZE
( £M )
472.15
112.13
5.02"
2.05"
5.43"
_____________ ____________ ___________ ___________ ____________
4.81
6.21"
3.50
5.35"
5.43"
68.54
52.12
1.03
0.59
0.62
_____________ ____________ ___________ ___________ ____________
LOG OF
TAGSIZE
2.92
2.98
0.31
0.29
0.62
BIDGEAR
0.28
0.29
0.44
0.45
0.85
Note:
1) " " Significant at 1%, 5%, 10% levels respectively, one tail test.
150
group" 8 . In Table 5.4 below we compare the size of goodwill
for bidders using merger accounting with bidders using
acquisition accounting while controlling for the method of
payment.
In Table 5.2 we can see that apart from goodwill the
only other variable which appears to affect the choice of
accounting method strongly is the size of the bidder.
Consistent with our a priori expectation based on the
political cost hypothesis (see Section 3.3.6) we find that
bidders using acquisition accounting are larger than those
bidders using merger accounting.
The size of the target (TAGSIZE) has no impact on the
choice of accounting policy. Since bidders are about 30
times larger than targets (see Table 4.3) it is probably
not surprising that the impact of anti-trust regulatory
pressure on the choice of accounting policy is more closely
related to the size of the bidder than the size of the
target.
Targets acquired by bidders using acquisition
accounting have a higher proportion of their nets assets
made up of depreciable fixed assets (DEPFA) . This is
consistent with the argument that the opportunities for
using provisions and write downs against the target's
assets to inflate post acquisition earnings (see Section
3.3.1) are greater the larger the proportion of the
target's assets which is composed of depreciable fixed
8 For evidence that the size of goodwill is positively related to
the level of bid premium see Nathan (1988) and Robinson & Shane (1990)
discussed in Section 3.4.
151
assets. However the statistical significance of this result
is weak.
Targets acquired by bidders using merger accounting
have higher levels of pre-acquisition distributable
reserves (]JISRES). Although this is consistent with our
prior expectation that bidders using merger accounting
value access to the target's pre-acquisition reserves (see
Section 3.3.3), the difference is not statistically
significant. If access to the target's pre-acquisition
reserves is important for bidders using merger accounting,
then we would expect such bidders to have high dividend
payout ratios (see Section 4.3.1) . We find that bidders
using merger accounting have lower dividend payout ratios
(PAYOUT) than bidders using merger accounting indicating
that access to the target's pre-acquisition reserves is not
a significant factor in the choice of accounting policy.
5.2.3. Merger accounting and acquisition accounting with
merger relief
In Table 5.3, we report the accounting method employed
categorised by the ability of the bidder to use merger
accounting. Out of the 439 (cases 2 to 5) bids in which the
bidder used acquisition accounting, the method of payment
precluded the use of merger accounting in 282 (cases 4 & 5)
of these bids. We therefore, focus on those cases where the
acquirers had a genuine choice between the two accounting
methods, ie, cases 1 to 3.
Out of a total of 211 such bids only 54 bids (case 1)
152
Table 5.3
Accounting method employed partitioned by the ability of
the bidder to use merger accounting.
A bidder is qualified to use merger accounting if PM 1 ^ 90%
and the bidder's pre bid shareholding in the target is less
than 20%. PM1 is the proportion of equity in the final offer
price.
Case Bids where the bidder
No of bids.
1
Used merger accounting
54
2
Qualified to use merger accounting
but used acquisition accounting
with_merger_relief
137
______
3
______
Qualified to use merger accounting
but used acquisition accounting
without_merger_relief
4
Not qualified to use merger
accounting but used acquisition
______ accounting with merger relief
5
_____
_____________
20
_____________
127
Not qualified to use merger
accounting but used acquisition
accounting without merger relief.
155
Total
493
actually resulted in the use of merger accounting. Of the
157 (cases 2 and 3) bids where the bidder could have used
merger accounting but didn't, 137 (case 2) resulted in the
bidder using acquisition accounting with merger relief. In
Table 5.4 we compare bids where the bidder used merger
accounting (merger group) with bids where the bidder was
qualified to use merger accounting but instead used
acquisition accounting (qualified acquisition accounting
group) in terms of the determinants of accounting policy
choice.
Although the goodwill write of f is still larger for
the "merger group" than the "qualified acquisition
153
Table 5.4.
Difference in means between the "merger group" and the
"qualified acquisition accounting group": tJnivariate tests.
The "merger group" represents bids where the bidder has
used merger accounting and the "qualified acquisition
accounting group" represents bids where the bidder
qualified to use merger accounting but instead used
acquisition accounting. The "merger group" has 54
observations and the "qualified acquisition accounting
group" has 157 observations. The means for the "merger
group" and the "qualified acquisition accounting group",
the t-statistics testing the difference in group means
assuming both unequal and equal group variances and the
non-parametric Mann-Whitney-Wilcoxon test are reported. The
variables are defined in table 4.1.
Variable
Mean For
Merger
Group
____________ ____________
Mean For
Qualified
Acquisition
Accounting
Group
t-Stat
Mannt-Stat
Assuming
Assuming
WhitneyUnequal
Wilcoxon
Equal
Variance
Test
Variance
___________ ___________ ___________
GWILL
2.51
1.27
1.28'
1 .59
3.1 6"
DISRES
0.35
0.21
1 .40'
0.87
0.22
PAYOUT
0.37
0.40
0.69
0.77
1.66'
DEPFA
0.77
0.99
2.23"
1.59'
1.28'
PROFIT
1.28
1.03
0.99
0.94
1.30'
BIDSIZE
( £M )
LOG OF
Bl DSIZE
TAGSIZE
( £M )
112.14
289.03
2.46"
3.97'"
3.21"
____________ ____________ ___________ ___________ ___________
3.50
4.49
4.28"
4.04'"
3.97"
42.57
52.12
0.69
0.77
0.52
____________ ____________ ___________ ___________ ___________
LOG OF
TAGSIZE
2.88
2.98
0.49
0.51
0.52
BIDGEAR
0.29
0.30
1.48'
1.60'
0.69
Note:
1)
Significant at 1 %, 5%, 10% levels respectively, one tail test.
154
accounting group" the statistical significance of the
difference is weak. Given the ability to use merger
accounting, a bidder's decision to do so is only weakly
influenced by the size of potential goodwill. This is
plausible f or at least two reasons. First, under SSAP 23,
acquisition accounting in conjunction with merger relief
provides some of the benefits associated with merger
accounting. Second, acquisition accounting may be combined
with provisions at acquisition time and this combination
has the effect of increasing the post-acquisition profits
of the bidder. These loopholes in the UK rules mean that
the distinction between merger and acquisition accounting
is often blurred.
The classic dilemma facing managers for example in the
USA, in choosing between merger and acquisition accounting
is the need to avoid the amortisation of goodwill through
the profit and loss account and its consequent adverse
impact on future earnings (see Section 3.3.1 for the
related discussion) . In the UK, the possibility of writing
off goodwill against net assets reduces the potency of this
choice. The main effect of goodwill write off is,
therefore, a reduction in reserves. Even this adverse
impact is mitigated to the extent that shares issued as
part of the consideration at a premium generate a merger
reserve to absorb the goodwill write off.
5.2.4. Summary of the univariate results
The univariate difference in means tests show that
155
1) the following variables are higher for bidders
offering equity as the method of payment: the potential
goodwill to be written of f as a proportion of the bidder's
net assets, gearing, the valuation ratio of the bidder, the
valuation ratio of the target, the level of the prebid
stock market index and pre-bid growth in the bidder's share
price;
2) the following variables are higher for bidders
offering cash as the method of payment:- liquidity, free
cash flow and size of the bidder relative to the target;
3) goodwill written off is higher for bidders using
merger accounting than for those using acquisition
accounting. However the size of the goodwill is only
slightly larger for those bidders using merger accounting
than for bidders who could have used merger accounting but
instead used acquisition accounting;
4) bidders using acquisition accounting are larger
than those using merger accounting.
5.3. Determinants of the payment method
The simultaneous equations (2A and 2B) in Section 4.4
are estimated by a two stage procedure employing the Two
Limit Tobit and Logit methodologies as outlined in Section
4.2.1. To assess whether a single equation model, rather
than a simultaneous equations model, is appropriate we also
estimate the former with payment and accounting methods as
dependent variables. Single equation modelling also allows
us to compare our results with those from earlier studies.
156
Table 5.5 presents the results for the payment method.
The Log-Likelihood ratio 9 shows that both the two stage
Tobit model (2STM) and the single equation Tobit (STM)
models are highly significant. The Wu-Hausman test
statistic (see Section 4.2.3) for the STM is not
significant suggesting that the related results may not be
tainted by simultaneity bias. All the variables significant
in one model are significant in the other with the
exception of ACCMET which is not significant in the 2STM
model. Each of the significant variables has the expected
coefficient sign. The results of the multivariate analysis
are consistent with the univariate difference in means
tests. This suggests that multicollinearity is not a
problem in the multivariate regressions.
Choice of accounting policy (ACCMET) has the correct
positive coefficient in the STM model. The difference
between the two models as regards ACCMET suggests that a
single equation model of the impact of accounting policy
choice on payment method may lead us to overstate such
impact. Thus such a model may be inappropriate. The nonsignificance of ACCMET in the 2STM model implies that
payment method choice is made independently of accounting
method choice. This could be plausibly due to the fact that
the availability of merger relief in the UK has blurred the
choice between merger and acquisition accounting (see
Section 5.2.3).
9 See Maddala (1989: p.84) for a description of the Log-Likelihood
ratio test.
157
Table 5.5.
Maximum likelihood estimates of Two Limit Tobit models
explaining the proportion of equity in the method of
payment.
The dependent variable (PM) is the proportion of equity in
the final method of payment. The explanatory variables are
defined in Table 4.1. The sample size is 471.
______________
Simultaneous Equation
Model (2STM)
Single Equation Model
(STM)
Variable
_________________
Expected
Sign
Coefficient
Estimate
t-statistic
Coefficient
Estimate
t-statistic
INTERCEPT
?
0.75"
5.14
0.68"'
5.30
ACCMET
+
0.10
0.94
0.82'"
5.49
GWILL
+
0.02
1.20
0.01
0.20
BIDGEAR
+
0.80'"
3.19
0.77"'
3.19
TAGGEAR
-
0.11
0.42
0.13
0.54
LOG (RELSIZE)
-
-0.12"
-4.15
-0.11...
-4.30
BIDDIRO TOOl
+
3.41"
2.80
3.57'"
3.05
BIDDIRO1TOO2S
-
- 1.99"
-1.88
-2.23"
-2.16
BIDDIR0ver026
+
- 0.51
-0.92
-0.42
-0.81
BIDLGE
-
-0.64'"
-2.86
-0.62'"
-2.94
RETMKT
+
1.17'"
3.33
1.11'"
3.30
LOG (VRBID)
+
0.09"
2.01
0.09"
2.03
NPDFLOW
-
-0.81"
-1.98
-0.86"
-2.20
BIDCASH
-
-0.46'"
-4.12
-0.43'"
-4.24
TAGCASH
-
0.01
0.09
0.07
0.69
HOSTILE
-
-0.04
-0.48
-0.02
-0.28
CGAIN
+
0.01
0.04
0.003
0.14
Log-Likelihood
-420.20'"
-404.29'"
Wu-Hausman
N/A
-0.40
Notes:
1) "' "' Significant at 1 %, 5%, 10% respectively, one tail test.
2) 2STM = Two stage Tobit model.
3) STM = Single equation Tobit model.
4) Reduction in the sample size is due to missing observations on some of the
independent variables.
158
The significant positive coefficient on the bidder's
gearing (BIDGEAR) is consistent with our a priori belief
that highly geared bidders are more likely to use equity
because the costs of raising the necessary funds to finance
a cash offer are positively related to the firm's level of
gearing. This is also consistent with Hansen's proposition
C that the probability of an equity offer increases with
the financial gearing of the bidder (see Section 2.4.1).
The relative size coefficient (RELSIZE) is negative
and significant at the l level. This supports the
signalling models of Hansen (1987) that large targets are
acquired in equity offers to reduce the valuation risk
faced by the bidder (see Section 2.4.1)
The significant positive coefficient on EIDDIP01,
indicates that at low levels of shareholdings directors'
concern about dilution is minimal, hence as directors'
shareholding increase there is an increase in the
propensity to use equity offers. The significant negative
coefficient on EIDDIR01025 shows that with shareholdings
between lO' and 25 directors believe that an equity of fer
will dilute their control, hence as shareholding increase
there is a decrease in the propensity to use equity offers.
The insignificant negative coefficient on EIDDIR over o 25 shows
that with shareholdings above 25 dilution of control is no
longer a major consideration, hence the payment currency is
not related to directors' shareholding. These signs are
consistent with our prior belief of a nonlinear
relationship between directors' shareholding in the bidder
159
(BIDDIR) and the proportion of the acquisition which is
equity financed (see Section 4.3.1) . It is also consistent
with the results from the univariate difference in means
test (see Section 5.2.1)
As an additional test to examine if the relationship
is non-linear with two slope changes, we estimated the
single equation model (STM) with BIDDIR in linear, squared
and cubic form. BIDDIR had a positive coefficient with a tstatistic of 2.038 (p-value = 0.021), BIDDIR2 had a negative
coefficient with a t-statistic of -1.919 (p-value = 0.028)
and BI]JDIR3 had a positive coefficient with a t-statistic of
1.495 (p-value = 0.068). Similar results were obtained for
the simultaneous equations model10.
The negative relationship between the proportion of
shares held by external blockholders (BIDLGE) and the
probability of an equity offer shows that the concern of
large shareholders about the dilution of their holding
influences the financing decision of firms.
The significant coefficient on the pre bid return on
the market index (RETMKT) is consistent with the existing
evidence in the literature that companies make equity
issues during rising stock markets when investors may
favour such issues. We re-estimated the single equation
model (STM) using the return on the bidder's equity
(RETBID) in the pre bid period as a proxy for market
10 The single equation model (STM) was also estimated with
directors' shareholding (IDDIR) in linear form only. BIDDIR had a tstatistic of -1.344 (p-value = 0.089) . While this result is weakly
significant, the results from the non-linear specification are much
more powerful.
160
conditions. RETBID had a positive coefficient with a tstatistic of 2.66 (p-value = 0.001) indicating that bidders
also timed the use of equity to coincide with increases in
their share price. This is consistent with the Myers &
Majiuf (1984) misvaluation hypothesis that bidders will
make equity offers when their shares are overvalued (see
Section 2.4.2). Since from Table 4.4, we know there is some
correlation between RETEID and RETMKT, we do not use both
variables in the same equation. The significant and
positive coefficients on RETMKT and RETBID support our a
priori belief that favourable market conditions can
influence the method of payment.
The valuation ratio of the bidder (VRBID) has a
significant positive impact on the proportion of equity in
the method of payment. This supports the Myers (1977) model
that firms with growth opportunities are more likely to use
equity as a method of financing new investments (see
Section 2.9) . Since an equity offer lowers the bidder's
debt/equity ratio our result is also consistent with the
evidence in Smith & Watts (1993) that the firm's valuation
ratio is negatively correlated with the debt/equity ratio.
To examine if growth opportunities in the target affects
the method of payment, we re-estimated the single equation
model (STM) using the target's valuation ratio (VRTAG) as
a proxy for growth options. VRTAG had a positive
coefficient with a t-statistic of 1.463 (p-value = 0.072)
This suggests that growth options in the target do affect
the choice of payment method but their influence is not as
161
strong as that exercised by growth options in the bidder.
Free cash flow (NPDFLOW) has a significant negative
coefficient. This shows that for low q bidders the
proportion of cash used in financing acquisitions increases
with their free cash flow. This is consistent with the
Jensen (1986) model of free cash flow. To check the
robustness of this result we re-estimated the single
equation model (STM) using alternative definitions of cash
flow (see Table 4.1). The variable WCOFLOW had a negative
coefficient with a t-statistic of -2.045 (p-value = 0.021)
The variable CFOFLOW had a negative coefficient with a tstatistic of -0.887 (p-value = 0.188) . The cash flow from
operations definition (CFOFLOW) which conceptually is the
most accurate is not significant. A possible explanation is
that this measure is subject to a greater degree of noise,
because it is more sensitive to accounting practices and
adjustments f or nonrecurring items (Lang, Stulz & Walkling,
1991)11. Results for the simultaneous equations model are
similar.
The bidder's liquidity (BIDCASH) has a significant
negative coefficient impact on the proportion of equity in
the final consideration, while the target's cash resources
do not influence the choice of payment method. This is
11 The adjustments required to calculate cash flow from operations
(see Section 4.3.1) are based on the assumption that any changes in a
non cash current account (ie, items classed as current assets or
current liabilities in the balance sheet) relates to an income or
expense already reflected in the P&L account (e.g, increase in debtors
is income not reflected in cash) . Drtina & Largay (1985) show that it
is not unusual for current accounts to change without any changes in
the current P&L account (e.g reclassification of long term liabilities
as current liabilities, inclusion of depreciation in stock under
absorption costing)
162
consistent with the results from the univariate difference
in means test discussed in Section 5.2.1 above.
The short run capital gains variable (CGAIN) is not
significant in explaining the choice of payment method.
When CGAIN is replaced by the shareholding of external
blockholders in the target (TAGLGE), a similar result is
obtained. In the single equation model, TAGLGE had a tstatistic of -0.653 (p-value = 0.257) . One of two
inferences can be drawn from this result. Either capital
gains tax does not affect the choice of payment method or
the proxies we are using are too imprecise to capture the
effect of CGT on the method of payment. The former
interpretation is in line with previous studies (Franks,
Harris & Mayer, 1988; Eckbo & Langhor, 1989; Niden, 1988;
Hayn, 1989) which have not found any significant
relationship between the announcement period abnormal
returns to target shareholders and the potential capital
gains tax payable.
5.3.1. Srnnmary of the determinants of the payment method
To summarise the above results, we found that:
1) the following variables have a significant positive
impact on the proportion of equity in the method of
payment: bidder's gearing (BIDGEAR), the recent return on
the market index (RETMKT), the recent increase in the
bidder's share price (RETBID), growth opportunities in the
bidder (VRBID) and the target (VRTAG);
2) the following variables have a significant negative
163
impact on the proportion of equity in the method of
payment: relative size of the bidder to the target
(RELSIZE), percentage of the bidder's shares held by
external blockholders (BIDLGE), bidder's free cash flow
(NPDFLOW) and bidder's liquidity (BIDCASH);
3) there is a non-linear relationship between
directors' shareholding in the bidder and the proportion of
equity in the method of payment. When the proportion of the
bidder's shares held by directors is low or high, then
acquisitions are more likely to be financed by equity.
5.4. Determinants of the accounting policy
Table 5.6 reports the results of estimating the
simultaneous equations model (see equation 2B in Section
4.4) for the accounting method choice with the two stage
Logit regression (2SLM) procedure outlined in Section
4.2.1. For comparison, the single equation logit model
results (SLM) are also presented in Table 5.6. The LogLikelihood ratio shows that both the simultaneous equations
model (2SLM) and the single equation model (SLM) are
significant. The Wu-Hausman statistic is not significant
suggesting that the single equation approach does not
suffer from simultaneity bias.
Both 2SLM and SLM are broadly similar with the same
significant variables. Payment method has a decisive impact
on accounting method choice with higher equity offers being
associated with merger accounting. This is consistent with
the 9O equity threshold required by SSAP 23 f or a bidder
164
Table 5.6.
Maximum likelihood estimates of 2 group Logit regressions
discriminating between the "merger group" and the
"acquisition group".
The "merger group" represents bids where the bidder has
used merger accounting and the "acquisition group"
represents bids where the bidder used acquisition
accounting. The dependent variable (ACCMET) is a dummy
variable equal to 1, if the bidder has used merger
accounting. The explanatory variables are defined in Table
4.1. The sample size is 47l, with 49 and 422 observations
in the "merger group" and "acquisition group" respectively.
Simuftaneous Equation
Model (2SLM)
Single Equation Model
(SLM)
Variable
_________________
Expected
Sign
Coefficient
Estimate
t-statistic
Coefficient
Estimate
t-statistic
INTERCEPT
?
-3.60'"
-2.79
-6.43'"
-3.23
PM
+
4.68"'
3.19
7.82"'
3.98
GWJLL
+
0.10'
1.63
0.08'
1.44
DISRES
+
0.30
0.55
0.17
0.28
PAYOUT
+
0.09
0.20
-0.37
-0.65
DEPFA
?
-0.21
-0.65
-0.37
-1.09
PROFIT
+
0.07
0.94
0.10
1.15
LOG (BIDSIZE)
-
-0.34"
-2.76
-0.41'"
-3.24
BIDGEAR
+
-1.98'
-1.50
-2.65"
-1.82
McFadden's R216.39%
29.21%
Log-Likelihood
-131 .48"
-111 .32"
Wu-Hausman
N/A
1 .33
Notes:
' Significant at 1 %, 5%, 10% respectively, one tail test.
1)
2) 2SLM = Two stage Logit model.
Single equation Logit model.
3) SLM
4) Reduction in the sample size is due to missing observations on
some of the independent variables.
165
to qualify for merger accounting.
The negative and significant coefficient on bidder's
size (BIDSIZE) is consistent with the Watts & Zimmerman
(1978) argument that large bidders would prefer to use
acquisition accounting to avoid the political costs of
increased regulation associated with reporting large profit
figures. To examine the impact of the size of the target on
the choice of accounting policy, we re-estimated the single
equation model (SLM) using the natural logarithm of target
size (TAGSIZE) as a proxy for the political cost of
increased anti-trust regulation. LOG (TAGSIZE) had a
negative coefficient with a t-statistic of -0.444 (p-value
= 0.329) . This is consistent with the univariate difference
in means test which showed that due to the smaller size of
targets relative to bidders, exposure to increased antitrust regulation was greater in bids involving large
bidders than in bids involving large targets.
Surprisingly we find a significant negative, but
relatively weak, relationship between the bidder's gearing
(BIDGEAR) and the use of merger accounting. This result
directly contradicts a significant body of evidence which
has tended to find a positive relationship between the
level of gearing and the choice of income increasing
accounting policies (see Section 3.3.7) . All of these
studies which have examined the relationship between
gearing and the choice of accounting policy are based on
American data. We are not aware of any UK studies which
have examined this issue.
166
It may be that the institutional environment in the UK
restricts the discretion of managers in highly geared firms
to choose income increasing accounting policies. A possible
explanation for this result could be that there is a
greater level of monitoring of highly geared bidders by the
external creditors, as a consequence of which managerial
discretion in choosing income increasing accounting
policies is constrained. The pressure from external
creditors for the firm to choose conservative accounting
policies may explain our results of a negative relationship
between the level of gearing and the choice of merger
accounting. This issue clearly calls for further
examination.
The positive and significant impact of size of
goodwill on choice of merger accounting is as anticipated.
Although the relative weakness of the impact (significant
only at the lO level) is again consistent with the
argument that merger accounting is not the only method of
avoiding the adverse impact of goodwill write off.
Other variables such as access to target's preacquisition reserves (DISRES), potentially high
depreciation charge due to a high level of depreciable
assets in the target (DEPFA) or the target's relatively
greater profitability (PROFIT) do not influence the
bidder's choice of accounting method.
167
5.4.1. Classificatory accuracy of the Logit discriminant
model
A separate issue in any analysis using discriminant
methodology is the ability of the model to classify
observations accurately into the relevant groups. To asses
the classificatory efficiency of the model we use the
proportional chance (PC) model as the bench mark (Joy and
Tollef son, 1975) . The PC model assigns observations to
groups with prior probabilities equal to group frequencies.
The bench mark classificatory accuracy (i.e, the rate
of correct classification) of the PC model is given by
= l
where
Q1 =
is the prior probability of group i membership
(this is equivalent to the proportion of the sample
belonging to group i).
The statistical significance of the classificatory
power of the model may be tested by the t-statistic
RdjSC -
=
R (1 - R )
n
N
where n = total sample size and R thSC = the proportion of the
sample correctly classified by the model.
The classificatory power of the model discriminating
168
Table 5.7.
Classification matrix for the Logit regressions presented
in Table 5.6.
Predicted group is merger accounting if Prob [ACCMET = 11
> 0.12.
Panel A: Classification matrix for the simultaneous equations model.
PredictedGroup ________________
Actual Group
_________________________
Acquisition
Accounting
Merger
Accounting
Total
________________
Acquisition Accounting
283
1 39
422
Merger Accounting
8
41
49
Total
291
180
471
Proportion correctly classified
68.79%
Bench mark based on the proportional chance
model
81 .36%
t-statistic
-5.42"
Panel B: Classification matrix for the single equation model.
PredictedGroup ________________
Actual Group
_________________________
Acquisition
Accounting
Merger
Accounting
Total
________________
Acquisition Accounting
299
T23
422
Merger Accounting
5
44
49
Total
304
1 67
471
Proportion correctly classified
72.82%
Bench mark based on the proportional chance
model
81 .36%
t-statistic
-4.77"
Notes:
1)
Significant at 1 %, 5% and 10% levels respectively.
2) The probability that an observation is classified as merger
accounting is set at 10%, to reflect the actual proportion of merger
accounting cases in the sample (49/471). This is used as an estimate
of merger accounting cases in the population
169
between the choice of merger and acquisition accounting
(See Table 5.7) is very weak. Both the simultaneous
equations model and the single equation model have correct
classification rates which were significantly lower than
the bench mark specified by the PC model. This results from
the high misclassification of acquisition accounting cases
as merger accounting. The high misclassification of
acquisition accounting cases reflects the fact about 35 of
bidders using acquisition accounting actually qualified to
use merger accounting (see Table 53)12 The high incidence
of bidders using acquisition accounting when they are
qualified to use merger accounting reflects , the fact that
in the UK the ability of bidders to combine elimination of
goodwill against reserves with merger relief has
sidestepped the classic merger versus acquisition
accounting choice (see Section 5.2.3)
The bidder's choice between merger or acquisition
accounting appears to based on a range of more subtle
issues not fully captured by our explanatory variables.
5.4.2. Results based on a truncated sample
Following some earlier studies (e.g. Higson, 1990b),
we also examine the choice of accounting policy while
holding the method of payment effect constant. We compare
bids in which the bidder used merger accounting with bids
where the bidder used acquisition accounting, but we focus
12 Table 5.3 and 5.7 have different sample sizes due to missing
observations on the independent variables.
170
on the subsample of bids in which the consideration
includes 9O or more of equity. The single equation Logit
regression results for this subsample of 199 takeovers are
shown in Table
5.8's.
These results are consistent with the SLM results in
Table
5.6. Large bidders use acquisition accounting to
avoid the political costs of large reported earnings. The
unexpected negative relationship between the use of merger
accounting and the level of the bidder's gearing is also
maintained.
Relative profitability is now a significant variable.
The univariate difference in means test (Table 5.2 and 5.4)
showed that targets were more profitable than bidders in
the merger accounting group. However the difference is not
significant using the parametric T-test, but significant
with the non-parametric MWW test. As this result is not
observed across the full sample (see Table
5.6) it is not
robust to sampling variation.
The sample used in Table
5.8 is directly comparable to
Higson's (l990b) . Consistent with Higson's results we find
that size and profitability are significant in
discriminating between users of merger and acquisition
accounting. Higson
did not test for the effect of gearing
on the choice of accounting method.
The classificatory power of the model in Table 5.8
(see Table 5.9) is better than that of the model in Table
'3mis subsample corresponds to cases 1 to 3 in Table 5.3. Missing
observations for some of the independent variables has reduced the
sample size from 211 to 199.
171
Table 5.8.
Maximum likelihood estimates of 2 group Logit regressions
discriminating between the "merger group" and the
"qualified acquisition accounting group".
The "merger group" represents bids where the bidder has
used merger accounting and the "qualified acquisition
accounting group" represents bids where the bidder was
qualified to use merger accounting but instead used
acquisition accounting. The dependent variable is a dummy
variable equal to 1, if the bidder has used merger
accounting. The sample size is 1992, with 49 and 150
observation in the "merger group 1' and the "qualified
acquisition accounting" group respectively. The independent
variables are defined in Table 4.1.
Variable
Expected Coefficient t-statistics
Sign
Estimate
_______________
____________________
INTERCEPT
?
l.56
1.73
GWILL
+
0.08
1.36
DISRES
+
0.14
0.24
PAYOUT
+
-0.79
-1.19
DEPFA
-0.53k
0.23**
-1.46
PROFIT
?
+
LOG (BIDSIZE)
-
_0.37***
-2.85
BIDGEAR
+
_3.08**
-1.92
1.81
McFadden's R2l2.06
Log-Likelihood
-97.672
Notes: m, S. * .
1) Significant at l, 5, l0 respectively, one
tail test.
2) Reduction in the sample size is due to missing
observations on some of the independent variables.
172
Table 5.9.
Classification matrix for the Logit regression presented in
Table 5.8.
Predicted group is merger accounting if Prob [ACCMET=l1 >
0 . 252
___________________
Actual Group
___________________
Predicted Group ___________
Acquisition
Accounting
Merger
Total
Accounting ___________
Acquisition
104
46
150
Accounting_______________ _____________ ____________
Merger
13
36
49
Accounting______________ ____________ ___________
Total
117
199
82
Proportion correctly classified
70.35%
Bench mark based on the
proportional chance model
62.88%
t-statistic
2.181
Notes:
Significant at 1%, 5%, 10% respectively, one
1)
tail test.
2) The probability that an observation is classified
as merger accounting is set at 25%, to reflect the
actual proportion of merger accounting cases in the
sample (49/199) . This is used as an estimate of merger
accounting cases in the population.
5.6 (see Table 5.7) and the bench mark specified by the
proportional chance model. Since bidders using acquisition
accounting may not be qualified to use merger accounting
the discrimination in Table 5.6 is broader than that in
Table 5.8'. The better classificatory power of the model
in Table 5.8 may reflect the reduction of noise in the
data.
14 Table 5.6 compares bidders using acquisition accounting with
those using merger accounting while Table 5.8 compares bidders using
acquisition accounting who qualified to use merger accounting with
bidders using merger accounting.
173
5.5. Robustness checks
The simultaneous equations model results in Tables 5.5
and 5.6 are based on a two stage Two Limit Tobit and Logit
methodology where one of the endogenous variables is doubly
censored and the other is dichotomous. The t-statistics
reported for this model may be inefficient, since they are
derived using an inefficient asymptotic covariance matrix
(see Section 4.2.2) . To assess the severity of this
problem, we dichotomise the doubly censored endogenous
payment method variable PM by restricting our sample to
only those observations where the method of payment is "all
cash" or "all equity". The simultaneous equations model is
then re-estimated using a two stage Logit methodology
(2SLM)
In Table 5.10 we report the results of the 2SLM model
for payment method choice (PM) as the dependent variable.
The corresponding results for accounting method (ACCMET) as
the dependent variable are provided in Table 5.11. In both
cases the single equation model (SLM) results are also
provided for comparison. The results obtained are similar
to those based on the two stage Two Limit Tobit and Logit
methodology (Tables 5.5 and 5.6) . This indicates that the
problem of inefficient t-statistics has no material impact
on our earlier conclusions.
There is some evidence on the importance of merger
reserve in determining the method of payment. While GWILL
has a significant and positive impact on the equity
proportion of the consideration (Table 5.10), it has no
174
Table 5.10.
Maximum likelihood estimates of 2 group Logit model of the
choice of the payment method based on the reduced sample of
observations either "all equity" or "all cash".
The dependent variable is a dummy variable equal to 1 if
the method is "all equity" and 0 if the method of payment
is "all cash". The explanatory variables are defined in
Table 4.1. The sample size is 252, with 84 and 168
observations in the "all cash" and "all equity" groups
_______________________________________
respectively.
______________ __________
Simultaneous Equation
Model_(2SLM)
Single Equation Model
(SLM)
Variable
________________
Expected
Sign
Coefficient
Estimate
t-statistic
Coefficient
Estimate
t-statistic
INTERCEPT
?
0.96
1.19
0.65
0.85
ACCMET
+
0.11
0.18
2.69"
2.41
GWILL
+
0.64"
2.35
0.54"
2.18
BIDGEAR
+
4.76m
2.85
4.81"
2.79
TAGGEAR
-
-1.24
-1.00
-0.67
-0.53
LOG (RELSIZE)
-
-0.51"
-3.21
-0.47"
-3.15
BIDDIROTOO1+
14.15"
2.06
14.76"
2.07
BIDDIRO , TOO25-
-8.30
-1.36
-10.63"
-1.72
BIDDIR0ver025
+
-2.79
-1.04
-1.74
-0.66
BIDLGE
-
-3.57"
-2.83
-4.15"
-3.13
RETMKT
+
3.95"
2.12
3.75"
2.00
LOG (VRBID)
+
0.09
0.52
0.05
0.32
NPDFLOW
-
-3.14
-1.18
-4.11
-1.45
BIDCASH
-
-2.55"
-3.54
-2.41"
-3.46
TAGCASH
-
0.29
0.44
0.27
0.43
HOSTILE
-
0.07
0.15
0.18
0.42
CGA1N
+
0.33
1.01
0.36
1.12
McFadden's R231
.96%
35.22%
Log-Likelihood
-109. 14"
-103.91"
Wu-Hausman
N/A
-0.08
Notes:
1) "'"' Significant at 1%, 5%, 10% respectively, one tail test.
2) 2SLM = Two stage Logit model.
3) SLM = Single equation Logit model.
175
Table 5.11.
Maximum likelihood estimates of 2 group Logit regressions
discriminating between the "merger group" and the
"acquisition group" based on the reduced sample of
observations either "all equity" or "all cash".
The "merger group" represents bids where the bidder has
used merger accounting and the "acquisition group"
represents bids where the bidder used acquisition
accounting. The dependent variable (ACCMET) is a dummy
variable equal to 1, if the bidder has used merger
accounting. The independent variables are defined in Table
4.1. The sample size is 252, with 42 and 210 observations
in the "merger group" and the "acquisition group"
respectively.
Simultaneous Equation
____________ _________
Model (2SLM)
Single Equation Model
(SLM)
Variable
______________
Expected
Sign
Coefficient
Estimate
t-statistic
Coefficient
Estimate
t-statistic
INTERCEPT
?
-0.52
-0.52
-1.72
-1.28
PM
+
1 .86"
3.09
3.22"
3.06
GWILL
+
0.09
1.24
0.08
0.99
DISRES
+
-0.02
-0.04
0.04
0.06
PAYOUT
+
-0.58
-0.94
-0.62
-0.92
DEPFA
?
-0.73"
-1.83
-0.71"
-1.79
PROFIT
+
0.26"
2.18
0.28"
2.14
LOG
( BIDSIZE )
-2.28
-2.62
-0.31"
-0.38"
_________ ___________ _________ ___________ _________
BIDGEAR
+
-2.90"
-1.79
-1.55
-2.53"
McFadden's R218.39%
22.82%
Log-Likelihood
-92.66'"
-87.63"'
Wu-Hausman
N/A
1.47
Notes:
1) "' " ' Significant at 1 %, 5%, 10% respectively, one tail test.
2) 2SLM = Two stage Logit model.
3) SLM = Single equation Logit model.
176
significant impact on the choice of accounting method
(Table 5.11) . This indicates that once again merger relief
and merger reserve provide enough flexibility in the
treatment of goodwill that merger accounting holds no
particular advantage over acquisition accounting even when
the potential goodwill is large.
The positive impact of goodwill on payment method
choice in Table 5.10 appears to be concentrated at the tail
ends of the sample distribution (ie, when payment is 100%
cash or lOOt equity). Across the whole sample goodwill does
not have a significant impact on the payment method (See
Table 5.5), while for the sub-samples of pure equity offers
and pure cash offers goodwill has a dramatic impact.
With the truncated sample, relative profitability of
the target (PROFIT) has a significant and positive impact
on the choice of merger accounting while the proportion of
depreciable assets in the target's net worth (DEPFA) has a
significant and negative impact on the accounting policy
choice (Table 5.11). These results are consistent with our
prior expectations but are not observed with the full
sample (see Table 5.6) and, hence, not quite robust to
sampling variations.
About 30's of the takeovers in our sample occurred
after the 1987 stock market crash. Since some of our
earlier results showed that the level of the market index
(see RETMKT in Table 5.5) influenced the choice of payment
method, it is reasonable to expect that the market crash of
1987 would cause a decline in the use of equity. We test
177
for the impact of the stock market crash on the choice of
payment method by re-estimating the models in Table 5.5 and
including a dummy variable for the stock market crash'5.
CRA.SH had a t-statistic of -3.167 (p-value = 0.001) and
-3.131 (p-value = 0.001) in the simultaneous equations
model and the single equation model respectively. This
confirms that the stock market crash of 1987 had an adverse
impact on the use of equity offers in takeovers.
5.6. Conclusion
In this chapter, we present the results of estimating
a simultaneous equations model in which the method of
payment and the choice of accounting policy are jointly
determined. In order to ensure the robustness of our
results and to enhance comparability with previous studies,
we also repeat our analysis with single equation models and
a truncated sample in which payment method is restricted to
l00 cash or l006 equity. We find that while the payment
method significantly influences the choice of accounting
method, the reciprocal impact of the latter is not
significant.
There are bid-specific control characteristics which
significantly explain the cross sectional variation in the
method of payment. Use of equity is positively related to
the relative size of the target, the bidder's gearing, the
bidder's market-to-book ratio, the target's market-to-book
' 5 CRASH is dummy variable equal to 1 if the takeover occurred in
1988, 1989 or 1990, otherwise the dummy is equal to 0.
178
ratio, a rising stock market and rising bidder share price
in the pre-bid period. Use of equity is negatively related
to the bidder's liquidity, free cash flow and the
shareholding of large investors.
Managerial concern about the dilution of control
affects the method of payment. Previous studies based on US
data found a linear negative relationship between
managerial shareholding and the probability of an equity
offer. We find a nonlinear relationship between management
shareholding and the use of equity.
Preference f or merger accounting is negatively related
to the size of the bidder. Our results show a negative
relationship between the bidder's gearing and the choice of
merger accounting. This is contrary to the existing
evidence from US studies that highly leveraged firms are
more likely to choose income maximising accounting
policies. We suggest that greater monitoring of highly
geared bidders by creditors in the UK could reduce
managerial discretion in choosing an income augmenting
accounting policy, i.e, merger accounting.
Consistent with previous US studies, goodwill does
have a positive impact on the choice of merger accounting
but very weakly so. We note that, in the US, there is a
clear-cut choice between merger and acquisition accounting
which is blurred in the UK by the ability of bidders to
eliminate goodwill by writing it off against reserves.
Distinction between merger and acquisition accounting is
further eroded by the UK merger relief provisions which
179
make available many of the benefits of merger accounting
even when acquisition accounting is used.
In this chapter, we have concentrated on the method of
payment decision from the bidder's perspective. Our
analysis of the types of payment methods used in takeovers
(see Table 4.2) showed that in 44 of takeovers, bidders
offer target shareholders the choice of accepting cash or
equity. In the next chapter, we investigate the method of
payment decision from the target shareholders' perspective.
By concentrating on those takeovers where the target
shareholders can choose cash or equity, we attempt to gain
an insight into how target shareholders make optimal
decisions about which method of payment to accept from the
bidder.
180
CHAPTER 6
TARGET SHAREHOLDERS' CHOICE BETWEEN CASH AND EQUITY
CONSIDERATION.
6.1. Introduction
On the 13th of August 1987, Christy Hunt Plc (CH)
intervened as a white knight, on behalf of Deritend
Stamping Plc (DS), in an ongoing hostile bid from Carclo
Engineering Plc (CE). Under the terms of the successful
offer from CH, shareholders in DS could either accept 20
shares in CH for every 3 DS or 600p in cash for each DS
share. The cash alternative offer was underwritten by Swiss
Bank Corporation who were the merchant bankers to CH.
Based on the closing share price for CH on 13/8/1987
(i.e. 127p per share), the equity offer was worth 847p per
DS share. When the offer was declared unconditional on
22/9/87, CH had a closing share price of 97p which implied
that the equity offer was worth 647p per DS share'. In this
bid, the equity offer was 41 and 8 higher than the cash
alternative on the announcement and unconditional dates
respectively. Despite the higher premium carried by the
equity offer, only 60.50 of DS shareholders accepted it.
This example raises two very interesting questions:
(1) why did some shareholders accept a cash offer which in
monetary terms was clearly inferior to the equity offer?
(ii) why should the equity and cash offers (which are
effectively offers for the same asset) have different
1 0n 22/9/87 Deritend Stamping had a closing share price of 610p.
181
monetary values?.
This chapter provides answers to these questions.
Offers in which the target shareholders are provided with
the opportunity to accept either cash or equity occur with
greater frequency in the United Kingdom than in the United
States (Franks, Harris & Mayer, 1988). This is partly due
to the differences in the institutional environment in
which takeover bids are conducted in the two countries. The
City Takeover Code can compel the bidder under certain
circumstances to offer cash as a method of payment.
Rule 9 of the City Takeover Code provides that if a
person or persons acting in concert acquire 30% or more of
the voting rights or when they already have more than 30%
acquire
1%2
in a 12 month period, then a full bid must be
made in cash or must include a cash alternative. Rule 11 of
the City code provides that where a person in the last 12
months preceding a general offer purchases 1O% or more of
the voting rights in a company for cash, then a subsequent
general offer must be for cash or must include a cash
alternative.
An illiquid bidder, required by rule 9 or 11 to make
a mandatory cash offer, can side step the problem by making
an equity offer but arrange f or its bankers to provide a
cash alternative. Under the cash alternative, the bidder's
2The relevant percentage during the period (1/1/80 to 31/12/90)
covered by the sample in this study was 2. The change from 2% to 1%
occurred on 3/3/93.
3The relevant percentage during most of the period (1/1/80 to
31/12/90) covered by the sample in this study was 15%. The change from
15% to 10% occurred on 26/6/89.
182
bankers will agree to buy from the target shareholders any
shares received under the offer at a pre-determined price.
Franks et al (1988) suggest that in addition to providing
cash for illiquid bidders the underwriters can act as a
means of signalling the true value of the bidder's equity
by an informed (or at least partially informed)
participant.
A liquid bidder could provide the cash alternative
from its own resources. There are several reasons why a
bidder with sufficient cash resources would choose to make
an offer including both equity and a cash alternative
rather than just cash:1) The equity offer can help the bidder reduce the
outflow of cash from the group resulting from the
acquisition;
-
2) The equity offer eliminates the need to include a
premium to compensate for capital gains tax in the overall
bid premium.
In explaining the choice between equity and cash faced by
a target shareholder, we present two alternative
perspectives on the nature of this decision.
If investors believe that all publicly available
information has been accurately impounded into the bidder's
current share price, then in making the decision whether to
accept the equity offer or the cash alternative, individual
shareholders should only be concerned with the relative
monetary values of the equity offer and the cash
alternative (i.e, the bidder's current share price should
183
form the sole basis of valuing the equity offer) . This is
the efficient market perspective which predicts that
acceptance of either the cash or equity offer will be
primarily influenced by the difference in value between the
two offers. Ceteris paribus, the efficient market argument
suggests that all target shareholders should accept the
offer with the higher monetary value.
Alternatively, investors may believe that the value of
the equity offer cannot be determined at the date of
consummation of the acquisition, since the returns to
shareholders accepting the equity offer is affected by the
post merger profitability of the enlarged group. Therefore,
in addition to the relative monetary values of the equity
and cash offers, variables concerned with the economic
fundamentals of the acquisition should play a part in the
decision models of the target shareholders. This is the
market mispricing perspective which suggests that target
shareholders will try to evaluate the extent to which
either offer reflects the size of post merger gains in
deciding whether to accept the cash or equity offer. The
market mispricing perspective predicts that acquisitions
which are perceived as offering significant post merger
returns will attract higher levels of equity acceptances.
In this chapter we empirically evaluate the
predictions of the efficient market and the market
mispricing arguments. Our results suggest that the
behaviour of individual investors is consistent with a
belief in market efficiency i.e, that capital markets
184
efficiently factor information concerned with the economic
fundamentals of an acquisition into share prices of both
the bidder and the target at the time of announcement of
the bid.
We approach the task of explaining the difference in
value between the cash and equity offers, by noting that
arguments based on risk and capital gains tax (CGT) are not
sufficient to justify any difference in value. Although the
equity offer is riskier than the cash offer, we show that
the higher risk of equity has already been priced by the
market and hence cannot affect the difference in value
between cash and equity. Similarly we show that, while the
cash offer is subject to CGT, an 'equity offer with a cash
alternative" is tax efficient and hence a tax compensating
premium is not required in the cash offer.
Using arguments derived from the options pricing
literature we show that the cash offer has incorporated
within it a put option written by the bidder, while the
equivalent contingent claim against the bidder in the
equity offer is an "option to exchange assets". Our results
suggests that the difference in value between the cash and
equity offers can be explained by the different types of
contingent claims against the bidder which are imbedded in
the two offers.
6.2. Theoretical background
Once the bidder has chosen an equity offer with a cash
alternative as the method of payment, then individual
185
target shareholders face two different decisions:1) is the present offer from the bidder acceptable?
2) if the present offer is acceptable, should I accept
the equity offer or the cash alternative?
In the extant literature the first decision facing the
target shareholders has been examined (Hoffmeister & Dyl,
1981; Walkling, 1985). In this study we focus on the second
decision confronting the target shareholders. This is
perhaps the first study which examines the factors likely
to influence the target shareholderst choice between equity
and cash.
In deciding whether to accept equity or cash, it is
inevitable that a wealth maximising shareholder will
compare the relative monetary values of the two choices. If
there is a difference in the monetary values of the cash
and equity offers then we expect that shareholders will be
inclined to accept the offer with the higher monetary
value.
Should rational shareholders consider any other
factors in making their choice apart from the basic issue
of the difference in value between the cash and equity
offers?. As already outlined there are two alternative
viewpoints on this issue: the market mispricing perspective
and the efficient market perspective.
6.2.1. Market mispricing perspective
The market mispricing perspective suggests that in
deciding whether to accept the equity offer or the cash
186
alternative, shareholders will be motivated by a variety of
considerations, which will include not only the difference
in value between the two choices, but also factors
concerned with the economic fundamentals of the
acquisition.
The starting point for this argument lies in the
nature of the two different methods of payment which are
being offered. The essential difference between the equity
offer and the cash alternative is that of valuation. The
value of the cash alternative is certain and is established
at the time of the acquisition, whereas the value of the
equity offer, which is a risky security, may depend on the
post merger performance of the enlarged group. As Weston,
Chung & Hoag (1990: p.688) observe:
"The key difference between a cash offer and a risky
securities offer is the contingent-pricing effect of
risky securities. When securities whose values are
related to the profitability of the target are
offered, the price of the target is actually
determined ex post. When the profitability of the
target turns Out to be high (low), the value of the
security will also be high (low), implying a higher
(lower) payment to the target than otherwise."
Implicit in the information signalling models,
developed by Fishman (1989), Berkovitch & Narayanan (1990)
and Brown & Ryngaert (1991), is the assumption that the
price paid for the target can only be determined after
completion of the acquisition. These models discuss the
role of the method of payment as a signal of the bidder's
valuation of the target's assets.
In these models, if the method of payment is uniform,
then the bidder cannot credibly communicate any information
187
which it possesses about the value of the target's assets.
Since both high and low value bidders 4 can undertake
acquisitions, it follows that all bidders will be valued as
belonging to the same class.
With heterogenous methods of payment, equity offers
should come from low valuation bidders while cash offers
should come from high valuation bidders. Since in an equity
offer, target shareholders participate in any post merger
profits, then f or any given terms of trade (ie, share
exchange ratio) the bidder finds that the cost of the
acquisition increases with the gains realised from the
merged firm under its management. For this reason an equity
offer is not used by a high valuation bidder who prefers to
make an immediate cash settlement. By this process the
method of payment can act as a signal by which high
valuation bidders can reveal themselves.
If the gains generated in the post merger period
affect the value of the equity offer, then given their
objective of wealth maximisation, target shareholders would
only accept the equity offer where the bidder is expected
to generate significant value increases in the post-merger
period.
The post merger performance of the bidder will be
influenced by the economic rationale underlying the
acquisition, managerial behaviour and agency conflict
(Slusky & Caves, 1991). If shareholders believe that the
4A high value bidder is one who values the target's assets highly,
ie, expects to realise a significant amount of post merger gains.
188
value of the equity offer is not determined until a future
date, then in making a choice between equity and cash at
the present time they will have to consider those
fundamental economic factors which are likely to affect the
future profitability of the group.
6.2.2. Efficient market perspective
The efficient market hypothesis in the semi-strong
form postulates that share prices should rapidly and
accurately reflect all publicly available information. If
this is true, then upon announcement of the terms and
conditions of an acquisition this information should be
impounded into the share price of the participating firms.
In an efficient market, we expect that after the
announcement of the bid terms the share prices of
participating firms in a takeover would very quickly
reflect information about the economic fundamentals of the
bid and any information which is signalled by the presence
or absence of an underwriter. Information impounded in the
bidder's share price affects the value of the equity offer.
If information concerned about the economic fundamentals is
impounded into share prices around the bid announcement
period, then the EMH implies that the values of both equity
and cash offers are determined at the time of announcement
of the bid.
A rational investor therefore has little further use
for information concerning the economic fundamentals of the
bid in deciding whether to accept cash or equity. The
189
investor should simply be concerned with the difference
between the monetary values of the cash and equity offers.
6.2.3. Relative monetary values of cash and equity offers
The second issue addressed in this chapter is whether
the equity and cash offers should have the same value. The
answer to this question is not immediately obvious, since
it raises issues similar to those addressed in the
literature on the irrelevance of dividend policy (see
Brealey & Myers, 1984: Chapter 16)
The value of the cash offer is certain while the value
of the equity offer is subject to future fluctuations in
share prices. It seems reasonable to conclude that the
equity offer is more risky and consequently that the value
of the equity offer should be higher since target
shareholders will demand a premium to compensate for the
higher risk of equity. This is similar to the so called
"bird in the hand" argument in the dividend irrelevance
controversy.
This argument is flawed because it fails to appreciate
the true nature of risk. The risk inherent in the bidder's
share has already been priced by the market. So long as the
bidder's investment opportunity set and future investment
policy are not affected by the decision of target
shareholders to accept the cash or the equity offer, it is
illogical to argue that a target shareholder accepting the
equity offer should be paid an additional compensation for
risk over the risk premium which has already been impounded
190
by the market into the bidder's share price.
Let us imagine a new risk averse investor who wishes
to buy shares in the bidder. This investor gives up cash
which is a safe asset for shares which are risky. In a well
functioning market this new investor will expect the future
returns from the investment to compensate for the higher
level of risk. The risk compensation to a new shareholder
is not demanded from existing shareholders since all
shareholders are bearing "the same level of risk. Target
shareholders who give up the option of accepting cash which
is a safer security are therefore not entitled to a
compensating premium from the existing shareholders.
Target shareholders accepting the equity offer are
indeed trading acceptance of a safe asset (ie, the cash
offer) for an uncertain future. However those target
shareholders who accept the cash offer are not safe because
of the decision they made, but because they have converted
their assets into cash in the bank. The same position could
have been achieved by a shareholder accepting the equity
offer, selling the bidder shares and putting the money in
the bank.
An alternative argument posits that shareholders
accepting the cash offer are subject to capital gains tax
(CGT), hence the cash offer should have a CGT compensating
premium which is not present in equity offers. Again this
argument is faulty, since the "cash or equity" offer is
already constructed to be tax efficient. Shareholders who
are liable to pay CGT on realised gains (if they accept
191
cash) are provided with an equity alternative, as a result
of which they cannot demand a CGT compensating premium from
the bidder.
It is important to realise that CGT is never avoided,
only postponed. All shareholders (except tax exempt
institutions) will eventually have to pay CGT on the
realisation of their assets. An unexpected bid could place
target shareholders in the position of having to realise
assets at an inopportune moment, hence the justification
for demanding a CGT compensating premium. If the bidder
provides an equity alternative then the target shareholders
are restored to the pre-bid position of being able to
choose the appropriate moment when any capital gains are
realised. It would be inappropriate once an equity
alternative is provided, for target shareholders who accept
the cash alternative . to demand a CGT compensating premium.
Since the above discussion shows that risk and CGT are
not sufficient to justify any difference in the monetary
values of the cash and equity offers, we are left with an
intriguing question:- are there any theoretically valid
reasons for the widely observed value difference between an
equity offer and its corresponding cash alternative in the
same bid?
Arguments developed from the options pricing
literature may offer a way of resolving this issue. During
the offer period (ie, the announcement date to the
unconditional date) target shareholders do not only own
equity in the target firm but also implicit options from
192
the bidder.
The wealth effect on target shareholders of a bidder's
offer can be viewed as composed of two separate parts.
There is an initial re-evaluation of the target's
underlying share caused by the new information released
during the bid process 5 . Additionally the offer from the
bidder creates a contingent claim against the bidding firm
during the offer period.
The cash and equity offers have distinct options
attached to them. In a cash offer a put option is created
by which target shareholders can sell their shares to the
bidder at the offer price. In an equity offer the
corresponding claim against the bidder is an option to
exchange two risky assets, ie, target shares for bidder
shares (Margrabe, 1978: p.184). During the offer period,
trade in the target's shares represents a trade in a
complex portfolio, ie, a share with a put option (cash
offer) and a share with an asset exchange option (equity
offer)
6.2.4. Valuation of the options implicit in the cash and
equity offers
The value difference between the cash offer and the
equity offer could be due to the different types of options
embedded in the respective offers. The value of the option
to exchange assets relative to the value of the put option
5 New information could concern the synergies to be realised by the
combination.
193
could explain why the cash and equity offers have different
monetary values. The following example illustrates the
valuation issues involved.
Suppose that company X bids for company Y. The terms
of the offer are either a cash price of C for every share
in Y or x shares in X for every y shares in Y. Provided Y
shareholders have decided to sell they have the option to
either exchange the risky Y shares for risky X shares or
accept the risk free cash offer.
Value of the put option
Implicit in the cash offer is a put option, which
gives Y shareholders the right to sell one share in Y at an
exercise price of C. This put option can be valued using
the Black-Scholes (1973) formula:
=C e
N(a fE - d) - P, N(-d)
(1)
Where
= price of the put option
C = cash offer per target share offered by the bidder
r = daily risk free rate of interest
t = number of days for which the offer remains open
cr
= standard deviation of Yts daily return
= share price of Y excluding the put option
N = cumulative normal density function
d = [log (Pr /C e)
/
°y Jti + 0.5 °y Jt
Unfortunately equation (1) cannot be directly applied
to compute a value for the put option because the price of
194
the underlying target share (P r ) cannot be observed. What is
observed during the offer period is the composite price of
a portfolio of the underlying target share and the put
option. Despite this problem, equation (1) provides an
insight into the factors that affect the value of this put
option i.e, the level of the share price relative to the
exercise price (P r /C), the volatility of the underlying
share (a,,), the time to expiration (t), and the risk free
rate of interest Cr)
p
pp = f (-, O), I t, .r)
(2)
Ehagat et al (1987) examined whether the put option
imbedded in a cash tender offer was valuable. Eased on the
prediction from option pricing theory that a portfolio of
a stock and a put should have a standard deviation and beta
lower than that of the stock itself, they argued that
during the tender period target firms should have lower
betas and standard deviations than in the pre or post
tender period.
For a sample of 295 cash tender offers made over the
period 1962 to 1980, they examined the changes in beta and
standard deviation of the target around the tender period.
The average beta for targets declined from 0.912 in the pre
tender period to 0.330 in the tender period, while the
decline in standard deviation was from 0.025 to 0.021. The
decline in the risk measures from pre tender to the tender
period was statistically significant. Average beta
increased from 0.330 to 0.729 between the tender and post
195
tender period (significant at 5d, while standard deviation
increased from 0.021 to 0.023 (not significant) over the
same intervals. This shows that put options, with
sufficient magnitude to affect the risk of the target, are
incorporated in cash tender offers.
Value of the option to exchange assets
Incorporated in the equity offer is an option to
exchange risky Y shares for risky X shares. The option to
exchange risky assets can be valued using Margrabe's (1978)
model:
=
e
t
[p N(d) - P N(d -
Gx
y
(3)
where
= price of the option to exchange assets
= share price of X
= share price of Y
r = daily risk free rate of interest
t = number of days for which the offer remains open
-
°xIy cr,
1,2
+
2
-
-,
P
ax
°y
= standard deviation of X's daily return
= standard deviation of Y's daily return
p = correlation of daily returns on X and Y
N = cumulative normal density function
d = [log (P/P) / a
t] + 0.5 a
Jt
Similar to the problem encountered in valuing the put
option above, the value of the option to exchange assets
cannot be calculated directly from equation (3) because the
196
price of the underlying target share (P r ) cannot be
observed.
Equation (3) shows that the value of the option to
exchange assets is a function of: the relative share price
of X and Y
(/)1
the volatility of the ratio of the share
price of X and Y (cr), the time to expiration (t) and the
risk free rate of return Cr).
= f
Py
Ox, t,
•j
r)
(4)
Value of the difference between the cash and equity offers
The value of the equity offer E is equivalent to x/y.P,
(i.e, the share exchange ratio multiplied by the share
price of the bidder). The gap between the value of the
equity and cash offers (i.e, E-C) is caused by the
difference in the value of the options embedded in both
offers, (i.e,
PePp) .
A comparison of equations (2) and (4)
shows that6:
E-C=f(c,
o)
y
(5)
We expect that the gap between the equity offer and
the cash offer will be:
(i) positively related to
a,
because the larger is
the volatility of the ratio of the share price of X and Y,
the more valuable is the option to exchange assets and so
6Although Pr /c and P/P should affect the gap between the cash and
equity offers, they are omitted from equation (5) because P cannot be
observed.
197
the more attractive is the equity offer relative to the
cash offer;
(ii) negatively related to o,,, because an increase in
the volatility of the return on Y, increases the value of
the put option, which reduces the gap between the equity
and cash offers.
6.3. Definition of variables
Under the market mispricing perspective of the choice
between equity and cash by target shareholders (see Section
6.2.1) the magnitude of potential post merger gains will
influence the willingness to accept the equity offer. In
the literature the extent of any post merger gains created
is dependent on the economic fundamentals of the
acquisition which is a function of the available sources of
synergy and the agency conflict between managers and
shareholders (Slusky & Caves, 1991)
6.3.1. Variables representing the economic fundamentals of
the acquisition
Below, we discuss and define several variables which
are used as proxies for the economic fundamentals of the
acquisition. Under the efficient market perspective, the
impact of these variables on merger gains would be
impounded into the share prices of the participating firms
when the acquisition is announced, hence would not
influence the choice of payment method accepted by target
shareholders (see Section 6.2.2).
198
Industrial relatedness (RELATE)
Related mergers offer opportunities for realising
synergies from economies of scale and scope (Singh &
Montgomery, 1987; Shelton, 1988), while unrelated mergers
offer opportunities for realising financial synergies
(Lewellen, 1971; Eruner, 1988) . The available empirical
evidence is inconclusive on whether related or unrelated
mergers offer the greatest scope for value creation. Elgers
& Clark (1980), Kusewitt (1985), Lubatkin (1987) and
Limmack & McGregor (1992) all provide evidence of superior
performance in unrelated mergers. Sicherman & Pettway
(1987), Scanlon et al (1989) and Datta et al (1992) provide
evidence of the superiority of related mergers. Westerfield
(1970) , and Seth (1990) find little difference between the
two merger types. In this study we leave unqualified the
direction of influence of relatedness on synergy in an
acquisition and on the target shareholders' choice of cash
or equity.
RELATE is a dummy variable equal to 1 if the bidder
and the target have the same Stock Exchange Industrial
Classification (SEIC) 7 and zero otherwise.
Difference in bidder and target gearing (GEARDIF)
Several authors have argued that there is a debt coinsurance in mergers, since after the merger the creditors
7 The SEI classification was obtained from the Risk Measurement
Service of the London Business School. The SEIC classification is
broadly similar to the 2-digit Standard Industrial classification
(Sudarsanam & Taffler, 1985)
199
of the participating firms now have the asset backing of
both firms. The reduction in the probability of default
will allow the merged firm to increase the level of debt in
its capital structure. Since debt is tax deductible, there
will be a consequent increase in the value of the firm
(Lewellen, 1971; Bruner, 1988) . The scope for financial
synergy in an acquisition is measured by GEARDIF which is
defined as the absolute difference in debt capacity between
the bidder and the target. Debt capacity is measured as the
ratio of total liabilities to total assets. We expect that
GEARDIF will be positively related to the magnitude of
financial synergy available in an acquisition and the
proportion of target shareholders accepting the equity
offer.
Complimentary fit of cash resources and growth
opportunities (MATCASE)
Opportunities for synergies can arise from a mismatch
between the cash resources and growth opportunities of the
participating firms. Where one party to an acquisition is
cash rich but has low growth opportunities and the other
party has high growth opportunities but lacks the cash
resources to exploit the growth options, then a merger can
creatively exploit this mismatch of resources (Palepu,
1986; Bruner, 1988; Myers & Majluf, 1984). The variable
MATCASH which captures this complimentary fit between the
200
bidder and the target is defined as:
(Bidder's expected sales growth rate - Target's expected
sales growth rate) x (Target's liquid assets - Bidder's
liquid assets)
Liquid assets are cash plus marketable securities at
market value at the last balance sheet prior to the
announcement date normalised by net assets. Expected sales
growth is proxied by the 4 year average growth rate
preceding the acquisition. We expect MATCASH to be
positively related to the gains realised in an acquisition
and the proportion of target shareholders accepting the
equity offer.
Relative performance of the bidder to the target (RELPERF)
Several studies have shown that takeover gains are
highest when well managed bidders acquire poorly managed
targets (Lang et al, 1989; Servaes, 1991). To measure the
performance of a firm we use the Valuation Ratio (VR) (see
Section 4.3.1). A large valuation ratio is viewed as an
indication of future growth opportunities and hence
superior performance. To measure the relative performance
of the bidder to the target we use the ratio of the
bidder's to the target's valuation ratio (RELPERF). We
expect that post merger gains and the proportion of target
shareholders accepting the equity offer will be positively
related to RELPERF.
Relative size of the bidder to the target (RELSIZE)
Relative size is used to capture the extent to which
201
the synergistic gains present in an acquisition may be
realised. The larger the relative size of the target, the
more difficult it is for the bidder to integrate it
(Scanlon, Trifts & Pettway, 1989; Hughes, 1989) . RELSIZE is
the ratio of the market value of the bidder's to the
target's equity (see Section 4.3.1). We expect that RELSIZE
will be positively related to the size of post merger gains
and to the proportion of target shareholders accepting the
equity offer.
Director's shareholdixig in. the bidder (BIDDIR)
The greater the level of managerial shareholding the
closer is the alignment between managerial and shareholder
objective functions (Jensen & Meckling, 1976) . However
Demsetz (1983) argues that high levels of managerial
shareholding would result in managerial entrenchment as
managers now control enough votes to defeat any challenges
to their authority. EI]J]JIR is the total of beneficial and
non-beneficial shares held in the company by directors of
the bidder at the acquisition announcement date (see
Section 4.3.1) . In accordance with the empirical evidence
of a curvilinear relationship between directors'
shareholding and the value of the firm (Morck et al, 1988;
McConnell & Servaes, 1990), we use BIDIJIR in both the
linear and quadratic form. At both low and high levels of
managerial shareholding we expect to observe large post
merger gains and high proportions of target shareholders
accepting the equity offer.
202
Presence of a large shareholder (BIDLGE)
The presence of a large shareholder increases the
external monitoring on the bidder's management and hence
reduces the scope for managers to take sub optimal
investment decisions (Shleifer & Vishny, 1986; Pound,
1988). BI]JLGE represents the total of all shareholdings
greater than 5 in the bidder (excluding the directors) as
reported in the last annual accounts before the
announcement of the bid (see Section 4.3.1) . We expect that
BI]JLGE will be positively related to the size of post
merger gains and the proportion of target shareholders
accepting the equity offer.
6.3.2. Variables representing the dynamics of the
acquisition
In addition to any possible influence which post
merger gains and the relative monetary values of the cash
and equity offers may have on the choice of payment method
by target shareholders, there are a variety of factors
concerned with the individual circumstance of each
shareholder and the characteristics of the bid which may
influence their choice.
Capital gains tax is an obvious example. Since the
cash offer is subject to CGT, tax exempt shareholders are
more likely to accept the cash offer than tax paying
shareholders (see Sections 1.2.1 and 2.5). In this section
we discuss some of these dynamic factors which might
influence the choice of payment method by target
203
shareholders.
Capital gains tax (CGAIN)
Shareholders accepting the cash offer are liable to
pay capital gains tax (COT) on the disposal of their shares
while roll over relief is available for the equity offer
(see Section 2.5) . Shareholders, liable to pay CGT on the
realised gains if they receive cash, can instead accept the
equity offer. We expect that the higher the potential COT
liability of target shareholders the higher the proportion
of target shareholders accepting the equity offer. CGAIN is
defined
CGAIN =
Pie-bid Mkt V alue Lowest Mkt V alue
Of The Target - Of The Target
Lowest Mkt V alue Of The Target
The lowest market value of the target over the one year
preceding the announcement of the bid is used as a base
cost for calculating the short run capital gains payable by
target shareholders.
Investor sentiment
(RETMKT & RETBID)
The general sentiment of investors could influence the
choice of accepting the equity offer or the cash
alternative. During a bull market, investor confidence is
likely to be high with a consequent willingness on the part
of investors to increase the equity holding in their
portfolios. In such a bull market it will be easier to
204
persuade target shareholders to accept the equity offer.
RETMKT measures the cumulative return on the market
index during the 80 trading days beginning 120 days before
the announcement of the bid. This variable is expected to
have a high value when investor confidence and hence the
willingness to accept equity are high.
Similarly bidders whose share prices have been rising
in the pre-bid period are likely to be more successful in
persuading target shareholders to accept their equity.
RETBID measures the cumulative unadjusted return on
the bidder's equity during the 80 trading days beginning
120 days before the announcement of the bid.
Presence of an underwriter (TJNWRITE)8
In an equity offer with an underwritten cash
alternative, the underwriters agree that whenever the
target shareholders elect to receive the cash alternative,
they will purchase the bidder's paper at a predetermined
price. The presence of an underwriter can provide a signal
to the market of the value of the bidder's paper and hence
reduce the information asymmetry problem faced by target
shareholders that the equity offered by the bidder is
overvalued (see Section 2.4.2) . UNWRITE is a dummy variable
8 1f dynamic variables are defined strictly as those factors
connected with the individual circumstance of a shareholder which
influences the choice of payment method, then the presence of an
underwriter isn't a dynamic variable. Additionally under the efficient
market perspective any information signalled by the presence of an
underwriter would be instantly impounded into the value of the equity
offer, suggesting that this variable may not influence shareholders in
their choice of payment currency. However since the presence of an
underwriter is not related to the economic fundamentals of the
acquisition, we have chosen to classify it as a dynamic variable.
205
equal to 1 if the cash offer was underwritten by a merchant
bank.
The definition of variables is summarised in Table
6.1.
6.4. Sample
The sample consists of successfully completed UK
takeovers in which the bidder made an equity offer with a
cash alternative and is a sub-set of the sample described
in Section 4.5. Over the sample period (1/1/80 to
31/12/90) , we identified an initial sample of 223 bids in
which the bidder offered "equity with a cash alternative"
as the method of payment (see Table 4.2). In order to study
the choice made by target shareholders with regard to
accepting either the cash or the equity offer, it was
necessary to collect data on the proportion of target
shareholders who accepted the equity offer. This
information is occasionally announced by the bidder in a
press release at the time the offer is declared
unconditional. This data which was collected from the Extel
News Cards was only available for 130 bids.
There is a possibility of a selection bias in the
sample. It is possible that information on the proportion
of target shareholders who accepted the equity offer is
only published in bids where the bidder has managed to
persuade a large proportion of the target shareholders to
accept the equity offer. To examine this possibility the
frequency distribution of the proportion of target
206
Table 6.1.
Choice of payment medium by target shareholders: Definition
of explanatory variables.
Variable
Definition
Variables representing economic fundamentals of the acquisition.
RELATE
Dummy variable = 1, if the bidder and the target have the same SEIC
classification otherwise zero.
GEARDIF
Absolute difference in gearing between the bidder and the target.
MATCASH Difference between the expected sales growth rates of the bidder and
the target multiplied by the difference between the liquidity of the
____________ target and of the bidder'
RELPERF
The bidder's valuation ratio divided by the target's valuation ratio2.
RELSIZE
Market value of the bidder's equity
____________ equity. Market value is at day -41.
BIDDIR
I market value of the target's
Proportion (%) beneficial & non beneficial shares held by the directors
in the bidder at the acquisition announcement date.
BIDLGE
Proportion (%) of all shareholdings greater than 5% in the bidder
____________ (excluding directors' shares) at the acquisition announcement date.
Variables representing bid dynamics
Dummy variable = 1, If the cash alternative offer was underwritten
UNWRITE
____________ by the bidder's merchant bankers otherwise zero.
(Market value of the target at day -41 - lowest market value of the
target over the preceding one year) / lowest market value of the
_____________ target.
CGT
RETMKT
Cumulative return in the market index during the 80 trading days
____________ beginning 1 20 days before the announcement of the bid.
RETBID
Cumulative raw returns earned on the bidder's equity during the 80
trading days beginning 1 20 days before the announcement of the bid.
Notes:
1) Expected sales growth is proxied by the logarithmic growth rate in sales over the five
years preceding the acquisition.
2) Valuation ratio is defined as (market value of equity at day -41 plus book value total
debt) I book value of total assets.
3) Unless specifically mentioned, all accounting data are drawn from the last financial
statement before the acquisition announcement data.
207
Table 6.2.
Frequency distribution of the variable PROEQUI.
PROEQtJI is the proportion of target shareholders who
accepted the equity offer.
Range
Sample
_______________________ Size
o
<
PROEQLJI
^
0.1
Frequency Cumulative
Frequency
___________
7
5.38%-
5.38%-
0.1
<
PROEQUI
^
0.2
10
7.69%-
13.08%-
0.2
<
PROEQUI
^
0.3
10
7.69%-
20.7Th
0.3
<
PROEQUI
^
0.4
l0
7.69%-
28.46%-
0.4
<
PROEQUI
^
0.5
15
11.54%-
40.00%-
0.5
<
PROEQUI
0.6
16
12.31%-
52.31%-
0.6
<
PROEQUI
^
0.7
14
10.77%-
63.08%-
0.7
<
PROEQUI
^
0.8
13
10.00%-
73.08%-
0.8
<
PROEQUI
^
0.9
12
9.23%-
82.31%-
1
23
17.69%-
100.00%-
130
100.00%-
100.00%-
0.9
<
PROEQUI
^
Total
shareholders accepting the equity offer called PROEQUI, is
given in Table 6.2.
The x2 goodness of fit that the number of observations
in each cell is equal had a value of 13.69 (prob-value
=
0.1338) . This shows that there is no evidence in Table 6.2
that the bidders who publish information on the proportion
of target shareholders accepting the equity offer have
higher levels of equity acceptances.
The difference in monetary value between the cash and
equity offer fluctuates with movements in the bidder's
share price. We measure this difference at different dates
during the period from the announcement date to the
unconditional date. The value of the equity offer is
208
calculated as the bidder's share price multiplied by the
share exchange ratio. The value of the cash offer is the
cash price per target share. The percentage difference
between the cash and equity offers is measured as
VA LDIF = V alue of equity offer - V alue of cash offer
V alue of cash offer
The difference is measured at intervals of 10 days from the
announcement date to the unconditional date. The
distribution of the difference between cash offers and
equity offers is given in Table 6.3.
Equity offers are on average 4.60 higher than cash
offers. The difference is 4.40 at the announcement date
and rises to 5.90 at the unconditional date. Between the
announcement date and day +30, the difference is fairly
stable. The increase in the difference between the
announcement and unconditional date could be due to a rise
in the bidder's share price as uncertainty about the
success of the bid is resolved. Table 6.3 shows that in
quite a significant number of bids (over 25) the value of
the cash offer is higher than the equity offer. The t-test
shows that the difference in value between the cash and
equity offers is statistically significant throughout the
offer period.
6.5. Results
6.5.1. Choice of payment method by target shareholders
In this section we report on the variables which
209
I W0)
-1
(j
-I
Q44)
4-IU)OjQ
u-I 0J
o (fl-H
_I_)W
-H
'd
(Li
Li
4)
OJE
.cl w
LI)
(0
0)0
U]U]W j
-I-I j
4-1
4-1
0
4-,
4-,
Co
N CO N
N (0 COO)
00 - LL) CO
-o
U]
a)
4-1
4-I
H
>-H
(0
0
>1
>-
.4-,
CU
(U
(1)
oE
a) 4-1 C]) (1)
.lJ
a)
H 0 -1-)
-H
(Li
>i'd
U
a)
a)
'4-I
('14-I
'rI
a)a)
OO'
0
ci,
0)
0
>.. C +-,
-- CU C (0
O o co
:i
nl(Li
)
b1 -..c
a)
a)
a)
(Li°
c:(Ti
rj E
0)
cv)
U)
a)
LU
CV)
a) (0
N
0)
- CV)
00'- CV) 0)
(00)
0)
00)
C°
cDC
j
-U
(Li
C
(U
U)
0
0
O - co
C') ' - 0)
00': - N
cd0dL9
(0
0)
(0
(V)
C4-'
C (0
00)-a
WE-4'd
C]) --H
C ) H -H
-.--p_la) (Li
(1) -I-H -1
0
-i 0JJ
-H
U) 4-1 H
jJ
•r4
tl-0E
-H
.l-)ti
Li) C])
'd.
Cl) 0 b0
0) (Li -H -I C)
E-'
ci)
>
4-,
0
ci)
a.
Cl)
a)
4.Cl,
0)
>
a)
0
I-
U,
4-.
4-4
-H ()
'0)
4-,
4- LI)
CO '4-
U]
-U
a)
0
-U .-
a)
0 - (0 co:
- CV) O
N (0
00
- (0
9- >'
.4-'
CO C (0
O a co o
-H Jj
a)
4-I 4-)
4_lu-i (Li
00 -I
U) 4
a)U]
H
a)C' U)
a)
CU
4-,
0)
C
0
4—'
cl)C
ci)
'.4-
rdb1d
r1
4-.
Cl)
a)
.4-.
CO
-o co-a
0)COE
0-H
-H
N
,—
0
rj
rti
C)
CO
dd0d
0)
00
Cl]
0)
Cl)
00
(Y)
O
U]
coN
'LL) CO
4-,
U]
0
004
0)4-1
U]
-H U] ,
-H .jJ
H
a) s-lU-I
4-iWO
04-I a) (1)
a)
a)
CO
0)
0)00)
-0)
o
D
o
U]
o
)
00
0)
LC
a)
-U W 4rLi-UuI
N
LQ)
9> 9<-a
4-,Coc
••4J4J
H
(ON
N
- (V) 0)
N
N
00
C) 0
CO
4-.
C
CO
0
'.4C,
(I)
E
Cf
(O
1
p.1.1J
210
E'°I
IA
4-' iii 0)
- ci)
El
I
E
9 9)<II0I9-0
-- 0)IF-0)I >010>
cJ)
I
0(01 (D CIw0IOa)
I
J
significantly affect the choice between equity and cash by
target shareholders.
Impact of the difference in value between cash and equity
on target shareholder choice of payment method
In Table 6.4 we relate the proportion of target
shareholders accepting the equity offer to the difference
in value between the equity offer and the cash offer. In
Panel A, the sample is divided into two groups: bids where
the cash offer is greater than the equity offer and bids
where the equity offer is greater than the cash offer. We
find that 65.6% of target shareholders accept the equity
offer when the equity offer is higher than the cash offer
compared to 37.7 when the cash offer is higher than the
equity offer. This difference is significant at better than
l level. This result is robust to possible deviations from
non-normality, since it holds with the non-parametric MWW
test.
To get another perspective on the data, in Panel B, we
turn the process around. The sample is ranked by the
proportion of target shareholders accepting the equity
offer. The average difference in value between the equity
and cash offers is calculated for each portfolio. This test
shows a consistently positive relationship between the
acceptance of the equity offer and the difference in value
between the equity and cash offers.
211
Table 6.4.
Impact of value difference between cash and equity offers
on target shareholder choice of payment method.
PROEQIJI is the proportion of target shareholders who
accepted the equity offer. VALDIF is the average of the
difference in value between the equity offer and the cash
offer measured at intervals of 10 days from the
announcement date to the unconditional date. Sample size is
130.
Panel A: Proportion of target shareholders accepting
the equity offer in bids where the value of the cash
offer is greater than the equity offer and in bids
where the value of the equity offer is greater than the
cash offer.
_______________________________ Sample Size
PROEQtJI
Cash offer > Equity offer
38
0.377
Equity offer > Cash offer
92
0.656
t-test
5.66
Mann- Whitney- Wilcoxon test
5.065
Panel B: Difference in value between the cash offer and
the equity offer for portfolios ranked by the
proportion of target shareholders accepting the equity
offer.____________
VALDIF
Portfolio ranking of PROEQUI Sample Size
0 to 25%
21
-0.007
25% to 50%
31
0.010
50% to 75%
38
0.056
75% to 100%
40
0.090
F-stat
7.56'
Note:
*1 *
Significant at l,
1)
respectively, one tail test.
•
0
212
5-s,
0
1O
0
levels
Impact of economic fundamentals on target shareholder
choice of payment method
In Table 6.5, we relate the acceptance of the equity
offer by target shareholders to measures of the potential
for realising synergy in the combination. We find that the
proportion of target shareholders accepting equity is not
significantly influenced by any of the variables measuring
the economic fundamentals of the acquisition. This is
consistent with the efficent market hypothesis that the
influence of these variables on the future profitability of
the enlarged group is priced by the market in both cash and
equity offers at the time of the announcement of the bid.
Target shareholders relying on the efficient impounding of
these variables into share prices have chosen to ignore
variables affecting the future profitability of the group
in their decision of whether to accept the cash or equity
offer.
Impact of bid dynamics on the choice of payment method by
target shareholders
In Table 6.6, we analyse the effect of bid dynamics
variables on the proportion of target shareholders
accepting the equity offer. None of the bid dynamics
variables has a significant influence on the choice made by
target shareholders to accept either the equity or the cash
offer.
In Chapter 5, we found that the conditions in the
capital markets (RETMKT and RETBI]J) have a significant
213
4i
4J
Ui
Ui
1
— — — — —
= = = = =
N
aD H I4
U
OH H H0
cl:i
'
Ui
p4
p4
o
o
-H
-----
o
-H
o
CU
cc
cc co
i-n io OLD
H H HO
o
H
0
4-1
4-1
0
ci)
H
(1)
Ui
O
H
-H
__
N
H
C/)
-H
(1) cr 'o
w
H
CO O't N N
G't LO H Co
-
-
H HN H0
('3
U)
U)
CU
ii
>1
H
___
-H
(I)
r .l
ci N IJ U
00 H
(N N 0
0
(I)
cI
0 0 0 0 0
w
p4
Ui
p4
o.
o
'3H
-H
-H
CU
-H
CU
H
in
1
o
E-
0
CUUi
H
LO
000
in
c-
Ui
H
-
O\O
0
H
___
0
rdu)
H
,cH
Ui
rd
Co LU H
N ci 0 H
H H H H
H
('3
0 0 0 0
i'
C-!:'
in
CU
.IJH
,
E
0H
0000
0
w
4JH
U-1
°r1
4
H N Co LO
co Lfl LU N
----OW H H cc 0
N ci ci
t'
_____
H
0
-H
-----
-H
P1H
Ui
4-)
•
LO4-4
LU
CU
HCU
000
E-
'-4
-
4J-1
(Ci
-
b
-H
p4
C/)
*+
*
+
OLOLON4J
rr
iI N
4-)
Q
Q
4-) 4-) 4-)
-Hp4
(1)
Hr-I
4 0
in
N
0
H
N
____ = = = = =
—————
214
4-il
ol
Table 6.6.
Impact of bid dynamics on target shareholder choice of
payment method.
PROEQUI is the proportion of target shareholders who
accepted the equity offer. The independent variables are
defined in Table 6.1.
Portfolio ranking of
No of
PROEQUIobs
UNWRITE
CGT
_______
RETMKT RETBID
________
0 to 25%
21
0.333
0.589
0.037
0.141
25% to 50%
31
0.516
0.463
0.093
0.063
50% to 75%
38
0.553
0.617
0.053
0.134
75% to 100%
40
0.550
0.488
0.077
0.119
F-Stat
130
1.05
0.59
1.59
0.72
Notes:
, Significant at 1%, 5%, and 10% levels respectively,, one tail
test.
positive impact on the decision of bidders to issue equity
(see Sections 5.2.1 and 5.3) . It is therefore surprising
that these variables have no impact on the choice of
payment currency by target shareholders.
Consistent the evidence from chapter 5 that capital
gains tax (CGT) does not influence the bidder's choice of
payment method (see Sections 5.2.1 and 5.3), we find that
the target's choice of payment method is also not affected
by CGT. This result probably reflects the fact that the
largest investors on the London stock market are
institutional investors who are exempt from CGT.
Additionally the existence of legitimate means for reducing
CGT by tax paying investors (see Section 1.2.1) has eroded
the importance of CGT in corporate acquisitions.
There is some weak evidence to support the role of
215
underwriters in signalling the value of the bidder's paper.
Only 33% of bids with the lowest level of equity acceptance
(0 to 25%) had an underwriter, while about 55% of bids with
equity acceptance greater than 25%- had an underwriter. The
t-statist±c comparing the lowest portfolio (0 to 25%-) wit.h
the other three portfolios (25%- to 100%) had a value of
1.80. (prob value 0.042) . This suggests that there is some
information signal given by the presence of an underwriter
but that this signal does "not monotonically increase the
level of equity acceptance.
Multivariate regressions
In Table 6.7, we relate the proportion of target
shareholders accepting the equity offer to the difference
in value between the equity and cash offers, measures of
synergy, agency conflict and bid dynamics variables. Since
the target shareholders face a binary choice of whether to
accept cash or equity, the regressions are performed using
the Logit methodology 9 . We estimate the following
multivariate logistic regression:
Proportion of
Target shareholders
=
accepting the
equity offer
(Difference in V alue between
IIcash and equity offers, Synergy1I
i
I
V ariables, A gency Conflict,
I
Bid Dynamics V ariables
)
9 Each individual shareholder faces a binary choice of whether to
accept cash or equity for each single share held. If the total number
of shares in the target in N and the holders of n shares accept the
equity offer, then the proportion of target shareholders accepting the
equity offer (n/N) is bounded by 0 and 1. As the underlying variable is
binary and not censored, the appropriate methodology is Logit and not
the Two Limit Tobit model. For a fuller description see Greene (1991:
p.666)
216
Table 6.7.
Maximum likelihood estimates of Logit models explaining the
proportion of target shareholders accepting the equity
offer.
Variables
____________
Expected
Sign
Model 1
Model 2
Model 3
Model 4
_____________ ____________ ___________ ____________
INTERCEPT
?
__________ _________
0.55
(0.27)
+
VALDIF
____________ __________
5.87***
0.52
(0.93)
0.08
(0.29)
-0.39
(-0.51)
5.80**
(2.72)
___________ __________
(2.25)
RELATE
___________ __________ ____________
0.10
(0.23)
__________
0.03
(0.07)
GEARDIF
+
___________ __________ ____________
-1.13
(-0.71)
__________
-0.88
(-0.52)
MATCASH
+
___________ __________ ____________
-0.35
(-0.73)
__________
-0.32
(-0.62)
__________
-0.17
(-0.52)
LOG OF
+
RELPERF__________ ____________
-0.12
(-0.39)
LOG OF
+
RELSIZE_________ ___________
-0.04
(-0.28)
_________
0.14
(0.81)
BIDDIR
+
-2.75
___________ __________ ____________ (-0.60)
__________
-2.38
(-0.50)
BIDDIR2___________ __________ ____________
2.69
(0.29)
__________
2.43
(0.26)
BIDLGE
+
_______________ _____________ ________________
1.78
(1.31)
_____________
1.48
(1.09)
UNWRITE
+
__________ _________ ___________ __________
0.29
(0.82)
0.61
(1.23)
CGT
+
__________ ________ __________ _________
0.01
(0.11)
0.02
(0.28)
RETMKT
+
___________ __________ ____________ ___________
0.95
(0.55)
0.36
(0.17)
0.05
0.00
0.23
-72.88
-88.16
-68.907
111
1 30
111
ADJ R20.19
Log- Likelihood
Sample Size 31
-84.39
30
i'Jotes:
1) " "' Significant at 1 %, 5% and 10% levels respectively, one tail test.
2) t-statistics are in brackets
3) Reduction in the sample size is due to missing observations on some of the
independent variables.
217
Consistent with the results in our earlier analysis,
we find that the only variable with any significant
influence on the choice of accepting equity or cash is the
difference in value between the two offers. This is
supports the efficient market perspective that investors
can rely on the market to impound all publicly available
information into security prices and simplify the decision
facing target shareholders to a comparison of the relative
monetary values of the equty offer with the cash offer.
The weak impact which the presence of an underwriter
has on the level of equity acceptance can be seen in model
4. The variable t.TNWRITE would be significant at the ll
level under a one tail t-test.
6.5.2. Difference in value between the cash and equity
offer
In this section we examine the determinants of the
difference in monetary value between the equity and cash
offers. We relate the difference in value between the cash
and equity offers to (i) the volatility of returns for the
target and (ii) the volatility of the ratio of the bidder's
share price to the target's share price (see Section
6.2.4) . In Table 6.8, we report the results of estimating
the following ordinary least squares regression:
Difference in
value
[V ariance of target's return,
Bidder's share price
variance
Target's share price
218
Table 6.8.
Regression of the value difference between the cash and
equity offer on explanatory variables.
The dependent variable is the difference in value between
the equity and cash offers. This is calculated as: {
(equity offer -cash offer) / cash offer}. The value of the
equity offer is calculated as the bidder's share price
multiplied by the share exchange ratio. The value of the
cash offer is the cash price per target share. The
difference is measured on the announcement date. TAGSTD is
the standard deviation of the target's returns during the
pre bid period. RATIOSTD is the standard deviation of the
ratio of the bidder's share price to the target's share
price during the pre bid period. The pre bid period
consists of 250 trading days beginning from 290 days before
the announcement date through to 41 days before. The sample
size is 130.
Variables
Expected Model 1
Model 2
Model 3
Sign __________ __________ ___________
O.057**+
0.013
?
0.016
________ (2.843)
(0.574)
(0.745)
_____________
INTERCEPT
__________
TAGSTD
-
_____________ __________
-0.574
(-0.779)
__________
^
-3.553
(-2.066)
1.070
(1.373)
3.708
(2.168)
1.10
11.87**
ADJ R2-0.003
0.011
0.081
F-STAT
2.427
6.676***
RATIOST]J
_____________ ___________ ___________
Breusch-Pagan Test
ll.024**
0.637
Notes:
' ' * Significant at the l, 5°-i, 10% levels
1)
respectively, one tail test.
2) t-statistics are in brackets.
219
The Breusch-Pagan statistic shows that the regression
residuals are affected by heteroskedasticity (Maddala,
1989: p.164). Hence the standard errors for the regression
coefficients are adjusted for heteroskedasticity using
White's (1978) procedure.
In Models 1 and 2, we find evidence which is
consistent with our prior expectation that the difference
in value between the cash and equity offers is negatively
related to the variance of the target's returns and
positively related to the variance of the ratio of the
bidder's share price to the target's share price. However,
the relationships are not very significant. In Model 3,
when we include TAGSTD and RA.TIOSTIJ in the same regression
we find a statistically significant relationship. This
suggests that the regressions in Models 1 & 2, may be
biased due to omitted variables.
Although, TAGSTD and RATIOST]J are highly correlated'°,
multicollinearity is not a serious problem. Maddala (1989:
Chapter 7) shows that high inter-correlation among
explanatory variables is not necessarily a problem unless
it. results in the regression coefficients having high
standard errors. Maddala suggests that an effective test
for the severity of multicollinearity is to examine the
stability of the estimated coefficients to random deletion
of observations from the sample. We randomly deleted 20
observations from the sample, and re-estimated Model 3 with
the reduced sample. Chow's (1960) predictive test for
10 The correlation co-efficient between TAGSTD and RATIOSTD is 0.7723.
220
stability had an F-statistic of 1.383 (prob-value =
0.1426) . This test rejects at the l0 level the hypothesis
that the estimated coefficients are sensitive to random
deletions of observations from the sample.
These results support our belief that during the offer
period, trades in the target's equity represent trades in
complex portfolios where the shares have put options and
options to exchange assets attached to them. The difference
in value between the cas'h offer and the equity offer
results from the difference in the types of claims against
the bidder which are incorporated into each offer.
6.5. Conclusion
The Efficient Market Hypothesis (EMH) postulates that
all publicly available information about a security can be
reduced into a single index: namely the share price. In
this chapter, we document behaviour by investors which
suggests that market participants behave in a manner
consistent with a belief in the EMH.
We examine the factors likely to influence target
shareholders in their choice of accepting cash or equity
when the bidder offered "equity with a cash alternative" as
the method of payment. Consistent with the theoretical
predictions based on the EMH, we find that the target
shareholders base their choice primarily on the difference
in value between the cash and equity offers. Target
shareholders appear to be ignoring information concerned
with the economic fundamentals of the acquisition and the
221
future profitability of the combination, because they
believe that all relevant information has been accurately
impounded into the share prices of the participating firms
at the time of announcement of the bid and hence
incorporated within the value of the equity offer.
This result suggests that the possession of private
information by the bidder about the value of its equity is
not relevant to the bargaining process between the bidder
and the target on the rne"thod of payment. If a bidder
attempts to exploit its information advantage about any
overvaluation of its equity to the detriment of the target
shareholders, the announcement of an equity offer reveals
such overvaluation to the market. Any revision in the
bidder's share price affects the value of the offer, with
the consequent result that target shareholders can make
efficient accept or reject decisions, without worrying
about the possession of private information by the bidder.
Additionally we find that the difference in value
between the cash and equity offers is negatively related to
the variance of the target's returns and positively related
to the variance of the ratio of the bidder's share price to
the target's share price. This is consistent with the
predictions from option pricing theory that the cash offer
effectively conveys a put option on target shareholders
while the equity offer has attached to it an option to
exchange risky assets.
In this chapter we have shown that, within the same
bid, the equity offer and the cash alternative do not have
222
the same value. One question that follows from this
observation is whether, in different bids the method of
paytnent has a heterogeneous impact on the wealth of
shareholders. In the next chapter we review the empirical
literature on this subject and in Chapter 8 we investigate
the underlying reasons for the documented result that
shareholder wealth is affected by the method of payment.
223
CHAPTER 7
THE IMPACT OF THE METHOD OF PAYMENT ON SHAREHOLDER
WEALTH:- THEORY AND EMPIRICAL EVIDENCE.
7.1 Introduction
There is a considerable body of literature which
examines the impact of takeovers on the share prices of
bidding and target firms. US and UK studies have been quite
consistent in showing that shareholders in target firms
gain large and significant abnormal returns around the
announcement of takeovers. For the US evidence, see Jensen
& Ruback (1983). For the UK evidence, see Franks, Harris &
Mayer (1988), Franks & Harris (1989), Limmack (1991) and
Sudarsanam et al (1993).
The evidence on the returns to shareholders in the
bidding firms is ambiguous. During the period immediately
surrounding the announcement of the bid, US studies such as
Servaes (1991), Franks, Harris & Titman (1991) and Stulz et
al (1990) report small abnormal losses to the bidder's
shareholders, while Loderer & Martin (1990), Hayn (1989),
Mitchell & Lehn (1990) and Franks, Harris & Mayer (1988)
report small abnormal gains. In the UK, Franks & Harris
(1989) report small gains to the bidder, while Limmack
(1991), Firth (1991), Limmack & McGregor (1992) and Higson
& Elliot (1993) report small losses to the bidder. Overall
the evidence suggests that while target shareholders
experience substantial wealth gains in takeovers, the
bidder shareholders at best experience small positive gains
224
and at worst small wealth losses (Jensen & Ruback, 1983)
Because a lot of the early research examining the
share price impact of takeovers was directed at answering
the question of whether takeovers were value creating,
researchers concentrated on the average wealth experience
of all shareholders. The more recent literature is now
attempting to disaggregate the data and examine whether
shareholders have different wealth experiences in different
types of acquisitions.
The theoretical and empirical literature suggests that
the method of payment is one of the main variables which
influence the announcement period abnormal returns. This
chapter discusses the theoretical reasons why the method of
payment is likely to influence shareholder wealth on the
announcement of a merger and reviews the relevant empirical
evidence.
7.2. The impact of the method of payment on shareholder
wealth
In a world characterised by perfect markets, symmetric
information and no taxes, the type of consideration offered
should not have an impact on shareholder wealth. With
perfect markets the share price response to a takeover
should only reflect the expected synergistic gains
resulting from the takeover. Because cash and equity offers
are treated differently for capital gains tax (CGT)
purposes (see Section 2.5) and the method of payment can be
used as a means by which managers can signal any
225
misvaluation in their shares to the market (see Section
2.4) there may be cross sectional differences in the wealth
experience of shareholders consequent on the type of
consideration offered.
The empirical evidence documenting such differences in
wealth experience is summarised in Tables 7.1 and 7.2.
The event study methodology is the main technique used
to measure the impact of takeovers on shareholder wealth
(Brown & Warner, l985). Shareholder wealth changes
resulting from the merger announcement is captured in the
abnormal return measure. Abnormal return is the difference
between the actual return and a control return which
measures what the return would have been in the absence of
a merger (see Appendix 8.1 below). To measure the impact of
an event (ie, the announcement of a takeover) on
shareholder wealth it is necessary to define an event
window. The event window represents the period during which
the impact of the event on the share price of the affected
firm is most concentrated. A short event window may not be
sufficient to capture all the impact of an event on share
prices, while a long event window may introduce noise into
the data by capturing price movements which are not
connected with the event being studied.
The evidence is quite strong that relative to equity
offers, cash offers result in higher returns to both the
bidder and the target. This result is robust to differences
in institutional environment, methodology, sample size and
time period.
226
Table 7.1
Cumulative abnormal returns to target shareholders around
the acquisition announcement date partitioned by the method
of payment.
Sample
Event
Window
Period &
___________________ Country ___________
Result for
Cash Offers
[%1
Result for
Equity Offers
[%]
Wansley, Lane &
1970-78
US
DayO
Yang (1983)
____________________ ____________ ____________
33.45
(102)
{na}
17.47
(87)
{na}
1977-82
Huang & Walkling
Days-i
US
and 0
(1987)
____________________ ____________ ____________
29.30
(101)
{na}
14.40
(32)
{na}
Franks, Harris &
1 955-84
Months -4
US
to +1
Mayer (1988)
__________________ ___________ __________
36.30
(476)
{24.8}
1 5.60
(577)
{14.9}
Franks Harris &
1955-85
Months -4
UK
to +1
Mayer (1988)
________________ _________ _________
30.50
(241)
{li.6}
18.20
(235)
{6.34}
Eckbo & Langohr
1966-82
Weeks -8
France
to +8
(1989)
_________________ __________ __________
28.50
(34)
{6.40}
3.90
(31)
{120}
Peterson & Peterson
1980-86
Days -30
US
to
(1991)
__________________ ___________ completion
10.12
(124)
{21.3}
4.88
(86)
{8.10}
1 972-87
Days 0 to
US
completion
_____________________ _____________ ____________
26.67
(408)
{na}
20.47
(1 80)
{na}
33.78
(156)
{16.9}
22.88
(128)
{12.1}
Study
Servaes (1 991)
Franks, Harris &
Titman (1991)
1 975-84
US
Days -5 to
+5
Notes:
refers to the sample size
{ } refers to the reported t-statistic
na means "Not Available"
227
Table 7.2
Cumulative abnormal returns to bidder shareholders around
the acquisition announcement date partitioned by the method
of payment.
Study
Event
Sample
Period &
Window
___________________ Country _____________
Resu}t for
Cash Offers
[%]
Result for
Equity Offers
[%I
Travlos (1987)
1972-81
Days-i and
US
0
___________________ ____________ ______________
0.24
(100)
(1.11)
-1.47
(60)
{-5.i}
Franks, Harris &
1 955-84
Months -4 to
Mayer (1988)
US
+1
________________ __________ ___________
2.60
(476)
{0.89}
0.60
(577)
{0.61}
Franks, Harris &
1955-85
Months -4 to
+1
UK
Mayer (1988)
________________ __________ ___________
4.30
(241)
{i.98}
1 .80
(235)
{i0.6)
Eckbo & Langohr
1966-82
Weeks -8 to
France
+8
(1989)
________________ ___________ ___________
-0.10
(34)
{0.10}
-3.60
(31)
{-0.70}
1981-83
Days-i and
Amihud, Lev &
Travlos (1 990)
US
0
_______________ __________ ___________
0.44
(83)
(1.04)
-1.19
(37)
{2.14}
1980-86
Days -30 to
Peterson &
Peterson (1991)
completion
US
________________ __________ ___________
0.26
(124)
{0.91)
-1 .01
(86)
(-0.8)
1981-86
Brown & Ryngaert
Days-i and
US
(1991)
0
________________ __________ ___________
-0.06
(166)
{-0.14}
-2.48
(62)
{-3.32}
Franks, Harris &
1975-84
Days -5 to
Titman (1991)
US
+5
________________ __________ ___________
0.83
(156)
{0.68}
-3.15
(128)
{-2.8}
1972-87
Days 0 to
US
completion
___________________ _____________ ______________
3.44
(172)
{na}
-5.86
(142)
{na}
-0.15
(96)
{1.63}
-2.96
(75)
(2.07)
Servaes (1991)
Limmack &
McGregor (1992)
1977-86
UK
Month 0
___________
Notes:
refers to the sample size
( } refers to the reported t-statistic
na means "Not Available"
228
A number of different arguments have been offered to
explain the higher returns to shareholders in cash offers
relative to equity offers. The most commonly cited
explanations include information asymmetry between bidder
managers and target shareholders, compensation for capital
gains tax and the transfer of wealth from shareholders to
bondholders.
7.3. Information asymmetry and shareholder wealth
There are two sources of information asymmetry which
may affect the returns to shareholders. Either the bidder
has superior information about the value of its own assets
or the target has superior information about its assets.
7.3.1. Overvaluation of the bidder and shareholder wealth
The Myers & Majiuf (1984) model predicts that the
bidder's managers will make an equity offer when they
believe that their shares are overvalued (see Section
2.4.2) . If target shareholders suspect that the bidder's
managers will behave in such an opportunistic manner, any
equity offer will be treated as bad news, while a cash
offer will be treated as good news. This information effect
implies that, other things being equal, abnormal returns to
bidders will be higher in cash offers than in equity
offers.
The downward revaluation of the bidder's shares
resulting from an equity offer should not influence the
returns to the target, since target shareholders can demand
229
additional shares from the bidder to compensate for the
fall in the bidder's share price.
The amount of information asymmetry existing between
bidder managers and target shareholders is not homogeneous
across all takeovers because the availability and
reliability of information are company specific. The
difference between the bidder managers' and the target
shareholders' perception of the value of the bidder will be
directly related to the amount of information asymmetry.
For any large firm with a substantial amount of high
quality information in the public domain, the difference
between the market valuation and managerial valuation of
the firm will be small, since managers and investors are
likely to be sharing the same information.
If information asymmetry about the bidder's shares is
high, ceteris paribus, the downward revaluation resulting
from an equity offer and the upward revaluation caused by
a cash offer will also be high. This leads to the following
hypothesis: 'Bidder's information asyiimietry hypothesis, 111': Returns to
bidder shareholders and the amount of information asymmetry
about the value of the bidder existing between bidder
managers and target shareholders will be positively related
in a cash offer and negatively related in an equity offer.
230
7.3.2. Uncertainty about value of the target and
shareholder wealth
Where the target has superior information about the
value of its assets, then the bidder will have to incur
real costs in carrying out a pre-merger appraisal of the
target (ie, the due diligence audit). If despite its premerger appraisal, the bidder is not fully informed about
the true value of the target's assets, the bidder will
prefer to make an equity offer, since the target's
shareholders will now share in the future risks of the
enlarged group. This reduces the bidders' adverse selection
problem (see Section 2.4.1)
If however the pre-merger appraisal establishes that
the target has a high value, then the bidder will prefer to
make a cash offer. The objective of the cash offer is to
capture as much of the takeover gains as possible by not
allowing target shareholders to participate in any post
merger gains (Fishman, 1989; Berkovitch & Narayanan, 1990;
Brown & Ryngaert, 1991). This is similar to the argument
advanced under the market mispricing perspective of the
choice facing target shareholders between equity and cash
(see Section 6.2.1)
As the results in Chapter 6 show, this argument is
flawed since the market discounts information about the
gains realisable in an acquisition into share prices at the
time of announcing the bid. If target shareholders believe
that they are not receiving a fair share of the merger
gains, then they can either demand a higher cash price from
231
the bidder or refuse to sell. Additionally, it is difficult
to believe that the bidder can use the method of payment as
a means of preventing anyone from participating in post
merger gains', since in an open market, target shareholders
can accept the cash offer and invest the proceeds in the
bidder's shares.
In Hansen's (1987) model uncertainty about the value
of the target will result in an equity offer as the bidder
seeks to minimise the valuation risk of discovering, expost, that it has overvalued the target (see Section
2.4.1) . Since the equity offer is a result of the bidder's
uncertainty about the value of the target, it is likely
that the premium offered to the target will be smaller as
result of this uncertainty.
If the bidder's uncertainty about the valuation of the
target's assets is high, this results in an equity offer
and a correspondingly low bid premium. When the bidder's
uncertainty about the valuation of the target's assets is
low this results in a cash offer. However there is no
reason to presume that the accompanying bid premium will be
affected by the reduction in the uncertainty surrounding
the value of the target. This leads to the following
hypothesis:
1 The bidding managers can only hope to prevent the target
shareholders from demanding a share of any merger gains, if all
information about gains realisable from the acquisition is not released
to the market at the time of announcing the bid. However, this
information cannot be selectively and secretly released to just the
bidder's shareholders existing at the date of announcing the bid.
232
'Target's information asymmetry hypothesis, 112': In an
equity offer, returns to the target shareholders are
negatively related to the amount of information asymmetry
between the bidder and target. In a cash offer, information
asymmetry has no effect on the returns to the target
shareholders.
7.4. Taxation and shareholder wealth
In a cash offer the CGT liability of target
shareholders crystallises immediately, while in an equity
offer the CGT liability can be deferred to a later date.
Hence bidders may have to pay a higher acquisition price in
a cash offer relative to an equity offer, in order to
compensate target shareholders for the associated CGT
burden (see Section 2.5).
In the UK the rollover of CGT applies to the equity
component of the total consideration irrespective of the
mix, while in the US at least 50% of the total
consideration must be in the form of equity in order for
the rollover relief to apply (Brown & Ryngaert, 1991). The
higher premium needed to compensate the target's
shareholders for capital gains tax implies that ceteris
paribus the target should experience higher returns in cash
offers than in equity offers, while the bidder should
experience lower returns in cash offers than in equity
offers.
In the US prior to the US Tax Reform Act of 1986
bidders made tax gains at the corporate level which offset
233
the higher premium required to compensate the target
shareholders. The principal source of this tax gain is the
step-up in the basis for tax depreciation when the target's
assets have a fair market value which exceeds the
undepreciated balance of those assets for tax purposes in
the target's books.
Under the step-up principle,
in a taxable
acquisition2 , the bidder can step up the target's asset to
the purchase price and depreciate this higher amount,
resulting in a reduction of the future tax liability of the
new group, since capital allowance deductions are higher
than would have been available to the two firms separately.
The tax gains resulting from asset step-up can offset the
higher premium paid to compensate target shareholders for
their CGT liability.
In the US, although CGT compensation implies that the
target would have a higher premium in a cash offer than in
an equity offer, this higher premium has not necessarily
been gained at the bidder's expense, since the asset stepup offsets this higher premium. There is no such offset in
the UK. Any higher premium which is offered to the target
shareholders must be at the expense of the bidder. In the
UK, if CGT were the only market imperfection we would not
2 1n the USA, the tax consequences of an acquisition for the bidder
depends on the tax status of the acquisition for the target's
shareholders. In a taxable acquisition the target's shareholders
realise a gain or loss on their shareholding in the year of the
acquisition. In a tax free acquisition the recognition of the gain or
loss is deferred until a future taxable event occurs. In a taxable
acquisition the bidder can step-up the value of the target's assets.
However step-up is not available in a tax free acquisition. See Hayn
(1989) and Njden (1988) for a summary of the relevant conditions to
qualify as a tax-free acquisition.
234
expect to see any cash offers.
If a CGT compensation premium is present in cash
offers, then this compensation is greater, the larger is
the realisable capital gain. Since equity offers do not
result in the realisation of CGT liability there is no
relationship between the returns to target shareholders and
the potential CGT realisable in an acquisition. This leads
to the following hypothesis:'Capital gains tax hypothesis, H3': In a cash offer, the
returns to the target shareholders are positively related
to realisable capital gains, while the returns to the
bidder shareholders are negatively related to realisable
capital gains. In an equity offer, realisable capital gains
have no impact on the returns to the bidder and target
shareholders.
7.4.1. Empirical evidence on the impact of taxation on
shareholder wealth
As the majority of the evidence in this area comes
from the US, most of the studies examine the impact of CGT
compensation and asset step up on shareholder wealth. The
most common approach is to regress the announcement period
abnormal returns on proxies for the potential CGT payable
by target shareholders 3 and the tax savings available from
asset step up by revaluing the target t s assets.
Carleton et al (1983) (reviewed in Section 2.12) found
3 The difficulties involved in measuring the potential CGT payable
by target shareholders have been discussed in Section 4.3.1.
235
evidence that the market-to-book value ratio of the target
was positively correlated with the probability of an equity
offer. They interpreted this as evidence consistent with
the tax compensation hypothesis on the assumption that the
market-to-book value ratio was a reasonable proxy for the
potential capital gains tax liability of target
shareholders. However, their interpretation of this result
is suspect. The book value of assets is a very weak proxy
for the CGT base cost of target shareholders, hence the
market-to-book value ratio is a very crude empirical proxy
for potential capital gains tax liability.
Auerbach & Reishus (1988) examined the role of tax
benefits as a motivating factor in takeovers. They used a
sample of 318 mergers over the period 1968-83. For each
merger in the sample, a corresponding "pseudomerger" was
created by randomly selecting a "pseudotarget" and a
"pseudoparent" from all firms in the same size class as the
target and the bidder respectively. They used a Logit model
to test whether the tax benefits available in actual
mergers were greater than the potential benefits available
in the control sample of "pseudomergers". None of the
variables in the model directly tested the hypothesis that
cash offers contain a CGT compensating premium. However
several of the variables were proxies for the potential
benefit from the asset step-up. They found that the tax
benefits from asset step-up was on average l.8 of the
target's value in actual mergers and 2.l in pseudomergers.
The difference between the two samples was insignificant.
236
In the multivariate Logit models, the asset step-up
variable was insignificant.
Auerbach & Reishus had to make a number of assumptions
in order to estimate the tax benefits asset from step-up.
Their estimates are likely to be subject to substantial
error which could understate the potential tax benefits in
cases where the target's depreciable assets have been held
for many years or where the assets have appreciated in
value at a rate exceeding the inflation rate. It is
possible that estimation errors could explain the low
explanatory power of the asset step up variable. Hayn
(1989) who uses a more reliable method to estimate asset
step-up benefits found more positive results4.
Hayn (1989) examined whether the announcement period
abnormal returns to both the bidder and the target was
affected by the tax consequences of the acquisition (i.e,
the CGT payable by target shareholders and the potential
step up in the target's assets). Her sample consisted of
116 firms involved in taxable acquisitions over the period
1970-85.She regressed the announcement period (days -30 to
completion) abnormal returns to the bidder and target
shareholders on proxies for asset step-up and the potential
4Auerijach & Reishus estimated the tax benefits from basis step-up
as the difference between the market value of the target's depreciable
assets and its cost. They estimated the market value of the assets by
multiplying the cost of the asset by the GNP deflator for the period
between the year of purchase and the year of merger. Hayn estimated the
tax benefits from basis step-up as the difference between the value of
the target's depreciable assets in the bidder's book and their value in
the target's book.
237
CGT liability of target shareholders 5 . She found that asset
step-up was significant and positively related to the
announcement period abnormal return for both the bidder and
the target. Potential capital gains tax was significant and
positive for target firms but negative and insignificant
for bidding firms. The coefficients on the CGT variable for
target and bidding firms were of about the same magnitude
but of opposite signs. Hayn argued that this suggested a
zero-sum effect for capital gains tax on the two groups of
shareholders (i.e, that the CGT compensating premium is
paid to the target out of the bidder's share of the merger
gains) . However, there is some doubt on this interpretation
given that bidders are larger than targets.
Peterson & Peterson (1991) examined the role of taxes,
type of offer and the medium of exchange in explaining
shareholder returns using a sample of 130 mergers over the
period 1980-86. They regressed the announcement period
abnormal returns (days -30 to completion) to the bidder and
target shareholders on proxies for the asset step-up and
CGT payable by target shareholders. To measure CGT payable
by target shareholders, they used an interaction variable
which incorporated the taxability of the offer and the tax
5 The proxy for CGT payable by target shareholders was derived as
the product of four components: (a) the difference between the target's
share price 40 days before the announcement date and the lowest share
price over the preceding six months, which captures short term capital
gains (b) one minus the average turnover ratio of the target's shares
over six months preceding the bid, which captures the likelihood that
a given gain will be recognised as short term. In the US, prior to the
1986 Tax Reform Act, gains accrued over a long holding period were
taxed at the lower capital gains rate, whereas gains realised over a
short period were taxed at the higher rate on ordinary income (c) the
difference between the marginal individual tax rate and the long term
capital gains tax rate (d) the number of outstanding shares. Hayn did
not provide any explanation for the choice of this variable.
238
status of the target's shareholders. This variable equals
0, if the offer was tax exempt (i.e, no CGT is payable) and
equals the percentage holding of non-institutional
shareholders otherwise 6 . Their hypothesis was that the
coefficient on this variable should be positive if the
target's shareholders are compensated f or the tax burden
associated with cash offers. The variable was significantly
positive at the 5 level.
However, it is difficult to accept the conclusion by
Peterson & Peterson that this evidence supports the CGT
compensation hypothesis 7 . We know that most cash offers are
also taxable offers. Hence the use of this interaction
variable to test the CGT compensation hypothesis is
unsatisfactory, since any other theory which predicts a
higher return to cash offers would also be consistent with
a positive coefficient.
Peterson & Peterson proxied the basis step-up with a
variable which equalled 0 for a non taxable acquisition and
the ratio of accumulated depreciation to the depreciation
expense in the target otherwise 8 . This variable had a
significant and positive effect on the target's returns but
6Since institutional shareholders are likely to be tax exempt,
this variable presumably measures the proportion of tax paying
shareholders in taxable acquisitions.
7In equity offer is likely to be a tax free offer, for which this
variable will have a value of zero. A cash offer is likely to be a
taxable offer, for which this variable has a non zero value (i.e,
proportion of shares held by a non-institutional shareholders) . Since
cash offers have a higher return than equity offers, then by its
construction, this variable must have a positive coefficient,
irrespective of how the non zero value is defined.
8 Peterson & Peterson provide no explanations for their choice of
variables.
239
an insignificant and negative sign on the bidder's returns.
It is doubtful whether this variable is capable of
capturing the potential f or asset step-up as we see no
relationship between historical cost depreciation and the
market value of the target's assets. Peterson & Peterson
concluded, contrary to Hayn (1989), that asset step-up had
an insignificant role in the distribution of merger gains.
It is possible that this different conclusion is driven by
the crude nature of the proxy employed by Peterson &
Peterson.
Franks, Harris and Mayer (1988) is the only study to
test the CGT compensation hypothesis for the UK. They
compared the target's bid premium for cash offers relative
to equity offers for the periods 1955-64 and 1965-85. Since
a full capital gains tax was first introduced in 1965, the
bid premia in cash offers should differ from equity offers
only after 1965. Contrary to this they found that the bid
premium was higher in cash offers than in equity offers
both before and after the introduction of CGT.
This empirical evidence does not support the CGT
hypothesis. This isn't surprising since in earlier
discussions (see Sections 1.2.1 and 2.5) we have shown that
the impact of CGT on the wealth of tax paying investors in
the UK, can be mitigated by the annual tax exemption limit,
the availability of indexation allowance, the use of
personal equity plans etc.
240
7.5. Debt co-insurance and shareholder wealth
An acquisition can result in wealth transfers between
shareholders and bondholders (see Section 1.3.3) . Higgins
& Schall (1975) show that where the total income stream 9 of
the merged firms are not perfectly correlated, there will
be some future state in which although one of the merged
firms is unable to pay its debt, the combined firm would be
able to meet its debt obligations. However no future state
can exist where the combined firm is unable to pay its
debts but one of the merged firms can still meet its own
obligations. This dominance condition exists because the
lack of correlation in cash flows results in a greater
probability that a given level of combined debt payments
will be met. Since the combined cash flows and hence the
value of the combined firm has not increased but the value
of debt has increased (due to the reduced risk of
bankruptcy) this implies that the value of equity must have
fallen.
Galai & Masulis (1976) establish the wealth transfer
effect using the Option Pricing Model (OPM). If the
correlation coefficient between the returns of the merging
firms is sufficiently low or negative, the variance of the
merged firm's returns will decrease. From the OPM, reduced
variability reduces the value of the option. Since equity
is an option on the firm's assets, its value will fall with
9 The
income stream to both shareholders and bondholders.
241
a decrease in the variance of the merged firm's returns'0
The correlation of returns of the merging firms can be
either positive or negative, hence the variance of the
combined firm's returns may be less than the variance of
the individual firms or greater than one of the firms and
less than the other firm. Shastri (1982) relaxes the Galai
& Masulis assumption that the variance of the combined
firm's returns is less than the sum of the variances of the
individual firms' returns. He shows that the effect of the
merger on the value of the combined firm's equity can be
ambiguous and depends on how the variance of the combined
firm's returns is related to the variance of returns of the
individual firms. The relationship between the variance of
returns of the merged firm, the variance of returns of the
individual firms and the value of equity is summarised in
Table 7.3.
In case 1, the variance of the combined firm's return
is less than the variance of firm A'S returns but greater
than the variance of firm B's returns. Hence the merger
reduces the risk of firm A's debt and increases the risk of
firm B's debt. The fall in risk increases the value of firm
A's debt while the increase in risk reduces the value of
firm B's debt. The effect of the merger on each individual
firm's equity would depend on the share exchange ratio,
hence the equity of firm C could either increase or
10 Equity is a call option written on the firm's assets with an
exercise price equal to the face value of debt. When the value of the
firm's assets exceed the face value of debt, equity is in the money,
otherwise it is worthless.
242
Table 7.3.
The effect of variance changes on the value of debt and
equity.
Case Variance of
returns
Debt A
Debt B
Equity C
1
OA >0C>aB
>0
<0
>Oor<0
2
oA <aC<aB
<0
>0
>Oor<0
>
> 0
> 0
< 0
< o• c
<0
<0
>0
3
o
4
°A '
°B
B
Notes:
1) A and B are the merging firms, while C is the
combined firm.
2) a A , a B and o represent the variance of returns to
firm A, B and C respectively.
decrease in value. In case 3, the variance of the combined
firm's returns is less than the variance of returns of both
firms A and B. The merger reduces the risk of both firms'
debt. The value of the debt f or both firms would increase
with the result that the value of the combined firm's
equity will fall. This is more easily accomplished in
conglomerate mergers where by definition the returns of the
merging firms are not correlated. In case 4, where the
merger increases the risk of debt for both firms' A and B,
the value of debt in both firms falls and the value of the
combined firm's equity rises.
Any redistribution of wealth from shareholders to
bondholders is more severe in equity offers than in cash
offers (Eger, 1983; Travios, 1987). In a cash offer,
significant resources leave the group hence reducing the
asset backing available for debt and eroding the
opportunities for a transfer of wealth from shareholders to
243
bondholders. Furthermore, the redistribution theory
requires that the cash flows of the two firms must be
combined. With resources leaving the group in a cash offer,
the asset base for generating operating cash flows falls
hence the post merger operating cash flows will change. In
a cash offer the cash flows of the combined firm, cannot be
viewed as a simple addition of the cash flows of the
merging firms.
If wealth redistribution from shareholders to
bondholders, as result of changes in default risk, is
larger in equity offers than in cash offers, this gives
rise to the following hypothesis:'Wealth transfer hypothesis, H4': The returns to both the
target's and the bidder's shareholders will be negatively
related to the change in risk of the merged firm's returns,
however the relationship between the change in the variance
of the merged firm's returns and the returns to the bidder
and target shareholders will be stronger in an equity offer
than in a cash offer.
7.5.1. Empirical evidence on the wealth redistribution
effect
Most of the evidence on the wealth redistribution
effect has been obtained by examining the announcement
period abnormal returns to the publicly quoted bonds of
firms participating in takeovers. If wealth is transferred
from shareholders to bondholders, then bondholders should
experience positive announcement period abnormal returns.
244
Asquith & Kim (1982) examined the returns to
bondholders involved in conglomerate mergers for a sample
of 50 mergers over the period 1960 to 1978. For the
announcement month they found statistically insignificant
positive abnormal returns of 1.0Th (t=l.66) for the entire
sample of bonds (both bidders' and targets' bonds). However
the Cumulative Abnormal Returns (CAR) diminish in the
months following the announcement month, suggesting that
while there may be a positive wealth effect in the
announcement month, it is not sustained. When the sample is
partitioned into bidders and targets similar results are
obtained. In the announcement month, the bidder's
bondholders gain l.08 (t=1.15), while the target's
bondholders gain l.05 (t=l.23). In the month following the
bid announcement, the bidder's bondholders lose -1.06 (t=1.21), while the target's bondholders lose -l.58 (t=0.93)
over the period months +2 to ^5.
If a wealth redistribution effect exists, it should be
larger for those bids where the cash flows of the
participating firms are not correlated. Asquith & Kim
calculated the Spearman rank order correlation coefficients
between bondholders returns and the correlation coefficient
of the returns for each pair of merging firms. For the
entire sample the Spearman correlation coefficient of
-0.0807 (p-value
=
0.328) was insignificant. In this test
Asquith & Kim did not make any adjustment for the relative
sizes of the merging firms and this may partly account for
their insignificant results.
245
Dennis & McConnell (1986) examined the returns to
convertible and non-convertible bondholders for a sample of
132 mergers over the period 1962 to 1980. Over the 13 day
event window (-6 to +6) the target's convertible bonds
earned a CAR of 8.92% (t=2.38) while the non-convertible
bonds earned a CAR of -0.28% (t=-0.29) . The bidder's
convertible bonds earned a CAR of 2.45% (t=l.46), while the
non-convertible bonds earned a CAR of -1.12% (t=-l.11).
These results indicate that contrary to the wealth
redistribution hypothesis, non-convertible bonds appear to
suffer losses around the date of merger announcement while
convertible bonds gain.
Dennis & McConnell attribute the difference in the
wealth experience of convertible and non-convertible bonds
to the difference in their treatment under the terms of the
merger. For the entire sample, convertible bonds were
exchanged f or some combination of equity, bonds and cash,
while non-convertible bonds were merely assumed by the
bidder and the bonds continued to be outstanding on the
same terms as they were prior to the merger. Dennis &
McConnell suggest that where the exchange of security for
a new class involves an "exchange premium" to induce
security holders to participate in the exchange, then
convertible bondholders would be able to participate in
some of the merger gains while non-convertible bondholders
could not.
A deficiency of the Asquith & Kim and Dennis &
McConnell studies is their sample selection. In both
246
studies the selected samples included other types of
consideration apart from equity offers. Since the wealth
redistribution effect is most concerned with equity offers,
non-equity bids could have introduced some noise into the
data and reduced the ability of these studies to detect the
wealth redistribution effect. Eger (1983) and Travios
(1987) address this problem by concentrating on pure equity
exchange offers.
Eger (1983) used a sample of 39 share exchange mergers
over the period 1958 to 1980. She found CAR for acquiring
company bondholders of 0.905%" over the 51 day period (-30
to +20). Over the same interval target bondholders gain
about 3%. Eger attributes the higher target bondholders'
returns to the fact that these bonds had a higher risk
rating and therefore stood to gain more from a reduction in
risk.
Travios (1987) examined the returns to non-convertible
bonds of bidders in 28 mergers (16 cash offers & 12 equity
offers) over the period 1972 to 1981. For the 11 day window
centred on the bid announcement date (-5 to ^5) bondholders
in equity offers had a CAR of -1.99% (t=-1.60), while cash
offers had a CAR. of 1.47% (t=1.94) . The mean difference
between the CAR in cash and equity offers was statistically
significant at the 0.10 level. This is inconsistent with
the argument that wealth redistribution is more severe in
equity offers.
While Eger (1983) found positive returns to the
11 t-statistics for
CAR
were not provided in the paper.
247
bidder's bondholders in equity offers, Travios (1987) found
negative returns. Although these results are inconsistent,
the small sample sizes involved in these studies, limit our
ability to draw general conclusions.
All the above studies have concentrated on the returns
to the bondholders. The wealth redistribution theory has
not been tested by examining the returns to shareholders
(i.e, relating the returns to shareholders to the changes
in risk resulting from th merger).
7.6. Other studies ou the method of payment and shareholder
wealth
Huang & Walkling (1987) and Franks et al (1988) tested
whether the higher abnormal returns earned by target
shareholders in cash offers could be caused by an
underlying relationship between cash offers and some other
bid characteristic.
Huang & Walkling (1987) argued that since most tender
offers and hostile offers are financed by cash, while
mergers and friendly offers are financed by equity
exchanges, the higher abnormal returns associated with cash
of fers could be measuring the premium associated with
either payment method, managerial resistance or form of
acquisition (i.e, tender offer or merger). For a sample of
204 acquisitions announced between April 1977 and September
1982, they regressed the target's announcement period
abnormal returns on, the method of payment, form of
acquisition and the attitude of the target's management
248
(i.e, hostile or friendly)
In separate regressions (ie, each variable considered
separately) they found that cash offers and tender offers
had significantly higher returns than equity offers and
mergers respectively. Resisted offers had insignificantly
higher returns than friendly offers. However, the
significant difference in abnormal returns between tender
offers and mergers disappears when the payment method and
the type of offer are included in the same multivariate
regression. This suggested that the higher returns observed
in tenders offers was partly attributable to the effects of
the method of payment (in the sample 8O of tender offers
used cash as the method of payment)
Franks et al (1988) performed a similar test to Huang
& Walkling. For a sample of 1,555 U'S acquisitions announced
over the period 1955 to 1984, they regressed the target's
announcement period abnormal returns on, the method of
payment, form of acquisition, the attitude of target
managers and bid revision (i.e, whether the initial offer
was revised or not) . They found that managerial resistance
and bid revision did not significantly influence returns to
the target. However, contrary to Huang & Walkling, Franks
et al found that returns to tender offers were still higher
than in mergers after controlling for the method of
payment.
A robust result established by Huang & Walkling and
Franks et al is that the higher returns to cash offers is
not related to either managerial resistance or the type of
249
takeover attempt (ie, tender offer or merger)
7.7. Conclusion
There is substantial empirical evidence that
shareholders of both the acquiring and acquired firms gain
higher returns in cash offers relative to equity offers
(see Table 7.1 and 7.2). However most of this evidence has
originated in the United States. There is very little UK
based evidence on the impact of the payment currency on
shareholder wealth. The only notable UK study is Franks et
al (1988), but it mainly documents the returns to bidder
and target shareholders given different methods of payment
and provides no insights into why the method of payment
influences shareholder wealth'2.
In this chapter, we discuss some of the theoretical
reasons why returns in cash offers should be higher for the
shareholders of both participating firms. The most commonly
cited explanations include information asymmetry between
managers and shareholders, compensation for capital gains
tax and transfer of wealth from shareholders to
bondholders. There is a very limited and inconclusive
empirical literature that tests these explanations which
have been advanced for the higher returns in cash offers.
Hayn (1989) and Franks et al (1988) examined the
capital gains tax compensation hypothesis. While Hayn found
12 While Franks et al (1988) show that capital gains tax cannot
explain the higher returns in cash offers they do not test any of the
other competing theories (ie, wealth transfer and information
asymmetry) or offer any other explanation for their result that returns
to bidder and target shareholders are higher in cash offers.
250
evidence that CGT affected the returns to target
shareholders in the US, Franks et al found that in the UK
CGT did not affect the returns to target shareholders.
With regard to the wealth transfer hypothesis, Eger
(1983) found that bondholders in both the bidder and the
target gained positive returns around the announcement of
a merger, while Travlos (1987) found that bondholder
returns around the announcement date were negative. There
is no UK evidence on the wealth transfer hypothesis.
In the next chapter, we investigate why the payment
method influences the returns to the shareholders of the
participating firms. Although our study has similarities
with some of the existing studies it has a much wider
conceptual scope.
1) By using suitable proxies for information asymmetry
between bidder managers and target shareholders, we test
the role of information asymmetry in explaining the higher
returns to cash offer. This issue has not been examined in
the literature.
2)
We test the wealth transfer hypothesis by
regressing the returns to bidder and target shareholders on
the post merger change in the risk of the combined firm. In
this way we can examine whether changes in risk dreates any
wealth transfer between shareholders and bondholders. The
relationship between changes in risk and the wealth of
shareholders has not been examined in the literature. This
is the first UK based study that tests the wealth transfer
hypothesis.
251
3) We test the CGT compensation hypothesis by
regressing the returns to bidder and target shareholders on
the potential CGT realisable by target shareholders.
Although there is some US evidence in this area (Hayn,
1989; Peterson & Peterson, 1991) , there is no corresponding
UK evidence. Differences between the US and UK tax
environment (eg, availability of asset step-up in the US)
implies that US evidence may not be broadly relevant to the
UK.
252
CHAPTER 8
DETERMINANTS OF THE DIFFERENCE IN THE BID PREMIUM BETWEEN
CASH OFFERS AND EQUITY OFFERS.
8.1. Introduction
There is significant evidence in the literature on
takeovers that shareholders have different wealth
experiences in different types of takeovers. A well
documented result in the literature is that bidders and
targets gain higher returns in cash offers than equity
offers (see Table 7.1 and 7.2) . However the source of this
gain to cash offers is a research question which has not
received much attention in the literature (see discussion
in Chapter 7).
In this chapter we examine how the method of payment
influences the returns to shareholders of both the bidder
and the target. We test three of the most popular
explanations that have been offered to explain the higher
returns in cash offers: (i) information asymmetry between
managers and shareholders (see Section 7.3), (ii)
compensation for capital gains tax (see Section 7.4) and
(iii) transfer of wealth from shareholders to bondholders
(see Section 7.5). This research goes beyond a replication
of earlier research which has examined the impact of the
method of payment on shareholder wealth, in that we seek to
provide evidence on the above theoretical arguments.
253
8.2. Methodology
We use standard event study methodology (Afshar et al,
1992; Brown and Warner, 1985) as described in Appendix 8.1.
to estimate the impact of takeovers on shareholder wealth.
The abnormal return measure (see Section 7.2) requires the
specification of a control rate of return. Various
alternatives are used to specify the control return':
1)
Market model with a thin trading adjustment
according to IJimson (1979): In this model, the control
return is equal to the predicted value obtained from an
ordinary least squares regression of the firm's return and
the market's return over the 250 trading days ending 40
days before the announcement of the bid;
2) Market adjusted model (assuming zero intercept and
beta equals 1 in the market model): In this model, the
control return is equal to the return on the market index;
3) Mean adjusted model: In this model, the control
return is equal to the average return of the firm over the
250 trading days ending 40 days before the announcement of
the bid.
Abnormal returns are cumulated over the period 40 days
before the announcement of the bid to 40 days after the bid
announcement date. The choice of event period is designed
to capture as much of the effects of the merger as possible
without including too much non-merger noise in the data. A
period like days -40 to +40 is considered adequate by most
1 There is a more detailed description of the control rate of
return in Appendix 8.1.
254
studies (Weston et al, 1990: p.288)
In trying to explain why cash offers generate higher
returns than equity offers we examine the interplay between
capital gains tax, information asymmetry and the debt coinsurance effects. To evaluate the relative importance of
all these factors in one single model, we estimate the
following regression model for bidders and targets
separately :(Capi tal gains tax,
Shareholder weal th
asymmetry,
I Information
gain from the takeover = Debt
co-insurance
)
8.3. Data
The sample2 is identical to the sample described in
Section 4.5.1. In any multivariate regression, the nine
methods of payment identified in Table 4.2 would require
the use of eight dummy variables. This would be an unwieldy
and cumbersome procedure, hence the nine methods of payment
in Table 4.2 have been grouped into four larger groups as
follows:
Cash offers: "All cash" and "Cash or debt";
Equity offers: "All equity";
Cash or equity offers: "Equity with a cash alternative
(underwritten)" and "Equity with a cash alternative (not
underwritten)";
2 Due to error in the market index return data one observation was
dropped from the analysis reducing the sample size to 504.
255
Cash & equity offers: "All cash or (equity plus cash)",
"All equity or (equity plus cash)", "Equity and cash" and
"others".
8.4. Wealth gains surrounding the bid announcement
In this section we report the effects of acquisition
announcements on the returns to bidder and target
shareholders.
8.4.1. Returns to target shareholders
Table 8.1 reports the wealth gains to the target over
different periods centred on the announcement date. The
results for targets are similar irrespective of the method
used in specifying the control rate of return. Over the
whole of the event period (-40 to ^40 days) targets have
significant wealth gains of about 28% irrespective of the
estimation model used, implying that the results are robust
to model specification. Our results are consistent with
those found in other studies for UK targets. Over
comparable event periods, Franks and Harris (1989) reported
abnormal returns to the target of 25.8, Limmack (1991)
reported returns of 3l.38, while Higson and Elliot (1993)
reported returns of 21.70%.
Over the pre-bid period (-40 to -1 days), targets have
significant positive abnormal returns of about 11%.
Positive returns in the pre-bid period suggest that the
market is somehow able to anticipate the bid. Shih & Suk
(1992) investigated whether the stock price runup in tender
256
Table 8.1.
Cumulative abnormal returns to targets over different
windows surrounding bid announcements.
Sample
size
=
Windows
(Days)
-40 to +40
______________
-40 to -1
______________
0
______________
+1 to ^40
_____________
504.
Dimson
Model
0 . 286*1*
a=32.38
b=41.25
c=87.30
0 .1 08*1*
a=l7.35
b=22.35
c=72.81
Market
Adjusted
0.288*1*
a=33.09
b=42.76
c=87.lO
0.111*1*
a=l8.18
b=23.62
c=73.21
0 .1 3 7 *1*
0.138*1*
a=l39.76
a=142.47
b=172.45
b=177.24
c=83.93
c=83.73
0 . 0 411*10.039*1*
a=6.7l
a=6.47
b=9.21
b=9.32
c=58.73
c=58.53
Mean
Adjusted
0.293*1*
a=31.11
b=41.52
c=85.9l
0.115*1*
a=17.29
b=23.13
c=72.82
0.138*1*
a=l31.25
b=169.93
c=81.94
0.041*1*
a=6.31
b=9.21
c=59.13
Notes:
at the 1, 5, l0
levels
* Significant
respectively.
a = t-statistic under the dependence assumption.
b = t-statistic under the independence assumption.
c = percentage of observations that are positive.
offers was a reflection of insider trading. They found that
the price runup for firms displaying insider net buying
activity was not different from those for firms displaying
insider net selling activity.
Niendorf & Huffman (1992) developed a Logit model to
predict the probability that a firm would be the target of
a takeover. The acquisition probabilities estimated by the
model were then used to explain the price reaction around
the announcement date. They found that the probability of
a firm being acquired estimated by their Logit model was
257
negatively correlated to the abnormal returns observed in
the announcement period. This suggests that the market is
able to impound publicly available information about the
probability of a firm being acquired into the share price.
The results from Niendorf & Huffman suggests that the
positive pre-bid abnormal returns are partly a reflection
of the market correctly predicting takeover targets.
Similarly in a study of UK targets, Holland & Hodgkinson
(1994) found that abnormal returns in the pre bid period
were related to the pre-announcement disclosure of bid
related news items.
In the post bid period, targets earn positive abnormal
returns of about 4%'. Statistically significant returns in
the post bid period could be explained by the flow of new
information in the post bid period which resolves any
uncertainty that the bid would be completed.
Figures 8.1-8.3 show the daily abnormal returns to the
target. Under all three models specifying different control
rates of return, we see that most of the valuation effect
of the bid occurs on the announcement date. The abnormal
return to the target shareholders on the announcement date
is around 14%'.
Figure 8.4 shows the cumulative daily abnormal returns
to the target. Consistent with the evidence of preannouncement information leak in Table 8.1, the cumulative
abnormal returns begin to rise about 25 days before the
announcement date. The cumulative abnormal returns continue
to rise until about 25 days after the announcement date.
258
Figure 8.1.
['Tote: Sample size is 504. Day 0 is the bid announcement
date. See text and Appendix 8.1 for a description of the
Dimson model.
01
4-J
(d
H
L
0
-o
fJ
01
—
c
-J
-D
0
E
L
0
-Q
>
(:1
0
- CT) C'.J
tfl
- - - v- -0 0 0 0 0
0
(N
Cfl
01
N CO fl
t 0
0 0 0 0 0 0
0 0
0 0 0 0 0 0 0 0 0
(%J
SU.Jfl -J
259
uiouq
0
0
0
Figure 8.2.
Sample size is 504. Day 0 is the bid announcement
te. See text and Appendix 8.1 for a description of the
rket Adjusted model.
rn
0)
L
(ii
H
-o
0
Ui
L
4-J
C
U)
a)
a)U£
0
:3
C
0
C
C
-o
U)
4)
(I)
4J
0
EL
.IJ
a)>
U)
L
a)L
L
0
(I)
>'
a
>\
C'J
Iii
(fl
C\i
T
N
lfl
Cfl
ci
10
- - - - - - 0 0 0 0 0 0 0 0 0
•
0
•
•
a a 0 0 0 0 0 0 0
o a a a a
C%J su..jn.ej
260
uJouq
0
0
0
Figure 8 .3.
Sample size is 504. Day 0 is the bid announcement
te. See text and Appendix 8.1 for a description of the
an Adjusted model.
En
4J
1)
L
(Ii
H
U)
0
0
En
4-J
C
E
C,
C
:3
0
C
C
ci)
0
L
:2
0
U)
((I
U)
-c
:3
0
D
4:
C
U)
U)
4J
1)
>
U)
E
3)
L
(I)
>'
U)
0
L
0
-Q
0
-
.
o
o
C)
.
c
C'.!
.
0 0
N
a)
a)
(1)
V
cn
r'J
- 0 0 0 0 0 0 0 0 0
0
.
.
.
.
0 0 0 0 0 0 0 0 0
C%J sun.ej
261
uiouq
0
0
0
Figure 8.4.
Sample size is 504. Day 0 is the bid announcement
,te. See text and Appendix 8.1 for a description of the
mson, Market Adjusted and Mean Adjusted models.
cn
a)
-o
H
•0
ci)
-4-J
(I)
L
0
:3
(0 C
0
(11
ci)
C
ci)
£
L
U
C
:3
4—J
a)
I
0
-
Cci)
CD
0
4- 0
0
4_J
E
L
0
4.J
ci)
ci)
> L
(tS
-Q
U)
>'
(00
D
0
a)
>
0
4i
0
DI
E
U
r.j
rn
m
•
o
w
to
i-
C\i
rJ
•
0
0
0
•
0
w
-
c' j rj r.j c' u
0
0
(%) suin;eJ
•
0
cc
-
r,i
-
c—0
0
•
0
0 0 0
wouq
262
cc
C\J
0
0
0 0 0 0
cn
a
This shows that the impact of the bid on the target's share
price is concentrated around the period -25 to +25 days.
With a window of -40 to +40 days we have captured almost
all of the wealth impact of bids on target shareholders.
8.4.2. Returns to target shareholders by the method of
payment
Table 8.2 shows the abnormal returns to targets over
days -40 to +40 partitioned by the method of payment.
Table 8.2.
Cumulative abnormal returns to targets over the period -40
to +40 days by the method of payment.
Dimson
Model
0.356*s*
Pure Cash
a=15.12
83
b=19.75
______________ _______ c=91.57
Market
Mean
Adjusted Adjusted
0.297
Cash or
a=21.68
Equity
b=28.71
_____________ _______ c=86.94
0.272***
Cash and
111
a=15.63
Equity
b=20.0l
______________ _______ c=88.29
0 .211***
Pure Equity
88
a=8.32
b=ll.48
______________ _______ c=82.96
3•49**
F-statistic.
504
0.305'
a=21.21
b=29.20
c=83.78
0.271**1
a=l4.97
b=19.31
c=90.09
Method of
Payment
Sample
Size
222
0.366
a=16.14
b=20.74
c=93.98
0.297*s*
a=21.96
b=29.65
c=86.04
0.277'
a=16.46
b=20.78
c=89.19
0 .202**4
a=7.99
b=1l.76
c=80.68
4.42'
Q375++
a=16.09
b=20.16
c=93.98
0 .2l4
a=8.09
b=ll.72
c=78.41
4.09k
Notes:*+. *s *
.
.
0
0
Significant at . the l, 5-i,
l0 0
levels
respectively.
a = t-statistic under the dependence assumption.
b = t-statistic under the independence assumption.
c = percentage of observations that are positive.
We find that targets in cash offers earn abnormal returns
263
of about 35%- while targets in equity offers earn returns of
about 21%-. This is consistent with the results in Franks et
al (1988) reported in Table 7.1. Additionally our results
of about 30%- and 27%- for "cash or equity" and "cash &
equity" offers respectively are consistent with Franks et
al's results of 28.4%- for "cash or equity" offers and 27.1%for "cash & equity" offers. The F-statistic for the
difference in group means shows that there are significant
differences in the abnoral returns to the different
methods of payment. However, Table 8.2 does not tell us
which group means are statistically different from one
another.
In Table 8.3 we perform a pair wise comparison of the
average abnormal returns to the target shareholders for
each of the different methods of payment. The returns to
equity offers and cash & equity offers are significantly
lower than the returns to cash offers.
"Cash or equity" offers reduce the detrimental tax
consequences of cash offers (see Section 2.5 and 7.4.1) . If
the CGT compensation hypothesis is correct, then the
returns to cash offers should be higher than the returns to
"cash or equity" offers. The difference in returns between
cash offers and cash or equity offers is significant at the
10%- level for the Market Adjusted Model and the Mean
Adjusted Model, but insignificant under the Dimson Model.
As this result appears to be sensitive to the choice of
control model and, where relevant, is only significant at
the 10%- level, we cannot reject the null hypothesis that
264
Table 8.3.
Pairwise comparison of the target's abnormal return over
the period -40 to +40 days for different methods of
payment.
t-statistics are calculated assuming unequal group
variances where the null hypothesis of equal group
variances is rejected at the l0 level. Otherwise tstatistics are calculated assuming equal group variances.
Panel A: Results based on the Dimson Model.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
1.54
1.95*
333"
0.70
2.25"
Cash or
Equity
Cash &
Equity
1.40
Panel B : Results based on the Market Adjusted Model.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
1.77*
2.lO**
3.76"
2.45 **
Cash or
Equity
0.56
1.81*
Cash &
Equity
Panel C: Results based on the Mean Adjusted Model
Cash Offer
Cash or
Equity
Cash &
Equity
1.71'
2.43"
Equity
Offer
359+**
0.90
2.24 **
Cash or
Equity
Cash &
Equity
1.35
Notes:
1) " " * Significant at 1, 5, 10% levels
respectively.
2) Positive t-statistic implies that the mean for the
payment method in the vertical column is higher.
265
there is no difference in the returns between cash offers
and 't cash or equity" offers. This suggests that while there
may be a CGT compensating premium in cash offers, it is not
sufficient to explain the higher returns observed in cash
offers.
Figures 8.5-8.7 partition the target's cumulative
abnormal returns according to the method of payment.
Consistent with Table 8.2, these show that the target bid
premia are distinctly highe"r in cash offers than in equity
offers. Offers which provide the vendors with the
opportunity to accept either cash or equity have lower
returns than cash offers, although as Table 8.3 shows the
statistical significance of the difference is weak.
Abnormal returns to cash & equity offers appear to be an
average of the results for pure cash and pure equity
offers.
Figures 8.5-8.7 show that the method of payment does
not have any effect on the returns to target shareholders
in the pre announcement period. This is understandable,
since we have no reason to believe that investors can
anticipate the method of payment prior to the announcement
of the terms of the offer. Further evidence of the absence
of any method of payment effect in the pre announcement
abnormal returns to target is provided in Tables 8.4 and
8.5, where we examine whether information leak in the pre
announcement period (-40 to -1 days) is related to the
method of payment. The F-statistic in Table 8.4 shows that
there are no significant differences in the average returns
266
Figure 8.5.
Mote: Day 0 is the bid announcement date. See text and
ppendix 8.1 for a description of the Dimson model. Sample
size:- Cash = 83, Equity = 88, Cash or Equity = 222, Cash
and Equity = ill.
4-,
4-J
3
uJ
E
>\
D
C
(0
a
L)
-C
U)
(0
0
>'
(0
•0
4j
-c
C
(1)
-c
0 L
0
4-J
o
.0
0
(1)
_-'\
-Q
0
0
•1-'
U)
E
(I)
L
ci)
(1)
> ci-.
—4-
0
(n
4J0
(ci
(1)4-'
L-
3
U)
(0
H
0
+
L
0
I
U)
L
ci)
I—
'4-
0
£
U)
a
0
•
o
In
f1)
•
0
Lfl
0.1
0
0
(%) SUfl
C'd
In
0
0
In
0
0
0
wAouqQ oA1Inwn3
267
0
In
0
0
0
Figure 8.6.
Note: Day 0 is the bid announcement date. See text and
Appendix 8.1 for a description of the Market Adjusted
model. Sample size:- Cash = 83, Equity = 88, Cash or Equity
222, Cash and Equity = 111.
4)
w
1)
E
>\
(1)
(s
U
0
>'
(1
C
I
-o
4J
-c
0
0 L
_I_J
°
6
c3)
C
0
Q)
I)
>
()
U)
L
ci)
I I-
-4—
Ri
-
L
G)
L
U)
>'w
Ri
0
H
+
L
C
U)
L
ci)
440
-C
U)
0
If)
C
cli
0
Cli
If)
c"J
-
0
0
J
GUAflJ
('Ii
V)
0
0
-
'-
WJouqQ
268
0
If)
0
Cl
Aj)IflWfl3
0
If)
0
C
0
Figure 8.7.
Note: Day 0 is the bid announcement date. See text an
Appendix 8.1 for a description of the Mean Adjusted model.
Sample size:- Cash = 83, Equity = 88, Cash or Equity = 222,
Cash and Equity = 111.
4J
w
E
>
U)
0
1)
a
-o
-o
a
C
0 L
o
£
.0
£
U
-c
4JQ
0
•1-
U)
L
tD
)
>
-44-)O
- >
Q) 4)
L-
(0
>\w
0
+
L
a
(1)
L
44-
0
-C
w
0
n
-
o
0
0
in
-
0
0
C'J
C')
0
0
(%) SUflJ
wouq
269
in
0
0
A1JIflWfl3
a
in
0
0
0
Table 8.4.
Cumulative abnormal returns to targets over the period -40
to -1 days by the method of payment.
Method of
Payment
Sample
Size
Pure Cash
83
______________ _______
Cash or
Equity
______________
Cash and
Equity
______________
Pure Equity
______________
F-statistic.
Dimson
Model
0.092*1+
a=5.53
b=7.59
c=67.47
0.l25
a=12.99
b=17.23
_______ c=73.87
222
0.097
a=7.94
b=11.lO
_______ c=75.66
0.093*1*
88
a=5.23
b=6.29
_______ c=71.59
111
504
1.11
Market
Adjusted
Mean
Adjusted
0.095
a=5.96
b=7.94
c=68.67
0.095k
a=5.80
b=7.56
c=68.68
0.l29
a=13.57
b=18.36
c=73.42
0.103*1*
a=8.66
b=ll.66
c=76.58
0.l37
a=13.59
b=18.65
c=72.97
0.0971*1
a=7.64
b=10.59
c=75.68
0.091
a=5.12
b=6.56
c=72.73
0.097
a=5.22
b=6.50
c=72.73
0.92
1.43
Notes:
Significant at the 1%.,
10%. levels
5%.,
respectively.
a = t-statistic under the dependence assumption.
b = t-statistic under the independence assumption.
c = percentage of observations that are positive.
*++ ** *
to the different methods of payment in the pre announcement
period. The t-statistic in Table 8.5 also shows that the
abnormal returns to the different methods of payment in the
pre bid period are comparable.
In Tables 8.6 and 8.7, we examine whether the abnormal
returns to targets in the post announcement period (^1 to
+40 days) is affected by the method of payment. Both the Fstatistic in Table 8.6 and the t-statistic in Table 8.7,
show that there is a method of payment effect in the
immediate post announcement period, with cash offers
270
Table 8.5.
Pairwise comparison of the target's abnormal return over
the period -40 to -1 days f or different methods of payment.
t-statistics are calculated assuming unequal group
variances where the null hypothesis of equal group
variances is rejected at the l0 level. Otherwise tstatistics are calculated assuming equal group variances.
Panel A: Results based on the Dimson Model.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
-1.40
-0.17
-0.06
0.98
1.31
Cash or
Equity
Cash &
Equity
0 . 11
Panel B : Results based on the Market Adjusted Model.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
-1.43
-0.25
0 . 14
0.97
1.59
Cash or
Equity
Cash &
Equity
0.36
Panel C: Results based on the Mean Adjusted Model
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
_1.68*
-0.07
-0.07
1.32
1.60
Cash or
Equity
Cash &
Equity
0.01
Notes:
.
.
.
0
1) *** *1 *
Significant
at 1,
respectively.
0
5,
l0-0
levels
2) Positive t-statistic implies that the mean for the
payment method in the vertical column is higher.
271
Table 8.6.
Cumulative abnormal returns to targets over the period +1
to ^40 days by the method of payment.
Method of
Payment
Sample
Size
Pure Cash
83
______________ _______
222
Cash or
Equity
______________ _______
Cash and
Equity
111
______________ _______
Pure Equity
88
______________ _______
F-statistic.
Notes:*1* ** +
504
Dimson
Model
O . 08 5 ***
a=5.20
b=6.72
c=61.45
0 . 0 2 9*1*
a=3.09
b=4.84
c=57.21
0.042*1*
a=3.51
b=4.55
c=60.36
0 . 028***
a=1.58
b=2 . 72
Market
Adjusted
0.090*1*
a=5.75
b=7.40
c=62.65
0.026*
a=2.73
b=4.58
c=57.21
c=57.96
2.19*
Mean
Adjusted
0.042
a=3.60
b=4.64
c=57.66
0.020*1*
a=l.15
b=2 . 64
c=59.09
0.lOf1
a=6.2l
b=7.8l
c=69.88
0.025*1*
a=2.50
b=4.31
c=58.11
0.041*1*
a=3.26
b=4.l6
c=54.96
0.027*1*
a=l.46
b=2 . 94
c=56.82
2.871*
3.36
.
0
0
.
Significant
at the l-,
lO-0
levels
5,
respectively.
a = t-statistic under the dependence assumption.
b = t-statistic under the independence assumption.
c = percentage of observations that are positive.
earning significantly higher abnormal returns than all
other methods of payment.
8.4.3. Returns to bidder shareholders
Table 8.8 shows the abnormal returns to bidder
shareholder. Bidders suffered significant wealth losses of
about 4. The negative wealth loss to bidders is consistent
with some of the extant results. Limmack and McGregor
(1992) reported returns to the bidder of -3.43 (t=2.49)
and Higson and Elliot (1993) reported bidder returns of 272
Table 8.7.
Pairwise comparison of the target's abnormal return over
the period +1 to +40 days by the method of payment.
t-statistics are calculated assuming unequal group
variances where the null hypothesis of equal group
variances is rejected at the l0 level. Otherwise tstatistics are calculated assuming equal group variances.
Panel A: Results based on the Dimson Model.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
2.27*!
1.70!
1. 90*
-0.77
0 .06
Cash or
Equity
Cash &
Equity
-0.62
Panel B : Results based on the Market Adjusted Model.
Cash or
Equity
Cash Offer
2.60
Cash or
Equity
Cash &
Equity
93*
Equity
Offer
-0.97
0.22
2 . 19*!
Cash &
Equity
0 .85
Panel C: Results based on the Mean Adjusted Model
Cash or
Equity
Cash Offer
2.91
Cash or
Equity
Cash &
Equity
2.42 **
Equity
Offer
-0.90
0.07
2 .28*!
Cash &
Equity
0.53
Notes:
* Significant at l,
1)
respectively.
5°-i,
10
levels
2) Positive t-statistic implies that the mean for the
payment method in the vertical column is higher.
273
Table 8.8.
Cumulative abnormal returns to bidders over different
windows surrounding bid announcements.
Sample size is 504.
Window
(Days)
-40 to +40
______________
-40 to -1
______________
0
_____________
+1 to +40
______________
Ijimson
Model
_0.038***
a=-4.42
b=-4.38
c=43.65
Market
Adjusted
_0.035*
a=-4.16
b=-4.01
c=42.06
0.008
a=1.32 "
b=2.18
c=49.80
0.010k
a=l.75
b=2.50
c=50.00
_0.012*
a=-12.43
b=-15.94
c=32.94
_0.034***
a=-5.78
b=-5.76
c=40.87
-0.0l2'
a=-12.30
b=-l5.47
c=32.54
_0.034+**
a=-5.74
b=-6.05
c=40.28
Mean
Adjusted
-0.035
a=-3.82
b=-3.65
c=46.82
0.014**
a=2.09
b=2.89
c=54.96
_0.012s**
a=-1l.39
b=-14.10
c=32.14
-0.037k
a=-5.80
b=-5.94
c=43.65
Notes:**$ ** +
.
.
0
0
Significant at . the 1, 5-s,
10 0
levels
respectively.
a = t-statistic under the dependence assumption.
b = t-statistic under the independence assumption.
c = percentage of observations that are positive.
5.30% (Z=2.70) . But these results are inconsistent with
Franks and Harris (1989) who report small significant
positive gains to bidders of 2.4 (t=2.28) and Limmack
(1991) who found bidder abnormal returns of -0.20
(t=0.22)
The size of the wealth loss suffered by bidders is
quite large over a relatively short event period. Most of
these losses occur in the immediate post announcement
period. In the post announcement period (+1 to +40 days)
Bidders suffered losses of _349, which is virtually
274
equivalent to the loss for the whole announcement period.
These negative post announcement abnormal returns are
puzzling. One possible explanation is that the negative
returns are caused by non-stationarity in the systematic
risk of the bidder. Franks et al (1991) provide some
evidence that these negative returns are due to errors in
the portfolio bench marks used to adjust for risk 3 . The
Franks et al result shows that the measurement of the
returns to the bidder can b sensitive to the model used to
control for risk. Since we are using a relatively short
event window and the result of negative performance for
bidders in the immediate post announcement period is
consistent across all three of our control models it is
doubtful whether our result can be attributed to errors in
the control bench mark.
Given the relatively short period of time over which
these negative post announcement returns have been
observed, variables such as the method of payment which
convey new information to the market could be significant
in explaining these negative returns (see Sections 2.4.2
and 7.3.1)
Figures 8.8-8.10 show the daily abnormal returns to
the bidder with the alternative models. The daily returns
to the bidder are more volatile than target daily returns
3 1n evaluating the post merger share price performance of bidder,
Franks et al (1991) calculated the control rates of return using both
a single factor bench mark (i.e, the market index) and multi factor
bench marks derived from the portfolio evaluation literature. They
found that abnormal returns in the post merger period were sensitive to
the number of factors included in the bench mark portfolio.
275
(see Figures 8.1-8.3) . This is consistent with the
prevailing view in the literature that it is more difficult
to estimate the wealth effects of a merger for bidders than
for targets (Jensen & Ruback, 1983)
Prior capitalisation of the merger gains in the
bidder's share price could affect our ability to detect any
merger benefits to bidders around the merger announcement
date. Usually targets can only be acquired once, whereas
bidders can engage in a prolbnged acquisition programme. If
the expected value of the acquisition programme is
capitalised in the bidder's share price when the
acquisition programme is announced, then the gains measured
around the merger announcement date would only partially
reflect the true merger gains. Schipper & Thompson (1983)
found some evidence that merger gains were capitalised in
the bidder's share price at the start of an acquisition
programme.
The larger size of bidders than targets also hampers
our ability to measure the wealth effects of the merger on
the bidder. Bidders in our sample are on average 30 times
larger than targets 4 hence the same € gains for bidders and
targets are translated into smaller abnormal returns for
bidders. The smaller returns to bidders caused by the
larger size of the bidders would limit the significance of
any wealth effects measured around the merger announcement
date (Asquith, Bruner & Mullins, 1983; Bruner, 1988)
Figure 8.11 shows the cumulative daily abnormal
4 See
the variable RELSIZE in Table 4.3 (chapter 4)
276
E'igure 8.8.
'Iote: Sample size is 504. Day 0 is the bid announcement
'ate. See text and Appendix 8.1 f or a description of the
Dimson model.
L
a)
-o
m
L
0
4J
Co
ci
/1\
w
ff
C
o
wE
a
I
L
0
-Q
o
o
•
o
1-
frjç'J
0 0 0
0
0
0
D 0 0
a
- ci o
- tn UJ N
0 0 0 0 0 0 0
0
0
0
0
0
0- 0
•
•
•
•
0 0 0 0 0 0 0
I
I
I
I
I
I
(%) sun.e
277
w.Jouq
- CJ C')
W U) c
0 0 0
0
0
0
0 0
•
•
a
•
0 0 I 0 C 0
I
I
I
I
I
E'igure 8.9.
Sample size is 504. Day 0 is the bid announcement
te. See text and Appendix 8.1 for a description of the
rket Adjusted model.
m
1)
-c
-o
I
a
-o
4J
C
a)
E
a)
U
C
0
C
C
a)
+J
-c
4J
0
4J
.4J
a)
a)
>
I-i
E
a)
L
a
L
(I)
>'
0
-Q
>\
1- m
O
0
rJ
0
0
0
- 0- 0- 0o o 0 0
m
0 c- CJ
0 0 0
0
IA ID N CD Dl
0 0 0 0 0 0
0 0 0 0- 0 0 0 0_ 0
- 0
0 0 0 0 0 0 0 0 0
I
I
I
I
C) sun;e
278
I
I
2w.ouq"
I
I
CJ C'1
c
I
0 0 0
-
0 0 0
I
I
I
Figure 8.10.
Sample size is 504. Day 0 is the bid announcement
te. See text and Appendix 8.1 for a description of the
an Adjusted model.
01
-o
-c
m
C
0)
£
U)
C)
C
:3
0
C
C
4J
U)
-c
(0
4J
0
cr
C
U)
>
4-J
0)
L
(I)
(0
0
L
0
-Q
>\
t1 i
C\J
-
o o o o
o
- o
- 0 0
o o 0 0
0
C\J çr
(\i Cfl
Ifl LD P'0)
0 0 0 0 0 0 0 0 0 0
0
0
0
0
0
0
0
0
0
0
0
cj
0 0 0 0 0 0 0 0 0 I 0 0 0
I
I
I
I
I
I
I
I
I
I
I
c
(%) sun.e
279
wouq
igure 8.11.
ote: Sample size is 504. Day 0 is the bid announcement
ate. See text and Appendix 8.1 for a description of thc
imson, Market Adjusted and Mean Adjusted models.
ri-i
L
-c
-o
0
0
ED
(0
0
(1)
L
4-j
a)
Q
-'
0
4)
(0
E
L
0
G)
>
1J+
L
-Q
(I)
>'
a)
>
£
0
w
E
4J
0
(0
ElI
Cd
0
C'J
(fl
0
I
0
I
0
0
U
O
0
0
I
(%J
SUfl.8J
wouq aAJ2pflWfl3
280
0
I
returns to the bidder. These are consistently positive
about 15 days before the announcement of the bid. This
result is in accordance with the extant evidence that
bidders generally tend to experience a positive run up in
their share price in the period immediately preceding a bid
(Mandelker, 1974; Langetieg, 1979; Asquith, 1983) . Assuming
that bidders wish to capitalise on the recent strong
performance of their shares, it is still doubtful whether
a positive abnormal return o 1 over the period -40 to -1
days would be sufficient to induce a bidder to undertake an
acquisition.
8.4.4. Returns to bidder shareholders by the method of
payment
Table 8.9 shows the abnormal returns to the bidders
partitioned by the method of payment. The returns of about
1 to cash bidders and about -2.5 to equity bidders is
consistent with the results in Franks et al (1988) and
Limmack & McGregor (1992). Franks et al reported returns of
0.7 (t=0.75) to cash bidders and -1.1 (t=-0.95) to equity
bidders, while Limmack & McGregor reported returns of
-0.l5 (t=0.09) to cash bidders and -2.96 (t=1.43) to
equity bidders. The most striking result in Table 8.9 is
the return to cash & equity offers of about -6.5%. Limmack
& McGregor found a similar result of -6.46% (t2.95) return
to cash & equity offers. Franks et al reported
insignificant gains of 0.3% (t=0.23) to cash & equity
offers. Given that cash bidders experience insignificant
281
Table 8.9
Cumulative abnormal returns to bidders over the period -40
to ^40 days by the method of payment.
Method of
Payment
Market
Adjusted
Mean
Adjusted
0.005
a=0.24
b=-0.07
_____________ _______ c=54.22
_0 . 0 4 0**
222
Cash or
a=-2.98
Equity
b=-2.57
______________ _______ c=43.24
_0 . 06 5 ***
Cash and
111
a=-4.07
Equity
b=-4.28
_____________ _______ c=36.04
-0.001
a=-0.03
b=-0.025
c=48.19
_0.037**
a=-2.76
b=-l.98
c=41.89
_0.065***
a=-4.24
b=-4.59
c=34.23
0.010
a=0.45
b=0.77
c=55.42
_0.037**
a=-2.52
b=-1.56
c=47.30
-0.074
a=-4.36
b=-4.95
c=35.14
-0.037
a=-l.60
b=-1.53
_______ c=44.32
1.98*
504
-0.025
a=-l.05
b=-l.05
c=46.59
-0.023
a=-0.89
b=-0.88
c=52.27
1.61
1.79
Pure Cash
Pure Equity
_____________
F-statistic.
Notes:
Sample Dimson
Size
Model
83
88
*
Significant at the 1%, 5%, 10% levels
respectively.
a = t-statistic under the dependence assumption.
b = t-statistic under the independence assumption.
c = percentage of observations that are positive.
positive returns and equity bidders experience
insignificant negative returns, the finding of significant
losses to cash & equity offers is rather puzzling. Further
analysis below suggests that these losses to cash & equity
offers results from the poor performance of bidders
offering cash & equity in the pre announcement period.
In Table 8.10 we perform a pair wise test of
difference in average abnormal returns to bidders using
different methods of payment. The only uniformly
significant difference is that between cash offers and cash
282
Table 8.10
Pairwise comparison of the bidder's abnormal return over
the period -40 to +40 days for different methods of
payment.
t-statist±cs are calculated assuming unequal group
variances where the null hypothesis of equal group
variances is rejected at the 10% level. Otherwise tstatistics are calculated assuming equal group variances.
Panel A: Results based on the Dimson Model.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
1. 9 7 *
2.58**
1.37
1.08
-0 .12
Cash or
Equity
Cash &
Equity
-0.93
Panel B : Results based on the Market Adjusted Model.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
1.52
2 . 55
0 . 79
1 .23
-0.44
Cash or
Equity
Cash &
Equity
-1.38
Panel C: Results based on the Mean Adjusted Model
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
1.52
2.69
0.88
1.35
-0.39
Cash or
Equity
-1.46
Cash &
Equity
Notes:
1)
' ' * Significant at 1%,
respectively.
5%,
10% levels
2) Positive t-statistic implies that the mean for the
payment method in the vertical column is higher.
283
& equity offers. This is not consistent with the argument
that the use of equity will lead to a downward revision in
the bidder's share price. If the market reacts unfavourably
to the use of equity, we would expect equity bidders to
show the largest losses. Contrary to this we find that
there is no significant difference in returns between
equity bidders and cash bidders, while bidders offering a
mixture of cash & equity suffer the largest losses.
Figures 8.12-8.14 partition the cumulative abnormal
returns to bidders according to the method of payment. It
is clear from these graphs that there is a method of
payment effect in the negative returns observed for the
post announcement period. There is a marked contrast in the
pre and post announcement abnormal returns behaviour of
equity and cash bids. In the pre announcement period
bidders who finance the acquisition substantially with
equity (i.e, equity offers and cash or equity offers)
experience an increase in their share price. In the post
announcement period while cash bids have consistently
positive returns, the returns to equity bids is
consistently negative.
The most unusual result in Figures 8.12-8.14 is the
post announcement performance of cash & equity offers.
Unlike the results for the targets (where the abnormal
returns to cash & equity offers falls between the abnormal
returns to cash offers and equity offers) the abnormal
returns to cash & equity offers are lower than the abnormal
returns to equity offers. This suggests that somehow cash
284
Figure 8.12.
note: Day 0 is the bid announcement date. See text and
Appendix 8.1 for a description of the Dimson model. Sample
size:- Cash = 83, Equity = 88, Cash or Equity = 222, Cash
and Equity = 111.
>1
.4)
w
w
0
C
E
>
(c
0
0
0
D >
(B
(cs
>'
-o
a
C
Q
o L
o C
C
0
a)
D
0
-c
-,'\ 0
-O
0
w
E
U)
b
a)
L
)
> tl_
-
U]
4) 0
('j
a,
L
-c
-o
Y
U)
>'w
0
m
L
0
+
U,
L
440
-c
(B
U
cii
ci 00 0
c
0
0
0
cii 0- 0in Crn CN
C'J 0
0 0
0
0
0
0
0 0 0
I
I
I
I
I
I
I
(%) sune
wouq
285
A!QInwn3
0
Figure 8.13.
Note: Day 0 is the bid announcement date. See text and
Appendix 8.1 for a description of the Market Adjusted
model. Sample size:- Cash = 83, Equity = 88, Cash or Equity
= 222, Cash and Equity = 111.
4-J
C
w
E
w
C
>\
U)
0
U
4-
0
-o
-o >
4J
C
w
0
C
4Jo
U
>\-?
0<
0
-I_J
.4J
(0
L
(fli
L
G) Q)
> II- '4-
4O
- >s
G)
L
-c
-o
m
4-
(I)
>'w
L
a
+
L
(1)
L
1)
'4-
4-
0
U)
0
(fl
0
(\J
0
0
0
d
d
b
b
0
(J
0
I
C%) SUJfl
iwjouq
286
(T)
0
d
I
T
0
I
(11
0
d
I
A1JIflUfl3
(0
0
I
N
0
0
Figure 8.14.
Note: Day 0 is the bid announcement date. See text an
Appendix 8.1 for a description of the Mean Adjusted model.
Sample size :- Cash = 83, Equity = 88, Cash or Equity =
222, Cash and Equity = 111.
Ui
-o
E
>
(c
-E
LU
(U
U
0
•1
0
(U
-o >
4J
-o
C
U)
0
L
0
0_C
U)
.I-J
0
-1-i(U
L
U) )
>4-44) 0
(U
L
a
Y
(U
0
+
to
L
0
L
4-
4-
C
£
LI
0
(U
r
0
in
0
OJ
0 0
d d
0
-
ri
t'i
0 0 0
d
I
I
C%J sUJflG.l
wouq
287
0
I
in
w
N
w
0 0 0 0
doo
I
I
I
I
AjJflWfl3
0
& equity offers send the worst signal to the market. To
explore this anomalous result further, below we analyse the
pre and post announcement behaviour of abnormal returns to
bidders using different methods of payment.
In Tables 8.11 and 8.12 we look at the pre
announcement (-40 to -1 days) abnormal returns behaviour of
bidders analysed by the different methods of payment. Pre
bid abnormal returns to offers with a substantial amount of
equity (i.e, equity offers and cash or equity offers) are
higher than those for offers with a large cash component
(i.e, cash offers and cash & equity offers) . This is
consistent with the results we found earlier in Chapter 5
(see Sections 5.2.1 and 5.3) that bidders experiencing a
short run increase in their share price are more likely to
use equity as a method of payment.
In Tables 8.13 and 8.14 we examine the bidder's
abnormal returns in post announcement period (+1 to +40
days) . Since the method of payment cannot be fully
anticipated by the market, the information impact of the
different payment methods is concentrated in the post
announcement period. The results show that all methods of
payment which have an equity component, incur significant
losses in the post announcement period. This difference in
abnormal returns between cash offers and offers with an
equity component over such a relatively short time period,
is consistent with the argument that the use of equity
conveys negative information to the market (see Sections
2.4.2 and 7.3.1).
288
Table 8.11
Cumulative abnormal returns to bidders over the period -40
to -1 days by the method of payment.
Method of
Payment
Market
Adjusted
Mean
Adjusted
-0.001
a=-0.08
b=-0.33
_____________ _______ c=48.19
0 . 017*
Cash or
222
a=l.75
Equity
b=2.72
______________ _______ c=50.45
-0.002
a=-0.l8
b=-0.4l
c=48.19
0.021**
a=2.20
b=3.35
c=50.90
-0.004
a=-0.08
b=-0.87
c=53.Ol
0.031*
a=1.75
b=4.64
c=58.56
-0.011
Cash and
111
a=-0.97
Equity
b=-0.37
_____________ _______ c=44.14
-0.014
a=-l.30
b=-l.03
c=45.05
0.027*
a=1.63
b=2.21
c=55.68
-0.021
a=-0.97
b=-1.82
c=46.85
0.030*
a=l.l2
b=2.43
c=57.96
3.26**
Pure Cash
Sample Dimson
Size
Model
83
0 . 0 1 8*
Pure Equity
a=1.12
88
b=l.65
______________ _______ c=56.82
F-statistic.
504
1.31
2.25k
Notes:
** * Significant at the 1, 5,
lO% levels
respectively.
a = t-statistic under the dependence assumption.
b = t-statistic under the independence assumption.
c = percentage of observations that are positive.
Combining the results from Tables 8.11 and 8.13 shows
that, while in the pre announcement period equity offers
had higher returns than cash offers, in the post
announcement period the returns to equity offers are lower
than the returns to cash offers. This shows that our
earlier result from Tables 8.9 and 8.10, that the valuation
effects of cash offers and equity offers were comparable
over the whole of the event period (-40 to +40 days) is due
to the fact that bidders offering equity had a run up in
their share prices in the pre bid period which when
289
Table 8.12
Pairwise comparison of the bidderts abnormal return over
the period -40 to -1 days for different methods of payment.
t-statistics are calculated assuming unequal group
variances where the null hypothesis of equal group
variances is rejected at the l0 level. Otherwise tstatistics are calculated assuming equal group variances.
Panel A: Results based on the Ijimson Model.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
-1.31
0.62
-0.99
1.70 *
-0.08
Cash or
Equity
Cash &
Equity
-1.39
Panel B : Results based on the Market Adjusted Model.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
-l.61
0.79
-1.46
2.35
Cash or
Equity
SI
-0.31
2.01
Cash &
Equity
Panel C: Results based on the Mean Adjusted Model
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
_2.13**
0.83
-1.41
2.79
Cash or
Equity
*5*
0.08
Cash &
Equity
1.99
Notes:
1)
Significant at l,
respectively.
*5* 5* *
•
0
5,
0
10
0
levels
2) Positive t-statistic implies that the mean for the
payment method in the vertical column is higher.
290
Table 8.13
Cumulative abnormal returns to bidders over the period +1
to +40 days by the method of payment.
Sample Dimson
Size
Model
Market
Adjusted
Mean
Adjusted
0.005
a=0.37
b=0.36
______________ _______ c=54.22
0.001
a=0.09
b=0.19
c=50.60
_0.048***
a=-5.19
b=-5.00
c=40.09
_0.030s**
a=-2.81
b=-3.20
c=34.23
_0.035**
a=-2.14
b=-2.44
c=42.05
0.016
a=l.04
b=l.31
c=62.65
Method of
Payment
Pure Cash
83
-0.047
Cash or
222
a=-5.05
Equity
b=-'5.24
_____________ _______ c=38.29
_0 . 033***
Cash and
ill
a=-2.98
Equity
b=-3.47
_____________ _______ c=32.43
_ 0.038 **
Pure Equity
88
a=-2.38
b=-2.61
_____________ _______ c=42.05
F-statistic.
504
3.47
+
2.68
+
-0.059k"
a=-5.80
b=-5.89
c=40.99
0.033
a=-2.82
b=-3.23
c=34.23
-0.036
a=-l.95
b=-2.50
c=44.32
3.66
+1
Notes:++* ++
levels
Significant
at the l, 5, 10
respectively.
a = t-statistic under the dependence assumption.
b = t-statistic under the independence assumption.
c = percentage of observations that are positive.
*
.
.
0
0
0
averaged with the post announcement losses results in a
total return which is comparable to the return in cash
offers.
The losses incurred by cash & equity offer and pure
equity offer bidders in the post announcement period are
similar. This suggests some explanation for the earlier
result that losses over the whole event period are larger
for cash & equity offers than for pure equity offers. The
pre announcement run up in the share price for equity
offers is alleviating some of the post announcement losses,
291
Table 8.14
Pairwise comparison of the bidder's abnormal return over
the period -40 to -1 days for different methods of payment.
t-statistics are calculated assuming unequal group
variances where the null hypothesis of equal group
variances is rejected at the lO level. Otherwise tstatistics are calculated assuming equal group variances.
Panel A: Results based on the Dimson Model.
Cash or
Equity
Cash &
Equity
Equity
Offer
Cash Offer
2.15**
2 . 04
Cash or
Equity
-1.02
-0.51
0.27
Cash &
Equity
Panel B : Results based on the Market Adjusted Model.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
2.90
1.79*
1 . 76
-1.24
-0.74
Cash or
Equity
0.26
Cash &
Equity
Panel C: Results based on the Mean Adjusted Model
Cash or
Equity
34Q***
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
2.33**
2.10d*
-1.40
-1.06
Cash &
Equity
0.11
Notes:
Significant at 1,
1)
respectively.
**$ +*
*
•
0
5,
0
l0-
0
levels
2) Positive t-statistic implies that the mean for the
payment method in the vertical column is higher.
292
while the absence of any run up in the share price for cash
& eqi.iity offer bidders results in significant losses over
the whole event period.
8.5. Role of underwriters
The results above show that the use of equity conveys
negative information to the market (see Tables 8.13 and
8.14). It has been suggested that the use of an underwriter
may help alleviate some of the information problems
associated with equity offers (Heinkel and Schwartz, 1985;
Slovin et al, 1990) . In a cash or equity offer, the
underwriter agrees ex-ante to buy back the bidder's shares
at a pre-determined price from those target shareholders
who elect to accept the cash alternative. The contractual
liability of the underwriter in such an agreement implies
that it should have a reasonable knowledge of the value of
the bidder's assets. Therefore the presence of an
underwriter can signal that the bidder's shares are not
overvalued, since underwriters would not normally agree to
buy back overvalued assets. Bidders using an underwriter
should therefore experience less negative abnormal returns
than bidders without an underwriter.
To examine this issue, in Table 8.15 we analyse the
returns to bidders offering "cash or equity" separating
cash offers which are underwritten from cash offers which
are not underwritten.
Bidders using an underwriter have higher preannouncement returns but lower post-announcement returns
293
Table 8.15.
Cumulative abnormal returns to bidders using "equity offers
with a cash alternative" as the method of payment.
Window
(Days)
-40 to -40
Window
(Days)
-40 to -1
Window
(Days)
+ 1 to + 40
-0.0485"
a=-2.54
b=-1.97
c=44.85
0.0242"
a=1.81
b=3.16
c=55.15
-0.0579"
a=-4.39
b=-4.60
c=35.29
-0.0272
a=-1.59
b=-1.65
c=40.70
0.0047
a=0.39
b=0.39
c=43.02
-0.0304"
a=-2.57
b=-2.63
c=43.02
0.83
1 .06
1 .77
-0.0438"
a=-2.32
b=-1.32
c=41.91
0.0344"
a =2.59
b=4.07
c=56.61
-0.0637"
a=-4.85
b=-4.56
c=37.50
-0.0263
a=-1.64
b =-1 .53
c=41.86
-0.0008
a=-0.07
b = 0.26
c=41.86
-0.0240"
a=-2.16
b =-2.30
c=44.19
0.59
1.70
2.21"
-0.0507"
a=-2.57
b=-1.25
_________ c=47.06
0.0450"
a=3.25
b=4.97
c=61.03
-0.0810"
a=-5.92
b=-5.51
c=38.97
-0.0152
a=-0.85
b =-0.93
c=47.67
0.0087
a =0.69
b = 1 .21
c=54.65
-0.0241"
a=-1.93
b =-2.53
c=44.19
0.99
1 .66
2.34"
Sample
Size
Panel A: Dimson Model
With an Underwriter
136
Withoutan
Underwriter
86
t-statistic
Panel B: Market Adjusted Model
With an Underwriter
136
Without an
Underwriter
86
t-statistic
Panel C: Mean Adjusted Model
With an Underwriter
Without an
Underwriter
t-statistic
136
86
Notes:
Significant at 1 %, 5%, 10% levels respectively.
a = t-statistic under the dependence assumption.
b = t-statistic under the independence assumption.
c = percentage of observations that are positive.
294
than bidders not employing an underwriter. Over the whole
event period the presence of an underwriter results in
greater losses for the bidder, although the difference is
not statistically significant.
The higher pre-announcement returns to bidders using
an underwriter suggest that these firms have stronger prebid performance and perhaps have less need for
underwriting. It is plausible that these are firms which,
having experienced a period of recent share price growth,
are trying to assuage investor fears that the use of equity
isn't an attempt to capitalise on this share price growth
by converting overvalued paper into real assets. If this is
the strategy, then it has obviously failed, since their
post-announcement losses are quite significant. There are
a number of plausible explanations for the larger negative
post-announcement abnormal returns to bidders using an
underwriter.
Significant fees are paid to the underwriters,
apparently without any benefits in terms of a reduction in
the information asymmetry problem. Oborne (1986) estimates
that in addition to fees paid for advisory services,
merchant banks involved in underwriting a cash alternative
can normally expect l/2 commitment commission paid to the
lead underwriter, l/2 commission paid to the subunderwriters, l/4 arrangement commission to the brokers.
These are paid irrespective of whether the offer is
successful. An additional 3/4 commission is payable to the
sub-underwriters if the offer is successful. Bidders using
295
an underwriter could end up paying 2 in underwriting fees.
Potentially, after the completion of the bid, the
underwriters could be left holding a large block of shares
in the bidder, if a significant number of target
shareholders accept the cash offer. This may create a fear
that eventually a large block of shares would be dumped in
the market in the aftermath of the bid. This fear of an
"overhang" from the bid may depress the bidder's share
price and so contribute to the observed negative abnormal
returns.
Further, there is some evidence of mean reversion in
the behaviour of share prices (]JeBondt & Thaler, 1985;
DeBondt & Thaler, 1987) ie, companies that have experienced
periods of exceptional share price increases, subsequently
experience significant share price falls and vice versa. It
is plausible that the higher pre-bid announcement rises in
the share price of bidders employing an underwriter through
the mean reversion process is contributing to the larger
post announcement losses.
8.6. Impact of capital gains tax, information asymmetry and
debt co-insurance on shareholder wealth
In this section we investigate the source of the
higher returns to cash offers. We regress the announcement
period abnormal returns on variables which control for the
impact of capital gains tax, information asyimnetry and debt
co-insurance.
By controlling for capital gains tax, information
296
asymmetry and debt co-insurance effects, we can examine
directly whether, the tax, information asymmetry or the
debt arguments explain the higher returns to cash offers.
Based on the discussion in Chapter 7, the following
specific hypotheses are tested:
'Bidder's information asymmetry hypothesis, Hi': Returns to
bidder shareholders and the amount of information asymmetry
about the value of the bidder existing between bidder
managers and target shareholders will be positively related
in a cash offer and negatively related in an equity offer.
'Target's information asymmetry hypothesis, H2': In an
equity offer, returns to the target shareholders are
negatively related to the amount of information asymmetry
between the bidder and target. In a cash offer, information
asymmetry has no effect on the returns to the target
shareholders.
'Capital gains tax hypothesis, H3': In a cash offer, the
returns to the target shareholders are positively related
to realisable capital gains, while the returns to the
bidder shareholders are negatively related to realisable
capital gains. In an equity offer, realisable capital gains
have no impact on the returns to the bidder and target
shareholders.
'Wealth transfer hypothesis, H4': The returns to both the
target's and the bidder's shareholders will be negatively
related to the change in risk of the merged firm's returns,
however the relationship between the change in the variance
of the merged firm's returns and the returns to the bidder
297
and target shareholders will be stronger in an equity offer
than in a cash offer.
The above hypotheses are stated in terms of just cash
offers and equity offers whereas the empirical analyses
below are conducted in terms of four different methods of
payment (cash offers, equity offers, cash and equity offers
and cash or equity offers) . Since "cash and equity" offers
and "cash or equity" offers are hybrid offers comprising
both cash and equity the effect of CGT, information
asymmetry and debt co-insurance on these two methods of
payment is likely to be an amalgamation of the impact which
these variables have in cash offers and equity offers
respectively. Consequently it is left to the data to
clarify the effects of CGT, information asymmetry and debt
co-insurance on "cash and equity" offers and "cash or
equity" offers.
8.5.1. Description of explanatory variables
Capital gains tax (CGAIN)
The tax based explanations for the higher returns to
cash offers suggest that target shareholders in cash offers
are compensated for the immediate crystallisation of their
CGT liability (see Sections 2.5 and 7.4) . While this
argument may explain the higher returns to target
shareholders, it is inconsistent with the evidence that
bidders also have higher returns in cash offers, since the
compensation premium paid is at the expense of the bidder.
The tax variable (CGAIN) which measures short run capital
298
gains iS:
Pre-bidMkt V alue Lowest Mkt V alue
CGA IN = Of The Target - Of The Target
Lowest Mkt V alue Of The Target
From hypothesis H3 we expect that in cash offers CGAIN
should be positively related to the target's CAR and
negatively related to the bidder's CAR., while in equity
offers CGAIN should not affect the returns to either
bidders or targets.
Information asymmetry (BIDSTD & TAGSTD)
An equity offer is likely to be associated with
negative information about the bidder and results in a drop
in the market value of the bidder (see Sections 2.4.2 and
7.3.1). The fall in the market value of the bidder should
be positively related to the level of information disparity
between the market and the firm. Uncertainty about the
value of the target results in an equity offer by the
bidder and a reduction in the premium paid to target
shareholders (see Section 7.3.2)
Direct measurement of information asymmetry is not
easy, since it is based on the existence of private
information which cannot be observed. Dierkens (1991)
presents a model which allows an indirect measurement of
information asymmetry. IJierkens suggests that the standard
deviation of the market adjusted returns can be used as a
suitable proxy for information asymmetry. The total
299
uncertainty about the value of a firm faced by an external
investor is made up of two components: uncertainty due to
market wide influences (e.g, interest rates, inflation etc)
and uncertainty due to firm specific variables (e.g, future
production plans, impact of competitors etc). Managers as
insiders acquire knowledge of firm specific factors before
the market does, although with the passage of time, firm
specific information known to managers will be passed to
the market.
If the variance of the market's return (a2 ) measures
the uncertainty about market wide factors, while the
variance of the firm's returns (a 2
) measures uncertainty
about both market wide and firm specific factors, then
cr2
minus a2 , measures the uncertainty about firm specific
factors faced by external investors5 . The volatility of the
firm's market adjusted return (standard deviation) is used
as a proxy for a 2 minus cr2.
The standard deviation of the firm's market adjusted
returns is measured over the 250 trading days ending 40
days before the announcement of the bid. This is referred
to as BI]JSTD and TAGSTD for bidders and targets
respectively.
From hypothesis Hi we expect that, in an equity offer,
the returns to the bidder will be negatively related to
BI]JSTD, while, in a cash offer, they will be positively
correlated.
5mis
implies an assumption that beta equals one.
300
From hypothesis H2 we expect that, in an equity offer,
the returns to the target will be negatively correlated
with TAGSTD, while in a cash offer, TAGSTD will not
influence the returns to the target.
Debt co-insurance effect (DEBTINS)
The debt co-insurance argument (see Section 7.5)
suggests that if a merger results in a fall in the variance
of the combined firm's returns, and hence the riskiness of
debt, then bondholders should gain while shareholders lose.
If the correlation of returns to bidders and targets is
zero, then the variance of the merged firm's returns is
simply the average of the variance of returns of the
individual firms weighted by their respective market
values, i.e:
1p=O
MV bidabid ^ MV tagOtag
where:
°bid
and
cr
2
g
are the pre-bid variance of the bidder t s and the
target's returns respectively;
MV
bjd
and MV are the market values of the bidder and the
target respectively;
p is the correlation of returns to the bidder and the
target.
With p equal to zero the merger has not changed the
risk profile of the merged
firm
and hence there is no
redistribution of wealth between bondholders and
shareholders. If p is not equal to zero, then the variance
301
of the merged firm's returns is given by:
0p*O -MVbldo bid + M1 agO ag +
If p is negative, then the variance of the merged
firm's returns will fall and wealth will be transferred
from the shareholders to bondholders, while a positive p
will increase the variance of the merged firm's returns and
result in a transfer of wealth from bondholders to
shareholders. The change in the variance of the merged
firm's return resulting from the correlation of returns on
the bidder and the target being non-zero Is:
DEBTINS = ___
0p*O
If p is negative and the merger reduces the variance
of the merged firm's returns, then ]JEBTINS will be greater
than one, while with a positive p the merger increases the
variance of the merged firm's returns and DEBTINS will be
less than one.
The variance of the pre-bid returns is measured over
250 trading days ending 40 days before the announcement of
the bid6 . Market value is measured 41 days before the
announcement of the bid.
6 The wealth transfer theory is concerned with changes in the total
return of the firm (ie, returns to debt and equity). Our proxy (which
is similar to that used by Eger, 1983 and Travlos, 1987) is deficient
to the extent that it only measures changes in the returns to equity.
Changes in the firm's total return are difficult to calculate because
corporate debt in the UK is generally unquoted.
302
From hypothesis H4 if there is a debt co-insurance
effect then the statistical relationship between DEBTINS
and the returns to the bidder and the target will be
stronger in equity offers than in cash offers.
The definitions of the variables discussed above are
summarised in Table 8.16.
Table 8.16.
Explanatory variables influencing the wealth experience of
shareholders in takeovers.
Variable Definition
(Target's day -41 market value - target's
lowest market value over the preceding one
_________ year) / the target's lowest market value.
CGAIN
BIDSTD
Standard deviation of the bidder's market
adjusted daily abnormal returns measured
over the 250 trading days ending 40 days
before the announcement of the bid.
TAGSTD
Standard deviation of the target's market
adjusted daily abnormal returns measured
over the 250 trading days ending 40 days
before the announcement of the bid.
DEBTINS
Variance of the merged firm's returns,
assuming the correlation of returns to the
bidder and target is zero / variance of the
merged firm's returns, if the correlation of
returns to the bidder and target is not
equal to zero
Table 8.17 reports the mean, median and standard
deviation for the explanatory variables, while Table 8.18
reports the results of the difference in means test
comparing the mean for each explanatory variable across the
different method of payments.
The F-statistic shows that potential short run capital
gains realisable by target shareholders (CGAIN) is not
significantly different across methods of payment. This is
303
Table 8.17.
Impact of the method of payment on shareholder wealth:
Descriptive statistics for the explanatory variables.
The variables are defined in Table 8.16.
MEAN
J
MEDIAN
STD DEV
No OF OBS.
Variable: CGAIN
Whole sample
0.546
0.332
1.276
504
Cash offers
0.410
0.273
0.616
83
Equity offers
0.431
0.270
0.441
88
Cash and equity
0.547
0.309
0.825
111
Cash or equity
0.644
0.400
1.768
222
F-Statistic
0.98
Variable: BIDSTD
Whole sample
0.020
0.018
0.010
504
Cash offers
0.020
0.018
0.008
83
Equity offers
0.022
0.018
0.012
88
Cash and equity
0.018
0.016
0.009
111
Cash or equity
0.020
0.018
0.010
222
2.79"
F-Statistic
Variable: TAGSTD
Whole sample
0.022
0.020
0.010
504
Cash offers
0.022
0.020
0.009
83
Equity offers
0.026
0.023
0.014
88
Cash and equity
0.020
0.018
0.009
111
Cash or equity
0.022
0.020
0.009
222
F-Statistic
6.57"
Variable: DEBTINS
Whole sample
0.943
0.970
0.078
504
Cash offers
0.970
0.991
0.062
83
Equity offers
0.914
0.944
0.099
88
Cash and equity
0.922
0.942
0.078
111
Cash or equity
0.955
0.974
0.067
222
F-Statistic
1 2.61
Note:
** * Significant at the 1%, 5% 10% levels respectively.
304
Table 8.18.
Difference in means test for the variables explaining the
impact of the method of payment on shareholder wealth.
t-statistics are calculated assuming unequal group
variances where the null hypothesis of equal group
variances is rejected at the l0 level. Otherwise tstatistics are calculated assuming equal group variances.
The explanatory variables are defined in Table 8.16.
Panel A: Results for the variable CGAIN.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
-1.71
-1.33
-0.26
0.68
1.66k
Cash or
Equity
Cash &
Equity
1.27
Panel B : Results for the variable BIDSTD.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
0.34
l.82
-1.15
1.68*
-1.48
Cash or
Equity
-2 . 60"
Cash &
Equity
Panel C: Results for the variable TAGSTD.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
0.39
l.95
-2.18"
Cash or
Equity
1.88
*
Cash &
Equity
Notes:
1)""
-2.67
*4*
-3.72
+4*
Significant at 1%, 5%, 10% levels respectively.
2) Positive t-statistic implies that the mean for the payment method in the vertical
column is higher.
305
Table 8.18. (Continued)
Difference in means test for the variables explaining the
impact of the method of payment on shareholder wealth.
t-statistics are calculated assuming unequal group
variances where the null hypothesis of equal group
variances is rejected at the 10% level. Otherwise tstatistics are calculated assuming equal group variances.
The explanatory variables are defined in Table 8.16.
Panel D: Results for the variable DEETINS.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
1.83*
4. 76w
3 74***
4 . 48
Cash or
Equity
Cash &
Equity
3 . 56
0.63
Notes:
1)
Significant at 1%, 5%, 10% levels respectively.
2) Positive t-statistic implies that the mean for the payment method in the vertical
column is higher.
consistent with the earlier evidence in Sections 5.2 and
5.3 that COT does not significantly affect the choice of
payment by the bidder. The CGT compensation hypothesis
predicts that potential capital gains (CGAIN) should be
lower in cash offers than in equity offers. The results
show that potential CGT (COAIN) is lower in cash offers
than in offers with an equity component. However this
evidence in support of the CGT compensation hypothesis is
weak and statistically insignificant.
The standard deviation of the pre bid market adjusted
abnormal returns (BIDST]J & TAGSTD) is about 0.02, however
the F-statistic shows that there are significant variations
306
across payment methods. Information asymmetry about the
bidder (BIDSTD) is higher in pure equity offers than in
offers with a cash component. This suggests that
information asymmetry about the value of the bidder is more
likely to result in an overvaluation of the bidder, hence
resulting in bidding managers offering equity as the method
of payment 7 (Myers and Majluf, 1984).
Information asymmetry about the value of the target
(TAGSTD) is significantly higher in equity offers than in
offers with a cash component. This is consistent with
Hansen's (1987) (see Section 2.4.1) model that where the
bidder is uncertain about the value of the target then an
equity offer is an effective means of sharing the valuation
risk faced by the bidder with the target shareholders.
Contrary to the implication of the prediction that
wealth is transferred from shareholders to bondholders
(Higgins & Schall, 1975; Galai & Masulis, 1976) we find
that the merger has increased the riskiness of the combined
firm (]JEBTINS is less one) 8 . This effect is similar across
all methods of payment with the largest increase occurring
in equity bids. This increase in riskiness is consistent
with the arguments of Smith & Warner (1979) and Jensen &
7 1f information asymmetry about the value of the bidder resulted
in an undervaluation of the bidder, then Myers and Majiuf (1984) (see
Section 2.4.2) predicts that the bidder would offer cash as the method
of payment. In this scenario information asymmetry (BIDSTD) would be
higher in cash offers than in equity offers.
8The average correlation coefficient between the target and the
bidder's pre-bid returns is 0.13. If the on average mergers were meant
to reduce the riskiness of the combined firm, then the correlation
coefficient between the bidder and the target's pre-bid returns should
be negative.
307
Meckling (1986), who suggest that the wealth redistribution
can flow from bondholders to shareholders. They suggest
that managers of firms with outstanding debt can seek to
transfer wealth from the bondholders to the shareholders by
investing in projects that increase the riskiness of the
firm. Since a merger is another form of corporate
investment, the increase in riskiness resulting from the
investment, would increase the default risk on the
outstanding debt, resulting in a transfer of wealth from
the bondholders to the shareholders.
8.5.2. Target returns and the method of payment
In Table 8.19 we report the findings from the cross
sectional regressions with target's CAR as the dependent
variable. To test for heteroskedasticity in the regressions
we use the Breusch-Pagan test (See Maddala, 1989: p.164)
As we shall see below a significant number of our
regression models suffer from heteroskedasticity. We use
White's (1978) procedure to correct for heteroskedasticity
in estimating the regression standard errors.
The impact of the explanatory variables on shareholder
wealth is dependent on the method of payment (see
hypothesis Hi, H2, H3 and H4) 9 , hence we estimate the
regression model for the entire sample and for the
different methods of payment. Unfortunately the results are
very disappointing.
9 For example, from hypothesis Hl, we expect that in cash offers
information asymmetry will have a positive impact on the bidder's
abnormal returns and a negative impact in equity offers.
308
'U
U]
a)
H
H
m
'D
N
a)
H
-H
-1
'ii
-H
::'
>1
>1
o•
UI
ri
(j
H
p
F'
cii
V
LU
l
'..D
U)
.
*
N
.4-i
H
H
cii
4)
U]
-
N
-1
çI.,
a)
H
-H
a)
a)
'1
-H
,qU)
0
a)
+
l
**
*+
+
l
o4
a)
4
V
H
-U 0
(i
o +
C.)
-U
a)
Ow
— -Ua)
a)-H
a)
c)4
cii
,
,
,
,
I-i
i-I
N
'N
a)
P-i
U]
—
LU
)N
OH H
O
N°'
r-1
ci)
0)
4)
UI
U] UI
v-1
HLfl
H Cc cc co
'
H
a)
ci)
H
j.,
rj
UI
o\°
O
H
0
E
O C/)
4-.)
fl
U)
a) 0
4-1
-H
•lj
j
o
'
0 HO O'rn co
H
UI
+ —
*Lt m +
Nm = N
O' 0 H
0 •H
- - H H
H
H 00
UI
rcj
cii
H
a)
-1
04)
C,)
Jl
C.)
z
'a)
H
H
'dcj
U]
•
Op
c -H
rj
-i
-U
* — * —.
U)
0
m
LI)
U]
•)
-H
.,-i
cii
O
a)
N
-,
El
*
a
LU
UI
-U
**
*
4)
-H
E
-H
0)-H
o
0
F'
CTLC CJ
N
,
On
a)O',
V
LO0
N H 0 N 0 H 0 m H
,
,
ci)
UI
a)
LU
-U
Q ow
-H
UI
4)
a) 1UI
O4i
:;'
-H
U]
w
.0
-r-44i
(ii
H
a)'
E
o
a)
.
4-i
o
•
(4-1
O
'-1w
Qr-4
UI.
-H a)
-1
UIa)
'r-I
.JUI
El
'd
H
4i H
Na) H
NN 'DN
O El HN
Na) NO Na)
C/)
•
OOCD0 OH 00
V
a) 0
..U] U
('I
a)
-1
U]
—
(1)
a)
UI
4-4
b1—
4-1' a)H a)N
ri
rdH
a)
Wa)
Wa) rj
-H
H
U]
0)4-i
(ii
U)
-i
ci)
-U
a)
UI
rI
4-1
-H
-U
-H
Cl)
-
a)
-H
a)U]
4-i
Q
U]
a)
HO
--H
o\0
0)
H
0
U]
UI
-H
i
(1)
N
'N
H
O
+
4 .'
:
b
•U]
UI
UIa) .a)
E-'r
E-'d
r
309
UI
U)
Cd
(1)
N
w
H
Cd
Oi
H
N
N
O 0 0 N 0
.
.
0
0
0
0
0
b
- --H
Cd
Cd
:::
>1
.1-)
0
U] -H
H
- W
- N.
H
'
m
m
UU)
H
H
Cd
__ __ __ __ __ __ H
-H
Cd
4*
(1]
0
H
i
0
rn
-1
C])
0 U
El N
N
m °
H N
a)
r
N
Cd
H
a)
-H
a
>1
*4
co
W
Cd
*
*
*
*
U
__
>1
H
U]
a) o
ji
0
o
+
U(I)
O
- U(])
o
U]
O jj
rdid
OP4
-H
-
o
Cd
E
a)
U
j
Cd
U
U]
-
a)
0
El
Ci
C)
,1
-1
Cd El
o a)-H
in
WN
U]
a)
Cd
U]
H
a)
>
H
0
a)- r ___
U
a) P4
Cd
H
-H
0 U
P4 UC])
W *N
* H :* N
H m H H 0 •
H i H
CN
0 CN 0 N
.-..- I—
H•
a) ___ ___ ___ ___ ___ ___
0
E Z
in
cm
4m *jjW
ri
HLfl Hm
CNN
El
00 00
H 0H
U
H
(
U
H
_
_.
ON
NW
Cd
Cd
0
Cd
0 0 0 t' H
0
•1llN NII NII
I
I
-
I
O\O
I
i
U
_P OU]
-H
Cd
U
a) I1
>U
-H
U]
U U]
-rlU
Cd
•H
a).
E HW
0
-Ha)
O\O
C])
El
NO CNN * H LflW CNN
NO mH
0 El HCN 00
C/)
• .
. .
. 'd Z OH 00 H 00 OH
a) 0
-'.
U] 0
Cd
a)
U
Cd
0
($]U
Cd.
-0
-H
Cl)
-4
U
H
>1
(1)
U]
r
4j
-H
U]
a)
E-
U
-H
-H
-d
a)
-iJ
-H
cow
P.I'd
U] (])a)
H
.Qbi
ida)
Cd
O
Cd
id
__
___
0
___
___
___
310
:*
Cd'
PIN
I
*
cn
—
H
C])
-H
Q)4
a)
EEl 'd
-1'd
1___
a)
oc:
rIO
• -H
qjN
-H U
t4
Cd
jiE CdW
4-)
>H
O
U q-i 4-)CCS
j
U
___
___
ZI
U)
Cl]
N
U]
U)
b
-H
N
N
V
N 0
0 0 0 0 0
t.O
H
o
r1
El
El
(I)
H
p4
*
.*
N
Cl]
*
**
*4
*
r-I
*
0
N
•
H
W
-
a)
p
w
Z :
*
O
(1)
a)
U
U]
a)
>1
H
a)
H
* — * - * -.. * -.. * —
*
* ..o *
+
Lfl
0.
+(J]
a)
oU]
.l)a)
0J-)
'a)
0
t4
rd
Op.,
-H
(1) -H
p4
rd
E
U]
Cl)
El
4 -.
El
a
-
—
NV 'DH
co• H
00 NO
NO\ r-ICfl
mm O'IH
mn
'-E
• —
cil
U]
U]'
-H .iJ
a).
r-IW
.Qrl
(.
-H a)
(da)
El
0
a)
U]
(ni
U]
.i-J
H
U]
a)
-H
U)
U
cDw
H
a)
-H
a)q-i
,a)
Eld
0
4-4
-U
U]
a)
flu
flu
0
U]
a)
flu
O\0
Lfl
a)
-U
a)
4J
a)
a)
H
ri
-H
-H
H
U)
0
U]
iii
On]
II
:
NO\ Hm
00 00
Lfl
HO NN 00
a)
4)
a)
-H
-U
0
cr:i
H
a)
U]
0
H
i'
l co N a
N c11 Lfl
H .N.H-0-
a)
p-I
U]
a)
a)
H
(1)
>1
0
0
m
if)
U]
a)
.4J
Oj
c
a)
til
Lfl
If)
a)
-H
H
if)
lC
>1
j_I
-
N
N
l
H
__ __ __ __ __ __ H
-H
-H
iii
.
+
flu
— . * -.
* a)
0 CO
NW ii;
-.- 4 —
N H n-I m
HLfl
O\O
ri
-ii
flu
-U
c;
0
-H
4-1
-H
a)
H
p-I—.
') 0
U)
a) II
I-I--0
U]
>1
.i)
>1
.4J
-H
U]
U)
a)
4-I
*
N
°co
r.
'-I
-H
Cl]
>!!
II
o
4
+
*
U]
U]
U
1I
cl-I
311
flu
U
U]
flu
U
"H
Cl]
a)
-U
0
0
-ii
U]
a)
4C
1rj
Information asymmetry about the target's assets
(TAGSTD) has a significant negative effect on the target's
returns, suggesting that bidders offer lower premia when
they are uncertain about the value of the target's assets.
However, this effect does not appear to be related to the
method of payment.
The change in the variance of the returns to the
merged firm (]JEBTINS) has a positive effect on the returns
to the target. This suggests that returns to target
shareholders are higher when the post merger risk of the
combined firm falls. This is inconsistent with the wealth
transfer hypothesis H4 predicts that the a fall in the
riskiness of the merged firm leads to a wealth transfer
from shareholders to bondholders. There is some support for
the prediction from the wealth transfer hypothesis H4 that
the equity offers provide the greatest opportunity for
wealth transfers between shareholders and bondholders. For
equity offers under the market adjusted model and the mean
adjusted model there is a negative relationship between
returns to the target shareholders and the change in the
risk of the merged firm's returns. However the result is
statistically insignificant and sensitive to the choice of
control model.
The effect of capital gains tax on the returns to the
target is very surprising. We find that the greater the
potential short run capital gains (CGAIN), the lower the
returns to the target. This result is uniform across
payment methods and is inconsistent with the capital gains
312
tax hypothesis H3. A possible explanation for this result
is that the short run capital gains in a target reflects
the efficiency with which the firm is presently being
managed. Consequently there is minimal scope for creating
post merger gains in a firm which is highly valued prior to
the bid. If there is a positive tax compensation effect in
the returns to the target, then it is dominated by the
inability of bidders to extract substantial gains from
firms which are presently well managed.
8.5.3. Bidder returns and the method of paynent
In Table 8.20, we report the results on the cross
sectional analysis of the bidder's returns.
There is support for hypothesis Hi that information
asymmetry results in investors reacting negatively to the
use of equity by the bidder. Across the whole of the
sample, we find that information asymmetry about the
bidder's shares (EIDSTD) has a negative effect on the
returns to the bidder. However the negative relation
observed between abnormal returns and the standard
deviation of the bidder's pre-bid market adjusted returns
is caused primarily by those bids which have an equity
component. In cash offers the level of information
asymmetry has a positive effect on the returns to the
bidder. This indicates that in situations where the market
believes that managers have a large amount of private
information, a cash offer is seen as a signal of positive
private information, while an offer with an equity
313
•
__
__
__
__
__
__
U)
U
w N
H 0 H 0
.
0 0 0 0 0
a)
0
H
b.
-H
-H
>1
:::'
E- 4
4_I
rj a)
' Ci
1
Ij
(1)
H
-
:*
u
m
i
H
•
m
•
cc
H
w
*
4_i
0
a)
N
H
4_
a)
4)
¼
*
__ __ __ __ __ __ H
-H
>4
Cl)
El
'Ci 'Ci
_
*
*
*
Hr
c'
m
b+
U)
*
*
+
0
4
+
N
Cl)
*
H
ç
H H N
_
_
_
_
a)
0
U)
04_i
N
a)
_
4-.
1-4
0
U)
Cl)
ftS
-
-H
-1c:
4-) a)-H
D4
(1)
H
ci)
a)
(j
C')
E
-1
o
a)-H
0
:>4)
,Q Ou
-H
I
.•-
El
4_i
a).
HO
,QH
U
-H CO
a)
4-)
L$4
0
.4
cda)
'H
4_i(
0 1El
a)
rç
NO
r1
cOO)
U)
C).i'Ci
Ci)
(i)U)
a)a)
i
4_i
0r-4
a)
(J
4-)
I
I-..-
U)
DLfl
'0
tlO•l
U)
a) rN
O HN
OH 00 H
"H
El
C/)
00
LI).
NH
Il
H
II
I
•m
I
O\°
LI)
o
a)
4_i
-
o\o
H
0
El
-i .orn H
0 F'
'Ci
U) 0
U) 0
H N
0
OH
0 r-1
I
I
I
I
_
00 or-i
I
I
00
ii
4)
U)
U)
a)
_
-H 4_ia)
4
U)
U)
-i
>1
4_i
-H
_i
U)
a)
ii
(
-
'
314
4_i
-H
Cl)
*
.
>.J1
,
••
Uç
•
Cfl
a)I
4)l
a)4
•tJ)
01
'
FEl V ___ ___ ___ ___ __ ___ ___ zi
'd
o'o
0
H
-
___ ___ ___ ___
H
a)
> ' Ci
(1)
H
H.
•N
a)
H
U)
HC:O:H
°co Nm
1. N N..
> 04_i
-H
U)
U)u
ni H4_i
H
0
4_i
0
U)
*
*
N
E
4_i
a)
oo
Qrfl0NOG'l
'Ci ___ ___ ___ ___ ___ ___
0
E
*.*.
0
4_I
OH
a)
I a)
H E
N
Ha)
___ ___ ___ ___ ___ ___ -1
()J
o
U)
'Ci
H
V
+
4_i
2
cc
°'
-H
4-)
*
U)
U)
4)
U)
4_ia)
E
__
U)
:
•
__
__
__
__
__
__
__
U]
U]
a)
r4
-1
O't
(N
b
-H
('j
H•
0
0
>1
4_i
flj
Ci)
H
000
0
0
0
*
*
E-
a)
Cl]
N
•
N
III
(j
H
•
H
>1
*
*
4_i
-1
fl
(N
(I)
4_i
__ __ __ __ __ __
w
H
-H
-H
_i
U]
(j
w
a)
Er-
ci)
-H
U]
>
a)
'ti
0
-i
O'
Oi
N
Oi
N
m
(N
a)
(N
>1
H
11)
Ui
4_i
0
4_i
0
U]
4_i
a)
H
HI!)
O(N H H
r OH
0
cl
+U]
a)
0(J)
.l_ia)
H
04_i
0
'a)
-H
1) -H
E
4_i
U]
a)
OJ4
(Li
4_i
a)
a)
-H
(Li
iii
0(l)
-'-I
a)
Cl)
I-I
04_i
(I]
(01
ID.
a):j
a)
H
0H 00
- * r
•
l HI-01: m
O0IN100IHt1 o•co
çrjcoj
(N •(N
-H a)
4_i,
4_i
0
04-i
_-. 0
0
(NO
--H
Ci) U]
(I)
a))
4_i iii
0
a)
U)
H
o
0
a)
(Li
4_i
' L' ' ' L' L
-H
a)
EIrd
11)
UI
0
>1
U)
0)
H
a)
a)
H
a)
(Li
0
4_i
>1
II
-H
C/)
*
**
4
*
4_i
(I)
CJ)LO
r-I
0
4_i
U]
a)
4-iJ
a)
-H
4-I
-H
to
a)
-H
a)4-i
lea)
rd
H
.4_i
oo rnu rn
'm o
H(N NW H (NH OH
0 -1 00 00 0 H 00
(0
4-)
a)
rd
('I
0
C')
Lfl
4_i
10
P.i'ti
Wa)
ILI
bi
(LI
p4
H
O\O
(Li
Uia)
4_i
(Li
Ui-
-H a)
U]
ci)
4_i
-H .i_i
HW
4J
H
4_i
4_i
iii
111
* -..'
N
p4
0
-H
4_i
E-
-_•
'H 0(N
0H 00 00
U]
H
a)
ot°
0
U]
a)
p4
(I)
ci)
a)
a)
op4
4_i
0
ci)
IHO\N
00 Ha)
- - 0.
og
rj
C']
z
I-I
0
-H
a)
9-i
4-i
°co
>
4_ifl
II
U]
0
-i
p4
315
U]
(Li
0
U]
ci)
4_i
0
(LI
(LI
ci
lxi
b
U]
a)
(1]
a)
H
N
'D
rn
O'
N 0 0 0
rj
0
0
r-1
2
El
a)
flj
H
LI
U)
rj
a)
N
0
0
*
**
**
ct'
N
•c11
N
>1
.1-)
N
0
U
r)
m
-H
U]
__ __ __ __ __ __ H
-H
ai
*
*
*
4_)
H.
Elk,
r
U] LO
cjl
CX)
H
Ui
a)
L*
,
0
a)
'-
El r
El
H
N
__ __ __ __ __
Cl)
m
__
>1
H
U]
r jI-1
o +
a)
oU]
-
D
0 4j
()
-
I
a)
(I]
H
a) U]
Z
H
rd
a)
E
'.D
-
OH
-
H
NN
NO
OH
El
00
r-
4J
(ii
U]
a)--i
O\
___ ___ ___ ___ ___ ___
-1
a)
a)
E
nj
Iii
0
o a)-H
a)
oU]
-H
a)
4)
a) ,
jJ
c
U]
U]
-H4J
—
a) •
'd
-"t'
a)
o
N :z
El
U)
W -
H
1
N
-1
0
'L
O\O
'.
El
Z
0
*
Lo
H
mr
OH NH mm HO
00 00 OH OH
-
C)
(i
o
04-1
4-)(t
— O
El
H
a)
N
c -,0
I
'
0
-H
-'-'
>1
a)
P'd
U]
U] a)a)
a)a) r j
>1
(1)
H
a)
••
0
H
Jl
4
a)
.1-)
Q
U]
-i
a)
tY)
'
Cl)
4-1
t44
II
H-
a)
-H
-H
(1)
0
a)
C•
(na)
r1
N
'L
H
El
,
r-1
NC
-H
-Ha)
.i..i,
rj
o
O\0
U]
rd
a)
U]
crr
rflH
t.1-H cc
U]
r1
(tI
U]
a)
-H
4J
4J
a)
__ __ __ __ __ __
r
-)
0
N
I_-
.L)
flj
,Q
**•
*
4J
0
>1
--i
WH
•-
•
'dH •N
(
II
-
•
II
I
2
-
b
N
U]I
4J1
a)4-i
..Qb
0
ci
(Q)
a)
ElEl 'd ___ ___ ___ ___ ___ ___ ___ ZI
316
'I
"l
U]
a)
N
N
component transmits a negative signal.
The change in the post-merger variance of the merged
firm's returns (DEBTINS) is positively related to the
bidder's abnormal returns. This is inconsistent with the
with the wealth transfer hypothesis H4. There is weak
support for a wealth transfer between shareholders and
bondholders under the mean adjusted model. With this model,
the abnormal returns to bidders offering equity are lower,
the bigger the fall in the post bid variance of the merged
firm's returns. However this result is unstable and
sensitive to the choice of control model.
The short run potential capital gains in the target
(CGAIN) has a negative effect on the bidder's returns, with
the strongest effect being observed in cash offers.
Although consistent with the capital gains tax hypothesis
H3, the evidence requires cautious interpretation given our
earlier result that short run capital gains have a negative
effect on the target's returns. A more realistic
interpretation is that short run capital gains have a
negative effect on the returns to the bidder because of the
limited scope for realising gains in well managed targets.
8.6. Conclusion
In this chapter, we analysed the impact which the
method of payment has on the wealth of both bidder and
target shareholders, using both univariate difference in
means tests and multiple regression analyses. Consistent
with existing evidence, we find that returns to both
317
bidders and targets are higher in cash offers than in
equity offers. We investigate the source of this higher
return to cash offers.
We find evidence that the method of payment is
associated with an information signalling effect, in which
the use of equity by the bidder conveys negative
information to the market, while the use of cash transmits
a positive signal.
The argument that compensation for CGT would result in
higher bid premium to target shareholders when cash is
offered as the method of payment is rejected. Contrary to
the CGT compensation hypothesis, we find that short run
capital gains for target shareholders lead to negative
returns in both the target and the bidder. Our results do
not support the argument that potential changes in the
variance of returns to the merged firm redistribute wealth
from shareholders to bondholders.
Bidders offering equity as the method of payment tend
to have experienced a significant increase in their share
price in the pre announcement period. However, the higher
this pre-announcement increase in the bidder's share price,
the greater is the post-announcement loss. The use of an
underwriter by the bidder as a signalling device does not
mitigate the post announcement losses of bidders offering
equity as the method of payment.
The implications of the results found in this study
and the directions for future research are discussed in the
next chapter.
318
APPENDIX 8.1.
EVENT STtJDY METHODOLOGY
For each company i, we define an abnormal return AR1 as
-
AR =
R1 is the continuously compounded realised return on
day t (dividend plus capital gains). This is calculated as'
"
R1
=Log
P1t +
.Pi,
D
t_i
= Price of company i's share at the end of trading
on day t.
D, = Dividends received on day t.
t is time defined relative to the bid announcement
date. The definition of the bid announcement date is given
in Section 4.5.
is the control rate of return which is what company
i's return would have been in the absence of the event. In
order to ensure that our results are not sensitive to the
models used in specifying the control rate of return, we
shall use three alternative models to determine C
1pd alternative method of calculating R, is to use discrete
returns, where
R
it --
+D -
pi,t-1
We have choose to use logarithmic returns, because when linking subperiods together to form longer event periods, logarithmic returns can
be simply added together, whereas discrete returns cannot be easily
manipulated in this way. Additionally logarithmic returns are more
likely to be normally distributed.
319
Model 1 : The market model
= a. + I1 jRm t + €jt
where
Rmt is the continuously compounded realised return on
day t for the market index.
o is the regression constant obtained from regressing
on Rmt. This measures the mean return over the estimation
period which is not explained by the market.
is the regression co-efficient obtained from
regressing
on Rmt . This is a measure of the sensitivity
of firm i's return to the market return (ie, systematic
risk)
€,
is an error term with a mean of zero and a constant
variance.
The values of c and € are obtained by regressing R 1 on
R for the 250 trading days (if possible) or fewer
observations (with a minimum of 120 days) beginning at t
= -290, where t = 0 is the event day.
Model 2 : The market adjusted model
The control rate of return for any firm in the event
period 2 is the return on the market index for that day.
C
= Rmt
2Event period refers to the number of days over which abnormal
returns centred on the event day are cumulated in order to estimate the
impact of the event on shareholder wealth. In this thesis event period
is -40 to -,-40 days centred on the acquisition announcement date.
320
This model is equivalent to the market model where for
all firms
c
= 0 and
f3
= 1.
Model 3 : The mean adjusted return model
The control rate of return for any firm for a day in
the event period is the mean daily return of the firm over
the estimation period (ie, -290 to -41 days)
t=-41
E
C - t=-290
it
250
R1
This model assumes that the expected return for company ±
is a constant that can vary across firms. The model would
be accurate if the risk free rate, risk premia 3 and a
company's systematic risk are constant over time.
Cumulative abnormal returns
For each day in the event period, the abnormal returns
are averaged across firms to produce the sample average
abnormal return for that day ARE.
N
A R = 1=1
N
3 me risk premia refers to the difference between the return on
the market index and the risk free rate. For a fuller discussion see
Copland & Weston (1983: Chapter 7)
321
Where N is the number of firms in the sample.
The average cumulative abnormal returns (CAR) for N
firms over a number of days from ti to t2 is calculated by
summing AR over the period from ti to t2.
t = t2
CARt:i
AR
t2 =
t=tl
The null hypothesis examined under the event study is
that AR = 0 and CAR = 0. The test statistic under the
null hypothesis can be calculated based on the assumption
of either cross-sectional independence or cross-sectional
dependence in the abnormal returns.
Test statistic assuming cross sectional independence
The test statistics described below assume that the
cross sectional correlation between the abnormal returns
for any two firms i and j equals zero. This assumption
would be valid where the event dates for the sample
takeovers are diffusely spread over a long period of
calender time.
Each abnormal return AR. 1 is divided by its estimated
standard deviation sD(AR) to yield the standardised
abnormal return SAR.
-
AR
SA Rr - sD(A R)
322
where
I t=-41
I E
SD(A R1) =
(AR1-)2
t=-290
249
and
t=-41
:
= t=-290
250
For a sample of N firms, the test statistic for day t
is the standardised abnormal return for day t SAR
N
SARtND 1=1
For a large N, SAR
has a normal distribution N(0,l).
For tests over the multi-day interval ti to t2, the
test statistic is the standardised cumulative abnormal
return SCARI2IND.
t= t2
E
SARtID
- ttl
b1,b2 - ______________
%/t2 - ti + 1
323
Test statistics assu.ming cross sectional dependence
Where the event dates for the takeovers are clustered
in calender time, or the firms concerned are in the same or
related industries, then the assumption of cross sectional
independence in the abnormal returns would be violated.
Test statistics which adjust for cross sectional dependence
are described below.
The test statistic for event day t is
AR
t
SD(A R)
where
t=-41
(AR-)2
SD(A R)
=
t=-290
249
and
t= -41
AR
= b=-290
250
For tests over the multi day interval ti to t2, the
test statistic is
324
CA Rt:it2
= __________________________
SD(A R) * /t2 - ti + i
The problem of thin trading
The market model estimates of beta can be subject to
a downward estimation bias if shares are thinly traded.
Thin trading occurs, when th price recorded at the end of
a trading day for a security relates to a transaction
occurring well before that trading day. IJimson (1979) shows
that the estimated betas of infrequently traded securities
rise as the returns measurement4 interval rises. This
implies that when using daily returns the market model
estimates of beta for thinly traded shares have a downward
bias, while for frequently traded shares the bias is
upward. Biased beta estimates will result in biased
estimates of abnormal returns and consequently misspecified
results in an event study. A number of approaches have been
suggested in the literature to correct f or such thin
trading bias. (Scholes and Williams, 1977; Dimson, 1979;
Fowler and Rorke, 1983)
Scholes and Williams (1977) show that under the
assumption that a transaction takes place in every
measurement interval (ie, a security does not have any
missing observation between day -1 and +1) a consistent
estimate of beta is
4 ie, betas calculated using monthly returns are higher than betas
calculated using daily returns for infrequently traded shares.
325
p
= p- 1
+ 30 + p+l
1 + 2 p1
where
/3'
is the slope coefficient in a simple regression of R
against
3O
is the slope coefficient in a simple regression of
against R
/3
is the slope coefficient in a simple regression of Rft
against Pmt+i
p1
is the first order serial correlation of the market
index.
The Dimson (1979) aggregated coefficient estimator
does not require that a transaction take place in every
measurement interval. The ]Jimson estimator is obtained by
regressing the security return on day t against leading,
synchronous and lagged returns of the market index, in
order to obtain a set of slope coefficients, f3 which are
then summed to give an unbiased estimate of true beta.
n
PDIM
E
t=-n
P
where j3, t = -n, . . .,0, . . . ,n are slope coefficients in a
multiple regression of the return on the security in period
t against the return on the market index in period tn,...,O,...,t-i-n.
326
Fowler and Rorke (1983) suggest a correction to the
Dimson aggregated co-efficient method to equate it to the
Scholes and Williams estimator. Fowler & Rorke show that
when a security skips a single price observation, the
correct beta estimate is
=
fl
-2
FR
+ p_i + 13° +
1
+
2(p
+
+
13+2
p2)
where
f
is the slope co-efficient in a simple regression of
the security return in period t on the return on the
market in period t+n.
p1
is the first order serial correlation coefficient of
the market index.
is the second order serial correlation coefficient of
the market index.
The
FR
expression can be generalised for securities that
skip two or more consecutive observations.
13
PFR
1
+
2(p
1 + p 2 + . . . .+
p)
To correct for thin trading in this study, we tried the
correction procedures suggested by Scholes and Williams
(1977), Fowler and Rorke (1983) and Dimson (1979) using
alternative combinations of lead and lagged market return
terms. The results of these various procedures are shown in
Table A8.l.
327
Table A8.l.
Average betas using different procedures to correct for
thin trading.
Number of Number of Targets
Lag Terms
Lead
Te rms
Bidders Targets and
Bidders
Scholes and Williams correction procedure.
1
1
0.6129
0.8483
Fowler and Rorke correction_procedure.
0.7306
____________
1
1
0.6537
0.9079
0.7808
2
2
0.6715
0.9238
0.7977
3
3
0.6752
0.9268
0.8010
4
4
0.6733
0.9151
0.7942
5
5
0.6779
0.9089
0.7934
Dimson_correction_procedure
__________ _____________
0
0
0.5152
0.7432
0.6292
0
1
0.4885
0.7270
0.6077
0
2
0.4857
0.7307
0.6082
0
3
0.4781
0.7252
0.6016
0
4
0.4569
0.7014
0.5792
0
5
0.4349
0.6742
0.5546
0
6
0.4309
0.6717
0.5513
1
0
0.6504
0.8761
0.7632
1
1
0.6257
0.8608
0.7432
1
2
0.6216
0.8662
0.7439
1
3
0.6106
0.8578
0.7342
1
4
0.5915
0.8321
0.7118
1
5
0.5697
0.8035
0.6866
1
6
0.5619
0.7978
0.6798
328
Table A8.1. (Continued)
Average betas using different procedures to correct for
cnincraaing.
__________
Number of Number of Targets
Bidders 1 Targets and
Lag Terms
Bidders
Lead
__________
Terms __________ __________ ____________
Dimson correction procedure
2
0
0.6867
0.9260
0.8064
2
1
0.6619
0.9116
0.7867
2
2
0.6'557
0.9173
0.7865
2
3
0.6454
0.9095
0.7774
2
4
0.6269
0.8828
0.7548
2
5
0.6049
0.8550
0.7299
2
6
0.5978
0.8474
0.7226
3
0
0.7035
0.9390
0.8213
3
1
0.6822
0.9274
0.8048
3
2
0.6762
0.9326
0.8044
3
3
0.6679
0.9247
0.7963
3
4
0.6511
0.8998
0.7755
3
5
0.6290
0.8723
0.7507
3
6
0.6203
0.8646
0.7424
4
0
0.7288
0.9607
0.8448
4
1
0.7066
0.9483
0.8275
4
2
0.7010
0.9526
0.8268
4
3
0.6924
0.9429
0.8177
4
4
0.6760
0.9187
0.7973
4
5
0.6542
0.8919
0.7731
4
6
0.6447
0.8845
0.7646
5
0
0.7524
0.9689
0.8607
5
1
0.7297
0.9555
0.8426
5
2
0.7236
0.9609
0.8423
329
Table A8.l. (Continued)
Average betas using different procedures to correct for
thin trading.
Number of Number of Targets
Lag Terms
Lead
Terms
Bidders Targets and
Bidders
Dimson correction procedure
5
3
0.7135
0.9512
0.8324
5
4
0.6964
0.9299
0.8132
5
5
0.6765
0.9039
0.7902
5
6
0.6675
0.8976
0.7825
6
0
0.7616
0.9678
0.8647
6
1
0.7402
0.9556
0.8479
6
2
0.7339
0.9617
0.8478
6
3
0.7229
0.9532
0.8381
6
4
0.7070
0.9332
0.8201
6
5
0.6872
0.9072
0.7972
6
6
0.6770
0.8991
0.7881
The final model was selected on the basis of maximum
average sample beta. The highest average beta for bidders
and targets occurs with the Dimson correction procedure
using 6 lags and 0 lead terms. The adjusted average betas
in our sample are 0.9678 for bidders and 0.7616 for
targets. This is comparable to the average betas in Franks
and Harris (1989), they reported thin trade adjusted betas
of 0.92 for bidders and 0.854 for targets. Targets have
lower betas than bidders probably because they are smaller
and hence more prone to the thin trading problem.
Choice of event window
In choosing the event window over which abnormal
330
Table A8.2.
Number of event days between the announcement date and the
unconditional date by method of payment.
________________ MEAN
MEDIAN STD DEV No OF OBS
Whole sample
33.19
30.71
13.47
504
Cash offers
34.98
32.14
14.64
83
Equity offers
30.23
29.64
9.63
88
Cash and equity
35.57
32.14
14.57
111
Cash or equity
32.50
30.00
13.54
F-Statistic
Note:
3.30
' '
Significant at 1%,
respectively.
5
222
**
and l0
levels
returns are cumulated we wish to choose a window which is
sufficiently wide to allow us to capture all the valuation
effects associated with the takeover announcement and yet
not too wide that it introduces noise into the data.
The distribution of the number of event days between
the announcement and the unconditional date is reported in
Table A8.2.
Under the City Takeover Code, the offer document must
be posted within 28 calendar days (20 event days) of the
announcement of an intention to bid. In general the bid
must be declared unconditional or defeat conceded within 60
calendar days (43 event days) of the posting of the offer
document. This implies that from the announcement date the
maximum length of time for the completion of a bid is 88
calendar days (63 event days).
Additionally offer terms cannot be revised after 45
calendar days (32 event days) from the posting of the offer
331
document. This implies that most of the information (ie,
any revision of terms) associated with the bid will have
been revealed by 73 calendar days (52 event days) from the
announcement date. Since the majority of bids are friendly
bids the time to completion is usually less than the
maximum of 88 calendar days.
The average time to completion is 33 event days. By
cumulating returns up to 40 event days after the
announcement date, there is a reasonable probability of
capturing any post announcement valuation effects of the
bid.
Previous studies (Shih & Suk, 1992; Niendorf &
Huffman, 1992; Holland & Hodgkinson, 1994) have shown that
there is some information leakage prior to the announcement
of the takeover. The CAR graph for the target (Figure 8.2)
shows that the target's share price begins to rise as early
as 25 days before the bid announcement date. In order to
ensure that we have a reasonable probability of capturing
any pre announcement information leakage we start
cumulating the abnormal returns 40 days before the
announcement date.
The F-statistic in Table A8.2 shows that the method of
payment does have an impact on the average time taken to
complete an acquisition. Table A8.3 reports statistical
significance of the difference in means f or the average
time to completion f or the different methods of payment.
Offers with a cash component take significantly longer to
complete than equity offers. This is contrary to the
332
Table A8.3
Pairwise comparison of the number of event days between the
announcement date and the unconditional date for different
methods of payment.
t-statistics are calculated assuming unequal group
variances where the null hypothesis of equal group
variances is rejected at the 10% level. Otherwise tstatistics are calculated assuming equal group variances.
Cash Offer
Cash or
Equity
Cash &
Equity
Equity
Offer
1.39
-0.29
2 .49"
-1.90'
1.66'
Cash or
Equity
Cash &
Equity
3 . 10"+
Notes:
1) " " ' Significant at 1%,
respectively.
5%,
10% levels
2) Positive t-statistic implies that the mean for the
payment method in the vertical column is higher.
evidence in the US, where cash offers were found to have a
shorter time interval between the announcement and
unconditional dates (Wansley, Lane & Yang, 1983).
333
CEAPTER 9.
SUMMARY, CONCLUSIONS AND IMPLICATIONS.
9.1. Introduction
As stated in Chapter 1, the broad objectives of this
study were to examine the determinants of the method of
payment in corporate acquisitions and its effects on
shareholder wealth. In the light of these objectives we
formulated (Section 1.4) three research questions:
1) What factors determine the method of payment used
by bidders in corporate acquisitions?
2) How do target shareholders choose between cash or
equity when the bidder has offered "equity with a cash
alternative" as the method of payment?
3) Why are bid premia higher in cash offers than in
equity offers?
We also wished to examine whether the method of
payment and the choice of accounting policy were jointly
and simultaneously determined. This implied that we would
have to study the determinants of the accounting policy in
corporate acquisitions.
In this chapter we summarise the results of our
analyses as regards our objectives and discuss the
implications of these results for investors, managers,
financial advisers, policy makers, related parties and
future research.
334
9.1. Determinants of the method of payment
In Chapters 4 and 5, we empirically examined the
determinants of the method of payment, using both a
simultaneous and single equations framework. From our
models we were able to identify a number of variables which
could explain the cross sectional variations in the types
of payment currency used by bidders. In this regard, we
consider that we have succeeded in our first objective of
understanding the factors which explain the method of
payment used in corporate acquisitions. We have been able
to demonstrate that the choice of payment method in
corporate acquisitions is not a random decision but one
which is motivated by rational economic considerations.
Using a simultaneous equations model, we find that
while the method of payment is a major determinant of the
accounting policy, the reciprocal effect of accounting
policy on the payment currency is not significant. With a
single equation model both effects are significant. This
result demonstrates the importance of recognising the
simultaneity problem between the method of payment and the
choice of accounting policy. Additionally this result
suggests that for UK bidders, the primary decision variable
is the financing method, not accounting policy.
The result that the choice of accounting policy does
not influence the method of payment reflects the fact that
the ability to write off goodwill against net assets and
the availability of merger relief in the UK has eroded the
distinction between acquisition and merger accounting. A
335
desirable characteristic for accounting rules is that they
should not influence the substantive commercial or
financing decisions of managers. We know that the method of
payment affects the wealth of shareholders. It is therefore
important that the accounting rules, which have no
substantive cash flow implications, do not alter
acquisition financing decisions which are based on economic
fundamentals. On this accounting neutrality test our
results show that UK accounting rules are presently
effective.
In the UK, the Accounting Standards Board (ASE) is
currently engaged in revising the accounting rules for
business combinations and for the treatment of goodwill
(ASB Discussion Paper issued in December 1993 "Goodwill and
Intangible Assets") . Two possible alternatives have
emerged:
(1) capitalise and amortise over a predetermined life;
(2) capitalise with annual review and only write off
if there is a permanent decline in the value of
goodwill.
Of these two proposals, the former with a regular charge to
the income statement is more likely to undermine accounting
neutrality if managers seek to avoid earnings dilution.
Under the second proposal, such dilution is of less
concern.
Our results also show that cash resources, free cash
flow, the investment opportunity set, insider control,
conditions in the capital markets and information asymmetry
336
all influence the choice of method of payment. These
further results are robust to whether simultaneous or
single equation framework is used and to sampling
variations. The evidence that these variables affect the
managerial preference on how to finance an investment
indicates that these factors can be used to enhance the
power of models which seek to explain the observed capital
structure of the firm.
The result that managerial control of votes influences
the acquisition financing decision can help managers and
investors resolve some of the information asymmetry problem
associated with seasoned equity offers. Given the potential
cost to managers of losing control as a result of issuing
new equity, the fact that a rights issue is undertaken by
a firm with high managerial ownership could signal to the
market that managers are not attempting to issue overvalued
equity. The size of managerial control of votes can
therefore act as a mechanism by which managers can signal
that a new issue of equity is being driven by fundamentally
sound economic decisions.
9.2. Determinants of the accounting policy
As an important by-product of studying the joint
determination between the method of payment and the choice
of accounting policy, we examined the determinants of
managerial preference between merger and acquisition
accounting. We were unable to develop a model which had
sufficient power to discriminate between the users of
337
merger and acquisition accounting. The data seem to
indicate that managers do not have strong preferences for
any particular accounting method in the UK (see Table 5.3,
where only 25 of bidders qualified to use merger
accounting actually chose merger accounting) . This evidence
questions the relevance and usefulness of Financial
Reporting Standard 6 (see Section 3.2.2)
The central objective of FRS 6 is to restrict the
circumstances under which merger accounting is used in
accounting for business combinations'. FRS 6 appears to
have been driven by a misguided belief that a large number
of companies violated the spirit of SSAP 23 by using merger
accounting in circumstances when its use was wholly
inappropriate. It is certainly true that some companies who
should have used merger accounting employed arrangements
such as vendor rights (see Section 3.2.1) to ensure
compliance with the SSAP 23 rules for using merger
accounting. However considering that the majority of
bidders who qualified to use merger accounting actually
chose acquisition accounting and the fact that only lO of
all bidders used merger accounting, we must question the
validity of the ASE's argument that the issue of FRS 6 is
justified by the widespread concern that the SSAP 23
conditions were too readily circumvented 2 . As there is very
limited use of merger accounting the violations of SSAP 23
could not have been widespread. It is highly unlikely that
1 See paragraph 1 of FRS 6.
2 See Financial Reporting Exposure Draft (FRED) 6.
338
the new rules contained in FRS 6 would lead to any
significant reduction in the number of companies using
merger accounting.
Managerial indifference between acquisition and merger
accounting methods can be attributed to the fact that in
the UK bidders are not compelled to amortise goodwill
through the P&L account. The neutrality of accounting rules
on financing decisions could be lost if writing off
goodwill through the balance sheet reduces the bidder's net
assets and causes a breach of loan covenants. We already
have evidence in Chapter 5 that the size of potential
goodwill to be written off has a weakly significant impact
on the choice of accounting policy (see Sections 5.2.2 and
5.5) . If the influence of goodwill on the choice of
accounting policy is to be minimised then lenders should be
encouraged to include goodwill written off on an
acquisition in the calculation of net worth. Citron (1992b)
found that 27 of loan agreements in his sample already
included such a provision. Given the significant proportion
(73%) of Citron's sample which did not provide for the
inclusion of goodwill written off in the calculation of net
worth, lenders need to be educated on the benefits of
maintaining the neutrality of accounting rules, so that
this practice becomes universal.
9.3. The choice of payment method by target shareholders
One of the issues confronting target shareholders in
an equity offer is the valuation of the bidder's paper. A
339
cash offer has a known value which is established at the
date of the acquisition, whereas, because shares are risky
securities the value of the equity offer may be affected by
the post merger performance of the merged firm. If the post
merger performance of the merged firm is high (low) then
the value of an equity offer will be high (low) . Target
shareholders face a dilemma in deciding whether to accept
a cash offer which has a known value or an equity offer
whose final value is subject to the uncertain future
performance of the bidder. In Chapter 6 we examined two
alternative perspectives on how target shareholders make
decisions on whether to accept cash or equity as the method
of payment.
The first perspective, based on the efficient market
hypothesis (EMH), suggests that target shareholders should
accept the present market price as a fair reflection of all
publicly available information and that uncertainty about
the post merger performance of the merged firm simply
reflects the normal risks associated with investing in any
security. Under this perspective, target shareholders
should use the present market price of the bidder as the
basis for valuing the equity offer. The choice by target
shareholders between cash and equity should be based
primarily on a comparison of the monetary values of the two
offers.
The second perspective, which implies some form of
mispricing of shares by the market, suggests the valuation
of the bidder's equity offer is only determined ex post
340
over time as information about the post merger performance
of the merged firm is revealed. Under this perspective,
factors which influence the post merger performance of the
merged firm, will influence the choice by target
shareholders of whether to accept the cash or equity offer.
The empirical validity of the two perspectives was
tested in Chapter 6 using a sample of 130 takeovers in
which the bidder had offered "equity with a cash
alternative" as the method of payment. The proportion of
target shareholders accepting the equity offer was
regressed on the difference in value between the equity and
cash offers and proxies for the economic fundamentals of
the acquisition. The EMH perspective predicts that while
acceptance of the equity offer should be positively related
to the difference in value between the equity and cash
offers it should not be influenced by the economic
fundamentals of the acquisition. The market mispricing
perspective predicts that acceptance of the equity offer
should be related to both the difference in value between
the equity and cash offers and the economic fundamentals of
the acquisition.
Our results show that target shareholders behave in a
manner which is consistent with a belief in the EMH i.e,
that investors believe that share prices are an accurate
reflection of economic realities. Since this test of the
efficiency of the stock market is not based on any
predictions about the relevance of the capital asset
pricing model (CAPM) this is evidence in favour of the EMH
341
which is not affected by the joint hypothesis problem3
which has weakened some of the previous tests of the EMil
(Summers, 1986)
This has the practical implication that target
shareholders can rely on market prices as the basis of
valuing equity offers. This indicates that the valuation of
equity offers is not as arduous or complex as the
information asymmetry literature would suggest. Information
signalling models (Fishman, 1989; Berkovitch & Narayanan,
1990) which suggest that the value of equity offers is
determined ex-post are probably based on assumptions which
do not reflect the beliefs of investors.
9.3.1. Difference in value between equity and cash offers
In Chapter 6, we examined the factors affecting the
difference in value between equity offers and the
associated cash alternative. This is an issue which has not
been examined in any previous study. Consistent with the
argument that during the offer period a put option giving
target shareholders the right to sell their shares to the
bidder is imbedded in a cash offer while in an equity offer
the equivalent claim against the bidder is an "option to
exchange" assets we found that the difference in value
between an equity offer and the cash alternative
iS:
positively related to the standard deviation of the
3 Previous tests which seek to show that trading rules cannot earn
abnormal returns are jointly testing the hypothesis of no abnormal
returns and the hypothesis that the control rate of return is correctly
specified.
342
ratio of the bidder's share price to the target's share
price and;
negatively related to the standard deviation of the
target's share price.
These results have implications for the conduct and
regulation of takeovers. Our results show that for both
equity and cash offers, the target shareholders have
valuable claims against the bidder. The City Code requires
that bids must be left open for a minimum period of 21 days
after the posting of the offer document. Additionally it is
possible, dependent on the bidder's actions that a maximum
period of 88 days can in general elapse between the
announcement date and the unconditional date. The longer
the period f or which the offer is kept open, the more
valuable are the options imbedded in the offer. Valuable
options given to target shareholders by the bidder
effectively represent part of the premium paid to acquire
control (Margrabe, 1978) . A long offer period increases the
value of these options hence increasing the bid premium.
The larger the bid premium paid to target shareholders the
smaller the share of merger gains accruing to bidder
shareholders. A long offer period could at the margin
render an acquisition unprofitable for the bidder by
increasing the costs of the bid. It is therefore important
that in setting any minimum offer period both bidders and
regulators should recognise this possible adverse effect
associated with the length of the offer period.
343
9.4. Difference in bid premium between cash and equity
offers
In Chapter 8, we tested some of the popular
explanations which have been advanced in the literature to
explain the widely observed higher returns to cash offers
than equity offers. The capital gains tax compensation
hypothesis and the wealth redistribution hypothesis were
rej ected.
We found that the method of payment does have an
information signalling effect. The use of equity by firms
whose managers the market believes have a high amount of
private information is treated as bad news by the market.
This is the first time that this effect has been documented
in the literature. Additionally, the use of underwriters by
bidders is ineffective as a method of signalling the
quality of private information possessed by managers. This
indicates that managers who wish to signal good news to the
market should rely on the method of payment rather than the
use of underwriters.
9.5. Issues for further research
In this study we have shown that it is possible to
identify systematic factors which influence the method of
payment used in corporate acquisitions. However in our
analysis of the determinants of the method of payment, we
have concentrated exclusively on the type of consideration
offered to target shareholders. In the majority of
acquisitions in our sample, the method of payment and the
344
method of financing are equivalent. In acquisitions where
the method of payment is cash which is funded primarily
from the bidder's own coffers, it is feasible that the cash
offer may have been financed by a previous issue of shares
or debt.
Since our results have already established that the
liquidity of the bidder is an important determinant of the
method of payment, a logical extension to the present study
would be an investigation of the source of financing for
cash offers. The results from such a study would complement
the findings of this project and enrich our understanding
of the factors which influence managerial preferences for
particular methods of financing investments.
In our study of the determinants of accounting policy
in corporate acquisitions we found that the use of merger
accounting was negatively correlated with the bidder's
gearing. This is a result which is inconsistent with the
accepted theory and empirical evidence that high levels of
gearing results in the choice of income increasing
accounting policies. We observe that the existing
literature on the relationship between gearing and the
choice of accounting policy is based largely on US data. We
suggest the possibility that the institutional
relationships between lenders and borrowers in the UK may
imply that managers of highly geared firms are constrained
by lenders in their flexibility to choose income increasing
accounting policies. While this explanation is plausible,
it requires further scrutiny.
345
Our results show that in the UK, the possibility of
writing off goodwill against reserves, the availability of
merger reserve and the use of a dangling debit all combine
to reduce managerial incentives for preference between
merger and acquisition accounting. This result combined
with the fact that about 6O (see Table 5.3) of takeovers
do not qualify f or using merger accounting, suggest that it
might be worth exploring whether a viable project can be
developed to examine if there are any economic incentives
for managers choosing among goodwill write off to reserves,
use of the merger reserve or the use of a dangling debit by
bidders who do not qualify for the merger accounting
method.
346
BIBLIOGRAPHY.
Afshar, K.A., Taffler, R.J. and Sudarsanam, p .s 1 (1992) 'The effect of corporate divestments on shareholder wealth:
The UK experience.', Journal of Banking and Finance 16 Pg
115 - 135.
Amihud, Y., Lev, B. and Travios, N.G., (1990) 'Corporate
control and the choice of investment financing: The case of
corporate acquisitions.', Journal of Finance 45 Pg 603 616.
Anderson, J.C. and Louderback, J.G., (1975) - 'Income
manipulation and purchase-pooling: Some additional
results.', Journal of Accouiiting Research Pg 338 - 343.
Archibald, T.R., (1972) - 'Stock market reaction to the
depreciation switch back.', The Accounting Review Pg 23 30.
Asquith, P. and Kim, E.H., (1982) - 'The effects of merger
bids on the participating firms' security holders.',
Journal of Finance 37 Pg 1209 - 1228.
(1983) - 'Merger bids, uncertainty and
Asquith, p .,
stockholder returns.', Journal of Financial Economics 11 Pg
51 - 83.
Asquith, P., Bruner, R.F. and Mullins, D.W., (1983) - 'The
gains to bidding firms from mergers.', Journal of Financial
Economics 11 Pg 121-139.
Asquith , P. and Mullins, D., (1986) - 'Equity issues and
offering dilution.', Journal of Financial Economics 15 Pg
61 -90.
Auerbach, A.J. and Reishus, D., (1988) - 'The effects of
taxation on the merger decision.', In Corporate Takeovers:
Causes and Consequences. Edited by Auerbach, A.J.
University of Chicago Press.
Ball, R., (1972) - 'Changes in accounting techniques and
stock prices.', Supplement to the Journal of Accounting
Research pg 1 - 38.
Balvers, R.J., McDonald, B. and Miller, R.E., (1988) 'tJnderpricing of new issues and the choice of auditor as a
signal of investment banker reputation.', The Accounting
Review 63 Pg 605 - 622.
Barclay, M. and Litzenberger, R., (1988) - 'Announcement
effects of new equity issues and the use of interday price
data.', Journal of Financial Economics 21 Pg 71 - 99
347
Bayless, M.E. and Diltz, J.D., (1991) - r The relevance of
asymmetric information to financing decisions.', Journal of
Business Finance & Accounting 18 Pg 331 - 344.
Beatty, R.P., (1989) - 'Auditor reputation and the pricing
of initial public offerings.' The Accounting Review 64 Pg
693 - 709.
Beatty, R.P. and Ritter, J.R., (1986) - 'Investment banking
reputation, and the underpricing of initial public
offerings', Journal of Financial Economics 15 Pg 213 - 232.
Beicher, A., (1988) - 'A review of techniques used in
merger evaluation.', Department of Economics and Management
Science. University of Keel. Working Paper No 88-34.
Beicher, A., (1989) - 'The event study methodology as
applied to merger evaluations: A review of recent
developments.', Department of Economics and Management
Science. University of Keele. Working Paper No 89-8.
Berkovitch, E. and Narayanan, M.P., (1990) - 'Competition
and the medium of exchange in takeovers.', The Review of
Financial Studies, 3-2, Pg 153 - 174.
Ehagat, S., Brickley, J.A. and Lowenstein, U., (1987) 'The pricing effects of interfirm cash tender offers.',
Journal of Finance 42 Pg 965 - 985.
Black, F. and Scholes, M., (1973) - 'The pricing of options
and corporate liabilities.', Journal of Political Economy
81 Pg 637 - 654.
Boocook, G., (1990) - 'The financing of mergers and
acquisitions - A review of recent developments.', Managerial
Finance Vol 16 Number 4.
Booth, J.R. and Smith, R.L., (1986) - 'Capital raising,
underwriting and the certification hypothesis.', Journal of
Financial Economics 15 Pg 261 - 281.
Bowen, R., Noreen, E. and Lacey, J., (1981) - 'Determinants
of the corporate decision to capitalise interest.', Journal
of Accounting and Economics 3 Pg 151 - 179.
Bowers, H.M. and Miller, R.E., (1990) - 'Choice of
investment banker and shareholders' wealth of firms
involved in acquisitions.', Financial Management Winter Pg
34 - 44.
Bradley, M., Desai, A. and Kim, E.H., (1983) - 'The
rationale behind interfirm tender offers: InformatIon or
Synergy?', Journal of Financial Economics 11 Pg 183 - 206.
348
Bradley, M., Desai, A. and Kim, E..H., (1988) - 'Synergistic
gains from corporate acquisitions and their division
between the stockholders of target and acquiring firm.',
Journal of financial Economics 22. Pg 3 - 40.
Brealey, R. and Myers, S.,
(1984) - 'Principles of
corporate finance, 2nd edition.', McGraw Mill Book Company.
Brickley, J.A., Lease, R.C. and Smith, C.W., (1988) 'Ownership structure and voting on antitakeover
amendments.', Journal of Financial Economics 20 Pg 267 291.
Brown, C. and Medoff, J.L., (1988) - 'The impact of firm
acquisitions on labour.', 131 Corporate Takeovers: Causes
and Consequences. Edited by Auerbach, A.J., University of
Chicago Press.
Brown, D .T. and Ryngaert, M.D., (1991) - 'The mode of
acquisition in takeovers: Taxes and asymmetric
information.' Journal of Finance 46 Pg 653 - 669.
Brown, S.J. and Warner, J.B., (1980) - 'Measuring security
price performance.', Journal of Financial Economics 8 Pg
205 - 258.
Brown, S.J. and Warner, J.B., (1985) - 'Using daily stock
returns. The case of event studies.', Journal of Financial
Economics 12 Pg 3 - 31.
Bruner, R.F., (1988) - 'The use of excess cash and debt
capacity as a motive for merger.', Journal of Financial and
Quantitative Analysis 23 Pg 199 - 217.
Canton, W.T., Guilkey, D.K., Harris, R.S. and Stewart,
J.F., (1983) -An empirical analysis of the role of the
medium of exchange in mergers.', Journal of Finance 38 Pg
813 - 826.
Choi, D., (1991) - 'Toehold acquisitions, shareholder
wealth, and the market for corporate control.', Journal of
Financial and Quantitative Analysis. 26 Pg 391 - 407.
Chow, G.C., (1960) - 'Tests of equality between subsets of
coefficients in two linear regression models.',
Econometrica Pg 591 - 605.
Chown, J., (1990) - 'Taxation and merger policy', The David
Hume Institute Occasional Paper No 26.
Christie, A.A, (1990) - 'Aggregation of test statistics:
An evaluation of the evidence on contracting and size
hypothesis.', Journal of Accounting and Economics 12 Pg 15
- 36.
349
Chung, K. and Charoenwong, C., (1991) - 'Investment
options, assets in place and the risk of stocks.',
Financial Management 20 Pg 21 - 33.
Citron, D.B., (l992a) - 'Financial ratio covenants in UK
bank loan contracts and accounting policy choice.',
Accounting and Business Research 22 Pg 322-336.
Citron, D.B., (1992b) - 'Accounting measurement rules in UK
bank loan contracts.', Accounting and Business Research,
Winter Pg 21-30.
Citron, D.B., (1992c) - 'Accounting covenants in UK public
debt contracts.', Unpublished Manuscript, City University
Business School.
Collins, D.W. and Kothari, S.P., (1989) - 'An analysis of
intertemporal and cross-sectional determinants of earnings
response coefficients.', Journal of Accounting and
Economics 11 Pg 143 - 181.
Copeland, T.E. and Weston, J.F., (1983) - 'Financial theory
and corporate policy, second edition.', Addison-Wesley
Publishing Company Inc.
Copeland, R.M. and Wojdak, J.F., (1969) - 'Income
manipulation and the purchase-pooling decision. ', Journal
of Accounting Research Pg 188 - 195.
Cotter, J.F. and Zenner, M., (1994) - 'How managerial
wealth affects the tender offer process.', Journal of
Financial Economics 35 Pg 63 - 97.
Cowan, A.R., (1992) - 'Nonparametric event study tests.',
Paper Presented To The 41st Annual Meeting of The Midwest
Finance Association, March 1992, Chicago, Illinois.
Daley, L.A. and Vigeland, R.L., (1983) - 'The effects of
debt covenants and political costs on the choice of
accounting methods: The case of accounting for R&D costs.',
Journal of Accounting and Economics 5 Pg 195 - 211.
Datta, D.K., Pinches, G.E. and Narayanan, V.K., (1992) 'Factors influencing wealth creation from mergers and
acquisitions: A meta-analysis.', Strategic Management
Journal 13 Pg 67 - 84.
DeBondt, W.F.M. and Thaler, R., (1985) - 'Does the stock
market overreact?', Journal of Finance 40 Pg 793 - 805.
DeBondt, W. F.M. and Thaler, R., (1987) - 'Further evidence
on investor overreaction and stock market seasonality.',
Journal of Finance 42 Pg 557 - 581.
350
Demsetz, H., (1983) - 'The structure of ownership and the
theory of the firm.', Journal of Law and Economics 26 Pg
375 - 390.
Demsetz, H. and Lehn, K., (1985) - 'The structure of
corporate ownership: causes and consequences.', Journal of
Political Economy Pg 1155 - 1177.
Dennis, D.K. and McConnell, J.J., (1986) - 'Corporate
mergers and security returns.', Journal of Financial
Economics 16 Pg 143 - 187.
Dhaliwal, D.S., (1980) - 'The effects of the firm's capital
structure on the choice of accounting methods.', The
Accounting Review Pg 78 - 81.
Dhaliwal, D.S., Salamon, G.L. and Smith, E.D., (1982) 'The effects of owner versus management control on the
choice of accounting methods.', Journal of Accounting and
Economics 4 Pg 41 - 53.
Dierkens, N., (1991) - 'Information asymmetry and equity
issues.', Journal of Financial and Quantitative Analysis 26
Pg 181 - 199.
Dimson, E., (1979) - 'Risk measurement when shares are
subject to infrequent trading.', Journal of Financial
Economics 7 Pg 197 - 226.
Dimson, E. and Marsh, P.R., (1983) - 'The stability of UK
risk measures and the problem of thin trading.', Journal of
Finance 38 Pg 753 - 783.
Drtina, R.E. and Largay, J.A., (1985) - 'Pitfalls in
calculating cash flow from operations.', Accounting Review
60 Pg 314 - 326.
Duke, J.C. and Hunt, H.G., (1990) - 'An empirical
examination of debt covenant restrictions and accounting
related debt proxies.', Journal of Accounting and Economics
12 Pg 45 - 63.
Eckbo, B.E. and Langhor, H., (1989) - 'Information
disclosure, method of payment and takeover premiums.',
Journal of Financial Economics 24 Pg 363 - 403.
Eckbo, B.E. and Giammarino, R.M., Heinkel, R.L., (1990) 'Asymmetric information and the medium of exchange in
takeovers: Theory and tests.', The Review of Financial
Studies 3 Pg 651 - 675.
Edwards, C.D., (1955) - 'Conglomerate bigness as a source
of power.', In Business Concentration and Price Policy.
Edited by Stigler, G.J., Princeton University Press.
351
Eger, C.E., (1983) - 'An empirical test of the
redistribution effect in pure exchange mergers.', Journal
of Financial and Quauntitative Analysis 18 Pg 547 - 572.
Elgers, F.T. and Clark, J.J., (1980) - 'Merger types and
shareholder returns: Additional evidence.', Financial
Management 19 Pg 66-72.
Elton, E.J. and Gruber, M.J., (1970) - 'Marginal
shareholders, tax rates and the clientle effect.', Review
of Economics and Statistics 52 Pg 68 - 74.
Fama, E.F. and Jensen, M.C., (1983) - 'Separation of
ownership and control.', Journal of Law and Economics 26 Pg
301 - 325.
Firth, M., (1991) - 'Corporate takeovers, shareholders
returns and executive rewards.', Managerial and Decision
Economics 12 Pg 421 - 428.
Fishman, M.J., (1988) - 'A theory of preemptive takeover
bidding.', R.A}D Journal of Economics 19 Pg 88 - 101.
Fishman, M.J., (1989) - 'Preemptive bidding and the role of
the medium of exchange in acquisitions.', Journal of
Finance 44 Pg 41 - 57.
Foster, G., (1986) - 'Financial statement analysis.',
Prentice-Hall International Edition.
Forbes, W. and Watson, R., (1993) - 'Managerial
remuneration and corporate governance: A review of the
issues, evidence and Cadbury committee proposals.',
Accounting and Business Research Pg 331 - 338.
Fowler, D.J. and Rorke, C.H., (1983) - 'Risk measurement
when shares are subject to infrequent trading: Comment.',
Journal of Financial Economics 12 Pg 279 - 283.
Franks, J.R., Harris, R.S. and Mayer, C., (1988) - 'Means
of payment in takeovers: Results for the united kingdom and
the united states.', In Corporate Takeovers: Causes and
Consequences. Edited by Auerbach, A.J. University of
Chicago Press.
Franks, J.R. and Harris, R.S., (1989) - 'Shareholder wealth
effects of corporate takeovers: The UK experience 1955 1985.', Journal of Financial Economics 23 Pg 225 - 249.
Franks, J.R., Harris, R.S. and Titman, S., (1991) - 'The
post merger share price performance of acquiring firms.',
Journal of Financial Economics 29 Pg 81 - 96.
Franks.J. and Mayer, C., (1993) - 'Hostile takeovers and
the correction of managerial failure.', Unpublished
Manuscript, London Business School.
352
Gagnon, J.M., (1967) - 'Purchase versus pooling of
interest: The search for a predictor.', Supplement To The
Journal of Accounting Research Pg 187 - 209.
Galai, ]J. and Masulis, R.W., (1976) - 'The option pricing
model and the risk factor of stocks.', Journal of Financial
Economics 3 Pg 53 - 81.
Gayer, J.J. and Gayer, K.M., (1993) - 'Additional evidence
on the association between the investment opportunity set
and corporate financing, dividend and compensation
policies.', Journal of Accounting and Economics 16 Pg 125 160.
Gemmill, G., (1993) - 'Option pricing: An international
perspective.', McGraw Hill Book Company Europe.
Greene, W.H., (1991) - 'Econometric analysis.', Macmillian
Publishing Company.
Grinyer, J.R., Russell,A. and walker, M., (1991) 'Managerial choices in the valuation of acquired goodwill
in the UK.', Accounting and Business Research 22 Pg 51 55.
Grossman, S. and Hart, 0., (1980) - 'Takeover bids, the
free rider problem, and the theory of the corporation.',
Bell Journal of Economics 11 Pg 42 - 64.
Hagerman, R.L. and Zmijewski, M.E., (1979) - 'Some economic
determinants of accounting policy choice.', Journal of
Accounting and Economics 1 Pg 141 - 161.
Halpern, P., (1983) - 'Corporate Acquisitions: A theory of
special cases? A review of event studies applied to
acquisitions.', Journal of Finance 38 Pg 297 - 317.
Hansen, R.G., (1987) - 'A theory for the choice of exchange
medium in mergers and acquisitions.', Journal of Business
60 Pg 75 - 95.
Harris, M. and Raviv, A., (1988) - 'Corporate control
contests and capital structure.', Journal of Financial
Economics 20 Pg 55 - 86.
Harris , M. and Raviv, A., (1991) - 'The theory of capital
structure.', Journal of Finance Pg 297 - 355.
Harris, M. and Stewart, J.F., Guilkey, D.K., Carlton, W.T.,
(1982) - 'Characteristics of acquired firms: fixed and
random probit analysis.', Southern Economic Journal. Pg 164
- 184.
Hausman,
J.A.,
(1978)
'Specification tests
econometrics', Econometrica 46 Pg 1251 - 1272.
353
in
Hayn, C., (1989) - 'Tax attributes as determinants of
shareholder gains in coporate acquisitions.', Journal of
Financial Economics 23 Pg 121-154.
Healy, P.M. and Krishna, G.P., Ruback, R.S., (1992) - 'Does
corporate performance improve after mergers.', Journal of
Financial Economics 31 Pg 135 - 175.
Heinkel, R. and Schwartz, E., (1985) - 'Rights versus
underwritten offerings: an asymmetric information
approach.', Journal of Finance 41 Pg 1 - 18.
Higgins, R.C. and Schall, L.D., (1975) - 'Corporate
bankruptcy and conglomerate merger.', Journal of Finance 30
Pg93 -113.
Higson, C.J., (1990a) - 'Determinants of the consideration
in UK mergers and acquisitions.', Unpublished Manuscript,
London Business School.
Higson, C.J., (1990b) - 'The choice of accounting method in
UK mergers and acquisitions.', Institute of Chartered
Accountants In England and Wales.
Higson, C. and Elliot, J., (1993) - 'The returns to
takeovers - The evidence.', London Business School, IFA
Working Paper 173-93.
Hoffmeister, J.R. and Dyl, E.A., (1981) - 'Predicting
outcomes of cash tender offers.', Financial Management
Winter Pg 50 - 58.
Holderness, C.G. and Sheehan, D.P., (1988) - 'The role of
majority shareholders in publicly held corporations.',
Journal of Financial Economics 20 Pg 317 -346.
Holgate, P.A.,
(1990) - 'Goodwill, acquisitions and
mergers.', Coopers & Lybrand Deloitte Accounting Guides.
Holland, K.M. and Hodgkinson, L., (1994) - 'The pre
announcement share price behaviour of Uk takeover
targets.', Journal of Business Finance and Accounting 21 Pg
467 - 490.
Holthausen, R.W. (1981) - 'Evidence on the effect of bond
convenants and management compensation contracts on the
choice of accounting techniques: the case of the
depreciation switch back.', Journal of Accounting and
Economics 3 Pg 73 - 109.
Hong, H., Kaplan, R.S. and Mandelker, G., (1978) - 'Pooling
vs Purchase: The effects of accounting for mergers on stock
prices.', The Accounting Review 53 Pg 3]. - 47.
354
Huang, Y.S.and Walking, R.A., (1987) - 'Target abnormal
returns associated with acquisition announcements: Payment,
acquisition form and managerial resistance.', Journal of
Financial Economics 2.9 Pg 329 - 349.
Hughes, A., (1989) - 'The impact of merger: A survey of
empirical evidence for the UK.', In Mergers and Merger
Policy. Edited by Fairburn, J.A. and Kay, J.A., Oxford
University Press.
Ibbotson, R.G. and Jaffe, J.F., (1975) - 'Hot issue
markets.', Journal of Finance 30 Pg 1027 - 1042.
Ingham, H. and Thompson, S., (1993) - 'Executive
compensation and deregulation in UK Building Societies.',
Accounting and Business Research Pg 373 - 383.
Ivancevich, D.M., (1993) - 'Acquisition and goodwill: The
United Kingdom and the United States.', International
Journal of Accounting 28 Pg 156 - 169.
Jalinvand, A. and Harris, R.S., (1984) - 'Corporate
behaviour in adjusting to capital structure and dividend
targets: An econometric study.', Journal of Finance 39 Pg
127 - 145.
Jarrell, G.A., Brickley, J.A. and Netter, J.M., (1988) 'The market for corporate control: The empirical evidence
since 1980.', Journal of Economic Perspectives Winter Pg 49
- 68.
Jarrell, G.A. and Paulsen, A.B., (1989) - 'The returns to
acquiring firms in tender offers: Evidence from three
decades.', Financial Management Autunn Pg 12 - 19.
Jensen, M.C., (1986) - 'Agency costs of free cash flow,
corporate finance, and takeovers.', american Economic
Review 79 Pg 323 - 329.
Jensen, M.C. and Meckling, W.H., (1976) - 'Theory of the
firm: managerial behaviour, agency costs and ownership
structure.', Journal of Financial Economics 3 Pg 305 - 360.
Jensen, M.C. and Murphy, K.J., (1990) - 'Performance pay
and top management incentives.', Journal of Political
Economy 98 Pg 225 - 264.
Jensen, M.C. and Ruback, R.C., (1983) - 'The market f or
corporate control: The scientific evidence.', Journal of
Financial Economics 11. Pg 5 - 50.
Jensen, M.C. and Warner, J.B., (1988) - 'The distribution
of power among managers, shareholders, and directors.',
Journal of Financial Economics 20 Pg 3 - 24.
355
Joy, O.M. and Toilet son, J.O., (1975) - 'On the financial
applications of discriminant analysis.', Journal of
Financial and Quantitative Analysis 10 Pg 723 - 739.
Kaplan, R.S. and Roll, R., (1972) - 'Investor evaluation of
accounting information: Some empirical evidence.', Journal
of Business Pg 225 - 257.
Krasker, W., (1986) - 'Stock price movement in response to
stock issues under asymmetric information.', Journal of
Finance 41 Pg 93 - 105.
Kusewitt, J.E., (1985) - 'An exploratory study of strategic
acquisition factors relating to performance.', Strategic
Management Journal 6 Pg 151 ,- 169.
Lambert, R. and Larcker, D., (1987) - 'An analysis of the
use of accounting and market measures of performance in
executive compensation contracts.', Journal of Accounting
Research 25 Pg 85 - 129.
Lang, L.H.P., Stulz, R.M. and Walking, R.A., (1989) 'Managerial performance, Tobin's Q, and the gains from
successful tender offers.', Journal of Financial Economics
24 Pg 127 - 154.
Lang,L.H.P., Stulz, R.M. and Walking, R.A., (1991) - 'A
test of the free cash flow hypothesis: The case of bidder
returns.', Journal of Financial Economics 29 Pg 315 - 335.
Langetieg, T.C., (1979) - 'An application of a three factor
performance index to measure stockholder gains from
merger.', Journal of Fianacial Economics 6 Pg 365 - 384.
Laster, M., (1991) - 'Determinants of the financing policy
of the firm: An empirical analysis.', Unpublished
Manuscript. City University Business School.
Levis, M., (1993) - 'Seasoned equity offerings and the
short and long run performance of initial public
offerings.', City University Business School, CERIFA
Working Paper No 94/1.
Lewellen, W.G., (1971) - 'A pure financial rationale for
the conglomerate mergers.', Journal of Finance Pg 521 545.
Lewellen, W., Loderer, C. and Martin, K., (1987) 'Executive compensation contracts and executive incentive
problems: An empirical analysis.', Journal of Accounting
and Economics 9 Pg 287 - 310.
Lilien, S. and Pastena, V., (1982) - 'Determinants of
intramethod choice in the oil and gas industry.', Journal
of Accounting and Economics 6 Pg 145 - 170.
356
Limmack, R.J., (1991) - 'Corporate mergers and shareholder
wealth effects: 1977 - 1986.', Accounting and Business
Research 21 Pg 239 - 251.
Limmack, R.J. and McGregor, N., (1992) - 'Industrial
concentration, structural factors and bidder returns.',
Unpublished Manuscript. University of Sterling.
Loderer, C.F. and Mauer, D.C., (1988) - 'Acquiring firms in
coporate mergers: The post merger performance.' Working
Paper No 944 Purdue University. West Lafayette, Indiana.
Loderer, C. and Martin, K., (1990) - 'Corporate
acquisitions by listed firms: The experience of a
comprehensive sample.' Financial Management Winter Pg 17 33.
Lubatkin, M., (1983) - 'Merger and the performance of the
acquiring firm.', Academy of Management Review 8 Pg 218 225.
Maddala, G.S., (1983) - 'Limited-dependent and qualitative
variables in econometrics.', Ca.mbridge University Press.
Maddala, G.S., (1989) - 'Introduction to econometrics.',
Macmillian Pu.blishing Company.
Majd, S. and Myers, S., (1984) - 'Valuing the government's
tax claim on risky assets.', Massachusetts Institute of
Technology Working Paper.
Mandelker, G., (1974) - 'Risk and return: The case of
merging firms.', Journal of Financial Economics 1 Pg 303 336.
Mangel, R. and Singh, H., (1993) - 'Ownership structure,
board relationships and CEO compensation in large US
corporations.', Accounting and Business Research Pg 339 350.
Margrabe, W., (1978) - 'The value of an option to exchange
one asset for another.', Journal of Finance 33 Pg 177 186.
Marsh, P.R., (1979) - 'Equity rights issues and the
efficiency of the UK stock market.', Journal of Finance 34
Pg 839 - 862.
Marsh, P.R., (1982) - 'The choice between equity and debt:
An empirical study.', Journal of Finance 37 Pg 121 - 144.
Masulis, R. and Korwar, A., (1986) - 'Seasoned equity
offerings: An empirical investigation.', Journal of
Financial Economics 15 Pg 91 - 118.
357
Mayer, W.J. and Walker, M.M., (1992) - 'An empirical
analysis of the choice of payment method in corporate
acquisitions during 1979-1990.', Paper presented to the
Midwest Finance Association 1992 Meeting In Chicago.
McConnell, J.J. and Servaes, H., (1990) - 'Additional
evidence on equity ownership and corporate value.', Journal
of Financial Economics 27 Pg 595 - 612.
McLaughlin, R.M., (1990) - 'Investment banking contracts in
tender offers.', Journal of Financial Economics 28 Pg 209 232.
Mikkelson, W. and Partch, M., (1986) - 'Valuation effects
of security offerings and the issuance process.', Journal
of Financial Economics 15 Pg31 -60.
Mikkelson, W.H. and Partch, M.M., (1989) - 'Managers,
voting rights and corporate control.', Journal of Financial
Economics 25 Pg 263 - 290.
Miller, M., (1977) - 'Debt and taxes,', Journal of Finance
Pg 261 - 275.
Mitchell, M.L. and Lehn, K., (1990) - 'Do bad bidders
become good targets?', Journal of Political Economy 98 Pg
372 - 398.
Modigliani, F. and Miller, M., (1958) - 'The cost of
capital, coporation finance and the theory of investment.',
American Economic Review Pg 261 - 275.
Modigliani, F. and Miller, M., (1961) - 'Dividend policy,
growth, and the valuation of shares.', Journal of Business
4 Pg 411 - 433.
Modigliani, F. and Miller, M., (1963) - 'Taxes and the cost
of capital: A correction.', American Economic Review Pg 433
-443.
Morck, R., Shleifer, A. and Vishny, R.W., (1988) 'Management ownership and market valuation: An empirical
analysis.', Journal of Financial Economics 20 Pg 293 - 315.
Morse, D., (1984) - 'An econometric analysis of the choice
of daily versus monthly returns in tests of information
content.', Journal of Accounting Research 22 Pg 605 - 623.
Murphy, K., (1985) - 'Corporate performance and managerial
remuneration: An empirical analysis.', Journal of
Accounting and Economics 7 Pg ii. - 42.
Myers, S., (1977) - 'Determinants of corporate borrowing.',
Journal of Financial Economics 5 Pg 147 - 175.
358
Myers, S. and Majiuf, N.S., (1984) - 'Corporate financing
and investment decisions when firms have information that
investors do not have.', Journal of Financial Economics 13
Pg 187 - 221.
Narayanan, M.P., (1988) - 'Debt versus equity under
asymmetric information.', Journal of Financial and
Quantitative Analysis 23 Pg 39 - 51.
Nathan. K., (1988) - 'Do firms pay to pool?: Some empirical
evidence.', Journal of Accounting and Public Policy 7 Pg
185 - 200.
Niden, C., (1988) - 'Role of taxes in corporate
acquisitions', Unpublished Dcctoral Thesis. Graduate School
of Business, University of Chicago.
Niendorf, B. and Huffman, S.P., (1992) - 'The ability of
prediction models to explain abnormal returns of acquired
firms.', Paper Presented to the Midwest Finance Association
Annual Meeting in Chicago.
Nurnberg, H. and Sweeney, J., (1989) - 'The effect of fair
values and historical costs on accounting for business
combinations.', Issues in Accounting Education 4 Pg 375 395.
Oborne, P., (1986) - 'Paying the piper: Fees paid for
acquisitions.', Acquisition Monthly, January 1986 Pg 28 30.
Palepu, K., (1986) - 'Predicting takeover targets: A
methodological and empirical analysis.', Journal of
Accounting and Economics 8 Pg 3 - 35.
Peterson, D.R. and Peterson, P.P., (1991) - 'The medium of
exchange in mergers and acquisitions.', Journal of Banking
and Finance 15 Pg 383 - 405.
Petit, R., (1977) - 'Taxes, transaction costs and the
clientle effects of dividends. 1 , Journal of Financial
Economics Pg 419 - 436.
Porter, M.E., (1979) - 'How competitive forces shape
strategy.', Havard Business Review March - April Pg
Porter, M.E., (1987) - 'From competitive advantage to
corporate strategy.', avard Business Review May - June Pg
43 - 59.
Pound, J., (1988) - 'Proxy contests and the efficiency of
shareholder oversight.', Journal of Financial Economics 20
Pg 237 - 265.
359
Ritter, J.R., (1984) - 'The "hot issue" market of 1980.',
Journal of Business 57 Pg 215 - 240.
Robinson, J.R. and Shane, P.B., (1990) - 'Acquisition
accounting method and bid premium for target firms.', The
Accounting Review 65 Pg 25 - 48.
Rosen, S., (1972) - 'Learning by experience as joint
production.', Quaterly Journal of Economics Pg 366 - 382.
Ross, S.A., (1977) - 'The determination of financial
structure: The incentive signalling approach.', Bell
Journal of Economics Pg 23 -40.
Sapienza, S.R., (1967) - ']iscussion of purchase versus
pooling of inetrests: The search for a predictor.',
Supplement To The Journal of Accounting Research Pg 205 209.
Scanlon, K.P., Trifts, J.W. and Pettway, R.H., (1989) 'Impacts of relative size and industrial relatedness on
returns to shareholders of acquiring firms.', Journal of
Financial Research 12 Pg 103 - 112.
Schipper, K. and Thompson, R., (1983) - 'Evidence on the
capitalised value of merger activity for acquiring firms.',
Journal of Financial Economics 11 Pg 85 - 119.
Scholes, M. and Williams, J., (1977) - 'Estimating betas
from nonsynchronous data.', Journal of Financial Economics
5 Pg 309 - 327.
Servaes, H., (1991) - 'Tobin's Q and the gains from
takeovers.', Journal of Finance 46 Pg 409 - 419.
Seth, A., (1990) - 'Value creation in acquisitions: A reexamination of performance issues.', Strategic Management
Journal 11 Pg 99 - 115.
Shastri, K., (1982) - 'Valuing corporate securities: Some
effects of mergers by exchange offers.', University of
Pittsburgh Working Paper 517.
Shelton, L.M., (1988) - 'Strategic business fits and
corporate acquisition: empirical evidence.', Strategic
Management Journal 9 Pg 279 - 287.
Shih, F.L. and Suk, D.Y., (1992) - 'Do stock price runups
on tender offer targets reflect insider trading?', Paper
Presented to the Midwest Finance Association 1992 Meeting
in Chicago.
Shleifer, A. and Vishny, R.W., (1986) - 'Large shareholders
and corporate control.', Journal of Political Economy 94 Pg
461 - 488.
360
Sicherman, N.W. and Pettway, R.H., (1987) - 'Acquisition of
divested assets and shareholders' wealth.', Journal of
Finance 42 Pg 1261 - 1273.
Singh, H. and Montgomery, C., (1987) - 'Corporate
acquisition strategies and economic performance.',
Strategic Management Journal 8 Pg 377 - 386.
Sloan, R.G., (1993) -'Accounting earnings and top executive
compensation.', Journal of Accounting and Economics 16 Pg
55 - 100.
Slovin, M.V., Sushka, M.E. and Hudson, C.D., (1990) 'External monitoring and its effects on seasonal common
stock issues.', Journal of Accounting and Economics 12 Pg
397 - 417.
Slusky, A.R. and Caves, R.E., (1991) - 'Synergy, agency and
the determinants of premia paid in mergers.', Journal of
Industrial Economics 39 Pg 277 - 296.
Smith, C.W., (1977) - 'Alternative methods for raising
capital: Rights versus underwritten offerings.', Journal of
Financial Economics 5 Pg 273 - 307.
Smith, C.W. and Warner, J.B., (1979) - 'On financial
contracting: An analysis of bond convenants.', Journal of
Financial Economics 7 Pg 117 - 161.
Smith, C.W. and Watts, R.L., (1993) - 'The investment
opportunity set and corporate financing, dividend and
compensation policy.', Journal of Financial Economics 32 Pg
263 - 292
Smith, D.C. and Jennings, R.H., (1993) - 'An empirical
investigation of the medium of exchange decision in a
signalling framework.', Unpublished Manuscript, Indiana
University.
Song, M.H. and Walking, R.A., (1993) - 'The impact of
managerial ownership on acquisition attempts and target
shareholder wealth.', Journal of Financial and Quantitative
Analysis 28 Pg 439 - 457.
SSAP 22 (Revised) (1989) - 'Accounting for goodwill.',
Accounting Standards Committee.
SSAP 23 (1985) - 'Accounting for acquisitions and
mergers.', Accounting Standards Committee.
Stapleton, R.C., (1982) - 'Mergers, debt capacity and the
valuation of corporate loans.', in Mergers and Acquisitions
Edited By Keenan, M., and White, L.J., Lexington Books.
3 6].
Strong, N., (1992) - 'Modelling abnormal returns: A review
article.', Journal of Business Finance and Accounting 19 Pg
533 - 553.
Stulz, R.M., (1988) 'Managerial control of voting rights:
Financing policies and the market for corporate control.',
Journal of Financial Economics 20 Pg 25 - 54.
Stulz, R.M., Walking, R.A. and Song, M.H., (1990) - 'The
distribution of target ownership and the division of gains
in successful takeovers.', Journal of Finance 45 Pg 817 833.
Sudarsanam, P.S. and Taffler, R.J., (1985) - 'Industrial
classification in the UK, capital markets.', Applied
Economics 17 Pg 291 - 308.
Sudarsanam, P.S., (1990a) - ' Merger gains to bidder and
target company shareholders: An empirical test of synergy
in United Kingdom mergers.', Paper Presented to the Annual
Conference of the European Accounting Association at
Budapest.
Sudarsanam, P.S., (1990b) - 'The efficiency of the market
for initial public offerings: The case of the UK unlisted
securities market.', Unpublished Manuscript. City
University Business School.
Sudarsanam, P.S., (1991) - 'Defensive strategies of target
firms in UK contested takeovers.', Managerial Finance 17 Pg
47 - 56.
Sudasanam, P.S., Holl, P. and Salami, A.R., (1993) 'Shareholder wealth gains in mergers: An empirical test of
the synergy and agency effects.', Paper Presented to the
European Accounting Association Annual Congress in Thrku,
Finland.
Sudarsanam, P.S., (1994a) - 'Defensive startegies of target
firms in Uk contested takeovers: A survey.', City
University Business School, CERIFA Working Paper No 94/1.
Sudarsanam, P.S., (1994b) - 'Determinants of outcome in UK
contested takeover bids: Defensive strategies, shareholding
structure and method of payment.', City University Business
School, CERIFA Working Paper No 94/2.
Summers, L.H., (1986) - 'Does the stock market rationally
reflect fundamental values?', Journal of Finance 41 Pg 591
- 601.
Sunders, S., (1973) - ' Relationship between accounting
changes and stock prices: Problems of measurement and some
empirical evidence.', Supplement to the Journal of
Accounting Research Pg 1 - 45.
362
Taggart, R.A., (1977) - 'A model of corporate financing
decisions.', Journal of Finance 32 Pg 1467 - 1468.
Taggart, R.A., (1988) - The growth of the "junk" bonds and
its role in financing takeovers.', In Mergers and
Acquisitions Edited by Auerbach, A.J. University of Chicago
Press.
Titman, S. and Trueman, B., (1986) - 'Information quality
and the valuation of new issues.', Journal of Accounting
and Economics 8 Pg 159 - 172.
Trautwein, F., (1990) - 'Merger motives and merger
prescriptions.', Strategic Management Journal 11 Pg 283 295.
Travios, N.G., (1987) - 'Corporate takeover bids, method of
payment and bidding firms stock returns.', Journal of
Finance 42 Pg 943 - 963.
Walking, R., (1985) - 'Predicting tender offer success: A
logistic analysis.', Journal of Financial and Quantitative
Analysis 20 Pg 461 - 478.
Wansley, J., Lane, W. and Yang, H., (1983) - 'Abnormal
returns to acquiring firms by the type of acquisition and
method of payment.', Financial Management 12 Pg 16 - 22.
Watts, R.L. and Zimmerman., J.L., (1978) - 'Towards a
positive theory of the determination of accounting
standards.', The Accounting Review Pg 112 - 134.
Westerfield, R., (1970) - 'A note on the measurement of
conglomerate diversification.', Journal of Finance 25 Pg
909 - 914.
Weston, J.F., Chung, K.S. and Hoag, S.E.,
(1990)
'Mergers, restructuring, and corporate control.', Prentice
Hall International, Inc.
White, H., (1978) - 'A hetroskedasticity consistent
covariance matrix estimator and a direct test for
hetroskedasticity.', Econometrica 48 Pg 817 - 838.
Wonnacott, T.H. and Wonnacott, R.J., (1977) - 'Introductory
statistics for business and economics, 2nd edition.', John
Wiley & Sons Inc.
Wyatt, A.R., (1967) - 'Discussion of purchase versus
pooling of inetrests: The search for a predictor.',
Supplement To The Journal of Accounting Research Pg 210 212.
Yadav, P.K., (1992) - 'Event studies based on volatility of
returns and trading volume: A review.', British Accounting
Review 24 Pg 157 - 184.
363
Zmijewski, M. and Hagerman, R., (1981) - 'An income
strategy approach to the positive theory of accounting
standard setting/choice.', Journal of Accounting and
Economics 3 Pg 129 - 149.
364