 

T535355

2.009

 

iBRARY
Michigan State

University '-

 

This is to certify that the
dissertation entitled

Essays in Corporate Finance

presented by

Tilan Tang

has been accepted towards fulﬁllment
of the requirements for the

Doctoral degree in Finance

 

W A. Jécéééé

Major F‘rofessor’s Signature
Mm] g. Z007

Date

MSU is an Afﬁrmative Action/Equal Opportunity Employer

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K‘lProg/AccaPres/ClRC/DateDue indd

 

ESSAYS IN CORPORATE FINANCE
By

Tilan Tang

A DISSERTATION
Submitted to
Michigan State University

In partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILSOPHY
Finance

2009

ABSTRACT

ESSAYS IN CORPORATE FINANCE
By
Tilan Tang

Over the past two decades, concerns with endogeneity in empirical ﬁnance
research have attracted a great deal of academic attention. Numerous papers have
proposed various approaches to control for the biases caused by endogeneity issues.
Inspired by previous studies, this dissertation aims to explore special identiﬁcation
strategies to answer the important questions in corporate ﬁnance.

The ﬁrst essay attempts to get around some well-known endogeneity issues in
evaluating bidder gains in mergers and acquisitions. The use of announcement returns to
assess the gains to acquiring-ﬁrm shareholders is problematic as the announcement of a
bid usually conveys more information than the synergy from combination. Using a
unique, hand-collected dataset on failed takeovers, we investigate termination returns in
deals canceled for reasons unrelated to the bidder's valuation. This methodology
addresses biases in conventional techniques and permits a cleaner assessment of value
improvements from acquisitions. We ﬁnd that bidder gains vary signiﬁcantly with the
type of target acquired: on average bidder returns are positive when acquiring private
targets or subsidiaries, while negative when acquiring public targets. In addition, we also
ﬁnd. no difference in returns for stock financed versus cash ﬁnanced deals, helping to
conﬁrm that the method-of—payment effect is likely due to the revelation of adverse
information about the bidder. Further evidence also suggests that the lack of liquidity in

private sellers contributes to the positive gains to bidders in transactions.

The second essay explores the role of access of internal cash in product market
competition dynamics. We exploit an exogenous shock to the ﬁnances of tobacco ﬁrms
arising from an excise tax in the early 19905. Using the event as a natural experiment, we
ﬁnd signiﬁcant positive returns to rivals who compete with non-tobacco segments in
tobacco ﬁrms and a signiﬁcant change in output behavior of those non-tobacco segments
after the shock. This suggests that access to cash is an important determinant of a ﬁrm’s
aggressiveness and success in the product market. The effects are even robust to a
number of estimation issues and identiﬁcation choices. Since the shock is exogenous to
the investment opportunities of non-tobacco industries, our evidence supports the
hypothesis that there is a causal relation between a ﬁrm's cash ﬂow and its product
market behavior. Moreover, we consider the determinants of the cross-sectional variation
in competitor abnormal returns and ﬁnd that the connection between cash ﬂow and
competitive performance is magniﬁed in competitive industries and when cash-

constrained ﬁrms only have a small market Share.

Copyright by
'Tﬂan'Tang

2009

ACKNOWLEDGMENTS

I would like to express my deepest appreciation to my adviser, Professor Charles
Hadlock, who taught me to have patience to search for the truth and an open mind to see
it. Without his guidance and help, this dissertation would not have been possible.

I would also like to thank my committee members, Professor Ted. Fee, Professor
Long Chen, Professor Andrei Simonov, and Professor Michael Mazzeo, for valuable
comments and fruitful discussions. I am extremely grateful for the assistance and advice I
received from Professor Thomas Bates, and Professor Richard Simonds. I also thank
other faculty members and doctoral students of Finance Department at Michigan State
University for support during my graduate studies.

Finally, my deepest gratitude goes to my loving and supportive parents: Bene Gao

and Guozhong Tang.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................ viii
l. A Re-examination of Bidder Gains in Acquisitions:
Evidence from Termination Announcements ............................................ l
1 .1 Introduction ........................................................................... 1
1.2 Data and Sample Selection ......................................................... 5
1.2.1 Sample Selection ......................................................... 5
1.2.2 Summary Statistics ...................................................... 12
1.3 Gains to Acquiring-ﬁrm Shareholders ........................................... 15
1.3.1 Cumulative Abnormal Returns upon Termination
Announcements .......................................................... 1 5
1.3.2 Concerns on Returns upon Merger Announcements ................ 18
1.3.3 Effect of Firm and Deal Characteristics on Bidder Gains ......... 19
1 .3.4 Robustness ................................................................ 25
1.4 Discussions on Gains to Acquirers of Non-public Targets .................... 27
1.4.1 Existing Alternative Explanations ..................................... 27
1 .4 2 Gains to Acquirers and Characteristics of Target Parents .......... 28
1 .5 Conclusion ........................................................................... 3O

2. Cash Flow and Product Market Competition:

An Empirical Analysis ......................................................................... 33
2. 1 Introduction ........................................................................... 3 3
2.2 Related Literature and Hypotheses Development ............................... 37
2.2.1 Financial Constraints, Internal Funds and Product Market

Competition. 3.7
2.2.2 Internal Capital Market in D1ver51ﬁed Firms ........................ 43

2.2.3 Hypotheses Development and Empirical Design ..................... 45
2.3 Sample Selection and Data Description ........................................... 49
2.3.1 Sample Selection .......................................................... 49
2.3.2 Summary Statistics ....................................................... 53

2.3.3 Industry Rivals and Customers ......................................... 54

2.4 Empirical Results ..................................................................... 55
2.4.1 Abnormal Returns to Industry Rivals .................................. 56

2.4.2 Output Changes of Non-tobacco Segments ........................... 59

2.4.3 Robustness ................................................................ 61

2.4.4 Cross-section Analysis ................................................... 65

2.4.5 Abnormal Returns to Customers of Industry Rivals ................. 66

2.5 Conclusion ............................................................................. 67
APPENDICES ...................................................................................... 70

vi

A Tables for Essay 1 ............................................................................. 71
B Tables for Essay 2 .............................................................................. 83

BIBLIOGRAPHY ................................................................................. 93

vii

A.1

A.2

A.2

A.2

A.3

A.4

A.5

A.6

A.6

A.7

A.8

B.1

B.2

B.3

B.4

B.5

B.6

8.7

8.8

B.9

LIST OF TABLES

Sample Selection ............................................................................. 72
Sample Summary Statistics ............................................................... 73
Continued ..................................................................................... 74
Continued ..................................................................................... 75
Cumulative Abnormal Returns upon Merger Termination ........................ 76
Cumulative Abnormal Returns upon Merger Announcement ..................... 77
Firm and Deal Characteristics: Sort by Target Status .............................. 78
Cross-sectional Regression Analysis of Bidder Gains ............................... 79
Continued ..................................................................................... 80
Termination Returns for Acquirers of Subsidiaries ................................. 81
Cross-section Analysis of Gains to Acquirers of Subsidiaries ..................... 82
Tobacco-dependent Firm Data ........................................................... 84
Segment Data for Tobacco-dependent Firms .......................................... 85
Statistics for Rivals to N on-tobacco Segments ......................................... 86
Abnormal Returns to Non-tobacco Rivals of Tobacco Firms ...................... 87
Output Changes Following Cash Shock ............... v ................................. 88
Robustness under Alternative Benchmarks ........................................... 89
Robustness under Additional Controls ................................................. 9O
Regressions of Abnormal Returns to Rivals ........................................... 91
Abnormal Returns to Customers of Industry Rivals ................................. 92

viii

 

CHAPTER 1

A Reéexamination of Bidder Gains in Acquisitions:
Evidence from Termination Announcements

1.1 Introduction

Mergers and acquisitions are economically important events that have attracted a
great deal of attention from ﬁnancial economists. Despite the importance of these events,
however, our understanding of the causes and consequences of corporate control activity
is still far from complete. Part of the problem arises from the fact that much of the prior
research on such activity focuses on studying market returns at the time of acquisition
announcements. 1 While this evidence is informative, announcement returns can be
difﬁcult to interpret because many different pieces of information may be revealed by the
announcement, and some of this information may have little to do with the value created
by the merger itself.2 For example, the decision to undertake a bid may convey to the
market that a ﬁrm’s internal growth prospects are poor, and the decision to use stock in
an acquisition may convey that the ﬁrm’s stock is overvalued.

Several researchers have attempted to overcome these interpretation problems.
Hietala, Kaplan, and Robinson (2003) and Bhagat, Dong, Hirshleifer, and Noah (2005)

use information relating to competing bids to extract information on the market’s

 

' Tremendous variation in bidder announcement returns is recorded in previous studies (see Jensen and
Ruback ( 1983)). To date, however, researchers are unable to fully explain this variation.

2 Grinblatt and Titman (2002) state that the stock return at the time of announcement reﬂects not just the
expected effect of the acquisition on ﬁrm proﬁtability but also information regarding the stand-alone value
of the ﬁrm. Furthermore, Hietala, Kaplan, and Robinson (2003) suggest that announcements contain
information about the potential synergies from the combination, the bidder’s stand-alone value, and
possible bidder overpayment.

 

assessment of the value of the merger to the original bidder. Taking a different approach,
Fuller, Netter and Stegemoller (2002) use repeated bids by the same acquirer to attempt
to remove the role of bidder-speciﬁc characteristics in announcement returns. While these
approaches offer additional insights on bidder gains in merger activity, some
interpretation issues remain. In particular, studies that exploit the presence of competing
bids assume the occurrence of a competing bid to be independent of the value of the
original bidder, while studies that consider repeat bidders cannot adjust for the possibility
of time-varying changes in underlying bidder characteristics.

In this paper we attempt to more cleanly identify the gains that accrue to bidders
as a result of acquisition activity by examining bidder returns at the time that a previously
announced acquisition is canceled for exogenous reasons. This approach depends
critically on identifying cancellations that have nothing to do with the value of the bidder
or the merged entity, that is, cancellations that arise due to factors such as regulatory
intervention, unexpected lawsuits, or the presence of competing bids.3 We therefore begin
with a comprehensive sample of canceled deals and select from this set only the
transactions for which we can determine that this exogeneity condition holds. The ﬁnal
sample includes 272 failed takeover bids occurring between 1990 and 2006.

For the sample as a whole, we ﬁnd that bidder returns upon announcement of deal
cancellation are on average insigniﬁcantly different from zero. However, this ﬁnding
masks considerable variation in the market reaction to different types of deals. In

particular, we ﬁnd that bidder returns are signiﬁcantly negative when a deal is canceled

 

3 While the sample that we use includes deals canceled because of the presence of competing bids, the
assumptions on these events are much weaker than the aforementioned studies that consider returns at the
time of the announcement of the competing bid. As a robustness check we also examine the sample after
excluding all deals canceled because of competing bids and the results still hold.

 

and the target is a private ﬁrm or a subsidiary of either a public or private ﬁrm. This
indicates that the market believes that these types of deals create value for bidders and
that this value is lost upon recognition that the deal will not be consummated. The fact
that these deals appear to create value for bidders is consistent with prior evidence that
bidders fare well when acquiring relatively illiquid assets (e. g., Ofﬁcer (2007)). When we
further compare acquisitions of public ﬁrm subsidiaries to acquisitions of private ﬁrm
subsidiaries, we ﬁnd that cancellation returns are relatively smaller in magnitude (still
negative in sign) for acquisitions of public ﬁrm subsidiaries, which suggests that private
sellers may be willing to sell at a lower price because of elevated ﬁnancial constraints
associated with not having easy access to public equity markets.

In contrast, for acquisitions of entire public ﬁrms, we ﬁnd that cancellation
announcement returns are on average signiﬁcantly positive. This suggests that the market
believes that these deals harm bidders on average and thus that deal cancellation is
viewed as positive news for bidder shareholders. Note that while many prior studies
suggest that bidders fare poorly when purchasing public ﬁrms, these studies are clouded
by the interpretation issues discussed above. The fact that we ﬁnd evidence of value
destruction for this large class of deals suggests that agency problems at the ﬁrm level
can generate a bias towards unproﬁtable growth, at least in the merger arena.

The experimental design discussed above also allows us to revisit some prominent
issues in the prior literature on bidder returns. In particular, an important ﬁnding in the
literature is that bidder returns tend to be more negative when the bidder uses equity as
the method of payment. The common interpretation of this result is that the choice of

stock ﬁnancing conveysa negative Signal regarding the value of the bidder. However, an

alternative possibility is simply that stock deals are systematically poorer deals for
bidders. In our sample, we ﬁnd no difference between cash and stock deals in terms of
announcement returns at the time of deal cancellation. This suggests that this alternative
possibility is unlikely to hold, strengthening conﬁdence in the more commonly held
interpretation of the role of stock ﬁnancing in bidder returns.

Finally, we examine the role of target size in bidder returns. Speciﬁcally, the
larger the target is relative to the bidder, the more positive is the announcement return at
the time of deal cancellation. This suggests that large targets are able to extract more
value from bidders, driving downwards bidder gains from merger activity, a ﬁnding that
is consistent with recent concerns raised in the business press about mergers of equals.‘I

Taken as a whole, this paper adds to our understanding of the value that accrues to
bidders when making acquisitions by introducing a cleaner approach to measuring the
market’s assessment of value creation in mergers. In particular, the approach of
considering bidder returns when deals are canceled allows one to estimate value
improvement in mergers without the confounding effect of contaminating information
that may have little to do with the proposed deal. Given this methodological improvement,
the results uncovered here add substantially to our conﬁdence in some prominent results
regarding bidder returns. In particular, the paper’s main results conﬁrm prior evidence
that bidders often do not beneﬁt when they acquire public ﬁrms, while they appear to fare
well when they purchase other types of entities including private ﬁrms and subsidiaries of
both public and private ﬁmis. In addition, this paper extends previous research by

providing additional evidence suggesting that the liquidity needs of owners of non-public

 

4 It has been argued that similarity of size may give the target more bargaining power in merger contests.
An article in The Economist (1999) also suggests that mergers of equals may be difﬁcult to consummate
successfully because of leadership issues.

assets help explain the higher gains to bidders of private ﬁrms or subsidiaries. This paper
also lends support to the interpretation that the previously documented method-of-
payment effect in bidder returns merely reﬂects a pure negative signal regarding the
underlying value of the bidder’s assets in place.

The remainder of the paper proceeds as follows. Section 1.2 describes the data
and sample selection process. In Section 1.3, we present ﬁndings on announcement
returns at the time of deal cancellation for the sample as a whole and for a variety of
selected subsamples. Section 1.3 also presents robustness tests. In Section 1.4, we
examine possible explanations for the differences in bidder returns. Section 1.5 concludes

the paper.

1.2 Data and Sample Selection

1.2.1 Sample Selection

To construct a sample of proposed mergers that end unsuccessfully, we begin
with all transactions posted in the Securities Data Company’s (SDC) Merger and
Acquisition database. This database includes information on 590,238 merger bids for US.
and non-US. companies made between January 1962 and July 2007. Since there were
important changes in the corporate control environment beginning in the early 19903, we
limit attention to all transactions occurring between January 1, 1990 and December 31,
2006. This sample period includes the large merger wave of the late 19905 and early

20005.5 For inclusion in the sample, we require that the acquirer controls less than 50% of

 

5 Several previous M&A studies investigate this time period and document the trends and characteristics of
mergers over the last century. See, e.g., Andrade, Mitchell, and Stafford (2001) and Moeller, Schlingemann,
and Stulz (2005).

 

the shares of the target as of the merger announcement date and seeks to obtain control of
all shares of the target. Moreover, we exclude all block purchases, exercises of previously
acquired options, leveraged buyouts, self tenders, and exchange offers, since these types
of transactions are economically distinct from the traditional mergers that we seek to
understand in this paper.

From the remaining transactions, we consider for further study all deals that were
ultimately canceled or withdrawn. We then impose the following sampling criteria, which
are fairly standard in studies of acquisition activities:

a The acquirer is a US. public ﬁrm traded on the NYSE, AMEX, or NASDAQ
and is listed on the Center for Research in Security Prices (CRSP) ﬁle with at
least three days of return data available around the deal termination date.

0 The target is an independent public or private ﬁrm or, alternatively, a
subsidiary of a public or private ﬁrm.

0 The deal’s value is greater than or equal to $1 million, with value deﬁned by
SDC as the total value of consideration paid by the acquiring ﬁrm, excluding
fees and expenses.6 In addition, the deal’s value is more than 1% of the market
value of the bidder’s assets, which is deﬁned as book value of assets minus
book value of equity plus market value of equity. These criteria ensure that the
deal has material value implications for the bidder.

0 Neither the acquirer nor the target is a utility or a ﬁnancial institution.7

0 Revised or repeated bids are excluded since these bids may reveal more

information about the acquirers and these deals are likely to yield

 

6 The dollar value would include the amount paid for all corrunon stock. preferred stock, debt, options,
assets, warrants, and any stake purchased within six months of the announcement date.
7 Industry classiﬁcation follows Fama and French ( 1997).

 

systematically different gains to bidders than contests with only a single bid
from the acquirer.8

o The deal status is recorded by SDC as withdrawn and the acquirer makes no

announcements about further merger or acquisition in the seven-day window
around the initial acquisition termination date.9

The above requirements yield a sample of 1,486 failed mergers. However,
because the success of this paper’s research approach hinges on identifying a group of
deals that are canceled due to reasons unrelated to the valuation of the acquirer, not all of
these deals are eligible for inclusion in the analysis. To identify such deals, for every
transaction in the failed deal sample, we search all available reports in Factiva and
attempt to investigate why the merger did not go through. This manual search requires
extensive attention since news often includes a reporter’s speculations or personal
opinions. Moreover, news headlines sometimes exaggerate the causes of deal failure or
understate other confounding events about the acquirer. Thus, the real causes of merger
termination are often obscure in news reports or even sometimes unspeciﬁed.

Based on the information collected from F activa, we ﬁnd that merger deals can
fail for various reasons. For example, an acquirer may terminate a deal due to bad market
reception (e. g., a decrease in the acquirer’s stock price), or a target may cancel the merger
after discovering severe ﬁnancial problems of the acquirer during the due diligence
review process. Such failures certainly convey some information related to the acquirer’s

stand-alone value and thus should be excluded from the sample. In addition, for a number

 

8 For example, cancellations of revised bids are more likely to be anticipated.

9 When we manually search the failed deals in Factiva. we ﬁnd that some transactions are misclassiﬁed as
withdrawn. These misclassiﬁed deals are ultimately completed as strategic alliances or as mergers at a later
date. We exclude such deals. Furthermore, two deals in the failed sample actually take the form of a
traditional Dutch auction in the bidding process. We exclude these transactions too.

of failed bids, multiple reasons are cited for deal failure, with at least one reason related
to the value of the acquiring ﬁrm. In order to control for the problem of multiple
information release at announcement, these controversial bids should also be excluded
from the failed deal sample.10

The ﬁrst set of excluded deals consists of those acquisitions that fail because the
acquirer experiences ﬁnancial problems or structural changes subsequent to the merger
announcement. Such causes of deal failure, which are explicitly or implicitly stated in the
termination announcement, mainly include signiﬁcant decreases in the acquirer’s stock
price, negative ﬁnancial performance of the acquirer, the inability of the acquirer to
obtain required ﬁnancing, or acquisitions of the bidder. We exclude these deals because it
is highly possible that for these deals the bidder’s abnormal return at termination would
reﬂect, at least in part, the change in the valuation of the acquirer, in which case inclusion
of these deals would signiﬁcantly bias the results. There are other cases in which a deal
fails when the acquirer receives strong opposition to the merger from its shareholders or
the market. This type of opposition may potentially signal poor growth opportunities for
the acquirer. Again, in such cases the cause of deal failure would be, if only indirectly,
related to the acquirer’s valuation, and thus we exclude these deals.

The second set of excluded deals consists of mergers that are terminated by the
acquirer due to worsening conditions at the target. For instance, the prospective acquirer
may discover problems at the target during the due diligence review process and call the

deal off. In this case, deal termination might seem to be unrelated to the acquirer’s

 

I0

Dong, Hirshleifer, Richarson, and Teoh (2006) ﬁnd that the acquirer’s valuation has no effect on the
probability of deal success. This evidence may suggest that the deal termination decision would not be
affected by the bidder’s valuation. However, observing that a good portion of sample deals fail after a
decrease in stock price, we decide to apply this screen to exclude deals that fail due to reasons related to the
valuation of the acquirer.

valuation. However, worsening target performance during the merging process can
negatively impact the potential synergy from combination and in turn reduce shareholder
gains from the acquisition. To ensure the termination return proxies for the bidder’s gains
from the original merger deal, these failed acquisitions are excluded.

Another signiﬁcant portion of deals fail because of either active rejection or lack
of response from the target. There are two ways to view these deals. One perspective
holds that in such cases the target is simply not satisﬁed with the merger offer or
managers are worried about job loss following combination. Given this interpretation,
failure of these bids is not related to the valuation of the acquirer or the deal and hence
these failures should be considered in the analysis. However, the other argument cautions
against inclusion. It is possible that the target refuses to accept the offer because it
discovers or foresees problems at the acquirer during the due diligence process. In this
case, the termination would correspond to negative information about the acquirer. The
same arguments apply to the group of deals that fail due to the two merging companies’
inability to reach agreement or conclude negotiation. To avoid potential bias, we
eliminate all these deals from the ﬁnal analysis.

Next, we exclude the failed bids that terminate due to changes in the form of
agreement. This includes a revision in merger agreement terms, a switch to an alliance
structure, a change of target, and an adjustment due to the resignation or death of key
managers. Because these changes could be, if indirectly, related to the acquirer’s
valuation, we exclude these from the analysis.

Furthermore, because market conditions usually have a material impact on a

ﬁrm’s growth options, we exclude deals that fail because of a changing (usually more

unfavorable) industry environment. To minimize the possibility of market anticipation of
deal failure, we also drop those deals that analysts might expect to fail right after the deal
is announced publicly.

Finally, we exclude those deals for which the real causes of termination are not
identiﬁed.

Table A.1 summarizes the sample selection process. After excluding any deal
whose failure is potentially related to the acquirer’s valuation or the synergy value of
combination, the ﬁnal failed sample contains 272 terminated deals, 81% of which are
domestic mergers. As can be seen from the table, exclusion of failed deals based on
reasons of deal termination dramatically reduces the size of the sample. The Exogenous
Failed Sample includes only those bids that fail because of disapproval by regulatory
agencies, occurrence of competing offers, or unexpected lawsuits or court rulings.

Regulatory objections can come from US. or foreign regulatory bodies and often
take the form of immediate or pending antitrust charges. The rejections included in the
sample come from the Federal Trade Commission, Department of Justice, Department of
insurance, European Union Commission, Food and Drug Administration, Department of
Transportation, British government, German regulatory authority, and state governors.ll
Most sample deals fail due to potential antitrust concerns, delays of regulatory approval,
or threats from unnamed regulatory agencies, with 67% of the ﬁnal deals categorized
under “regulatory rejection.”

Competing bids often emerge subsequent to the initial acquirer’s ﬁrst offer and

may ultimately win the bidding war. Slightly less than 30% of the exogenous deals’

 

H One exception is the deal terminated by the Securities and Exchange Commission because of accounting
issues.

10

failure is listed as due to “competing bids.” Note that the occurrence of competing offers
sometimes may be predicted, especially when the target is actively seeking a “white
knight.” The ﬁnal sample therefore excludes deals associated with reports of active
contact between the target and a white knight company, even if the deals did not close in
the end due to competing offers. Some concern remains about the inclusion of deals that
fail as an occurrence of competing bids. We discuss some treatments of this issue in the
robustness checks included below.

Unexpected legal actions include actual or potential lawsuits posed by rival ﬁrms
or local governments. Most of those threats contain possible antitrust charges claimed in
federal or state court. The exceptions are eight deals that were stopped by bankruptcy
court rulings, which stated concerns about an insufﬁcient offer or about the market power
of the combination.

The main analysis of this paper relies on the accuracy of the acquirer’s abnormal
returns at termination announcement as a proxy for the bidder’s gain from the acquisition.
There are two assumptions underpinning this approach. First, acquisition expenses, such
as advisory or legal fees and the expense of information exchange, opportunity costs, and
management effort, are not substantial enough to impact the acquirer’s valuation, that is,
these expenses are assumed to be relatively small compared to the acquirer’s market
value. In analyzing acquirer reports around acquisition termination, we note that very few
companies record merger-related expenses as a loss in ﬁnancial statements. This suggests
that the ﬁrst assumption is well founded. Second, the acquirer is not anticipated to
participate in future mergers. Given that the sample requires no announcement about

future mergers in a seven-day window around deal temiination, this assumption seems

ll

reasonable. For those deals that fail due to exogenous reasons, we also ﬁnd that less than
7% of acquirers make acquisition offers to a different target within three years of the

failed bid, which provides additional support for the second assumption.

1.2.2 Summary Statistics

Table A.2 provides summary statistics on deal and ﬁrm characteristics for the
exogenous failed deals. Panel A shows that exclusion of failed deals based on reasons of
deal termination sharply increases the mean and median deal value of the sample
acquisitions: for the exogenous failed deals, the mean (median) deal value is 1.73 billion
(190.28 million), while the average (median) for all failed deals is 1.53 billion (170.54
million).12 Further, the deal value as a fraction of the acquirer’s market value of equity
(MVE) or market value of assets (MVA) at the end of the ﬁscal year of the acquisition
announcement is larger for exogenous failed deals than for all failed deals. For example,
the average ratio of deal value to MVE is 0.6645, much higher than the average ratio of
0.2181 for all failed deals. This indicates that the target ﬁrms are economically
signiﬁcant to sample acquirers. This conjecture ﬁnds further support when we examine
the relative size of target to acquirer. The ratio of target to acquirer market value is much
larger for the exogenous failed deals. Moreover, acquisitions in the exogenous failed
sample are more likely to have multiple bidders than those in the failed sample (0.2935 vs.
0.1398). In sum, these results show that the sample of exogenous failed deals contains
large targets and that the sample acquisitions are economically signiﬁcant to the acquirers.

Panel A of Table A.2 also reports evidence on a deal’s method of payment and

 

l2 . . . . . .
Because the main analysrs of this paper focuses on the exogenous failed sample, the summary statrstrcs
for all failed deals are not reported in the tables.

12

other deal characteristics. Following Martin (1996), we sort method of payment into three
categories: 1) cash ﬁnancing, which includes cash, debt, or the combination of cash, debt,
and liabilities; 2) equity ﬁnancing, which comprises payments with common stock,
ordinary shares, or a combination of stock, options, or warrants; and 3) combination
ﬁnancing, which includes a combination of common stock, cash, debt, preferred stock,
convertible security, and others forms of payment listed by SDC. We use “equity (cash)
in payment,” deﬁned as the percentage of equity (cash) in consideration, and “Pure equity
(cash) deals,” deﬁned as the percentage of pure equity (cash) ﬁnanced deals, to measure
how much and how often stock (cash) ﬁnancing is employed by the acquirer. The panel
shows that the exogenous failed deals are less likely to be ﬁnanced with stock and more
likely to be ﬁnanced with cash. The exogenous failed deals are also more likely to be
acquisitions of public ﬁrms than private ﬁrms.

Turning to acquiring ﬁrm characteristics, Panel B further shows that acquirers in
the exogenous failed deals are large. Whether measured by book value of assets, market
value of assets, or market value of equity, acquirers in the exogenous failed sample are
much larger than those in the failed sample. Moreover, approximately 80% of the sample
acquisitions are proposed by large ﬁrms, which we deﬁne as those with a market
capitalization above the 25th percentile of NYSE ﬁrms (downloaded from Kenneth
French’s data library) in the month prior to the acquisition announcement.13 Given that
most deals in the exogenous sample fail due to regulatory concerns that the acquisition
may decrease market competition, it is not surprising to ﬁnd the existence of large targets
and acquirers and highly valued acquisitions.

In addition, Panel B also shows that sample acquirers are high-growth ﬁrms, as

 

'3 This evidence is not tabulated but available upon request.

13

can be seen from the high Tobin’s q, which is proxied by the market-to-book ratio.l4
There is also evidence that acquirers in the exogenous failed sample do not have very
strong shareholder rights (as measured by the governance index), signiﬁcant operating
income, or higher leverage.15 These factors examined as standard may help explain much
of the variation in bidder gains from acquisitions.

Panel C of Table A.2 presents, by industry, the number of the bidders and targets
in the exogenous failed sample. Industry data are organized using the four-digit
classiﬁcation codes in Fama and French (1997). The last column reports data on the
number of bids originating in the bidder’s own industry. It is evident that the takeover
activities cluster strongly by industry, and that the percent of own-industry bids is high
for some industries such as telecommunications and transportation. This ﬁnding is in line
with previous research on the distribution of mergers by (industry and may further suggest
that regulatory control on market competition in certain industries is active.‘6

Note that the distribution of the exogenous sample over calendar years is also

 

'4 The market-to-book ratio is deﬁned as the market value of equity plus the book value of debt and
preferred stock divided by the sum of book value of equity, debt, and preferred stock one year prior to the
bid. Book value of equity is computed as stated in Cohen, Polk, and Vuolteenaho (2003). Book equity is
deﬁned as stockholders’ equity, plus balance sheet deferred taxes (data item 74) and investment tax credit
(data item 208; if available), plus post-retirement beneﬁt liabilities (data item 330; if available), minus the
book value of preferred stock. Depending on availability, we use redemption (data item 56) for the book
value of preferred stock. Stockholders’ equity used in the above formula is calculated as follows. If
available, we use the stockholders’ equity number in COMPUSTAT (data item 216). If this ﬁgure is not
available, we measure stockholders’ equity as the book value of common equity (data item 60), plus the par
value of preferred stock. If common equity is not available, we compute stockholders’ equity as the book
value of assets (data item 6) minus total liabilities (data item 181). All data items are available in
COMPUSTAT.

'5 The governance index data is from Gompers et al (2003). Operating cash ﬂow is deﬁned as sales minus
the cost of goods sold, sales and general administrative costs, and the change in working capital,
normalized by MVA. As suggested in Bhandari (1988), the debt-to-equity ratio is computed as the book
value of assets minus the book value of equity divided by the market value of equity of the acquirer at the
end of the ﬁscal year prior to the acquisition announcement.

'6 Previous research shows that deregulation became a dominant factor in takeover activities after the late
19805 and accounts for nearly half of the mergers since then. Due to the signiﬁcant regulatory changes for
the Transportation and Telecommunications industry since the 19805, the cluster of mergers in these
industries in our sample is not surprising.

14

investigated but not reported.l7 There is clear evidence that the acquisitions included in
the exogenous failed deal sample ﬂuctuate over time. The number of transactions peaks
in the period between the late of 19905 and 2001. Approximately 68% of the exogenous
failed transactions occur during 1995-2001, as compared with 32% distributed over the
remaining years. This is consistent with the survey paper by Andrade, Mitchell, and
Stafford (2001) that also records similar ﬂuctuations in aggregate merger activities in the
19905. This suggests that restricting the sample to only exogenous failed deals does not
induce selection bias with respect to the distribution of deals, although acquirers/targets

in the exogenous sample are, on average, large ﬁrms.

1.3 Gains to Acquiring-ﬁrm Shareholders

1.3.1 Cumulative Abnormal Returns upon Termination Announcements

We use standard event study methods to estimate the abnormal percentage returns
of acquirers in the exogenous failed deal sample when deal termination is announced
(e.g., Brown and Warner (1985)). Daily return data for the sample of acquirers and the
data for market index returns are mainly collected from CRSP. We estimate acquirer
abnormal returns over the three-day event window (-1, +1) surrounding deal termination
announcements using market model benchmark returns with the CRSP value-weighted
index returns.18 The parameters for the market model are estimated over a period from

250 to 50 days before the initial merger attempt announcement for the target. T-statistics

 

'7 The part of result is available upon request.

‘8 We also calculate the abnormal returns of acquirers included in the sample using the equally-weighted
CRSP market returns in estimation of the market model. The results are almost the same. Further, when we
estimate the abnormal returns by subtracting the value-weighted market return from the acquiring ﬁrm’s
return in the three-day event window around termination announcement, the results become even more
signiﬁcant.

15

are based on tests that standardized prediction errors are equal to zero, as in Campbell, Lo,
and MacKinlay (1997).

Table A.3 presents the three-day cumulative abnormal returns upon termination
announcements of acquirers associated with exogenous deal failures. '9 The value-
weighted abnormal termination return for the sample of exogenous failed deals is 0.22%
and insigniﬁcantly different from zero.20 On average, therefore, shareholders of acquiring
ﬁrms do not beneﬁt from acquisitions. However, after differentiating returns on the basis
of target status and payment type, we ﬁnd that the termination returns are quite different
across deals. The mean termination return of acquirers is signiﬁcantly positive (2.19%)
for public targets, but signiﬁcantly negative (—4.16%) for private targets, and is
signiﬁcantly negative (-2.60%) for subsidiary targets. Moreover, the mean termination
return of acquirers of public targets is roughly 500 basis points higher than the average
termination return for private targets and the difference is statistically signiﬁcant at the
1% level. However, there is no signiﬁcant difference between the termination returns for
acquirers of private targets and acquirers of subsidiaries.

Under the assumption that the exogenous failed sample effectively controls for
information about the acquirer’s valuation at deal termination, any ﬂuctuation in the
bidder’s stock price should reﬂect the effect of the proposed merger for the acquirer
shareholders. Since the termination returns above are measured upon recognition that the
deal will not be consummated and the synergy value will be lost, the results above

indicate that the acquirer shareholders gain on average 3.21% when purchasing private or

 

'9 Hereafter, we abbreviate “cumulative abnormal returns upon termination announcements" to
"termination returns.” '

20 In previous literature, research investigating bidder announcement returns also ﬁnds similar results. For
example, Andrade et al. (2001) report insigniﬁcant negative abnormal returns from 1973 through 1998.

16

subsidiary targets while lose —2.l9% when purchasing a public ﬁrm.” These results are
consistent with prior theoretical and empirical research that implies bidder gains from
acquisitions vary with the type of assets acquired.22 In particular, these results are in line
with the literature on asset sales that suggests owners of private ﬁrms and subsidiaries
may be more likely to accept a discounted deal for liquidity needs or restructuring
reasons.23

Next we analyze the relation between termination returns and method of
payment.24 Table A.3 also presents the termination returns of acquires by method of
payment. Evidence shows that regardless of payment method, the mean termination
returns of acquirers of private or subsidiary targets are always signiﬁcantly negative
while the termination returns of acquirers of public ﬁrms are all signiﬁcantly positive.
This result suggests that the observed positive wealth effect in takeovers of non-public
targets cannot be fully explained by the payment effect. Moreover, it is evident that for
all deals, cash offers generally do not correspond to higher retums than equity offers. One

exception is that when the target is a subsidiary, the bidder’s termination return is

signiﬁcantly higher when the bid is ﬁnanced with cash versus stock. This suggests that

 

2' There is some concern about whether deal termination is anticipated before the public announcement.
Speciﬁcally, since acquisitions of public targets usually receive more news coverage, termination
announcements might be more anticipated for acquisitions of public targets than for those of private targets.
As a result, the termination returns of acquirers of public targets would be pulled toward to zero compared
to those of acquirers of private ﬁrms. If this were to the case, the termination returns for acquirers of public
targets would be even more negative.

22 Related papers include Grossman and Hart (1980), Zingales (1995), Hansen and Lott (1996), Chang
(2004), Fuller, Netter, and Stegemoller (2002), Moeller, Schlingemann, and Stulz (2004). However,
different from these prior papers, the methodology in this study addresses biases in conventional techniques
and permits a cleaner measure of the market assessment of value improvements from acquisitions. Thus,
our results are more statistically signiﬁcant and economically important.

23 See, e.g., Maksimovic and Phillips (2001 ), Poulsen and Stegemoller (2008), Warusawitharana (2008) and
Yang (forthcoming).

3’ Earlier research suggests that the method of payment plays an important role in explaining the abnormal
returns of acquirers. See. for example, Travlos (1987). Chang (1998), and Fuller, Netter, and Stegemoller
( 2002).

17

equity offers may yield higher gains to bidders when acquiring subsidiaries. Moreover,
the difference in termination returns of acquirers of various types of targets becomes even
larger when the deal is ﬁnanced with equity. For example, termination returns for
acquirers of public targets are roughly 400 basis points higher than those for acquirers of
non-public targets in cash offers, but the difference rises to approximately 800 basis

points when equity is used.

1.3.2 Concerns on Returns upon Merger Announcements

Table A.3 above shows how acquirer termination returns vary by the type of
target acquired and the method of payment. There is some concern, however, about
whether the same patterns would be obtained upon acquisition announcements for the
exogenous failed deals. Examination of cumulative abnormal returns upon acquisition
announcements for the sample acquisitions could offer additional insights on the
effectiveness of the methodology employed in this paper.25

Table A4 shows the acquirer announcement returns by target type and form of
payment. We ﬁnd that the value-weighted abnormal announcement returns for the sample
of exogenous failed deals is insigniﬁcantly negative. On average, therefore, shareholders
of acquiring ﬁrms do not beneﬁt from acquisitions. Consistent with previous studies, we
ﬁnd that acquirers of public targets generally do not gain, but lose signiﬁcantly when
using stock as payment. The mean announcement return for acquirers of private targets is
generally positive and signiﬁcant. However, for acquisitions of subsidiaries, no

signiﬁcant positive announcement return is found regardless of payment method. Also

 

25 ' ‘5 ' ‘ ‘ ' Q!
Hereafter, we abbrevrate cumulative abnormal returns upon acqursrtron announcements to

“announcement returns.”

18

consistent with prior literature, we ﬁnd that acquirers in stock deals experience more
negative returns than those in cash deals. Based on this table alone, one might conclude
that stock deals are systematically poorer deals for bidders. However, combined with the
evidence presented in Table A.3, these results actually Show a different story, that is, the
previously documented lower returns to equity offers merely reﬂect the negative
revelation of the bidder’s stand-alone value, not the gains from combination. Moreover,
different from the evidence in Table A4, there is no signiﬁcant variation in acquirer
announcement returns by different types of target. This shows the problems in
conventional techniques that focus on examining announcement returns only.

As discussed above, acquisition announcements usually release more information
about the stand—alone value of the acquiring ﬁrm than the synergy of combination. This
may cause the unexplained variation in bidder announcement returns recorded in
previous research. However, the methodology introduced in this paper to examining the
termination returns for a speciﬁed failed sample permits a cleaner assessment of bidder
gains from acquisitions. We attribute this improvement to the research design that

controls for biases due to multiple information release.

1.3.3 Effect of Firm and Deal Characteristics on Bidder Gains

Univariate evidence on the effects of target type and payment form on bidder
gains is presented above in Table A.3. However, as documented in prior research, various
ﬁrm and deal characteristics help contribute to bidder gains. In order to further explore
the relation between bidder gains in acquisitions and various ﬁrm and deal characteristics,

we ﬁrst examine how those characteristics vary across different types of targets and then

19

discuss possible impacts of various characteristics on acquirer termination returns in
multivariate regressions.

Table A5 reports ﬁrm and deal characteristics for sample acquirers sorted by
target type. Panel A shows that the transaction value is much larger for acquisitions of
public targets than acquisitions of private targets or subsidiaries. This is not surprising
since private targets are, on average, much smaller than public targets. Moreover, relative
to bidders’ market value of assets (equity), public targets are much larger. The relative
size of the target to the bidder is also signiﬁcantly different between deals with public
targets and those with non-public targets. The effect of relative size is documented by
prior work, though there is tremendous variation in this effect. For example, Asquith,
Bruner, and Mullins (1983) ﬁnd that bidder returns increase with the relative size of
target market capitalization to bidder market capitalization. However, using a different
sample, Moeller, Schlingemann, and Stulz (2004) report that relative size is unrelated to
the returns of bidders acquiring public ﬁrms. Thus, further investigation in the paper may
help shed light on this debate.

Turning to additional results on deal characteristics, Panel A of Table A5 also
shows that deals with multiple bids are more frequent for acquisitions of public ﬁrms
than for those of private ﬁrms.26 Equity is used more frequently in the acquisitions of
public ﬁrms, especially compared to takeovers of subsidiaries. Further, we ﬁnd that
offers for public targets are twice as likely to be hostile as those for non-public targets

(though few offers are hostile in our sample), and acquirers are more likely to make

 

3" However, this measure may suffer from the possibility that many potential bidders are not included.
especially when private auctions take place (see, e.g.. Boone and Mulherin (2007)). Thus, the number of
actual bidders for the target may be understated.

20

tender offers when purchasing public targets.27 Finally, we ﬁnd that acquisitions of public
or subsidiary targets, compared to offers for private targets, are half likely to be
conglomerate deals.28 Prior work has documented that the variation in deal characteristics
may help explain the differences in bidder gains, therefore the further multivariate
examination is necessary.29

Panel B of Table A5 indicates that, for the exogenous failed sample, acquirers of
public targets are much larger than other bidders in terms of book value of assets or
market value of equity. Even when compared to the lower 25th percentile of NYSE ﬁrms
in the same year, acquirers of public ﬁrms are half as likely to be small. Moreover, the
ratio of operating cash ﬂow to market value of assets is higher for acquirers of public
targets. Free cash ﬂow theory predicts that ﬁrms with poor investment opportunities and
excess cash are more likely to make poor acquisitions. For this reason, we also compare
the acquirer’s Tobin’s q, deﬁned as the market value of total assets divided by the book
value of assets. Existing evidence argues that ﬁrms with higher q values make better
acquisitions. The panel shows that acquirers of public targets have lower Tobin’s q values.
As mentioned above, the differences in these characteristics for acquirers of different
targets may help explain the variation in abnormal returns of acquirers in our sample.

The comparisons in Table A.3 only consider how bidders’ abnormal returns vary
with method of payment and target status. To take into account other determinants of
acquirers’ abnormal returns, we estimate cross-sectional regressions on a set of control

variables for ﬁrm and deal characteristics.

 

27 Prior work suggests that acquirers achieve lower returns when ﬁnancing acquisitions with stock and that
the hostile offers have lower bidder’s returns (e.g., Schwert (2000)).

28 Conglomerate deals are deﬁned as acquisitions involving a target with a different two-digit SIC code
from that of the bidder.

29 Other related papers include Morck, Shleifer and Vishny (1990), Maquieria. Magginson, and Nail (1998).

21

Table A.6 presents the results of multivariate regressions. The dependent variable
in all regressions is negative one times the acquirer’s 3-day percentage abnormal returns
upon termination announcement. Since the termination return measures the bidder’s
value change around the deal cancellation, the positive (negative) abnormal return to the
termination announcement implies that acquiring-ﬁrm shareholders actually lose (gain)
from the proposed merger if the deal goes through. Thus, using the negative of the
termination return as the dependent variable, we can examine the relation between gains
to acquiring ﬁrms and various ﬁrm and deal characteristics that are known to affect the
bidder returns from acquisitions. All regressions also use year and industry ﬁxed effects
at the two-digit SIC code level.

Regressions (1)-(3) use all the acquisitions in the exogenous failed sample,
regardless of the status of target acquired, and control for both acquiring-ﬁrm and deal
characteristics. To capture the size effect, we use a dummy variable that equals one when
the acquirer is smaller than the 25th percentile of NYSE ﬁrms in same year, the natural
logarithm of the acquiring ﬁrm’s market value of equity, and the natural logarithm of the
acquirer’s book value of assets. All else equal, acquirers of private ﬁrms and subsidiaries
have signiﬁcantly larger gains than acquirers of public targets. Controlling for ﬁrm and
deal characteristics, we ﬁnd that bidder returns in acquisitions increase by roughly 6%
when ﬁrms choose to buy private ﬁrms or subsidiaries as opposed to public ﬁmis. This
ﬁnding is consistent with results in previous studies that also show signiﬁcant differences
in gains to acquirers when purchasing public targets versus private targets or subsidiaries.
Whether an acquisition is paid for with equity or how much equity ﬁnancing is used in

consideration is not correlated with the bidder’s gains from acquisitions when the entire

22

exogenous sample is examined.

To adjust for the impact of an acquisition on the market capitalization of the
acquirer, regressions of bidder returns generally control for the relative size of the target
to the acquirer. In existing literature the relative size is often signiﬁcant, but the sign of
the coefﬁcient varies across studies. Asquith et a1. (1983), for example, show that the
bidder’s gains increase with the relative size of the target to the bidder, while Travlos
(1987) ﬁnds the opposite. Here we deﬁne the relative size as the transaction value of the
target divided by the market capitalization of the acquirer as of one month prior to the
acquisition announcement. The coefﬁcient on this variable is signiﬁcantly negative (-0.5 8)
and is not sensitive to the alternative speciﬁcations of the acquirer’s size. This result
suggests that acquisitions of small targets by large bidders will tend to generate greater
gains per dollar spent on acquisitions than combinations of similar-sized ﬁrms. One
plausible explanation for this is that so-called “mergers of equals” are hard to implement
successfully.3O

Though the signs of the coefﬁcients on the other control variables are similar to
those of earlier studies, with some exceptions, none of these coefﬁcients turns out to be
signiﬁcant. Evidence from conventional studies that are centered on acquisition
announcement returns generally shows that equity offers, hostile offers, or diversifying
offers are associated with lower bidder returns. However, our results indicate that these
effects merely reﬂect differences in signals about the bidder’s stand-alone value, not
differences in the gains from acquisitions. For example, cash offers on average are

associated with higher bidder returns than equity offers or mixed-payment offers. In

 

3“ On January 9, 1999, The Economist mentions that “ ...... mergers of equals seem to be especially tricky,
perhaps because they disrupt two strong corporate cultures. and they often throw up intractable problems of
leadership.”

23

contrast, based on the new methodology used in this paper, which effectively controls for
the biases caused by the release of information on the acquirer’s valuation, equity offers
do not lead to lower gains to bidders than cash or mixed offers. This ﬁnding suggests that
the apparent superiority of cash offers in creating shareholder value is an illusory
consequence of a more negative revelation effect for equity or mixed offers.

The proxy for Tobin’s q has a negative and insigniﬁcant coefﬁcient, which is a
little surprising since earlier studies (e.g., Lang et a1. (1989)) show higher returns to
bidders with higher q values. However, the effect as shown is economically trivial. The
coefﬁcient on leverage, deﬁned as the ﬁrm’s total debt over the ﬁrm’s market value of
equity at the end of the ﬁscal year prior to the acquisition announcement, is insigniﬁcant
for all deals in the exogenous sample, although the coefﬁcient becomes signiﬁcantly
positive when only acquisitions of public ﬁrms are examined. The governance index has
a negative coefﬁcient, which suggests that strong shareholder rights (lower governance
index) are linked with higher bidder gains from acquisitions, but the effect is insigniﬁcant.

We further estimate the regression for samples sorted by the organizational form
of the target acquired. The relative size effect still prevails in acquisitions of public or
private ﬁrms but is very weak in acquisitions of subsidiaries. The coefﬁcients on the
other variables remain the same as in the regressions for all deals in the sample. However,
the coefﬁcient on equity in consideration, deﬁned as the percentage of equity in
consideration, becomes signiﬁcantly positive for acquisitions of subsidiaries only. This is
consistent with Hansen (1987), who shows that the bidder should use stock ﬁnancing for

better returns when there is greater uncertainly about the target’s valuation.

24

1.3.4 Robustness

Using a methodology that effectively controls for the biases caused by multiple
information release, we ﬁnd that the gain to acquiring ﬁrms is related to the
organizational form of targets. The main ﬁnding is that acquirers gain signiﬁcantly when
purchasing private ﬁrms or subsidiaries and lose signiﬁcantly when acquiring public
ﬁrms, after controlling for variables that are known to predict bidder returns.

In this subsection we consider a few tests designed to verify the robustness of this
ﬁnding. To conserve space, the robustness test results are not tabulated but are available
from the author upon request.

The exogenous failed sample contains all the deals that fail because of
disapproval by regulatory agencies, occurrences of competing offers, or unexpected
lawsuits or court rulings. There is some concern about including deals that fail because of
competing bids in the analysis. One may argue that the occurrence of a competing bid
reveals negative information about the initial bidder. If the market is efﬁcient,
information about the initial acquirer should be fully incorporated by the market at the
time of the release of the originally proposed merger, in which case the exit of the initial
bidder following a higher bid from a rival ﬁrm may merely reﬂect the acquirer’s
unwillingness to overpay for the target. Under this interpretation, such deals should be
considered in the analysis.” However, if market is not fully efﬁcient and there is a slight
chance that an emerging competing bid causes the termination of the initially proposed
merger, revealing negative information about the ﬁrst acquirer, our results may

understate the shareholder’s returns from acquisition. To check for this, we estimate all

 

3' By searching the statements made by acquirers at the time of deal termination, most ﬁrms explicitly
express that their unwillingness to overpay is the main reason for calling off the deal.

25

the regressions after excluding those deals that fail due to competing bids. We ﬁnd that
the results are not sensitive to the inclusion of such deals, although the statistical
signiﬁcance drops somewhat due to decreased sample size.

Another potential concern with respect to sample selection bias is that a deal
failure might signal an adverse industry shock, which in turn could negatively affect the
abnormal return of the acquirer. We attempt to investigate the merger and acquisition
activities in the acquirer’s industry after the failure of the sample deals. The evidence
shows that there is no signiﬁcant drop in the number of announced merger offers in the
same industry two years following the deal failure.

As shown above, the coefﬁcient on relative size is signiﬁcantly negative, which
implies that the greater the target’s market capitalization is relative to the acquirer’s
market size, the less the gain to bidders from acquisitions. There is also a possibility that
the relation between bidder gains and relative size of targets to acquirers may not be
linear.

In order to better capture the relative size effect, we ﬁrst add to the regressions a
variable measuring the square of relative target size and ﬁnd that the coefﬁcient on the
variable is signiﬁcantly negative while the coefﬁcient on the relative size variable
continues to remain negative. This result suggests that the relation between bidder returns
from acquisitions and the relative size of the target probably follows an inverse U-shape.

Next, we estimate the regressions including a dummy variable that equals one if
the relative size of the target is above the sample median. We ﬁnd that the coefﬁcient on
this relative size dummy is statistically negative and the coefﬁcients on the other

variables are similar to those above. This result is consistent with the previous ﬁnding

26

that combinations of similarly sized ﬁrms generate less gain to acquiring-ﬁrm

shareholders.

1.4 Discussions on Gains to Acquirers of Non-public Targets

1.4.1 Existing Alternative Explanations

The results above show that the bidders gain signiﬁcantly in takeovers of private
targets or subsidiaries, which echo previous ﬁndings.32 However, we still know little
about why there are fundamental differences in bidder gains between acquisitions
involving public and private targets. The two main possible explanations for the
signiﬁcant gains to bidders when acquiring non-public ﬁrms are as follows.

The ﬁrst plausible explanation is that managers of private ﬁrms or subsidiaries are
willing to trade at a discounted price for access to liquidity. Private ﬁrms and subsidiaries
are not as easily bought or sold as are publicly traded ﬁrms. The lack of liquidity makes
investments in those non—public assets less attractive and thus less valuable than similar,
more liquid investments. According to this argument, private target managers might want
to sell at a discounted price due to a desire to cash out quickly or a desire to facilitate the
ﬁrm’s transition from private to public. Ofﬁcer (2007) shows that non-public targets are
acquired at a 15% to 30% discount relative to the average acquisition multiple paid for
comparable publicly held targets. The liquidity discount is caused by the need for, and the
availability of, the liquidity provided by the acquirer. The bidder’s signiﬁcant gain from
acquisitions of non-public ﬁrms, therefore, is the retum on offering liquidity service.

Related to this argument, while sales of public targets are typically auction-like in

 

32 See, e.g.. Chang (1998). Fuller, Netter, and Stegemoller (2002). and Faccio, McConnell. and Stolin
(2006)

27

nature, with full disclosure required by the SEC, the sales process can vary substantially
for private targets or subsidiaries. Most likely, the sales of private targets or subsidiaries
often go through limited auctions or with few interested bidders. In this case bidders are
likely to have a bargaining advantage in the acquisitions of non-public targets, which
enable them to gain more from the transactions.

The other explanation comes from the literature on assets sales. One reason
suggested for why a ﬁrm sells a subsidiary is the gain from increased focus. This implies
that diversiﬁed ﬁrms might accept a relatively lower price for an asset sale than a non-
diversiﬁed ﬁrm. The diversiﬁcation discount, therefore, can help explain the bidder’s

gains from acquisitions of subsidiaries.

1.4.2 Gains to Acquirers and Characteristics. of Target Parents

As discussed above, liquidity needs or the bargaining power of targets may have a
signiﬁcant impact on the gains to acquirers of non-public targets. To investigate the effect
of liquidity on the gains to acquirers of non—public targets, while we note that it is hard to
ﬁnd an accurate quantitative measure of liquidity needs or the bargaining power of
targets, in this paper we use the public status of selling parent ﬁrms as a proxy for their
needs for liquidity. This analysis requires information on parents of non-public
acquisition targets.

We ﬁrst use a dummy variable that equals one if the parent of the target is private.
The liquidity discount is caused by the need for, and the availability of, the liquidity
provided by the acquirer. Private parents are less liquid compared to public parents, so

that in sales of targets, keeping everything else the same, private parents are likely to

28

accept a greater discount for access to liquidity offered by acquirers. Thus, the private
parent dummy can help explain the effect of the liquidity discount on bidder gains. Since
parents of private ﬁrms are always private, the deals studied here include only those
offers for subsidiaries. The public status of parents of subsidiaries comes from SDC. To
avoid possible data error in SDC reports, we also manually check the organizational
forms of parents of subsidiaries.

Next, we construct a dummy variable that equals one if the parent of the target
subsidiary hires ﬁnancial advisors in the process of the merger deal. If the subsidiaries
hire ﬁnancial advisors, they may be able to more readily promote an auction-like
transaction, with participation by a large number of qualiﬁed bidders. The bargaining
power of subsidiaries increases with the participation of more acquirers and therefore
decreases bidder gains from acquisitions of subsidiaries.

Finally, we use a third dummy that indicates whether the parent of the target
subsidiary is diversiﬁed or not. A diversiﬁed parent is deﬁned as a parent whose two-
digit SIC code is different from that of the subsidiary. As mentioned above, diversiﬁed
ﬁrms might accept a relatively lower price for an asset sale than a non-diversiﬁed ﬁrm. If
this is the case, the bidder’s gain should be higher when buying a subsidiary from a
diversiﬁed ﬁrm than when buying a subsidiary from a non-diversiﬁed ﬁrm.

Table A7 shows the acquirer’s abnormal returns at termination announcements
for acquisitions of subsidiaries, sorted by the target-parent characteristics above and the
method of payment. The results Show that the bidder’s abnormal return upon termination
announcement is more negative and signiﬁcant when the subsidiary target has a private

parent. When parents of subsidiaries hire no ﬁnancial advisors to promote the deal, the

29

acquirer’s termination return is more negative and signiﬁcant. However, there is no
signiﬁcant difference in acquirer termination returns based upon whether parents of
targets are diversiﬁed or not.

Table A8 presents the cross-sectional regression results. The dependent variable
is negative one times the acquirer’s 3-day percentage abnormal return upon termination
announcement, which is a proxy for acquiring-ﬁrm gains from takeovers. The results
show that after controlling for ﬁrm and deal characteristics, the bidder’s gain from
acquisitions of subsidiaries is higher when the assets acquired are owned by a private
ﬁrm. This suggests that private parents might tend to sell at a more discounted price so
that the bidders beneﬁt more from takeovers of those discounted assets. Under the
assumption that private parents are generally less liquid compared to public parents, this
ﬁnding supports the hypothesis that the liquidity need of non-public targets could be the

reason underlying higher bidder gains from acquisitions of non-public ﬁrms.

1.5 Conclusion

Over the past three decades, takeover activity has attracted a great deal of
academic attention. A fundamental debate among researchers and practitioners concerns
the impact of takeovers on the wealth of acquiring-ﬁrm shareholders. To date the
literature has offered inconclusive results about bidders’ value improvement from
takeovers and researchers have been unable to successfully explain the tremendous
variation observed in bidder returns around takeover announcements. The main goal of
this paper is to re-examine whether acquisitions create value for acquiring-ﬁrm

shareholders.

30

 

— b _—- 4 w .......

The conventional approach, which is centered on the abnormal acquisition
announcement returns in mergers, is complicated by the fact that the announcement of a
takeover reveals more information than the potential synergy from the transaction. It is
therefore difﬁcult to interpret the announcement return for acquiring ﬁrms and to achieve
unbiased estimates of the value gains of bidders from takeover activities. To avoid the
problem of multiple information release at the time of announcement, we introduce a
methodology that employs a sample of takeovers that fail due to exogenous reasons and
uses termination returns as a proxy for bidder gains. Since the sample construction
excludes deals whose termination is related to the valuation of the acquirer, the abnormal
return of an acquiring ﬁrm at the time of deal termination reﬂects the bidder’s potential
gain from the acquisition if it goes through. Using the termination return as a proxy for
bidder gains from acquisitions, we can also examine the relation between acquiring—ﬁrm
gains and various ﬁrm and deal characteristics.

Speciﬁcally, we use a hand-collected sample of 272 takeovers that failed due to
exogenous reasons, and ﬁnd that the bidder’s shareholders gain from acquisitions when
purchasing a private ﬁrm or a subsidiary, while they lose when the acquisition involves a
public target, even after controlling for ﬁrm and deal characteristics. Bidders acquiring
private or subsidiary targets perform signiﬁcantly better than those acquiring public
targets. This result is robust to a variety of speciﬁcation choices and sample selection
criteria. In addition, bidder gains are not signiﬁcantly lower when stock is offered. This
evidence suggests that the previously recorded lower announcement returns in stock
offers are merely a reﬂection of market adjustment to bidders’ over—valued stock. We

also ﬁnd that the relative size of the target to the bidder is a signiﬁcant determinant of

31

bidder gains: the greater is the relative size of the target, the less gain to bidders from
acquisition activity. This ﬁnding offers support to recent concerns about so-called
“mergers of equals.”

This paper also addresses possible explanations for the considerably positive gain
of bidders when buying non-public targets and analyzes bidder gains in several cases in
which the characteristics of the targets’ owners are different. The evidence shows that
bidder gains are signiﬁcantly higher when bidders acquire assets from private owners
than when they acquire assets from public owners. This supports the idea that the
liquidity discounts are the price paid by private owners or corporations for the liquidity
provided by acquirers. The acquirers, therefore, beneﬁt from providing this liquidity
service to the owners of non-public targets.

In summary, the current paper adds to the existing literature on bidder returns
from acquisitions by providing a new methodology to control for the problem of multiple
information release at the time of announcement. The evidence we present extends the
important ﬁndings of Fuller, Netter, and Stegemoller (2002) and Faccio, McConnell, and
Stolin (2006) by showing that bidder gains vary with the status of targets acquired in
takeovers. This paper also offers some evidence that bidder gains are greater when the
relative size of the target to the bidder is smaller. In addition, the ﬁndings add to the body
of evidence indicating that the price paid to access liquidity by selling a part, or the whole,
of a ﬁrm is reﬂected in the discounted sale price, and potentially in the higher gains

accruing to the acquirer.

32

CHAPTER 2

Cash Flow and Product Market Competition:
An Empirical Analysis

2.1 Introduction

The nature and extent of the interaction between ﬁnancial markets and product
markets is of great importance to corporate ﬁnance research. Numerous theoretical works,
starting with Titman (1984) and Brander and Lewis (1986), have been devoted to
analyzing how ﬁnancial choices affect output market behavior. While a large part of
ﬁnance research has considerably enriched our understanding of the link between debt
ﬁnancing and ﬁrms’ competitive strategies, existing empirical research offers little
evidence on the potential effect of ﬁrms’ cash on product market competition. This paper
attempts to bridge part of that gap.

From an intuitive as well as a theoretical viewpoint, cash may inﬂuence a ﬁrm’s
product market choices and that of its competition. Several studies, starting from Bolton
and Scharfstein (1990), have argued that cash-rich ﬁrms may take actions to decrease the
output price and drive their ﬁnancially constrained competitors out of business. More
generally, rich cash enables a ﬁrm to employ a number of altemative competitive policies
other than pricing against rivals such as active advertising, commercial promotions, large
distribution networks, or productive staffs. Accordingly, cash is important for product
market competition. In addition, recent evidence shows that cash and debt are not

equivalent with existing ﬁnancial frictions. The imperfect substitutability implies that

33

cash may play a distinct role in inﬂuencing a ﬁrm’s competitive outcomes.l

Previous empirical assessments on cash ﬂows are often trapped by the fact that a
ﬁrm’s cash ﬂow is correlated with proﬁtability. A common approach to addressing this
methodological problem is to identify variations in cash ﬂow that are independent of
investment opportunities. For example, Blanchard, Lopez-de-Silanes, and Shleifer (1994)
and Rauh (2006) take advantage of exogenous shocks from unique events or institutional
features that change the cash position of ﬁrms. Alternatively, Lamont (1997) exploits the
presence of the internal capital market and assumes that a cash ﬂow shock to one division
of a ﬁrm may be exogenous to the investment opportunities in other divisions.

Inspired by the latter approach, we start the investigation on the effect of cash on
competition by identifying a unique negative shock to cash ﬂows in the tobacco industry
due to the potentially higher federal excise tax. This event can be used as an ideal natural
experiment for three reasons. First, tax changes are generally regulated and released by
federal or state authorities and hence are unambiguously exogenous to any individual
ﬁrm’s actions. Second, excise tax on tobacco products should have no effect on the
proﬁtability of investments in other industries, while expected cash ﬂows in tobacco
ﬁrms would be decreased. This promises that the cash flow shock is exogenous to ﬁrms’
non-tobacco segments. Last, tax events usually have a signiﬁcant impact on the ﬁrms’
performance, which allows for a powerful test.

To examine the connection between cash and product market competition, we
ﬁrst examine the share price response of industry rivals that compete with non-tobacco
divisions of tobacco ﬁrms that experience a negative cash shock. After properly

controlling for investment opportunities in non-tobacco industries, if a decrease in cash

 

' See. for example, Stein ( 2003).

34

has no causal effect on product market competition, we would expect to observe no
abnormal returns to the industry rivals. In a sample of 436 industry rivals that compete in
non-tobacco industries with tobacco ﬁrms, we ﬁnd that on average those rivals
experience positive and signiﬁcant abnormal returns. This indicates that the market
believes a negative cash shock to one ﬁrm is good news for its rivals. The result suggests
that internal funds may affect a ﬁrm’s product market behavior signiﬁcantly and that a
decrease in cash may “soften” the product market competition. Since the cash shock is
exogenous to the investment opportunities of non-tobacco industries, our evidence
supports the hypothesis that there is a causal relation between a ﬁrm’s cash ﬂow and its
product market competition.

To reinforce the strength of our results, we then proceed to employ a number of
robustness checks and explain the abnormal returns of rivals as a function of ﬁrm and
industry characteristics. The results are robust to a variety of estimation issues and
identiﬁcation concerns. In addition, we ﬁnd that the competitive effect of cash is
magniﬁed in competitive industries and when cash-constrained ﬁrms only take a small
market share. These ﬁndings provide evidence on whether and how industry competition
characteristics affect cash-competition sensitivity.

Next, we study how output behavior varies with the exogenous shock to cash that
can be contributed to ﬁnance its production. In particular, we examine the raw and
industry-adjusted variations in the output level of the non-tobacco segments in tobacco
ﬁrms that were affected by the negative cash shock between the year prior to the shock
and each of the ﬁve years following the shock. we observe that the industry-adjusted

sales and cash ﬂows of the non-tobacco segments in tobacco ﬁrms drop signiﬁcantly over

35

the ﬁve years after the shock. This is consistent with the idea that a decrease in internal
funds may increase the probability of liquidation for any given level of production and
hence causes the ﬁrm to produce less. Moreover, it is evident that the number of non-
tobacco segments that tobacco ﬁrms operate decreases following the cash ﬂow shock.
This suggests that when facing tightening ﬁnancial constraints, ﬁrms may produce less
and even exit certain business markets.

Lastly, we examine the abnormal customer returns to those industry rivals that
compete with non-tobacco divisions of tobacco ﬁrms. This step may help to better
understand how ﬁrms change their competitive strategies after a negative shock to cash.
On average, customers of rival ﬁrms experience positive average returns at the
announcement of cash shock to the tobacco industry. This is inconsistent with the idea
that overall product price increases after the cash shock affects the competition. Actually,
it suggests the possibility that some customers may beneﬁt from the new market
equilibrium.

Overall, this paper contributes in two dimensions. First, this study complements
the existing evidence relating ﬁnancial decisions and corporate strategy. A group of
theoretical literature, notably Titman (1984) and Brander and Lewis (1986), examines the
interactions among ﬁrms in output markets and their ﬁnance choices. However, a large
portion of empirical research in this area focuses on the association between debt and
product market strategy (e.g., Chevalier (1995a, 1995b), Phillips (1995), Zingales (1998),
etc). Certainly, by demonstrating that cash ﬂow of ﬁrms affects product market actions,
the present work “points out a more complex relation between ﬁrms’ ﬁnancial and

operating decisions.

36

Second, this paper extends our understanding of the implications of corporate
cash. The observed stockpiles of cash in US. ﬁrms draw academic attentions to
rationales for such cash-rich status. The earliest explanations offered by academic
research were based on trade—offs motivated by transaction costs. In particular, Baumol
(1952) and Meltzer (1963) suggest that ﬁrms hold cash to avoid the cost of a potential
short in liquid assets. Furthermore, Opler, Pinkowitz, Stulz, and Williamson (1999) and
Almeida, Campello, and Weisbach (2004) argue that the recent stockpiles of cash are
results of increases in precautionary motives to hold cash. Different from these previous
studies, our ﬁndings highlight the substantial impact of corporate cash on competitive
outcomes, and suggest that ﬁrms can hold cash for the incentives created by their needs
in product market competition. Thus, the strategic effect of cash can also be a
determining factor for managers in assessing the optimal level of cash holdings for ﬁrms.

The remainder of this paper is organized as follows. Section 2 reviews the related
literature and develops the main hypothesis. Section 3 describes empirical testing
methods, data, and sample selection. The main analysis results are presented in Section 4.

Finanlly, conclusions and possible extensions are discussed in Section 5.

2.2 Related Literature and Hypotheses Development

2.2.] Financial Constraints, Internal Funds and Product Market
Competition
A large portion of the literature in corporate ﬁnance and economics is devoted to

understanding how ﬁnancial constraints affect a ﬁnn’s output market behavior and that of

its competitors (see Maksimovic (1995); Parson and Titman (2007)). Two ideas play a

37

central role. First, ﬁnancial constraints alter a ﬁrm’s incentive to compete in the product
market. As discussed in Gale and Hellwig (1985) and Bolton and Scharfstein (1990), a
ﬁnancially constrained ﬁrm has a limited access to funds, and hence may have an
incentive to behave more cautiously in its output market. In contrast, Brander and Lewis
(1986) and Hendel (1996) suggest that a ﬁrm incurring debt has an incentive to mitigate
the risk of bankruptcy and thus may adopt, ex post, more aggressive output market
behavior in the form of high output or low prices.

Second, ﬁnancial constraints affect a ﬁrm’s interaction with its competitors.
Tesler (1966) and later Bolton and Scharfstein (1990) argue that cash-rich ﬁrms may take
actions to decrease the output price, and drive their ﬁnancially constrained competitors
out of business by reducing their rivals’ cash ﬂow. Consequently, liquidity constraints of
a cash-poor ﬁrm may prompt cash-rich rivals to adopt “predation” behavior and compete
more aggressively in the product market. Moreover, in a similar vein, a ﬁrm’s liquidity
situation may signal the possibility of future competitive behavior, and thus has an
indirect inﬂuence on competitors’ strategy. For example, consistent with the predation
theory, Benoit (1984) argues that cash-rich incumbents may be capable of detening entry
if potential entrants face ﬁnancial constraints. Contrary to the traditional view, in a recent
paper, Hege and Henessy (2007) suggest that unleveled incumbents may actually prompt
entry since they prefer to acquire entrant assets instead of practicing predation.
Regardless of the different arguments, these papers imply that a ﬁrm’s ﬁnancial
constraints may have a causal impact on the competition behavior of its own and of its
industry rivals.

Surprisingly, given the broad theoretical interest in the relationship between ﬁrm

38

liquidity and product market competition, most prior empirical work solely concentrates
on linking a ﬁrm’s output and pricing behavior to its debt ﬁnancing. Highly indebted
ﬁrms are assumed to have a constrained capacity of raising additional funds, which in
turn distorts their competitive strategy, especially against unlevered rivals. In such a
context, some research documents that high leverage leads to poor performance in the
product market, such as sales and market share decline, lower product price, or exit from
a previous market (see Chevalier (1995a, 1995b), Zingales (1998), Khanna and Tice
(2000) and Campello (2003)), while other studies ﬁnd that indebtedness increases ﬁrms’
aggressiveness in the output market competition (see Lyandres (2006) and Campello
(2006)).

In these studies, most rely on natural experiments involving shocks to either a
ﬁrrn’s leverage ratio or its product market environment. For instance, Phillips (1995) and
Chevalier (1995a, 1995b) empirically investigate ﬁrms’ competitive responses to sharp
increases in leverage. Subsequently, Chevalier and Scharfstein (1996), Zingales (1998),
Karma and Tice (2000), and Campello (2003) analyze shocks to competitive
environments, examining how differences in ex ante capital structure are associated with
differential responses and competitive outcomes. Unfortunately, while those papers
document the interaction between a ﬁrm’s debt and its competitive outcomes, the causal
connection between the two is hard to establish, since leverage is usually chosen in
advance and the shocks examined just ameliorate to some extent the endogeneity problem.

Given the existing empirical evidence, our understanding of the relationship
between ﬁnancial constraints and product market competition is far from complete. There

is very little work that examines whether or how cash (whether measured as a ﬂow or a

39

stock) affects ﬁrrns’ output behavior. It is surprising given the surging interest in the role
of ﬁrms’ cash. In particular, recent evidence from several perspectives calls for
indispensable attention on the role of cash in explaining product market performance and
actions. This observation motivates the empirical tests in this paper.

First and most importantly, internal and external funds are different. If there are
no various agency and information problems, Modigliani and Miller (1958) predict no
differential costs of internal and external ﬁnance. Nonetheless, as Stein (2003) discusses,
a cost wedge between internal and external funds may often come from information
asymmetry as discussed in Myers and Majluf (1984) and Greenwald, Stiglizt, and Weiss
(1984), or incentive and agency problems as in Jensen and Meckling (1976), Stulz (1990),
and Hart and Moore (1995). Those frictions can result in an outcome that internal and
external funds are not perfect substitutesz Furthermore, Acharya, Almeida, and Campello
(2007) show evidence from a hedging perspective that cash is not equivalent to debt in
presence of ﬁnancing frictions. Consequently, it is likely that cash and debt may play
distinct roles in inﬂuencing a ﬁrm’s output behavior, and hence the supply of internal
funds may have a signiﬁcant impact on ﬁrms’ competitive outcomes. However, if one
believes on a priori ground that cash and debt are valid substitutes (at least to some
extent), research on the effect of cash on output behavior can also add our understanding
of the literature on the relationship between debt and product market competition.

Second, it is recorded that US. corporations hold signiﬁcant amounts of cash on
their balance sheets and the holdings have been recently increasing. Despite a growing

literature focusing on corporate liquidity, our understanding of the causes and

 

2 Other related papers include Grossman and Hart (1986). Gertner. Scharfstein, and Stein (1994). and Stein
(1997)

40

consequences of corporate cash holdings is still far from complete. The existing empirical
investigation has attempted to explain cash by agency considerations (Jensen (1986) and
Blanchard, Lopez-de-Silanes, and Shleifer (1994)), transaction costs (Mulligan (1997)),
precautionary motives (Opler, Pinkowitz, Stulz, and Williamson (1999) and Bates, Kahle,
and Stulz (2008)), or tax costs (Foley, Hartzell, Titman, and Twite (2007)). However,
little attention has been paid to the incentive created by the potential impact of a ﬁrm’s
cash on product market competition.3 As suggested by Campello (2006), ﬁrms can
implement a number of competitive strategies besides pricing to improve their market
performance, such as research and development spending, plant or store location,
distribution network, advertisement, or mergers of business competitors. Firms’ cash,
other than debt, is a great source for funding these strategic policies. Thus, a better
understanding on the relation between cash and product market competition would help
us explain the recent increase in ﬁrms’ cash from a whole new perspective.

Moreover, from an intuitive viewpoint, the linkage between cash and output
market outcomes is possible. Recent studies show that business risk is empirically
correlated with corporate cash level (Bates, Kahle, and Stulz (2008)). Generally, as
competitive interaction in the product market is a key determinant for industry risk, the
recent evidence suggests a plausible connection between cash and product market
competition.4 More directly. Chevalier and Scharfstein (1996) argue in their model that

“deep-pocketed” ﬁrms may decrease products’ prices to secure long-term market share at

 

3 A few papers present evidence on the implication of ﬁrms’ cash in other areas. For instance, Harford
(1999) studies acquisitions by ﬁrms with unusual cash holdings and documents that managers with weaker
incentives tend to spend cash inefﬁciently. In a similar spirit, Dittmar and Mahrt-Smith (2006) and Harford.
Mansi, and Maxwell (2006) ﬁnd that poorly-governed ﬁrms tend to dissipate their cash in ways that destroy
ﬁrm value. In contrast, Kim, Mauer, and Sherman (1998) and Opler, Pinkowitz, Stulz, and Williamson
(1999) document that persistent cash holdings do not hinder proﬁtability.

4 Other related papers include Kim, Mauer, and Sherman (1998), Mikkelson and Partch (2003), and
Faulkender and Wang (2006).

41

the expense of short-term proﬁts. By and large, such investment in market share building
would affect cash-poor rivals’ decisions to stay and compete and hence alter market
outcomes. Recently, Schroth and Szalay (2007) and Fresard (2008) show empirical
evidence that conﬁrms to a certain extent a strong connection between cash and product
market success, while the nature of the linkage remains unclear. Further investigation
may help unveil the causal relation between the two.

Like most research that attempts to shed light on the inﬂuence of debt on product
market competition, the existing studies on cash effects are directly or indirectly trapped
by the fact that both cash ﬂow and market performance are endogenously driven by
underlying shocks to proﬁtability.5 One can never be certain that it is a ﬁrm’s cash that
causes the change in the ﬁrm’s output strategy or promises the ﬁrm’s success in product
market competition. It may be that a common factor, such as a change of investment
opportunities, leads the ﬁrm to both alter its cash holdings and change its competitive
behavior. Financial economists have explored various solutions to mitigate the potential
endogeneity problem. In most cases, researchers take advantage of a plausibly exogenous
shock that is unlikely to correlate with investment opportunities, although the
identiﬁcation of the exogenous variation in cash ﬂow is not easy. Blanchard, Lopez-de-
Silanes, and Shleifer (1994) focus their study on a small sample of ﬁrms that receive cash
windfalls from lawsuits that do not change the investment opportunity sets. More recently,
Rauh (2006) identiﬁes, in a large sample, the exogenous variation in internal ﬁnancial
resources arising from mandatory contributions to pension plans. Alternatively, Lamont

(1997) exploits the presence of internal capital market in industrially diversiﬁed ﬁrms

 

5 The survey paper by Parsons and Titman (2007) highlights the endogeneity problems encountered in
corporate ﬁnance study and discusses approaches researchers have taken to address those particular
problems.

42

and identiﬁes a cash ﬂow shock that affects one sector of a ﬁrm and is exogenous to the
performance of other sectors of the ﬁrm.6 The latter study suggests a good approach to
ﬁnding an exogenous instrument for cash. we adopt this approach here to identify
variations in cash ﬂows of ﬁrms that are not correlated with their investment

opportunities.

2.2.2 Internal Capital Market in Diversiﬁed Firms

An important part of the capital allocation process takes place in internal capital
markets in which corporate headquarters allocate capital to their diversiﬁed business
segments. Different from external ﬁnance, internal funds are self-provided and do not
need to be secured by speciﬁc assets as collateral. Without many of the frictions arising
in the governance of ﬁrms ﬁnanced with external capital, internal funds are less costly
and hence are largely used by corporations to ﬁnance most of their capital outlays.

Under the hypothesis that the internal market plays a non-trivial role in allocating
capital among various business segments within one ﬁrm, all segments of the ﬁrm are
ﬁnancially interdependent. In such. a context, a ﬁnancial shock to one segment can
tighten ﬁnancial constraints in another segment.7 Thus, like the role of banks or other
external capital markets in transmission of business cycles, internal capital markets in
diversiﬁed ﬁrms can work as another channel to transmit a ﬁnancial shock from one

sector of a ﬁrm to others, even if the investment Opportunities of other sectors of the ﬁrm

 

° To deﬁne exogenous cash ﬂow shocks to some parts ofa ﬁrm. Lamont (1997) requires that shocks to cash
are not correlated (at least not positively correlated) with the proﬁtability of investments in those parts of
the ﬁrm.

7 The ﬁnancial constraints can be tightened for two reasons. First. the available cash ﬂow may be decreased
due to the ﬁnancial shock. Second, the value of assets used as collater‘als may decrease after the shock that
may increase the cost of ﬁnance.

43

are not affected by the shock.

As suggested by Lamont (1997), the presence of internal capital markets offers a
reliable way to ﬁnd an exogenous instrument for cash. By focusing on a group of
diversiﬁed ﬁrms, one needs to unambiguously identify a certain cash shock that affects
one part of the ﬁrm. Through the internal capital market, such a cash ﬂow shock may be
transmitted to other parts of the ﬁrm, even if the investment opportunity sets of other
parts of the ﬁrm are not affected by the shock at all. If the market performance or
competitive behavior of the other parts of the ﬁrm changes in response to the shock, then
a causal connection between cash and product market competition can be established
with certain conﬁdence, since the cash shock is exogenous to the proﬁtability of
investment in the other parts of the ﬁrm.

Correspondingly, the alternative hypothesis assumes that there is no internal
capital market and all corporate segments function as stand-alone units. On such a ground,
a shock to the cash in one sector of a ﬁrm would not affect cash in other sectors of the
ﬁrm if the investment opportunities in the other sectors are not changed by the cash shock.
Hence, if no responses are observed in product market behavior of the other sectors
following the cash shock, one can conclude that either the internal capital market does
not exist or there is no relation between cash and product market competition.

In order to exploit the presence of internal capital markets to identify exogenous
cash shocks, this paper relies on the corporate segment-level data. In the US, public
ﬁrms are required, pursuant to the various accounting standards, to disclose certain

ﬁnancial information for any industry segment that comprised more than 10% of its

44

consolidated yearly sales, assets, or proﬁts.8 Although the segment-level data has its own
econometric and conceptual weakness, it offers a valid way to systematically detect the
various business segments of a ﬁrm and trace their performance, and hence is used by a

growing number of academic papers.

2.2.3 Hypotheses Development and Empirical Design

The central empirical strategy in this paper is to examine the product market
responses of ﬁrms to exogenous shocks in cash. The success of the strategy relies on the
identiﬁcation of a type of cash shock for which the investment opportunities of a ﬁrm
should be exogenous. Cash shocks due to certain industry-speciﬁc tax increase seem to ﬁt
this description well. The industry-speciﬁc tax event can be used as a natural experiment
for identifying an exogenous cash shock for three reasons. First, tax rate changes are
generally regulated and released by federal or state authorities and hence are
unambiguously exogenous to any individual ﬁrm’s actions.9 Second, a speciﬁc industry
tax event should affect expected cash ﬂows to all the ﬁrms that operate business
segments in that industry. A shock like that would promise a decent number of ﬁrms for
study. Moreover, an industry-speciﬁc tax change should not affect the proﬁtability of
investment in other industries, which promises that the cash ﬂow shock is exogenous to
ﬁrms’ segments operating in those unaffected industries. Third, tax events usually have a
signiﬁcant impact on ﬁrms’ performance and expected cash ﬂows, which allows a

powerful test for the effect of cash on competition.

 

8 Segment report requirement was initially established by Statement of Financial Accounting Standards
No.14 (PAS 14) issued by the Financial Accounting Standards Board in 1976. The requirement was
amended in 1979 by PAS 30 and both PAS 14 and F AS 30 were superseded by FAS 131 in 1997.
9 Large corporations may hire lobbyists to inﬂuence the congress in making policies for their own sake.
However, it is difﬁcult to imagine that any certain tax policy is passed for a single ﬁrm’s beneﬁt.

45

To examine the connection between cash and product market competition, this
paper ﬁrst examines the market response of industry rivals to the business segments that
experience an exogenous cash shock. Absent changes in the investment opportunities of
the industries in which business segments operate, there is no reason to believe that the
market performance of industry rivals would change if the cash shock has no effect on
competition behavior of the segments. However, if the exogenous cash shortfall severs
ﬁnancial constraints of the business segments and hence in turn changes their competitive
behavior in the product market, then the performance of industry rivals may be affected,
which will be reﬂected in the change of their stock prices. Thus, under the assumption of
an efﬁcient market, abnormal returns to industry rivals can be used as meaningful
indicator of the effect of cash on product market competition.

Two strands of models suggest how the industry rivals may be affected by the
change in product market competition. One class of models, notably Chevalier and
Scharfstein (1996), predicts that a tightening ﬁnancial constraint may induce business
units to “soften” the competition by decreasing output.'0 This may in turn beneﬁt the
industry rivals of those units. Moreover, Chevalier (1995b) suggests the incentive for
industry rivals to actively prey on ﬁnancially weak ones and drive them out of the market.
Under the rational hypothesis that industry rivals will only prey on those ﬁrms if the
long-term beneﬁts exceed the short-run costs, returns to industry rivals may be increased
as a result of predatory actions. An alternative class of models, often associated with
Brander and Lewis (1986), suggests the possibility that a decrease in cash may make the

business unit more aggressive in competition, such as to increase output. This may

 

'0 In Coumot competition, ﬁnancially constrained units will decrease the output in competition with others,

which pushes up the industry price in equilibrium. Thus the performance of the industry rivals can simply
improve from the high output price.

46

consequently “toughen” product market competition and hence drag down the
performance of industry rivals. In addition, literature on the beneﬁts of internal markets
implies that when one unit of a ﬁrm suffers poor performance, the corporate headquarter
may have strong incentive to efﬁciently redeploy assets and put more effort and resources
into competition in units in which investment opportunities are not negatively affected.
This also suggests that competition with ﬁnancially constrained ﬁrms in certain markets
may become more intense, which may hurt the industry rivals’ performance in the
product market.

In tests to investigate market response of industry rivals to business units that are
affected by exogenous cash shock, we examine the abnormal returns of industry rivals as
a whole and in some sub-samples speciﬁed by industry characteristics. Abnormal returns
to portfolios of rivals are also examined to control for the potential correlations of stocks
in the same industry. All of these tests would provide the main evidence of this paper on
the nature and extent of how cash affects the product market competition.

Next, this paper turns to studies on how output behavior varies with the
exogenous shock to cash. There are two theorized effects that may drive the potential
change in output after the decrease in cash. First, the “cost effect” predicts that a decrease
in internal funds may increase the probability of liquidation for any given level of
production because the ﬁrm must borrow more. Thus the cash shortfall may increase the
marginal cost of output expansion, which induces the ﬁrm to produce less. The second is
the “revenue effect,” which suggests that producing a high output may allow the ﬁrm to
generate revenue that it needs to repay the loan. Consequently, the ﬁrm may have the

incentive to increase the output when facing a negative shock to cash.

47

In the tests that follow, we examine both the raw and industry-adjusted variations
in the output level of business segments affected by an exogenous cash shock between
the year prior to the shock and each of the ﬁve years following the shock. Industry
adjustment is necessary to control for industry-wide changes in the proﬁtability of
investment. The tests also control for time effect, since all dollar values are deﬂated by
average inﬂation index in that year. These tests would offer more direct evidence on the
effect of cash on competition.

To provide further support for the results, we then proceed to examine the
robustness of the ﬁndings to a number of estimation issues and identiﬁcation choices. To
better understand the determinants of abnormal returns to industry rivals, we further
estimate models explaining the abnormal returns as a function of ﬁrm and industry
competition characteristics. This provides some evidence on how the industry
competition level may affect the cash-competition sensitivity.

After presenting the evidence that establishes a connection between cash and
product market competition, we consider how the competition change caused by the cash
shock affects customers. As discussed before, the severity of ﬁnancial constraints may
induce ﬁrms to decrease output in the product market. Consequently, product price may
be increased, which causes customers to suffer from negative returns. This suggests that a
decrease of cash may “soften” product market competition and rivals will beneﬁt at the
expense of their customers. Alternatively, a decrease of internal funds may make ﬁrms
produce high output, which in tum drives down the product price. In this case, industry
output increases after ﬁrms’ cash decreases and customers experience positive returns at

the cash shock to ﬁrms. Hence, this examination on customers’ retums may provide us

48

more evidence on how ﬁrms change their competitive strategies after a negative shock to
cash. No paper has attempted to exploit the customers’ returns to show the change in
competitive behavior of ﬁrms constrained by a decrease of cash.ll

The empirical design to identify an exogenous instrument for cash is closely
related to research by Lamont (1997). However, the analysis in this study is quite distinct
from his in that this paper takes a step forward to investigate how ﬁrms’ cash affects
product market competition rather than the relation between cash ﬂow and investment,
and that the exogenous cash shock is identiﬁed based on certain policy changes rather
than an economic event. Moreover, to our knowledge, this paper is the ﬁrst to empirically
examine the causal impact of cash ﬂow on ﬁrms’ competitive outcomes and produce
results that provide insight into a more complete picture of the relationship between

ﬁnancial constraints and product market competition.

2.3 Sample Selection and Data Description

2.3.1 Sample Selection

The selection process that leads to ﬁnd the industry-speciﬁc tax events examined
in this paper is as follows: we start by searching three main news sources; The Wall
Street Journal, the New York Times and the Reuters News, under the key words “higher
”12

taxes” from January 1, 1980 to December 31 , 1997 through the search engine “Factiva.

We end the sample in 1997 to avoid segment deﬁnitions inconsistent over the sample

 

H Some previous studies have relied on pricing data to examine the changes in rival ﬁrms’ competitive
behavior. Thus, these papers are often limited to speciﬁc industries or special settings for which price data
can be available.

'2 Segment data from COMPUSTAT is available from 1979. In order to match previous research, we
require the sample period starting from 1980.

49

period as there is a change in segment reporting requirements effective in 1998. In order
to obtain the most relevant reports in the search, we limit the subject of reports to be in
the “corporate\industrial” category with a concentration on ﬁrms’ performance.
Republished news is also excluded to determine more accurately the release dates of tax
events to the public. Those initial search steps turn out a sample with 1639 reports which
contain information directly or indirectly related to tax increases during the selected
period.13

After carefully reviewing each report, we then narrow the initial sample using the
following criteria. First, we exclude reports that reveal no real information about tax
events, as Factiva works to retrieve all reports that contain the words “higher” and
“taxes” separately. A large number of reports are excluded by this step. Next, we exclude
those reports that claim or imply certain tax-increase events effective for all the US.
corporations or for ﬁrms in a large group of industries.'4 Tax events effective in certain
geographic areas (e.g., city taxes) are excluded too as they inﬂuence a limited number of
ﬁrms or business units within the same industries. Events related to tax changes in
ﬁnancial and regulated industries are also excluded as standard. After these steps, only
seven reports containing information about tax increases in the speciﬁc industry remains.
Finally and most importantly, we require that the impact of the certain tax event be
quantitatively large in the speciﬁc industry so that it can cause a signiﬁcant negative
shock to the cash ﬂow of all business units in that industry. While the amount of cash loss

is hard to estimate, the abnormal returns of those business units at the release date of the

 

'3 We only search for events related to tax increases Since in this paper we are interested in studying the
tightening of ﬁnancial constraints.

'4 A great deal of news on tax changes is ambiguous regarding the deﬁnitions of affected industries. For
instance, a potential tax raise in transportation industry was reported brieﬂy while the news released no
information about which speciﬁc transportation industry it was.

50

shock are reliable proxy for the signiﬁcance of the tax event. In addition, signiﬁcant stock
reactions also indicate that certain tax event was not previously predicted by the market.
This makes ex ante anticipation of such a negative shock unlikely for those affected ﬁrms.
To select the signiﬁcant tax event(s), we require that equally-weighted abnormal returns
for portfolios of all ﬁrms in the same speciﬁc industry are less than -lO%. Firms’ four-
digit SIC codes are used for industry classiﬁcation and industry deﬁnitions are from
Fama and French (1997). All ﬁrms examined are publicly traded US. ﬁrms with
sufﬁcient data from CRSP to calculate abnormal returns at the release date of events.
These selection steps in the end leave one report that claims a signiﬁcant shortfall
of expected cash ﬂow in the tobacco industry due to worries of potential increases in
federal excise tax.'5 The event took place on April 2, 1993, the day that Philip Morris
(P.M.), the leading company in the tobacco industry, warned the market that “operating
earnings from its US. tobacco business could be down as much as 40 percent.” The
company blamed the disappointing net on the expected higher excise tax and an inability
to raise cigarette prices. Analysts and asset managers followed Philip Morris’ statement
and commented that “lingering fears of higher excise taxes” could result in the “rout” in
the tobacco industry. The market generally worried that by increasing the federal excise

tax the governance may be attacking “the most proﬁtable industry.” As a result, tobacco

 

[5 The tax rate on tobacco products remains unchanged from 1951 to 1982. The rate was ﬁrst increased as
part of the Tax Equity and Fiscal Responsibility Act of 1982 and then increased in two stages under the
Revenue Reconciliation Act of 1990. One half of the increase took effect on January 1, 1991 with second
half in effect as of January 1, 1993. The most recent increase in tax rate on tobacco products occurred in the
Balanced Budget Act of 1997 and took effect on January 1, 2000 and 2002, respectively. To our surprise,
the tobacco stocks did not experience signiﬁcant changes (more than 5%) on any of these dates. One main
reason is that the formation process of those acts is time-consuming and the market gradually adjusted for
the expected change in tax rates. Another reason may be that the tobacco companies can always lessen the
effect of tax on performance by charging higher product prices. However, in early 1993, the tobacco
companies feared both the threats of higher taxes and price pressures. Thus the market responded
signiﬁcantly.

51

stocks dropped signiﬁcantly on the event day.

Although there was no public announcement from authorities on the same day
about the deﬁnite tax increase, the market always anticipates the movement of policies
and reﬂects the information in the stock prices. This event can be used a natural
experiment to identify negative shock to cash ﬂow of ﬁrms in the speciﬁc industry due to
tax changes. Since all the tobacco companies traded in US. markets experienced
signiﬁcant drops in stock prices, it is difﬁcult to imagine P.M.-speciﬁc news that would
affect the whole industry and would arouse market worries on the effect of higher taxes.

To gather a sample of ﬁrms likely to be affected by the negative shock in the
tobacco industry identiﬁed above, we extract every diversiﬁed ﬁrm that operates a
segment with primary business in the tobacco-related industry in 1992-1993 period and
had at least 10 percent of its sales from the segrnent’s business. The primary business of
each segment is identiﬁed either by the segment name or the primary 4-digit SIC industry
code. A tobacco-related industry is deﬁned as any industry that was involved in the
farming of tobacco or in the production, reﬁning, transportation or sale of tobacco
products, in the service of the tobacco industry, or in the production of substitutes for
tobaccos. The line of business descriptions of each segment are collected from the
COMPUSTAT segment database. We then select those ﬁrms that had at least one
segment in the non-tobacco—related industry. This process selects a sample of twelve

. . . . . . 1(
diversrﬁed ﬁrms operating in tobacco busrness, specrﬁed as tobacco-dependent ﬁrms. ’

 

'6 The non-tobacco-related industries are deﬁned as those in which the investment opportunities would not
be affected by the negative cash shock in the tobacco industry. Financial and utility industries are excluded,
since these industries have complex accounting variables or are often regulated by special requirements.
The judgment of those non-tobacco-related industries depends on the examination of their SIC codes or the
line-of—business descriptions.

52

2.3.2 Summary Statistics

Table 31 describes characteristics of those tobacco-dependent ﬁrms. Most of
these ﬁrms were quite large measured by sales of 1992. Their average annual sales were
around $11 billion, far larger than a typical COMPUSTAT ﬁrm. Although those ﬁrms
operate in multiple divisions, on average more than half of their sales were from the
tobacco-related business. Thus, on the day when the threat of higher tax caused a
negative shock to the tobacco industry, all the ﬁrms in the sample experienced
signiﬁcantly negative daily returns with a mean of -12%, and the ex post change in cash
ﬂow from the tobacco business between 1992 and 1993 reaches to nearly -6%. All these
results conﬁrm the signiﬁcance of the identiﬁed shock to cash ﬂow in the tobacco
industry.

Table 8.2 follows to present detailed segment data for those tobacco-dependent
ﬁrms. Segments included for analysis in this paper are those which operate in non—
tobacco-related industries and generate no less than $50 million in sales in 1992. We also
exclude segments with incomplete information and segments that are involved in utility
and ﬁnancial service industries.17 It is shown that those non-tobacco segments are on
average large in Size, with a mean dollar value of $2,311 million, and that they generally
take an important percent of the whole ﬁrm sales. In addition, the ex post change in the
segment cash ﬂow to sales ratio between 1992 and 1993, expressed as percentage point,
turns out positive (nearly 3%). It is possible that the investment opportunities of the
industries in which those segments operate may improve after the negative shock to cash

in the tobacco industry. Another possibility is that tobacco-dependent ﬁrms may change

 

I7 66 9.9 ' '
Hereafter, those segments selected are referred to as non-tobacco segments in this paper.

53

their product market competition behavior in those non-tobacco business units and hence
bring up the proﬁtability in assets which are not affected by the negative shock. We will

examine this possibility below.

2.3.3 Industry Rivals and Customers

As discussed, the analysis of this paper requires identifying the rivals and
customers to the segments that are not affected by the negative cash shock. Industry
rivals to those non-tobacco segments, speciﬁed above, are identiﬁed as any ﬁrm with a
segment that has the same four-digit SIC industry code as the non—tobacco segments of
tobacco-dependent ﬁrms. All the segment data are available in the COMPUSTAT
database. The rivals are also required to be publicly-traded ﬁrms with sufﬁcient data from
CRSP to calculate the abnormal returns on the day of negative cash shock to the tobacco
industry.

In total, 436 industry rivals to non-tobacco segments are identiﬁed. Table B.3
classiﬁes those industry rivals to non-tobacco segments by industries. Compared to those
industries listed in Table 8.2, non-tobacco-related industries in Table B.3 exclude few in
which no rivals are identiﬁed or information for rivals is incomplete. On average the
industry rivals are much smaller than the tobacco-dependent ﬁrms in size. However,
compared to the competing segment of the tobacco-dependent ﬁrms in the same industry,
the industry rivals are not much different in size. It suggests that those identiﬁed rivals
may be strong competitors in the same industry. There is also some concern about the
selection of the rivals in the sample. Due to differences in sales channels, targeted

consumers or product specialties, it is possible that two ﬁrms share the same industry

54

code but compete remotely. Below we will discuss this possibility and examine a group
of more closely competing rivals.

Table B.3 also includes the summary statistics for the industry rivals to non-
tobacco segments in tobacco-dependent ﬁrms. Among different industries, we ﬁnd no
heterogeneity in leverage, Tobin’s q or operating cash ﬂow of those industry rivals.
Nevertheless, some industries are clustered with lots of competitors while most industries
have less than ten rivals identiﬁed.

Although various ﬁnancial accounting standards require ﬁrms to disclose the
identity of any customer representing more than 10% of the total sales of the ﬁrm, data
on the customers of ﬁrms is not publicly available. Segment ﬁles contain some
disclosures from ﬁrms about their customers. However, these ﬁles only list an
abbreviation for customers’ names. By employing a text-matching procedure to link the
customer abbreviations with company identities, Fee and Thomas (2004) successfully
identify a group of customers and suppliers to disclosed ﬁrms. Under the authors’
permission, we use their customer data and identify 101 customers to those industry
rivals included in Table B.3. Those customers selected do not include those operating in
the tobacco-related industry. This allows that the market performance of those customers
is not affected by the change of investment opportunities in the tobacco industry.
Unfortunately, we can not link any meaningful customer to those non-tobacco segments
in tobacco-dependent ﬁrms since the few customers available for tobacco ﬁrms operate at

least some tobacco-related business.

2.4 Empirical Results

55

2.4.1 Abnormal Returns to Industry Rivals

The main interest in this paper is to address how a ﬁrm’s cash affects its product
market behavior. Previous models (e.g., Chevalier and Scharfstein (1996)) predict that
cash-poor ﬁrms may be less likely to invest in market share building by decreasing the
price. This suggests that a decrease of internal funds may induce a ﬁrm not engaged in a
price war. As a result, industry rivals beneﬁt from the “soft” product market competition.
Moreover, the possibility of predation (Chevalier (1995b)) may threaten the ﬁnancially
weak ﬁrms to exit a certain market, which helps the industry rivals to earn more in
competition. Alternatively, other models (Brander and Lewis (1986)) predict more
aggressive competitive behavior from ﬁnancially constrained ﬁrms, such as lower price
or higher output, which in turn hurts the rivals in the same market. Although pricing data
is not generally available, the investigation of abnormal performance of industry rivals
sheds some light on the effect of cash on product market competition, and offers another
part of the major results of this paper.

In this section, we examine the share price response of industry rivals to those
non-tobacco segments in tobacco ﬁrms that experience a negative cash shock. Abnormal
returns to industry rivals are calculated using a standard market model. The market model
is estimated over the 200 trading days from 240 to 41 days before the announcement day
of cash shock to the tobacco industry. Abnormal returns are calculated over the event
window centered on the announcement day of the shock and all tests of signiﬁcance are
performed when standardized abnormal returns are equal to zero.

Table B.4 reports the abnormal returns to industry rivals of non-tobacco segments

of tobacco ﬁrms on the announcement of the negative cash shock to the tobacco industry.

56

For the entire sample of 436 industry rivals, on average they earn positive abnormal
returns (1.02%) on the day of announcement.18 This is signiﬁcant at the 1% conﬁdence
level. Moreover, a sign test indicates that signiﬁcantly more industry rival ﬁrms show
positive (230) than negative (158) abnormal returns. This ﬁnding of positive average
returns is consistent with the idea that internal funds may affect a ﬁrm’s product market
behavior signiﬁcantly, and that a decrease in cash may “soften” the product market
competition.

Campbell, Lo, and MacKinlay (1997) discuss the “clustering” problems in
calculating abnormal returns as returns to the ﬁrms in same industry may be correlated, at
least to some extent. To further investigate the statistical signiﬁcance of the observed
positive returns to industry rivals, we regress the return of the rival against a constant,
and adjust the White standard error estimates for clustering by industry. The coefﬁcient
on this regression naturally equals the mean return of 1.02%, and the adjusted t-statistic is
2.40. Thus the univariate ﬁnding of positive abnormal returns for industry rivals is robust
to potential correlations within returns for same-industry ﬁrms.

Table B.4 also reports further results for several subsamples of industry rivals for
which the competitive effect of cash is expected to be prominent. First, to gauge the cash—
competition effect in highly competitive industries, we identify a subsample of industry

rivals that compete in industries with the pre-announcement industry Herﬁndahl index

 

'8 While those tobacco ﬁrms included in the paper are larger companies compared to a typical
COMPUSTAT ﬁrm. their stocks take less than 6% weight in the S&P 500 index according to the report of
“Index Component Weights of Stocks in the S&P 500” offered by mvw.imlexarb.c0m. Speciﬁcally, the
largest tobacco ﬁrm. Philip Morris International. takes only 0.91% weight in the index. Thus, these tobacco
ﬁrms’ stocks take a small percentage in the benchmark and would not bias the abnormal returns estimated
using the market model. Actually on the announcement day of the cash shock to the tobacco industry, the
value index of S&P 500 drops only around 2%. The following robustness check shows that the results are
not affected by the choice of benchmark.

57

larger than 2000.19 The Herﬁndahl index is a measure for market concentration and is
calculated as the sum of the squared market share of all ﬁrms operating a segment in the
industry. 20 A higher Herﬁndahl index indicates greater market concentration in the
industry. The subsample of industry rivals in concentrated industries shows signiﬁcantly
positive abnormal returns (0.73%) but the returns are lower than what other rivals in
competitive industries receive. This result suggests that the competitive effect of cash
may be magniﬁed in highly competitive industries, as high business risk due to severe
product market competition may increase the possibility of bankruptcy in competitive
industries, which in turn induces ﬁnancially constrained ﬁrms to decrease output more.
Next, we identify a subset of industry rivals that compete in industries where the
segments in tobacco ﬁrms take less than 10% market share. The abnormal returns to this
group of rivals are signiﬁcantly positive and much higher than the average returns to all
rival as a whole. Moreover, the difference test between rivals in market with non-tobacco
segment market share less than 10% and more than 10% shows that the size of the
segment that suffers from negative cash shock has a signiﬁcant effect on the rivals’ return.
This shows that industry rivals beneﬁt more from competition when a smaller market
participant experiences a decrease in internal funds. It is consistent with the idea that
smaller ﬁrms may be more ﬁnancially constrained and hence a decrease of internal funds
affects those ﬁrms more. However, it is also possible that ﬁnancially constrained ﬁrms
are more likely to exit certain minor business markets at the threat of possible predation

from rivals.

 

'9 Following the literature, we choose the Her‘ﬁndahl of 2000 as a benchmark. Generally, an industry in
which there are ﬁve market competitors each with an equal market share has an industry Herﬁndahl of
2000.

20 In this paper, the Herﬁndahl is calculated on the segment lever basis using four-digit SIC codes for
industry classiﬁcations.

58

As a further test, we examine the abnormal returns to rivals’ portfolios on the
announcement of negative shock. Equal-weighted portfolios are formed here for analysis.
The effect of cash shock on industry rivals may be difﬁcult to detect if rivals operate in
multiple lines of business. Thus, we perform the analysis on both the overall group of
rivals (single- and multi-segment rivals) and on those rivals that operate business in only
one four-digit SIC code industry. Table B.4 presents results for rival portfolios comprised
of single-segment rivals only, as well as rivals’ portfolios comprised of single- and multi-
segment rivals. For portfolios of the entire sample, industry rivals exhibit positive and
signiﬁcant abnormal returns. In the two subsamples, rival portfolios also show signiﬁcant
stock price increases. Moreover, we ﬁnd that the size and statistical signiﬁcance of the
abnormal returns to rivals are greater in the tests using portfolios of single-segment rivals.
In addition, abnormal returns to small rivals may be large. We ﬁnd that for rival
portfolios comprised of only small rivals (market share less than 5%), abnormal returns
are also positive and larger than the portfolios of all rivals, although the signiﬁcance level
drops. These results strengthen the idea that a decrease of cash strategically inﬂuences

product market competition, which in turn beneﬁts the rivals.

2.4.2 Output Changes of Non-tobacco Segments

In this section, we examine how the output behavior changes for those non-
tobacco segments when they experienced an exogenous negative cash shock. Although
the shock was not correlated with the investment opportunities in non-tobacco industries,
through the internal capital market, the cash shock may be transmitted to the non-tobacco

segments. The decrease in internal funds may sever the ﬁnancial constraints of those

59

segments, thus affects their output behavior in product market competition.

Our investigation builds on and ties into previously documented hypotheses how
a ﬁrm’s output varies with the internal funds that are used to ﬁnance investment projects.
First, there is “cost effect”: a decrease in internal funds requires more borrowing from the
outside and hence increases the probability of bankruptcy at any given level of
production. This increases the marginal cost of output expansion, which induces a ﬁrm to
produce less. The second is “revenue effect”: a high output allows a ﬁrm to generate
revenue that it needs to repay the loan. Thus this provides the ﬁrm an incentive to
increase output when facing a tightening ﬁnancial condition.

To trace the changes in output of those non-tobacco segments before and after the
negative cash shock to the tobacco industry in 1993, we calculate the change in sales and
cash ﬂow for every non-tobacco segment between the year prior to the cash shock and
each of the ﬁve years following the shock. Although the true output data is not available,
we believe that sales or cash ﬂow generated from a segment are closely (positively)
related to its output level. It is important to control for time effect, because the investment
opportunities of industries may change over time. Thus, in the tests that follow, the dollar
value of the sales and cash ﬂow for each non-tobacco segment is deﬂated by the average
consumer price index in that year.

Table 85 presents raw and industry-adjusted changes in sales and cash ﬂow of
non-tobacco segments, and displays some of the major results of this paper. Industry-
adjustment is necessary to control for industry-wide changes in the investment
opportunities. The method used is fairly standard in the corporate ﬁnance literature. For

each observation of AS (ACF), we subtract the mean value of AS (ACF) from the group of

60

all segments that were in the same industry.2| The changes, both raw and industry-
adjusted are economically and statistically signiﬁcant. In ﬁve years after the negative
cash shock, sales (cash ﬂow) of those non-tobacco segments were increasing. Based on
this result alone, one might conclude that those segments increase their output after the
decrease of internal ﬁinds. The close examination of industry-adjusted changes, however,
tells a different story. Those non-tobacco segments performed much lower than the
industry mean and the difference was increasing over time. One possible explanation for
these results is that the proﬁtability of investment in those non-tobacco industries
increased after 1993 while those non-tobacco segments in tobacco ﬁrms had to produce
much less than the rivals due to the constraints of their internal ﬁnds. The
underperformance of output can be explained for two reasons. The ﬁrst is “cost effect,”
which induces ﬁrms to produce less. The second is the possible predation behavior of
industry rivals after seeing the non-tobacco segments ﬁnancially constrained. In addition,
the number of the segments that tobacco ﬁrms operated was decreasing after the negative
cash shock, which suggests that tobacco ﬁrms may pull out of certain markets. This result
corresponds to the explanation above that ﬁrms tend to decrease output with a decrease of
internal funds. Of course, the exit from certain markets can be caused by a voluntary drop

in output or forced leave by rivals’ predation actions.

2.4.3 Robustness

The main ﬁnding above is that cash ﬂow is signiﬁcantly related with the product

 

2' In a previous version of this paper, we use a different method of industry-adjustment (such as to use the
median value instead of the mean, or to use the mean (median) value from the group of the segments that
are in the same industry, but were owned by ﬁrms that were not affected by the negative cash shock) and
the results are not changed.

61

market competition: a decrease of cash can inﬂuence a ﬁrm’s output behavior and affect
the competition outcomes that beneﬁt the industry rivals. In this section, we consider a
number of robustness checks concerning the results.

In estimating the abnormal returns to industry rivals, we include the normal value-
weighted market returns as a benchmark in the standard market model. This choice of
benchmark may be problematic, considering the fact that tobacco ﬁrms are generally
large corporations and represent a certain percent of the whole market value. If the
tobacco stocks were down sharply on one day, the dramatic drop may bring down the
market slightly, which may explain why the rivals over-performed the market even if the
cash shock did nothing to affect the competition in a way to beneﬁt the rivals.

To get around this benchmark concern, we construct two value-weighted market
portfolios with all the tobacco stocks excluded. First, we retrieve the historic constituents
of the S&P 500 index every month within the time period of two years before the shock
event to one year after. From those constituents, we exclude all the tobacco-dependent
ﬁrms identiﬁed above. The rest of the ﬁrms will form as the S&P market portfolio
without the tobacco stocks. Second, for every day included in the time period, we select
the largest 500 publicly-traded ﬁrms from CRSP with all tobacco ﬁrms excluded and
form the CRSP value-weighted portfolio. Using these two new benchmarks in the
estimation of the market model, we can assure that the previous results are not just driven
by an artiﬁcially low benchmark on the event day.

Table B.6 reports the main results using the two speciﬁed benchmarks in
estimating the abnormal returns to industry rivals. It is shown that on average rivals do

exhibit positive and signiﬁcant abnormal returns and the Size of the abnormal returns to

62

rivals are greater for the subsamples of rivals. Notably, the difference test between rivals
in the market with the non-tobacco segment market share less than 10% and rivals in the
market with that share more than 10% shows that there is signiﬁcant difference in the
rivals’ abnormal returns. All the results are consistent with what is reported above. Thus,
it appears that the ﬁndings above are not sensitive to the choice of benchmarks. The
results are consistent with the idea that a decrease of cash “softens” the competition in the
product market and the event is actually good news for the rivals.

As discussed above, P.M. was the ﬁrst to warn the market about the negative
shock to expected cash ﬂow in the tobacco industry and predicted that the shock from the
potential increase in federal excise tax could be signiﬁcant in decreasing the cash ﬂow.
External parties comment on its statements and agreed that lingering fears of higher tax
would affect the tobacco industry more than what the market anticipated previously. Thus,
all the tobacco stocks dropped signiﬁcantly on the announcement day.

Yet the estimation of value effect from special events is always challenged by the
revelation bias, since the abnormal returns on the announcement date possibly reﬂect not
only the value from the event but also news about the ﬁrm itself. It is often observed in
market that ﬁrms may make certain announcements to shift attention away from other
less favorable development. Based on market responses, it is difﬁcult to imagine that the
drop in cash ﬂow was not, to some extent, caused by the shock to the tobacco industry. It
is possible that the decrease in its expected cash ﬂow was not solely from the shock,
since P.M. may just have revealed the shock to divert the attention from its problems in
other segments, thus the inclusion of P.M. other segments into the sample may bias the

results.

63

To assure that this sample selection issue has no substantial effect on the ﬁndings,
we recheck the segments’ output behavior and abnormal returns to rivals after excluding
the other segments in the P.M. ﬁrm. The results are presented in Table B7 As reported,
output from those non-tobacco segments increases slightly after the shock but is still
much less than what the rivals produce. Abnormal returns of rivals to non-tobacco
segments in other tobacco ﬁrms are also positive and signiﬁcant. Thus, the ﬁndings do
not appear to be sensitive to this sample selection concern.

The identiﬁcation process employed above to ﬁnd industry rivals to those non-
tobacco segments relies heavily on the four-digit SIC industry code. Thus all the ﬁrms
with business units operating in same industry as tobacco ﬁrms do are treated equally as
rivals. However, ﬁrms classiﬁed under the same industry code are not always directly
competing with each other due to variations in market locations or product differentiation.
For example, it is difﬁcult to believe that all ﬁrms in the food industry produce the same
product and face the same customer markets. The previous selection of industry rivals
may include some ﬁmis that are not meaningful market competitors. In addition, the
previous empirical investigation treats all rivals equally. Yet it seems like that the closer
the competitor, the larger should be the effect if the connection between cash and product
market competition is valid. Hence, the identiﬁcation of close rivals seems necessary.

The measure of how closely rivals compete with tobacco ﬁrms is not easy to
establish. In this section, we assume that the correlation of segment sales between a
tobacco ﬁrm and its rival that compete in the same industry is larger if they are close
competitors. By examining the correlation of sales between rivals and competing tobacco

ﬁmis in ﬁve years before the cash shock, we identify “closely-related rivals” as those

64

with at least one segrnent’s sales strongly correlated with sales in non-tobacco segment
that operates in the same industry.22 For this group of close rivals, we ﬁnd that they
exhibit slightly higher returns than the others at the announcement date of negative cash
shock to the tobacco industry, although the difference test between two groups shows no
signiﬁcant difference. The result suggests, at least to certain extent, that the effect of cash
on product market competition may be larger for close competitors. To conserve space,

these abnormal returns are not reported in tables but are available upon request.

2.4.4 Cross-section Analysis

Table B.8 presents multivariate regressions that explain abnormal returns to
industry rivals as a function of ﬁrm and industry characteristics. Given the strength of
product market competition on the performance of ﬁrms, most of the multivariate
analysis focuses on linking the abnormal returns to rivals with variables that describe the
industry competition characteristics. In column 1, we include as an independent variable
the industry market share of the non-tobacco segments that suffered from the negative
cash shock. The estimated coefﬁcient of this variable is negative and signiﬁcant. As in
the univariate tests, the result suggests that the rivals beneﬁt more from competition
change when a small market participant is affected by a cash shock. It is consistent with
the idea that small ﬁrms may be more ﬁnancially constrained and hence may be more
likely to pull out of the market when suffering a decrease in cash.

In Column 2, we add the industry Herﬁndahl as another independent variable to

test the effect of competition on the cash-competition sensitivity. The coefﬁcient on the

 

22 Strong correlation is deﬁned when the correlation coefﬁcient is larger than 0.5.

65

Herﬁndahl index variable is negative and weakly signiﬁcant. This suggests that high
industry competition may amplify the cash-competition effect. Column 3 adds the market
share of the rival ﬁrm into the regression to account for the size of the rival. Column 4
also adds a number of commonly used variables to control for ﬁrm-speciﬁc
characteristics. In these two regressions, the size and signiﬁcance of coefﬁcients on the

Herﬁndahl index and market share of non-tobacco segments are almost unchanged.

2.4.5 Abnormal Returns to Customers of Industry Rivals

As outlined above, the examination of abnormal customer returns to industry
rivals would help to better understand how ﬁrms change their competitive strategies after
a negative shock to cash. Table B.9 presents results for the abnormal returns of customers
to industry rivals as identiﬁed above. For the entire sample of customers, we ﬁnd that
customers of rival ﬁrms experience positive average returns at the announcement of cash
shock to the tobacco industry. This is inconsistent with idea that overall product price
increases after the change of competition due to cash shock. Actually, it suggests the
possibility that some customers may beneﬁt from the new market equilibrium. Combined
with the evidence of positive rival returns at the time of announcement, this result is also
inconsistent with the hypothesis that industry rivals may gain from the change of market
competition at the expense of their customers.

The full sample is then divided into four subsamples based on whether the
industry rivals are in industries with a Herﬁndahl value less than 2000 or greater than
2000, and whether the industry rivals are in industries with a non-tobacco segment

market share less than 10% or greater than 10%. Abnormal customer returns show a

66

distinct pattern in those subsamples. For rivals in industries with a lower market
concentration (higher competition), their customers experience signiﬁcant positive
returns. Alternatively, signiﬁcant negative returns are observed for customers of rivals
that are in industries with a higher market concentration (lower competition). This result
suggests that ﬁrms in various industries may respond to cash shocks with different
pricing and output behavior and in turn affect their customers differently. Unfortunately,
without further examination on the price and output level in the new industry equilibrium,
it is hard to conclude how rivals change their competitive strategies and then affect

customers in aggregate. It opens up a puzzle for future research.

2.5 Conclusion

In this paper, we empirically investigate the connection between cash ﬂow and
product market competition. Using a tax-induced negative shock to expected cash ﬂows
in the tobacco industry as a natural experiment, we uncover that industry rivals to non-
tobacco segments in tobacco ﬁrms experience positive and signiﬁcant abnormal returns
at the announcement of the cash ﬂow shock. In addition, we ﬁnd a signiﬁcant change in
output behavior of those non-tobacco segments following the shock. This suggests that a
ﬁrm’s cash ﬂow plays an important role in inﬂuencing its competitive outcomes in the
product market. Moreover, the effect of cash ﬂow on competitive performance is
magniﬁed in competitive industries and when cash—constrained ﬁrms only have a small
market share. These results are robust to a number of estimation issues and identiﬁcation
choices. Since the shock is exogenous to the investment opportunities in non-tobacco
industries, our ﬁndings support the hypothesis that there is a causal relation between a

ﬁrm’s cash ﬂow and its product market competition.

67

In the end, we examine the customers of industry rivals to non-tobacco segments
in tobacco ﬁrms and ﬁnd that on average customers experience positive and signiﬁcant
returns on the shock announcement date. Nonetheless, there is considerable variation in
the market reactions of customers in different subsample. Customers of the ﬁrms that
operate in competitive industries seem to exhibit signiﬁcant stock price increases while
customers of the ﬁrms that compete in concentrated industries experience negative and
signiﬁcant abnormal returns. This suggests that some customers may beneﬁt while others
suffer from a new industry equilibrium after the negative shock. Although it is hard to
ﬁgure out the change of pricing and output in a new equilibrium, our results at least rule
out the possibility of overall increases in product price.

This paper adds to the growing literature on the connections between ﬁnance
choice and corporate strategy. By establishing the causal link between cash ﬂow and
product market competition, we provide a more complete picture of the relationship
between ﬁrms’ ﬁnancial and operating decisions beyond mostly studied association
between debt ﬁnancing and competitive strategy. This paper also complements the
literature on corporate liquidity by suggesting the substantial effect of cash on product
market outcomes. The evidence we present helps to shed some light on the implication of
corporate cash. Lastly, our results add to the body of evidence on the interactions
between ﬁrms, rivals and customers, which helps to better understand the nature and
extent oflinkage among ﬁnns.

While our results support the interactions between ﬁnancial and strategic
decisions, they do not answer the question of how the ﬁnancial choices (e.g., debt and

cash) interact to inﬂuence the real competitive strategy. This is an important issue for

68

ﬁrms to seek optimal ﬁnancing package that secures the effectiveness of strategy. Further

research on this topic can be insightful.

69

APPENDICES

70

APPENDIX A

Tables for Essay 1

71

This table records the steps in selecting the exogenous failed sample for the period from 19.90
to 2006. The original transaction data are from the SDC M&A database and the reasons for
deal termination are hand-collected from Factiva. All acquirers are. publicly traded ﬁrms listed
on the NYSE, N asdaq, or AMEX. Targets are comprised of public, private, or subsidiary firms.
Panel A contains the number of domestic and foreign deals after each query step. Panel B
lists the number of domestic and foreign transactions that fail due to various acquirer—related
or target-related matters. After screenng out any deals for which the cause of termination
may be, even if indirectly, related to the acquirer’s valuation. the exogenous failed sample

Table A. 1.
Sample Selection

consists of 272 merger bids.

Panel A: Selection of Failed Sample

 

 

Search in SDC

 

Sarrrple Size

Query Description

 

Domestic Foreign

 

SDC M&As announced between 1/1/1090 and 12/31/2006

 

85876 1.1712 Acquirer: US. public trading firm
Percent of Shares Acquirer is Seeking to Own
61.915 1169.9 after Transaction: 50% or higher
3144 4.34 Deal Status: ﬁrst offer in cycle but uncornplctcd
1593 270 Deal Value is 1 million dollars or more.
1.3.34 237 Neither Acquirers Nor Targets are utility or financial firms
1235 2:31 Acquirers are identiﬁed in CRSP
1235 251 Failed Sample

 

 

Panel B: Selection of Exogenous Failed Sarrrple

 

 

Hand Collected in Factiva

 

 

 

Domestic Foreign Reasons for Deal Failure
1235 251 All unsuccessful merger bids
-'12 —‘2 2'\cquisition of bidder
—16.1 -‘28 Inability to reach agreement/Inability to conclude negotiation
-38 -10 Inability to obtain financing
~56 -7 Fall in acquircr's stock price/Finalrcial problems in acquirer
—12 —7 .-'\cquirer shareholders Opposition/Bad market reception
-ll -1 Management ctnillicp/(‘hz—rnge of rriarragcr's
-l.()l —‘2'l Change of merger form/Change of target
—'23 —l Ilill'crcnce in corporate growth strategy or culture
—lil'2 —'~.ZT Target rejection or lack of response or delay
—7-1 -‘21 \Vorscning conditions in target/lbw diligence revelation
43!) -!1 ."\cquircr“s or 'l‘nr‘gct's failure to fulfill certain conditions
- )0 —7 (l'lranging rnarkct condition or industry environment.
393 3).! No reason given/No enough inforrnatit)n/ctimpounding events
2‘21) ”)2 Exogenous Failed Sample

 

 

7‘2

Table A.2
Sample Summary Statistics

This table reports descriptive statistics for the exogenous failed sample. Panel A shows the
deal characteristics. The deal value ($ million) is from SDC and represents the total value of
consideration paid by the acquirer, excluding fees and expenses. The market value of equity
(3? million) is the shares outstanding times the stock price at the ﬁscal year-end prior to the
merger announcement. The market value of assets ($ million) is the book value of assets
minus the book value of equity plus the market value of equity. Relative size is the deal value
divided by the equity market capitalization of the acquirer at the end of the month prior to
the acquisition announcement. Days to termination measure the number of calendar days
between announcement and termination dates. The cash (equity) in payment is the percent
cash (equity) payment of the deal value. Pure cash (equity) deals are when 100% of the
consideration is cash (equity). Acquisitions are deﬁned as competed deals, hostile deals, or
tender offers as reported by SDC. Conglomerate deals involve targets with a two-digit SIC
code other than that of the bidder. Panel B summarizes the characteristics of acquirers.
Following Bhandari (1988), debt-to-equity ratio is deﬁned as the book value of assets minus
the book value of equity divided by the market value of equity of the acquirer at the end
of the ﬁscal year prior to the acquisition announcement. Tobin’s q is deﬁned as the market
value of equity plus the book value of debt and preferred stock divided by the sum of the.
book value of equity, debt, and preferred stock at the end of the ﬁscal year before the merger.
Governance index is the closest reported Governance Index Score from Gompers et al. (2003)
for the acquirer. Operating cash flow is deﬁned as sales minus the cost. of goods sold, sales
and general administrative costs, and the change in working capital. Small dummy is equal
to one if the acquirer’s market capitalization is less than the market capitalization of the 25th
percentile of NYSE ﬁrms in the same year. Panel C shows the distribution of the exogenous
failed sample by industry using the Fama and French (1997) industry classiﬁcation. Columns
2 through 5 report the number of acquirers and targets, respectively, in a particular industry.
Column 6 reports the number of bidders acquiring targets in their own industry.

73

Table A.2. Continued

Panel A: Deal Characteristics

 

Exogenous Failed Deals (11:272)

 

 

Mean Median
Deal Value (DV) 1730.35 190.28
DV / Market Value of Assets 0.3569 0.1863
DV / Market Value of Equity 0.6645 0.5713
Relative Size 0.8176 0.2766
Days to Termination 79.46 53.15
Cash in Payment (W) 47.15
Equity in Payment (‘77:) 40.77
Pure Cash Deals ((70) 50.00
Pure Equity Deals (‘7?) 30.15
Competed Deals (%) 29.35
Hostile Deals (%) 11.39
Tender Offers (%) 15.81
Conglomerate Deals (‘70) 40.44
Foreign Deals ((70) 19.11
Public Target (”/c) 63.60
Private Target (%) 14.34
Subsidiary Target (V) 22.06

 

 

Panel B: Acquirer Characteristics

 

 

Exogenous Failed Deals (112272)

 

 

l\'lean Median
Book Value of Assets 1624.35 716.31
Market Value of Assets 6224.98 712.30
Market Value of Equity 2836.64 322.50
Debt/Equity 0%) 97.27 47.67
Tobin‘s q 3.3157 1.7058
Governance Index 9. 18 10
(5)C..'F/.\Iarket Value of Assets 0.2692 0.1981

Small

0.2096

 

 

Panel C: Distribution by industry

Table A.2. Continued

 

 

 

 

Target i\' umber of own
Industry Acquire Target Public Private Sub. industry bids
Agriculture 1 3 2 0 1 1
Aircraft 5 0 0 0 0 0
Apparel 4 5 3 2 0 3
Automobiles 4 -1 4 0 0 2
Business services 25 29 21 4 4 16
Business suppliers 4 3 3 0 0 2
Candy and soda 2 2 1 0 1 2
Chemicals 6 5 3 0 2 4
coal 0 1 0 0 1 0
Computers 10 10 8 0 2 5
Construction 1 0 0 0 0 0
Construction materials 3 5 4 0 1 1
Consumer goods 4 5 2 1 2 3
Defense 1 2 1 0 1 0
Electrical equipment 5 7 6 l 0 3
Electronic products 10 17 12 3 2 5
Entertainment 15 10 4 4 2 5
Food products 10 8 3 0 5 7
Healthcare 9 6 3 , 2 1 5
Insurance 10 9 4 4 1 7
Machinery 11 5 4 0 1 3
Measuring and control equipment 8 2 2 0 0 1
Medical equipment 13 14 10 4 0 9
Miscellaneous 1 1 1 0 0 0
Nonmetallic mining 2 2 2 0 0 2
Personal Service 1 2 l 1 0 1
Petroleum and natural gas 7 7 6 0 l 6
Pharmaceutical 7 7 4 2 1 4
Precious metals 3 3 2 0 1 3
Printing and publishing 1 0 0 0 0 0
Real estate 1 2 0 0 2 0
Recreational products 4 l 1 0 0 1
Restaurants. motels. hotels 11) 13 9 2 2 6
Retail 13 13 9 1 3 9
Rubber and plastics 0 l l (l 0 0
Shipbuilding. railroad 1 1 4 0 0 1
Shipping containers 1 0 0 (l 0 0
Steel works '1 l 3 0 1 (l
Telecoimmmicat ions 29 26 14 l 8 25
Textiles l I l (I (V) 0
Transportation 20 21 7 2 12 19
\Vliolesale ") '2 7 '2 3 3

 

 

Table A.3.
Cumulative Abnormal Returns upon Merger Termination

This table presents the cumulative abnormal returns around termination announcement for
acquirers who did not close merger deals due to exogenous reasons. The acquirer’s cumulative
abnormal return upon deal termination is measured during (-1, +1) trading days around
the termination date. The single-index normal market model is used to measure abnormal
returns. The estimation period uses 200 daily observations from 250 days before to 50 days
before the merger announcement date. Similar to Martin (1996), the method of payment
is grouped into three categories. Cash ﬁnancing includes cash only or a mixture of cash
and debt. Stock financing includes common stock only or a combination of common stock,
options, or warrants. Combination ﬁnancing comprises combinations of common stock, cash
debt, preferred stock, convertible securities, and methods classiﬁed as ”others” by SDC. The
t-statistics are in brackets and the number of bids is reported beltwv the t-st atistics. Difference
tests are based 011 the t-test for equality in means.

Cumulative abnormal returns upon termination by status of targets and method of payment

 

 

 

 

 

 

 

 

All Cash Stock Combo Difference Tests

(1) <2) (3) (1H2) (1H3) (2)43)
All deals 0.22% 0.08% 0.04% 0.84% 0.04% -0.76% -0.80%
[0.52] [0.15] [0.05] [0.83] [0.04] [-073] [-058]

272 136 82 54
Public Targets 2.19%?" 2.18‘70‘" 1.92%" 2.60%"* 0.27% -0.41% -0.68%
[1.03] [3.00] [2.25] [3.00] [0.24] {-0.36} [0.53]

173 69 64 40
Private Targets -4.l6%"* —2.93%' —6.22%M -4.02% 3.29% 1.08% -2.20%
[3.17] {-1.81] [-249] [1.00] [1.10] [0.32] [0.51]

39 19 12 8
Sub. Targets -2.60%*" 4.77%” -7.47% -4.41‘/c 5.70%" 2.64% -3.06%
[3.15] [-255 [-170] [1.03] [2.33] [1.09] [-050]

60 48 6 6
Priv.+Sub. Targets 45.21717" 32.10%”, 45.64907“ 4.19% 4.54%" 2.09% 2.45%»
[-117] [3.12] [3.00] [-150] [2.07] [1.10] [0.72]

99 67 18 I4

 

Difference Tests

 

Public vs. Priv.

0.30% W

5.12%”"

8.14%”‘

0.02%

 

 

 

 

[5.1.3] [3.1.0] [3.62] [2.07]
Public vs. Sub. 4.79%7” 3.96717” 940%,.” 7.02%"
[0.1 1] [3.32] [3.05] [2.50]
l’riv. vs. Sub. 4.30% 4.10% 1.20% 0.10%
[1.00] [41.78] [0.27] [0.07]
I“ Denotes significance at the 1% level.

Denotes significance

Denotes significance

at the 5% level,
at the 10% level.

Table A.4.
Cumulative Abnormal Returns upon Merger Announcement

This table presents the cumulative abnormal returns around merger announcement for bidders
who in the end have to terminate the merger due to exogenous reasons. The acquirer’s
cumulative abnormal return upon merger announcement is measured during (-1, +1) trading
days around the merger announcement date. The single-index normal market model is used to
measure abnormal returns. The estimation period uses 200 daily observations from 250 days
before to 50 days before the merger attempt announcement date. Similar to Martin (1996),
the method of payment is grouped into three categories. Cash ﬁnancing includes cash only
or a mixture of cash and debt. Stock ﬁnancing includes common stock only or a combination
of common stock, options, or warrants. Combination ﬁnancing comprises combinations of
common stock. cash debt, preferred stock, convertible securities, and methods classiﬁed as
”others” by SDC. The t-statistics are in brackets and the number of bids is reported below

the t-statistics. Difference tests are based on the t-test for equality in means.

Cumulative abnormal returns upon announcement by status of targets and method of payment

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

All Cash Stock Combo Difference Tests
(1) (2) (3) (1H?) (ll-(3) (21-(3)
All deals -0.76% -0.18% 41.90% 4.99%," 0.72% 1.81%7 1.10%
[1.37] [0.30] [0.01] [1.93] [0.55] [1.70] [0.5.1]
272 136 82 54
Public Targets -2.47‘7o -0.75% -3.72%* 3.45% 2.97%" 2.70% -0.27%
[1.08] [1.10] [1.89] [1.21] [2.08] [1.23] [0.15]
173 69 64 40
Private Targets 4.33%” 4.04% 1.42% 2.26% -2.54%* 3.30% 4.96%
[1.87] [0.53] [1.52] [0.03] [1.81] [0.88] [1.58]
39 19 12 8
Sub. Targets 0.87% 0.98% 4.08% 2.03% 2.05%} 4.06% -3.l0%
[1.25] [1.42] [0.29] [0.75] [0.89] [0.19] [0.07]
60 48 6 6
Priv.-+810). Targets 2.24% 0.41% 1.13% 2.16% 4.81%" —l.76% 4.45%
[1.20] [0.55] [1.11] [0.95] [0.07] [0.92] [1.13]
90 67 l8 14
Difference Tests
Public vs. Priv. 45.81% 0.20% $809“ —5.72%*
[1.05] [0.18] [1.71] [2.00]
Public vs. Sub. 43.35% 4.73%“ $2.649? $1.489?“
[-l.0'2] [4.72] [41.61] {-1.86}
Priv. vs. Sub. 3.53% 32.07% 2.63% 0.31%
[1.00] [0.87] [1.10] [005]
Public vs. l’l‘iv. 1- Sub. 4.71% 4.1697 42.8367" 3.62%
[1.21] [1.15] [1.89] [1.19]
""1 Denotes significance at the 1% level,

Denotes significance at the 71% level.

Denotes significance at the 10'2”? level.

RI
\1

Table A.5.
Firm and Deal Characteristics: Sort by Target Status

This table reports descriptive statistics for the exogenous failed sample sorted by target status.
The sample contains all failed mergers and acquisitions that did not close due to exogenous
reasons between 1990 and 2006 as listed by SDC where the publicly traded acquiring ﬁrm
seeks to gain control of a public, private, or subsidiary target whose deal value is at least
$1 million and 1% of the acquirer’s market value. Panel A and B present deal and acquirer
characteristics, respectively. Variables are deﬁned as in Table 11.

Panel A: Deal Characteristics

 

 

 

 

 

 

 

 

Public Target Priv. Target Sub. Target

N=173 N=39 N=60
Deal Value (DV) 2355.68 802.38 530.51
DV/Market Value of Assets 0.5985 0.3189 0.3284
DV/Market Value of Equity 0.8368 0.5169 0.3355
Relative Size 0.9628 0.6644 0.4986
Days to Termination 78.69 78.97 82.02
Cash in Payment. ("/7') 41.16 52.24 45.00
Equity in Payment (‘77) 50.15 38.12 15.46
Pure Cash Deals (‘2’?) 34.68 48.72 80.00
Pure Equity Deals (‘%) 36.99 30.77 10.00
Competed Deals (‘77 ') 36.99 15.38 30.00
Hostile Deals (W) 15.03 2.56 6.67
Tender Offers (‘77:) 21.97 7.69 3.33
Conglomerate Deals (‘77) 38.15 61.54 33.33
Foreign Deals (‘77) 19.07 20.51 18.33
Panel B: Acquirer Characteristics

Public Target Priv. Target Sub. Target

(N=173) (N239) (N=60)
Book Value of Assets 5068.61 3597.26 3661.01
Market Value of Assets 6469.72 6161.10 1650.84
Market Value of Equity 4452.25 2675.31 2719.75
Debt /Equity (9? ) 88.81 72.37 137.85
Tobins q 3.28 1.89 1.88
Governance Index 9.64 8.87 9.10
()CF/Markct Value of Assets 0.2988 0.1829 0.2399
Small Dummy 0.1792 0.1872 0.1167

 

 

383:5 :M @3598 v.8 mosmsﬁm-» “EB dommmgwm: .395 E @3535 E: Beebe :86 Remix: 6:: .5ch .2533 E. E 22::30
:25 mm Boom 08?: 8:929:80 .Eﬁmao @5203 E mm:a:o 95 6800 925865353 122% “:5 wide. .28 team .3 Ewes. .2: 5:5:
wine. we Umpﬂsoﬁmu mm Bow ammo wEumEQO .338 mo 9:? 9938:: as» .3 cove/:8 ﬁasco :o:::ou mo 03.9.» zoom 2: 35:: 3.7.1.1. :33 .3
egg x00: of we. :8:ch mm mmgmgq 5:088:5055 $92: 23 8 :25 Emamﬁﬁm Reagan :58: $2: of :5: .45? 1;._.:....E.a 1::
“for .335 m0 .5352 #003 “0 8:5 23 :3 @833: 2.00% Umbﬂma 9:: ”Eu: mo wig x00: .2: 2:: 3:25 9.5.55.5 2: .3 2:13 9.2.5::
.2: mm :2va E U @5205 .u:c:$o::o::m Swag: .23 3 :25 5:9: 0:0 go mm 355% .2: we 55.851373 $53: 2: ..3 7.53.5.
2:5» :oﬂucm:ms 9: m: mm? 3533 EH. .:oS§me:oo E :8: zoos». mo 83:5:ng 23 mm :oﬁxSEéou E .Aﬁiom— ...z._;.E...:?u.: 683:;
.25 :af 29: mm 0:2: 2:33 :m 6:: ,UQm 3 9:?on 2:8: mm :9: Juno 53:8 a m: :3... .5558 .2: .3 :2: :3: :32: 35: U?
ﬂirésﬁ a E v.55 we .29». :5: 5:5 @954 we czﬁacuc: 5mm 2: 2575 v: :cﬁxszzﬁaec 3:55 89:: 5:5 t5 .m.:.:S.E£_:.L .7. .55:
$835 ho $8563an :8 o:o mo miss. 05 8:3 use. 833.51; 55:38 :a 28 682550 98. .258: £328 Jot: 81:8 ..35:2?wa5
AREA .iﬁcgsm “88.3.5 8:13 ”.8255 95:538. 2: .5 05H was :BEE ; 3.1.2 ac. E 2:; 1%: 833.3 gumbo.» 2:32:32; :3 .3825
.232: m .3 33:8 :Ew 3 $68 BE @5938. move: A3225: 8 95:3 .OQm :3 68m: we meow 9:: egg :ccBZ: 2:76;: 55:39.8
3 :3. owe—u as: E: 3:: mSw.§: 3:3 2: 2:3:8 29:8 27H 4:22:55“: Sw.§: we :§:wo::2:§ do 2:: 2: :8 :.:.:..: 122.55?
cux:8952 7387M 9.83:5: 8:5 25 gimme: mm 03:28..» 33:88: 2: 0.85.: 255556: E:2..So.n7mx3s we 75:5: 2: 2:.7....:_. ..33 5:7

9:90 nowvmm mo mmmbwcdw :ommmmuwom 3:238:880
65¢. 03:8

79

.3er 0&2 2: E ouErEM—rai wretioa

F
4&5— o\om. E: as 3.2.551? $3.75 3
€53 .XL 2: .2 ckEéEm? .r..:::;2 in

 

 

 

 

 

w; mo> we.» 3% 9%» far 85:: ~r..x_.._ €253
mo> mm> mm> m®> wm> m¢> numtm MVP/AL 22TH
.NExm $5.3 9am: x38 $3.8 m3? E :_,.n_.,.:_._.<
8 2,. m: EN EN Em 2% 93:3.
5.??3? _.$.c-_8.o- 5.38.? 39:3- 3.13.3- 83-5 .3- 5:5 ...5§.:.5o
3.25.0 5.33.0 Edie ridge 5.55.3 3.25.3 23:2; iiimuo
$35.3 5.1%.? Lgafg 5.23m: $.23: 322%.: exaszj
3.3-5.3- 5.3-33.3- 33.33. 5.123- 3m.o-_3.e- $3-55- : is;

Sa.c-_cn.c-
L313:
82-5.7
ﬁcimv?
:5.o-§.m-
5.1%;
33-53-

.._£.ﬁ-_£.c-
5.3::
_;.N.o-:.n.n-
$.23:
going-
30.???
33-33-

.:_:m.m-§.c-

Mm_.,..o:s.c
$343.7
Soﬁa?
53-33-
3.2%..“
3.0-5.9-

:;_:._l.m-§..c-
31:3:
:Sﬁﬁ-
:;.2-§.s-
33-52-
80:2;
$3-3”?
33.133-

5:5:

5:3;

...ra.m.-_$.c-
$3-? ._-
in???
:aﬁﬁf-

going-

:uvgmcd

KNEW;

,.*_M..A..n-_wn.c-
Sufff
55-5.3-
7:733-

35-333-

..me ...:._:‘__;:
:.:,:...Z_:.f.:cu E >..::rm
Exec 7.221223
3:73:

.3325

.52: .Etzﬁ.
31.._.:::_w:30

3:2: fins :4

 

€5-33- :Ndiao- 312.9- :z.o-:o.o- 3,2,:5; 22.3 5

Esra: :Ezm

533$ 33%? ...;:m.m_§.o ..zzixia

..Law153 5.2.23.“ .1535 a 3.2.5

2: A5 CV Q; Amv CV $355.; :E—zrxfti
:zm .Em 23$ =< :< :4.

 

.‘:,...::3.=:—C::£ C36 13:25.! ESQ: . ... .r. ..EHSLVLQ T: _ X I H ..L A 42.35" .:L Ier;
. . . . . . ﬂ .

 

 

32.5.80 6.4 £an

80

49:: “NA: 23 an 92:35:wa 3.27.5

at

. . ‘ .. D 1 4
3?: Km 2: E $28,525? wBEED

II.

.75.: XL 22 a: a..:~....uc_:mi $5232

'3‘“!

 

 

 

 

 

o v. mm. mm
<2 5.: an»; :51
0 3553.7 1933.7 3:52.” manor/6c. o: 5.2: Ears 4:; 2.5.3
.o m 3 mm.
:5: :51 an; :31
$2.6. $8.3- “53;. 25:6- WEEZE 5:: $35 .L:./_ :23,
n m. mm. “n
$.93 :51 5.1 25.2
”55$- ﬁned- 1&34- iﬁuiﬁ- EmSm coEmEZES: s .3: .3m.:: :5 2.5.5
n w 9... mm
:5: 33.. E: 3.2
$635- ..Xbmﬁ- $3.7 L2: .M... :52; Eciéit : xx; ..uw...~:.. .25. :2;
n m. 2 Z...
391 an; $53 :52
353.3- gmad- ...KNETM- isoxﬁmd- €8.52 83:5 6. as; 5m._§ 417.. 2.3:»
n ~ on an
5.1 <2 :31 3.1
mNLwA- ﬁne.” $2.7 053.7 3.9:“; 332$ .2 .53 3.3:; 4:7. 2.5.5
3:250 53m :30 :<

 

 

 

$03,353?“ 23 3333 m“ £9.92: mo Spas: .3: USU. mpexpms E v,.:.:..9: Es 331.331.-“
0:8 .38: 338.5 Una Edam Soc UoEmEO Ea mecca :oSqummEu .3632: .bmzxmna 2: .3 ﬁg: mm 255 .5: E 05 «3273.5
.2..::EE 91.2.3 Egan $255695 a ma recast E £8me t..£m$3t < £86ch we $55.2: “E: miEE; heat; as I“:7._._;_~.E.::_; .3
water. £3ng .ﬁmzxmézm Mo 32563 .5“ 5058:5025“ :OS‘QEEHB 832: com: 2:38 3555? v.25s325o of 3:57.. 223 E;

mmmammvmmpﬂm mo mumumsvoda new mannamm comamcmaumﬁ
.54 @358

81

Table A.8.
Cross-section Analysis of Gains to Acquirers of Subsidiaries

This table reports the results of multivariate regressions for acquisitions of subsidiaries in the
exogenous failed sample. The dependent variable is negative one times the acquirer’s 3-day
percentage abnormal returns upon merger termination. Private parent, diversiﬁed parent,
and use of advisors are dummy variables that take the value of one if the target subsidiary’s
parent is private, if the target subsidiary’s parent is diversiﬁed, and if the parent of the target
subsidiary hires ﬁnancial advisors, respectively. The public status of parents of subsidiary
ﬁrms and information on ﬁnancial advisors come from SDC. A diversiﬁed parent is deﬁned
as a parent whose primary two-digit SIC is not same as that of the subsidiary. Industry
classiﬁcation codes are obtained from Fama and French (1997). The t-statistics are reported
in brackets.

 

 

Dependent variable
Independent variables = (-1) x the acquirer’s CAR upon termination announcement

 

Private Parent
Diversiﬁed Parent

2.49[1.83]*

3.15[l.26]

2.83[1.81]*

3.13[i.30]

 

 

Use of Advisors 1.02[0.40] 0.75[0.30]
In Equity (market) -0.06[—O.21] -0.12[~0.43] —0.01[-0.0.‘3] —0.13[—O.42]
Conglomerate l.26[0.47] 2.26[0.90] 1.95[0. 78] 2.69[1 .06]
T(_'n(_ler Offers 2.29[1.:52] 2.13[1.03] 3.14[1.(54] 2.83[l.3(5]
Foreign -3.22[-0.88] -l.32[—0.40] -4.28[~l.34] -2.75[—0.90]
Hostile -4.48[-0.91] -3.95[—0.81] -5.22[-1.10] -4.93[-1.01]
Competed Deals -1.89[-0.6‘2] -1.33[—0.45] -2.04[-0.68] -l.45[-0.48]
Equity in Consideration 0.05[1.25] 0.06[1.60] 0.05[1.48] 0.07[l.91]
Relative Size O.64[l.20] O.46[O.90] 0.66[l.28] 0.51[1.00]
Tobin‘s q 0.88[0.83] 0.53[O.53] 0.72[O.7l] O.50[O.48]
Leverage 0.6-4[1.0‘2] 0.41[O.74] 0.45[0.82] 0.39[O.64]
OCF/Assets (market) -1.09[-0.17] -2.10[-0.37] —0.81[-0.15] 0.09[0.01]
Governance Index -().56[—1.‘27] -0.35[—0.87] -().5‘2[—1.21] -0.33[—0.79]
Sample Size 60 60 60 60
.»'\djnsted R2 31. 12% 32.72% 28.91% 29.29%
Time Fixed Effect. Yes Yes Yes Yes
Industry Fixed Effect Yes Yes Yes Yes

 

 

tilt

t1

Denotes signiﬁcance at the 10% level.

Denotes signiﬁcance at the 1% level,

Denotes signiﬁcance at the 5% level.

82

APPENDIX B

Tables for Essay 2

83

 

 

 

 

 

a- E - .52. can 7 E E a .53 a .n..:._....,<
a- a..- 052- exam- a. E n :E 1: :55 ._3....../
<2 a ex. ex: 7 3 an m E: 2; em;
3 7 s2- .52- mm E n $3 3.6 15.525
Nu m .u Raﬁ: TXLHHa Fm Cm. m. anNA 3.7V 1227722222; Thai—2am
S- 2- 2.2- 052. E S m :62 $5.3: :55 3.7.51.2
E- 8. $3- $2. an I a 33.2 2.30 7.3.3
.- 2- ,5- .5”. cm. : E 5...: .5 5.1.2::
3- B- as- 5:- on an .o 33% 2; £55 .55;
m..- o 0%- $2- 8. a. m :cé ea 4? 5.5;... 27:22
n: a :ALI ﬂNUNu wN mm N me. ICCALJQNCJA 233.7126.
m. N- .53. x2- 8 Q a 2:43 a: 35.5 2:2
H- a- Na- 95. mm .5 m 2: a; :5 ...5 .._..:_._e2.
35064 .3254. m< pom EEO .253 £30 52m 5205QO Am :2; >.:E::c.u
misExzm 33535 mc\vC\naaH as 39.5% we Evian we 5:2 xizm N2:

 

E ..xzmgu gmoaém >52 Eozagzsozzxr. £2225 Oeuoxnorh #5 :EE

 

 

 

Acute 2:: no.“ 25095 8 $9». 15.1.: :2: 5 Two 3.5.3»: 2: .3
@333. «:1. 35%: 5:33 m2: :< .EECQ emcﬁaoouca 7.x 13?:va 8a memzxzu "Vassar NE: E 31?. .xE::&% ~..c:w_;,T:.:E_:_
.3 Casewzzc: .Bcc 2mg 32.. mica. .mﬁrzzucm CEELTCPEQE 5 magic 5:3? E: is mi: 2 m2: :3: >5: “7...... 1:: 73:7. 5
memzmzu amomkm .53 $585 82?; mo mama o... 25 x5365 83:15 3 52?. :55 923519: .3 .28:53:32:: 2: :3 £25.:
~§:.5:Qm 22 89.02% 22%? .muzmﬁmem 85 :w no exam 9: mo 55m .2: c3 Amammmmv 81?. .mE.E&cm w.§..€.7:...:::z 2: $53.5.
..3 $83955 ﬂ mmwﬁmzp ﬂooﬁxeaéukﬁg one 5.“ wcﬁxm we ”Eugen“ Nag E mace 63.57% as $295QO :5 .3 2:23 31?. Z: .3 5%
of $55: amicm mod? .cvd£:.Et Echcm B<Hmbn=200 Ecc Ea v.55 we aceEmcm :33 we v..E:;E..m.€ mm.::.x.:; .t. .5: .EH
£23565 vauﬂanooomﬁoaéo: E 8:2:me a £32 as was hose can 3?th c8£2§5m€5 23 :5: Bars NE: E 31% :2: we
ﬁEuBQ 3 $.22 use. a #5323. v5 math 637mm? 5 2:5 «deﬁcmamcéuownoa UmEmSEU .2: no.“ 5:? SE 7.6.39: LEE 15H

Suﬁ Shh unmﬁcmamvooQMQoﬁ
.H.m mEmE

8—1

Table B.2.
Segment Data for Tobacco-dependent Firms

This table shows data for the non-tobacco segments in tobacco-(leperident ﬁrms. Firms
are classiﬁed as being tobacco-dependent if at least 10 percent of their sales in 1992 were
from the tobacco-related industry and they had at least a segment in non—tobacco-related
industries. Segments shown here are restricted to those operating in non-tobacco-related
industries. Segments that operate in utility or ﬁnancial services industries are deleted. I also
exclude segments with incomplete information and segments that 1992 sales were less than
$50 mil. SIC code is the primary segment industry code according to Standard Industrial
Classiﬁcation reported by COMPUSTAT segment database. Percent of Sales is computed
by dividing the segments sale by the sum of the sales of all segments in the ﬁrm. AC'F/S
is the change in the ratio of segment cash ﬂow to sales between 1992 and 1993, expressed
as percentage points. Cash ﬂow equals pretax operating proﬁt plus depreciation. All 1993
dollar ﬁgures are deﬂated by the average CPI in that year.

 

 

1992 Size Percent of

 

 

Company Segment Industry SIC Code (mil $) Sales ACF/S
Alliance One Intl Inc Flowers 5193 360 33 —0.65
Altria Group Inc Food Products 2000 29.048 49 0.26
Altria Group Inc Beer 2082 3.976 7 1.87
American Maize Corn Processing 2046 392 72 -2.78
Fortune Brands Inc Home Improvemt 3432 1,014 11 -2.21
Fortune Brands Inc Specialty Business 5990 2.214 5 1.13
Fortune Brands Inc Distilled Spirits 2085 716 8 5.71
Fortune Brands Inc Ofﬁce Products 2782 1,003 11 0.60
Hanson Plc Consumer Products 2834 1.158 10 0.31
Hanson Plc Coal Mining 1221 1.801 16 -0.50
Hanson Plc Lumber 2411 394 3 3.57
Hanson Plc Industrial Products 3824 1.429 12 1.10
Hanson Plc Chemicals 2816 967 8 2.17
Hanson Plc Housebuilding 1531 1,963 17 1.13
Hanson Plc Material Handling 3537 A160 4 -6.33
Hanson Plc. Gold Mining 1041 189 2 NA
Loews Corp Hotels 7011 N A NA NA
Loews Corp Timing Device 509-1 181 l —-'1.-1‘2
Loews Corp Drilling 1381 218 2 15.26
Nabisco Group Foods 2052 6.707 43 -0.45
Standard Comnwrcial \\'ool 2209 3‘19 28 0.65
Universal Corp Building Products 5031 370 12 0.13
Universal Corp .-\gri—l’roducts 51-19 38-1 13 —().18
UST Inc “inc 2081 75 7 1.67
Vector Group Ltd Collectible Pictures 1759 87 12 55.78
A vertige 2.3] l 17 3.22

 

 

Table B.3.
Statistics for Rivals to Non-tobacco Segments

This table records statistics for rivals to non-tobacco segments in tobacco-dependent ﬁrms.
Non-tobacco segments are those as shown in Table II, excluding those with no rivals available
or with incomplete information. Industry rivals are identiﬁed as ﬁrms with at least one
segment operating in the same four-digit SIC industry as the non-tobacco segments of the
tobacco companies, as identiﬁed in the COMPUSTAT segment database. Size denotes the
average market value of rival ﬁrms within the same industry. Market value of a ﬁrm in one
ﬁscal year is share outstanding times the stock price at the ﬁscal year end. Leverage denotes
the average debt-to—equity ratio for all rivals, calculating as the book value of total asset minus
the book value of equity divided by the market value of equity. Tobin’s q (OCF) is the mean
market-to—book ratio (operating cash ﬂow) for rivals within the same industry, calculated
as standard. The industry Herﬁndahl index is calculated as the sum of the squared market
shares of all ﬁrms that at least operate a business segment in that industry. The last row of
the table shows the average statistics for rivals across all non-tobacco—related industries. All
ﬁgures are from COMPUSTAT consolidated and segment database.

 

 

Rivals to Tobacco Firms

 

 

Non—Tobacco—Related Size OCF
Segment Industry SIC Num (mil 53) D/E Tobin's q (mil 5%) Herf. Index

Gold Mining 1041 70 698 0.44 2.08 17.23 504
Coal Mining 1221 12 2,192 1.05 1.49 18.43 2069
Drilling 1381 29 1.094 0.66 0.37 15.92 697
Housebuilding 1531 45 990 2.21 1.38 17.30 733
Food Products 2000 6 8.325 0.14 2.84 20.46 2881
Corn Processing 2046 5 1,997 0.70 1.93 18.61 3058
Foods 2052 2 419 0.03 1.52 17.41 8495
Beer 2082 6 4.590 0.13 1.73 18.06 3123
W'ine 2084 7 5.899 0.40 1.51 17. 75 5031
Distilled Spirits 2085 2 5,807 0.17 1.54 20.08 5899
\\'001 2299 1 122 0.30 1.58 16.88 6892
Lumber 2411 2 774 0.44 1.54 18.72 4425
Collectible Pictures 2759 3 165 0.55 2.10 17.03 3567
Ofﬁce. Products 2782 5 1,015 0.05 2.3-'1 27.110 3161
Chemicals 2816 6 6,213 0.88 1.85 18.67 3558
Consumer Products 283-1 160 3.681 0.11 1.17 17.93 397
Home Iinproveint 3432 4 1.298 0.29 1.06 17.75 4754
Material Handling 3537 11 205 0.62 0.79 16.94 1771
Industrial Products 3821 7 2.018 0.12 1.50 15.57 4802
Building Products 5031 l 255 2.55 0.89 17.19 2658
Timing Device 509—1 1 578 0.23 2.87 17.07 6626
Agri—Products 5119 i 1.536 2.19 1.86 18.12 8797
Hotels 7011 11 979 2.83 1.11 16.76 2912
A verage 2.075 0.80 1.32 17.36

 

 

86

._ .3

.‘t .

Table B.4.
Abnormal Returns to Non-tobacco Rivals of Tobacco Firms

This table presents the abnormal returns of industry rivals to non-tobacco segments in
tobacco-dependent ﬁrms. Industry rivals are identiﬁed as ﬁrms with at least one segment
operating in the same four—digit SIC industry as the non-tobacco segments of the tobacco
companies, as identiﬁed in the COMPUSTAT segment database. Single-segment. rivals are
those that report only one industry segment. Small rivals are those with market share less
than 5%. A standard market model in event study is employed to calculate abnormal returns.
The market model is estimated over the 200 trading period from 240 to 41 days before the
announcement of negative cash shock to tobacco industry. Abnormal return is calculated
over the one-day event window centered on the announcement date of the shock. Abnormal
returns to rival portfolios are calculated for the equally-weighted portfolios of rivals in same
industry. T-statistics are based on tests that standardized abnormal returns are equal to
zero. The adjusted t-statistic is from the regression that adjusts the White standard error
estimates for clustering by industry. Signiﬁcance of the number of positive versus number
of negative is assessed using a sign test. Difference tests are based on t-test for equality in
means. The t-statistics for difference tests are shown in brackets. Subsample of the rivals are
selected for those that compete in the industry with the pre—announcement Herﬁndahl Index
larger than 2000, or those that. compete with non-tobacco segments with market share less
than 10%.

 

 

Subsample of Rivals

 

All Industry Non-tobacco Seg.
Herf. Index 1,2000 market Share 1 10%

 

Abnormal returns to individual rivals

Abnormal return 1.02% 1.00% 1.60%
t—statistics 760*" 7.36"" 5.96“”
Adjusted t (clustering) 2.40” 2.47" 2.40"
Positive, negative 230,158’" 150.1135" 157,893."
Difference Test. [0.94] [2.91“*]

 

Abnormal returns to rival portfolios: single segimznl only

Abnormal return 066% 0.23% 2.54%
t-statistics 2.08” 0.79 2.92"
Positive, negative 12.7 3.2 5.0M

 

:lln'iormal returns to rival portfolios: small rivals

 

 

Abnormal return 0.68% 0.50% 2.53%

'l—statist ics 1.64 1.76 2.57M

Positive. negative 11.9 2.3 5.0"
it?

Denotes significance at 1‘77. level.

" Denotes significance at. 5‘}? level,

* Denotes significance at 10% level.

I," h .

:‘t‘

 

 

 

 

 

 

 

S: a: a: 3.: a: 6: G: S:

9.32. :92- as: $.me Essa- 3.82. 3% :15 s+ 3 T
a: a: a: a: Q: a: a: a:

3.4.2- semi- ES Raw 2.3.3- 8.3a- was? wees 4+ 3 2.
Am: 9: a: a: a: a: a: :3

95 mi 3? 5.5 3.34. 332.- 3.3» 2.25 f. 2 T
S: a: Q: 3: a: :3 a: :3

43s 3.2.. 8.3 4.3% was- is. £43.. 3.5.. ...+ 2 a-
s: a: a: a: so as so so

23 2.5a 2...: 3.2. Es- 43- a: s3 2+ 3 T
Q S E E Q E E E 55>

voumsgxx 33m $33.25 3cm

 

 

£03 .83 :so

 

$in $2225 .5:st

 

 

955% 3%.: ﬁx? 5 ~90 am??? .2: ..3 waists:

zoosm :mmO wnmspozom mew—3:0 «59.50

.m.m 2pr

v.5 magma 3:30 9: :< .mwwnﬁﬂo Ewﬁmcw B<HmDnHEOU Eat 0328\6 was Sad 23 =< .magﬁzcaxa 2: E :82? E #527.
035mm: .2: “up? So» magozﬂ Some 5 madam ooomnou E maceawmm 8523-:0: m0 535:: 25. .soEaEBﬁ: 8335322 no £152
mﬁvzoamoboo o: 9%.: ﬁt: 358me $2? mmpiuxm A3 55:8 233 E 226%. E wmusomoa muaoamom 0856.720: 2: :1. 255:3
:V 55:00 8.8233 Sm measuﬁmnzm no 2552603 93 5 no 55¢.sz 8353 mo @53an 2: E 5 .mﬁztoa 39:53 .3 0:7. .5
:oSmﬁoamch .wﬁcceu ,cozoswoa on... 5 go .8238 m0 9:5me 9: E «5302: we? ﬁx: 5.52:: b8 E 55:15 :8..s—eaéyzsé5
< .mwiumswﬁ vmuﬂeauovoenofco: E 3325 233 madam 08.38 we ESQ mam umﬁ $05. 0.8 musmﬁwem 8.5.38-5: .231 2513::
25% on“. E 053 $5 muses—mom 2s mo macaw 23 89c EOQV mQ mo made» 535 3: .EOQV mQ mo 203.3?530 some .5.“ .uscﬁpzx
E E tom: Costa acoEumzﬁtauxSmscﬁ .455 2: mcsczc.“ Ease». c>¢ of mo Lode can :53 cu xcczm 39%;: m::_Smcxc of E
SE use» 23 "~vaqu mucoEwem ooownouéo: mo Bow £30 98 moﬁmm E mewcsso gigginsmsvﬁ ~35 >2: masts: 22.5 7.38

88

Table B.6.
Robustness under Alternative Benchmarks

This table presents the results for robustness checks on abnormal returns to industry rivals.
Industry rivals are identiﬁed as firms with at least one segment operating in the same four-
digit SIC industry as the non-tobacco segments of the tobacco companies, as identiﬁed in
the COMPUSTAT segment database. In estimation of the standard market model, two new
benchmarks are employed. First is based on the S&P 500 market portfolio with all the
tobacco ﬁrms excluded and the other is based on the CRSP value—weighted market portfolio
with largest. 500 public non—tobacco-dependent ﬁrms. The market model is estimated over
the 200 trading period from 240 to 41 days before the announcement of negative cash shock
to tobacco industry. Abnormal return is calculated over the oneday event window centered
on the announcement date of the shock. T—statistics are based on tests that standardized
abnormal returns are equal to zero. Signiﬁcance of the number of positive versus number
of negative is assessed using a sign test. Diﬁ’erence tests are based on t-test for equality in
means. The t-statistics for difference tests are shown in brackets.

 

 

Subsample of Rivals

 

All Industry Non-tobacco Seg.
Herf. Index {.2000 market Share i 10%
5641’ 500 market portfolio with all tobacco firms excluded is used:

 

 

Abnormal returns to individual rivals

Abnormal return 066% 0.65% 1.22%
t-statistics 4.70”" 4.5?" 4.04""
Positive, negative 213.1%?" 156.124"* 143.101?"
Difference Test. [0.72] [2.82***]

 

CRSP value-weighted market. portfolio with all tobacco ﬁrms ercluded is used:

 

Abnormal returns to individual rivals

Abnormal return 0.54% 0.50% 1.10%
t—statistics 318*“ 3.11”” 3.18””
Positive. negative 202.186“ 149.131" 134,112"
Difference Test [0.74] [2.89"]

 

 

*" Denotes significance at "1% level.
**

Denotes signiﬁcance at S‘X; level,

Deliotes significance at l0% level.

89

 

 

 

 

T..La.m_ 7&2 .7.qu ¢.:E._;.::.:
***awg~m: *..mw:,pom .**mm:,vmm Q>e:mﬁz_,¢>rzmcm
Emma 33.» is: Ecéii
“N994 05:04 Mare; .25.: ._:E.:::_<
m.~c_‘.~ai Naxﬁtzﬁs S nigfg ViEgezev‘
05A: _ 9:25. €322: :EEwum 08:38-:0: :32: 332553 5:265 :<

 

£qu .3 ingxmnqﬁ

 

 

 

 

 

 

 

 

 

 

 

:: a: a: 2: a: :3 a:
£5 MEN no.3 Neda- 9.57 madam... 3:.an m+ 3 T
Q: a: a: :3 :5 :3 $3
3.2 at: no.» 3.3.- 3.3. 3.2.: 23m..- Tr E T
as :V A: Auv A: A»; :V Izrwr
c338»: 6333;: and

+

E03 32: $5

 

AWQV am:§_0 wixm

 

 

:5» 35¢ $2: 5 7H0 $5.52.: of .3
Cayman: S: myswm 8:0: 9:5 .252: E .Ssmsvm .5.“ “may-“ :0 woman 98 3mg 8:98mwn— .umﬁ :mﬁ a mHHMm: Cmmmmmns 2. $31.3:
.3 #55:: 35:5 9,360: :0 Hun—8:: 9: .Ho 8:305:me .23 3 12:5 :5 9:32 3:295: maﬁvbzéﬁm :2: 5.4,: :3 cows:
.2: moﬁmsgmuh .xuocm m5 m0 3st EmEouSSEE 2: :o @88ch BOCEB E95 kmmcéso 9: .55 tug—:35 E 55.: 12,5254.
$50553: v5 5 :39? mm xuonm o>5ame 9: Sim Sum 9:30:an 5% E 2:5 08:28 E mEoEmmm 08:32-22: ..5 .r:.:::: BEL
.:2.§:5ﬁ: 32:88:: :0 Edi: w:€:ommvtg o: 9%.: R9: 3:2:wmm mmoﬁ 8:588 55:5 Amv 5:38 32.5 125;: z: £:.E&.7.
38:23-:0: 2: 2m 553:8 A: 5:300 .95 wESZ 925% E mamawmﬁ 2: @5259wa 52:56:: 1:EETDBEB:3: E 4:53;:
2:33 maﬁa 09:58 we 53: PE HS: .365 E: mEczacm EBEELE: EEK .mﬁEEmcm :c £2.33... _¢:£E€: 5:5: EF.:::% .72:
3 3:3: We 3:38 35:93: 95 3:25:me 08:38-:0: We 09:55 gage :0 .3605 $252275 :3 £318 of $.39: :33 x29

29:25.0 anmﬁﬁvdw .865: mmmaumznom

Knm 03:5

90

Table B.8.
Regressions of Abnormal Returns to Rivals

This table presents multivariate regressions that explain the abnormal returns to imlustry
rivals. The dependent variable is the percentage abnormal returns to the industry rival. The
ﬁrst independent variable is industry market share of the non-tobacco segments that suffered
from the negative cash shock. Herﬁndahl Index is calculated as the sum of the squared market
share of all firms operating a segment in the industry. Market share of the rival ﬁrm is the
percentage of rival ﬁrm’s sale to the whole market sale. Leverage is the average debt—to—equity
ratio calculated as the book value of total asset minus the book value of equity divided by
the market value of equity. And Tobin’s q is the mean market-to—book ratio as standard.
Multi-segment dummy is a binary variable that takes value one when the rival ﬁrm operating
on multiple lines of business. The t-statistics are reported in brackets.

 

 

Speciﬁcation (1) (‘2) (3) (4)
Dependent variables

 

 

Rival's abnormal returns

 

market share of -0.05 -0.06 -0.06 -0.06
non-tobacco seg. [225]" [-‘2.68]M [264]" [228]"
Herfindahl index ~0.03 -U.03 -0.04
[-1.64]* [-1.80]* [~1.85]*
market share of rival 0.03 0.03
[1.19] [1.06]
Leverage 0.00
[1.29]
Tobin’s Q 0.00
[1.30]
Multi-segment dummy -0.00
[-0.50]
Intercept 0.01 0.01 0.01 0.01
[445?” [3.62]"* [3.62]"* [0.92]
Adjusted R2 0.01 0.01 0.01 0.03
Num of Observations 436 436 435 410

 

 

ti$

11*

*

Denotes signiﬁcance at. 5‘5?

Denotes signiﬁcance at. 1% level.

6 level,

Denotes signiﬁmmre at 10% level.

91

 

. a: o. ..E ¢.::+.:.:&:m 1.93:;
m m 5:: ._. . 2

i

.75.: exam ..3. 3:55:35 5::52

. PS .9. is 9:5: Em? no. 3:...
\o . . .

A i

*I‘.

 

 

 

24m ismdm 9K ...iomdn 1.3me 923%.: .cﬁﬁmcm
3.: . 2...: them- :8: ...3. Ecési
$33.: :33 .2 Samoa- ﬁnnd RENO $53.: _§:.5:.§.

WkQEmeib Ndﬁwzagpwﬁtk. 3w mgrtztk. N3.~:L.C.:.c T

 

 

. . .V .
so: so:
.wom 09538-:0: .mem 08323-:02 2
.8 9:21 “MESS 2:3 mo 93:.» “3:2: 2:3 game who: fr: 25% ..«EE 4:? (.
7.223.565 5. £12.: :8 7.2.3.565 5. £33.: :8 52265 E is»: :8 $2.12:on E $1.3: :3 :<

 

32:32: 0 a 2:234 :
_ .

 

 

.pmop :mﬁ a wEm: vommommm m: Piedmo: m0 555:: 95:9, oiﬁmo: mo 52:5: 2: .3 9:239:me .oSN
8 35m .95 2:30.: $8.85? wommwaewﬁﬁw ﬁx: mung :0 women mum moﬁmsﬁmé. .xoozm 23 mo 35: 32:3:525: of :3 t.o.:..::eo
30:53 E98 mattio 2: :95 toad—:33 mm 552 E8552 25:3: 8258 o: :ﬂtozm ammo we as: EEEEEEEc 2:
88.02 was: 3 o: ovm Sch v2.8: chm: com 9: :25 8.853% m: 3:2: :25: EH. @533 7:50:25 8:,ng o... coho—:25
m: 33:? :55 E 7:58 €325: €535? < swag :55 seesaw .3 0%: :55 £3 aﬁaﬁmem 08323.::: a: 9:27. “vii: 4:3
mewsmzva: 5 m8 35>: 5:965 23 .5523 95 doom. :wﬁ Salaam :o ooom 5:: mmﬂ 2:? 36.9553 2:? mavsmzci E E: 7.3::
..Eaﬁﬁﬁ E: .852? :c tome: 33:85:; 58 SE #513? Cm? mm 295?. :3 SC. 6E :EEmcm 2.35:3: .ﬁ/ﬁmeE/XVU :5:
BEEEE use $85830 A: 2an E «59:83 332 3:365 .3 295% m5 :8 3:52 $50.35 €5,253: 2: 3971.3 $32 7:5.

Enid muamzﬁam mo 2050350 3 main—”Em 3:20:34“
.m.m 03:8

BIBLIOGRAPHY

93

BIBLIOGRAPHY

Acharya, V., H. Almeidab, and M. Campello, 2007, Is cash negative debt? A hedging
perspective on corporate ﬁnancial policies, Journal of Financial [ntermediation 16, 515-
554.

Almeida, H., M. Campello, and MS. Weisbach, 2004, The cash ﬂow sensitivity of cash,
Journal of Finance 59, 1777-1804.

Andrade, G., M. Mitchell, and E. Stafford, 2001, New evidence and perspectives on
mergers, Journal of Economic Perspectives 15, 103-120.

Agrawal, A., J.F. Jaffe, and G.N. Mandelker, 1992, Post-merger performance of
acquiring ﬁrms: A re-examination of an anomaly, Journal of Finance 47, 1605-1621.

Asquith, P., R.F. Bruner, and D.W. Mullins, Jr., 1983, The gains to bidding ﬁrms from
merger, Journal of Financial Economics 11, 121-139.

Bae, K.H., J.K. Kang, and J .M. Kim, 2002, Tunneling or value added? Evidence from
mergers by Korean business groups, Journal of Finance 57, 2695-2740.

Barber, B., and J. Lyon, 1997, Detecting long-run abnormal stock returns: The empirical
power and speciﬁcation of test statistics, Journal of Financial Economics 43, 341-372.

Bates, T.W., K.M. Kahle, and RM. Stulz, Why do US. ﬁrms hold so much more cash
than they used to?, Journal of Finance, forthcoming

Baumol, W.J., 1952, The transactions demand for cash: An inventory theoretic approach,
Quarterly Journal of Economics 66, 545-556.

Benoit, J ., 1984, Financially constrained entry in a game with incomplete information,
RAND Journal of Economics 15, 490-499.

94

 

 

Bhagat, S., M. Dong, D. Hirshleifer, and R. Noah, 2005, Do tender offers create value?
New methods and evidence, Journal of Financial Economics 76, 3-60.

Bhandari, LC, 1988, Debt/Equity ratio and expected common stock returns: Empirical
evidence, Journal of Finance 43, 507-528.

Blanchard, 0.1., F. Lopez-de-Silanes, and A. Shleifer, 1994, What do ﬁrms do with cash
windfalls?, Journal of Financial Economics 36, 337-360.

Bolton, P. and D. S. Scharfstein, 1990, A theory of predation based on agency problems
in ﬁnancial contracting, American Economic Review 80, 93-106.

Boone, AL, and H.J. Mulherin, 2007, How are ﬁrms sold?, Journal of Finance 62, 847-
875.

Bradley, M., A. Desai, and EH. Kim, 1988, Synergistic gains from corporate acquisitions
and their division between stockholders of target and acquiring ﬁrms, Journal of
Financial Economics 21, 3-40.

Brander, J.A. and T. R. Lewis, 1986, Oligopoly and ﬁnancial structure: The limited
liability effect, American Economic Review 76, 956-970.

Brown, D.T., and MD. Ryngaert, 1991, The mode of acquisition in takeovers: Taxes and
asymmetric information, Journal of Finance 46, 653-669.

Brown, S.J., and J .B. Wamer, 1985, Using daily stock returns: the case of event studies,
Journal of Financial Economics 14, 3-31.

Campbell, J.Y., AW. Lo, and AC. MacKinlay, 1997, The econometrics of ﬁnancial
markets, Princeton University Press.

Campello, M., 2003, Capital structure and product markets interactions: Evidence from
business cycles, Journal of Financial Economics 68, 353-378.

95

Campello, M., 2006, Debt ﬁnancing: Does it boost or hurt ﬁrm performance in product
markets?, Journal of Financial Economics 82, 135-172.

Chang, S., 1998, Takeovers of privately held targets, methods of payment and bidder
returns, Journal of Finance 53, 773-784.

Chang, S., and D. Suk, 1998, Failed takeovers, methods of payment, and bidder returns,
Financial Review 33, 19-34.

Chatterjee, S., 1992, Sources of value in takeovers: Synergy or restructuring -
Implications for target and bidder ﬁrms, Strategic Management Journal 13, 267-286.

Chevalier, J .A., 1995a, Capital structure and product-market competition: Empirical
evidence from the supermarket industry, American Economic Review 85, 415-435.

Chevalier, J .A., 1995b, Do LBO supermarkets charge more? An empirical analysis of the
effects of LBOs on supermarket pricing, Journal of Finance 50, 1095-1112.

Chevalier, J.A., and D.S. Scharfstein, 1995, Liquidity constraints and the cyclical
behavior of markups, American Economic Review 85, 390-396.

Chevalier, J .A., and D.S. Scharfstein, 1996, Capital-market imperfections and
countercyclical markups: Theory and evidence, American Economic Review 86, 703-715.

Cohen, R.B., C.Polk, and T. Vuolteenaho, 2003, The value spread, Journal of Finance 58,
609-642.

Davidson, W.N. 111, D. Dutia, and L. Cheng, 1989, A re-examination of the market
reaction to failed mergers, Journal of Finance 44, 1077-1083.

Dcllas, H. and A. chandes, 2007. Finance and competition, unpublished working paper.

Dittmara, A. and J. Mahrt-Smith, 2007, Corporate governance and the value of cash
holdings, Journal of Financial Economics 83, 599-634.

96

 

Dodd, P., 1980, Merger proposals, Managerial discretion and stockholder wealth, Journal
of Financial Economics 8, 105-137.

Dodd, P. and J .B. Warner, 1983, On corporate governance: A study of proxy contests,
Journal of Financial Economics 11, 401-438.

Dong, M., D. Hirshleifer, S. Richardson, and SH. Teoh, 2006, Does investor
misvaluation drive the takeover market?, Journal of Finance 61,725-762.

Eckbo, BE, 1992, Mergers and the value of antitrust deterrence, Journal of Finance 47,
1005-1029.

Eckbo, B.E., R. Giammarino, and R. Heinkel, 1990, Asymmetric information and the
medium of exchange in takeovers: Theory and evidence, Review of Financial Studies 3,
651-676.

Eckbo, BE, and KS. Thorbum, 2000, Gains to bidder ﬁrm revisited: Domestic and
foreign acquisitions in Canada, Journal of Financial and Quantitative Analysis 35, 1-25.

Faccio, M., and J. McConnell, and D. Stolin, 2006, Returns to acquirers of listed and
unlisted targets, Journal of Financial and Quantitative Analysis 41, 197-220.

Fama, E., abd K. French, 1993, Common risk-factors in the returns on stocks and bonds,
Journal of Financial Economics 33, 3-56.

Fama, E., abd K. French, 1997, Industry costs of equity, Journal of Financial Economics
43, 153-193.

Faulkender, M. and R. Wang, 2006. Corporate ﬁnancial policy and the value of cash,
Journal of Finance 61, 1957-1990.

Fee, CE. and S. Thomas, 2004, Sources of gains in horizontal mergers: Evidence from
customer, supplier, and rival ﬁrms, Journal of Financial Economics 74, 423-460.

97

Franks, J ., R. Harris, and S. Titman, 1991, The post-merger share-price performance of
acquiring ﬁrms, Journal of Financial Economics 29, 81-96.

Fresard, L., 2008, Financial strength and product market behaviors: The real effects of
corporate cash holdings, unpublished working paper.

Foley, C.F., J .C. Hartzell, S. Titman, and 6.]. Twite, 2007, Why do ﬁrms hold so much
cash? A tax-based explanation, Journal of Financial Economics 86, 579-607.

Fuller, K.R., J. Netter, and M. Stegemoller, 2002, What do returns to acquiring ﬁrms tell
us? Evidence from ﬁrms that make many acquisitions, Journal of Finance 57, 1763-1793.

Gale, D. and M. Hellwig, 1985, Incentive-compatible debt contracts: The one-period
problem, Review of Economic Studies 52, 647-663.

Gertner, R., D. Scharfstein, and J. Stein, 1994, Internal versus external capital markets,
Quarterly Journal of Economics 109, 1211-1230.

Gompers, P.A., J .L. Ishi, and A. Metrick., 2003, Corporate governance and equity prices,
Quarterly Journal of Economics 1 18, 107-155

Grinblatt, M., and S. Titman, 2002, Financial markets and corporate strategy, 2nd edition,
McGraw Hill Irwin, New York.

Grossman, S., and 0. Hart, 1980, Takeover bids, the free rider problem and the theory of
the corporation, Bell Journal of Economics 10, 20-32.

Grossman, S. and 0. Hart, 1986, The costs and beneﬁts of ownership: A theory of
vertical and lateral integration, Journal of Political Economy 94, 691-719.

Hansen, R., 1987, A theory for the choice of exchange medium in mergers and
acquisitions, Journal of Business 60, 75-95.

98

 

Hansen, R.G., and J .R. Lott, Jr., 1996, Extemalities and corporate objectives in a world
with diversiﬁed shareholder/consumers, Journal of Financial and Quantitative Analysis
31 , 43-68.

Harford, J ., 1999, Corporate cash reserves and acquisitions, Journal of Finance 54, 1969-
1997.

Harford, J., S. Mansi and W. Maxwell, 2008, Corporate governance and ﬁrm cash
holdings in the US, Journal of Financial Economics 87, 535-555.

Haushaltera, D., S. Klasab and W.F. Maxwellb, 2007, The inﬂuence of product market
dynamics on a ﬁrm's cash holdings and hedging behavior, Journal of Financial
Economics 84, 797-825.

Heckman, J ., 1979, Sample selection bias as a speciﬁcation error, Econometrica 47, 153-
161.

Hege, U. and C. Hennessy, 2007, Acquisition values and optimal ﬁnancial (in)ﬂexibility,
unpublished working paper.

Hendel, 1., 1996, Competition under ﬁnancial distress, Journal of Industrial Economics
44, 309-324.

Hertzel, M.G., M.S. Ofﬁcer, Z. Li, and K]. Rodgers, 2008, Inter-ﬁrrn linkages and the
wealth effects of ﬁnancial distress along the supply chain, Journal of Financial
Economics 87, 374-387.

Hietala P., S.N. Kaplan, and D.T. Robinson, 2003, What is the price of hubris? Using
takeover battles to infer overpayments and synergies, Financial Management 32, 5-31.

Hite, (EL. and J.E. Owers, 1983, Security price reaction around corporate spinoff
announcements, Journal of Financial Economics 12, 409-436.

Holl, P., and D. Kyriazis, 1996, The determinants of outcome in UK takeover bids,
International Journal of the Economics of Business 3, 165-184.

99

Holmstrom, B., and SN. Kaplan, 2001, Corporate governance and merger activity in the
united states: Making sense of the 19808 and 19903, Journal of Economic Perspectives 15,
121-144.

Jarrel, G.A., J. Brickley, and J. Netter, 1988, The market for corporate control: The
empirical evidence since 1980, Journal of Economic Perspectives 2, 49-68.

Jarrel, GA, and AB. Poulsen, 1989, The returns to acquiring ﬁrms in tender offers:
Evidence from three decades, Financial Management 18, 12—19.

Jensen, M.C., 1986, Agency costs of free cash flow, corporate ﬁnance, and takeovers,
American Economic Review 76, 323-329.

Jensen, M.C., 1988, The takeover controversy: Analysis and evidence, in Coffee, J ., L.
Lowenstein and SR. Ackerman (editors) --Knights, raiders and targets: The impact of
hostile takeovers, Oxford University Press.

Jensen, M.C., and R. Ruback, 1983, The market for coeporate control: The scientiﬁc
evidence, Journal of Financial Economics 1 1, 5-50.

Jovanovic, B., and S. Braguinsky, 2004, Bidder discounts and target premia in takeovers,
American Economic Review 94, 46-56.

Kaplan, SN, and MS. Weisbach, 1992, The success of acquisitions: Evidence from
divestitures, Journal of Finance XLVII, 107-138.

Khanna, N. and S. Tice, 2000, Strategic responses of incumbents to new entry: The effect
of ownership structure, capital structure, and focus, Review of Financial Studies 13, 749-
779.

Kim, C ., D.C. Mauer and A.E. Sherman, 1998, The determinants of corporate liquidity:
Theory and evidence, Journal of Financial and Quantitative Analysis 33, 335-359.

Kothari, SB, and J. Warner, 1997, Measuring long-horizon security price performance,
Journal of Financial Economics 43, 301-339.

100

Kovenock, D. and G. Phillips, 1995, Capital structure and product-market rivalry: How
do we reconcile theory and evidence?, American Economic Review 85, 403-408.

Kovenock, D. and G. Phillips, 1997, Capital structure and product market behavior: an
examination of plant exit and investment decisions, Review of Financial Studies 10, 767-
803.

Lamont, 0., 1997, Cash flow and investment: Evidence from internal capital markets,
Journal of Finance 52, 83-109.

Lang, L.H.P., R. Stulz, and R.A. Walkling, 1989, Managerial peformance, Tobin's q and
the gains from successful tender offers, Journal of F inanical Economics 24, 137-154.

Lang, L.H.P., R. Stulz, and R.A. Walkling, 1991, A test of the free cash ﬂow hypothesis:
The case of bidder returns, Journal of F inanical Economics 29, 315-336.

Loderer, C., and K. Martin, 1990, Corporate acquisitions by listed ﬁrms: The experience
of a comprehensive sample, Financial Management 19, 17-33.

Luo, Y., 2005, Do insiders learn from outsiders? Evidence from mergers and acquisitions,
Journal of Finance 60, 1951-1982.

Lyandres, E., 2006, Capital structure and interaction among ﬁrms in output markets:
Theory and evidence, Journal of Business 79, 2381-2421.

MacKinlay, A.C., 1997, Event studies in economics and ﬁnance, Journal of Economic
Literature 35, 13-39.

Maksimovic, V., 1995, Financial structure and product market competition, chapter 27.
Handbooks in Operations Research and Management Science by Jarrow, R.A., V.
Maksimovic, and W.T. Ziemba, vol 9. Finance, Elsevier North Holland.

Maksimovic, V and G. Phillips, 2001, The market for corporate assets: Who engages in
mergers and asset sales and are there efﬁciency gains?, Journal of Finance 56, 2019-
2065.

101

Manne, HQ, 1965, Mergers and the market for corporate control, Journal of Political
Economy 73, 110-120.

Maquieria, C.P., W.L. Megginson, and L. Nail, 1998, Wealth creation versus wealth
redistributions in pure stock-for-stock mergers, Journal of Financial Economics 48, 3-33.

Marris, R., 1964, The economic theory of managerial capitalism, London, Macmillan.

Martin, K.J., 1996, The method of payment in corporate acquisitions, investment
opportunities, and management ownership, Journal of Finance 51, 1227-1246.

Meltzer, AH, 1963, The demand for money: A cross-section study of business ﬁrms,
Quarterly Journal of Economics 77, 405-422.

Mikkelson, W. and M. Partch, 2003, Do persistent large cash reserves hinder
performance?, Journal of Financial and Quantitative Analysis 38, 275-294.

Mitchell, M and K. Lehn, 1990, Do bad bidders become good targets?, Journal of
Political Economy 98, 372-398.

Moeller, S.B., F.P. Schlingemann, and RM. Stulz, 2004, Firm size and the gains from
acquisitions, Journal of Financial Economics 73, 201 -228.

Moeller, S.B., F .P. Schlingemann, and RM. Stulz, 2005, Wealth destruction on a
massive scale? A study of acquiring-ﬁrm returns in the recent merger wave, Journal of
Finance 60, 757-782.

Moeller, S.B., F.P. Schlingemann, and RM. Stulz, 2007, How do diversity of opinion
and information asymmetry affect acquirer retums?, Review of Financial Studies 20,
2047-2078.

Morck, R., A. Schleifer, and R.W. Vishny, 1990, Do managerial objectives drive bad
acquisitions?, Journal of Finance 45, 31-48.

102

Mulligan, C., 1997, Scale economies, the value of time, and the demand for money:
Longitudinal evidence from ﬁrms, Journal of Political Economy 105, 1061-1079.

Myers, SC, and NS. Majluf, 1984, Corporate ﬁnancing and investment decisions when
ﬁrms have information that investors do not have, Journal of Financial Economics 13,
187-221.

Ofﬁcer, MS, 2007, The price of corporate liquidity: Acquisition discounts for unlisted
targets, Journal of Financial Economics 83, 571-598.

Ofﬁcer, M.S., A. Poulsen, and M. Stegemoller, 2008, Target—ﬁrm information asymmetry
and acquirer returns, Review of Finance 1, 1-27.

Opler, T., L. Pinkowitz, R. Stulz, and R. Williamson, 1999, The determinants and
implications of corporate cash holdings, Journal of Financial Economics 52, 3-46.

Opler, T. and S. Titman, 1994, Financial distress and corporate performance, Journal of
Finance 49, 1015-1040.

O'Sullivan, N., and P. Wong, 1998, The impact of board composition and ownership on
the nature and outcome of UK takeovers, Corporate Governance: An International
Review 6, 92-100.

O'Sullivan, N., and P. Wong, 2001, The determinants and consequences of abandoned
takeovers, Journal of Economic Surveys 15, 145-186.

Parsons, C. and S. Titman, 2007, Capital structure and corporate strategy, unpublished
working paper.

Phillips, G.M., 1995, Increased debt and industry product markets: An empirical analysis,
Journal of Financial Economics 37, 189-238.

Poulsen, A., and M. Stegemoller, 2008, Moving from private to public ownership: Selling
out to public ﬁrms vs. initial public offerings, Financial management 37, 81-101.

103

 

Povel, P. and M. Raith, 2004, Financial constraints and product market competition: Ex
ante vs. ex post incentives, International Journal of Industrial Organization 22, 917-949.

Rauh, J.D., 2006, Investment and ﬁnancing constraints: Evidence from the funding of
corporate pension plans, Journal of Finance 61, 33-72.

Ruback, R., 1983, Assessing competition in the market for corporate acquisitions,
Journal of Financial Economics 1 1, 141-153.

Savor, P. and Q Lu, Do stock mergers create value for acquirers?, Journal of Finance,
forthcoming.

Scherer, F .M., 1988, Corporate takeovers: The efﬁciency argument, Journal of Economic
Perspectives 2, 69-82.

Schipper, K., and R. Thompson, 1983, Evidence on the capitalized value of merger
activity for acquiring ﬁrms, Journal of Financial Economics 11, 85-119.

Schlingemann, FR, 2004, Financing decisions and bidder gains, Journal of Corporate
Finance 10, 683-701.

Schroth, E. and D. Szalay, 2007, Financially constrained innovation races: Evidence from
pharmaceutical patents, unpublished working paper.

Schwert, G.W., 2000, Hostility in takeovers: In the eyes of the beholder?, Journal of
Finance 55, 2599-2640.

Servaes, H., 1991, Tobin's Q and the gains from takeovers, Journal of Finance 46, 409-
419.

Stein, J., 1997, Internal capital markets and the competition for corporate resources,
Journal of Finance 52, 111-133.

104

Stein, J ., 2003, Corporate investment policy, appearing as Chapter 3 in the Handbook of
the Economics of Finance Volume 1A, Elsevier North Holland: Amsterdam, edited by
GM. Constantinides, M. Harris, and RM. Stulz.

Telser, LG, 1966, Cutthroat competition and the long purse, Journal of Law and
Economics 9, 259-277.

Titman, S., 1984, The effect of capital structure on a ﬁrm's liquidation decision, Journal
of Financial Economics 13, 137-152.

Travlos, N.G., 1987, Corporate takeover bids, methods of payment, and bidding ﬁrms'
stock returns, Journal of Finance 42, 943-963.

Walker, M., 2000, Corporate Takeovers, strategic objectives, and acquiring-ﬁrm
shareholder wealth, Financial Management 29, 53-66.

Warusawitharana, M., 2008, Corporate asset purchases and sales: Theory and evidence,
Journal of Financial Economics 87, 417-497.

Weston, J .F ., J .A. Siu, and BA. Johnson, 2001, Takeovers, restructuring, and corporate
governance, 3rd edition, Prentice Hall International, London.

Yang, L., 2008, The real determinants of asset sales, Journal of Finance 63, 2231 - 2262.

Zingales, L., 1995, Insider ownership and the decision to go public, Review of Economic
Stuides 62, 425-448.

Zingales, L., 1998, Survival of the ﬁttest or the fattest? Exit and ﬁnancing in the trucking
industry, Journal of Finance 53, 905-938.

105

   

"'ll'llllliiijlﬂlﬁlﬂlﬂiﬂljlllllllllls