Content uploaded by Sanjai Bhagat
Author content
All content in this area was uploaded by Sanjai Bhagat on Feb 21, 2014
Content may be subject to copyright.
SANJAI BHAGAT
University of Colorado
ANDREI SHLEIFER
University of Chicago
ROBERT W. VISHNY
University of Chicago
Hostile Takeovers in the 1980s:
The Return to Corporate
Specialization
HOSTILE TAKEOVERS invite strong reactions, both positive and negative,
from academics as well as the general public. Yet fairly little is known
about what drives these takeovers, which characteristically involve sig-
nificant wealth gains to target firms' shareholders. The question is where
these wealth gains come from.
We examine the sample of all 62 hostile takeover contests between
1984 and 1986 that involved a purchase price of $50 million or more.
In these contests, 50 targets were acquired and 12 remained indepen-
dent. We use a sample of hostile takeovers exclusively to avoid using
evidence from friendly acquisitions to judge hostile ones, as many
studies have done. We examine such post-takeover operational changes
as divestitures, layoffs, tax savings, and investment cuts to understand
how the bidding firm could justify paying the takeover premium. We
also examine the possibility of wealth losses by bidding firms' stock-
holders as the explanation for target shareholder gains.
The analysis of post-takeover changes is complicated because once
the target and the bidding firms are merged, it becomes impossible to
attribute to the target the changes recorded in joint accounting data. As
a consequence, we do not use such data, but rather focus on discussion
in annual reports, 1OK forms, newspapers, magazines, Moody's and
We would like to thank Dennis
Carlton, Eugene Fama,
Kenneth
Frenchy,
Gregg Jarrell,
Michael Jensen, Steve Kaplan,
Rene Stulz, and Lawrence Summers for comments
and
the
Bradley Foundation for financial
support.
2 Brookings Papers: Microeconomics 1990
Value Line reports, and other such sources. Our approach is similar to
the one recently employed by Bhide (1989). The advantage of this
design is that we can attribute the changes we examine, such as layoffs
and selloffs, to the target firm. The disadvantage is that most changes
we examine are biased downward because some may not be reported.
Our calculations suggest that, on average, taxes and layoffs each
explain a moderate fraction of the takeover premium. Layoffs, which
disproportionately affect high-level white-collar employees, explain
perhaps 10-20 percent of the average premium, although in a few cases
they are the whole story. Tax savings are usually somewhat smaller
than savings from layoffs (although they are significant in a larger
number of cases), since debt is typically repaid fairly fast. But tax
reductions are very large in management buyouts, acquisitions by part-
nerships, and acquisitions by firms with tax losses. Large investment
cuts occur infrequently in our sample, and do not appear to be an
important takeover motive. Wealth declines of the bidding firms' share-
holders, similarly, while important in a few cases, are usually small
and cannot be a systematic source of target shareholders' gains.
Our most significant finding is that most hostile takeover activity
results in allocation of assets to firms in the same industries as those
assets. In most hostile takeovers, the bidding firm is in the same business
as, or a business closely related to, that of the target firm. Similarly,
the majority of selloffs, which amount to 30 percent of the acquired
assets, are to buyers in the same business as the assets they acquire.
Overall, of the assets that changed hands in our sample, 72 percent
ended up owned by corporations with other similar assets. By and large,
hostile takeovers represent the deconglomeration of American business
and a return to corporate specialization.
These findings have significant implications for explaining the sources
of gains in hostile takeovers. First, they suggest that the places to look
for the gains are cost savings from joint operations, market power, or
possibly overpayment by buyers of divisions and whole companies.
Some of these gains might be from eventual layoffs that we document,
but others we might not be able to capture. In any event, changes that
result from consolidation of industries are essential for understanding
takeover gains. Second, the findings suggest that incentive-intensive
organizations, such as management buyout teams, investment compa-
nies, or raiders, are not very important in the long run. In our sample,
Sanjai Bhagat,
Andrei
Shleifer,
and Robert
Vishtny 3
only 20 percent of the assets ended up under control of such organi-
zations after two to three years, and this fraction would surely dwindle
if we looked at the assets over a longer period. Control by raiders or
by MBO teams is often a transitory arrangement used to allocate assets
to corporations managing other similar assets.
Potential Sources of Takeover Gains
What is the source of target shareholders' gains in hostile takeovers?
The literature offers a wealth of theories of the sources of takeover
gains in general. One possibility is simply that the stock market under-
prices the target, so that no operational changes are actually needed for
the bidder to profit from the acquisition. Another possibility is that
bidding firms overpay for their targets, perhaps because acquisitions
serve the objectives of managers and not of shareholders. In these cases
the target shareholders' gains are the bidding shareholders' losses. Con-
sistent with this view is the evidence of David Ravenscraft and F. M.
Scherer that the earnings of acquired lines of business in the friendly
takeovers of the 1960s and 1970s did not rise.1
Although underpricing and bidder overpayment might be important,
they are probably not the whole story in hostile takeovers. For example,
Steven Kaplan shows that cash flow (net of capital expenditures) rises
significantly in his sample of LBOs.2 Moreover, substantial anecdotal
evidence also indicates that hostile takeovers are followed by large
operational changes in many firms. Accordingly, we first deal with the
role of wealth changes in bidding firms' shareholders, and then describe
some potentially important changes that can justify takeover premiums,
as well as discuss how these sources fit into existing takeover theories.
Wealth Change of Bidding Shareholders
Target shareholders in hostile takeovers clearly gain significant wealth,
but less is known about bidding shareholders. If they gain as well, then
the analysis of operational changes must come up with greater savings
1. Ravenscraft and Scherer (1987).
2. Kaplan (1990).
4 Brookings Papers: Microeconomics 1990
to account for the wealth gains. If, on the other hand, bidding share-
holders lose, a smaller shareholder wealth increase is left to explain.
Many argue that competition between actual and potential bidders,
as well as the ability of target shareholders to "free ride" on gains,
ensures that most of the gains in takeovers accrue to the target firm's
shareholders. Bidding shareholders gain from a takeover to the extent
that there is a component to the value gain that is not lost through
competition and cannot be appropriated by target shareholders. Such
bidder gains might be particularly large when the bidder and the target
are in the same industry, and special opportunities to the given merger
are not available to other bidders. Considerable evidence, however,
shows that many mergers are driven by managerial rather than share-
holder objectives, which make bidding firms willing to overpay for the
acquisition targets. Overpayment, of course, leads to negative returns
to bidding shareholders.
The evidence is that the bidders just about break even, but the findings
vary by time period and the type of acquisition. No studies examine
bidder returns in hostile takeovers; the only proxy for such evidence is
the finding for tender offers. In the 1980s, bidders in tender offers lost
small amounts of wealth on average.3 This evidence suggests that many
acquisitions are driven by the objectives of managers rather than of
shareholders, so that managers are willing to overpay for the targets to
pursue their own goals. In a sample of mostly friendly acquisitions,
Randall Morck, Andrei Shleifer, and Robert Vishny find that bidding
shareholders are more likely to lose when acquisitions serve managerial
objectives, such as diversification and pursuit of growth.4 Since there
are many "strategic" acquisitions in our sample, we examine the changes
in the wealth of bidding shareholders as one potential source of gains
of the target firms' shareholders.
Strategic Acquisitions
As noted, hostile takeovers often involve the acquisitions of firms
closely related to the bidding firm. Gains from related acquisitions are
likely to come from operating efficiencies (either pure efficiency gains
or wealth transfers) or from increased market power. Gains from market
3. Bradley, Desai, and Kim (1988).
4. Morck, Shleifer, and Vishny (1990).
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 5
power are commonly believed to have driven the horizontal mergers
preceding World War II, before the strict antitrust enforcement of the
1960s and 1970s made such mergers more difficult.5 With the relaxation
of antitrust enforcement during the 1980s, many horizontal combina-
tions once again became possible. Some argue that the gains in some
strategic acquisitions come from this ability to restrict output and raise
price.
Some evidence on the importance of such acquisitions comes from
the findings of Espen Eckbo and Robert Stillman that competitors' stock
prices rise when horizontal mergers are announced, consistent with the
simple model of declining competition.6 However, stock prices of com-
petitors do not rise nearly as much as those of the target, as a simple
oligopoly model would predict. Furthermore, competitors' stock prices
sometimes also rise when the government challenges the initial merger,
which seems contrary to the oligopoly theory. Although this evidence
is still being debated, it suggests that market power is not the whole
story.
Joint operating efficiencies might come from combining research and
development, procurement, distribution, marketing, and headquarters
operations. Some of these operating efficiencies are reflected in layoffs
and other cuts, but others are harder to capture. These gains might be
all the greater if the target is not well run to begin with and is acquired
by a firm with better managers who find more ways to reduce costs.
With the exception of labor cost savings (discussed later), the evi-
dence on efficiency improvements after hostile takeovers is still indirect.
Morck, Shleifer, and Vishny find that targets of hostile takeovers have
low Tobin's q's relative to their industry peers and also that industries
with significant hostile activity have low q's.7 Henri Servaes finds that
bidding firms in tender offers have higher q's than do the targets.8 This
evidence points to better performers taking over poor performers. Ser-
vaes also finds that premiums are higher when the target's q is lower,
suggesting that there are greater efficiencies to realize. Unfortunately,
it can also mean that targets are undervalued or that bidding firms are
overvalued and so have a low cost of capital.
5. Stigler (1950).
6. Eckbo (1983) and Stillman (1983).
7. Morck, Shleifer, and Vishny (1988, 1989).
8. Servaes (1989).
6 Brookings Papers: Microeconiomics 1990
We do not have the data to identify the sources of gains in strategic
acquisitions, other than layoffs. Accordingly, we simply identify ac-
quisitions that appear to have an important strategic element, and try
to determine how prevalent they are in the hostile takeover process.
We present some examples in which operating efficiencies were realized
as well as some examples with the potential for increased market power.
The question of the sources of gains in general remains open, and it is
not clear to us that any simple description of synergistic gains in related
acquisitions will fit the data. We base this conclusion on an examination
of all the accounts (qualitative and quantitative) of our sample takeovers
that we could find. The takeovers are summarized in the appendix.
Labor Cost Savings
Labor costs are by far the largest component of costs in most cor-
porations. Labor cost savings can therefore be one of the most effective
ways to increase cash flow. Such savings can take a number of forms,
including layoffs, early retirements, hiring freezes, wage reductions,
reductions in future pension benefits, and other cuts in compensation.
If some of the employees in the firm are paid more than their marginal
product, then laying them off or cutting their pay can increase the cash
flow and so justify some of the premium.
Previous studies have examined the extent of labor cost savings.
Joshua Rosett considers wage reductions of union employees and finds
that they can explain, at most, 9 percent of the takeover premium.9
Interestingly, up to 21 percent of the premium can be explained for the
subsample in which the chief executive officer changes after the take-
over. However, the wage changes are not reliably different from zero
in most specifications. Andrei Shleifer and Lawrence Summers present
evidence of substantial wage reductions in one hostile takeover-Icahn's
acquisition of TWA-that are large enough to more than justify the
takeover premium. 10
They have only one famous case, however.
Wages, of course, are not the only form of compensation. Jeffrey
Pontiff, Andrei Shleifer, and Michael Weisbach present evidence of
reversions of excess pension assets following hostile takeovers, and
9. Rosett (1989).
10. Shleifer and Summers (1988).
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 7
document the greater incidence of such reversions following hostile
takeovers than following friendly acquisitions. I They estimate that
reversions can explain about 13 percent of the takeover premium in
cases where they actually occur, but they occur in only 10 percent of
cases. Pontiff, Shleifer, and Weisbach argue that pension asset rever-
sions represent a cut in future benefits to the employees. Although these
reversions are a potentially important source of takeover gains in a few
cases, on average they are relatively unimportant.
Overall, the available evidence on compensation cuts of existing
workers does not suggest that such cuts are an important source of
takeover gains in many cases.
A second way to save on labor costs is to reduce employment. Charles
Brown and James Medoff examine control changes in small Michigan
companies and find no evidence of employment reductions. 12 They do
not appear to have a single hostile takeover in their sample. In an early
version of his paper, Steve Kaplan finds some evidence of employment
reductions after defensive MBOs that follow hostile bids. 13 There is no
evidence of employment reductions following friendly MBOs. Kaplan
does not estimate the value of savings from these reductions. Frank
Lichtenberg and Donald Siegel find evidence of substantial employment
reductions surrounding (and mostly preceding) ownership changes-
and not necessarily hostile takeovers.14 They also find that such re-
ductions are primarily among white-collar employees. All these findings
point to the potential importance of employment reductions, although
the studies do not estimate how much money is saved and do not draw
an adequate distinction between hostile and friendly takeovers.
In discussing layoffs, it is important to keep in mind the normative
interpretation of employment reductions. One view, taken by Lichten-
berg and Siegel, is that employment reductions of headquarters per-
sonnel in particular represent a pure efficiency gain.15 Michael Jensen
also argues that employment reductions improve efficiency.'16 Shleifer
11. Pontiff, Shleifer, and Weisbach (1989).
12. Brown and Medoff (1988).
13. Kaplan (1990).
14. Lichtenberg and Siegel (1988).
15. Lichtenberg and Siegel (1988).
16. Jensen (1988).
8 Brookings Papers: Microeconomics 1990
and Summers, in contrast, stress the importance of long-term implicit
contracts in employment relationships. 17 They do not deny that ex post
layoffs might be efficient; rather, they suggest that if continued em-
ployment is part of the implicit contract, layoffs represent a breach of
trust that transfers future wages in excess of marginal product from
employees to shareholders. Moreover, the efficiency gain from moving
workers across firms may be much smaller than their lost wages.
We cannot determine what happens to laid-off workers after the
takeover and how much their compensation falls. The transfer com-
ponent should be measured by the present value of the difference be-
tween current wages and alternative wages. In this paper we attempt
to measure the wage savings from takeovers and to compare them with
the takeover premium. We cannot determine what fraction of these
savings is an efficiency gain and what fraction is a transfer. The con-
clusion that everything is an efficiency gain is premature, given the
universal employee insistence that takeovers reduce their welfare.
Divestitures
Most hostile takeovers are followed by significant divestitures: sales
of divisions of the target companies to other firms, investment com-
panies, or management teams in MBOs. Unlike layoffs, debt increases,
or investment cuts, divestitures do not necessarily imply operational or
financial changes, although such changes may be made by the acquirers
of divisions following the divestitures. We do not identify these changes
since we cannot follow what happens to the acquired businesses. How-
ever, it is useful to describe the extent of divestitures and to note how
they should be interpreted.
Bidding firms might sell off divisions of the target firms simply to
pay off some of the debt incurred in the acquisition. Such a move,
however, does not explain how divestitures fit into the bidders' plans
to justify the premium. For if the target was valued fairly before the
takeover, and if the buyers of divisions pay fair prices for them and do
not make any changes in the operations, then divestitures do nothing
to explain the takeover premium. This section suggests several ways
in which divestitures help pay the premium.
17. Shleifer and Summers (1988).
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 9
First, the target might have been underpriced in the stock market to
begin with, and so if a bustup occurs at fair market prices, takeover
organizers can profit by unbundling undervalued divisions. This theory
of divestitures based on undervaluation of conglomerates has been tested
by Dean LeBaron and Lawrence Speidell.18 They compute breakup
values of conglomerates by applying price-earnings multiples of un-
diversified firms in the same industry as each division of the conglom-
erate to the earnings of that division and adding up the divisions. They
find that this breakup value of conglomerates is typically higher than
the market price. They do not take into account the possibility that
divisions of conglomerates might be getting lower multiples because
they grow less fast or invest in projects with negative net present value.
Nonetheless, their analysis is suggestive.
Second, if the diversified company sold for a fair market value under
its old management, it must be that the pieces are worth more to the
buyers than they are to the takeover artist or under the old management.
One such group of buyers, incentive-intensive organizations such as
management buyout teams or investment companies, can improve the
cash flow by reducing tax payments, cutting investment and employ-
ment, and taking other steps to reduce costs. In these cases, the higher
cash flow comes from more effective management and not from com-
bining two related firms. To be sure, we still cannot conclude in these
cases that the primary source of value gains is efficiency improvements
rather than wealth transfers.
A second type of buyer to whom divisions of a target firm could be
especially valuable is a strategic buyer with his own operating company
who can either exploit the combination of the two firms or is simply
willing to overpay for the division. Such strategic buyers are in fact
much more common than divisional MBOs, and selloffs to them are
probably the main reason that bustups are profitable. Just as with initial
acquisitions by strategic buyers, it is not known whether these selloffs
to strategic buyers improve efficiency or just redistribute wealth away
from consumers, suppliers, or other firms in the industry.
Tax Savings
Reductions in taxes can come from a variety of sources, the first and
most obvious of which is the savings from merging a profitable company
18. LeBaron and Speidell (1987).
10 Brookings Papers:
Microeconomics
1990
with one that has tax losses. The combination can realize the tax benefits
sooner than the owner of tax losses can alone, and therefore these
benefits are more valuable to the combined entity. Alan Auerbach and
David Reishus have shown, and we confirm with our data, that these
tax benefits are important in some 5 percent of the cases. 19
The second source of tax savings before the 1986 tax reform was
the general utilities doctrine combined with accelerated depreciation,
which enabled the acquirer to redepreciate the target's assets in some
cases without the selling company's having to pay capital gains taxes
on the assets sold. Unfortunately, no available studies gauge the mag-
nitude of this source of tax savings.
The third source of tax savings has been conversion of the target,
or some part of it, into a partnership, such as a master limited part-
nership. As a result, double taxation of profits-at both the corporate
and the individual level-could be avoided. This loophole is largely
closed now, but was available during our sample period.
Perhaps the most important potential source of tax gains in takeovers
is increases in leverage and deductibility of interest payments on debt.
These gains can be achieved if the target was underleveraged to begin
with, so that increases in leverage create corporate-level tax benefits
not offset by significantly higher expected costs of financial distress or
personal taxes paid by bondholders. In the case of leveraged buyouts,
Steven Kaplan shows that tax savings from leverage explain at least 50
percent and perhaps more than 100 percent of the takeover premium
on average.20 This number makes all the other sources of gains pale
by comparison. To estimate the value of the tax shield, we would need
information on how fast the debt is repaid. Our rough estimates suggest
that the potential for increasing value through higher leverage in hostile
takeovers is indeed substantial but still much smaller than Kaplan's
estimate for the MBOs.
Investment Cuts
Jensen argues that takeovers stop target firms from investing their
surplus cash in negative net present value projects.2"
The takeover gains
19. Auerbach and Reishus (1988).
20. Kaplan (1989).
21. Jensen (1986).
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 11
are realized because the money previously wasted is now distributed
as interest or dividends. Jensen cites oil exploration expenses and di-
versification by conglomerates as examples of wasteful expenditures.
Part of the gain in a takeover, then, is a commitment to stop such
investments.
Inferring waste from post-takeover investment cuts is not as simple
as it looks. First, cuts in investment can simply be a response to low
availability of internal funds rather than a source of value. Kaplan finds
evidence of significant investment cuts in management buyouts, but it
is not clear whether these cuts eliminate waste or productive invest-
ment.22 Second, the divestitures and fast debt repayment that typically
follow hostile takeovers relax the cash constraint and therefore might
actually make further wasteful investment easier even in Jensen's model.
His model does not explain why firms try to get back so quickly to the
level of cash flow at which they can invest. Despite these reservations,
Jensen's free cash flow theory is a tenable explanation of hostile take-
overs in several industries, and we try to evaluate its empirical rele-
vance.
Summary
These potential sources of target shareholders' wealth gains in take-
overs-bidding shareholders' losses, gains from strategic combina-
tions, layoffs, divestitures, tax savings, and investment cuts-are germane
to a variety of takeover theories. Unfortunately, one can rarely infer
from the evidence about these operational changes whether takeover
gains are dominated by efficiency improvements or by wealth transfers.
Strategic takeovers probably increase operating efficiency but might
hurt consumers, suppliers, or industry rivals as well. Layoffs both raise
efficiency and transfer wealth from workers earning substantial rents
at their jobs. Divestitures raise efficiency if firms are sold to better
management teams, but they also reflect transfers as underpriced com-
panies are busted up or as divisions are sold to overpaying growth-
oriented managers. In the latter case, efficiency might fall. Tax savings
are largely transfers from the government; they are not efficiency gains.
Finally, investment cuts can eliminate wasteful overinvestment, but
22. Kaplan (1990).
12 Brookings Papers. Microeconomics 1990
they can also transfer wealth from the investment sector (for example,
oil field services). Moreover, investment cuts necessitated by high le-
verage might actually decrease total wealth. The result is that many of
the changes following hostile takeovers cannot be unambiguously in-
terpreted as wealth transfers or efficiency improvements-they prob-
ably have elements of both. Nonetheless, the question remains of how
important some of these changes are empirically.
Description of the Data
As noted, the sample in this paper consists of all 62 firms that were
targets of hostile takeover offers of more than $50 million, as reported
by W. T. Grimm for the period 1984-86.23 Like most others, W. T.
Grimm classifies hostile takeovers as those in which the target's board
at least initially expressed opposition, if only to raise the price. The
sample stops in 1986 to allow a look at changes in the two to three
years following the takeover. The sample includes successful hostile
takeovers, hostile takeover attempts in which the actual acquisition was
completed by a "white knight" (including defensive MBOs), and un-
successful hostile takeover bids in which the target remained indepen-
dent. Whenever possible, we point out systematic differences between
outcomes for targets that are successfully acquired by hostile bidders
and outcomes for those that are ultimately acquired by white knights
or remain independent. A substantial coercive element is probably pres-
ent in almost all of these deals.
Table 1 presents the distribution of our sample of hostile takeover
attempts, by outcome, during the three years of our sample. The number
of attempts nearly tripled. The ratio of takeovers by white knights
relative to straight hostile takeovers also rises substantially, perhaps
because of the increased availability of investment resources on the part
of large firms that usually act as white knights over the course of the
economic expansion.
23. We used the $50 million cutoff because too little information was available for
smaller deals. This restriction
eliminates
few observations,
since being
a target
of a hostile
bid is typically
a privilege
of only very large
firms with diffuse
ownership
structures.
Sanjai Bhagat, Andrei
Shleifer,
and Robert
Vishny 13
Table 1. Distribution of Sample of Hostile Takeover Attempts in the United States,
1984-86
Outcome 1984 1985 1986
Successful takeover 7 12 10
White knight acquisition 1 8 12
Unsuccessful bid 2 4 6
Total 10 24 28
Source: Andrei F. Rhoads, ed., Mergetstat Review 1986 (Chicago: W. T. Grimm, 1986); and authors' calculations.
The most important constraint on the analysis of post-takeover ex-
perience is that the bidder and the target are merged. That is why
Ravenscraft and Scherer's line-of-business sample is in many ways more
informative than our own, although it comes from an earlier period and
hence has few hostile takeovers.24 For the same reason, Kaplan can get
much better information in some respects by focusing on MBOs.25 Our
research is limited by our inability to separate the target and the bidder,
and our reluctance to use the joint data.
Throughout the analysis, we rely on multiple sources of data, in-
cluding the Center for Research and Security Prices (CRSP) stock price
data; bidder and target annual reports; 10 K forms; Moody's Industrial
Manual; Value Line Investment Survey; the Wall Street Journal; the
New York Times; business periodicals, especially Business Week and
trade publications; and DATEXT. We try to piece together the infor-
mation from all these sources because post-takeover information typi-
cally concerns the merged firm, making the attribution of changes to
the target difficult. Most important, for our analysis of layoffs, selloffs,
investment cuts, and tax savings we use only quantitative information.
Our numbers should not be confused with journalistic opinions. Because
reporting can be incomplete, however, most of our measures of changes
are probably biased downward. We measure only what is reported. In
interpreting our results, it is crucial to keep in mind these data limi-
tations.
24. Ravenscraft and Scherer (1987).
25. Kaplan (1990).
14 Brookings Papers: Microeconomics 1990
Wealth Changes of the Shareholders
The first step in documenting the wealth changes of shareholders is
to calculate the gain, or the premium, paid to the target firm's share-
holders.26 We compute the premium as follows. First, we take the (often
long) period from 20 days before the first bid for the target is announced,
to the day when the target accepts or defeats the final bid. We call the
first date Datel and the second Date2. We then estimate the market
model for each target firm from 260 trading days before Datel to 60
trading days before Datel. Using that market model, we forecast what
the price of each target would be on Date2 given its actual value on
Datel and the return on the market between Datel and Date2. The
premium is the difference between the price paid (or offered in the last
bid in the case of unsuccessful takeovers) for the target and the predicted
price on Date2. In other words, the premium is the difference between
what was paid (offered) and what the price would have been on the
resolution date had no takeover activity occurred. We use the long
interval to take account of market movements during the period of
negotiation. At the same time, we do not use any market prices, other
than the price of 20 days before the very first bid, to compute the
premium because market prices reflect a variety of market beliefs and
hence do not reflect the premium alone. Using such prices generally
leads to lower estimates of the premium than our procedure.
Computing changes in the wealth of the bidding firm's shareholders
is more complicated because there is no equivalent of the price paid.
Also, using a long interval in this case creates significant problems,
because many large acquirers have dramatic value changes over long
intervals that might have nothing to do with the acquisition. Accord-
ingly, to evaluate the change in the wealth of the bidding shareholders,
we define Date3 as the date of the first bid by the actual acquirer. Date3
always falls between Datel and Date2, and it often coincides with one
26. By focusing on shareholders, we ignore other financial claimants such as bond-
holders. There is no systematic analysis of the returns to bondholders in hostile takeovers.
The analyses of takeovers as a whole show that bondholders unprotected by covenants lose
small amounts of wealth, whereas protected bondholders do not lose-see Asquith and
Wizman (1989). The magnitude of bondholder wealth losses is quite small relative to
shareholder wealth gains.
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 15
of them. We then estimate the market model for the bidding firm from
260 days before to 60 days before Date3 and use this model to compute
the abnormal change in bidding shareholders' wealth from 3 days before
to 3 days after Date3. We define the return to bidding shareholders as
this abnormal wealth change.
We do not estimate bidding shareholders' returns in unsuccessful
bids because we have found no satisfactory procedures. Accordingly,
in these cases we ignore the bidding shareholders and try to explain
the premium offered to the target shareholders in terms of changes
implemented after the bid.
The results for the takeover premium and for the increase in the
wealth of the bidding shareholders are presented in table 2. Throughout
this paper, we deal with dollar rather than percentage wealth changes
since it is dollar values that we try to explain. The premiums in this
sample vary significantly, and some of them are very small when ad-
justed for market movements. These are typically the cases of bids
highly contested by the target's management where the bidder gained
control in part by buying shares in the open market. In general, the
premiums are somewhat smaller than those one would obtain without
the market correction because the market rose during this period.
The results on bidders are often missing. We do not compute the
returns to bidders in unsuccessful takeovers. Furthermore, many of the
bidders are not listed on CRSP because they are foreign companies
(Hanson Trust, Campeau, First City Financial) or private companies
often owned by the raiders (Reliance Capital, Maxxam, James Gold-
smith, Asher Edelman, and so forth) or leveraged buyout specialists.
As a result of these omissions, we report changes in bidding share-
holders' wealth in only 30 cases.
The results are similar to the usual findings for bidders, except we
find that bidder returns are negative in more than half the cases. On
average, the bidders lose $15 million, a tiny fraction of the average
acquisition price of $1.74 billion (including debt)'. Unavailability of
data might bias these results toward finding poorer performance by
bidders, since raiders typically bid through private firms and their re-
turns are more likely to be positive. Note, however, that Irwin Jacobs,
in two acquisitions by Minstar, earned a negative market-corrected
return in both cases.
Bidder returns are very negative in some related acquisitions, such
Table
2.
Changes
in
the
Wealth
of
Target
and
Bidding
Shareholders,
Sample
of
Hostile
Takeover
Attempts,
1984-86
Millions
of
dollars
Change
in
Target
Bidder
Takeover
premium
bidder
wealth
Total
1.
Aegis
Minstar
22
-
1
21
2.
American
Motor
Inn
Prime
Motor
Inn
60
19
79
3.
Allied
Stores
Campeau
1,106
n.a
n.a.
4.
American
Natural
Resources
Coastal
763
117
880
5.
AMF
Minstar
74
-25
49
6.
Anderson
Clayton
Quaker
Oatsa
108
-
165
-57
7.
Atlas
Van
Linesb
Wesraya
27
n.a.
n.a.
8.
Avondale
Mills
Walton
Monroe
Millsa
38
n.a.
n.a.
9.
Carter
Hawley
Hale
The
Limitedc
397
.
.
.
10.
Carter
Hawley
Hale
The
Limitedc
1,162
...
...
11.
CBS
Turner
Broadcastingc
2,754
.
.
.
...
12.
Chesebrough-Ponds
Unilevera
1,440
43
1,483
13.
Cluett
Peabody
Westpoint
Pepperella
77
-
14
63
14.
Crown
Zellerbach
James
Goldsmith
193
n.a.
n.a.
15.
Easco
Equity
Group
27
n.a.
n.a.
16.
Frigitronics
Revlon
41
-27
14
17.
Frontier
Holdings
People
Expressa
117
13
130
18.
Fruehaufb
Fruehauf
Holdinga
414
n.a.
n.a.
19.
Gillette
Revlonc
1,483
.
.
.
...
20.
Great
Lakes
International
Itel
14
9
23
21.
Gulton
Mark
IV
Industries
36
-6
30
22.
Hammermill
Paper
International
Papera
418
11
429
23.
Hook
Drugs
Krogera
74
-26
48
24.
Houston
Natural
Gas
Coastalc
961
...
25.
Imperial
Continental
Gas
Gulf
Resourcesc
n.a.
n.a.
n.a.
26.
Informatics
General
Sterling
Software
28
-4
24
27.
John
Blair
Reliance
Capitala
165
n.a.
n.a.
28.
Jonathan
Logan
UMM
16
6
22
29.
Joy
Technologiesb
Joy
Manufacturinga
229
n.
a.
n.
a.
30.
Masland
Burlington
Industries
42
-79
-37
31.
Mayflower
Laidlawc
63
n.a.
n.a.
32.
McGraw-Edison
Cooper
Industries
410
-41
369
33.
Medford
Amalgamated
Sugar
49
n.a.
n.a.
34.
MidCon
Occidental
Petroleuma
1,406
-365
1,041
35.
National
Gypsumb
Management
Groupa
325
n.a.
n.a.
36.
NL
Industries
Harold
Simmons
-77
n.a.
n.a.
37.
Owens
Coming
Fiberglass
Wickesc
528
.
.
.
.
.
38.
Pacific
Lumber
Maxxam
256
1
257
39.
Phillips
Petroleum
Carl
Icahnc
1,219
.
.
.
...
40.
Ponderosa
Edelman
68
n.a.
n.a.
41.
Prentice
Hall
Gulf
and
Western
254
-59
195
42.
Quotron
Citicorp
137
109
246
43.
Revlon
Inc.
Pantry
Pride
256
n.a.
n.a.
44.
Richardson-Vicks
Procter
and
Gamblea
776
-
165
611
45.
Ryan
Homes
HV
Homes
64
n.a.
n.a.
Table
2.
(continued)
Change
in
Target
Bidder
Takeover
premium
bidder
wealth
Total
46.
Safewayb
KKRa
1,512
n.a.
n.a.
47.
Saga
Marriott
148
-
162
-
14
48.
Sanders
Associates
Lockheeda
614
119
732
49.
SCM
Hanson
Trust
386
n.a.
n.a.
50.
Scovill
First
City
Properties
212
n.a.
n.a.
51.
Southland
Royalty
Burlington
Northern
126
-97
29
52.
Sperry
Burroughs
1,223
n.a.
n.a.
53.
Strawbridge
&
Clothier
Berryc
55
n.a.
n.a.
54.
Tull
Industries
Bethlehem
Steela
26
52
78
55.
Unidynamics
Cranea
99
-
15
84
56.
Union
Carbide
GAFC
1,344
.
..
...
57.
Uniroyalb
Clayton-Dubiliera
234
n.a.
n.a.
58.
Unocal
Mesa
Partnersc
2,206
.
.
.
...
59.
U.S.
Industries
Hanson
Trust
196
n.a.
n.a.
60.
Van
Dusen
Air
APL
Partnership
14
1
15
61.
Westchester
Financial
Service
Marine
Midland
Banka
35
8
43
62.
White
Consolidated
Electrolux
164
n.
a.
n.a.
Source:
See
table
1.
a.
White
Knight.
b.
LBO.
c.
Unsuccessful.
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 19
as Occidental's buying Midcon and Marriott's buying Saga. On the
other hand, other strategic acquisitions, such as Citicorp's buying Quo-
tron and Coastal's buying American Natural Resources, result in in-
creases in the wealth of bidding shareholders. We cannot conclude from
this sample that related acquisitions are systematically good or bad for
the bidders, although earlier work by Morck, Shleifer, and Vishny
suggests that in this period related acquisitions are better for the bidders
than unrelated ones.27
The last column of table 2 presents the combined wealth change of
the bidder and the target for the available observations. The change is
positive in all but three cases, each of which is a strategic acquisition
by a firm extending its product line. In general, most of the gains go
to the target, and the bidder wealth change is relatively small, just as
the other studies find. It is not the case in this sample that target gains
can often be explained as bidder losses.
Layoffs and the Takeover Premium
One of the most direct ways for the acquirer to justify the takeover
premium is to lay off employees and save on labor costs. The oppor-
tunities for layoffs are considerable: consolidation of headquarters, white-
collar employment cuts due to selloffs, closing of plants, consolidation
of production, and many others. Because labor costs are so high relative
to profits, the effect of such savings on the market value can be sub-
stantial.
Our measure of layoffs is the sum of layoffs and early retirements
from the retained divisions that can be attributed to the target company,
as reported by any of our sources. We use documented evidence of
early retirements in only two cases (Gillette and Owens Corning); the
vast majority of observations are layoffs. All types of layoffs are grouped
together-those from plant closings, staff reductions, consolidations,
and so forth. However, in most cases we can distinguish between white-
and blue-collar layoffs. Most of the information on layoffs comes from
the Wall Street Journal, although other sources are sometimes used. In
two cases, where we saw reports of layoffs but did not have the numbers,
27. Morck, Shleifer, and Vishny (1990).
20 Brookings Papers: Microeconomics
1990
we called the bidding company and got the numbers from them. We
include all post-takeover layoffs for three calendar years, starting with
the year of the takeover.
For several reasons our post-takeover layoff numbers are biased
downward. First, we do not include the layoffs from the bidding firm,
although one could argue that these layoffs helped to pay the premium
and hence can be included in the calculation of cost savings. Second,
we look at actual layoffs and not at employment reductions due to
selloffs. Nor do we follow sold-off divisions to check what happened
to employment there. If the cuts occurred after the selloffs, we miss
them. For these reasons, we do not use Compustat data on employment
changes because bidder layoffs and especially selloffs are often the
causes of reduced Compustat employment. Third, if no publication
reported a layoff, we assume that none has occurred. Wherever we have
data on both employment and layoffs, our layoff number is much smaller
than the decline in measured employment, which reflects selloffs. Hav-
ing mentioned these sources of downward bias in the layoff numbers,
we should also mention that in some cases our numbers probably over-
estimate the number of layoffs precipitated by the takeover because the
layoffs would have occurred anyway. More on this later.
To compute the value of labor cost savings, we assume that the after-
tax cost of a blue-collar worker is $20,000 a year, that of a white-collar
worker $50,000 a year. Labor costs of course include all benefits (in-
cluding social security), which run at 30 percent or more of wages.
These numbers imply an annual pretax labor cost of $30,000 to $40,000
for blue-collar workers and $70,000 to $ 100,000 for white-collar work-
ers, depending on whether the tax rate is 34 percent or 46 percent.
The U. S. Department of Labor's publication, Employment and Earn-
ings (January 1987), reports December 1986 weekly earnings for pro-
duction workers in durable
goods manufacturing
of $443, or approximately
$23,000 on an annual basis. A study by Felicia Nathan published in
the Monthly Labor Review (October 1987) reports that wages are ap-
proximately 70 percent of total compensation in U.S. manufacturing.
Using these figures, we get an average blue-collar pretax employment
cost in durable manufacturing in 1986 of $32,600. It is a little more
difficult to get a handle on the employment cost for the average man-
agerial worker laid off. Nathan's study reports that total compensation
per hour for executive, administrative, and managerial workers is ap-
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 21
proximately 1.8 times that for a typical blue-collar worker. We have
assumed an employment cost for managerial workers 2.5 times that for
blue-collar workers, adjusting Nathan's figure upward because the man-
agerial layoffs in our sample are concentrated at corporate headquarters.
In the few cases where we have both the number of laid-off workers
and the annual cost saving reported by the firm, plugging in our esti-
mates of labor savings gave us answers very close to the savings reported
by the firms.
Of course, labor cost savings should be interpreted as the difference
between the cost of hiring these workers and their marginal product.
The laid-off workers might be producing only slightly less than they
are paid, in which case the annual savings are much smaller. Unfor-
tunately, we have no estimates of the marginal product and so cannot
estimate the true savings. It is easier to believe that the marginal product
of laid-off workers is lower for white-collar than for blue-collar work-
ers, since white-collar functions of the target can be more easily carried
out by the bidding firm without extra costs. We propose our calculation
as a useful benchmark that should be adjusted on the basis of one's
views about output loss.
The second question is how to compute the present value of labor
cost savings. We assume that the real risk-adjusted discount rate ap-
plicable to this calculation is 10 percent, consistent with a real riskless
rate of 4 percent and a risk premium of 6 percent. We use the real rate
because in principle wages rise with inflation. We make the calculation
assuming that the labor cost saving will last five years, after which
perhaps the people would have been laid off or retired without replace-
ment even without a control change. We also make the calculation
assuming that the labor cost saving is permanent. To simplify the com-
putations, we calculate the savings from layoffs on the assumption that
all the layoffs are done right after the takeover. For both calculations
of the present value, we also compute the ratio of the present value of
labor cost savings to the takeover premium. The computation of the
premium is described in the previous section. This gives us estimates
of the fraction of the takeover premium that can be explained by labor
cost savings.
The riskiness of savings created by layoffs is hard to determine. We
would argue that the savings are risky in the sense that they might have
been realized anyway if the firm did poorly but not if the firm did well.
22 Brookings Papers:
Microeconomics
1990
On the other hand, if the savings are closer to risk-free, perhaps a 5
percent discount rate might be appropriate. Use of a 5 percent discount
rate would increase the five-year saving by only about 14 percent but
would double the savings calculated as a perpetuity.
Table 3 presents the information on the 28 of our 62 firms for which
there is direct evidence of layoffs. Recall that the sample includes
successful hostile takeovers, white knight acquisitions, and unsuccess-
ful takeovers. For the remaining firms, there is no evidence of layoffs.
This is not necessarily to say that no layoffs occurred, but rather that
the sources we have do not mention layoffs. For 7 of our 28 firms, the
sources mentioned that layoffs had occurred but did not specify the
numbers. Imperial Continental Gas was excluded from this analysis
because it is foreign, and very little information on it is available. The
averages are thus computed using 21 firms with layoffs and 33 firms
for which no layoffs are mentioned and hence are assumed to be zero.
For the 21 targets that experienced layoffs, the number laid off ranges
between 120 and 6,148, or between 0.1 percent and 24.3 percent of
the labor force of the target firm. The average layoff among the firms
is 1,262 employees, or 5.7 percent of the firms' work force. These
numbers, of course, are cut by about 61 percent to take account of the
fact that 33 of the 62 firms experienced no layoffs. Even so, layoffs
are not trivial for affected firms: a takeover-successful or not-costs
on average 500 jobs, assuming that these people would not have been
laid off had the takeover not occurred. At the same time, the total job
loss in the economy because of companies directly affected by hostile
takeovers is trivial. In our whole sample over three years, fewer than
30,000 workers were laid off.
Some evidence on the question of how much of the premium these
layoffs can justify is also presented in table 3. Under our assumptions
about wages of white-collar and blue-collar workers, the annual after-
tax labor cost savings range from $6.0 million to $200 million, with
an average of $37.5 million. If these savings last five years, they explain
an average of 27. 1 percent of the premium in cases where layoffs
actually occur, with the range between 2.1 percent and 89.9 percent.
In 9 out of the 21 cases these savings explain over a quarter of the
premium. If these savings are permanent, they can justify 65.5 percent
of the premium on average and more than a quarter of the premium in
15 out of 21 cases. Moreover, in 6 cases the savings can justify more
Table
3.
Employee
Layoffs
Subsequent
to
Takeover
Bid,
Sample
Hostile
Takeover
Target
Firms,
1984-86
Estimated
savings
from
layoffs*
Millions
of
dollars
Layoffs
Present
value
at
Present
value
as
percent
Percent
of
10
percent
of
premium
Firm
Number
work
force
Annually
Five
years
Perpetuity
Five
years
Perpetuity
1.
Allied
Stores
2,625a
4.3
70.5
267.9
705
24.2
63.7
2.
American
Natural
Resources
400b
3.6
12.6
47.9
126
6.3
16.5
3.
AMF
350c
1.9
17.5
66.5
175
89.9
236.5
4.
Anderson
Clayton
200d
1.1
10.0
38.0
100
35.2
92.6
5.
CBS
1,492e
5.0
62.7
238.3
627
8.7
22.8
6.
Chesebrough-Ponds
400f
1.7
8.0
30.4
80
2.1
5.6
7.
Crown
Zellerbach
n.a.
n.a.
n.a.
n.a.
n.a.
n.
a.
8.
Fruehauf
730h
2.8
14.6
55.5
146
13.4
35.3
9.
Gillette
2,400'
8.0
48.0
182.4
480
12.3
32.4
10.
Houston
Natural
Gas
n.a.i
na.
n.a.
n.a.
n.a.
n.a.
n.a.
11.
Informatics
General
120k
4.6
6.0
22.8
60
81.4
214.3
12.
Mayflower
Group
125
5.0
6.3
23.8
63
37.8
100.0
13.
McGraw-Edison
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
14.
MidCon
2,000-
23.9
40.0
152.0
400
10.8
28.5
15.
NL
Industries
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
16.
Owens
Coming
Fiberglass
8530
3.0
42.65
162.1
426.5
30.7
80.8
17.
Phillips
Petroleum
6,148P
24.3
123.0
467.4
1,230
38.3
100.9
18.
Ponderosa
120
0.5
6.0
22.8
60
33.5
88.2
19.
Revlon
n.a.
n.a.
n.a.
na.
na.
n.a.
n.a.
20.
Safeway
300r
0.1
15.0
57.0
150
3.8
9.9
Table
3.
(continued)
Estimated
savings
from
layoffs*
Layoffs
Millions
of
dollars
Present
value
as
percent
Present
value
at
of
premium
Percent
of
10
percent
Firm
Number
work
force
Annually
Five
years
Perpetuity
Five
years
Perpetuity
21.
Sanders
Associates
165s
2.1
8.3
31.4
83
5.1
13.5
22.
Scovill
n.a.t
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
23.
Southland
Royalty
n.a.u
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
24.
Sperry
2,370v
3.0
47.4
180.1
474
14.7
38.8
25.
Union
Carbide
4,000w
15.0
200.0
760.0
2,000
56.5
148.8
26.
Uniroyal
700X
3.5
29.0
110.2
290
47.1
124.0
27.
Unocal
600Y
3.0
12.0
45.6
120
2.1
5.4
28.
U.S.
Industries
400Z
3.1
8.0
30.4
80
15.5
40.8
Average
1,261.8
5.7
37.5
142.4
375
27.1
71.3
Source:
See
table
1.
n.a.
Not
available.
a.
Six
hundred
employees
from
Allied's
headquarters;
Campeau
cut
2,025
of
Allied's
(blue-collar)
workers.
b.
ANR
had
a
takeover
agreement
with
its
acquirer.
Coastal
Corp..
to
keep
ANR's
headquarters
in
Detroit
and
honor
ANR's
employment
contracts.
Four
hundred
jobs
(including
153
management
positions)
were
eliminated
from
Coastal's
subsidiary,
ANR
Freight.
in
1987.
To
estimate
savings
from
layoffs.
we
assumed
that
153
white-collar
and
247
blue-collar
employees
were
laid
off.
c.
Three
hundred
and
fifty
of
AMF's
400-member
senior
management
and
corporate
staff
were
fired
in
August
1985.
d.
Quaker
Oats'
PR
Department
notes
that
200
employees
from
Anderson
Clayton's
headquarters
were
laid
off.
e.
Seventy-four
TV
news
staff
(9/85),
8
economic
analysis
staff
(9/85),
40
record
group
staff
(11/85).
700
broadcast
group
employees
(7/86).
300
from
a
New
Jersey
plant
(7/86).
30
corporate
staff
(9/86),
14
publishing
group
staff
(10/86),
70
secretarial
pool
and
medical
staff
(10/86).
26
pages
(10/86).
230
news
staff
(3/87).
To
estimate
savings
from
layoffs,
we
assumed
that
30
white-collar
and
1,462
blue-collar
employees
were
laid
off.
f.
Plant
closed-400
workers
affected
(10/87).
g.
Closed
two
plants.
discontinued
other
operations.
and
consolidated
technical
services:
exact
number
of
reduction
in
personnel
not
known.
h.
These
730
employees
are
presumed
to
be
blue-collar
workers.
Also,
reduction
of
administrative
personnel
and
closure
of
two
plants.
i.
Gillette's
board
approved
a
restructuring
plan
that
would
reduce
Gillette's
worldwide
work
force
by
2,400
(8
percent)
through
retirements,
attrition,
and
firings.
j.
After
the
HNG/Internorth
merger,
work
force
was
reduced
by
1,670
(19
percent).
Unclear
how
many
are
ex-HNG
employees.
k.
Acquirer
(Sterling
Software)
cuts
corporate
staff.
1.
Seven
McGraw-Edison
plants
closed.
A
McGraw-Edison
unit's
headquarters
shut
down.
Exact
numbers
of
employees
are
not
known.
m.
We
do
not
know
if
the
2,000
laid-off
workers
are
blue-collar
or
white-collar
employees.
In
computing
the
estimated
annual
savings
from
layoffs.
we
assumed
that
all
these
2.000
employees
are
blue-collar
workers.
n.
Number
of
NL
employees
in
the
United
States
decreased
from
3,200
(12/80)
to
820
(12/87).
Although
there
are
no
selloffs.
it
is
unclear
how
much
of
this
decrease
can
be
attributed
to
layoffs.
o.
OCF
laid
off
480
of
its
970
research
employees.
Another
373
white-collar
employees
took
early
retirement.
p.
Layoffs:
2,378
(4/86).
2,180
(12/87),
1,590
(4/88).
To
estimate
savings
from
layoffs,
we
assumed
that
these
were
all
blue-collar
employees.
q.
Layers
of
management
and
legal
staff
eliminated.
Exact
number
not
known.
r.
These
are
25
percent
of
the
headquarters
staff
(8/86).
Also,
union
leaders
at
Safeway
are
concerned
about
the
effect
on
their
members
of
Safeway's
spinoffs.
s.
To
estimate
savings
from
layoffs,
we
assumed
that
these
are
white-collar
employees.
t.
Scovill
cuts
corporate
staff
by
30
percent
and
divisional
staff
by
10
percent
(12/85).
Exact
number
is
not
known.
u.
Acquirer
(Burlington
Northern)
reduces
its
natural
gas
pipeline
work
force
by
30
percent
(8/86).
Unclear
how
many
of
these
are
ex-Southland
Royalty
employees.
v.
One
thousand
five
hundred
and
seventy
Sperry
workers
were
fired
from
its
Bristol,
Tennessee,
facility
and
800
from
its
Eagan.
Minnesota,
facility.
Two
other
Sperry
plants
(in
Voorhes.
New
Jersey.
and
Jackson,
Minnesota)
were
closed;
however,
we
do
not
have
specific
information
on
layoffs
from
these
two
plants.
w.
These
4,000
workers
are
white-collar
employees.
x.
To
estimate
savings
from
layoffs,
we
assumed
that
500
white-collar
and
200
blue-collar
employees
were
laid
off.
y.
To
estimate
savings
from
layoffs,
we
assumed
that
these
are
blue-collar
employees.
z.
To
estimate
savings
from
layoffs,
we
assumed
that
these
are
blue-collar
employees.
*To
estimate
savings
from
layoffs.
we
assumed
that
laid-off
blue-collar
(white-collar)
employees
were
earning
quasi
rents
worth
$20,000
($50,000).
26 Brookings Papers: Microeconomics 1990
than the whole premium. Although the means should be cut by about
61 percent for the sample of all takeovers because in 33 cases there is
no evidence of layoffs, the five-year labor cost savings can explain
about 11 percent of the premium in an average takeover and permanent
labor cost savings can explain perhaps 26 percent. With a 5 percent
rather than a 10 percent discount rate, these numbers would be 12.5
percent and 52 percent, respectively. On the other hand, if the marginal
product of laid-off workers is one-half of their wage rather than zero,
the savings should be cut in half. Labor cost savings are thus only a
moderate source of takeover gains.
In table 3 white- and blue-collar workers are grouped together. They
are separated in tables 4 and 5. For 13 companies out of 21 we can
identify white-collar layoffs, and for 12 out of 21 we can identify blue-
collar layoffs. In cases where we could identify them, blue-collar layoffs
average 1,493 workers, or 6.5 percent of the firm's total work force,
and save an average of 11. 1 percent of the premium using a five-year
horizon and 29.2 percent if the gains are permanent. Among the 12
firms for which we have numbers, white-collar layoffs average 660
employees, or 3.2 percent of the firm's total work force, which can
justify 33.6 percent of the premium on average using a five-year horizon
and 88.3 percent using a perpetuity. Since white-collar wages are as-
sumed to be higher than blue-collar wages, the estimated savings from
white-collar layoffs are higher than those from blue-collar layoffs even
though the layoffs themselves are smaller.
White-collar layoffs are smaller in number and as a percentage of
the total labor force than blue-collar layoffs, but of course the white-
collar labor force is on average much smaller than the blue-collar labor
force of a firm. Virtually all laid-off white-collar workers in our sample
are managerial and professional, a group that constitutes 25 percent of
U.S. manufacturing employment, compared with 61 percent for pro-
duction workers and 14 percent for clerical and sales. If we assume
that our sample firms have a similar occupational structure, we can
infer that the probability of layoff for white-collar workers is about 20
percent higher than for blue-collar workers.28 Because we have several
outliers, it might be better to look at median layoff numbers that are
28. That
is, 0.2 = (0.61/0.25)(3.2/6.5) - 1.
Table
4.
White-Collar
Employee
Layoffs
Subsequent
to
Takeover
Bid,
Sample
Hostile
Takeover
Target
Firms,
1984-86
Estimated
savings
from
layoffs
Layoffs
Millions
of
dollars
Percent
of
Percent
of
Present
value
at
10
Present
value
as
total
white-collar
percent
percent
of
premium
Firm
Number
work
force
work
force
Annually
Five
years
Perpetuity
Five
years
Perpetuity
1.
Allied
Stores
600
1.0
60.0
30.0
114.0
300
10.3
27.1
2.
American
Natural
Resources
153
1.4
n.a.
7.7
29.1
77
3.8
10.1
3.
AMF
350
1.9
87.5
17.5
66.5
175
89.9
236.5
4.
Anderson
Clayton
200
1.1
n.a.a
10.0
38.0
100
35.2
92.6
5.
CBS
1,096
3.7
n.a.
54.8
208.2
548
7.6
19.9
6.
Informatics
General
120
4.6
n.a.
6.0
22.8
60
81.4
214.3
7.
Mayflower
Group
125
5.0
n.a.
6.3
23.8
63
37.8
100.0
8.
Owens
Coming
Fiberglass
853
3.0
n.a.
42.65
162.1
427
30.7
80.8
9.
Ponderosa
120
0.5
n.a.
6.0
22.8
60
33.5
88.2
10.
Safeway
300
0.1
25.0
15.0
57.0
150
3.8
9.9
11.
Sanders
Associates
165
2.1
n.a.
8.3
31.4
83
5.1
13.5
12.
Scovill
n.a.
n.a.
n.a.b
n.a.
n.a.
n.a.
n.a.
n.a.
13.
Union
Carbide
4,000
15.0
n.a.
200.0
760.0
2,000
56.5
148.8
14.
Uniroyal
500
2.5
n.a.c
25.0
95.0
250
40.6
106.8
Average
660
3.2
57.5
31.3
125.4
330
33.6
88.3
Source:
See
table
1.
n.a.
Not
available.
a.
CBS's
1986
annual
report
notes
that
nearly
35
percent
of
corporate
staff
was
cut.
b.
Scovill
cut
corporate
staff
by
30
percent.
c.
"Company
says
that
of
20,000
employees
before
the
LBO
19,300
will
still
be
working
.
.
.
although
a
headquarters
staff
in
Middlebury,
Connecticut,
that
totaled
500
people
a
year
ago
will
just
about
disappear."
Table
5.
Blue-Collar
Employee
Layoffs
Subsequent
to
Takeover
Bid,
Sample
Hostile
Takeover
Target
Firms,
1984-86
Estimated
savings
from
layoffs
Layoffs
Millions
of
dollars
Present
value
at
10
Present
value
as
Percent
of
percent
percent
of
premium
to
ta
l_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
Firm
Number
work
force
Annually
Five
years
Perpetuity
Five
years
Perpetuity
1.
Allied
Stores
2,025
3.3
40.5
153.9
405
13.9
36.6
2.
American
Natural
Resources
247
2.2
4.9
18.8
49
2.5
6.4
3.
CBS
396
1.3
7.9
30.1
79
1.1
2.9
4.
Chesebrough-Ponds
400
1.7
8.0
30.4
80
2.1
5.6
5.
Fruehauf
730
2.8
14.6
55.5
146
13.4
35.3
6.
Gillette
2,400
8.0
48.0
182.4
480
12.3
32.4
7.
MidCon
2,000
23.9
40.0
152.0
400
10.8
28.5
8.
Phillips
Petroleum
6,148
24.3
123.0
467.4
1,230
38.3
100.9
9.
Sperry
2,370
3.0
7.4
180.1
474
14.7
38.8
10.
Uniroyal
200
1.0
4.0
15.2
40
6.5
17.1
11.
Unocal
600
3.0
12.0
45.6
120
2.1
5.4
12.
U.S.
Industries
400
3.1
8.0
30.4
80
15.5
40.8
Average
1,493
6.5
26.5
113.5
299
11.1
29.2
Source:
See
table
1.
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 29
2. 1 percent of the labor force for white-collar workers and 3 percent
for blue-collar workers. These medians imply that the probability of a
layoff is 70 percent higher for white-collar workers.29 In fact, since
most of the laid-off workers are corporate staff, and these workers are
an even smaller fraction of the labor force, the odds of a layoff of a
corporate staff member are considerably higher than for any other class
of worker.
The most common reason for white-collar layoffs is consolidation
of headquarters after a takeover. In some cases, however, such as those
of CBS and Union Carbide, white-collar layoffs are a large source of
cost savings after unsuccessful takeover attempts. Blue-collar layoffs,
in contrast, typically have to do with major retrenchments, as in the
case of Phillips Petroleum. In some cases, such as that of Fruehauf,
plant closings also lead to blue-collar layoffs.
Table 6 compares layoffs in successful, unsuccessful, and white
knight takeovers. We discuss first the results for a subsample in which
layoffs actually occurred and then the unconditional results for the
whole sample.
Of the 21 target firms for which we could quantify layoffs, 7 were
successfully taken over, 7 were acquired by white knights, and 7 re-
mained independent. Both in terms of raw numbers and as a percentage
of the labor force, firms that remain independent have the highest layoffs
for blue- and white-collar employees combined. The fraction of the
takeover premium that can be explained by savings from layoffs is
highest for targets of successful takeovers-37.9 percent, compared
with 16.8 percent for firms acquired by white knights, and 26.6 percent
for firms that remained independent. The much higher fraction of the
premium accounted for by layoffs after successful takeovers is some
evidence in favor of breach of trust, as argued by Shleifer and Summers.
In the table 6 subsample in which blue- and white-collar layoffs are
separated, we again find that white-collar layoffs are most significant
as a fraction of the labor force in targets that remained independent,
but that savings from such layoffs are largest among targets of successful
takeovers. With blue-collar layoffs, targets that remained independent
again had the largest fraction of the labor force laid off, but the dif-
29. That
is, 0.7 = (0.61/0.25)(2.1/3.0) - 1.
Table
6.
Post-Takeover
Attempt
Layoffs
Classified
by
Outcome
of
Hostile
Takeover
Bida
Item
Successfulb
White
knightc
Independentd
Total
number
of
sample
target
firms
30
20
12
Number
of
target
firms
for
which
we
found
no
public
documentation
of
layoff
information
10
8
4
Number
of
firms
that
experienced
some
employee
layoff
subsequent
to
the
takeover
bid
13
7
8
Firms
whose
employee
layoffs
we
were
able
to
document
exactly
Number
of
firms
7
7
7
Average
number
of
workers
laid
off
912
642
2,231
Average
percent
of
work
force
laid
off
3.0
5.0
9.0
Average
estimated
savings
from
layoffs
annually
(millions
of
dollars)
24.0
17.8
70.7
Present
value
of
estimated
savings
from
layoffs
(millions
of
dollars)
91.2
[240]
67.8
[178]
268.5
[707]
Average
of
present
value
of
estimated
savings
from
layoffs
as
a
percent
of
premium
offered
for
the
target
37.9
[99.81
16.8
[26.7]
26.6
[70.2]
Firms
whose
white-collar
employee
layoffs
we
were
able
to
document
exactly
Number
of
firms
5
4
4
Average
number
of
white-collar
layoffs
268.6
291
1,519
Average
percent
of
total
work
force
laid
off
1.9
1.5
6.7
Average
percent
of
white-collar
work
force
laid
off
73.8e
25.0
n.a.
Average
estimated
savings
from
white-collar
layoffs
annually
(millions
of
dollars)
13.4
14.6
75.9
Present
value
of
estimated
savings
from
white-collar
layoffs
(millions
of
dollars)
51.0
[134.4]
55.4
[146]
289
[759]
Average
of
present
value
of
estimated
savings
from
white-collar
layoffs
as
a
percent
of
premium
offered
for
the
target
43.8
[115.2]
21.2
[55.7]
33.2
[87.4]
Firms
whose
blue-collar
employee
layoffs
we
were
able
to
document
exactly
Number
of
firms
4
4
4
Average
number
of
blue-collar
layoffs
1,261
832
2,386
Average
percent
of
total
work
force
laid
off
2.9
7.4
9.2
Average
estimated
savings
from
blue-collar
layoffs
annually
(millions
of
dollars)
15.2
16.7
47.7
Present
value
of
estimated
savings
from
blue-collar
layoffs
(millions
of
dollars)
95.8
[252]
63.3
[167]
181
[477]
Average
of
present
value
of
estimated
savings
from
blue-collar
layoffs
as
a
percent
of
premium
offered
for
the
target
11.6
[30.6]
8.2
[21.6]
13.4
[35.4]
Source:
See
table
1.
n.a.
Not
available.
a.
The
sample
target
firms
constitute
an
exhaustive
sample
of
targets
of
hostile
takeover
bids
of
U.S.
firms
during
1984-86
where
the
value
of
the
initial
offer
was
more
than
$50
million.
Calculations
are
for
five-year
savings
from
layoffs;
calculations
in
brackets
are
for
perpetual
savings.
b.
The
bidding
firm
is
successful
in
acquiring
the
target.
c.
The
target
is
acquired
by
a
white
knight.
d.
The
target
remains
independent.
e.
Based
on
2
observations.
32 Brookings Papers: Microeconomics 1990
ferences in percentage of the premium accounted for by layoffs are no
longer as large. These results provide mixed evidence for the breach
of trust hypothesis. On the one hand, that white knights seem to make
fewer layoffs than successful hostile acquirers, particularly when mea-
sured as a percentage of the premium saved, suggests that white knights
breach fewer contracts than do the hostile acquirers. This is as predicted
by the theory. On the other hand, we find that managers of firms that
remain independent themselves lay off a lot of workers-both white-
and blue-collar. Perhaps targets that remain independent, unlike white
knights, continue to be under pressure, which explains the greater lay-
offs.
One further nice piece of evidence illustrates the difference between
successful acquirers and white knights. In the 30 instances of successful
hostile takeovers, we have evidence of 10 closings or sales of head-
quarters; in the 20 instances of white knight acquisitions, we have no
evidence of closings or sales of headquarters. If the sale of headquarters
reflects breach of trust, this evidence shows clearly that hostile acquirers
breach trust but that white knights do not.
The findings just described are for the subsample in which the layoffs
actually occurred and can be measured. To understand the importance
of layoffs in all takeovers, we must correct for the fact that in more
than half the cases there is no evidence of layoffs. With the correction,
the finding that targets that remain independent laid off most aggres-
sively is strengthened. They laid off an average of 6.3 percent of the
labor force, compared with 1.75 percent for white knights, and 0.9
percent for successful acquirers. The unconditional fraction of takeover
premium explained by savings from layoffs is 18.6 percent for targets
that remained independent, 5.9 percent for white knights, and 11 percent
for successful acquirers. Thus the evidence on white knights as against
successful acquirers is mixed, but the aggressive layoffs by targets that
remained independent indicate that they continue to be under the pres-
sure of takeovers.
An objection to this analysis of layoffs is that we do not compare
post-takeover layoffs with pre-takeover layoffs in the same firms or
with layoffs in other firms in the same industry. If layoffs reflect only
industry trends, then takeovers cannot be responsible for them. But this
argument is flawed. If fear of takeover is responsible for industry lay-
offs, then layoffs might well be the source of gains in acquired firms.
Sanjai Bhagat, Andrei
Shleifer,
and Robert
Vishny 33
Advocates of takeovers often credit takeover pressure for eliminating
inefficient investment throughout the oil industry and not just in ac-
quired firms. They do not give takeover pressure credit for industrywide
employment reductions, but the logic is the same.
High costs of collecting data prevent us from conducting our analysis
on the industries of acquired firms, but we made a few comparisons.
First, we looked at the Wall Street Journal for evidence of layoffs for
our 62 sample firms in the two years before the takeover attempt. The
Wall Street Journal is not the only source of layoff information for our
sample, but it is by far the dominant source. We have found evidence
of layoffs in 8 firms before the takeover period, compared with 28 in
the post-takeover period. Of these, 6 were firms that also had post-
takeover layoffs. The average layoff was 275 employees, or 0.6 percent
of the affected firm' s employment. This translates into the unconditional
average layoff of 35 employees, or less than 0. 1 percent of the labor
force, which is much smaller than the post-takeover unconditional av-
erage. Layoffs after the takeover are clearly greater than layoffs before.
This evidence supports the idea that managers of hostile takeover targets
are reluctant to breach implicit contracts unless thrown out or forced
to do so under takeover pressure.
To examine industry layoffs, we matched each firm in the sample to
a similar-sized firm in the same industry. We then used the Wall Street
Journal to examine layoff practices of control firms over the same period
as the sample firms. We have found that 11 control firms experienced
layoffs during the relevant period, averaging 539 employees, or 5.2
percent of their work forces. The implied unconditional expected layoff
for the whole control sample is 100 employees, or 0.5 percent of the
work force-obviously much less than for the post-takeover sample.
Although these results are biased down by the use of the Wall Street
Journal only, layoffs in control firms measured in this way are very
significantly smaller. The implication is that takeovers do cause layoffs.
In addition to layoffs, hostile takeovers often result in pension plan
terminations accompanied by reversions of excess pension assets to the
acquirer. Pontiff, Shleifer, and Weisbach document the increase in
reversion activity after takeovers, and the greater frequency of such
reversions following hostile takeovers.30 In our sample, there is evi-
30. Pontiff, Shleifer, and Weisbach (1989).
34 Brookings Papers:
Microeconomics
1990
dence that 14 firms reverted pension plans after the takeover. Although
in most cases these reversions are small, in a few, such as those of
Union Carbide, NL Industries, and Jonathan Logan, reversions are a
large fraction of the premium. Even if one takes the estimate of Pontiff,
Shleifer, and Weisbach that only a third of the excess pension assets,
on average, would have gone to the employees were it not for the
reversion, one still gets nontrivial transfers in a few cases. This evidence
confirms that pension reversions are common in hostile takeovers but
rarely justify a large fraction of the premium.
In sum, layoffs after takeovers are common and can explain 10-20
percent of the premium. Moreover, layoffs seem to be much more
common in our sample firms than in other firms in their industries. At
the same time, layoffs are clearly not the whole story behind hostile
takeovers, and it is hard to believe that plans for future layoffs constitute
an important takeover motive. The direct consequences of takeovers
for U.S. employment are trivial. Layoffs are a common by-product of
the hostile takeover process but do not appear to be the driving force
behind it.
The Importance of Selloffs after Takeovers
In this section, we compute the value of divestitures following hostile
takeovers. As with layoffs, we look at all post-takeover divestitures
during three calendar years starting from the year of the takeover. We
followed the same strategy as with layoffs, making sure that the sold-
off assets are from the target. When it proved impossible to attribute
the divestiture to the target, we counted it as zero. This was the biggest
problem when the target and the bidder are in the same industry and
have relatively homogeneous assets, such as gas pipelines. The main
sources of information on divestitures are Moody's and the Wall Street
Journal, although annual reports and lOKs also proved useful. In most
cases we found prices of divestitures; otherwise, the divestiture is not
counted. The restriction that divestitures must be identified with the
target and have a reported price biases our count of divestitures down-
ward.
In computing the total value of divestitures, we did not correct for
market movements and simply added up the realized prices. This is an
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 35
aggressive calculation during a period of rising stock prices, although
it is not clear that a rising stock market always raises the value that a
rational buyer is willing to pay for a division. We were conservative
in taking the debt assumed by a buyer of a division to be zero, thus
reducing the price that might have been effectively paid. Given these
approximations, as well as the fact that some divestitures might have
gone unreported, our estimates should be viewed as rough.
The results of this calculation, presented in table 7, reveal a signif-
icant number of selloffs in most takeovers. The average portion of the
acquisition price realized through selloffs is 29.6 percent and the median
is 16.6 percent. In only 20 cases out of 62 were there no quantifiable
selloffs, and even in some of these cases, such as that of Aegis, there
were selloffs, but we could not find the price. In 17 cases, more than
half the acquisition price was regained through selloffs, and in 3 cases
more than 100 percent was regained. It is absolutely clear from these
data that selling off divisions is one of the most pervasive consequences
of hostile takeovers.
Who Makes Divestitures?
There are a variety of selloff styles. Some cases are classic bustups
in which the bidding firm does not seem to retain much of the target
at all. Examples include Revlon's acquisition of Frigitronics and First
City's acquisition of Scovill. Not surprisingly, such complete bustups
are typically done by the raiders. We do not know whether the purpose
is to take advantage of the underpricing of the target, to sell divisions
to strategic buyers at inflated prices, to enable eventual acquirers to
implement significant changes, or all three. Even in complete bustups,
the takeover artist usually gets some benefits from headquarters layoffs.
Another, and perhaps more interesting, case of nearly complete bust-
ups are acquisitions of conglomerates for the purposes of retaining some
pieces of them. For example, when Quaker Oats bought Anderson
Clayton, it retained the Gaines dog food business that it had previously
sought and sold off everything else. Another example is James Gold-
smith, which bought Crown Zellerback and sold off everything but the
timber properties. A third example is Minstar, which bought AMF and
Aegis and sold off most assets other than the boat divisions that it
wanted. In these examples, the acquirer got the segment of the company
Table
7.
Selloffs
as
a
Fraction
of
the
Acquisition
Price
Millions
of
dollars
except
as
noted
Ratio
of
Price
selloffs
Target
Bidder
(debt
+
equity)
Selloffs
to
price
1.
Aegis
Minstar
79.00
0
. .
.
2.
American
Motor
Inn
Prime
Motor
Inn
305.00
383.70
1.258
3.
Allied
Stores
Campeau
4,364.00
2,500.00
0.573
4.
American
Natural
Resources
Coastal
3,525.00
0
...
5.
AMF
Minstar
715.00
511.00
0.715
6.
Anderson
Clayton
Quaker
Oatsa
824.00
535.00
0.649
7.
Atlas
Van
Linesb
Wesraya
76.60
0
...
8.
Avondale
Mills
Walton
Monroe
Millsa
150.60
18.00
0.120
9.
Carter
Hawley
Hale
The
Limitedc
1,621.00
333.00
0.205
10.
Carter
Hawley
Hale
The
Limitedc
2,650.00
213.00
0.080
11.
CBS
Turner
Broadcastingc
5,781.00
3,440.00
0.595
12.
Chesebrough-Ponds
Unilevera
4,270.00
1,800.00
0.422
13.
Cluett
Peabody
Westpoint
Pepperella
499.00
0
.
.
.
14.
Crown
Zellerbach
James
Goldsmith
1,848.00
1,146.00
0.620
15.
Easco
Equity
Group
265.00
13.00
0.049
16.
Frigitronics
Revlon
123.00
120.00
0.976
17.
Frontier
Holdings
People
Expressa
342.00
0
.
..
18.
Fruehaufb
Fruehauf
Holdinga
1,580.00
732.00
0.463
19.
Gillette
Revlonc
4,556.00
50.00
0.011
20.
Great
Lakes
International
Itel
200.00
31.00
0.155
21.
Gulton
Mark
IV
Industries
122.00
0
22.
Hammermill
Paper
International
Papera
1,410.00
0
23.
Hook
Drugs
Krogera
174.00
0
24.
Houston
Natural
Gas
Coastalc
3,080.00
544.00
0.177
25.
Imperial
Continental
Gas
Gulf
Resourcesc
1,205.00
717.00
0.595
26.
Informatics
General
Sterling
Software
127.00
34.00
0.268
27.
John
Blair
Reliance
Capitala
596.00
377.00
0.633
28.
Jonathan
Logan
UMM
214.00
17.00
0.079
29.
Joy
Technologiesb
Joy
Manufacturinga
716.00
252.00
0.352
30.
Masland
Burlington
Industries
136.50
0
...
31.
Mayflower
Laidlawc
320.00
25.00
0.078
32.
McGraw-Edison
Cooper
Industries
1,400.00
177.00
0.126
33.
Medford
Amalgamated
Sugar
145.00
38.00
0.262
34.
MidCon
Occidental
Petroleuma
3,900.00
1,470.00
0.377
35.
National
Gypsumb
Management
Groupa
1,720.00
441.00
0.256
36.
NL
Industries
Harold
Simmons
900.00
0
...
37.
Owens
Coming
Fiberglass
Wickesc
2,640.00
595.00
0.225
38.
Pacific
Lumber
Maxxam
916.00
351.00
0.383
39.
Phillips
Petroleum
Carl
Icahnc
11,340.00
2,000.00
0.176
40.
Ponderosa
Edelman
328.00
17.00
0.052
41.
Prentice
Hall
Gulf
and
Western
708.00
0
.
..
42.
Quotron
Citicorp
687.00
0
...
43.
Revlon
Pantry
Pride
2,285.00
2,060.00
0.902
44.
Richardson-Vicks
Proctor
and
Gamblea
1,830.00
106.00
0.058
45.
Ryan
Homes
HV
Homes
348.00
0
Table
7.
(continued)
Ratio
of
Price
selloffs
Target
Bidder
(debt
+
equity)
Selloffs
to
price
46.
Safewayb
KKRa
5,300.00
3,200.00
0.604
47.
Saga
Marriott
607.00
306.00
0.504
48.
Sanders
Associates
Lockheeda
1,199.00
0
...
49.
SCM
Hanson
Trust
1,188.50
920.00
0.774
50.
Scovill
First
City
Properties
627.00
680.70
1.086
51.
Southland
Royalty
Burlington
Northern
1,034.00
0
.
.
.
52.
Sperry
Burroughs
5,470.00
1,800.00
0.329
53.
Strawbridge
&
Clothier
Berryc
480.00
0
...
54.
Tull
Industries
Bethlehem
Steela
106.00
0
.
.
.
55.
Unidynamics
Cranea
251.00
0
.
.
.
56.
Union
Carbideb
GAFC
8,000.00
4,500.00
0.563
57.
Uniroyalb
Clayton-Dubiliera
1,001.00
935.00
0.934
58.
Unocal
Mesa
Partnersc
10,500.00
40.00
0.004
59.
U.S.
Industries
Hanson
Trust
572.00
178.00
0.311
60.
Van
Dusen
Air
APL
Partnership
77.00
101.50
1.318
61.
Westchester
Financial
Service
Marine
Midland
Banka
102.00
0
...
62.
White
Consolidated
Electrolux
923.00
0
.
.
.
Average
values
1,735.70
539.34
0.296
Source:
See
table
1.
a.
White
Knight.
b.
LBO.
c.
Unsuccessful.
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 39
it wanted for an extremely low price and managed to get most of its
money back through divestitures. These near-bustups suggest that even
apparent bustup takeovers might be driven by the strategic objectives
of the acquirers, with selloffs being an incidental consequence of the
pursuit of particular businesses.
In about a third of the cases there are no selloffs at all, and in a few
selloffs are minor. Examples of such low-selloff takeovers include Coastal
acquiring American Natural Resources (both are pipelines), Walton
Monroe Mills acquiring Avondale Mills (both are textile firms), West-
point Pepperell acquiring Cluett Peabody (both are textile firms), Gulf
and Western acquiring Prentice Hall (both are in publishing), and Ci-
ticorp acquiring Quotron (to pursue the alleged complementarity of
banking and provision of stock quotations). As we argue later in the
paper, such strategic acquisitions of whole or parts of companies are
an extremely important part of the hostile takeover process. Without
these wholly strategic bids, the average fraction of assets sold off after
hostile takeovers would be even larger.
Several insights about the reasons for selloffs can be gained by
looking at different types of transactions. When we compare LBOs with
non-LBOs, we find that the mean portion of the acquisition price rea-
lized through selloffs in the two cases are 43.5 percent and 28. 1 percent,
respectively, and the medians are 40.7 percent and 12.3 percent. The
incidence of selloffs is higher after LBOs, especially judging by the
medians, which suggests that the pressure from debt is a reason for
some selloffs. When we compare successful and unsuccessful takeo-
vers, we do not find much difference in the fraction of the acquisition
price recovered through selloffs. This finding suggests, as did our results
on layoffs, that firms escaping the takeover often do most of the things
that the acquirer would have done anyway. Finally, we do find fewer
selloffs by white knights than by hostile acquirers. In the means, the
portion is 24.3 percent for the former and 36.5 percent for the latter.
In the medians, the ratio is 8.9 percent for white knights and 20.9
percent for hostile acquirers. The picture is similar to that with layoffs
and headquarters closures: white knights do less than the hostile ac-
quirers, but firms remaining independent have to make more changes
than those acquired by a white knight. The explanation that targets
remaining independent continue to be under pressure applies here as
well.
40 Brookings
Papers: Microeconomics
1990
Who Are the Buyers in Divestitures?
Perhaps the most interesting question for understanding the function
of divestitures is who the buyers are. If the function of hostile takeovers
is to create organizations with appropriate incentives, such as high
management ownership and high debt, then one would expect a lot of
divestitures to MBO teams or to investment companies. If the function
of hostile takeovers is to allocate businesses to strategic buyers, then
one would expect most divestitures to be to strategic buyers. To examine
these issues, we divide divestitures into those to strategic buyers (those
in a business related to that of the acquired division), those to MBOs
and investment companies, and those to unrelated acquirers. We also
have two residual categories. The first is a small category consisting
of headquarters buildings, stocks, and other liquid assets that are sold
off. The second is the selloffs for which we cannot identify the buyer.
Table 8, which presents the results for each firm, shows that most
selloffs go to acquirers in the same industry. The total volume of selloffs
in this sample is $33.7 billion. Of those, $23.70 billion, or 70 percent,
were selloffs of divisions to firms with lines of business the same as
or closely related to those of the divisions. Compared with that, $5.4
billion, or 16 percent, were sold off to MBOs or investment companies,
and $2.8 billion, or 8 percent, to unrelated acquirers. We could not
identify the buyers in 4 percent of the sold-off assets, and 2 percent
were headquarters and marketable assets selloffs.
Selloffs, clearly, go mainly to related acquirers. Such selloffs are
four times more important than those to MBOs and other incentive-
intensive arrangements. Even if we assume that all of the 4 percent of
asset sales for which we could not identify the buyer are to an MBO
or an investment company, we still get the result that selloffs to related
buyers are three and one-half times more important.
These results present a different view of divestitures than that sug-
gested by Jensen, who views the large incidence of divestitures as
evidence that mergers do not increase concentration.31 Divestitures in
fact seem to increase concentration (using industry-level measures em-
ployed by antitrust authorities), because the buyers of divisions are
typically not the raiders or MBO boutiques but large firms in the same
31. Jensen (1988).
Table
8.
Selloffs,
by
Type
of
Acquirer
Millions
of
dollars
Total
Headquarters,
Not
Target
Bidder
selloffs
Related
Unrelated
MBOs
securities
identified
1.
Aegis
Minstar
0
0
0
0
0
0
2.
American
Motor
Inn
Prime
Motor
Inn
384
66
0
318
0
0
3.
Allied
Stores
Campeau
2,500
2,425
0
75
0
0
4.
American
Natural
Resources
Coastal
0
0
0
0
0
0
5.
AMF
Minstar
511
100
0
411
0
0
6.
Anderson
Clayton
Quaker
Oatsa
535
235
0
100
0
200
7.
Atlas
Van
Linesb
Wesraya
0
0
0
0
0
0
8.
Avondale
Mills
Walton
Monroe
Millsa
18
0
0
0
18
0
9.
Carter
Hawley
Hale
The
Limitedc
333
333
0
0
0
0
10.
Carter
Hawley
Hale
The
Limitedc
213
213
0
0
0
0
11.
CBS
Turner
Broadcastingc
3,440
3,440
0
0
0
0
12.
Chesebrough-Ponds
Unilevera
1,800
1,800
0
0
0
0
13.
Cluett
Peabody
Westpoint
Pepperella
0
0
0
0
0
0
14.
Crown
Zellerbach
James
Goldsmith
1,146
800
0
246
100
0
15.
Easco
Equity
Group
13
13
0
0
0
0
16.
Frigitronics
Revlon
120
100
0
20
0
0
17.
Frontier
Holdings
People
Expressa
0
0
0
0
0
0
18.
Fruehaufb
Fruehauf
Holdinga
732
627
0
0
0
105
19.
Gillette
Revlonc
50
0
0
0
0
50
20.
Great
Lakes
International
Itel
31
20
0
0
11
0
21.
Gulton
Mark
IV
Industries
0
0
0
0
0
0
22.
Hammermill
Paper
International
Papera
0
0
0
0
0
0
23.
Hook
Drugs
Krogera
0
0
0
0
0
0
24.
Houston
Natural
Gas
Coastalc
544
489
0
55
0
0
25.
Imperial
Continental
Gas
Gulf
Resourcesc
448
448
0
0
0
0
Table
8.
(continued)
Total
Headquarters,
Not
Target
Bidder
selloffs
Related
Unrelated
MBOs
securities
identified
26.
Informatics
General
Sterling
Software
34
20
0
0
0
14
27.
John
Blair
Reliance
Capitala
377
241
0
136
0
0
28.
Jonathan
Logan
UMM
17
0
0
0
17
0
29.
Joy
Technologiesb
Joy
Manufacturinga
252
252
0
0
0
0
30.
Masland
Burlington
Industries
0
0
0
0
0
0
31.
Mayflower
Group
Laidlawc
25
0
0
20
0
5
32.
McGraw-Edison
Cooper
Industries
177
105
0
45
0
27
33.
Medford
Amalgamated
Sugar
38
0
0
0
0
38
34.
MidCon
Occidental
Petroleuma
1,470
1,470
0
0
0
0
35.
National
Gypsumb
Management
Groupa
441
348
0
27
0
66
36.
NL
Industries
Harold
Simmons
0
0
0
0
0
0
37.
Owens
Corning
Fiberglass
Wickesc
595
240
0
267
0
88
38.
Pacific
Lumber
Maxxam
351
0
0
320
31
0
39.
Phillips
Petroleum
Carl
Icahnc
2,000
1,890
0
0
110
0
40.
Ponderosa
Edelman
17
17
0
0
0
0
41.
Prentice
Hall
Gulf
and
Western
0
0
0
0
0
0
42.
Quotron
Citicorp
0
0
0
0
0
0
43.
Revlon
Pantry
Pride
2,060
1,185
0
875
0
0
44.
Richardson-Vicks
Procter
and
Gamblea
106
0
66
40
0
0
45.
Ryan
Homes
HV
Homes
0
0
0
0
0
0
46.
Safewayb
KKRa
3,200
3,035
0
165
0
0
47.
Saga
Marriott
306
0
0
306
0
0
48.
Sanders
Associates
Lockheeda
0
0
0
0
0
0
49.
SCM
Hanson
Trust
920
884
0
0
0
36
50.
Scovill
First
City
Properties
689
229
460
0
0
0
51.
Southland
Royalty
Burlington
Northern
0
0
0
0
0
0
52.
Sperry
Burroughs
1,800
1,200
0
600
0
0
53.
Strawbridge
&
Clothier
Berryc
0
0
0
0
0
0
54.
Tull
Industries
Bethlehem
Steela
0
0
0
0
0
0
55.
Unidynamics
Cranea
0
0
0
0
0
0
56.
Union
Carbide
GAFC
4,500
890
1,420
1,260
340
590
57.
Uniroyalb
Clayton-Dubiliera
935
100
835
0
0
0
58.
Unocal
Mesa
Partnersc
40
0
0
0
40
0
59.
U.S.
Industries
Hanson
Trust
178
64
0
114
0
0
60.
Van
Dusen
Air
APL
Partnership
102
102
0
0
0
0
61.
Westchester
Financial
Services
Marine
Midland
Banka
0
0
0
0
0
0
62.
White
Consolidated
Electrolux
0
0
0
0
0
0
Total
33,717
23,650
2,781
5,400
667
1,219
Source:
See
table
1.
a.
White
Knight.
b.
LBO.
c.
Unsuccessful.
44 Brookings
Papers:
Microeconomics 1990
lines of business. In fact, the role of the raiders and MBO boutiques
seems to be largely to take diversified firms, bust them up, and sell the
divisions to other firms in the same business. Management buyout or-
ganizers and raiders thus serve as brokers working for the ultimate
purpose of increased concentration. Their goal is to create a temporary
organization that facilitates the allocation of assets to related buyers.
Of course, debt and concentrated ownership create significant incentives
for them to do a bustup. The evidence is clear that the goal of divestitures
is typically to realize gains from industry consolidation, not to improve
performance through an incentive-intensive organizational form.
Analysis of Sources of Target Shareholders' Wealth Gains
In table 9 we summarize the motives and the sources of gains that
might have been important in each transaction. We look at the impor-
tance of strategic factors, selloffs, layoffs, tax savings, and investment
cuts.
In three cases, we have not been able to find any clear sources of
gain or takeover motives. The first is Revlon's bid for Gillette, which
probably was intended to result in a bustup. However, Gillette did not
make significant selloffs after defeating the bid, and its employment
cuts, while large, could not have justified a large fraction of the proposed
premium. As a result, we cannot clearly infer what Revlon was up to.
The second case is Berry's unsuccessful bid for Strawbridge and Cloth-
ier, which was rapidly defeated by a controlling family and resulted in
no changes. This observation is hard to fathom. Finally, we did not
identify important sources of gains in Hanson Trust's acquisition of
U.S. Industries. It is often said that operations are improved after Han-
son's acquisitions, but we do not have adequate measures of that, and
the measures we do have did not register significant changes.
The Importance of Strategic Acquisitions
Table 9 shows that 34 out of 62 acquisitions were strategically mo-
tivated. In most cases, the bidder was interested in the majority of assets
of the target company. Such acquisitions thus reflect the same phenom-
enon that appears to underlie most friendly takeovers in the 1980s: firms
Table
9.
Classification
of
Transactions
by
Sources
of
Gains
Key
source
of
gain
Strategic
Layoffs
Tax
Investment
Target
Bidder
factors
Bustupa
or
pensionsb
savings'
cuts
1.
Aegis
Minstar
S
Debt
2.
American
Motor
Inn
Prime
Motor
Inn
S
B
Partn
3.
Allied
Stores
Campeau
S
B
->
S
W,
B
4.
American
Natural
Resources
Coastal
S
Debt
Maybe
5.
AMF
Minstar
S
B
W
6.
Anderson
Clayton
Quaker
Oatsd
S
B
W
7.
Atlas
Van
Linese
Wesrayd
Debt
8.
Avondale
Mills
Walton
Monroe
Millsd
S
9.
Carter
Hawley
Hale
The
Limitedf
S
10.
Carter
Hawley
Hale
The
Limitedf
S
11.
CBS
Turner
Broadcastingf
S
B
--
S
12.
Chesebrough-Ponds
Unileverd
S
13.
Cluett
Peabody
Westpoint
Pepperelld
S
Debt
14.
Crown
Zellerbach
James
Goldsmith
B
>
S
15.
Easco
Equity
Group
B
16.
Frigitronics
Revlon
B
--
S
17.
Frontier
Holdings
People
Expressd
S
18.
Fruehaufe
Fruehauf
Holdingd
B
19.
Gillette
Revlonf
20.
Great
Lakes
International
Itel
NOLs
Table
9.
(continued)
Key
source
of
gain
Strategic
Layoffs
Tax
Investment
Target
Bidder
factors
Bustupa
or
pensionsb
savingsc
cuts
21.
Gulton
Mark
IV
Industries
Debt
22.
Hammermill
Paper
International
Paperd
S
23.
Hook
Drugs
Krogerd
S
24.
Houston
Natural
Gas
Coastalf
S
Maybe
25.
Imperial
Continental
Gas
Gulf
Resourcesf
B
S
26.
Informatics
General
Sterling
Software
S
W
Debt
27.
John
Blair
Reliance
Capitald
S
B
S
28.
Jonathan
Logan
UMM
S
P
NOLs
29.
Joy
Technologiese
Joy
Manufacturingd
Debt
30.
Masland
Burlington
Industries
S
31.
Mayflower
Group
Laidlawf
S
W
Debt
32.
McGraw-Edison
Cooper
Industries
S
33.
Medford
Amalgamated
Sugar
Yes
34.
MidCon
Occidental
Petroleumd
S
35.
National
Gypsume
Management
Groupd
Debt
Yes
36.
NL
Industries
Harold
Simmons
37.
Owens
Coming
Fiberglass
Wickesf
W
Debt
Yes
38.
Pacific
Lumber
Maxxam
Yes
39.
Phillips
Petroleum
Carl
Icahnf
B,
P
Debt
Yes
40.
Ponderosa
Edelman
w
41.
Prentice
Hall
Gulf
and
Western
S
42.
Quotron
Citicorp
S
43.
Revlon
Pantry
Pride
B
S
44.
Richardson-Vicks
Procter
and
Gambled
S
Debt
45.
Ryan
Homes
HV
Homes
S
Partn
46.
Safewaye
KKRd
B
>
S
47.
Saga
Marriott
S
B
48.
Sanders
Associates
Lockheedd
S
49.
SCM
Hanson
Trust
B
->
S
50.
Scovill
First
City
Properties
B
51.
Southland
Royalty
Burlington
Northern
S
Maybe
52.
Sperry
Burroughs
S
B
53.
Strawbridge
&
Clothier
Berry'
54.
Tull
Industries
Bethlehem
Steeld
S
NOLs
55.
Unidynamics
Craned
56.
Union
Carbide
GAFf
B
W,
P
57.
Uniroyale
Clayton-Dubilierd
B
B,
W
58.
Unocal
Mesa
Partnersf
Partn
Yes
59.
U.S.
Industries
Hanson
Trust
60.
Van
Dusen
Air
APL
Partnership
B
->
S
Partn
61.
Westchester
Financial
Services
Marine
Midland
Bankd
S
62.
White
Consolidated
Electrolux
S
Total
34
17
(10)
14
19
9
Source:
See
table
1.
a.
B:
at
least
50
percent
of
the
acquisition
price
was
recouped
through
selloffs.
B
S:
bustup
where
significant
selloffs
are
to
strategic
buyers.
b.
P:
significant
pension
plan
withdrawals.
W:
white-collar
layoffs
can
explain
over
25
percent
of
the
premium
(computed
at
halfpoint
between
five
years
and
perpetuity).
B:
blue-collar
layoffs
can
explain
over
25
percent
of
the
premium
(computed
at
halfpoint
between
five
years
and
perpetuity).
c.
NOLs:
net
operating
losses
can
explain
over
15
percent
of
the
premium.
Debt:
tax
savings
from
debt
tax
shield
can
explain
over
25
percent
of
the
premium.
Partn:
tax
savings
from
conversion
of
target
into
a
partnership.
d.
White
knight.
e.
LBO.
f.
Unsuccessful.
48 Brookings Papers: Microeconomics 1990
Table 10. Movement of Assetsa
Millions
Movement of dollars Percent
Assets that
changed
hands 68,743 100
Assets that went to strategic
buyers 49,660 72
Strategic
acquisitions
net of selloffs 26,010 38
Selloffs to strategic
buyers 23,650 34
Assets that went to MBOs 10,234 15
Direct
MBOs net of selloffs 4,834 7
Selloffs to MBOs 5,400 8
Assets that
stayed
with initial
nonstrategic
bidders 3,810 5.5
Assets that went to unrelated
acquisitions 3,154 4.5
Direct
unrelated
bidders 373 0.5
Selloffs to unrelated
bidders 2,781 4
Selloffs of headquarters
and other
assets 667 1
Unidentified
selloffs 1,219 2
Source: See table 1.
a. Total value of offers in the sample is $108.5 billion. The value of assets that did not change hands was $39.7 billion.
buying other firms in the same or a closely related industry. In the cases
where the initial acquirer did not want the majority of the assets of the
target company but only some divisions, we see a combination of a
strategic acquisition and a bustup. (In these cases, there is an S in the
first column and a B in the second column of the table.) Thus many
apparent bustups turn out to be strategic in nature as well.
The first column of table 9 does not capture the full extent of strategic
acquisitions. It does not consider the bustups that resulted in the sale
of a significant fraction of the assets to other strategic buyers when the
original acquirer was not strategically motivated. This adds another
seven cases to the list of ultimately strategic transactions. Most of these
are cases of takeover artists buying diversified firms and then selling
off the pieces to strategic buyers.
When we add the number of strategic acquisitions to the number of
nonstrategic acquisitions where the selloffs were largely to strategic
buyers, we end up with 41 cases out of the total 62, or 66 percent.
These results are confirmed by value-weighted evidence in table 10 on
who the eventual holders of the assets in our sample are. The total
value of offers in our sample is $108 billion.
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 49
Of this value of assets, $39 billion did not change hands. These
assets were initially owned by targets that remained independent, and
were not sold off by these targets. The remaining $69 billion of assets
changed hands, and the question is where they ended up.
As we see it, the assets could go to strategic buyers, to unrelated
buyers, to MBOs or investment firms, to nonstrategic initial acquirers
other than MBOs, to buyers of headquarters and other marketable assets
in selloffs, and to unspecified buyers in selloffs. An asset can end up
with a strategic buyer either if it is bought initially by a strategic buyer
and then not sold off or if it is sold off by anyone to a strategic buyer.
Similarly, there are two ways to end up owned by an MBO or held by
an unrelated buyer: by being retained or by being acquired in a selloff.
As table 10 shows, of the $69 billion of assets that changed hands,
72 percent ended up in the possession of strategic buyers. That 72
percent consisted of 38 percent that was retained by initial strategic
buyers and 34 percent that was sold off to strategic buyers. By com-
parison, only 15 percent of assets ended up with MBOs or investment
companies, 4.5 percent ended up with unrelated acquirers, 5.5 percent
stayed with initial nonstrategic buyers, 2 percent went to unidentified
buyers in selloffs, and 1 percent went to buyers of sold-off headquarters
and other marketable assets.
The results again demonstrate the overwhelming importance of stra-
tegic acquisitions in this sample. If we assume, not implausibly, that
the assets that ended up with MBOs or investment companies as well
as the assets that were retained by nonstrategic buyers were managed
by a new organizational form rather than by a corporation, we end up
with 20.5 percent of all the assets that changed hands. That share is
only 28 percent of the assets that went to strategic buyers. It is very
clear that a movement to a new organizational form, where incentives
are crucial, is not the steady state of most assets acquired in hostile
takeovers. Rather, the steady state of an asset is being owned by a large
public corporation in the same line of business as that asset. The primary
motive behind hostile takeovers is not a change in incentives but an
increase in the concentration of asset holdings.
The predominance of this allocation of resources suggests a relatively
small role for incentive-intensive organizational forms, except perhaps
for the purposes of implementing selloffs. It also suggests that unrelated
acquisitions have become rare. There are only a few cases of unrelated
50 Brookings Papers: Microeconomics 1990
diversification by buyers of divisions in this sample: Ralston Purina
buying Eveready Batteries from Union Carbide, a British conglomerate
buying Yale Lock from Scovill. There are also two cases of initial
unrelated acquisitions, although in both cases the acquisitions are
vaguely related. The time of unrelated diversification seems to be past.
The idea is rather to bust up the conglomerates that have previously
pursued unrelated diversification and to allocate divisions to strategic
buyers. In fact, Bhide finds in his sample of post-takeover divestitures
that only 3 out of 81 were companies divesting, directly through selloffs,
a business they started; the rest were cases of divesting past acquisi-
tions 32
This evidence raises the obvious question of where the gains in
strategic acquisitions come from. In some cases, they may come from
increases in market power. This, for example, seems quite possible in
consolidation of pipelines following deregulation (American Natural
Resources and Coastal, Midcon and Occidental Petroleum). Quaker
Oats' acquisition of Anderson Clayton to keep Gaines dogfood was an
explicit move to raise market share, although the link from market share
to market power is not always evident. Minstar's acquisition of boat
divisions of several companies also gave it a significant market share.
Mergers of paper companies probably do not increase competition ei-
ther. Many selloffs also raise concentration. KKR's selloff of some
stores of Safeway was to buyers who already had significant market
shares in the areas where the acquired stores operated. In these cases,
declines in competition are potentially important.
Cost savings due to joint economies in management, production,
distribution, and purchasing are also obviously important in strategic
acquisitions. The classic example is the Burroughs takeover of Sperry
that resulted in significant economies accompanied by massive layoffs.
Some of the better deals negotiated with the suppliers and some of the
gains from layoffs probably reflected wealth transfers rather than pure
efficiency improvements. Significant cost savings were realized in Ster-
ling Software's acquisition of Informatics as well. Some of the mergers
we mentioned in discussing market power probably also led to cost
cuts. These cuts might be another source of gains in strategic acqui-
32. Bhide (1989).
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 51
sitions, which may represent an efficiency gain rather than a transfer
from consumers.
Last but not least, many strategic acquisitions result in declines in
the value of acquiring firms, indicating the market's skepticism that the
bid serves the interest of shareholders rather than the managers. We
have presented some evidence on bidder value declines, although these
numbers are subject to considerable error in estimation. The increase
in combined values of bidders and targets in strategic acquisitions means
either that large joint profit gains that we do not measure are realized
or that the stock market in the 1980s, like the stock market in the 1960s,
was excessively bullish on takeovers.
Bustups
In 17 of 62 cases, or about a quarter of the total, proceeds from
selloffs amounted to at least 50 percent of the purchase price. In about
half the 62 cases, the bustup was a side effect of a bidder's buying the
whole company and keeping only the part that was wanted. Another
significant fraction of bustups was engineered by raiders-for example,
James Goldsmith-who specialize in selling the parts to strategic buy-
ers. Bustups also occured after MBOs. After Clayton-Dubilier's selloffs
at Uniroyal, the only retained division was tires, and even that division
was eventually sold to a tire company. By and large, then, bustups fit
very closely into the picture of strategic acquisitions. Either the original
buyer in a hostile takeover keeps the parts it wants, often selling the
others to strategic buyers as well; or the company is broken up and sold
off largely to strategic buyers. In fact, 48 percent of assets that ended
up with strategic buyers did so as a result of a selloff rather than of the
original acquisition. The willingness of strategic buyers to pay a lot for
these assets must be an important source of gains in bustups.
Layoffs
Table 9 marks 14 cases in which layoffs (or pension terminations)
can explain at least 25 percent of the premium using the midpoint of
the five-year and the perpetuity cost saving estimates. In 8 cases, the
laid-off employees were only white-collar, in 3 cases they were only
blue-collar, in 2 cases they were both, and in 2 cases pension plan
terminations can account for a large chunk of the premium. Savings
52 Brookings Papers: Microeconomics 1990
from white-collar employment cuts can account for a larger fraction of
the premium.
White-collar employment cuts typically resulted from headquarters
consolidations, although in a few cases (Union Carbide, Allied) there
were massive white-collar layoffs. Even in the cases where headquarters
reductions led to large cost savings, the takeovers were primarily stra-
tegic and would probably have occurred even if such savings were not
possible. In only one case (Owens Corning Fiberglass), layoffs were
of R&D staff, and it is sometimes argued that R&D expenditures in
this company were excessively high. Treating layoffs as the primary
motive for takeovers appears to be inappropriate, although they often
are a significant source of gains.
Taxes
The source of gains that we have not so far treated systematically,
but that we still believe to be important, is taxes. Table 9 notes 19
cases in which tax gains can probably account for at least 25 percent
of the premium offered. Of these, 4 are cases in which the target became
a partnership, 3 are cases of bidders with tax losses, and the rest are
cases of large-debt tax shields. Of the 12 cases with significant debt
tax shields, 6 are cases of successful takeovers by large corporations
(all but 1 are clearly strategic) rather than leveraged recapitalizations
by firms remaining independent or of LBOs.
In many cases in which a large debt was incurred in the acquisition,
the firm proceeded to make large selloffs and pay back most of that
debt. Since the value of the tax shield is limited by the duration of the
debt, we try to get an estimate of how quickly the debt is paid down.
We calculate the change in debt based on the firm's debt level two to
three years after the acquisition for the purposes of the debt tax shields
calculation. We take the present value of the tax benefits to be 0.2
times the additional debt. The 0.2 multiplier is consistent with either
some Miller-type personal tax effects lessening the net tax advantage
to debt or with gradual repayment of the debt over seven to ten years.
We also had no handle on other tax benefits, such as accelerated de-
preciation under the general utilities doctrine. These issues require con-
siderable further investigation. Our conjecture is that tax gains will
prove to be somewhat important-probably about as important as lay-
offs. Tax gains might also help explain the puzzle of the increase in
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 53
the combined value of the bidder and the target in some strategic ac-
quisitions with few measured post-takeover changes.
Investment Cuts
Our analysis of investment cuts is severely limited by the scarcity
of target-specific data. Table 9 marks nine cases in which investment
cuts might be important. We note a few things about each of them.
In three cases the cuts were in the petroleum industry (Unocal, Phil-
lips, and Southland Royalty) during the time of declining oil prices.
These cases fit nicely Jensen's free cash flow theory, according to which
oil companies continued to explore even when value maximization dic-
tated a significant curtailment of exploration activities. Hostile takeover
threats enforced this reduction in exploration (as well as adoption of
limited partnership organizational form to save taxes). Although firms
that were not targets of hostile takeovers also cut back exploration, they
might have done it under the threat of takeover. The free cash flow
story seems very plausible for oil.
Another industry in which severe investment cuts took place in the
period of declining oil prices was gas pipelines, presumably because
gas is a substitute for oil and its price falls when oil prices fall. If it is
rational to curtail investment when the gas price declines, the observed
cuts must be rational. In the case of gas pipelines, we do not believe
the free cash flow story explains the takeovers. During this period the
industry experienced deregulation and consolidation, with the effect
that there were many mergers assembling pipeline systems to increase
reliability of supplies as well as market power. There is virtually no
public discussion of the objective of pipeline mergers being to stop
investment rather than to create pipeline systems. Our view also fits
with the mass of often successful antitrust suits following these mergers.
The investment cuts after the takeovers thus do not seem to be driving
these takeovers in the first place or even to be a by-product of these
takeovers.
The third group of takeovers with investment cuts-or rather in-
creases in disinvestment-is in timber. Most notable of these is Maxxam's
takeover of Pacific Lumber, but there is also Amalgamated Sugar's
takeover of Medford. In these two cases, the cutting of trees was sharply
increased. The targets were apparently cutting too few trees given the
interest rate, the growth rate of trees, and price path for timber. Pacific
54 Brookings Papers: Microeconomics 1990
Lumber, in particular, had a huge forest of old redwoods that were not
growing. The company was nonetheless not cutting them despite high
interest rates, perhaps because it expected a secular rise in the price of
old redwoods. Maxxam came in and cut the trees. Assuming away the
possibility of disagreement over the future prices of these trees (market
underpricing), we have a case in which cutting the trees raises effi-
ciency. The Medford story is similar. This is not really a free cash flow
case, since there was no cash that the target reinvested wastefully, but
it is nonetheless a case in which disinvestment and partial liquidation
were enforced by a takeover.
The final two cases are an LBO and a leveraged recapitalization
where investment was cut and the debt raised very sharply (National
Gypsum and Owens Corning). These may be free cash flow cases, but
they may also be cases of cuts in useful investment because of limited
access to capital. In both cases, much of the takeover premium can be
explained with the debt tax shield, and in the case of Owens Corning
there were large white-collar layoffs also. In a similar vein, Kaplan
finds large investment reductions following MBOs more generally.33
In sum, the evidence on the free cash flow theory is mixed. It appears
relevant in the case of oil. It might be relevant in the case of gas pipelines
and in the case of LBOs, although the evidence is not compelling,
especially with pipelines. In the case of timber, an amended version of
the free cash flow theory might apply as well, although there is no free
cash flow. Investment cuts are thus potentially important in 9 cases out
of 62. In most other cases, particularly with strategic acquisitions,
investment cuts are either ignored in all the discussions and reports we
found, or else are not there. However, we do not think it is likely that
lack of reporting is the whole story, since in the case of the oil industry,
reports of investment cuts are very common. The conspicuous absence
of such discussions for most other industries might indicate that in-
vestment cuts are simply unimportant.
Summary and Conclusions
Although we do not have the final answer on what drives hostile
takeovers, several conclusions do emerge from the analysis.
33. Kaplan (1990).
Sanijai
Bhagat, Andrei Shleifer, and Robert Vishny 55
First, hostile takeovers largely allocate businesses to firms owning
other related businesses. Firms acquire related firms both directly in
the takeover and in selloffs after the takeover. Of the $69 billion in
assets that changed hands in our sample, 72 percent ended up in the
hands of firms managing other similar assets. We do not know whether
the gains in related acquisitions come from improvements in operating
efficiency, increases in market power, or other sources.
Second, by comparison with strategic reallocation of assets, re-
allocation to MBO teams, investment groups, raiders, or other incen-
tive-intensive organizations is only 20 percent of the total that changed
hands. Raiders and MBO teams appear largely to serve the temporary
function of brokering the transfer of assets toward related acquirers.
High debt levels and concentrated ownership give these organizations
a strong incentive to implement a bustup, but the task of subsequent
management is left to others.
Third, layoffs are an important but not a dominant source of hostile
takeover gains, accounting perhaps for 11 percent to 26 percent of the
premium on average. Layoffs are disproportionately targeted at white-
collar employees, many of them in the course of consolidation of head-
quarters.
Fourth, selloffs are a pervasive consequence of hostile takeovers,
and in many cases result in a liquidation or a near-liquidation of the
target. A key objective of selloffs is to foster related acquisitions: over
two thirds of the selloffs are to firms in a related business. Selloffs are
smallest when the acquirer is a white knight.
Fifth, tax savings are important in some cases, particularly in LBOs,
but the benefits of the debt tax shield are significantly reduced by the
rapid repayment of debt. Tax losses and conversion to partnerships are
less common but provide large benefits when they do take place.
Sixth, among the explanations of the sources of takeover gains that
are important in some, but relatively few, cases are bidding shareholder
losses and cuts of wasteful investment by the target firm. These are
much less common sources of gains than layoffs or tax savings.
These findings suggest the following picture of the hostile takeover
process of the 1980s. In the postwar period, aggressive antitrust en-
forcement prevented significant consolidation of U.S. industry and per-
haps even encouraged the formation of conglomerates in the 1960s.
Experience from the 1970s showed that conglomerates are probably not
56 Brookings Papers: Microeconomics 1990
the most profitable way to run businesses.34 The failure of conglom-
erates is revealed by the finding of Porter and of Kaplan and Weisbach
that 40 percent to 60 percent of unrelated acquisitions were subsequently
divested.35
In the 1 980s, the economy has again experienced a corporate liquidity
boom. This increased liquidity has come in two forms. The first is
greater internally generated cash flows in large corporations that want
to make related (strategic) acquisitions. The second is a newfound
ability to issue speculative-grade ("junk") bonds as a temporary fi-
nancing tool while assets are being sold off. Junk bonds facilitated
acquisitions by so-called "raiders" as well as by some strategic buyers.
In addition, the most lenient antitrust enforcement in decades allowed
the large pent-up demand for related acquisitions to be satisfied. Finally,
heightened foreign competition and the deregulation of several key
industries such as transportation, banking, and oil and gas increased
the gains from intraindustry mergers. The takeover wave became the
wave of related acquisitions. Much of this wave is reflected in the
increased pace of friendly related acquisitions. But in addition, the pace
of hostile takeovers skyrocketed. Many of the hostile takeovers, as well
as acquisitions by white knights, aimed to deconglomerate large cor-
porations and to allocate their various divisions to related acquirers.
Raiders and MBO organizers in particular acquired diversified firms
and sold off the parts to related acquirers, profiting handsomely from
performing this brokerage function.
Although the expansion into related businesses drove the takeover
process, several opportunities to raise the value of firms must have also
encouraged hostile takeovers. The opportunities to cut overhead during
consolidations by cutting headquarters staff as well as some common
functions allowed nontrivial cost savings that certainly helped the pro-
cess, although they did not drive it. Tax subsidies to debt clearly helped
the raiders and MBO organizers to realize some gains, although the
opportunity to unload the assets on strategic buyers must have been
crucial. The fact that the typically strategically motivated white knights
reap fewer gains from layoffs and selloffs than hostile acquirers suggests
34. Ravenscraft and Scherer (1987).
35. Porter
(1987) and Kaplan and Weisbach (1990).
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 57
that these changes are not the most essential source of gains from
consolidation.
Our evidence also shows what hostile takeovers are not. They are
not typically a reflection of change in the internal organization of the
firm. Management buyouts and acquisitions by raiders are often a tem-
porary step in the reallocation of assets; they are not a new permanent
organizational form. The eventual holders of assets are large public
corporations, which are not about to be eclipsed.
This interpretation of hostile takeovers in the 1980s offers several
tentative conclusions for policy analysis. First, hostile takeovers do not
result in massive employment cuts in acquired companies. State anti-
takeover laws that aim to stop takeovers to protect blue-collar workers
are misguided. Since such laws probably stop some takeovers that foster
specialization of corporations, they are more likely than not to reduce
efficiency. Second, Reagan's lenient antitrust enforcement of the 1980s
indirectly fostered deconglomeration of the U.S. economy. Since the
experience with conglomerates seems almost uniformly disappointing,
the move toward specialization probably on balance raises efficiency.
In some cases, competition is probably reduced, but our case studies
suggest that there are many business reasons for related acquisitions
other than to raise prices. Unfortunately, we do not have the highly
disaggregated market share data necessary to evaluate the precise scope
for increased market power in our sample of acquisitions. On balance,
however, the evidence suggests to us that the Reagan antitrust stance
has had a positive influence on the economy.
Our analysis leaves open the key question: where do the value gains
in strategic acquisitions come from? The fact that in a typical strategic
acquisition the combined wealth change of the bidding and target share-
holders is positive suggests that the market believes these gains do
exist. We have identified some sources of efficiency improvements and
other gains, such as headquarters layoffs and tax savings, but they
clearly are not the whole story. There may be many efficiency gains in
production, procurement, and distribution that we have no way of cap-
turing. Gains from market power might also be relevant, but the evi-
dence on the existence of these gains is inconclusive. Finally, the value
gains recorded may simply reflect the market's overestimation of the
value of strategic combinations, just as the market overestimated the
gains to conglomerate mergers in the 1960s. This may be reflected
58 Brookings
Papers: Microeconomics
1990
either in too low a share price for the pre-takeover target firm or in too
high a price for the post-takeover acquirer or those buying divested
divisions of the target at high prices. In part, the market may currently
be underpricing conglomerates relative to undiversified companies. All
of the above sources of gains are potentially important. We appear to
have reduced the scope of the problem to that of related acquisitions,
but that problem remains wide open.
Appendix: Brief Summaries of Deals
In this appendix, we present a brief summary of what we believe to
be important sources of target shareholder wealth gains in each one of
our 62 hostile takeover attempts. We bring in the earlier evidence on
bidding shareholder wealth changes, layoffs, and selloffs when such
evidence is pertinent, but also discuss taxes, investment changes, and
possible stock market underpricing when we believe these are impor-
tant. In addition, we document the cases where strategic motives appear
to play a role, that is, how often the bidder and the target are in the
same or in closely related businesses. Equity refers to the purchase
price of equity, debt to the pre-takeover value of debt, and premium
to the premium computed in the text.
1. Minstar buys Aegis. Equity $59 million; debt $20 million; pre-
mium $22 million. Selloff of Cherco Compressors and Turboscope, the
latter bought as part of AMF, for $150 million, but don't know how
much is from Aegis. Some operations discontinued, but no clear layoffs.
$3 million taken from the pension fund. $53 million in additional debt
is taken on by Minstar so taxes may play a role. Primarily, this seems
to be a strategic acquisition. Minstar makes boats and is interested in
the boat division of Aegis.
2. Prime Motor Inn buys American Motor Inn. Equity $238 million;
debt $67 million; premium $60 million. Selloffs of $383.7 million, with
12 Holiday Inns retained. Over half of selloffs are to limited partnerships
with Prime leasing some of them back. There was an increase in in-
vestment (refurbishing). No layoffs. This is partly a strategic acquisition
with Prime now managing many of the former AMI properties. It is
Sanjai Bhagat, Andrei
Shleifer,
and Robert
Vishny 59
also a bustup, motivated largely by the tax benefits of the limited part-
nership form of organization (only one layer of tax).
3. Campeau buys Allied Stores. Equity $3.7 billion; debt $664 mil-
lion; premium $1. 11 billion. Selloffs of at least $2.5 billion. Remaining
divisions generate about 50 percent of revenue. Assumes about $3
billion in additional debt but with very rapid payback through selloffs.
Layoffs of at least 2,625: 600 headquarters, 2,025 blue-collar; savings
of between $270 million and $705 million in present-value terms. This
appears to be a strategic bid for Allied's shopping centers (Campeau
is a shopping center developer), but the premium may derive from
profitable asset selloffs and savings from layoffs.
4. Coastal buys American Natural Resources. Equity $2.46 billion;
debt $1,065 billion; premium $763 million. No selloffs. Assumes $1.9
billion in additional debt with fairly slow pay down. Some job reduction
and a sharp decrease in investment, but this is true throughout the
industry. Appears to be a strategic bid stemming from the post-dere-
gulation consolidation in the pipeline industry. The market seems to
like Coastal's strategy; its market value rose by $117 million around
the announcement of its bid.
5. Minstar buys AMF. Equity $545 million; debt $170 million; pre-
mium $74 million. Selloffs of at least $511 million (although Minstar
CEO Jacobs says $535 million). Retained several AMF businesses ac-
counting for approximately 20-25 percent of pre-acquisition operating
profit and $350 million in book value. In particular, kept boat division,
Hatteras, which fits into Jacobs' strategy of acquiring boat manufac-
turers. Cut 350 staff positions with present value of saving between
$67 million and $175 million. Strategic bid for part of company com-
bined with a profitable bustup and large savings from headquarters
layoffs.
6. Quaker Oats buys Anderson Clayton as a white knight. Equity
$805 million; debt $19 million; premium $108 million. Sells off all but
Gaines dogfood for $535 million. Anderson Clayton had $240 million
in cash. This means $50 million for Gaines. Previously offered $250
million for Gaines alone. Headquarters staff reduction with present-
value of savings of at least $145 million. Strategic bid for part of
company, combined with a very profitable bustup and headquarters
layoffs. Market value of Quaker Oats declines by $165 million when
60 Brookings
Papers:
Microeconomics 1990
acquisition is announced, suggesting that market did not anticipate such
a profitable bustup.
7. Wesray buys Atlas Van Lines in a white knight LBO. Equity $72
million; debt $4.6 million; premium $27 million. Assumed $70 million
additional debt. No selloffs. Evidence that employment declines from
over 600 to 524 over two years, but no mention of layoffs. Debt tax
shield can explain most of the premium.
8. Walton Monroe Mills buys Avondale Mills as a white knight.
Equity $1 13 million; debt $37.6 million; premium $38 million. Intended
selloffs of about $18 million. No layoffs, but information is scarce
because the acquirer is private. $4.5 million from pension fund. Largely
a strategic acquisition; Avondale is the exclusive marketer of Walton
Monroe products and also buys a large quantity for its own account.
9. The Limited unsuccessfully tries to buy Carter Hawley Hale (1984).
Equity $1. 1 billion; debt $521 million; premium $397. Selloffs of $333
million. No evidence of layoffs. Appears to be a strategically motivated
acquisition attempt possibly with bustup elements.
10. Retail Partners (including the Limited) unsuccessfully tries to
buy Carter Hawley Hale (1986). Equity $1.95 billion; debt $700 mil-
lion; premium $1.16 billion. Selloffs of $213 million. However, also
spin off most valuable properties to shareholders, including Neiman-
Marcus, Bergdorf-Goodman, Contempo Casuals. No evidence of lay-
offs. Again, seems to be strategic, with elements of a bustup.
11. Turner Broadcasting tries unsuccessfully to buy CBS. Equity
$5.41 billion; debt $371 million; premium $2.75 billion. Defensive
response by CBS is a leveraged recapitalization with selloffs of at least
$3.44 billion, but the most valuable assets remain. Partly these selloffs
are profitable because of a rise in the market as a whole over the two
years following the Turner bid. Present value of savings from layoffs
between $238 million and $627 million. Capital expenditures in the
broadcast division are cut by $30 million. This appears to have been a
strategic bid by Turner for part of the company (broadcasting), but the
bid may have also been motivated by gains from bustup and layoffs.
Seems to have elements of everything.
12. Unilever buys Chesebrough-Ponds as a white knight. Equity
$3.1 billion; debt $1.17 billion; premium $1.44 billion. Selloffs are
more than $1 .8 billion, largely consisting of Stauffer Chemical that CP
Sanjai Bhagat, Andrei
Shleifer, and Robert Vishny 61
bought shortly prior to being itself acquired. Layoffs save between $30
million and $80 million in present value terms. Strategic bid for the
core business, which is personal hygiene and health care products. This
fits with Unilever's strategic objective to gain a larger U.S. presence
in these products for purposes of exploiting its superior R&D.
13. Westpoint Pepperell buys Cluett Peabody as a white knight.
Equity $375 million; debt $124 million; premium $77 million. No lay-
offs. No selloffs. West Point assumes approximately $125 million in
additional debt, which it does not pay down quickly. Appears to be a
strategic acquisition motivated by Westpoint's desire to acquire more
brand names including several international brands of Cluett.
14. James Goldsmith buys Crown Zellerbach (no white knight found).
Equity around $1.2 billion; debt $648 million; premium $193 million.
Selloffs of at least $1.15 billion, mostly to strategic buyers. Goldsmith
is after Crown Zellerbach timberland, which he considers undervalued.
This is clearly a bustup, possibly with elements of underpricing by the
stock market of timber assets or too slow harvesting by Crown as in
the Pacific Lumber case (see below).
15. Equity Group buys Easco. Equity $175 million; debt $90 million;
premium $27 million. Selloffs of $13 million. Acquirer is interested in
the hand tool business of Easco as evidenced by its buying 100 percent
of this business from remaining shareholders of Easco. Hand tool busi-
ness is in the midst of dramatic turnaround clearly started before the
bid (but not finished until after). There is a reduction of employment
from 3,065 to 2,184 with a large rise in sales per employee. At least
350 of these 881 layoffs take place before the bid, and the manager
brought in to lead the turnaround is kept on by the acquirer. Layoffs
of this magnitude can easily explain the premium paid. The only ques-
tion is whether the acquisition was needed to speed up the turnaround
or whether the acquirer was simply more willing than the market to bet
on the results. Easco Hand Tools goes public again in 1987.
16. Revlon buys Frigitronics after a failed LBO. Equity $115 mil-
lion; debt $8 million; premium $41 million. Asset selloffs of at least
$120 million, largely to strategic buyers. No layoffs. Complete bustup
and liquidation-both planned from the start.
17. People Express buys Frontier Holdings as a white knight. Equity
$307 million; debt $35 million; premium $117 million. No layoffs by
62 Brookings
Papers:
Microeconomics
1990
agreement with People's. No asset selloffs until bankruptcy. People
tries to turn Frontier into a no-frills airline but the strategy does not
work out as Frontier is badly hurt by fare wars in its major markets.
After that, People tries to sell Frontier to United for $146 million, but
fails. Instead, United buys hangars, gates, etc., for $50 million. Even-
tually bought by Texas Air for $174 million, including assumption of
debt. Initial acquisition is clearly strategic. Part of People's ill-fated
expansion program.
18. Fruehauf LBO as defense against Edelman. Equity $ 1.1 billion;
debt $480 million; premium $414 million. Selloffs of at least $732
million, but kept automotive business and domestic truck trailers di-
visions representing over 80 percent of pre-acquisition operating profit.
Close to bankruptcy soon after the buyout. Assumed $500 million in
additional long-term debt and $400 million-$500 million in additional
short-term debt at time of buyout. After 1.5 years had $400 million of
the additional long-term debt remaining. Layoffs of 730 to save between
$55 million and $146 million in present-value terms. Basically, man-
agement forced into bustup/restructuring LBO, but probably overpay-
ment by the LBO group combined with industry shocks to the trailer
and container businesses.
19. Revlon unsuccessfully tries to buy Gillette. Equity $4.12 billion;
debt $436 million; premium $1.48 billion. Selloffs of $50 million.
Planned layoffs will save between $182 million and $480 million in
present-value terms. The bidder probably intended a much more radical
bustup and restructuring than actually occurred as evidenced by low
level of Gillette's share price after fending off Revlon.
20. Itel Corporation buys Great Lakes International. Equity $170
million; debt $30 million; premium $14 million. Selloffs of $31 million.
Itel has large tax loss carryforwards, and Great Lakes is showing profit.
Also, Itel may be betting on legislation passed shortly after acquisition,
which is a huge boon to Great Lakes' offshore dredging business. Small
premium can easily be explained by taxes.
21. Mark IV Industries buys Gulton. Equity $113 million; debt $9
million; premium $36 million. Insubstantial selloffs. Appears to take
on approximately $100 million in additional long-term debt, which is
not paid down quickly, although it is difficult to attribute all that debt
to Gulton acquisition. Some headquarters layoffs, but cannot document
Sanjai
Bhagat, Andrei Shleifer,
and Robert Vishny 63
numbers. Mark IV assembles a diversified portfolio of small high-tech
businesses in various industries and simply reallocates capital among
them. Mark IV market value declines by $6 million on announcement
of bid.
22. International Paper buys Hammermill Paper as a white knight.
Equity $1.1 billion; debt $310 million; premium $418 million. Small
selloffs. No layoffs. Strategic.
23. Kroger buys Hook Drugs as a white knight. Equity $161 million;
debt $13 million; premium $74 million. No selloffs initially. Hook is
supposed to be well run. Subsequently Kroger sells off Hook and its
own drug chain, Superex, to management team for $490 million. Not
clear how much is Hook, but conjectured $100 million-$200 million.
Kroger is said to lose a little on Hook overall. Initial intention is clearly
strategic; alleged synergies in pharmaceutical purchasing and private
label procurement. Does not appear to have worked out. Kroger's mar-
ket value falls by $26 million on announcement of its bid.
24. Coastal tries unsuccessfully to buy Houston Natural Gas. Equity
$2.72 billion; debt $360 million; premium $961 million. Defensive
response by HNG is leveraged recapitalization/share buyback plan. In-
itially, $700 million in additional debt is assumed. Selloffs of more
than $544 million. No evidence of layoffs at this stage. HNG acquires
two additional pipelines in 1984. Strategic bid 'a
la Coastal/ANR; post-
deregulation consolidation in pipeline industry. Within two years, HNG
is acquired by Internorth in a friendly acquisition.
25. Gulf Resources unsuccessfully tries to buy Imperial Continental
Gas (a British Company). Equity 753 million pounds; some debt; pre-
mium 155 million pounds. Defensive response of IC Gas is to be split
into two companies, the first containing the firm's sought after Belgian
assets. After spinoff the Belgian assets were acquired for 448 million
pounds. The second piece of IC Gas later rejects a bid for 821 million
pounds, but then spins off a subsidiary valued at 300 million pounds.
A clear attempted bustup.
26. Sterling Software buys Informatics General. Equity $126 mil-
lion; debt $1 million; premium 28 million. Selloffs of at least $34
million. Sterling assumes $100 million in additional debt. Initial goal
is to reduce debt to $50 million, but two years later only reduced by
$20-30 million. Eliminate entire corporate staff (60 domestic and 50
64 Brookings Papers:
Microeconomics 1990
international employees). Present value of the after-tax savings is $22-
60 million. Layoffs suffice to explain most of the premium, although
debt tax shields may explain a large part of the premium. Also, this is
clearly a strategic acquisition.
27. Reliance Capital buys John Blair as a white knight. Equity $356
million; debt $240 million; premium $165 million. Selloffs at least
$377 million. Keep Spanish TV stations that Reliance also has. No
evidence of layoffs. Strategic bid for part of company.
28. UMM buys Jonathan Logan (over unsuccessful defensive LBO).
Equity about $196 million; debt $18 million; premium $16 million.
Selloffs of $17 million. Substantial layoffs but most appear to be at
UMM rather than Jonathan Logan. UMM did phase out ladies' apparel
division of JL. Pension transfer of $15 million. UMM has some tax
losses. Pension transfer along with better use of tax losses can explain
the small premium here, although the original motives seem to have
been strategic.
29. Joy Technologies makes a defensive LBO of Joy Man4facturing.
Equity $620 million; debt $96 million; premium $229 million. Selloffs
are at least $252 million plus assumption of some liabilities. Retained
divisions represent at least 75 percent of pre-acquisition operating in-
come. No evidence of layoffs. Approximately $400 million in additional
debt taken on, so tax benefits may be important. Motivation for LBO
is to remain independent; premium can probably be explained by prof-
itable partial bustup coupled with tax benefits.
30. Burlington Industries buys Masland. Equity $117 million; debt
$19.5 million; premium $42 million. No selloffs. No layoffs. Strategic
acquisition. After Burlington LBO in 1988, Masland division put on
the block. It is sold for $79 million plus $66 million in preferred stock.
Burlington lost $80 million around the announcement of its bid.
31. Laidlaw tries unsuccessfully to buy Mayflower. Equity $260 mil-
lion, debt $60 million; premium $63 million. Defensive response of
Mayflower is a leveraged recapitalization. Approximately $150 million
in additional long-term debt after two years. Selloffs of $25 million.
Layoffs of 125 people bring savings of $24 million-63 million in pres-
ent-value terms. Also, may be exploiting unused debt capacity in a
low-risk mature firm (school bus operator). Layoffs and debt tax shields
can probably explain the premium. Strategic element is also important
Sanjai Bhagat, Andrei
Shleifer,
and Robert Vishny 65
as Laidlaw and Mayflower are the two biggest firms in the school bus
business.
32. Cooper Industries buys McGraw-Edison. Equity $1.1 billion;
debt $300 million; premium $410 million. Selloffs of $177 million that
can be identified; Moody says $260 million. In either case, keep vast
majority of assets. Some layoffs at a McGraw-Edison unit headquarters
and some plant shutdowns. Strategic acquisition with some evidence
of restructuring.
33. Amalgamated Sugar (Harold Simmons) buys Medford. Equity
$108 million; debt $37 million; premium $49 million. Selloffs of $38
million. Large stepup in harvesting trees. Probably a similar story to
Pacific Lumber. Either underpriced or underutilized timber resources.
Is there a general violation of Hotelling pricing of timber or just a lot
of heterogeneity in types of timber, tree growth rates, and so forth?
34. Occidental Petroleum buys MidCon as a white knight. Equity
$3.1 billion; debt $800 million; premium $1.4 billion. Selloffs of at
least $1.47 billion. Layoffs of 2,000, saving between $152 million and
$400 million in present-value terms. Approximately $80 million cut in
capital expenditure. Not clear these layoffs and capital spending cuts
are abnormal given what is happening in the industry right around the
time of the acquisition. Appears to be a strategically motivated acqui-
sition with Oxy trying to acquire pipelines to transport its gas. Later
Oxy loses civil lawsuit saying that tried to monopolize regional gas
distribution through MidCon acquisition. Oxy's market value declines
by $365 million on announcement of Midcon bid; fits with Oxy's typical
pattern.
35. National Gypsum LBO (defensive against Wickes). Equity $1.6
billion; debt $120 million; premium $325 million. Assume approxi-
mately $1.3 billion in additional long-term debt. Fairly slow debt re-
duction over time. Selloffs of at least $441 million. No layoffs we can
identify. Cut investment from $91 million to $24 million. Not clear
that this is wasteful investment. Tax savings from higher debt can justify
much of the premium. Free cash flow story may be important.
36. Harold Simmons buys NL Industries after unsuccessful bid by
Coniston. Offers $900 million for the firm (plus debt of $231 million),
but ends up buying control of the firm on the open market. Firm consists
of very profitable chemicals business and not very profitable oil rigs
66 Brookings Papers:
Microeconomnics
1990
business. As a defensive response NL has spun off NL Chemicals with
most of its cash flow going to a new class of preferred stock. Initially
Simmons purchases 51 percent of the parent firm (supposedly obtaining
control and cash flow claims to oil business) and 20 percent of the
preferred stock with cash flow claims to chemical business. Premium
in this initial purchase is negligible or negative. Ultimately, Simmons
is after the chemical business and over time accumulates a majority of
the preferred stock. His attempts to sell off chemicals to his own firm
fail despite having 51 percent of the votes in the parent. In the meantime,
the chemicals business (primarily titanium dioxide) is taking off due to
shortages, and a subsequent LBO offer of $915 million for chemical
subsidiary alone is rejected as too low. No selloffs. Pension plan ter-
minated: $81 million received after tax. Employment falls very sharply
in the petroleum business, but this is typical of the whole industry.
Pension transfer can explain the premium, but most likely the company
was significantly underpriced because the potential of chemical business
was not appreciated by the market. At least this seems to be what is
motivating Simmons.
37. Wickes tries unsuccessfully to buy Owens Corning Fiberglass.
Equity $2.1 billion; debt $540 million; premium $528 million. Defen-
sive response of Owens is leveraged recapitalization/share buyback.
Take on $2 billion in additional debt, but pay down about $1 billion
of that within one to two years. After three years still have approxi-
mately $800 million in additional debt. Selloffs at least $501 million;
$679 million according to Moody's. Layoffs of 480 R&D personnel;
also early retirement of 373 at headquarters. Estimated present value
of saving from research personnel layoffs alone is between $91 million
and $240 million. There is some evidence that Owens has been spending
a lot on research without significant results. Early retirements save
perhaps another $70 million-$186 million. Also significant capital
spending cuts. Can trace approximately $30 million a year of those cuts
to nondivested businesses. Motivation of Wickes is a little unclear.
Owens defensive restructuring works to boost share price even a little
above the level of the bid. Research and headquarters employment
reductions, some tax savings from increased debt, and elimination of
possibly wasteful capital expenditures can probably explain this.
38. Maxxam buys Pacific Lumber. Equity $870 million; debt $46
million; premium $256 million. Selloffs of $351 million. Withdrawal
Sanjai Bhagat, Andrei
Shleifer,
and Robert Vishny 67
of $50 million from pension plan. Sharply step up cutting old redwood
trees, which the previous management left untouched. Employment
rises to accommodate higher output. Subsequent appraisal of timber
assets at $2. 1 billion. A little unclear how much this is case of timber
assets being underpriced by the stock market and how much the old
management was wasting by harvesting too slowly. There is some
evidence for the view that the old management's slow harvesting policy
was suboptimal.
39. Pickens tries unsuccessfully to buy Phillips Petroleum (Icahn
also involved). Equity $8.5 billion; debt $2.84 billion; premium $1.22
billion. Defensive response of Phillips is leveraged recapitalization/
share buyback plan. Assumed approximately $3.7 billion of additional
debt. Three years later still had $2 billion in additional debt. Selloffs
at least $2 billion. Sharp exploration cutbacks. Large layoffs (saving
between $467 million and $1.23 billion in present-value terms). Pension
excess of $400 million. Possibly utilization of unused debt capacity
(tax shields). Premium can be explained by layoffs and cutbacks in
wasteful exploration, although there is a big question about how much
would have been cut even without takeover threat. Post-recapitalization
projections for exploration spending do not contain severe cuts but talk
about the possibility of such cuts if oil prices fall substantially (which
happened in the year after the recap). Taxes may also be important
here: not only exploitation of unused debt capacity, but also Pickens
may have been interested in spinning off oil and gas properties into a
partnership or trust to avoid double taxation.
40. Edelman buys Ponderosa after another takeover bid is defeated.
Equity $280 million; debt $48 million; premium $68 million. Assumed
about $270 million in additional debt. Selloffs of $17 million not in-
cluding sale of office building, corporate jet, and art collection; other
planned selloffs failed. Headquarters layoffs save between $22.8 mil-
lion and $60 million in present value. Cut capital spending from $63
million to $20 million. Sold whole company to Metromedia in 1988 at
a small loss rather than put in a mere $8 million-$10 million more in
equity capital. Apparently intended more piecemeal selloffs, although
layoffs and debt tax shields should easily explain the premium with
unclear value enhancement role for capital spending cuts.
41. Gulf and Western buys Prentice Hall. Equity $705 million; debt
$3 million; premium $254 million, but G&W declines by $59 million.
68 Brookings Papers:
Microeconomics
1990
No layoffs, no selloffs. Strategic bid: both companies are publishers.
42. Citicorp buys Quotron. Equity $680 million; debt $7 million;
premium $137 million. No selloffs. No layoffs. Strategic acquisition
fits Citicorp's plan to move into information services.
43. Pantry Pride buys Revlon. Equity $1.8 billion; debt $485 mil-
lion; premium $256 million. Selloffs of assets are at $2.06 billion
largely to strategic buyers in the health care business, but the main
cosmetics business is kept. Rejects offer to sell cosmetics for $905
million. Revlon also has $420 million in cash. Classic bustup although
may have also improved cosmetics operations. Allegedly the cosmetic
business received the attention and resources (advertising budget) it
needed after health care business was sold off.
44. Procter and Gamble buys Richardson-Vicks as a white knight.
Equity $1.66 billion; debt $170 million; premium $776 million, but
P&G loses $165 million on the announcement. Approximately $1 billion
in additional long-term debt is assumed by P&G and this is not reduced
in the two years after the acquisition. Selloffs at least $106 million.
No layoffs even though there was some talk about efficiencies from
combining sales forces. Take on $1.5 billion in debt. Clearly a strategic
acquisition. P&G wants international distribution channels of Richard-
son-Vicks to increase sales of its products overseas and wants to use
its own bargaining power in purchasing advertising to benefit R-V prod-
ucts in U.S.
45. HV Homes (a limited partnership) buys Ryan Homes. Equity
$330 million; debt $18 million; premium $64 million. No selloffs. No
layoffs. Conversion to limited partnership has large tax benefits, as
does the increase in debt. Also, acquirer argued that strategic consid-
erations are important. Tax savings can explain the premium, however.
46. KKR buys Safeway as a white knight defense against Dart Group
bid. Equity $4.2 billion; debt $1.1 billion; premium $1.5 billion. Layoff
of 300 HQ staff for a saving of between $57 and $150 million in PV
terms. Selloffs of at least $3.2 billion. Stores representing at least 70
percent of sales are kept. Sold-off stores seem to be the higher labor
cost stores (depending on local union and labor market conditions).
Some concerns that union workers will be squeezed. Also, each regional
piece of the firm seems to be sold to other players with lots of stores
in that region (for example, Vons Grocery in Southern California),
Sanjai
Bhagat, Andrei
Shieifer,
and Robert
Vishny 69
perhaps with the intention of raising market power. Taxes and layoffs
alone cannot explain the premium; bustup must be important.
47. Marriott buys Saga. Equity $502 million; debt $105 million;
premium $148 million, but Marriott loses $162 million in value. Selloffs
of $306 million. Combined operations and support staff but no layoff
numbers are available. Keeps foodservice business. Strategic bid for
part of the company.
48. Lockheed buys Sanders Associates as a white knight. Equity
$1. 18 billion; debt $19 million; premium $614 million. Some layoffs,
but small relative to the premium. No selloffs. Strategic bid, both do
defense.
49. Hanson Trust buys SCM. Equity $927.5 million; debt $261 mil-
lion; premium $386 million. Selloffs of at least $900.2 million mostly
to strategic buyers, but kept typewriter division and titanium dioxide
assets representing almost 50 percent of pre-acquisition operating in-
come. No evidence of layoffs. Seems like a clear bustup.
50. First City Properties buys Scovill. Equity $540 million; debt
$87 million; premium $212 million. Selloffs of at least $680.7 million,
but a division with book value of assets of over $100 million and at
least 25 percent of pre-acquisition operating profit is retained. Cuts in
corporate staff of 30-35 percent as well, exact numbers are not known.
The crown jewels of Scovill (Nutone and Yale Security) were sold off
about 2.5 years after the acquisition near the market's peak (and ac-
counted for $480 million of the proceeds from selloffs).
51. Burlington Northern buys Southland Royalty (after the latter
tries to reorganize as a limited partnership to reduce tax bill). Equity
$695 million; debt $339 million; premium $126 million, but BN loses
$97 million in value. No selloffs. Big reductions in BN pipeline work
force (30 percent), unclear how many from Southland Royalty. Some
investment reductions. Layoffs and investment cuts are quite typical of
the entire industry around this time. Motive appears to be strategic;
Southland is a big supplier to BN's pipelines. Burlington falls by almost
full amount of premium paid for Southland. Eventually, Burlington
Northern spins off Southland Royalty and other natural resources into
Burlington Resources, a limited partnership when it itself becomes a
target.
70 Brookings
Papers:
Microeconomics
1990
52. Burroughs buys Sperry. Equity $4.44 billion; debt $1.03 billion;
premium $1.22 billion. Selloffs of at least $1.8 billion. Layoffs of at
least 2,370 for a saving of $180 million to $474 million. Actual em-
ployment reductions are much bigger than layoffs, at least 10,000 through
early retirement, attrition, and layoff. That means total savings of $760
million to $2 billion. Employment cutbacks are a big source of gain.
Also, other efficiencies, including a purchase of inputs. Allegedly save
$100 million per year before tax by getting the better of each of the
firm's previous deals with individual input suppliers. Layoffs and pur-
chasing efficiencies realized can clearly explain the premium.
53. Berry tries unsuccessfully to buy Strawbridge & Clothier. Equity
$375 million; debt $105 million; premium $55 million. No selloffs. No
layoffs. Bidder has a board seat already and wants specific changes,
such as financing of receivables and expansion of the discount chain,
but the family in power refuses to do it. Looks like potential efficiency
improvement opposed by the founding family. Berry may just be trying
to put Strawbridge in play; there is evidence that he had little prospect
of lining up necessary financing.
54. Bethlehem Steel buys Tull Industries as a white knight. Equity
$96 million; debt $10 million; premium $26 million. No selloffs. No
layoffs (there are layoffs at Bethlehem, but they are probably not from
Tull). Motives are probably Bethlehem's tax loss carryforwards and
Tull's profitability, as well as strategic. However, a year after the
acquisition, Tull is sold for $100 million to Inland Steel, the original
suitor, for $100 million. Inland Steel also has tax losses. This is largely
a tax deal with some strategic overtones.
55. Crane buys Unidynamics as a white knight. Equity $188.5 mil-
lion; debt $62.5 million; premium $99 million, and Crane loses $15
million in value. Some selloffs; amounts not reported. No layoffs.
Pension reversion of $18 million. Diversification. Crane is after defense
business of Unidynamics to augment its own and diversify away from
cyclical high-fixed-cost industries in which it mostly operates.
56. GAF unsuccessfully tries to buy Union Carbide. Equity $5.6
billion; debt $2.4 billion; premium $1.34 billion. Defensive response
of Carbide is leveraged recapitalization/share buyback. Selloffs (mostly
of sought after consumer brand names) are at least $4.5 billion. Present
value of savings from white-collar layoffs is $760 million to $2 billion.
Sanjai Bhagat, Andrei
Shleifer,
and Robert
Vishny 71
Remove $500 million in excess assets from the pension plan. Interest-
ingly, UC first buys back its equity at a high price and issues debt, but
several months later issues new equity at low prices and retires debt.
This has a massive cost to the company and shareholders who stayed
on. White-collar layoffs and pension transfer can probably explain the
premium, but selloffs of consumer business were probably profitable
as well.
57. Clayton-Dubilier buys Uniroyal Tire in an LBO. Equity $746
million; debt $255 million; premium $234 million. Also $300 million
underfunding in the pension plan. Selloffs of at least $935 million, but
the main tire division is kept. The tire division accounts for roughly
one-third of operating profit in the pre-acquisition firm. Present value
of savings from layoffs is $112 million-$240 million. A bustup, with
important savings from layoffs.
58. Mesa Partners (Pickens) unsuccessfully tries to buy Unocal.
Equity $9.2 billion; debt $1.3 billion; premium $2.21 billion. Defensive
response by Unocal is a discriminatory (against Pickens) leveraged
recapitalization/share buyback. Take on $4.5 billion in additional debt.
Three years later, still has over $3 billion in additional debt. Spin off
45 percent of oil and gas properties into master limited partnership to
avoid corporate tax. Selloffs of under $40 million. Layoffs of 600
employees, not counting early retirements, has the present value of
savings between $45 million and $120 million. With early retirements
and a hiring freeze, Forbes estimates $60 million per year before tax,
which is between $1 10 million and $300 million in present-value terms.
Capital expenditures cut $500 million per year, although this is quite
typical for the whole industry during this period. Debt tax shield along
with tax benefits from limited partnership can explain large part of
premium given that Unocal is not paying down its debt quickly through
asset selloffs, although the investment cut is probably also important.
59. Hanson Trust buys US Industries (response to MBO proposal
at very low price). Equity $511 million; debt $61 million; premium
$196 million. Selloffs of $178 million. Lays off 400 employees for
saving of $30 million-$80 million in present value terms. This is a
partial bustup with important savings from layoffs and possibly unused
debt capacity (tax shields).
60. APL Partnership buys Van Dusen Air. Equity $63 million; debt
72 Brookings
Papers:
Microeconomics
1990
$14 million; premium $14 million. No layoffs. Sell for $101.5 million
an aviation service division that is 80 percent of revenues. It is sold to
a strategic buyer. Keep remaining assets and put them in tax advantaged
limited partnership. This is a bustup-tax savings situation.
61. Marine Midland Bank buys Westchester Financial Services. Eq-
uity $102 million; premium $35 million. No selloffs. No layoffs. Stra-
tegic.
62. Electrolux buys White Consolidated. Equity $743 million; debt
$180 million; premium $164 million. No selloffs. No layoffs. Strategic
acquisition whereby Electrolux tries to enter U.S. appliance industry.
Comments
and Discussion
Comment by Gregg Jarrel: As George Stigler told me at the University
of Chicago, "The plural of anecdote is data. " This paper is an excellent
illustration of that statement. Several conclusions come out of the au-
thors' discussion that bear emphasis. The first is that loosening antitrust
policies was the basic regulatory spur to the merger-and-takeover boom
of the 1980s. I do not know if that is a valid conclusion, but it is the
conclusion of this paper, and I agree with it.
The second conclusion is that the theory free cash flow receives is
trivial. The authors concede that in the oil and gas industry the theory
has some explanatory power, but that is as far as it goes. Michael
Jensen's investigation of the industry motivated him to come up with
the theory in the first place. He will be disappointed in the conclusion
here, especially since in a recent article he argued that KKR is going
to take over the United States.'
Those are my broad comments, but I have some specific questions
and criticisms. Is the rate or the amount of selloffs directly related to
the burden of leverage? How precisely is that measured? How far did
the authors go to test that obvious empirical prediction?
I am also troubled by the measurement of premiums of the firms that
remain independent. That is not nearly as easy as measuring the pre-
miums of the takeover targets that are successfully bought out. There
is a big bag of money, and it can be measured and the measurement
expressed as a fraction of something. So, the researcher goes back in
time to find the right base.
1. Jensen (1988).
73
74 Brookings Papers. Microeconomics 1990
A firm that remains independent sometimes does so using a lever-
aged-recap transaction. But sometimes it defeats the offer by legal
means. I wonder whether the cases here represent the former or the
latter and whether the authors had any difficulties with them?
Also, did the authors find that the white knight cases had more
frequent negative returns to bidders relative to the target gains? That
finding comes out of the literature and is normally expressed as the
"winner's curse" hypothesis-that the bidders that win auction con-
tests for control do worse than the bidders that had negotiated agree-
ments. Normally white knights, almost by definition, win auctions, at
least as auctions are defined in this particular area.
A related question: are the conclusions sensitive to whether the au-
thors used the takeover premiums to table 2 or the total in table 2? The
total in table 2 takes account of the bidder returns, so the total takeover
premium is measured. That is one potential measure as a base. The
other measure nets out the return to the bidders.
It is my experience that bidder returns are measured with a great
deal more error than are target returns. The best a researcher can do is
hope to get large samples. With this particular approach in table 2 the
authors are asking a great deal from their statistical methods on the
bidder's side, to go in on a case-by-case basis and try to understand
what the market had to say about the bidder's reaction to a particular
event and then netting that out of the takeover premium.
One approach would be to say, "I do not know in any particular
case if the market has any idea what it is doing in revaluing the bidder;
I am going to assume the excess bidder return is zero and use the
takeover premiums and the targets as a base. " It is a substitute approach.
It would be nice to know if the conclusions remain valid under this
substitute approach.
A related question is what is the relative size in these particular
cases? How big are these bidders relative to the targets? There are a
lot of data problems on the bidders' side. There are a lot of bidders
that are not measurable because they are shell firms that do not really
trade. The bidders that are measurable sometimes are very large relative
to the target. A reader does not know what is being discounted with
the bidders relative to the target. Maybe the bidders have been looking
for firms that are like this target before, so the market has already
discounted much of the information. You know all the old problems.
Sanjai Bhagat,
Andrei
Shleifer,
and Robert
Vishny 75
Comment by Lawrence Summers: Sanjai Bhagat, Andrei Shleifer,
and Robert Vishny are to be congratulated on an impressive paper that
reveals a great deal of hard work. They have examined 62 takeovers,
studying carefully the journalistic accounts of what took place.
Because I harbor the suspicion that Business Week and the New York
Times may be more accurate in the coverage of corporate America than
they are in their coverage of the nation's economics departments, it is
with some scepticism that I think one has to take their results. Clearly
if one wants to study 62 companies, there is no alternative to the
approach the authors pursue, but my confidence in the results would
have been greatly enhanced if they were also able to report on actual
discussions and actual investigations with the people involved in some
of their transactions so as to corroborate the method that was used.
The paper presents information that is new, bearing in one way or
the other on three potential sources of value in takeovers: tax advan-
tages, layoffs, and selloffs. I want to discuss their analysis of each and
then offer some broader observations about the desirability or lack
thereof of hostile takeovers as an agent of change.
The authors attribute some importance to the effects of tax advan-
tages, stressing the role of debt and the interest deductability as a source
of value in premiums and noting that the analysis by Auerbach and
Reishus dismissed the effect of taxes.1 The analysis is largely suspect
because Auerbach and Reishus focused on only a single tax effect: the
ability to combine losses with profits. Everybody agrees that is not of
great importance. I suspect that Bhagat, Shleifer, and Vishny here
understate the importance of taxes for four reasons. First, contrary to
their assertion, it does not really matter whether the debt involved in
a takeover is paid back or is not paid back. Imagine that an acquisition
is financed with debt and that the assets are then sold to some company
that pays for them with cash out of its treasury. The government loses
the interest deductions that used to cost the government revenue. The
government now loses that revenue in the form of smaller interest
earnings from the treasury of a company that made the acquisition. It
is only if new equity is issued to finance acquisitions and is used to
1. Auerbach and Reishus (1988).
76 Brookings Papers: Microeconomics 1990
pay back debt that it is inappropriate to use the full permanent value
of the debt in measuring tax advantages.
Second, modern financial technology permits interest deductions that
substantially exceed true interest payments on high-technology bonds
and other sophisticated financial instruments that carry high yields. The
tax yield in the early years substantially exceeds the true interest cost.
To that extent, the tax benefit of the deduction is understated by the
authors' procedure of simply using the value of the debt. To think about
that, just take the example of a 20-year junk bond that carries an 18
percent interest rate but which, nonetheless, has a negligible probability
of defaulting in the first year. The true interest cost in the first year is
9 percent. The true deduction that is taken is a certain fraction of 18
percent. The value of the deduction exceeds the tax rate times the value
of the actual interest payment. To that extent the tax saving is under-
stated.
Third, the authors-because they cannot say much about it-slide
quickly over the General Utilities aspect and the step-up of bases, which
is regarded by practitioners as being of very considerable importance
in a number of transactions.
Finally, they make no mention of the avoidance of dividend tax
liabilities that takes place as a consequence of these repurchases and
in a period when capital gains are preferentially taxed. They are a device
for getting cash out of the corporate sector without paying the dividend
taxes.
Taking all of those features together, I suspect the taxes are subsi-
dizing these transactions to a rather greater extent than the authors
suggest.
The authors come next to the question of layoffs. That is a slightly
odd word for what is being discussed. It has to be acknowledged that
if one believed that these transactions were substantially increasing
efficiency, the way in which they would do that is by getting the same
amount done with fewer people working. So, the reallocation of re-
dundant employees would be another term that one could use.
The authors do not really pursue the idea that Andrei Shleifer and I
had discussed in our earlier work, which was that there was an element
of expropriation of stakeholders involved. The extent to which that is
true depends on whether the displaced employees get new jobs at the
same wage, or get new jobs at lower wages, or do not get new jobs at
Sanjai
Bhagat, Andrei
Shleifer,
and Robert Vishny 77
all. It would have been interesting to have had some hint or information
about the extent to which the efficiency gained from these layoffs was
matched by a loss to those who had been laid off.
Second, I suspect the authors understate the amount of reduced em-
ployment that is a consequence of these transactions. I take it that to
qualify in their sample a firm has to announce in the Wall Street Journal
that it is going to have a layoff. If it simply increases the generosity
of its early retirement plan or increases employment less rapidly than
it had intended, it does not qualify and probably will not get in the
Wall Street Journal with that news. And to that extent it will be missed
in their analysis.
Third, the authors' procedure for discounting the layoffs is odd. They
use a 10 percent discount rate, and they say, "Well, you can get to
that conclusion two ways. You can suppose that there is no inflation,
that wages will not grow at the rate of inflation; or you can suppose
that an 8 percent risk premium is appropriate." However risky General
Electric may be, I would assume that the cost saving from firing 1,000
people is roughly constant. It knows what that cost saving is, and it is
not uncertain about what the benefit from that cash flow will be in the
future. To that extent, the layoff should be discounted at a riskless rate.
Since we are talking about that, changing the discount rate from 10
percent to 3 percent triples the amount, and that would be enough to
substantially alter the authors' conclusions.
All of that said, what is on the other side? The authors acknowledge
that they do not do much with the fact that presumably even the em-
ployees that could be laid off were contributing something beforehand.
So, assuming that these employees were purely deadwood and all of
the labor costs were treated as an efficiency improvement is probably
not appropriate.
The authors also take no account of another possible expropriation,
namely growth in wages that is slower than would otherwise have taken
place, similarly for bargaining with suppliers and the like.
In the end there are substantial biases going in both directions, so it
is difficult to know which way one would want to correct the authors'
estimates. My own best guess would be that the value of the labor-
saving costs is substantially greater than the authors suggest. However,
there are probably also costs to not having that labor around any more,
which the authors do not take any account of.
78 Brookings Papers: Microeconomics 1990
The central conclusion of the paper, and by far the most interesting
part, is the analysis summarized in table 8. It is a commentary on both
table 8 and tables 1 through 7 that about 60 percent of the value added
of the paper is contained in table 8, which illustrates that a large fraction
of the assets that are involved in takeover transactions end up getting
sold to other public corporations in industries closely related to those
industries where the assets were already being deployed.
I think Gregg Jarrell was correct to suggest that the eclipse of the
public corporation perhaps has lasted about as long as the typical eclipse
lasts. The authors' analysis does suggest that these transactions do not
primarily represent disincorporation. And there is a great deal of an-
ecdotal evidence to suggest the the type of effect they are talking about
is correctly observed. I recall the CEO of one major bank observing
that "of course, our bank would be worth more broken up, but fortu-
nately banking regulations make a hostile acquisition impossible, so
we will remain a valuable organization for the foreseeable future."
Other chief operating officers are happy to explain in private that
they have engaged in the exercise of shopping around individual di-
visions of their company and that when they do that, the combined
value of individual divisions is two or three times the current value of
their company. I usually refrain from asking them what they think about
the market's implicit verdict on their managerial ability, sitting on top
of these valuable assets.
It does seem as if busting up assets and allocating them separately
to different people is a major source of value. That conclusion, I think,
has not received enough attention and is the major contribution of the
paper. The question one has to ask is: Why is this so? Why is it that
reallocating assets produces such substantial contributions, apparently,
or increases in value?
Let me just comment on four possible explanations. The first is what
I would call the " winner's curse" aspect. Imagine that the Metropolitan
Museum of Art could be put on sale in two different ways. In one case
the whole museum would be put on sale. In the other case the paintings
would be put on sale separately. I would imagine that the value fetched
would be substantially greater in the second case because the people
who liked a given painting would compete to buy it, and the total of
these individual sales would be greater than that of a block purchase.
So, too, if there are optimists about the steel industry or optimists about
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 79
the oil industry. Selling steel assets and oil assets separately would
increase the total value of the assets that can be realized. That may be
why conglomerates carry such low values. There is no pure-play effect.
Nobody is going to bet that everything IT&T is doing is going to be
good, but one can bet that car rental will be good or that some other
division of the corporation will be good.
There is, however, an interesting conflict because it is perfectly
possible that maximizing value by finding optimistic buyers for each
class of assets will lead to a different conclusion than the conclusion
of who can manage the assets best. It may well be that breakups that
allow optimists about each individual part to own that part will increase
the value at which the assets can be sold, even though people who buy
them may not be those who are best able to operate them. Think about
the sale of the paintings from the Metropolitan Museum. Some of the
people who buy could be very poor at appreciating art.
A second hypothesis suggested by the conglomerate experience is
that some of these transactions may involve transfer of earnings from
situations in which they carry a high price-earnings ratio to situations
where they carry a low ratio. I have no evidence on this point, but it
does have the virtue that one could potentially get evidence on it.
Supporting the conglomerate boom, in significant part, was that these
conglomerates had high price-earnings ratios and other companies had
low price-earnings ratios. When the conglomerates purchased compa-
nies and got more earnings, those extra earnings carried the same old
price-earnings ratio that the conglomerate had had before. It is the same
factor that explains why any American company that can will have a
subsidiary issue equity in Japan right now.
Perhaps what is going on is that these transactions are taking the
form of brokering earnings from places where they have a low price-
earnings ratio to places where they have a high ratio. Those differences
in price-earnings ratios may reflect rational expectations about the abil-
ity of management to increase earnings. They may simply reflect market
conventions for valuing different companies that are slow to change.
The same CEO who admitted to me that if his company sold its divisions
separately the total price would be two or three times the company's
current value also confessed to be spending some amount of his time
endeavoring to speak with Standard and Poor about getting his company
reclassified from one industry to another because the industry he wanted
80 Brookings Papers: Microeconomics 1990
to be in had a price-earnings ratio that was 30 percent greater than the
industry that he was now in. I think it is unlikely that that was an
entirely futile exercise, if he could have pulled it off. And, to that
extent, reallocations may be in order.
A third reason that reallocating assets produces increases in value is
that the transactions are all coming at the hands of the people who put
up the debt. Any idiot who called himself a condo minimum developer
in Boston in the four years before a year and a half ago, or who called
himself any kind of real estate developer and bought properties and
painted them or did almost anything, made a spectacular rate of return.
During the period when these takeovers yielded very substantial value,
the stock market roughly tripled. If one simply bought the S & P 500
stocks on 25 percent margins, one would have seen the money increase
tenfold. Not even the KKR buyout fund has turned in such a perfor-
mance.
I suspect that a large part of the success in buying divisions and
selling those divisions separately has been partly the result of getting
on a fast-moving escalator. But if the escalator slows down, who loses?
It is the people who put up the debt and will not be in a position to
collect. There is substantial evidence that, at least in the last six months,
people who had put up debt in the preceding year are finding that action
a very expensive mistake. So it may well be that a source of value in
these transactions is that people have been allowed to borrow on margin
to buy stock at an excessively favorable rate.
Finally, there is the question of market power, and here I would just
highlight an observation that Michael Salinger made years ago. If you
really think that firms have roughly constant returns of scale and that
a little bit of market power will translate into a great deal of value,
then let me give a simple example. It is not wildly wrong to say that
the typical firm has a market value roughly equal to the value of its
sales. Imagine that its elasticity of demand changes from five to six.
That is not a change of spectacular magnitude. The change in the markup
will be about 3.33 percent of sales. This percentage, for most com-
panies, is a substantial share-perhaps a third, perhaps more-of prof-
its, so the implied increase in market value would then be one-third to
one-half. Thus even very small changes in the relevant elasticity of
demand can account for very substantial increases in market value. And
that lends some credence to the monopoly-power interpretation.
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 81
I would be very surprised if a closer investigation than the one the
authors carry out would suggest that a large fraction of the divisional
sales that they report would in fact have been stopped by the antitrust
authorities at the average moment from between 1950 and 1980. The
suggestion that antitrust relaxation as a major force behind what has
happened is unsupported by evidence does not really accord with my
intuition.
Finally, about the public-policy tilt that all of this suggests: econ-
omists generally believe that interference with the markets should not
be based on how good or how bad some trading commodity is but only
on the external consequences of various activities. So, I looked at the
analysis in this paper with a view to trying to determine what the external
benefits and costs were, as distinct from the benefits and costs that were
realized directly by the parties to these transactions. As best I could
judge, whatever efficiency consequences there were in the form of
economies of scale would be captured by the parties to takeover trans-
actions. On the other hand, potential other sources of value-lost tax
revenue, reductions in the value of human capital, sales people who
were not smart and paid the wrong price-would all be imposing neg-
ative, external costs to those who were not involved in the transaction.
And to that extent and to the extent that the authors have not suggested
any important external benefits to these transactions, it seems to me
that the analysis pointed in the direction of a policy tilt toward reducing
the number of hostile takeovers.
General Discussion: An issue that particularly interested the par-
ticipants was the increased postmerger profits that seemed to result from
a horizontal acquisition. Lawrence White alleged that it is often difficult
to tell whether these higher profits come from increased market power
or decreased costs. This was a serious issue for many of the participants
because of the implications for antitrust enforcement. According to
White the authors claim that antitrust enforcement virtually stopped in
the 1980s. If these increased profits are, in fact, coming from enhanced
market power, this brings into question the policy of reduced antitrust
enforcement. If, on the other hand, they are coming from decreased
costs-from efficiency gains-then the new antitrust policies might be
performing as they were intended to. White claimed that the 1982 and
1984 merger guidelines rationalized antitrust enforcement, forcing it to
82 Brookings Papers: Microeconoinics 1990
focus on where the threats of market power really came from. Andrei
Shleifer said that the data used for the paper did not allow for a dis-
tinction to be made between monopoly power and efficiency gains.
Michael Whinston wanted more information about what features of
hostile takeovers made them different from friendly ones, since the
paper pointed out that, by and large, both occur for strategic reasons.
He was interested in a more substantial definition of the term "strategic
takeover." According to Whinston, the authors apply this term to take-
overs occurring for seemingly different reasons: monopolization, ver-
tical integration, R&D advantages, and so forth. Shleiffer replied by
saying that the overwhelming majority of things that the paper defines
as "strategic" involve horizontal moves, with only one or two con-
cerned with vertical integration.
Whinston was also interested in the industrial organization of the
brokerage market. He wondered if the final buyers of the firms involved
in takeovers made money on the transaction, or if the brokers were able
to successfully extract all of the rents. He suggested that the type of
strategic acquisition-that is, monopolization or strategic advantage-
might have an effect on the amount of the rents that brokers would be
able to extract.
Oliver Williamson said that the evidence brought out in the paper
supports the theory of complacent management. He said that it is dif-
ficult for management to "run a tight ship" for an extended period of
time. When aggressive capital markets see this, a takeover occurs, and
belt-tightening is instituted, which accounts for the takeover premiums.
Richard Caves said that the takeovers appear to be a mechanism by
which businesses are transferred from corporations with which they
have a bad fit to corporations that can better manage them. Shleiffer
agreed with this and said that the evidence presented in the paper was
consistent with the idea of improving efficiency by transferring assets
to better managers in the same line of business as those assets.
Margaret Blair said that more attention must be given to free-cash-
flow theory. She asserted that free cash flow was tied to the authors'
claim that the abrogation of implicit employment contracts is a source
of takeover premiums. Blair said breaking such implicit contracts would
be more likely to occur when those contracts become inconsistent with
profit maximization, that is, when they represent inappropriate use of
free cash flow.
Sanjai Bhagat, Andrei Shleifer, and Robert Vishny 83
Martin Baily asked if the journalistic accounts used in the paper
constituted data. Sanjai Bhagat responded by saying that they had used
all publicly available published reports on layoffs, selloffs, and so forth.
These included newspapers, popular and trade magazines, proxy state-
ments, annual reports, lOK's, and other sources.
References
Asquith, Paul, and Thierry
A. Wizman. 1989. "Event Risk, Wealth Redis-
tribution
and the Return
to Existing Bondholders
in Corporate
Buyouts."
Mimeo, Massachusetts
Institute of Technology (February).
Auerbach,
Alan, and David Reishus. 1988. "The Effects of Taxation
on the
Merger Decision." In Corporate Takeovers:
Causes and Consequences,
edited by A. J. Auerbach.
University
of Chicago Press.
Bhide, Amar. 1989. "The Causes and Consequences
of Hostile Takeovers."
Continental
Bank
Journal of Applied Corporate
Finance 2:36-59.
Bradley, Michael, Anand Desai, and E. Han Kim. 1988. "Synergistic
Gains
from Corporate
Acquisitions and Their Division between the Stockholders
of Target
and Acquiring Firms." Journal
of Finanicial
Economics
21 (May):
3-40.
Brown, Charles, and James L. Medoff. 1988. "The Impact
of Firm Acqui-
sitions on Labor." In Corporate Takeovers. Causes and Consequences,
edited by A. J. Auerbach.
University
of Chicago Press.
Bureau of Labor Statistics. 1987. Employment
and Earnings. Washington:
Department of Labor. January.
Eckbo, B. Espen. 1983. "Horizontal Mergers, Collusion, and Stockholder
Wealth." Journal of Financial Economics 11 (April):241-74.
Jensen, Michael
C. 1984. "Takeovers:
Folklore
and Science." Harvard
Busi-
ness Review 84 (November-December): 109-21.
. 1986. "Agency Costs of Free Cash Flow, Corporate
Finance, and
Takeovers." American
Economic
Review
76:323-29.
. 1988. "Takeovers: Their Causes and Consequences." Journal of
Economic
Perspectives 2 (Winter):21-48.
Kaplan, Steven N. 1989. "Management
Buyotuts:
Evidence on Taxes as a
Source of Value." Journal of Finanice 44 (July):611-32.
. 1990. "The
Effects
of Management
Buyouts
on Operations
and
Value."
Journal of Finanicial
Economics
(forthcoming).
Kaplan, Steven N., and Michael S. Weisbach. 1990. "Acquisitions
and Di-
vestitures:
What is Divested and How Much Does the Market
Anticipate?"
Mimeo, University
of Chicago.
84 Brookings Papers: Microeconomics 1990
LeBaron, Dean, and Lawrence Speidell. 1987. "Why Are the Parts Worth
More than the Whole? 'Chop Shop,' A Corporate Valuation Model." In The
Merger Boom. Federal Reserve Bank of Boston.
Lichtenberg, Frank, and Donald Siegel. 1988. "The Effect of Takeovers on
the Employment and Wages of Central Office and Other Personnel." Mimeo,
Columbia University.
Morck, Randall, Andrei Shleifer, and Robert W. Vishny. 1988. "Character-
istics of Targets of Hostile and Friendly Takeovers." In Corporate Take-
overs: Causes and Consequences, edited by A. J. Auerbach. University of
Chicago Press.
. 1989. "Alternative Mechanisms for Corporate Control." American
Economic Review 79 (September):842-52.
. 1990. "Do Managerial Objectives Drive Bad Acquisitions?" Journal
of Finance 45 (March):31-48.
Nathan, Felicia. 1987. "Analyzing Employers' Compensation Costs." Monthly
Labor Review (October):3- 11.
Pontiff, Jeffrey, Andrei Shleifer, and Michael Weisbach. 1989. "Reversions
of Excess Pension Assets after Takeovers. " Mimeo, University of Rochester.
Porter, Michael E. 1987. "From Competitive Advantage to Corporate Strat-
egy. " Harvard Business Review, 65 (May-June):43-59.
Ravenscraft, David J., and F. M. Scherer. 1987. Mergers, Selloffs and Eco-
nomic Efficiency. Brookings.
Rosett, Joshua. 1989. "Do Breaches Explain Takeover Premiums? The Evi-
dence on Union Wages." Mimeo, National Bureau of Economic Research
(November).
Servaes, Henri. 1989. "Tobin's Q, Agency Costs, and Corporate Control."
Mimeo, Center for Research in Security Prices, University of Chicago.
Shleifer, Andrei, and Lawrence H. Summers. 1988. "Breach of Trust in Hostile
Takeovers." In Corporate Takeovers: Causes and Consequences, edited by
A. J. Auerbach. University of Chicago Press.
Stigler, George J. 1950. "Monopoly and Oligopoly by Merger." American
Economic Review (Papers and Proceedings, 1951) 40 (May):23-34.
Stillman, Robert. 1983. "Examining Antitrust Policy towards Horizontal
Mergers." Journal of Financial Economics 11 (April):225-40.
