the development of the takeover auction process: the...
TRANSCRIPT
The Development of the Takeover Auction Process:
The Evolution of Property Rights in the Modern Wild West*
Tingting Liu
Heider College of Business
Creighton University
2500 California Plaza, Omaha, NE 68178
J. Harold Mulherin
Georgia Athletic Association Professor
Terry College of Business
University of Georgia
Athens, GA 30602
William O. Brown Jr.
Department of Accounting and Finance
Bryan School of Business and Economics
University of North Carolina at Greensboro
Greensboro, NC 27402
March 2018
*We thank Julian Atanassov, Leonce Bargeron, Eric De Bodt, Jonathan M. Karpoff, seminar participants
at the Midwest Finance Association 2018 Annual Meeting, Creighton University, University of Georgia,
and University of Nebraska–Lincoln for helpful comments.
The Development of the Takeover Auction Process:
The Evolution of Property Rights in the Modern Wild West
Abstract
We study the interaction between the legal system and the takeover auction process during the
1981 to 2015 time period. We associate the strengthening of the property rights of target boards
after the 1989 Time Inc. decision with fundamental changes in the takeover auction process.
Following the 1989 decision, takeover auctions have moved from public and often hostile battles
to a more behind the scenes, underground process where target boards control both the number
of bidders and the flow of information. Target boards are more likely to initiate the auction
themselves and the time between deal initiation (in private) and public deal announcement has
significantly lengthened. Accounting for the increasing length of the deal initiation process,
target premiums have significantly risen since 1989. Our property rights interpretation contrasts
with the entrenchment story that argues that the decline in hostile takeovers has protected poorly
performing management at the expense of target shareholders. Our results are consistent with the
fundamental proposition of Coase (1959, 1960) that with well-defined property rights, resources
flow to their highest valued use.
1
Introduction
Ronald Coase (1959, 1960) insightfully noted that with well-defined property rights, resources
flow to their highest valued use. In the Coasian framework, property rights are an essential
prelude to market transactions because they identify who to contact to use a given resource.
Well-defined property rights and common law precedent also reduce costly legal disputes. But
given the delimitation of property rights, the transfer and use of rights to assets takes place via
market transactions. Using a cave as an example, Coase (1959) stated that the legal granting of
the property right either to (a) the owner of the land at the entrance to the cave or (b) the owner
of the land over surface of the cave will be irrelevant to the highest-valued use of the cave
whether it be, say, to store bank records or to grow mushrooms.
The property rights framework of Coase (1959, 1960) inspired analysis of the emergence
and adaptation of property rights to economic, political and technological change. Demsetz
(1967) applied the property rights paradigm to the movement from communal to private property
in reaction to the arrival of European fur traders on the new continent. Umbeck (1981) analyzed
the formation of property rights during the California gold rush. Anderson and Hill (1975)
studied how barbed wire and other innovations were associated with the evolution of property
rights in the American west.
In this paper, we apply the property rights paradigm to a more recent setting: the
corporate takeover market in the United States. Observers such as Jensen (1993) argue that the
major corporate control activity beginning in the 1980s in the United States was tied to political,
economic and technological shocks that shook the existing structure of American industry.
Indeed, innovations in junk bond financing and the relaxation of antitrust laws made large firms
the object of corporate takeovers for the first time. Hence, major corporations found “Barbarians
2
at the Gate” (Burrough and Helyar (1989)) in a modern Wild West setting. In our analysis, we
document how the takeover auction process underwent fundamental changes as an outgrowth of
the sea change in the market for corporate control in the 1980s.
This increased susceptibility of large firms to takeovers was followed by major
alterations in the legal and governance setting related to the corporate takeover market. Lobbied
by the large firms in their states, legislators passed antitakeover laws. Major corporations
themselves instituted poison pills and other takeover hurdles. These legal and governance
changes induced lawsuits between bidders and targets to determine the degree to which the
boards of target firms could “defend the corporate bastion” and repel the “barbarians.” Alchian
(1965) defines a property right as protection against the use of a resource against the owner’s
will. Using Alchian’s terminology, the courts in Delaware and other states were faced with
determining the appropriate property rights to grant to target boards in the new takeover
environment: when could a target board just say no to a prospective bidder? While the legal
process evolved iteratively and often did not please academic commentators such as Gilson
(2001), the courts in Delaware ultimately clarified the property rights of the boards of directors
of target firms.
A focal case amid this changing takeover environment was the 1986 Revlon decision
where the court ruled that once a target firm is up for sale, the duty of the target board changes
from a defender of the corporate bastion to an auctioneer. However, the Revlon decision did not
clearly define what an auction was. This ambiguity induced further lawsuits. But through an
evolutionary litigation process, as in the model of Rubin (1977), common law cases clarified
when and whether target boards must conduct an auction, the provision of information during an
auction, and the allowability of takeover hurdles during the auction process. By 1989, in the
3
Time Inc. decision, the courts had clarified the property rights during corporate control
transactions by moving to broad deference via the business judgement rule that implied relatively
strong property rights to target boards.
In our empirical work, we compare and contrast the takeover auction process in the
Revlon period of 1986 to 1989 with later time periods following the Time Inc. decision. We
document the following fundamental changes in takeover auctions:
After the Time Inc. decision, takeover auctions have moved behind the scenes, or
underground, where a significant portion of the bidding process occurs prior to public
revelation of the takeover.
In conjunction with this movement to underground auctions, target boards are much
more likely to initiate the takeover process rather than waiting for an initial bid.
The time period between the initiation of a deal in private and the public
announcement of a transaction has lengthened significantly.
Accounting for this longer period between deal initiation and deal announcement,
target premiums have increased significantly since the Time Inc. decision.
Our results have important implications for research on corporate takeovers. A number of recent
papers such as Cain, McKeon and Solomon (2017) ask the question: Do takeover laws and the
legal setting matter? By the term “matter,” this research tends to ask whether takeover laws and
the legal setting impact hostile takeovers. Yet our analysis suggests that the emphasis on hostile
takeovers is too narrow and points to more fundamental changes in how takeover auctions are
being conducted. Consistent with our property rights story, target boards are more in control of
the takeover auction process and this control has benefited target shareholders.
4
A related thread of research such as Bebchuk, Coates and Subramanian (2002) laments
the apparent negative impact of takeover hurdles such as poison pills and staggered boards on
hostile takeovers. The concern of these authors is that the decline in hostile takeovers has
weakened the monitoring of shirking target management. Such authors call for a lessening of the
property rights of target boards. One problem with this line of research is that Schwert (2000)
fails to find systematic performance differences between targets of hostile and friendly targets.
Schwert (2000) instead argues that the choice of a hostile takeover is a tactical decision in the
bargaining between targets and bidders. Consistent with Schwert’s (2000) argument, we find that
the heyday of hostile takeovers in our sample was the period between 1986 to 1989 when the
Revlon decision created ambiguous property rights that induced significant lawsuits by bidding
firms as part of a hostile strategy to force an auction of the target. After the 1989 Time Inc.
decision clarified the property rights of target boards, bidder lawsuits and hostile deals both
declined significantly. But target premiums increased.
Our work is also related to some of the seminal research of the effect of the legal setting
on target premiums. Jarrell and Bradley (1980) find that the delay caused by the provisions of the
Williams Act is associated with higher target premiums. In our more recent sample, we find that
the lengthening of the takeover process between deal initiation and deal announcement after the
Time Inc. decision is also associated with higher target premiums. Indeed, our results suggest
that the now-standard event study framework that follows Schwert (2000) and applies a common
(-63, +126) window to all deals in all periods will underestimate the target premiums for deals
done after the 1989 Time Inc. decision.
The paper proceeds as follows. Section 1 defines a takeover auction, surveys how
takeover auctions have been modeled in law and financial economics, and sketches the treatment
5
of takeover auctions in the courts. Section 2 describes our sample. Section 3 presents evidence of
bidder lawsuits over time in our sample to illustrate the intertemporal change in the property
rights of target boards. Section 4 presents evidence on how the changing property rights have
affected the takeover auction process. Section 5 provides related evidence on how the initiation
of takeovers by target boards has changed over time. Section 6 reports evidence on target
premiums. Section 7 performs robustness analysis using some alternative samples. The final
section offers a summary and conclusion.
1. What is a Takeover Auction?
In our empirical analysis, we will report the fundamental changes that have occurred in the
takeover auction process over time. To provide a benchmark for our analysis, Part A of this
section sketches the key components of the takeover auction process. We then use this
framework to describe how takeover auctions have been modelled in the legal and financial
economics literature (Part B) and how takeover auctions have been treated in the courts,
especially in conjunction with the 1986 Revlon decision (Part C). This background sets the stage
for our empirical analysis.
A. Schematic of the Takeover Auction Process
Table 1 provides a schematic of the takeover auction process. Information to describe takeover
auctions is taken from the conceptual model of Hansen (2001) and the empirical work in Boone
and Mulherin (2007, 2009) which studies deals in the 1990s. As we will report below in our
empirical work, the duration of the entire takeover auction process and the specific steps
sketched in Table 1 vary over our sample period of 1981 to 2015.
6
A given takeover auction can be initiated by either the target itself or a potential bidder.
This initiation is usually done behind the scenes or “underground” via private communication
and interaction between the target and prospective bidders well before the news of a takeover
becomes public. Targets often internally initiate an auction by considering their strategic
alternatives and, in consultation with their legal and financial advisors, deciding how many
potential bidders to contact. Alternatively, a bidding firm may approach with a preliminary offer
which may follow the acquisition of a toehold stake in the target. The target then reacts by either
agreeing to negotiate with the bidder or resisting the offer, establishing takeover defenses and
considering other bidders. The bidder may make counter moves and engage in lawsuits to enjoin
any takeover hurdles instigated by the target. The filing of lawsuits by the bidder can move the
auction from a behind the scenes interaction to a battle in the public sphere.
Once a deal is initiated, a crucial aspect of a takeover auction is the receipt of information
by potential bidders and an iterative bidding process. As emphasized by Hansen (2001), target
firms conducting an auction tend to restrict both the number of bidders and the flow of
information to those bidders. Selected bidders receive confidential information, but only after
signing a standstill contract whereby the bidder agrees not to make an offer without approval
from the target board of directors.
A subset of the bidders that sign confidentiality/standstill agreements make non-binding
indications of interest. This information is used by the target firms and their legal and financial
advisors to invite a subset of bidders for further due diligence. The next step is formal bids with
the ultimate outcome being a takeover agreement with the high bidder that includes terms such
as the deal price and method of payment as well as deal protection devices such as terminations
fees in the event another bidder offers an even higher price before the deal is completed.
7
In the final stage of the auction process, the takeover agreement is publicly announced,
subject to shareholder and regulatory approval. There is a potential for a higher bid in the period
between deal announcement and deal completion. In a hostile bid with a white knight, there may
be protracted bidding once an initial bid is announced. The deal is completed with the acquisition
of the target by the winning bidder.
B. Conceptual and Theoretical Analysis of Takeover Auctions
With the schematic in Table 1 as a framework, we next discuss the depiction of takeover
auctions in the legal and financial economics literature. These models differ in the definition of
an auction, the assumptions regarding the property rights of the target board, and the information
costs inherent in the takeover auction process. As summarized in Table 2, the questions
addressed include: whether and when to auction, the interaction of preemptive bids and access to
information, and the effect of hurdles such as poison pills on takeover auctions.
B.1. Whether and When to Auction
As reported in Table 2, some of the earliest analysis of takeover auctions was a rather
diametric debate in the legal literature between Easterbrook and Fischel (1981) and Bebchuk
(1982). The analysis focused on one specific step illustrated in Table 1: how a target board
should react to an unsolicited bid. Easterbrook and Fischel (1981) argued for a mandatory
passivity rule where target boards would refrain from auctioning a firm that was the subject of a
takeover bid. Their proffered reason was that the use of an auction would compromise incentives
for the initial bidder to incur search costs to identify undervalued targets. In effect, the
proscription of Easterbrook and Fischel (1981) rules out hostile takeovers, as target boards never
reject a bid in their setting.
8
Bebchuk (1982) countered with a mandatory auction rule where target boards would
always attempt to auction a firm faced with an initial takeover bid. Bebchuk (1982) argued that
an auction would find the highest bidder and thereby bring the greatest return to target
shareholders. Moreover, the initial bidder could be compensated for initial search by obtaining a
pre-bid toehold in the target firm.
Critiques of this early debate on takeover auctions included the implicit assumptions
regarding the property rights of the target board as well as the impact of search costs on the
chosen bidding process. Haddock, Macey and McChesney (1987) point out that the Easterbrok
and Fischel (1981) rule implies extremely weak property rights that impose a first possession
rule and a race to capture the target firm. The Bebchuk (1982) mandatory auction rule places
property rights in the hands of individual shareholders which creates a communal property
problem (DeAngelo and Rice (1983)). Macey (1990) notes that both Easterbrook and Fischel
(1981) and Bebchuk (1982) offer rather rudimentary depictions of takeover auctions that fail to
delve into the costs and benefits of the complex bidding process depicted in Table 1 and thereby
ignore the fact that the appropriate sales procedure in a given takeover can vary with target and
deal characteristics.
Much of Macey’s (1990) critique of the early auction analysis was based on the model of
takeover bidding in French and McCormick (1984). That model adopts a setting envisioned by
both Easterbrook and Fischel (1981) and Bebchuk (1982) where potential bidders bear costs to
evaluate a potential target, but delves much deeper into the implications of search costs. In a very
Coasian (1960) result, French and McCormick (1984) find that in equilibrium the target firms
ultimately bear the cost of search incurred by any bidders in the price they receive. Hence, a
deeper auction as measured by a more extensive canvassing of potential bidders brings both costs
9
and benefits and target firms have the incentive to economize on the number of bidders invited to
a particular auction. French and McCormick (1984) suggest that target firms can impact the
number of bidders by auction rules such as entry fees and can affect search costs by themselves
providing some information to selected bidders. Of course, the ability of the target firm to set the
auction rules implicitly assumes strong property rights by the target board.
B.2. Preemptive Bids and Access to Information
While French and McCormick (1984) conclude that target firms have the incentive to control the
number of bidders to economize on search costs, some bidders may choose to “jump in line” by
making a preemptive bid which is often accompanied by a toehold in the target firm. Bulow,
Huang and Klemperer (1999) model this scenario and conclude that a toehold can give an initial
bidder an advantage in a takeover auction. Hence, a toehold may not simply provide
compensation for search costs but may also lead to a lower price paid for the target. Hence, like
French and McCormick (1984), Bulow, Huang and Klemperer (1999) state that target firms may
want to adapt the auction rules to mitigate such toehold advantages. Moreover, Berkovitch,
Bradley and Khanna (1989) point out that the winning bidder can be compensated for search via
deal protection devices such as termination fees and lock-up options that lessen the reason for the
acquisition of a toehold by the bidder.
Fishman (1988) models the target and bidder tactics related to preemptive bids in a
takeover auction. A central result in Fishman’s (1988) model is that the elimination of
preemptive bids raises the target’s revenue. How might a target firm prevent such preemptive
bidder tactics? Fishman (1988) states that target firms could withhold confidential information
until all bidders have signed standstill agreements. These results and insights of Fishman (1988)
10
are quite prescient and serve well to explain the relatively complex set of steps depicted in the
due diligence stage of takeover auctions reported in Table 1.
B.3 Takeover Hurdles
Much of the conceptual and theoretical research on takeover auctions addresses the role of the
potential takeover hurdles imposed by state and federal law as well as firm specific devices such
as poison pills. Easterbrook and Fischel (1981) and Bebchuk (1982) both argue adamantly
against takeover hurdles. Easterbrook and Fischel (1981) view takeover defenses by target
management as prima facie evidence of entrenchment. Bebchuk (1982) opposes any device that
seems to dampen a full-fledged auction. The formal model of takeover auctions by Bulow and
Klemperer (1996) also concludes that impediments to auctions reflect agency costs of target
management.
However, several commentators express less opposition to takeover hurdles. DeAngelo
and Rice (1983) provide a model in which takeover hurdles such as shark repellents can bolster
the property rights of the target board and thereby act in target shareholder interests. Here, a
property right is the protection against the use of target firm resources by unsolicited bidders
against the will of the target board (Alchian (1965)). The benefit of greater property rights for the
target board is attained by centralizing the decision making when faced with a takeover bid.
Related empirical research suggests some of the benefits that takeover hurdles can have
for target shareholders. Jarrell and Bradley (1980) report that the delay enabled by the Williams
Act was associated with higher target premiums. Jarrell (1985) finds that target resistance to an
initial bid via lawsuits against a hostile bidder induces auctions by multiple bidders. Comment
and Schwert (1995) conclude that poison pills improve the bargaining power of target boards.
11
However, more recent analysis by Bebchuk, Coates and Subrmanian (2002) argues that poison
pills in conjunction with staggered boards of directors provide a lethal combination against a
viable takeover auction.
C. Legal Cases Related to Takeover Auctions
We next discuss how takeover auctions have been treated in the courts. In particular, we report
how the auction requirement of the 1986 Revlon decision induced further legal cases to pinpoint
what exactly a takeover auction was and what actions of the target board were allowed during the
bidding process. For legal analysis directly pertinent to the 1986 Revlon decision, see Herzel and
Shepro (1989) and Gilson and Kraakman (1990). For a broader overview, see Bainbridge (2006).
The cases that we highlight are presented in Table 3. Note that all the cases involve target
firms in our sample. Full citations to the legal cases follow the reference section of the paper.
The seminal Mesa decision in 1985 posed the duties of the target board of directors as a
question of process that was evaluated in the context of the business judgement rule. But the
Revlon decision the next year famously stated that in cases where the target firm is up for sale,
the target board’s duties switch from defense of the corporate bastion to the role of an auctioneer.
However, like much of the conceptual and theoretical research discussed previously, the Revlon
court did not clearly define exactly what an auction was. This lack of clarity raised several
questions such as (i) can a preemptive bidder force an auction of the target, (ii) to what extent do
the Revlon duties require a level playing field across all bidders and (iii) what role do takeover
hurdles such as poison pills play in the course of an auction? Such questions induced further
lawsuits by hostile bidders.
12
Arguing for a level playing field, many hostile bidders sued the target to get access to
confidential information. In the 1988 deal involving Koppers, the hostile bidder, BNS, argued
that it should be given access to confidential information about the target. Similarly, in the 1988
J.P. Stevens deal, a hostile bidder sued to gain access to confidential information without having
to contractually commit to cease its hostility by signing a standstill agreement. In such cases
related to information access, the courts tended to defer to the target board and not force
revelation of information to hostile bidders and attested to the validity of standstill agreements.
(See, for example, the discussion in Rhodes (1991) and Kidd (2003).) Hence, the courts
empowered the target board with the property rights to information about the target firm and
thereby enabled the target board to control access to bidding along the lines suggested by the
model of Fishman (1988).
The role of poison pills within the takeover auction process was greeted with less clarity
immediately following the 1986 Revlon decision. This created uncertainty in the property rights
of target boards and resulted in numerous lawsuits by hostile bidders. See, for example, Gayle
(1989) and Yablon (1989).
Following Revlon, hostile bidders often sued to enjoin a target’s poison pill as interfering
with the takeover auction. In some cases, the courts considered poison pills to be viewed as part
of the overall takeover auction process and rejected the bidder lawsuits. Facet Enterprises in
1988 stated that a poison pill provided a gavel to run an auction. Similarly, in the 1988 deal
involving Federated Department Stores, the court ruled that the target board could leave the
poison pill in place for the life of the takeover auction. However, in deals such as Moore
McCormack Resources in 1988 and Pillsbury in 1989, the courts concluded that the poison pills
interfered with the auction process and forced the target boards to redeem the pills.
13
Much of the uncertainty created by the auction duties of the 1986 Revlon decision was
resolved by the 1989 case involving Time Inc. Time had proposed a merger with Warner but
then became the object of a hostile bid by Paramount. The court ruled that Time could just say
no to Paramount, concluding that the directors of the target firm were not obliged to abandon a
deliberate corporate strategy. In effect, a hostile bidder could not force an auction on a target
board. As noted by Velasco (2002, p.390), “Although there was a point at which it appeared that
the Delaware courts would be willing to mandate the redemption of the poison pill in the face of
a non-coercive tender offer, Paramount Communications Inc. v. Time Inc. appears to have
eliminated any such hopes.” The Time decision clarified the relatively strong property rights held
by the target board of directors. Indeed, the “just say no” ruling corresponds directly to Alchian’s
(1965) concept that a property right enables protection from actions by others that are against the
will of the possessor of the property right.
The 1995 case involving Wallace Computer Services appears to have continued the
strong property rights of the target board. Bebchuk, Coates and Subramanian (2002) emphasize
this legal decision as locking in the power of boards via a combination of poison pills and
staggered boards. Lipton (2005), however, suggests that there might be some weakening of target
board property rights wrought by the federal mandates of the Sarbanes Oxley Act in 2002.
D. Summary and Motivation for Research
The material presented in this section can be summarized as follows. The takeover auction
process depicted in Table 1 is quite complex. The conceptual and theoretical analysis sketched in
Table 2 captures this complexity in varying degrees and raises issues on the impact of takeover
hurdles such as poison pills on the depth of takeover auctions. The legal cases reviewed in Table
3 indicate that the property rights to target boards during takeover auctions have varied over
14
time, with the 1986 Revlon decision creating uncertainty related to the board duties in takeover
auctions and the 1989 Time Inc. decision clarifying the deference of the courts to the target
board when considering strategic decisions.
The variation in property rights over time motivates our analysis. We aim to study the
impact of the changing legal setting on the depth and nature of takeover auctions over time. Our
data covers the 1981 to the present. The cases reviewed in Table 3 suggest the following
apportionment of sub-periods of analysis:
Pre-Revlon: 1981 to 1985
Revlon: 1986 to 1989
Post-Time Inc: 1990-1995
Post-Wallace Computer Services: 1996-2001
Post-Sarbanes Oxley: 2002 to 2015
Our analysis of the changing takeover auction process over time will also jointly compare the
three post-1990 time sub-periods to the Revlon period of 1986 to 1989. Given the large number
of court cases and the complicated nature of the legal decisions, one can argue about the extent to
which any one case impacted property rights or whether and when all legal uncertainty was
resolved. However, the lack of 100 percent definitive breakpoints for the allocation of property
rights only serves to bias the results of documenting measurable differences across the
designated sub-periods.
2. Sample Formation and Description
In our analysis, we start with a set of major, publicly traded U.S. corporations at the beginning of
the 1980s. We then determine the firms within this base set that become the object of a takeover
15
between the 1980s and the present. We use this sample of takeover activity to study the changes
in the takeover auction process over time.
As our base set of firms, we start with 1,064 firms that were listed on the Value Line
Investment Survey in the 4th quarter of 1981. This is the same set of firms that formed the base
for Mitchell and Mulherin (1996) in their analysis of takeover activity in the 1980s. As noted by
Mitchell and Mulherin (1996), this set of firms represented 60 percent of the value of listings on
the NYSE, AMEX, and NASDAQ at year-end 1981.
We use this sample from Value Line, which is comprised of relatively large U.S.
corporations, for several reasons. For one, the use of Value Line emphasizes large takeover
targets and captures the sea change in the takeover market beginning in the 1980s as documented
by Mitchell and Mulherin (1996). These same large firms often were the prime lobbyists for
changes in state antitakeover laws (Romano (1988). Relatedly, consistent with the legal model of
Rubin (1977), the large corporations were directly in the litigation tied to corporate takeovers.
We note that while this paper emphasizes the relatively large takeover targets derived from
Value Line, we find and report below that our results are robust to random sample from the same
overall time period.
For each of the 1,064 firms, we use CRSP delist codes and information from the
Securities Data Corp (SDC) to determine those firms that are acquired in a completed deal that
was announced between 1981 and 2015. We rely on both CRSP delist codes and SDC to ensure
completeness of the sample. For our sampling, we find that SDC is very accurate in identifying
deals throughout the full sample period. We initially identify 691 completed deals.
16
For the completed deals, we then obtain information on the takeover auction process from
SEC takeover documents as in Boone and Mulherin (2007). For the mid-1990s to the present, the
SEC documents are available on SEC EDGAR. For earlier periods, we obtain SEC documents
from Lexis-Nexis, Thomson One Financial and microfiche. From these various sources, we are
able to obtain SEC takeover documents for 679 of the completed deals.
We also use SDC to identify targets that were the object of a takeover that was later
withdrawn and the target was not acquired. For each of the firms in the base set, we identify
those identified by SDC as being in a withdrawn deal in the 1981 to 2015 period. We then drop
any withdrawn deals where the target firm was acquired by another bidder at the same time and
therefore is in the completed deal sample. We identify 182 withdrawn deals where the target is
not acquired at that time. Of these, we find 108 withdrawn deals that have SEC documents on the
takeover process, with the deals lacking documents tending to be those that were quickly ended
by the target and/or bidder without substantive takeover bidding.
Our final sample includes 787 takeovers announced between 1981 and 2015. There are
679 completed deals and 108 withdrawn deals. In 2015 dollars, the total deal value for the full
sample is $3.4 trillion.
Table 4 reports the time series distribution of the sample. Panel A reports the distribution
by year. The reported variables include the number of deals in a given year and the percent of the
sample occurring in a given year. The final variable is the rate per year, defined as the number of
deals announced in a year divided by the available firms, where the available firms are the
original 1,064 firms less the firms that have been delisted due to acquisitions and bankruptcies in
previous years.
17
Panel B reports the distribution for each of the five sub-periods in our analysis: 1981 to
1985, 1986 to 1989, 1990 to 1995, 1996 to 2001, and 2002 to 2015. For the full sample and the
completed and withdrawn sub-samples, the Revlon period of 1986 to 1989 has the greatest rate
of takeover activity per year.
Panel C of Table 4 reports regression analysis where the dependent variable is the rate of
takeover activity per year and the independent variables are dummy variables for the different
sub-periods. Note that to facilitate comparisons we make the Revlon period of 1986 to 1989 the
intercept in these regressions. The regression analysis confirms that the Revlon period of 1986 to
1989 has the greatest rate of takeover activity, as the dummy variables for the other four sub-
periods tend to be negative and statistically different from zero. Moreover, the rate of activity in
the 1986 to 1989 sub-period is also greater than in the later three time periods, as evidenced by
the joint F-test at the bottom of Panel C of Table 4.
Table 5 reports some summary statistics for the sample of takeovers. As reported in Panel
A for the full sample, the variables include deal characteristics such as the fraction of cash deals,
tender offers and hostile takeovers. There are also variables related to corporate governance and
takeover hurdles such as the fraction of targets incorporated in Delaware, the fraction of deals
covered by a state antitakeover law, the fraction of targets with a poison pill and a law/pill
dummy for deals covered by a state antitakeover law and/or where the target has a poison pill.
Note that the appendix has data on these variables for each of our five sub-periods.
Panel B of Table 5 reports regression analysis where the dependent variables are the deal
and governance characteristics and the independent variables are the dummies for the sub-
periods. The intercept represents the Revlon period of 1986 to 1989. The regressions in Panel B
of Table 5 indicate that the Revlon period of 1986 to 1989 was distinctly a time of hostile cash
18
tender offers, confirming the “Barbarians at the Gate” depiction of that era (Burrough and Helyar
(1989)). The regressions also indicate a growing presence of state antitakeover laws and poison
pills over time with near ubiquitous coverage by a state law and/or a poison pill since 1990,
consistent with analysis such as Comment and Schwert (1995).
3. Lawsuits and the Evolution of Property Rights
The summary statistics in Panel B of Table 5 indicate that the Revlon period of 1986 to 1989 was
noted for its sharp increase in hostile takeovers along with targets that had growing coverage by
state antitakeover laws and use of poison pills. In this section, we link these changes in the
takeover auction setting to the changing property rights of target boards of directors. Our
analysis focuses on lawsuits by bidders in our sample of takeovers. We emphasize the effect that
the 1986 Revlon case and subsequent decisions had on bidder lawsuits to gauge the intertemporal
variation in the delineation of property rights in takeover auctions.
Research on corporate lawsuits suggests that legal action is a negative sum game. Cutler
and Summers (1988) study the legal battle between Texaco and Pennzoil related to the
acquisition of Getty Oil and find that any gains to the plaintiff Pennzoil were well offset by the
losses to the defendant Texaco. Bhagat, Brickley and Coles (1994) offer similar findings in a
more systematic analysis of corporate lawsuits, many of which entail disputes between targets
and bidders in corporate takeovers.
So why do corporations engage in lawsuits amongst each other rather than simply settle
their disputes? In their survey of research on legal disputes, Cooter and Rubinfeld (1989)
attribute lawsuits to the vagueness of laws. They state (p.1092) that “Vague laws cause
litigation.” They further argue (p. 1093) that the legal process will iteratively clarify vague laws:
19
“Laws whose inefficiency derives from their vagueness will tend to be litigated until the courts
achieve a clear allocation of the underlying entitlement.” Rubin (1977) offers a similar
evolutionary model to indicate why the common law is efficient.
For our sample of takeovers, we determine which deals had a lawsuit by at least one of
the bidders in the transaction. Our sources of information for bidder lawsuits include SEC
takeover documents, media stories and LexisNexis. As suggested by the examples in Table 3,
many bidder lawsuits are tied to access to confidential information during the auction process as
well as the interference of poison pills in the auction process. Other notable reasons for bidder
lawsuits include the validity of state antitakeover laws and the use of deal protection devices
such as termination fees and lock-up options during the auction process.
Applying the Cooter and Rubinfeld (1989) model to our setting, our primary inquiry is
how the 1986 Revlon decision requiring auctioneering impacted bidder lawsuits. Our prediction
derived from this model is that the vagueness of auction requirement in the 1986 decision
actually induced an increase in lawsuits. Examples are reported in Table 3. Through these
subsequent lawsuits, the courts clarified when and whether a given target firm was subject to the
Revlon auctioneering duties. As noted by Policastro (1991, p. 189)), “The broad and ambiguous
language of Revlon left subsequent courts with the task of articulating the circumstances under
which a corporate board of directors would be charged with the Revlon duty.”
Table 6 reports our findings on bidder lawsuits in the 679 completed deals. For the full
sample, 11.5 percent of the deals have a bidder lawsuit. As reported in Panel A, the Revlon
period of 1986 to 1989 has the greatest fraction of bidder lawsuits. In this sub-period, nearly one
quarter of the sample has a bidder lawsuit.
20
Panel B of Table 6 confirms that the rate of lawsuits in the Revlon period of 1986 to 1989
is statistically greater than both the earlier 1981 to 1985 period as well as each of the subsequent
three sub-periods. Moreover, the rate of lawsuits in the Revlon 1986 to 1989 period is also
jointly greater than the three post-1990 sub-periods as reported by the F-test at the bottom of
Panel B. These results hold for our basic specification as well as a regression that controls for
any industry effects in the tendency for bidder lawsuits.
We interpret these results as consistent with the argument that the 1986 Revlon decision
created vagueness in the takeover law regarding auctions especially as pertaining to the property
rights of the target board. This led to a significant increase in lawsuits by hostile bidders using
preemptive bids and toehold strategies in an attempt to force takeover auctions and attain a target
firm at a relatively lower premium. These lawsuits enabled the courts to clarify the duties of the
target board of directors. This iterative process in the courts culminated in the 1989 Time
decision that returned strong property rights to the target board of directors tied to the deference
of the courts via the business judgement rule.
Our interpretation of the evolution of takeover law is distinctly different from the
conventional view that associates the 1989 Time case with target management entrenchment.
(See, for example, Bebchuk, Coates and Subramanian (2002).) One trouble with the
entrenchment view is that it presumes that hostile takeovers are a means to remove poorly
performing target managers. Yet Schwert (2000) finds no systematic measurable differences in
economic performance for hostile targets vis-à-vis targets acquired in a friendly transaction. Our
interpretation is consistent with Schwert’s (2000) conclusion that hostile takeovers are related to
the bargaining choices of targets and bidders in response to the legal environment rather than the
performance of the target firm.
21
Note also that our sample period starting in 1981 is also quite useful in discerning the
evolution of takeover law and legal cases. The facts show it is not simply that lawsuits declined
after the 1989 Time case. But they also had increased dramatically after the 1986 Revlon
decision. This is consistent with ambiguous property rights between 1986 and 1989.
4. Property Rights and the Takeover Auction Process
In this section, we report evidence on how the changing property rights of target boards wrought
by legal decisions have impacted the depth and nature of takeover auctions over time. In our
analysis, we use two measures of auctions. The first is the auction process in the public sphere
depicted in Panel C of Table 1 and studied in research such as Schwert (1996, 2000). The second
is the less visible bidding process depicted in Panel B of Table 1 and studied in research such as
Boone and Mulherin (2007). We find that the impact of the changing property rights of target
boards has had distinctly different effects on these two measures of takeover auctions.
For our first measure of a takeover auction, we follow Schwert (1996, 2000) and define
an auction as a takeover that has two or more bidders competing publicly for a target. Using this
definition, research by Schwert (2000) and Andrade, Mitchell and Stafford (2001) reports that
auctions have declined between the 1980s and 1990s.
In our sample, we find a similar decline over time in this first measure of auctions. Table
7 reports our results for completed deals. In Panel A, the fraction of our sample deals conducted
as auctions averages 24 percent for the full sample. But there is a significant change in auctions
over time. The rate of auctions is greater than 30 percent in the first two sub-periods from the
1980s but falls three-fold to roughly 10 percent in the latter three sub-periods. The regression
analysis in the first column of Panel B of Table 7 confirms a significant decline in the rate of
22
auctions in the latter three sub-periods relative to the Revlon period of 1986 to 1989. Hence,
using this first measure of auctions, based on publicly reported bidding, takeover competition
appears to have declined following the legal decisions granting greater property rights to target
boards.
Boone and Mulherin (2007), however, find that for their sample of deals from the 1990s,
the number of firms engaged in publicly reported bidding is only the tip of the iceberg of the
actual competition for a target. As depicted in Panel B of Table 1, Boone and Mulherin (2007)
describe a deep and active bidding process that occurs behind the scenes prior to any public
announcement of a given deal. In this process, targets sequentially contact potential bidders,
provide confidential information and receive indications of interest and then formal bids.
Appendix C provides summary statistics for these various stages of the auction process for the
deals in our sample.
Boone and Mulherin (2007) refer to this behind the scenes bidding as the “private auction
process”. We use the term “underground auction” to avoid confusion with the type of bidder,
namely a publicly traded synergistic bidder or a private equity bidder engaged in a financial
transaction. Consistent with Boone and Mulherin (2007), we define an underground auction as a
deal where two or more potential bidders sign a confidentiality/standstill agreement with the
target. We extend the empirical findings of Boone and Mulherin (2007) by providing novel
evidence on the underground auction process during the 1980s as compared to the 1990s and
later time periods.
Panel A of Table 7 reports our results on underground auctions. For the full sample, the
rate of underground auctions is roughly 50 percent. But unlike the first measure of auctions
based on publicly reported bidding, the rate of underground auctions remains relatively steady
23
over time. As reported in the regression in the second column of Panel B of Table 7, there is no
significant decline in the rate of underground auctions between the Revlon period of 1986 to
1989 and the later three sub-periods. Hence, the increased property rights to target boards has not
actually reduced takeover competition but has instead shifted the takeover auction process
underground.
These results cannot be underemphasized as they indicate a fundamental change in the
takeover auction process. The increased property rights to target boards has moved takeover
auctions out of the public realm and is now organized in a structured auction that is controlled by
the target board behind the scenes along the lines suggested by Fishman (1988). But rather than
lessening takeover competition and thereby entrenching target management, the movement to
underground auctions has maintained an active bidding process in recent sub-periods that is as
deep as earlier sub-periods.
Our results shed further light on Schwert’s (2000) finding of a strong association between
hostile deals and auctions in the public sphere. In addition to linking hostile deals with public
bidding, Schwert (2000) found that, rather than being related to target performance, hostile deals
were distinguished by the use of publicity in the media by bidders and targets during the deal.
Similarly, in our sample, we find that 72 percent of the deals conducted as auctions in the public
sphere are hostile in nature. Hence, consistent with Schwert’s (2000) conclusion that the use of
hostile deals reflects bargaining choices made by bidders and targets, our findings of the
movement to underground auctions and the associated decline in hostile deals jointly reflect the
changes in the property rights of target boards.
5. Deal Initiation and the Length of the Auction Process
24
In this section, we delve further into the bidding process depicted in Table 1. We analyze how
the choice to initiate a deal by the target board has changed over time. We also study changes in
the length of time between deal initiation and deal announcement (the underground auction in
Parts A and B of Table 1), and the length of time between deal announcement and deal
completion (the auction in the public sphere depicted in Panel C of Table 1). We relate the
changes in the initiation decision and the length of the bidding process to changes in the property
rights of target boards from legal decisions spanning from Revlon (1986) to Time Inc. (1989).
The auctioneering rules in the 1986 Revlon decision would make target boards hesitant to
initiate a deal as that would force the target board to ultimately sell the firm. In effect, under the
auctioneering rule the target board would not be able to set an implied reserve price that would
be the minimum required to sell the firm. Following the 1989 Time Inc decision, however, the
courts returned to a business judgement rule that granted deference to target boards in terms of
strategy. In reaction, target boards often chose to “consider strategic alternatives” where the
target board, in conjunction with their legal and financial advisors, considered policies to
improve shareholder returns including the outright sale of the firm. The 1989 Time Inc. decision
enabled the target board to retract the consideration of alternatives if the board determined that a
shareholder enhancing policy was not available.
To determine the fraction of deals in which the target board initiated the takeover auction,
we consult the SEC takeover documents for each completed deal in our sample. The results are
reported at the top of Panel A in Table 8. For the full sample, 33.8 percent of the sample deals
are initiated by the target board. But there are measurable changes in the rate of target initiation
over time, with the sub-periods of the 1980s having less than 20 percent and the sub-periods
following the 1990 Time Inc decision having double the rate of target initiation.
25
The regression analysis in column (1) of Panel B of Table 8 confirms a significant
increase in the rate of target initiation over time. The dependent variable in the regression is a
dummy variable equal to one when the target board initiates the deal. The independent variables
are the sub-period dummy variables based on legal cases and the regression also controls for
industry fixed effects. Compared to the intercept which represents the Revlon period of 1986 to
1989, the post-Time Inc sub-periods of 1990 to 1995, 1996 to 2001 and 2002 to 2015 all have a
significantly greater rate of target initiation. This is consistent with our argument that the
deference to strategy in the Time Inc decision downplayed the mandated auctioneering rules of
the 1986 Revlon decision and induced target boards to initiate more takeover auctions.
This dramatic change in the rate of initiation by target boards is related to the measurable
decline in hostile takeovers over time. In our sample, only a small fraction (7 percent) of the
deals initiated by the target board turn hostile. The significant movement toward target initiated
deals, which reflects greater property rights and control over the auction by target boards,
thereby lessens hostile activity.
The movement toward target initiated deals and away from hostile transactions also has
had a significant impact on the length of time of the components of the takeover auction process
depicted in Table 1. Using SEC documents, we determine the private initiation date for each deal
and then estimate the length of the two components of the auction process: (1) the length of days
in calendar time between deal initiation and deal announcement (the underground auction in
Parts A and B of Table 1), and (2) the length of days in calendar time between deal
announcement and deal completion (the auction in the public sphere depicted in Panel C of Table
1).
26
The results on the length of the two components of the takeover auction process are
reported in the last three parts of Panel A of Table 8. For the full sample, the length of the time
between deal initiation and deal announcement (the underground auction) has a mean of 125
days and a median of 101 days. But the length of the underground auction varies noticeably
between the first two sub-periods from the 1980s and the latter three time sub-periods from 1990
and later. For the 1981 to 1985 and the 1986 to 1989 sub-periods, the mean length of the
underground auction is less than 90 calendar days, or less than 3 calendar months. By contrast,
the mean length of the underground auction from deal initiation to deal announcement is nearly 6
calendar months or twice as long.
The regression in Column (2) of Panel B of Table 8 confirms a significant increase in the
length of time of the underground auction between deal initiation and deal announcement
following the Time Inc decision in 1989. Hence, rather than the hostile deals of the 1980s that
quickly became auctions in the public sphere, the legal setting following the Time Inc decision
enabled target boards to take more time in the bidding process before announcing a deal to the
public.
Table 8 also reports symmetric changes in the length of the second component of the
takeover auction process in the public sphere between deal announcement and deal completion.
As reported in Panel A of Table 8, the mean length of time between deal announcement and deal
completion is 155 calendar days. But as reported in regression (3), the length of the public
auction process falls following the Time Inc. decision as compared to the deals in the Revlon
period of 1986 to 1989. This reduction is directly related to the decline in the publicly hostile
deals following the Time Inc decision.
27
The final variable reported in Table 8 is the total length of time between deal initiation
and deal completion. For the full sample, the mean length of the entire takeover process is 280
calendar days. As reported in Regression (4) in Panel B of Table 8, the total length of time of the
takeover process has increased following the Time Inc decision.
6. Takeover Premiums over Time
In this section, we estimate target takeover premiums for our sample. The changes in the legal
setting that we document can be expected to have an impact on the level of premiums. Moreover,
the changing length of the time period between deal initiation and deal announcement that we
find suggests that care must also be taken in choosing the event window in which to measure
takeover premiums.
Prior research has found that changes in the legal setting related to takeovers has had a
measurable impact on target takeover premiums. Jarrell and Bradley (1980) found that the delay
induced by the disclosure and minimum tender period requirements of the 1968 Williams Act
was followed by a significant increase in target takeover premiums. Similarly, Jarrell (1985)
found that the delay enabled by target resistance tactics such as lawsuits was associated with
more auctions and correspondingly higher target premiums. To address the possible impact of
the legal setting in our sample, we measure whether there are any changes in target takeover
premiums over time in conjunction with the lengthier time period between deal initiation and
deal announcement following the 1989 Time Inc. decision.
We also analyze whether the changing length of time between deal initiation and deal
announcement has implications on the appropriate event widow with which to gauge target
takeover premiums. A commonly used estimate of target premiums is Schwert’s (2000) measure
28
of target abnormal returns over the (-63, +126) event window which includes a run-up period of
roughly 3 calendar months (63 trading days) prior to deal announcement and 6 calendar months
(126 trading days) following deal announcement. Schwert’s (2000) framework for measuring
target returns is widely used in research in financial economics such as recent work by Cain,
McKeon and Solomon (2017).
Schwert’s (2000) sample of takeovers is from the 1975 to 1996 period. He chose his run-
up period from an ocular examination of the average trends of target returns prior to deal
announcement. In an earlier study of deals from the 1975 to 1991 period, Schwert (1996) used a
shorter run-up period starting two calendar months (42 trading days) prior to deal announcement.
Our data from Table 8 indicate the run-up period prescribed by Schwert (1996, 2000),
while appropriate for the deals from the 1980s that make up much of his sample, is less
appropriate for deals in the 1990s and beyond. In prescribing his run-up window, Schwert’s
(1996, 2000) intention, as discussed on pages 155 to 156 of Schwert (1996), was to capture the
time period in which a target firm and potential bidders engage in pre-announcement bidding, as
depicted in Parts A and B of Table 1 of our paper. Referring to Table 8 of our paper, the Schwert
(2000) run-up period starting 3 calendar months prior to deal announcement appears to capture
the period between deal initiation for our sub-periods of 1981 to 1985 and 1986 to 1989, as the
median time between deal initiation and deal announcement is less than two calendar months.
However, for the three later sub-periods in our analysis, much of which falls outside the years
studied by Schwert (1996, 2000), the run-up window of 3 calendar months fails to capture the
period from deal initiation to deal announcement. Hence, the Schwert (2000) measure of target
premiums based on the (-63, +126) event window may underestimate the gains to target firms in
the latter three sub-periods in our sample.
29
To graphically illustrate the importance of the event window for the measurement of
target premiums in the 1980s versus later sub-periods, we first replicate Figure 2 in Schwert
(1996). As reported in our Figure 1, we plot the cumulative average abnormal returns from 126
trading days (i.e., 6 calendar months) before the public merger announcement (day 0) for the
deals from the 1980s and deals in the 1990s and beyond, separately. Consistent with Schwert
(1996), our Figure 1 indicates that for deals announced in the 1980s, target price starts to move
around day -42 (i.e., two calendar months prior to the public announcement). However, for deals
announced in the 1990s and beyond, the target price starts to move around day -84, two calendar
months earlier than deals from the 1980s. The plot shown in Figure 1 confirms our conjecture
that the standard windows starting from day -42 or -63 may underestimate the gains to target
shareholders in later periods.
To further demonstrate how the choice of event windows may affect the measure of
wealth effects, we estimate two target premiums. The first is CAR (-63, +126) which follows
Schwert (2000) in using an event window from 63 trading days prior to deal announcement
through 126 trading days after deal announcement. The second is CAR (Private Initiation, + 126)
which starts the event window of a given deal on the day the deal is privately initiated, as
determined from SEC takeover documents. This latter estimate is more likely to capture the
effects of the lengthening over time of the period between deal initiation and deal announcement
that we document.
Our results for target takeover premiums are reported in Table 9. Panel A of Table 9
provides means and medians for the full sample and for each of the five sub-periods. Panel B of
Table 9 reports regression analysis of target premiums where the explanatory variables include
30
time period dummy variables, deal and target characteristics and industry fixed effects. The
intercept represents the Revlon period of 1986 to 1989.
In regression (1) in Panel B of Table 9, using the Schwert (2000) measure based on the (-
63, +126) window, target premiums are roughly the same across sub-periods and the F-test at the
bottom of Panel B indicates no significant difference between target premiums in the latter three
sub-periods vis-à-vis the Revlon period of 1986 to 1989. By contrast, in regression (2) using the
(private initiation, +126) window, target premiums are greater in the latter three periods. The
coefficients on the dummy variables for 1990 to 1995, 1996 to 2001, and 2002 to 2015 are all
significantly different from zero and the F-test also indicates a significant difference between
target premiums in the latter three sub-periods vis-à-vis the Revlon period of 1986 to 1989.
These results on target premiums indicate why it is important to account for the changing
takeover auction process over time. The takeover environment following the Time Inc. decision
has clearly changed, with target boards having greater property rights and more control of the
auction process. Not only has there been a decline in hostile takeovers, but the auction process
has gone underground. Moreover, the length of this underground process where the target
engages in pre-announcement bidding with potential suitors has increased. Our estimates of
target premiums that capture these changes indicate that target firms receive higher premiums in
the sub-periods following the Time Inc. decision. These results resemble those of Jarrell and
Bradley (1980) regarding the impact of the Williams Act on target premiums.
Our interpretation of the increasing premium over time is that it reflects greater
bargaining power of target boards where a more valuable bidder match can be found. An
alternative story is that the higher premium reflects a greater entrenchment hurdle that
unsolicited bidders must overcome. To test this alternative story, we contrast the premiums for
31
deals that are and are not initiated by the target board. Presumably, deals initiated by the target
board would be less subject to entrenchment. Inconsistent with the entrenchment story, we find
that adding a variable for target board initiation does not alter our finding of increased target
premiums to target firms in the three later sub-periods. Moreover, we find no difference in
premiums for deals that are and are not initiated by the target board, which is inconsistent with
an entrenchment story that target boards initiate deals to find preferred bidders who are favorable
to management but unfavorable to target shareholders.
7. Robustness Analysis
In this section, we provide robustness analysis using alternative samples and sub-samples. We
first study a random sample of firms from the sample time period from the 1980s through the
present. We then study a sub-sample of our main sample where we match a target firm with a
withdrawn takeover and then a completed deal at a later time period.
A. Analysis of a Random Sample
The sample for this paper is based on the takeover activity between 1981 and 2015 for a set of
firms listed on the Value Line Investment Survey as of the fourth quarter 1981. This sample is
weighted toward relatively large takeover targets. As a check for the robustness of our results,
we replicate our analysis using the data from Liu and Mulherin (2018) which studies a random
sample of takeovers from the 1981 to 2014 time period.
Table 10 reports a summary of this robustness analysis using the random sample which
replicates the regressions in Tables 6 to 9 of this paper for five dependent variables: bidder
lawsuits (Table 6), underground auctions (Table 7), target initiation (Table 8), length of the
period from deal initiation to deal announcement (Table 8), and target premiums using the event
32
window from deal initiation to 126 days after deal announcement (Table 9). In all of the
regressions, the intercept represents the Revlon period of 1986 to 1989.
The results of the regressions using the random sample are markedly similar to the results
from the main sample of this paper. In the three sub-periods following the 1989 Time Inc.
decision, there is:
A significant decline in bidder lawsuits
A similar level of underground auctions
A significant increase in deals initiated by the target
A significant lengthening of the time between deal initiation and deal announcement
An increase in target premiums
Hence, the results from the main sample for this paper, which emphasizes large takeover targets,
are robust to a sample of relatively smaller takeover targets.
B. Matched Sample Analysis of the Takeover Process
As reported in Table 4, our full sample includes both completed deals and withdrawn deals that
were announced but not completed at that time. We found 34 hostile withdrawn deals that
entailed a target firm that was in a completed deal at a later date. We used this sub-sample of
deals to perform a matched sample analysis of the changes in the takeover auction process over
time. Our focus was on bidder lawsuits, underground auctions, and target initiation of deals.
The results of this matched sample analysis are reported in Table 11. Consistent with our
overall analysis, there is a sharp decline in bidder lawsuits. Underground auctions increase
significantly as does the fraction of deals initiated by the target. Hence, for the same target in
different time periods, we see similar time series patterns as for the full sample.
33
C. Matched Sample Analysis of Target Performance
Our interpretation of the changes in the takeover auction process over time is that it reflects
stronger property rights of the target board after the Time Inc. decision. But authors such as
Bebchuk, Coates, and Subramanian (2002) view the decline in hostile takeovers as evidence of
entrenched management. A maintained hypothesis of the entrenchment story is that target
management that rejects a hostile takeover is performing poorly.
In Table 12, we test the entrenchment hypothesis with our matched analysis of targets
that were first the object of a hostile deal that was withdrawn and then were acquired in a
friendly completed deal. The first tests in Panel A match on the same identical firm. The second
tests take a given target of a hostile withdrawn deal and match on industry and size with a later
completed deal. Our performance variables for the target firm are taken from Schwert (2000).
In both Panels in Table 12, there is generally no significant difference in performance for
target firms between the hostile withdrawn deals and the friendly completed deals. Hence,
consistent with Schwert (2000), we find no support for the entrenchment hypothesis for hostile
takeovers. This contrasts with the strong support for the impact of changing property rights of the
target board on the takeover auction process.
8. Summary and Conclusion
We have provided new evidence on the changing nature of corporate takeover auctions over
time. Our main results document a fundamental alteration in the manner in which takeovers are
conducted in the 1981 to 2015 period. In particular, following the 1989 Time Inc. decision:
Takeover auctions have moved from the public sphere to behind the scenes in what we
label underground auctions.
34
The target board is much more likely to initiate a takeover auction.
The length of time between deal initiation and deal announcement has increased.
Accounting for the lengthening of the deal initiation period, takeover premiums to target
firms have increased.
Our results have important implications for research on corporate takeovers and corporate
governance. Research of increasing sophistication such as Cain, McKeon and Solomon (2017) is
making ongoing attempts to determine whether and which state laws and legal cases matter for
the market for corporate control. The gauge of “matter” focuses on the impact of laws and legal
cases on the rate of hostile takeovers. Yet our results point to a more first order effect on the
manner in which corporate takeover auctions are conducted. Target boards are in more control of
the auction process. While competition within such auctions has not abated, it has moved out of
the public sphere of hostile deals and is instead conducted in a more sequential underground
process.
Moreover, these fundamental alterations in the takeover auction process have not
occurred in response to a single type of state law or a specific court decision. Instead, the
changing nature of takeover auctions iterated with a series of court decision between the 1986
Revlon decision and the 1989 Time Inc. case. More importantly, these cases themselves were a
reaction to imposition of takeover hurdles that reacted to the sea change in the corporate takeover
market wrought by shocks such as antitrust relaxation and junk bond financing that made large
firms vulnerable to “the barbarians at the gate.” As in the model of Rubin (1977), these changing
conditions led to lawsuits that clarified the property rights of target boards when defending the
corporate bastion.
35
In a Coasian vein, we interpret the decline of hostile takeovers following the Time Inc.
decision as directly due to the clarification of property rights. Indeed, the burst of hostile
takeovers that we document following the 1986 Revlon decision can be viewed as stemming
from the ambiguity of the auction duties prescribed in that case. Our view contrasts with the
standard argument that the Time Inc. decision entrenched incumbent management. The
entrenchment view implies that hostile deals are associated with poorly performing targets which
is at odds with the results in Schwert (2000). Moreover, the view that target management could
become entrenched and refuse to do value enhancing deals is distinctly at odds with the main
point in Coase (1959, 1960) that well-defined property rights facilitate exchange.
Indeed, the extant literature's (over) emphasis on hostile takeovers seems somewhat out
of place when studying takeover auctions. In any setting with scarcity, markets and private
property rights direct conflicts over the use of scarce resources (Alchian and Demsez (1973)).
Hence, any contest for the control of a corporation could be deemed “hostile” in the sense that
competing parties aim to attain control. But the incumbent party controlling the corporation need
not be a shirking set of managers. The resilience and long-term survival of the modern
corporation modeled by Alchian and Demsetz (1972) is simply that if the stock price under the
incumbents is lower than that deemed appropriate by an outside competing set of possible
mangers, then competition for control will ensue. Our main result on takeover auctions over time
is that this competition has moved from outwardly visible contests to more behind the scenes
auctions.
Our results also have implications for the measurement of target premiums in corporate
takeover transactions. Most research (e.g., Cain, McKeon and Solomon (2017)) relies on the
measure of target premiums developed by Schwert (2000) based on a (-63, +126) event window
36
for all deals. The bulk of Schwert’s (2000) sample come from the 1970s and 1980s. But our
analysis indicates that the deal initiation period has lengthened in more recent decades and that,
accounting for this lengthening period, target premiums have increased after 1990, exactly when
takeover hurdles have increased.
Our results also have implications for the measurement of target price run-up prior to the
public announcements. A recent study by Guercio, Odders-White, and Ready (2017) examines
target abnormal returns in a run-up period 20 days prior to the public announcement for
takeovers in the 1981 to 2012 time period. They report that the level of this measure of run-up is
significantly smaller in more recent years compared to that in the 1980s. The authors interpret
this result as being consistent with increased fear of prosecution in the more recent decades.
However, our results on the lengthening period prior to public announcement after 1990 suggest
that using a fixed window of 20 days to compare run-up in the 1980s and the post-1990 might
not be appropriate. Indeed, as shown in our Figure 1, price run-up in the later periods appears to
occur earlier and is likely not captured by the 20 day window. In untabulated results, we confirm
this conjecture and find that consistent with Guercio, Odders-White, and Ready (2017), run-up is
significantly smaller (at the 1% level) in the post-1990 period compared to the 1980s using the
same 20 day window. However, this run-up difference disappears if we instead measure run-up
from deal initiation until one day prior to the public announcement. This confirms our overall
theme that one must be careful in estimating and comparing takeover premiums between the
1980s and later decades.
Finally, our results suggest the importance of further modeling of the takeover auction
process depicted in Table 1. The process itself is rather complex and often does not fit standard
auction models (Hansen (2001)). We show fundamental changes in both the initiation of auctions
37
and the steps in the process over time. Hence, future work can pursue questions such as how
deals are initiated (Gorbenko and Malenko (2015)) and how the underground bidding process
generates information (Quint and Hendricks (2017)).
38
References
Alchian, Armen A., 1965. “Some Economics of Property Rights.” Il Politico 30, 816-829.
Alchian, Armen A. and Harold Demsetz, 1972. “Production, Information Costs, and Economic
Organization.” American Economic Review 62, 777-795.
Alchian, Armen A. and Harold Demsetz, 1973. “The Property Rights Paradigm.” Journal of
Economic History 33, 16-27.
Anderson, Terry L. and P.J. Hill, 1975. “The Evolution of Property Rights: A Study of the
American West.” Journal of Law and Economics 18, 163-179.
Andrade, Gregor, Mark Mitchell and Erik Stafford, 2001. “New Evidence and Perspectives on
Mergers.” Journal of Economic Perspectives 15, 103-120.
Bainbridge, Stepehn M., 2006. “Unocal at 20: Director Primacy in Corporate Takeovers.”
Delaware Journal of Corporate Law 31, 769-863.
Bebchuk, Lucian A., 1982. “The Case for Facilitating Competing Tender Offers.” Harvard Law
Review 95, 1028-1056.
Bebchuk, Lucian Ayre, John C. Coates IV, and Guhan Subramanian, 2002. “The Powerful
Antitakeover Forces of Staggered Boards: Theory, Evidence and Policy.” Stanford Law Review
54, 887-951.
Berkovitch, Elazar, Michael Bradley and Naveen Khanna, 1989. “Tender Offer Auctions,
Resistance Strategies and Social Welfare.” Journal of Law, Economics and Organization 5, 395-
412.
Bhagat, Sanjai, James A. Brickley, and Jeffrey L. Coles, 1994. “The Costs of Inefficient
Bargaining and Financial Distress: Evidence from Corporate Lawsuits.” Journal of Financial
Economics 25, 221-247.
Boone, Audra L., and J. Harold Mulherin, 2007. “How Are Firms Sold?” Journal of Finance 62,
847-875.
Boone, Audra L., and J Harold Mulherin, 2009. “Is there One Best Way to Sell a Company?
Auctions versus Negotiations and Controlled Sales.” Journal of Applied Corporate Finance 21,
28-37.
Bulow, Jeremy, Ming Huang and Paul Klemperer, 1999. “Toeholds and Takeovers.” Journal of
Political Economy 107, 427-454.
Bulow, Jeremy and Paul R. Klemperer, 1996. “Auctions vs. Negotiations.” American Economic
Review 86, 180-194.
Burrough, Bryan, and John Helyar, 1989. Barbarians at the Gate: The Fall of RJR Nabisco
(Harper: New York).
39
Cain, Matthew D., Stephen B. McKeon and Steven Davidoff Solomon, 2017. “Do Takeover
Laws Matter? Evidence from Five Decades of Hostile Takeovers.” Journal of Financial
Economics 124, 464-485.
Coase, R.H., 1959. “The Federal Communications Commission.” Journal of Law and Economics
2, 1-40.
Coase, R.H., 1960. “The Problem of Social Cost.” Journal of Law and Economics 3, 1-44.
Comment, Robert, and G. William Schwert, 1995. “Poison or Placebo? Evidence on the
Deterrence and Wealth Effects of Modern Antitakeover Measures.” Journal of Financial
Economics 39, 3-43.
Cooter, Robert D., and Daniel C. Rubinfeld, 1989. “Economic Analysis of Legal Disputes and
Their Resolution.” Journal of Economic Literature 27, 1067-1097.
Cutler, David M., and Lawrence H. Summers, 1988. “The Costs of Conflict Resolution and
Financial Distress: Evidence from the Texaco-Pennzoil Litigation.” Rand Journal of Economics
19, 157-172.
DeAngelo, Harry and Edward M. Rice, 1983. “Antitakeover Charter Amendments and
Shareholder Wealth.” Journal of Financial Economics 11, 329-360.
Del Guercio, Diane, Elizabeth R. Odders-White, and Mark J. Ready, 2017. “The Deterrent Effect
of the Securities and Exchange Commission’s Enforcement Intensity on Illegal Insider Trading:
Evidence from Run-up before News Events.” The Journal of Law and Economics 60, 269-307.
Demsetz, Harold, 1967. “Toward a Theory of Property Rights.” American Economic Review 57,
347-359.
Easterbrook, Frank, and Daniel Fischel, 1981. “The Proper Role of a Target’s Management in
Responding to a Tender Offer.” Harvard Law Review 94, 1161-1204.
Fishman, Michael J., 1988. “A Theory of Preemptive Takeover Bidding.” Rand Journal of
Economics 19, 88-101.
French, Kenneth R., and Robert E. McCormick, 1984. “Sealed Bids, Sunk Costs and the Process
of Competition.” Journal of Business 57, 417-441.
Gayle, Wendy B., 1989. “The Defensive and Offensive Use of Poison Pills within the Business
Judgement Rule.” University of Richmond Law Review 24, 127-150.
Gilson, Ronald J. and Reineer Kraakman, 1990. “What Triggers Revlon?” Wake Forest Law
Review 25, 37-59.
Gilson, Ronald J., 2001. “Unocal Fifteen Years Later (And What We Can Do About It).”
Delaware Journal of Corporate Law 26, 491-513.
Gorbenko, Alexander S. and Andrey Malenko, 2015. “Auctions with Endogenous Initiation.”
Working paper available on Google Scholar.
40
Haddock, David D., Jonathan R. Macey, and Fred S. McChesney, 1987. “Property Rights in
Assets and Resistance to Tender Offers.” Virginia Law Review 73, 701-746.
Hansen, Robert G., 2001. “Auctions of Companies.” Economic Inquiry 39, 30-43.
Herzel, Leo, and Richard W. Shepro, 1989. “Negotiated Acquisitions: The Impact of
Competition in the United States.” Business Lawyer 44, 301-322.
Jarrell, Gregg A., 1985. “The Wealth Effects of Litigation: Do Interests Diverge in a Merge.?”
Journal of Law and Economics 28, 151-177.
Jarrell, Gregg A., and Michael Bradley, 1980. “The Economic Effects of Federal and State
Regulations of Tender Offers.” Journal of Law and Economics 23, 371-407.
Jensen, Michael, C., 1993. “The Modern Industrial Revolution, Exit and the Failure of Internal
Control Systems.” Journal of Finance 48, 831-880.
Kidd, Brian K., 2003. “The Need for Stricter Scrutiny: Application of the Revlon Standard to the
Use of Standstill Agreements.” Cardozo Law Review 24, 2517-2573.
Lipton, Martin, 2005. “Twenty Years after ‘Takeover Bids in the Target’s Boardroom: Old
Battles, New Attacks and the Continuing War.” Business Lawyer 60, 1369-1382.
Liu, Tingting, and J. Harold Mulherin, 2018. “How has takeover competition changed over
time?” Journal of Corporate Finance 49, 104-119.
Macey, Jonathan R., 1990. “Auction Theory, MBOs and Property Rights in Corporate Assets.”
Wake Forest Law Review 25, 85-119.
Mitchell, Mark L., and J. Harold Mulherin, 1996. “The Impact of Industry Shocks on Takeover
and Restructuring Activity.” Journal of Financial Economics 41, 193-229.
Policastro, Portia, 1991. “When Delaware Corporate Manages Turn Auctioneers: Triggering the
Revlon Duty after Paramount.” Delaware Journal of Corporate Law 16, 187-240.
Quint, Daniel and Kenneth Hendricks, 2017. “A Theory of Indicative Bidding.” American
Economic Journal Microeconomics forthcoming.
Rhodes, Melissa J., 1991. “The White Knight Privilege in Litigated takeovers: Leveling the
Playing Field in Discovery.” Stanford Law Review 43, 445-475.
Romano, Roberta, 1988. “The Future of Hostile Takeovers: Legislation and Public Opinion.”
University of Cincinnati Law Review 57, 457-505.
Rubin, Paul H., 1977. “Why is the Common Law Efficient?” Journal of Legal Studies 6, 51-63.
Schwert, G. William, 1996. “Markup Pricing in Mergers and Acquisitions.” Journal of Financial
Economics 41, 153-192.
Schwert, G. William, 2000. “Hostility in Takeovers: In the Eyes of the Beholder?” Journal of
Finance 55, 2599-2640.
41
Umbeck, John, 1981. “Might Makes Right: A Theory of the Formation and Initial Distribution of
Property Rights.” Economic Inquiry 19, 38-59.
Velasco, Julian, 2002. “The Enduring Illegitimacy of the Poison Pill.” Journal of Corporation
Law 27, 381-423.
Yablon, Charles M., 1989. “Poison Pills and Litigation Uncertainty.” Duke Law Journal 1989,
54-91.
42
List of Legal Cases (Sorted by Target Firm)
Facet Enterprises: Facet Enterprises v. Prospect Group Inc., WL 36140 (Del Ch 1988)
Federated Department Stores: CRTF v. Federated Department Stores, 683 F. Supp. 422 (S.D.
NY 1988)
J.P. Stevens: West Point Pepperell Inc v. J.P. Stevens & Co., 542 A.2d 770 (Del Ch 1988)
Koppers Co.: BNS Inc v. Koppers Co., 683 F. Supp. 458 (d. Del 1988)
Moore McCormack Resources Inc: Southdown Inc. v. Moore McCormack Resources Inc., 686
F. Supp. 595 (S.D. Tex 1988)
Pillsbury Co.: Grand Metropolitan PLC v. Pillsbury Co., 558 A.2d 1049 (Del Ch 1988)
Revlon: Revlon v. MacAndrews and Forbes, 506 A.2d 173 (Del 1986)
Time Inc.: Paramount Communications v. Time Inc., 571 A.2d 1140 (Del 1989)
Unocal: Unocal Corp v. Mesa Petroleum Co, 493 A.2d (Del 1985)
Wallace Computer Services: Moore v. Wallace Computer Services, 907 F. Supp. 1545 (D. Del
1995)
43
Table 1. Schematic of the Takeover Auction Process
This table provides a schematic of the takeover auction process.
A. Deal Initiation
Target
Consider strategic alternatives
Contact potential bidders
Bidder
Preemptive bid / toehold?
Target reaction
o Friendly / other bidders?
o Hostile
Target defenses / other bidders/
Bidder counter moves / lawsuits?
B. Due Diligence / Iterative Bidding
Confidentiality / Standstill Agreement(s)
Bidder(s) granted access to confidential information
But must agree via contract not to make offers not okayed by target board
Indications of Interest
Subset of bidders make non-binding range of offers
Formal Bids
Takeover Agreement with High Bidder
Deal price / method of payment
Deal protection / fiduciary out
C. Deal Announcement
Possible Additional Bidding
Regulatory and Shareholder Approval
Deal Completion
44
Table 2. Theory on Takeover Auctions
This table presents selected theoretical and empirical studies on takeover auctions.
A. Whether and When to Auction?
Easterbrook and Fischel (1981) Mandatory passivity / Initial bidder search costs
Bebchuk (1982) Mandatory auction / Toehold to compensate search
French and McCormick (1984) Target bears costs / Economize on bidders
B. Preemptive Bids and Access to Information
Bulow, Huang and Klemperer (1999) Toehold helps initial bidder / Target alter rules
Fishman (1988) Target revenue higher by eliminating preemptive
Withhold confidential info until after standstill
Berkovitch, Bradley and Khanna (1989) Deal protection to compensate bidder search
C. Takeover Hurdles
Easterbrook and Fischel (1981) Not allow / Entrenchment
Bebchuk (1982) Not allow / Interfere with auction
Bulow and Klemperer (1996) Impediments to auctions reflect agency costs
DeAngelo and Rice (1983) Can centralize property rights with target board
Jarrell and Bradley (1980) Williams Act associated with higher target premium
Jarrell (1985) Target resistance delays and induces auctions
Comment and Schwert (1995) Poison pill can increase bargaining power
Bebchuk, et al (2002) Staggered boards and pills lethal combination
45
Table 3. Legal Cases on Takeover Auctions
This table reports selected legal cases included in our sample. For each legal case, we report the
name of the target firm, the year of the lawsuit, the main topic of the lawsuit, and the court
decision.
Target Year Topic Court Ruling
Unocal 1985 Board duties Business judgement about the process
Revlon 1986 Board duties Switch from bastion defender to auctioneer
Koppers 1988 Information access Target board can impact the playing field
J.P. Stevens 1988 Information access Use of confidentiality/standstill agreement
Facet Enterprises 1988 Poison pill Gavel to run an auction
Federated Stores 1988 Poison pill Can keep in place for life of auction
Moore McCormack 1988 Poison pill Must redeem pill
Pillsbury 1989 Poison pill Must redeem pill
Time Inc 1989 Board duties Not obligated to abandon deliberate strategy
Wallace Computer 1995 Poison pill Embellishes just say no defense
Sarbanes Oxley 2002 Board duties Mandates board independence
46
Table 4: Sample Distribution by Year and by Sub-Periods.
This table reports sample distribution by year and by five sub-periods. Panel A reports the sample
distribution by year. Percent of sample in each year is the number of deals announced in that year divided
by the total number of deals during the sample period. Rate per year is defined as the number of deals
announced in each year divided by the number of available firms in that year, where the number of available
firms is the original sample size of 1,064 firms minus the number of firms that have been delisted due to
acquisition or other reasons. Columns (2) to (4) report results for the full sample. Columns (5) to (7) report
results for completed deals and Columns (8) to (10) report results for withdrawn deals. Panel B reports
sample distribution by five sub-periods and Panel C reports OLS regression analysis results. For Panel C,
the dependent variable is rate per year and the independent variables are sub-period dummies. The intercept
represents the period of 1986-1989. P81_85 is a dummy variable equal to 1 if the deal is announced from
1981 to 1985. P90_95 is a dummy variable equal to 1 if the deal is announced from 1990 to 1995. P96_01
is a dummy variable equal to 1 if the deal is announced from 1996 to 2001. P02_15 is a dummy variable
equal to 1 if the deal is announced from 2002 to 2015. The last row of Panel C reports a joint test that the
three coefficients of periods P90_95, P96_01, and P02_15 are equal to zero. Robust t-statistics are reported
in parentheses. ***, **, * correspond to statistical significance at the 1, 5, and 10 percent levels,
respectively.
Panel A: Sample distribution by year
Year
# of
deals
% of
sample
Rate per
year
Comp-
leted
% of
sample
Rate per
year
With-
drawn
% of
sample
Rate per
year
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
1981 5 0.64% 0.47% 5 0.74% 0.47% 0 0.00% 0.00%
1982 42 5.34% 3.95% 37 5.45% 3.48% 5 4.63% 0.47%
1983 35 4.45% 3.40% 33 4.86% 3.20% 2 1.85% 0.19%
1984 68 8.64% 6.84% 54 7.95% 5.43% 14 12.96% 1.32%
1985 66 8.39% 7.05% 57 8.39% 6.09% 9 8.33% 0.85%
1986 82 10.42% 9.37% 68 10.01% 7.77% 14 12.96% 1.32%
1987 53 6.73% 6.58% 40 5.89% 4.96% 13 12.04% 1.22%
1988 76 9.66% 10.12% 63 9.28% 8.39% 13 12.04% 1.22%
1989 40 5.08% 5.87% 32 4.71% 4.70% 8 7.41% 0.75%
1990 12 1.52% 1.87% 10 1.47% 1.56% 2 1.85% 0.19%
1991 8 1.02% 1.30% 8 1.18% 1.30% 0 0.00% 0.00%
1992 3 0.38% 0.50% 2 0.29% 0.33% 1 0.93% 0.09%
1993 5 0.64% 0.85% 4 0.59% 0.68% 1 0.93% 0.09%
1994 16 2.03% 2.78% 14 2.06% 2.43% 2 1.85% 0.19%
1995 24 3.05% 4.23% 23 3.39% 4.06% 1 0.93% 0.09%
1996 21 2.67% 3.90% 19 2.80% 3.53% 2 1.85% 0.19%
1997 29 3.68% 5.65% 27 3.98% 5.26% 2 1.85% 0.19%
1998 28 3.56% 5.81% 25 3.68% 5.19% 3 2.78% 0.28%
1999 34 4.32% 7.62% 31 4.57% 6.95% 3 2.78% 0.28%
2000 30 3.81% 7.26% 27 3.98% 6.54% 3 2.78% 0.28%
2001 14 1.78% 3.68% 13 1.91% 3.42% 1 0.93% 0.09%
2002 3 0.38% 0.85% 3 0.44% 0.85% 0 0.00% 0.00%
2003 7 0.89% 2.08% 5 0.74% 1.49% 2 1.85% 0.19%
47
2004 4 0.51% 1.23% 4 0.59% 1.23% 0 0.00% 0.00%
2005 12 1.52% 3.76% 12 1.77% 3.76% 0 0.00% 0.00%
2006 8 1.02% 2.62% 8 1.18% 2.62% 0 0.00% 0.00%
2007 16 2.03% 5.39% 14 2.06% 4.71% 2 1.85% 0.19%
2008 8 1.02% 2.89% 7 1.03% 2.53% 1 0.93% 0.09%
2009 5 0.64% 1.90% 5 0.74% 1.90% 0 0.00% 0.00%
2010 5 0.64% 1.97% 5 0.74% 1.97% 0 0.00% 0.00%
2011 6 0.76% 2.42% 4 0.59% 1.61% 2 1.85% 0.19%
2012 5 0.64% 2.07% 5 0.74% 2.07% 0 0.00% 0.00%
2013 8 1.02% 3.45% 6 0.88% 2.59% 2 1.85% 0.19%
2014 6 0.76% 2.61% 6 0.88% 2.61% 0 0.00% 0.00%
2015 3 0.38% 1.35% 3 0.44% 1.35% 0 0.00% 0.00%
Total 787 100.00% 679 100.00% 108 100.00%
Panel B: Sample distribution by sub-periods
Period
# of
deals
% of
sample
Rate per
year
Comp-
leted
% of
sample
Rate per
year
With-
drawn
% of
sample
Rate per
year
1981_1985 216 27.45% 4.06% 186 27.39% 3.50% 30 27.78% 0.56%
1986_1989 251 31.89% 6.70% 203 29.90% 5.42% 48 44.44% 1.28%
1990_1995 68 8.64% 1.66% 61 8.98% 1.49% 7 6.48% 0.17%
1996_2001 156 19.82% 4.59% 142 20.91% 4.17% 14 12.96% 0.41%
2002_2015 96 12.20% 1.80% 87 12.81% 1.64% 9 8.33% 0.17%
Average
per year 22.5 2.86% 3.82% 19.4 2.86% 3.34% 3.1 2.86% 0.29%
Panel C: Regression of rate per year on sub-period dummies
(1) (2) (3)
Dependent variable Rate per year
Full sample Completed deals Withdrawn deals
Intercept 0.080*** 0.065*** 0.011***
(9.58) (8.84) (10.22)
p81_85 -0.036*** -0.027*** -0.006***
(-3.26) (-2.78) (-3.81)
p90_95 -0.061*** -0.047*** -0.010***
(-5.64) (-5.02) (-7.15)
p96_01 -0.023** -0.013 -0.009***
(-2.17) (-1.39) (-6.38)
p02_15 -0.055*** -0.042*** -0.011***
(-5.84) (-5.09) (-8.53)
Observations 35 35 35
R-squared 0.624 0.589 0.739
Joint test of p90_95,
p96_01, and p02_15 =0
F=16.39
(Prob > F = 0.000)
F=14.22
(Prob > F = 0.000)
F=25.38
(Prob > F = 0.000)
48
Table 5 Summary Statistics
This table reports summary statistics for the full sample (Panel A) and OLS regression analysis for sub-
periods (Panel B). See Appendix A for variable definitions. For Panel B, the intercept represents the period
of 1986-1989. P81_85 is a dummy variable equal to 1 if the deal is announced from 1981 to 1985. P90_95
is a dummy variable equal to 1 if the deal is announced from 1990 to 1995. P96_01 is a dummy variable
equal to 1 if the deal is announced from 1996 to 2001. P02_15 is a dummy variable equal to 1 if the deal is
announced from 2002 to 2015. The last row of Panel B reports a joint test that the three coefficients of
periods P90_95, P96_01, and P02_15 equal to zero. Robust t-statistics are reported in parentheses. ***, **,
* correspond to statistical significance at the 1, 5, and 10 percent levels, respectively.
Panel A: Summary statistics for the full sample period.
Variable Mean Median Std Mean Median Std Mean Median Std
All deals Completed deals Withdrawn deals
Cash 0.64 1.00 0.48 0.62 1.00 0.48 0.73 1.00 0.45
Tender offer 0.48 0.00 0.50 0.49 0.00 0.50 0.41 0.00 0.49
Hostile 0.36 0.00 0.48 0.30 0.00 0.46 0.78 1.00 0.42
Delaware 0.53 1.00 0.50 0.52 1.00 0.50 0.56 1.00 0.50
Poison pill 0.44 0.00 0.50 0.43 0.00 0.50 0.52 1.00 0.50
State law 0.61 1.00 0.49 0.62 1.00 0.48 0.53 1.00 0.50
Law/pill dummy 0.68 1.00 0.47 0.68 1.00 0.46 0.67 1.00 0.47
Panel B: OLS regression analysis for sub-periods
(1) (2) (3) (4) (5) (6) (7)
Dep. Var. Cash
Tender
offer Hostile Delaware
Poison
pill State law
Law/pill
dummy
Intercept 0.741*** 0.653*** 0.554*** 0.546*** 0.530*** 0.610*** 0.773***
(26.71) (21.68) (17.59) (17.31) (16.77) (19.73) (29.14)
p81_85 -0.079* -0.172*** -0.170*** -0.037 -0.484*** -0.494*** -0.624***
(-1.86) (-3.78) (-3.71) (-0.79) (-13.93) (-13.05) (-17.38)
p90_95 -0.168** -0.168** -0.230*** -0.046 0.102 0.317*** 0.197***
(-2.53) (-2.48) (-3.54) (-0.67) (1.54) (7.15) (5.89)
p96_01 -0.273*** -0.288*** -0.381*** -0.007 0.227*** 0.333*** 0.201***
(-5.60) (-5.87) (-8.70) (-0.14) (4.84) (9.21) (6.85)
p02_15 -0.106* -0.466*** -0.387*** -0.015 -0.072 0.359*** 0.195***
(-1.87) (-9.31) (-7.83) (-0.24) (-1.19) (10.07) (6.13)
Observations 787 787 787 787 787 787 787
R-squared 0.042 0.091 0.102 0.001 0.276 0.477 0.531
Joint test of
p90_95,
p96_01, and
p02_15 =0
F=10.97
(Prob > F
= 0.000)
F=31.25
(Prob > F
= 0.000)
F=31.48
(Prob > F
= 0.000)
F=0.15
(Prob > F
= 0.927)
F=11.14
(Prob > F
= 0.000)
F=36.22
(Prob > F
= 0.000)
F=16.78
(Prob > F
= 0.000)
49
Table 6: Bidder Lawsuits over Time for Completed Deals
This table reports the analyses of bidder lawsuits over time. The sample period is from 1981 to 2015 and
includes completed deals only. Panel A reports summary statistics for the full sample and five sub-periods.
Panel B reports the OLS regression analysis. See Appendix A for variable definitions. For Panel B, the
intercept represents the period of 1986-1989. P81_85 is a dummy variable equal to 1 if the deal is
announced from 1981 to 1985. P90_95 is a dummy variable equal to 1 if the deal is announced from 1990
to 1995. P96_01 is a dummy variable equal to 1 if the deal is announced from 1996 to 2001. P02_15 is a
dummy variable equal to 1 if the deal is announced from 2002 to 2015. The last row of Panel B reports
joint test that the three coefficients of periods P90_95, P96_01, and P02_15 equal to zero. Industry effects
are controlled in model (2) using the Fama French 48 industry classification. Robust t-statistics are reported
in parentheses. ***, **, * correspond to statistical significance at the 1, 5, and 10 percent levels,
respectively.
Panel A: Summary statistics
Period Mean Median Std N
Full sample 0.115 0.000 0.319 679
1981_1985 0.086 0.000 0.281 186
1986_1989 0.241 0.000 0.429 203
1990_1995 0.098 0.000 0.300 61
1996_2001 0.042 0.000 0.202 142
2002_2015 0.011 0.000 0.107 87
Panel B: OLS regression analysis
(1) (2)
Dependent variable Bidder lawsuit
Intercept 0.241*** 0.177***
(8.01) (4.56)
p81_85 -0.155*** -0.153***
(-4.25) (-4.21)
p90_95 -0.143*** -0.143***
(-2.94) (-2.80)
p96_01 -0.199*** -0.197***
(-5.76) (-5.40)
p02_15 -0.230*** -0.223***
(-7.13) (-6.70)
Industry effect NO YES
Observations 679 679
R-squared 0.074 0.117
Joint test of p90_95, p96_01, and p02_15 =0
F=17.64
(Prob > F = 0.000)
F=15.47
(Prob > F = 0.000)
50
Table 7: Takeover Competition over Time for Completed Deals
This table reports the analyses of auctions and underground auctions over time. The sample period is from
1981 to 2015 and includes completed deals only. Panel A reports summary statistics for the full sample and
five sub-periods. Panel B reports the OLS regression analysis. See Appendix A for variable definitions. For
Panel B, the intercept represents the period of 1986-1989. P81_85 is a dummy variable equal to 1 if the
deal is announced from 1981 to 1985. P90_95 is a dummy variable equal to 1 if the deal is announced from
1990 to 1995. P96_01 is a dummy variable equal to 1 if the deal is announced from 1996 to 2001. P02_15
is a dummy variable equal to 1 if the deal is announced from 2002 to 2015. The last row of Panel B reports
a joint test that the three coefficients of periods P90_95, P96_01, and P02_15 equal to zero. Industry effects
are controlled in all models using the Fama French 48 industry classification. Robust t-statistics are reported
in parentheses. ***, **, * correspond to statistical significance at the 1, 5, and 10 percent levels,
respectively.
Panel A: Summary statistics for auction and underground auction.
Period Mean Median Std N
Variable: Auction
Full sample 0.240 0.000 0.427 679
1981_1985 0.328 0.000 0.471 186
1986_1989 0.355 0.000 0.480 203
1990_1995 0.115 0.000 0.321 61
1996_2001 0.106 0.000 0.308 142
2002_2015 0.092 0.000 0.291 87
Variable: Underground auction
Full sample 0.492 0.000 0.500 679
1981_1985 0.446 0.000 0.498 186
1986_1989 0.586 1.000 0.494 203
1990_1995 0.426 0.000 0.499 61
1996_2001 0.430 0.000 0.497 142
2002_2015 0.517 1.000 0.503 87
Panel B: OLS analysis for auctions and underground auctions
(1) (2)
Dependent variable Auction Underground auction
Intercept 0.083 0.422
(1.15) (1.61) p81_85 -0.014 -0.117**
(-0.29) (-2.20)
p90_95 -0.202*** -0.092
(-3.64) (-1.24)
p96_01 -0.205*** -0.096*
(-4.50) (-1.68)
p02_15 -0.220*** -0.031
(-4.60) (-0.46)
Industry effect YES YES
Observations 679 679 R-squared 0.146 0.088
Joint test of p90_95, p96_01, and p02_15 =0 F=9.27
(Prob > F = 0.000) F=1.12
(Prob > F = 0.338)
51
Table 8: Takeover Process and Target Deal Initiation over Time for Completed Deals
This table reports the analyses of the length of takeover process and target deal initiation over time. The
sample period is from 1981 to 2015 and includes completed deals only. Panel A reports summary statistics
for the full sample and five sub-periods. Panel B reports the OLS regression analysis. See Appendix A for
variable definitions. For Panel B, the intercept represents the period of 1986-1989. P81_85 is a dummy
variable equal to 1 if the deal is announced from 1981 to 1985. P90_95 is a dummy variable equal to 1 if
the deal is announced from 1990 to 1995. P96_01 is a dummy variable equal to 1 if the deal is announced
from 1996 to 2001. P02_15 is a dummy variable equal to 1 if the deal is announced from 2002 to 2015.
The last row of Panel B reports a joint test that the three coefficients of periods P90_95, P96_01, and
P02_15 equal to zero. Industry effects are controlled in all models using the Fama French 48 industry
classification. Robust t-statistics are reported in parentheses. ***, **, * correspond to statistical significance
at the 1, 5, and 10 percent levels, respectively.
Panel A: Summary statistics for target initiation and the length of takeover process.
Period Mean Median Std N
Variable: target initiation
Full sample 0.338 0.334 0.324 679
1981_1985 0.167 0.000 0.374 186
1986_1989 0.167 0.000 0.374 203
1990_1995 0.410 0.000 0.496 61
1996_2001 0.380 0.000 0.487 142
2002_2015 0.379 0.000 0.488 87
Variable: days between deal initiation and public announcement
Full sample 124.943 101.000 116.856 679
1981_1985 79.812 60.500 81.049 186
1986_1989 88.631 53.000 103.547 203
1990_1995 160.148 140.000 119.125 61
1996_2001 171.282 147.500 118.434 142
2002_2015 205.839 191.000 129.924 87
Variable: days between public announcement and completion/withdrawn
Full sample 154.720 127.000 102.411 679
1981_1985 157.957 119.000 109.974 186
1986_1989 166.266 133.000 111.107 203
1990_1995 135.869 127.000 77.698 61
1996_2001 146.831 125.000 99.729 142
2002_2015 146.954 130.000 80.003 87
Variable: days between deal initiation and completion/withdrawn
Full sample 279.663 259.000 151.014 679
1981_1985 237.769 207.500 143.408 186
1986_1989 254.897 227.000 149.069 203
1990_1995 296.016 266.000 141.262 61
1996_2001 318.113 299.000 146.781 142
2002_2015 352.793 326.000 144.307 87
52
Table 8. Panel B: OLS regression of target initiation and the length of takeover process
(1) (2) (3) (4)
Dependent variable
target
initiation
days(initiation,
announcement)
days(announceme
nt, completion)
days(initiation,
completion)
Intercept -0.098 -36.972 146.609*** 109.637***
(-1.49) (-1.01) (5.30) (5.98)
p81_85 0.035 -3.494 -14.658 -18.152
(0.86) (-0.35) (-1.29) (-1.21)
p90_95 0.299*** 87.154*** -31.884** 55.270**
(4.23) (4.91) (-2.34) (2.45)
p96_01 0.252*** 85.912*** -24.763** 61.148***
(4.95) (6.68) (-2.13) (3.74)
p02_15 0.226*** 120.903*** -18.510 102.393***
(3.75) (7.57) (-1.59) (5.33)
Industry effect YES YES YES YES
Observations 679 679 679 679
R-squared 0.149 0.236 0.101 0.156
Joint test of p90_95,
p96_01, and p02_15 =0
F=12.78
(Prob > F =
0.000)
F=27.26
(Prob > F =
0.000)
F=2.40
(Prob > F =
0.067)
F=10.56
(Prob > F =
0.000)
53
Table 9: Takeover Premiums over Time for Completed Deals
This table reports the analyses of takeover premiums over time. The sample period is from 1981 to 2015
and includes completed deals only. Panel A reports summary statistics for the five sub-periods. Panel B
reports the OLS regression analysis. See Appendix A for variable definitions. For Panel B, the intercept
represents the period of 1986-1989. P81_85 is a dummy variable equal to 1 if the deal is announced from
1981 to 1985. P90_95 is a dummy variable equal to 1 if the deal is announced from 1990 to 1995. P96_01
is a dummy variable equal to 1 if the deal is announced from 1996 to 2001. P02_15 is a dummy variable
equal to 1 if the deal is announced from 2002 to 2015. The last row of Panel B reports a joint test that the
three coefficients of periods P90_95, P96_01, and P02_15 equal to zero. Industry effects are controlled in
all models using the Fama French 48 industry classification. Robust t-statistics are reported in parentheses.
***, **, * correspond to statistical significance at the 1, 5, and 10 percent levels, respectively.
Panel A: Summary statistics for target premiums.
Period Mean Median Std N
Variable: CAR (-63, + 126)
Full sample 0.338 0.317 0.305 679
1981_1985 0.325 0.326 0.254 186
1986_1989 0.358 0.346 0.315 203
1990_1995 0.354 0.350 0.306 61
1996_2001 0.348 0.304 0.323 142
2002_2015 0.292 0.272 0.350 87
Variable: CAR (deal initiation, +126)
Full sample 0.338 0.334 0.324 679
1981_1985 0.318 0.311 0.262 186
1986_1989 0.337 0.342 0.302 203
1990_1995 0.368 0.344 0.381 61
1996_2001 0.352 0.361 0.345 142
2002_2015 0.342 0.293 0.413 87
54
Table 9. Panel B: OLS regression of target premiums
(1) (2)
Dependent variable CAR (-63, +126) CAR (Private Initiation, +126)
Intercept 0.219*** 0.217***
(2.81) (3.15)
p81_85 -0.004 0.022
(-0.13) (0.68)
p90_95 0.033 0.080*
(0.80) (1.68)
p96_01 0.075** 0.111***
(2.10) (2.98)
p02_15 0.033 0.093**
(0.83) (2.07)
Tender offer 0.098*** 0.088***
(4.13) (3.40)
Target size -0.008
(-0.97) Target size (initiate) -0.011
(-1.24)
Hostile 0.078*** 0.054**
(2.95) (2.00)
Market/book -0.014** -0.012**
(-2.44) (-2.03)
Debt/equity 0.006 -0.002
(0.47) (-0.14)
ROE -0.001 -0.026
(-0.02) (-0.47)
Price/earnings -0.000 0.000
(-1.63) (0.57)
Sales Growth 0.053 -0.021
(0.83) (-0.33)
Industry YES YES
Observations 656 656
R-squared 0.146 0.141
Joint test of p90_95, p96_01,
and p02_15 =0
F=1.47
(Prob > F = 0.223)
F=3.35
(Prob > F = 0.018)
55
Table 10: Robustness Test: Evidence from a Random Sample
This table reports OLS regression analysis for a random sample of 388 observations. The sample is from
Liu and Mulherin (2018). See Liu and Mulherin (2018) for details of the construction of the random sample.
Appendix A provides variable definitions. The intercept represents the period of 1986-1989. P81_85 is a
dummy variable equal to 1 if the deal is announced from 1981 to 1985. P90_95 is a dummy variable equal
to 1 if the deal is announced from 1990 to 1995. P96_01 is a dummy variable equal to 1 if the deal is
announced from 1996 to 2001. P02_15 is a dummy variable equal to 1 if the deal is announced from 2002
to 2015. The last row of the table reports a joint test that the three coefficients of periods P90_95, P96_01,
and P02_15 equal to zero. For Model (5), there are additional control variables included in the regression:
tender offer, target size, hostile, market/book, debt/equity, ROE, price/earnings, and sales growth. For
brevity, the coefficients of these additional control variables are not reported. Robust t-statistics are reported
in parentheses. ***, **, * correspond to statistical significance at the 1, 5, and 10 percent levels,
respectively.
(1) (2) (3) (4) (5)
Dependent
variable
Bidder
lawsuit
Underground
auction
Target
initiation
days(initiation,
announcement)
CAR
(Initiation, +126)
Intercept 0.142** 0.100 0.779*** 97.329*** 0.408***
(2.41) (1.14) (9.63) (3.63) (3.57)
p81_85 -0.036 -0.204* -0.058 -61.439** -0.014
(-0.42) (-1.79) (-0.59) (-2.17) (-0.17)
p90_95 -0.135** -0.183 -0.027 14.645 0.168*
(-2.05) (-1.60) (-0.28) (0.44) (1.66)
p96_01 -0.142** -0.100 0.221*** 114.671*** 0.170**
(-2.41) (-1.14) (2.74) (4.27) (2.05)
p02_15 -0.151*** 0.074 0.142* 147.512*** 0.195**
(-2.62) (0.84) (1.76) (5.04) (2.32)
Industry YES YES YES YES YES
Observations 388 388 388 388 344
R-squared 0.258 0.152 0.161 0.270 0.286
Joint test of
p90_95, p96_01,
and p02_15 =0
F=2.31
(Prob > F
= 0.07)
F=0.71
(Prob > F =
0.39)
F=4.49
(Prob > F =
0.00)
F=13.55
(Prob > F =
0.00)
F=2.01
(Prob > F =
0.10)
56
Table 11: Robustness Test: Matched Sample Takeover Process
This table reports the takeover process for 68 observations consisting 34 hostile withdrawn deals and 34
matched firms. The sample period is from 1981 to 2015. Hostile withdrawn targets are matched by
themselves when they later became completed and friendly deals. We report Bidder lawsuit, underground
auction, and target initiation for the withdrawn deals and the matched completed deals. Columns (3) – (5)
report the difference of the operating performance. See Appendix A for variable definitions. Pr > |t| is the
probability of student’s t test. Pr >= |S| is the probability of the signed rank test. ***, **, * correspond to
statistical significance at the 1, 5, and 10 percent levels, respectively.
Variable Withdrawn Completed
Difference
(withdrawn-
completed) Pr > |t| Pr >= |S| N
(1) (2) (3) (4) (5) (6)
Bidder lawsuit 0.486 0.000 0.486 0.000*** 0.000*** 68
Underground auction 0.114 0.353 -0.239 0.018** 0.010** 68
Target initiation 0.029 0.353 -0.324 0.000*** 0.000*** 68
57
Table 12: Robustness Test: Matched Sample Performance Analysis
This table reports the operating performance for hostile withdrawn deals and their matched firms. The
sample period is from 1981 to 2015. Panel A reports hostile withdrawn targets matched by themselves
when they later became completed and friendly deals. Panel B reports hostile withdrawn deals matched
by completed and friendly deals in the same industry and with the closest firm size prior to merger
announcement. Columns (3) – (5) in both Panel A and Panel B report the difference of the operating
performance. All variables are detrended by regressing the variable by time and by taking the residual
plus the sample average. Pr > |t| is the probability of student’s t test. Pr >= |S| is the probability of the
signed rank test. ***, **, * correspond to statistical significance at the 1, 5, and 10 percent levels,
respectively.
Panel A: matched by the same firm
Variable Withdrawn Completed
Difference
(withdrawn-completed) Pr > |t|
Pr >=
|S| N
(1) (2) (3) (4) (5) (6)
Market/book 1.751 2.805 -1.054 0.113 0.742 64
ROE 0.040 0.110 -0.070 0.516 0.980 68
Sales Growth -0.002 0.095 -0.097 0.087* 0.142 68
Price/earnings 24.265 31.353 -7.088 0.394 0.143 42
Debt/equity 0.625 0.884 -0.259 0.288 0.855 64
Panel B: matched by industry and firm size
Variable Withdrawn Completed
Difference
(withdrawn-completed) Pr > |t|
Pr >=
|S| N
(1) (2) (3) (4) (5) (6)
Market/book 1.663 2.021 -0.358 0.153 0.572 140
ROE 0.025 0.067 -0.042 0.457 0.578 144
Sales Growth 0.000 0.042 -0.042 0.246 0.552 144
Price/earnings 27.681 28.648 -0.966 0.907 0.304 100
Debt/equity 0.662 0.638 0.024 0.871 0.251 140
58
Figure 1: Target abnormal returns, 1980s versus the 1990s and later
This figure plots the cumulative average abnormal returns from 126 trading days before the public merger
announcement (day 0) through 21 trading days after the public announcement for deals announced in the
1980s (i.e., 1981-1989) and in the post 1990s (i.e., 1990-2015). Deals announced in the 1980s are shown
with a dashed line and deals announced in the 1990s and beyond are shown with a solid line. Following
Schwert (1996), we use CRSP value-weighted portfolio for days -379 to -127 to estimate market model
parameters to define abnormal returns.
-0.05
0.05
0.15
0.25
0.35
-126-121-116-111-106-101-96 -91 -86 -81 -76 -71 -66 -61 -56 -51 -46 -41 -36 -31 -26 -21 -16 -11 -6 -1 4 9 14 19
Average cumulative returns- comparing 80s and post 90s
1980s Post 1990s
59
Appendix A: Variable Definitions
This appendix reports the definitions of our main variables and data sources for the variables. Variables
are listed in alphabetical order.
Variable Definition Data Source
Auction A dummy variable equal to 1 if two or more potential
buyers announced a bid publicly.
SEC documents, media
stories
Bidder lawsuit A dummy variable equal to 1 if there is a lawsuit filed
by a bidder in a given deal.
SEC documents, media
stories
CAR(-63,+126)
Cumulative abnormal returns around the event window
(-63, 126), where abnormal returns are net of market
returns, proxied by the CRSP value-weighted index, and
day 0 is the public merger announcement date.
CRSP
CAR(Private
Initiation, + 126)
Cumulative abnormal returns from the private initiation
date until 126 trading days after the public
announcement, where abnormal returns are net of
market returns, proxied by the CRSP value-weighted
index.
CRSP
Cash A dummy variable equal to 1 if the bidder uses cash as
the only method of payment. SEC documents
Confidential
The number of potential buyers that signed a
confidentiality/standstill agreement with the selling
firm.
SEC documents
Contact The number of potential buyers contacted by the selling
firm and its financial advisor. SEC documents
Days
(Announcement,
Completion)
The number of days between the public announcement
date and the deal completion date. SEC documents
Days (Initiation,
Announcement)
The number of days between the private initiation date
and the public announcement date. SEC documents
Days (Initiation,
Completion)
The number of days between the private initiation date
and the deal completion date. SEC documents
Deal size The transaction value adjusted by inflation. Deal value
reflects the 1980 dollar value. SDC, SEC documents
Delaware A dummy variable equal to 1 if the target firm is
incorporated in Delaware SEC documents
Hostile
A dummy variable equal to 1 if the deal involves a
hostile bidder (not necessarily the winning bidder)
defined as when the target board rejects the bid.
SEC documents, media
stories
Indication of
interest
The number of potential buyers that submitted an
indication of interest of the price range offered to the
selling firm.
SEC documents
Law/pill dummy A dummy variable equal to 1 if a firm has a poison pill,
or is covered by a state antitakeover law.
SEC documents, proxy
statements, News
stories, Bebchuk and
Ferrell (2002), Bertrand
and Mullainathan
(2003), Bryan (1991)
60
Market/book
The ratio of the year-end market value of common stock
to the book value of equity for the fiscal year prior to the
merger announcement.
Compustat
Poison pill A dummy variable equal to 1 if the target has a
shareholders rights plan (poison pill). SEC documents
Price/earnings
The ratio of the year-end stock price to earnings per
share for the fiscal year prior to the merger
announcement.
Compustat
Private bidders The number of potential buyers that submitted a private
written offer. SEC documents
Public bidders The number of potential buyers that announce a formal
bid for the firm publicly. SEC documents
ROE
Return on equity, measured as the ratio of earnings to
average equity for the fiscal year prior to the merger
announcement.
Compustat
Sales growth The proportional change in sales over the fiscal year
prior to the merger announcement. Compustat
State law
A dummy variable equal to 1 if the target state of
incorporation has anti-takeover laws implemented at the
time the deal is announced.
Bebchuk and Ferrell
(2002), Bertrand and
Mullainathan (2003)
Target initiation A dummy variable equal to 1 if the deal is initiated by
the target. SEC documents
Target size
The log value of the target stock price times shares
outstanding 3 months prior to the merger announcement
date.
CRSP
Target size
(initiate)
The log value of the target stock price times shares
outstanding one day prior to the private initiation date. CRSP
Tender offer A dummy variable equal to 1 if the deal is a tender offer SEC documents
Underground
Auction
A dummy variable equal to 1 if two or more potential
buyers sign a confidentiality/standstill agreement. SEC documents
Withdrawn
A dummy variable equal to 1 if the target firm was not
taken over by any other firm within a year of the merger
announcement.
SDC, SEC documents
61
Appendix B: summary statistics for deal characteristics by sub-periods
This appendix reports summary statistics for deal characteristics by five sub-periods. In each period,
mean, median and standard deviation of deal characteristics are reported for all deals, completed deals,
and withdrawn deals. See Appendix A for variable definitions.
Sample period: 1981-1985
Variable Mean Median Std Mean Median Std Mean Median Std
All deals (N=216)
Completed deals
(N=186)
Withdrawn deals
(N=30)
Cash 0.66 1.00 0.47 0.64 1.00 0.48 0.80 1.00 0.41
Tender offer 0.48 0.00 0.50 0.49 0.00 0.50 0.40 0.00 0.50
Hostile 0.38 0.00 0.49 0.33 0.00 0.47 0.70 1.00 0.47
Delaware 0.51 1.00 0.50 0.50 0.50 0.50 0.57 1.00 0.50
Poison pill 0.05 0.00 0.21 0.04 0.00 0.20 0.07 0.00 0.25
State law 0.12 0.00 0.32 0.12 0.00 0.33 0.07 0.00 0.25
Law/pill dummy 0.15 0.00 0.36 0.16 0.00 0.36 0.10 0.00 0.31
Sample period: 1986-1989
Variable Mean Median Std Mean Median Std Mean Median Std
All deals (N=251)
Completed deals
(N=203)
Withdrawn deals
(N=48)
Cash 0.74 1.00 0.44 0.74 1.00 0.44 0.73 1.00 0.45
Tender offer 0.65 1.00 0.48 0.70 1.00 0.46 0.44 0.00 0.50
Hostile 0.55 1.00 0.50 0.48 0.00 0.50 0.88 1.00 0.33
Delaware 0.55 1.00 0.50 0.53 1.00 0.50 0.60 1.00 0.49
Poison pill 0.53 1.00 0.50 0.50 0.00 0.50 0.67 1.00 0.48
State law 0.61 1.00 0.49 0.62 1.00 0.49 0.58 1.00 0.50
Law/pill dummy 0.77 1.00 0.42 0.76 1.00 0.43 0.81 1.00 0.39
Sample period: 1990-1995
Variable Mean Median Std Mean Median Std Mean Median Std
All deals (N=68)
Completed deals
(N=61)
Withdrawn deals
(N=7)
Cash 0.57 1.00 0.50 0.57 1.00 0.50 0.57 1.00 0.53
Tender offer 0.49 0.00 0.50 0.51 1.00 0.50 0.29 0.00 0.49
Hostile 0.32 0.00 0.47 0.25 0.00 0.43 1.00 1.00 0.00
Delaware 0.50 0.50 0.50 0.48 0.00 0.50 0.71 1.00 0.49
Poison pill 0.63 1.00 0.49 0.62 1.00 0.49 0.71 1.00 0.49
State law 0.93 1.00 0.26 0.95 1.00 0.22 0.71 1.00 0.49
Law/pill dummy 0.97 1.00 0.17 0.97 1.00 0.18 1.00 1.00 0.00
Sample period: 1996-2001
Variable Mean Median Std Mean Median Std Mean Median Std
All deals (N=156)
Completed deals
(N=142)
Withdrawn deals
(N=14)
Cash 0.47 0.00 0.50 0.45 0.00 0.50 0.64 1.00 0.50
Tender offer 0.37 0.00 0.48 0.37 0.00 0.48 0.36 0.00 0.50
Hostile 0.17 0.00 0.38 0.13 0.00 0.34 0.57 1.00 0.51
Delaware 0.54 1.00 0.50 0.55 1.00 0.50 0.43 0.00 0.51
Poison pill 0.76 1.00 0.43 0.75 1.00 0.44 0.86 1.00 0.36
State law 0.94 1.00 0.23 0.94 1.00 0.23 0.93 1.00 0.27
Law/pill dummy 0.97 1.00 0.16 0.97 1.00 0.17 1.00 1.00 0.00
62
Sample period: 2002-2015
Variable Mean Median Std Mean Median Std Mean Median Std
All deals (N=96)
Completed deals
(N=87)
Withdrawn deals
(N=9)
Cash 0.64 1.00 0.48 0.62 1.00 0.49 0.78 1.00 0.44
Tender offer 0.19 0.00 0.39 0.16 0.00 0.37 0.44 0.00 0.53
Hostile 0.17 0.00 0.37 0.11 0.00 0.32 0.67 1.00 0.50
Delaware 0.53 1.00 0.50 0.54 1.00 0.50 0.44 0.00 0.53
Poison pill 0.46 0.00 0.50 0.45 0.00 0.50 0.56 1.00 0.53
State law 0.97 1.00 0.17 0.97 1.00 0.18 1.00 1.00 0.00
Law/pill dummy 0.97 1.00 0.17 0.97 1.00 0.18 1.00 1.00 0.00
63
Appendix C: Summary Statistics for the Takeover Process Variables by Sub-Periods
This table reports summary statistics for the takeover process variables by five sub-periods. Contact is the
number of potential buyers contacted by the selling firm and its financial advisor. Confidentiality is the
number of potential buyers that signed a confidentiality/standstill agreement with the selling firm.
Indications of Interest is the number of potential buyers that submitted an indication of interest indicating
the price range offered to the selling firm. Private bidder is the number of potential buyers that submitted
a private written offer. Public bidder is the number of potential buyers that announce a formal bid for the
firm publicly.
Period Mean Median Std N
Variable: Contact
1981_1985 5.952 2.000 14.838 186
1986_1989 10.158 3.000 24.876 203
1990_1995 7.312 1.000 18.428 61
1996_2001 8.239 2.000 19.667 142
2002_2015 17.058 3.000 36.896 87
Variable: Confidentiality
1981_1985 2.366 1.000 4.034 186
1986_1989 5.103 2.000 12.002 203
1990_1995 3.639 1.000 8.635 61
1996_2001 3.669 1.000 7.331 142
2002_2015 6.736 2.000 13.074 87
Variable: Indications of interest
1981_1985 1.651 1.000 1.121 186
1986_1989 1.980 1.000 1.715 203
1990_1995 1.639 1.000 1.291 61
1996_2001 1.634 1.000 1.748 142
2002_2015 2.632 1.000 3.458 87
Variable: Private bidder
1981_1985 1.473 1.000 0.699 186
1986_1989 1.616 1.000 0.856 203
1990_1995 1.377 1.000 0.860 61
1996_2001 1.394 1.000 0.842 142
2002_2015 1.276 1.000 0.564 87
Variable: Public bidder
1981_1985 1.398 1.000 0.626 186
1986_1989 1.438 1.000 0.667 203
1990_1995 1.115 1.000 0.321 61
1996_2001 1.113 1.000 0.339 142
2002_2015 1.103 1.000 0.342 87