managerial performance, boards and takeover bidding
TRANSCRIPT
of Fillallr.~ I(199J) 63-91)
Hirshleifcr *.", Thakor " Atr,lrr!:,:: (irudrrurr Sclrool of ,fnf~a~[~f?rr,~r, UCLA. Li,s A ~ r g ~ ~ l ~ ~ r . CA VOO2.t. U M
"t.lrool ~~fBrrsirrt.sr. Lrdiurrcr U~lirrr.sity, Blr~o.*!irrcro~r. 1,V J7J05. U.W
Octoher 199.3)
modcls thc thc simultanc- intcrnal cxtcrnal corporate mn t ro l mcchanisms-hoard
takemcrs. examine the thc the affccts thc hchavior thc and
thc Thc of intcrcsts. snxctimcs takeover. oppositinn Sc ncv's
attcmpl hc rate lakcovcr attempt advcrsc information the
managcr. there i s prc~hability lhat thc incffcctive. thcn the can ncws
sffcct dominatc when adverse perfomac-.e highcr er ante
ineffective. exarnplc. the board managemcnt-
* aulhnr. based rarlier wilh differrrlt tiilc. Ws Itlanit Sulran Franks.
Richard Dennis Shechdn. Cllcster Spalt Sheridan Titman t o r useful discusrians Sugato Chakravarty Milbourn rescarch assislance. the
editor. Ken Lehn. helpful benefitted frt~rn cornnientr ;I! !lie Auslralasian Cc~nfermrc Sydncy. Austrdliii piiriir-ular. lrslrn
the C;eratd Garveyl Ihe Washinglon. at Purdue Uni.dersity. ctf Minnrsota.
USC. nolwilhstanding, responsihility for the
IM29-11Yc~/W/$07.00 0 IYW Plscvicr rights rcscrvcd S.$Il/ 0929. I t o ' ~ ~ ' ~ 3 ~ E l l t ~ l l s . 7
Journal of CORPORATE FINANCE
Journal Corporate
Managerial performance, boards of directors and takeover bidding
David Anjan V.
(Final version received
Abstract
This paper maintenance o f managemcnt quality through ous functioning o f and dis-missals and Wc how information sets o f board and acquiror are noisily aggrcgatcd, and how this of board
acquiror. board directors. acting in shareholders' wi l l oppose a and this csn good for the firm. An unsuccess-ful takeover may followed hy a high of managcmcnt turnover, bccausc a conveys possessed by the hidder about
I f a hoard is a forced resignation of managcr be either good or bad for the firm. A positive
is predicted t o thcrc is more public information available about the manager's and when there is a probability that the board is for i f is dominated rather than outsider-dominated.
Corresponding This paper i s in part on an paper a
Roll. and in developing these ideas. and Todd fur and
for comments. We have also received Third Finance and Banking in (in
discussant. and American Finance Association Meeting in DC as well as in seminars University of Illinois. University and This we hear sole contents.
Science B.V. Al l
0. Hirrhleifer, A.V. Thakor /Journalof 1 (1994)
K q Takcovers; pcrformancc; Information
classificarion:
Schoelhorn corne-
quadntp!ing pisfiis torrted
Schoelhorn-the
is.
two sl~areholders'
k g . DeLong, 1990). a d
even
I entitled. Chiel' k i n g Out tk~ilrds'. the lVall Stmt J~irmral (h/h/91)
heen cert;rinly wholes;~le curpiwale CEOs.
w)mewhat Elncorp. lnc.. General Corp., Grumman Soiltheart 1l:lnliing Corp. Ahhot! L~horatorics.
gnwinp l i \ t ihr ecimon~ic hard inrt i luri~~nal itnd the impi~rlant k ~ i t r d cc>mmittees outnamher nsiden a1 Pb of ItX) hy execulivc SpencerSluart. decade 6:o. R c of those 4.m-l ratlo outsiders douhled the past deciide. irverage r;~tio 3-to-I. 2-to-l 1980.'
64 Corporate Finance 63-90
words: Board of directors: Managerial aggregation
JEL G34
Robert A. appeared to be the model chief executive. As CEO of Abbott Laboratories. the former sales rep had overseen a major back, nearly tripling sales and during the 1980s. He was
as an Executive of the Year by a business magazine in 1986 and invited on the boards of more than half a dozen corporations and trade groups.
But few people knew much about the less public grandiose man who ruled over his company as if it were a private fiefdom. Few people, that until Abbott's board kicked him out last March.
Cover Story: 'CEO Disease,' Business Week, April 1, 1991.
I. Introduction
Despite an apparent decline in the last decades of the role of the hoard of directors as the watchdog to monitor management performance, boards have traditionally played an important part in the functioning of public corporations With the recent proliferation of regulatory boards are expected to become ing. ' Of course. to the extent that boards pick up the slack by being tougher on inefficient managers, the cost of these impediments to shareholders is reduced.
In a recent article 'More Executives Are Forced By Tougher reports.
'Chief executives have long pampered. powerful and paid plenty. And while there's no revolution going on against it's also clear that their
perches are getting more precarious. lust ask the recently departed chiefs of First City of Texas Data Corp.. Circle K Corp..
see legal impediments to (hostile) takeovers,
more significant in corporate monitor-
and The partly reflects times. I t also shows the incieased power of
investors rising vigilance of outside directors. These directors now run and large U.S.companies surveyed
recruiters That's up from HI a firm says the number hoards with a or gicater of to insiders har to 40 in
The at hip corporations now i s up from in
Hirshleifer. K Thakor /Jarmu1 (IW41 6-3-90 65
by
that takeover market
(e.g. Grossman Fishman,
Kini (1993) txrnwer
Bricklep empirical aad
disciplinaly
to
generate firm's
igilores
pricc hwever, !o
board's
acfept by
er acquiror
We model external corporate control mechanisms as between the board and the
7 - rvidencc takrwrr-rclatrd see
McConnell (1991). Klein Rosenfeld (19881, survey by Jenxn (1988). .I (hostile) takeoven
targel firm's
D. A. of Corporate Finance I
As this discussion suggests, there is an interplay between the internal mechanism for corporate control (as represented actions of the board of directors) and the external mechanism for corporate control (as represented by actions of an acquiror). The takeover market is less important when the board is up to the task of removing inefficient managers. Conversely, the board's behavior is likely to be affected by its knowledge the
is active. This interplay between the board and acquirors has not received attention in the theoretical literature on corporate control which has, for the most part, considered takeovers in isolation and Hart, 1980; 1988). The importance of the interplay between board and acquirors is highlighted by the evidence of et al. that CEO
after takeovers is inversely related to prior target among targets with insiderdominated boards, but not for those with out-sider-dominated boards. See also and James (1987) whose evidence in banking suggests that the market for takeovers boards of directors are substitute devices in controlling managerial behavior.
We develop a model in which the roles of the board of directors and takeovers are considered simultaneously. Our purpose is to provide a theoretical framework within which interpret the existing evi-dence on the relation between these internal and external control mecha-nisms, and to additional predictions. 2
We assume in our analysis that the target board can be either lax (aligned with management) or vigilant (aligned with shareholders), and that only the board knows its own type. A lax hoard internal signals of managerial performance and seeks to protect the manager's job to the extent possible. Thus, it never dismisses the manager. A sufficiently high offer may exert enough 'price pressure' on it, succumb to the takeover bid. A vigilant incentive, on the other hand, is more closely aligned with that of the shareholders, although it too may display some aversion to a takeover. Such a board may dismiss the manager, and even if it retains him, it may a takeover bid, based on a rational processing of its own information about the manager in conjunction with information conveyed the bid. We also assume that it is uncertain ante whether an will arrive with a more efficient manager to replace the incumbent in the target firm.
the interaction between the internal and a sequential interaction
tor related empirical on management turnover. Marlin and and and the and Warner
Such aversion may arise from the fact the often result in changes in the
composition of the board of directors.
acquiror.
ability is ia
privately firm.
the ~Tnveys i~formation.
removed rejected,
unsucccss;ul foilawed
board abcut (nnisily)
about managerial
t y (i.e., (Klein
1988). (1993)
information/expenise in less subject to some practitioners. For example, in its cover Busitless Week (7/3/89) states that
'.. .management
less threat of the large.'
Furthcrmorc. (1991) rcports targets have lower ownership avcrage thcir tRan
The board moves first in making a decision to either retain or fire the manager. This decision is based on a (noisy) signal of the manager's
that is privately observed by the board, and also on whether the board or vigilant. If the board fires the manager and replaces him with
another manager, the game ends. Otherwise, the acquiior !if one is present) observes another signal of the manager's ability and then decides
whether or not to make a bid for the The bid price depends in part on acquiror's beliefs about the board's vigilance. Although the board does
not directly observe the acquiror's signal, the bid price The board can either accept or reject the bid, based on the noisy aggregation of its own signal with its inference of the acquiror's information. If the bid is accepted, the incumbent target manager is by the acquiror. If the bid is the board can decide once again whether to retain the manager or fire him based on the information communicated by the bid.
The model has several implications. We show that takeover attempts may be by a high frequency of management turnover. The reason is that even though a takeover attempt is rejected because the offer price is inadequate, the infers from the takeover bid that the bidder possesses adverse information the manager. i'his, when aggre-gated with its own information about the manager, leads the board to have less favorable posterior beliefs ability than it did prior to the offer. This explains the otherwise puzzling finding that targeted repur-chases of shares held potential takeover bidders greenmail) are frequently associated with lop management changes and Rosenfeld,
Similarly. Franks and Mayer find that unsuccessful hostile takeover bids-even those not involving greenmail-are followed by high rates of executive dismissals.
Another result is that when the board acts to maximize shareholder wealth, an active takeover market tends to substitute for internal dismissal by the board. When the takeover market is active, board members become more lenient in marginal cases because it becomes more desirable to exploit the
of the outside acquiror to determine whether a change management is desirable. Conversely, firms with active boards should be
takeovers. This agrees with the view of story.
at many companies does need to be made more accountable. If more hoards took charge, management would be under
attack-from raiders, from foreign rivals, from sharehold-ers. and from public at
Shivdasani that directors in hostile tekeover stakes on in firms do
D. Hinl~leifer, A.V. 771aknr/Journnl Finance (19%)
contrul :o t~ostile
occrlrs
retention has a
information acquiror the board's Vow,
information
demand tesult
foi
firm
ex ante
havc
grabability er
cx ante vieilant scch that
be~ng
takeover The Paper most closely Hirshleifer and Thakor
(1991). which also :he inte:action bemeen the actions of bards and acquirors. he:ween that and this Paper are as follows. Thakor allow the acquiror to dismiss the as well rise to what is referred to .. a 'kick-in-the-pants' board may be overly aggressive in firing the manager to for vigilance and thereby
of Corporate I 63-90 67
directors in a sample. This is consistent with the view that firms with inactive boards are more prone becoming targets.
Nevertheless, the information aggregation that between the bidder and :he board is far from perfect. An obvious problem is that managerial
may be due to the laxity of the board. Furthermore, even if a board good reputation for vigilance, the bidder does not always know it the
board regards a retained manager as a marginal or superior performer. To the extent that the bidder is uncertain, this pooling reduces his incentive to replace a marginal manager. Conversely, even if the hoard has favorable
about the manager, the ignorance of the potential about information will sometimes lead to a takeover attempt.
even a vigilant board will oppose the takeover attempt if its own about the manager is sufficiently favorable to warrant retention despite the adverse information conveyed by the bid. Or, it may a price so high that the bidder is not willing to acquire. Hence, we have the that board resistance to a takeover attempt can be good news shareholders.
We show that if there is a probability that the board is lax, a dismissal can be either good or bad news for the firm. The reason is that on the one hand i t may signal that the has been doing poorly, but on the other may reveal that the board is effective in removing a bad manager. Hence, the positive effect is more likely to dominate when public information about the man-ager's performance is more adverse and when there is a higher probability that the board is lax (for example, if the board is management versus outsider dominated or if directors lower percentage equity ownership).
Our analysis also indicates that the price offered in a takeover bid depends on the bidder's beliefs about the board's vigilance, because vigilance affects the of offer success. Somewhat surprisingly, for a range of sufficiently high values of the ante probability that the board is vigilant, the bid price is decreasing in the probability that the board is vigilant. Thus, target shareholders may benefit from the bidder believing that there is a likelihood that the board is lax. However, there is a critical value of the
probability of the board being a lower probability may lead to no takeover bid made. Thus, a sufficiently strong suspicion that the target board is lax can hurt the target shareholders in that there are even fewer attempts relative to the social optimum.
related to our work here is theoretically examines
The three key differences First, Hirshleifer and
board as the manager. This gives effect, so that the
enhance its own reputation
68 D. HinhIo'fc~ K ~ ~ k o r / J o u m a l Corporute I (1994:
Second,
Hirsh-
takeoven. focils
has
types good (Gt (T). Firm
belief y E (0.1) i y
rnanage- (type).
that r i.e..
~ r ( x a i l ~ ) ~ P r ( x ~ f l T ) a n d P r ( y a ~ l G ) ~ P r ( y > f l T ) . (1)
P x, x^ x
~'igilant
the a as well decision
accept a takeover processi~ig its information the manager as well coilveyed by the bid.
A la* board ignores manager's perfor- milncc and ncvcr the extent possible, it even resists any
A. of Finance 63-W
protect its members' job security. Hirshleifer end Thakor assume that all the surplus from the takeover accrues to the target firm's sharehold-ers, whereas the bid price is determined endogenously here. Third, leifer and Thakor do not examine price reactions to board firings and takeovers. Our principal focus is on the determination of the bid price and on the price reactions to board firings as well as the management turnover implications of This is motivated by a growing body of empirical work in this area.
The rest of the paper is organized as follows. Section 2 describes the model. Section 3 the analysis and tbe results. Section 4 interpret? the equilibrium and relates it to the empirical evidence. Section 5 concludes. All proofs are in the Appendix.
2. The model
The basic setup is as follows. There are two of managers: and terrible value is higher with a good manager than with a terrible manager, but the manager's type is a priori unknown. The commonly known and shared prior is that there is a probability that the incumbent manager is good and a probability - that he is terrible. We assume that the board and the acquiror privately receive signals of rial ability Let x and y be noisy indicators of managerial ability that are privately observed by the board and by the potential outside acquiror, respectively. Let R versus D denote retention versus a board dismissal of the manager, and let A denote an acquisition attempt. We assume and y arc positive indicators.
for every in the range of with strict inequalities for every except at the end-points of the range of values of x. Further, neither nor y is a sufficient statistic for the other.
The average stock price reaction to a dismissal as well as the likelihood and size of a takeover bid may depend on the perceived vigilance of the board of directors. We allow for two categories of boards: and lax. The objective of a vigilant board is to maximize the expected value of cash flows accruing to shareholders, net of a random, non-pecuniary cost suffered by board if there is takeover. It therefore makes its initial dismissal/ rctcntion decision as its subsequent of whether or not to
bid based on a rational of own as that potentially
informatior: about the incumbent dismisses him. To
Hirshleifer. V: Thakor/Joumal Corporate (I%)
exampie,
lower Kravis
know
inside- 45 w
1 - w
o commcn kx~wledge. G > 0
f(S) [S, 81 F(6)) beviewed
target the t 0
k t 7(P) P
P, T(P) <PC r(P) > T'(P) 0 > PC.
6, (G), p then R (T), 0.
x = a ~ r ( t a r g e t -uPr(G) PPdtarget =PPdG).
* rea~tion the appoinltnent h a r d mvnlbers
D e of Kini al. (1993)
' might of wuld contxding
lo mnndgers reluctsnt terminiite unprofitahle
D. A. of Finance I 63-90 69
takeover attempts, but there may be a critical price level above which an offer will sometimes succeed. We assume that ihis success probability against a lax board is increasing in the bid. The idea is that even a board that seems to be allied with management may side with an opposing bidder (owing, perhaps, to the threat of lawsuit). For in the RJR-Nabisco LBO, the board was viewed as subservient to CEO Ross Johnson, but eventually the board accepted an arguably offer from Kohlberg, and Roberts' instead of Johnson's buyout offer. The success of an offer is assumed to be stochastic to reflect the fact that bidders and investors often do not with certainty whether a board is willing to remove the top executive.
A high versus low probability that a board is lax is intended to correspond at least partially with Weisbach's (1988) categorization of boards as (management-) versus outside-dominated. Let be the prior probability that the board is vigilant, and the prior probability that it is lax. Neither pstential acquirors nor investors know the board's type, but the value of is
We assume that even a vigilant board has a random cost (with density function uniform over and cumulative density function
of being acquired. This may as a control premium, which reflects the fact that a takeover may also displace the board of an acquired target. We assume that this cost is bidder-specific, so that the board and the bidder assign same probability distribution at = over 6. The board does not incur this cost if it fires the manager itself.
be the probabiiity that a lax board will accept a takeover bid in which a price is offered for the firm. We envision that there is a critical price such that = 0 for all P and 0, > for all P That is, for a possibly large range of prices, a lax board will staunchly resist any takeover attempt, but will tend to capitulate when faced with higher bid prices.
Firm value has two components. The first, A, arises from past decisions. The second, arises from new decisions. If the manager is good then B = and =a. if the manager is terrible then A = B = The expected values of these two components are: manager is good) and = manager is good)
Rosenstein and Wyatt (1990) find a positive stock price to of outside board members, consistent with a role for outside distinct from that of insiders.
evidence ct referred to in the Introduction also supports the notion that board vigilance increases with outside domination.
There are other reasons why a board be reluctant to fire. For example, members a board could lose reputation by tiring a manager, since this be tantamount to that the board made an error in the original hiring decision. This is similar Boot's (1992) prediction thnt will be to projects.
The board obsewes x, potential acquiror Y but x.
i.e.. y that
retaia/dismiss o$%irving x 0, 1 p
Br 3' Ex',
(x, Y). R I x). 3" E ( B x), ps E(A I .t,y) = aPdG I x,y). E r s E(BI x,y) PPdCI x,y). ENM
BNM = py. x i.i.d.
x Pdx=OIC)=p, , Pdx- l I G ) = p l , P d x = 2 1 G ) = 1 -p,-p,, P d x = 017') =q,, Pdx 1 I T) =ql , Pdx 2 ) T ) = -q, -9,.
x nffilliared Milgrom Weher, 1952).
(i) PI, <o,,. (ii) PI <q l , (iii) p,[l -PI, -p1 l <q,,[l -q,, - q l l and
(iv) P,[I - ~ " - ~ l l > q l [ ~ -%I-%I. (R-1)
(R-1) 0 < ( < z
I 6 have (de nola)
6 >
> 0.
delay !he B i ~ n The of ?\'aI:ing
dcciiiions the
first after which a observes not If the board dismisses the incumbent manager, it replaces him with a random draw from the available pool of managers, there is a probability
the replacement is a good manager. The board does not know y when it makes its decision after x. We simplify by assuming that and y can take on only the values of or 2. Let and
denote the expected values of the two components of firm value, given x. Similarly, let these expected values be and conditional on the pair
That is, = E(A = I = and = = In addition, we let
denote the expected value of the second component of firm value under a new manager. That is.
Conditional on the manager being good or terrible, and y are random variables. Thus, the following distribution of also applies to y:
= and = 1 Note that this implies that and y are random variables (see and
We impose the following restrictions on these exogenous parameters:
Throughout. 'R' will refer to a set of parametric restrictions. We will explain the need for later.
There is a probability 1 that the acquiror will have good man-ager, and a probability - that it will an average manager, whose ability equals the expected value of the ability of a randomly drawn manager. The bidder knows whether its manager is good or average, and makes its bid to maximize his expected profit from acquisition. We will assume shortly that it never pays to make a bid if the bidder's manager is average. so this assumption is equivalent to assuming that there is a probabil-ity that a potential bidder will arrive.
The bidder faces two costs associated with takeovers: a bidding cos: 5 0, and a control transfer cost K The bidding cost prevents the bidder from making a frivolous bid even though its own signal reveals that the target manager is very good and hence that its bid will be rejected with high probability. The control transfer cost. if sufficiently large, ensures that it does not pay to acquire a firm which has just fired its manager.
We assume that there is a cost of in removing a terrible manager. cost is that the expected value arising from new
is lower when a terrible manager is retained longer. Specifically, if board fires the manager before a takeover bid arrives and replaces him
to
de nolso B IS BNM; aftcr unsuccessfu! iakeover
W E ( I .r, [ I w]BN* ' , ( 2 ) ~(0.1) . game
0 man&ger thc m. the
0. 1, cbserved
x, D) (R)
the ebserves whethcr whether
S > O
accept 1% acwpts
y. i n f ~ .
biuder
strategie~ strateg D type
board will manager observed x hP!ow
jcnds target D 1, c + l i
x , realized
8,, :hat, ( x , y), will be accepted if 2nd oniy 8 g gx,. Perfect Bayesian
with a manager, the expected value of but if the manager is only replaced an attempt, the expected value is
B bid price P ) + -where w The structure of the is as follows. There are four dates. At date the selects a project. If the manger is good, project will succeed and yield a payoff of A = If manager is terrible, the project will fail and yield a payoff of A = At date a signal about the manager (or equivalently about project success) is by the board of directors. Based on the board, if vigilant, decides either to force a resignation (dismissal, or to retain the manager. A lax hoard always retains manager. At date 2 the acquiror his own manager is good or terrible, observes y, and decides to submit a takeover bid of price P to the target's board. The bidder incurs a cost b if he makes a bid. At date 3, the vigilant target board observes its private cost
of being acquired. The vigilant board aggregates its own x with its inference about y drawn from the acquiror's bid price P and makes its decision to or reject the offer, conditional on the realized 6. The board's action depends only on the bid price P. If the target board the offer, then the bidder incurs a cost of control transfer K and acquires the firm.
We have assumed that the board does not observe It will, however, in equilibrium be able to y from the presence or absence of a takeover bid and the size of the bid. If there is no bid (and the board has not dismissed the manager previously), then the (vigilant) board will still retain the man-ager, since the failure of an acquimr to make an offer conveys favorable information about the manager. 'The will make an offer only if there is a gain to removing the manager that exceeds the total cost K + b.
3. The analysis
3.1. Conjectured equilibrium
The of targets and bidders are interdependent. The target board's at date 1 with regard to or R depends on its own aria its signal x. We will show that a vigilant dismiss the if its
signal falls a critical value x * . The bidder's strategy at date 2 de on the board's decision of or R at date as weli as its
signal y. Finally, the target board's strategy at date 3 depends on its signal the bidder's offered price P and the target's 8. It should be remembered that the acquiror does not observe x.
Let be the critical value of 8 such given the pair the bid if Then in the proposed
72 D. Hirshleifer, K 771akor/Jmmral Corpomr~ fIW4i
h
( i ) 0 D (ii) x If
observes 0 P(0) date - 1 P = P ( l ) P(Q) S < S,,,, P(1)
S g St,; noiqo If y
(iii) x 1 Acquiror 0 P(0). 6 < 6,);
othewise nor0 I , P=P(l) ,
and
(iv) s(P).
srrarcgies
Sincc gaod Furthermore,
Subcase: x l (:aigi:ii~~ii ig~ore 0 bidsrr hr_.n:d's utiiity
.TN' wE" w]BNM, (3) E(B x P(0)) E( x = 0) =El0 ,
infcr P(0) 0. (3) captures ;he ther- i s waiting manifested B"') the II the uti1i.p
/j-lIl * jjlll, (4)
'' dcscrihc hclirfs from idi-cquilihrium hills given later.
A. of Finance I 65-90
Equilibrium, we have the following. In this we assume the acquiror has a good manager.
Vigilant board observes x = at date I and chooses Vigilant board observes = 1 at date 1 and chooses R. the acquiror
y = for target manager, it bids F= at 2. If acquiror observes y for target manager, it bids at date 2. Target board accepts bid of at date 3 only if accepts bid of at date 3 only if otherwise it rejects bid and dismisses the manager, replacing him with a de manager. acquiror observes = 2, it does not make a bid. Vigilant board observes = 2 at date acd chooses R. observes y = for target manager, and bid P = Board accepts bid if
it rejects the bid and replaces the manager with a de manager. If acquiror observes y = he bids in which case board rejects bid retains manager. If acquiror observes y = 2, he does not make a bid. The acquiror bids only if it has a good manager. A lax board never dismisses the manager and its probability of accepting a bid of P is
3.2. The bidders
the bidder's decision is trivial unless it has a good manager, we will restrict attention to a bidder with a manager. the fact that the proposed equilibrium is separating in the bidder's strategies implies that the (vigilant) board will be able to infer accurately the bidder's signal from the observed bid price. We will deal separately with each possible value of y.
Case I: v = 0 x = 1. Suppose first that = is observed by the
board. (We can x = here since the does not come into play in that case?. The if it rejects the bid and dismisses the manager is
+ + [ I -where I = 1, P = = B I = 1, y since the board can
from that y = The expression in notion that a cost of io dismiss the manager, which is higher (as
in a lower if manager is perceived to have lower ability. the board rejects the bid and retains manager, its is
We hcrc and actions arising equilibrium bid levels. A discussion of is
I>. Hi&lpifer, A, TIiakor/Jorrrnal Corporarr Firranee f lW41 6 3 4 0
(3) (4). i.e., ~ N M T~IU. (R-2')
Eli' pPdG I x 1, y 0)
piIp, < q,,q, (R-1).
fi,,, S i.e.,
5,,, = P(O) .hl" WE"' [ I w]BNM. (5) P(0) 0,
x
ill
incumb~lit ( 5 ) (6) c2cdi:i~iiai
'6-81 I - ' [ ( H , +H2i-6)P(O) - [p(0)IZ- H , ( H , +&)I, - (7)
=XI0 wEiU [I-w]ENM,
H 2 = x l " + 13 Suka~e: x 2 x
A"'+ WE2i' [ I - W]BuM. (8) rhe menager,
A-2n D2fl. (9) thc cquilihrium
B N M > EZi1. ( R-2)
73 II!
To ensure that it is better for the board to dismiss the manager if the bid is rejected (consistent with our equilibrium). we will find parameters such that
is greater than , Note that = = = where
since by Now the board will accept the bid if
Let be the value of for which the above holds an equality,
- - - -So, if the bidder bids when y = his expected payoff, conditional on
= 1 and on the board being vigilant, is
where the the braces above appears without a probability because the bidder bids only when he is sure that it has a good manager with whom to replace the target manager.
Substituting in yields the bidder's expected payoff
-where
H , + + -K. Now suppose the board observes = 2. If the board =
rejects the bid and dismisses the manager, its utility is
+
If it rejects the bid and retains iis utility is +
In order for the board to follow prcscrihed strategy of dismissing the manager after rejecting the bid, we need
D Hirrhleifer, A.V. Shokor /Journal Corporole Finance 1 (1994)
B2"=pPr(I;Ix=2, y - 0 )
po[l - po - pl1 < qdl - qtl - ql] (R-1). (R-2). trans- (R-2')
-NM = P(0) -KZ" wBB' [ I w ] B (10)
= 0
[E-s_]- '[(H,+ H, + 8 ) ~ - - P2- HJ(HJ+fi]], (11)
H,IX~'+WE")+[I-W]INM, H ~ = X ~ [ ~ + P - K .
btdder's uptimiration manxgc: srrd aiso
qipnal p 2, i ; i ~ Pdboard
1 R ) = w,(R. - 0). most
T( PIP K E(A = O)] , where E ( A l y * y * ) r ~ r ( ~ = ~ [ Y = y * ) . . ~ " ~ ' - ~ p F ( . ~ = I l y = - j . * ) $ l . * +
-7 .. Pdx l y y * ) A -' - CssSlaiiig fact (7) (11) inaximization observed 5-
+ [ : - & , , ] T ( P ) [ ~ - K + E ( A ~ Y = O ) ] (12)
&,, wC(R. 11 = 0). Obsclvc Pr(x 1 y = 0) < Pdw I 0). The P(O)
rcvcal -0. u~mputing
of 63-9074
Note that
since by Also, given it is parent that is redundant.
As in the previous case. we obtain 8, as a solution to
- - - . The expected payoff of the bidder with a price of P, given x 2. y =
and a vigilant board, is
where
The Now, having observed that the board did not dismiss the
having observed the - bidder revises its prior belief about the board's vigilance. The bidder's posterior belief is given by vigilant y =0, board chose y Explicit expressions for this and other posteriors are given in the Appendix. Now, if the board is lax. the bidder's expected payoff from bidding P is - + I y
= 2 = ihis with and yields the problem of the bidder who has y 0:
- b .
whcrc that = 2 = I y = bidder is choosing to maximize his expected profits, taking as given
that his bid will him to have the signal y In his
P(0) P(1) 0
E (0.1) P( j ) E (0.1) j # Moreover,
P ( 0 ) P(O)
(12). ' P(0) aY,,/aP(O) 0,
H , H , 6 > H , 8, S,, < S ,,,, turn, progability x
x 1. the because tl!? v ~ k c iiijiilg
high;;, snd retaining
makc P(0) i.e.,
Case P(1)
I , =,$I WE'' [ I - ~ I B N M , l2 p K ,
-- ' tllosc 1.f
Sobel. 1987, exmnple) not bccausc we examtne
expected profits, the bidder takes into account his posterior belief that the board is lax. This may be viewed as follows. In the proposed equilibrium there are separating bid levels of and for bidders with signal values of and 1, respectively. To ensure that this is an equiiibrium, we will check the incentive compatibility conditions that a bidder with signal y = i does not bid with j and i. if we stipulate beiiefs such that off-equilibrium bids below are interpreted as having been made by a bidder with y = 0 , then must be the solution to the optimization problem of
Now, must satisfy the first order condition = which yields
Note that + + H , + + implying that which implies, in that the of acceptance is lower when = 2 than when = Intuitively, a target board with a more favorable signal about its . .
manager is more inclined to reject bid, both from old decisions i s because there is a greater gain to the old manager.
We must ensure that the bidder finds it profitable to an offer. This follows if (12)with P replaced by is positive,
2: y = 1 The bidder bids in equilibrium. Define
+ + + -
Note that the standard refinements of sequential equilibrium (such as Banks and for are applicable here, a game of two-sided
asymmetric informarion.
D. Hirshleifer, I.: 7hakor Corporate (19941 63-90
x 1 P(1)
Furthemore, x P(1)
BNM <gl*, By,
(R-1) p1[l -P, -PI ] > 40 -9'1.
/ 3 - B m > ~ + 8 , -
( i i ) p - B I 1 > ~ + 8 , -
p B*' 6. - (R-5)
(R-51, (i) (ii) even S - S, c~mbiiia-
signals. Conditic?n {iii! iriip:ies rnat x 1,
of
x
9 d 1 P
i, = o,(R, 1) = Prtboard lboard P( l ) a 9 1 / a P
0.
76 A. /Journal of Finance 1
To make it optimal for the board to dismiss the manager at date 3 if = and a bid of is rejected, we need
in equilibrium the board rejects the bid and retains the manager at date 3 if = 2 and a bid of is received. For this, the necessary condition is and using the definition the following condition is sufficient for the above condition to be satisfied
- More-This is true since asserts that over, to be consistent with the conjectured equilibrium, we make the follow-ing assumptions:
(i)
(iii) - <K + In and imply that if all the surplus were to accrue to the
target and it would pay to displace the manager for those tions of when = 2 and y = the manager is not displaced even by a combination of the best offer by the bidder and a realization the lowest possible S.
Now, proceeding as before, we have
Since the bid is rejected by a vigilant board if = 2, the bidder's objective function becomes
Max
where y = vigilant retained manager, y = I) . Let be the bid price that satisfies the first order condition =
Thus,
Hirsltleifer, A.V. Tlrakor/Jonr~rol 1 11994)
( 1 , 1 ) (2 , O), that 0
> H I > H , . (13) B(1) P(0).
Incentirv strategy
.. P(I), 1 , determined. (IC)
obsewes 0 P(0) P(1).
PJ1) P(1) (R-7) PJ1) tound P!!!.
1C hidder observes P(1) P(O), i.e.,
P,,(1) P(1) equalie. l
D. of Corporate Finance 63-90 77
Now let us assume that is a more favorable signal than as would be the case if it is very unlikely a signal of would come from a good manager. This means that
Note that I , and I , Then comparing and (16) we see that >
3.3. compatibility of the bidder's
w e .
will first discuss how the equilibrium bid price when y = should be One incentive compatibility condition that must be satisfied is that a bidder who y = will bid and not
Now let denrte the value of for which holds as an equality. Thus, provides a lower on the equilibrium value of The other condition is that a who y = 1 must choose over
Let denote the value of for which (R-8) holds as an In other words, the y = bidder maximizes profits by making the lowest
D. Hirslrleifer, A.V. Thakor/JoumalofCorporale Fitlance flP(h4163-90
1. P,(1) P(1).
PJ1) < P,(I). !?I) g P,(l), P(1) PJI). :Thus,
(R-8) i i P(1) E (P,(l), P,(l)), P(1) k?l), (R-7) (R-8)
1C
op(R, = 2) Pdboard R). (R-9)
P(1) tiider P(0) y Finai:);,
check y bidder
rarget).
S!rureh~ of ir t*igilunr target
1: x 0 Firs:, miisi ensure there will bid ii tile manager
reo!rlc!lm :ha; is suificient
[ T u + p - K - 8 - 6 1 - - [ X ~ ' + B N M ] < = 0 . (R-11)
x 0,
A7' = aPr(G l x 0). biddcr - 0 I
x - 3 , EU(R I x 0).
EU(R x
78 I
possible offer consistent with his being viewed as having y = Thus, provides an upper bound on the equilibrium value of I! can be verified that
Now, if then we set the equilibrium bid = (R-7) holds 2s an equality and is slack. On the other hand,
then we set = and both and are slack. Another condition is that the bidder will abstain from bidding i f y = 2. That is,
where y vigilant I y = 2, board chose The exoge-nous parameter restrictions implied by ensure that the bidder will find it unprofitable to bid when y = 2. It can be verified, by monotonicity, that the will also find it unprofitable to bid when = 2. we need to that if = !, :he makes a positive profit by bidding iconditional on having a good manager to replace the incumbent in the
This participation constraint is
3.4. board
Case = wc ihat be nu is dismissed.
The parametric for this is
Now, given = if the board dismisses the manager its utility is
where = If the board retains the manager, a bid is received if the observes y or and has a good manager. The expected utility of the board from retaining the manager. given
= is I = 0 )
D Hirshleifer, Y. Tl~akor / l o ~ ~ m a l of Fina~ice 1 (1994)
+ [ r ~ P r ( ~ = 2 I x = o ) + ~ - t ] [ R " + w P n + [ l - w ] B N M ]
+cPr( y 1 l x O)[I F ( s , , , ) ] [ A ~ ~ ~ +WE"' + [ l W ] # N M ]
-NM +.fPr(y = O l x=O)[1 -- F ( s , , ) ] [ ~ ~ + wBW w ] B 1, (I81 2" B"
given
- = aPr (G I x ,
B" E PPr(G I or
x
- - A"" 5"". ohser:ir.g iliai rnere - - ,run ; ,,,s::z
,7 i l w ] g N M . ENM > son, will (R-2),
BNM p. x 0 P
P(O), S G 6 > 6,
6,,=P(O) -2'"-WE"- [ I -w]BNM, (I9)
- ~ ~ - a P r ( G I x = 0 , y=O), - EM = -- -pPr (GIx=O, = 0 ) .
0, -NM X"+B > E U ( R I x = O ) .
x
XI ENM (20)
x' = a P d G l 1). i?
A-1" w@n [ I - ,,,]jjNM < ~ - l n gin,
BNM <El,, .
A. Corporate 63-90 79
= - - -
where and are the components of expected firm value arising from old and new decisions of the incumbent, respectively, that x = O and that no bid was received. That is,
A" y = 2 or bidder did not have good manager),
x , y = 2 bidder did not have good mriiager).
What should the (vigilant) board do if it retains the manager despite = 0, and then fails to receive a bid? If it retains the manager. its utility is
+ If i t dismisses him after is no bid, its utility is + - Thus, if it pay for the board to
dismiss the manager even if no bid is received. Given it is easy to verify that >
+ [ I -
If the board retains the manager with = and then receives bid of it will accept the bid if 6,. Given our preceding arguments, it will
reject the bid and dismiss the manager if 6,. Here satisfies
where
y
For incentive compatibility of the board's dismissal of the manager when x = we need
(R-12)
Case 2: = I If the board dismisses the manager, its utility is
+
where x = If the board initially retains the manager, should later dismiss him if no bid appears? We shall assume that the board will
retain the manager in this case. A sufficient condition for this is
+ + +
which reduces to
80 Hiishlrifer, A.V. 'Thakor/lournd Fmmce (IW)
EU(R I x - 1). EU(RIx=l)
-NM +CPr(y = O l x - l)[l - F ( s , , ) ] [ ~ ' + w ~ " + [ ~ - w ] B ] +f Pr(! ! 1 r -- l)[l F ( s , , ) ] [ ~ ' WE'' [I w]E"M] + [ ~ P ~ ( Y - z I x - I ) + 1 - , ~ I [ A ~ ~ + B ' ~ I . (21)
1C
EU(R <xi BNM.
Case x =
l~!I!if; i;
EU(R 1 2), EU(R x =
Pr( y - 1 I I 2)([X2' wE21 (1 w)ENM]
+?r(y = 01 x = 2)([1-~(6,)][.4~ iiiB2% [ ~ - w ] B ~ ' ~ j -N M + [ ? r ( y - ? l x = 2 ) ( + I - ~ I [ A ' Z " + ~ ~ ~ " + [ l - ~ l B 1. (u)
IC wndition,
E U ( R I X = ~ ) > ~ + B ~ ~ .
the main
exists ualucs (R-1MR-15) unlues,
equiIibrium is a
m
D. ofCorporate I 63-90
The board's expected utility from retaining the manager, is
= - + + -
The condition which guarantees that the board will choose to retain the manager conditional on x = 1 is
(R-14)I x = 1) +
3: 2 Given our earlier parametric restrictions, it will obviously not pay for the
board to dismiss the manager if the manager is initially retained and then no bid is received. If the board dismisses the manager. its
If it retains the manager. its expected utility, x = is given by I 2 )
+ = + + -+
To ensure that the board will not dismiss the manager when it observes x = 2, we need the following
(R-15)
We can now state result of this section.
Proposition 1. that restrictions conjectured
There an open set of exogenous parameter are satisfied. For these parameter
Perfect Bayesian equilibrium.
such the
An important aspect of this proposition that the price offered in the takeover attempt reveals the bidder's private information about managerial
performance. underlvicg !hc iiicenrive fslliiws. When
x 1 1, = 1 x 0 the
x = y thc (1, 1).
that 1-igilant
1,
1 )I - 0
(P(1) > P(0))
equilibrim ir,
F~position exists opt1 culues such criticalprobability ~~igilnnce, w,(0,1),
P(y) o o E (o,,l) E (0,l). calues takeover attempt w
s
~ f f b;.
removs!, alinost (Brickley
expected
McWilliams
Jarrell
The intuition compatibility of such a scheme is as a potential acquiror observes that the manager was not fired, he revises upward his probabilistic assessment that the target board observed = or 2. Moreover, when he observes y = he assesses the probability of x to be higher than the probability of = or probability of 2. This is because x and are affiliated random variables. Now, conditionnl on pair of signals the potential acquiror knows
he can increase the probability of acceptance of the bid by a board by bidding higher. Moreover, the probability of acceptance by a lax board is weakly increasing in the bid price. Thus, having observed y = a potential acquiror knows that the probability of acceptance of the bid is strictly greater with a higher bid. The key is that the bidder has a stronger incentive to bid more when he has y = than when he has because the expected value of the target firm due to old decisions is larger in the former case. Thus, separation through the bid price can occur as part of a Perfect Bayesian Equilibrium.
We will now examine some of the properties of this next result.
2. There an set of exogenous parameter
our
that, for some of board the equilibrium bid price is decreasing in for all and y There also exist exogenous parameter such that no is made for sufficiently low.
An interesting aspect of the model is that suspicion on the part of potential acquiror that the target board may be lax may actually make the target board's shareholders better iiiducing a higher bid. The intuition is that the bidder is setting his price to overcome the takeover resistance of a lax board, so that a stronger suspicion of laxity causes the bidder to offer a higher price. This is consistent with the stylized fact that anti-takeover charter amendments, which reduce the probability of takeover (Pound, 1987) and thereby protect the board from hostile are always voluntarily approved by shareholders et al., 1988). Of course, in our model if the suspicion of laxity becomes strong enough, the payoff to the bidder from submitting a bid may be negative, given bidding and control transfer costs. In this case, there are on average too few takeover attempts relative to the social optimum. This is consistent with the evidence that proposals of some anti-takeover amendments are associated with nega-tive stock price reactions, whereas others are not. For instance, (1990) finds a significant positive stock price reaction to the proposal of anti-takeover amendments when managers own less than 10% of the target's shares, but a negative reaction when managers have larger holdings, see also
and Poulsen (1987).
D. Hirsllcifcr. A.K 77mkor/lwmal ofCorpomre (1W)
mb
conserve
straightfonuard fcature neither x y. model thcn
discusion.
Tumor-er
annlyriz f ~ i e succcssiui takeover
change. 'L'his McConnell
evidence Klein Rosenfeld (i.e..
becomes
x 1 0,
x = 0. assumed pr im
y y
82 Finance I 63-90
4. Interpretation of equilibrium relation to empirical evidence
Our equilibrium has numerous implications which we now discuss and relate to empirical evidence. Although to on algebra we have not computed stock price reactions to the equilibrium strategies of the board and the aquiror, it is to add this by assuming that investors know nor The yields price reactions consistent with the ensuing
4.I . implications
There are three principal implications of our management turnover in targets. First, always. triggers a management
implication is consistent with the evidence in Martin and (1991) that the turnover rate for top managements of targets of
tender offers increases significantly after the takeovers. Second, and even more interestingly, even an unsuccessful takeover attempt leads to higher management turnwer on average in the target. This is consistent with the
of and (1988) that top management changes often follow targeted repurchases of shares held by potential bidders green-mail). Third, since a lax board always resists a takeover and never dismisses the manager, we should not expect all failed takeover attempts to trigger internally precipitated management changes. Rather, there will sometimes be firings and sometimes the incumbent will be retained.
4.2. Management performance implications
In the equilibrium we have described, a firm a takeover target when a potential bidder observes a signal of the target manager's perfor-mance that is so adverse that it more than offsets the favorable information conveyed by the fact that the board chose not to dismiss. The manager of a takeover target has = and y = a combination that leads the potential acquiror to a lower posterior assessment of his ability than the average ability of the pool of retained manages. That is, his firm in retrospect will be viewed as doing poorly relative to its cohort group. The same is true for firms in which managers are dismissed by their boards, since in those cases In our model we have that the commonly held on the incumbent are the same as those on a manager. In practice, however. investors' priors about the incumbent manager are likely to be influenced by prior managerial performance that was publicly observable. Hence, if the firm has been doing poorly, investors will have a relatively low about the incumbent. whereas their on his replacement will be higher. Our model implies a greater likelihood of takeover in this case, an observation that is
(1985) (1986) perfomnance
Tobin's Coughlan (1985) ihat
reaction !c z &ismissal positivp m zegaiive. y , --- . n~ i t ao i i t t y y
Moreover, o, combination
the
' y o,
ex
"ccentcating information
lax multiperiod setting,
the tirm examp!e. (1988) fraclions
between lax prim seems (1988)
2nd ;\ possihle model. resignations ofcur
that uncor:clated and as
intcqreiation ir consistent benveen f3r
after
consistent with the findings of Hasbrouck and Palepu who report that takeover targets tend to have poor prior as mea-
and Schmidt find lowsured by q. Similarly. stock returns lead to relatively high management turnover.
4.3. Marker reaction to dismissals
Our analysis implies that the market of the CEO can be either To see this, suppose that the prior
that the incumbent manager is good is very low, relative to the
for his replacement. suppose that the prior probability that the board is vigilant, is also very low. This implies that public information about the manager as well as the board is quite adverse. The firm's stock price will reflect this pessimism. So when the manager is fired, investors view as quite significant expected improvement in productivity achieved by replacing the incumbent with a new manager. Thus. the more adverse the public information about the manager, the greater is the positive stock price reaction to his dismissal. On the other hand, if the market's prior belief reflected high values for and then a dismissal constitutes a 'negative surprise' since it indicates poor managerial performance and the incidence of dismissal costs. It is also intuitive that the higher the ante probability that the board is lax, the greater will be the positive stock price reaction to the dismissal. The greater the investors' initial pessimism about the board's vigilance, the more pleasantly surprised they are at the dismissal of the incumbent.
this effect is the favorable conveyed by the dismissal that the board is vigilant. Although not explicitly modeled. i t would be easy to include a value component that is higher for a vigilant board than for 3 one. This would implicitly rrflect the fact that in a
learning that the board is vigilant provides favorable information about how well will be managed in the future.
For Weisbach classifies boards based on the of boards that are managers versus outside directors as insidedominaled, mixed, or outside-dominated boards. The implication of a positive link the probability that a board is a dismissal inconsistent with the evidence provided by Weisbach returns on announcement of resignation are positive (with significance sensitive to the event window) for outside mixed boards and close to zero for inside boards. explanation for this is that, contrary to our do even when the board is management dominated. These may be voluntary resignations are with poor performance which may not convey much news relevant for the stock price. This
with Weisbach's evidence that the relationship prior poor performance and the probability of CEO turnover is stronger outside boards than for inside boards. It would be of interest to examine the price reaction to resignations for outside versus inside boards normalizing for the prior performance of management.
and the reaction to that the abnormal
el (1988)
findings (i9!?8? and Eruner 11989) thc !he pulu;iiiy periormance oE
represen!ation 1991) vigilance
intertemporal
knowledge.
rakeoc3er resistance
good Gi finn's aiways
6 "
being
Brickley
insiders. Brickley al.'s
tzrget (e.g. 19881,
(1985)
average Poillscn (i988)
Handa (1989)
1%) Huang Walkling (1987) Rnd ahnorn;al that opposed hy th2n
This implication is consistent with the findings of Warner al. that a forced resignation can be either good news or bad news. It is also consistent with the of Weisbach and Bonnier that more adverse available information about the the manager. the more positive is the stock price reaction to forced resigna-tion.
The increase in outsider on boards in the last decade (see Lublin. suggests an increase in board over this period. Our analysis would then imply an weakening of the stock price reaction to a dismissal. This time series prediction has not yet been tested, to the best of our
4.4. Marker reaction to
Our analysis implies that takeover resistance can be either bad news for the target shareholders. Since a vigilant board acts in its shareholders' best interests, any takeover resistance on its part conveys the good news that the manager is probably good. so that the value from old decisions is high, and there may be further benefit from retaining the old manager. Thus, if the target firm's board is generally perceived to be vigilant and its is sufficiently low, the stock price of the iarget will increase when its board resists a takeover. On the other hand, if the board is perceived as
lax or controlled by the incumbent manager, takeover resistance by it will lead to a price decline because the resistance strengthens investors' beliefs about the board's type and the low ability of management. Consistent with this prediction, et al. (1992) provide evidence that the enact-ment of poison pills leads to a positive stock price reaction when the majority of the board consists of outsiders. and a negative stock price reaction when the majority of the board consists of (Our prediction predates
et evidence and independent modeling of this phenomenon). While some takeover defensive measures are bad news for share-
holders poison pills; see Ryngaert, 1988; Malatesta and Walkling, others are not. For example, Jarrell argues that his evidence on defensive litigation is consistent with shareholder interest. Partch (1987) finds that the stock price reactions to dual voting class recapitalization is not on
negative. However, Jarrell and do find a negative reaction in a more recent sample oi dual-class recapitalizations. Moreover,
and Radhakrishnan find that defensive leveraged recapitaliza-tions are associated with positive target abnormal returns.
and that announcement day returns are slightly higher takeover hids are management for unopposed ones.
Bladers
sharchcklers
(Brickley a].,
upposition g d information
takeover
occnrs i~~!ormation bid&
insider-
new. If la,
anre
et 1988).
Our model predicts that the bid premium in takeovers is not monotoni-cally related to the probability that the board is vigilant. must beiieve that this probability is sufficiently high in order to be induced to bid. However. for all values of the vigilance probability above a critical value, the bid premium declines as the vigilance probability increases. Thus, the target
may be made better off by suspicion in the acquiror's mind that the board is lax, which may perhaps explain why shareholders almost always approve anti-takeover charter amendments which protect the board from hostile removal
5. Conclusion
We have examined the interaction between the internal mechanism for corporate control, as represented by actions cf the board of directors, and the external mechanism for corporate control, as represented by takeovers. Our objective has been to study the determination of the bid price as well as the behavior of the board and that of the acquiror in this setting. The principal results and implications of our analysis are summarized below.
1. The board of directors, acting in shareholders' interests, will sometimes oppose a takeover, and this can be news for the firm, because the board has favorable private about firm value. Such resistance can sometimes cause the bid to fail.
2. Unsuccessful takeover attempts may be followed by a high frequency of management turnover. This in our model because a takeover attempt conveys adverse possessed by the about the target's manager.
3. Investors' perception of the board's vigilance will influence the market's reaction to takeover resistance. If the board is viewed as being dominated-ana hence more likely to be lax-then defensive measures against a takeover are bad news. On the other hand, for boards perceived to be outsider-dominated, they are good
4. the stock market believes that the board may be then a forced resignation can be either good or bad news for the firm. A positive effect is predicted to dominate when there is more adverse public information available about the manager's performance and when there is a higher ex
probability that the board is lax; for example, if the board is management rather than outsider dominated.
5. For sufficiently high prior beliefs that the board is vigilant, a decrease in the prior probability that the board is vigilant can increase the price offered in a takeover bid.
Our
actions since these
The
Proposition
q,, = 0.02, p,=0.16, q,=0.17,S=O, 8 = o=0.8, b=10-" ww0.8, K=0.109, y =
fi 6 0.35, s (P ) ;= [10.72]-1~21,,,,2, ,)(P), I(.) fenction 11.
(R-1) (R-151,
x
P + {P(O), P(1)). x 1 P E
([O, P(0)) u (P(O), P(1 ))I, ir (x = 2. P(I),
{x = 1. x cven
P > P(1). x some S:(P),
Pr(y 21 P > P(1)) S <6:(P). P,,, will between
(iii) (R-5) S:( P,,) < 8. - D
Poroposiirion
It (13) aP(O)/aw < P(I) P(I)VU E (o,. 1). T'( P ) > rl'( P)
3-0 (16) nP(l)/aw <
for P(1) =li(t).
I
analysis highlights the importance of the board of directors in corpo-rate control. Even with a well-functioning takeover market, the wealth of the firm's shareholders depends on the reputation and of the board
determine whether a bid is made for the firm and the price at which the bid is made. link this suggests between the 'qualities' of boards of directors and the health of financial capitalism is an indication of the need for further study of board composition and motives.
6. Appendix
Proof of
Consider the following exogenous parameter values: p,,= 0.01, I,
0.65, = 0.3, = where is the indica-tor on the interval (0.28. It can be verified that all of the restrictions on exogenous parameters, through are satisfied. Thus, the set of exogenous parameters for which the conjectured strategies and beliefs constitute a Nash equilibrium is nonempty. We will now show that this is also a Perfect Bayesian Equilibrium. Since can never be observed by any party other than the board itself, the only out-of-equilibrium move is for the bidder to bid Given an out-of-equilibrium bid P, suppose the board believes that y = 2. Then, if = and
is obvious that the board will reject the bid since the combination y = 1) leads it to reject a bid of and this combination implies the same firm value as y = 2). Clearly if = 2, the board has an stronger incentive to reject the bid.
Now consider a bid of Suppose = I. It is possible that there exists 6, call it such that the board would accept such a bid, wnditional on a belief that = = I and a realized value
Now consider the maximum P, call it at which the bidder make a bid. This is the bid price that renders him indifferent
making a bid and not making a bid. Then, condition of implies that Hence, we can rule out this out-of-equilibrium move too.
Proof of
is apparent from that 1. Suppose exogenous parameter values are such that = Assume that 0.
2
for the relevant range of P values. It is clear now from that I. It is straightforward to verify that there do exist exogenous
parameter values which (Take, for instance, the parameter
Hirslrleifer, A.K Thakor/Jurdr~~ol Finanre 1 09941
1.) not q w O d P ) OVP G F,,,,,, P,,,
grn o >
Pr(y 0, lboard X Pr(board =
( ~ r ( 0, X Pr(board
\ 0, x Pr(board
Pr (x= 1Iy = Pr(x= 1 IG)Pr(G I Pr(x 1 IT)Pr(TI - O ) ,
with p<x= 1IG)-p,, P d x = i IT)=q , . P~(GIY=O)=PI,Y(P,,Y +qo[l- y ] ) , PdT I Y = 0) =9,,[l - Y](PI,'Y + 911[l - Y]).
P r ( G l x = l , y = l ) = P:Y < y , p , < q l .
P:Y + 9 X I - y l (A-5)
[ ~ P , + ( I - Y ) ~ , ] [ Y ( ~ -pol + ( ~ - ~ ) ( ~ - 9 l l ~ l ~ op(R, y = 1) =
[ Y P , + ( ~ - Y ) ~ , I [ Y ( ~ -p,,) + ( l - q 0 ) I w
+[YPI (1 Y I ~ l l [ l -01 I ' (A-6)
D. of Corporate 63-90 87
values in the proof of Proposition To see that there may be any bids if is too low. consider the case in which = and = and
arbitrarily high. Then, the bidder's expectea payoff from a takeover is clearly negative if is high enough. By continuity, the bidder's expected payoff will be negative for 0 small enough.
Details of posterior probabilities
= board chose R vigilant) vigilant)
y = board chose R lboard vigilant) vigilant)\
+ Pr( y = board chose R lboard lax) lax)
Note that = 0 )
y = 0 ) + = y (A-2)
Thus,
Similarly,
since
+ -
Hinhkifer, Thokor/Jorrn~ol Curporarr Fif~afrc (19941
Pr(G I 0, p gclod
Pr( CJ I .V - I . y = 2 havc m a n a g r )
- - - - P I [ € ( l - P I , - / ) , ) + 1 - .< lY I - € I ? +[,l lF(l - 9 t ! - ( / l ) 1 - E l i 1 - 7 1
r l )
J,, n j ( 6 ) d R r r - ' ] / 2 [ 8 81 -
88 D. A.V. of 1 63-90
x = = 2 or bidder did not have manager)
(A-IS)
or hiddcr does not a good
+ +
(A-17'
= [h' - -
Pr(G i x
Sohel. Econometrica 647-661.
stock dirtrcssed o f
A.. IYY?. Ioserh: Di..estitures (Sept.). 1401- 1423.
structure: hailking. (Apr.). lbl-180. Br~ckley. 1988. Ownership
;~ntivakeover ::mesdments. Ecu~nomics 20 (Jan./Mar.). 267-242. Drccklcy. Coles enaclmen!
poiso!: Ullpuhlisned Dec. A T . 19x5.
(Apr.). 43-hh. Dehng. 1990. men pcrrpective
va rc i . Franks. 1993. correction
iililurc. paper. S.J. IYXO. hids.
Bell Journal II. 42-64, Handx. P, AII investi~ation recapitaliza-
lions: neu takecrver Mimen. Hashrouck. I.. 1985.
Finance Y (Sept.). 351-362. 1091. control h a r d ~akcovers.
Y.-S. R.A. 19x7. ahnormal with acqu~sition ;~nnouncemcnts: resisthnce.
19 (Dec.). 329-350. G.A.. 1985. litigatian Dcl interests diverge
Ec(rnl~mics 28 (Apr.). 151-178. Jarrell. A.B. Pnulsen. 1987. rcpellcnts stock effect> 111
antitilkrnvrr amendmentr 1980. Jo~~rnal of Fin;~nci;~l Economic> I?. 117-168. Jarrrll. L A , Poulscn. 1988. Dual.class
recent evidence. ZO (Jan./Mar.). 129- J;nsen. M.C. 1988. of powcr anlong corpor;ltc rn:cnagers.
rharcholders i ~ n d dirurtors. Journ;~l of En~nt~rnics ?[I lJan./M:~r.). 3-24,
= 2, y = 2 = or bidder does not have a good manager)
7. References
Banks. J. and J. 1987. Equilihrium selection in signaling games. 55 (May).
Bonnier. K.-A. and R.F. Bruner. 1989. An analysis of price reaction to management change in firms. Journal Accounting and Ecunomics 11. 95-106.
Boot. Why hang on lo and takemers. Journal of Finance 47
Brickley. J.A. and C. James. 1987. The takeover market, corporate hoard composition. and ownership The case of Journal of Law and Economics 30
J.A.. R.C. Lease and C.W. Smith. Jr.. structure and voting on Journal of Financial
J.A.. J.L. and RL. Terry. 1992. The Board of Directors and the of pills, manuscript,
Coughlan, and R.M. Schmidt. Executive compensation. management turnover. and firm pcrfcrrmance: An empirical investigalinn. Journal of Accounting and Economics 7
B., Did J.P. Morgan's add value'! A historical on financial capitalism. NBER working Feb.
J. and C. Mayer. Hostile takeovers in the U.K. and the of management Working London Business School. July.
Crossman. and O.D. Hart. Takeover the free-rider prohlcm, and the theory of the corporation. of Economics
and A.R. Radhakrishnan. 1989. empirical of leveraged A defense strategy. New York University. May.
The characteristics of takeover targets: q and other measures. Journal of Banking and
Hirshleifer. D. and A.V. Thakor, Corporate through dismissal? and UCLA working paper. June.
Huang. and Walking. Target returns associated Payment, acquisition form and managerial Journal of Financial
Economics Jarrell. The wealth effects of hy targets: in a merger?
Journal of Law and G.A. and Shark and prices: The
since and A.B. recapitalizations as antitakeover mechanisms:
The Journal of Financial Economics 152. and J.B. Warner. The distribution
Financial
Kini, 0.. Kracaw S. Mian. IW3. Smeal Buriness
Klein. Rosenfeld. 1988, management (Jan./Mar.). 493-506.
1991.
Malatesta. 1988. ownenhip of Economics (Jan./Mar.). ;47-376.
McConnell. 46 671-688.
McWilliams. IWO. the (Dec.). 1627-1640.
Milgrom. Weher. 1982. ct)mpetitive hidding. Emnometrica 1W9-
Palepu, takeover h u n t i n g 8,
Economics 18.313-339. 1987.
(Oct.). 353-367. Rosenstein. S. 1990. shi~reholdrr
Jwrnal Fronomics 26, 175-191. Ryngaert. 1988. cffect of poison shareholder
20 (Jan./Mar.). Sllivdasani. A.. 1991.
Dec. Shleifer. R.W. I98h. Lurge
4hl-48% J.B.. Wruck. IYM.
Economics 20 (Jan./M;~r.), UI-4Ml. Weishach. .MS.. IVW. turntlver. of Fi~ancial 20
(Jan./Mar.). 4hl-492.
W. and Corporate takeovers. firm performance. and hoard composition. Manuscript. College of Administration. Pennsylvania State University. Feh.
A. and J. Targeted share repurchases and top changes. Journal o f Financial Economics 20
Lublin. J.S., Tense times: More chief executives are being forced out by tough boards. Wall Street Journal. June 6.
P.H. and R.A. Walking, Poison pill securities: Stockholder wealth. profitability and structure, Journal Financial 20
Martin. K.J. and J.J. 1991. Corporate performance. corporate takeovers and man-agement turnover. Journal of Finance (June).
V.B.. Managerial share ownership and stock price effects of antitakeover amendment proposals. Journal o f Finance 45
P. and R. A theory of auctions and SO. 1122.
K.G. 19116. Predicting targets: A methodological and empirical analysis. Journal of and Economics 3-25.
Partch. M.. 1987. The creation of a class of limited voting common stock and shareholder wealth. Journal of Financial
Pound. J., The effects o f antitakeover amendments on takeover activity: Scme direct evidence. Journal of Law and Economics 30
and J. Wyatt. Outside directors. hoard independence. and wealth. of Financial
M.. The pill securities on wealth. Journal of Financial Economics 377-417
Board composition. ownership structure. and hostile takeovers. Working paper. Faculty of Management. University of Calgary.
A. and Vishny. shareholders and corporate control. Journal of Political Economy 94.
Warner. R.L. Watts and K.H. Stock prices and top management changes. Journal of Financial
Outside directors and CEO Journal Economics