Вы находитесь на странице: 1из 27

Journal of

CORPORATE
FINANCE
ELSEVI ER Journal of Corporate Finance 3 (1997) 113 139
Boards of directors and capital structure:
Al ternati ve forms of corporate restructuring
Erns t Ma u g *
Fuqua School of Business, Duke Unit,ersity, P.O. Box 90120, Durham, NC 27708, USA
Ab s t r a c t
This paper discusses a model that combines internal and external control mechanisms in
a firm in which assets can have alternative uses that are in some states more profitable than
the current one. However, restructuring a firm in order to realize the gains from alternative
uses affects managers adversely since they invest in firm-specific human capital. Managers
can be motivated to restructure the firm through their compensation scheme. Alternatively,
investors can acquire costly information on the firm and interfere with managers' decisions.
The main focus is on independent directors, who review and monitor contracts and
managers' compensation. If information is not too costly, directors are the optimal
institution to check managerial discretion and the degree of managerial entrenchment
depends on the compensation of independent directors. However, if directors fail to exercise
control over management properly, takeovers or creditor control become second-best
solutions. If information is costly to transfer, unchecked managerial control may be optimal.
JEL classification." G32; G33:G34
Keywords." Capital structure; Corporate restructuring; Internal control mechanisms; Boards of directors
: Takeovers; Agency theory
* Tel.: (+1) 919-660.7753; fax: (+1) 919-660.8038; e-mail: maug@mail.duke.edu; web:
ht t p: / / www. duke. edu/ ~ maug.
0929-1199/97/$17.00 Copyright 1997 Elsevier Science B.V. All rights reserved.
PII S0929- 1199( 96) 00010- 7
114 E. Maug/Jour nal q[Corporate Finance 3 (1997) 113 139
" The trade of a j oi nt stock company is al ways managed by a court of
directors. This court, indeed, is frequently subj ect (...) to the control of a
general court of proprietors. But the greater part of these propri et ors sel dom
pretend to understand any thing of the business of the company; and when
the spirit of factions happens not to prevail among them, give t hemsel ves no
trouble about it, but recei ve cont ent edl y such hal f yearl y or yearl y di vi dend,
as the directors think proper to make to t hem. " Adam Smith (1925, p. 262)
1. I n t r o d u c t i o n
How can managers in a large corporat i on be held accountable, so that they are
responsi ve to the interests of sharehol ders and still have sufficient discretion over
business deci si ons? This paper investigates the role of i ndependent directors lom
an opt i mal cont ract i ng perspect i ve, and tries to show how this institution can be
underst ood as an answer to this question. The aim is to contrast i ndependent or
outside directors with alternative mechani sms that regulate the agency rel at i onshi p
bet ween execut i ve management and shareholders, such as the market for corporat e
control, or the control exerci sed by creditors through debt contracts.
Adam Smith (1925) bel i eved that the agency probl em bet ween owners and
managers posed such an obst acl e as to make it quest i onabl e whet her the modem
corporat i on with its separation of ownershi p and control is a vi abl e institution.
However, the separation of ownershi p and control has proved to be viable, thereby
i ndi cat i ng that the i mpl i ed agency probl em can be regul at ed effectively.
The model di scussed bel ow focuses on a corporat e restructuring probl em in
order to compare alternative governance structures. The assets of the firm have
alternative uses to the current one, and in some states it is opt i mal to real l ocat e the
assets (e.g. a divestiture). In this case managers have to write off a part of their
fi rm-speci fi c human capital. Hence, the model is set up to cover two el ement s of
restructuring: (1) managers are pot ent i al l y opposed to real l ocat i ng the assets since
they have a private stake in their current use; and (2) if managers rely on i mpl i ci t
contracts which can be breached, then rat i onal l y anticipated restructurings can lead
to underi nvest ment in human capital. Then managers are prot ect ed from such a
breach by gi vi ng them all rights of control and a hi gh-powered compensat i on plan
that mot i vat es them to make optimal use of the company' s assets. Managers can
thus be prot ect ed by an expl i ci t contract that cannot be breached, but this is not
only a very expensi ve way to induce correct decisions, it also fails to solve the
underi nvest ment probl em. Out si de investors (creditors, t akeover raiders) or a
supervi sory board of directors can pot ent i al l y al l evi at e this probl em. However, in
order to check managers' deci si ons effect i vel y they have to acquire costly
information.
Independent directors are anal yzed as agents separate from execut i ve manage-
E. Mau g / Journal of Corporate Finance 3 (1997) 113-139 115
ment whose main function is to monitor and renegotiate contracts, but who are not
part of the productive activity of the firm as such. In particular, they have no
vested interests in specific operating decisions. They are also different from other
outside shareholders by being able to negotiate bilaterally with executive man-
agers. Directors' ability to negotiate over the remuneration of executive managers
is crucial to the model here. They can induce an optimal use of the company' s
assets without breaching implicit contracts. The argument below emphasizes the
essential role of their independent j udgement. Weak directors who are not
independent do not only harm shareholders through excessive pay awards - an
aspect usually emphasized in the public debate - but also because they reduce the
link between pay and performance.
Capital structure can be used to transfer control from managers to security
holders: either to creditors in bankruptcy, or to raiders in takeovers. These
mechanisms can be effective in reallocating resources. However, they can breach
implicit contracts on which managers and other stakeholders in the corporation
rely when making firm-specific investments. Conversely, less disruptive mecha-
nisms like friendly takeovers may fail to provide sufficient information to out-
siders so that the restructuring potential of the firm is not exploited.
The following section relates the argument of this paper to the literature. The
remainder of this paper proceeds by developing the model and deriving the
first-best allocation as a benchmark case (Section 3). Section 4.1 discusses the
case where managers are not constrained by other agents and subj ect only to a
compensation contract. Section 4.2 is devoted to the analysis of corporate boards
and presents the main results of the paper. Sections 4.3 and 4.4 investigate to what
extent capital structure can serve as a substitute for independent directors and
analyze takeovers and debt contracts. Section 5 discusses empirical implications
and concludes.
2. D i s c u s s i o n o f t h e l i t e r a t u r e
Berle and Means (1932) provided the first comprehensive analysis of the
organizational solutions that have been developed in order to balance the necessity
of managerial discretion against the hazards of unchecked managerial power. In
their seminal paper Jensen and Meckling (1976) stress the control rights of
different members of the corporation in the context of their ' nexus of contracts' -
approach. This view was further elaborated by Fama and Jensen (1983a,b) who
relate the structure of residual claims to the organisational structure of the firm.
Their distinction between ' decision maki ng' as an executive function, and ' deci-
sion control' as a supervisory task corresponds to the division of tasks between
executive managers and independent directors in the model presented in this paper.
This distinction is also emphasized by Lorsch and McIver (1989). Formal discus-
sions of the role of corporate boards as distinct from executive management
116 E. Maug / Journal of Corporate Finance 3 (1997) 113-139
appeared only very recently (Hirshleifer and Thakor (1991, 1994), Warther
(1994)). However, these papers emphasize the role of boards in dismissing
underperforming managers, whereas the model discussed in this paper emphasizes
the role of the board as a negotiator over managerial pay. Berkovitch and Israel
(1996) analyze the board as an institution representing different stakeholders.
The literature on the market for corporate control is vast and will not be
discussed here. 1 The most important aspect here is the hypothesis of Shleifer and
Summers (1988) that takeover premia are largely explained by transfers of wealth
from corporate stakeholders to shareholders and a breach of implicit contracts (see
Schnitzer (1995) for a formalization of this argument). This paper emphasises the
ex ante incentives to invest in specific skills. These incentives are impaired if
implicit long term contracts are vulnerable to breach in future reorganisations of
the firm, even though the transfers caused by such a breach may be small relative
to other gains from a takeover. 2 The paper adopts the ' substitute hypothesis'
which holds that boards are an alternative mechani sm to hostile takeovers. 3
The analysis of debt contracts follows a more recent approach that views debt
as a mechani sm that allocates control contingent on the state between insiders
(managers) and outside investors in a world where contracts are incomplete (cf.
Aghi on and Bolton, 1992; Zender, 1991). This literature has opened up perspec-
tives which view debt as a device constraining managerial discretion (see Hart,
1993; Hart and Moore, 1995; Jensen, 1986). However, such a use of debt causes
problems similar to those with takeovers, since control is assumed by security-
holders who are not bound by implicit contractual arrangements.Berkovitch et al.
(1994) show that bankruptcy law can mitigate this problem.
3. De s c r i pt i o n o f t he mo d e l
Consider a proj ect which has been undertaken in the past and yields a state
cont i ngent payoff y~ where s denotes the state of nature. The payoff y~ depends
on the state which can take on three values, namel y H, M, and B
( hi gh/ mi ddl e/ bad) . The assets of the proj ect have alternative uses which are not
known until after the state of nature has been revealed. Realizing the returns from
alternative uses requires maj or restructuring. Restructuring involves redeployment
of the assets and their use under a different management. This restructuring
decision will be referred to as a choice between proj ect closure and proj ect
t See Jarrell et al. (1988) for a survey of the empirical literature and Weston et al. (1990, appx. A),
for a textbook discussion of major theoretical contributions.
2 Empirical studies have not found much evidence for the claim that such transfers of wealth can
explain takeover premia (e.g. Lichtenberg and Siegel, 1990; Rosett, 1990).
3 Cf. Kini et al. (1995) and Shivdasani (1993) for empirical support and Hirshleifer and Thakor
(1991) for a discussion of the alternative 'complementarity hypothesis'.
! i
E. Maug / Journal of Corporate Finance 3 (1997) 113 139 117
continuation. The returns from restructuring are denoted by L~ (then Ys = L~); if
the proj ect is continued it yields x s (then Ys = Xs). All values are expressed in
prices of stage 1.
The firm is run by managers who make human capital investments. The human
capital investments can be either general or specific to the firm. The investment in
human capital is not modelled explicitly here. This assumes that the total amount
of investment in human capital (general plus specific) is given, so that only the
choice of the specificity of human capital needs to be considered. This specificity
of human capital is represented by one variable h, where h lies in the closed unit
interval and can be thought of as an index. The following assumption describes the
technology of the proj ect: the specificity of human capital h increases expected
profits, and profits in each state are affected by the closure decision. The last
aspect is detailed further in Assumption 2.
As s u mpt i o n l ( Pr o d u c t i o n ) .
(i) Ys depends on the continuation decision:
{LII if continued,
Y~ = if closed.
(ii) The specificity of managers' human capital determines the likelihood of
more profitable states:
YH with probability p, ( h) ,
3(, = YM with probability I - - P H( h ) - - PB, ,
YB with probability PB
where p, ( h ) is increasing and concave, h ~ [0; 1] and p~/(0) = ~c, p~(1) = 0.
(iii) x H > x g > x~.
Parts (i) and (ii) of Assumption 1 summarize the previous discussion and
impose some regularity conditions. Note that the probability of the low state B is
independent of h. This assumption facilitates proofs and is for convenience only.
The defining characteristic of states is the ranking of continuation profit levels x s
in (iii). 4 Managers are risk neutral and maximize expected utility. They negotiate
contracts with the original owners of the firm. These contracts specify payments
/z(y.~) contingent on profits. The managers do not care about the specificity of
their human capital investment h as long as the proj ect is continued, and the effort
for investing in some kind of human capital does not need to be considered here.
However, investing in a high level of firm-specific human capital implies that their
4 There is no restriction on the ranki ng of L B, L M and L H ; cf Assumpt i on 2 below.
118 E. Maug / Journal t)f Corporate Finance 3 (1997) 113-139
val ue in the outside l abour-market has decreased. 5 If the proj ect is discontinued,
they suffer a utility loss D(h), which is continuous, increasing and convex in h.
One interpretation of D( h ) is that managers have l ower career-prospect s if the
human capital they have acquired is specific to the firm rather than of general
appl i cabi l i t y. Al t ernat i vel y, the utility loss may be vi ewed as a search cost that
managers i ncur i f they have to l ook for new empl oyment . 6 Yet another interpreta-
tion is that managers' aversion to reallocating the assets is rel at ed to the commi t -
ment and skill they devot e to the pr oj ect ' s devel opment . Manager s' obj ect i ve can
be rewritten as
U M = E( t ~( y~) ) - P ( l ) D( h ) ( 1)
where E(.) denotes the expect at i ons operat or and P( I ) the probabi l i t y of cl osi ng
the proj ect under pol i cy 1. l represents the set of states in which the proj ect is
cl osed (e.g. 1 = {B} indicates that it is cl osed if and only if s = B; then P(1) = Pn;
si mi l arl y, i f 1 = {B, M}, then P ( 1 ) = 1 - P H) .
The fol l owi ng assumpt i on further defines states and is made in order to
introduce the possi bi l i t y of inefficient deci si ons about the use of the assets.
As s umpt i on 2 (St at es).
( i ) L H- - x H<O
( i i ) D(O) > L M- x M > O
(i i i ) 0< D( I ) < L~- xn
( i v) ( pn( l ) - - P u( 0) ) ( xn- - xM) <ps ( L , - x , ) .
(st at e H) ,
(st at e M) ,
(st at e B) ,
In state H, nobody woul d consi der cl osi ng a proj ect. In state M, it is pri vat el y
but not soci al l y opt i mal to restructure since the gains do not exceed the pecuni ary
equi val ent of the ut i l i t y-l oss they inflict on managers. In state B, it is pri vat el y and
soci al l y opt i mal to redepl oy the assets. Onl y these three situations matter since
they descri be the conflict of interest bet ween owners and managers. Outsiders who
make deci si ons cont i ngent on pecuni ary returns only will restructure whenever
L - x > 0 even if this di fference is rel at i vel y small.
Assume that h is observabl e but not contractible. Then managers cannot insure
t hemsel ves against proj ect closure by writing compensat i on payment s into their
contracts cont i ngent on h. The onl y way in which they can protect t hemsel ves is
by investing in general human capital in order to l i mi t their exposure to restructur-
i ng-deci si ons taken by outsiders (i.e. keepi ng their opt i ons open on the manageri al
5 For academics, research is usually a non-institution specific activity, whereas teaching skills are
more institution specific.
6 I am grateful to Robert Heinkel for this interpretation.
E. Maug / Journal of Corporate Finance 3 (1997) 113 139 119
2. 3. 4. 5.
I I I I , tPoe
Owners determine Manager chooses State of nature s Party in control Payofts).,, realized;
governance structure: specificity of revealed decides about parties paid according to
capilal struclure human capital h project closure contracts; managers
managerial contract paid ,u(y, )
decision making rights
Fig. 1.
l abour market ). Thi s will, however, make the proj ect less successful . Moreover,
condi t i on (i v) ensures that the rest ruct uri ng gai n is suffi ci ent l y i mpor t ant rel at i ve
to pot ent i al ef f i ci ency- gai ns from human capi t al i nvest ment s to make cl osure in
state B al ways attractive. The t i mi ng of the model is di spl ayed i n Fig. 1.
The next proposi t i on deri ves the fi rst -best al l ocat i on ( l *, h* ) as a benchmar k
for furt her results. It is defi ned from the maxi mand:
max 1I = ~ p, ( h) y., - P ( l ) D( h) . ( 2)
{ I , h} s ~ { H, M, B }
Recal l that P(1) denot es the probabi l i t y of r edepl oyi ng the assets under pol i cy
I. The first t erm gi ves profits as a funct i on of h and fol l ows i mmedi at el y from
Assumpt i on 1. The second t erm is the expect ed di sut i l i t y manager s suffer (cf. Eq.
(1)). Thi s obj ect i ve can be i nt erpret ed as the pr of i t - maxi mi zat i on- pr obl em of an
owner under the assumpt i on that h and / are cont ract i bl e at stage 1. The f ol l owi ng
proposi t i on charact eri zes the first-best, where h* denot es the fi rst -best i nvest ment
in proj ect -speci fi c human capital: 7
Proposi t i on 1 (Fi rst -best ). The solution o f pr ogr am (2) whi ch charact eri zes the
f i rst -best is to liquidate a project i f and onl y i f it is in state B. The f i rst -best val ue
h* E (0, 1) is i nt eri or and uni que and soh, es:
p~ ( h" )( x H - x M) - pa D' ( h *) = 0. ( 3)
I assume t hroughout the paper that weal t h-const rai nt s are suffi ci ent l y severe
such that out si de f i nance is requi red and cont rol has to be del egat ed. I also assume
that h is not cont ract i bl e. Thi s i nt roduces a di fference bet ween the ex ant e
obj ect i ve (2), and the ex pos t perspect i ve where out si ders have an i ncent i ve to
di sregard manager s' i nvest ment in human capital.
7 All proofs are available in the long version of this paper and are not included here to economize on
space. The lull version of the paper can be downloaded from the web page ' http://www.duke.edu/ ~
maug/research/boarddir/'. You can also obtain a hardcopy from the author upon request. Please see
title page for mailing address.
120 E. Maug / Journal (~f O~rporate Fi nance 3 (1997) 113 139
4. Analysi s of alternati ve organi zati onal forms
The question of which organizational form is optimal is analyzed in the
remainder of this paper. The basic distinction here is between managers and
outsiders. Managers are in control and by virtue of their involvement in the
operations of the firm have sufficient information to identify any restructuring
potential to increase shareholder value. Alteruatively, outside parties can be given
control rights, but any outsider other than the incumbent management needs to
acquire additional costly information in order to realize the restructuring potential
of the firm:
As s umpt i on 3 (Res t r uct ur i ng b! [brmat i on). An outside party can realize the
restructuring potential of the firm only if it has invested in additional information.
This information is obtained with probability T. information acquisition is costly
with information acquisition costs C(~-). The cost function satisfies C' (~-)> 0,
C"(-r) > 0 and C(0) = 0. Managers observe this information without incurring any
additional costs.
Note that state s is still assumed to be common knowledge. However, in order
to realize the difference L, - x,., more detailed operating knowledge of the firm is
required. Whereas managers, who operate the firm on a day to day basis are
assumed to possess this knowledge, any outside party who wants to improve the
operations of the firm has to acquire it. Each organizational structure must
determine two elements: managerial compensation and the implied closure policy,
which depends also on the incentives of outsiders to acquire information. Organi-
zational forms are characterized as follows:
Managerial control
Directors
Debt
Takeovers
Managers' compensation contract is fixed by the initial orga-
nizational designer and only managers have control over
closure. (Section 4.1)
Managers decide on closure policy; directors have control
over managerial remuneration and can use this to affect
closure-policy. (Section 4.2)
If the firm is solvent, only managers have control. In
bankruptcy states only debt holders have control over the
assets. Debt holders do not get involved in determining man-
agerial pay. (Section 4.3)
A takeover raider can acquire control of the assets of the firm
and can offer managers severance payments. (Section 4.4)
Each of these organizational forms is analyzed as an alteration of stage 4 of the
extensive form of the game (cf. Fig. 1).
, i , ~ I i , , i T " l' rl I ,
E. Ma u g/ J o u r n a l (?['Corporate Fi nance 3 (1997) 113 139 121
4.1. Expl i ci t cont ract ual prot ect i on: Pur e manageri al cont rol
Pure manageri al control is the si mpl est organi zat i onal form and resembl es a
classical pr i nci pal - agent situation, Managers are fully prot ect ed by an expl i ci t
contract since they exerci se all control rights over the use of the assets. Then the
onl y way to obtain the gains from restructuring is a compensat i on-scheme that
gives managers an i ncent i ve to cl ose the proj ect in the low state. Manager s'
compensat i on can be any profit cont i ngent contract /z(ys). Managers have zero
initial wealth so that the contract cannot i nvol ve negat i ve payment s in any state,
i.e. /~(y~) > = 0 for all y~,. Since managers have no wealth they will not be made
residual cl ai mant s; since they are risk neutral this must be ruled out, otherwise the
solution of the probl em woul d be trivial. Denot e by h E the degree of speci fi ci t y of
human capital for the case of expl i ci t contractual covenants, and by r the rent
managers extract, defi ned as the di fference bet ween their expect ed utility and their
reservat i on utility, assumed to be zero.
Proposi t i on 2 ( Pu r e manager i al cont rol ). (i ) The opt i mal manageri al cont ract
under pur e manager i al cont rol has:
{ ~
( hE) i f y.~ = LB.
/ ~(y~) = (XH) > 0 i f y , =X u . ( 4)
i f y.~ 4= L B , x, ,
wher e 0 < I ~(X#) <x ~ - x g. i ~(x H) and hF, sat i s~:
~ ( x u ) p' ( hE) = pe D' ( hF) ( 5)
f r om ma n a ge r s ' f i r s t -o r d e r conditions. This i nduces rest ruct uri ng i[ and onl y (f
s = B and under i nves t ment in project -speci f i c human capital: h E < h*.
(i i ) Manager s obt ai n a strictly posi t i ve rent:
r = I ~ ( x H) P H( h E ) . ( 6)
Hence, the opt i mal contract compri ses (1) a severance payment in case of
liquidation (a ' gol den parachut e' ), and (2) a bonus payment in the high state H.
The latter coul d be i mpl ement ed through a stock opt i on scheme with exerci se
price X equal to Max( L B, x M) and granting t ~ ( XH) / ( X H - - X ) options. /~(Z~) is
posi t i ve if it induces the manager either to cl ose the proj ect or to invest in more
specific human capital. It fol l ows from Assumpt i on 2 that l i qui dat i on is opt i mal
whenever s = B, hence the manager is pai d for liquidating in this state. An
i ncrement al increase in h i ncreases profits by p' u ( h ) ( x H- x M) from Assumpt i on
1, so that profit sharing in state s = H benefits initial owners; nevertheless,
managers are not made residual cl ai mant s, ot herwi se owners woul d surrender all
benefits to them, so that /,z(x n ) < x H - x g and h E < h*. Since managers have no
122 E. Maug / Journal (~[" Corporate Finance 3 (1997) 113-139
resources, they cannot pay for the val ue of this share and obt ai n a rent. In vi ew of
Proposi t i on 2, profits / / F in the case of pure manager i al cont rol are:
//F~ = y' p~( he ) x., + p~( L,~ - xB) - r. ( 7)
. s'
Hence, in setting / x(x H) owners trade off the benefi t from hi gher profits of the
fi rm agai nst the rents t hey have to pay manager s in order to i mpl ement a cert ai n
val ue for h. Si nce these rents i ncrease in h, a subopt i mal val ue h e < h* is chosen.
The weakness of prot ect i ng manager s by expl i ci t cont ract s is that manager s do not
si gn an empl oyment cont ract cont i ngent on h and have to be mot i vat ed by a bonus
payment ~( x, ) in the hi gh state. The next sect i on i nt roduces an ownershi p-st ruc-
ture whi ch pot ent i al l y over comes this probl em.
4. 2. I n d e pe n d e n t d i r e c t o r s
Thi s ownershi p-st ruct ure i nt roduces anot her agent whi ch I call a supervi sory
board of di rect ors or si mpl y i ndependent directors. For the purpose of this
ar gument it is i mmat eri al whet her i ndependent directors are a part of a uni t ary
board, or form a separate supervi sory board in a t wo-t i er syst em, s Unl i ke
managers, directors do not i nvest ill fi rm-speci fi c human capital, but t hey have to
be mot i vat ed to acqui re i nf or mat i on about al t ernat i ve operat i ng pol i ci es accordi ng
to Assumpt i on 3. Di rect ors maxi mi ze expect ed utility. Thei r compensat i on- schemes
must be l i near payment s whi ch may be cont i ngent on the net profits of the fi rm net
of manager s' compensat i on. They nmst be of the form
7( Y, - #(Y. ~)) ( 8)
Li ke managers, directors have no initial weal t h, hence they cannot pay for t hei r
shares out of their own wealth. 9 Di rect ors' mot i vat i on can be underst ood i n three
ways. First, t hey may be obl i ged to hol d "y shares of the fi rm whi ch ent i t l es t hem
to a fract i on 7 in the di vi dend y.~ - / , ( y, ). m Second, t hey may be under a
fi duci ary dut y to sharehol ders to act on t hei r behalf. In this cont ext such a dut y
woul d obl i ge t hem to maxi mi ze total payout s. Last, t hey may want to acqui re a
good reput at i on as i ndependent moni t ors in order to gai n more di rect orshi ps in the
future. ~t The model l i ng approach here relies excl usi vel y on the mot i vat i on of
directors t hrough sharehol di ngs or i ncent i ve cont ract s conf or mi ng to (8). Wher eas
s One might argue that separate supervisory boards are more likely to be independent. See Monks
and Minow (1995) for a textbook treatment of corporate governance institutions.
9 Situations where directors are or represent large shareholders would have to be covered by a
different analysis. The discussion here refers to the typical situation of externally appointed directors
whose wealth is small relative to the size of the firm.
10 Cf. Linn and Martin (1994) for an empirical documentation of the stock ownership of directors.
ii Cf. Fama (1980) for a detailed exposition of this hypothesis and Kaplan and Reishus (1990) and
Gilson (1990) for supporting empirical evidence.
E. Mtmg / Journal of Corporate Finance 3 (1997) 113-139 123
all three approaches imply that directors' incentives will be aligned with those of
shareholders, only the granting of free shares to directors implies that this comes at
a direct cost to initial shareholders, motivating the double agency problem
analyzed here. The assumption on linearity is a simplification. It will be shown
below (Section 4.3) that granting directors options or levering the firm would
improve the allocation. However, linearity does not affect any of the qualitative
insights discussed in this section.
Under this ownership-structure, managers have the right to decide about the use
of the assets. Directors have the right to review managers' salaries. For clarity,
denote the original managerial contract agreed at stage 1 by i~o(Z,) and the
compensation offered by directors by /2. Directors can influence closure decisions
by offering managers additional compensation for divesting. Hence, managers
negotiate with directors over the closure decision and over adequate compensation.
4.2.1. Ful l cont rol by i ndependent di rect ors
The maintained assumption of this subsection (to be relaxed below) is that
directors have all the bargaining power when they negotiate managers' compensa-
tion. Whenever directors want to redeploy assets they offer managers a compensa-
tion /2 in return for their consent to this decision. The extensive form of stage 4
(see Fig. 1 above) is then the following stylized bargaining game:
(a) Directors offer managers a compensation /2 subject to their consent to asset
reallocations.
(b) Managers accept or reject this offer.
(c) If they reject, the original managerial contract is implemented, managers
receive ~0(Y~) and no asset can be sold. If they accept, they receive /2 and
directors decide whether to liquidate or not.
The firm is fully equity financed, but equity holders have no decision making
rights other than the initial appointment of directors. Proposition 3 describes the
allocation.
Proposi t i on 3 (Di rect ors). Assume cont rol is del egat ed to managers and direc-
tors. Di rect ors can identiD" rest ruct uri ng possi bi l i t i es at a cost as def i ned in
Assumpt i on 3. They retain all bargai ni ng powe r in negot i at i ons with management ,
but managers have a veto ri ght over the use o f the assets. Then:
(i ) Di rect ors hol d a f r act i on t~ D o f the equi t y and identify the rest ruct uri ng
pot ent i al o f the f i r m wi t h probabi l i t y "r D. I f di rect ors are informed, managers
recei ve a t ransf er /2 = D( h* ) in state s -- B and zero t ransf ers in the ot her t wo
states. There are no prof i t -cont i ngent payment s, i.e. /~0(Y~) = 0 .for all Z,. The
opt i mal val ues "r D and YD are uni que and rel at ed as
YD = ps ( L 8 _ x8 _ D( hD) ) " (9)
124 E. Maug /Journal ~[ Colporate Finance 3 (1997) 113-139
The specifici~, o f human capi t al ht~ and the set~erance pay me nt to manager s
D(hl>) are decreasi ng in ~-L~, and "rt> is i ncreasi ng in YD.
(i i ) I f C' ( I ) and C"(1) are smal l and PB l arge enough, this mechani s m
i mpl ement s t he f i rst -best al l ocat i on with ho = h *. I f C'(~'t~) is suf f i ci ent l y l arge
and pe sufficiently small, t hen this mechani s m l eads to a l ower ex ant e payoff" to
initial sharehol ders than pur e manager i al control.
St andard bargai ni ng theory i mpl i es that directors will offer managers al ways
exact l y D( h ) if they propose to close the proj ect, so that the net economi c gain
from closure, L R - x B - D( h) accrues entirely to shareholders, whereas managers'
utility remai ns unchanged through bargai ni ng and is unaffected by closure.
Therefore, directors choose the liquidation pol i cy which maxi mi zes the net gain,
and managers' severance payment al ways exact l y offsets their utility-loss if the
proj ect is closed, thereby gi vi ng them full insurance against expropri at i on. As a
result, the l i kel i hood of the closure deci si on has no i mpact on their deci si on to
invest in specific human capital, and they can invest the equi l i bri um amount ht~
without a positive bonus payment . One way of i mpl ement i ng such a negot i at i on
game woul d have one i ndependent di rect or less than the number of votes required
to deci de on asset sales. Observe that directors act as agents of shareholders in that
they maxi mi ze ex ant e sharehol der wealth. Ex post , shareholders woul d prefer to
renege on the contract and close in state s = M as well.
However, the benefits of restructuring come at a cost to shareholders, since
they have to mot i vat e directors to become informed. From Eq. (9), the key
vari abl e is the margi nal cost of information, since this det ermi nes the degree to
whi ch directors become i nformed for a given fraction of the equity they hold. If
the margi nal cost for becomi ng i nformed is rel at i vel y modest, then y~ and the
costs to shareholders, which are YD ~, P.~ Y.~, are low enough, so that directors will
be given incentives to become perfect l y i nformed (T = 1). This requires also that
the cost function is not too convex ( C" ( h) is small), since the ex ante t radeoff is
bet ween an incremental increase in ~- and the benefits of supervision, and an
incremental increase in costs, (i.e. an increase in y and therefore C'(~-)). Con-
versely, if the costs of provi di ng directors with incentives are too large, then the
best solution which can be achi eved with this mechani sm is still domi nat ed by
pure manageri al control. Whereas shareholders pay the costs of control through
directors in all states of the world, they obtain the benefits only in the bad state
s = B. Hence, directors become more advant ageous i f PH is larger. Al so, if
directors have stronger incentives, then (1) the i nci dence of restructuring increases
because ~- increases, and therefore (2) the i nvest ment in specific human capital by
the manager is reduced. ~-~
12 This result is related to the entrenchment argument by Shleifer and Vishny (1989). However, here
managers respond to the possible breach of implicit contracts. There is no sense in which they actively
entrench themselves by making themselves indispensable.
. . . . . . T . . . . . . . . T . . . . . . . .
E. Maug / Journal o[' Corporate Finance 3 (1997) 113-139 125
The first-best result is somewhat special but instructive, since it reveals the
essential elements of the director-mechanism very clearly. In order to understand
better the distinctive features of the director-mechanism and the importance of
negotiation, compare directors with any randomly chosen large shareholder who
owns a share o- of the firm and has the right to trigger liquidation of the proj ect
but is not allowed to change managers' remuneration. Managers are protected by a
severance package which gives them a golden parachute G if they are dismissed.
This setup is similar to the director-mechanism except that the negotiation element
is removed. This leads to the following:
Proposi t i on 4 ( R a n d o m sharehol der). As s ume a randoml y chosen sharehol der
owns a share o" o f the f i r m and has t he ri ght to initiate l i qui dat i on and manager s
hat,e a set~erance package whi ch pays t hem a gol den par achut e G ~f t hey are
di smi ssed. Then the al l ocat i on can be impro~,ed f o r all possi bl e par amet er s by
al so gilding this sharehol der the right to renegot i at e manageri al compensat i on.
Note that this result does not assume that the golden parachute G and the stake
of the random shareholder cr are in any sense optimally chosen. This result
illuminates why directors' discretion over managers' severance package, and
hence the negotiation aspect of the director-mechanism is crucial: it makes
severance pay contingent on h, so that the allocation is dependent on the
non-contractable variable h. A managerial contract which automatically grants an
ex ante fixed severance payment in case of dismissal cannot replace this since it
replaces a payment which is cont i ngent on h by one which is not. This phe-
nomenon is familiar from the literature on the hold-up problem. 13
Note that several assumptions about directors are important for the result to
obtain which define a director as an institution which combines elements of
managers and elements of ordinary shareholders. They are similar to shareholders
in that they aim to increase pecuniary payoffs to shareholders only. 14 They are
different from shareholders because they are conferred additional property rights
(determine operating decisions, managerial compensation) and they bear additional
costs to make informed decisions. Directors are different from executive manage-
ment because they have no vested interests in particular operating decisions.
Specifically, they have no specific human capital at stake in case of proj ect
closures. This may be regarded as the def i ni t i on of an independent director.
Directors whose human capital is so specialized that it gives them a stake in a
13 Cf. Hermalin and Katz (1991) for a proof of a general theorem for a principal agent contract.
Aghion et al. (1994) and Hart and Moore (1995) discuss alternative renegotiation settings and the
implications for the underinvestment problem. Other contributions include Huberman and Kahn (1988),
Chung (1991) and MacLeod and Malcomson (1993).
14 Byrd and Hickman (1991) and Linn and Martin (1994) find that shareholdings of independent
directors increase the performance of corporate boards.
126 E. Maug / Journal ~/' Corporate Finance 3 (1997) 113-139
particular decision are not independent and should be classed as executive
management. Directors who are representatives of affiliated companies are also
not independent in the sense required (see Section 4.2.2 below). Lastly, directors
are assumed to have all the negotiating power vis a vis managers. This is a strong
assumption and its relaxation is the subj ect of the next section.
4.2.2. Managerial negotiating power
The positive result about implementation of the first-best with the director-
mechanism is obviously extreme as it depends on all bargaining-power being
given to directors. However, as directors need managers' consent in order to force
restructuring and because they have to renegotiate their salary, managers may be
to some degree successful in pressing for a higher compensation than is needed to
make them j ust indifferent between continuation and closure of the project. This
section therefore analyzes the same ownership-structure as the previous section
under alternative assumptions on bargaining power. Hence, the extensive form of
stage 4 of the game is now:
(a) Directors and managers bargain over the decision to close the firm and over
managers' salary.
(b) If they reach an agreement, managers receive a transfer/2 and implement the
agreed restructuring decision. If they do not agree managers decide and
receive the payment from their original contract /~0.
15
The bargaining outcome is determined from:
Assumption 3 (Bargaining). If managers and directors enter a bargaining game
at stage 4, then the outcome is determined according to a generalized Nash-
bargaining solution which gives directors bargaining-power 1 - ce and managers
bargaining-power c~.
It is almost impossible to prescribe in any kind of explicit contract the
bargaining-positions of insiders who work in close relationships, let alone formal-
ize such a bargaining-process. The approach advocated here is to take ' bargaining
power' as exogenously given rather than to make it part of the formal model.
Researchers of corporate governance have analyzed determinants of the relative
positions of directors versus executives. Lorsch and McIver (1989) identify in
particular (1) the nomination process of directors (selected by the CEO or an
independent committee), (2) their legal accountability, and (3) their access to
15 For a t ext book t reat ment of the gener al i zed Nash bar gai ni ng- sol ut i on see Bi nmor e (1992, ch. 5).
The cooper at i ve sol ut i on chosen here is di rect l y rel at ed to non- cooper at i ve bar gai ni ng sol ut i ons and the
sol ut i on bel ow has an i nt erpret at i on in t ernl s of al t ernat i ng offer games; cf. Osbor ne and Rubi nst ei n
(1990, ch. 4) for a di scussi on.
.... T ~' "~" - - ~ ~'~r "-' t -*
E. Maug / Journal ~)~ Corporate Finance 3 (1997) 113 139 127
i ndependent i nformat i onal sources. The last argument woul d i mpl y that costs C(~-)
and c~ are posi t i vel y correlated. Baysi nger and Butler (1985) emphasi ze the
di fference bet ween affiliated outside directors who represent organi sat i ons with
financial or busi ness rel at i onshi ps with the firm and i ndependent directors. They
find that firms with a stronger i ndependent el ement perform on average better. ~6
Di rect ors recei ve in case of closure T( L~- / 2) and in case of no closure
y ( x , - t xo(x, )). Si mi l arl y, managers get /x0(X~.) in case of cont i nuat i on and
/ 2 - D( h ) in case of closure. Hence, the maxi mand for the general i zed Nash
Bargai ni ng game is
q~(h, / 2, a ) {T(L,. x~ / 2+ " , x, (, ,~, = - - tZol, X , ) ) j * { / 2- / . %( x ) - O( h) } "
(lO)
Then:
/2( o~, h) - argmax ~p( h, / 2, c~). (1 1)
#
The result is expressed in the fol l owi ng proposi t i on.
Proposi t i on 5 (Bar gai ni ng Power ). (i ) Suppose Assumpt i on 3 holds and man-
agers have bargai ni ng powe r c~. Then the f i r m wi l l be cl osed i f and onl y (f
di rect ors have acqui red i nf ormat i on and the f i r m is in state s = B. Di rect ors
awar d managers a t r ans f er / 2 cont i ngent on c and h gi ven by
/2( ce, h) = , , , ( xs ) + D( h) c~( L~ - x R - D( h ) ) (12)
and /2 = tXo in all ot her states.
(i i ) I f c~ > O, the di rect or-mechani sm does not i mpl ement the f i rst -best allo-
cation anymore f o r any f or m o f C(T). The payoff" to ori gi nal owners is monot oni -
cal l y decreasi ng in c~ and there exists a cri t i cal val ue f o r e~, ceMc , wi t h
0 < c~Mc < 1 such t hat the di rect or-mechani sm domi nat es pure manageri al con-
trol i f 0 < c~ <_ C~Mo and vice versa i f ce Mc < c~ < 1.
It may be noted from Eq. (12) that the mi ni mum payment which managers
recei ve if they have no bargai ni ng-power (c~ = 0) is D(h ) as in Proposi t i on 3.
However, if they have bargai ni ng-power, they recei ve not only a compensat i on for
their disutility but also a fraction a of the surplus. If ce = 1, they obtain the whol e
surplus, a st andard result for Nash-bargai ni ng-games. Hence, Proposi t i on 5 i mpl i es
that directors with bargai ni ng power less than 1 - O~Mc are worse for the share-
hol ders than del egat i on to managers ' no strings at t ached' .
16 See also Byrd and Hickman (1991), Rosenstein and Wyatt (1990), Lee et al. (1992), Weisbach
(1988) for similar results. Agrawal and Knoeber (1996) find that outside directors have a negative
effect, but do not distinguish between affiliated outside and independent directors.
128 E. Maug / Journal qf Cotporate fTnance 3 (1997) 113 139
The intuition behind the second part of the result is that directors who do not
have sufficient bargaining-power against management do not only become less
useful, but positively harmful since they worsen managers' incentives to invest in
proj ect-specific human capital. This is somewhat surprising because managers'
compensation in restructuring-states i ncr eas es with their bargaining-power. 17
Inspection of Eq. (1 1) reveals why this leads to an inferior allocation. First-order
conditions for choosing h in the director-mechanism are:
~ ( x H) ~ , ; ( h) = ~ r pB D' ( h ) . ( 1 3)
If managers have more bargaining power, their compensation depends more on
their negotiating skills than on their performance, i.e. their investment in specific
human capital. This removes the advantage of the director-mechanism of making
their compensation contingent on their investment. Then the director-mechanism is
similar to pure managerial control in terms of incentives but total transfers are
larger since managers receive larger rents. It is this double hazard which directors
with low bargaining-power introduce: they award excessive salaries to managers,
causing a direct loss to security-holders, and they weaken the link between pay
and performance, thereby causing an indirect loss in the form of productive
inefficiencies. The two taken together lead to the uniform verdict expressed in
Proposition 5(ii).
4.3. Cont rol and capi t al st ruct ure: Debt and cr edi t or cont rol
The previous section has shown that independent directors have the potential to
implement the first best if their negotiating position is very strong, otherwise an
independent supervisory board may be a mixed blessing. Under such circum-
stances, managerial discretion may be checked by creditors or raiders. The
circumstances under which they gain control of the firm are determined by the
capital structure and the corporate charter. These mechanisms are discussed in this
section in order to investigate altertmtives to the internal control structures
analyzed above. The first subsection shows how leverage can constrain managerial
discretion, whereas the second subsection focuses on the market for corporate
control as an external disciplining device.
Assume now that ownership rights can be transferred to security-holders by
using profit-contingent debt. i.e. a financial contract which specifies that control of
the firm is transferred to creditors if profits fall below a certain value (see below).
i f creditors have the right to liquidate, they will always liquidate at t = 1 if
L, - x,. > 0, i.e. they will only consider pecuniary returns. Suppose the face value
of debt F satisfies x B < F< xa4. Then in state B, the firm is bankrupt and
creditors gain control over the assets, whereas in state M the firm is still solvent.
17 See e.g. Berkovitch et al. (1994) where the first best is implemented only if the manager
(entrepreneur) has all the bargaining power.
. . . . . T . . . . T"
E. Maug / Journal of Corporate Finance 3 (1997) 113-139 129
What happens in case of creditor control depends critically on the information
creditors have in case of bankruptcy. Suppose that the debt is owned by a bank.
Suppose the bank expects to gain control of the assets if and only if s = B. Then
the equilibrium level of solves
C ' ( ~ ) =p~ ( L B - x 8 - D( h ) ) . (14)
Note that this is analogous to Eq. (9) above. Evidently, borrowing from a bank
will be costly in this model, since the bank will need to recover the information
costs in order to break even. The following proposition presents the main result on
creditor control; h o denotes the level of h chosen by the manager if owners have
chosen the optimal debt level.
Proposi t i on 6 (De bt ). (i ) Two debt -pol i ci es can be ex ant e opt i mal f o r own-
ers."
1. l = {B}: t he. f i rm wi l l be cl os ed in st at e s = B;
2. 1 = {M, B}: the f i r m wi l l be cl osed in st at es s = M and s = B.
(i i ) I f the cost s o f acqui ri ng i nf ormat i on C(~') are suf f i ci ent l y low, t hen this
sol ut i on gi ves i ni t i al owner s hi gher prof i t s than pur e manager i al cont rol and is
domi nat ed by the di r ect or -mechani s m as char act er i zed in Proposi t i on 3; the
sol ut i on i nvol ves under i nves t ment in pr oject -s peci f i c human capi t al : h D < h*.
(i i i ) I f the cost o f i nf ormat i on acqui si t i on are suf f i ci ent l y high, t hen debt is
domi nat ed by pur e manager i al control.
(i v) If" the pr obabi l i ~ o f rest ruct uri ng in the l ow st at e PB is suf f i ci ent l y low,
t hen debt domi nat es the di rect or-mechani sm.
Even though owners can reduce the rent managers can extract relative to pure
managerial control, investment in specific human capital remains inefficient,
Recall from Proposition 2 that under managerial control managers obtain a strictly
positive rent (given by Eq. (6)). Hence, whenever owners determine the optimal
debt-policy, they have to observe two constraints: first, managers' incentive
compatibility constraint which requires:
tX( Xn ) P'n( hD ) = P ( I ) D' ( hD) (15)
and second, their participation constraint which amounts to their rent r being
non-negative. From Assumption 2(ii), it can never be profitable to liquidate in
state s = M if this requires an additional payment D( h D) to the manager in that
state. However, suppose managers receive a strictly positive rent if 1 = {B}. Then
liquidating in state s = M as well gives owners pM( L M- - x M) minus the effi-
ciency loss from a higher probability of proj ect closure. This efficiency loss is
P' n(h)(x H - x M) Ah to a first-order approximation, where Ah is the reduction in h
because managers expect a higher probability of proj ect closure. No assumption on
the parameters of the model prevents the net gain from being positive, and
1 = { M, B} is an optimal policy if the net gain is positive and managers' participa-
130 E. Maug/ . l our nal r~/O)rporate Fi nance 3 (1997) 113-139
tion constraint is still satisfied. Note that this is a purely redistributional argument.
Allocative efficiency may be reduced by too frequent restructurings if h is reduced
further below its first-best value. Debt commits initial shareholders to an ex ant e
inefficient restructuring policy relative to the first-best solution because they trade
off their payoff against managers' rents.
The argument for debt in favor of pure managerial control in part (iii) of the
proposition relies on the rent managers obtain (see Proposition 2 above). Under
managerial control, managers need to be given a bonus payment if they restructure
the firm. With creditor control, this bonus payment is no longer necessary, since
managers' discretion is constrained by debt. Consider a managerial contract which
gives managers the same expect ed payment, but offers them a higher bonus
payment in the high state rather than a bonus for restructuring. If managers would
continue to choose the same level of investment in human capital, the allocation
and profits to shareholders would be unchanged. However, since their incentives
are improved they choose a higher investment in specific human capital, leading to
a superior allocation. The only disadvantage of debt is the additional cost of
information acquisition, hence this cannot be too large for the conclusion to hold.
The comparison of debt and the director-mechanism (part iv) hinges again on
information costs. Recall that the director-mechanism can implement the first-best
allocation if these costs are sufficiently low. Since debt lacks the negotiation
element over managerial compensation, creditor control leads to underinvestment
in human capital, hence directors clearly dominate debt for low supervision costs.
However, if the likelihood of restructuring is small, then the costs of the
director-mechanism are large even for modest costs of information acquisition,
since directors would receive dividends on their shares in the non-restructuring
states as well, whereas the bank receives a payoff only if it successfully restruc-
tures the firm. Hence, supervision through creditor control is preferable if restruc-
turing is less likely.
There is another way in which debt can compl ement rather than substitute the
director-mechanism since higher leverage concentrates the shareholdings of direc-
tors for any given dollar-investment directors have in the firm. This leads
immediately to
Proposi t i on 7 (Let , erage): [f debt is i nt roduced into the capi t al structure, and
all ot her contracts" are as in the di rect or-mechani sm des cr i bed in Proposi t i on 3,
t hen sharehol der t;alue is strictly i mproued i f the .[+~ce ualue c~f debt is strictly
positiL'e. Also, the opt i mal f a c e val ue o f debt is" F *= Min(xM, x R) and Yt)
i ncreases in F *
Increasing leverage reduces the costs of hiring directors, since they obtain the
same incentives (based on y L ~ ( L , - x ~ ) ) with a reduced expected payment to
directors for profits which do not depend on their activity ( y t ) ~ , p, x , . ) . Since
higher leverage makes higher shareholdings of directors cheaper tbr the initial
' " ' 1 ~" ~+ " l +~ "' ~ '
I
E. Maug /Journal of Corporate Finance 3 (1997) 113-139 131
owners, they will choose a higher y if they choose a higher leverage F, hence
director' s shareholdings and leverage should be positively correlated. It should be
observed that implementing this solution with debt is not the only possibility.
Directors could equally well receive options or warrants with sufficiently high
exercise prices to generate the same result.
Another literature on debt has emphasized the ability of lenders to renegotiate
fixed claims, thereby increasing the flexibility of debt contracts. ~s However, debt
renegotiation improves efficiency mostly through avoiding ex post inefficient
liquidation, whereas the model here implies liquidations which are ex ante, but not
ex post inefficient. Also, debt renegotiation typically serves to adjust financial
claims against the company to allow continuation as a going concern, and not the
adj ustment of executive compensation through creditors.
4.4. TakeoL~ers and the role o f the corporat e chart er
Assume the firm is fully equity financed and at least a certain part of the
outstanding shares have voting rights. The number of outstanding shares is
normalized to 1. Suppose after stage 3 a potential raider can take over the
company and bid for the voting rights. Moreover, assume the raider can acquire a
fraction ~b of the firm at the current market price P, where P is the ex dividend
value of the shares at stage 3 before a bid is announced. Evidently, bidders are
only interested in pecuniary payoffs, leaving managers vulnerable to interventions
by outsiders who do not take into account the loss they may suffer if their human
capital is depreciated. Consider therefore the inclusion of takeover defenses in the
corporate charter which owners decide at the first step of the game. Hence, stage 4
of the game described in Section 2 has the following extensive form:
(a) A potential raider decides whether to become informed about the firm and
take it over or not. If not, managers decide about the use of the assets.
(b) If the raider becomes informed and decides to take over he purchases ~b
shares on the stock market at the current market price P and announces the
conditions of the bid.
(c) Managers decide whether they use takeover defenses if such defenses are
available to them.
(d) If managers defend the bid successfully, they stay in control and decide about
the use of the assets. If they do not defend the bid, shareholders decide
whether to tender their shares,
(e) If the bid succeeds, the raider gains control and decides how to use the assets.
Otherwise managers stay in control and make decisions.
Only states where L~ > x~ are candidates for takeovers. The raider will close
the firm, redeploy the assets and obtain L~. Hence, the post announcement value
is Cf. Bolton and Scharfstein (1996) and Rajan (1992).
132 E. Ma u g/ J o u r n a l q['Corporate Fi nance 3 (1997) 113-139
of the share is L~. Sharehol ders will t ender their shares only i f the raider offers
L~, ot herwi se they woul d prefer to stay as mi nori t y shareholders in the firm and
the bid woul d fail. ~9 Denot e by q the probabi l i t y that a t akeover bi d is made if
the rai der is informed. This is also the mar ket ' s expect at i on of a bi d in equilib-
rium. Then, before the announcement of the bid the value of the company and the
market price of the share is P = q'rL, + (1 - q~')x~. The rai der' s profits are
R = 4) ( Ls - P ) = 05(1 - q r ) ( L , - x , ) . (16)
Hence, the raider will take over whenever R >_ 0 which requires L,. > x~. Note
that R > 0 i mpl i es that the raider al ways bids and q = 1. The rai der obtains the
profit R with probabi l i t y r. This gives the first result on hostile or undefended
takeovers:
Proposi t i on 8 (Hos t i l e t akeovers): I f t akeovers are undef ended, a bi d will
al ways be made whe ne ve r t he rai der has become i nf ormed and s = M or s = B.
The rai der becomes i nf ormed with probabi l i t y ~-R < 1 / 2 whi ch sol ves:
Ma x ~ &Te ( 1 - - Tn ) ( ps ( L s - - x B ) +pM( L M- - X M) ) - - C ( Te ) . ( 17)
Manager s recei ve onl y a pay me nt ~[" the st at e is hi gh ( x H - x M > t z(x n ) > 0).
One i mpl i cat i on of this result is that there may be too few t akeovers and the
restructuring potential L ~ - x B may not be real i zed if the stake the raider can
acquire in the market is too small. Fi rst -order condi t i ons are:
05(1 -- 2 Tn ) ( pB( L n -- x s ) + pM( LM -- XM ) ) = C' ( TR) (18)
and i mpl y that r e < 1/ 2. If 05 is small, r R is going to be small and the i nci dence
of t akeovers is low.
If the raider is sufficiently informed, t akeovers lead to excessi ve asset real l oca-
tion because this is opt i mal after stage 3 when the deci si on about human capital is
sunk and managers cannot bribe potential raiders. This situation provi des a
rationale for t akeover defenses which give managers some control over the
t akeover process. There are numerous strategies to defend bids. 20 Since t akeover
defenses are not the focus of this discussion, no anal ysi s of the mechani cs of
different devi ces is undertaken here. Assume for the remai nder of this section that
managers have access to some effective defense (e.g. a poi son pill), and that a
raider can only acquire control of the assets by obt ai ni ng managers' consent for
the transaction. Al so, suppose the corporat e charter coul d include a clause
speci fyi ng that any severance pay recei ved by the manager in the event of a
t akeover is pai d out of the company' s cash flows and that there is sufficient
19 This is fami l i ar from the anal ysi s of the free rider problem, cf. Grossman and Hart (1980).
~0 See West on et al. (1990, ch. 20) for a t ext book t reat ment of t akeover defenses, and Herzel and
Shepro (1990, ch. 1.8), for a di scussi on from a legal point of view.
E. Maug / Journal of Corporate Finance 3 (1997) 113 139 133
competition in the market for corporate control. This could be achieved with
exclusion devices like restricted offers of the kind discussed in Grossman and Hart
(1980). Alternatively, the bidder could offer the target shareholder shares in the
merged firm instead of cash (a 100% paper offer). This would also pass on the
bidding costs to the target shareholders.
Pr opos i t i on 9. As s ume a s ucces s f ul bi dder is per mi t t ed to gr ant manager s
s euer ance pay out o f t he t arget f i r m' s f unds and t here are at l east t wo pot ent i al
bi dders. Moreouer, i f t here are seL, eral bids, t he hi ghes t bi d mus t be accept ed.
Then"
( i ) The pa y o f f to t he rai der is
~p,.T(1 - ~-)(L,. - x , - D( h ) ) - C(-r ) (19)
and manager s receil~e tx( x ~ ) --- O. Then t he al l ocat i on i mpl ement ed wi t h f r i e ndl y
t akeover s has a l owe r i nves t ment o f i nf ormat i on by t he raider, a l ower i nci dence
o f t akeot , ers and a hi gher i nt ' est ment in human capi t al t han t he al l ocat i on wi t h
host i l e takeot;ers. The compar i s on o f s har ehol der ~'alue is ambi guous.
(i i ) I f p~ is suf f i ci ent l y smal l , t hen f r i e ndl y takeoL, ers domi nat e t he di rect or-
mechani s m.
The competition requirement is essential here. If there was only one potential
bidder, he could always offer managers a bribe b > D( h ) and shareholders a price
L B - b. Competition ensures that a bid of L 8 - D( h ) and a payment equal to the
disutility is the unique equilibrium.
Effectively, this implements a similar renegotiation-mechanism as the board of
directors, where raiders take the place of directors. They initiate the process and
have all bargaining power. The ' bargai ni ng' position of the raider is strengthened
by competition in the market for corporate control. Managers exercise veto-power
with the potential use of takeover-defenses. Since the raider takes the severance
pay out of the fi rm' s funds, he will bid only if the post takeover value of the firm
L, - D( h ) exceeds the value under the incumbent management xs, i.e. whenever
L~ - x~ >_ = D( h ) which implies that friendly takeovers can only happen in state
s = B. Hence, compared to hostile takeovers, the post-takeover value of the firm
and the likelihood of bidding are reduced, thereby reducing the incentives to
acquire information for the raider and the incidence of takeover bids even further.
Then it is optimal for the manager to invest in more firm-specific human capital,
so that friendly takeovers lead to larger investments in firm-specific human capital.
The implications for shareholder wealth are ambiguous, because two countervail-
ing effects operate here. Hostile takeovers lead to a larger incidence of restructur-
ing, but a lower investment in firm-specific human capital. Inefficient restructur-
ings in state s = M (relative to the first best) can be efficient in a second best for
similar reasons as in Proposition 6. Moreover, the fact that the firm is restructured
in many states implies stronger incentives to acquire information for the raider,
[ 34 E. Maug / Journal ~/" Corporate Fi nance 3 (1997) 113-139
which may help to overcome the probl ems of a low threshold 4). Sharehol ders
benefit more from friendly t akeovers then directors because raiders do not need a
compensat i on in non-t akeover states since they are not subj ect to wealth con-
straints. As a result, the benefits of fri endl y t akeovers do not depend on PB,
whereas those of directors do.
The result shows that the t akeover process can be understood as a mechani sm
whi ch renegot i at es the manageri al contract. However, unlike Scharfstein (1988)
who uses a si mi l ar notion, the key el ement of a t akeover is the acquisition of
control over the company' s assets, and it was shown (rather than assumed) how
this leads to a change in managers' compensat i on. This whole section confi rms an
intuition al ready stated by Shleifer and Summers (1988): if managers rely on an
implicit, l ong-t erm contract with the firm when they make far-reachi ng decisions,
the possi bi l i t y of hostile takeovers has a social cost. It was shown here that
fri endl y t akeovers provi de an alternative, even though under somewhat restrictive
assumpt i ons, and that the associ at ed and pot ent i al l y large separation payment can
serve an effi ci ency-enhanci ng purpose.
5. Di s c us s i on and c onc l us i on
This paper has i nvest i gat ed a model in which the assets of the firm have
alternative uses which are somet i mes more profi t abl e than the current one.
Managers of the firm invest in fi rm-speci fi c human capital which becomes
worthless in case the firm is restructured and proj ects closed. Human capital
deci si ons are not contractable, so that managers are given either all residual rights
of control, thus prot ect i ng them from outside interference, or they rely on i mpl i ci t
contracts with the firm whi ch are somet i mes breached. Outside interference
requires that an agent other than management acquires costly i nformat i on about
the firm. The analysis has shown how i ndependent directors can be underst ood as
an institution that regul at es the rel at i onshi p bet ween shareholders and managers in
a worl d where contracts are incomplete. They are different from execut i ve
management since they do not part i ci pat e di rect l y in the product i ve activities of
the firm, and accordi ngl y do not invest in fi rm-speci fi c human capital. However~
even though they act on sharehol ders' behalf, they are not like a large maj ori t y
sharehol der who exerci ses all residual rights of control, since directors have to
respect managers' prerogat i ve over the allocation of assets. This mechani sm works
well i f directors retain an i ndependent j udgement and can acquire information
about opt i mal operat i ng deci si ons at rel at i vel y low cost. The i mport ant insight is
that in wi del y held compani es control has to be del egat ed to two institutions or
agents: managers, who are in charge of the day to day running of the finn, and
directors who moni t or and revi ew contracts and use this right as an instrument to
influence business decisions.
E. Maug / Journal of Carporate Finance 3 (1997) 113-139 135
Strong Directors, (Friendly Takeovers)
>
(Hostile Takeovers), Debt
>
Managerial Control
>
Weak Directors
II HII II
Fig. 2.
The analysis has also provided partial rankings of the alternative governance
structures discussed above. If supervision costs are sufficiently low and the
possibility of restructuring sufficiently high, then the rankings can be summarized
as in Fig. 2. Here the labels ' Strong' and ' weak' refer to directors' bargaining
position vi sa vis executive management. Independent directors can secure gains
from restructuring and simultaneously sustain the commitment of the company to
implicit contracts. However, if directors are weak, managers' severance pay is
largely determined by their bargaining strength and they are able to extract
excessive compensation payments in restructuring. Then shareholders lose twofold:
directly, since they lose the extra compensation paid to managers, and indirectly,
since the link between managers' remuneration and their performance (here: their
investment in specific human capital) is reduced. Empirically, bargaining strength
of the board could be approximated by relating it to board composition. Many
researchers found that larger proportions of independent directors (as opposed to
affiliated outside directors) on corporate boards improve performance. 2w More
specifically, the analysis here predicts that the pay-performance relationship is
stronger for managers of companies who have more independent directors on the
board.
Institutions associated with the capital structure can provide second-best solu-
tions that improve on unconstrained managerial control of the corporation: explicit
contractual protection is no alternative to implicit contracts that can be breached.
Bargaining between a takeover raider and management about the takeover can also
implement a second-best solution if regulations on share purchases in the market
prior to a bid are not too stringent. However, these rankings depend on the cost of
acquiring information through outsiders not being too high. In case information is
proprietary to management and difficult to communicate, unchecked managerial
control is the dominant mechanism, since any outside control would be unjustifi-
el Cf. the literature referred to in note 15.
136 E. Maug / Journal ~![ Cot porat e Fi nance 3 (1997) 113- I 39
I I IIIIIIII III III
Costs
Low High
Low Friendly Take-
overs, Debt
Restructuring
Potential
Managerial Control
High Directors, plus debt Managerial Control
I
Fig. 3.
ably costly. Lastly, all considerations depend on the expected restructuring poten-
tial, particularly the probability of a state where the alternative uses of the assets
are significantly more profitable then the current use. Fig. 3 gives the optimal
ownership structure for different categories of information costs and restructuring
potential. Hence, the analysis predicts that independent directors are an optimal
solution only if two conditions are satisfied. Firstly, assessing managers' decisions
by obtaining independent information must be possible at a sufficiently low cost.
Therefore, one would not expect to see supervisory boards with majorities of
independent directors in companies where the key information for decisions is
proprietary to management and cannot be obtained or communi cat ed to outsiders.
Secondly, the expected restructuring potential must be large. This implies that the
director-mechanism is more likely to be chosen if the firm and its industry go
through a critical period where restructuring is likely then if the firm and its
environment are relatively stable. Hermalin and Weisbach (1988) have found that
firms tend to have more independent directors if (1) their performance deteriorates
or (2) if the firm leaves a product market. Both variables can be seen as proxies
for PB, the likelihood of profitable alternative uses of the assets. An extreme form
are LBOs, where sometimes LBO-specialists acquire large equity stakes and the
right to be presented on the board. In these transactions, the possibility of
profitable restructuring is a near-certainty..~2
Conversely, interventions by investors (creditors, raiders) are likely to be
optimal mechanisms if the likelihood of restructuring is small. This supports the
view that restructuring through takeovers or reorganization in bankruptcy are seen
as the exception rather than the rule. This prediction is consistent with the findings
_*2See e.g. the LBO of Saf eway' s in 1985, where KKR acted as a buyout speci al i st and acqui red
al most 90% of the equi t y and three seats on the board, The company went subsequent l y through a
period of si gni fi cant asset sales; of. Deni s (1994).
E. Ma u g/ J o u r n a l of Corporat e Fi nance 3 (1997) 113 139 137
of John et al. (1992) that most restructurings following performance declines are
vohmtary and initiated internally rather than being forced on management by
out~ider~.
Acknowledgements
This paper is adapted from chapter 3 of my PhD thesis at the LSE and I am
indebted to my supervisor, Margaret Bray, and to David Webb for numerous
discussions and to an anonymous referee and Kenneth Lehn (the editor) for advice.
I am also grateful to Daron Acemoglu, Patrick Bolton, Vittoria Cerasi, Leonardo
Felli, Julian Franks, Michel Habib, Oliver Hart, Robert Heinkel, Martin Hellwig,
John Moore, Kjell Nyborg and Ulf Schiller for helpful comments and suggestions.
All remaining errors are my own responsibility. I am indebted to the Institute of
Finance and Accounting at the London Business School and the Financial Markets
Group at the London School of Economics for financial support. Earlier versions
of this paper were presented (under a different title) at seminars at the University
of British Columbia, University of Frankfurt, London Business School, London
School of Economics, Stanford University, Wharton School, the 1993 symposium
at Gerzensee (Switzerland) and the 1992 congress of the Verein fOr Socialpolitik,
Oldenburg (Germany).
References
Aghion, Philippe and Bolton, Patrick, 1992, An 'incomplete contract" approach to financial contract-
ing, Review of Economic Studies 59, 473-494.
Aghion, Philippe, Mathias Dewatripont and Patrick Rey, 1994, Renegotiation design with unverifiable
information, Econometrica 62, 257-282.
Agrawal, Anup and Charles R. Knoeber, 1996, Firm performance and mechanisms to control agency
problems between managers and shareholders, Journal of Financial and Quantitative Analysis,
forthcoming.
Baysinger, Barry D. and Henry N. Butler, 1985, Corporate governance and the board of directors:
Performance effects of changes in board composition, Journal of Law, Economics, and Organiza-
tion.
Berkovitch, Elazar and Ronen Israel, 1996, The design of internal control and capital structure, Review
of Financial Studies 9, 209-240.
Berkovitch, Elazar, Ronen Israel and Jaime F. Zender, 1994, The design of bankruptcy law: A case for
management bias in bankruptcy organizations, Mimeo.
Berle, Adolf A. and Gardiner C. Means, 1932, The modern corporation and private property
(Transaction Publishers, New Brunswick, N J).
Binmore, Ken, 1992, Fun and games: A text on game theory (Heath, Lexington, MA).
Bolton, Patrick and David Scharfstein, 1996, Optimal debt structure and the number of creditors,
Journal of Political Economy 104, 1-25.
Byrd, John W. and Kent A. Hickman, 1991, Do outside directors monitor managers? Evidence from
tender offer bids, Journal of Financial Economics 32, 195-221.
138 E. Maug / Journal (~f Corporate Finance 3 (1997) 113-139
Chung, Tai-Yeong, 1991, Incomplete contracts, specific investments, and risk sharing, Review of
Economic Studies 58, 1031 - 1042.
Denis, David J., 1994, Organizational form and the consequences of highly leveraged transactions:
Kroger' s recapitalization and safeway' s LBO, Journal of Financial Economics 36, 193 224.
Fama, Eugene F., 1980, Agency problems and the theory of the firm, Journal of Political Economy 88,
288 307.
Fama, Eugene F. and Michael C. Jensen, 1983, Agency problems and residual claims, Journal of Law
and Economics 26, 327-349.
Fama, Eugene F. and Michael C. Jensen, 1983b, Separation of ownership and control, Journal of Law
and Economics 26, 301-325.
Gilson, Stuart C., 1990, Bankruptcy, boards, banks, and bondholders: Evidence on changes in corporate
ownership, Journal of Financial Economics 27, 355-387.
Grossman, Sanford J. and Oliver D. Hart, 1980, Takeover bids, the free rider problem, and the theory
of the corporation, Bell Journal of Economics 11,253 27/).
Hart, Oliver D., 1993, Theories of optimal capital structure: A managerial discretion perspective, In:
Margaret M. Blair, ed., The deal decade: What takeovers and leveraged buyouts mean for corporate
governance (Brookings Institution, Washington, DC).
Hart, Oliver D. and John Moore, 1995, Debt and seniority: An analysis of the role of hard claims in
constraining management, American Economic Review 85, 567-585.
Hermalin, Benjamin E. and Michael L. Katz, 1991, Moral hazard and verifiability: The effects of
renegotiation in agency, Econometrica, 59. 1735-1753.
Hermalin, Benjamin E. and Michael S. Wcisbach, 1988, The determinants of board composition and
direct incentives on firm performance, Rand Journal of Economics 19, 95 112.
Herzel, Leo and Richard W. Shepro, 1990, Bidders and targets: Mergers and acquisitions in the U.S.
(Basil Blackwell, Cambridge, MA and Oxford, UK).
Hirshleifer, David and Anjan V. Thakor~ 1991, Corporate control through board dismissals and
takeovers, Mimeo.
Hirshleifer, David and Anjan V. Thakor. 1994, Managerial performance, boards of directors and
takeover bidding, Journal of Corporate Finance 1, 63 90.
Huberman, Gut and Charles Kahn, 1988, Limited contract enforcement and strategic renegotiation,
American Economic Review 78, 471-484.
Jarrell, Gregg A., James A. Brickley and Jeffi'y M. Netter, 1988, The market for corporate control: The
empirical evidence since 1980, Journal of Economic Perspectives 2, 49-68.
Jensen, Michael C., 1986, Agency costs of free cash flow, corporate finance, and takeovers, American
Economic Review 76, 323 329.
Jensen, Michael C. and William H. Meckling, 1976, Theory of the firm: Managerial behavior, agency
costs and ownership structure, Journal of Financial Economics 3, 305-360.
John, Kose, Larry H.P. Lang and Jeffrey Netter, 1992, The voluntary restructuring of large firms in
response to performance decline, Journal of Finance 47, 891-917.
Kaplan, Steven N. and David Reishus, 19911, Outside directorships and corporate performance, Journal
of Financial Economics 27, 389 410.
Kini, Omesh, William Kracaw and Shehzad Mian, 1995, Corporate takeovers, firm performance, and
board composition, Journal of Corporatc Finance 1, 383-412.
Lee, Chun I., Nanda Rosenstein, Nanda Rangan and Wallace N. Davidson, III, 1992, Board composi-
tion and shareholder wealth: The case of management buyouts, Financial Management 21, 58 72.
Lichtenberg, Frank R. and Donald Siegel, 1990, The effects of leveraged buyouts on productivity and
related aspects of firm behavior, Journal of Financial Economics 27, 165-194.
Linn, Scott C. and Kenneth J. Martin, 1994, The effects of outside director stock incentive compensa-
tion plans on stockholder wealth, Mimeo (University of Oklahoma).
Lorsch, Jay W. and Elisabeth Mclver, 1989, Pawns or potentates: The reality of America' s corporate
boards (Harvard Business School Press, Boston, MA).
E. Maug / Journal of Corporate Finance 3 (1997) 113-139 139
MacLeod, W. Bentley and James M. Malcomson, 1993, Investments, holdup, and the form of market
contracts, American Economic Review 83, 811-837.
Monks, Robert A.G. and Nell Minow, 1995, Corporate governance (Basil Blackwell, Oxford).
Osborne, Martin J. and Ariel Rubinstein, 1990, Bargaining and markets (Academic Press, New York).
Rajah, Raghuram G., 1992, Insiders and outsiders: The choice between informed and arm' s-length
debt, Journal of Finance 47, 1367-1400.
Rosenstein, Stuart and Jeffrey G. Wyatt, 1990, Outside directors, board independence, and shareholder
wealth, Journal of Financial Economics 26, 175-191.
Rosett, Joshua G., 1990, Do union wealth concessions explain takeover premiums? The evidence on
contract wages, Journal of Financial Economics 27, 263 282.
Scharfstein, David, 1988, The disciplinary role of takeovers, Review of Economic Studies 55,
185-199.
Schnitzer, Monika, 1995, Breach of trust in takeovers and the optimal corporate charter, Journal of
Industrial Economics 43, 229-259.
Shivdasani, Anil, 1993, Board composition, ownership structure, and hostile takeovers, Journal of
Accounting and Economics 16, 167-208.
Shleifer, Andrei and Lawrence H. Summers, 1988, Breach of trust in hostile takeovers, In: Alan J.
Auerbach, ed., Corporate takeovers: Causes and consequences (University of Chicago Press,
Chicago, IL).
Shleifer, Andrei and Robert W. Vishny, 1989, Management entrenchment: The case of manager-specific
investments, Journal of Financial Economics 25, 123-139.
Smith, Adam, 1925, An inquiry into the nature and the causes of the wealth of nations (Original
version: 1776) (Bell and Sons Ltd, London).
Warther, Vincent A., 1994, Board effectiveness and board dissent: A model of the board' s relationship
to management and shareholders, Mimeo.
Weisbach, Michael S., 1988, Outside directors and CEO turnover, Journal of Financial Economics 20,
431 460.
Weston, J. Fred, Kwang S. Chung and Susan E. Hoag, 1990, Mergers, restructuring, and corporate
control (Prentice Hall, Englewood Cliffs, NJ).
Zender, Jaime F., 199 I, Optimal financial instruments, Journal of Finance 46, 1645-1663.

Вам также может понравиться