Journal of Banking & Finance 33 (2009) 709718

Contents lists available at ScienceDirect

Journal of Banking & Finance

journal homepage: www.elsevier.com/locate/jbf

Irreversible investment, managerial discretion and optimal capital structure

Andreas Andrikopoulos
University of the Aegean, Department of Financial and Management Engineering, 31 Fostini Str., 82 100 Chios, Greece

a r t i c l e

i n f o

Article history:
Received 18 June 2008
Accepted 6 November 2008
Available online 19 November 2008
JEL classication:
D81
D92
E22
G31
J20
Keywords:
Agency conicts
Irreversible investment
Managerial compensation
Capital structure

a b s t r a c t
We explore the signicance of employee compensation and alternative (reservation) income on investment timing, endogenous default, yield spreads and capital structure. In a real-options setting, a managers incentive to under(over)invest in a project is associated to labor income he has to forego in
order to work on the project, the managers salary, his stake on the projects equity capital and his subsequent income, should he decide to terminate operations. We nd that the optimal level of coupon payments decreases with managerial salary and ownership stake while it is increasing in the managers
reservation income. Yield spreads (optimal leverage ratios) are increasing (decreasing) in the managers
salary and ownership stake, while they are decreasing (increasing) in the managers reservation income.
Exploring agency costs of debt as deviations from a value-maximizing investment policy, we document a
U-shaped relationship between agency costs of debt and the managerial compensation parameters: the
managers reservation income, salary and ownership share.
2008 Elsevier B.V. All rights reserved.

1. Introduction
Being the cornerstone of corporate nance, the question of capital-structure relevance to rm value has generated a still inconclusive debate which has long been enriched with the
operational exibility insights of the real-options paradigm. This
stream of literature has extensively discussed investment timing
and optimal capital structure as emerging out of the conicts of
interest between shareholders and creditors. In this paper, our
objective is to make a contribution to this debate by exploring
the incentives of a manager who makes the nancing and investment decisions and thus by redening agency costs of debt in a
real-options context.
The real-options account of the interaction between investment
and nancing decisions is now over thirty-years-old. Drawing on
the then recently developed option pricing theory, Merton (1974)
derived a valuation result for corporate bonds and employed this
result in constructing a proof on capital-structure irrelevance.
Brennan and Schwartz (1978) used the real-options approach to
reach a result on optimal capital structure and incorporated bankruptcy costs and default probability in their derivation of an optimal capital structure. It is fair to say that the now well-applied
real-options analytical platform on agency costs and capital structure can be largely attributed to the work of Mello and Parsosns
E-mail address: apa@fme.aegean.gr
0378-4266/$ - see front matter 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jbankn.2008.11.002

(1992) who examined the case of a hypothetical mining investment with operational and nancial exibility and were the rst
to measure agency costs of debt. Mauer and Triantis (1994)
produced probably the richest operational exibility setting in
the debate on optimal capital structure and suggested that while
operational exibility signicantly affected the nancing decisions,
nancing exibility had little effect on the operational decision
making. The real-options account of the optimal capital-structure
debate was further advanced by Leland (1994) who introduced
the exibility of debt renegotiation, Leland and Toft (1996) who
explored the fundamental issue of optimal debt maturity and with
Leland (1998) who incorporated the choice of corporate risk and
provided a most comprehensive account of the agency-theoretic
agenda within the real-options framework. This platform has since
been enriched with market structure considerations (Lambrecht,
2001), nancing constraints (Boyle and Guthrie, 2003), stochastic
growth opportunities (Childs et al., 2005), incomplete markets
(Hugonnier and Morellec, 2007) and a handful of further modelling
extensions. However, despite the rich variety of models on capital
structure and agency costs, this debate has yet to fully incorporate
a classic agency problem: the ownermanager conict. Most of
these models have discussed corporate settings in which a creditor
signed a contract with an entrepreneur who was also the sole
owner, thus prompting the need to include managerial
discretion in investment decision making, within the real-options
framework.

710

A. Andrikopoulos / Journal of Banking & Finance 33 (2009) 709718

Cadenillas et al. (2004) built on the option-like features of corporate securities and explored shareholdermanager conicts as
well as the effect of managerial compensation on capital structure.
In this modelling framework, managers were only rewarded with
stock and decided on corporate risk and the level of effort they
were to exert, while the choice of leverage and of the level of managerial compensation was made by the shareholders. In a corporate
setting where stock ownership is the only compensation of risk
averse managers which face costly effort, the capital-structure
choice would depend on the managers efciency, momentum as
well as company size. Grenadier and Wang (2005) revisited the
question of investment timing for an option to invest, in the context of ownermanager contracts in an all-equity rm. In their
agency-theoretic contribution to the problem of optimal investment timing, they explored the issue of asymmetric information
and costly effort. Decomposing the option to invest into a managers option and an owners option, they found that both underinvestment and overinvestment can hold in equilibrium. Mauer and
Sarkar (2005) examined the effect of the conict between shareholders and creditors on capital structure and on investment timing. Based on the tax-deductibility of interest income, they found
that the shareholders choice of investment and nancing will generate deviations from value maximization and will create agency
conicts and costs. In their model, the shareholders incentive to
overinvest imposes agency costs on yield spreads and the resulting
cost of capital determines nancing as well as investment
decisions.
Making a contribution to the ongoing debate on the interaction
between investment timing and capital structure, this paper is
motivated by empirical ndings that the ownercreditor agency
conict alone cannot explain the observed variations in deviations
from value-maximizing investment (Parrino and Weisbach, 1999)
and that managerial compensation is signicant in determining
capital structure (e.g. Smith and Watts, 1992) as well as corporate
value (e.g. Coles et al., 2001). We explore over-and-under investment with respect to the managers compensation and interactions
with corporate investors, in the framework of a levered rm.
Investment timing, real option value, optimal capital structure
and agency costs of debt are estimated for various structures of
managers compensation and reservation income. For the purposes
of our analysis on investment timing, we employ the investment
nancing setting of Mauer and Sarkar (2005), where corporate
decision makers decide on investment timing and optimal capital
structure, which is implemented at the moment when the project
starts. Section 2 lays out our real-options framework, Section 3
analyzes some analytical and numerical results and Section 4 concludes the paper, indicating directions for future research.
2. An investment setting
We explore the case of a rm that has a monopolistic, perpetual
right to implement an investment project. The projects cost is I.
The decision to invest is made by the rms manager.1 The managers compensation consists of an equity stake a on the rm and
a xed salary Cp which is collected per unit of time.2 In order for
1
In this approach, implicit is the assumption that the shareholders cannot manage
the operations of theproject themselves and this is why the manager is hired. This
assumption matters in the discussion ofthe managers deviations from equity
maximizing investment in Section 3: the shareholders are worseoff in the sense that
investment and nancing choices are not aligned with their objectives but, on
theother hand, running the project on their own is more inefcient (e.g. substantial
decline in operatingprotability).
2
One could argue here that the managers compensation is taxed. Imposing a
personal tax rate wouldnot change our conclusions on investment timing and capital
structure. Alternatively, one could thinkof Cp as the after-tax managers income from
salary.

the project to start, the manager will have to give up previous

employment that yields income PI. The rms income from projects
operations is taxed at a tax rate sc. Projects operations generate stochastic revenue P and incur a xed cost C per unit of time. The managers salary is part of the xed cost, hence we must have C > Cp. We
also assume that the dynamics of P can be replicated by forming a
portfolio of traded assets in a risk neutral, no-arbitrage economy
and satisfy the following stochastic differential equation

dP r  dPdt rPdz;

where r is a risk free interest rate, r is the standard deviation of annual returns on a portfolio that perfectly replicates the dynamics of
P, d is a convenience yield that can be earned by holding the production output in inventory, dt is an increment of time and dz is the
increment of a Wieners process. When the option to invest is exercised, the capital-structure choice is implemented and the investment project could be partly nanced with a debt contract, in
which case the amount of debt nancing is K and the shareholders
contribute the rest I  K. This nancing arrangement can happen in
the setting of a line-of-credit contract (or a loan commitment), in
which external funds are committed up to a contractually specied amount to be available in a future point in time for the rms
needs. The amount of debt nancing K is determined in a contract
signed by the manager and the creditor. If debt nancing is provided, the rm faces a xed coupon payment R per unit of time. Project abandonment and bankruptcy are decided upon by the
manager. Should the rm go bankrupt, the creditor will take over
the rms assets, suffering bankruptcy costs b (0 < b < 1). In the
event of bankruptcy, the manager will nd employment in another
job that yields an income of RI that can be considered as a kind of
reservation income.
2.1. Solving for the unlevered asset value
After the project has started, the operational exibility that is
available to the managers discretion consists of the option to shut
down the project. In this all-equity case, well-known portfolio replication arguments can be used to show that the managers wealth
MU(P) satises

0:5r2 P 2 M UPP r  dPMUP  rM U aP  C1  sc C p 0:

The general solution of (2) is of the form



P C
Cp
1  sc A1 Pb1 A2 Pb2 for P > PUA ;
Mu P a

d r
r

where A1 and A2 are constants to be determined and P UA is the optimal abandonment level, maximizing the managers wealth. The
solution to the homogeneous part of the equation yields

s

2
rd 1
2r

2 > 1 and
2
2
r
r
s

2
1 r  d
r  d 1
2r
b2 


2 < 0:
2
2
r2
r2
r

1 r  d
b1 

2
r2

Since the abandonment option is decreasing in P, we need

A1 = 0.
The equation for the value of managers wealth in the case of
the unlevered rm is subject to the following boundary conditions.



P C
Cp
1  sc
lim M U P a

P!1
d r
r
MU PUA RI and

@MU 
0

@P  U
PPA

4:1
4:2
4:3

711

A. Andrikopoulos / Journal of Banking & Finance 33 (2009) 709718



P CR
ML P a

1  sc
d
r

Applying (4.1)(4.3), (4.1), we get



P C
M U P a  1  sc
d r
"
#
!
!b2
Cp
P UA C
Cp
P
 1  sc  RI
;
 a
r
r
d
r
PUA

!
!b 2
!
Cp
PLD C R
Cp
P
1  sc  RI
 a

r
r
d
r
PLD

where

and

PLD
Cp
r

db a Cr 1  sc  RI
PUA 2
:
a1  sc 1  b2



C
db2 a CR
1  sc rp  RI
r

6
subject to the boundary conditions

7:1

and

0:5r2 P2 ELPP r  dPELP  rEL P  C  R1  sc 0

14:1

EL PLD 0:

14:2

The solution yields

E P
7:2

The solution of (6) yields

V P

the abandonment trigger PUA being known from the solution to the
managerial wealth problem. These differences in the operating
choices of corporate decision makers and residual claimants seen
here in the case of default and investment triggers create the need
for an effective system of corporate control (Fama and Jensen,
1983).
This valuation result will be needed to calculate the asset value
accruing to the creditor in the event of default.
2.2. Calculating managerial wealth in a levered rm
In the case of a levered rm, the value of managers wealth
ML(P) satises

0:5r2 P2 MLPP r  dPM LP  rM L aP  C  R1  sc C p 0:

9
The general solution of (9) is of the form



P CR
Cp
1  sc A3 Pb1 A4 Pb2 for P > P LD ;
M L P a

d
r
r
10
where A3 and A4 are constants to be determined and PLD is the optimal default trigger, as seen from the managers point of view.
Since the abandonment option is decreasing in P, we need
A3 = 0. The equation for the value of managerial wealth in the case
of the unlevered rm is subject to the following boundary
conditions.

M L PLD RI and

@M L 
0

@P  L
PP D

If we apply (11.1), (11.2) and (11.3) to (10) we get

!
!b 2

P CR
PLD C R
P
1  sc 
1  sc :


d
r
r
d
P LD
15

PLD

!
!b 2

P C
PUA C
P
1  sc

1  sc 

d r
r
d
PUA



P CR
Cp
lim ML P a

1  sc
P!1
d
r
r

13



P CR
Cp
lim EL P a

1  sc
and
P!1
d
r
r

subject to the boundary conditions

0:5r2 P2 V UPP r  dPV UP  rV U P  C1  sc 0 for P > PUA

V U PUA 0:

a1  sc 1  b2

As for the shareholder, the value of equity EL(P) satises the following partial differential equation

As for the shareholders, the value of the unlevered rm VU(P) satises the following partial differential equation



P C

1  sc
lim V U P
P!1
d r

12

being known from the solution to the managerial wealth

problem.
Following e.g. Leland (1994), the value of debt D(P) after the
exercise of the investment option, is given by

DP

R
A5 Pb1 A6 Pb2
r

16

and constants A5 and A6 are determined by the boundary conditions

R
and
r
L
DPD 1  bV U PLD :
lim DP

17:1

P!1

17:2

Implicit in (17.2) is the assumption that, if the rm goes bankrupt

and the creditor becomes the sole owner, the creditor will hire a
manager to run the then unlevered rm and the manager will of
similar type as the one we have discussed so far. This can be a legitimate assumption since it is more efcient for the new owners to
continue operating the project with managers that have more
rm-specic knowledge and value (Douglas, 2001). Substituting
for (17.1) and (17.2) in (16) yields

!b 2


R
R
P
U
L
DP 1  bV PD 
:
r
r PLD

18

Summing debt and equity and rearranging the components of the

value of the levered rm yields

sc R @
V P V P
1
L

P
PLD

!b 2 1
A  bV U PL
D

P
PLD

!b 2
:

19

(19) demonstrates that the value of levered rm is equal to the

sum of the value of unlevered rm and the tax benets of debt
minus the value of bankruptcy costs.3

11:1
11:2
11:3

3
The effect of the tax advantages of debt and of the bankruptcy costs on the choice
of capital structurehas recently been conrmed in a survey on the decision-making
criteria of European CFOs (Brounen et al., 2006) as well as in empirical work on
international evidence on the capital structure choice (de Jong et al., 2008; Wu and
Yue, 2009).

712

A. Andrikopoulos / Journal of Banking & Finance 33 (2009) 709718

2.3. Pricing the option to invest

The solution of (24) is of the form

The manager has the right to choose the time of project implementation. Upon exercise of this real option, the manager will start
receiving salary Cp, he will pay a fraction a of the equitys contribution to investment cost (since he is rewarded with this portion of
the rms equity) but he will also have to part with his previous income PI, which could be taken to be the present value of a perpetual stream of salary payments that come from competing
professional engagements, even within the same rm.
Let PM be the exercise trigger for the investment option and K
the amount of debt nancing. The value of the managers exibility
M before the option is exercised satises

1 2 2
r P MPP r  drPMP  rM 0;
2

P < PM :

20

P < PM ;

MP A7 Pb1 A8 Pb2 ;

25

where constants A9 and A10 are calculated with the boundary

conditions

FPF TV L PF  I  PI

26

and



@F 
@V L 


@PPPF
@P 

27

PPF

TV given in (23). (27) can be used to numerically solve for the

investment trigger PF, in the value-maximizing setting of (25) and
(26). We dene agency costs of debt as

ACM

The solution of (20) is of the form

P < PF ;

FP A9 Pb1 A10 Pb2 ;

FPF  F M PM
FP F

28

where constants A7 and A8 are determined with the boundary

conditions

As a benchmark, we can calculate agency costs of debt in the

case of an equity-maximizing investment, abandonment and
nancing policy. Agency costs of debt due to equity maximization
are thus dened as

MPM M L PM  aI  K  PI

ACE

21

22:1

and



@M 
@M L 


@P PPM
@P 

22:2
PP M

can be used to numerically solve for PM.

Upon exercise, the creditors will be willing to supply capital
only equal to the equilibrium value of debt under an investment
policy that maximizes the managers wealth. Therefore, at exercise,
we will have K = D(PM). Yield spreads are dened as the difference
R
and
between the effective rate that is charged on corporate debt DP
the risk free rate of interest r. We introduce yield spreads in our
analysis of managerial compensation and capital structure based
on the theoretical prediction of John and John (1993) that yield
spreads are expected to be increasing in managerial ownership,
since managerial ownership affects the outcome of agency conicts and may lead to deviations from value and equity
maximization.
Let FM be the value of the option to invest as a rms asset, if the
investment policy aims at maximizing managerial wealth. FM can
be calculated if we measure the value of M with Cp, RI, PI equal
to zero and a equal to 1, maintaining however, PM as an investment
trigger. If we set Cp, RI, PI equal to zero and a equal to 1 and further
use this parameter set to estimate an investment trigger we can
nd an equity-maximizing investment trigger PE.
In the same vein, we can calculate the value of the option to invest under a policy that aims at maximizing the value of all involved stakeholders: debt, equity and managerial wealth. Total
value is given by
L

TVLTV D 1  aE M :

23

Since the managers wealth ML includes a portion a of equity

ownership, we have to subtract aEL, in order to avoid a double calculation of this part of equity value. Accordingly, total investment
cost includes not only I, but also PI, which is the income that the
manager foregoes in order to get involved in the project.
In this rst-best case the option to invest satises

1 2 2
r P F PP r  drPF P  rF 0;
2

P < PF :

PF being a rst-best, value-maximizing investment trigger.

24

FPF  F E PE
FPF

29

where FE is the value of the option to invest in the setting of an equity-maximizing investment policy.
Numerical results in the following discussion on investment
timing can help discuss investment timing and capital structure
as outcomes of the decision makers employment conditions: high
salary, low salary and large ownership stake.
3. Numerical results
In this section we discuss the effect of the managers reservation
income, salary and ownership stake on investment timing, optimal
capital structure, yield spreads and agency costs of debt. Numerical
results on the effect of the other parameters interest rate, bankruptcy costs, volatility, etc. can be found in previous work on real
options and capital structure (e.g. Leland, 1998; Mauer and Ott,
2000). Without any loss of generality, we will assume that RI  PI,
that is the income the manager will have to forego, should he decide
to implement the project, is the same as his reservation income.
Throughout the numerical analysis of this section we employ
the same reference set of parameters: P = $1, R = $0.8, C = $0.75,
RI = $1, Cp = $0.04, I = $3, r = 0.05, d = 0.02, a = 0.03, b = 0.35,
sc = 0.2, r = 0.3.
3.1. Investment timing and capital structure: The effect of managerial
compensation
We explore the effect of the parameters which are related to the
managers compensation (reservation income, salary and ownership share) on investment timing and capital structure. It is the extent of deviation from value-maximizing investment which affects
the size of the yield spread imposed by the creditor and it is this
cost of debt that affects the managers choice over coupon size
and capital structure.
3.1.1. Managerial compensation and the interaction between
investment and nancing decisions: Managers reservation income
The managers reservation income affects the managers investment and nancing decisions and therefore affects the value of the
real option to invest and the magnitude of the yield spread on

713

A. Andrikopoulos / Journal of Banking & Finance 33 (2009) 709718

Option value to the manager vs coupon payment for varying reservation income

Investment trigger vs Reservation income

3. 5

0.85

RI=1.1
0.75

Investment trigger

Option value to the manager

RI=1

0.8

RI=1.2

0.7

0.65

0.6

FB

1. 5
EB

0.55

0. 5
0. 6

0. 7

0. 8

0. 9

1. 1

1. 2

1. 3

Reservation income
0

0.5

1.5

2.5

Coupon
Fig. 1a. Sensitivity of option value to coupon payments, for varying levels of the
managers reservation income. Output price (P) is $1, operating cost (C) is $0.75,
managers salary (Cp) is $0.04, project cost (I) is $3, risk free rate of interest (r) is
0.05, dividend yield (d) is 0.02, managers ownership share (a) is 0.03, bankruptcy
costs are 0.35 (b), tax rate for corporate income (sc) is 0.2, and standard deviation of
annual returns (r) is 0.3.

corporate debt. Our numerical solutions can be synopsized in the

following result:
Result 1: The managers reservation income has a positive effect
on the level of the investment trigger and it has a negative effect on
the value of the managers real option to invest as well as on the
level of the yield spread.
The value of the managers option to invest and the subsequent
value of managerial wealth after the real option has been exercised
are decreasing in the managers reservation income (Fig. 1a); higher levels of reservation income essentially delay investment and
precipitate default thus reducing the prospective value of income
out of project implementation.4
The effect of the reservation income on the managers option to
invest stems from the impact of reservation income on investment
timing and the capital-structure choice. If the reservation income is
very low, the manager will be very eager to start the project in order to earn the salary and benet from the tax shield of debt, thus
ending up to overinvest, compared to both the rst-best (FB) and
the equity-maximizing (EB) investment policies (Fig. 1b). On the
other end, if the managers reservation income is high, he will
not start the project until protability is high enough to compensate him for the sacrice of the reservation income, thus resulting
in an exercise policy of underinvestment. For intermediate levels of
reservation income RI ranging from $0.8 to $0.92 the manager
will have an incentive to overinvest with respect to the FB policy
and underinvest with respect to the EB policy.
Deviations from value-maximizing investment -project start
and termination- have their impact on the yield spreads. The managers reservation income has a decreasing effect on yield spreads
(Fig. 1c); High levels of reservation income weaken the managers
equity-driven incentive to overinvest and this effect outweighs the
managers motive for early default. These conditions make the
creditors decrease yield spreads and hence the cost of debt. Lower
yield spreads -associated with higher reservation income- will in4
Differentiating the default trigger can show that the effect of managerial
compensation on the level ofthe default trigger is similar in spirit to the effects on
investment timing; high salary makes projectrelatedemployment more attractive and
delays default, whereas high reservation income makesalternative employment
opportunities more attractive and thus precipitates default.

Fig. 1b. Sensitivity of investment trigger to the managers reservation income. The
investment trigger is chosen so as to maximize the value of the managers option to
invest (MB). The dashed line is the rst-best investment trigger (FB) which
maximizes value for all parties (shareholder, creditor and manager) and the dotted
line is the investment trigger in an operating policy which maximizes equity value
(EB). Output price (P) is $1, coupon payment (R) is $0.8, operating cost (C) is $0.75,
managers salary (Cp) is $0.04, project cost (I) is 3$, risk free rate of interest (r) is
0.05, dividend yield (d) is 0.02, managers ownership share (a) is 0.03, bankruptcy
costs (b) are 0.35, tax rate for corporate income (sc) is 0.2 and standard deviation of
annual returns (r) is 0.3.

Yield spread vs Reservation income

160
150
140
130

Yield spread

0.5

MB

2. 5

120
110
100
90
80
70
0.8

0.9

1.1

1.2

1.3

Reservation income
Fig. 1c. Sensitivity of the yield spread to the managers reservation income. Yield
spread is dened as the difference between the effective rate on corporate debt and
the risk free rate (r). The effective rate is calculated as the ratio of coupon R to debt
value D(P). The spread is measured in basis points. Output price (P) is $1, coupon
payment (R) is $0.8, operating cost (C) is $0.75, managers salary (Cp) is $0.04,
project cost (I) is $3, risk free rate of interest (r) is 0.05, dividend yield (d) is 0.02,
managers ownership share (a) is 0.03, bankruptcy costs (b) are 0.35, tax rate for
corporate income (sc) is 0.2 and the standard deviation of annual returns (r) is 0.3.

crease the level of the optimal coupon payment as well as the optimal leverage ratio5, since debt is cheaper and also because increased
tax shields are required to induce the manager to give up his reservation income and start the project. As far as I know, there has been
no previous research to associate the managers income from alternative employment opportunities with the capital-structure question in a dynamic setting of irreversible investment.

5
We dene the optimal leverage ratio as the prevailing leverage ratio when the
coupon payment is at its optimal level.

714

A. Andrikopoulos / Journal of Banking & Finance 33 (2009) 709718

In the case of our numerical example (Fig. 1b), the optimal size
of the coupon ranges from $1.02 for the RI = $1 case to $1.29 for the
RI = $1.2 case. However, due to the fact that optimal coupon payment increases with the reservation income, the decline in option
value starts at lower coupon levels in the case of low reservation
incomes and this is the reason that after the optimal coupon payment has been reached and before option exercise option value
for low reservation income can briey be less than option value
for options with higher reservation income. For our parameter
set, option should be immediately exercised for coupon values on
the right of the kink in the coupon-option graph. After the option
to invest has been exercised, we are essentially addressing an issue
of the effect of coupon payments on managers wealth when he is
in charge of an ongoing project and the negative association between reservation income and the value of the managerial option
still holds for the reasons discussed above.
In the setting of our numerical example, we observe that, as we
expected, optimal leverage is increasing in the managers reservation income; optimal leverage ratio is 0.43 for R = $1 and it increases to 0.52 for R = $1.2. It is the lower yield spreads for high
levels of reservation income that make increased leverage more
attractive in the case of higher reservation income.
3.1.2. Managerial compensation and the interaction between
investment and nancing decisions: Managers salary
Having examined the impact of the managers reservation income on investment timing and the capital-structure choice, we
proceed to explore the effect of the managers compensation contract that consists of salary and a share of equity ownership. Our
numerical examination of the effect of the managers salary on
the interaction between investment and nancing decisions has
yielded the following result.
Result 2: The managers salary has a negative effect on the level
of the investment trigger and it has a positive effect on the value of
the managers real option to invest as well as on the level of the
yield spread.
High levels of salary naturally make the managers option worth
more and also increase managerial wealth in the case of managing
an ongoing levered company after the project has started (Fig. 2a).
Furthermore, the higher the managers salary, the stronger the
managers motive to start the project and hence to exercise the real

option at a lower level of P (Fig. 2b). For low salary levels, the
manager does not want to give up his reservation income and
get involved in the project, thus leading to phenomena of underinvestment with respect to both FB and EB policies.
Associated with the managers motive to overinvest, the effect
of the managers salary on the yield spreads is positive. An increased salary will strengthen the managers incentive to overinvest so as to start getting paid and will also make him want to
(suboptimally) keep the project alive for as long as possible, so as
to maintain the xed income ow. This overinvestment motive
leads to high spreads in the cases of high salaries (Fig. 2c). Higher
costs of debt for higher levels of salary mean that, as the managers
salary increases, the optimal coupon size and the optimal leverage
ratio decrease; optimal coupon size is $1.29 and optimal leverage
is 0.52 for a salary of $0.03, while optimal coupon size is $0.75
and optimal leverage is 0.27 for a salary of $0.05).
Our result on the effect of managers salary on investment timing and capital structure extends the discussion in Mauer and Sarkar (2005) on the determinants of investment and nancing
decisions in a real-options framework.
3.1.3. Managerial compensation and the interaction between
investment and nancing decisions: Managers ownership stake
Managerial ownership has often been employed as a contractual mechanism in the direction of aligning the objectives of the
managers with the ones of the owners. The following result summarizes the ndings of our analysis on effect of managerial ownership on the managers choice over investment timing and capital
structure.
Result 3: The managers ownership share has a negative effect
on the level of the investment trigger and it has a positive effect
on the value of the managers real option to invest as well as on
the level of the yield spread.
The effect of the managers ownership share on the value of his
option to start a project is, of course, positive (Fig. 3a). The higher
the compensation involved in running the project, the more valuable will the option be and the more valuable his wealth will be

Investment trigger vs Manager's salary

3.5
MB

Option value to the manager vs coupon payment for varying manager's salary

Cp =0 .0 5

0. 85

Option value to the manager

Investment trigger

0. 9

0. 8
0. 75

2.5
FB

1.5
EB

Cp =0 .0 4

0. 7

1
0. 65
Cp =0 .0 3

0. 6

0.5
0.02

0. 5

0.025

0.03

0.035

0.04

0.045

0.05

0.055

0.06

Manager's salary

0. 55

0. 5

1. 5

2. 5

Coupon
Fig. 2a. Sensitivity of option value to coupon payments, for varying managers
salary. Output price (P) is $1, operating cost (C) is $0.75, managers reservation
income (RI) is $1, project cost (I) is $3, risk free rate of interest (r) is 0.05, dividend
yield (d) is 0.02, managers ownership share (a) is 0.03, bankruptcy costs are 0.35
(b), tax rate for corporate income (sc) is 0.2 and standard deviation of annual returns
(r) is 0.3.

Fig. 2b. Sensitivity of investment trigger to the managers salary. The investment
trigger is chosen so as to maximize the value of the managers option to invest (MB).
The dashed line is the rst-best investment trigger (FB) which maximizes value for
all parties (shareholder, creditor and manager) and the dotted line is the investment
trigger in an operating policy which maximizes equity value (EB). Output price (P) is
$1, coupon payment (R) is $0.8, operating cost (C) is $0.75, managers reservation
income (RI) is $1, project cost (I) is 3$, risk free rate of interest (r) is 0.05, dividend
yield (d) is 0.02, managers ownership share (a) is 0.03, bankruptcy costs (b) are
0.35, tax rate for corporate income (sc) is 0.2 and standard deviation of annual
returns (r) is 0.3.

715

A. Andrikopoulos / Journal of Banking & Finance 33 (2009) 709718

Yiled spread vs Manager's Salary

160
150
140

Yield spread

130
120
110
100
90
80
70
0. 02

0. 025

0. 03

0. 035

0. 04

0. 045

0. 05

Manager's salary
Fig. 2c. Sensitivity of the yield spread to the managers salary. Yield spread is
dened as the difference between the effective rate on corporate debt and the risk
free rate (r). The effective rate is calculated as the ratio of coupon R to debt value
D(P). The spread is measured in basis points. Output price (P) is $1, coupon payment
(R) is $0.8, operating cost (C) is $0.75, managers reservation income (RI) is $1,
project cost (I) is $3, risk free rate of interest (r) is 0.05, dividend yield (d) is 0.02,
managers ownership share (a) is 0.03, bankruptcy costs (b) are 0.35, tax rate for
corporate income (sc) is 0.2 and the standard deviation of annual returns (r) is 0.3.

after the exercise of the investment option (on the right of the
kink).
The level of the investment exercise trigger is decreasing in the
level of the managers ownership share; for low levels of managerial ownership share, the manager is less eager to leave his previous (reservation) income in order to start the project and gain a
small portion of corporate prots and tax shields and therefore
he tends to underinvest, compared to both the rst-best investment trigger and the equity-maximizing investment policy. For
sufciently high levels of ownership share, he is more willing to
part with his reservation income and benet from the tax shields
of debt and this is why he will end up overinvesting (Fig. 3b).
Our nding that managerial ownership can be associated with

deviations from equity-maximizing and also from value-maximizing investment is in line with the intuition of John and John (1993)
that managerial ownership affects the agency costs of both equity
and debt. It is interesting that, for high levels of a, additional ownership does not offer improved alignment with equity-maximizing
investment policies. This is a result of the fact that for high levels of
managerial ownership, the sensitivity of the investment trigger to
the managers salary and reservation income weakens and the
deviations from equity-maximization cannot be further restrained
by granting the managers an additional stake of corporate ownership. The weak sensitivity of the investment trigger with respect to
managerial ownership agrees with empirical evidence in Singh and
Davidson (2003) that increased managerial ownership cannot
eliminate managerial motives to exercise discretion and deviate
from equity maximization.
Fig. 3c shows that the higher the managers stake on equity
ownership, the higher the yield spread will be. This is because
the higher the managers equity stake, the stronger the tendency
of the manager to abide by an equity-maximizing policy calling
for overinvestment and late default in order to rip off the tax
shields of debt. Therefore, in the case of increased managerial ownership, the increased risk of debt contracts can lead to an increased
yield spread. This result is in accordance with the theoretical prediction of John and John (1993) as well as the empirical evidence of
Strock Bagnani et al. (1994) and Dadyvenko and Strebulaev (2007)
that yield spreads should be increasing in the ownership component of managerial compensation.
The positive effect of managerial ownership on yield spreads
and hence on the cost of debt means that higher levels of managerial ownership will be associated with lower levels of optimal coupon payment and also with lower optimal leverage ratios; as the
managers ownership stake increases, the size of the optimal coupon decreases from $1.02 for a = 0.03 to $0.87 for a = 0.07. For this
parameter range, the optimal leverage ratio falls from 0.43 to 0.27
because the equity-driven incentive to overinvest leads to increased yield spreads and therefore to a lower optimal coupon

Investment trigger vs Manager's ownership share

2.6

2.4

Option value to the manager vs coupon payment for varying ownership shares
2

2.2

Investment trigger

a=0.7

Option value to the manager

1.8

1.6

1.4
a=0.5

MB

1.8
FB
1.6

1.2
1.4
EB

1
0

0.8

a=0.3

0.5

1.5

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Manager's ownership share

2.5

Coupon
Fig. 3a. Sensitivity of option value to coupon payments, for varying managers
ownership share. Output price (P) is $1, operating cost (C) is $0.75, managers salary
(Cp) is $0.04, managers reservation income (RI) is $1, project cost (I) is $3, risk free
rate of interest (r) is 0.05, dividend yield (d) is 0.02, bankruptcy costs are 0.35 (b),
tax rate for corporate income (sc) is 0.2 and standard deviation of annual returns (r)
is 0.3.

Fig. 3b. Sensitivity of investment trigger to the managers ownership share. The
investment trigger is chosen so as to maximize the value of the managers option to
invest (MB). The dashed line is the rst-best investment trigger (FB) which
maximizes value for all parties (shareholder, creditor and manager) and the dotted
line is the investment trigger in an operating policy which maximizes equity value
(EB). Output price (P) is $1, coupon payment (R) is $0.8, operating cost (C) is $0.75,
managers reservation income (RI) is $1, managers salary (Cp) is $0.04, project cost
(I) is 3$, risk free rate of interest (r) is 0.05, dividend yield (d) is 0.02, bankruptcy
costs (b) are 0.35, tax rate for corporate income (sc) is 0.2 and standard deviation of
annual returns (r) is 0.3.

716

A. Andrikopoulos / Journal of Banking & Finance 33 (2009) 709718

Yield spread vs Manager's ownership share

Agency costs of debt vs Reservation income

155

0. 04

150

0.035
0. 03

Agency costs of debt

Yield spread

145

140

135

130

0. 02
0.015
AC E

0. 01

125

120
0. 05

AC M

0.025

0.005

0. 1

0. 15

0. 2

0. 25

0. 3

0. 35

0. 4

0. 45

0
0. 7

0. 5

0. 8

0. 9

and leverage ratio. This result is in accordance with ndings in Jensen et al. (1992) that higher insider ownership is associated with
lower levels of debt.
3.2. The effect of managerial compensation on agency costs of debt
Agency costs of debt were dened in (28) as the percentage difference of option value under an investment policy that maximizes
total value (i.e. equity, debt, managerial wealth) and option value
under an investment policy that aims at maximizing managerial
wealth. Our numerical solutions on the effect of managerial compensation on agency costs have produced the following nding.
Result 4: There is an U-shaped relationship between agency
costs of debt and the managerial compensation parameters: the
managers reservation income, salary and ownership share.
We see in Fig. 4a that there exists an U-shaped pattern with respect to the effect of the managers reservation income to the
agency costs of debt. For very low values of reservation income
there is a strong incentive to overinvest and this leads to deviations
from value maximization. As the reservation income increases,
overinvestment incentives weaken and agency costs of debt decrease. However, beyond a point, high values of reservation income
decrease the attractiveness of project-related compensation to the
manager and thus he tends to underinvest and underinvestment
leads to increased agency costs of debt. The ACE line in Fig. 4a
shows agency costs of debt under an equity-maximizing setting,
as dened in (28). We see that the agency costs of debt are lower
under the equity-maximizing investment and nancing policies,
in cases reservation income is too low (worse managerial overinvestment) or too high (worse managerial underinvestment). For
intermediate values of reservation income, we see that the agency
costs of debt under an equity-maximizing policy are higher than
under a policy that maximizes managerial wealth. This means that
the creation of an agency conict generates an outcome that is closer to the rst-best value maximization under the managers than
under the shareholders wealth maximization. This result can be
considered as an extension of the intuition of Brander and Poitevin
(1992) and John and John (1993), in the direction of dynamic
investment choice.

1. 1

1. 2

1. 3

1. 4

Fig. 4a. Sensitivity of agency costs of debt to the managers reservation income.
Agency costs of debt (ACM) are dened as the percentage difference between option
value under a nancinginvestment policy that maximizes managerial wealth and
one that maximizes value for all parties (shareholder, creditor and the manager).
The ACE line plots agency costs of debt in an equity-maximizing investment
nancing policy. Output price (P) is $1, coupon payment (R) is $0.8, operating cost
(C) is $0.75, managers salary (Cp) is $0.05, project cost (I) is $3, risk free rate of
interest is (r) 0.05, dividend yield (d) is 0.02, managers ownership share (a) is 0.03,
bankruptcy costs (b) are 0.35, tax rate for corporate income (sc) is 0.2 and standard
deviation of annual returns (r) is 0.3.

We now proceed to discuss the effect of managers salary on the

agency costs of debt (Fig. 4b). We see that there exists a U-shaped
relationship between managerial salary and the agency costs of
debt. For very low salary levels the manager will tend to underinvest because he will not have sufcient motivation to part with
the reservation income and default too early because the lower
the managers salary the greater the attractiveness of the reservation income in the case of default. As salary increases, these agency

Agency costs of debt vs manager's salary

0.04
0.035
0.03

Agency costs of debt

Fig. 3c. Sensitivity of the yield spread to the managers ownership share. Yield
spread is dened as the difference between the effective rate on corporate debt and
the risk free rate (r). The effective rate is calculated as the ratio of coupon R to debt
value D(P). The spread is measured in basis points. Output price (P) is $1, coupon
payment (R) is $0.8, operating cost (C) is $0.75, managers reservation income (RI) is
$1, managers salary (Cp) is $0.04, project cost (I) is $3, risk free rate of interest (r) is
0.05, dividend yield (d) is 0.02, bankruptcy costs (b) are 0.35, tax rate for corporate
income (sc) is 0.2 and the standard deviation of annual returns (r) is 0.3.

Reservation income

Manager's ownership share

ACM
0.025
0.02
0.015

ACE

0.01
0.005
0
0.02

0.025

0.03

0.035

0.04

0.045

0.05

0.055

Manager's salary
Fig. 4b. Sensitivity of agency costs of debt to the managers salary. Agency costs of
debt (ACM) are dened as the percentage difference between option value under a
nancinginvestment policy that maximizes managerial wealth and one that
maximizes value for all parties (shareholder, creditor and the manager). The ACE
line plots agency costs of debt in an equity-maximizing investmentnancing
policy. Output price (P) is $1, coupon payment (R) is $0.8, operating cost (C) is $0.75,
managers reservation income (RI) is $1, project cost (I) is $3, risk free rate of
interest is (r) 0.05, dividend yield (d) is 0.02, managers ownership share (a) is 0.03,
bankruptcy costs (b) are 0.35, tax rate for corporate income (sc) is 0.2 and standard
deviation of annual returns (r) is 0.3.

717

A. Andrikopoulos / Journal of Banking & Finance 33 (2009) 709718

effects weaken and the managers policy approaches value-maximization choices. Beyond a certain point of salary increase however, the manager will have an incentive to invest too early so as
to earn the high salary and will also have the incentive to default
too late so as not to give up the substantial salary income. These
managerial incentives cause the agency costs of debt to increase
in the case of high salary compensations. Between these two extreme cases, agency costs of debt in an equity-maximizing regime
are higher, compared to agency costs of debt under a policy that
maximizes managerial wealth.
Agency costs of debt have a U-shaped relationship with the
managers ownership stake. For low levels of managerial ownership, the manager will underinvest since compensation from the
project may be less attractive compared to the reservation income.
This underinvestment motive weakens as the managers ownership stake increases, shortening the deviations from a value-maximizing investment policy. However, beyond a point, increased
ownership renders the project increasingly attractive, leading to
overinvestment and an increase in deviations from value-maximizing policies. We observe in Fig. 4c that agency costs of debt under a
policy that maximizes managerial wealth (ACM) are lower than
agency costs of debt in a setting of equity maximization (ACE). This
is because the overinvestment incentives of the shareholders are
partly offset by the effects of the managers reservation income.
This effect of managerial ownership on the agency costs of debt
is more prevalent when managerial ownership is at low levels.
When the managers stake on equity capital is large, we have seen,
in Section 3.1, that the marginal effect of managerial ownership on
the decisions to invest and default is rather small. As a result of
this, when managerial ownership is large, the agency costs of debt
will not be substantially affected by the extent of managerial
ownership.
Deviations from value maximization are a result of the debt and
employment contracts. Given that the employment contract determines managerial wealth taking into account both salary and stock
ownership for a given reservation income there is a plethora of
contracts that will provide the manager with the same level of ex-

pected wealth. Assuming a reference level of managerial wealth

(real option value) of $1.2 as a working example, Fig. 4d demonstrates the family of contracts yielding the same level of expected
wealth for the manager. As expected, there is a tradeoff between
salary and stock ownership along the compensation contracts that
yield the same expected wealth of $1.2 to the manager. If the manager is rewarded with higher levels of stock ownership, there has
to be a decrease in the salary in order to keep the managers wealth
at the same level. Furthermore, we see in Fig. 4d that increased levels of the managers reservation income warrant increased compensation in the employment contract of the project for the same
level of expected managerial wealth. As we saw in Fig. 1a, higher
levels of reservation income decrease expected managerial wealth
from the project (i.e. the managers real option). Therefore, as the
reservation income increases, we must also increase managerial
compensation in order to achieve the same level of expected managerial wealth. Observing the wide range of contracts that corresponds to a given level of managerial wealth, welfare
considerations raise the question of an optimal contract; one that
makes it incentive compatible for the manager to implement the
investment policy that maximizes total value. If we solve numerically for the optimal contract, we nd that a contract that grants
the manager with a salary of $0.041 and an ownership stake of
0.044, is consistent with both the benchmark level of $1.2 of expected managerial wealth and also value maximization as it induces rst-best investment timing.
All these contracts that yield the same level of managerial
wealth correspond to varying operating policies and hence to varying levels of yield spread. Fig. 4e shows how the dependence of the
managers salary on the ownership stake in the setting of the $1.2
benchmark level of managerial wealth affects the level of the
investment trigger. The investment trigger increases along this
family of employment contracts. As we increase managerial ownership, the investment trigger is affected by two contrasting effects. Increased stock ownership induces overinvestment but
decreased salary depicted in Fig. 4d provides a motive for
underinvestment and the latter effect prevails, leading to the

Agency costs of debt Manager's ownership share

Employment contracts that yield the same level of managerial wealth

0.015

0.08
ACE

0.07

Manager's salary

Agency costs of debt

RI=1

RI=1.1

0.06

0.01

0.005
ACM

0.05
0.04
0.03
0.02

0
0.02

0.01
0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

Manager's ownership share

Fig. 4c. Sensitivity of agency costs of debt to the managers ownership share.
Agency costs of debt (ACM) are dened as the percentage difference between option
value under a nancinginvestment policy that maximizes managerial wealth and
one that maximizes value for all parties (shareholder, creditor and the manager).
The ACE line plots agency costs of debt in an equity-maximizing investment
nancing policy. Output price (P) is $1, coupon payment (R) is $0.8, operating cost
(C) is $0.75, managers salary (Cp) is $0.05, managers reservation income (RI) is $1,
project cost (I) is $3, risk free rate of interest is (r) 0.05, dividend yield (d) is 0.02,
bankruptcy costs (b) are 0.35, tax rate for corporate income (sc) is 0.2 and standard
deviation of annual returns (r) is 0.3.

0
0.03

0.035

0.04

0.045

0.05

0.055

Manager's ownership share

Fig. 4d. Family of contracts that yield the same expected wealth to the manager, for
varying levels of reservation income (RI). For each value of the stock-ownership
parameter a in the horizontal axis, we nd the respective level of the managers
salary that yields an expected level of managerial wealth equal to $1.2. Output price
(P) is $1, coupon payment (R) is $0.8, operating cost (C) is $0.75, project cost (I) is $3,
risk free rate of interest is (r) 0.05, dividend yield (d) is 0.02, bankruptcy costs (b) are
0.35, tax rate for corporate income (sc) is 0.2 and standard deviation of annual
returns (r) is 0.3.

718

A. Andrikopoulos / Journal of Banking & Finance 33 (2009) 709718

Investment trigger vs Managerial compensation (for a given level of managerial wealth)

orient the capital-structure discussion to an analysis of irreversible

investment in incomplete capital markets.
Acknowledgment

Investment trigger

3.5

I thank an anonymous reviewer for many helpful suggestions.

3
2.5

References

2
1.5
1
0.5
0
0.08
0.06

0.055
0.05

0.04

0.045
0.04

0.02

Manager's salary

0.035
0

0.03

Manager'sownership share

Fig. 4e. Sensitivity of investment trigger to the parameters of the employment

contract which are compatible with a given level of expected managerial wealth.
We rst nd the employment contracts a and Cp that yield expected managerial
wealth equal to $1.2 and then we calculate the resulting investment timing policy
for this family of contracts. Output price (P) is $1, coupon payment (R) is $0.8,
operating cost (C) is $0.75, managers reservation income (RI) is $1, project cost (I) is
$3, risk free rate of interest is (r) 0.05, dividend yield (d) is 0.02, bankruptcy costs (b)
are 0.35, tax rate for corporate income (sc) is 0.2 and standard deviation of annual
returns (r) is 0.3.

upward sloping line of Fig. 4e. It is the investment timing of Fig. 4e

and the respective default policies that affect rm value and the level of the yield spread.
4. Concluding comments
Introducing ownermanager conicts to the agency-theoretic
analysis of the option to invest, we investigated the effect of managerial compensation and reservation income on investment timing and nancing decisions. We found that the investment
exercise trigger is negatively associated to the managers salary
and ownership share while it is positively associated to his reservation income. The manager will choose a coupon payment that
will maximize his wealth and this optimal coupon level is increasing in the managers reservation income and decreasing with the
managers salary and ownership share. As for the agency costs of
debt, we documented a U-shaped relationship between the managers reservation income and agency costs of debt. A similar
U-shaped pattern was found for the case of managers salary and
the managers ownership share. Causally related to the comparative statics of the yield spreads, optimal leverage ratios are
decreasing in the managers salary and ownership share but they
are increasing in his reservation income.
Future work in this area should address the issue of trading frictions and information asymmetries in the market for nancial and
human capital and their impact on the managerial decision on
when to start and how to nance a project. Moreover, the option
to invest could be decomposed into the component owned by
the manager and the one owned by the equity holder. Finally, a
more realistic contractual setting of managerial compensation
would call for the inclusion of employee-stock options. The decision makers risk attitude and effort aversion (Palmon et al.,
2008) and the non-transferability of these option contracts would

Boyle, G.W., Guthrie, G.A., 2003. Investment, uncertainty and liquidity. Journal of
Finance 58, 21432166.
Brander, J.A., Poitevin, M., 1992. Managerial compensation and the agency costs of
debt nance. Managerial and Decision Economics 13, 5564.
Brennan, M.J., Schwartz, E.S., 1978. Corporate income taxes, valuation and the
problem of optimal capital structure. Journal of Business 51, 103114.
Brounen, D., de Jong, S., Koedijk, L., 2006. Capital structure policies in Europe:
Survey evidence. Journal of Banking and Finance 30, 14091442.
Cadenillas, A., Cvitanic, J., Zapatero, F., 2004. Leverage decision and manager
compensation with choice of effort and volatility. Journal of Financial
Economics 73, 7192.
Childs, P.D., Mauer, D.C., Ott, S.H., 2005. Interactions of corporate nancing and
investment decisions: The effect of agency conicts. Journal of Financial
Economics 76, 667690.
Coles, J.W., McWilliams, V.B., Sen, N., 2001. An examination of the relationship of
governance mechanisms to performance. Journal of Management 27, 2350.
Dadyvenko, S.A., Strebulaev, I.A., 2007. Strategic actions and credit spreads. Journal
of Finance 62, 26332671.
de Jong, A., Kabir, R., Nguyen, T.T., 2008. Capital structure around the world: The
roles of rm- and country-specic determinants. Journal of Banking and
Finance 32, 19541969.
Douglas, A.V.S., 2001. Managerial replacement and corporate nancial policy with
endogenous manager-specic value. Journal of Corporate Finance 7, 2552.
Fama, E.F., Jensen, M.C., 1983. Separation of ownership and control. Journal of Law
Economics 26, 301326.
Grenadier, S.R., Wang, N., 2005. Investment timing, agency and information. Journal
of Financial Economics 75, 493533.
Hugonnier, J., Morellec, E., 2007. Corporate control and real investment in
incomplete markets. Journal of Economic Dynamics and Control 31, 17811800.
Jensen, G.R., Solberg, D.P., Zorn, T.S., 1992. Simultaneous determinants of insider
ownership, debt and dividend policies. Journal of Financial and Quantitative
Analysis 27, 247263.
John, T.A., John, K., 1993. Top-management compensation and capital structure.
Journal of Finance 48, 949974.
Lambrecht, B.M., 2001. The impact of debt nancing on entry and exit in a duopoly.
Review of Financial Studies 14, 765804.
Leland, H.E., 1994. Corporate debt value, bond covenants and optimal capital
structure. Journal of Finance 49, 12131252.
Leland, H.E., 1998. Agency costs, risk management and capital structure. Journal of
Finance 53, 12131243.
Leland, H.E., Toft, K.B., 1996. Optimal capital structure, endogenous bankruptcy and
the term structure of yield spreads. Journal of Finance 51, 9871019.
Mauer, D.C., Ott, S.H., 2000. Agency costs, underinvestment and optimal capital
structure: The effect of growth options to expand. In: Brennan, M.J., Trigeorgis,
L. (Eds.), Project Flexibility, Agency and Competition. Oxford University Press,
New York, pp. 151180.
Mauer, D.C., Sarkar, S., 2005. Real options, agency conicts and optimal capital
structure. Journal of Banking Finance 29, 14051428.
Mauer, D.C., Triantis, A.J., 1994. Interactions of corporate nancing and investment
decisions: A dynamic framework. Journal of Finance 49, 12531277.
Mello, A.S., Parsosns, J.E., 1992. Measuring the agency cost of debt. Journal of
Financial Economics 47, 18871904.
Merton, R.C., 1974. On the pricing of corporate debt: The risk structure of interest
rates. Journal of Finance 29, 449470.
Palmon, O., Bar-Yossef, S., Chen, R-R., Venezia, I., 2008. Optimal strike prices of stock
options for effort-averse executives. Journal of Banking and Finance 32, 229
239.
Parrino, R., Weisbach, M.S., 1999. Measuring investment distortions arising from
stockholderbondholder conicts. Journal of Financial Economics 53, 342.
Singh, M., Davidson, W.N., 2003. Agency costs, ownership structure and corporate
governance mechanisms. Journal of Banking and Finance 27, 793816.
Smith, C.W., Watts, R.L., 1992. The investment opportunity set and corporate
nancing, dividend and compensation policies. Journal of Financial Economics
32, 263292.
Strock Bagnani, E., Milonas, N.T., Saunders, A., Travlos, N.G., 1994. Managers, owners
and the pricing of risky debt: An empirical analysis. Journal of Finance 49, 453
477.
Wu, L., Yue, H., (2009). Corporate tax, capital structure, and the accessibility of bank
loans: Evidence from China. Journal of Banking and Finance 33, 3038.