Leland 1994 JF

Corporate Debt Value, Bond Covenants, and Optimal Capital Structure
Hayne E. Leland
The Journal of Finance, Vol. 49, No. 4. (Sep., 1994), pp. 1213-1252.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28199409%2949%3A4%3C1213%3ACDVBCA%3E2.0.CO%3B2-0
The Journal of Finance is currently published by American Finance Association.
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/afina.html.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic
journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,
and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take
advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].
http://www.jstor.org
Wed Feb 27 16:10:37 2008
THE JOURNAL O F FINANCE VOL. XLIX, NO. 4 SEPTEMBER 1994
Corporate Debt Value, Bond Covenants,

and Optimal Capital Structure
HAYNE E. LELAND*
ABSTRACT
This article examines corporate debt values and capital structure in a unified
analytical framework. It derives closed-form results for the value of long-term risky
debt and yield spreads, and for optimal capital structure, when firm asset value
follows a diffusion process with constant volatility. Debt values and optimal lever-
age are explicitly linked to firm risk, taxes, bankruptcy costs, risk-free interest
rates, payout rates, and bond covenants. The results elucidate the different behavior
of junk bonds versus investment-grade bonds, and aspects of asset substitution,
debt repurchase, and debt renegotiation.
THEVALUE OF CORPORATE debt and capital structure are interlinked variables.

Debt values (and therefore yield spreads) cannot be determined without
knowing the firm's capital structure, which affects the potential for default
and bankruptcy. But capital structure cannot be optimized without knowing
the effect of leverage on debt value.
This article examines corporate debt values and optimal capital structure
in a unified analytical framework. It derives closed-form results relating the
value of long-term corporate debt and optimal capital structure to firm risk,
taxes, bankruptcy costs, bond covenants, and other parameters when firm
asset value follows a diffusion process with constant volatility.
Traditional capital structure theory, pioneered by Modigliani and Miller
(1958), holds that taxes are an important determinant of optimal capital
structure.lAs leverage increases, the tax advantage of debt eventually will be
offset by an increased cost of debt, reflecting the greater likelihood of finan-
cial d i ~ t r e s s .While
~ identifying some prime determinants of optimal capital
*Haas School of Business, University of California, Berkeley. The author thanks Ronald
Anderson, Fischer Black, Arnoud Boot, Michael Brennan, Philip Dybvig, Julian Franks, Robert
Gertner, William Perraudin, Matthew Spiegel, Suresh Sundaresan, Ivo Welch, and especially
Rob Heinkel and Klaus Toft for helpful comments. The referee and the editor, Ren6 Stulz,
provided many valuable suggestions.
'personal a s well as corporate taxes will affect the tax benefits to leverage (Miller (1977)).
Disagreement remains as to the precise value of net tax benefits.
he costs of financial distress include bankruptcy costs and agency problems associated with
risky debt. See, for example, Altman (1984), Asquith, Gertner, and Sharfstein (1991), Harris and
Raviv (1991), Jensen and Meckling (1976), Myers and Majluf (19841, Titman and Wessels (1988),
and Warner (1977).
1214 The Journal of Finance
structure, this theory has been less useful in practice because it provides
qualitative guidance onlya3
Brennan and Schwartz (1978) provide the first quantitative examination of
optimal leverage. They utilize numerical techniques to determine optimal
leverage when a firm's unlevered value follows a diffusion process with
constant volatility.4 Although an important beginning, the Brennan and
Schwartz analysis has three limitations.
First and most importantly, their numerical approach precludes general
closed-form solutions for the value of risky debt and optimal leverage. Numer-
ical examples suggest some possible comparative static results but cannot
claim generality.
Second, their analysis focuses on the special case in which bankruptcy is
triggered when the firm's asset value falls to the debt's principal value. This
provision approximates debt with a positive net-worth covenant. But it is by
no means the only-or even the typical-situation.5 We shall show that
alternative bankruptcy-triggering conditions, including endogenously deter-
mined ones, lead to very different debt values and optimal capital structure.
Finally, Brennan and Schwartz (1978) consider changes in financial struc-
ture that last only until the bonds mature. A maturity date is required for
their numerical algorithm; permanent capital structure changes are not
explicitly analyzed.6
This article considers two possible bankruptcy determinants. The first is
when bankruptcy is triggered (endogenously) by the inability of the firm to
raise sufficient equity capital to meet its current debt obligations. The second
is the Brennan and Schwartz case with a positive net-worth covenant. Debt
with such a covenant will be termed protected debt.
We can derive closed-form results by examining corporate securities that
depend on underlying firm value but are otherwise time independent. Yet
debt securities generally have a specified maturity date and therefore have
time-dependent cash flows and values. Time independence nonetheless can be
justified, perhaps as an approximation, in at least two ways. First, if debt has
sufficiently long maturity, the return of principal effectively has no value and
3 ~ a x t e r(19671, Kraus and Litzenberger (19731, and Scott (1976) offer general analyses
balancing tax advantages with the costs of financial distress, but their results have not provided
directly usable formulas to determine optimal capital structure. For a n alternative view on the
determinants of capital structure, see Myers (1984).
4 ~ i m (1978) also presents numerical examples of optimal capital structure, based on a
mean-variance model. His model is less parsimonious, as knowledge of the joint distribution of
market and firm returns is required.
' ~ i n i m u mnet-worth requirements are not uncommon in short-term debt contracts, but are
rare in long-term debt instruments (also see Smith and Warner (1979)). In a later and more
complex model, Brennan and Schwartz (1984) offer some examples with alternative bankruptcy
conditions.
'Brennan and Schwartz do look a t some examples when T becomes large. The relative
insensitivity of these examples to T, as T exceeds 25 years, suggests that our limiting closed-form
results for infinite maturity debt will be good approximations for debt with long but finite
maturity.
Debt Value, Bond Covenants, and Optimal Capital Structure 1215
can be i g n ~ r e d Very
. ~ long time horizons for fixed obligations are not new,
either in theory or in practice. The original Modigliani and Miller (1958)
argument assumes debt with infinite maturity. Merton (1974) and Black and
Cox (1976) look a t infinite maturity debt in an explicitly dynamic model.
Since 1752 the Bank of England has, on occasion, issued Consols, bonds
promising a fixed coupon with no final maturity date. And preferred stock
typically pays a fixed dividend without time limit.
An alternative time-independent environment is when, at each moment,
the debt matures but is rolled over at a fixed interest rate (or fixed premium
to a reference risk-free rate) unless terminated because of failure to meet a
minimum value, such as a positive net-worth covenant. As we discuss later,
this environment bears resemblance to some revolving credit agreements.
Time independence permits the derivation of closed-form solutions for risky
debt value, given capital structure. These results extend those of Merton
(1974) and Black and Cox (1976) to include taxes, bankruptcy costs, and
protective covenants (if any). They are then used to derive closed-form
solutions for optimal capital s t r u c t ~ r eThe
. ~ analysis addresses the following
questions:
How do yield spreads on corporate debt depend on leverage, firm risk,
taxes, payouts, protective covenants, and bankruptcy costs?
Do high-risk ("junk") bond values behave in qualitatively different ways
than investment-grade bond values?
What is the optimal amount of leverage, and how does this depend on
risk-free interest rates, firm risk, taxes, protective covenants, and
bankruptcy costs?
7 ~ o30-year
r debt, the final repayment of principal represents 1.5 percent of debt value when
the interest rate is 15 percent, and 5.7 percent of value when the interest rate is 10 percent.
Recently, a number of firms have issued 50-year debt, and one firm (Disney) has issued 100-year
debt.
' ~ e c e n t l I~ have become aware of important related work by Anderson and Sundaresan
(1992), Longstaff and Schwartz (19921, and Mella and Perraudin (1993). Anderson and
Sundaresan (1992) focus on risky debt in a binomial framework. Using numerical examples, they
examine the choice of debtors to discontinue coupon payments prior to bankruptcy and show that
this may explain the sizable default premiums found in bond prices (see Jones, Mason, and
Rosenfeld (19841, and Sang and Warga (1989)). They do not examine optimal capital structure.
Longstaff and Schwartz (1992) derive solutions for risky debt values with finite maturity and
with stochastic risk-free interest rates. Their key assumption is that bankruptcy is triggered
whenever firm value, V, falls to a n exogenously given level, K, (our V,), which is time
independent. This is a strong assumption for finite maturity debt, whose debt service payments
are time dependent. Equation (14) below shows that V, depends on the risk-free interest rate,
suggesting that an endogenously determined K should depend upon the (stochastic) interest
rate. Longstaff and Schwartz (1992) do not consider optimal capital structure.
Mella and Perraudin's approach more closely parallels this article, with endogenously deter-
mined bankruptcy levels. However, firm value is driven by a random product selling price whose
drift as well a s volatility must be specified, a s must the firm's cost structure. (See also Fries,
Miller, and Perraudin (1993)). Like Anderson and Sundaresan (1992), the article considers an
endogenous decision to continue service debt.
How does a positive net-worth covenant affect the potential for agency
problems between bondholders and stockholders?
When can debt renegotiation be expected prior to bankruptcy, and can
renegotiation achieve results that debt repurchase cannot?
The model follows Modigliani and Miller (1958), Merton (1974), and
Brennan and Schwartz (1978) in assuming (i) that the activities of the firm
are unchanged by financial structure, and (ii) that capital structure decisions,
once made, are not subsequently changed.
Much of the recent literature in corporate finance examines possible vari-
ants to assumption (i): see, for example, the survey by Harris and Raviv
(1991). A particularly important variant is the "asset substitution" problem,
where shareholders of highly leveraged firms may transfer value to them-
selves from bondholders by choosing riskier activities. If the appropriate
functional form were known, feedback from capital structure to volatility
could be captured in an extension of our model, at the likely cost of losing
closed-form result^.^ But a simpler model that ignores such potential feed-
back still serves some important purposes:
1) Taxes and bankruptcy costs will importantly condition optimal capital
structure even if asset substitution can occur; knowing these relation-
ships in a basic model will provide useful insights for more complex
situations.
2) The potential magnitude of the asset substitution problem can be identi-
fied by knowing how sensitive debt and equity values are to the risk of
the activities chosen.
3) Bond covenants may directly limit opportunities for firms to alter the
risk of their activities. In other cases, bond covenants may indirectly
limit asset substitution by reducing potential conflicts of interest be-
tween stockholders and bondholders. Section VII below shows that a
positive net worth requirement can eliminate the firm's incentive to
increase risk.
Our second major assumption is that the face value of debt, once issued,
remains static through time. This is not as unreasonable as it might appear.
In Section VIII, we show that additional debt issuance will hurt current
debtholders; it is typically proscribed by bond covenants. We further show
that marginal debt reductions via repurchases will hurt current stockholders.
These considerations may preclude continuous changes in the outstanding
amount of debt, even if refinancing costs are zero.
e el lo and Parsons (19921, using a numerical approach similar to Brennan and Schwartz
(1978) but including operating decisions of a (mining) firm, contrast decisions that maximize
equity value with those that maximize the total value of the firm. They associate the difference
in resulting values with agency costs and present a n example showing the effect of these costs on
optimal leverage. Mauer and Triantis (1993) also use the Brennan and Schwartz (1978) approach
to examine the interaction of investment decisions and corporate financing policies.
However, large (discontinuous) debt repurchases via tender offers may

under certain circumstances benefit both stock and bondholders, if refinanc-
ing costs are not excessive. A dynamic model of capital structure capturing
these possibilities is desirable but considerably more difficult. First steps in
this direction have been made in important work by Kane, Marcus, and
McDonald (1984) and Fischer, Heinkel, and Zechner (1989). Their analyses
pose several difficulties, which we avoid by adopting the static assumption
shared with earlier authors.1°
The structure of the article is as follows. Section I develops a simple
dynamic model of a levered firm, and derives values for time-independent
securities. Sections I1 and I11 consider debt value and optimal leverage when
bankruptcy is determined endogenously. Sections IV and V consider debt
value and optimal leverage when bankruptcy is triggered by a positive
net-worth covenant. Section VI considers some alternative assumptions about
tax deductibility, cash payouts by the firm, and the absolute priority of
payments in bankruptcy. Section VII addresses agency problems and asset
substitution, while Section VIII considers aspects of debt repurchase and
renegotiation. Section IX concludes.
I. A Model of Time-IndependentSecurity Values

Consider a firm whose activities have value V which follows a diffusion
process with constant volatility of rate of return:
where W is a standard Brownian motion. We shall refer to V as the "asset

value" of the firm.'' The stochastic process of V is assumed to be unaffected
by the financial structure of the firm. Thus any net cash outflows associated
with the choice of leverage (e.g., coupons after tax benefits) must be financed
by selling additional equity.12
Following Modigliani and Miller (1957), Merton (19741, Black and Cox
(1976), and Brennan and Schwartz (1978), we assume that a riskless asset
10
In Fischer, Heinkel, and Zechner (1989), the value of a n unlevered firm (their A) cannot be
exogenous, since it depends on the optimally levered firm value less costs of readjustment (see
their p. 25). Since closed-form solutions are not available for the restructuring boundaries, they
do not offer closed-form equations for risky debt value and optimal capital structure.
11
We leave unanswered the delicate question of whether V, which could be associated with the
value of a n unlevered firm, is a traded asset. An alternative approach is to note that if equity, E,
is a traded security, its process could be used to define a process, V, through equation (13) below,
using Ito's Lemma. Our assumption that V has constant volatility will restrict the permissible
process of E.
his is consistent with bond covenants that restrict firms from selling assets. Brennan and
Schwartz (1978) also make this assumption, although Merton (1974) does not. In Section VI.B,
we consider how our results are affected by relaxing this assumption.
exists that pays a constant rate of interest r. This permits us to focus on the
risk structure of interest rates directly.13
Now consider any claim on the firm that continuously pays a nonnegative
coupon, C, per instant of time when the firm is solvent. Denote the value of
such a claim by F(V, t). When the firm finances the net cost of the coupon by
issuing additional equity, it is well known (e.g., Black and Cox (1976)) that
any such asset's value must satisfy the partial differential equation
with boundary conditions determined by payments at maturity, and by

payments in bankruptcy should this happen prior to maturity.14 In general,
there exist no closed-form solutions to equation (2) for arbitrary boundary
conditions. Hence Brennan and Schwartz (1978) resort to computer analysis
of some examples. However, when securities have no explicit time depen-
dence, the term Ft(V, t) = 0 and equation (2) becomes an ordinary differen-
tial equation with F(V) satisfying
Equation (3) has the general solution
where
and the constants Ao, A,, and A, are determined by boundary conditions.
Any time-independent claim with an equity-financed constant payout C 2 0
must have this functional form. We turn now to examining specific securities.
Debt promises a perpetual coupon payment, C, whose level remains con-
stant unless the firm declares bankruptcy. The value of debt can be expressed
as D(V; C). For simplicity, however, we will suppress the coupon as an
argument and simply write debt value as D(V). Let V, denote the level of
asset value at which bankruptcy is declared. (Note that we again suppress
the argument C.) If bankruptcy occurs, a fraction 0 s a I 1 of value will be
l 3~ x t e n s i o n sof numerical bond valuation to include interest rate risk are provided in Brennan
and Schwartz (1980) and Kim, Ramaswamy, and Sundaresan (1993). They find that the yield
spreads between corporate and Treasury bonds are quite insensitive to interest rate uncertainty.
14
More generally, if net payouts by the firm not financed by further equity issuance are
denoted P(V, t), and C(V, t ) represents the payout flow to security F, then
Note that u2(V, t ) could be of the form u2[C(V, t),V, tl, reflecting possible asset substitution.
Equation (2) requires that V, or a n asset perfectly correlated (locally) with V, such as equity,
be traded. See also footnote 11.
lost to bankruptcy costs, leaving debtholders with value (1 - a)VB and

stockholders with nothing.15
Later we show how the bankruptcy value, VB,is determined, given alterna-
tive debt covenants. For the moment regard it as fixed. Since the value of
debt is of the form in equation (41, we must determine the constants A,, A,,
and A,. Boundary conditions are:
Condition (6ii) holds because bankruptcy becomes irrelevant as V becomes

large, and the value of debt approaches the value of the capitalized coupon
(and therefore the value of risk-free debt).
From equation (4), it is immediately apparent using equation (6ii) that
A, = 0. Because VpX -t 0 as V + m, this with equation (6ii) implies that
A, = C/r. Finally, A, = [(I - a)VB - C/r]V$, using equation (6i). Thus
where p, - +
Equation (7) can also be written as D(V) = [ l - pB](C/r) pB[(l - a)Vi],
( V / V ~ ) - h~as the interpretation of the present value of $1
contingent on future bankruptcy (i.e., V falling to VB).16
Equation (7) represents a straightforward extension of Black and Cox
(1976) to include bankruptcy costs.17 But we shall see later that taxes affect
the value, VB, when bankruptcy is determined endogenously. Both taxes and
bankruptcy costs are important determinants of debt value in this case.
Debt issuance affects the total value of the firm in two ways. First, it
reduces firm value because of possible bankruptcy costs. Second, it increases
firm value due to the tax deductibility of the interest payments, C. The value
of both these effects will depend upon the level of firm value, V, and are time
independent. Therefore they can be valued as if they were time-independent
securities.
First, consider a security that pays no coupon, but has value equal to the
bankruptcy costs aVB at V = VB. This security has current value, denoted
15we focus on bankruptcy costs that are proportional to asset value when bankruptcy is
declared. Alternatives such as constant bankruptcy costs could readily be explored within the
framework developed. Deviations from absolute priority (in which bondholders do not receive all
remaining value) can also be incorporated in the boundary conditions; we do so in Section V1.C.
Franks and Torous (1989) and Eberhart, Moore, and Roenfeldt (1990) document deviations from
the absolute priority rule.
l6 ore exactly,
where f(t;V, V,) is the density of the first passage time from V to V,, when the process for V
has drift equalto the risk-free interest rate, r.
17
Merton (1974) derives a different formula for the case where cu = 0. This is because he
assumes the firm liquidates assets to pay coupons.
BC(V), that reflects the market value of a claim to aVB should bankruptcy
occur. Because its returns are time independent, it too must satisfy equation
(4) with boundary conditions
In this case equation (4) has solution
BC is a decreasing, strictly convex function of V. Again, note the reinter-

pretation of equation (9) as BC = pB[ffVB]: the current value of bankruptcy
costs is their magnitude if bankruptcy occurs, times the present value of $1
conditional on future bankruptcy. Subsequent expressions will have similar
interpretations.
Now consider the value of tax benefits associated with debt financing.
These benefits resemble a security that pays a constant coupon equal to the
tax-sheltering value of interest payments (TC) as long as the firm is solvent
and pays nothing in bankruptcy. This security's value, TB(V), equals the
value of the tax benefit of debt. It too is time independent and therefore must
satisfy equation (4) with boundary conditions
Equation (10i) reflects the loss of the tax benefits if the firm declares
bankruptcy. Equation (10ii) reflects the fact that, as bankruptcy becomes
increasingly unlikely in the relevant future, the value of tax benefits ap-
proaches the capitalized value of the tax benefit flow, TC. Using equation (4)
and the boundary conditions above gives
Tax benefits are an increasing, strictly concave function of V.

Note that the value of tax benefits, equation ( l l ) , presumes that the firm
always benefits fully (in amount TC) from the tax deductibility of coupon
payments when it is solvent. But under U.S. tax codes, to benefit fully the
firm must have earnings before interest and taxes (EBIT) that is at least as
large as the coupon payment, C.18An alternative approach, in which EBIT is
related to asset value, V, and tax benefits may be lost when the firm is
solvent (but close to bankruptcy), is considered in Section V1.A.
he losses associated with interest payments exceeding EBIT may be carried forward, but
lose time value, and may lose all value if the firm goes bankrupt. (Reorganizations under
Chapter 11of the Bankruptcy Code may carry forward some tax benefits. This could be modeled
by a boundary condition, equation (lOi), with a positive value.)
The total value of the firm, v(V), reflects three terms: the firm's asset
value, plus the value of the tax deduction of coupon payments, less the value
of bankruptcy costs:
v(V) = V + TB(V) - BC(V)
Note that v is strictly concave in asset value, V, when C > 0 and either a > 0
or r > 0. Note also that if a > 0 and r > 0, then v(V) < V as V + VB, and
v(V) > V as V -t 03. This coupled with concavity implies that u is (pro-
portionately) more volatile than V at low values of V and is less volatile at
high values.
The value of equity is the total value of the firm less the value of debt:
E(V) = v(V) - D(V)
= V - (1 - r)C/r + [ ( I - r)C/r
v B ] [ v / v B l x . (13)
-
We see from equation (14) below that when VB is endogenously deter-

mined, [(I - r)C/r - VB] > 0, implying that E(V) is a convex function of V.
This reflects the "option-like" nature of equity, even when debt has an
infinite horizon. When VB is determined by a positive net worth requirement,
however, we show in Section V that equity may be a concave function of V.
This has important ramifications for agency problems associated with asset
substitution, which are examined in Section VII. Finally, Ito's Lemma can be
used to show that the volatility of equity's rate of return declines as V (and
therefore E ) rises. Stock option pricing models would need to reflect this
nonconstant volatility, as well as the possibility that E reaches zero with
positive probability.
Equations (7) and (13) indicate the importance of VB in determining the
values of debt and equity. In the following sections, we consider alternative
bankruptcy-triggering scenarios.
11. Debt with No Protective Covenants: The Endogenous

Bankruptcy Case
If the firm is not otherwise constrained by covenants, bankruptcy will occur
only when the firm cannot meet the required (instantaneous) coupon pay-
ment by issuing additional equity: that is, when equity value falls to zero.lg
However, any level of asset value, V,, that triggers bankruptcy will imply
19
In continuous time, the coupon ( C d t ) paid over the infinitesimal interval, d t , is itself
infinitesimal. Therefore the value of equity simply needs to be positive to avoid bankruptcy over
the next instant. In discrete time, where the time between periods, 6 t , is of a fixed size, the value
of equity a t each period must exceed the coupon ( C a t ) to be paid that period.
It is sometimes assumed that bankruptcy is triggered by a cashflow shortage. This can be
criticized, because, if equity value remains, a firm will always be motivated and able to issue
additional equity to cover the shortage, rather than declare bankruptcy. Positive equity value
rather than positive cashflow seems to be the essential element when bankruptcy is endoge-
nously determined.
that the value of equity is zero at that asset value, given the absolute priority
rule.
When VB can be chosen by the firm (rather than imposed by a covenant
such as positive net-worth requirement), it can be seen from equation (12)
that total firm value, u , will be maximized by setting VB as low as possible.
Limited liability of equity, however, prevents VB from being arbitrarily small:
E(V) must be nonnegative for all values of V 2 VB.From equation (131, E(V)
is strictly convex in V when VB < (1 - r)C/r. Thus the lowest possible value
for VB consistent with positive equity value for all V > VB is such that
dE/dVlv= v, = 0: a "sinooth-pasting" or "low contact" condition at V = VB.
(This choice of bankruptcy level can also be shown to maximize the value of
equity at any level of V: dE/dVB = 020). Differentiating equation (13) with
respect to V, setting this expression equal to zero with V = VB, and solving
for VB gives
where the second line uses equation (5). Since VB < (1 - r)C/r, equity is
indeed convex in V.
Observe that the asset value, VB, a t which bankruptcy occurs
a) is proportional to the coupon, C;
b) is independent of the current asset value, V;
c) decreases as the corporate tax rate, r , increases;
d) is independent of bankruptcy costs, a ;
e) decreases as the risk-free interest rate, r , rises; and
f) decreases with increases in the riskiness of the firm, a 2 .
The results above also describe the behavior of total firm value at bank-
ruptcy, vB = u[VB]= (1 - cu)VB, except that uB falls as bankruptcy cost, a ,
increases. The fact that asset value, V, does not affect uB means that the
bankruptcy level of total firm market value can be estimated from the
coupon, C, (plus parameters r , a,, a , and r), without needing to know the
firm's current asset value."
Substituting equation (14) into equations (7), (12), and (13) gives
'Osee also Merton (1973; footnote 60). The equivalence of the two conditions suggests that the
endogenously set VB is incentive compatible in the following sense. Ex ante (before debt
issuance), stockholders will wish to maximize firm value subject to the limited liability of equity.
The ex ante optimal VB achieves this by satisfying the smooth-pasting condition. Ex post, equity
holders will have no incentive to declare bankruptcy a t a different V, since VB also satisfies the
ex post optimal condition for maximizing equity value.
21
Knowledge of the market value of equity, E, and debt, D, in addition to C , combined with
equations (71, (131, and (14), permits calculation of a unique V and a given r , 7 , and a .
Alternatively, a and a can be recovered, given r , 7 , and V .
where
The interest rate paid by risky debt, R(C/V), can be derived directly from
dividing C by D(V), giving
where
The interest rate depends positively on the ratio of the coupon, C, to firm
asset value, V. Note K(C/V) has the interpretation of a risk-adjustment
factor (multiplying the risk-free rate) that the firm must pay to compensate
bondholders for the risks assumed. The yield spread is R(C/V) - r =
r(C/VIXk/[1 - (C/V)Xk].
The values above are derived for an arbitrary level of the coupon, C.
Section I11 examines the optimal choice of coupon (and leverage) for unpro-
tected debt. But first, we examine the behavior of unprotected debt values
and yield premiums for an arbitrary coupon level.
A. The Comparatiue Statics of Debt Value (D(V))

Equation (15) extends Black and Cox's (1976) results to include the effects
of taxes and bankruptcy on debt value. Row 1 of Table I summarizes the
comparative statics of debt value. Not surprisingly, larger bankruptcy costs
decrease the value of debt. Less obvious is that an increase in the corporate
tax rate will raise debt value, through lowering the bankruptcy level, VB.22
More surprising still are the results when taxes or bankruptcy costs are
positive and firm asset value, V, nears the bankruptcy level, VB. Table I
indicates that the effects of increases in the coupon, firm riskiness, and the
risk-free rate become reversed from what is expected. An increase in coupon
can lower debt value. An increase in firm risk can raise debt value, as can an
increase in the risk-free rate. Thus the behavior of "junk" bonds (or "fallen
angels") differs significantly from the behavior of investment-grade bonds
when bankruptcy costs and/or taxes are positiue.23
To understand these results, first consider the presence of positive
bankruptcy costs. If V is close to VB, the value of debt will be very sensitive
to such costs. Lowering VB will raise the value of debt since bankruptcy costs
22
These comparative static results presume that other parameters (including V ) remain a t
their current level, the usual ceteris paribus assumption. Note, however, that a change in the
corporate tax rate might affect V as well.
23
The ratios of V/V, (or C/V) a t which the various behaviors are reversed are not identical.
Of course, these ratios may not correspond to Wall Street's definition of " j u n k bonds.
Table I
Comparative Statics of Financial Variables: Unprotected Debt
This table describes properties of the equations describing debt value, D, the interest, R , paid on debt, the yield spread of the debt over the risk-free
rate ( R - r ) , the total firm value, u , and the value of equity, E, when debt is not protected by a positive net-worth covenant. V is the firm's asset
value, VB is the endogeneously determined value a t which bankruptcy is declared, C is the coupon paid on debt, u 2 is the variance of the asset
return, r is the risk-free interest rate, a is the fraction of asset value lost if bankruptcy occurs, and r is the corporate tax rate.
Limit As Sign of Change in Instrument for a n Increase in: 3

Variable Homogeneity Shape V-+m V + VB C u2 r a r V
D Degree 1 Concave C/r C(l - a)(l - r ) > 0; < 0; < 0; <O >O >O
3
in V, C in V ,C ( r + 0.5u2) < 0* a s V + VB > 0* a s V + VB > 0* as V + VB 9
n
Pu
R Degree 0 Convex r +
( r 0.5u2) >0 > 0; > 0; >O <O <O %
in V , C in V / C (1 - ax1 - r ) <O*asV+VB <O*asV+VB 3
J
R -r Degree 0 Convex 0 [0.5u2+ >0 > 0; <0 >O <O <O n
in V , C in V / C r ( a + r - ar)]/ < 0* a s V V, J
[ ( l- a x 1 - 711
+
2
v Degree 1 Concave V + rC/r C(1 - a ) ( l - T ) > 0; < 0; < 0; <O >O >O
in V , C in V ,C +
( r 0.5u2) < 0* a s V + VB > 0* a s V + VB > 0* as V + VB
E Degree 1 Convex V - (1 - r)C/r 0
in V , C in V , C
*Sign reversal a s V -, VB only if a and/or T > 0.

will be less imminent. From equation (14), higher asset volatility, higher
risk-free interest rates, or lower coupon, C, will all serve to lower VB. For
values of V close to VB, this positive effect on D(V) will dominate. Even if
there are no direct bankruptcy costs, the event of bankruptcy causes the
value of the tax shield to be lost when r > 0, and the previous conclusions
continue to hold.
The fact that D(V) is eventually decreasing as the coupon rises implies
that debt value reaches a maximum, Dmax(V),for a finite coupon, Cmax(V).
We can naturally think of Dm,, as the debt capacity of the firm. Differentiat-
ing equation (15) with respect to C, setting the resulting equation equal to
zero and solving for C gives
Substituting this into equation (15) and simplifying gives
The debt capacity of a firm is proportional to asset value, V, and falls with
increases in firm risk, a', and bankruptcy costs, a. Debt capacity rises with
increases in the corporate tax rate, r , and the risk-free rate, r.
Figures 1and 2 show the relationship between debt value and the coupon
Figure 1. Debt value as a function of the coupon, when debt is unprotected. The lines
plot the value of unprotected debt, D, a t varying coupon levels C , for three levels of asset
volatility, a : 15 percent (open square), 20 percent (filled diamond), and 25 percent (solid line).
It is assumed that the risk-free interest rate r = 6.0%, bankruptcy costs are 50 percent
( a = 0.51, and the corporate tax rate is 35 percent (T = 0.35).
The Journal of Finance
C
Figure 2. Debt value as a function of the coupon, when debt is unprotected. The lines
plot the value of unprotected debt, D, a t varying coupon levels, C, for three levels of bankruptcy
costs: 0 (solid line), 50 percent (filled diamond), and 100 percent (open square) ( a = 0, 0.5, and
1, respectively). It is assumed that the risk-free interest rate r = 6.0 percent and the corporate
tax rate is 35 percent (T = 0.35).
for varying firm volatility and bankruptcy costs, when V = $100 and r = 6
percent. Our normalization implies that the coupon level (in dollars) also
represents the coupon rate as a percentage of asset value, V. Note that at
high coupon levels, the debt of riskier firms has higher value than that of less
risky firms. The peak of each curve indicates the maximum debt capacity,
D,,,, with corresponding leverage level. Figure 3 repeats Figure 1,but with
leverage, [ D/v], rather than coupon level on the x-axis. The reversals seen in
Figure 1do not appear in Figure 3. This is because leverage itself depends on
the value of debt.
B. Yield Spreads: The Risk Structure of Interest Rates

Rows 2 and 3 of Table I indicate the behavior of risky interest rates and
yield spreads. Increasing the coupon, C, always raises the yield spread. An
increase in bankruptcy costs, a , also raises the spread, although a rise in the
corporate tax rate will lower the spread because debt value will rise. Related
to our earlier discussion, we observe the surprising result that junk bond
yield spreads may actually decline when firm riskiness increases. Of course,
this holds only for junk bonds: the yield spread on investment-grade debt will
increase when firm risk rises. Also note that junk bond interest rates may
L
Figure 3. Debt value as a function of the leverage, when debt is unprotected. The
lines plot the value of unprotected debt, D, a t varying leverage ratios, L, for three levels of asset
volatility, u : 15 percent (open square), 20 percent (filled diamond), and 25 percent (solid line).
It is assumed that the risk-free interest rate r = 6.0%, bankruptcy costs are 50 percent
( a = 0.5), and the corporate tax rate is 35 percent (7 = 0.35).
actually fall when the risk-free rate increases. Figures 4 and 5 plot yield
spreads against coupon level and leverage, respectively, as asset value risk
changes.
Observe that R ( C / V ) + ( r + 0.5u2) as V + V' when a = T = 0. That is,
long-term risky debt will never have a yield exceeding the risk-free rate by
more than 0.5u2 if there are no bankruptcy costs or tax benefits to debt.24
Observing a yield spread greater than this on corporate long-term debt
implies the presence of bankruptcy costs, taxes, or both.25
C. The Comparative Statics of Firm Value (v(V)) and Equity Value (E(V))
Row 4 of Table I indicates the comparative statics of total firm value. Again
observe the perverse behavior of total firm value for firms with junk debt. In
the presence of bankruptcy costs and/or corporate taxes, total firm value
'*A firm whose asset value has a n annual standard deviation of 20 percent, for example, would
have debt whose yield spread never exceeds two percent. I t has been argued that the tax
advantage to debt may be nil (Miller (1977)). For arguments that bankruptcy costs may be small,
see Warner (1977) (who focuses on direct costs only) and Haugen and Senbet (1988).
25 when the firm has several debt issues, junior debt could have higher rates. But the weighted
average cost of debt will be limited to r + 0.5u2 in this case.
Figure 4. Yield spreads on unprotected debt as a function of the coupon. The lines
plot the yield spread, YS (in basis points/year), the amount the firm's debt yield exceeds the
risk-free rate, a s a function of the coupon, C, for varying levels of asset volatility, u :15 percent
(open square), 20 percent (filled diamond), and 25 percent (solid line). I t is assumed that the
risk-free interest rate r = 6.0 percent, bankruptcy costs are 50 percent ( a = 0.5), and the
corporate tax rate is 35 percent (T = 0.35).
may rise as firm riskiness increases. Rising risk-free rates may also lead total
firm value to increase. The values of firms with investment-grade debt will
not exhibit such behavior. Figure 6 and Figure 7 illustrate total firm value, v,
as a function of the coupon level C and the leverage D / u , respectively.
Optimal leverage is the ratio at which each curve reaches its peak.
Row 5 of Table I indicates the behavior of equity value. Unlike debt, there
are no reversals of comparative static results when V is close to V,. The fact
that bankruptcy costs do not affect equity value is perhaps surprising, but it
reflects the fact that, given the coupon, C, debtholders bear all bankruptcy
costs. In Section I11 we show that the optimal coupon and debt-equity ratio
do depend upon a , and that initial equity holders ultimately are hurt by
greater bankruptcy costs.
111; Optimal Leverage with Unprotected Debt

Consider now the coupon rate, C, which maximizes the total value, v, of the
firm,given current asset value, V. Differentiating equation (16) with respect
to C, setting the derivative equal to zero and solving for the optimal coupon,
Figure 5. Yield spreads on unprotected debt as a function of the leverage. The lines
plot the yield spread, YS (in basis points/year), the amount the firm's debt yield exceeds the
risk-free rate, as a function of the leverage, L, for varying levels of asset volatility, a:15 percent
(open square), 20 percent (filled diamond), and 25 percent (solid line). It is assumed that the
risk-free interest rate r = 6.0 percent, bankruptcy costs are 50 percent ( a= 0.51, and the
corporate tax rate is 35 percent (T = 0.35).
C*, as a function of asset value, V, gives:
Note that h > k, implying C*(V) < C,,,(V). Substituting C*(V) into
equations (151, (16), (18), and (14) gives
Table I1 indicates the comparative statics of these variables plus optimal

leverage L* = D*/v* and equity E* = v* - D*. While most results are con-
sistent with what is expected, a few merit comment.
The optimal coupon C* is a U-shaped function of firm riskiness, as illus-
trated in Figure 8. Firms with little business risk, or very large risk, will
optimally commit to pay sizable coupons. Firms with intermediate levels of
risk will promise smaller coupons. However, the optimal leverage ratios of
riskier firms will always be less than those of less risky firms, as can be seen
C
Figure 6. Total firm value as function of the coupon, when debt is unprotected. The
lines plot total firm value, v, a t varying coupon levels, C , for three levels of asset volatility, (r: 15
percent (open square), 20 percent ( f l l e d diamond), and 25 percent (solid line). It is assumed
that the risk-free interest rate r = 6.0 percent, bankruptcy costs are 50 percent ( a = 0.5), and
the corporate tax rate is 35 percent ( 7 = 0.35).
by observing the maximal firm values in Figure 7. The potential gains in

moving from no leverage to optimal leverage (where v = v*)are considerable.
For reasonable parameter levels, optimizing financial structure can increase
firm value by as much as 25 to 40 percent over a firm with no leverage.
Our results confirm Brennan and Schwartz's (1978)observation that opti-
mal leverage is less than 100 percent even when bankruptcy costs are zero.
Too high leverage risks bankruptcy-and while there are no bankruptcy
costs, the tax deductibility of coupon payments is lost.
Leverage of about 75 to 95 percent is optimal for firms with low-to-
moderate levels of asset value risk and moderate bankruptcy costs.26 Even
firms with high risks and high bankruptcy costs should have leverage on the
order of 50 to 60 percent, when the effective tax rate is 35 percent. Optimal
2 6 ~ist of interest that many of the leveraged buyouts of the 1980s created capital structures
that had 95 percent leverage or more. And targets were often firms with relatively stable value
(low (r 2). Our analysis indicates these firms will reap maximal benefits from increased leverage.
Subsequent leverage reduction by many of these firms could in part be explained by the
substantial fall in interest rates, which reduces the optimal leverage ratio.
L
Figure 7. Total firm value as function of the leverage, when debt is unprotected. The
lines plot total firm value, v, a t varying levels of leverage L, for three levels of asset volatility, cr:
15 percent (open square), 20 percent (filled diamond), and 25 percent (solid line). It is assumed
that the risk-free interest rate r = 6.0 percent, bankruptcy costs are 50 percent ( a = 0.51, and
the corporate tax rate is 35 percent (T = 0.35).
leverage ratios drop by 5 to 25 percent when the effective tax rate is 15

percent, with the more pronounced falls at high volatility levels.27Variations
of our assumptions that lead to lower optimal leverage ratios are discussed in
Section VI.
The behavior of the yield spread at the optimal leverage ratio exhibits one
surprise. Increased bankruptcy costs might be thought to increase interest
rates. Indeed they do-but only if the coupon is fixed. As bankruptcy costs
rise, the optimal coupon C* falls. The probability of bankruptcy is then less
and the yield spread decreases. Figure 9 illustrates yield spreads at the
optimal leverage as a function of bankruptcy costs, a , and asset risk, a.
Higher risk-free interest rates might also be expected to reduce the optimal
amount of borrowing, but they do not: the added tax shield when interest
rates are high more than offsets the greater costs of borrowing. This could be
destabilizing, since supply would normally be expected to decrease as interest
27~ollowingMiller (1977), if the effective personal tax rate on stock returns (reflecting tax
deferment) were 20 percent, the tax rate on bond income were 40 percent, and the corporate tax
rate 35 percent, the effective tax advantage of debt is [ l - ( 1 - 0.35X1 - 0.20)/(1 - 0.40)l =
0.133, or slightly less than 15 percent.
Table 11
Comparative Statics of Financial Variables at the Optimal
Leverage Ratio: Unprotected Debt
This table describes the behavior of the coupon, C * ,that maximizes firm value, and the debt, D*,
leverage, L*, interest rate, R * , yield spread, R* - r, total firm value, u*, equity value, E * , and
bankruptcy value, VB*,a t the optimal coupon level, for unprotected debt. (V is the firm's asset
value, a is the variance of the asset return, r is the risk-free interest rate, a is the fraction of
asset value lost if bankruptcy occurs, and T is the corporate tax rate.
Sign of Change in Variable for a n Increase in:

Variable Shape a r a T
Linear in V < 0, a 2 small;

> 0, a large
Linear in V <0
Invariant to V <0
Invariant to V >0
Invariant to V >0
Linear in V <0
Linear in V >0
Linear in V <0
aNo effect if a = 0.
20Oo/
Sigma
Figure 8. The optimal coupon as a function of firm risk and bankruptcy costs. The
surface plots the optimal coupon, C * ,a t varying levels of firm risk, u ,and bankruptcy costs, a . I t
is assumed that the risk-free interest rate r = 6.0 percent and the corporate tax rate is 35
percent ( T = 0.35).
Figure 9. The yield spread as a function of firm risk and bankruptcy costs. The
surface plots the yield spread, YS, the difference between the yield on the firm's debt (at the
optimal coupon, C * ) and the risk-free interest rate, r, for varying levels of firm risk, a , and
bankruptcy costs, a. It is assumed that the risk-free interest rate r = 6.0 percent and the
corporate tax rate is 35 percent (7 = 0.35).
rates rise.28 Despite the greater borrowing, the yield spread a t the optimal
leverage actually falls slightly as the risk-free interest rate increases.
IV. Positive Net-Worth Covenants and the Value of Protected
Debt
Consider now the case in which debt remains outstanding without time
limit unless bankruptcy is triggered by the value of the firm's assets falling
beneath the principal value of debt, denoted P. We presume the principal
value coincides with the market value of the debt when it is issued, denoted
Do. Thus V, = Do."
28
Note that a n increase in r might well cause a decline in V. If so, it is possible that the
desired amount of borrowing (which is proportional to V) could decline even though optimal
leverage rises.
29
It must be verified that the VB = Do is consistent with the value of equity remaining positive
a t all levels V r VB. This requires that Do exceed the level in equation (14) satisfying the
smooth-pasting conditions. In fact, this is always the case a t the optimal protected debt coupon
level, and is satisfied a t all but extremely high initial coupon levels. We limit our examination of
protected debt to coupon levels for which the minimum net-worth requirement (rather than
equation (14)) is the determinant of V,.
Are there contractual arrangements in which this is a realistic description

of bankruptcy? One possibility would be long-term debt as examined previ-
ously, with a protective covenant stipulating that the asset value of the firm
always exceed the principal value of the debt: a positive net-worth require-
ment. Such covenants are not common in long-term bond contracts, however.
An alternative contractual arrangement approximating this case would be
a continuously renewable line of credit, in which the borrowing amount and
interest rate are fixed at in~eption.~'At each instant the debt will be
extended (''rolled over" a t a fixed interest rate) if and only if the firm has
sufficient asset value, V, to repay the loan's principal, P; otherwise bank-
ruptcy occurs.31 Thus the roll-over process proxies for a positive net-worth
requirement. With this latter interpretation, the differences between the
unprotected debt analyzed above and protected debt analyzed below may
capture many of the differences between long-term debt and (rolled over)
short-term financing.
From equation (7) with VB = DO,we can write the value of protected debt
as a function of the value of assets, V,, a t the time the debt is initiated:
Except when CY = 0, closed-form solutions for the function Do(Vo) satisfy-

ing equation (26) have not been found. However, we can easily solve this
equation numerically to determine the value, Do, of the debt, given initial
values, Vo and C (as before we suppress the argument C). Note that the
function Do(Vo) is homogeneous of degree one in Vo and C. Also note that
equation (26) gives the value of protected debt only at the initial asset value,
Vo. Equation (7) with VB = Do(Vo)gives protected debt value as a function of
asset value, V.
Figures 10 and 11 illustrate the behavior of protected debt value as the
coupon and leverage change, for V = Vo = 100. They should be compared
with Figures 1and 3. We observe that the surprising behavior of unprotected
"junk" debt does not hold for protected debt, even when the debt exhibits
considerable risk. Unlike the unprotected case, the value of debt increases
with the coupon a t all levels of C. And increased firm risk or a higher
risk-free interest rate always lowers debt value. This is because the
bankruptcy-triggering value, VB, is determined exogenously rather than en-
dogenously.
30 w e assume the firm will never choose to borrow less than the stipulated credit line amount.
The fact that most credit lines are tied to a floating rate is not important here, since the risk-free
rate is assumed to be constant. It is important that the interest rate paid by the firm be
independent of the firm's asset value V (providing V 2 VB)after the initial agreement is reached.
31
Many lines of credit have a "paydown" provision, requiring that the amount borrowed must
be reduced to zero a t least once per year. A firm will fail to meet this provision if its (market)
value of assets is less than the loan principal. Also note that Merton (1974) requires V 2 P a t
maturity to avoid default on a pure-discount bond: the firm must have positive net worth a t
maturity or bankruptcy occurs.
C
Figure 10. Debt value as a function of the coupon, when debt is protected by a
minimum net-worth requirement. The lines plot the value of protected debt, D, a t varying
coupon levels, C , for three levels of asset volatility, a: 15 percent (solid line), 20 percent (filled
diamond), and 25 percent (open square). It is assumed that the risk-free interest rate r = 6.0
percent, bankruptcy costs are 50 percent ( a = 0.51, and the corporate tax rate is 35 percent
(7 = 0.35).
When there are no bankruptcy costs ( a = 01,

a) Protected debt is riskless and pays the risk-free rate, r.
b) For any C, the value of the tax shield with protected debt is less than
the tax shield with unprotected debt.
c) For any C, the bankruptcy-triggering value of assets, VB, for protected
debt exceeds the VB for unprotected debt.
Protected debt is riskless when a = 0 because the firm's asset value is
constantly monitored. Should asset value fall to the principal value, bank-
ruptcy is declared and, because there are no bankruptcy costs, debtholders
receive their full principal value. In this case, for a given coupon, C, the value
of protected debt always exceeds that of unprotected debt. Further, VB = P =
Do(Vo) = C/r. This exceeds the bankruptcy-triggering value, equation (141, of
assets for unprotected debt, and implies smaller tax benefits from equation
(11).
When bankruptcy costs are positive ( a > 01, the results change markedly.
For a given coupon, C, protected debt may have a lesser value than unpro-
tected debt (and therefore may pay a higher interest rate). This follows
because bankruptcy will occur more frequently when debt is protected,
because VB is higher in the protected case, and bankruptcy costs will be
Figure 11. Debt value as a function of the leverage, when debt is protected by a
minimum net-worth requirement. The lines plot the value of protected debt, D, a t varying
leverage ratios, L, for three levels of asset volatility, u :15 percent (solid line), 20 percent (filled
diamond), and 25 percent (open square). It is assumed that the risk-free interest rate r = 6.0
percent, bankruptcy costs are 50 percent ( a = 0.51, and the corporate tax rate is 35 percent
(T = 0.35).
incurred when a > 0. Figure 12, when compared with Figure 5, shows yield
spreads to be substantially higher for protected debt when a = 0.5, except at
extreme leverage ratios.
V. Optimal Leverage with Protected Debt

We now use a simple search procedure to find the coupon, C*, that
maximizes the total value, u , of the firm with protected debt. Figure 13,
compared with Figure 7, illustrates that maximal firm value occurs at lower
leverage when debt is protected.
For a reasonable range of parameters, we find that
a) Optimal leverage for protected debt is substantially less than for unpro-
tected debt.
b) The interest rate paid a t the optimum leverage is less for protected debt,
even when bankruptcy costs are positive ( a > 0).
C) The maximum value of the firm (and therefore the benefit from lever-
age) is less when protected debt is used.
Figure 12. Yield spreads on protected debt as a function of the leverage. The lines plot
the yield spread, YS (in basis points/year), the amount the firm's debt yield exceeds the risk-free
rate, a s a function of the leverage, L, for varying levels of asset volatility, a : 15 percent (solid
line), 20 percent (filled diamond), and 25 percent (open square). It is assumed that the risk-free
interest rate r = 6.0 percent, bankruptcy costs are 50 percent ( a = 0.5), and the corporate tax
rate is 35% (T = 0.35).
d) The maximal benefits of unprotected over protected debt increase as:

Corporate taxes increase
Interest rates are higher
Bankruptcy costs are lower
A closer examination of numerical results reveals that the optimal bank-
ruptcy level Vz is the same for both protected and unprotected debt, when
bankruptcy costs are zero. We know, however, the closed-form solution for
unprotected debt's optimal bankruptcy level, VB, from equation (25). Since
Do = VB,this in turn suggests a closed-form solution for the optimal value of
protected debt and related values when bankruptcy costs are zero and
v = v,,:
Because protected debt is risk free when a = 0, it also follows that

L
Figure 13. Total firm value as function of the leverage, when debt is protected. T h e
lines plot total firm value, u , at varying levels o f leverage, L, for three levels of asset volatility,
u : 15 percent (solid line), 20 percent (filled diamond), and 25 percent (open square). I t i s
assumed that the risk-free interest rate r = 6.0 percent, bankruptcy costs are 50 percent
( a = 0.5), and t h e corporate tax rate is 35 percent (T = 0.35).
Recall that equations (27) to (29) hold only for the protected debt case with
no bankruptcy costs. We have not been able to find closed-form solutions
when a > 0. Equation (28) implies that [(I - 7)(C*/r) - Vg*] = -7C*/r <
0, when a = 0. From equation (13), this implies that equity is a strictly
concave function of V. By continuity, equity will be concave when a is close to
zero. And in the numerical example considered in Section VI, equity is
strictly concave in V for all a.
The observed comparative statics of optimal protected debt value (and
related values) are given in Table 111. There are some important differences
with the comparative statics of optimal unprotected debt value. The debt
yield and the yield spread at the optimum rise rather than fall as bankruptcy
costs rise. The yield spread also increases as the risk-free interest rate rises,
although the magnitude is small. The optimal leverage ratio, (D*/u*), de-
clines as the corporate tax rate increases, when bankruptcy costs are low.
Optimal debt, D*, increases with 7, but (unlike the unprotected debt case)
increases less rapidly than u *.
VI. Discussion and Variations: Debt Value and Capital
Structure
Our analysis has determined optimal leverage ratios and associated yield
spreads in a variety of environments, for both long-term unprotected debt
Table I11
Comparative Statics of Financial Variables at the Optimal
Leverage Ratio: Protected Debt
This table describes the behavior of the coupon, C*, that maximizes firm value, and the debt, D*,
leverage, L*, interest rate, R*, yield spread, R* - r, total firm value, u * , equity value, E*, and
bankruptcy value, V ,: a t the optimal coupon level, for debt protected by a positive net-worth
covenant. V is the firm's asset value, u 2 is the variance of asset returns, r is the risk-free
interest rate, cu is the fraction of asset value lost if bankruptcy occurs, and r is the corporate tax
rate.
Sign of Change in Variable for a n Increase in:

Variable Shape (T r (Y r
C* Linear in V < ob >0 <0 > 0"
D* Linear in V <0 >0 <0 > 0"
L* Invariant to V <0 >0 <0 < 0, cu smallb;
> 0, large
(Y
R* Invariant to V > 0" >0 > ob > 0"

R* - r Invariant to V > 0" > ob > ob > Oa
u* Linear in V <0 >0 <0 >0
E* Linear in V >0 <0 >0 > 0, a smallb;
< 0, a large
vi? Linear in V <0 >0 <0 > Oa
"No effect if (Y = 0.
b ~ e p r e s e n tds ifferent behavior from unprotected debt.
and protected (or continuously rolled-over) debt. It is of interest to compare

these results with typical leverage ratios and yield spreads in the United
States. Leverage in companies with highly rated debt is generally less than
40 percent. Yields of investment-grade corporate bonds have exceeded Trea-
sury bond yields by a minimum of 15 basis points (bps), and a maximum of
215 bps from 1926 to 1986. The average yield spread over this period was 77
bps.32 These spreads reflect finite-maturity debt and also reflect the fact that
corporate debt typically is callable. Call provisions may add about 25 bps to
the annual cost of corporate debt.33Subtracting 25 bps from the average yield
spread of 77 bps to eliminate the impact of call provisions gives an adjusted
historical yield spread of about 52 bps.
We examine a base case where the volatility of the firm's assets is 20
percent, the corporate tax rate is 35 percent, the risk-free rate is 6 percent,
and bankruptcy costs are 50 percent. In this case, optimal leverage with
unprotected debt is 75 percent and the yield spread is 75 bps. The optimally
3 2 ~reported
s by Kim, Ramaswamy, and Sundaresan (1993); see also S a n g and Warga (1989).
33
Kim, Ramaswamy, and Sundaresan (1993) estimate a call premium of 22 bps using a
numerical example.
levered firm's equity is volatile, with a 57 percent annual standard deviation.

Reducing the effective tax rate would reduce optimal leverage and the yield
spread. For example, with an effective tax rate of 15 percent, optimal lever-
age is 59 percent, and the yield spread is 35 bps. Equity volatility is lower,
but still substantial.
It is clear that, based on our assumptions thus far, the analysis of unpro-
tected debt suggests optimal leverage considerably in excess of current
practice. This could be construed as a criticism of current management rather
than the model. Managers may be loath to pay out "free cash flow" (see
Jensen (1986)); the wave of leveraged buyouts in the late 1980s suggests that
firm value may be raised by using greater leverage (see Kaplan (1989) and
Leland (1989)). However, the model's predicted yield spreads seem low given
the suggested high leverage.
Optimal leverage ratios and yield spreads for protected debt are more
consistent with historical ratios. In the base case, optimal leverage is 45
percent and the yield spread is 45 bps. Equity has a 34 percent annual
standard deviation, which is a bit higher than the historical average equity
risk of a single firm of about 30 percent.
We now consider how variations in the assumptions may affect the nature
of optimal leverage and yield spreads, in both the unprotected and protected
cases.
A. No Tax Shelter for Interest Payments When Value Falls

We have assumed that the deductibility of interest payments generates tax
savings at all values above the bankruptcy level. But as firm asset value
drops, it is quite possible that profits will be less than the coupon payout and
tax savings will not be fully realized (or will be substantially postponed). If
lesser tax benefits are available, the optimal leverage ratio declines.
In Appendix A, we extend the analysis to allow for no tax benefits when-
ever V < VT, where V, is an exogenously specified level of firm asset value.34
In the base case considered above, optimal leverage falls from 75 to 70
percent, and the yield spread at the optimal leverage rises from 75 to 87 bps,
when VT = 90, i.e., 90 percent of the current asset value.
A possible criticism of the above approach is that VT does not depend upon
the amount of debt the firm has issued. Consider an alternative scenario in
which higher profit is needed if higher coupon payments are to be fully
deductible. For example, assume that the rate of EBIT is related to asset
value as follows:
EBIT = (V - 60)/6. (30)
Thus gross profit before interest drops to zero when V falls to 60 and
represents one-sixth of asset value in excess of 60. Further, assume that
34 w e do not allow tax loss carryforwards in this analysis, since they would introduce a form of
time (and path) dependence. Thus, we may overstate the loss of tax shields: the "truth" perhaps
lies somewhere between the previous results and the results of this analysis.
coupon payments, C, can be deducted from profit (for tax purposes) only if
EBIT - C 2 0. (We ignore partial deductibility.) It then follows that
V, = 60 + 6C. (31)
In contrast with the previous scenario, greater debt now has a greater
likelihood of losing its tax benefits.35 Optimal leverage falls to 65 percent. The
yield spread falls to 61 bps, reflecting the lesser leverage. Equity volatility
remains high a t 51 percent. In the case of protected debt, the loss of tax
deductibility has a much smaller effect on optimal leverage and yield spread.
As expected, the loss of tax deductibility reduces the maximum value of the
firm in all cases.
B. Net Cash Payouts by the Firm

Following Brennan and Schwartz (1978) and others, we have focused on
the case where the firm has no net cash outflows resulting from payments to
bondholders or stockholders. We now change this assumption.36 Net cash
outflows may occur because dividends are paid to shareholders, and/or
because after-tax coupon expenses are being paid, without fully offsetting
equity financing. In this latter case, assets are being liquidated and the scale
of the firm's activities is clearly affected by the extent of debt financing.
To keep matters analytically tractable, we consider only cash outflows that
are proportional to firm asset value, where the proportion, d, may depend on
the coupon paid on debt. Equation (3) is replaced by
with general solution

F(V, t ) = A, +A , v ~+ ~A 2 V X , (33)
where
2
( r - d - 0.502) - [ ( r - d - 0.502) + 2 0 2 r ]1/ 2 o 2 (35)
Boundary conditions remain unchanged, implying A, = 0 as before since

Y I- 1. Therefore, solutions for all security values will have exactly the
same functional form as before, but with the exponent, X, given by equation
(34) rather than equation (5).
When d = 0.01, a one-percent payout on asset value (equivalent to approxi-
mately a 3 percent dividend on equity value, given the leverage of the base
n base case the optimal coupon falls to $5.08, implying VT is about 90, as above.
3 5 ~ the
3fi
The reader may wonder how equity value could be positive if the firm never pays dividends.
But our earlier assumption is not that firms never pay dividends-rather, there is no net cash
outflow: any cash dividends must be financed by issuing new equity. Like Black and Scholes
(1973), our model is a partial equilibrium one, and simply assumes the process for V.
case), the optimal leverage falls from 75 to 74 percent, and the yield spread
rises from 75 to 86 bps. But what if payouts also depend upon the coupon
being paid to debtholders? Consider the case where the proportional payout is
sufficient to cover the after-tax cost of debt when it is initially offered.37
Normalizing the initial value of V to 100 implies a payout d = (1 - ~)C/100,
or 0.0065C in the above example. Any dividend payout would be in addition
to this amount. For the base case above, we search over coupon levels, C, that
maximize v, subject to the constraint that d = 0.0065C + 0.01. This reduces
optimal leverage from 75 to 64 percent and increases the yield spread at the
optimum from 75 to 124 bps. The volatility of equity falls from 57 to 42
percent. In the case of protected debt, optimal leverage falls from 45 to 36
percent, the yield spread increases from 45 to 49 bps, and the volatility of
equity falls from 34 to 29 percent.
The maximum firm value drops from $128.4 to $122.0 with unprotected
debt, and from $113.3 to $110.0 with protected debt. This decrease in maxi-
mal value reflects the fact that bankruptcy is more likely with cash payouts,
with a resulting loss of tax benefits. Therefore, ex ante, shareholders (as well
a s bondholders) benefit from a covenant that prevents the firm from selling
assets to meet coupon payments. It is not surprising that many debt instru-
ments have such a preventive covenant. But if such a covenant cannot be
written (or cannot be enforced), shareholders will benefit (at bondholders'
expense) from the firm selling assets to pay coupons after the debt has been
issued. Recognizing this incentive, debtholders will pay less for debt and the
optimal leverage will fall as indicated above.
C. Absolute Priority Not Respected

We have assumed that debtholders receive all assets (after costs) if
bankruptcy occurs, and stockholders none: the "absolute priority" rule. Now
consider a simple alternative, where debtholders receive some fraction (1 - b)
of remaining assets, (1 - a)VB, while equity holders receive b(l - a)VB.38
This will affect debt value in two ways: debtholders will receive less value if
bankruptcy occurs, and bankruptcy will occur at a different level VB.
It can readily be shown that equation (7) will be replaced by
3 7 ~ o tthat
e as value falls, the proportional payout will no longer completely cover the after-tax
coupon-some equity financing becomes necessary. This may not be unreasonable, since bond-
holders will become increasingly sensitive to liquidation of assets a s firm value approaches the
bankruptcy level.
38
Franks and Torous (1989) estimate that deviations in favor of equity holders in Chapter 11
reorganizations are only 2.3 percent of the value of the reorganized firm. Eberhart, Moore, and
Roenfeldt (1990) estimate average equity deviations of 7.8 percent for their sample of Chapter 11
firms. We choose a 10 percent deviation as a n upper bound for this effect.
and equation (14) will be replaced by

V, = (1 - ~ ) C / [ r ( l- b + ab)][X/(l +X)]. (37)
For the base case with unprotected debt, deviations from absolute priority
of 10 percent (b = 0.1) cause the optimal leverage ratio to fall from 75 to 72
percent. The yield spread remains at 75 bps. The effect of deviations from
absolute priority are also minor when debt is protected: leverage remains
unchanged a t 45 percent, while the yield spread rises from 45 to 51 bps.
D. All of the Above

As a final exercise, consider the base case where, in addition, (i) dividends
equal 3 percent of equity value; (ii) after-tax coupon payments are not
initially financed with additional equity; (iii) coupon payments are not tax
deductible when V < V, = 60 + 6C; and (iv) there is a 10 percent deviation
from absolute priority (b = 0.1). When these conditions hold simultaneously,
the optimal leverage with unprotected debt falls to 47 percent and the yield
spread is 69 bps. The annual standard deviation of equity is 36 percent. For
protected debt, the optimal leverage falls to 32 percent, the yield spread is 52
bps and the standard deviation of equity is 29 percent. These last numbers
seem quite in line with historical yield spreads, leverage ratios, and equity
risks.
VII. Protected versus Unprotected Debt: Potential Agency
Problems
Our results suggest that optimal leverage ratios are lower when debt is
protected, and that the maximal gains to leverage are less. This raises a key
question: why should firms issue protected debt? The answer may lie with
agency problems created by debt, and asset substitution in particular. Jensen
and Meckling (1976) argue that equity holders would prefer to make the
firm's activities riskier, ceteris paribus, so as to increase equity value at the
expense of debt value. The expected cost to debtholders will be passed back to
equity holders in a rational expectations equilibrium, through lower prices on
newly issued debt.
Higher firm asset risk tends to benefit equity holders when equity is a
strictly convex function of firm asset value, V. And equity is strictly convex in
V when debt is unprotected. In Section V, however, it was shown that equity
may be a strictly concave function of V when debt has a positive net-worth
covenant. With protected debt, stockholders may not have an incentive to
increase firm risk at debtholders' expense.
To illustrate our point, consider the base case above with different levels of
asset volatility. If debt is unprotected, the optimal coupon is $6.50, firm value
is $128.4, and VB is $52.8. If debt is protected, the optimal coupon is $3.26,
firm value is $113.3, and VB is $50.6. Assume that, ex post, managers can
raise the risk of the firm's assets from the current annual standard deviation
Table IV
Values of Protected and Unprotected Debt and Equity for
Different Levels of Risk
This table gives the values of debt and equity for both unprotected and protected debt, when the
coupon (in each case) is chosen to maximize total firm value given a 20 percent asset volatility,
but asset volatility may be increased by management to higher levels.
Unprotected Debt Protected Debt

Asset Volatility Debt Value Equity Value Debt Value Equity Value
(%) ($1 ($) ($1 6)
of 20 percent-with no change in current asset value V. Will they be

motivated to engage in such "asset substitution"? Using equations (7) and
(13), and recalling from equation (14) that V, will change when debt is
unprotected but not when debt is protected, gives the results reported in
Table IV.
Debtholders are hurt by higher risk. In the case of unprotected debt, equity
value is enhanced by greater risk. Without covenants to prevent such a
change, it will always be in the interest of equityholders to increase risk. But
the opposite is true when debt is protected by a positive net worth covenant: in
this case, increasing risk lowers equity value as well as debt value.39
In the absence of protective covenants, investors recognize that sharehold-
ers will wish to raise asset volatility to the maximum (60 percent). They will
pay only $52.6 for the debt, and total firm value will be $111.7. If the firm
offers protected debt, investors recognize that shareholders will have no
interest in increasing firm risk, and total firm value will be $113.30. The firm
maximizes value by issuing protected rather than unprotected debt. (Even if
the firm initially chose to issue the amount of unprotected debt optimal for a
60 percent volatility, the total firm value would be $112.1-less than the
maximal value with protected debt.)
A reevaluation of the belief that asset substitution is always advantageous
for equity holders seems warranted. It is true for unprotected debt, but it is
false in the case examined here, when debt is protected by a positive
net-worth covenant. Both debt and equity are concave functions of asset value
39
The difference in behavior as a changes reflects the convexity (concavity) of equity in V
when debt is unprotected (protected). In addition, VB changes in the unprotected case as u
changes. This latter effect explains the curious result in Section II.A, that (for V close to VB) an
increase in firm risk can raise unprotected debt value. Thus, there is yet a further anomaly with
unprotected debt: a t the brink of bankruptcy (and only there), both debtholders and stockholders
wish to increase firm risk!
in this case.40 The greater incentive compatibility of protected debt may well
explain its prevalence (or the prevalence of short-term financing), despite the
fact that, ceteris paribus, it exploits the tax advantage of debt less effectively.
VIII. Restructuring via Debt Repurchase or Debt
Renegotiation: Some Preliminary Thoughts
The preceding analysis has assumed that the coupon, C, of the debt issue is
fixed through time. In the absence of transactions costs, restructuring by
continuous readjustments of C would seem to be desirable to maximize total
firm value as V fluctuates. However, we shall see that continuous readjust-
ments of C by debt repurchase (issuance) may be blocked by stockholders
p debt holder^).^' Debt renegotiation may be required to maximize total firm
value in these cases.
To prove this contention, first consider the firm selling a small amount of
additional debt, thereby increasing the current debt service by dC. This will
change the total value of debt by
But this total value change will be shared by current and new debtholders.
New debtholders will hold a fraction dC/C of the total debt value, leaving
current debtholders with value
(ignoring terms of O(dC2)). The change from D, the current debtholders'

value before the debt issuance, is
[(aD/dC) - (D/C)IdC < 0 for dC > 0, (40)

with the inequality resulting from the concavity of D in C and the fact that
D = 0 when C = 0. This "dilution" result holds for arbitrary initial V and C,
implying current debtholders will always resist increasing the total coupon
payments through additional debt issuance, even though such sales may
increase the value of equity and the firm. This resistance is frequently
codified in debt covenants that restrict additional debt issuance a t greater or
equal ~eniority.~'
40
This, of course, is consistent with the earlier result, equation (121, where total firm value, v ,
is concave in V. Concavity of u follows from the concavity of tax benefits and convexity of
bankruptcy costs (which are subtracted).
41
We consider debt issuance/repurchase for capital restructuring only. Any funds raised by
debt issuance will be used to retire equity, and vice-versa. Debt raised for new investment, or
retired by asset sales, are asset-changing decisions that are not considered here.
42
Our analysis assumes a single class of debt, implying that newly issued debt has the same
seniority in the event of bankruptcy. Even if the newly issued debt is junior to the current debt,
it will reduce the value of the current debt by raising VB.A full analysis of multiple classes of
debt securities is beyond the scope of the present article.
A related result on debt repurchase is perhaps more surprising: current

shareholders will always resist decreasing the coupon, C, by repurchasing
current debt (in small amounts) on the open market. To prove this, consider a
small decrease, dC < 0, and its effect on current shareholders. The total
value of equity will change by
The cost of retiring debt will equal the value of the fraction of debt retired, or
-D(dC/C). This cost must be financed with newly issued equity, whose
value is included in the change in total equity value above. Current share-
holders will therefore have equity value
implying a change in value to current shareholders of
using equation (41). For unprotected debt, it follows from equations (15) and
(17) that [(dE/dC) + (D/C)] > 0, implying that the change in equity value
to the original shareholders is negative when dC < 0. This result holds for
arbitrary initial V and C. Therefore, it will never be optimal for the firm's
shareholders to restructure by retiring unprotected debt via small open
market repurchases financed by new equity.43 In Appendix B, we show that
the result also holds for small repurchases of protected debt, when the coupon
is near its optimal level.
To illustrate the arguments above, consider the base case with unprotected
debt. With V = $100, the optimal coupon is $6.50 and V, = $52.80. Assume
this coupon level has been chosen by the firm. Now let V drop from $100 to
$90. Using equations (7) and (13) to compute the current values of debt and
equity gives: D = $91.79, E = $23.14, and v = $114.93. The firm's total value
can now be increased by reducing debt. The firm should cut its coupon by 10
percent to $5.85, since C* is proportional to V, which has fallen from $100 to
$90. This would increase the total firm value from $114.93 to $115.60.
But consider the firm repurchasing 10 percent of its debt to achieve the
new optimal leverage. The coupon is reduced from $6.50 to $5.85, and V, falls
by 10 percent to $47.52. The firm must pay (at least) $9.18 to retire 10 percent
of the bonds whose value is $91.79 prior to r e p ~ r c h a s e .It~ will
~ raise this
amount by issuing additional stock worth $9.18. Again using equations (7)
and (13) to compute debt and equity values with the lower coupon gives:
D = $86.65, and E = $28.95.
4 3 ~ h i debt
s repurchase result holds even if there are multiple classes of debt. Stockholders
might benefit from retiring debt via asset sales, but this would violate the assumption that the
asset value, V, is independent of the firm's capital structure.
44
Note that debt becomes more valuable per unit, as the coupon is reduced. We are assuming
here that the entire amount of repurchase can be effected a t the lowest (i.e., current) price. Any
higher price would magnify the losses to equity holders.
Debtholders are clearly better off, having received payments of $9.18 to

retire 10 percent of their holdings, plus retaining holdings worth $86.65. The
original equity holders have had their stock diluted: $9.18 of stock-the
amount raised to pay the debtholders-now belongs to new shareholders,
leaving the original shareholders with stock worth $28.95 - $9.18 = $19.77.
This is less than the $23.14 value of their shares prior to repurchase.
Although the total value of the firm would be increased by the restructuring,
equity holders cannot benefit from the repurchase, and will want to block such
refinancing. This problem results from an externality: when debt is reduced,
its "quality" is improved. Investors who continue to hold the firm's debt
receive a windfall gain from the debt repurchase.
The example shows that restructuring through debt repurchases or sales
may not be possible, although such changes could increase total firm value.
To capture such potential increases, changes in the terms of the debtholders'
securities (or "side payments") will be required. These types of restructurings
will be labeled debt renegotiation. In our example, replacing current debt
with convertible debt may be used to achieve the optimal coupon level. By
agreeing to exchange the current debt for debt with coupon $5.85 (worth
$86.65), plus a convertibility privilege into stock worth (say) $5.50, debthold-
ers receive a security worth $92.15. This exceeds the $91.79 value of the
current debt paying a $6.50 coupon, so bondholders will benefit. Stockholders
will also benefit by the rise in the equity value of $5.81 ($28.95 - $23.14) less
the $5.50 value of the convertibility option given bondholders.
Renegotiation of unprotected debt is particularly simple when bankruptcy
is imminent (V is close to V,), and C > C,,,(V). In this case, a small
reduction in the coupon will increase the value of both debt and equity-with
no further compensation to bondholders (such as the convertibility privilege)
being required. The firm may be able to reduce its coupon payment all the
way to C*(V) with no additional payments to bondholders if the value of debt
D*(V) at the optimal coupon is greater than the value of debt D(V) when the
renegotiation begins. This assumes stockholders can credibly make a "take-it-
or-leave-it" offer to bondholders. Note that the firm may wait until the brink
of bankruptcy before renegotiating, since this will minimize D(V).
M.Conclusion
By assuming a debt structure with time-independent payouts, we have
been able to develop closed-form solutions for the value of debt and for
optimal capital structure. This permits a detailed analysis of the behavior of
bond prices and optimal debt-equity ratios as firm asset value, risk, taxes,
interest rates, bond covenants, payout rates, and bankruptcy costs change.
The analysis examines two types of bonds: those that are protected by a
positive net-worth covenant, and those that are not. The distinction is critical
in determining when bankruptcy is triggered, which in turn affects bond
values and optimal leverage. To be rolled over, short-term financing typically

requires that the firm maintain positive net worth. Therefore short-term
financing seems to correspond to our model of protected debt. Long-term debt,
in contrast, rarely has positive net-worth covenants; it seems closer to our
model of unprotected debt.
Our results indicate that protected debt values and unprotected "invest-
ment grade" debt values behave very much as expected. Unprotected "junk"
bonds exhibit quite different behavior. For example, an increase in firm risk
will increase debt value, as will a decrease in the coupon. Such behavior is
not exhibited by protected " j u n k bonds.
Two curious aspects of optimal leverage are observed. First, a rise in the
risk-free interest rate (increasing the cost of debt financing) leads to a greater
optimal debt level. Higher interest rates generate greater tax benefits, which
in turn dictate more debt despite its higher cost. Second, the optimal debt for
firms with higher bankruptcy costs may carry a lower interest rate than for
firms with lower bankruptcy costs. This is because firms will choose signifi-
cantly lower optimal leverage when bankruptcy costs are substantial, making
debt less risky. This result does not hold for protected debt: higher bankruptcy
costs imply higher interest rates a t the optimal leverage.
Optimal leverage, yield ratios, and equity risk are well within historical
norms for protected debt. But optimal leverage seems high (and/or yield
spreads seem low) for unprotected debt. Variants of the basic assumptions,
discussed in Section VI, are needed for unprotected debt to fall within
historical norms. The most important modification is dropping the require-
ment that payouts to bondholders be externally financed.
Issuing debt without protective net-worth covenants yields greater tax
benefits and would seem to dominate issuing protected debt. However, this
conclusion may be reversed if firms have the ability to increase the riskiness
of their activities through "asset substitution." Increasing risk will transfer
value from bondholders to stockholders when debt is unprotected, leading
cautious bondholders to demand higher interest rates even when the firm
currently has low risk. But such costs typically are not incurred when firms
issue protected debt: stockholders will not gain by increasing firm risk when
debt is protected by a positive net-worth covenant, and bondholders will not
need to demand higher interest rates in anticipation of riskier firm activities.
Protected debt may be the preferred form of financing in these situations,
despite having lower potential tax benefits.
Our results offer some preliminary insights on debt repurchases and on
debt renegotiations. The former cannot be used to adjust leverage continu-
ously to its optimal level: bondholders will block further debt issuance, and
shareholders will block (marginal) debt reductions. Debt renegotiation can
achieve simultaneous increases in debt and equity value. But the costly
nature of renegotiation suggests it would be suboptimal to do so continuously
(see Fischer, Heinkel, and Zechner (1989)). Our analysis shows that it may be
desirable for shareholders to wait until the brink of bankruptcy before
renegotiating. When bankruptcy is neared, a reduction in coupon payments to
Debt Value, Bond Covenants, a n d Optimal Capital Structure 1249
the optimal level may benefit both stockholders and bondholders, without
additional side payments.
Although we have not emphasized equity values, our analysis also provides
some interesting insights. Equity return volatility will be stochastic, chang-
ing with the level of firm asset value, V. This (and the possibility of
bankruptcy) has important ramifications for option pricing.45
The model can be extended in several further dimensions. Multiple classes
of long-term debt can be analyzed, recognizing that payments to the various
classes of debtholders when bankruptcy occurs are determined by seniority.
More difficult extensions will include finite-lived debt, dynamic restructuring,
and a stochastic term structure of risk-free interest rates.
Appendix A
We assume in this case that instantaneous tax benefits = 0 whenever
V < VT. There are no carryforwards. Differential equation (3) with C = 0 has
solution:
Differential equation (3) with instantaneous tax benefit TC realized has

solution:
TB(V) = ( T C / ~ )+ B2VpX, V 2 VT. (45)
TB(V) must satisfy:
Solutions:
A, = (7C/r)(X/(X + l))(l/VT) (49)
A, = -(TC/~)(X/(X + l))(V$+l/VT) (50)
Substituting for tax benefits from equation (44) into equation (13) for equity
gives, for V IVT,
To find VB, we again set dE/dV lv= v, = 0:

dE/dV
45~reliminary
work on this question has been done by Klaus Toft (1993).
Evaluating equation (53) at V = VB:
Substituting for A, and A, gives
D can be computed from equation (7); and
Note we can rewrite the expression for VB as
with the last inequality holding since VT > _V,, where _V, satisfies equation
(14).
Appendix B
Parallel to the discussion following equation (43), we know that sharehold-
ers will reject a buyback of debt (i.e., dC < 0) if [(dE/dC) + (D/C)] > 0.
Since E = u - D,
Define V* as the firm asset value at which the current coupon would be
optimal for protected debt, i.e., for which du(V*)/dC = 0. From equation
(401, it follows that [(dD/dC) - (D/C)] will be strictly negative, and there-
fore equation (58) will be strictly positive when dv/dC = 0. Continuity
implies that there exists a neighborhood of values, V, around the value V*,
for which equation (58) is strictly positive. For all V < V* in this neighbor-
hood, firm value, u, would be increased by lowering the coupon, since the
optimal coupon is decreasing in V. But because equation (58) is positive in
this neighborhood, current stockholders' equity value will fall when dC < 0
and shareholders will resist reducing the coupon to its optimal level.
REFERENCES
Altman, E., 1984, A further empirical investigation of the bankruptcy cost question, Journal of
Finance 39, 1067-1090.
Anderson, R., and S. Sundaresan, 1992, Design and valuation of debt contracts, Working paper,
Universite Catholique de Louvain.
Asquith, P., R. Gertner, and D. Scharfstein, 1991, Anatomy of financial distress: An examination
of junk bond issuers, Working paper, University of Chicago.
Baxter, N., 1967, Leverage, risk of ruin, and the cost of capital, Journal of Finance 22, 395-403.
Black, F., and J. Cox, 1976, Valuing corporate securities: Some effects of bond indenture
provisions, Journal of Finance 31, 351-367.
Black, F., and M. Scholes, 1973, The pricing of options and corporate liabilities, Journal of
Political Economy 81, 637-654.
Brennan, M., and E. Schwartz, 1978, Corporate income taxes, valuation, and the problem of
optimal capital structure, Journal of Business 51, 103-114.
, 1980, Analyzing convertible bonds, Journal of Financial and Quantitative Analysis 15,
907-932.
, 1984, Optimal financial policy and firm valuation, Journal of Finance 39, 593-607.
Eberhart, A., W . Moore, and R. Roenfeldt, 1990, Security pricing and deviations from the
absolute priority rule i n bankruptcy proceedings, Journal of Finance 45, 1457-1469.
Fischer, E., R. Heinkel, and J. Zechner, 1989, Dynamic capital structure choice: Theory and tests,
Journal of Finance 44, 19-40.
Franks, J., and W . Torous, 1989, An empirical investigation of firms i n reorganization, Journal
of Finance 44, 747-779.
Fries, S., M. H. Miller, and W . Perraudin, 1993, Pricing bond default premia i n a competitive
industry equilibrium with costs of entry and exit, Working paper, University of Cambridge.
Harris, M., and A. Raviv, 1991, The theory o f capital structure, Journal of Finance 44, 297-355.
Haugen, R., and L. Senbet, 1978, The insignificance of bankruptcy costs to the theory o f optimal
capital structure, Journal of Finance 33, 383-392.
Jensen, M., and W . Meckling, 1976, Theory o f the firm: Managerial behavior, agency costs, and
ownership structure, Journal of Financial Economics 4, 305-360.
Jensen, M., 1986, The agency costs of free cash flow: Corporate finance and takeovers, American
Economic Review 76, 323-329.
Jones, E., S. Mason, and E. Rosenfeld, 1984, Contingent claims analysis of corporate capital
structures: A n empirical investigation, Journal of Finance 39, 611-625.
Kane, A,, A. Marcus, and R. McDonald, 1984, How big is the tax advantage to debt?, Journal of
Finance 39, 841-852.
Kaplan, S., 1989, Management buyouts: Evidence on taxes as a source of value, Journal of
Finance 44, 611-632.
Kim, E. H., 1978, A mean-variance theory o f optimal capital structure and corporate debt
capacity, Journal of Finance 33, 45-64.
Kim, I., K. Ramaswamy, and S. Sundaresan, 1993, Does default risk i n coupons affect the
valuation of corporate bonds?: A contingent claims model, Financial Management 22,
117-131.
Kraus, A., and R. Litzenberger, 1973, A state-preference model of optimal financial leverage,
Leland, H., 1989, LBOs and taxes: No one to blame but ourselves?, California Management
Review 32, 19-28.
Longstaff, F., and E. Schwartz, 1992, Valuing risky debt: A new approach, Working paper,
University o f California, Los Angeles.
Mauer, D., and A. Triantis, 1993, Interactions of investment decisions and corporate financing
policies, Working paper, University o f Wisconsin.
Mella, P., and W . Perraudin, 1993, Strategic debt service, Working paper, University of Cam-
bridge.
Mello, A,, and J. Parsons, 1992, The agency costs of debt, Journal of Finance 47, 1887-1904.
Merton, R., 1973, A rational theory of option pricing, Bell Journal of Economics and Manage-
ment Science 4, 141-183.
, 1974, O n the pricing of corporate debt: The risk structure o f interest rates, Journal of
Finance 29, 449-469.
Miller, M., 1977, Debt and taxes, Journal of Finance 32, 261-275.
Modigliani, F., and M. Miller, 1958, The cost of capital, corporation finance and the theory o f
investment, American Economic Review 48, 267-297.
Myers, S., 1984, The capital structure puzzle, Journal of Finance 39, 575-592.
, and N. Majluf, 1984, Corporate financing and investment decisions when firms have
information that investors do not have, Journal of Financial Economics 5, 187-221.
Sarig, O., and A. Warga, 1989, Some empirical estimates of the risk structure of interest rates,
Scott, J., 1976, A theory of optimal capital structure, Bell Journal of Economics and Manage-
ment Science 7, 33-54.
Smith, C., and J. Warner, 1979, On financial contracting: An analysis of bond covenants, Journal
of Financial Economics 7, 117-161.
Titman, S., and R. Wessels, 1988, The determinants of capital structure choice, Journal of
Finance 43, 1-20.
Toft, K., 1993, Options on leveraged equity with default risk, Working paper, Haas School of
Business, University of California, Berkeley.
Warner, J., 1977, Bankruptcy costs: Some evidence, Journal of Finance 32, 337-347.
LINKED CITATIONS
- Page 1 of 12 -
You have printed the following article:

Corporate Debt Value, Bond Covenants, and Optimal Capital Structure
Hayne E. Leland
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28199409%2949%3A4%3C1213%3ACDVBCA%3E2.0.CO%3B2-0
This article references the following linked citations. If you are trying to access articles from an
off-campus location, you may be required to first logon via your library web site to access JSTOR. Please
visit your library's website or contact a librarian to learn about options for remote access to JSTOR.
[Footnotes]
1
Debt and Taxes
Merton H. Miller
The Journal of Finance, Vol. 32, No. 2, Papers and Proceedings of the Thirty-Fifth Annual Meeting
of the American Finance Association, Atlantic City, New Jersey, September 16-18, 1976. (May,
1977), pp. 261-275.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28197705%2932%3A2%3C261%3ADAT%3E2.0.CO%3B2-D
2
A Further Empirical Investigation of the Bankruptcy Cost Question
Edward I. Altman
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198409%2939%3A4%3C1067%3AAFEIOT%3E2.0.CO%3B2-L
2
The Theory of Capital Structure
Milton Harris; Artur Raviv
The Journal of Finance, Vol. 46, No. 1. (Mar., 1991), pp. 297-355.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28199103%2946%3A1%3C297%3ATTOCS%3E2.0.CO%3B2-L
NOTE: The reference numbering from the original has been maintained in this citation list.
LINKED CITATIONS
- Page 2 of 12 -
2
The Determinants of Capital Structure Choice
Sheridan Titman; Roberto Wessels
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198803%2943%3A1%3C1%3ATDOCSC%3E2.0.CO%3B2-A
2
Bankruptcy Costs: Some Evidence
Jerold B. Warner
1977), pp. 337-347.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28197705%2932%3A2%3C337%3ABCSE%3E2.0.CO%3B2-V
3
Leverage, Risk of Ruin and the Cost of Capital
Nevins D. Baxter
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28196709%2922%3A3%3C395%3ALRORAT%3E2.0.CO%3B2-4
3
A State-Preference Model of Optimal Financial Leverage
Alan Kraus; Robert H. Litzenberger
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28197309%2928%3A4%3C911%3AASMOOF%3E2.0.CO%3B2-3
3
The Capital Structure Puzzle
Stewart C. Myers
The Journal of Finance, Vol. 39, No. 3, Papers and Proceedings, Forty-Second Annual Meeting,
American Finance Association, San Francisco, CA, December 28-30, 1983. (Jul., 1984), pp.
575-592.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198407%2939%3A3%3C575%3ATCSP%3E2.0.CO%3B2-%23
LINKED CITATIONS
- Page 3 of 12 -
4
A Mean-Variance Theory of Optimal Capital Structure and Corporate Debt Capacity
E. Han Kim
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28197803%2933%3A1%3C45%3AAMTOOC%3E2.0.CO%3B2-O
5
Analyzing Convertible Bonds
Michael J. Brennan; Eduardo S. Schwartz
The Journal of Financial and Quantitative Analysis, Vol. 15, No. 4, Proceedings of 15th Annual
Conference of the Western Finance Association, June 19-21, 1980, San Diego, California. (Nov.,
1980), pp. 907-929.
Stable URL:
http://links.jstor.org/sici?sici=0022-1090%28198011%2915%3A4%3C907%3AACB%3E2.0.CO%3B2-1
8
Contingent Claims Analysis of Corporate Capital Structures: An Empirical Investigation
E. Philip Jones; Scott P. Mason; Eric Rosenfeld
611-625.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198407%2939%3A3%3C611%3ACCAOCC%3E2.0.CO%3B2-A
8
Some Empirical Estimates of the Risk Structure of Interest Rates
Oded Sarig; Arthur Warga
The Journal of Finance, Vol. 44, No. 5. (Dec., 1989), pp. 1351-1360.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198912%2944%3A5%3C1351%3ASEEOTR%3E2.0.CO%3B2-5
9
Measuring the Agency Cost of Debt
Antonio S. Mello; John E. Parsons
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28199212%2947%3A5%3C1887%3AMTACOD%3E2.0.CO%3B2-S
LINKED CITATIONS
- Page 4 of 12 -
9
Corporate Income Taxes, Valuation, and the Problem of Optimal Capital Structure
M. J. Brennan; E. S. Schwartz
The Journal of Business, Vol. 51, No. 1. (Jan., 1978), pp. 103-114.
Stable URL:
http://links.jstor.org/sici?sici=0021-9398%28197801%2951%3A1%3C103%3ACITVAT%3E2.0.CO%3B2-Z
9
Stable URL:
10
Dynamic Capital Structure Choice: Theory and Tests
Edwin O. Fischer; Robert Heinkel; Josef Zechner
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198903%2944%3A1%3C19%3ADCSCTA%3E2.0.CO%3B2-C
12
Stable URL:
12
On the Pricing of Corporate Debt: The Risk Structure of Interest Rates
Robert C. Merton
The Journal of Finance, Vol. 29, No. 2, Papers and Proceedings of the Thirty-Second Annual
Meeting of the American Finance Association, New York, New York, December 28-30, 1973.
(May, 1974), pp. 449-470.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28197405%2929%3A2%3C449%3AOTPOCD%3E2.0.CO%3B2-%23
LINKED CITATIONS
- Page 5 of 12 -
13
1980), pp. 907-929.
Stable URL:
15
An Empirical Investigation of U.S. Firms in Reorganization
Julian R. Franks; Walter N. Torous
The Journal of Finance, Vol. 44, No. 3, Papers and Proceedings of the Forty-Eighth Annual Meeting
of the American Finance Association, New York, New York, December 28-30, 1988. (Jul., 1989),
pp. 747-769.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198907%2944%3A3%3C747%3AAEIOUF%3E2.0.CO%3B2-G
15
Security Pricing and Deviations from the Absolute Priority Rule in Bankruptcy Proceedings
Allan C. Eberhart; William T. Moore; Rodney L. Roenfeldt
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28199012%2945%3A5%3C1457%3ASPADFT%3E2.0.CO%3B2-D
17
Robert C. Merton
(May, 1974), pp. 449-470.
Stable URL:
24
Debt and Taxes
Merton H. Miller
1977), pp. 261-275.
Stable URL:
LINKED CITATIONS
- Page 6 of 12 -
24
Jerold B. Warner
1977), pp. 337-347.
Stable URL:
27
Debt and Taxes
Merton H. Miller
1977), pp. 261-275.
Stable URL:
31
Robert C. Merton
(May, 1974), pp. 449-470.
Stable URL:
32
Stable URL:
36
The Pricing of Options and Corporate Liabilities
Fischer Black; Myron Scholes
The Journal of Political Economy, Vol. 81, No. 3. (May - Jun., 1973), pp. 637-654.
Stable URL:
http://links.jstor.org/sici?sici=0022-3808%28197305%2F06%2981%3A3%3C637%3ATPOOAC%3E2.0.CO%3B2-P
LINKED CITATIONS
- Page 7 of 12 -
38
pp. 747-769.
Stable URL:
38
Stable URL:
References
A Further Empirical Investigation of the Bankruptcy Cost Question

Edward I. Altman
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198409%2939%3A4%3C1067%3AAFEIOT%3E2.0.CO%3B2-L
Leverage, Risk of Ruin and the Cost of Capital

Nevins D. Baxter
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28196709%2922%3A3%3C395%3ALRORAT%3E2.0.CO%3B2-4
Valuing Corporate Securities: Some Effects of Bond Indenture Provisions

Fischer Black; John C. Cox
The Journal of Finance, Vol. 31, No. 2, Papers and Proceedings of the Thirty-Fourth Annual
Meeting of the American Finance Association Dallas, Texas December 28-30, 1975. (May, 1976),
pp. 351-367.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28197605%2931%3A2%3C351%3AVCSSEO%3E2.0.CO%3B2-Z
LINKED CITATIONS
- Page 8 of 12 -
The Pricing of Options and Corporate Liabilities

Fischer Black; Myron Scholes
The Journal of Political Economy, Vol. 81, No. 3. (May - Jun., 1973), pp. 637-654.
Stable URL:
http://links.jstor.org/sici?sici=0022-3808%28197305%2F06%2981%3A3%3C637%3ATPOOAC%3E2.0.CO%3B2-P
Stable URL:

1980), pp. 907-929.
Stable URL:
Optimal Financial Policy and Firm Valuation

593-607.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198407%2939%3A3%3C593%3AOFPAFV%3E2.0.CO%3B2-5
Stable URL:
LINKED CITATIONS
- Page 9 of 12 -
Dynamic Capital Structure Choice: Theory and Tests

Edwin O. Fischer; Robert Heinkel; Josef Zechner
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198903%2944%3A1%3C19%3ADCSCTA%3E2.0.CO%3B2-C

pp. 747-769.
Stable URL:
The Theory of Capital Structure

Milton Harris; Artur Raviv
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28199103%2946%3A1%3C297%3ATTOCS%3E2.0.CO%3B2-L
The Insignificance of Bankruptcy Costs to the Theory of Optimal Capital Structure

Robert A. Haugen; Lemma W. Senbet
The Journal of Finance, Vol. 33, No. 2. (May, 1978), pp. 383-393.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28197805%2933%3A2%3C383%3ATIOBCT%3E2.0.CO%3B2-5
Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers

Michael C. Jensen
The American Economic Review, Vol. 76, No. 2, Papers and Proceedings of the Ninety-Eighth
Annual Meeting of the American Economic Association. (May, 1986), pp. 323-329.
Stable URL:
http://links.jstor.org/sici?sici=0002-8282%28198605%2976%3A2%3C323%3AACOFCF%3E2.0.CO%3B2-M
LINKED CITATIONS
- Page 10 of 12 -
Contingent Claims Analysis of Corporate Capital Structures: An Empirical Investigation

E. Philip Jones; Scott P. Mason; Eric Rosenfeld
611-625.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198407%2939%3A3%3C611%3ACCAOCC%3E2.0.CO%3B2-A
How Big is the Tax Advantage to Debt?

Alex Kane; Alan J. Marcus; Robert L. McDonald
841-853.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198407%2939%3A3%3C841%3AHBITTA%3E2.0.CO%3B2-0
Management Buyouts: Evidence on Taxes as a Source of Value

Steven Kaplan
pp. 611-632.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198907%2944%3A3%3C611%3AMBEOTA%3E2.0.CO%3B2-C
A Mean-Variance Theory of Optimal Capital Structure and Corporate Debt Capacity

E. Han Kim
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28197803%2933%3A1%3C45%3AAMTOOC%3E2.0.CO%3B2-O
A State-Preference Model of Optimal Financial Leverage

Alan Kraus; Robert H. Litzenberger
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28197309%2928%3A4%3C911%3AASMOOF%3E2.0.CO%3B2-3
LINKED CITATIONS
- Page 11 of 12 -
Measuring the Agency Cost of Debt

Antonio S. Mello; John E. Parsons
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28199212%2947%3A5%3C1887%3AMTACOD%3E2.0.CO%3B2-S

Robert C. Merton
(May, 1974), pp. 449-470.
Stable URL:
Debt and Taxes

Merton H. Miller
1977), pp. 261-275.
Stable URL:
The Cost of Capital, Corporation Finance and the Theory of Investment

Franco Modigliani; Merton H. Miller
The American Economic Review, Vol. 48, No. 3. (Jun., 1958), pp. 261-297.
Stable URL:
http://links.jstor.org/sici?sici=0002-8282%28195806%2948%3A3%3C261%3ATCOCCF%3E2.0.CO%3B2-3
The Capital Structure Puzzle

Stewart C. Myers
575-592.
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198407%2939%3A3%3C575%3ATCSP%3E2.0.CO%3B2-%23
LINKED CITATIONS
- Page 12 of 12 -

Stable URL:
The Determinants of Capital Structure Choice

Sheridan Titman; Roberto Wessels
Stable URL:
http://links.jstor.org/sici?sici=0022-1082%28198803%2943%3A1%3C1%3ATDOCSC%3E2.0.CO%3B2-A

Jerold B. Warner
1977), pp. 337-347.
Stable URL:

Leland 1994 JF

Uploaded by

Copyright:

Available Formats

Leland 1994 JF

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Leland 1994 JF

Uploaded by

Copyright:

Available Formats

Corporate Debt Value, Bond Covenants, and Optimal Capital Structure

The Journal of Finance is currently published by American Finance Association.

Corporate Debt Value, Bond Covenants,

THEVALUE OF CORPORATE debt and capital structure are interlinked variables.

problems between bondholders and stockholders?

When can debt renegotiation be expected prior to bankruptcy, and can

renegotiation achieve results that debt repurchase cannot?

However, large (discontinuous) debt repurchases via tender offers may

I. A Model of Time-IndependentSecurity Values

where W is a standard Brownian motion. We shall refer to V as the "asset

with boundary conditions determined by payments at maturity, and by

Equation (3) has the general solution

lost to bankruptcy costs, leaving debtholders with value (1 - a)VB and

Condition (6ii) holds because bankruptcy becomes irrelevant as V becomes

In this case equation (4) has solution

BC is a decreasing, strictly convex function of V. Again, note the reinter-

Tax benefits are an increasing, strictly concave function of V.

We see from equation (14) below that when VB is endogenously deter-

11. Debt with No Protective Covenants: The Endogenous

A. The Comparatiue Statics of Debt Value (D(V))

Comparative Statics of Financial Variables: Unprotected Debt

Limit As Sign of Change in Instrument for a n Increase in: 3

*Sign reversal a s V -, VB only if a and/or T > 0.

Substituting this into equation (15) and simplifying gives

B. Yield Spreads: The Risk Structure of Interest Rates

111; Optimal Leverage with Unprotected Debt

C*, as a function of asset value, V, gives:

Table I1 indicates the comparative statics of these variables plus optimal

by observing the maximal firm values in Figure 7. The potential gains in

leverage ratios drop by 5 to 25 percent when the effective tax rate is 15

Comparative Statics of Financial Variables at the Optimal

Leverage Ratio: Unprotected Debt

Sign of Change in Variable for a n Increase in:

Linear in V < 0, a 2 small;

IV. Positive Net-Worth Covenants and the Value of Protected

Are there contractual arrangements in which this is a realistic description

Except when CY = 0, closed-form solutions for the function Do(Vo) satisfy-

When there are no bankruptcy costs ( a = 01,

V. Optimal Leverage with Protected Debt

d) The maximal benefits of unprotected over protected debt increase as:

Because protected debt is risk free when a = 0, it also follows that

VI. Discussion and Variations: Debt Value and Capital

Comparative Statics of Financial Variables at the Optimal

Leverage Ratio: Protected Debt

Sign of Change in Variable for a n Increase in:

R* Invariant to V > 0" >0 > ob > 0"

and protected (or continuously rolled-over) debt. It is of interest to compare

levered firm's equity is volatile, with a 57 percent annual standard deviation.

A. No Tax Shelter for Interest Payments When Value Falls

B. Net Cash Payouts by the Firm

with general solution

Boundary conditions remain unchanged, implying A, = 0 as before since

C. Absolute Priority Not Respected

and equation (14) will be replaced by

D. All of the Above

VII. Protected versus Unprotected Debt: Potential Agency

Values of Protected and Unprotected Debt and Equity for

Different Levels of Risk

Unprotected Debt Protected Debt

of 20 percent-with no change in current asset value V. Will they be