debt tax gains kolari mwfa

8/8/2019 Debt Tax Gains Kolari MWFA

1/43

1

On the Debt Tax Shield Controversy in Corporate Valuation:Discounting at the Levered Cost of Equity

James W. Kolari* Texas A&M University

Abstract

The value of debt tax shields in foundational valuation models by Nobel Laureates Modiglianiand Miller (MM) continues to be a controversial issue that is central to our understanding of corporate finance. This paper contributes to this pivotal debate by proposing that the appropriatediscount rate to value interest tax deductions available to levered shareholders is the levered costof equity, rather than riskless and risky costs of debt or the unlevered cost of equity in previousstudies. Assuming no bankruptcy risk and no personal taxes, the proposed revised tax modelyields the following results that are more consistent with MMs original tax model than theirlater corrected tax model: (1) relatively small debt tax shields, (2) a flat-based, inverse U-shapedfirm value function with an interior optimal capital structure, (3) a wide relevant range of capitalstructures that captures most debt tax shields, (4) a flat-based U-shaped cost of capital function,and (5) debt tax shields and capital structures that are sensitive to riskless debt rates, unleveredequity rates, and corporate tax rates. Implications to corporate capital structure decisions andprevious literature are discussed. Also, analyses are extended to Millers personal tax extensionof the MM model.

JEL classification: G12; G31; G32

Keywords: Value of tax shields; Capital structure; Firm valuation; Share valuation

*The author is JP Morgan Chase Professor of Finance, Texas A&M University, FinanceDepartment, College Station, TX 77843-4218Office phone: 979-845-4803Email address: [email protected]

I have benefited from numerous discussions over the years with many colleagues, including Will

Armstrong, Jaap Bos, Michele Caputo, Paige Fields, Donald Fraser, Johan Knif, MichaelKoetter, Seppo Pynnnen, Joseph Reising, Hwan Shin, Soenke Sievers, Sorin Sorescu, AnttiSuvanto, and Marilyn Wiley, in addition to participants at the 2008 Financial ManagementAssociation European conference in Prague, Czech Republic. All remaining errors are theresponsibility of the author.


2/43

2

On the Debt Tax Shield Controversy in Corporate Valuation:Discounting at the Levered Cost of Equity

After more than 50 years of debate and extensive empirical evidence 1, controversy

continues regarding now famous corporate valuation theories by Nobel Laureates Modigliani and

Miller (hereafter MM) (1958, 1963). The main dispute focuses on the present value of interest

tax deductions on debt. Initiating the debate, in 1958 MM used the unlevered equity rate to

discount interest tax deductions but in 1963 corrected their valuation results by discounting

interest tax deductions at the riskless debt rate to yield large debt tax gains. Motivated by

excessively large debt tax shields that did not coincide with observed corporate practice, Miller

(1977) extended MMs tax correction valuation model to personal taxes and proposed that

interest tax gains on debt have little or no value for most firms. Excluding personal taxes, some

authors subsequently returned to MMs original tax model approach by discounting interest tax

deductions at the unlevered equity discount rate, including Miles and Ezzell (1980, 1985), Harris

and Pringle (1985), Modigliani (1988), and Kaplan and Ruback (1995). Consistent with

approach, Grinblatt and Liu (2008) have viewed the debt tax shield as a derivative of the

underlying unlevered asset (and its cash flows) and, subsequently, have used option pricing

methods to derive its value. Other researchers have employed the cost of risky debt to discount

interest tax deductions, such as Myers (1974), Luehrman (1997), Damodaran (2006), and others.

In general, these alternative discounting approaches 2 tend to diminish the present value of tax

1 For example, see Myers (1974), DeAngelo and Masulis (1980), Masulis (1980), Fama and French (1998), Ruback,(1995, 2002), Graham (2000, 2008), Graham and Harvey (2001), Brealey and Myers (2003), Arzac and Glosten(2005), and many others. Graham (2003) provides an excellent review of this literature.2 For studies debating different discount rates, see Miles and Ezell (1980, 1985), Harris and Pringle (1985),Modigliani (1988), Damodaran (1994), Kaplan and Ruback (1995), Ruback (2002), Fernandez (2004, 2007), Arzacand Glosten (2005), Cooper and Nyborg (2006), and others. For example, Miles and Ezell (1985) have argued that,if firms maintain a constant level of debt as assumed by MM, the riskless cost of debt is the appropriate discount ratefor the debt tax shield. However, if firms instead maintain a constant debt-to-firm value ratio, debt tax shieldsshould be discounted using the unlevered equity rate.


3/43

3

gains on debt and thereby soften MMs 1963 conclusions. Importantly, the debt tax shield

controversy is central to our understanding of corporate finance, especially firm, equity, and debt

valuation as well as the cost of capital.

Relevant to this controversy and the present paper, Ross (2007) has recently advocated

that, even though interest tax shields are riskless, they should be discounted using an equity rate.

In his words, A flaw in the traditional approach occurs at the first step where the system is

closed by valuing the cash flow of the tax shield at the riskless debt rate. The tax shield is, in

fact, assumed to be riskless, but that does not mean that it should be valued at the riskless rate on

bonds As Miller (1977) argues, the tax shield goes to the equity holders and not thebondholders and the discount rate at which it should be valued should depend on the marginal

tax rate of equity holders and not debtholders. (2007, p.7) Consistent with Ross and Miller, this

paper argues that interest tax deductions are cash flows available to shareholders. 3 Rather than

discounting interest tax deductions with debt rates or the unlevered cost of equity as in previous

papers cited above, the valuation implications of discounting at the levered cost of equity are

examined. It is argued that interest tax deductions are levered cash flows that arise from the use

of debt and are available to levered shareholders, as opposed to debt cash flows available to

debtholders or unlevered cash flows available to unlevered shareholders. Moreover, as shown

later, since equity valuation in the MM framework discounts interest tax shields available to

levered shareholders at the levered cost of equity, firm valuation should discount interest tax

shields available to levered shareholders at the levered cost of equity also. By solving for the

levered cost of equity used for discounting purposes in this joint equity/firm valuation problem, a

revised tax model with new firm, equity, and debt valuation results is derived. To the authors

3 Standard textbooks in finance commonly note that interest tax deductions increase the cash flows of shareholders.For example, Grinblatt and Titman (1998, p. 511) observe that, if a firm issues debt to buy back equity, shareholdersbenefit from an increase in cash flow associated with interest tax deductions on debt.


4/43

4

knowledge no previous papers have investigated this logically appealing levered equity rate

discounting approach. Upon doing so, the results appear to help mitigate a number of criticisms

of MMs tax models, contribute to reconciling theory and evidence on capital structure and firm

valuation, and provide new insights into the role of debt tax shields in firm valuation.

The main findings of the proposed revised tax model are plausibly consistent with

empirical evidence: (1) much smaller debt tax shields than in MMs tax correction model; (2) a

flat-based, inverse U-shaped firm value function with an interior optimal capital structure as

opposed to an all-debt corner solution; (3) a relevant range of debt/value ratios spanning a fairly

broad interval within which most debt tax shields are captured; (4) a flat-based U-shaped cost of capital function; and (5) sensitivity of debt tax shields and optimal capital structures to riskless

debt rates, unlevered costs of equity (i.e., business risk), and corporate tax rates. Importantly, the

revised tax model provides results that are more akin to real world capital structure and firm

valuation observations. Also, the results tend to support for MMs original 1958 tax model in

which firm valuation was independent of capital structure under Proposition I, which Modigliani

(1988) later claimed in a retrospective article was the definitive truth.

MM's valuation models are briefly overviewed in the next section. Section 2 discounts

interest tax deductions in both share value and firm value models at the levered cost of equity.

The resultant revised tax model yields new firm and share values that are markedly different

from MMs tax models. Section 3 discusses implications to corporate capital structure decisions

and related literature that bode well for the revised tax models valuation of debt tax shields.

Following Miller (1977), Section 4 extends the analyses to personal taxes. Section 5 concludes.


5/43

5

1. MM's Valuation Models

According to Proposition I of Modigliani and Miller's (MM) (1958) original valuation

models with and without corporate taxes (but no personal taxes), firm valuation is independent of

capital structure. However, MM (1963) later corrected their original tax model by utilizing a

lower discount rate for tax deductible interest payments of the firm, which resulted in a large tax

gain on debt usage to the firm and its shareholders. In the absence of bankruptcy costs,

Proposition I was rejected in favor of an extreme all-debt capital structure corner solution. In a

1988 rejoinder to the MM tax correction paper, Miller (1988) commented that: "We must admit

that we too were somewhat taken aback when we first saw this conclusion (i.e., an all-debtcapital structure) emerging from our analysis." (1988, p. 112) 4 As noted by Green and Hollifield

(2003), because an all-debt corner solution is inconsistent with observed capital structures of

firms, MMs tax correction model is not useful for policy purposes. Nonetheless, these authors

conceded that it is a foundational theoretical model in capital structure research and practice.

In an effort to assuage large tax gains on debt, researchers introduced bankruptcy costs to

trade-off tax benefits of debt against rising costs of debt as firms increase financial leverage (see

Kraus and Litzenberger (1973), Scott (1976), Chen and Kim (1979), Bradley and Kim (1984),

Myers (1984), Shyam-Sunder and Myers (1998), and others). However, empirical evidence has

revealed that these costs are relatively small (see Warner (1977), Andrade and Kaplan (1998),

Frank and Goyal (2008), and others). Referring to the lopsided trade-off of excessively large tax

gains on interest deductions compared to small expected bankruptcy costs, especially for big

firms with high credit quality, Miller (1977) criticized MMs tax correction model as a horse

and rabbit stew. Motivated in large part by this problem, Miller extended MMs tax correction

4 See also Gordon (1989) and Weston (1989).


6/43

6

model to personal taxes. Assuming negligible personal taxes on equity (due to deferring capital

gains) and personal taxes on interest income equal to the corporate tax rate at the margin in the

bond market, he found that there is no tax gain from financial leverage for the corporate sector as

a whole (i.e., Proposition I holds).

It is interesting that Modigliani (1988) later made a reversal: "Personal taxation aside,

the definitive truth was all in MM (1958) ..." (1988, p. 153) As in the original valuation model

with corporate taxes, "... the levered firm can be seen to be proportional to that of the unlevered

firm ... This proportionality of returns in turn implies that the market value of the levered firm

VL must also be proportional to that of the unlevered firm, V U ..." (1988, p. 152) The reason for

this reversal is that Modigliani returned to using a risky rate of return to discount interest tax

deductions similar to the discount rate on the unlevered firm's equity earnings, which are

discounted at a riskless rate of interest in MMs tax correction model. The larger discount rate

on interest tax gains decreases the present value of debt tax shields and diminishes their potency

as a factor driving corporate capital structure decisions. Closely related to Modiglianis

retrospective opinion, Miles and Ezzell (1980, 1985), Harris and Pringle (1985), Kaplan and

Ruback (1995), Grinblatt and Liu (2008), and others have argued that debt tax shields are as

risky as the free (unlevered) cash flows generated from the firms assets, which suggests that the

unlevered equity discount rate is appropriate. However, as shown in the next section, while

application of higher unlevered equity discount rates diminishes the magnitude of debt tax

shields, it does not necessarily overturn the extreme all-debt capital structure implication of

MMs tax correction model.

MMs no tax and tax models of firm valuation are well known. Assuming no corporate

or personal taxes, no potential bankruptcy and nondebt tax shields, capital markets are


7/43

7

frictionless, symmetric information, complete contracting, complete markets, and all cash flow

streams are perpetuities, MM derived the following original model of corporate valuation with

no taxes: 5

U LV=

X

=r

rD +

rD)-X(=D+S =V

i, (1)

where V L = the market value of the levered firm, S = the market value of levered equity, D = the

market value of riskless debt, = the expected value of risky stream that is divided between

the firms shareholders and debtholders as rD)-X( and rD, respectively, r = the riskless debt rate,

= the market rate of capitalization with no taxes (i.e., required rate of return or cost of

unlevered equity), i = the levered cost of equity with no taxes, and V U = the market value of the

unlevered (or no debt) firm. Known as Proposition I, this model implies that the value of the

firm is independent of its financial leverage and that unlevered and levered weighted average

costs of capital (WACC) are equal to one another. Using equation (1), we can compute the value

of equity alternatively as S = V L - D = V U - D. That is, given the value of unlevered equity, or

VU, we can substitute in values of debt, or D, within the range of 0 D VU to obtain the full

range of equity values, or S.

With corporate taxes MM (1958, p. 272) originally posited that firm value is proportional

to the expected after-tax earnings of the firm, or , as follows:

,rD

+V=rD

+)-(1X

=X

=V UL

(2)

5 Stiglitz (1969) has shown that MMs original valuation model with no taxes holds under more general conditions,such that these assumptions are not restrictive for the most part.

X X~

X


8/43

8

where LV = the market value of the levered firm after taxes, = the corporate tax rate, X (1-)

= the expected unlevered returns after taxes, = the appropriate after-tax market capitalization

rate for unlevered equity, rD = the interest payments on riskless debt D paying the riskless debt

rate r, and UV = the after-tax value of the unlevered firm. As already mentioned, Modigliani

(1988) later noted that, since the unlevered equity rate applied to discounting interest tax

deductions rD is higher than the riskless debt rate r, the present value of the debt tax shield is

substantially reduced.

In a major correction to their 1958 paper, MM (1963, p. 436) altered their after-tax

argument by rejecting Proposition I. The expected after-tax earnings stream defined as

= )-rD)(1-X( + rD = )-(1X + rD was discounted as follows in their tax correction

model:

D,+V=rrD

+

)-(1X =V U

L (3)

where the lower riskless discount rate r is used to discount the interest tax deduction rD

available to shareholders. This change produced a large tax gain equal to D for the firms

shareholders. 6 MM created arbitrage proofs to show that, in equilibrium, levered firm value

cannot be more or less than the sum of the unlevered firm value plus the tax gain D, such that

equation (3) holds. The value of the firm is no longer proportional to its earnings after taxes, as

firm value is dependent on the firm's capital structure. Indeed, excluding bankruptcy costs and

other imperfections, their interpretation of this tax correction model was that the firms

shareholders will seek a corner solution of nearly all debt. The value of levered equity is

6 See Myers (1974), Miles and Ezzell (1980, 1985), Fernandez (2004), and Cooper and Nyborg (2006) fordiscussion and different viewpoints on computing the present value of tax shields.

X


9/43

9

computed simply as =S LV - D =UV - D + D, which shows that shareholders reap the large

tax gain on debt interest. As in the original no tax valuation model, by varying the level of debt

within the range 0 D LV , the full range of equity values, or ,S can be computed.

2. Revised Tax Model

This section proposes a revised tax model that discounts interest tax deductions available

to levered shareholders at the levered cost of equity in both equity value and firm value models.

We seek to find the levered cost of equity that simultaneously solves this joint equity/firm

valuation problem at a particular financial leverage level. The MM tax models do not have this

simultaneity problem due to fixing the discount rate at the riskless rate or unlevered equity rate,

both of which are invariant to financial leverage. Based on simple numerical examples

comparing MMs tax models to the proposed revised tax model, discounting debt tax shields at

the levered cost of equity changes the valuation results of MMs tax models to be more in line

with Proposition I in their original 1958 paper. As quoted in the previous section, Modigliani

(1988) later considered Proposition I in their original paper to be the definitive truth even

without personal taxes.

2.1 Derivation of firm, equity, and debt values

The value of the firm is the sum of the following general expressions for equity and debt

values:

,r

rD+rD+rD-)-(1X=r

rD+)-(1rD)-X( =D+S=V

L iii(4)

where i = the rate of return on equity with taxes or levered cost of equity, and other terms are as

defined in the previous section. Equation (4) decomposes the earnings available to levered


10/43

10

shareholders, or )-rD)(1-X( , into the component parts rD-)-(1X and rD. This

decomposition makes clear that the present value of levered shareholders cash flows before

interest tax deductions but after paying debt interest is [ rD -)-(1X ]/ i , and the present value of

levered shareholders interest tax deductions is rD/ .i 7 These cash flows distributed to

shareholders as either capital appreciation or dividends are aggregatively discounted at the

levered cost of equity . We do not decompose these cash flows by discounting X(1- ) at the

unlevered equity rate and discounting rD and rD at the riskless debt rate r. Herein lies the

flaw in the MM tax approaches for which Ross (2007) was quoted in the introduction. In their

original tax and tax correction models of firm valuation shown in equations (2) and (3), the

interest tax deduction rD is discounted at the unlevered equity rate

and riskless debt rate r,

respectively. However, this discount approach in their firm value models is inconsistent with

discounting the interest tax deduction rD at the levered equity rate i in the general equity value

model in equation (4). Since levered shareholders compute equity value by discounting levered

cash flows such as interest tax deductions at the levered cost of equity i rather than the riskless

debt rate, interest tax deductions in the firm value model (which represent levered cash flows

available to levered shareholders) should be discounted at i also. Because i > > r when D

> 0, the firms debt tax shield rD/ i is smaller than MMs original tax model value of rD / in

equation (2) and considerably smaller than MMs tax correction model value of rD/r in equation

(3).

7 Note that ,rrD

+r

rD -

)-(1X

rD+

rD -

)-(1X =S

iiiwhich shows that shareholders do not discount

components of their cash flows at different rates, such as the unlevered cost of equity and riskless debt rate.

i


11/43

11

Upon setting the present value of debt interest payments equal to rD/ i in both share

valuation equation (4) and firm valuation equation (5), based on these two equations, the revised

tax model implies that the unlevered value UX(1- ) - rD rD

V = .ri

+ This important condition of

the proposed revised tax model can be written more precisely as:

U NTG NTG

X(1- ) - rD rDV = V = S D = ,

ri

+ + (5)

where NTGV = the levered value of the firm with no tax gain (denoted NTG), NTGS

= the levered

equity value of the firm in the case of no tax gain, and other notation is as before. Equation (5)

shows that MMs Proposition I with no taxes holds with taxes when there is no tax deductibility

of debt interest. This implication of the revised tax model is obviously true and is a condition

that supports its structural validity. The value of levered equity is the sum of its value with no

deductibility of interest plus the value of the tax gain available to levered shareholders, or

NTG

rDS = S .

i

+

Combining these results, the value of the firm can be alternatively written as

L NTG U

rD rDV S D S D V .

i i

= + = + + = +

Notice that shareholders tax gain equals the firms

tax gain, or rD/ i , such that there a symmetry between the shareholder and firm valuation

equations.

By contrast, according to MMs tax model, the value of the firm can be alternatively

written as L NTG UV S D S D D V D,

= + = + + = + where NTGS S D

= + . At high debt levels, it

is possible for the unlevered equity value with no debt tax gain NTGS will be negative. For

example, given UV = 7, D = 7.37, and = 0.30, we get D = 2.21, LV

= 9.21, S = LV - D = 1.84,

and NTGS = S - D = -0.37. Of course, it is not possible for equity value to be negative, as zero is


12/43

12

a lower bound due to limited liability. Hence, MMs tax model cannot be reconciled with its

form with taxes but no deductibility of taxes. This shortfall suggests that MMs tax model based

on their 1963 correction is flawed.

MM argue that the sure stream of interest tax deductions rD should be capitalized at the

riskless rate r. Since there is no bankruptcy risk, the tax savings of rD are generated for certain

and, therefore, should be discounted at the riskless rate. Unfortunately, the cash flow of the firm

is not traded as a separate asset to observe its market valuation. However, if it were traded, it

should be valued as a derivative instrument with underlying assets based on shares and debt

issued by the firm. Hence, the value of the firm is dependent on the how equity and debt cashflows are valued by shareholders and debtholders, respectively. Since shareholders value interest

tax deductions as rD/ i , this tax shield valuation must be used in valuing the firm derivative

instruments interest tax deductions. In this regard, it is not possible to create a silk purse out of

a sows ear by repackaging the cash flows of underlying assets, as proven by the recent home

securitization debacle that triggered the 2008-2010 financial crisis. The cash flows from low

quality subprime and Alt-A U.S. home loans were repackaged into mortgage-backed bonds with

inappropriate investment grade ratings, which resulted in coincident very low interest rates and

very high valuations on such bonds, and then sold into world financial markets to unsuspecting

investors. Applying this principle to MMs tax model, a small interest tax deduction value equal

to rD/ i in the underlying equity asset cannot be transformed into a large interest tax deduction

value equal to rD/r = D in the derivative firm asset due to using a riskless discount rate (i.e.,

analogous to a AAA investment grade rating) that shareholders cannot use to value their tax

shield cash flows.


13/43

13

We next derive valuation models under the consistent assumption that interest tax

deductions are discounted at the levered cost of equity in both levered equity value and firm

value models. The condition in equation (5) must hold in our valuation models. We proceed to

solve for the levered cost of equity for these joint equity/firm valuation discounting purposes by

means of the following steps. Rearranging the equity value model in equation (4), the general

expression for the levered cost of equity can be written as:

. S

)-rD(1-

VS

)V-(1X=

SrD-rD

- S

)-(1X =

S)-rD)(1-X(

= U

U

i (6)

Since levered shareholders receive debt tax shields and discount them as levered cash flows at

the levered cost of equity, we obtain the following revised tax value model of the firm:

.rD

+V=rD

+

)-(1X =

X

=V U

L ii

(7)

Using simple algebra, this model can be alternatively written (for purposes to be explained

shortly) as:

.)]

VD

(r

-1[

V =V

L

U

L

i

(8)

It is interesting that we can similarly re-write MMs tax correction model in equation (3) as

V

= V

[1 r D / V

)], which shows that a fixed debt policy in this model can be equivalently

specified in terms of a fixed debt/value ratio policy.


14/43

14

Writing firm value equation (7) as U rD/ +V=D+S i and rearranging as

U rD/ -D+S=V i , this unlevered equity value can be substituted into the numerator of the

right-hand expression in equation (6) to get the following levered cost of equity function: 8

S)D-r(1

- SV

)Dir/ -)(1-(1X +

V)-(1X

= S

)-1(r -

VS

)irD/ -D+)(S-(1X =i

U

U

U

D , (9a)

which can be simplified using the unlevered cost of equity defined as = U)/V-(1X to get

.)(D/S r

-)(D/S)]-r(1-[=

+i

i (9b)

Notice that the debt tax shield from the firm value model in equation (7) enters into the

numerator of equation (9a)s levered cost of equity. Hence, equation (9a) simultaneously

discounts debt tax shields in both firm value and equity value models at the same levered cost of

equity. 9 It is important to observe that equation (9b) collapses to MMs tax correction result for

the levered cost of equity when the riskless rate r is used in the denominator of the term

)(D/S r

iinstead of i (i.e., the interest tax deduction rD available to shareholders in the

firm value model is discounted at r), or .)r)(D/S-)(-(1+= i If the unlevered equity

rate is instead used in the denominator of the term )(D/S r

iin line with Modigliani and

Miller (1958), Miles and Ezzell (1980, 1985), Harris and Pringle (1985), Modigliani (1988),

Kaplan and Ruback (1995), and others (i.e., the interest tax deduction rD available to available

8 This procedure for deriving the levered cost of equity is based on MM (1958, 1963).

9 That is, equation (9a) can be equivalently written as

rD+

rD -

V

)rD/ -D+)(S-(1X =S

U

iii

i, which explicitly

shows interest tax deductions discounted at the levered cost of equity in both share value and firm value models.


15/43

15

to shareholders in the firm value model is discounted at ), we get ,SD

r)-(+=

i which

is the same as MMs levered cost of equity in their original tax valuation model. Thus, the MM

tax models can be readily derived from equation (9b) by discounting interest tax deductions with

different discount rates. However, these discount approaches are subject to question due to the

fact that, as shown in equation (4), levered shareholders discount levered cash flows such as

interest tax deductions using the levered cost of equity, rather than the unlevered cost of equity

or riskless debt rate. Also, as discussed in the previous section, MMs tax model suffers from

limit problems with the present value of interest deductions as riskless rates approach zero,

negative equity valuation when tax deductions are disallowed, and over estimation when

derivative valuation of firm cash flows is considered. The riskless (risky) debt rate is appropriate

for discounting riskless (risky) cash flows available to debtholders, while the unlevered cost of

equity is appropriate for discounting unlevered cash flows available to unlevered shareholders.

Since interest tax deductions available to levered shareholders arise from levered cash flows, the

levered cost of equity should be used to discount their value. Also, the proposed revised tax

model does not suffer from the aforementioned MM tax model problems associated with

diminishing riskless rates, negative equity with no interest deduction but taxes, and derivative

firm cash flow valuation.

It should be noted in equation (9b) that the levered cost of equity is both on the left- and

right-hand-sides due to its use in discounting interest tax deductions in the equity value and firm

value models, respectively. We seek to find the levered cost of equity that solves both valuation

models at a particular debt/equity ratio, or .D/S 10 To isolate the levered cost of equity in

10 Miles and Ezzell (1985) assume that the debt/equity ratio is fixed, rather than the dollar amount of debt as inMMs original and tax correction models. They show that this assumption causes interest tax deductions to have


16/43

16

equation (9b), both sides can be multiplied by the levered cost of equity and rearranged to get the

following quadratic equation:

.0=)r(D/S+)}(D/S)]-r(1-[+{-)( 2 ii (9c)

Solving for the roots of this equation, we obtain

=i [ 0.52 )}(D/Sr4-])}(D/S)]-r(1-[+{[-{+)(D/S)]-r(1-[+ ] / 2. (9d)

Given a fixed unlevered equity rate, riskless debt rate, and corporate tax rate, as well as a

particular D/S leverage ratio, the levered cost of equity can be computed using equation (9d).

The next step is to utilize the levered cost of equity solution in equation (9d) to discount

interest tax deductions in equation (7)s revised tax firm value model. However, it is not

possible to implement this step, as it is not known what amount of debt, or D, to employ in

equation (7), i.e., U

L

rD +V =V

i, that corresponds to a particular debt/equity ratio, or D/S ,

used in equation (9d) to compute the levered cost of equity. Although this impasse would seem

to make the revised tax model intractable, a simple solution exists. Since a debt/value ratio of

0.50 corresponds (of course) to a debt/equity ratio of 1 in any firm valuation model, the

relation debt/value = debt/(equity + debt) can be rearranged as debt/value =

[1/(1 + debt/equity)]debt/equity. Using this relationship between debt/value and debt/equity

ratios regardless of the valuation model being applied, the value of the firm can be computed

equity risk and, therefore, should be discounted at the unlevered equity rate (after the first period is discounted at thedebt rate). However, as forthcoming discussion demonstrates, the present setup differs from Miles and Ezzell inthat a fixed dollar amount of debt in MMs tax correction model (and revised tax model) can be associated with afixed debt/value ratio and, in turn, fixed debt/equity ratio.


17/43

17

with the alternative revised tax model of firm value in equation (8), or)]

VD

(r

-1[

V =V

L

U

L

i

.11

Hence, for any particular debt/equity ratio used to compute the levered cost of equity in equation

(9d), a corresponding debt/value ratio can be calculated and, subsequently, the value of the firm

can be obtained by means of equation (8).

We now have sufficient results to compute the value of debt and equity. Rearranging the

revised tax models firm valuation in equation (7), we can write the value of debt as follows:

).V-(Vr

=D U

L

i (10)

Equation (10) can be computed using the value of the firm per equation (8), levered cost of

equity per equation (9d), unlevered firm value, riskless debt rate, and corporate tax rate. Finally,

the value of equity can be computed as follows:

D.-V=S L (11)

In sum, the values of the firm, equity, and debt in the revised tax model are much

different from MMs original tax and tax correction models due to simultaneously discounting

interest tax deductions using the levered cost of equity discount rate in both share value and firm

value models. The MM tax models are special cases of this debt tax shield solution that fix its

discount rate at the riskless debt rate or unlevered cost of equity which are invariant to changes

in financial leverage. To the authors knowledge, no previous studies have applied the levered

cost of equity to discounting interest tax deductions within MMs corporate tax case valuation

framework. It is not known why this logically appealing valuation approach has not been

11 Hence, for any fixed dollar amount of debt, a fixed debt/value ratio and fixed debt/equity ratio is implied by theMM tax correction model. The simple relationships between these leverage variables are utilized to solve for therevised tax model.


18/43

18

investigated in extant literature. One possibility is that the above simultaneous equity value and

firm value solution may have eluded previous researchers. In the next section the effects on

firm, equity, and debt values of discounting interest tax deductions at the levered cost of equity

are comparatively examined with respect to MMs tax correction and original tax models.

2.2 Simple numerical examples

Using the revised tax model equations and following the steps outlined above, Table 1

gives simple numerical examples that compare its results to those of MMs tax correction model.

The following assumptions are made: X = 1, = 0.30, r = 0.04, = 0.10, and UV = 7 (i.e.,UV =

X (1- )/ = 1(1-0.30)/0.10 = 7). Figure 1 plots data provided in Table 1 to show that the levered

cost of equity for the revised tax model 12 (upper bold line) is somewhat higher than for MMs tax

correction model 13 (lower thin line). This result can be explained by the general expression for

the levered cost of equity in equation (4), wherein the value of equity in the denominator in the

revised tax model is lower than in MMs tax correction model due to the lower debt tax gain

available to shareholders.

A much larger disparity between these two models is apparent in Figure 2s valuation

results. The revised tax model does not yield the all-debt corner solution of MMs tax correction

model; instead, a flat-based, inverse U-shaped firm value function emerges with a maximum

firm value of 7.27 at an interior optimal debt/value ratio of about 0.55. This maximum firm

value exceeds the unlevered value of the firm equal to 7 by only 0.27, which represents a debt

tax shield of just 3.9 percent of unlevered equity value, as shown in the last column in Table 1

12 The levered cost of equity was computed using equation (9d) and then checked against the general levered cost of

equity formula in equation (5), or .)/S-rD)(1-X(= i 13 As defined in the previous section, MMs levered cost of equity in their tax correction model is

.)r)(D/S-)(-(1+= i


19/43

19

labeled rD/ i . By comparison, at a debt/value ratio of 0.55, MMs tax correction model yields

the much larger debt tax shield rD/r = D of 1.38, or 19.7 percent of unlevered equity value.

Also, the debt tax shield continues to increase and even accelerate beyond this debt/value ratio.

By comparison, the inverse U-shape of the firm value function in Figure 2 is relatively flat

throughout the range of debt/value ratios. This result is more consistent with Proposition I in

MMs (1958) original tax model, which Modigliani (1988) later believed to be the truth, than the

large tax gain in MMs (1963) tax correction paper.

If a higher debt rate is assumed, the size of the debt tax shield increases to some extent.

For example, all else the same, if the debt rate is raised to 0.08 (i.e., closer to the unlevered

equity rate of 0.10), the maximum value of the firm increases to 7.80 at an optimal debt/value

ratio of about 0.60. Thus, the debt tax shield jumps from 0.26 to 0.80, or from 3.9 percent to

11.4 percent of unlevered equity value, as the debt rate increases from 0.04 to 0.08. Figure 3

graphs the valuation results using the higher riskless debt rate of 0.08, which shows that the debt

tax shield is noticeably higher than in Figure 1 (i.e., the firm value function in more inverse U-

shaped). By contrast, the present value of debt tax shields in MMs tax correction model does

not change as riskless debt rates increase. This result seems counter-intuitive, as higher riskless

debt rates provide increasingly greater interest tax deductions that are available to shareholders,

and vice versa. The next section discusses this issue further.

What if the unlevered (rather than levered) cost of equity is used to discount all expected

after-tax earnings of the firm including interest tax deductions as in MMs original tax model? 14

14 Kaplan and Ruback (1995) dubbed discounting all capital cash flows available to both debt and equity holders atthe unlevered cost of equity the Compressed Adjusted Present Value Technique (Compressed APV). The intuitionis that capital cash flows are comprised of all after-tax cash flows, including interest tax deductions. They noted thatthis approach is equivalent to the Adjusted Present Value (APV) method of discounting interest tax deductions at the


20/43

20

Figure 4 repeats the above analyses under this assumption. The thin line represents MMs tax

correction model. The bold lines graph MMs original tax model discounting debt tax shields at

the unlevered cost of equity using riskless debt rates of 0.04 and 0.08, all else the same. At the

lower riskless debt rate of 0.04, the maximum debt tax shield is about 0.95 at near all debt, but it

increases considerably to about 2.21 at the higher debt rate of 0.08. Comparing the results of the

lower riskless debt rate of 0.04 with the revised tax model in Table 1, at an optimal debt/value

ratio of about 0.55, the value of the firm is 7.49, which implies a larger debt tax gain of 0.49

compared to 0.27 for the revised tax model. Hence, discounting interest tax deductions at the

levered cost of equity results in lower debt tax gains than discounting at the unlevered cost of equity. This difference in debt tax gains increases beyond the revised tax models optimal

debt/value ratio. Additionally, as the spread between the unlevered cost of equity and riskless

debt rate decreases (e.g., due to lower business risk for a firm or good economic times with lower

general business risk that tend to lower the unlevered cost of equity), the difference in firm

values between MMs original and tax corrected tax models decreases. Figure 4 shows that

using the higher debt rate of 0.08 yields large tax gains on interest deductions that approach

MMs tax correction model. These results show that discounting debt tax shields at the

unlevered cost of equity: (1) still implies an all-debt capital structure, and (2) does not

necessarily diminish debt tax gains to small values relative to MMs tax correction model under

low business risk conditions. Also, while unlevered cash flows should be discounted using an

unlevered equity rate, interest tax deductions arising from the issuance of debt are paid out to

levered shareholders and, therefore, should be discounted with a levered equity rate.

unlevered cost of equity. This approach assumes that interest tax deductions have the same systematic risk as thefirms unlevered cash flows, which are associated with the business risk of the firm.


21/43

21

2.3 Cost of capital

The cost of capital under the revised tax model differs from MMs tax correction model.

The weighted average cost of capital for a levered firm can be generally defined as:

WACC = i ( L /VS ) + r(1- )( LD/V ), (12)

where r(1- ) = the after-tax cost of debt. Substituting equation (9b)s levered cost of equity into

equation (12), the WACC can be written as

WACC = [ 1 -

ri

( LL /VD )]. (13)

If the levered equity rate i is replaced with the riskless debt rate r, MMs tax correction models

cost of capital, or WACC = (1 - LL /VD ), is obtained. Equation (13) implies that, as the

riskless debt rate increases, the cost of capital function becomes more U-shaped (and related firm

value function more inverse U- shaped), all else the same. Figure 5 plots the cost of capital for

the revised tax model in equation (13) versus MMs tax correction model using data from

Table 1s numerical examples. The cost of capital for MMs tax correction model falls

continuously until it reaches a relatively low minimum at all debt, where firm value is

maximized. However, the minimum cost of capital for the revised tax model is slightly U-

shaped with a minimum value of about 0.096 at an optimal debt/value ratio of about 0.55, where

firm value is maximized at 7.27. The relatively flat shape of the cost of capital function is again

consistent with Proposition I of MMs original tax model and Modiglianis later opinion. The

fact that the cost of capital does not decrease substantially as more low cost debt is employed by

the firm is attributable to discounting interest tax deductions at the levered cost of equity, which

is evident in equation (13). Increased debt results in higher interest tax deductions but a higher

levered cost of equity also.


22/43

22

3. Implications of the Revised Tax Model

The revised tax model has a number of important implications to corporate capital

structure decisions and related literature that tend to affirm its feasibility as a plausible valuation

theory. First, the revised tax model implies much smaller debt tax shields than MMs tax

correction model. In the comparative examples in Table 1, given an unlevered equity value of 7,

the maximum debt tax shield is 0.27 compared to 3 for MMs tax correction model at all debt. In

the context of the trade-off theory balancing expected tax gains on interest deductions against

expected bankruptcy costs, the revised tax model implies a fox and rabbit stew, in contrast to

Millers (1977) observation that MMs tax correction model envisages a horse and rabbit stew.Consistent with modest debt tax gains, a number of U.S. and international empirical studies (e.g.,

MacKie-Mason (1990), Rajan and Zingales (1995), Booth, Aivazian, Demirg-Kunt (2001),

Huizinga, Laeven, Nicodeme (2008), and others) confirm that the relationship between firm

leverage and taxation is relatively weak. Also, a number of studies find that debt tax benefits

have a smaller effect on firm value than MMs tax correction would predict (e.g., Fama and

French (1998), Graham (2000, 2008), and others). 15

Second, the results in Table 1 indicate that there is a fairly wide range of debt/value ratios

within which most of the present value of debt tax shields is captured. For example, when

0.30 < D/V L < 0.80, the debt tax gains available to shareholders reach levels in the range of 0.18

to 0.27 assuming r = 0.04 and in the range of 0.49 to 0.80 assuming r = 0.08. It appears that

differences in the levels of debt tax shields within this debt/value range are relatively small,

especially for lower riskless debt rates relative to unlevered equity rates. We infer that the

15 For example, Graham (2000) finds that the tax benefit of debt is about 9-10 percent of firm value in the period1980-1994. Another study by Graham (2008) estimates a range of debt tax benefits from 7.7 percent to 9.8 percentin the period 1995-1999. These estimates are comparable to the higher debt rate example for the revised tax modelin the previous section shown in Figure 3 (i.e., the tax benefit was about 0.80 at the optimal debt/value ratio or about10.3 percent of firm value).


23/43

23

revised tax model implies a relevant range for capital structure in the sense that most tax gains

available to shareholders are achieved. This implication agrees with empirical evidence and

actual practice. In this regard, Kane, Marcus, and McDonald (1984) have reported that the actual

range of debt-to-value ratios for U.S. firms is from zero to 60 percent. More recently, Frank and

Goyal (2008) have documented that the average debt-to-value ratios of the aggregate U.S.

nonfarm, nonfinancial business sector ranged from 0.27 to 0.45 in the period 1945 to early

2000s. Related to this discussion, the revised tax model implies that even moderate debt-to-

value ratios are able to capture the lions share of shareholder tax gains. Hence, it is possible that

debt-to-value ratios could vary considerably among similar firms that obtain most tax gainsavailable to shareholders an implication that helps to mitigate a previously unaddressed

problem in the static trade-off theory of capital structure observed by Myers (1984).

Third, the revised tax model implies that even firms with low probabilities of financial

distress and, in turn, low expected bankruptcy costs may employ fairly conservative debt-to-

value ratios. This implication helps to reconcile another problem inherent in the trade-off theory

raised by Myers (2001). That is, many highly profitable and big firms with investment grade

credit ratings use conservative levels of debt even though the trade-off theory predicts that such

firms should use higher levels of interest tax shields (e.g., Graham (2000, 2008), Kemsley and

Nissim (2002), and others). 16 In view of the revised tax model, even firms with no expected

16 Problems with the trade-off theory in part motivated Myers (1984) to propose the pecking order theory of capitalstructure (see also Myers and Majluf (1984)). This theory predicts that firms do not have target or optimal capitalstructures. Instead, driven by adverse selection costs derived from asymmetric information between managers andthe financial market as well as debt capacity, firms initially utilize internally generated funds, then debt, and lastlyequity in a pecking order. Another theory of capital structure with no target leverage ratio focuses on market timing.Baker and Wurgler (2002) find that firms time equity issues to take advantage of high market values, which affectfirms short- and long-run capital structures.


24/43

24

bankruptcy costs would not likely use high levels of interest tax shields. Most tax gain

advantages of debt are exhausted in the lower half of the relevant debt/value range.

Fourth, the revised tax model implies an inverse U-shaped firm value function. The

hallmark of the trade-off theory balancing debt tax gains against expected bankruptcy and other

potential debt-related costs is an inverse U-shaped firm valuation function. However, the revised

tax model yields this intuitively attractive and popular firm valuation function without these

balancing costs. It is the nonlinearly increasing levered cost of equity as debt increases that

serves as a countervailing force to offset rising interest deductions and produce the inverse U-

shaped firm value function.Fifth, returning to the impact of changes in riskless debt rates discussed in the previous

section, as riskless debt rates increase relative to a fixed unlevered cost of equity, Figure 3 shows

that debt tax shields increase and optimal debt/value ratios are somewhat higher than for lower

riskless debt rates in Figure 2, all else the same. In the limit at a zero interest rate, there is no

debt tax shield, and the value of the firm is independent of financial leverage in line with

Proposition I of MMs original no tax model. We can infer that, as riskless debt rates fall to very

low levels that approach zero, trivial debt tax shields imply that Proposition I holds for the most

part. On the other hand, as riskless debt rates increase narrowing the difference from the

unlevered cost of equity, the debt tax shield function becomes more inverse U-shaped indicative

of significant potential debt tax shields and, therefore, rejection of Proposition I. Hence, the

revised tax model may or may not reject Proposition I depending on the level of riskless debt

rates (relative to the unlevered cost of equity) that affects the shape of the firm value function.

Starkly contrasting with these riskless debt rate implications, in MMs tax correction model very

low interest rates have no effect on debt tax shields (i.e., equal to D for all positive riskless debt


25/43

25

rates). The same problem affects discounting debt tax shields with the unlevered equity rate in

the case of firms with low business risk with unlevered equity rates close to the riskless rate.

Sixth, assuming a fixed riskless debt rate, the revised tax model results in Figures 2 and 3

imply that firms with higher business risk have lower debt tax shields and optimal debt/value

ratios, all else the same. Higher business risk increases the unlevered cost of equity and,

therefore, widens its spread over the fixed riskless debt rate. Conversely, firms with lower

business risk have narrower spreads between the unlevered cost of equity and riskless debt rate,

which makes the firm value function more inverse U-shaped. Following this logic, in good

economic times with lower general business risk, the present value of the debt tax shield wouldbe more inverse U-shaped across firms, and vice versa in bad economic times (i.e., unlevered

equity rates increase in recessions and riskless debt rates tend to decrease due to stimulative

monetary policy efforts by central banks, which would increase their spread and tend to flatten

the firm value function). It can be inferred that Proposition I is more likely to hold for high than

low business risk firms and for bad as opposed to good economic times. Also, differences

between high and low business risk firms decrease as riskless debt rates decrease and are

virtually eliminated as riskless debt rates and therefore debt tax shields approach zero as

observed above.

Seventh, the debt tax shield function is affected by the corporate tax rate. As corporate

tax rates increase, the debt tax shield available to shareholders increases, all else the same. This

tax effect on corporate capital structure is consistent with empirical work by MacKie-Mason

(1990), Trezevant (1992), Graham (1996), and Graham, Lemmon, and Schallheim (1998), who

have found that high-tax-rate firms use more debt than low-tax-rate firms. Also, other papers by

Gordon and MacKie-Mason (1990), Givoly, Hahn, Ofer, and Sarig (1992), Rajan and Zingales


26/43

26

(1995), Booth, Aivazian, Demirg , Maksimovic (2001), and Huizinga, Laeven, and Nicodeme

(2008) have found that changes in corporate (and personal) tax rates lead to predicted changes in

debt ratios among firms.

Eighth, and last, a general implication based on the above discussion is that Proposition I

can approximately hold under a variety of conditions, including relatively low riskless debt rates,

high unlevered costs of equity (i.e., business risk), and low corporate tax rates, as well as

combinations of these conditions. An inverse U-shaped firm value function exists but its shape

is relatively flat under these conditions. This general implication is consistent with Modiglianis

(1988) later reversal of the 1963 MM tax correction model in favor of Proposition I in the

original 1958 MM paper. Under other conditions, Proposition I may not hold. Perhaps the most

obvious assumption to relax is riskless debt. Most firms issue debt at an interest rate in excess of

comparable maturity government debt (e.g., U.S. treasury rates). These higher debt costs would

tend to boost interest tax deductions as well as narrow the spread from unlevered equity rates

and, therefore, work against Proposition I.

Together, the aforementioned implications lend support for the revised tax model by

helping to reconcile theory and evidence on capital structure and firm valuation.

4. Personal Taxes and Miller

Millers (1977) renowned work on capital structure, firm valuation, and personal taxes

proposed dramatically lower debt tax shields compared to MMs tax correction model. This

section extends the revised tax model to the Miller valuation model incorporating both corporate

and personal taxes. Tax terms are defined as follows: ps = the personal tax rate on equity

earnings, pb = the personal tax rate on debt earnings, and = [1 - (1- )(1- ps)/(1- pb)].

According to Miller, due to equilibrium tax conditions in the bond markets, the value of the firm


27/43

27

with corporate and personal taxes is D.+V=V UL Particularly relevant to the present paper,

cash flows to shareholders are exposed to personal equity tax rates as

(X - rD)(1- )(1 - tps) X(1 - )(1 - tps) rD(1 - tps) rD(1 - tps), = + where the last term clearly shows

that the interest tax deductions are shareholders cash flow due to its association with the

personal equity tax rate. In the introduction Ross was cited for observing this linkage.

Given that is the unlevered cost of equity after both corporate and personal equity

taxes, the levered cost of equity in the revised tax model becomes

,)(D/S

r -)(D/S)]-r(1-[+=

ii

(14)

where the value of the firm used in deriving equation (14) is

.rD

+V=rD

+)ps-)(1-(1X

=X

=V UL

ii

(15)

With no personal taxes equations (14) and (15) collapse to their corporate tax only counterparts

in equations (9b) and (7), respectively. As before, the levered cost of equity can be solved using

the quadratic form of equation (14), and the value of firm can be subsequently computed by re-

writing equation (15) as

,)]

VD

(r

-1[

V =V

L

UL

i

(16)

where the D/S ratio in equation (14) corresponds to a particular D/ LV in equation (16) (i.e., as

before, debt/value = [1/(1 + debt/equity)]debt/equity). Rearranging equations (15) or (16), the

value of debt is

).V-(Vr

=D UL

i

(17)


28/43

28

Finally, the value of equity is computed as =S VL - D.

Under Millers equilibrium corporate and personal tax conditions wherein = 0, the debt

tax shield in firm value equation (16) is zero and Proposition I holds. However, if 0 for an

individual firm, then leverage changes the value of the firm. Since it is likely for individual

firms that < , previous results for the revised tax model under low tax rates are most

applicable. Thus, conditional on the assumption that the spread between the unlevered cost of

equity and debt rate is not relatively low, Proposition I generally holds for low levels of for an

individual firm but may be rejected as increases. Also, even if = due to national tax

policies (e.g., for = ps = pb or other tax regimes), Proposition I can approximately hold under

conditions discussed in the previous section. Hence, Millers introduction of personal taxes

affects MMs tax model results but much less dramatically in the proposed revised form than in

their tax correction form. Relating these revised results to published empirical work, they may

help to explain the weak effect of national tax regimes on firm leverage found in international

empirical studies by Rajan and Zingales (1995), Booth, Aivazian, Demirg , Maksimovic

(2001), Huizinga, Laeven, and Nicodeme (2008), and others.

5. Conclusion

The value of debt tax shields available to shareholders in the Modigliani-Miller (MM)

corporate valuation models is a long-standing subject of debate that is central to our

understanding of corporate finance. MM initiated the controversy by reversing Proposition I in

their original 1958 paper that posited no debt tax shields and proposing a tax correction model in

1963 that implied large debt tax shields. Subsequently, researchers have alternatively discounted

interest tax deductions using different debt and unlevered equity rates. This paper contributes to

this debate by considering the valuation implications of discounting interest tax deductions


29/43

29

available to levered shareholders at the levered cost of equity. According to Ross (2007), the

traditional approach of discounting interest tax shields using debt rates is inappropriate due to the

fact that the tax shield is paid out to shareholders rather than bondholders. Since shareholders

discount interest tax deductions at the levered cost of equity in the equity valuation model,

interest tax deductions in the firm valuation model should be discounted at the levered cost of

equity also. This revised tax model approach is not prone to a number of previously

unrecognized problems inherent in MMs tax model, including a negative equity with no interest

tax shield problem (i.e., inability to reconcile tax valuation results with and without interest tax

deductions), and a firm cash flow derivative instrument valuation problem (i.e., firm cash flowstraded as a derivative security must be valued based on how underlying debt and equity cash

flows are valued, wherein interest tax shields are discounted at the cost of equity by

equityholders).

To discount interest tax deductions at the cost of equity in both share valuation and firm

valuation models, we solve MMs system of valuation equations simultaneously to derive the

values of the firm, debt, and equity at a particular financial leverage level. We show that any

fixed debt/equity implies a fixed debt/value ratio and fixed debt level, and vice versa.

Interestingly, new valuation results from this model are much closer to Proposition I in MMs

original 1958 tax model than their 1963 corrected tax model. Also, our results appear to be more

consistent with the large body of empirical evidence on taxation and capital structure than MMs

corrected tax model. The following reasonable results are obtained: (1) relatively small debt tax

shields; (2) a flat-based, inverse U-shaped firm value function with an interior optimal capital

structure; (3) a broad relevant range of debt/value ratios within which most debt tax shields are

captured; (4) a flat-based U-shaped cost of capital function, and (5) debt tax shields and optimal


30/43

30

capital structures that are sensitive to riskless debt rates, unlevered costs of equity (i.e., business

risk), and corporate tax rates. In contrast to the latter finding, MMs tax correction model

discounting interest tax deductions at the riskless debt rate predicts no change in the value of

large debt tax shields for different riskless debt rates. This also occurs using the unlevered

equity rate to discount debt tax shields for firms with low business risk, in which case the

unlevered equity rate is close to the riskless debt rate. According to the revised taxed model,

under conditions of low riskless debt rates, low corporate tax rates (or corporate and personal tax

factors proposed by Miller), and higher unlevered costs of equity (i.e., business risk), as well as

combinations thereof, Proposition Is firm value and capital structure independence assertionholds for the most part as Modigliani (1988) retrospectively believed. However, as these

variables interact to produce higher debt tax shields that are economically meaningful to

shareholders, Proposition I may be rejected.

The revised tax model helps to explain the relatively weak relationship between financial

leverage and national tax policies reported by numerous empirical studies. Also, it addresses

some previous criticisms of the trade-off theory of capital structure by Miller (1977), Myers

(1984, 2001), and others rooted in large tax gains under MMs tax correction model relative to

expected bankruptcy costs. With respect to these criticisms, comparative static analyses of

shareholder tax gains suggest that: (1) Millers criticism of MMs tax correction model as a

horse and rabbit stew can be less severely parodied as a fox and rabbit stew; (2) it is

possible that similar firms could have widely divergent leverage ratios; and, (3) large and

profitable firms with low probabilities of financial distress may reasonably utilize fairly

conservative debt-to-value ratios. Graham (2003, 2008) has cited the latter underlevered or

conservative leverage puzzle as an unresolved issue in capital structure. In this regard, the debt


31/43

31

tax shield results of this paper do not support aggressive use of leverage even for large, profitable

firms with low business risk (even though they have greater potential debt tax shields than high

business risk firms) and, therefore, help to explain more moderate approaches to corporate debt

financing observed in the real world. Based on these findings and implications, and with

reference to Green and Hollifields (2003) criticism of the MM tax correction model noted

earlier in Section 1, MMs tax model in revised form should play a greater role in corporate

capital structure pedagogy and policy than previously believed.

Since the revised tax model predicts that the present value of debt tax shields is not large,

other factors potentially affecting firm value, including bankruptcy costs, agency costs,asymmetric information, non-debt tax shields, market timing, etc., may be important to capital

structure decisions by firms (e.g., for citations and discussion of this extensive literature, see

Graham (2003, 2008), and Hovakimian, Hovakimian, and Tehranian (2004)). How these other

factors precisely affect the inverse U-shaped debt tax shield function is left for future research.


32/43

32

References

Andrade, G., Kaplan, S. N., 1998. How costly is financial (not economic) distress? Evidence

from highly levered transactions that became distressed. Journal of Finance 53, 1443-1493.

Arzac, E. R., Glosten, L. R., 2005. A reconsideration of tax shield valuation. European

Financial Management 11, 453-461.

Baker, M., Wurgler, J., 2002. Market timing and capital structure. Journal of Finance 57, 1-32.

Booth, L., Aivazian, V. Demirg , A., Maksimovic, V., 2001. Capital structures in developing

countries. Journal of Finance 56, 87-130.

Bradley, M., Jarrell, G., Kim, E. H., 1984. On the existence of an optimal capital structure. Journal of Finance 39, 857-878.

Brealey, R.A., Myers, S. C., 2003. Principles of Corporate Finance . McGraw-Hill, New York.

Chen, A. H., Kim, E. H., 1979. Theories of corporate debt policy: A synthesis. Journal of

Finance 34, 371-384.

Cooper, I. A., Nyborg, K. G., 2006. The value of tax shields IS equal to the present value of tax

shields. Journal of Financial Economics 81, 215-225.

Damodaran, A., 1994 . Damodaran on Valuation , 2nd ed., John Wiley & Sons, New York.

DeAngelo, H., Masulis, R., 1980. Optimal capital structure under corporate and personal

taxation. Journal of Financial Economics 8, 3-29.

Fama, E., French, K., 1998. Taxes, financing costs, and firm value. Journal of Finance 53, 819-

843.

Fernandez, P., 2004. The value of tax shields is NOT equal to the present value of tax shields.

Journal of Financial Economics 73, 145-165.


33/43

33

Fernandez, P., 2007. A more realistic valuation: Adjusted present value and WACC with

constant book leverage ratio, Journal of Applied Finance 17, 13-20.

Frank, M. Z., Goyal, V. K., 2008. Trade-off and pecking order theories of debt, in B. Espen

Eckbo (ed.), Handbook of Corporate Finance: Empirical Corporate Finance ,Vol.2,

Handbooks in Finance Series, Elsevier/North-Holland, Ch. 12.

Givoly, D., Hahn, C, Ofer, A., Sarig, O., 1992. Taxes and capital structure: Evidence from

firms' response to the Tax Reform Act of 1986. Review of Financial Studies 5, 331-355.

Gordon, M. J., 1989. Corporate finance under the MM theorems. Financial Management 18,

19-28.Gordon, R. H., MacKie-Mason, J. K., 1990. Effects of the Tax Reform Act of 1986 on corporate

financial policy and organizational form, in J. Slemrod (ed.), Do Taxes Matter ?, MIT Press,

Cambridge (1990), 91-131.

Graham, J. R., 1996. Debt and the marginal tax rate. Journal of Financial Economics 41, 41-73.

Graham, J. R., 2000. How big are the tax benefits of debt? Journal of Finance 45, 1901-1941.

Graham, J. R., Harvey, C. R., 2001. The theory and practice of corporate finance: Evidence

from the field. Journal of Financial Economics 60, 187-243.

Graham, J. R., 2003. Taxes and corporate finance: A review. Review of Financial Studies 16,

1075-1128.

Graham, J. R., 2008. Taxes and corporate finance, in B. Espen Eckbo (ed.), Handbook of

Corporate Finance: Empirical Corporate Finance , Vol. 2, Handbooks in Finance Series,

Elsevier/North-Holland, Ch. 11 (2008).

Graham, J. R., Lemmon, M., Schallheim, J., 1998. Debt, leases, taxes, and the endogeneity of

corporate tax status. Journal of Finance 53, 131-162.


34/43

34

Green, R. C., Hollifield, B., 2003. The personal-tax advantages of equity. Journal of Financial

Economics 67, 175-216.

Grinblatt, M., Liu, J., 2008. Debt policy, corporate taxes, and discount rates. Journal of

Economic Theory 141, 225-254.

Grinbatt, M., Titman, S., 1998. Financial Markets and Corporate Strategy. McGraw-Hill, New

York.

Harris, R. S., Pringle, J., 1985. Risk-adjusted discount rates: Extensions from the average risk

case. Journal of Financial Research 8, 237244.

Hovakimian, A., Hovakimian, G., Tehranian, H., 2004. Determinants of target capital structures:the case for dual debt and equity issues. Journal of Financial Economics 71, 517-540.

Huizinga, H., Laeven, L., Nicodeme, G., 2008. Capital structure and international debt shifting.

Journal of Financial Economics 88, 80-118.

Kane, A., Marcus, A. J., McDonald, R. L., 1984. How big is the tax advantage of debt? Journal

of Finance 39, 841-853.

Kaplan, S., Ruback, R. S., 1995. The valuation of cash flow forecasts: An empirical

analysis. Journal of Finance 50, 1059-1093.

Kraus, A., Litzenberger, R. H., 1973. A state-preference model of optimal financial leverage.

Journal of Finance 28, 911-922.

Kemsley, D., Nissim, D., 2002. Valuation and the debt tax shield. Journal of Finance 57, 2045-

2073.

Luehrman, R., 1997. Using APV: A better tool for valuing operations. Harvard Business

Review 75, 145-154.


35/43

35

MacKie-Mason, J. K., 1990. Do taxes affect corporate finance decisions? Journal of Finance

45, 1471-1493.

Masulis, R. W., 1980. Stock repurchase by tender offer: An analysis of the causes of common

stock price changes . Journal of Finance 35, 305-319.

Miles, J., Ezzell, J. R., 1980. The weighted average cost of capital, perfect capital markets, and

project life: A clarification. Journal of Financial and Quantitative Analysis 15, 719-730.

Miles, J., Ezzell, J. R., 1985. Reformulating the tax shield: A note. Journal of Finance 40,

1485-1492.

Miller, M., 1977. Debt and taxes. Journal of Finance 32, 261-275.Miller, M., 1988. The Modigliani-Miller propositions after thirty years. Journal of Economic

Perspectives 2, 99-120.

Modigliani, F., Miller, M. H., 1958. The cost of capital, corporation finance and the theory of

investment. American Economic Review 48, 261-297.

Modigliani, F., Miller, M. H., 1963. Corporation income taxes and the cost of capital: A

correction. American Economic Review 53, 433-443.

Modigliani, F., 1988. MM -- past, present, and future. Journal of Economic Perspectives 2,

149-158.

Myers, S. C., 1974. Interactions of corporate financing and investment decisions -- Implications

for capital budgeting. Journal of Finance 29, 125.

Myers, S. C., 1984. The capital structure puzzle. Journal of Finance 34, 575-592.

Myers, S. C., 2001. Capital structure. Journal of Economic Perspectives 15, 81-102.

Myers, S. C., Majluf, N., 1984. Corporate financing and investment decisions when firms have

information the investors do not have. Journal of Financial Economics 32, 187-221.


36/43

36

Rajan, R. G., Zingales, L., 1995. What do we know about capital structure choice? some

evidence from international data. Journal of Finance 50, 1421-1460.

Ross, S. A., 2005. Capital structure and the cost of capital. Journal of Applied Finance 15, 5-23.

Ruback, R. S., 1995. A note on capital cash flow valuation. Harvard Business School Note

9-295-069, Harvard Business School, Boston.

Ruback, R. S., 2002. Capital cash flows: A simple approach to valuing risky cash flows.

Financial Management 31, 85-103.

Scott, J. H., 1976. A theory of optimal capital structure. Bell Journal of Economics 7, 33-54.

Shyam-Sunder, L., Myers, S., 1998. Testing static trade-off against pecking-order models of capital structure. Journal of Financial Economics 5, 219-244.

Stiglitz, J. E., 1969. A re-examination of the Modigliani-Miller theorem. American Economic

Review 59, 784-793.

Trezevant, R., 1992. Debt financing and tax status: Tests of the substitution effect and the tax

exhaustion hypothesis using firms responses to the Economic Recovery Tax Act of 1981.

Journal of Finance 45, 1557-1568.

Warner, J. B., 1977. Bankruptcy costs: Some evidence. Journal of Finance 32, 337-347.

Weston, J. F., 1989. What have MM wrought. Financial Management 18, 29-38.


37/43


38/43

26

Figure 1. Levered Cost of Equity: MMs Tax Correction Model (lower thin line) and theRevised Tax Model (upper bold line)

Assumptions: unlevered equity rate = 0.10, riskless debt rate = 0.04, and corporate tax rate =0.30.


39/43

27

Figure 2. Values of Debt, Equity, and the Firm: MMs Tax Correction Model (thin lines)and Revised Tax Model (bold lines)

Assumptions: unlevered value of the firm = 7, unlevered equity rate = 0.10, riskless debt rate =0.04, and corporate tax rate = 0.30.


40/43


41/43

29

Figure 4. Firm Values: MMs Tax Correction Model Discounting Debt Tax Shields at theRiskless Debt Rate (thin line) Compared to MMs Original Tax Model Discounting DebtTax Shields at the Unlevered Equity Rate with Alternative Riskless Debt Rates of r = 0.04and r = 0.08 (bold lines)

Assumptions: unlevered value of the firm = 7, unlevered equity rate = 0.10, and corporate taxrate = 0.30.


42/43

30

Figure 5. Cost of Capital: MMs Tax Correction Model (thin line) and the Revised TaxModel (bold line)

Assumptions: unlevered value of the firm = 7, unlevered equity rate = 0.10, riskless debt rate =0.04, and corporate tax rate = 0.30.


43/43

This document was created with Win2PDF available at http://www.win2pdf.com.The unregistered version of Win2PDF is for evaluation or non-commercial use only.This page will not be added after purchasing Win2PDF.
http://www.win2pdf.com/http://www.win2pdf.com/http://www.win2pdf.com/http://www.win2pdf.com/http://www.win2pdf.com/

debt tax gains kolari mwfa

Documents