This action might not be possible to undo. Are you sure you want to continue?

Welcome to Scribd! Start your free trial and access books, documents and more.Find out more

**A Theory of the Firm’s Cost of Capital
**

How Debt Aﬀects the Firm’s Risk, Value, Tax Rate and the Government’s Tax Claim

This page intentionally left blank

Tax Rate and the Government’s Tax Claim University of Texas at Austin. Value.A Theory of the Firm’s Cost of Capital How Debt Aﬀects the Firm’s Risk. USA Ramesh K S Rao Eric C Stevens USA World Scientiﬁc NEW JERSEY • LONDON • SINGAPORE • BEIJING • SHANGHAI • HONG KONG • TA I P E I • CHENNAI .

ISBN-13 978-981-256-949-3 -. 3. Corporate debt. cm. Inc. Eric C. Rao & Eric C. Danvers. S. Ramesh K. tax rate. value. 4.. may not be reproduced in any form or by any means. Ltd. II. Typeset by Stallion Press Email: enquiries@stallionpress.ISBN-10 981-256-949-9 1. Ltd. Singapore 596224 USA office: 27 Warren Street.S. . A theory of the firm’s cost of capital : how debt affects the firm’s risk. NJ 07601 UK office: 57 Shelton Street. 1962– . and the government’s tax claim / by Ramesh K. This book. including photocopying. or parts thereof. Pte. Copyright © 2007 by World Scientific Publishing Co. USA. 2. All rights reserved. 222 Rosewood Drive. Financial leverage. Includes bibliographical references. Capital costs. London WC2H 9HE Library of Congress Cataloging-in-Publication Data Rao. Stevens. In this case permission to photocopy is not required from the publisher.com Printed in Singapore.R366 2007 338. Stevens. p. recording or any information storage and retrieval system now known or to be invented. please pay a copying fee through the Copyright Clearance Center. Corporations--Finance. electronic or mechanical. Hackensack.6'04101--dc22 2006052555 British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. Suite 401-402. 5 Toh Tuck Link. Pte. Covent Garden. Title. I. without written permission from the Publisher. MA 01923. Capital assets pricing model.Published by World Scientific Publishing Co. HG4026.. For photocopying of material in this volume. 5.

the ﬁrst relating to the eﬀects of leverage on the ﬁrm risk and cost of capital. the ﬁrm’s capital structure is unimportant. 6(1). When this intuition is extended to include corporate taxes. In essence. RKS and EC Stevens (2006). This book is concerned with the ﬁrst. and it is a slightly expanded version of our paper that was published by the Berkeley Electronic Journals in Economic Analysis and Policy. Topics in Economic Analysis & Policy. The ﬁrm’s cost of capital.* Our original motivation for this research was the “pie-slicing” analogy that is the core intuition of modern corporate ﬁnance theory. Thus. v ∗ . published by Berkeley Electronic Press. available at http://www. the ﬁrm’s investment decision determines the size of the economic pie that the ﬁrm creates.bepress. it does not matter. in frictionless capital markets. and the second to the ﬁrm’s optimal capital structure (mix of debt and equity). its eﬀective marginal tax rate. and debt and equity are simply two diﬀerent claims on this pie. and the value of the government’s tax claim. how this pie is sliced. as MM argued. the size of the pie is determined by the ﬁrm’s after-tax cash ﬂows and.com/bejeap/topics/vol6/iss1/art3.Preface Modigliani and Miller’s (MM) seminal analyses spawned two broad research strands in corporate ﬁnance. thanks to the government’s Rao. This article is adapted here with permission. in this case. Article 3.

the value of the ﬁrm is consistent with the standard perspective that the ﬁrm’s after-tax output is distributed between the debt and the equity. is a framework for better measuring the ﬁrm’s cost of capital while. A second research goal. The outcome of this eﬀort. As we worked on this research. it became clear that with risky debt and corporate taxes it is critical to understand how the risks of the ﬁrm’s depreciation and the debt tax shields change with leverage.vi A Theory of the Firm’s Cost of Capital subsidy of the ﬁrm’s interest payments. and also with the view that the pre-tax output is apportioned among three risky claims. was to model how the tax shields’ risks are aﬀected by leverage. maximizing debt becomes optimal. the risk of the tax shields had not been adequately formalized in the research. with the tax authority being the third claimant. therefore. identifying the marginal eﬀects of debt policy on market values.g. The literature’s focus was on “two claims models” of the ﬁrm. there are now three claimants to the economic pie— stockholders. which is presented here. and authors have relied on various ad hoc assumptions about the tax shields’ risks. Although this intuition was well known. bondholders and the tax authority. thus. at the same time. risks. and expected rates of return. was to develop a theoretical framework that can identify how the value of the government’s tax claim varies with corporate borrowing. between the borrowing rate and the tax shields) within a simple analytical framework allows us to illustrate the model with numerical examples and graphical illustration. Thus. The ability of our model to capture several important economic interdependencies (e. in principle. Our primary goal. the model can be used to generate better estimates of the ﬁrm’s cost of capital and marginal . be able to value the ﬁrm as the sum of the values of three claims. To our knowledge. we did not see a satisfactory formal analysis of the “three claims view of the ﬁrm” with risky debt and corporate tax eﬀects. In the analysis that is presented here.. one should. With taxes. As we discuss.

interest rates. We also thank our spouses for their support. USA Eric C. We are grateful to the Berkeley Electronic Press for permission to reproduce our earlier paper in modiﬁed form. the market price of risk) on the ﬁrm’s economic balance sheet and on the value of the government’s claim on output.g. Texas. namely Juliet Lee.. Utah. tax laws. USA September 2006 . Venkatesh Sandhya. and thus may be useful for studies of tax and public policies. S.Preface vii corporate tax rates. and the colleagues that have provided feedback on various drafts of the manuscript. Ramesh K. it provides a conceptual framework for evaluating the implications of exogenous market forces (e. Rao Austin. we thank the staﬀ of World Scientiﬁc. In addition. Finally. Stevens Salt Lake City. Chean Chian Cheong and Hooi-Yean Lee for their eﬀorts at bringing this book into the present form.

This page intentionally left blank .

Introduction Model Setting Distributional Assumptions Model Solution Procedure Discussion of Results Extension to s × s states Numerical Illustration Conclusion v xi xiii 1 5 19 23 33 45 47 57 63 77 85 89 Appendix A Appendix B References Index ix . VIII. II. VII. VI. IV. V. III.Contents Preface List of Figures List of Tables I.

This page intentionally left blank .

cost of debt. Par yield (r) and the cost of debt (kD ) for the numerical examples. 5. r(1 − M T R) as proxies for kD . Risk of the tax shields for the numerical examples. 3. Cost of equity. 4. 2. kD . 6.List of Figures 1. and WACC computed using r(1 − T ). 10 38 50 52 53 54 55 xi . kE . Expected post-ﬁnancing MTR for the numerical examples. and the WACC for the numerical examples. 7. WACC from the numerical examples. Impact of an incremental debt dollar on levered ﬁrm risk and value. Output apportionment diagrams for the tax shields and claims.

This page intentionally left blank .

6. Relative magnitude of risks of the debt tax shield. 5. Post-ﬁnancing expected MTR computed as the expected value of the applicable tax rate. 8 9 25 28 30 31 31 48 70 76 xiii . tax status. A. 2. 8. Value of the levered ﬁrm and the marginal value impact of debt. and ﬁnancial solvency ˜ for diﬀerent levels of output X and of debt D in relation to assets A.1. Illustration of valuation of the depreciation tax shield. 3. Risk of the tax shields and claims for each pricing case (Table 3).2. Numerical illustration: parameters assumed. Debt pricing for the 2 × 2 model. for the 2 × 2 model with θx > 0. computed using Equation (12) and the output apportionment formulas (Table 2). Output apportionment for tax shields and claims. 7.List of Tables 1. the unlevered ﬁrm and the debt. Tax shield use. A. Pricing cases for the 2 × 2 model. 4.

78 81 . for seven debt levels. Results for the 5 × 5 example. Results for the 2 × 2 example.1.xiv A Theory of the Firm’s Cost of Capital B. B. for seven debt levels.2.

and is the sum of the products of the tax rates (tax payment divided by taxable income) in each state of nature multiplied by the relevant state probability. regulation. The ‘cost of capital’ has been incorporated into both conventional macroeconomic models and intertemporal general equilibrium models of the impacts of tax policy.1 In economics. .” 1 . 3) notes that the cost of capital “is now a standard variable in the analysis of macroeconomics and of investment behavior at the ﬁrm. The concepts of the ‘cost of capital’ and its associated measure of a ‘marginal tax rate’ have generated a voluminous literature in the economics of taxation. 2 Lau (2000. and welfare analysis. . It has also become a standard tool for the assessment of economic impacts of changes in tax policy.Chapter I Introduction The cost of capital is perhaps the most fundamental and widely used concept in ﬁnancial economics. Also see Graham (1996b) and Graham and Lemmon (1998). rate regulation. Fullerton (1984) provides a taxonomy of various deﬁnitions of the eﬀective tax rate in economics. p. holding investment ﬁxed.2 1 The MTR is the expected eﬀective tax rate on an incremental dollar of taxable income arising from debt ﬁnancing. and bankruptcy valuation. the cost of capital and the MTR are central to the research on tax policy. Business managers and regulators routinely employ estimates of the ﬁrm’s weighted average cost of capital (WACC ) and the marginal tax rate (MTR) for investment decisions. industry and economy-wide levels. restructuring activities.

and the MTR are intertwined. and several private companies (e. They do not. The tax shields’ risks and values depend on interactions between the debt and non-debt deductions (Zechner and Swoboda. At the same time. 1990). even with a ﬁxed statutory corporate tax rate. This alters r. cost of capital theory assumes riskless 3 Bulow and Summers (1984) criticize the treatment of depreciation in the extant research. due to increased default risk. Since r reﬂects the tax shields’ value. the tax shields and ﬁrm (debt plus equity) value change. . This is because the related theory has developed along two broad research strands. Increasing debt increases interest payments. 2001). Second. in turn. MackieMason. explore the link between the risk of the deprecation tax shields and ﬁrm value. not just because the ﬁrm is borrowing more. Thus. Brattle Group) generate WACC and MTR estimates for external use. the MTR may be reduced. The borrowing interest rate (coupon rate. Graham and Harvey. it inﬂuences and is. these interactions.3 Thus a circularity arises—as debt increases and r changes. the risks of the non-debt (depreciation) and debt (interest) tax shields. may change the tax shields’ magnitudes and risks. 1986). however. First. which. and they must be determined together. but not modeled. This book develops a theory of the ﬁrm’s WACC and its MTR with risky debt and potentially redundant debt and non-debt tax shields. Ibbotson Associates. increasing debt also increases the probability that some tax shields will be unusable (DeAngelo and Masulis.. the WACC. capital structure research uses state-pricing to examine the combined value of the debt and equity.2 A Theory of the Firm’s Cost of Capital A majority of ﬁrms use a single company-wide WACC for analyzing investments (Bierman. inﬂuenced by the MTR. The WACC and the MTR are endogenous to the ﬁrm’s debt policy.g. Prior research has noted. nor interdependencies between the depreciation and interest deductions. r). 1980. but also because creditors will require that each debt dollar pay a higher r. 1993. each employing a diﬀerent research strategy. pointing out that it is important to recognize the stochastic nature of tax depreciation. in turn.

. our related WACC and MTR ﬁndings have potentially important implications for low. An implication is that managers. Our model also identiﬁes the correct discount rate for valuing the tax shields and yields implications for estimation of ﬁrms’ MTR. to simplify. but taxes pose technical problems. Ross. We compare our results to standard textbook and industry cost of capital formulations that are derived from the riskless debt assumption. 1994). Second. we determine r endogenously. Collectively. can alter their debt capacity. 1978. and the theory thus implemented. and this violates the CAPM ’s assumptions). and Scholes and Wolfson (1992). The model parameters can be estimated from historical data. we assume that priced risk is the standardized covariance of returns with an exogenous factor generating economy-wide shocks and that a single-factor version of the approximate arbitrage pricing theory (APT ) holds (a later chapter elaborates).g. 1987. we do not model bankruptcy costs. Evidence indicates that acquiring external funds is not always easy (Graham and Harvey.and high-debt ﬁrms. to capture interdependencies between the tax shields’ risks and the MTR. 4 The options approach (e. to the extent that they can alter characteristics of their assets.g. 1976) admits kinked payoﬀs. and our model provides managers insights into the determinants of debt capacity. Majd and Myers (1985). These challenges are compounded with depreciation tax shields and. First.4 This research employs two innovations. This research strategy provides better cost of capital estimates. 2001).Introduction 3 debt (e. The diﬃculties associated with admitting interest tax shields in the options theory are discussed in. continuous time models typically assume zero coupon debt and abstract from depreciation (e. Following the tradition in the cost of capital theory. for example. Long (1974). Leland. 1977) and uses capital asset pricing model (CAPM ) pricing (default results in “kinked” equity payoﬀs.. Galai and Masulis. We also derive the ﬁrm’s debt capacity—the maximum that the ﬁrm can borrow irrespective of the interest rate that it is willing to oﬀer. Hite.. We ﬁnd that debt capacity depends on characteristics of the ﬁrm’s investment and on exogenous economic variables.g. Brennan and Schwartz. .

. Chapter II describes our accounting. As it turns out.4 A Theory of the Firm’s Cost of Capital The result that the ﬁrm maximizes borrowing with riskless debt and tax beneﬁts (and no bankruptcy costs) is the classic result of Modigliani and Miller (MM. the MM all-debt result is preserved. Copeland (2002) argues that for corporate ﬁnance theory to be useful to managers. Appendix A addresses technical issues. and the riskless rate) aﬀect the market values of both the private (debt and equity) and public (tax) claims. Finally. and somewhat surprisingly. The ﬁrm’s realized cash ﬂows are two ﬁrm-states. and pricing assumptions. Given the exogenous policy variables. Chapter III describes our distributional assumptions. its WACC. Chapter IV presents a four-step solution procedure that yields the cost of capital. our methodology allows us to value the government’s (tax) claim across alternative debt levels. tax. tax rules. current theory does not lend itself to such numerical speciﬁcation. Chapter V discusses results of the binomial model and provides the intuition for our ﬁndings. it is important to be able to illustrate the theory with a complete numerical example. Appendix B illustrates the eﬀect of each marginal debt dollar on the ﬁrm’s economic balance sheet. initially a joint binomial assumption (2 × 2 = 4 states) that yields analytical expressions for the risks and the costs of capital. 1963). 5 As is well known. We specify numerically how policy variables (tax rate. It also derives the ﬁrm’s debt capacity. Chapter VII contains numerical illustrations. and the MTR. The ﬁnal chapter closes.5 The book has eight chapters. We examine how this result changes with risky debt and risky tax shields. and the return on the factor generating economy-wide shocks takes on two macro-states. Chapter VI generalizes the model to s × s states. we ﬁnd that the ﬁrm will still optimally maximize debt. We consider all possible tax and default/solvency states in this single-period four-state economy and identify the risks of the tax shields and of the ﬁrm. thus providing a potentially useful conceptual framework for tax and interest policy debates.

creditors (if debt is used). As noted. for this purpose. and equity holders. considers ˜ debt ﬁnancing. we abstract from bankruptcy costs. such as licenses or patents) required to generate an uncertain end-of-period output. To be viable. we assume the ﬁrm ˜ covers its variable costs so that X is nonnegative with probability 6 The research routinely assumes that the ﬁrm is a perpetuity. Out-of-model tax shield risk assumptions also aﬀect economic linkages between the MTR and the WACC. Operations yield a t = 1 net output X (gross mar˜ gin on a cash ﬂow basis plus the liquidation value of A). but it obscures the tax eﬀects of default and necessitates outof-model assumptions about the tax shields’ risks. An investment in an amount A represents all assets (physical or otherwise. 5 . Output X is apportioned among the tax authority. and consistent with limited liability.6 The capital markets are perfect (frictionless) except for corporate taxes. The owner runs the ﬁrm to maximize personal wealth and.Chapter II Model Setting An entrepreneur/owner sets up the ﬁrm/project at t = 0 and dismantles it at t = 1. This makes the analysis tractable.

the rents) as an immediate dividend. . The interest rate is thus endogenous. (1985). he/she may pay himself/herself a t = 0 dividend DIV0 equal to his/her wealth increase. 1963) showed that the ﬁrm can borrow “up to ﬁrm value”—which. ∆W = VE − E0 . is the market value of A plus the ﬁrm’s economic rents (NPV ). The tax code speciﬁes the corporate tax rate and other tax rules. agreeing to pay interest at a rate r and to repay principal at t = 1. and bondholders will not object.e. This is also the tax rate on the capital gain on the t = 1 7 This assumption is consistent with the standard capital structure literature. The value of the tax claim is VT . r. We assume par debt. The initial investment A is fully depreciable for tax purposes at t = 1. but without personal taxes. If E0 < A. MM (1958. Zechner and Swoboda. the ﬁrm borrows A − E0 to ﬁnance A. and the coupon rate.. 1986). so that D = VD . the owners can withdraw all cash in excess of A (i. E0 = A. The owner provides E0 of initial equity (0 < E0 A) and receives a t = 1 equity cash ﬂow with a t = 0 market value of VE . The pre-tax net present value of the investment plan is: ˜ NP VA = VX − A. and the maximum they will lend (debt capacity).7 Creditors examine the investment plan and general economic variables and specify the ﬁrm’s borrowing schedule—the rates r that they require for diﬀerent borrowing amounts. When D > A. The tax rate on corporate income is T . For a ﬁrm without debt. The value of the debt claim is VD . 1985. Our accounting setup is similar to that of Green and Talmor (1985) and Talmor et al. The ﬁrm issues coupon debt with face value D. Since cash has no positive role in the ﬁrm valuation theory. The asset can be fully or partially ﬁnanced by the owner at t = 0. They are fully aware of this possibility.. the ﬁrm’s future cash ﬂows protect their claims. (1) ˜ where VX is the risk-adjusted present value of X. reﬂects this possibility.6 A Theory of the Firm’s Cost of Capital densityfX . and it may borrow more than A. The “interest ﬁrst” doctrine applies for payments to creditors in default (Talmor et al. The owner’s t = 0 wealth increase from setting up the ﬁrm is the diﬀerence between the value of his/her holdings and his/her contributed capital. by deﬁnition.

Kraus and Litzenberger. since X ∗ = D(1 + r). D. respectively. and A. ˜ Let Φi be the t = 1 cash ﬂow to i. 1980. DeAngelo and Masulis. To preclude negative taxes. rD. 1987) leads to internal inconsistencies (see Talmor et al. which implies 1−T 1−T D > A. The interest expense. ΦN T S is the cash The working assumption that both interest and principal are tax deductible (e. Depreciation is senior to the interest deduction... Depending on the debt amount. 1979. if D > A.8 The US tax code allows ﬁrms to deduct interest even on borrowing that exceeds A. D(1 + r) rD+A ⇒ D A. D(1+r)−(rD+A)T > rD + A. T (X ∗ − rD − A) 0 ⇒ X ∗ rD+A. T (X ∗ − rD − A) > 0 ⇒ X ∗ > Dr + A. 1963. 1973. we assume that taxes are paid only if taxable income is positive. Ross. 8 . The depreciation and the debt tax shield are denoted by subscripts NTS and DTS and their market values by VN T S and VDT S . 0] D(1 + r). Baron.9 Table 1 illustrates tax shield utilization and debt default/ ˜ solvency possibilities for diﬀerent levels of X. 1985. and. X ∗ . The “just solvent” or ﬁnancial breakeven level of output. the ﬁrm is solvent only if taxable income is positive. ˜ The ﬁrm is ﬁnancially solvent when output X is suﬃcient to ˜ fully pay interest and repay principal. and the output realization X. is not).Model Setting 7 liquidation value of A. If solvency occurs when taxable income is positive. 1−T If D A. and since X ∗ = D(1+r)−(rD+A)T . the ﬁnancial breakeven ˜ level of output X ∗ . if D A. Rubinstein. the borrowing amount D. when X − T · MAX ˜ [X − rD − A. 1975).g. that is. is tax deductible (the principal. 1985. 0] = D(1 + r). (2) X ∗ = D(1 + r) − (rD + A)T . the coupon rate r. The extent to which the tax shields are utilized depends on the magnitudes of A. the ﬁrm may be solvent even with negative taxable income. 1973. D. We use the following notation: ˜ ˜ ΦDT S is the cash ﬂow from the debt tax shield. Turnbull. there are two expressions for X ∗ : D(1 + r). is deﬁned by X ∗ − MAX[T (X ∗ − rD − A). 9 If solvency occurs when taxable income is negative. MM. If D > A.

Since these output apportionment formulas are well known. by D(1+r)−(rD+A)T ∗ Equation (2). ΦT is the cash ﬂow to the tax claimant. D A and hence X ∗ = D(1 + r). 1−T ˜ ﬂow from the non-debt (depreciation) tax shield. tax status. we do not discuss them further. Table 2 deﬁnes payouts from the tax shields and to the various claimants. . and ΦD+E is the cash ﬂow to the levered ﬁrm (debt plus levered equity). X* 0 A: X* NTS partially used DTS unused Taxable income < 0 Default Solvent A DTS partially used A+rD NTS fully used DTS fully used Taxable income > 0 ~ X D 0 A< X*: A NTS partially used DTS unused X* NTS fully used DTS partially used Taxable income < 0 A+rD ~ X DTS fully used Taxable income > 0 Solvent Default A< D: 0 NTS partially used DTS unused DTS partially used A A+rD NTS fully used X* ~ X DTS fully used Taxable income > 0 Solvent Taxable income < 0 Default Note: When X ∗ A or D A < X ∗. Tax shield use. ˜ ˜ ΦD is the cash ﬂow to the debt holders. ΦE is the cash ﬂow to ˜ equity. Figure 1 provides a graphical representation of the output apportionment formulas. and ﬁnancial solvency for diﬀerent levels of ˜ output X and of debt D in relation to assets A.8 A Theory of the Firm’s Cost of Capital Table 1. ΦT T S is the cash ˜ ﬂow from total tax shields. When A < D we have X = .

AT ˆ ˆ` ´ ˜˜ ˜ ˜ ΦDT S = MAX 0. X(1 − T ) + (A + rD)T 9 . ˆ ˜ ˜ ˜ ΦNT S = MIN XT. X − A − rD T ( ˆ ˜ ˜ D A : MIN X. rDT ˆ ˜ ˜ ˜ ΦT T S = MIN XT. X − X ∗ (1 − T ) ˆ ˜ ˜ ˜ ˜ ΦD+E = MIN X. MIN X − X ∗ . X(1 − T ) + AT − D(1 + r(1 − T )) ˜ ΦE = ˆ ` ´ ˜ ˜ D > A : MAX 0. X(1 − T ) + (A + rD)T. (A + rD)T ˆ ` ´ ˜ ˜ ˜ ΦT = MAX 0. Depreciation tax shield Debt tax shield Total tax shield Tax claim Debt Equity The ﬁrm Output apportionment for tax shields and claims.Model Setting Table 2. MIN X − A T. X ∗ ˜D = Φ ˆ ˜ ˜ ˜ D > A : MIN X. D(1 + r) ( ˆ ˆ ˜˜ ˜ ˜ D A : MAX 0.

10 A Theory of the Firm’s Cost of Capital Net output ~ ΦX 1 1 ~ X Depreciation tax shield ~ Φ NTS A T 1 A Figure 1. Table 2). ~ X Output apportionment diagrams for the tax shields and claims (using .

Model Setting

11

**Debt tax shield
**

~ Φ DTS

rDT

T 1

A

rD+A

~ X

**Total tax shield
**

~ Φ TTS

T(rD+A)

T 1

rD+A

Figure 1. (Continued)

~ X

12

A Theory of the Firm’s Cost of Capital

Tax claim

~ ΦT

T

1

~ X

rD+A

Figure 1. (Continued)

Model Setting

Debt

~ ΦD , D A ~ ΦD, D> A

D(1+r)

1−T

rD+A D(1+r)

1 1 1

~ X

1

1

~ X

D(1+r)

Figure 1. (Continued)

rD+A

X*

13

D > A A Theory of the Firm’s Cost of Capital 1−T A−D 1 1 1 1−T ~ X 1 ~ X D(1+r) rD+A Figure 1. (Continued) X* . D A ~ ΦE .14 Equity ~ ΦE .

(Continued) .Model Setting 15 The firm ~ Φ D+ E 1−T rD+A 1 1 1 ~ X rD+A Figure 1.

1963) showed that if the ﬁrm. If the ﬁrm can default on its debt. Kim. 1969. 1977). assumed to be a perpetuity. the tax deductibility of interest lowers the after-tax cost of debt. Rubinstein. but only when D A. the debt tax shield is fully utilized. The depreciation tax shield is equivalent to a long position in a riskless discount bond with principal TA plus a short position in 1 − T put options with ˜ strike price A. and hence the ﬁrm’s WACC. MM (1958. The debt claim is the famil˜ iar riskless bond plus short put on X with strike price D(1 + r). When D A. the depreciation tax shield (NTS ) provides ˜ a tax savings of T dollars. chooses to ﬁnance its investments with default-free debt. When A < X. For rD + A < X. When D > A. short T call options with strike price rD + A (total deductions) and short 1 − T call options with strike price X*. for each additional dollar of ˜ ˜ X in the range A < X < rD + A. To estimate the cost of capital. The payoﬀ to the ˜ ﬁrm (D + E) is equivalent to a long position in X plus short T call ˜ options on X with strike price rD + A.. For each additional dollar of X ˜ in the range 0 < X < A. generating tax savings of rDT. However. but only when D > A. X (not the ﬁrm’s assets). The “underlying asset” for all claims is ˜ ˜ net output. the ˜ debt tax shield goes unused if X < A. the early research (Hamada. the equity claim is a ˜ long call option on X with strike price D(1 + r) plus a short position ˜ in T call options on X with strike price rD + A. the debt claim is a long position in ˜ X.16 A Theory of the Firm’s Cost of Capital The apportionment formulas in Table 2 can be reinterpreted using option payoﬀ language. the depreciation tax shield is fully utilized and the cash ﬂow from it is AT. 1978) extended the MM analyses to the single-period CAPM. 1973. Because depreciation is deducted ﬁrst. the CAPM cannot be used if the debt is risky (Gonzales et al. the equity payoﬀs become “kinked” (piecewise linear) and thus inadmissible in the pricing theory. The equity claim ˜ is the familiar long position in 1 − T call options on X with strike price X*. The options . The debt tax shield (DTS ) is a bull spread on X. the debt tax shield provides a tax ˜ savings of T dollars. or long T call options with strike price A and short T call options with strike price rD + A.

uses CAPM pricing to generate risk-return metrics. Dammon and Senbet. as noted. and E(˜) = E(˜) = E(˜ε ) = 0 (this holds e ε e˜ automatically when asset returns are projected linearly onto random variable ˜). does not adequately handle tax eﬀects. Talmor et al.. 1985. as we subsequently discuss. and these studies fail to model the payoﬀs precisely. However. These papers examine the (combined) value of the debt and equity claims. computing covariance with kinked payoﬀs is diﬃcult. ˜ is a pervasive.. n×1 1×1 n×1 (3) ˜ where R is the asset returns vector. 1994) relies on the state-pricing approach. Thus. The linear projection of asset returns onto random variable ˜ is: e n×1 ˜ ˜ R =E R n×1 e ˜ + β · ˜ + ε . . We assume k is small relative to m and that m 10 The capital structure research (e. 1985. taking care to correctly model the payoﬀs arising from default and corporate taxes. β is the vector of asset sensitivities to ˜. Mauer and Triantis. we invoke the approximate APT. Lewis. economy-wide e risk factor. For this reason. where m is the number of ﬁrms and k is the number of distinct types of traded corporate claims. neither the CAPM nor the options theory is adequate for estimating the after-tax cost of capital for the (typical) ﬁrm with risky and potentially redundant debt and non-debt tax shields. 1980. Asset values are the expectation (with respect to martingale probabilities) of cash ﬂows discounted at the riskless rate. 1988. and this research cannot identify the marginal eﬀects of debt ﬁnancing on the risks and values of the ﬁrm’s claims. on the other hand. The cost of capital research.10 The economy has n traded assets. e ε ˜ n×n (4) The n assets include m × k corporate claims. The covariance matrix of returns is = β · E ˜2 · β + E(˜ · ε ). Ross. DeAngelo and Masulis. Asset values are the expectation of cash ﬂows (with respect to statistical probabilities) capitalized at the appropriate risk-adjusted discount rate. The tax shields’ risks and the ﬁrm’s cost of capital are not central to this literature. ε is the idiosyne ˜ cratic returns vector.g.Model Setting 17 approach. The random variables ˜ and ε are assumed to have ﬁnite e e ˜ variance. 1990.

provided the idiosyncratic returns correlations are suﬃciently small. re · (VAR(˜e ))−1 = βi · Vi . and the cash ﬂow beta of asset i is: ˜ ˜ r βi = COV Φi . Since the claims on the output of a particular ﬁrm are deterministic functions of its output. with equality.. (5) where E(˜e ) is the expected return on risk factor “e. the tax claim) is permissible in the linear factor framework. The existence of non-traded assets (e. (Also see Connor and Korazcyk. 1 + rz (8) (7) (6) . 1983. where Vi is the risk-adjusted present value of i: Vi = ˜ E Φi ˜ 1 + E Ri = ˜ E Φi − βi · (E(˜e ) − rz ) r . The idiosyncratic returns matrix E(˜· ε ) is thus non-diagonal. re (VAR(˜e ))−1 .18 A Theory of the Firm’s Cost of Capital is small relative to n (k m n). and that idiosyncratic returns are diversiﬁable in the sense of CR (1983). 1993.) We assume the CR conditions on the eigenvalues of as n → ∞ hold. namely. tax shields.” and rz is the r expected return on the zero-beta asset. The approximate APT pricing expression is: ˜ r E Ri ≈ rz + βi · (E(˜e − rz ). Thus. This means a single factor is pervasive. ε ˜ and exact APT pricing (Ross. kinked payoﬀs are admissible. We assume markets price all assets (including corporate claims) according to Equation (5). the idiosyncratic returns on these claims tend to be highly correlated. and the assumptions regarding the return generating process apply only to the traded assets (Grinblatt and Titman. Chamberlain and Rothschild (CR. Our assumption that k m n justiﬁes this. The returns beta (or priced risk) of asset i is: ˜ ˜ r βi = COV Ri . 1976) is precluded. 1983) show that approximate APT pricing is valid in this case. 1985).g. that has a single divergent eigenvalue as n → ∞.

in the interest of tractability we do not address this here. We later generalize the framework to the s × s (“joint s-nomial”) setting. but use numerical methods to compute the debt coupon rate.o Pe.p poo pop ppo ppp Pe. Φi .o Φi.o re.˜e with joint probability ˜ r matrix P : 11 2×2 Φi. Pp (9) A similar enumeration for an s × s model (s > 2) would yield an unmanageably large number of cases.Chapter III Distributional Assumptions We assume four states of nature at t = 1.p Po . and the return on the priced risk factor re have joint binomial probability distribution fΦi . and. where the cash ﬂow ˜ ˜ on asset i.p re. 19 11 .

ppp . .o .p (11) 14 Dividing the The state “o” risk-neutral probability for asset i is πi .p Pe. pop .12 The statistical (not risk-neutral) probabilities are poo . Pe.o −re. Pe. An alternative estimation strategy is to iterate on poo . and re.p can be estimated using maximum likelihood methods. Po = poo +pop .o . Pe. Pe.p = θi · Φi.o .o Φ −Φi.20 A Theory of the Firm’s Cost of Capital Subscripts “o” and “p” denote optimistic and pessimistic states.o +Pe. If historical data on Φi and re are available. re. re.p „ « pop poo (re. 0 <´Pe. . ppo .p re.p .o + (1 − πi ) · Φi.o − Φi.p .o −re. re. re. (10) Pe. 0 < rz . Φi. The cash ﬂow beta of asset i is13 : ˜ Bi = Φi.p = pop +ppp . pop .p πi · Φi. the parameters poo .o − re. re. re .p = 1.p to produce desired ˜ ˜ ˜ ˜ ˜ ˜ values of E Φi .o we have 0 < πi < 1 by inspection. Φi.p Equation (10) is crucial for our results since it simpliﬁes the valuation of kinked payoﬀs.p · θi .p < rz < E re < re. ppo . .o Pe. and re.o − Φi.p where θi = poo ppp − pop ppo .o − re.p (12) ` ´ We assume Φi.o − rz ) + (rz − re. Φi.p` and 0 < Pe.o −re.o − Φi. 1 + rz where πi = Po − βi = Φi. ppo . and COV Φi .p .o Pe. VAR re .o = poo +ppo .o + (1 − πi ) · Φi.o − re. and ˜ ˜ ppp .p . cash ﬂow beta in Equation (10) by the value of asset i in Equation (11) yields the returns beta of asset i: Vi = πi · Φi.p re. 0 ˜ pij 1.p Φi. VAR Φi .p θi θi · (1 + rz ) · = · .o´` ` ´` ´ 12 13 14 Bi = ˜ ˜ COV Φi re VAR re ˜ ` ´ = poo ppp −pop ppo Since πi can be written as 1 · πi = re.p ) . Pp = ppo +ppp . pop .p ´2 re. Po + Pp = 1.p Pe. re.o Φi.p ` i.o . ppp .p Φi. Vi re.o .o − re. Equation (8) and (10) yield a risk-neutral valuation (RNV) expression for asset i: θi · E re − rz ˜ .o − re. E re .o −Φi.

p i and any other asset that has the same joint binomial probability matrix as asset i. re )} = sign {Bi }. If P is symmetric about ˜ ˜ either diagonal.p = Pe. CORR(Φi .p and θi = CORR(Φi . re ).o Pe. and since these claims have the same joint binomial probability matrix P.o Pe. re ) = p Po Pp Pe. they can all be valued using θx or πx . such as a contingent claim on asset i. . The under˜ lying asset for all corporate claims is X. 15 ˜ θi is akin to the correlation between Φi and the return on the priced risk factor re : ˜ ppo ppp − pop ppo ˜ ˜ . r Po . then Po Pp Pe.p ˜ ˜ Note that sign {θi } =p sign{CORR(Φi .Distributional Assumptions 21 The risk adjustment to asset i’s state “o” probability of occurrence.15 The probability πi can be used to value asset re. is θi ·(E(˜e )−rz ) .o Pe.o −re.

This page intentionally left blank .

D. Table 1 shows that the extent to which the tax shields are utilized depends on the magnitude of the depreciable assets A. 23 . The numerical examples in a later chapter implement this solution methodology. and r. the breakeven output level X ∗ [from Equation (2)]. and this rate. The Relevant Tax States The ﬁrst step involves identiﬁcation of the relevant tax states. deﬁnes the tax states. In this chapter we focus on the procedure by which the model is solved. The tax states also depend on the borrowing rate. in turn. Step 1. This highlights the importance of endogenizing r. An interpretation of these results is deferred to the next chapter.Chapter IV Model Solution Procedure The ﬁrm’s cost of capital and the market value of all claims on the ﬁrm’s output are determined in four steps. the distribution of the ˜ output X.

D. It is useful to illustrate. with reference to a speciﬁc situation. and the value of the tax shields. The columns in Table 3 indicate where Xo falls and the rows ˜ indicate where Xp falls. and in cases 13–20 it is strictly negative. Four cases are possible: . The relevant pricing case depends on the magnitudes of the borrowing rate. ˜ Table 3 enumerates all possible distributions of binomial X across all tax. the ﬁrm. This is interesting because it establishes an economic linkage between an investment’s economic viability and the cost of capital. they serve only to ensure that the resultant cost of capital theory is consistent in every one of these cases. Our interest in this research is not on these cases per se. our analysis accommodates both positive and negative NPV ﬁrms. the investment’s pre-tax NPV. default/solvency and tax shield utilization/redundancy situations. The reader will recognize that the standard cost of capital theory. It is not easy to generalize the implications of the pricing cases in Table 3.1. in cases 9–12 it is strictly positive.24 A Theory of the Firm’s Cost of Capital Given the assumed tax rules. the debt. is silent about the signiﬁcance of the ﬁrm’s pre-tax NPV. because of the tax treatment of interest and depreciation. and the value of the ﬁrm’s physical assets. As seen from this table.1 shows that in cases 1–8 NPV A may be positive or negative. the situation where A < D. Thus. with its focus on ﬁrm value. Note that the cash ﬂows to the debt and equity depend. Appendix A. In every case X must exceed X ∗ at least in state “o” (= Xo ) so that the debt can be issued at par. Consider. for example. on the relation between the amount borrowed. and the tax claim. A. how Table 3 is used in developing our cost of capital results. in Table A. Table 2 shows how the cash ﬂows are apportioned among the various claims. Each pricing case in Table 3 implies a corresponding risk for the tax shields. there are 20 possible cases.

The rows depict Xp . NPV A+/− : Sign of NPV A may be positive or negative. Xo } across the tax and solvency states in Table 1. tax shield risk status. DTS. NTS: Depreciation (non-debt) tax shield is risky. NPVA+ rz . NTS. DTS. rD . NPV A− 14: rz . N P VA− 15: rz . NTS. NPV A− A + rD < Xo 5: rD . DTS: Debt tax shield is risky.1) are indicated for each case. DTS. NTS. X∗ X ∗ A: XP < X ∗ X ∗ Xp A A < Xp A + rD A + rD < Xp Xo A A < Xo A + rD 1: 2: 3: 4: A + rD < Xo rD . NPV A− : NPV A is negative or zero. NPV A− 18: rz . NTS. NPV A+/− 8 : rz . DTS. DTS. 25 . and the sign of NPV A (see Appendix A. DTS. NPV A+/− 6: rD . and r11 ). N P VA+/− rz . NPV A+/− 7: rz . r11 : Debt is risky and par yield = r11 . NPVA+ r11 . DTS. NPV A+/− Model Solution Procedure 19: rD . NPVA+ rD . NPV A− X∗ Xo A < D: Xp A A < Xp A + rD A + rD < Xp < X ∗ X ∗ Xp 9: 10: 11: 12: rD . NTS.˜ Table 3. NTS. DTS. N P VA− 13: rD . Twenty cases exist. NPV A− 17: rD . NPVA+ rz : Debt is riskless and par yield = rz . NPV A+/− rz . NPV A+ : NPV A is positive. Enumerates possible distributions of X = {Xp . DTSW. NTS. The par yield (see Table 5 for rz . DTS. DTSW. DTS. DTS. DTS. DTSW: Debt tax shield is worthless. rD : Debt is risky and par yield = rD . DTS. DTS. NPV A+/− X∗ D A < X ∗: Xp A A < Xp < X ∗ X ∗ Xp A + rD A + rD < Xp Xo A + rD 16: rD . the columns Xo . NTS. NTS. NPV A+/− rz . NPV A− 20: rz . Pricing cases for the 2 × 2 model.

Xp partially unused. the depreciation deduction is possible ranges. NPV A is positive in cases 9–12 and is zero . all tax shields are fully utilized. The debt is risky in cases 9–11. but after-tax cash ﬂow is insuﬃcient to fully repay debt principal. since it provides a tax shield of rDT in both states of nature. cash ﬂow from the depreciation tax shield is AT in state “o” but only XP T < AT in state “p. taxes are paid. The depreciation tax shield is riskless in cases 10–12. and after-tax funds are suﬃcient to fully repay creditors.” Xo is suﬃciently large in cases 9–12 that all tax shields are utilized.” Similarly. the debt tax shield is riskless in cases 11 and 12. all tax shields are fully utilized and the ﬁrm pays taxes. since state “p” entails partial default on interest or principal or on both. and the debt tax shield is fully redundant (completely unused).” Xp may fall in one of four A. As noted earlier. the depreciation deduction is fully taken while the debt tax shield is partially unused. in cases 9 and 10 it is risky. the ﬁrm pays taxes.26 A Theory of the Firm’s Cost of Capital Case 9: 0 Xp A A+rD X* Xo ~ X Case 10: 0 A Xp A+rD X* Xo ~ X Case 11: 0 A A+rD Xp X* Xo ~ X Case 12: 0 A A+rD X* Xp Xo ~ X In state of nature “o. In case 9. In case 11. since in state of nature “p” the cash ﬂow from the debt tax shield is (Xp − A)T < rDT (case 10) or zero (case 9). In case 9. In case 12. In state of nature “p. and debt principal is fully repaid. since the cash ﬂow from the depreciation deduction is a constant AT in ˜ both states of nature (see ΦNTS in Table 2). The debt is riskless in case 12. since creditors are fully paid in both states of nature. In case 10.

Creditors examine the investment plan (A. Three groups of cases exist for the depreciation tax shield. and is risky. and 18) the depreciation shield is a (riskless) constant AT in both states. The par yield is the coupon rate r . We next discuss details of how the debt is priced and r determined.” and is risky. and 16) the depreciation deduction is partially unused in state of nature “p. 5. The risk of the depreciation tax shields is unaﬀected by borrowing (see βN T S in Table A. The value of the non-debt tax shield (VN T S ) is computed for each group of cases using the RN V expression (11). They demand a coupon rate that equates their expected return with their opportunity cost ˜ [E(RD ) = kD ] so that VD = D. 14. NPV A may be positive or negative. if Xp < Xo . ΦD . rz ) to assess ˜ the risk of the t = 1 debt cash ﬂow. Now consider. The debt is valued in a manner similar to the depreciation tax shield except that r must be simultaneously computed such that the debt is fairly priced at par. it must be computed before valuing the debt tax shield. In the second group (cases 3. 9. Equation (12) yields the returns beta (βN T S ) for each group of cases. for purposes of illustration. In the third group (cases 19 and 20) the depreciation deduction is partly unused in both states. Cases are combined where cash ﬂows are identical. the depreciation tax shield. 10–12. Finally.1). and general economic conditions (πX . the proposed ˜ ˜ borrowing level D. and the equity claim. Table 4 illustrates the risk and valuation calculations for the depreciation tax shield. In the ﬁrst group (cases 1. 15. 6–8.rc ). First. In cases 1–8. 17. Determine the Par Yield Since r aﬀects the cash ﬂows to the tax authority and to equity. 4. This is a direct implication of the assumption that depreciation is senior to the interest deduction. 13. the tax claim. fΦX . Step 2.Model Solution Procedure 27 or negative in cases 13–20. cash ﬂow from the depreciation deduc˜ tion (ΦN T S ) is computed for states of nature “o” and “p” for each of the 20 pricing cases in Table 3. 2.

9. 20 AT −Xp T πX AT +(1−πX )Xp T · θX (1+rz ) re. 2. 17. 17.p πX ΦNTS. 5.p = = βX .AT ] 1+rz = VNT S by pricing case. AT ]. 18 19. From Table 2. 5.p Xo T −Xp T πX Xo T +(1−πX )Xp T AT −AT πX AT +(1−πX )AT =0 · θX ·(1+rz ) re. 2.p βNTS by pricing case.” “p”) 1. 6–8.o −ΦNTS. 16 3. 5.o −re. 14. 16 3.p Xo −Xp ˜ E πX (X) · θX (1+rz ) re.o −re.p = ΦNTS. 6–8. 13. 9. Cases ˜ ˜ ΦNT S. 13. 14. 18 19. 17. 10–12.p VNTS · θX re. using Equation (12) 1.o −ΦNTS. 4.o +(1−πX )ΦNTS.p · θX (1+rz ) re. 16 3. 15. 20 Cases VNTS = AT AT Xo T ˜ E πX (ΦN TS ) 1+rz Xp T AT Xp T ˜ E πX MIN [XT. 10–12. 9.28 Table 4. output apportionment formula for ˜ ˜ depreciation tax shield is ΦNT S = MIN [X T. Illustration of valuation of the depreciation tax shield. 6–8. 4. using Equation (11) 1. 14. 4. 10–12. AT ] ˜ ˜ ΦNT S.o = MIN [Xo T. 2.p = · A−Xp πX A+(1−πX )Xp θX (1+rz ) re. 18 19. 13. 15.p = MIN [Xp T.p · θX (1+rZ ) re.o −re.o −re. 15.o −re. AT ] ˜ ΦNT S by pricing case (Table 3) and by state of nature (“o. 20 Cases βNTS = πX AT +(1−πX )Xp T 1+rz πX AT +(1−πX )AT 1+rz A Theory of the Firm’s Cost of Capital = AT 1+rz πX Xo T +(1−πX )Xp T 1+rz ΦNTS.o −re.o −re.

eliminate r using the par yield expressions from Table 5. or costs of capital. 17 To express the cost of debt kD in terms of fundamental parameters. equity. tax claim and the ﬁrm (D + E). Step 4. Step 3. Merton’s (1974) debt supply analysis ignores taxes. Table A. Determine Expected Rates of Return and Market Values The approximate APT expression (5) links betas to expected rates of return. Knowing the expected cash ﬂows and the cost of capital. we can identify the risks of the tax shields and the other claims. the ﬁrm’s value is implied by Equation (11). Table 7 contains the relevant expressions for the value of the levered ﬁrm. 16 . Table 5 derives the par yield for the 20 pricing cases in Table 3. Miller’s (1977) debt supply arguments emphasize personal taxes and general equilibrium—tax shield risk is not the focus. Table 6 compares the magnitude of the debt tax shield risk to that of the unlevered ﬁrm and that of the debt. D.Model Solution Procedure 29 that solves:16 VD = ˜ E π X (ΦD (r)) =D 1 + rz (13) for a proposed debt amount. Following this procedure.1 presents the returns betas of the tax shields.17 The levered ﬁrm returns beta (βD+E ) and Equation (5) together yield the WACC. Determine the Risks of the Tax Shields and the Claims Once the par yield is determined. Substituting the returns betas from Table A.1 into Equation (5) yields the expected rates of return. and the marginal impact of debt on these betas.

8. 6. 17. 15.30 Table 5. Cases VD = ˜ E πX (ΦD (r)) 1+rz =D Par yield „ “rD ”= rz + (1 − πX) πX D − Xp D « Debt is risky. The par yield is computed for each pricing case using the output apportionment ˜ formula for debt holders ΦD (Table 2) and the RNV formula (11). 13. but no taxes are paid in state “p” (case 11) Debt is riskless (cases 2–4. 10. 18. rz . 19) Debt is risky. The risk-neutral formula is inverted for the par yield r. and the riskless rate. 16. 14. 5. 12. and ﬁrm pays taxes in states “o” and “p” (cases 1. 7. 9. Three distinct par yield expressions arise: rD . Debt pricing for the 2 × 2 model. r11 . 20) πX D(1 + r) + (1 − πX)Xp =D 1 + rz A Theory of the Firm’s Cost of Capital πX D(1 + r) + (1 − πX)(Xp (1 − T ) + (A + rD)T ) =D 1 + rz πX D(1 + r) + (1 − πX)D(1 + r) =D 1 + rz « „ D − (Xp (1 − T ) + AT ) rZ + (1 − πX) D “r11 ” = πX + (1 − πX)T rz .

6. 12) DTS worthless (19. 10) DTS riskless (4. 6. 13–18) Both DTS and D are riskless (4. 9. 12 11 13–20 Value of the levered ﬁrm and the marginal value impact of debt. 8. 5. 10 2. 5–7. 8. the unlevered ﬁrm and the debt. 3. 2. β U βDT S > β U Sign varies βDT S = 0 < β U βDT S is null βDT S vs. 10. Description (cases) Both DTS and D are risky (1. 20) Description (cases) Both DTS and D are risky (1–3.Model Solution Procedure 31 Table 6. D risky (11) DTS worthless (19. VD+E πX (Xo (1 − T ) + (A + DrD )T ) + (1 − πX)Xp 1 + rz πX (Xo (1 − T ) + (A + Drz )T ) + (1 − πX)Xp 1 + rz πX ˜ E (X)(1 − T ) + (A + Drz )T 1 + rz ˜ E πX (X)(1 − T ) + (A + Dr11 )T 1 + rz ˜ E πX (X) 1 + rz ∂VD+E ∂D πX rD T ∂rD · >0 1 + rz ∂D πX r z T >0 1 + rz rz T >0 1 + rz r11 T ∂r11 >0 · 1 + rz ∂D 0 . 9. 20) βDT S vs. 8. for the 2 × 2 model with θX > 0. 9. Relative magnitude of risks of the debt tax shield. 12) DTS riskless. βD βDT S > βD βDT S = βD = 0 βDT S = 0 < βD βDT S is null Table 7. 7. 7 4. 5. 13–18) Both DTS and D are risky (3. Cases 1. 11.

This page intentionally left blank .

as Table 5 shows. and r11 ). This is because the bondholders face increasing default risk as leverage increases. With risky debt the par yield is increasing in D (Appendix A. it permits an assessment of the risk of the tax shields. three par yield expressions emerge (rz . on the magnitude of the borrowing rate. Table 5 also shows that the magnitude of r. rD . depends on whether or not the ﬁrm pays taxes in default. When the debt is riskless the borrowing interest rate is r = rz . More interestingly.Chapter V Discussion of Results The Borrowing Rate Determines the Firm’s Tax Obligations The tax states depend. not surprisingly. the after-tax cost of risky debt. for a particular amount of risky debt. If the ﬁrm pays taxes in default state the borrowing rate is r11 (case 11). As will be seen.2). and the value of the government’s tax claim at any level of borrowing. the par yield is rD . If the ﬁrm pays zero taxes in the default state. This knowledge is important for at least two reasons. 33 .

34 A Theory of the Firm’s Cost of Capital The Cost of Risky Debt Consider the textbook question: What is the correct way to calculate the ﬁrm’s after-tax cost of risky debt when the tax shields are risky? The answer to this question has remained unclear. As long as their expected return is adequate. the creditors’ expected return on the debt. for a higher tax rate. Grinblatt and Titman. 2002. in an accounting sense. pp. the ﬁrm has incurred a net cash outﬂow of r(1 − T ) and this is the “book cost” of borrowing. although the ﬁrm pays taxes. The standard 1 − T tax adjustment is justiﬁed on the grounds that interest is tax deductible. where r is the borrowing rate and T is the ﬁrm’s tax rate. in case 11 the tax rate appears in the par yield expression: In state “p” debt principal is only partially repaid. If the argument is that a ﬁrm paying a riskless rate r has a tax shield of rT .. which is routinely used as a proxy for kD . 20 Of course. and no further tax adjustments are needed in determining the cost of debt. an alternative formulation consistent with the cost of capital theory has not been formally developed. 18 . the after-tax cost of debt is kD . For example. creditors do not care whether the ﬁrm can deduct interest payments for tax purposes. However.19 This is because the par yield aﬀects the debt’s risk (note the dependence of βD on r in Table A. then. and it depends on the expected return on alternative investments with the same risk.20 No explicit tax adjustments are needed for estimating the cost of debt. and there is no clear answer about what discount The limitations of using the standard riskless debt formulation r(1 − T ). less after-tax funds are available to repay creditors. 19 The cost of debt capital formula takes on two diﬀerent forms depending upon whether one thinks of the ﬁrm’s accounting cost of debt capital or creditors’ opportunity cost.g. r(1 − T ). tax adjustments are already included in kD . e.. 485).1) and hence creditors’ expected return. there is no consensus in the research about the tax shields’ risks. are well known (see. Thus.18 In our framework. Risk of the Tax Shields and the Firm’s Claims As is well known. p. The economic cost of debt capital is the creditors’ opportunity cost. 1988. kD . tax eﬀects are relevant to creditors. 456–459. and not the after-tax par yield. Ross et al. Intuitively.

It is diﬃcult to generalize.. and Kaplan and Ruback (1995) discount the debt shield at the unlevered cost of equity. Miles and Ezzell (1980.21 Our interest in this research is not on the tax shields’ risks per se. however. Luehrman (1997) maintains that tax shields should be discounted at a rate that lies between the cost of debt and the cost of unlevered equity. and ﬁnancing details. Taggart. 1985) assume that the risks of the ﬁrm’s total assets. Copeland et al. debt. its depreciable asset base. is that when both the debt and the interest tax shield are risky. Green and Talmor (1985).g. As can be seen. Fernandez (2004) assumes that the risk of the non-debt tax shields is the same as that of the unlevered ﬁrm and that the levered ﬁrm’s tax shields have the same risk as the ﬁrm’s equity. and the government’s tax claim) are summarized in Table A. and the debt shields are about the same. (1985). Identifying r endogenously allows us to specify the tax shields’ risks correctly.1.. less than. Contrary to the routine assumption that the debt tax shield is as risky as the debt..Discussion of Results 35 rate for the tax beneﬁt of depreciation and interest is correct (e. 1974. 21 The common view that the debt shields are as risky as the interest payments (e. 2000). the unlevered assets. 1991) is ad hoc. which compares the risk of the debt tax shields with that of the unlevered ﬁrm and the risks of the debt tax shields with the risk of the debt. Harris and Pringle (1985) argue that all tax shields should be discounted at the unlevered cost of equity. the relative magnitudes of the debt tax shield risk versus the unlevered ﬁrm risk. What we can say from Table 6. the ability to model their risks analytically allows us to better understand how borrowing aﬀects the ﬁrm’s total risk (risk of debt plus equity). and the risk of the debt relative to that of the debt tax shield risk are case speciﬁc. As seen in Table 6. Rather. Also see Talmor et al. the interest tax shield is riskier than the debt. . Myers. The risks of the ﬁrm’s other claims (equity. and Ehrhardt and Daves (2002). the risk of the interest tax shields matches that of the ﬁrm’s assets. We ﬁnd that the tax shields’ risks depend on the magnitudes and characteristics of the ﬁrm’s cash ﬂows.g. This knowledge is critical for correctly estimating the WACC and MTR. we ﬁnd that risky debt tax shields may be greater than. In Ruback (2002). or equal to the risk of both the debt and the unlevered ﬁrm. Our ﬁndings are summarized in Table 6.

θX > 0. borrowing will.1 into Equation (5). The impact of borrowing on these claims is either zero. is because r is endogenous and is a function of the ﬁrm’s tax shields. D + VE D + VE VE D · r (1 − T ) + · kE . or it depends on the sign of θX . 22 The WACC can also be obtained by substituting the levered ﬁrm returns beta βD+E from Table A. D + VE D + VE with kD . This. The sign of θX is positive (negative) when the ﬁrm’s cash ﬂows are positively (negatively) correlated with the exogenous economic shock. we have: WACC = VE D · kD + · kE . Thus. one minus the tax rate does not appear in the WACC as a multiplicative coeﬃcient on the debt coupon rate.22 The Marginal Eﬀects of Borrowing on Firm Risk and the WACC The prevailing view is that with corporate taxes (and no bankruptcy costs). This belief rests on the assumption that the debt tax shield is riskless.36 A Theory of the Firm’s Cost of Capital the return betas for these claims depend on their tax treatment. βD+E ). kE . for the typical ﬁrm. increase the risk of these claims. increasing debt lowers ﬁrm risk (the risk of D + E. Tax eﬀects are impounded into the market values and required rates of return on the debt and equity. The WACC Since no explicit tax adjustments are required to compute kD . . as in the familiar expression: WACC = Instead. One would expect that for the typical ﬁrm. and VE all non-linear functions of the tax rate. again.

then their returns betas are equal. When the total tax shield is risky.23 Thus.” the returns beta equals πX re. This is the case considered in the standard textbook result wherein debt lowers ﬁrm risk and hence the WACC. The risks of the total tax shields and of the ﬁrm increase. the interest rate on the debt rises from r to r . for a ﬁrm with θX > 0. we ﬁnd that βD+E and the WACC may fall.24 In scenario 2.1. an In Table A. which depicts three scenarios that are obtained in regard to the incremental tax shields from borrowing. an extra dollar of borrowing may increase ﬁrm risk.1 shows if the total tax shield is riskless and θX > 0. This generates an incremental tax shield in state “o. It can be shown that if in state “p. depending on the extent to which the ﬁrm can utilize the incremental deductions. The kinked lines in these ﬁgures show the cash ﬂow to D + E for a given level of operating cash ﬂows. Consider ﬁrst scenario 1. The intuition for these results is provided in Figure 2. ΦD+E . When the ﬁrm increases borrowing by $1 to a new level D . if the incremental tax shield would not be fully utilized in all states.5 helps explain why the returns betas of both the debt tax shield and the ﬁrm increase in this scenario. Table A. or even rise with leverage increases. if the ﬁrm could utilize the incremental tax shield in all states. an extra dollar of borrowing lowers ﬁrm risk. 23 . increasing debt lowers the risk of the ﬁrm. which in turn aﬀects the risk of ˜ the ﬁrm cash ﬂow.p two claims have the same joint probability matrix with the priced risk factor and their cash ﬂow distributions diﬀer only by a multiplicative constant.o − re. where A + rD lies between the “up” and “down” state operating cash ﬂows. 24 Appendix A.” and the ﬁrm’s after-tax cash ﬂow increases. but so does ﬁrm value. remain constant.Discussion of Results 37 With risky debt and potentially redundant tax shields. the marginal impact of borrowing on ﬁrm risk is either zero or has the same sign as θX . To the extent that most ﬁrms’ cash ﬂows are positively correlated with the economy (θX > 0). note that whenever cash ﬂow is positive in state “o” and zero 1 θX · (1 + rz ) · .

5–7.38 A Theory of the Firm’s Cost of Capital ~ Φ D+E ΦD+E. r r ) generates incremental tax shield in state “o. Impact of an incremental debt dollar on levered ﬁrm risk and value. ~ Φ D+E ΦD+E.” If θX > 0. 8. o ΦD+E. An incremental debt dollar generates incremental debt tax shield in both states. βT T S and βD+E increase (see Table A. o 1−T 1 Φ D+E. p 1 1 Xp A+rD A+r’D’ Xo ~ X Scenario 1. An incremental debt dollar (D increased to D and par yield may increase. If θX > 0. but so do VT T S and VD+E . 11. This situation arises in cases 1–3. p A+rD A+r’D’ Xp Xo ~ X Scenario 2.5). βD+E falls and VD+E rises.1 and Appendix A. This situation arises in cases 4. 9. and 10: Xp A + rD < Xo . . Figure 2. and 12: A + rD < Xp .

This reduces the returns beta and increases ﬁrm value. in fact. βD+E and VD+E are unaﬀected. This situation arises in cases 13–20: Xo A + rD. the ﬁrm’s risk and its WACC can. As it turns out. The incremental borrowing yields no tax beneﬁt.Discussion of Results 39 ~ Φ D+E ΦD+E. increase with borrowing. (Continued) incremental debt dollar generates tax shields in both the “o” and the “p” states. Since ﬁrm cash ﬂow is unaltered. causing levered ﬁrm cash ﬂows to increase by an equal amount. since the incremental tax shield from the debt increase is unused in both states of nature. we ﬁnd that the ﬁrm . Our framework allows us to examine how this result changes with risky debt and risky tax shields. An incremental debt dollar produces no incremental tax shield. p Xp Xo A+rD A+r D ~ X Scenario 3. In scenario 3 the debt increase does not aﬀect ﬁrm risk. This question is particularly interesting because as just seen. Figure 2. and somewhat surprisingly. o ΦD+E. The Marginal Eﬀects of Debt on Firm Value MM (1963) showed that with riskless debt and tax beneﬁts (and no bankruptcy costs) the ﬁrm maximizes debt. βD+E is unchanged.

) Our model identiﬁes the ﬁrm’s debt capacity (Appendix A. it is not an economic explanation for the borrowing limit. In scenario 2. to a ﬁrm with uneconomic investments (N P VA < 0. is increasing in D for risky debt (Appendix A. however. Dmax .40 A Theory of the Firm’s Cost of Capital will still optimally maximize debt. and that the MM all-debt result is preserved. a viable ﬁrm optimally borrows up to debt capacity. however. Scenario 3 corresponds. since it says little about the debt level or ﬁrm value at which this upper bound is reached. r can only be computed up to a certain 25 Internal optimal capital structures with corporate taxes were obtained in Brennan and Schwartz (1978) and Turnbull (1979) because the borrowing interest rate is held ﬁxed for all borrowing levels and because of bankruptcy costs. cases 13–20). (Also see Myers. however. is the highest level of D for which the debt can be fairly priced at par and at which the ex dividend equity value approaches zero from above. is a legal implication of equity’s limited liability (equity cannot have a negative value) and. the risk of the tax shields and of the ﬁrm increases with the marginal debt dollar in scenario 1.25 To see why.3)— the debt level at which it can borrow no more. Debt capacity. borrowing does not aﬀect the tax shields’ risk or ﬁrm value. but ﬁrm value also increases because the ﬁrm’s cash ﬂows increase proportionately more than enough to compensate for the increase in risk. As noted in the ﬁgure. 1977. The par yield. even when its cost of capital is increasing in the amount borrowed. In scenario 3. Thus. as noted. Debt Capacity MM (1963) showed that the ﬁrm can borrow “up to ﬁrm value.” Their argument. revisit Figure 2 and consider a ﬁrm with θX > 0. . This result is obtained because the increase in the ﬁrm’s expected cash ﬂow arising from the marginal tax shields outweighs the eﬀects of any increase in the WACC.2). debt lowers ﬁrm risk and hence increases ﬁrm value.

if N P VA ≤ 0 the ﬁrm can borrow no more than the value of its cash ﬂows. the statutory tax rate.3 and A. Turnbull (1979.Discussion of Results 41 debt amount. and the ﬁrm reaches debt capacity. the supply of additional credit evaporates. The characteristics of the investment and exogenous economic (the riskless rate and the market risk premium) and policy variables (the riskless rate. irrespective of r.g. 938) argues that the adjusted present value criterion rests on ad hoc assumptions about the functional form of the debt capacity constraint.. 26 . Appendix A. An implication is that we may be able to better understand ﬁrms’ borrowing decisions by shifting our focus away from the rightand onto the left-hand side (the assets side) of the ﬁrm’s economic balance sheet. 1974) that the ﬁrm’s limit to debt usage is exogenously given. but our credit supply side route to debt capacity links the borrowing limit to characteristics of the output distribution and exogenous economic variables. In this case. Although r increases with D. and the risk-neutral probabilities that depend on the economic variables. Our result is consistent (as it should be) with limited liability. the debt cannot be rationally priced.3) together determine this upper limit to borrowing. Myers. p. Beyond this point. It cannot borrow any more.26 If the ﬁrm’s investment is viable on a pre-tax basis (N P VA > 0) then at debt capacity the ﬁrm is tax-paying and D > A (Appendix A. the ﬁrm borrows more than the value of the physical assets generating the cash ﬂows. creditors’ expected rate of return eventually falls short of their required rate of return. The tax authority is also a It is often assumed (e.4). the cash ﬂow distribution. The “Three Claims View” of the Firm The standard theory views the ﬁrm as consisting of two claims on the output—the equity and the debt. irrespective of how high a coupon rate that it is prepared to oﬀer. just as in MM. However.

numerically. and hence to the endogenous determination of the MTR.. in principle.42 A Theory of the Firm’s Cost of Capital claimant on output (Miller. The sensitivity of interest rate and tax policy changes on ﬁrms’ economic balance sheets. with rationally 27 Various authors (e. and tax): VE +VD = (1 − T ) ·VX + VN T S + VDT S = VX − VT . We ﬁnd that for a positive N P VA ﬁrm at debt capacity. one needs a theory that can link these policy changes to the economic (market) values of the various claims.4). 1985) have noted that. debt. This is often interpreted to mean that managers are “leaving money on the table. 2000).” Our model provides an explanation for this empirical ﬁnding. the value of the tax claim is positive (Appendix A. The Value of the Tax Claim The literature routinely assumes that the ﬁrm can borrow until it altogether eliminates its tax liability. be evaluated in our model.27 Appendix B illustrates how our framework can be used to identify. Firm value is the same irrespective of whether it is assumed that after-tax output is distributed between two claims (debt and equity). Our model de facto treats the government as a third claimant. The advantage of the “three-claims” view of the ﬁrm is that it lends itself to numerical illustration. 1988). to evaluate the implications of taxes and tax policy changes. Majd and Myers. However. but this intuition has not been formalized with risky debt [Galai (1998) assumes riskless debt]. Moreover. how ﬁrms’ balance sheets change across all feasible borrowing levels given exogenous economic variables such as the corporate tax rate and the riskless interest rate. or whether the pre-tax output is assumed to be apportioned among three claims (equity. and hence inventors’ wealth can.g. this is inconsistent with the evidence that corporations do not use debt up to the point that tax shields are maximized (Graham. it permits analysis of the wealth implications of alternative tax and interest rate policy decisions. That is. .

and values of the claims have been identiﬁed. It is a numerical result that is inferred after the coupon rate. the ﬁrm cannot fully eliminate its tax obligation by issuing additional debt. the debt cannot be rationally priced. Eﬀective MTR The ﬁrm’s MTR is an implication of modeling the ﬁrm’s borrowing. the risks. the current value of the tax claim must also be positive. As a residual claim. in which case. equity cannot have zero value in a viable ﬁrm. we will discuss the MTR later. The ﬁrm must therefore have taxable income in at least one state. Otherwise.Discussion of Results 43 priced debt. . For this reason. The intuition is straightforward. in the context of our numerical examples.

This page intentionally left blank .

. Using the identity vector i = [ 1 1 · · · 1 ] .Chapter VI Extension to s × s States The probability matrix (9) can be expanded to s × s states (s > 2). .. the state probabilities of the priced risk factor are [ Pe.2 ··· ··· ··· . . Consider a joint probability matrix P containing elements pi.1 Pe. and the state cash ﬂow probabilities are [ P1 P2 · · · Ps ] = (Pi) . .2 · · · Pe.1 p11 p21 .1 re. . ps1 Pe. .j : s×s Φ1 Φ2 . Φs re.s ] = i P. With single-factor pricing.s P1 P2 . . . . ··· ··· re.2 p12 p22 . . pss Pe. Ps (14) where Φ = [ Φ1 Φ2 · · · Φs ] are the state cash ﬂows and r s×1 s×1 = [ re. ps2 Pe.s ] are the state returns on the priced risk factor. . .1 re.s p1s p2s .2 · · · re. the certainty-equivalent valuation expression [the s × s-state 45 s×1 .

Once r is computed. . the remaining claims are valued using Equation (15). Returns betas are computed using s × s-state analogs of Equations (5)–(8).46 A Theory of the Firm’s Cost of Capital analog to Equation (8)] is: iPΦ− VΦ = i P∗ [Φ − i P Φi] [r − i Pri] i [i Pr − rz ] [i P] [[r − i Pri] ∗ [r − i Pri]] . debt capacity is computed by iterating on D and r simultaneously until the s × s-state analog of Equation (13) holds with equality and the equity cash ﬂow approaches zero from above. 1 + rz where “*” denotes element-wise multiplication (Hadamard product). Using π Φ . 1 + rz where (15) s×1 [P − (Pi) (i P)] [r − i Pri] i Pr − rz . [i P] [[r − i Pri] ∗ [r − i Pri]] The term π Φ is equal to the state cash ﬂow probability vector minus a vector of risk adjustments. Collecting terms in Φ yields the RNV expression [the s × s analog to Equation (11)]: VΦ = π Φ = Pi − πΦ Φ . the par yield is computed for a given D by iterating the s × s-state analog of Equation (13) on r until the expression holds with equality.

28 Additional computed values are shown in Tables B. Figure 3 shows how the coupon rate changes with debt. and in the 5 × 5 example the debt is risky at borrowing amounts above $76.000. The corporate tax rate is T = 30%. For the 5 × 5 example the ˜ elements of the X and re vectors and the 5 × 5 probability matrix are ˜ ˜ ˜ ˜ ˜ E(˜e ). and the riskless rate is rz = 8%.719.Chapter VII Numerical Illustration This chapter illustrates the model numerically and graphically. VAR(˜e ). VAR(X). and COV (X.2. r(1 − T ) (borrowing rate multiplied by one minus the tax rate). The debt becomes risky in the 2 × 2 example at D = $46. r closely match the corresponding values in the 2 × 2 example. yields a generally biased proxy for kD . re ) r chosen such that E(X). Note that the standard deﬁnition of the after-tax cost of debt.296. Table 8 lists the parameter values assumed for the 2 × 2 and 5 × 5 examples.000 produces an ˜ uncertain X with expected value $135. Given this information. 47 .28 The initial investment of A = $100.29.30.1 and B.

A Theory of the Firm’s Cost of Capital (Continued ) .48 Table 8. Numerical illustration: parameters assumed.

6078 122. by Equation (11) Pre-tax net present value of investment plan. Initial investment. A Corporate tax rate.1869 113. VAR(˜e ) r Covariance of output with the priced risk factor. by Equation (1) 49 .55E + 3 0. ˜ E πX (X) Cash ﬂow beta of output. ˜ ˜ COV (X.0051 1. E(˜e ) r ˜ Variance of output. by Equation (10) Returns beta of output.08 2 × 2 example 135.039E + 5 2. by Equation (10) Expected output under the risk-neutral probabilities.6078 – 122.039E + 5 2.3 0. re ) Correlation of output with the priced risk factor.0051 1.275E + 9 0. ˜ ˜ CORR(X.65 13.672 0. VAR(X) Variance of the return on the priced risk factor.743.1200 1.14 3. T Expected return on a zero-beta asset. by Equation (12) Required rate of return on output.Numerical Illustration Table 8.743. BX .843.000.000 0.65 5 × 5 example 135.743.843. VX .275E + 9 0.55E + 3 0.1869 113.00 0.743.672 0. kX Value of output. E(X) Expected return on the priced risk factor. re ) Risk-adjustment term.71 13.71 ˜ Expected output.6078 0.20 3. θX . N P VA . r Computed values: (Continued) 100.12 1.000 0. βX .

000 80.000 80.000 60.000 120. When the debt is riskless.50 A Theory of the Firm’s Cost of Capital 2 x 2 example 35% 30% 25% 20% 15% 10% 5% 0% 0 20.000 120.000 100. For risky debt.000 r r(1−MTR) r(1−T) kD D Figure 3.000 40. . the par yield and the cost of debt are equal.000 100. Two proxies for kD are also shown for comparison: the par yield times one minus the tax rate (the dotted line) and the par yield times one minus the expected MTR. Par yield (r) and the cost of debt (kD ) for the numerical examples (data from Table 8 and Appendix B).000 40.000 r r(1−MTR) r(1−T) kD D 5 x 5 example 70% 60% 50% 40% 30% 20% 10% 0% 0 20.000 60. r > kD .

30 Figure 7 and Table A. the WACC increases for the 5 × 5 example from 16. Debt capacity is $110. Even with riskless debt.Numerical Illustration 51 r(1 − M T R) is closer to kD than r(1 − T ) for low debt levels but is a worse proxy for high debt levels. except at very low levels of debt.661.082 and $635 (for the 2 × 2 and 5 × 5 examples. In the 5 × 5 example. The risk of the depreciation tax shield βN T S is ﬁxed. due to our assumption about the seniority of depreciation over interest deductions. The cost of debt does not increase much because as debt increases. its contribution to the WACC is dampened by the fact that the value of the equity as a proportion of total capitalization falls rapidly. even at debt capacity.75 for the 2 × 2 example and is $113. the WACC in our model is signiﬁcantly lower than that implied by using either r(1 − T ) (the standard textbook kD proxy) or r(1 − M T R). and as D is increased r(1 − T ) ﬁrst understates and then overstates kD . respectively). r(1 − T ) = kD .22%. thus providing additional protection for the debt.29 These results have potentially important implications for low and highly-levered ﬁrms. Figure 5 plots the costs of debt and equity and the WACC for different debt amounts. 29 . which is commonly used in the industry as a proxy for kD . The WACC does not increase much. because as debt increases.41).108.22% at debt capacity (D = $113.108. 30 At debt capacity the tax claim is worth $3. the value of the tax shields also increases.56% (at D = 0) to 18.3. The WACC. although the cost of equity increases rapidly. Figure 6 shows that in highly leveraged ﬁrms. The value of the tax claim is positive. The risk of the debt tax shield is relatively insensitive to borrowing at low debt levels but rises dramatically at high debt levels.40 in the 5 × 5 example. The formulas for computing debt capacity are presented in Appendix A.56% to 18. Figure 4 shows how the risk of the tax shields change with debt.2 show the Although not apparent in Figure 5. the WACC increases with debt from 16. can be signiﬁcantly aﬀected by how one deﬁnes the ﬁrm’s cost of debt.

βDT S = debt tax shield.000 100.000 βDTS β TTS βNTS D Figure 4.000 40.000 100.000 80. βNT S = depreciation tax shield.52 A Theory of the Firm’s Cost of Capital 2 x 2 example 5 βDTS 4 3 βTTS 2 βNTS 1 0 0 20.000 60.000 120. and βT T S = total tax shield. .000 120.000 D 5 x 5 example 10 9 8 7 6 5 4 3 2 1 0 0 20.000 60.000 40.000 80. Risk of the tax shields for the numerical examples (data from Appendix B).

Numerical Illustration

53

2 x 2 example

30%

25%

kE

20%

WACC kD

15%

10%

5%

0% 0 20,000 40,000 60,000 80,000 100,000 120,000

D

5 x 5 example

120%

100%

80%

60%

40%

kE

20%

WACC

kD

80,000 100,000 120,000

0% 0 20,000 40,000 60,000

D

Figure 5. Cost of equity, kE , cost of debt, kD , and the WACC for the numerical examples (data from Appendix B).

54

A Theory of the Firm’s Cost of Capital

2 x 2 example

25%

WACC using r(1−MTR) WACC using r(1−T)

20%

WACC using kD

15%

10%

5%

0% 0 20,000 40,000 60,000 80,000 100,000 120,000

D

5 x 5 example

60%

**WACC using r(1−MTR)
**

50%

40%

**WACC using r(1−T)
**

30%

20%

WACC using kD

10%

0% 0 20,000 40,000 60,000 80,000 100,000 120,000

D

Figure 6. WACC from the numerical examples, and WACC computed using r(1 − T ), r(1 − M T R) as proxies for kD (data from Table A.2 and Appendix B).

Numerical Illustration

55

30%

2 x 2 Example

25% 20% 15% 10% 5% 0% 0 20,000 40,000 60,000 D 80,000 100,000 120,000

5 x 5 Example

Figure 7. Expected post-ﬁnancing MTRs for the numerical examples. Data from Table 8 and Table A.2.

expected post-ﬁnancing MTR for various debt amounts. The MTR is computed (using the state probabilities in Table 8) as the sum of the products of applicable tax rates (tax payment divided by taxable income) in each state of nature, times the state probabilities. In the 2 × 2 example, since taxable income is negative in state “p” for any debt amount, the applicable tax rate in that state is zero, hence the marginal expected tax rate is less than the statutory tax rate of 30%. Multiplying the relevant binomial probabilities (Po = 0.85 and Po = 0.15 from Table 8) and summing yield the 25.5% marginal expected tax rate. In the 5 × 5 example, for low debt levels taxes are paid in all states except one; as debt is increased, taxes paid fall to zero in additional states, causing the MTR to drop from 23% to 7%.31

31 Graham (1996a) computes the expected MTR as the present value of taxes paid on an incremental dollar of taxable income using multi-period Monte Carlo simulation, an increasing tax rate schedule, tax loss carry forwards, investment tax credits, and the alternative minimum tax. Since ours is a one-period model, for simplicity we do not discount the tax payment when computing the expected marginal tax rate, though we could have, since we have the proper discount rate, kT .

provides cost of capital estimates using MTR estimates generated using Graham’s (1996a) Monte Carlo methodology. . In our 5 × 5 example.56 A Theory of the Firm’s Cost of Capital Ibbotson Associates. we ﬁnd that the discount rate for the tax claim exceeds the coupon rate32 . in its Cost of Capital Yearbook. and he discounts tax payments using the average corporate debt yield. Graham does not model the discount rate for valuing the tax shields. the industry approach may thus be underestimating the true discount rate for the tax payments. 32 In Appendix B. in the 5 × 5 example we have kT > r for all debt levels.

Chapter VIII Conclusion How ﬁrms estimate their WACC and MTR is no trivial matter. Rosen (1998) argues that the cost of capital aﬀects entrepreneurs’ investment decisions. For this reason. 57 .3.. through the WACC and MTR. Also see Auerbach (1983) and Auerbach and Hassett (1992). Our methodology integrates the 33 Congressional tax policy analysis presumes that the average corporate tax liability and the present value of tax depreciation are two of four critical factors aﬀecting investment [“Joint Committee on Taxation. Reinschmidt (2002) discusses how the cost of capital is used for public sector project analysis. 1998). the evidence reviewed in Caballero (1999) corroborates the long-term link between investment and cost of capital. Consistent with this. we have developed a cost of capital theory with risky debt and corporate taxes. ﬁrms’ investment decisions.33 In this research. In a diﬀerent context.” US Congress. A few percent in capital costs can mean a swing in billions of expenditure dollars (Bruner et al. Overview of Work of the Staﬀ of the Joint Committee on Taxation to Model the Macroeconomic Eﬀects of Proposed Tax Legislation to Comply with House Rules XIII. US policy makers evaluate alternative tax policies by studying how they can inﬂuence.(h)(2)(JCX105-03). 2003].

58 A Theory of the Firm’s Cost of Capital state pricing and beta pricing frameworks in a single-period approximate APT model. coupon rate) as a function of debt. The net result of this integration is an internally consistent WACC with corporate taxes. and (v) yields the expected post-ﬁnancing MTR. expected rates of return. (iii) identiﬁes the risks. generalizations about tax shields’ risks are diﬃcult. numerically. no further tax adjustment is necessary. This is not surprising but. debt. With this integration. We have also shown that the standard textbook deﬁnition of the after-tax cost of debt (r multiplied by one minus the tax rate) and the formula used in industry (r multiplied by one minus the MTR) are both biased estimates of the cost of debt. and the government’s claims. we have shown that the creditors’ opportunity cost already impounds the implications of tax eﬀects. and tax) on the ﬁrm is invariant to the valuation methodology employed—it is the same using either the martingale (state pricing) or the risk-adjusted discount rate (beta pricing) approach. and values of the tax shields and of the debt. . the value of any claim (equity. it identiﬁes. the equity. several interesting and useful results are obtained. more interestingly. and captures interactions between the WACC and the MTR. the model (i) generates the ﬁrm’s borrowing schedule—the borrowing rate (r. and possibly redundant tax shields. risky debt. and a case-by-case analysis is required. Precise statements require knowledge of the magnitudes of the various ﬁrmspeciﬁc and economy-wide parameters. The ﬁrm’s after-tax cost of debt is the creditors’ opportunity cost. (ii) identiﬁes debt capacity (the maximum feasible borrowing) in terms of the ﬁrm’s investment characteristics. Speciﬁcally. given details of the ﬁrm’s investment plan and exogenous economic variables. the marginal eﬀects of debt for any feasible borrowing level. Nevertheless. The model yields a “fully speciﬁed” cost of capital theory in the sense that. When the debt is risky and the ﬁrm has potentially redundant tax shields. (iv) shows how the WACC changes with borrowing.

Our analysis shows that the ﬁrm’s debt capacity is determined by characteristics of its investment and exogenous economic variables. decrease. again. . Contrary to the riskless debt case wherein the ﬁrm’s risk and WACC decrease with leverage. In our numerical examples. Speciﬁcally. also impounded into our WACC formula. the latter formulation is closer to the true cost of debt than the textbook deﬁnition for low debt levels. and ﬁnd that this rate exceeds the coupon rate for all debt amounts. Which of these situations are obtained depends on the relevant magnitudes in the analysis. tax loss carry forwards. through the cost of debt. and the extent to which the tax shields are usable in each state.34 His empirical methodology discounts the tax payment using the average corporate debt yield. Graham (1996a) uses Monte Carlo methodology to estimate ﬁrms’ MTRs.Conclusion 59 Tax eﬀects are. we have shown that with risky debt the WACC may increase. We ﬁnd that in high-debt ﬁrms the WACC is signiﬁcantly lower than that implied using both the textbook and the industry proxies for the cost of debt. We determine the correct discount rate for tax payments. These results have potentially important implications for valuation and other analyses of low and highly-levered ﬁrms. exogenous economic factors (riskless rate and the market risk premium and the risk-neutral probabilities that depend on the economic variables) and policy variables 34 Graham’s methodology accommodates multi-period tax code features (an increasing tax rate schedule. no multiplicative tax adjustment (one minus the tax rate) is necessary. or remain constant. Ibbotson Associates provides cost of capital estimates (these are reported in their Cost of Capital Yearbook) using MTR estimates generated with the Graham methodology. and the alternative minimum tax) but he does not model the discount rate for the tax shields. the cash ﬂow distribution. investment tax credits. We have provided graphical intuition for these results. but is a worse proxy for high debt levels. and our ﬁndings suggest that these industry estimates may underestimate the true discount rate.

how each debt dollar aﬀects the entire economic balance sheet. and hence aﬀect appropriate macroeconomic policies. However.35 The need to extend the model analytically to a multi-period setting and to admit dynamic features of the tax code is clear. and tax and interest rate details. Our analytic tools are going to have to increasingly focus on changes in asset values and resulting balance sheet variations if we are to understand these important economic forces. that even with risky debt and possibly redundant tax shields. the ﬁrm optimally maximizes leverage. are going to have to be able to ascertain how changes in the balance sheets of economic actors inﬂuence real economic activity.” (New Challenges for Monetary Policy. 1999). Central bankers. August 27. .3) together determine debt capacity. exogenous economic factors. the evolution of alternative organizational forms. surprisingly. numerically. This result holds with risky debt because the increase in the ﬁrm’s expected cash ﬂow arising from the marginal tax shields outweighs the eﬀects of any increase in the WACC. Appendix A. A distinctive feature of our model is that it shows. compensation structures in the economy. even when its cost of capital is increasing in the amount borrowed. We ﬁnd. and business reorganizations. the statutory tax rate. The model shows how leverage alters not just the value of the debt and equity. in particular. Also see Scholes and Wolfson (1992) for a discussion of how tax and interest rate polices can have far-reaching implications for pension planning. An implication is that we may be able to better understand ﬁrms’ capital structure decisions by shifting our focus onto the left-hand side (the assets side) of ﬁrms’ economic balance sheets. FRB of Kansas City Symposium. retirement planning. but also the government’s tax claim. The marginal eﬀects of borrowing depend on the ﬁrm’s investment characteristics.60 A Theory of the Firm’s Cost of Capital (the riskless rate. Jackson Hole. . 35 Alan Greenspan has recently articulated the need for this: “. The ﬁrm optimally borrows up to debt capacity. . and the MM “all debt” result is sustained. It thus provides a potentially useful tool for examining the relation between interest rate and tax policy changes and resultant changes in ﬁrms’ economic balance sheets.

. The single-period APT is inadequate. tax credits. Monte Carlo methods) to a multi-period setting can enrich the framework to accommodate tax loss carry forwards. The researcher attempting this generalization must model the ﬁrm’s borrowing interest rate in a multi-period world wherein the coupon rate will be determined endogenously by dynamic features of the tax code. and oﬀers the potential for a better understanding of the ﬁrm’s ﬁnancing and investment decisions. perhaps. a multi-period formulation will be required. Nevertheless.Conclusion 61 this is a formidable task. personal taxes. subsequent reﬁnements and extensions of this model (with. and bankruptcy costs.

This page intentionally left blank .

16. 15. DrD + A. In cases 14. Substituting + A.1 Sign of NPV A Consider the pricing cases in Table 3. and since D VX we have A < VX . D for rD . D A and Xo Drz + A. and 18. and 17. Now Xp Xo ⇒ (1 − πX )Xp (1 − πX )Xo ⇒ πX Xo + (1 − πX )Xp ˜ ⇒ E πX (X) Drz + A Arz + A ⇒ VX Arz + A = A ⇒ NPVA 1 + rz 0. Xo In cases 13. hence NPVA > 0 by Equation (1). Xo D πX A and Xo rz + (1 − πX ) 63 D − Xp D . the sign of NPVA may vary. Cases 9–12 involve A < D. For cases 1–8.Appendix A Discussion of Various Results A.

3 Debt Capacity D can be increased until X ∗ = Xo . 0. We . 1 − πX ∂rD = sign · Xp > 0. and the cash ﬂow to equity approaches zero from above. At that point. with riskless debt. the par yield is constant. the par yield is increasing in D when the debt is risky. making creditors owners ex ante.64 A Theory of the Firm’s Cost of Capital Multiplying by πX and adding (1 − πX )Xp . ∂D sign > 0. From Table 3. 0. D cannot be increased further. Xo < A. X ∗ = Xo can occur only in cases 9–12 for an NPVA > 0 ﬁrm or in cases 16–20 for an NPVA 0 ﬁrm. ∂D πX 1 − πX ∂r11 = sign · (Xp (1 − T ) + AT ) sign ∂D πX + (1 − πX ) T and ∂rZ = 0. since the equity cash ﬂow would then be zero in both states. state “o” output is just suﬃcient to fully repay creditors. A. Since Xp ˜ E πX (X) X0 ⇒ ˜ E πX (X) 1 + rz Xo and πX 1.2 The Par Yield is Non-Decreasing in D From the par yield expressions (Table 5). πX Xo + (1 − πX )Xp D(1 + rz − πX ) + πX A πX (A − D) ⇒ VX D + 1 + rz πX πX A+ 1− D A ⇒ NPVA ⇒ VX 1 + rz 1 + rz In cases 19 and 20. Xo < A ⇒ VX < A ⇒ NPVA A. That is.

and .9−10 (1 − T ) The debt is priced as DMAX .9−10 + (1 − πX ) Xp .11 = DMAX .11 (1 − T ) − AT = Xo 1−T Xo (1 − T ) + AT − DMAX . 1−T For cases 9 and 10.11 + (1 − πX ) Xp (1 − T ) + A + DMAX . Xp < D(1 + rD ) = Xo . In case 11.9−10 = . DMAX . and substituting for rDMAX .11 (1 − T ) The debt is priced as DMAX . In case 12.9−10 = πX DMAX . In cases 16 and 17. at debt capacity we have Xp < DrD + A < X ∗ = Xo .9−10 . at debt capacity we have Xp < Dr11 + A < X ∗ = Xo . ⇒ rDMAX . X∗ = DMAX .9−10 (1 − T ) − AT = Xo 1−T Xo (1 − T ) + AT − DMAX . X ∗ or 11 is obtained. 10.9−10 . ⇒ rDMAX .Appendix A 65 D (1 + r) + T (D−A) > Dr + D > Dr + A). DMAX .11 rDMAX .11 = “ ” ” ” “ “ πX DMAX . Now DMAX .11 = ˜ E πX (X) + AT 1−T + (1 − πX )(Xo − Xp )T T 1−T 1 + rz + . and D can be increased until cases 9.11 . Xp .11 .9−10 1 + rDMAX .11 T 1 + rz .11 1 + rDMAX .9−10 = DMAX . 1 + rz ˜ E πX (X) + 1 + rz + πX AT 1−T πX T 1−T consider these nine cases.11 1 + rDMAX . and utilize the results that X ∗ is increasing in D and A < D ⇒ Dr +A < X ∗ (note that X ∗ = D(1+r(1−T ))−AT = 1−T and substituting for rDMAX .9−10 1 + rDMAX .

Case 19 involves Xp < D(1 + rD ) Xo . DMAX .11 . . 1 + rz In case 20. or 19. Xp Substituting for DMAX .19 = πX DMAX . the debt is priced as DMAX . consider whether case 9.16−17 = πX DMAX . DMAX .66 A Theory of the Firm’s Cost of Capital the debt is priced as DMAX . for a ﬁrm with NPVA and occurs in case 16. Finally.11 · rDMAX . (Xo − Xp )(1 − T ) Hence (Xo − Xp ) 1 − T + πX T 1−T ˜ E πX (X) − A = NPVA .9−10 (case 9. 10. 0.16−17 (1 + rD ) + (1 − πX )Xp 1 + rz πX Xo + (1 − πX )Xp = = VX . Substituting for In case 11. 10) or DMAX . When Xp < D(1 + rD ) = Xo . For a ﬁrm with NPVA > 0. 1 + rz Xp Xo A + rz D. D(1 + rz ) increased until case 16 or 17 is obtained. D(1 + rz ) Xp Xo . debt capacity is VX Summarizing. 1 + rz ˜ E πX (X) + AT 1−T + (1 − πX )(Xo − Xp )T T 1−T 1 + rz + − A.11 − A.11 + A rDMAX .11 (case 11). 17. debt capacity is DMAX . Xp .19 (1 + rD ) + (1 − πX )Xp 1 + rz πX Xo + (1 − πX )Xp = = VX .11 Xo (1 − T ) + AT − DMAX . and D can be In case 18. and D can be increased until case 19 is obtained. or 11 is obtained for the NPVA > 0 ﬁrm.11 (1 − T ) ⇒ (Xo − Xp )(1 − T ) DMAX .11 +A DMAX .11 .

A. which DMAX expression is obtained depends on the magnitude of NPV A : DMAX ˜ E πX (X)(1 − T ) + πX AT D MAX . 1 + rz (1 − T ) − (1 − πX )T if 0 < NPVA < ((1 + rz )(1 − T ) + πX T )(Xo − Xp ) .9−10 Xo (1 − T ) + AT − DMAX . 1 + rz πX ˜ = DMAX . if 1 + rz DMAX .11 = E (X)(1 − T ) + AT + (1 − πX )(Xo − Xp )T (1 − T ) .17.9−10 (1 − T ) ⇒ DMAX . so taxes of .16.19 = VX . 1 + rz (1 − T ) ((1 + rz )(1 − T ) + πX T )(Xo − Xp ) NPVA . Xp < DMAX . output exceeds available deductions in state “o” (Dr + A < X ∗ = Xo ).9−10 DMAX .9−10 . Substituting for rDMAX . VT > 0 For an NPVA > 0 ﬁrm at debt capacity.9−10 = .9−10 + A.Appendix A 67 In cases 9 and 10. Xp < DMAX .9−10 . ˜ E πX (X) + 1 + rz + Hence (Xo − Xp ) 1 − T + πX T 1−T > ˜ E πX (X) − A = NPVA .9−10 − A < (Xo − Xp )(1 − T ). if NPVA 0 .4 For an NP VA > 0 Firm at Debt Capacity. In summary. +A Substituting for DMAX .9−10 · rDMAX . 1 + rz πX AT 1−T πX T 1−T − A < (Xo − Xp )(1 − T ).

p + δ) · βi = πi (Φi.p θi (1 + rz ) · βi = πi (Φi.p Φi.p ∂Φi.p ∂βi ⇒ sign = sign{Φi.o − re.o − re.o + δ − Φi.o θi ∂Bi ⇒ re.o + δ) + (1 − πi )(Φi. A. Bi = ((Φi.p ) e.o + δ) + (1 − πi )(Φi.o − Φi.o − (Φi.p .p ∂βi θi (1 + rz ) = ⇒ · 2 r ∂δ (πi (Φi.o + δ) + (1 − πi )Φi.o + δ) − Φi. Bi = ((Φi. ˜ ∂δ ∂ Φi θi ∂Bi ⇒ re.p −(Φi. = re.o + δ) − (Φi.p ∂δ ∂Bi θi ⇒ sign = sign{θi }.o − re. ∂Φi.o Φi. we have VT > 0 by Equation (11).p ) ∂βi θi (1 + rz ) = ⇒ · 2 r ∂δ (πi (Φi.p re.p + δ)) · θi ∂Bi = 0.p ∂δ θi (1 + rz ) (Φi.o − re.p ∂δ ∂Bi ⇒ sign = −sign{θi }.5 Sensitivity of Bi and βi to an Increase ˜ in Φi ˜ If Φi is incrementally increased by δ > 0 in both states.o + δ) + (1 − πi )Φi.p If only the state “o” cash ﬂow is incrementally increased. Bi = (Φi.o − re.p + δ) re.p · θi }.o − re.68 A Theory of the Firm’s Cost of Capital (Xo − Dr − A)T > 0 are paid in that state. Since πX > 0.o + δ) − (Φi.o − re.o − re.p ) · θi } .p ∂βi ∂βi = sign ⇒ sign = −sign{(Φi. ⇒ re.o − re.p + δ)) e.p ) · If only the state “p” cash ﬂow is incrementally increased.o − Φi. ∂Φi.p + δ)) · = −θi re.

p + δ) re. θX > 0.p ∂βi ⇒ sign = −sign{Φi.o + (1 − πi )(Φi.o and Φi.Appendix A 69 βi = θi (1 + rz ) Φi.p only sign ∂Bi ˜ ∂ Φi 0 + − sign ∂βi ˜ ∂ Φi − + − For the speciﬁc case of Φi.o · θi }.p + δ)) e.p = 0 < Φi.o ∂βi θi (1 + rz ) = ⇒ · 2 r ∂δ (πi Φi.o only Φi. Cash ﬂow(s) incremented: Φi.p − δ · πi Φi.o only Φi.o .o and Φi.o − re.p Φi.p only sign ∂Bi ˜ ∂ Φi sign ∂βi ˜ ∂ Φi 0 sign {θi } −sign {θi } −sign {(Φo − Φp ) · θi } sign {Φp · θi } −sign {Φo · θi } For the speciﬁc case of 0 < Φi. ∂Φi.p Φi.p < Φi.o − Φi.o + (1 − πi )(Φi.o only sign ∂Bi ˜ ∂ Φi + sign 0 ∂βi ˜ ∂ Φi .o . θX > 0.o − re. Cash ﬂow incremented: Φi.p ˜ The sensitivity of Bi and βi to an increase in Φi can be summarized as follows: Cash ﬂow(s) incremented: Φi.p −Φi.

p θX · (1 + rz ) Xo − Xp · ˜ re. Risk of the tax shields and claims for each pricing case (Table 3). 13. 12) DTS partially used in state “o. 7) βDT S 0 1 θX · (1 + rz ) · πX re.o − re. 20) Debt tax shield (DTS ) Description (cases) DTS fully used in both states (4. computed using Equation (12) and the output apportionment formulas (Table 2). 18) NTS fully used in state “o. 14.p sign ∂βDT S ∂D 0 0 0 sign{θX } (Continued) ﬀ βNT S 0 θX · (1 + rz ) A − Xp · πX A + (1 − πX )Xp re. 8. 9. 5.o − re. 15. 4. 16) DTS partially used in both states (15. debt riskless (3.70 Table A. 16) NTS partially used in both states (19.p Xo − Xp θX · (1 + rz ) = βX · ˜ re. 2. 5. 14.” partially used in state “p”.p E πX (X) sign ∂βNT S ∂D 0 0 0 ﬀ A Theory of the Firm’s Cost of Capital . 2. Depreciation tax shield (NTS ) Description (cases) NTS fully used in both states (3.p E πX (X) − A rz D − (Xp − A) θX · (1 + rz ) · πX rz D + (1 − πX )(Xp − A) re. 13.” zero in state “p” (1. 9.” partially used in state “p” (1.1. 6–8. 10–12.o − re.o − re. 17. 17.o − re. 18) DTS fully used in state “o. 11.

p sign ∂βT T S ∂D 0 0 sign{θX } sign{θX } (Continued) ﬀ 71 .o − re. debt riskless (2. 11. debt risky (6. 5. 10) DTS zero in both sates (19.p E πX (X) θX · (1 + rz ) A + rD D − Xp · πX (A + rD D) + (1 − πX )Xp re.o − re. Description (cases) DTS fully used in state “o.o − re. 6.p Null sign ∂βNT S ∂D ﬀ sign{θX } Null Total tax shield (TTS) Description (cases) TTS fully used in both states (4. 7) βT T S 0 Xo − Xp θX · (1 + rz ) · = βX ˜ re. debt risky (1. 8.p θX · (1 + rz ) A + rz D − Xp · πX (A + rz D) + (1 − πX )Xp re.” partially in state “p”. 3. 9.Appendix A Table A.” partially in state “p”.1.” partially used in state “p”.o − re. 10) TTS fully utilized in state “o. 20) (Continued) βNT S rD D − (Xp − A) θX · (1 + rz ) · πX rD D + (1 − πX )(Xp − A) re. 12) TTS partially used in both states (13–20) TTS fully utilized in state “o.

9. 8.p (X) 1 θX · (1 + rz ) · πX re. 8. 15. 9.72 Table A.p Null Description (cases) Riskless debt (2–4. 12.p E πX (X) − (A + rz D) E πX sign ∂βT ∂D Taxes paid in both states. 5–7.p (Continued) . 5. 13.p sign ∂βD ∂D ﬀ 0 sign{θX } D(1 + r11 ) − (Xp (1 − T ) + (A + r11 D)T ) θX · (1 + rz ) sign{θX } · πX D(1 + r11 ) + (1 − πX )(Xp (1 − T ) + (A + r11 D)T ) re. 17. 7.o − re. no tax paid in state “p” (1. Tax claim (T ) (Continued) ﬀ Description (cases) βT Xo − Xp θX · (1 + rz ) · ˜ re. 6.o − re. 14. 12) Taxes paid in both states.1. 20) Default in state “p”.o − re.” taxes paid in state “p” (11) βD 0 θX · (1 + rz ) D(1 + rD ) − Xp · πX D(1 + rD ) + (1 − πX )Xp re. 18. 10) Zero tax paid in both states (13–20) Debt claim (D) sign{θX } sign{θX } 0 Null A Theory of the Firm’s Cost of Capital θX · (1 + rz ) Xo − Xp · ˜ − (A + r11 D) re. 16.o − re.o − re. debt riskless (4. 10. 19) Default in state “p. debt risky (11) Taxes paid in state “o” but not in state “p” (1–3.

o − re. riskless debt (2. 17. 5. risky debt (1. riskless debt (4. 9–11. 15. 18.p Xo − Xp ˜ E πX (X) + AT 1−T sign ∂βE ∂D ﬀ 0 sign{θX } − D(1+rz (1−T )) 1−T · θX · (1 + rz ) re. 16. 12) No taxes paid.o − re. 20) βE 1 θX · (1 + rz ) · πX re. 3.p (X) sign{θX } (Continued) 73 . 6. 13.o − re.1. riskless debt (14. 8.p sign{θX } E πX Xo − Xp θX · (1 + rz ) · ˜ − D(1 + rz ) re. 19) Taxes paid in state “o” but not in state “p”. Equity claim (E) (Continued) Description (cases) Equity ﬂow zero in state “p”. 7) Taxes paid in both states.Appendix A Table A.o − re.p θX · (1 + rz ) Xo (1 − T ) + (A + Drz )T − Xp · πX (Xo (1 − T ) + AT − D(1 + rz (1 − T ))) + (1 − πX )(Xp − D(1 + rz )) re.

10) TTS risky.p E πX (X) sign{θX } 0 (Continued) .p πX Xo (1 − T ) + (A + Drz )T − Xp θX · 1 + r z ) · πX (Xo (1 − T ) + (A + Drz )T ) + (1 − πX )Xp re. 5. risky debt (1. 12) TTS riskless. fully used in state “o. fully used in both states.” partially used in state “p”.o − re.1.p (X) 1−T sign ∂βD+E ∂D ﬀ E πX −sign{θX } −sign{θX } A Theory of the Firm’s Cost of Capital sign{θX } Xo − Xp θX · (1 + rz ) · ˜ + (A+Dr11 )T re.p Xo − Xp θX · (1 + rz ) = βX · ˜ re.” partially used in state “p”. riskless debt (2. fully used in both states. 7) TTS risky.o − re. riskless debt (4. 6.o − re. partially used in both states (13–20) βD+E Xo − Xp θX · (1 + rz ) · ˜ + (A+Drz )T re.o − re. 8.o − re.p E (X) 1−T θX · (1 + rz ) Xo (1 − T ) + (A + DrD )T − Xp · πX (Xo (1 − T ) + (A + DrD )T ) + (1 − πX )Xp re. 9. risky debt (11) TTS risky. 3. Levered ﬁrm (D + E) (Continued) Description (cases) TTS riskless.74 Table A. fully used in state “o.

20) ∗ (Continued) βU Xo − Xp θX · (1 + rz ) · AT ˜ re.o − re.p E πX (X) + 1−T Xo (1 − T ) + AT − Xp θX · (1 + rz ) · πX (Xo (1 − T ) + AT ) + (1 − πX )Xp re.Appendix A Table A. 2. 4. 5. 14. Case numbers here are for an otherwise identical but levered ﬁrm.o − re. 15.p Xo − Xp θX · (1 + rz ) = βX · ˜ re.p E πX (X) ˜ Unlevered ﬁrm risk. 16) NTS partially used in both states (19. D and r are irrelevant. 17. 9. depends only on A and X. β U . 18) NTS fully used in state “o. 75 .” partially used in state “p” (1. Unlevered ﬁrm (U ) Description (cases*) NTS fully used in both states (3. 6–8. 10–12.1.o − re. 13.

3 0.3 0.3 0 0 0 0 0.3 0.000 105.3 0.3 0 0. Debt amount (D) 2 × 2 example 0 DMAX = 40.000 110.3 Tax rate.255 0.1772 0. the tax rate paid on taxable income is either 30% or zero.3 0.2317 40.3 0.3 0 0.2317 0.000 105.3 0.000 46.255 5 × 5 example Tax rate.3 0 0.000 100.2317 0.108. state 4 Tax rate.2317 0.3 0.3 0. state 5 E(tax rate) = MTR 0 0.3 0.296.2.3 0.3 0.76 A Theory of the Firm’s Cost of Capital Table A.000 113.716.3 0 0.3 0.255 0.000 100. state “p” 0 0 E(tax rate) = MTR 0. Since we assume a ﬂat statutory tax rate of 30%.3 0.3 0 0.255 Tax rate.40 0. state “o” 0.3 0.3 0 0.3 0 0.75 0. state 1 Tax rate. Post-ﬁnancing expected MTR computed as the expected value of the applicable tax rate (tax payment divided by pre-tax income).3 0.29 80.0683 .30 80.3 0 0.3 0 0 0. state 2 Tax rate.3 0.255 0.3 0 0.2317 0.255 0.661.3 0. state 3 Tax rate.3 0 0.000 76.255 0.

Appendix B Examples Illustrating the Firm’s Balance Sheet Using the Data in Table 8 77 .

036. D(1 + r) A Theory of the Firm’s Cost of Capital Debt tax shield ΦDTS .040.000.1.67 0 8.970.78 Table B.68 2.08 0 βD kD VD Par yield.08 0 0.000 0.11 0 944.623.610.420 0.430.3051 137.94 0 7.661.000 1 99.00 43.994.01 1.12 0 5.1768 100.75 0.30 0.506 0.2602 749.296.866. Debt amount (D) 0 40.08 46.430.169.00 50.000.200.01 105.296.o ΦDT S.506 0.49 46.000 0.111.970.000.036.00 131.2602 5.2602 4.90 4.296.474.2602 6.08 40.34 50.15 8.519 0.2602 7.o ΦD.21 (Continued) Pricing case (Table 3) Debt ΦD.06 2.679.00 43.623.506 0.506 0.42 80.p ˜ E(ΦD ) 2 0 0 0 Null 0.00 0 816.44 4.66 10.1848 110.000 0.30 DMAX = 110.08 43.00 0 0.45 4.07 4.000. The debt is risky when D > $46.00 0 0.34 9.000 0.89 2.27 4.2602 647.000 2 43.08 0 0 0 0 Null 0.39 50.3142 145.553.85 100.000 5 129.200.506 0.661.506 0.430.000 9 137.00 123.991.90 0 8.2962 129.000.2496 99.898 0.980.200.000.200.15 50.886.00 92.00 4.482.p ˜ E(ΦDTS ) βDT S kDT S VDT S .1559 80. Results for the 2 × 2 example.30.39 5.75 9 145.00 960.092.08 50.00 50.30 1 50.621 0. for seven debt levels.1808 105.66 50.116.00 117.000.035.83 1.000. r Promised pmt.

00 0 12.610.00 15.10 0 4.883.99 VT T S Taxes ΦT.000.2602 3.1796 31.00 ΦT T S.p 27.113.000.657.488.000 46.o ΦT.275 0.00 E(ΦNT S ) ˜ 1.506 0.566.o 15.196.919.960.968 0.p ˜ E(Φ) βT kT VT 15.27 2.041.046.123.000.000.00 6.580.45 2.89 0 11.750.75 Depreciation tax shield (for any feasible D) 30.506 0.978 0.000.00 1.886.33 0 3.888.117.934.93 4.008.506 0.00 4.73 4.569.694.000 100.25 38.67 15.09 39.61 31.56 4.41 13.69 35.o 15.469.11 15.10 30.2602 4.00 1.005.000.1.296.000.1560 kNTS 24.90 2.00 32.506 0.Appendix B Table B.30 80.899 βT T S 0.661.00 35.89 (Continued) 79 .1781 30.000.2602 6.805.389.00 35.90 15.1560 kT T S 24.430.420 0.453 0.00 36.653.07 2.29 5.2602 10.2602 9.94 15.506 0.000.842.1768 30.506 0.076.005.991.99 VNTS Total tax shield 30.00 28.00 28.p ˜ E(Φ) 27.84 9.2602 9.755.12 15.506 0.00 ΦNT S.367.000 DMAX = 110.00 ΦNT S.55 4.00 4.2602 3.49 14.00 0 11.1710 28. (Continued) Debt amount (D) 0 40.000.634.06 0 5.000.44 1.111.10 4.490 0.88 0 7.80 40.000.750.00 ΦT T S.081.000 105.750.20 4.1591 24.040.303.899 βNTS 0.1587 24.616.

506 0.661.75 0.0000.430.000.672 βD+E+T 0.603 0.611 0.661.020.65 VD+E+T .703.2602 27.44 4.065 0.76 Debt + equity + tax (for any feasible D) 150.506 0.743.000 100.p ˜ 135.00 129.00 127.00 144.108.000.07 2.00 2.493 = β U 0.620.11 0 73.00 123.00 = E(ΦU ) 2.12 50.44 2.2602 9.01 0 0.90 50.39 20.0000.000.56 0 6.00 130.375.000.04 86.111.960.00 50.000.800.45 2.438.2602 0.00 ˜ 122.667.55 7.00 E(ΦD+E+T ) 2.000.000.621 0.1827 107.626.194.00 122.1803 104.991.96 0 136.000 DMAX = 110.67 50.000.62 4. 250.1.96 41.505 0.000.1797 103.866.568 0.886.661.574.667.38 4.667.39 0 140.11 50.506 0.296.375.1802 104.27 2.620.55 0 9.760.39 14.116.00 7.1848 110.54 0 135.419.23 4.867.274.506 0. 626.1845 110.342.00 3.1869 kD+E+T 113.274.73 0 34.80 Table B.00 123.00 131.263.80 0 12.66 53.00 50.610.90 2.o ΦE.p ˜ E(ΦE ) βE kE VE E0 Wealth gain.o ΦD+E.507 0.76 10.00 ΦD+E+T.84 5.626.55 0 143.000 46.04 60.2602 5.30 80.250.108.o 50.04 0 135.124.00 4.2602 58.00 6.803.274.000.70 4.506 0.p ˜ E(ΦD+E ) βD+E kD+E = WACC VD+E = V L 40.000.54 92.079.0000.94 50.00 4.00 2.108.194.54 = V U 100.000.84 0 10.661.1841 109.1797 = k U 103.01 0 10.2426 64.75 135.111.066. ∆W DIV0 Debt + equity ΦD+E.000.84 A Theory of the Firm’s Cost of Capital 145.00 79.00 50.000 105.493 0.000. (Continued) Debt amount (D) 0 Equity ΦE.620.01 4.00 ΦD+E+T.

29 0.715.63 88.00 43. for seven debt levels.2511 3.6049 181.000 DMAX = 113. Results for the 5 × 5 example.27 120.1 ΦDTS .00 124.853.51 10.60 82.005.09 2.49 0 10.27 120.000 0.4 ΦDTS .88 6.173.60 86.200.000.513.60 0 0.60 4.54 1.452.173.60 88.528.00 43.27 130.841.04 8.728.63 88.000 0.176 0.95 4.35 4.452.853.841.576.599 0.173.2450 130.29 80.27 82.11 7.853.2511 1.84 4.3 ΦDTS .00 43.60 82.172.108.99 13.718.000 100.88 6.000 105.000 76.173.718.810.841.278 0.27 120.853.0870 80.728.716.40 0.452.282 0.60 133.961.66 1.841.Table B.09 2.853.95 0.00 43.108.853.09 0 1.728.853.63 120.60 118.000 0.08 76.136.452.27 4.40 7.2 ΦD.2556 4.00 0 0 0 0 0 0 Null Null 0 960.19 0 1.40 181.421.278 0.1313 105.1 ΦD. r Promised pmt.173.35 145.08 82.994.728.556 0. The debt is risky when D >$76.08 40.44 2.2 ΦDTS .63 82.60 111. D(1 + r) Debt tax shield φDT S.501.984. Debt amount (D) 0 Debt ΦD.893.498.200.3 ΦD.59 0.00 130.5 ´ ` ˜ E ΦDTS βDTS kDTS VDTS 40.00 960.08 0 43.08 43.00 7.498.716.00 0 71.19 1.00 124.853.50 20.2511 592.994.19 1.000 0.720.576.88 6.500.200.1198 100.169.297.853.172.200.525.5 ´ ` ˜ E ΦD βD kD VD Par yield.172.92 (Continued) 81 .2511 1.00 130.2.40 7.00 82.29.45 2.60 82.576.88 0 4.853.278 0.716.49 0 8.718.19 1.200.389 0.40 7.853.200.00 43.09 2.1822 113.98 82.528.60 82.783.172.2058 120.00 0 0.576.35 Appendix B 0 0 0 0 0 0 Null 0.576.1022 88.4 ΦD.51 6.278 0.63 88.00 960.766.392.00 960.02 135.08 0 0.60 82.4240 7.996 0.200.98 82.

498.189.960.39 50.00 ΦNT S.856.1 30. (Continued) Debt amount (D) 0 40.000.49 24.519 βNT S 0.675 0.000.19 31.09 32.00 ΦNT S.1007 26.501.00 ΦT T S.00 30.88 36.000 DMAX = 113.43 31.000 105.1041 26.51 40.000.38 0.2.34 32.09 24.841.2 30.280 0.19 31.00 ΦNT S.000.602 0.856.3 30.841.856.1070 27.000.172.960.98 1.19 31.27 0.856.828.08 ΦNT S.718.189.856.3 30.856.569.500.29 80.88 24.00 ΦNT S.40 36.49 24.703.856.000 76.19 24.00 ΦT T S.000 100.172.220.4 24.960.519 0.5 ´ ` ˜ 28.716.452.43 E ΦT T S βT T S kT T S VT T S 30.08 33.00 37.108.5 ´ ` ˜ 28.4 24.08 34.960.748.40 37.718.326.452.487.89 VNT S Total tax shield 30.08 30.00 ΦT T S.1198 30.172.46 2.996 0.08 30.109 0.1090 27.08 29.722.452.88 36.08 ΦT T S.718.43 E ΦNT S 0.250.1244 30.724 0.828.40 37.08 39.172.78 0.03 0.81 (Continued) A Theory of the Firm’s Cost of Capital 0.00 24.905.595.525.841.856.89 .2 30.000.00 37.82 Table B.1712 33.1007 kNT S 26.782.000.498.841.40 Depreciation tax shield (for any feasible D) 30.000.40 37.00 30.09 32.09 32.88 36.000.00 ΦT T S.99 43.1 30.452.00 30.

777.40 5.13 10.91 5.525.22 10.73 10.108.540.62 8.29 80.46 31.48 0 25.046.340.500.143.108.701 0.41 13.490.493.403.000 76.40 (Continued) 26.11 43.796.865 0.00 6.100.000.002.503.538.4 ΦT.40 0 9.88 8.865 0.283.328.896.59 36.93 0 23.97 11.279.108.62 14.04 7.000 105.3 ΦT.73 11.403.51 9.49 82.648 0.48 = V U 100.7873 3.110.0141 635.498.658.346 0.43= E ΦU 2.00 Equity ΦE.599.03 23.60 ` ´ ˜ 123.328.22 161.225 0.54 23.078 0.006 0.620.716.302.324.11 2.000 DMAX = 113.4002 27.59 0 18.653.230.57 9. (Continued) Debt amount (D) 0 Taxes ΦT.00 6.496.00 117.2 ΦE.97 38.000.08 50.3 ΦE.0141 0.00 0 19.280.783.10 17.Appendix B Table B.992.81 5.183.5231 6.691.59 0 9.6746 4.8938 5.500.49 0 10.810.51 124.525.00 7.447 0.660.751.000.946.71 75.093.1656 = k U 105.52 5.762.671.40 0 0 10.48 131.1 ΦT.64 23.049.867.682 0.501.990.63 4.12 1.93 0 0 0 0 1.324.61 3.933.03 0 24.325.000.465.49 0 11.00 5.01 0 13.91 1.217.08 0 41. 810.352 1.03 60.260.83 11.3859 30.657.00 7.620.4 ΦE.10 0 16.491.71 6.327.00 75.67 7.32 8.4890 7.425.139 = β U 0.12 0 0 0 0 0.5 ´ ` ˜ E ΦE βE kE VE E0 Wealth gain.30 0 10.484.352 1.94 45.781.781. ∆W DIV0 83 .61 0 6.73 37.000 100.6746 9.94 0 34.02 17.49 39.48 89.2 ΦT.249.930.11 46.92 13.32 20.541.2179 66.5 ´ ` ˜ E ΦT βT kT VT 40.51 82.778.2.263.5080 6.047.4712 7.1 ΦE.971.60 0 0 5.774.096.617.256.60 80.853.78 79.904.68 118.00 20.487.419.98 5.42 8.123.78 3.92 12.98 0 10.79 14.781 0.51 10.869.30 0.

676.669.1 145.51 125.456.19 119.03 162.503.41 ΦD+E.02 135.720.40 124.000 76.716.60 125.1668 107.1 40. (Continued) Debt amount (D) 0 Debt + equity ˜ ΦD+E.843.000 100.71 126.108.496.000.005.720.324.27 2.853.36 137.58 82.454.853.1822 113.40 ΦD+E+T.525.00 124.994.27 2.463.60 126.853.000.60 124.235 0.67 133.344.51 124.1869 kD+E+T 113.60 128.47 145.93 163.09 119.672.750.4 82.98 82.500.000 105.853.68 82.57 133.069.00 118.48 131.962.151 0.853.46 2.403.955.88 139.620.98 A Theory of the Firm’s Cost of Capital Debt + equity + tax (for any feasible D) 187.002.222.00 167.91 132.000 DMAX = 113.00 ΦD+E+T.1660 106.218.460.169 0.146.528.49 82.419.810.1685 109.743.3 124.98 82.4 ΦD+E.20 ΦD+E+T.1656 105.60 ΦD+E+T.005.556 0.853.00 E ΦD+E+T 2.49 82.37 82.38 2.3 ΦD+E.78 2.88 123.108.71 VD+E+T .2 ΦD+E.213 0.59 168.48 132.2.60 123.337.952.29 80.162 0.1665 106.98 ΦD+E+T.60 124.1694 110.2 135.48 161.03 2.328.994.095.994.853.5 ´ ` ˜ E ΦD+E βD+E kD+E = WACC VD+E = V L 161.221.84 Table B.39 130.000.40 181.853.60 133.00 117.175.672 βD+E+T 0.946.341.5 ´ ` ˜ 135.948.139 0.60 130.43 2.

Journal of Business. Capital budgeting in 1992: A survey. JB Taylor and M Woodford (eds. 3rd Ed. Connor. 22–24. Taxation. 85 . G and RA Korazcyk (1993). Financial Practice and Education. 1263–1291. Journal of Finance. valuation and the problem of optimal capital structure. Valuation: Measuring and Managing the Value of Companies. RJ (1999). Bierman. Brennan. 905–940. Firm valuation. and mean-variance analysis on large asset markets.References Auerbach. Aggregate investment. 20–39. 51 (September). 48(4). JI and LH Summers (1984). Copeland. Econometrica. T Koller and J Murrin (2000). Corporate income taxes. 1251–1264. M and E Schwartz (1978). New York: Wiley. 13–28. Journal of Economic Literature. AJ and K Hassett (1992). G and M Rothschild (1983). Journal of Political Economy. KM Eades. pp. RF. Financial Management. Arbitrage. corporate ﬁnancial policy and the cost of capital. A test of the number of factors in an approximate factor model. R Harris and RC Higgins (1998). 30(5). Bruner. Chamberlain. Journal of Public Economics. Amsterdam: North Holland. 21(3). 103–114. Best practices in estimating the cost of capital: Survey and synthesis. DP (1975). In Handbook of Macroeconomics. factor structure. Bulow. 8. Caballero. TE. HJ (1993). Journal of Finance. Auerbach. 1281–1304. The taxation of risky assets. AJ (1983). Baron. 141–170. 22. 51(1). 813–862.) (1B). Tax policy and business ﬁxed investment in the United States. 92(1). corporate taxes and default risk. 47(2).

86 A Theory of the Firm’s Cost of Capital Copeland. Measuring corporate tax rates and tax incentives: A new approach. Graham. Irwin: McGraw-Hill. Dammon. R Litzenberger and J Rolfo (1977). Factor pricing in a ﬁnite economy. R and LW Senbet (1988). Financial Markets and Corporate Strategy. The theory and practice of corporate ﬁnance: Evidence from the ﬁeld. Taxes. 43(2). Journal of Empirical Finance. Journal of Financial Economics. The option pricing model and the risk factor of stock. The adjusted present value: The combined impact of growth and the tax shield of debt on the cost of capital and systematic risk. 5–12. Journal of Financial Economics. 60. What do practitioners want? Journal of Applied Finance. 55(October). Gonzales. European Financial Management. 42. . 3(1/2). 357–373. The structure and incentive eﬀects of corporate tax liabilities. Journal of Financial Economics. 41–73. 187–243. Journal of Finance. 12. The value of tax shields is not equal to the present value of tax shields. 187–221. 1901–1941. Journal of Financial Economics. Journal of Financial Economics. Ehrhardt. 28(1). Galai. D (1984). D (1998). 12(1). Grinblatt. H and R Masulis (1980). 145–165. On mean variance models of capital structure and the absurdity of their predications. 165–179. JR and CR Harvey (2001). M and S Titman (1983). 11(Spring) 54–65. P (2004). N. How big are the tax beneﬁts of debt? Journal of Finance. 23–41. Graham. Graham. 497–507. 41. 3–27. 12(2). 2nd Ed. RC and E Talmor (1985). Galai. Grinblatt. Optimal capital structure under corporate and personal taxation. JR (2000). 73(1). Fullerton. 1095–1114. 8(1). DeAngelo. Green. Journal of Applied Corporate Finance. JR and ML Lemmon (1998). Which eﬀective tax rate? The National Tax Journal. D and RW Masulis (1976). 53–81. Graham. Proxies for the corporate marginal tax rate. Journal of Financial Economics. M and S Titman (2002). Graham. 40(4). 4(2). The eﬀects of taxes and depreciation on corporate investment and ﬁnancial leverage. Journal of Financial and Quantitative Analysis. Debt and the marginal tax rate. Fernandez. JR (1996a). M-M propositions and the government’s implicit cost of capital in investment projects in the private sector. 143–157. TE (2002). MC and PR Daves (2002). JR (1996b). Journal of Finance. Journal of Financial Economics.

Do taxes aﬀect corporate ﬁnancing decisions? Journal of Finance. MA. Journal of Finance. 2(4). 15(September). 1213–1252. J and J Ezzell (1980). NBER Working Paper 1553. The valuation of cash ﬂow forecasts: An empirical analysis. Journal of Finance. MA: MIT Press. 1959–1093. 719–730. 911–922. A and RH Litzenberger (1973). Journal of Economic Perspectives. 449–470. Kaplan. Reformulating tax shield valuation: A note. M and S Titman (1985). Cambridge. Kraus. Interactions of corporate ﬁnancing and investment decisions: A dynamic framework. Leland. Econometrics: Econometrics and the Cost of Capital. 40(5). Corporate debt value. Harris. CM (1990). (1995). Kim. Vol. Miles. 25–42. On the pricing of corporate debt: The risk structure of interest rates. 237–244. Cambridge. JB (1974). 1471–1493. 177–202. 1485–1492. Merton. Journal of Finance. 13–31. Journal of Financial and Quantitative Analysis. 1367–1373. LJ (2000). . A multiperiod theory of corporate ﬁnancial policy under taxation. HE (1994). 29(2). Journal of Finance. Portfolio analysis. 24(1). 45–63. Journal of Financial and Quantitative Analysis. market equilibrium and corporate ﬁnance. Journal of Financial Economics. Lau. MH (1988). The weighted average cost of capital. Leverage. J and J Ezzell (1985). 1253–1277. Majd. Risk adjusted discount rates: Extensions from the average risk case. Journal of Financial Research. Valuing the government’s tax claim on risky corporate assets. 8(Fall). 49. RS and JJ Pringle (1985). Journal of Finance. and optimal capital structure. perfect capital markets. RS (1969). SN and R Ruback. Journal of Finance. Hite. GL (1977). A mean-variance theory of optimal capital structure and corporate debt capacity. and project life: A clariﬁcation. output eﬀects and the M-M theorems. Miller. 33(1). 29. Journal of Finance. Mackie-Mason. D and A Triantis (1994). Journal of Finance. Miles. Lewis. bond covenants. Long. A state preference model of optimal ﬁnancial leverage. 485–488. 28(September). Hamada. JK (1990). EH (1978). 40. Journal of Finance. 25(March). 4(2). 2. Discussion. 49(4). Approximate factor structures: Interpretations and implications for empirical tests. RC (1974). Journal of Finance. 45(5). S and SC Myers (1985).References 87 Grinblatt. Mauer. The Modigliani–Miller propositions after thirty years. 99–120. 50(4).

Financial Management. F and MH Miller (1958). Consistent valuation and cost of capital expressions with corporate and personal taxes. 13(3). American Economic Review. Talmor. MH (1977). Modigliani. Ross. The determinants of corporate borrowing. Reinschmidt. 191–202. Do taxes aﬀect entrepreneurs’ investments? 41st Alex G. corporation ﬁnance and the theory of investment. F and MH Miller (1963). RA (1991). 307–312. Financial Management. 167–181. 5(2). 931–940. SA. Journal of Finance. A mean-variance synthesis of corporate ﬁnancial theory. 147–175. The arbitrage theory of capital asset prices. IL: Irwin. Modigliani. Taggart. 29(1). 53(3). Englewood Cliﬀs. 327–341. 28(1). Summer. 40. 48. Journal of Finance. 8–20. present and future. 2(4). 261–297. 10. 6th Ed. HS (1998). 85–103. 341–360. Journal of Political Economy. Zechner. Modigliani. The critical implicit tax rate and capital structure. F (1988). M (1973). R Westerﬁeld and JF Jaﬀe (1988). SM (1979). E. Ross. The value of the tax subsidy on risky debt. Homewood. SC (1977). Ruback. Corporate Finance. Journal of Finance. Rubinstein. Ross. Journal of Business.88 A Theory of the Firm’s Cost of Capital Miller. Rosen. Debt and taxes. 261–275. Myers. Journal of Banking and Finance. KF (2002). SA (1987). Debt capacity. SA (1985). March 1998. Journal of Financial Economics. SC (1974). Scholes. Interactions of corporate ﬁnancing and investment decisions: Implications for capital budgeting. RS (2002). 95(2). Aggregate social discount rate derived from individual discount rates. 433–443. McKenna Economic Education Series Lecture. M and M Wolfson (1992). Ross. Autumn. 34. 48(3). 149–158. Management Science. The cost of capital. NJ: Prentice-Hall. American Economic Review. R Haugen and A Barnea (1985). Capital cash ﬂows: A simple approach to valuing risky cash ﬂows. . 58(2). J and P Swoboda (1986). Corporate income taxes and the cost of capital: A correction. Journal of Finance. 32(2). SA (1976). Taxes and Business Strategy. MM-past. Debt and taxes under uncertainty. Journal of Economic Perspectives. 1–25. Myers. Arbitrage and martingales with taxation. Journal of Finance. 371–392. Journal of Economic Theory. 637–657. Turnbull.

2 debt. 17 maximum likelihood. 29 government’s (tax) claim. vi cost of debt. 1–4 debt capacity. 33. 18 ﬁrm risk. 3 borrowing interest. 19 kinked payoﬀs. 4 Monte Carlo simulation. 34 adjusted present value. 40 accounting cost of debt. 33 default/solvency states. 1963). 18 arbitrage pricing theory (APT ). 41 approximate APT. 3. 3 cash ﬂow beta. 55 . 2 borrowing rate. 4 debt tax shield. 6 ex dividend equity. 21 cost of capital while. vii eﬀective MTR. 35–40 ﬁrm’s value. 20 Modigliani and Miller (MM. 20 leverage. 17 default risk. 3 economic balance sheet. r. 16 capital asset pricing model (CAPM ). 3 89 discount rate. 7 bull spread. 43 eigenvalues. 17. vii marginal tax rate (MTR). 7. 1 martingale probabilities. 16 call options. vi. 46 Ibbotson Associates. 41. 18 congressional tax policy. 4 marginal corporate tax rates. 34. 37 macro-states. 4 depreciation. 9. 2 joint binomial assumption. 42 “interest ﬁrst” doctrine. 16. Brattle Group. 4 Hadamard product. 57 correlation. 2. 4 joint binomial probability distribution. vi breakeven level of output.Index “Three Claims View”. 35 coupon rate.

36 net present value. 18 . vi. 16–18 shields’ risks. 2. 20 state-pricing. 21 public sector project analysis. 2 statutory tax rate. 6. v. 6 non-debt. 19. 27–30 probability matrix. vi states. 17 non-debt tax shield. 5. 8. 29. 30 perpetuity. 27. vi policy. 21 pricing cases. 3 par yield. 55 tax tax tax tax tax claim. 3 priced risk factor. 29. 16 redundant tax shields. 2–4. 18 risk-neutral probability. 1 zero-beta asset. 20 risk-adjusted present value. 31 weighted average cost of capital (WACC ). 16 priced risk. v. 19. 25. 24. 23 unlevered ﬁrm. 25. 20 risk-neutral valuation (RNV).90 Index multiplicative coeﬃcient. 18 optimal capital structures. 7–10. 5. 57 put options. 37 return on risk factor. 18 returns beta. 18. 27 non-traded assets. 40 options approach. 42 shield.

- MBA_Curriculum_Syllabus_0.pdf
- Carbon Credit Final
- Financial Structuring
- ICICI Gold ETF Application Form
- Master Thesis Maurits Kruithof
- Hindalco Project
- FAC1502 Property, Plant and Equipment Notes
- (Phase 1) Q&A Process - Understanding the Data
- Responsibility Centres1
- Assessing the Ethics of Implementing Performance Appraisal Systems (2010)
- Bank Financial Management
- Finance Summaries
- 37045713 Depreciation Accounting
- AFRB 223
- Audit
- PWC All Clear Eu Introduces Clearing and Reporting Regime for Otc Derivatives
- 09435BBA305 - Notes (Module 1)
- hris ppt ma
- F7_2014_Dec_Ans
- Basileia III e Os Novos Indicadores de Liquidez - Jorg Hashagen
- Assignment 1 Solution SC 2010
- Controller or Director of Finance or General Manager
- Research Project on Invst Prefrence
- Sores 397
- SSRN-id322700
- Dynamic ALM
- Lecture 01
- Cost and Management Accounts
- Negwa ResumeMBA
- Ratio Analysis of Ncl by Shrawan Kumar Dwivedi

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue listening from where you left off, or restart the preview.

scribd