FOURTH EDITION

s

STUDENT SOLUTIONS MANUAL
Thomas E. Copeland J. Fred Weston Kuldeep Shastri
Managing Director of Corporate Finance Monitor Group, Cambridge, Massachusetts Professor of Finance Recalled, The Anderson School University of California at Los Angeles Roger S. Ahlbrandt, Sr. Endowed Chair in Finance and Professor of Business Administration Joseph M. Katz Graduate School of Business University of Pittsburgh

Reproduced by Pearson Addison-Wesley from electronic files supplied by author. Copyright © 2005 Pearson Education, Inc. Publishing as Pearson Addison-Wesley, 75 Arlington Street, Boston, MA 02116 All rights reserved. This manual may be reproduced for classroom use only. Printed in the United States of America. ISBN 0-321-17954-4

1 2 3 4 5 6 OPM 07 06 05 04

Contents
Preface...............................................................................................................................................v Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Chapter 7 Chapter 8 Chapter 9 Chapter 10 Chapter 11 Chapter 12 Chapter 13 Chapter 14 Chapter 15 Chapter 16 Chapter 17 Chapter 18 Chapter 19 Introduction: Capital Markets, Consumption, and Investment................................1 Investment Decisions: The Certainty Case..............................................................6 The Theory of Choice: Utility Theory Given Uncertainty ....................................13 State Preference Theory.........................................................................................32 Objects of Choice: Mean-Variance Portfolio Theory............................................44 Market Equilibrium: CAPM and APT...................................................................60 Pricing Contingent Claims: Option Pricing Theory and Evidence........................77 The Term Structure of Interest Rates, Forward Contracts, and Futures ................90 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis ..........97 Efficient Capital Markets: Theory .......................................................................119 Efficient Capital Markets: Evidence....................................................................125 Information Asymmetry and Agency Theory......................................................128 The Role of the CFO, Performance Measurement, and Incentive Design ..........133 Valuation and Tax Policy ....................................................................................137 Capital Structure and the Cost of Capital: Theory and Evidence ........................140 Dividend Policy: Theory and Empirical Evidence ..............................................160 Applied Issues in Corporate Finance ...................................................................166 Acquisitions, Divestitures, Restructuring, and Corporate Governance ...............172 International Financial Management ...................................................................184

The student is asked to combine the elements of what he has learned into something slightly different — a new result. But it is also necessary for the layman to learn a little mathematics. He must think for himself instead of just regurgitating earlier material. He’ll know what to do!” It was almost as if the scientist was being equated with a magician or witchdoctor. Needless to say. From the crowd on screen an individual is heard to shout. mathematics cannot solve every problem. mathematical programming. In order to narrow the gap between the rigorous language in academic Finance journals and the practical business world it is necessary for the academician to translate his logic from mathematics into English. We are hoping that the reader will meet us halfway. v . By the way — the movie scientist did know what to do. Additionally. statistics and econometrics. In fact. transistor. The usual approach to the question of “why” is to build simple mathematical models. Most theory texts in Finance do not have end-of-chapter questions and problems.Preface The last forty years have seen a revolution in thought in the field of Finance. variance is a precise measure of risk and yet almost everyone has an intuitive grasp for its meaning. The basic questions remain the same. ordinary calculus. Problem sets are useful because they help the reader to solidify his knowledge with a hands-on approach to learning. probability theory. How are real and financial assets valued? Does the market place provide the best price signals for the allocation of scarce resources? What is meant by risk and how can it be incorporated into the decision-making process? Does financing affect value? These will probably always be the central questions. but it does force us to use more precise language and to understand the relationship between assumptions and conclusions. and car are familiar throughout the globe. Such extrapolative questions ask the student to go beyond simple feedback of something he has just read. stochastic calculus. In most articles. The words computer. Unfortunately. We assume a basic understanding of algebra and simple calculus. but little more. This is already happening. this is infrequently the case in the real world. Students learned about the way things were rather than why they came to be that way. “Go get a scientist. This solutions manual and the textbook which it accompanies represent an effort to bridge the gap between the academic and the layman. Technical words in English can be found unchanged in almost every language throughout the world. Notable exceptions were Fama’s Foundations of Finance and Levy and Sarnat’s Capital Investment and Financial Decisions. differential equations. technical terms are becoming a world language. Today the emphasis is on answering the question — why have things come to be the way we observe them? If we understand why then we can hope to understand whether or not it is advisable to change things. he can usually understand the introduction and conclusions. problems can be used to stretch the reader’s understanding of the textbook material by asking a question whose answer cannot be found in the text. This has the effect of widening the gap between theory and application. The more scientific and more mathematical Finance becomes the more magical it appears to the layman who would like to understand and use it. In Finance. This bewildering set of applied mathematics makes the best academic journals in Finance practically incomprehensible to the layman. However. A serious student of Finance must seek prerequisite knowledge in matrix algebra. The mathematics employed here is at a much lower level than in most academic journals. Forty years ago the field was largely descriptive in nature. In their efforts to gain better understanding of complex natural phenomena. We remember a quote from an old Japanese science fiction movie where a monster is about to destroy the world. On the other hand it is at a higher level than that which the layman usually sees. academicians have adopted more and more complex mathematics. the answers to them have changed dramatically in the recent history of Finance.

comments. If our efforts stimulate you. and the MBA students at UCLA. Copeland Monitor Groups Cambridge. to other new ideas. It cannot be effectively employed if kept behind locked doors as an instructor’s manual. CA 90024 vi . we highly recommend that the solutions manual be made available to the students as an additional learning aid. Consequently. Students can order it from the publisher without any restrictions whatsoever. PA J. MA 02141 Kuldeep Shastri University of Pittsburgh Pittsburgh. This is also the objective of the end-of-chapter problems in our text. Fred Weston Anderson Graduate School of Management University of California Los Angeles. criticisms and corrections. We think the users will agree that we have broken some new ground in our book and in the end-ofchapter problems whose solutions are provided in this manual. We wish to express our thanks to the following for their assistance in the preparation of this solutions manual: Betly Saybolt. we will welcome your suggestions. Thomas E. the user.The objective of education is for each student to become his own teacher. Any kinds of communications will be welcome.

e.1 Fisher separation for the lender case equals the market rate of interest at point B. P1). Consumption.. and Investment 1.Chapter 1 Introduction: Capital Markets. Finally. W0: W0 = y0 + y1 1 + rf The individual will take on investment up to the point where the marginal rate of return on investment Figure S1. Using the market rate. at point A. the present value of his endowment is his current wealth. Assume the individual is initially endowed. in order to achieve his maximum utility (on indifference curve U1) the individual will lend (i. C1 which has a present value of W0∗ = C∗ 0 + ∗ C1 1 + rf . At this point his ∗ optimal consumption is C∗ 0 . This determines the optimal investment in production (P0. with current income of y0 and end-of-period income of y1. consume less than P0) along the capital market line until he reaches point C.

3 Market rate cannot fall below net rate from storage .3. then it increases to W0′ . rate. from U L to U L′ .2 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Originally it is at W0. 3. After the decrease in interest rate. Figure S1. This implies a capital market line with a 45° slope (a slope of minus 1) as shown in Figure S1. Borrowers originally chose levels of current consumption to the right of A. current wealth is the intercept of the capital market line with the C0 axis. Assuming that there are no opportunity costs or spoilage costs associated with storage. and experience an increase in utility from U L 1 to UB ′1 . then the rate of return from storage is zero. The case for those who were originally lenders is ambiguous. 2 2 (b) Because borrowers and lenders face the same investment opportunity set and choose the same optimal investment (at A before the interest rate decreases and at A’ afterward). The remaining lenders experience a decrease in utility. their utility has increased unambiguously from UB to U′ B. lower. Fourth Edition 2. Some individuals who were lenders become borrowers under the new. Figure S1. (c) The amount of investment increases from I to I’.2 An exogenous decline in the interest rate (a) An exogenous decrease in the interest rate shifts the capital market line from the line through AW0 to the line through A′W0′.

4. no one will be willing to do so because storage at a zero rate of interest is better than lending at a negative rate. These dominant alternatives are represented by the heavy lines in Figure S1. and Investment 3 Also shown is a line with lower absolute slope. If his time preference is such that he desires to save some of his current consumption and store it.3 are infeasible. points along line segment YZ in Figure S1. Clearly.3. In order to borrow at a negative rate it is necessary that someone lend at a negative rate.Chapter 1 Introduction: Capital Markets. it may not be possible. which represents a negative borrowing and lending rate. If next year’s coconut supply does not mature until then. y1) rather than lending (to the left of y0). The conclusion is that the market rate of interest cannot fall below the storage rate. Consequently. if the individual wishes to consume more than his current supply of coconuts in order to move to point B. he will do so and move to point A in Figure S1. consuming tomorrow’s endowment today).e. If we were not assuming a Robinson Crusoe economy.4 Storage as the only investment On the other hand. He would also prefer to borrow at a negative rate rather than storing backward (i. Any rational investor would choose to store forward from his initial endowment (at y0. An individual who wished to consume more than his current allocation of wealth could contract with other individuals for some of their wealth today in return for some of his future wealth. Assume that Robinson Crusoe has an endowment of y0 coconuts now and y1 coconuts which will mature at the end of the time period. Consumption. . then exchange would make it possible to attain point B.. In this case he is storing forward. Figure S1. However.4. it may be impossible to store coconuts backward. one of them is not feasible.

6 Investment opportunity set 6. Figure S1.6.000 5. all of which have the same rate of return.000.000 1.000 7.000.30 1.000.e.04 Outlay for the ith Project $3.000 2.000.000 1.000 4. This is done below.000.000.20 1. consumption foregone along the C0 axis) and the extra consumption provided at the end of the investment period. For example.5 All investment projects have the same rate of return The resultant investment opportunity set is a straight line with slope –(1 + R*) as shown in Figure S1.5 shows a schedule of investments.. The marginal rate of substitution between C0 and C1 is a constant. In order to graph the production opportunity set.08 1. Fourth Edition * 5. R . first order the investments by their rate of return and sum the total investment required to undertake the first through the ith project. Figure S1.4 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.000 Sum of Outlays $3.000.000 The production opportunity set plots the relationship between resources utilized today (i. Figure S1.000. if only project D were undertaken then $3 million in current . Project D B A C One Plus the Rate of Return 1.

9 million in end-of-period consumption. Figure S1.6364 million 1. The answer to part b of the question is found by drawing in a line with a slope of −1.Chapter 1 Introduction: Capital Markets. This is graphed below in Figure S1.1 C1 Figure S1.7 If we aggregate all investment opportunities then $7 million in consumption could be foregone and the production opportunity set looks like Figure S1. The present value of this decision is W0 = + C0 1+ r 5.7. Consumption.1 and finding that it is tangent to point B.8 Production opportunity set .3 × ($3 million) = $3.1 = + 3 = $7. Hence the optimal production decision is to undertake projects D and B. and Investment 5 consumption would be foregone in order to receive 1.8.

000 = 158.000 100.000 $18.000 10. equation 2-13a can be used: CF = ∆NI + ∆dep + (1 − τ c )∆ k d D = 18.4)(0) = 28. (a) Earnings before depreciation.4) + .4(10.000 + (1 − .000 (b) Net present value using straight-line depreciation NPV = ∑ (Revt − VCt )(1 − τ c ) + τ c (dept ) − I0 (1 + WACC)t t =1 = (annual cash inflow) (present value annuity factor @12%.000 = (5.000)(1 − . 10 years) − I 0 N $140.000) = 28. (a) Cash flows adjusted for the depreciation tax shelter Sales = cash inflows Operating costs = cash outflows Earnings before depreciation.Chapter 2 Investment Decisions: The Certainty Case 1.200 − 100.200 .000 + 10.000 4.800 $7.000 − 100.000) − 100.000 = 58.000 12.200 2.650)(28.000 Alternatively. interest and taxes Depreciation (Dep) EBIT Taxes @ 40% Net income Using equation 2-13: CF = (∆Rev − ∆VC)(1 − τ c ) + τ c ∆dep = (140.000 40.000 12.000 30.000 10. interest and taxes Depreciation (straight-line) EBIT Taxes @ 40% Net income $22.

200 + 7.545.797 .29 4.000 22. .180 − 100.02 9.000 22.000 22.200 + 6.000 22.893 .484.000 22.000 (3) Dept (10/55)100.540.200 + 2.181.363.000) = 17.712 .200 + 5.000 (4/55)100.4) + .58 100.958.000 (5/55)100.214.18 13.09 13.54 13.507 . where T = 10 55 In each year the cash flows are as given in the table below: (1) Year 1 2 3 4 5 6 7 8 9 10 (2) Revt − VCt 22.636 .633.820 (b) NPV using sum-of-years digits accelerated depreciation In each year the depreciation allowance is: Dep t = T +1− t N ∑i i =1 T = T + 1− t .82 13.000 = −2.452 .000 (8/55)100.272.536.958.200 + 3.000 (6/55)100.4(10.91 13.27 (5) PV Factor .200 + 2.14 15.64 13.45 13.72 13.000 22.282.27 Notice that using accelerated depreciation increases the depreciation tax shield enough to make the project acceptable.58 8.454.31 6.000 (7/55)100.000 = 958.000)(1 − .281.000 (9/55)100.Chapter 2 Investment Decisions: The Certainty Case 7 Net present value using straight-line depreciation CF = (∆Rev − ∆VC)(1 − τc ) + τc ∆dep = (22.000 (3/55)100.404 .200 NPV = ∑ CFt − I0 t + (1 WACC) t =1 = (annual cash flow) (present value annuity factor @ 12%.94 11.27 NPV = PV of inf lows − I 0 NPV = 100.200 + 4.737.26 5.200(5.636.000 22.000 (4) (Revt − VCt)(1 − τc) + τcdep 13. 10 years) − I 0 = 17.000 (1/55)100.03 7.200 + 1.090.000 22.322 (6) PV 18.000 (2/55)100.000 22.909.12 13.361 .27 − 100.290.650) − 100.818.200 + 5.958.200 + 727.000 = 97.36 13.567 .

936 . Replacement with salvage value Amount before Tax Outflows at t = 0 Investment in new machine Salvage value of old Tax loss on sale Inflows.400 –2.000 = $17. 4.000 1–8 1–8 1–8 8 Using the NPV rule the machine should be replaced.968 –9.500 Amount after Tax $100.000 Year 0 0 0 PVIF @ 12% Present Value 1.600 5.859 4.176 $31.000 –5.405 4.968 $92.968 24. years 1–8 Savings from new investment Tax savings on depreciation $100.245 Net present value = $117. VC = cash savings from operations = −3.00 –15. Fourth Edition 3.176 − 75.000 31. years 1–8 Savings from new machine Depreciation saving on new Depreciation lost on old Salvage value of new $100.000 12.000 τc = the tax rate = .600 4.000 1–8 4.968 92.000 –10.00 –10.000 –15.0 Present Value $100. Replacement Amount before Tax Outflows at t = 0 Cost of new equipment Inflows. 5.000 1.000 Amount after Tax $100. buy the new machine.000 12.000 18.000 $18.840 Present value of inflows = $117.405 1–8 4.8 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.848 Net cash inflows = $109.245 − 100.000 12.000 = $34.000 –25.000 –15.176 Net present value = $109. There is no change in revenues.245 If the criterion of a positive NPV is used.404 4.000 1.00 $100.000 Net cash outlay = $75.968 21.000 4.000 11. The correct definition of cash flows for capital budgeting purposes (equation 2-13) is: CF = (∆Rev − ∆VC) (1 − τc) + τc ∆dep In this problem Rev = revenues.000 Year 0 PVIF @ 12% 1.000 .4 dep = depreciation = 2.

Figure S2. Note also that the function is undefined at IRR = −1.4(2.83 344. it should be rejected (even though the IRR is greater than the cost of capital). At an opportunity cost of capital of 10 percent.1 The internal rate of return ignores the order of cash flows 1 There is a second IRR at −315.600 The net present value of the project is NPV = −10.000 −200 @ 10% 363.000 + 2. the annual net cash flows for years one through five are CF = 3.41 Figure S2.75%.64 330. This is an interesting example which demonstrates another difficulty with the IRR technique.52 −640.46 297.10 1 Discounted Cash Flows @ 15% @ 16% 347. therefore.78 −578.223. The NPV at different positive rates of return is @ 0% 400 400 −1.991) = −2.Chapter 2 Investment Decisions: The Certainty Case 9 Therefore.44 −7.58 −751.000) = 2. the project has a negative NPV. but it has no economic meaning. The IRR on this project is approximately 15.4) + .66 1.27 −657.83 302. .8 percent.32 −57.40 Therefore. namely.23 @ 20% 333. 6. that it does not consider the order of cash flows.33 277. the project should be rejected.600(2.1 graphs NPV versus the discount rate.000(1 − .70 32.

8. the discount factor for a three year annuity paid in arrears (at 10%). we see that the IRR is −200%.4) + .4(400) = 336 + 160 = 496 NPV = 496 (PVIFa: 10%. it would be “double counting” to include financing costs (or the tax changes which they create) in cash flows for capital budgeting purposes.2 An IRR calculation internal rates of return for the project. as well as their costs.55 The project should be accepted. Fourth Edition 7. By inspection. 3 years.5(180) = 145 + 90 = 235 * Note: PVIFa: 10%. 3 years)* – 1. . First calculate cash flows for capital budgeting purposes: CFt = (∆Revt − ∆VCt )(1 − τ c ) + τ c ∆dep = (0 − (−290))(1 − . The cash flows are: (∆Rev − ∆VC − ∆FCC − ∆dep)(1 − τ c ) + τ c ∆dep = (200 − (−360) − 0 − 0)(1 − .2 plots NPV against various discount rates for this particular set of cash flows.5) + . These rates are the Figure S2. But how is this determined? One way is to graph the NPV for a wide range of interest rates and observe which rates give NPV = 0. 9. Sources of financing. We are told that the IRR for these cash flows is −200%.200 = 496 (2.200 = 33.10 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. All of the information about the financing of the project is irrelevant for computation of the correct cash flows for capital budgeting. Therefore. Figure S2. are included in the computation of the cost of capital. These are the cash flows for project A which was used as an example in section E of the chapter.487) – 1.

909 . B + C is preferred. at the end of the first time period. NPV does not. Because C is an independent choice. Using the method discussed in section F. NPV does obey the value additivity principle.00 .91 0 . Now A + C is preferred. we have 5.826 . Project C has a three-year payback.75 . project B would be preferable according to both capital budgeting techniques. NPV = $1. On the other hand. if projects A and B are mutually exclusive.00 0 1.Chapter 2 Investment Decisions: The Certainty Case 11 Next.000 and expects to earn IRR. Project B has a one-year payback.25 1.000(1 + IRR) During the second period the firm borrows from the project at the opportunity cost of capital. Once Project C is combined with A or B. The net present values are calculated below: Year 0 1 2 3 PVIF 1.00 = 890. Project (B + C) has a three-year payback.00 = −9.65 −. B was preferred. Its NPV is simply the sum of the NPV’s of B and C separately. Therefore.91 Project A has a two-year payback.00 0 0 2.15 NPV(B + C) = 1.25 A+C −2 0 2 2 B+C −2 1 0 4 5 NPV(A + C) = 1.12 The project should be rejected because it has negative net present value. the results change if we use the payback criterion. this is not true.15. Therefore. Payback violates value additivity. it should be irrelevant when considering a choice between A and B.000 − 5.66 C −1 0 0 3 PV (C) −1. in the first year the firm invests $5.10 B −1 1 0 1 PV (B) −1.91.89 − 900.3 of this chapter. Therefore. with payback. k. 11.000(1 + IRR)) . However. Previously.751 A −1 0 2 −1 PV (A) −1. 5 years) − I 0 = 235(3. NPV = $1. The amount borrowed is (10.000 . 10. Project (A + C) has a two-year payback. calculate the NPV: NPV = ∑ CFt − I0 t t =1 (1 + WACC) = (CFt ) (present value annuity factor @ 10%.75 −.791) − 900.

000(1 + IRR)) (1 + k) The firm then lends 3.000 = (10. Fourth Edition By the end of the second time period this is worth (10.000 − 5.10 − 10.10) Solving for IRR.000 − 1 = IRR = 45.000 at the end of the second time period: 3.12 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. we have 3.000(1 + IRR)) (1.000 1.000 − 5.45% −5.000 .

we need to know the marginal rate of substitution between return and risk. U’ (E + σZ). 2. Since the integral in the numerator is preceded by a minus sign. MU(W) > 0. the same f(Z)) negative returns. dMU(W)/dW > 0. .e.19: dE − ∫ U′(E + σZ)Zf(Z)dZ = dσ ∫ U′(E + σZ)f(Z)dZ (3. In order to know the shape of a risk-lover’s indifference curve. a risk lover has positive marginal utility of wealth. +Z.19) The denominator must be positive because marginal utility. look at Figure S3.2. i. This leads to indifference curves like those shown in Figure S3. is always higher than the marginal utility of equally likely (i. The minimum set of conditions includes (a) • • • • • The five axioms of cardinal utility complete ordering and comparability transitivity strong independence measurability ranking (b) Individuals have positive marginal utility of wealth (greed). of any level of wealth is positive. Therefore.Chapter 3 The Theory of Choice: Utility Theory Given Uncertainty 1.e. the result must be positive. In order to see that the integral in the numerator is positive. −Z.6. To do so. is positive and because the frequency. As shown in Figure 3. when all equally likely returns are multiplied by their marginal utilities. and summed. f(Z). look at equation 3. the entire numerator is negative and the marginal rate of substitution between risk and return for a risk lover is negative. E[U(W)] < U[E(W)]. (c) The total utility of wealth increases at a decreasing rate (risk aversion).. which increases with increasing wealth..1 on the following page. (d) The probability density function must be a normal (or two parameter) distribution. The marginal utility of positive returns. matched.

5(8.2 Indifference curves of a risk lover 3.5ln(6.000) = . the individual would be indifferent between the gamble and $4.4967825 eln W = W e8.98 for sure.14 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.29405) + . Fourth Edition Figure S3.000) + .26178 = $3.898.02 for insurance.02. is $4.000) = .98 = W Now the individual would be willing to pay up to $127.26178 eln W = e8. Therefore. . he would not buy insurance for $125. (b) The second gamble.5ln(5.1 Total utility of normally distributed returns for a risk lover Figure S3.872.5(8.000. This amounts to a risk premium of $101.5(8. (a) E[U(W)] = .517193) = 8. Its expected utility is E[U(W)] = . Since insurance costs only $125.4967825 = $4.000 plus or minus $1.98 = W Therefore.898.699515) = 8.5ln(3.006368) + .5(8. given his first loss.5ln(4. he will buy it.000) + .

Chapter 3

The Theory of Choice: Utility Theory Given Uncertainty

15

4. Because $1,000 is a large change in wealth relative to $10,000, we can use the concept of risk aversion in the large (Markowitz). The expected utility of the gamble is E(U(9,000,11,000; .5)) = .5 U(9,000) + .5 U(11,000) = .5 ln9,000 + .5 ln11,000 = .5(9.10498) + .5(9.30565) = 4.55249 + 4.652825 = 9.205315 The level of wealth which has the same utility is

ln W = 9.205315 W = e9.205315 = $9,949.87
Therefore, the individual would be willing to pay up to $10,000 − 9,949.87 = $50.13 in order to avoid the risk involved in a fifty-fifty chance of winning or losing $1,000. If current wealth is $1,000,000, the expected utility of the gamble is E(U(999,000, 1,001,000; .5)) = .5 ln 999,000 + .5ln1,001,000 = .5(13.81451) + .5(13.81651) = 13.81551 The level of wealth with the same utility is

ln W = 13.81551 W = e13.81551 = $999,999.47
Therefore, the individual would be willing to pay $1,000,000.00 − 999,999.47 = $0.53 to avoid the gamble. 5. (a) The utility function is graphed in Figure S3.3.

U(W) = − e −

aW

16

Copeland/Shastri/Weston • Financial Theory and Corporate Policy, Fourth Edition

Figure S3.3 Negative exponential utility function The graph above assumes a = 1. For any other value of a > 0, the utility function will be a monotonic transformation of the above curve. (b) Marginal utility is the first derivative with respect to W. U′(W) = dU(W) = −( −a)e −aW > 0 dW

Therefore, marginal utility is positive. This can also be seen in Figure S3.3 because the slope of a line tangent to the utility function is always positive, regardless of the level of wealth. Risk aversion is the rate of change in marginal utility. U′′(W) = dMU(W) = a( − a)e − aW = −a 2 e − aW < 0 dW

Therefore, the utility function is concave and it exhibits risk aversion. (c) Absolute risk aversion, as defined by Pratt-Arrow, is ARA = − ARA = − U′′(W) U′(W) −a 2 e − aW =a ae − aW

Therefore, the function does not exhibit decreasing absolute risk aversion. Instead it has constant absolute risk aversion. (d) Relative risk aversion is equal to RRA = W(ARA) = −W = Wa Therefore, in this case relative risk aversion is not constant. It increases with wealth. U′′(W) U′(W)

Chapter 3

The Theory of Choice: Utility Theory Given Uncertainty

17

6. Friedman and Savage [1948] show that it is possible to explain both gambling and insurance if an individual has a utility function such as that shown in Figure S3.4. The individual is risk averse to decreases in wealth because his utility function is concave below his current wealth. Therefore, he will be willing to buy insurance against losses. At the same time he will be willing to buy a lottery ticket which offers him a (small) probability of enormous gains in wealth because his utility function is convex above his current wealth.

Figure S3.4 Gambling and insurance 7. We are given that Also, we know that Transposing, we have A>B>C>D U(A) + U(D) = U(B) + U(C) U(A) − U(B) = U(C) − U(D) ∂U >0 ∂W Therefore, from (1) and (2) we know that U(A) − U(B) U(C) − U(D) < A−B C−D Using equation (3.1), equation (3.3) becomes 1 1 < A−B C−D A−B>C−D A+D>C+B 1 1 1 1 A+ D> C+ B 2 2 2 2 1 1 1  1  U  (A) + (D)  > U  (C) + (B)  2 2 2  2  In general, risk averse individuals will experience decreasing utility as the variance of outcomes increases, but the utility of (1/2)B + (1/2)C is the utility of an expected outcome, an average. (3.3) ∂2U <0 ∂W2 (3.1)

Assuming the individual is risk-averse, then and (3.2)

18

Copeland/Shastri/Weston • Financial Theory and Corporate Policy, Fourth Edition

8. First, we have to compute the expected utility of the individual’s risk. E(U(W)) = ∑ pi U(Wi ) = .1U(1) + .1U(50,000) + .8U(100,000) = .1(0) + .1(10.81978) + .8(11.51293) = 10.292322 ln W = 10.292322 W = e10.292322 W = 29,505 The maximum insurance premium is Risk premium = E (W) – certainty equivalent = $85,000.1 − $29,505 = $55,495.1 9. The utility function is

Next, what level of wealth would make him indifferent to the risk?

U(W) = − W −1
Therefore, the level of wealth corresponding to any utility is W = –(U(W))
–1

Therefore, the certainty equivalent wealth for a gamble of ±1,000 is W. W = −[.5(−(W + 1,000)−1 ) + .5(−(W − 1,000)−1 )]−1 The point of indifference will occur where your current level of wealth, W, minus the certainty equivalent level of wealth for the gamble is just equal to the cost of the insurance, $500. Thus, we have the condition W − W = 500     1 W − −  = 500 −1 −1  .5  + .5   W 1,000 W 1,000 + −       −1 W−  = 500 −W   2   W = 1,000,000    −W2 − 1,000,000  W+  = 500 −W   2 2 W − W + 1,000,000 = 500W W = 2,000

50 .693 ln 4 = 1. Table S3. Consequently.130  N (1/2) N+1 2 N . no heads). no one has ever paid an infinite amount to accept the gamble. The expected utility of the gamble is E(U) = ∑ ( 1 2)i +1 ln 2i i=0 N E(U) = 1 2 ∑ ( 1 2)i i ln 2 i =0 N E(U) = 1 2 ln 2 ∑ i =0 N i 2i Proof that ∑2 i =0 ∞ i i = 2 follows: First. The reason is that people are usually risk averse.173 .000 . expected payoffs. we have ∑2 i=0 ∞ 1 i = 1 + 12 + 14 + 1 + ⋅⋅⋅ 8 = 1+ 1/ 2 =2 1 − 1/ 2 .Chapter 3 The Theory of Choice: Utility Theory Given Uncertainty 19 Therefore.5) probability of ending after the second flip (one head and one tail). Below that level of wealth you will pay for the insurance while for higher levels of wealth you will not.50 ln 2 = N ln 2 N N ln 2 N 1 = 0 2 + 2 The gamble has a .1 shows the payoffs. note that the infinite series can be partitioned as follows: ∑ 2 =∑ i=0 i ∞ i 1+ i −1 ∞ 1 ∞ i −1 =∑ i + ∑ i 2i i=0 i=0 2 i =0 2 ∞ Evaluating the first of the two terms in the above expression.079  E U(Payoff) . Table S3.50  U(Payoff) ln 1 = . which is infinite.e.50 .173 .386 ln 8 = 2.000 ln 2 = . they would be willing to pay an amount whose utility is equal to the expected utility of the gamble.50 . and utility of payoffs for n consecutive heads. a (. you will be indifferent.1 Number of Consecutive Heads = N 0 1 2 3  Probability = (1/2)n+1 1/2 1/4 1/8 1/16  Payoff N =2 1 2 4 8  E(Payoff) $. if your current level of wealth is $2. 10. and so on.5 probability of ending after the first coin flip (i. The expected payoff of the gamble is the sum of the expected payoffs (column four)..000. However.

an individual with a logarithmic utility function will pay $2 for the gamble. we have the desired proof that i ∞ 1 ∞ i −1 =∑ i + ∑ i = 2 + 0 = 2 ∑ i i=0 2 i=0 2 i=0 2 ∞ Consequently.98) + 5.000(.005) + 30.20 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.005) = $350 . we have i −1 1 1 1 1 = −1 + + + + + ⋅⋅⋅ i 2 4 8 16 i=0 2 ∞ i −1 = −1 + 1 = 0 ∑ i i=0 2 ∑ ∞ Adding the two terms. since i 2 i=0 N ∑2 i=0 N i i = 2 If the expected utility of wealth is ln2.98) + 5.000 insurance) = 0(. we have ∑ i=0 ∞ i −1 1 2 3 4 = −1 + 0 + + + + + ⋅⋅ ⋅ i 2 4 8 16 32 The above series can be expanded as −1 = −1 1 1 1 1 1 + + + + ⋅⋅⋅ = 4 8 16 32 2 1 1 1 1 + + + ⋅⋅⋅ = 8 16 32 4 1 1 1 + + ⋅⋅⋅ = 16 32 8 1 1 + ⋅⋅⋅ = 32 16  Therefore. we have E(U) = 1/ 2 ln2∑ i = ln 2.000(.005) + 50.000(. (a) First calculate AVL from the insurer’s viewpoint.000(.000 insurance) = 0(.000(.98) + 5.01) + 10.000 insurance) = 0(. since the insurer sets the premiums.005) = $300 AVL 3 ($50.005) = $250 AVL 2 ($40.005) + 40. 11.000(. the corresponding level of wealth is U(W) = ln2 eln2 = W = $2 Therefore. Fourth Edition Evaluating the second term.01) + 10. AVL1 ($30.000(.000(.000(.01) + 10.

995) + 1.07) ≅ 7.0374(1.14 − 1.9598 With $40.01) 5 + 2.14 0. find the expected utility of wealth for each amount of insurance.07) = 7.000 loss (P = .005) 5 + 2.0 ≅ 6. calculate the insuree’s ending wealth and utility of wealth in all contingencies (states).005) = 1.6312 $20.000 insurance: E(U(W)) = 1. The utility of each ending wealth can be found from the utility function U(W) = ln W.000’s) U(W) = ln W With no insurance 1.000 loss (P = .14 − .8148 $10.000 insurance 1.9582 With $30. Contingency Values Of Wealth And Utility of Wealth (Savings = $20.11 With $40.11 1.995) 5 + 2.0 = 2.005) 5 + 2. Casadesus is $50.000) Utility of End-of-Period Wealth Wealth (in $10.9608 No loss (P = .000 $40.000 insurance 1.03 − 2 ≅ 5.01) + 1.6312(.105 With $50.14 − .995) 5 + 2.14 1.000 loss (P = .9657 No loss (P = .07) ≅ 7.0327(1.64 1.005) = 1.14 − .) Finally.8148(.000 insurance: E(U(W)) = 1.14 − .8931 $5.995) + 1.8931(.5 = 6.Chapter 3 The Theory of Choice: Utility Theory Given Uncertainty 21 We can now calculate the premium for each amount of coverage: Amount of Insurance $30.8091(.105 1.000 $50.98) 5 + 2(1.035 − 1.005) 5 + 2. (See Table S3.14 With $30.7608 $50. given his utility function.000 insurance 1.2a.9601 Therefore.000 Premium 30 + 250 = $280 27 + 300 = $327 24 + 350 = $374 Next.9615(.8091 $10.000 insurance: E(U(W)) = 1.2a With no Insurance: E(U(W)) = 1.0 = 6.000.10 Table S3.005) = 1.9601 No loss (P = 1.14 − .9615 No loss (P = .0) 5 + 2.07) ≅ 7.000 loss (P = .7608(.98) + 1.9608(.000 loss (P = . .9657(.0280(1. the optimal insurance for Mr. Assume he earns 7 percent on savings and that premiums are paid at the beginning of the year.005) + 0.9600 With $50.005) 5 + 2. E(U(W)) = ∑ Pi U(Wi ) i and choose the amount of insurance which yields the highest expected utility.14 − 5.14 − .

016 – . so we can calculate the expected utilities for each amount of insurance directly.995(1.4) – .24 − 1.005U(21.005(1.07) ≅ 39.000 loss (P = .005) $50.035 − 1.21 5 + 34.3140 1.3404) = 1.4)) + .0) 5 + 32.005(200/61.3543) + .07) ≅ 39.3413 1.07) = 39.2b above for these calculations.005(200/21.4) – .047 = –2. Fourth Edition Table S3.24 − 5.000’s) With no insurance No loss (P = .000 insurance No loss (P = .3413) + .07) ≅ 39.000 insurance: E(U(W)) = .24 − .745 – .005(U(61.366162 = 1.366071 With $50. (c) The end-of-period wealth for all contingencies has been calculated in part a).005(1.0327(1.74 5 + 34.030 – .2308) = 1.000 insurance: E(U(W)) = .24 − .24 − .24 − .000) Utility of Wealth U(W) = ln W (Wealth in $100.3404 1.0374(1.98) $5.005) With $30.366038 With $40.838 With $30.0 ≅ 38.3661 End-of-Period Wealth (in $10.3671 1.995(1.205 5 + 34. while a less wealthy individual may choose maximum coverage. the analysis illustrates that a relatively wealthy individual may choose no insurance.3662 1.22 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.028(1.005 (1.98(U(71.005(1.24 5 + 34.366092 = 1.3663) + .4)) + .3662) + .24 5 + 34.01(U(66.24 − .005) With $50.01) $10. only with $320.) With no Insurance: E(U(W)) = .01(1.366092 The optimal amount of insurance in this case is no insurance at all.000 loss (P = .2308 1.24 − .3140) .20 (b) Follow the same procedure as in part a).21 5 + 34.000 loss (P = .995) $10.000 insurance No loss (P = .98(200/71.0 = 38.2b Contingency Values of Wealth and Utility of Wealth(Savings = $320.000 insurance No loss (P = 1.4) = –2.98(1.000 insurance: E(U(W)) = (1)1.000 in savings instead of $20.0 = 34.000’s) 1.00(1.995) $20. Although the numbers are close with logarithmic utility.000.000 loss (P = .4)) + .3543 1.24 5 + 34.005) With $40. (See Table S3.4) – .3663 1.01(200/66.5 = 38.4) = –. With no insurance: E (U(W)) = .03 − 2 ≅ 37.3671) + .205 5 + 34.000 loss (P = .

05) – .000 insurance: E(U(W)) = . Case 1 (a) Second order dominance—B dominates A because it has lower variance and the same mean.819 With $40.995(200/71. It lies to the right of B.8172 With $50.995(U(71.1) – .005(200/51.799 – . with this utility function.1)) = – . Eventually.05)) = – .8169 Hence. he would prefer not to take out any insurance.1)) + .995(U(71. Mr. make his savings account = $400.000 insurance: E(U(W)) = (1)( −200/71) = −2.005(200/61. To see this.05) = –2.05) + . 1 Properties of this utility function. Casadesus would renew his policy for $50.8008 – . Because returns are normally distributed. (b) First order dominance—There is no dominance because the cumulative probability functions cross.0164 = –2. the mean and variance are the only relevant parameters. Case 2 (a) Second order dominance—A dominates B because it has a higher mean while they both have the same variance. 12. as his savings account is increased he prefers to bear greater and greater amounts of risk.000W < 0 −MU′ W ARA = = 2W-1 > 0 MUW -3 nonsatiation risk aversion ∂ARA = −2W-2 < 0 ∂W RRA = W(ARA) = 2 > 0 ∂RRA =0 ∂W decreasing absolute risk aversion constant relative risk aversion Since the individual has decreasing absolute risk aversion.000 insurance: E(U(W)) = .1) = –2.005(U(61.000W-2 > 0 MU′ W = −400.005(U(51. . U(W) = −200.Chapter 3 The Theory of Choice: Utility Theory Given Uncertainty 23 With $30.020 = –2.000W− : MUW = 200.995 (200/71. (b) First order dominance—A dominates B because its cumulative probability is less than that of B.000. once his wealth is large enough.000.

0 3.36) = 6.5(7.29) = 3.2(31.69) = 2.5 First order dominance not possible 13. (b) Second order stochastic dominance may be tested as shown in Table S3.0 2.4 3.4 .1 Y pi Yi . as shown in Figure S3. (a) Prob X .1 .8 30 3.888 .49) = 59.7 −1.6 5.4 Xi − E(X) −16. .2 X −10 5 10 12 pi Xi −1.89) = .272 var (X) = 37. it is not possible for B to dominate A according to the first order criterion.1(268.6 pi(Xi − E(X)) .2 .3 on the following page.24 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.3 .7 24. there is no second order dominance.4 2 3 1. Because Σ(F − G) is not less than (or greater than) zero for all outcomes.6.7 −2.7 Yi − E(Y) −3.896 .784 .010 2 X is clearly preferred by any risk averse individual whose utility function is based on mean and variance.0 E(Y) = 5.2(2.0 2.4 E(X) = 6.3(12.645 .4(1. Figure S3.5 4 .96) = .5 shows an example.2(13. Figure S3. Fourth Edition Case 3 (a) Second order dominance—There is no dominance because although A has a lower variance it also has a lower mean.96) = 26.2 .4 −1.1(590.5 .049 var(Y) = 66.578 . because X has a higher mean and a lower variance than Y.738 .96) = 3. (b) First order dominance—Given normal distributions.3 pi (Yi − E(Y)) .840 2 Prob Y .

Chapter 3

The Theory of Choice: Utility Theory Given Uncertainty

25

Table S3.3 Outcome −10 −9 −8 −7 −6 −5 −4 −3 −2 −1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Prob(X) .1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 .4 0 0 0 0 .3 0 .2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1.0 Prob(Y) 0 0 0 0 0 0 0 0 0 0 0 0 .2 .5 .2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 .1 1.0 Σ Px = F .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .5 .5 .5 .5 .5 .8 .8 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 Σ Py = G 0 0 0 0 0 0 0 0 0 0 0 0 .2 .7 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 1.0 F−G .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 −.1 −.6 −.8 −.4 −.4 −.4 −.4 −.4 −.1 −.1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 0 Σ (F − G) .1 .2 .3 .4 .5 .6 .7 .8 .9 1.0 1.1 1.2 1.1 .5 −.3 −.7 –1.1 −1.5 –1.9 –2.3 –2.4 –2.5 –2.4 –2.3 –2.2 −2.1 –1.9 –1.8 –1.7 –1.6 –1.5 –1.4 –1.3 –1.2 –1.1 –1.0 –.9 –.8 –.7 –.6 –.6

Because Σ (F − G) is not less than (or greater than) zero for all outcomes, there is no second order dominance.

26

Copeland/Shastri/Weston • Financial Theory and Corporate Policy, Fourth Edition

Figure S3.6 Asset X is preferred by mean-variance risk averters 14. (a) Table S3.4 shows the calculations. Table S3.4 pi .1 .2 .4 .2 .1 Co. A 0 .50 1.00 2.00 3.00 Co. B −.50 −.25 1.50 3.00 4.00 pi Ai 0 .10 .40 .40 .30 1.20 pi Bi pi [A − E(A)]2 .144 −.05 .098 −.05 .016 .60 .128 .60 .324 .40 .710 1.50 σA = .84 σB = 1.44 pi[B − E(B)]2 .4000 .6125 0 .4500 .6250 2.0875

E(A) = 1.20, E(B) = 1.50,

(b) Figure S3.7 shows that a risk averse investor with indifference curves like #1 will prefer A, while a less risk averse investor (#2) will prefer B, which has higher return and higher variance.

Figure S3.7 Risk-return tradeoffs (c) The second order dominance criterion is calculated in Table S3.5 on the following page. 15. (a) False. Compare the normally distributed variables in Figure S3.8 below. Using second order stochastic dominance, A dominates B because they have the same mean, but A has lower variance. But there is no first order stochastic dominance because they have the same mean and hence the cumulative probability distributions cross.

Chapter 3

The Theory of Choice: Utility Theory Given Uncertainty

27

Figure S3.8 First order stochastic dominance does not obtain (b) False. Consider the following counterexample. Table S3.5 Return −.50 −.25 0 .25 .50 .75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 (Problem 3.14) Second Order Stochastic Dominance Prob(A) 0 0 .1 0 .2 0 .4 0 0 0 .2 0 0 0 .1 0 0 0 0 1.0 Prob(B) .1 .2 0 0 0 0 0 0 .4 0 0 0 0 0 .2 0 0 0 .1 1.0 F(A) 0 0 .1 .1 .3 .3 .7 .7 .7 .7 .9 .9 .9 .9 1.0 1.0 1.0 1.0 1.0 G(B) .1 .3 .3 .3 .3 .3 .3 .3 .7 .7 .7 .7 .7 .7 .9 .9 .9 .9 1.0 F−G −.1 −.3 −.2 −.2 0 0 .4 .4 0 0 .2 .2 .2 .2 .1 .1 .1 .1 0 Σ (F − G) −.1 −.4 −.6 −.8 −.8 −.8 −.4 0 0 0 .2 .4 .6 .8 .9 1.0 1.1 1.2 1.2

Because Σ (F − G) is not always the same sign for every return, there is no second order stochastic dominance in this case. Payoff $1 $2 $3 $4 Prob (A) 0 .5 .5 0 1.0 Prob (B) .3 .1 .3 .3 1.0 E(A) = $2.50, E(B) = $2.60, F (A) 0 .5 1.0 1.0 G (B) .3 .4 .7 1.0 G (B) − F(A) .3 −.1 −.3 0

var(A) = $.25 squared var(B) = $1.44 squared

The cumulative probability distributions cross, and there is no first order dominance.

28

Copeland/Shastri/Weston • Financial Theory and Corporate Policy, Fourth Edition

(c) False. A risk neutral investor has a linear utility function; hence he will always choose the set of returns which has the highest mean. (d) True. Utility functions which have positive marginal utility and risk aversion are concave. Second order stochastic dominance is equivalent to maximizing expected utility for risk averse investors. 16. From the point of view of shareholders, their payoffs are Project 1 Probability Payoff .2 0 .6 0 .2 0 Project 2 Probability Payoff .4 0 .2 0 .4 2,000

Using either first order or second order stochastic dominance, Project 2 clearly dominates Project 1. If there were not limited liability, shareholder payoffs would be the following: Project 1 Probability Payoff .2 −4000 .6 −3000 .2 −2000 Project 2 Probability Payoff .4 −8000 .2 −3000 .4 2,000

In this case shareholders would be obligated to make debt payments from their personal wealth when corporate funds are inadequate, and project 2 is no longer stochastically dominant. 17. (a) The first widow is assumed to maximize expected utility, but her tastes for risk are not clear. Hence, first order stochastic dominance is the appropriate selection criterion. E(A) = 6.2 E(B) = 6.0 E(C) = 6.0 E(D) = 6.2 E(E) = 6.2 E(F) = 6.1

One property of FSD is that E(X) > E(Y) if X is to dominate Y. Therefore, the only trusts which might be inferior by FSD are B, C, and F. The second property of FSD is a cumulative probability F(X) that never crosses but is at least sometimes to the right of G(Y). As Figure S3.9 shows, A > C and D > F, so the feasible set of trusts for investment is A, B, D, E.

10.6 −1.8 −1.7 0 .8 1.0 .5 .2 0 .8 0 1.5 .8 0 1.0 . they are also inferior by SSD.6 1. Σ(F(X) − G(Y)) are presented in Table S3.0 0 .1 −.2 −.0 −1.4 0 1.8 −1.2 B<E SSD (ED) 0 −.6 0 1.1 0 . If the sum of cumulative differences crosses the horizontal axis.1 0 . and E > D.6 .4 2.2 1. so the optimal investment is E.2 0 0 D<E * cumulative probability ** SSD calculated according to Σ (F(X) − G(Y)) where F(X) = cumulative probability of X and G(Y) = cumulative probability of Y.1 0 0 .3 . there is no second order stochastic dominance.1 −.1 .1 0 .0 1.7 1.9 .0 1. Table S3.0 1. E > B.2 0 . By SSD.0 1.6 on the following page and graphed in Figure S3.6 0 .0 .1 0 0 −.2 0 .4 0 . .4 0 .4 −.8 .2 −.8 1.5 .7 −1.9 −1.4 0 .0 .0 .4 −1.1 −. The pairwise comparisons of the remaining four funds.8 1.0 .4 0 .2 no 2nd order dominance SSD (EB) −.0 −1.6 .9 .5 −.0 .4 −1.6 1.9 First order stochastic dominance (b) The second widow is clearly risk averse.2 1.0 −2.2 0 .4 .2 −. so second order stochastic dominance is the appropriate selection criterion.1 −.2 −.2 −.8 0 . E > A.6 .2 −.6 Ret.4 0 1.2 1.0 .0 1.3 −.2 −.8 −.6 1.1 0 −.0 0 1.4 .4 0 1.0 .4 0 1.2 −.3 . Since C and F are eliminated by FSD.1 .5 .8 .0 1.1 0 0 −.5 −1.7 −.8 1.5 1. −2 −1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Second Order Stochastic Dominance SSD** P(A)* P(B) P(D) P(E) (BA) 0 .3 1.6 −1.4 −.1 −1.0 .2 A>B SSD SSD (DA) (EA) 0 0 .8 −.9 .4 .2 0 .8 1.0 .0 .5 .2 . as in the comparison of B and D.2 1.4 1.Chapter 3 The Theory of Choice: Utility Theory Given Uncertainty 29 Figure S3.9 −2.2 0 0 0 0 0 A>D A<E SSD (DB) −.6 −.4 −.0 .5 0 1.7 .

F. Means and variances of the six portfolios are shown in Table S3. D has the highest mean of the four remaining trusts.7. C. or higher mean and lower or equal variance. B. but also the highest variance.30 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. C. X must have higher or equal mean and lower variance than Y. To dominate Y.10 Second order stochastic dominance . E > A. Fourth Edition 18. D. F and A > B. normal). (a) Mean-variance ranking may not be appropriate because we do not know that the trust returns have a two-parameter distribution (e. The next in rank cannot be determined. By mean-variance criteria..g. D. Figure S3. The only other unambiguous dominance is C > B.

00 28.2 6. the rankings of suboptimal portfolios are not consistent across the two selection procedures. C. B. F.1 var(X) 1. D.2 6.2 6.7 A B C D E F E(X) 6. D > F A > B.89 (b) Mean-variance ranking and SSD both select trust E as optimal.Chapter 3 The Theory of Choice: Utility Theory Given Uncertainty 31 Table S3.0 6. C > B FSD SSD M-V . Optimal A. E E E Dominance Relationships A > C.0 6.96 26.36 0. A > D A > B. However.80 2. D.36 26.

20 P1 = . B) Pj = price of pure security j (j = 1.Chapter 4 State Preference Theory 1. 30P1 + 10P2 = 5 20P1 + 40P2 = 10 Multiplying the first equation by 4 and subtracting from the second equation.10 .20 = . (a) State 1 $30 $20 Payoff State 2 $10 $40 Price PA = $5 PB = $10 Security A Security B (b) The prices of pure securities are given by the equations below: P1QA1 + P2QA2 = PA P1QB1 + P2QB2 = PB Qij = dollar payoff of security i in state j Pi = price of security i (i = A. 20P1 + 40P2 = 10 2 + 40P2 = 10 40P2 = 8 P2 = .10 P2 = . 2) Substituting the correct numbers. 20P1 + 40P2 = 10 −[120P1 + 40P2 = 20] 100P1 = 10 P1 Substituting into the first equation.

Pi. 24P1 + 40P2 = 44 −[24P1 + 10P2 = 20] 30P2 = 24 P2 = 24 / 30 = . 6 St.8. Substitution of payoffs and prices for securities j and k in the situation given yields 12P1 + 20P2 = 22 24P1 + 10P2 = 20 Multiplying the first equation by two. 20 .5 and P2 = . Pi = P1Qil + P2Qi2 = . 10 4P1 + 36P2 = St.8) = 22 12P1 = 22 – 16 P1 = 6/12 = .8(10) =3+8 = $11. solve two equations with two unknowns: 6P1 + 6P2 = St.5 (b) The price of security i. and the prices of pure securities for states 1 and 2. 12P1 + 20(. (a) The payoff table is: Nova Nutrients = j Galactic Steel = k S1 = Peace St. P1 and P2. and subtracting the second equation from the first. can be determined by the payoff of i in states 1 and 2.Chapter 4 State Preference Theory 33 2.8 Substituting . and Pj is the price of security j. P1 = . P1 is the price of a pure security which pays $1 if state 1 occurs. are given below: P1Qj1 + P2Qj2 = Pj P1Qk1 + P2Qk2 = Pk where Qj1 is the payoff of security j in state 1. 6 St. 4 S2 = War St. P1 and P2.8 for P2 in the first equation. Thus.5(6) + . 36 To find the price of pure securities. (a) The equations to determine the prices of pure securities.00 3. From part a) we know the prices of pure securities.

): If he sells –nj shares of j.S.S.N. 20 −[36P1 + 36P2 = St. Let W1 = his final wealth if peace prevails.000) 20 –2nj + 200 = –6nj nj = –50 shares of j (N.000/10 = 100 nk = W0/Pk = 1.800 St.34 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. 1. he receives –njPj.) . he can buy 50 shares. Fourth Edition Multiplying the first equation by six. W2 = nkQk2 = 50(36) = 1. and with his initial W0 he will have –njPj + W0. and subtracting it from the second equation. (c) For sales of j (N.) and purchases of k (G. he must pay out at most –njQj1. Therefore. If he buys G.N. which will return at least [(–njPj + W0)/Pk]Qk1.4167 (b) Let nj = number of Nova Nutrients shares and nk = number of Galactic Steel shares.: W1 = njQj1 = 100(6) = 600 St.N. 4P1 + 36P2 = St.25) + 6P2 = 10 P2 = . If he buys only Galactic Steel.000/20 = 50 If he buys only Nova Nutrients. 60] −32P1 = − 40 P1 = St. W2 = njQj2 = 100(6) = 600 St. he can buy 100 shares.: W1 = nkQk1 = 50(4) = 200 St. If he buys N. the minimum –nj is determined by −n j Pj +W0 Pk Qk1 = −n jQ j1 4 = −6n j (−10n j + 1. Then nj = W0/Pj = 1.25 6(1. and W2 = his final wealth if war prevails. With this he can buy at most (–njPj + W0)/Pk shares of k.

Therefore.) (d) Let Pa = price of Astro Ammo. and with his initial W0 he will have –nkPk + W0. the minimum –nk is determined by −n k Pk + W0 Q j2 = −n k Q k2 Pj (−20n k + 1.2 W2 ÷ .2W1. which will return at least [(–nkPk + W0) /Pj]Qj2.S.8 W2−.8W1−.Chapter 4 State Preference Theory 35 For sales of k and purchase of j: If he sells –nk shares of k.25(28) + .2 .4167(36) = 35 + 15 = 50 St.2 .8 W2. (e) See Figure S4. he must pay out at most –nkQk2. (f) The slope of the budget line must equal the slope of the utility curve (marginal rate of substitution) at optimum.000) 6 = −36n k 10 –12nk + 600 = –36nk nk = –25 shares of k (G. as given in the equation below: −∂W2 / ∂W1 = −[∂U / ∂W1 ÷ ∂U / ∂W2 ] With utility function U = W1. he receives –nkPk.8 = 4W1−1 W2 = 4W2 / W1 . this equality results in . With this he can buy at most (–nkPk + W0)/Pj shares of j. Then Pa = P1Qa1 + P2Qa2 = 1.1 on the following page.

1.36 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. 4W2 = 3W1 W1 = (4/3)W2 The wealth constraint is: W0 = P1W1 + P2W2 Substituting the correct numbers.1 State payoffs in peace and war In equilibrium. ∂W2 / ∂W1 = P1 / P2 4W2 / W1 = P1 / P2 = (5/ 4) /(5/12) = (12 / 4) = 3 Therefore. Fourth Edition Figure S4.000 = (5/4) (4/3)W2 + (5/12)W2 = (20/12)W2 + (5/12)W2 = (25/12)W2 W2 = (1.000)(12/25) = $480 W1 = (4/3)480 = $640 .

solve the two simultaneous equations W1 = njQj1 + nkQk1 W2 = njQj2 + nkQk2 Substituting the correct numbers. and nk the number she can buy if she buys only k. with this she can buy at most (–njPj + W0)/Pk shares of k which will return at least [(–njPj + W0)/Pk]Qk2. 640 = 6nj + 4nk 480 = 6nj + 36nk Subtracting the second equation from the first yields 160 = –32nk nk = –5 Substituting –5 for nk in equation 2 gives a value for nj: 480 = 6nj – 36(5) = 6nj – 180 660 = 6nj nj = 110 Hence (nj = 110. Then W W 1.200 1. n k = 0 = = 100 (a) Pj 10 Pk 12 If she buys j: W1 = njQj1 = 120(10) = $1.440 final wealth in state 2 If she buys k: W1 = nkQk1 = 100(20) = $2. she receives –njPj.200 nj = 0 = = 120. Let nj = the number of shares the investor can buy if she buys only j. in this case the investor buys 110 shares of Nova Nutrients and issues five shares of Galactic Steel.Chapter 4 State Preference Theory 37 To find optimal portfolio.000 final wealth in state 1 W2 = nkQk2 = 100(8) = $800 final wealth in state 2 (b) For sales of j and purchases of k: If she sells –nj shares of j. 4. the minimum –nj is determined by: −n j Pj +W0 Pk (Qk2 ) = − n jQ j2 −10n j + 1.200 final wealth in state 1 W2 = njQj2 = 120(12) = $1. 400 = −36n j n j = −150 . she must pay out at most –njQj2. Therefore. 200 (8) = −12n j 12 − 20n j + 2. nk = –5) is the optimum portfolio. and with her initial wealth W0 she will have –njPj + W0.

50)12 =2+6 =8 (e) (The state contingent payoffs of a portfolio invested exclusively in security i are plotted in Figure S4. and subtracting the second equation from the first equation.) If the investor places all of her wealth in i.000 State 2: –150(12) + 225(8) = 0 Final wealth for sales of k and purchases of j: State 1: 300(10) – 150(20) = 0 State 2: 300(12) – 150(8) = 2. 20P1 + 24P2 = 20 −[20P1 + 8P2 = 12] 16P2 = 8 P2 = . and with her initial wealth W0 she will have –nkPk + W0. she must pay out at most –nkQk1.38 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.200 (10) = − 20n k 10 n k = − 150 Final wealth for sales of j and purchases of k: State 1: –150(10) + 225(20) = 3.2. she receives –nkPk. solve two equations for two unknowns as follows: 10P1 + 12P2 = 10 20P1 + 8P2 = 12 Multiplying the first equation by two. Therefore.40)5 + (.50 Substituting . with this she can buy at most (–nkPk + W0)/Pj shares of j.400 (c) To find the price of pure securities.40 (d) The price of security i is given by Pi = P1Qi1 + P2Qi2 = (. Fourth Edition For sales of k and purchases of j: If she sells –nk shares of k.50 for P2 in equation 1. the minimum –nk is determined by: −n k Pk + W0 (Q j1 ) = −n k Q k1 Pj −12n k + 1. 10P1 + 12(.5) = 10 P1 = . which will return at least [(–nkPk + W0)/Pj]Qj1. the number of shares she can buy is given by ni = W0 1.200 = = 150 Pi 8 .

5W2 / W1 Figure S4. j. This portfolio plots as the W2 intercept in Figure S4.5 = 0.4 and substituting the correct numbers. ∂W2 = (. her wealth in state one will be zero. (f) Set the slope of the budget line equal to the slope of the utility curve in accordance with the equation below: ∂W2 / ∂W1 = (∂U / ∂W1 ) ÷ (∂U / ∂W2 ) Given utility function U = W1.6W1−.4W2−6 W1.5W2 = 0.5W2/W1 = .4/.8W1 W1 = 1. resulting in zero wealth in state two.4 ) ÷ (.6 W2.800 If the investor sells k to purchase j.4 W2. The W1 intercept is the portfolio of j shares sold to buy k.2 on the following page.6 ) ∂W1 = 1. and k In equilibrium: dW2/dW1 = P1/P2 1.8 1.2 State payoffs for securities i.Chapter 4 State Preference Theory 39 Her wealth in state one would be niQi1 = 150(5) = $750 Her wealth in state two would be niQi2 = 150(12) = $1.875W2 .

we can use them to determine its equation.800 = 60nj + 3.800 – 10nj) ÷ 20] 4.200 = .800 Since both of these payouts lie on the budget constraint (see Figure S4.875(960) = 1.5W2 W2 = 1. the number of shares of each security to be purchased. (a) If we know the maximum payout in each state.200/1. S2 = $800 S1 = $900.800 – 10nj nk = (1.400. 5. Substituting the correct numbers. The equation for the line is W2 = a + bW1 . W1 = 1.40 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. and substituting this value into equation two: 20nk = 1.600 – 20nj 1. S2 = $1. Fourth Edition Wealth constraint: W0 = P1W1 + P2W2 1.800 Optimal portfolio: Solve the two simultaneous equations for the final wealth in each state: W1 = njQj1 + nkQk1 W2 = njQj2 + nkQk2 Solve for nk and nj.200 = 40nj nj = 30 nk = (1.800 – 10nj) ÷ 20 960 = 12nj + 8 [(1.3 on page 42).875W2) + . he can buy $720 = 90 shares with payout $8 If he spends $720 on security k. it will be possible to determine what an equal payout will be. If the individual uses 100 percent of his wealth to buy security j.25 = 960 W1 = 1.800 = 10nj + 20nk W2 = 960 = 12nj + 8nk Solving equation one for nk in terms of nj. he can obtain $720 = 80 shares with payout $9 S1 = $2.4(1.800 – 10nj)/20 nk = 75 The investor should buy 30 shares of j and 75 shares of k.

800 = a − 2 (900) 3 a = 2. The maximum wealth in state one is 2 W1 3 3/2(2.440 = W2 5/ 3 If you buy nj shares of asset j and nk shares of k.400 − W1 = 2.500) − Therefore.400 = $1.400)] 1. the slope is − 1.500 3 2 and the intercept is 3 1. Therefore.400 − W2 = W1 2 W1 (the budget constraint) 3 (equal payout) 2 W1 3 Substituting the second equation into the first. we want to find the point along the budget line where W2 = W1.Chapter 4 State Preference Theory 41 Substituting in the values of the two points. “a” and “b.8 and nj = 57. the payout of the risk-free asset is W1 = 2.400) = W1 = $3.440 nk = 28.440 Multiplying the first equation by 2 and subtracting.600 0 = 2. we have nj 20 + nk 60 = 2. your payout in states one and two will be State 1: State 2: nj10 + nk30 = 1.800 = a + b(900) –[800 = a + b(2.000 = b(–1.6 . regardless of the state of nature which occurs.880 −[n j 20 + n k 10 = 1. we obtain two equations with two unknowns.400 The maximum wealth in state two is $2.” 1.400.440] n k 50 = 1.440 nj20 + nk10 = 1.000 2 = b =− 1.400 − A risk-free asset is one which has a constant payout. We now have two equations and two unknowns W2 = 2. which we have already determined.

6Q1 – . Therefore.1).000 – .8 shares of asset k. the risk-free rate of return is 720 = 1 + rf = 1.2Q1 – .000 – .1) (4. Fourth Edition Figure S4.6Q1 + .4Q 2 3Q1 (4.4Q 2 3Q 2 or 1.42 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.440 for sure. In general.4 1 + = 0 50.2) or 1.2Q2 = 50.000 We can solve for C in (4.6Q1 – .8Q1 = 100.4Q2 (4.6Q1 − . 000 −.000 − .8Q2 (4.4Q2 = 50.6Q1 − .440 =2 720 rf = 100% (c) It would be impossible to find a completely risk-free portfolio in a world with more states of nature than assets (if all assets are risky).3) Take the partial derivative with respect to Q2 and set it equal to zero: −. It costs $720 and returns $1. 6. it is necessary to have at least as many assets as states of nature in order for complete capital markets to exist.3 The budget constraint (b) The risk-free portfolio contains 57.6 shares of asset j and 28. We to solve Max[log C + 2/3 log Q1 + 1/3 log Q2] subject to C + .000 – 1.6 2 + =0 50.440 1 + rf 1.4Q2) + 2/3 log Q1 + 1/3 log Q2] Take the partial derivative with respect to Q1 and set it equal to zero: −. Any attempt to solve the problem would require solving for three unknowns with only two equations. Max[log (50.4) .2) and substitute for C in (4. No feasible solution exists.

and 20.833.78 (a) The risk-averse individual will purchase 27.2) and (4. the investor divides his wealth equally between current and future consumption (which we would expect since the risk-free rate is zero and there is no discounting in the utility functions).8Q1 = 2. C = $50.666.000 also from (4.3) and (4. Substituting into (4.2).000 Q2 = 20.4Q2. but he buys more of pure security 1 (because its price per probability is lower) than of pure security 2.33 = $25.3333Q2 .67 – $8.000 – $16.333.2Q2 = 25.Chapter 4 State Preference Theory 43 Together.666.33. or Q1 = 1.4) imply 1.777. (b) From (4.4Q2 = 50.000 Hence.4).60 each for a total of $16. C = 1.3) yields 2.833.333. (4.78 units of pure security 1 at $0.67.777.33 units of pure security 2 at $0.33 hence Q1 = 27.40 each for a total of $8.

2.1 Skewed distribution of stock prices The reason stock prices are skewed right is because theoretically there is no upper bound to the price level a stock can attain. The equation for correlation between x and y. Given y = a – bx.y = cov(x. var(y) = var(a – bx) = E{[a – bx – E(a – bx)]2} = E{[a – bx – a – E(–bx)]2} = E{[–bx + bE(x)]2} = (–b)2E{[x – E(x)]2} = b2 var(x) Therefore. var(y). y).Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 1. with limited liability. while. the probability distribution is bounded on the left by P = 0. because the standard deviation must be positive σ y = bσ x and σ x σ y = bσ x 2 . Figure S5. and cov(x. rx. y) σx σy requires the calculation of var(x).

y) − b var(x) = = −1 σx σy b var(x) Thus.25 ri –.30) = .25) = 1/2(.002 .10 .E(r))2 = .16 0 .009 ∑ = . a – bx) = E[(x – E(x)) (a – bx – E(a – bx))] = E[x(a – bx) – xE(a – bx) – E(x) (a – bx) + E(x)E(a– bx)] = E[ax – bx2 – ax – x(–b) E(x) – E(x)a – (–b)xE(x) + aE(x) – b(E(x))2] = E[(–b)(x2 – 2xE(x) + (E(x))2] = –b E(x – E(x))2 = –b var(x) Substitution into the correlation equation yields rxy = cov(x.15 .30 –.20 . y) = cov(x.028 Range (r) = rmax – rmin = .016 0 .30 .045 –.20 – (–.000 .16)) = .20 piri –.014 .10 – (–.1 Prob.75 – r.10 .028 2 E(r) = Var(r) = ∑p r i i i = . (a) Table S5.50 Semi-interquartile range = 1/2(r.050 ∑ = .003 . . 3.020 . P0 P0 = 50 pi(ri – E(r)) .13 .009 i ∑ p (r .009 where r = P1 − P0 .Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 45 cov(x. x and y are perfectly negatively correlated.

009 Range = rmax – rmin = .14 .30 . Only n. y) = E[(x – E(x))(y – E(y))] Multiplying the factors in brackets on the right-hand side.004 –. we have cov(x.20 .015 –.38 – (–1.019 ∑ = . Fourth Edition (b) Table S5.10 . y) = E(xy) – E(x)E(y) – E(y)E(x) + E(x)E(y) cov(x.10 .23 –.000 — — — — ∑ = .023 E(r) = ∑piri = . recall the definition of covariance.009 pi (ri− − E(r))2 (semivariance) .01 .001 .016 0 .75 – r. • Financial Theory and Corporate Policy.05 .021 .16 0 . Semi-variance is given by E[(x i − − E(x)) 2 ] where xi − = xi if x i < E(x). First. because future possible prices are drawn from the same probability distribution as in part a).16)) = 0.05 ri –1.005 . E(xy) = cov(x.003 .46 Copeland/Shastri/Weston.004 . Semi-variance is important to risk-averse investors who are more concerned with downside risk (losses) than gains.010 .38 piri –.016 –. semi-variance = ∑ pi (ri− − E(r))2 = 0.05 .010 .15 .13 The expected return and semi-interquartile range are the same.020 .38 Semi-interquartile range = 1/2(r.010 –. 4. cov(x.02 . y) + E(x)E(y) .20 .20 –.00 –.0) = 1. . the number of observations. The range increases with n.2 Prob. has increased.25) = 1/2(.30 –. y) = E[(xy – xE(y) – yE(x) + E(x)E(y)] and taking the expectation of the right-hand side we have cov(x.07 .10 – (–. x i + = 0 if x i ≥ E(x) In the example above. y) = E(xy) – E(x)E(y) Therefore.023 .

Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 47 5.78 3 3 3 . the variance is 2 var = σ2 x − 2 cov(x. The variance of a perfectly hedged portfolio is 2 2 var = a 2 σ2 x + 2ab cov(x. Let σ2 x be the variance of the stock. 6. x) + b σ x The long position means 100% in X and the short position means –100% in X. the variance is V = X' Σ X where V = variance X = vector of weights X’ = transpose of the vector of weights Σ = the variance-covariance matrix (a) Therefore. Therefore. the variance of an equally weighted portfolio is  24 −10 25 1/ 3   V = [1/ 3 1/ 3 1/ 3]   −10 75 32  1/ 3   25 32 12   1/ 3  − 10 / 3 + 25/ 3  8  V = [1/ 3 1/ 3 1/ 3]  −10 / 3 + 25 + 32 / 3    25/ 3 + 32 / 3 + 4    13   V = [1/ 3 1/ 3 1/ 3]  321/ 3   23   V = 13 321/ 3 23 + + = 22. a=1 b = –1 Therefore. Using matrix notation. x) + σ x 2 2 σ2 x − 2σ x + σ x = 0 and the position is perfectly risk free.

8 2 pi[Xi – E(X)][Yi – E(Y)] –9.298 4.800 81.875 σp 6.0 var(X) = 28.6 1.0 9.8 .0 2.1 .475 19.2 var(Y) = 50.0 Yi – E(Y) –5 –8 10 7 –4 [Yi – E(Y)] 25 64 100 49 16 2 Xi – E(X) 9.0 pi[Yi – E(Y)] 5.0 12.4 cov(X.00 9.2 . (a) pi .0 –4.15   27.8 1.49 7.80) + .25   cov = [.4 .2 Xi 18 5 12 4 6 piXi 3.2 1.8 .0 3.0 –7.0 1.00 9.0 9.127 9.25   cov = [.0 2.0 7.0 –3.25    cov = [.4 .466 7.00 2 pi[Xi – E(X)] 16.00 25.875 50.2 E(X) = 9.2 3.8 3.48 Copeland/Shastri/Weston.292 4.0 2.00 16.1(26.2 E(Y) = 5.0 5.8 2 Yi 0 –3 15 12 1 piYi 0 –. Y) = –1.048 The opportunity set is shown in Figure S5.0 6.80    26.8 20.100 29.0 8.2 on the following page.8 5.2 . .905 5.675 28.25)] = –14.2 (b) Mean = aE(X) + (1 – a)E(Y) 2 2 Variance = a 2 σ2 x + 2ab cov(X.370 5.4 6.25) – .000 18.1]   −24.2 .0 –5.1 . Y) + (1 − a) σ y % in X 125 100 75 50 25 0 –25 % in Y –25 0 25 50 75 100 125 E(Rp) 10.1(27. Fourth Edition (b) The covariance is  24 −10 25 1.8(24.0 var(Rp) 47.0 [Xi – E(X)] 81. • Financial Theory and Corporate Policy.10    25 32 12    −.1]   −10 75 32   −.2 .0 6.6 3.0 4.

a* = = σ2 y − cov(x. The portfolio mean and variance at this point are: E(R p ) = . y) 2 σ2 x + σ y − 2 cov(x.8) = 11.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 49 (c) The minimum variance portfolio is given by equation 5.36)2 (50.64)(.584 = 17. y) = rxyσxσy and substituting cov(x. a* = σ2 y − rxy σ x σ y 2 σ2 x + σ y − 2rxy σ x σ y Recognizing the relationship between rxy and cov(x.183 .553 + 6. cov(x.56 var(R p ) = (.2) + (.64(9.0 + 50.2) 28.64) 2 (28.0) + 2(. y).8 − 2(−1.36(5.2) = .469 − .0) + . y) into equation 5.36) ( −1.21.8 − (−1.64 Figure S5. y) 50.2 The opportunity set Thus. the minimum variance portfolio has 64 percent invested in asset x and 36 percent in asset y.5 σp = 4.21.0) = 7.

36]     −1. It can be proven that the covariance between the global minimum variance portfolio and any portfolio along the efficient set is a constant.) (g) The variance of the global minimum variance portfolio is  28.64 .25 cov(Vmin .75  6.2 50.7  = [.8  The covariance between the minimum variance portfolio.0 −1.1  = [.8  . portfolio B is 25 percent in X and 75 percent in Y. B) = [. (a) Using the definition of expected value. The variance of the minimum variance portfolio equals the covariance between the minimum variance portfolio and any other portfolio along the efficient set. Vmin.75(6. and portfolio B is  28.36  17.75 . 8.36]   = 17.0 −1. and portfolio A is  28.75 .2  . (See Merton (1972).25(37.64 . A) = [.36]     −1.2 50.25 20.025 (e) The covariance between the minimum variance portfolio. B) = [.25]     −1.64 . Vmin.1) + . E(ax + by) = ∑ p (ax + by ) = ∑ p (ax ) + ∑ p (by ) = a ∑p x + b ∑p y i i i i i i i i i i i = aE(x) + bE(y) . The covariance between the two is  28.8 = .5 37. Fourth Edition (d) Portfolio A is 75 percent in X and 25 percent in Y.50 Copeland/Shastri/Weston.0 −1.36]   = 17.5 17. this is not a coincidence.8  .5 11.75 cov(Vmin .1  = [.64 .2 50.2  .64 .488  = [.75  6.0 −1.520  Again.8 (f) It is no coincidence that the covariance between the minimum variance portfolio and portfolios A and B is the same (except for rounding error).8) = 14.36]   = 17.36]     −1.64 .25 cov(A.2 50. • Financial Theory and Corporate Policy.2  .25]   37.64 var(Vmin ) = [.8  .2  .8  .

y) = rxyσxσy Therefore. cov(x. y) cov(x. if rxy = 1. then σ(x + y) = σx – σy: 2 var(x + y) = σ2 x + σ y − 2σ x σ y σ2 (x + y) = (σx − σy )2 σ(x + y) = σ x − σ y . var(ax + by) = ∑ p [(ax + by ) − E(ax + by )]2 = ∑ p [(ax + by ) − (aE(x) + bE(y))] = ∑ p [a(x + E(x)) + b(y − E(y))] = ∑ p [(a (x − E(x)) + 2ab(x − E(x)) (y i i i i i i i i 2 i i i 2 i 2 2 i i i − E(y)) + b (yi − E(y)) ] 2 2 = a 2 ∑ pi (x i − E(x)) 2 + 2ab ∑ pi (x i − E(x)) (yi − E(y)) + b2 ∑ p (y i i − E(y)) 2 = a 2 var(x) + 2ab cov(x. if rxy = –1. x) = E(x 2 ) − [E(x)]2 var(x) = E(x2 ) − [E(x)]2 E(x 2 ) = [E(x)]2 + var(x) (e) Prove that if rxy = 1. x) = E[(x − E(x))(x − E(x))] = E[x 2 − E(x)x − xE(x) + E(x)E(x)] = E(x 2 ) − E(x)E(x) − E(x)E(x) + E(x)E(x) cov(x. cov[(ax + bz). y) + b cov(z. 2 var(x + y) = σ2 (x + y) = σ2 x + σ y + 2σ x σ y σ2 (x + y) = (σx + σ y ) 2 σ(x + y) = σ x + σ y Prove that. y) + b 2 var(y) (c) Using the definition of covariance. y) (d) Using the definition of the covariance of x with itself. then σ(x + y) = σx + σy: var(x + y) = var(x) + var(y) + 2 cov(x. y] = ∑ p (ax + bz − E(ax + bz ))(y − E(y)) = ∑ p [a(x − E(x)) + b(z − E(z))](y − E(y)) = a ∑ p (x − E(x)) (y − E(y)) + b ∑ p (z − E(z)) (y − E(y)) i i i i i i i i i i i i i i i i = a cov (x.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 51 (b) Using the definition of variance.

063 1.67)(. • Financial Theory and Corporate Policy.3.50% σ(Rp) = 11.36 36.01)]1/ 2 = (. a* = σ2 y − rxy σ x σ y 2 σ2 x + σ y − 2rxy σ x σ y where rxy σ x σ y = cov(x.00 4.000 1. Fourth Edition 9.250 σ(Rp) 26.33(.01) .01)1/ 2 = 10% E(R p min ) = .02 + .31 11.93% 14.00 10.03) + .09 If we put two-thirds into asset 1.05) + 2(.33)2 (. line 4: E(Rp) = 5.3 150 100 75 50 25 0 –50 Opportunity Set Computations % in Asset 2 –50 0 25 50 75 100 150 E(Rp) 0.50% 3.18 16.05 − 2( −.05 − (− .01) .33)(−.3.65% The minimum variance portfolio is plotted in Figure S5. (a) Table S5. y) then a * = .06 = = .01 22.67(.21.67 .18% .25 5.50 Var(Rp) 7. the portfolio’s standard deviation is σp min = [(.14 10.52 Copeland/Shastri/Weston.67)2 (.250% 2.08) = 4.75 8.563 5. (c) From Table S5.000 13.02) + (.250 2.50 6.40 % in Asset 1 Figure S5.3 The opportunity set in mean-standard deviation space (b) Using equation 5.

30) + .1(.35)2 + .21.1(.3250 .30) + .1350 .1(.0225 (b) The table on the following page gives the mean and standard deviation of returns for various portfolios.35 Var(R1 ) = .1( −1.1125 .1(1.2025 − 2(.1(1.65 − .35) 2 + .0225) + .5) + .30) 2 + . we also need to know the mean and variance of the minimum variance portfolio.30) 2 + .0225) + .15 − .2025 E(R 2 ) = .35)2 = .1(.5 − .2025 + .30 Var(R 2 ) = .8(.65) = .5) = .8(−.8(.1(1.3500 .36 .30) = . Table S5.2025 − .8(.2025 Cov(R1 .1350 .1(.8225) + .367 .450 Because the mean-variance efficient set is the upper half of the opportunity set.8225) = . (a) E(R1 ) = .1(.450 .0) + .15) + .2025 σ(Rp) .35) (.0225) .1(. R 2 ) = .5 − .0225) + .5 .0225 .1(−1.8(.18 = = .30) 2 = .367 .5 − .1(−1.2025) + .1(.3000 σ2(Rp) .35) (.3375 .8(. we have a* = = σ2 y − rxy σ x σ y 2 σ2 x + σ y − 2rxy σ x σ y .1(.15 − .5 − .3125 .8(.8(.15 − .0 − .1(.2025) = .15) + . Using equation 5.335 .0225) + .35) (1.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 53 10.2025 .0225) = .65 − .0 − .4 % in 1 100 75 50 25 0 Computing The Opportunity Set % in 2 0 25 50 75 100 E(Rp) .15 − .

5 below shows the new efficient set as the solid line. i. can be written as Var(R p ) = ∑∑ w i w jσij i j Because the returns are independently distributed.5 Efficient set with no riskless borrowing 11. Asset 2 enters into efficient portfolios because of the diversification benefits it provides since it is not perfectly correlated with asset 1. all of the covariance terms are equal to zero.6). . Figure S5. then asset 1 dominates asset 2 because it has a higher mean and the same variance. 50 percent in 2) and ends with 100 percent in asset 1 (assuming no short sales). Fourth Edition The opportunity set is graphed below. a risk-averse investor will simply choose the set of weights.e. which minimizes variance. (e) Figure S5. wi. This means that the variance becomes Var(R p ) = ∑ w i2 σii i Because the assets also have identical means (as shown in Figure S5. • Financial Theory and Corporate Policy. σij = 0 for i ≠ j. Intuitively. (d) If mean and variance are the only relevant decision criteria (even though the returns are not jointly normally distributed). It starts with the minimum variance portfolio (50 percent in 1.. Portfolio variance. in general. Figure S5.54 Copeland/Shastri/Weston.4 The opportunity set (c) The efficient set is the solid line in the above figure.

we have Ψ = The first-order conditions are ∂Ψ = 2w i σii − λ = 0. the optimal weight is 1/N. 0. the marginal rate of substitution between risk and return is dE − ∫ U ' (E + σZ)Zf(Z.. the rate of return always equals its expected value. and since R = E + σZ . we see that the minimum variance is achieved if all of the weights are equal. the standard deviation of return is zero. i = 1. 1)dZ When an asset is risk-free. ∂w i ∂Ψ = ∂λ ∑w σ 2 i ii − λ ( ∑ w i − 1) .e. If there are N securities. Going back to Chapter 3 (Utility Theory). Proof is given by minimizing the variance subject to the constraint that the weights add to 100 percent. N ∑w i −1 = 0 The ratio of the partials with respect to wi is 2w i σii λ = 2w jσ jj λ Therefore wi =1 wj Thus.1)dZ = dσ ∫ U' (E + σZ)f(Z. Forming a Lagrangian. 0. 12. i. N equations.6 Optimal portfolio composition this implies putting an equal amount of wealth in each security. we have the result that when returns are normally distributed. ∑ wi = 1.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 55 Figure S5.

7 No risk-averse investor will hold 100% of his wealth in the risk-free asset 13. therefore the slope is b = E(X) − E(Y) 10 − 8 = = .15 The simplest way to solve the problem is to make use of the fact that the opportunity set for a portfolio of two perfectly correlated assets is a straight line. Therefore. Figure S5. this implies that the slope of the risk-averter’s indifference curve is always zero at the y-axis.0 Xi – E(X) 20 10 5 0 –60 [Xi – E(X)] 400 100 25 0 3600 2 pi[Xi – E(X)] 40 20 10 0 360 2 VAR(X) = 430 Using the probability properties.0 –5. the slope of the capital market line must be positive at the y-axis.74 σY = 17. The linear equation for the opportunity set is E(Rp) = a + bσ(RP) We know the coordinates of the points X and Y. If we find the expected return represented by its intercept then we can determine the correct weights for the zero variance portfolio. dE/dσ.0 6. the standard deviations of X and Y are σX = 430 = 20.1206 σX − σY 20.0 4.2 . pi .2 = 4. E(Y) = 6 + . we can immediately write the expected value and variance of asset Y.2 .74 − 4. No risk averter will hold all of his wealth in the risk-free asset. The opportunity set is graphed in Figure S5. is zero.8.1 1.04(430) = 17.1 .2) 2 VAR(X) = .2 Also.56 Copeland/Shastri/Weston.0 Xi 30 20 15 10 –50 piXi 3.0 E(X) = 10.4 . First compute the expected returns and standard deviations for asset X. As shown in the figure below.15 . Therefore.0 2. Fourth Edition we have the result that Z = 0. • Financial Theory and Corporate Policy. a tangency between a risk-averter’s indifference curve and the CML at the y-axis is impossible. As long as the ex-ante market rate of return is greater than the risk-free rate.2(10) = 8 VAR(Y) = (.2E(X) = 6 + . the marginal rate of substitution between expected return and risk.

1206(20. To confirm this result we can plug the appropriate weights into the definition of the variance of a portfolio of two assets as follows: VAR(R P ) = α 2 VAR(X) + 2α(1 − α)rXY σX σY + (1 − α)2 VAR(Y) = ( −.0.25)(1.1206σX 10.25 Therefore. wi = Vi .25)(1.74)(4.5%.79 + 26.0 = a + .74) a = 7.0)(20. Vi.2) = 26. (a) The variance of the shareholders’ equity position recognizes that the shareholders’ portfolio has positive weights (long positions) in assets and negative weights (short positions) in liabilities. Vi < 0 for liabilities S . therefore each weight is defined as the value of the ith asset or liability.25)2 430 + 2( −.88 − 53. we should sell short 25% of our wealth in asset X and go long 125% in asset Y.5 = α(10) + (1 − α)8 α = − .Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 57 Figure S5.8 Opportunity set for two perfectly correlated assets Using the coordinates for X and the slope we have E(X) = a + .5 The zero variance portfolio has an expected return of 7.88 ≈ 0 14. divided by the market value of equity. S. The return on equity can be written as RS = wSTARSTA + wUSRUS + wLRL – wSTLRSTL – wDRD We require that the sum of the weights add to 1. If we let α be the percent invested in X then the proportions of X and Y in this portfolio are E(R p ) = α E(X) + (1 − α)E(Y) 7.15) + (1.25)2 (17.

03) − − .0     −.TBσD PTB σTB Expanding this expression to fit the current problem.08) .02) 2  0 0 0 0   2 (.07) 0 (.33003 = .07) −850(.0 7.0     7.00040  0 .58 Copeland/Shastri/Weston.0 − .00036 .04) 700(.5]  .07) (. we have N=− N=− VUS rUS.3(.09(.3)(.0 2.8(.33003 The standard deviation is the square root of the variance σ= . 403 + 1.000420    0 0 .TB σL PTB σTB − 200(.0 wL 7.0 w US 2.00020    −.063 N = −3.5] 0 0 0 0  .03) (.03)    0 0 (.00040 0   0  .03)   w' Σ w = [1.33) is the result of taking the derivative of the shareholders’ portfolio variance (given that it has been augmented with futures contracts) with respect to the number of futures contracts.08) .0 2.2(.04) (.3(.8(.00490 0 .2(.0 −w D ] − 8.000 − 3. • Financial Theory and Corporate Policy.TB σi PTB σTB −VD rD.03)   0 Σ =  0 .00224 .01582    = [1.000900   0   1. N = −∑ i Vi ri.5  w' Σ w  .340 .04) .07) 2 0 .000360     0 .09(.08) N = −1.07) 0 .0   2.00160 . Fourth Edition The variance of the shareholders’ position is Var(R S ) = w' Σ w where: w' = [ w STA = [1.00224 0 .04) (.0 − w STL − 5.9)(.04) (.574482 (b) The position in T-bond futures contracts is taken in order to minimize the standard deviation of the shareholders’ claim.TBσ US PTB σTB − VL rL.5] (.02) 2 0  0  2  0  .09(.0 − .5 − 8.03) .04) (.00042 0 .5)(.00040   .00399   w' Σ w = .0 7.03521     −.07) (.5    −8.5 − 8. Equation (5.

0]    −.0 2.253605 The new standard deviation is .0 S 100 million and the variance of the revised shareholder position is w' Σ w = [1.00040  0   0   0  0    0 0 .000420 .00288   .0 − .0 − .000360 .00004   w' Σ w = .00490 0 .00280   0 .00036 .5 − 3.5 − 8. The new shareholders’ position is augmented with futures as follows.00040 0 0  .000720 .00040   .5 − 8.0] .0     .340 futures contracts. .00640    1.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 59 The conclusion is that the variance of shareholders’ return can be minimized by short selling 3. RS = wSTARSTA + wUSRUS + wLRL – wSTLRSTL – wDRD + wTBRTB The market value weight of the futures position is wF = NPTB −3340(.5 − 3.0 7.0     7.000900 . and this will offset the declines in the value of the bank’s loan portfolio and government bond holdings.00718     .00042 0 .00615      .09 million) = = −3.503592.00224 0 .00288 0 0 0 0  .0     −.00020   −.5     −3.00072   .00224 0 .0 2. The T-bond futures position will rise in value when interest rates fall.00280 0 .00160 .5   −8.0   2.0 7.02681  w' Σ w = [1.

Rm) = .00225 .R m ) σ 2 m E(R m ) = .01 (a) E(Rm) = 10% (b) σ2 σm = 10% m = 1%.2 –. The conditions are those necessary for two-fund separation. namely (in addition to those already assumed) • • • • 2.015 .00675 .1 .0525 Rm pi R m Rm − Rm –.30 .01125 .3(–.10) = .00 .02150 βj = cov(R j .Chapter 6 Market Equilibrium: CAPM and APT 1.0100 pi (R j − R j )(R m − R m ) –.35 .05 .20 .35)(.00100 .10 pi (R m − R m )2 .0215 = 2.015 .50 E(R* j ) = . (c) E (R* j ) = 15% (d) cov(Rj.05)(.00625 .05) = .45)(–.03 .00100 .05) = .15 .15 .15 .45 –.02025 .060 .05 .3 .00200 σ2 m = .00 .4 .00700 cov (Rj.06]β j .20 –.06 + [.4(.15)(–.10 = .040 .15 pi R j –.25 –.15% (e) E(R j ) = R f + [E(R m ) − R f ]β j E(Rj) = .08 .15 .00100 .02450 σ2 j = .2(.10 – . Rm) = 2.00075 .25) = . pi homogeneous expectations asset markets frictionless with costless information simultaneously available to all individuals all assets marketable and perfectly divisible no market imperfections such as taxes or restrictions on selling short Rj –.1(–.05 .05 .10 Rj − Rj pi (R j − R j )2 .

10 cov(R j .05 –.021 5 σ m = 13.Chapter 6 Market Equilibrium: CAPM and APT 61 (f) The required rate of return on Donovan Company is E(R j ) = .0225 0 .25 .15) = 14.06 + .6 ∑ (R j − R j )(Rm − Rm ) = .0225 1978 1977 1976 1975 1974 1973 . E(X) = ∑X N i var(X) = ∑ (X i − E(X))2 N −1 Note that the sample variance is calculated by dividing the sum of mean deviations squared by N – 1. We subtract one because we lose one degree of freedom in estimating the mean.06 + [.27 ∑ R m = .0225 .0300 .R m ) σ2 m .20 . Because Donovan earns more than is required.6 Rj − Rj (R j − R j )(R m − R m ) Rm − Rm .20 .12 –.018 5 ∑R j .15 . cov(X.0225 0 .15 .1% . we can use statistics formulae for sample mean and variance.72 .15 0 –.6%.105 = .01 = .0300 ∑ = .04] cov(R j .30 ∑ R j = .15 –. R m ) = N −1 6 = (a) E(Rm) = 12% (b) var(Rm) = 1. we might expect its price to increase so that in equilibrium its expected rate of return equals 14. Rm) = 2.05 –.8%.72 = .27 .105 ∑(R m − R)2 = .10 − . 3.09 = .10 .03 . Now we are dealing with sample data.05 .0215 .15 (R m − R m )2 .12 –. therefore.42% (c) E(Rj) = 10% (d) cov(Rj. E(R* j ) = 15% (see c above).0225 0 . the number of observations minus one.15 –.0225 0 . Covariance calculations also are divided by N – 1.06] = .12 6 var(R m ) = ∑ (R m − R m )2 N −1 = .03 .15 0 –.06 + [.09 E(R m ) = E(R j ) = ∑R N N m = = .6% The expected return.04(2. .Y) = Year Rm ∑ (X i − E(X))(Yi − E(Y)) N −1 Rj .05 –.

1034 5. See table S6. Other than the usual perfect market assumptions. Panel A..8318 . then the market . 8) = .1. the key assumptions are that 1) investors have homogeneous expectations and 2) all assets are perfectly divisible and marketable.Rm ) var(R m ) (f) Required return (Rj) = .1. expected-end-of-period-utility-of-wealth maximizers. 7) = .0801 = .9% The market requires 12. we would expect the Milliken company to decline in value.0107 10 − 1 9 (c) σ(R m ) = . R m ) var(R m ) = .0107 = . ∑ (col. 5) = . Given the two above assumptions.e. ∑ (col. See table S6.62 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.054] (1. and all assets are held. 5) = 1. If every individual holds an efficient portfolio. 4. Fourth Edition (e) E(Rj) = Rf + [E(R m ) − R f ] cov(R j .066 + [.0089 = .7796 ≅.1022 (a) R j = 10 10 (b) var(R j ) = ∑ (col.9% and the expected (i. Therefore.0964 = .078 (a) R m = 10 10 (b) var(R m ) = ∑ (col. It is not necessary that there be risk-free assets. and given that we have rational.0294 10 − 1 9 (c) cov(R j .2644 = . risk-averse. 7) = .0089 10 − 1 9 cov(R j . Panel B.1667) = 12.0107 6.0221 = .R m ) = (d) β j = ∑ (col. all individuals will perceive the same efficient set and all assets will be held in equilibrium. anticipated) rate of return is only 10%.

0054 .0534 .2662 .0587 .1384 –.0020 .04 .17 85.Chapter 6 Market Equilibrium: CAPM and APT 63 Table S 6.1947 .0208 .38 69.0307 .1030 .0269 .06 .0070 .0864 –.0059 .1471 –.0770 .0201 .0397 .0137 –.05 .05 .05 .1043 .1376 .84 83.05 .0300 .0320 .0043 –.0736 –.1993 –.70 97.26 91.22 (8) .0004 .0113 .0317 .0971 –.0004 .0738 .1556 –.0356 –.87 81.3709 .03 .1212 .0250 .0007 .05 .0340 .0264 –.0801 (1) 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 (2) 48 49 52 74 90 102 87 78 81 74 70 (5) .0298 .04 .0782 .0002 .0242 .1102 .0612 .0136 .0221 (6) –.0337 .06 .0383 (5) .0050 –.2644 (6 × 6 [Panel A]) (8) –.05 .4731 .85 66.4231 .1494 (4) .0358 .0541 Calculation of Beta for General Motors Percent Return Rjt (3 + 4) Deviation of Returns (R jt − R j ) (5 − R j ) Dividend Yield Dt Pt (4) .0324 .03 .93 98.0192 .05 .0964 R f = .1333 –.44 /10 = .0385 –.0041 1.37 88.1891 Market Variance (Rmt − Rm )2 (62) (7) Bish-free rate Rf (1) 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 (2) 55.0010 .0087 –.1201 .0885 –.1646 .0010 .0708 .1063 .07 .1136 .0013 .0010 .0237 –.1 Year (t) S&P 500 Price Index Pt Panel A: Estimates of Market Parameters % Change Dividend Percent Return in Price Yield Return Deviation Rmt Dt Pt (Rmt − Rm ) −1 Pt Pt −1 (3 + 4) (5 − Rm ) (3) .03 .2162 .0314 .06 .0330 .0066 .1518 .27 62.0263 –.1833 –.0191 .0029 .05 .0106 .044 Panel B: GM Year Price (t) Pt % Change in Price Pt −1 Pt −1 (3) .0811 –.05 .0153 –.0165 .1286 –.1034 .0543 –.1866 –.0190 .0301 .0274 .0836 –.0029 .1640 .05 Variance of Returns (R jt − R j )2 Covariance with Market (R jt − R j )(R mt − R m ) (62) (7) .04 .7796 (6) .1167 .2164 –.0322 .1111 .

0 βp = a The greater the percentage. σp = βp σm . With risk-free borrowing and lending. a. β is the measure of systematic risk. Algebraically. . the less risk averse the investor is.64 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. the variance in any efficient portfolio would be 2 2 var[a(R m ) + (1 − a)R f ] = a 2 σ2 m + (1 − a) σ R f + 2(a)(1 − a)σ m. and if you can observe any one of the following: 1) the percent of the investor’s portfolio held in the market portfolio.e. Because the risk-free asset has no variance. Thus. 2) the β of the investor’s portfolio. Finally. a = β p. i. as proven in problem 6.7. all efficient portfolios are linear combinations of the risk-free asset and the market portfolio. βp = aβm and since β m = 1. total variance in every efficient portfolio is contributed solely by the market portfolio. you can tell how risk averse he is. Systematic risk is defined as market risk. is equal to the portfolio’s β p. βp = βf (1 − a) + βm a where β f = 0 is the beta of the risk-free asset. of the market portfolio held in the investor’s total portfolio. These three measures are equivalent. all risk is systematic. Given a riskless asset. This result does not change in a world without a risk-free asset because the zero-β portfolio may be substituted for the risk-free asset without changing any results. a. Therefore. two-fund separation obtains. To see this. a portfolio’s variance that can be explained in terms of market variance. The percentage. 8. Therefore. Thus.2. note that E(Rp) = Rf + [E(Rm) – Rf] β p E(R p ) = R f (1 − βp ) + R mβp Also E(R p ) = R f (1 − a) + R m a where a is the percent invested in the market portfolio. The capital market line is made up of a linear efficient set as in Figure 6.R f = a 2 σ2 m +0+0 where a represents the percentage of the efficient portfolio invested in the market portfolio. 7. Yes.. Fourth Edition portfolio must also be efficient because it is merely the sum of all efficient portfolios held by all individuals. or 3) the expected return on the investor’s portfolio.

Chapter 6 Market Equilibrium: CAPM and APT 65 9. (a) We know by the CAPM: R j = R f + [E(R m ) − R f ]β j .05 Total risk is defined by equation 6.05 = 2. σ j = β jsm = 1.0)2 (. equation 17) is rjm = To find cov (Rj.5(. Rm). σ2 ε 2 2 2 σ2 k = β k σ m + σε 2 .06 Since rjm = 1.Rm ) σ jσ m cov(R j . we again have the result that efficient portfolios are perfectly correlated with the market (and with each other).12) is 2 2 2 σ2 j = β j σ m + σε and since the unsystematic risk of an efficient portfolio. β k is its systematic risk.25 − . Therefore.15 − .2) = . we can find the β k for Rapid Rolling.52 = (2.36 or σε = . i.05 = β j = 1.0 . that all portfolios lie on the security market line. Assuming equilibrium. is zero.8β j β j = .10 = .5 .06)β j . all assets must be priced so that they fall along the security market line.2) = cov(Rj.e.5 The CAPM assumes that the market is in equilibrium and that investors hold efficient portfolios. σ2 ε .14 − .6 11.52 − 4(. σ2 j = . The total risk (equation 6. E(R j ) − R f + [E(R m ) − R f ]β j . .04 = .2) = .5 (.12. We can use it to find Rapid Rolling’s unsystematic risk. use the defination of β j bj = 2 cov(R j .2)2 + σε 2 σ2 ε = .0.3 or 30%.R m ) σ2 m 1.09 or 9% The definition of correlation (from Chapter 5.15 − .. E ( R k ) = R f + [E(R m ) − R f ]βk βk = . 10. Rm) = .05 We know that efficient portfolios have no unsystematic risk.06 + (.2 − .

Also. is E(R j ) = E(P1 ) − P0 P0 E(P1 ) − P0 = R f + [E(R m ) − R f ]β j P0 where E(P1) = $100.2P0 1. the one period rate of return.11) The security market line 12.66 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.18 − . Figure S6. we have $100 − P0 = .14 = . (a) A zero-beta portfolio has zero covariance with the market portfolio. we have E(R j ) = Substituting in the appropriate numbers and solving for P0. Fourth Edition (b) Let “a” be the percent invested in the risk-free asset.5 a= −.125 13. Portfolio return is the point on the security market line where E(R p ) = aR f − (1 − a)E(R m ) 10% = a(6%) + (1 − a)(14%) . by definition.10 − . Rj. Assuming that the company pays no dividends. (1 – a) = 50%.08P0 + .08]2.e. Using the CAPM. i.08 + [.1 (Problem 6.. the individual should put 50% of his portfolio into the market portfolio.28P0 = $100 P0 = $78.08 Therefore. using matrix notation. the covariance between two portfolios is cov = W1'Σ W2 .0 P0 $100 − P0 = .

Chapter 6

Market Equilibrium: CAPM and APT

67

where W1' = the row vector of weights in the zero -β portfolio Σ = the variance-covariance matrix of two risky assets W2 = the column vector of weights in the index portfolio By setting cov = 0, we can solve for W1' . cov = [X1 0  .5 .01 X2 ]    = 0  0 .0064  .5 0  .005 + =0 X2 ]  + .0032   0  .005X1 + .0032X2 = 0 In order to have two equations and two unknowns, we also use the fact that the weights must always sum to one, i.e., X1 + X2 = 1 Solving this system of equations we have .005X1 + .0032X 2 = 0 .005X1 + .0050X 2 = .005 −.0018X2 = −.005 X2 = .0050 = 2.78 .0018

[X1

This implies putting 278 percent of the portfolio wealth into asset 2 and –178 percent into asset 1. Therefore, the expected return on the zero-β portfolio is E(R p ) = −1.78E(R1 ) + 2.78E(R 2 ) = −1.78(.2) + 2.78(.1) = −.078 (b) The vector of weights in the minimum variance portfolio can be found by using equation 5.21. a =
*

σ2 y − rxy σ x σ y
2 σ2 x + σ y − 2rxy σ x σ y

From the variance-covariance matrix, Σ , we know that σ2 x = .01
σ2 y = .0064

cov(X, Y) = 0

Therefore, rxy = 0, and a* = .0064 − 0 = .390 .01 + .0064 − 0

Thus, the minimum variance portfolio consists of 39 percent in asset 1 and 61 percent in asset 2.

68

Copeland/Shastri/Weston • Financial Theory and Corporate Policy, Fourth Edition

(c) Using the covariance definition again, we have cov = W1' Σ W2 0   −1.78 .01 = [.39 .61]     0 .0064   2.78   −.0178 = [.39 .61]    .0178  = –.006942 + .010858 = .003916 (d) The equation of the market line is E(R j ) = E(R z ) + [E(R I ) − E(R z )]β jI where th E(Rj) = the expected return on the j asset E(Rz) = the expected return on the zero– β portfolio E(RI) = the expected return on the efficient index portfolio β jI = the covariance between the returns on the jth asset and the index portfolio, standardized by the variance of the index portfolio Substituting in the estimated parameters, we have E(R j ) = −.078 + [.15 − ( −.078)]β jI E(R j ) = −.078 + .228β jI 14. Using matrix notation, the definition of covariance is cov( X, Y ) = Wx' Σ Wy where Wx' = a row vector of weights in portfolio A Wy = a column vector of weights in portfolio B Σ = the variance-covariance matrix of A and B Substituting in the appropriate numbers, we have  .01 cov(X, Y) = [.1 .9]   −.02  .006 = [.1 .9]   −.012  −.002  = [.1 .9]    .004  = –.0002 + .0036 = .0034 15. Using the definition of the correlation coefficient, we have
rk,m = cov(k, m) σ k σm

−.02  .6    .04   .4  − .008  + .016  

Chapter 6

Market Equilibrium: CAPM and APT

69

Substituting the correct numbers and solving for cov(k, m),
.8 = cov( k, m) (.25)(.2)

cov(k, m) = .8(.25)(.2) = .04 Using the definition of β we can calculate the systematic risk of Knowlode
βk = cov(k,m) .04 = = 1.0 σ2 (.2)2 m

The systematic risk of a portfolio is a weighted average of the asset’s β ’s. If “a” is the percent of Knowlode, β p = (1 – a) β f + a β k 1.6 = (1 – a) (0) + a(1.0) a = 1.6 or 160% In this case the investor would borrow an amount equal to 60 percent of his wealth and invest 160 percent of his wealth in Knowlode in order to obtain a portfolio with a β of 1.6. (This analysis assumes investors may borrow at the risk-free rate.) 16. First, we need to know E(Rm) and Rf. Since each of the assets must fall on the security market line in equilibrium, we can use the CAPM and E(Ri). Asset 1: E(Ri) = Rf + [E(Rm) – Rf] β i 7.6 = Rf + [E(Rm) – Rf].2 Asset 2: 12.4 = Rf + [E(Rm) – Rf] .8 This gives two equations with two unknowns. Multiplying the first equation by 4 and subtracting the second equation, we have 4(7.6) = 4Rf + [E(Rm) – Rf] .8 −[12.4 = R f + [ E(R m ) − R f ].8] 18.0 = 3R f R f = 6 (i.e.,6%) Substituting the value of Rf into equation 1, we have 7.6 = 6 + [E(Rm) – 6] .2 7.6 − 6 + 6 = E(R m ) .2 E(Rm) = 14 (i.e., 14%) To double check, if the values of Rf and E(Rm) are used in order to obtain E(Rj) for assets 3 and 4, they should also fall on the security market line. For asset 3, we have E(R3) = 6 + [14 – 6]1.2 = 15.6

70

Copeland/Shastri/Weston • Financial Theory and Corporate Policy, Fourth Edition

and for asset 4 we have E(R4) = 6 + [14 – 6] 1.6 = 18.8 The expected rate of return and β for the current portfolio are E(R p ) = ∑ Wi E(R i )
i −1 5

= .1(7.6) + .1(12.4) + .1(15.6) + .2(18.8) + .5(6) = 10.32 βp = ∑ Wiβi
i =1 5

= .1(.2) + .1(.8) + .1(1.2) + .2(1.6) + .5(0) = .54 In order to achieve a new portfolio with a 12 percent expected rate of return, we subtract X percent from our holdings in the risk-free asset and add X percent to new holdings in the market portfolio. 12 = E(Rp) = .1(7.6) + .1(12.4) + .1(15.6) + .2(18.8) + (.5 – X)(6) + X(14) 12 – 7.32 = 3 + 8X 1.68 = X = .21 8 Therefore, the new portfolio holdings will be Return on Asset i 1 2 3 4 Rf Rm 7.6% 12.4% 15.6% 18.8% 6.0% 14.0% βi .2 .8 1.2 1.6 0 1.0 Percent in Asset i 10% 10% 10% 20% 29% 21%

The expected return of the new portfolio is 12 percent and its β is .75. If you hold only the risk-free asset and the market portfolio, the set of weights that would yield a 12 percent expected return can be determined as follows: W1 = amount in risk-free asset W2 = amount in market portfolio .06W1 + .14W2 = .12 W1 + W2 = 1

13 percent. P1.08 β j = . Multiplying the second equation by .14W2 = .75 If the rate of return covariance with the market portfolio doubles.06 and subtracting from the first gives .20 If the expected future price is still $45.20/1.07 + (.13) = $45.98 .75 = 1 W1 = .19 –1 (σ jm )0 σ2 m σ2 m where (σ jm )0 = original cov(E(Rj). was P0(1 + r) = P1 40(1.06W1 + .07 = βj .06) .5 = . E(Rm)) 2(σ jm ) 0 = $37. so that the required return is 19 percent.08W2 = .06W1 + .Chapter 6 Market Equilibrium: CAPM and APT 71 We have two equations with two unknowns.13 = .20 but the covariance with the market has doubled.06W2 = . E(Rm)) where 2(σ jm )0 = subsequent cov(E(Rj). the expected future price.08) β j .06 W2 = .75 W1 + .13 − . and given a present price of P0 = $40. We know from the CAPM E(Rj) = Rf + [E(Rm) – Rf] β j Substituting the correct numbers for β j . then the present price will be P0 = P1(1 + r) = 45.19 = 19% At the original rate of return.08)1.12 −(.07 + (. 17. β j will also double since β j0 = β j1 = Therefore β j1 = 2 β j0 so the new required return for the security will be Required Rj = .25 A portfolio with 25 percent in the risk-free asset and 75 percent in the market portfolio has an expected return of 12 percent.

First.01 is E(R j ) = R f + [E(R m ) − R f ]β j = .05 = .08).05 − .2 The ex post security market line .08 + [R m − . the expected rate of return and the estimated risk could have been overestimated.2. P* is a 1.08 + (. To find the percentage change in the security’s price that would result in a return of 10 percent * instead of 12 percent.01 = .1 Thus. Fourth Edition 18. reverse the investment decision.04 The ex post security market line is depicted in Figure S6. (b) There are two possibilities for the low ex post return.01818 P0 1.82 percent.16 − .12 = = 1.25 (. The second possibility is that after the fact.25 R m = [.10) P* 1. investment in the security would be advisable.08 = −. solve for P in terms of P0: P0 (1.72 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. the market unexpectedly fell. E(Rm)) = . (a) The required rate of return on a security with cov(E(Rj).12) = P*(1.08]4 + .08].82 percent increase in price over P0.10 where βj = σ jm σ 2 m = .25 E(R j ) = . If the security’s price increases more than 1. The ex post market rate of return which would have resulted in a 5 percent rate of return for the security in question is R j = R f + [R m − R f ]β j . Figure S6.2)2 Since the expected rate of return is greater than the required rate of return.

R J ) = [1 0]    . and 3) the portfolio of risky nonmarketable assets.0025  0  . 20.5 = . This may be interpreted as a form of three-fund separation.0081 COV (R I .R H )] This implies that the individual will take into consideration not only the covariance of an asset with the market portfolio.00405 = [.0081 COV(R A . Every investor will hold one of three funds: 1) the riskfree asset.75(.00125 = . (a) Given that some assets are nonmarketable.5 .R m ) + cov(R j .00405 0  . 2) the market portfolio.5]   .5]    .R I ) .0025  0  .0045563 + .0025  0  . and 3) a hedge portfolio chosen to hedge against unforeseen changes in the future risk-free rate.28 E(R j ) = rf + γ1[E(R m ) − rf ] + γ 2 [E(R N ) − rf ] Once again.0081 COV(R J .26. three-fund separation obtains.002025 + . (b) If the risk-free rate is nonstochastic.R I ) = [1 0]    . then we can use Merton’s (1973) continuous time model as shown in equation 6.000625 = .75 .75 .5 .25 = .5 .25]    . where the three funds are 1) the risk-free asset.0047126 Therefore.25]   . We have separation because the market price of risk (the marginal rate of substitution between risk and return) is independent of individuals’ utility functions.002650 .289 COV (R I . E(R j ) = R f + λ[Vm cov(R j . investor I computes a β of βI = COV (R A .75 .5(.5 .0001563 = .25 .0081) = .0047126 Repeating the exercise for investor J we have 0  . the equilibrium pricing equation is given by Mayers (1972) in equation 6.R I ) = [.006075 The variance of investor I’s index is 0  .R I ) .0025  0  . Separation is still valid in the sense that the marginal rate of substitution between risk and return is independent of individuals’ utility preferences. but also its covariance with the portfolio of nonmarketable assets.006075 = [.000625 = .006075 = = 1.5 . (a) The covariance between investor I’s index and asset A is 0  .0081) = . 2) the portfolio of risky marketable assets.75 .Chapter 6 Market Equilibrium: CAPM and APT 73 19.0081 COV (R A .R J ) = [.

R I ) = Wz' Σ WI = 0 The appropriate numbers for investor I are: COV(R z .0081)Wz1 + . as shown below .85% Therefore. we need be concerned only with the zero covariance condition. The zero-beta portfolio is the minimum variance portfolio which has zero covariance with the index portfolio.25 Solving.1885]βI .528 COV(R J .00405 = = 1.1147(30%) + 1. Fourth Edition And investor J estimates a β of bJ = COV(R A . In order to numerically demonstrate this result the zero-beta portfolios and security market lines for each investor are computed below.1147(20%) = 18. they will perceive the same expected rate of return on any portfolio. but both portfolios are efficient.R I ) = [Wz1 0  .75(. Wz2 = 1.2750 − . If we define Rz as the return on the zero-beta portfolio and Wz as its vector of weights. Since all portfolios lie on the minimum variance opportunity set in this two-asset world.75 . then we require that COV(R z . we have Wz1 = −. Any combination of them will lie on the mean variance opportunity set. Consequently.006075Wz1 + . The expected return on asset A is 30%.002650 (b) Both investors will require the same rate of return on asset A. (c) The key to understanding this problem is Roll’s critique.R J ) .0081 Wz2 ]   =0 .1147. even though the investors perceive different security market lines and different betas.0025)Wz2 = 0 However. R J ) .0025  0  . SML for I . The two investors choose different index portfolios. They must.1147 Thus.000625Wz2 = 0 Wz1 + Wz2 = 1 Solving. we have one equation and two unknowns . the security market line perceived by investor I can be written as follows: E(R j ) = E(R z ) + [E(R I ) − E(R z )]βI = . They have homogenous expectations therefore they perceive the same risk-return combination.74 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. for investor I the expected return on his zero-beta portfolio is E(R z ) = Wz1E(R A ) + Wz2 (R B ) = −.1885 + [.25(. we also know that the weights must add to one Wz1 + Wz2 = 1 This gives two equations and two unknowns. They must be because this (over-simplified) world has only two risky assets.

528 = 30% This numerical example demonstrates that although investors I and J estimate different betas.4464. we can now verify that investor I expects a 30% rate of return on asset A.R j ) = Wz' Σ WJ = 0 COV(R z .2% for her portfolio and using the APT equation.50(. Finally. we repeat the same procedure for investor J. Wz2 = 1. E(R A ) = .05)(−.1885 + [. the expected return on her portfolio is E(R i ) = R f + [E(R M ) − R f ]βi = .2750 − .08 + [.50(.0081 Wz2 ]    . (a) Using the CAPM.1554 + [. they will have the same expected return for asset A. The only difference is that investor J uses a different index portfolio.289 = 30% Next.0081)Wz1 + .4464(20%) = 15.54% Investor J’s security market line equation is E(R j ) = .142 = .5) = βi2 .289 for asset A.0025)Wz2 = 0 Wz1 + Wz2 = 1 and their solution is Wz1 = −.05](−.05]βi1 + [.2500 − .08 − (.1554 + [.11]βi2 .08 + [.5) + [.08 + [.1554]β j .1885]1.2500 − .50  The expected rate of return on investor J’s zero-beta portfolio is E(R z ) = −.7909 = βi2 .0025  0  . we have E(R i ) = .Chapter 6 Market Equilibrium: CAPM and APT 75 Recalling that investor I estimated a β of 1. recalling that investor J estimated a β of 1.4464(30%) + 1.11 . SML for J.R J ) = [Wz1 The two simultaneous equations are .50  .11]βi2 .142 = .4464 0  .528 for asset A.0 = .062]1. Investor J’s zero-beta portfolio will be COV(R z . we see that his expected return on asset A is E(R A ) = .142 Using the fact that E(Ri) = 14. 21.1554]1.142 − .

Fourth Edition (b) Writing the APT equations once more.08 + .08 = βi1 . we have .142 − .76 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.11)β i2 Substituting in the value of bi2 as zero.142 = .05βi1 .24 = β i1 . we have E(R i ) = .142 = .08 + (.05)β i1 + (.05 1.

1566 = .5 d2 = –.36 pricing European calls. we have d1 = d1 = ln(28/40) + .40335 – ( .1829) C = 9.3434 N(d2) = .40335 .Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 1.5 – .25 = –.5) .5( .06(. C = SN(d1 ) − Xe− rf T N(d 2 ) where d1 = ln(S/ X) + rf T + (1/ 2 )σ T σ T d2 = d1 – σ T Substituting the correct values into d1.03 + .5 .5) (.5) + .1829) Solving for P0.5)( . We know from put-call parity that the value of a European put can be determined using the value of a European call with the same parameters.1829 Substituting these values back into the Black-Scholes formula.52 2. We can use the Black-Scholes formula (equation 7. and converting the formula to continuous rather than discrete compounding.5 −.5) = –.3434) – 40e = $2.3171 = . according to equation 7.8 C0 – P0 = (1 + rf )S0 − X (1 + rf ) –. we have C = 28(.5 – .356675 + .06(.970446)(. we have N(d1) = .6152 – 40(.5)( .90335 Using the Table for Normal Areas. we have P0 = C0 − S0 + X − rf T e . the present value of a put.

we have d1 = ln(20 / 20) + (.08) (.4243 = –.12172 = .96 .5) + (1/ 2)(.5 + .5)(20)(.6)( .6)( .216 = $2. P = 3.744 – 20 + (20)e–(.08)(.6 .45224 Substituting these values into the Black-Scholes formula yields C = 20(.45224) = 12.256 + 19.5) (.62172) – e–(.36) C = SN(d1) – e − rf T XN(d2) where: d1 = Solving for d1.5) ln(S/X) + rf T + (1/ 2)σ T σ T = .5 – . Fourth Edition First.1179 From the Table of Normal Areas. according to the Black-Scholes formula (equation 7.3064 – .5 = .744 Solving for the value of a put.04776 = .5) = –16.08)(.434 – 8.78 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.3064 and d2 is equal to d2 = d1 – . calculate the value of the corresponding call option.690 = 3.62172 N(d2) = . we have N(d1) = .

P1 = C1 – S + Xe = P0 + a If you sold one call and one put for prices of C0 and P0. Given the inside information. Figure S7. (a) Figure S7. –rt = C0 + a – S + Xe–rt Figure S7. selling one put (–P).1a shows the payoffs from selling one call (–C).1b shows the payoffs from buying one call (+C). and from the combination (–C–P). Suppose the value of the call increases by increases. the value of the put also increases by a.Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 79 3. If the instantaneous variance of the stock ∂C > 0. the portfolio strategy would be to buy both the put and the call (at P0 and C0) for a gain of 2a. C1 = C0 + a Then by put-call parity. selling one put (–P) and from the combination (C – P). (b) Figure S7. the value of the call increases.1b Payoffs from buying one call and selling one put . and the options’ true values were C0 + a and P0 + a. since ∂σ some amount a > 0. It earns a positive rate of return if the stock price does not change much from its original value. this represents an opportunity loss to you of –2a.1a Payoffs from selling one call and one put The portfolio (–C–P) is the opposite of a straddle.

18 – ( .07) (156/365) N(d2). X1) b) P(S.07)(.07)(156 / 365) + (1/ 2)( .31)(.6538) (. ln(44. The value of the call option can be calculated directly by the Black-Scholes formula for a European call. Using the correct values from the problem. C = 44.6538) = . C = 44.5714) – 45(. we can determine N(. Compare the payoffs at maturity of two portfolios.0961)( 156 / 365) −. and the second is a European put option written on the same stock.4847 = $3.3559 – 21.4920) = 25.375(. The first is a European put option with exercise price X1.375N(. with the same time to maturity. since.0961) (. Substituting these values into the formula yields C = 44. Because portfolio A has a value either greater than or equal to the value of portfolio B in every possible state of nature.03 Using the Table of Normal Areas. C0 + a – (P0 + a) = C0 + P0. Therefore. T.87 5. X2) Comparative Value of A and B S < X2 X1 – S X2 – S VA > VB if X2 < X1.31)(. 4.1 Portfolio a) P(S. Fourth Edition The return from this portfolio (C – P) remains unchanged by an increase in the instantaneous variance. T. but with exercise price X2 < X1.02).1. the strategy of this portfolio is neither advantageous nor disadvantageous.4274) + (1/ 2)(. the put with a higher exercise price is more valuable. the increase in the value of the long position in the call option is exactly offset by an increase in the value of the shorted put option.6538) = –. by put-call parity.375 / 45) + (. In the algebra of part (a) above. X1) > P(S. T. T. The payoffs are given in Table S7.18 d2 = .375N(d1) – 45e First calculate d1 and d2: d1 = = – (.18) – 45e N(–.80 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.9704)(.0961)( 156 / 365) ( .32. P(S. equation 7. X2) Table S7.02) –.02 Substituting these values into the Black-Scholes Formula.18) and N(–. X1 ≤ S 0 0 VA = VB State Contingent Payoffs of Put Portfolios X2 ≤ S < X1 X1 – S 0 VA > VB .01398 + (. given inside information of an increase in the instantaneous variance.

6065)(.000.05.399 = $892.795) = 5.42507) – 4.000(.5 + .350 – 517.5 )( 10 ) Note that T.Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 81 6.107.2231 + 1) 2.236 + 1. the market value of the debt.2231) + (.000. is V=S+B We can use the OPM to determine the market value of the stock (expressed as a call option on the firm).000e = 4.000.5) + 1.107.000.44) – 4. ten years. is the time until the firm’s debt matures.5)( 10) ( .118 2. $5MM.236 = –. d1 = (.000 (.5)( 10) –(.05)(10) N(–.601.399 The value of the debt is $5.5)( 10) = 1.5 – .44 – 2. we have S = VN(d1) – D e − rf T N(d2) where S = market value of the stock V = market value of the firm D = face value of the debt d1 = ln(5MM / 4MM) + (.118 = 1.236 = 1.05)(10) + (1/ 2 ( .625.000.951 = $4. we have d1 = = ln(5MM / 4MM) + (.1)(10) + (1/ 2) ( . For Rf = .665 d2 = d1 – ( .5 )( 10 ) (.2865) ( .795 S = 5.000 – 4. and subtract this from $5MM to find B.665 – 2. For rf = .236 = –.10.571 .000N(1.44 d 2 = d1 – σ T = 1. The total value of the firm. the time to maturity.

rf = 6%.025 . (S = 1).0 .156 .10 .153 .2 S/X = .1)(10) N(–.050 .061 .3 C/S as a Function of rf Assume: S = 1.041 .890 Therefore.18 .2 C/S as a Function of T Assume: S = 1.000.131 .16 .9 .000.123 .1 . T = 6 months. The three tables below show the call price as a percentage of the stock price.06 .2.060 .085 .188 .039 .028 .098 .131 .110 . 7.110 The value of the debt decreases to –(.118 .142 S/X = 1.121 .760.016 .1 .118 .000e = 4. and the standard deviation (in Table S7. In order to standardize the results we assume that the stock price is equal to 1.571) = 5.107 S/X = 1.0 . This is what we would expect.110 = $657.4).000.101 .02 .000(.072 .04 . the rise in rf benefits the stockholders. σ = .045 .083 .2 S/X = . The relationships given by the tables are graphed in Figures S7.3).12 . and S7.000.005 .175 rf . σ = .198 ∂C > 0.665) – 4.034 .037 .5 – .031 .000 (. and the ∂rf T in Months 2 4 6 8 10 12 14 16 18 Table S7.132 .024 .342.071 .21603) $5.08 .081 .4.342.000 – 417. Fourth Edition S = 5.028 .82 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.5 + .049 S/X = 1.038 .000.167 .14 .452) – 4.091 S/X = 1.890 = $4.9 .2).104 .089 . plotted against the time to maturity (in Table S7.095 .072 .142 . S7.056 .066 . since stockholders in this example hold a call option on the value of the firm. Table S7.138 .000 – 4.3679)(.3.167 .077 .178 . the risk-free rate (in Table S7.000N(1.160 .146 .

118 .141 .4 σ . rf = 6% S/X = .068 .000 .000 .25 .112 .40 C/S as a Function of σ Assume: S = 1.126 S/X = 1.041 .033 .045 .15 .119 .028 .005 .131 .054 .118 .0 .3 C/S as a function of rf .072 .2 C/S as a function of T Figure S7.163 .9 .35 . T = 6 months.05 .1 .030 .152 .20 .015 .082 S/X = 1.175 Figure S7.124 .10 .099 .058 .Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 83 Table S7.01 .30 .085 .

BS. The above condition can also be rewritten as follows: S < X(1 − e − rf (T − t) ) If the stock price falls below a predetermined level on any given calendar date. it pays to exercise the put option early. Therefore. there is no difference between having only common stock or having subordinated debt and common stock outstanding.84 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. if you hold the option to maturity it is worth X dollars at most. then the Figure S7. If the date of exercise is t and the time of maturity is T. the price cannot fall below zero. However. Therefore. As far as the senior debt is concerned.4 C/S as a function of σ 8. equals the value of the firm minus the call option: BS = V – [VN(dS) – DS e − rf T N(dS – σ T)] dS = ln(V/DS ) + rf T σ T + (1/ 2)σ T . The owner of a put option profits when the stock price falls. This fact limits the maximum profit from holding a put. the value of senior debt. At any point in time the put can be exercised for X – S dollars. In words. Alternatively. this condition says that you can gain more from exercising immediately and putting the proceeds in a risk-free asset than you could from holding the put to maturity. Fourth Edition Figure S7. In either case the senior debt-holders own the firm but have written a call option held by the junior securityholders to sell the firm for DS at the * calendar date. the condition for early exercise is (X − S) e rf (T − t) > X. 9. t .5 The early exercise of an american put option future value if you exercise now is (X – S) e rf T − t .

the correct ratio of all contracts per share of stock is [1/N(d1)].5 percent. there is no difference between senior or junior * debt. are out of line relative to each other. This is always equal to the face value of all the debt DS + DJ. the option which can be exercised at any time is more valuable than the corresponding option which may be exercised only at maturity. If these two positions are combined. It really doesn’t matter which price is out of line. 10. As shown in problem 7. we can form a risk-free hedge by (1) writing 1/( ∂ C1/ ∂ S) August 45 calls and buying one share of stock and (2) buying 1/( ∂ C2/ ∂ S) August 35’s calls and selling short one share of stock. all that matters is the exercise price they must pay to repurchase the firm at t .Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 85 where T is the remaining time to maturity. This implies either that (1) the August 45’s are overpriced by the market or (2) the August 35s. which provided the estimate of implicit variance of 51. 11.15. the value of the subordinated debt must be . If the August 45’s are overpriced relative to the August 35’s. The Black-Scholes formula is C = SN(d1) – X e − rf T N(d2) The partial derivative with respect to S is ∂C = N(d1 ) ∂S Therefore. The Black-Scholes model estimated the price of the August 45’s options to be $0. both written on the same asset. An American put is always worth at least as much as the equivalent European put because the American put can be exercised before its maturity date. Therefore.09 while the actual market price was $0. All that matters is that two options. the stock holdings net out. all we need to do is write 1/( ∂ C1/ ∂ S) August 45s and buy 1/( ∂ C2/ ∂ S) August 35’s. there are conditions where it pays to exercise an American put early. are undervalued by the market. According to the BlackScholes model d1 = ln(S/ X) + rf T + (1/ 2)σ T σ T + (1/ 2)σ T Since the value of the firm equals the value of its securities. So the value of the stock is S = VN(dJ) – (DS + DJ) e − rf T N(dJ+S – σ T) where dJ + S = ln(V /(DS + DJ )) + rf T σ T BJ = V – BS – S The subordinate debtholders own a call option to purchase the firm from the senior debtholders for an exercise price of DS but have written a call option owned by the stockholders allowing the stockholders to buy the firm at an exercise price of DS + DJ.8. Therefore. From the point of view of the common stockholders.

0171)(30 / 365) (.112212 + . we can solve for N(d1).5 – . Fourth Edition Using the facts applicable to Krespy Kreme August 45s we have ln(34.515)( 30 / 365) (.515)( 30 / 365) = and from the Table of Normal Areas ' N( d1 ) = .9793 = 1.1477 = –1.. sell) [1/( ∂C / ∂ S)] call options. 12.4793 = .2614 + .515)( 30 / 365) . if we write [1/N(d1)] call contracts.046 Repeating this exercise for the KK August 35s.75/ 40) + (. we have ' d1 = ln(46.1476 + (1/ 2)(. we have ∂C = N(d1) ∂S Therefore.5 + .0014 + .75/ 35) + (. we will have created the correct hedge.6865 and from the Table of Normal Areas N(d1) = . d1 = ln(S/ X) + rf T + (1/ 2)σ T σ T ln(44.9793 Therefore.0738 = –1.9131966 + .04 .02 of the August 45 calls.0683619 = .0014 + .7603 + .0738 = 2.65 / 45) + (.31)( 71/ 365) (.0171)(30 / 365) d1 = + (1/ 2)(. (a) Black and Scholes show that a riskless hedge can be created if we continuously adjust our portfolio so that for every share of stock we write (i.86 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.046 = 21.0126438 + .e.1367239 = .0683619 .515)( 30 / 365) = −. the hedged portfolio will be to write 1/( ∂ C1/ ∂ S) = 1/N(d1) = 1/.74 of the August 35 calls and buy ' 1/( ∂ C2/ ∂ S) = 1/N( d1 ) = 1/.2815 + .981559 = . Given the facts of the problem.31)( 71/ 365) . The BlackScholes formula for the value of a European call is C = SN(d1) – X e − rf T N(d2) By taking the partial derivative with respect to S.454 = .0738 .065)(71/ 365) = + (1/ 2(.

we have C – P = S0 – X e − rf T Solving for P: P = X e − rf T – S + C Substituting the Black-Scholes formula for the value of a European call.135 put options 1 − N(d 1 ) . 1 share = 6. we can obtain an exact solution to the problem by using put-call parity. Since N(d1) = . for each share of stock you hold. if you owned one share you would buy 6. you would buy 1/(1 – N(d1)) put options.815 shares.837.163 shares = 1 put .Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 87 From the Table of Normal Areas. a change in the stock price is offset by an equal and opposite ∂P : change in the value of the put options held.815 shares = 5 puts .1947 Therefore.10.135 puts 1 shares = 1 put 6. a risk-free hedge can be formed if you purchase enough put contracts so that ∂P/∂S = –dS.163 If you own five put options.337 = .135 . we have N(d1) = .5 + . From the equation for a put.837 The inverse of N(d1) is the number of call options to write against each share. you would buy . 1/(∂C/∂S) = 1/N(d1) = 1. the solution is only approximate.135 puts. that is. we have P = X e − rf T – S + SN(d1) – X e − rf T N(d2) = X e − rf T (1 – N(d2)) – S(1 – N(d1)) If you own one share of stock. if you hold 100 shares of stock you should write 119. (b) If we assume that Honeybear options are European options. If we write equation 7. we can find ∂S ∂P = –(1 – N(d1)) ∂S Therefore. 1 1 = = 6. If the option is an American option.47 call options. the put-call parity formula in continuous time.

Otherwise they would not enter into the contract. In an efficient market. if we hold one call contract. this is 1⋅S– 1 1 C – (1 ⋅ S + P) = VH N(d1 ) N(d1 ) The position in common stock nets out. . Either way you receive $X at the end of 5 years (assuming that the contract will be honored no matter what happens). It is only after the fact (i. Initially they both anticipate returns commensurate with the systematic risk of their position. will also be riskless.e.88 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. The contract is really an application of put-call parity. To answer this question. the option would be priced to reflect the perceived probability of winning or losing. after the option expires) that one investor wins and the other loses. At maturity the only possible events are: (1) The store is worth more than $X at the end of 5 years. If this happens she must buy it for $X. C= = = 1/(1 − N(d1 )) P 1/ N(d1 ) N(d1 ) P 1 − N(d1 ) .6 Frequency function for price of S 14. The call writer will have a negative β of equal magnitude and opposite sign. The difference between the riskless position for stock and calls. we should buy 5. and you receive $X.135 puts. (2) The store is worth less than $X at the end of 5 years. In other words. while the call buyer “wins” if the share price is above X.163 = 5. all we need to do is appeal to the logic of the CAPM. leaving us with the necessary position in call and put contracts. neither knows what the final outcome will be.837 P . the call writer “wins” if the share price is below X. if we buy 1/N(d1) calls.6 shows one possible instantaneous frequency function for the price of the underlying asset. if we assume that the puts and calls are European options. The only question is how to establish the exercise price. A priori. Given an exercise price of X. Figure S7. Mathematically. S. The call buyer will have a positive β . $X. we should also buy 1/(1 – N(d1)) puts.135P 13. In this case she buys the store for $X and you receive $X in cash. so that you earn 20% per year.. Therefore. we can find an exact solution. and the position for stock and puts. as shown below. Figure S7. Fourth Edition (c) Again.

Thus.000. therefore S0 + P0 – C0 = $80.Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 89 If we assume discrete compounding.2) 5 X (1 + r ) 5 .000 = and the exercise price is X = $199.065. .60 X (1. the put-call parity relationship is S0 + P0 – C0 = The amount lent is $80. 80.000.

93%.5209 = = 1.1703 = 1 + 2 f 3 (1. There is mean reversion in the process specified in equation (8. Forward Contracts. T) = A(t.2996 Therefore. T)r(t) . we can solve the problem by computing the following ratio: (1 + 0 R 3 )3 (1 + 0 R1 )(1 + 1f 2 )(1 + 2 f 3 ) = = 1 + 2f3 (1 + 0 R 2 ) 2 (1 + 0 R1 )(1 + 2 f 2 ) (1.5): B(t.4): dr = ν(µ – r)dt + σdz where µ is the long-term mean of the short-term rate. in turn. impliea a negative (positive) drift. implies that. 8. as compared to the long-run mean.140) 2 1. The price at time t of a zero-coupon bond that pays $1 at time T is then given by (Eq.155)4 1.14) 2 1. ν is the speed of adjustment of the short-term rate to the long-term mean and σ is the instantaneous standard deviation.Chapter 8 The Term Structure of Interest Rates. r. 8.15)3 1. an average. The formula is given below: (1 + 2f3 ) (1 + 3f 4 ) = = (1 + 0 R 4 ) 4 (1 + 0 R 2 ) 2 (1 + 0 R 1 )(1 + 1f 2 )(1 + 2 f 3 )(1 + 3f 4 ) (1 + 0 R 1 )(1 + 1f 2 ) = (1. 2. 8. and Futures 1.3693 (1. T)e –b(t.2) is valid.02%.7796 = = 1. the implied forward rate for the third year is 2 3 f = 17.3693). the short-term rate will decrease (increase) towards the long-run mean.2996 The two-year rate of interest is 36.4) since a higher (lower) current short-term rate.03% (b) The rate of interest on a bond held from the beginning of the third year and held to the beginning of the fifth year is a two-year rate of interest equal to the product of the expected forward rates during the third and fourth years. Vasicek (1977) assumes the following mean-reverting process for the short-term rate (Eq. which.5 –1 = 17. µ. (a) Assuming that the unbiased expectations hypothesis (as given by Eq. The average one-year rate is (1.

ν.6) suggests that the entire yield curve can be obtained as a function of r(t) once the three process parameters: the long-term mean of the short-term rate. (8.T)  σ   −   4ν  ν2        A (t.3767 1. the current short-term rate.T)e − b(t. where tRT is the interest rate at time t for a term of (T – t). Cox. µ.9443 0.7): dr = ν(µ − r)dt + σ( r )dz The price of a zero-coupon bond in this framework is given by the equation (8. In the model. T) = 1 − e− v(T − t) v    2 σ2   ( b(t. Forward Contracts. T)] 0. the short-term interest rate process is Eq. is 10% and the rate of adjustment.9953 2. T) = e− (t RT )(T− t) .1617 A(t. we get: t RT = 1 [b(t. the speed of adjustment.58% 2. σ are specified. the long-run mean.87% 3. Ingersoll and Ross (1985) (CIR) propose a mean reverting process for the short rate where the standard deviation of the changes in interest rates are proportional to the square root of the level of the rate. These set of parameters result in the following term structure of interest rates: Time to Maturity in Years (T – t) 1 2 3 4 5 Value of a ZeroCoupon Bond [B(t.08% b(t.9707 0.T)r(t) − A(t. T) = e Since B(t. is 0.00% 3.8951 3. In this question.9464 0. ν. σ.T)r(t) . µ.9907 0.8848 0. r(t) is 2%. The term structure can be upward-sloping. downward-sloping or humped depending on the values of the various parameters.T) = A(t.06% 3.T)] T−t Equation (8.4.7470 1. the instantaneous standard deviation. and the instantaneous standard deviation.9208 0.8572 Interest Rate [tRT] 2.9745 0.8): B(t. T) 0.9139 0.8242 1.Chapter 8 The Term Structure of Interest Rates. and Futures 91 where r(t) = the short rate at t b(t.T )−T + t)  ν µ−   2 2  2 b (t. T) 0. is 6%.

65% 3. In this question.9373 0. T) 0. Fourth Edition where b(t.9734 0. Ho and Lee propose a binomial model for bond prices of all following form: . T) =   2 γ + ( ν + γ ) (e γ (T− t) − 1)    γ = ν 2 + 2σ 2 The interest rate from t to T.T)] T−t where b(t.8963 0. T) are as defined above.9279 0.9634 0.3705 1. would be given by Equation (8.97% 4. the current short-term rate is 2%.9): t RT = 1 [b(t. These set of parameters result in the following term structure of interest rates: Time to Maturity in Years (T – t) 1 2 3 4 5 Value of a ZeroCoupon Bond [B(t. the instantaneous standard deviation is 10% and the rate of adjustment is 0.23% b(t.4. the long-run mean is 6%.1284 A(t. tRT.8231 1.T)r(t) − A(t.92 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.8531 0.70% 3.8447 4.24% 3. T) 0. T) and A(t.7326 1.9712 2. T) = 2 γ + (ν + γ )(e γ (T− t) − 1) 2 νµ/σ 2 2 (er(T−t) − 1)   2γe( ν+γ )(T−t)/2  A(t.9895 0.8874 0.8095 Interest Rate [tRT] 2. T)] 0.

0. 0.10b.5B(1.T) = h(T − t) B(t − 1. 8. Combining the no arbitrage condition with the perturbation equations yields the following constraint on the perturbation function (Eq. i.5 and 0.i. the value of a bond that pays $1 in 2 periods is going to be 0.9496 and B(1.053 1. T) is the initially observed term structure of bond prices.9021 in one period. These inputs would result in the following values for h(T).1. T) is the price of a bond at time t and state i that pays $1 at time T.11): B(t.T) h * (T − t) B(t − 1.10a.1. 0.5B(1.T) B(t. 2).i. h*(T) and B (0.T) = [πB(t + 1. This implies that with a perturbation of h or h*. 2 and 3-year rates are at 4.T) + (1 − π)B(t + 1. t + 1) where π is the probability associated with the perturbation h(τ).2)]B(0.105 1.1.i.9.2) = 0.9496 or 0.157 h*(T) 0.947 0.i.i + 1. 8. and Futures 93 where B(t. 1.0.10c) B(t − 1.0.9070 0.2) h*(1) Substituting values from the table into the above two equations yields B(1. 0.T) h(T − t) = B(t.895 0. respectively. 5 and 6%.1) The perturbation conditions imply that: B(1. 0. 8.2) h(1) = B(1. In this question.i.14): h( τ) = δτ 1 and h*( τ ) = π + (1 − π)δτ π + (1 − π)δ τ for some constant 0 ≤ δ ≤ 1. 0.T)]B(t. 2) = [0. T) h*(T − t) The no-arbitrage condition implies that (Eq.i. 8. h(τ) and h*(τ) have to satisfy the following conditions(Eq.9021. 2) = 0. 8. 8.Chapter 8 The Term Structure of Interest Rates. In addition.i.8396 Applying the no-arbitrage condition to B(0. Note that B(0.i + 1. 8.2) + 0.t) h(0) = h* (0) = 1 The above equations when divided imply that (Eq.i.i + 1. T) 0.0. respectively.i.13): πh(τ) + (1 − π)h*(τ) = 1 Ho and Lee also show that for path independence to hold such that the order of the perturbations is not important. π and δ to be 0.9615 0.t) B(t.T) = B(t − 1. T): Maturity (T) 1 2 3 h(T) 1. Forward Contracts. we get: B(0. .843 B(0. The evolution of bond prices is based on perturbation functions h(τ) and h*(τ) such that (Eqs.12) B(t.

000(1. (a) The first step is to compare the product of short-term rates against the longer-term rate.94 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Fourth Edition Applying the same logic and process at different nodes yields the following tree of bond prices: 5.012541) 1.045205 > (1.10) 167/360 > (1.125) 90/360 (1.000. answer is that hedgers may have no need to protect themselves against price fluctuations more than 18 months in the future.000 face value 90-day T-bill on March 22. There are several possible answers to this question.045205 > 1. (b) To make a riskless arbitrage profit we need to borrow enough to purchase one futures contract that will deliver a $1.000.83 –167/360 = $1. there may be no market for long-lived contracts because their prices would not fluctuate enough to create any trading volume. As a result.30).042799 The results tell us that a higher yield can be obtained if we are long in the 167 day T-bill and short in the futures contract and the short-term (77-day) T-bill.000. using the insight provided by Samuelson in Eq. it costs PV = $1.749. (8. Far distant contracts have relatively lower variance because autoregressive prices have a long interval to correct themselves. If the 167-day T-bill yields 10 percent. and related.95674983) .06) 77/360 1. A second.10) = $956. we note that the variance of futures contracts declines for longer-lived contracts.000(. 6. First. (1.029884) (1.

83(1. and the variance of the futures price will decrease as we get closer to maturity.49 Simultaneously.00 – $31.07 = $10.93 . In return.5) = $42. On March 22. to cover our short position. (b) If you believe the stock price will be $42 six months from now.07(.5) 77/360 = $956.05 7. with the same time to maturity and exercise price. on the underlying asset with maturity T and exercise price X equal to the forward price 0 FT. a.000 face value T-bill that now has 90-days to maturity.21 is greater than one.00 – $30.80 in order to have zero cash outlay.000.54 Loan payment 968.749. in Eq.00 – $30. P. we receive the following amount of cash PV = $1. The put-call parity equation is C0 – P0 = S0 – X e − rf T and the exercise price must equal the forward price.000(. we deliver our $1. (a) We can create a synthetic forward contract by buying a European call.80 Consequently.2 (b) Trends in spot prices can affect the variance of futures prices. our profit (on March 22) is Cash received $970. C0 – P0 = S0 – 0 FT e− rf T C0 – P0 = 30 – 35e –. (a) Trends or seasonality in spot prices have no effect on futures prices.000. we will sell short a T-bill futures contract with a $1.000. Therefore.00e.012541) = 30 – 35(. you profit at delivery will be Profit = ST – 0FT = $42.54 After repaying the loan. and Futures 95 We will borrow this amount at 6 percent for 77 days.000.983.749.125)–90/360 = $1. the autoregressive coefficient.965605) = 30 – 33. Forward Contracts. Since spot prices are expected to rise from February through May.000 face value.03562) = $42.49 Pre-tax profit $ 2.83(1.07(.Chapter 8 The Term Structure of Interest Rates. on March 22 we will repay FV = $956.79619 = –3.97098354) = $970.235.00 – S0e rf T = $42.06) = $968.748. it would be necessary to borrow $3. See Section D. 8.748. and simultaneously writing a European put. 8.000(1.983.00(1. therefore. C.

The risk of changes in expected convenience yields and carrying costs beyond the March 1987 contract expiration date are assumed to be small. If so.105171 – 1. The general idea is to minimize shareholders’ risk by augmenting their current position with T-bond contracts. If the two contracts are highly correlated.3 + 421.08) – Ste (3/12)(. From Chapter 5. the result is N = −∑ − i =1 N ρ i.884.30)(. take the derivative with respect to the number of futures contracts.684 . An effective strategy. would be to sell long-term T-bill futures and buy shortterm T-bill futures. you will make a profit. if the spot price of silver is greater than $4.684 per ounce. we have N=– (. If the yield spread narrows you will lose money. similar to what was covered in Chapter 5 (Portfolio Theory). (b) The profitability of your position depends on the spot price of silver.398 = St = $4.F = the correlation between the ith asset (liability) and futures contracts σi = the standard deviation of the ith asset (liability) PF = the price of a futures contract Vi = the market value of the ith asset (liability) Substituting in the facts of the problem. write out the variance of the shareholders’ position. Hence the carrying costs and convenience yields should be the same for your expected 3-month holding period. the difference in the values of the two contracts is t F3/88 – tF3/87 = Ste (15/12)(.9 = –1. Fourth Edition 9. 11. a short position is required.08(. According to Eq. Since T-bonds are positively correlated with the market value of loans.306. and set the result equal to zero. then close out your position after the yield spread widens. called a time spread.15)(.08(. you are planning to close out your positions in March of 1987 when the shorter-lived contract expires.4 contracts Therefore. This effect is counterbalanced by the positive correlation between T-bonds and demand deposits (a liability).008 – $5. (a) Presumably. If the market index falls.08 mm) = –2. As interest rates rise. Note that this is not a hedged position.06)(820 mm) (. (8.610 = St (1.020201) $.02)(−900 mm) − . To find the optimal number of contracts.08 mm) .96 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. losses in your portfolio will be offset by gains in your stock index position (and vice versa).08) $6.34) the price of a T-period futures contract is t FT = St e t r T + t WT − t CT Your position is the difference between two futures contracts which will be held for the same period of time. you bear little risk until March 1987. To alleviate timing or market risk. 12. you can sell short stock index futures contracts. the optimal hedge is to go short 1.884.084970 Consequently.Fƒ i 'i ƒ F !F where: N = the optimal number of futures contracts ρi. the market value of fixed rate loans falls. but would be offset by gains from a short position in T-bonds.4 T-bond futures contracts. 10. This is a cross-hedging problem. .

16 Figure S9.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 1. Without the option to expand the NPV of the project is 10. The up and down movement for the binomial tree are estimated as follows: Annual Volatility (σ) = 15% u = eσ d= t = e 0.16 1 1 = = 0. (a) First.86 u 1.15 1 = 1. what is the project’s NPV without the option to expand? NPV cannot capture the flexibility of the project.1A Present value event tree .

(a) First. we must estimate the NPV of the two mutually exclusive projects.53. The real option to expand turns out to be valuable because the project can be optimally expanded under certain scenarios. Fourth Edition (b) What is the project’s ROA value with the option to expand? (Prove your answer). The ROA value of the project is $11.2A NPVS of the projects and PV event trees . which is higher than its NPV of $10 Figure S9. Present Value Event trees and objective probabilities for each project are given below.98 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.1B ROA decision tree 2. The NPV of both projects is $20 Figure S9.

2B Decision tree for the projects Using Decision Tree Analysis (DTA).Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 99 Using a Decision Tree Analysis (DTA). the option to contract should be executed next year if their present values continuously decline. the option should also be executed at the end of the second year if the present value is at its current level Figure S9. answer the following questions: Which project should the company select? Since the DTA value of Project 1 is higher and positive it should be selected for execution. With Project 2. answer the following question: when and under what conditions would the options to contract be executed with each project? With Project 1 and Project 2. Note that DTA discounts using the WACC even though risk has changed. Figure S9.2C Optimal execution using DTA .

which project should the company select? As the ROA value of Project 2 is higher and positive. its value is equal to the difference between the DTA value and the NPV Figure S9.2D Value of the option to contract using DTA (b) Using Real Option Analysis (ROA).100 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Fourth Edition Using a Decision Tree Analysis (DTA). it should be selected for execution Figure S9.2E Real options NPVs . answer the following question: What is the value of the option to contract with Project 1 and Project 2? As the option to contract is the only flexibility accounted for in the DTA valuation.

contract if the PV of the project drops during the first period. contract at the end of year two if the present value has been falling successively. This optimal execution is different when using ROA versus DTA. Figure S9. With Project 2. answer the following question: When and under what conditions would the options to contract be executed with each project? With Project 1.2G The value of the option to contract using ROA .2F Optimal execution gives RDA What is the value of the option to contract with Project 1 and Project 2 using ROA? As the option to contract is the only flexibility accounted for in the ROA valuation.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 101 Using a Real Option Analysis (ROA). or if the present value is at its current level at the end of the second period. its value is equal to the difference between the ROA value and the NPV Figure S9.

Figure S9. The risk-free rate is 5%. One of the companies wants to buy the right from the other to acquire its 50% share in a year for $15 million.102 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. With DTA we wrongly select Project 1. Figure S9. Two companies are developing a 50/50 joint venture with an NPV of $25 million.2H Project 1 and Project 2 valuations 3. The annual volatility of the venture is 20%. Fourth Edition (c) Do the DTA and ROA valuation suggest the same optimal execution for the options. and with ROA we correctly select Project 2.3A Two period per year event tree for 50% of the joint venture . the same value for each of the two projects? The DTA and ROA valuations suggest different optimal execution of the option to contract for Project 1.

but the NPV remains the same. the real option to acquire half of the joint venture in a year is worth 0.3C Three period per year event tree for 50% present value of the joint venture . what is the maximum price the company should be ready to pay for the option? Valued with two period per year.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 103 (a) Using a two period model (six months per period.46 million Figure S9.3B PV of the right to acquire 50% of the JV (b) Given 3 periods per year.). Figure S9. the estimate of u and d changes.

104

Copeland/Shastri/Weston • Financial Theory and Corporate Policy, Fourth Edition

Using a three period model (four months per period) how does the option price change? Valued with three periods per year using the replication method, the real option to acquire half of the joint venture in a year is worth 0.41 million

Figure S9.3D Different Estimate of value, assuming 3 Binomial Trials per year (c) How can we use the Black-Scholes formula to solve this problem? What is the option price if we use the Black-Scholes formula? As the option to acquire half of the joint venture in a year is a European Call Option we can use the Black-Scholes formula to value it. The value is 0.41 million Inputs for Comparison Calculations • Present value of underlying (V0): 12.5 • Risk-free rate (rf): 5% • Exercise price (X): 15 • Annual Volatility (σ):20% • Time Span (T): 1 Step 1: Calculate d1 and d2 for Black-Scholes formula d1 = d1 = d1 = Step 2: ln(V0 /X) + rf T σ T 1 + σ T 2

ln(12.5/15) + 0.05 * 1 1 + 0.20 1 2 .20 1 −0.1823 + 0.05 + 0.1 = −0.562 .20

Calculate cumulative normal probabilities N(d1) = N ( –0.562) = 0.287 d2 = d1 − σ T = −0.562 − .2 = –0.762 N(d2) = N ( –0.762) = 0.223

Chapter 9

Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis

105

Step 3:

Calculate option value C0 = V0 N(d1 ) − Xe-rf T N(d 2 ) = 12.5*0.287* – 15e
–0.05

(0.223)

= 3.588 – 3.182 = 0.41 (d) Which of the three prices would you use to make a decision and why? What price would you use if the buyer wants the right to buy the share at any time during the year? The Black-Scholes formula provides the most correct valuation and so 0.41 million is the correct price (e) We can use the same price if the option is American and could be exercised at any time. As the joint venture does not pay dividends the option should always be exercised at the end of the year. 4. (a) First create on event tree for the underlying risky asset

Figure S9.4A Event tree for underlying asset (b) Execution of the option must occur at the end of the decision tree (it is a European put) Use of replicating portfolios to value the project with the abandonment put option

Figure S9.4B ROA of the European put

106

Copeland/Shastri/Weston • Financial Theory and Corporate Policy, Fourth Edition

5. (a) Event tree for the underlying risky asset. Note that the value of the underlying ex dividend is ten percent less when the dividend is paid at the end of each period

Figure S9.5A Event tree with divided payment (b) Note, in this case, we have optimal execution of the option at the end of the decision tree and we used replicating portfolios to value the call option

Figure S9.5B Valuation of the American call

Chapter 9

Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis

107

6. (a) The first step is to create the event tree for the underlying risky asset

Figure S9.6A Event tree for the underlying risky asset (Non-recombining) (b) The second step is to use replicating portfolios to value the call option, assuming optimal execution of the option at the end of the decision tree

Figure S9.6B ROA value

we have made the following assumptions: • If the mine is open to begin with. Fourth Edition 7.7A shows the calculations.000 tons) at the rate of 4.000 tons per year. at node D you can expand. the values of these alternatives are respectively $2545. To solve the problem. contract or keep your options open. Figure S9. so that if the mine stays open it will require 2 time periods to be exhausted. • If the mine should be closed at time zero.108 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. $1600 and $2554.000 tons immediately. it produces 4. the best action at node D is simply to keep your option open. If closed (at a cost of $30. there is no production a time zero (inventory is not reduced) and the closing cost is paid immediately. Figure S9. For example.000) there is no production.7A 8. . Therefore. This is a switching option.000) the mine depletes the inventory (12. If open (at a cost of $20.

all cash flows (FCFC) and the present value (PVC) are zero.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 109 Price and free cash flow event trees with objective probabilities for the extraction project are given in Figure S9. Figure S9.8B Value event tree (open mine) . Of course.8A. as shown in Figure S9. Figure S9. if the project is not in operation. assuming the mine is open.8B.8A Price and cash flow event tree Since all decisions are based on value. we need to construct the present value event tree for the project in operation.

110 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Each decision compares the marginal cost of closing/opening (lost cash flow from current operations and closing cost) with the present value of keeping the mine open/closed 3. Identify optimal action for each node on the event tree entering in “Open” and “Closed” mode. the cash flow in the valuation of the mine is zero even though it may be reopened in the future if the price is sufficiently high 4. The optimal reopening of the mine in the future should be decided by solving the same type of problem at each future node while deposits remain. starting from the back of the tree 2. ROA value of the extraction project entering in each state (price level) in mode “Closed” Figure S9. If we close the mine at a node with remaining deposits in the ground. Fourth Edition The Solution Logic is the following 1.8C Present value assuming the mine starts in a “Closed” mode .

8E Solution to switching option .Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 111 ROA value of the extraction project entering in each state (price level) in mode “Open” Figure S9.8D Precent value assuming the mine starts in as “Open” mode The final solution starts by comparing the period 2 values of starting the period open or closed. changes the optimum (enter open) then works back to period 1—using a “backward forward” algorithm Figure S9.

If the project is not in operation. Figure S9. Fourth Edition How does the answer change if the price of the mineral is currently $26? Price and free cash flow event trees with objective probabilities for the extraction project given the current price of $26 per ton and shown in Figure S9. Figure S9.8F.8F Price and CF event trees Present value event tree for the project in operation. all cash flows (FCFC) and the present value (PVC) are zero. assuming a current price of $26 per ton.8G Value tree .112 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.

8H Values of the project entering in mode “Closed” Figure S9.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 113 Figure S9.8I Values of the project entering in mode “Open” .

114 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.8J New solution.370 Figure S9. give price of $26 . Fourth Edition The higher price per ton changes the optimal starting mode of operation to “open” and raises the value of the project to $67.

9B values the equity. therefore the present value of the debt is $516 and its yield to maturity is 15.9A. Its exercise price is the $800 face value of the zero-coupon debt and its maturity date is 3 years. The value of the equity is $484.9B Equity valued as on option on the levered firm . Figure S9. The binomial event tree for the underlying unlevered form is given in Figure S9.7% Figure S9.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 115 9. This is an example of a simultaneous compound option. Figure S9.9A Binomial event tree for the unlevered firm The equity of the levered firm is a call option on the firm.

the equity is the underlying asset for the American call that has an exercise rice of $400. Figure S9.10 shows the underlying.116 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. The first option. Its value is calculated in Figure S9.9C Call on equity of a levered firm 10. chronologically.10a Firm value event tree. therefore the option values of the second option are the underlying for the first. Figure S9.9C as $260 Figure S9. The problem statement failed to provide a risk-free rate. is contingent on the value of the second. . therefore we assume it is 8%. Fourth Edition Finally.

The option value at nodes F and G can be calculated in the same way. no investment will be made at node E.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 117 Figure S9. is $1.128 − 800 = 328 m(1.10b Valuing the Second Option Therefore. if we use the replicating portfolio method. Because the option is worth $29 more if it is not exercised. the option is worth m u 2 V0 + B = 1(1.271 − $800 = $471. The value if the option is exercised.433 −833 = = −771 1.08 1.128) = 305 305 =1 m= 305 mu3 V + (1 + rt )B = 633 1.10b shows the valuation of the second option.89% and B = −583.08)B = 633 B= 633 − 1. we have mu3 V + (1 + rt )B = 1. . At node H. At node E.433 + (1. m = 0.433 − 800 = 633 mu 2 dV + (1 + rt )B = 1.433 − 1.271) − 771 = 500 at node E. for example.08 Figure S9.

08B = 0 21 − 0 21 m= = = . the payout of the first option is the value of the second option chronologically.08B = 21 m 201 − 1. we have m 421 − 1.08 Finally.10c Valuing the First Option and solving for B.10 (393) − 19 = 16 .10 421 − 201 220 Figure S9. we have m (421) − 1. Again.118 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. the value of the compound option at the zero is mV + B = . If exercised at the end of the first year. we calculate the value of the first option as shown in Figure S9. using the replicating portfolio approach at node A. Fourth Edition Next.08B = 21 21 − 42 = −19 B= 1.10c.

The fact that some investors enjoy higher returns than others is the result of the uncertainty in stock returns. the firm is actually earning worse than normal returns. it is impossible to earn an arbitrage return because everyone knows what the price will be. If we want to graph the price of Clark Capital common stock over time. the average (and expected) return on the market is positive. Figure S10. The rate of return earned on the stock would be the same as the risk-free rate. (a) Clark Capital Corp. stock is a fair game because in the long run the expected payoff is equal to the actual payoff. in fact. we know that just after the ex dividend date. . Except for very bad years. brokerage fees may wipe out any trading profits. Even so.05 The day before it goes ex dividend it is worth $200 plus the $10 dividend. some observations will lie above the mean and some will lie below. which would make the investment a submartingale. Given any probability distribution. The market is efficient in its strongest form. This is considered a normal return. If the computer costs are less than the 3 percent profit. If the computer costs exceed the excess 3 percent profits from stocks. 2. Also.Chapter 10 Efficient Capital Markets: Theory 1.1 The price of Clark Capital Corporation common stock over time (b) At t the price at t + 1 is known with certainty. semi-strong capital market efficiency may be refuted. in this example the expected and actual payoffs are precisely equal at every instant. it is worth S0 = D1 $10 = = $200 rf . The price is graphed in Figure S10. (a) The information which enables the brokerage firm to earn a consistent 3 percent abnormal profit is not costless. because all relevant information is known and impounded into the price of the stock instantaneously. (b) The hypothesis of an efficient capital market is not contradicted. It is also a fair game.1.

062 ∑ p(e|m )U(a 1 2 . all relevant publicly available information has not been instantaneously incorporated into stock prices. benefits from insider information are accommodated under this hypothesis.18 e 2 −1. In the example given.1) (0) = − .1. Fourth Edition (c) Semi-strong (as well as weak) capital market efficiency is contradicted. we need to find optimal actions given a specific message.00) = .18 −. e) is a2 a1 e1 .5 If the message is m1 : for a1: for a2: ∑ p(e|m )U(a . nearly costless.00 Notice the benefit function includes the opportunity cost.8 q(m2) = .5 . 3.9) (.162 .e) = (. U(a. price information that enables you to forecast future prices with better-than-random accuracy.2 e1 = customer pays loan e2 = customer defaults a1 = First National grants the loan a2 = First National refuses the loan The information structure is m1 e1 . where apparent abnormal profits were available to insiders. e) − V(η0 ) a e where m1 = favorable credit check m2 = unfavorable credit check q(m1) = . which are calculated according to the formula max ∑ p( e|m) U (a.00 0. –18 percent. You have discovered a trading rule based on past.1: V(η) = ∑ q (m) max ∑ p (e|m)U(a.18) + (. strong form efficiency is refuted but semi-strong and weak efficiency are unchallenged.1 The benefit function. thus. In order to find the correct values for the last sum in equation 10. Thus. (d) The semi-strong form of efficient capital markets assumes publicly available information is instantaneously impounded into prices. of not granting a loan when it would have been paid.18) + (.9) ( − . e) a m2 .1) ( − 1.9 e 2 . The value of credit check information can be calculated according to equation 10.e) 1 1 = (.120 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.

and the value of the optimal actions with information. V (η) – V(η0).18) + (. With information From previous calculations. = 3. E(R) = –9 percent.09 given m2.1(0) = 0.16 – (–3. If we cancel the product our payoff is determined by multiplying the payoffs in various market conditions by their prior probabilities as follows: V(cancel) = ∑p i (payoff)i = . Recall that the loan program has been losing money. the bank would choose always to grant the loan (a1). Finally.2 percent.8 (.18) + (. the optimal action is a1 (grant the loan) for an expected return of 6. with an expected payoff of $5 million.00) = −.1 as follows: V(η) = .82) (.3(0) + .e) = (. given m1.9(0) + .6(0) + .5) ( − 1.5) (0) = −. The expected payoff from going nationwide (without any test market) is V(nationwide) = .9(–10) + .41 )U(a 2 . we know ∑ p(e|m) U(a. the credit check should be implemented. This represents the optimal action given that the firm does not use a test market.3(10) + .e) = (. Substituting these values into equation 10.18) + (. If the message is m2 : for a1: for a2: ∑ p(e|m ∑ p(e|m 2 2 )U(a1 .24 percent. in the absence of information. for an expected loss of 3.062 given m1. we need to estimate the value of the optimal action without any test market. (a) If the firm is risk neutral it will maximize expected pro-profits. the optimal action given m2 is a2 (refuse the loan) with an expected opportunity loss. calculate the difference between the value of the optimal action without information.09) = 3.24) = 6.16% Since the cost of information = 5 percent.Chapter 10 Efficient Capital Markets: Theory 121 Therefore. Without information value of a1 = (.18) (0) = –14.1(80) = 5. Of these two actions it is best to go nationwide.18) (–1.82) (–.00) = –3. and the value of information. First. If we decide to test market the Kidwich product we can determine the optimal action given the outcome of the test marketing experience.062) + .6(–10) + .1(0) + 0(0) = 0 V (nationwide|no acceptance) = . The calculations are shown below: V (stop| no acceptance) = . If the experiment predicts no acceptance for Kidwich the optimal action is to cancel the product.1(10) + 0(80) = –8 .76% Therefore. 4. e) is . and –.24% value of a2 = (.4 percent.18) + (.2 (–.5) (.5) ( − .09 Therefore.

6(0) = 0 V (nationwide|success) = . if the experiment predicts success the optimal action is to go nationwide. Given that U(W) = ln (W + 11) the payoff matrix.1(0) + .6(80) = 50.1(0) + . expressed in utiles.3(0) + .7(0) + .1(2.398 U (nationwide) = .6(0) + . the optimal decision is to go nationwide.6 – 5 = $6.1(50) = 11. .1(–10) + .1(2. On the other hand.3(3.1(0) + .398 3.398) = 2.045) + . U (stop) = .398) + .3(2.2(0) = 0 V (nationwide|marginal) = .” you are correct. Nevertheless.045) + .1(4. it is interesting to see how risk aversion might affect the value of information.2(80) = 22. is Action Cancel Go Nationwide 2.3(22) + .620. Finally. e) m a e Having determined the optimal action given the test market results the value given the test market results is V(η) = . if there is a test market and if it predicts success. V (stop|success) = .122 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.511) = 1. The value of the optimal action with no information at all was $5 million.6(2. as shown below.398) + .3(3. The value of information is defined as V(η) = ∑ q(m) MAX ∑ p(e|m) V (a.511 Market Conditions No acceptance Marginal Success If there is no test market the optimal action is to cancel the product because it has higher expected utility.398) + .398 0 2.6 million.045 2.398) = 2.6.398 4.6(0) + . U (stop|success) = . therefore the value of the test market information is V(η1) – V ( η0) = 11. (b) If you said to yourself “There’s no such thing as a utility function for a firm.6(2.1(–10) + .511) = 3. the optimal action is to go nationwide. Fourth Edition If the test market indicates marginal acceptance. at least for this simple example.6(4.398 U (nationwide|success) = .3(10) + .3(2.365. The cost of the test market is $5 million therefore the net value of the information is $1.398) + . V (stop|marginal) = . The major difference is that the payoff matrix must first be converted into utiles so that expected utility can be computed.7(10) + .6 million.

at the margin.034) + . The corresponding level of wealth is V(η2) = $4.3(3. And since the optimal action given no information (i.398) + . cancelling the product) has a value of zero. Given the information provided by a test market and the optimal action given each experimental outcome the expected utility given the information is U(η2) = .04 million. V ( η0)=0. more individuals would enter the arbitrage business until it became true. If our expect to lose 83 percent of the time and if.9(2.511) = 3.Chapter 10 Efficient Capital Markets: Theory 123 If the experiment predicts marginal success the best action is to go nationwide. The options market would be a martingale if expected returns were equal to zero.04 million. However. One would expect that. the game has been fair.7(3. revenue from arbitrage just equals costs. 5.1(3.6(2. Costs include the cost of gathering information and a fair rate of return on physical and human capital. Martingale: [Prob (loss)] ⋅ [Expected value (loss)] + [Prob (gain)] ⋅ [Expected value (gain)] = 0 . And if the experiment predicts no acceptance it is best to cancel the product.511) = .1(3.398 U (nationwide|no acceptance) = .e.045) + 0(4.305.2(4. (a) A fair game means that. the value of information is V(η2) – V(η0) = 4.1(0) = ..398) + .398) + .398) + . on the average. 6.04 – 0 = $4. (b) The options market is a fair game so long as investors’ expected returns are equal to their long-run average returns.9(0) + . There is nothing in the efficient markets hypothesis which says that arbitrageurs cannot make profits. If options are expected to expire unexercised 80 percent of the time. If this were not true. then the options market would follow a martingale if and only if the expected net present value of an exercised option was four times as great as the expected investment loss. Because the test market costs $5 million. U (stop|no acceptance) = . it is important to look at their net profits after costs. over a period of time.620) = 2.2(2. outcomes conform to expectations.034. U (stop|marginal) = .398) = 2.711.1(2.96 million.045) + .398) = 2.398 U (nationwide|marginal) = . you do in fact lose 83 percent of the time. It would not be undertaken in the risk average case because its net value is $4.7(2.398) + 0(2.1(2.04 – $5 = $-.

.8/.. However. whether or not product markets are efficient. a martingale holds if –(. i. but we cannot determine whether or not the value of the firm reflects excess profits above a competitive return in the product market. This will continue.e. the NPV of any security is zero. that although the model may be used to test for abnormal returns in an efficient capital market. then the options market would follow a submartingale. driving up the price of A relative to B. This is just another way of saying that you get what you pay for. Given a world with investors of this type. In other words. Fourth Edition In this example. namely. It reflects one of the “weaknesses” of equilibrium models like the CAPM or the APM. namely that the starting market value of the firm reflected no excess profits (or monopoly rents). then it is possible to measure ex post excess returns via the CAPM or the APM. we can test for excess profits to investors in the capital market. 7. 9.2)(4L + a) – (.2a > 0 This result is a submartingale with E(R) > 0. if the assumptions of the CAPM or the APM are a reasonable approximation of reality.8)(L) = . If we assume a is some positive amount. In an efficient securities market.2) L G = 4L for L>0 If the expected NPV of a gain were more than four times the value of an investment loss. since all securities are priced according to their perceived appropriate risk. it cannot tell us whether or not the initial market value of a firm reflects abnormal returns. This has the effect of altering the expected distributions of returns until there is no second-order stochastic dominance in equilibrium.2) (G) = 0 G = (. Thus. and NPV’s are discounted according to (the same) appropriate risk. until the market price of A relative to B is high enough to make the marginal investor indifferent between A and B. Assuming that the market portfolio is known to all investors so that systematic risk can be appropriately measured.8) (L) + (. The second-order stochastic dominance criterion assumes that all individuals have positive but decreasing marginal utility. If we make a second assumption.124 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. they are greedy and risk averse. But the higher price for A (and lower price for B) implies lower expected returns for A (and higher expected returns for B). 8. if asset A dominates asset B and if they have the same present value. then the residuals from the security market line will reflect the presence or absence of any abnormal returns above those which could be expected from a perfectly competitive firm of equivalent risk. then E (R) = (. the second assumption is unrealistic. The net result is NPV = 0. then every investor will sell B and buy A.

Ibbotson’s results can be interpreted either as 1) selection of an inefficient index. 3. What happens when the market is not in equilibrium? Suppose new information is revealed such that the market must adjust toward a new equilibrium which incorporates the news. Should an individual—for example. If an individual can employ a strategy (using the –4. almost all of the “information” in an annual report has already been discounted into the security price before the annual report is released. he can earn a risk-adjusted abnormal return after transactions costs.56 percent trading rule of Dann. or 2) detection of abnormal performance during disequilibrium. (a) The securities market can be efficient even though the market for information is not. Given such a situation. Mayers and Raab) where he purchases shares at the block price. All that is required for efficient securities markets is that prices fully reflect all available information. Or suppose that a new security is introduced into the marketplace. then the market cannot reflect the information because it is not publicly available. Before the disequilibrium situation as well as afterward. it is unlikely that the latest copy of the annual report will allow an investor to earn an abnormal return. although it may cause portfolio adjustments that leave prices unchanged. This is not a deficiency in the securities market. a corporate insider—have monopoly access to valuable information. However. 4. . If it has no impact on prices. then sells at the closing price. it is largely irrelevant. (a) As shown by the Ball and Brown study (1968). Therefore. the abnormal performance of an asset can be measured by the arbitrage profits available as its price is adjusted to a new market equilibrium. The evidence may or may not be inconsistent with the strong form of the efficient markets hypothesis. (b) Information is material if it has an impact on securities prices when it becomes publicly available for the first time. 2. Roll’s critique (1977) is based on the assumption that capital markets are in equilibrium. Roll’s critique applies and we cannot expect to observe any abnormal performance relative to an efficient index. because once the block trade becomes public information. In this way. the adjustment process itself can be used to detect abnormal performance relative to the market index prior to equilibrium. but rather a deficiency in the market for information. It may be the case that the return is actually a fair return for the liquidity services rendered to the individual who sold the block. the “abnormal return” is defined relative to the risk-adjusted return which would be predicted by the capital asset pricing model. The quote taken from the special committee of the Securities and Exchange Commission is correct when it says that the efficient markets theory is silent as to the optimum amount of information required—that issue can be decided by better understanding of the supply of and demand for information.Chapter 11 Efficient Capital Markets: Evidence 1. However. The empirical evidence on block trading is consistent with the semi-strong form of the efficient markets hypothesis. as was the case of new issues studied in the Ibbotson (1975) paper. (b) The evidence on block trading indicates that it is highly unlikely that anyone can react fast enough to make a profit from the ticker tape announcement of a block trade. it is not possible to earn an abnormal return.

000 investors. the probability of wining every year is (1/2) ≅ .770 of them will have beat the market every year for ten years. A solicitation is mailed out to 270. Ninety thousand are told the market will move up by more than 10 points.” the first individual would be paid $20 per month out of his original $1. However.000977 10 If there are 10 million investors in the market.” the time until bankruptcy would be T= 1. his earnings would be 20(1 + Rf) T–1 + 20(1 + Rf) T–2 + · · · + 20 .126 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Thus. Brennan and Copeland [1987] model splits as costly signals to show that they can convey information in a rational expectations equilibrium. splits per se can be interpreted as good news about the future prospects of the firm. Ponzi frauds follow the adage “you can fool some of the people some of the time. a large new issue is also a large block. this “abnormal” return may be simply a liquidity premium charged for services provided by buyers of the stock to the owners of the ABC Company. It would operate as follows: If the first participant put in $1. 6. it is possible to earn large risk-adjusted abnormal returns if you buy a new issue at the offering price and sell at the end-of-month market price.” It goes without saying that the “success” of the solicitation in predicting the market for three months straight has absolutely nothing to do with their ability to forecast the next month’s index. and immediately reinvested all dividend payments in the risk-free asset. then over a ten year interval.” The plan described has paid out 2 percent per month for 18 months. If X were the amount extorted by the founders of the Ponzi fund. (d) As indicated by the Ibbotson (1975) study. as with block trades. by the end of the third month there remains a group of 10. then approximately 9.000 − X ≤ 50 20 If the investor participated in the Ponzi Fund. If he purchased it. Consider the following scheme. And finally. This would take 50 months if the perpetrators of the scheme kept their portion of the capital equal to zero. his first dividend would be paid out of the money he put in. The investor’s loss can be determined as follows: If he had originally invested his $1. After all. If no one else joined the “fund. Mr. The same thing can be said of individual investors who point to their past success in the market as evidence that they are clairvoyant.000. and the remainder are told it will fall by more than 10 points. ninety thousnd are told it will stay within a 10 point range. A should not purchase the service. If there is a 50–50 chance of “winning” in the market during a year’s time.000. he could be the victim of fraud.000 in the risk-free asset. until the money was gone and the fund was declared officially bankrupt. Fourth Edition (c) Empirical evidence of announcement abnormal returns for a pure split sample of stocks that had not dividends indicates a significant positive return (Grinblatt.000 investors who have received three successive correct “predictions. After the first month the group of ninety thousand which received the correct “forecast” is split into three groups. and $20 per month was paid as “dividends. 5. he would have $1.000(1 + Rf) T where T = time until declaration of the Ponzi Fund’s bankruptcy. Masulis and Titman [1984]).

then one would expect it to have higher returns also. then they should know that last year’s performance is typically unrelated to this year’s performance. represents his last dividend. 7. any differential effects of taxation are unaccounted for. 1971. An individual investor who could accurately predict the time until bankruptcy. 9. In the aggregate. on average. . net of taxes and transactions costs). The mystery is why. Either way. It is not possible to discern whether or not the portfolios’ actual performances are consistent with the efficient securities market until we test for abnormal performance based on net cash flows (i.Chapter 11 Efficient Capital Markets: Evidence 127 where the first term represents the earnings from his first dividend paid. and the last term. (a) A selection bias discredits any results implied by this study. it could go on indefinitely. by definition. Mutual funds do. the fund’s investors lose money because no investments in real assets are being made. then. The major problem is that the portfolios’ performances in Figure Q11. Firms continuously listed on Compustat tapes from 1953 to 1973 are firms that.000(1 + Rf) + T ∑ t =1 T 20(1 + Rf) < 0 T–t However. if enough new people joined the Ponzi Fund over time. at any point in time their choice of a mutual fund should be a matter of indifference. have not failed. When the scheme could not attract new participants.8 have not been adjusted for risk. What should be reported is risk-adjusted abnormal rates of return. Fourth. the chart does not say whether or not the returns recorded there include dividends as well as capital gains. If the information is predictive. But it isn’t. If portfolio 1 is riskier than the other portfolios. the risk involved is his ability to predict the fraud’s demise. April. Any model can be worked over to fit a set of data if enough variables are allowed. may be able to profit from the investment if the dividends paid to him (after his initial investment was recovered) exceeded the amount he would have received had be initially invested in a legitimate asset of equal risk. Therefore. Thus. In this instance.. 8. capital gains plus dividends. it would quickly go bankrupt. by Niederhoffer. The relevant test would be in the model’s predictive ability. (d) The evidence is not creditable because the model has not been tested on independent date. of course.e. in fact. 20 = 20(1 + T–T Rf) . (c) The results of this research provide evidence against capital market efficiency because past. If investors are rational. their behavior is not inconsistent with efficient capital markets. T. A study of this nature. It would be surprising if their preferences can be used to predict rate of return differences. attract more customers in a year which follows abnormal performance. publicly available data provides predictive power for future price changes. A second problem is that returns are reported without subtracting the transactions costs involved in readjusting the portfolios each week. his total loss would be –$1. (b) It is not surprising that stock brokers’ preferences are correlated with rate of return differences across industries. since the records of bankrupt firms are completely deleted from the tape. It simply doesn’t make any difference. The flows of dollars which they provide to mutual funds will not affect the way in which security prices adjust to new information. is reported in The Journal of Business. Third. brokers can beat the market.

D is the face value of debt issued by the firm at time 0 and C is a penalty imposed on the manager if V1 < D. the time 0 values of the two types of firms would be given by V0a = and V0b = V1b 150 = = 136. investors perceive the firm to be bad. the minimum cost of false signalling should be 100. . investors perceive the firm to be good and if D ≤ D*. γ1 = 0. since the marginal payoff from telling the truth is greater than that from a lie.2)(250) = 80 if D ≤ D *  The manager of a good firm would have the incentive to issue the correct signal (choose a debt level higher than D*) as long as her compensation from signaling correctly is greater than her compensation based on an false signal.2. Therefore.27 1 + r 1. Assume that D* is the maximum amount of debt a bad firm can carry without going into bankruptcy.2.2)(150) + (0. her compensation would be (0. good firms have a value. In this question.2C < 60 or if C > 100. The compensation of the manager of a bad firm is given by (0.1 V1a 250 = = 227. at time 1.2)(250) = 100 if D* < D ≤ 250 A M (D) =   (0. If there is no uncertainty in the market and pricing is risk neutral. she will give the correct signal. Further assume that if D > D*.1 where: r = the risk-free interest rate = 10% The manager’s compensation schedule is given by V1 if V1 > D  M = (1 + r) γ 0 V0 + γ1   V1 − C if V1 ≤ D  where γ0 = 0.2)(150) = 60 if D ≤ D *  The manager of a bad firm will have the incentive to signal correctly if 80 – 0. In this case. that is 100 > 80.2)(150 − C) = 80 − 0. greater than bad firms with value Vb = 150.Chapter 12 Information Asymmetry and Agency Theory 1. V1 is the value of the firm at time 1.2)(250) + (0. Based on the signal (debt level) chosen by the manager of a good firm.2)(150) + (0. Va = 250.2C if D* < D ≤ 250  B M (D) =   (0.2)(250) + (0.36 1 + r 1.

8412 0 . V(D) schedules are D*(t) = At and V[D*(t)] = (τP + β A)D*(t) Where A is given by:  τ   1 + r   τP   1 + r  A = – P   +     β  1 + 2 r   β  1 + 2 r  1+ β(1 + 2r ) 2 τP (1 + r )2 Where τP is the personal tax rate.25) 2 (1.2  A = –  1.2   0.8  . τP = 25%.5  1.4   1+ (0.25) 2 (1. D*(t).7494))(374. Therefore. In this case.5)(1.05 If the tax rate is 30% (0.25  1.  0.45 and V[D*(t)] = (0. The equilibrium dividend. and value response.4)  0.70 and V[D*(t)] = (0.7    1.25 + (0.4     0. V(D) schedules are D*(t) = (0.2) 2  0.25  1.7347       (0.7869 (0. r is the after tax rate of interest.4   0. V(D) schedules are D*(t) = (0.5)(0.3  1.Chapter 12 Information Asymmetry and Agency Theory 129 2.4  1 + (0.2   0. D*(t).7)(0.5)(0. and value response.4)  0.84 If the shortfall cost is 70%.25) 2 (1.59 and V[D*(t)] = (0.3) 2 (1.4  + A = – 1+ = 0.8412)(500) = 420.4   0.8412))(420. and value response. V(D) schedules are D*(t) = (0.70) = 290.7494 (0.2  A = –     +     1+ = 0. D*(t).4  +  0.5)(1. β = 50% and t = 500.59) = 282.7869))(393.3  1.7494)(500) = 374.8   0.23 If the discount rate is 40% (0.8)  0.25  1.2  A = – +  0.4) 2  0. (0.25  1.45) = 272.2) 2 = 0.7)(1.3 + (0.5  1.25 + (0. r = 20%.7869)(500) = 393.4) = 0.25  1. 5       The equilibrium dividend. β is the shortfall penalty and t is the upper bound on the cash flows. and value response.5  1.5  1.25  1.5)(1.5  1.2) 2 The equilibrium dividend.7    1. D*(t).4  The equilibrium dividend.2   0.

Fourth Edition The equilibrium dividend. Therefore. the value of the probability q is updated to either 0 or p.7347)(500) = 367. medium types receiving bH with probability p and bad types receiving bH with probability q (q < p < 1).33 and V[D*(t)] = (0. For example. Equity issuance makes sense for a bad firm since the equity issue helps it avoid the costs of financial distress. the debt is overpriced by I – [qI + (1 – q) bL] = (1 – q) (1 – bL). Under these circumstances. that is. consider a situation in which the bad firm issues straight debt. convertible debt that matures at time 2 but can be called to force conversion at time 1 at a predetermined conversion ratio and equity. On the other hand. Since the bad firm raises an amount I from the debt issue. the medium firm bears the cost of underpricing if it issues equity but gets no offsetting gain. The firm has three financing options at time 0—straight debt that matures at time 2. Since this conversion value always exceeds the call price K. Each firm receives a cash flow from the investment of either bL or bH at time 2 (bL < I < bH). Each firm has access to a project with required investment at time 0 of I and expected net present value of B. the expected cost of financial distress is greater than the underpricing. 1 and 2) with three types of firms (good. To see this. V(D) schedules are D*(t) = (0.25 + (0.33) = 226. If it does so. that is. the amount required for investment has to be raised from external sources and the firm is completely owned by its manager prior to the infusion of capital. Now consider the situation in which a medium firm issues straight debt.79 3. a call price K. it does not pay for a medium firm to mimic a good firm. The expected costs of financial distress are (1 – q)c. it does not pay for a bad firm to mimic a good firm. The expected costs of financial distress are (1 – p)c. the firm can force conversion with certainty and is never left with any debt. this implies that (1 – p) c > (1 – p) (I – bL). The costs of financial distress are such that c > (I – bL). Firm types are private information at time 0 and the true value of bad firms is volatile between time 0 and time 1.5)(0. Therefore. the firm type is revealed and for bad firms. the debt is overpriced by I – [pI + (1 – p) bL] = (1 – p) (I – bL). all agents are risk-neutral. Since it has been assumed that c > (1 – bL). the firms should not want to − p)bL mimic each other. Finally. there is a separating equilibrium I in which good firms issued debt with face value I. The discount rate is assumed to be zero. If it does so. this implies that (1 – q)c > (1 – q) (1 – bL). D*(t). Since the medium firm raises an amount I from the debt issue. At time 1. bL < K < I and is I convertible to pbH + (1 of equity. it issues debt with a true value of qI + (1 – q) bL. medium and bad). The probability of deterioration is assumed to be z and for consistency q = (1—z)p. This follows from the fact that the medium firm cannot reduce the costs of financial distress by issuing equity since it perceives these costs to be zero. bad firms issue a fraction qbH + (1 − q)bL of equity and medium firms issue convertible bonds with face value F > bL. and value response. Since it has been assumed that c > (I – bL).130 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Debt financing is associated with a potential for costly financial distress where a deadweight cost of c is imposed on the owner-manager. it issues debt with a true value of pI + (1 – p)bL. Firms differ in the ex ante probability of receiving bH with good types receiving bH with certainty. consider the situation in which a medium firm chooses to issue equity. note than the medium firm knows that the conversion value of the convertible bond will always be I at time 1. . For this to be a separating equilibrium. Stein considers a three-date model (time 0. the expected cost of financial distress is greater than the underpricing.7347))(367.

λ is a constant. c is the contract between the principal and agent. V is the agent’s utility function.p)] where U is the principal’s utility function. p ∂U is a performance measure. In this question both utility functions are of the form –αie α . U' = ∂ W –W/ i ∂V and V' = ∂ .p)] =λ V' [c(s.Chapter 12 Information Asymmetry and Agency Theory 131 4. The optimal sharing rule is: U' [s − c(s. i = agent. principal and αi W p and V = e− WA /. s is the final outcome.

is i’s risk tolerance. A . U = e Substituting these expressions in the optimal sharing rule yields: ’ − Wp/. Therefore. where WP = s – c and WA = c.

V is the agent’s utility function. The optimal sharing rule is this case is: U' [s − c(s)] f (s | a) =λ+µ a ' V [c(s)] f(s | a) where U is the principal’s utility function. s is the final outcome. V(c) = [1/(1 – γ)](δ0 + δ1c) In this case. p is a performance measure. a is the agent’s action. λ is ∂f(s | a) a constant. V (c) = δ1/(δ0 + δ1c)γ and the sharing rule can be written as: ’ (1–γ) and fa (s | a) is linear in s. ’ e e − (s −c)/α p − c/α A α  α  = λ or e  A s + c  =λ   α   αAαp  where α = αA + αP. the principal is risk neutral. f( s | a ) is the density function for the outcome conditional on the effort. f(s | a) 1 fa (s | a) γ   (δ0 + δ1c) = λ + µ f(s | a)  . Taking natural logs of both sides and some simple algebra yields: α c=β+  A  α  s    ln   α α where β =  A P  α 5. c is the contract between the principal and agent. W ∂W In this question. fa( s | a ) = ∂a ' ∂U ∂V = U' = ∂ V .

1  Solving the above equation for the compensation function yields: γ δ  c(s) = –  0  + (δ1)  δ1  1   −1  f (s | a)  γ  = λ + µ a  f(s | a)   1 Therefore. the compensation function is a convex function of s. if 0 < γ < 1. . It is linear if γ is one and concave if γ is larger than 1.

Given a tenure of T and an opportunity cost of i. where r* is given by: r* = i 1 − (1 + i)− T The following table provides the relation between the tenure of the manager and the hurdle rate employed for an opportunity cost of 25 percent (=i).44 37. Fourth Edition 6.01 25.91 25. the manger will use a hurdle rate of r* for project accept/reject decisions.29 25 . Manager Tenure (T years) 2 5 10 15 20 40 Project Hurdle Rate (r* percent) 69.18 28.132 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.

(Note that I0 = $10. therefore the economic profit approach is isomorphic to the DCF approach (before or after taxes). 2. Let’s take a look at the data provided in Table S13.) Table S13.2. Assume that the cash flows of the firm are a constant perpetuity. not $1.000 0 — — — $11.000 as found in early printings and the corresponding sale price is $11. (where It = new plus replacement investment) this reduces to V0 = E(EBITt ) WACCt The economic profit approach definition of value is ( I t −1 is beginning-of-year invested capital).Chapter 13 The Role of the CFO.2 Data for Q13. In this case the DCF definition of the entity value of the firm is V0 = E(EBITt − dept − I t ) WACCt and since a perpetuity has It = dept. and Incentive Design 1.000 400 500 500 500 — (b) Cut Costs 10% $10. V0 = V0 = E(ROICt − WACCt )ICt −1 + IC t −1 WACCt E(ROICt )IC t −1 (WACCt )ICt −1 E(EBITt ) − + ICt −1 = WACC t WACC t WACC t Note the last two terms cancel.000 WACC I0 ∆I0 ∆CF1 ∆CF2 ∆CF3 Sale Price .000 $0 $300 $300 $300 — (a) Bottleneck 10% $10.2 Base Case 10% $10.000.000. and that E(ROICt) ICt −1 = E(EBITt).000 0 350 350 350 — (c) Sell for 10%Premium 10% $10. Performance Measurement.

therefore the NPV is NPV(cost cutting) = $50 $50 $50 + + = 45 + 41 + 38 = 124 > 0 1.000 (remember that the basic value of $10. rises to only 3.000 as after tax proceeds). Its net present value is NPV(debottlenecking) = −400 + 200 200 200 + + 1.12 1.1 1.1 NPV (sale) = $11.1 300 = 3.12 1.000 .134 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.13 Again.13 = −400 + 182 + 165 + 150 = 97 > 0 Traditional NPV analysis would favor accepting this project even though the business unit ROIC is only $300/$10.000 . traditional NPV would cause us to accept the project although the ROIC. The incremental investment is zero and incremental cash flows increase by $50 per year.000 = 3% in the base case and $500/$10. (c) If the sale takes place immediately and is not taxed (or if we interpret $11.1 1.000 is a sunk cost). Fourth Edition (a) The first alternative is debottlenecking. Note that this is a lower business unit ROIC but higher NPV than debottlenecking. It will require an incremental investment of $400 and will bring in additional cash flows of $200 for three years.8% after the debottlenecking.1 350 NPV (cost cutting perpetuity) = = 3. By way of contrast.400 = 4.600 . if the cash flows of the other two alternatives were perpetuities their NPVs would be NPV (debottlenecking perpetuity) = −400 + NPV (base case perpetuity) = 500 = 4.5%. then this is by far the best alternative with NPV of $11. originally at 3%. (b) The second alternative is cost cutting.500 .

0 million $9. the stock price would fall even more because expected earnings would fall from $13. Table S13.0% — 200 2 1.5 million..2% 7. its stock price will fall because the expected total EBIT will fall from $13.0% 1421 Revenue − Operating Costs = Earnings before interest and taxes − interest on debt (5%) = Earnings before taxes – taxes (50%) = Net income Invested Capital Return on invested capital (after tax) Return on new capital (after tax) Debt 4. Let’s assume that ROIC is after tax.e.5 to $10.5 million $12.7% 6.6% 862 5 1. Project B should be taken. In this alternative the shareholders would be presumed to earn the opportunity cost of capital. Any new investment should be accepted when its expected return in greater than the cost of capital. Table S13.8% 620 4 1.5.8% 3.5 Announced Capital $5 million $30 million $5 million $30 million Announced ROIC 30% 40% 30% 40% Revised ROIC 30% 30% 30% 10% Expected EBIT Announced Revised $1.3% 400 3 1.063 548 –71 477 –238 239 3221 8.611 –1.3% per year. but omit expectations.5 million.9% 1128 6 1.5 million Project A Project B Total Project A No Project B Total WACC 10% 10% 10% 10% If the company makes public its revised estimate of the return on Project B. New investment that earns less than the cost of capital destroys value. Therefore.0 million 3. EPS (earnings per share) and the growth in EPS are uncorrelated with the total return to shareholders because they ignore balance sheet information and because they fail to capture the changes in expectations that drive TRS. invested capital and the cost of capital. Table S13. net income grows at an average of 3. . and Incentive Design 135 3.210 –755 455 –31 425 –212 213 2420 9. Performance Measurement.5 million 12. If so.5 to $4.000 –600 400 –10 390 –195 195 2000 10.5 million 1. i.464 –951 513 –56 457 –228 229 2928 8. yet ROIC declines every year because incremental return on new capital is less than the cost of capital (assumed to be 10%). The facts are summarized in Table S13.0 million $13.Chapter 13 The Role of the CFO.5 million $10. Consequently. EVA® and the change in EVA® contain information about earnings. The alternative is to announce that Project B will not be taken.5% 6.100 –675 425 –20 405 –202 203 2200 9. they also have a very low correlation with TRS. 5.3 ROIC falls while the Top and Bottom Lines both Grow 1 1. 10%.0 million $13.5 million 1.5 million $1.3 provides an example where revenue growth averages 8.4% 6.4%.5 million $4.331 –839 492 –43 449 –224 225 2662 9.

136 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. For tax policy see Chapter 14.6 below. Fourth Edition 6. Fixed salary provides a floor and non-salary incentives all provide gains that increase with performance.6 Tax considerations are not discussed in the answer to this problem. Figure S13. . The structure of all three incentives is roughly the same. However. Panel C is the only one that stipulated performance is based on expectations and that will be correlated with TRS. the graphed relationships break down if the measure of performance is not highly correlated with the stock price. The 3 alternative types of compensation are illustrated in Figure S13.

6 because the forecast has assumed that dividends are 65% of net income. and that (other than providing an interest tax shield) no shareholder value results from issuing debt. The value of equity is then derived by subtracting the market value of debt from the enterprise value.736/17. Although the current amount of excess cash and marketable securities is added to the DCF of operating cash flows. it is assumed that (assuming trivial transactions costs) the firm receives $1 of cash for each $1 of debt that it issues. called unscheduled debt.Chapter 14 Valuation and Tax Policy 1.9% of current assets and by 2010 it is 7. Simply. is combined with long term debt in Table 14.165 = 45. (b) There are two ways to bring excess cash into line. This entity approach assumes that the act of issuing debt (or not issuing it) is independent of the forecast of operating free cash flows. Free cash flow to equity include after-tax interest income and expense.7).830 = 6. This stream of free cash flow to equity is then discounted at the cost of equity. This does not make sense until we realize that it does not affect the DCF value of the firm and it is merely a plugged number (whose counterpart. the equity approach is also appropriate for valuing insurance companies. and they earn a positive spread that creates value for shareholders. and increases in deposit liabilities. One is to forecast growing dividend payments. (a) Banks are different because of their franchise to provide demand deposits and certificates of deposit. However. For this reason we recommend an equity cash flow approach for valuing banks. that has already been done in Table 14. These liabilities provide (insured) liquidity service to customers of the bank who demand less than their opportunity cost. 2.1%. Consequently these liabilities are an important part of the operations of the bank. (a) In 2003 excess cash is 472/6. future amounts are not because they are assumed to have zero NPV. Therefore.8 and subtract an amount from excess cash each year and add the same amount to “decrease in common and treasury stock. (b) Insurance companies also create shareholder value on the liabilities side of the balance sheet because their insurance reserves (the present value of their expected insurance benefits) yield investment rates of return below the opportunity cost to policyholders. (c) Neither dividend payments nor share repurchases affect the value of the firm (unless there is a signaling effect) because both are ways of delivering value—not creating value. The second approach is to go to financing flows in Table 14.” In this case you are assuming that the excess cash is used in a share repurchase program. . The standard industrial valuation model (see Chapter 14) discounts free cash flows at a weighted average cost of capital to value the enterprise.

837 = 45.553.33)] + 10   .33) − .8 in 2000 to 18.000) + .1%/year 34. K = g/r therefore g = rK and g = .80 10.33(6. 2) Second the finite supernormal growth model assumes ROIC = . The WACC assumptions are the same.3b 2000 $128. Table S14.33)(.000 + 11.3a provides the calculations of the multiples and Table S14. Note that the NOPLAT growth rate is 12.4% assumptions during 2000–2010 with those of the perpetuity (years 11 and on). 4. (b) The spreadsheet value estimate used EBIT (1 − T) = $2.35 for N = 10 years.000(1 − .265  = 31.680—a small difference.083)   . . then declines to equal WACC (on new capital invested). we get an estimate of continuing value as of year 10 V0 = 4.1% for the next 10 years then g = 5% thereafter.52(1 − .716 + 2.52 (1 – .3a Multiple Calculations Entity Value EBIT Multiple Table S14.91 3.083(1.52 (1 – . What creates the difference are the following: 1) First.15) = 5.8 Assumptions Perpetuity Assumptions g r 5%/year 30% 2010 $194. The spreadsheet assumed r = 30% forever.33)  .9 billion and the finite supernormal growth formula has V0 = $45.15[4.716 + 2.44 18. (a) Equation 14. not large enough to explain why the spreadsheet indicates that V0 = $128.138 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. These assumptions are consistent with a multiple that declines from 34.020    .21) Substituting in the suggested parameters.0845  .1% during the explicit forecast.090  = 31. Fourth Edition 3.487 in 2000 whereas the formula above assumes EBIT (1 − T) = $2.083  + .33) = .5 billion.3b compares the forecast Table S14.000 + 4.2% whereas the spreadsheet assumes g = 12.000(1 − .7 Averages for 2000–2010 NOPLAT Growth ROIC 12.7 in 2010.21 is the finite supernormal growth model: V0 = EBIT1 (1 − Tc ) r − WACC   + Tc B + K[EBIT1 (1 − Tc )]N   ku  WACC(1 + WACC)  (14. The ROIC also comes down—but it is assumed to remain well above the cost of capital. but more than halves from then to infinity.71 34.

167) 182. Clearly.8 = = = 9.02 .15[1.040(1 − . 14.12 − .15/. the new retention rate will be K = g/r K = . 10. becomes $4.12  .21.12 forever) and the result of the finite supernormal growth model (that assumes r = 1. in this case.083(1.33(6.12(1.062/15 = 41. EBIT.4)(1 − . .2% and assuming that K remains at 15%. 5.3% after tax. ROIC rises from 35% to r = g/k = .12)  = 9.12  + .0845  .4%.15[4.939. While possible for some companies to have negative working capital (e.083  4.4)]4   .00 for 4 years then r = WACC thereafter. This is an increase of $3.6)(.02 If the growth in NOPLAT increases from 10% to 15%. Tc = 0.386 million or 7.083)  = 32.15/. let’s perform the valuation as of year 10 and use the valuedriver model V10 = EBIT11 (1 − Tc )(1 − g/r) WACC − g Note: NOPLAT = EBIT(1 − Tc ).1)(1 − . that ROIC goes up to achieve the higher growth that K does not change. so let’s use the finite supernormal growth model assuming that N = 4 years.1)(1 − . the firm cannot grow at 15% forever.000 = 1.834(.667(1 − Tc ).040(1 − .g.413 − .33)]10   .15 = 100% and the new valuation becomes infeasible because WACC – g = . magazine publishers who receive payment before delivering the goods) it is more likely.906 = 48.033 + 2.138. we have V0 =  .138 .10 and r = .169 + 4.322 = 13.Chapter 14 Valuation and Tax Policy 139 (c) To simplify. Because we have NOPLAT in the 10th year. let’s assume that an extra revenue growth of 1% translates into 1% EBIT growth.2% to 6.1)(1 − . Growth in NOPLAT rises from 5.040.000 + 14.4)  1.667(1.000) + . If so r = g/k r = . Substituting these into Eq.12 – .12 = 1.10 / .667(1 + .667(1. and ROIC stays at 12%.10 .12) 1.00 − .4 = 1.33) + . Thus.15 < 0 in the denominator. Note that it is not appropriate to compare the result of the value driver formula (that assumes g = .25 = 125% This implies that the company can grow with negative capital requirements. V10 = V10 = EBIT11 (1 − Tc ) r − WACC   + K[EBIT11 (1 − Tc )]N   WACC  WACC(1 + WACC)  1. N stays at 10 years..

3. and with the same assumptions. the total cash flows paid by the firm to shareholders. VL = EBIT (1 − T) (1 − τc ) τc k d D Xτc rp FP + + ρ kb kp where: ρ = the cost of equity for an all-equity firm k b = the cost of debt k P = the cost of preferred This equation is equal to VL = VU + τc B + Xτc P .Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 1. preferred dividends and taxes k d = the coupon rate on debt D = the face value of debt rp = the preferred dividend rate FP = the face value of preferred stock X = the percent of preferred dividends which may be expensed EBIT (1 − T) = the cash flows from operations The right-hand side of the above equation may be rearranged as follows: = EBIT (1 − T) (1 − τc ) − k d D + τc k d D + k d D − Xrp FP + Xτc rp FP − rp FP + Xrp FP + rp FP = EBIT (1 − T) (1 − τc ) + τc k d D + Xτc rp FP Each of these cash flow streams is discounted at the appropriate risk-adjusted rate in order to obtain the value of the levered firm. debtholders and holders of preferred stock are: NI + k d D + rp FP = (EBIT (1 − T) − k d D − Xrp FP) (1 − τc ) − (1 − X) rp FP + k d D + rp FP where: NI = cash flows after interest. (a) Following an approach similar to the Modigliani-Miller derivation of equation 15.

5 .06]2. k0.06 + [.18) in order to obtain the 100 percent equity cost of capital for Carternut.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 141 where: VU = the value of an unlevered firm = B = the market value of debt = kdD kb EBIT(1 − T) (1 − τc ) ρ kp (b) Given the above result for the value of the levered firm. The object is to find the appropriate weighted average cost of capital for the project assuming that it has 50 percent debt. use the MM definition of the cost of equity (equation 15. its unlevered β and its 100 percent equity cost of capital will be the same as that of the project. 2. the weighted average cost of capital. may be written directly as B P   − Xτc k 0 = ρ  1 − τc  B+P+S B + P + S  P = the market value of preferred = rp FP Note that if 100 percent of the preferred dividends may be expensed.06 .30 = ρ + (ρ − .5 5. the cost of equity is given by the CAPM.4 + .18 − . For Carternut (assuming it has risk-free debt).30) 9ρ = + ρ − .06) (1 − . B S 1 . since 9 9(. X = 1 and the third term becomes τc If X = 0. it is equal to zero.0 = 30% P P+B+S Next.74% 19 k s = ρ + (ρ − k b ) (1 − τc ) B 1 = B + S 10 .06 = ρ = 28.5) . k s = R f + [E(R m ) − R f ]β = . Assume that Carternut has the same operating risk as the project. If so.

5 (. The treasurer’s reason for accepting the project is highly suspect.09 = = . Fourth Edition Next. Therefore. the new WACC is WACCτ = . accept the project since it earns 25 percent while the opportunity cost of capital. the cost of equity if the firm were all equity would be B   WACC = ρ  1 − τc B + S   (15.5) .000 of debt to the firm’s capital structure.1125 1 − .) 3.13).000 Therefore. in order to solve for the weighted average cost of capital if the project has 50 percent debt.0825 Because the project earns more than the new WACC. (Note: There are at least three other ways to obtain the same answer.000 = = .5)) = 21.55 percent. B   WACCτ = ρ  1 − τc B + S   = .667)) = .09 . assuming the MM theorems apply. if a more reasonable debt capacity were assumed for the project.000. the WACC is 9 percent.55% Therefore. .8 B 1. We need to know the appropriate weighted average cost of capital for the project.142 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.2874 (1 − .4(.500. However. Before the project. use the MM definition of the weighted average cost of capital (equation 15.667 B + S 1.4(. it should be accepted. No project has 100 percent debt capacity. then it is reasonable to expect the project to have a lower WACC than the firm. is only 21. and if the project has the same operating risk as the firm but higher debt capacity. a better approach is to calculate the new weighted average cost of capital for the firm after the addition of $500.13) ρ= The new capital structure is .1125(1 − . WACCτ. One should be immediately suspect that 100 percent debt financing is unrealistic. Consequently.

24) If the corporate tax rate decreases.1. the gain from leverage is  (1 − τc )(1 − τPS )  G = B 1 −  (1 − τPB )   The gain is zero if the following condition is met: (1 − τc) (1 − τPS) = (1 − τPB) Both changes proposed by Congress affect the tax rate on income from common stock. there will be no optimal capital structure from the firm’s point of view. then there will be no effect on the equilibrium condition and no effect on optimal capital structure from the point of view of an individual firm. τc 5. there can be no systematic change in optimal capital structure from the point of view of a single firm. If the reduction in the personal tax on dividend income exactly offsets the increase in capital gains tax for every individual. then in equilibrium the gain from leverage will be zero. corporations will repurchase debt and the aggregate amount of debt will decrease. Even if the tax effects are not exactly offsetting. . the gain from leverage will decrease. the gain from leverage is  (1 − τc )(1 − τPS )  G = B 1 −  (1 − τPB )   (15.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 143 4.1 Changes in the Aggregate Amount of Debt Given a Decrease in the Corporate Tax Rate. (a) Using the Miller (1977) argument. ceteris paribus. Hence. Therefore. both before and after the change in the corporate tax rate. Therefore. (b) Given the assumptions of Miller (1977). Equilibrium will be such that the gain from leverage will be zero both before and after the change. (a) According to Miller (1977). as shown in Figure S15. Figure S15. there is no optimal capital structure is either case. τPS.

or offset. The result would be to * reduce the aggregate amount of debt in equilibrium from B to B1.2 Supply and Demand Curves for Aggregate Debt If the effective tax on income from shares decreases. assuming a market in equilibrium.2 shows the supply and demand curves for aggregate debt. WACC.06 (.12 because B is an all-equity firm. an increase in * aggregate debt from B to B2. Hence.06) (1) = .19:  B   S  + ks  WACC = (1 − τc )k b     B + S  B + S ks can be found by the equation for the security market line. if the effective tax on income from shares increases. is the more likely effect of the proposed tax changes because the capital gains tax can be deferred.06 + (.144 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.5) = . is given by equation 15. then we might expect the aggregate demand for debt to decline because equity capital is relatively more attractive.15 (1 − τc) kb = Rf (1 − τc) = . (a) The cost of capital.12 − .06 + (. This. in fact. Of course.12 − . we would observe the opposite result. Fourth Edition (b) Figure S15. 6.03 .06)1. one might expect the reduction in the dividend tax to be overriding.5 = .12 WACC (B) = . Figure S15. where ks = Rf + (E(Rm) – Rf)β For firm B: ks = . For firm C: ks = . namely.

E (R m )) var (R m ) rjm σ jσ m σ2 m (. The income stream should be capitalized at the rate of the weighted average cost of capital for firm C.5) = 180 (.12) = .03) + .09. B/S. transform the target debt-equity ratio. the equation for the value of a perpetuity can be used. βu j = cov (E (R j ).0144) = = . rmj.5 (.000 (b) Since we know each project’s correlation between its unlevered free cash flows and the market.09 = 1.6)(.5 B+S Then. The value of firm B can be computed using equation 15. u we can calculate the β appropriate to the project and use this to determine the project’s cost of equity and WACC.1: VU = where E (EBIT1 (1 − τc )) = 180 (1 − . to B/(B + S): B =1 S B=S B = .5 (.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 145 To find the WACC. Of course. WACC (C) = .10)(.5) = 90 Therefore. VL = = E (EBIT1 ) (1 − τc ) WACC 90 . For project 1.15) = .5 (. since E (EBIT1 (1 – τc)) is a perpetual stream. the WACC is the required rate of return on the project.12 E (EBIT1 )(1 − τc ) ρ = 750 For firm C. VU = 90 .

taxes a leveraged firm will always have a lower required rate of return on projects than an unleveraged firm.09).75 = .03(.5)  B + S  B + S = .03 The expected return for project 1 is E(R j ) = = E (EBITj (1 − τc )) cost j 9 = .12 − . the unlevered firm. . but no personal.2.75 Using the CAPM to compute the cost of the levered equity in firm C.5 [1 + (1 − .105 The required rate of return for debt financing on project 1 is Rf (1 − τc ) = . In order to compute the cost of equity given that firm C uses 50% debt financing. for firm B (all-equity). is higher than the required return. and 4 are summarized in Tables S15.06 + (.09 This is also the WACC for firm B.06). and the project is profitable.0675 Therefore the expected return. This is because the differential tax structure in effect subsidizes debt. In a world with corporate. We know that B u  βL e = β 1 + (1 − τc ) S    = . Similar calculations for projects 2.06 + [. we need to L estimate the levered βe for the project then use the CAPM. Project 2 is rejected by firm B. the required return on project 1 is equal to the expected return (. There is a gain from leverage.5) + . and the project is marginally acceptable. but accepted by firm C with 50% leverage. For firm C.146 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.06] . we have L k s = R f + [E(R m ) − R f ] βe = .1 and S15.12 − .105(.5)1] = .5 = .09. 3.09 100 Therefore. the total required return on project 1 is  S   B  ks  + kb    = . . Fourth Edition The required rate of return given 100% equity financing of project 1 is Required E(Rj) = Rf + (E (Rm) – Rf)β j = .

0916 .03 .6416 .104 = .0985 . βL j . Project Evaluation for Firm B (All-Equity) Expected Rate of Return on the Project  Ñ j  − τc  =  cost j     .1950 Required Return on Debt Financing .1050 . it will have $1.15 Decision marginally acceptable reject accept reject Project Evaluation for Firm C B    LEVERED : B + S = .25 Required Return on Equity Financing .5 .12 Decision accept accept accept accept 7.13 1 − . It will still have $500.108 . we can compute the all-equity cost of capital. or b) that since it does not change the operating risk of the firm.1125 Expected Return: E (EBITj (1 − τc )) cost j .250. Therefore.2.1125 .2 2.000 in assets.0675 . which is relevant for valuing the project.5) If we had adopted the first assumption.4 percent.250.5    Project j 1 2 3 4 βL j .Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 147 Table S15. The second assumption is more reasonable.1125 .09 . B+S WACC .8 1.09 .000 in debt. is B 500. If the firm’s current WACC is 10. then this would be the required rate of return.03 .9624 1. We also assume that the MM model in a world with only corporate taxes is appropriate. There are two possible assumptions about the appropriate capital structure: a) that it cannot support any debt. Proceeding on the basis of the second assumption. that the project cannot support any debt.09 . the new debt to equity ratio. ρ. we have ρ= .5 Required Rate on Equity Financing Equals the WACC .1178 .03 .0810 .000 = = .1.03 Required Return Given B = .000 . we see that after the firm takes on the project.0739 . namely.75 .4 B + S 1. from equation 15.5. it can support a debt to total assets ratio of 50 percent. B   WACC = ρ  1 − τc  B + S  Substituting in the appropriate numbers and rearranging.13.4(.1320 .12 Project j 1 2 3 4 Table S15.0916 .

05] β f = 10% Next. Varhard’s recommendation is obviously wrong. Fourth Edition The new WACC can be computed from equation 15. Using the standard equation for the weighted average cost of capital.3). 9. Reject the project.09   + . The investment decisions of the firm remain unchanged while the financing decision increases the firm’s weighted average cost of capital.05 + [E(Rm) – Rf] (β f 1. β i = 1. =18% Therefore. When growth is zero.05] β f [E(Rm) – . (β f = the beta for the firm.18   = 15%  12   12  Next.3).4(. the expected rate of return on the project. is less than the required rate of 18 percent.13 (1 – .1092 8. using the CAPM for the project. we have an estimate of the return that is required: where β i = beta for the project E(Ri) = Rf + [E(Rm) – Rf] β i.05 + [E(Rm) – .3β f since [E(Rm) – . WACC = .05] β f = .148 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. assume away the effect of growth. = .15 = . the appropriate valuation model is VL = E (EBIT(1 − T)) (1 − τ c ) k D + τc d ρ kb since. The value of the firm will decline if there is any gain to leverage.05 + (. For a numerical answer. use the CAPM and solve for variables E(Rm) and β f. In a MM world with only corporate taxes VL = VU + τc B Reducing leverage reduces the value of the debt tax shield.10 VL = where E (EBIT(1 − T)) + τc B ρ E (EBIT (1 – T)) = E(EBIT (1 – T)) (1 – τc ) B= kdD kb . 17 percent.13.4)) = . = . we have (in a world without taxes) WACC = k b B S + ks B+ S B+ S  4  8  = . It is the rate of return which the project must earn in order to be acceptable.) E(R) = Rf + [E(Rm) – Rf] β f .10) (1.

823 . The new price-earnings ratio will be 37.08(500)  + .200 . 10(50) + 1.50 20 20 Therefore.000 23. the appropriate discount rate is the weighted average cost of capital.000 50.08(1.864 .1) + .08   ρ= 50 = 5% 1.0468 The net present value of the project is the sum of the NPV of the cash flows: Year 0 1 2 3 4 5 PVIF at 5% 1.5)  .800 31.9) = . the cost of equity for the unlevered firm.035 39.250 − 500 750 = = $37.000 .000 45.50 = 25 1. we can determine the value of the levered firm after the swap.11 (.000 Using this fact. and thus is in the same risk class.50 10.250. VL = 100(1 − . The correct weights of debt and equity to use in the WACC formula are from the target debt/equity ratio: B = . where n = number of shares outstanding n n 1. the price/share will fall from $50 to $37.0468 (.000 25.48) = .000)  +.9 B+S Then.5)  .5   = 1.0275) (1 – τc) = (.000 45.  S   B  + kb  WACC = k s     B + S  B + S = .000 NPV –500. the value of the firm has fallen to $1.000 = 100(1 − .907 .000 35. The per share price after the swap will be VL − B S = .08  Therefore.250 .50. Since the new branch is an extension of Community Bank’s business.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 149 Substituting in the facts of the problem.784 CF –500.952 . kb = (.09) (1 – . we can solve for ρ.5   ρ .880 37.0625 + .745 38.05312 ∼ 5% where kb is the total marginal cost of debt financing.05  .

800 + 31. If the firm is not levered the project will be accepted because it increases shareholders’ wealth (although marginally). which is identical to the value of the all-equity firm. The current IRS code allows individuals to deduct donations to not-for-profit organizations from their gross income before arriving at taxable income.745 + 38. From the option pricing model we know that the greater diversification created by the new investment will lower the rate of return variance of the firm’s assets. .660 Because the NPV > 0. If the decrease is large enough to offset the small positive NPV of the project. plant and equipment β al β a2 Liabilities debt βB equity βS Let Wai = the percent in asset i. this lowers the after-tax opportunity cost of capital to such individuals. shareholders’ equity may be thought of as a call option on the value of the firm. it will be rejected by shareholders of the levered firm.05 and discount five years to the present: 1. However. yet the project is 100% equity financed. The answer implicitly assumes that the firm does not change its leverage. WACC = Rf + [E (Rm) – Rf] (walβ al + Wa2β a2) Once taxes are introduced into our set of assumptions. and Wj = the percent in liability j.000 + 23. 12. Consequently. In a world without taxes. This assumption is valid if the bank intends to maintain its long-term capital structure at a target of 90% debt.000 (.000 The NPV of the branch is –500.880 + 37.000. This question focuses on the fact that the firm’s cost of capital is an opportunity cost appropriate for the riskiness of the firm’s assets.000.784) = 784.000 = 454. Assets liquid assets property. the systematic risk of the firm’s portfolio of assets must equal the systematic risk of its portfolio of liabilities. with a levered firm. even though a not-for-profit organization has no debt or equity. use the perpetuity formula.000 r . The project has positive NPV. the lower variance of return on the assets of the firm will decrease the value of the call. the correct answer is difficult to determine. the branch should be opened. Then we have Walβ al + Wa2β a2 = WBβ B + WSβ S The cost of capital can be determined using the CAPM for the assets side of the balance sheet.000 = = 1. CF 50. Fourth Edition For the years after 5.150 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.200 + 784. Presumably.035 + 39. 11.

3 and S15. shareholders benefit if the value of the underlying asset exceeds the exercise price. .e. with face-value debt = DA. Obviously.3 Probability Density Functions for Future Earnings of Firm A (Low Leverage) Figure 15. S15. Figures S15. i. Firm A in Figure S15. and hence is detrimental to shareholders. the probability that future earnings will cover the facevalue of debt (avoidance of bankruptcy) is higher for firm A than for firm B. Thus. Vt+1 > D An increase of variance for Firm A assigns more probability to the outcome Vt+1 < D.3 is a low-leverage firm. with exercise price equal to the face value of the debt..4 shows a highly leveraged firm. a 20 percent increase in variance is more apt to benefit the shareholders of a highly leveraged firm.4 Probability Density Functions for Future Earnings of Firm B (High Leverage) Regarding the stock as an option on Vt+1 . and is beneficial to shareholders.4 depict the high-variance and low-variance probability distributions of two firms’ future value. an increase in variance results in a higher probability that Vt+1 > D.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 151 13. Figure 15. For Firm B.

every individual in the economy will be indifferent because his wealth is unaffected.11 million.3 d 2 = 1.040 = .919) − 10e− (. the market value of equity is $17. during the year.54) = 22. There are.06) 1 (1. (a) We can use the Black-Scholes option pricing model and the facts of the problem to solve for the market value of the firm’s equity.11 .975 − 5. two implicit assumptions: 1) we are dealing in a world without taxes (or in a world where there is not tax effect to leverage) and 2) the firm makes no disbursements of capital.89 = $7.5 + . The market portfolio is composed of proportions of every asset held according to its market value weight. such as dividend payments. we have S = (25) (.06 + .06) (1) (.419 = . Any action which leaves the market value of the firm unchanged will not affect the value of the market portfolio. even if the value of equity increases at the expense of an equal and offsetting decrease in the value of debt.3 1 = .086 = 17. B=V−S = $25 − 17.5 + .89 million and the market value of debt is $7.4 − 1. however.3) 1 = . The BlackScholes formula applied to the pricing of corporate equity is S = VN(d1 ) − D e− rf T N (d2 ) where d1 = ln (V/D) + rf T σ Τ + (1/2) σ T d2 = d1 − σ T Substituting in the facts of the problem. we have N(d1) = . 15.3) 1 + (.4 1.91629 + .540 and substituting into the option pricing formula.919 N(d2) = .1 From the Table of Normal Areas.5)(1. every individual holds the market value fractions of the debt and equity of every firm. Fourth Edition 14. we have d1 = ln (25/10) + (.65 = 1.89 Therefore.152 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. If we live in a world where markets are perfect. then two-fund separation implies (among their things) that everyone holds the market portfolio of risky assets. Therefore. Therefore.

5) = 1. then N(d1) = 1 in the OPM.919)   (1.89   25  βB = (1 − . k s = R f + N (d1 ) (R v − R f ) V S .12 − .48 for the cost of equity capital. E (Rj) = Rf + [E (Rm) − Rf] β j For equity.93  17. the firm) is a virtual certainty.919)   (1. The CAPM applies to any risky asset.11  ∂S V βV .1758 and for debt we have kb = . the equity) will equal the value of the underlying asset (here.5) = .06 + [.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 153 (b) In order to determine the cost of debt and equity we can use the CAPM. If we assume an all-equity firm.42 and 15.12 − . because the probability that the value of the option (here. From equations 15.43) = . we have ks = . (a) True.49 we have the systematic risk of equity and debt respectively: βS = where ∂S = N(d1 ) ∂V ∂B = 1 − N(d1 ) ∂V Using the facts of the problem. The second term of the option pricing model drops out since the exercise price (here. we can use the CAPM to determine the cost of equity and debt.06] (1. ∂V S βΒ = ∂B V βV ∂V B And finally.0858 16. face value of the debt) is zero.06] (.93) = . Thus S = V N(d1 ) − D e − rf T N(d2 ) becomes S=V The OPM was used to obtain equation 15. we have  25  βS = (.43  7.06 + [.

02 . Given the Black-Scholes option pricing model.000) (.68 – 773.68 – (.5 + .06) (4) (.99434 N(d2) = .06) (4) (1. Thus structure is. k s = R f + (R v − R f ) = R f + (R v − R f ) = R v + (R v − R f ) V S B S + (R v − R f ) S S B S which is the Modigliani-Miller definition of the cost of equity capital in a world without taxes. this reduces to ks = Rv + (Rv − Rf) (0) ks = RV (b) If N(d1) = 1 in the OPM.49434 = .66437 = $1. Fourth Edition substituting N(d1) = 1. For an all-equity firm.99434) − e− (.24 + .48352 = .98352) S = 1988.000/1.154 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. 17. then S = V as shown in (a) above.000) + (.000)(.2 = 2. we have S = V N(d1 ) − e − rf T D N(d2 ) ∂S = 1.133 Using the Table of Normal Areas.2) 2 = 2. by definition. S = (2.5)(. N(d1) = .215. and the firm’s capital ∂V where d1 = ln (V/D) + rf T σ Τ + (1/ 2)σ Τ d2 = d1 − σ Τ Substituting in the parameters of the question. we have d1 = ln (2.533 .5 + .2) 4 d1 = .52) S = 1988.98352 Substituting these values into the call formula gives us the market value of equity.533 − (.4 d2 = 2.693147 + .786628) (983.2) 4 + (. 100 percent equity. since equity is a call option written on the value of the firm.

000) + (. the empirical work frequently uses cross-section regressions which are likely to have highly correlated residuals.85525 (. Changes in the value of the firm may be attributed to investment policy. or differences in risk? Finally. Consequently. In addition to this fundamental problem.75). 20.3) 4 + (1/ 2)(. If the firm is above its optimal amount of debt. For example. This weakens the power of the statistical significance tests. there are problems of measurement error. The main empirical problem is that firms typically change their capital structure at the same time that they change their portfolio of assets by taking on new investment.06) (4) (.5 + .895 S = (2.468 = .968) − e− (. how can one measure future growth in the firm.395 = .000) (. then the project should be financed with a higher percentage of equity.65 1215 19. or both.25525 N(d2) = .5 + .936 − 704 = $1.693147 + .3)(2) N(d1) = . at least to the point where the marginal gain from debt is offset by the marginal loss from potential bankruptcy.000 – 1215.06) (4) (1.02 = $784. then the value of the firm will increase by taking on more debt. these cash flows. the project should be financed with approximately 75 percent debt (since B/(B + S) . or expected rates of return. The new debt-equity ratio is 768 = . when discounted at the appropriate risk-adjusted rate.98 18.895) S = 1.3) 4 d1 = . As given in the problem. Therefore.62 1232 The old debt/equity ratio was 785 = . the market value of equity has increased and the market value of debt has decreased. (b) The earnings per share ratio has nothing to do with the economic value of the firm.3) 4 = 1. Only the aftertax operating cash flows provided by the project are relevant.85525 − (.3 = 1.000/1. will increase the value of the firm by $100 (NPV = $100).Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 155 The market value of debt is B=V–S = 2.24 + . d1 = ln (2. the effects are indistinguishable. Each of the three reasons given by the president may be criticized. financial policy. the project should be accepted. We can apply the Black-Scholes formula.232 Therefore. . If the firm is already at its optimal capital structure.968 d2 = 1.000) (. (a) If we assume that there really is a gain from leverage.

then equity should be issued to finance investments. The cash flows are operating cash flows and are.0315 = 14. The statement that “equity markets are currently depressed” has no meaning.1125 + (.06 + [. Knowledge of the risk-free rate. then we solve for the appropriate discount rate assuming it has the same leverage as the firm.1125 − .18 − .06] .156 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. First. he probably does not really believe that the equity of the firm is undervalued.06 + .25% The required rate of return on equity invested in the project assuming it has the same leverage as the firm is: k s = ρ + (ρ − R f )(1 − τc ) B S (15. If anyone knew that equities were undervalued in the market.4375 ρ = E (Rj) = Rf + [E (Rm) – Rf]β j = . D+E (a) Project 1.4)1 = .18) = .0 D ratio of 50 percent and a beta of . and the market beta allows us to determine the market price of risk (the slope of the security market line). then that individual would take advantage of his nonpublic information by placing his wealth in equity securities while prices were artificially low. Because Southwestern will finance the new project with 50 percent debt. therefore. the market equilibrium risk-free rate is also 6 percent.1125 + . E (R m ) − R f = slope = The firm has a 18% − 6% = 12% 1. the before-tax cost of debt is k b = Rf = 7.5 = 6% 125 Since corporate debt is assumed to be risk-free. 21.06) (1 − .5.0525 = 11. Unless the firm’s president has all of his wealth in the common stock of the firm.4% . Fourth Edition (c) If the project requires equity financing in order to bring the firm’s capital structure back to its target leverage ratio.4375 = . we find ρ for project 1. the expected return on the market. unlevered. βU j (the levered β ) is . It is important to recognize that the market equilibrium rate should not be adjusted by the tax rate.

5) = .75 = .06) (1 − .5) = .06 (1 − .4) (.15 + (.09) + .06 = 7.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 157 The correct discount rate on the project is its after-tax weighted average cost of capital.4% The correct discount rate is WACC = k b (1 − τc ) B S + ks B+ S B+ S B S = .8% .5 (.06 + .18 − .5) + .5) + .06 + [. k s = E (R j ) = ρ + (ρ − R f ) (1 − τc ) = . WACC = k b (1 − τc ) B S + ks B+ S B+ S = .12) (. which has an unlevered βU j = .204 (.5) ks = 12% Therefore.06) . WACC (joint project) = . ρ = Rf + [E (Rm) – Rf] β j = .06 + (.5) + .06 (1 − .09 = 15% Now find the required rate of return on equity if the project has the same leverage as the firm as a whole. ks = E (Rj) = Rf + [E (Rm) – Rf]β j = .06] (.12 (.054 = 20.5% (b) The security market line gives us the required rate of return on the equity of the firm. the required rate of return is simply their weighted average. the weighted average cost of capital for the levered firm is  B   S  WACC = k b (1 − τc )  + ks     B + S  B + S = .12) = 10.072 = 9% Repeat the same procedure for project 2.5) = .018 + .4) (.018 + .75.5) = .018 + .18 − .06 (1 − .5 (.15 + .06 + (.4) (.4) 1 = .102 = 12% Since the projects are of equal size.144 (.15 − .

105)−10  PV31 = 80   .50 Years 31–40  1 − (1.7412 .07 Add all of the present values to obtain the final NPV: NPV = −783.105   = 110 (8.20 = 40 (.69 783. do not accept the joint project. Interest expenses should not be included as part of the outflows because the discount rate already includes the interest factor. Fourth Edition (c) In order to determine the acceptability of the joint project.0500) = 24.015) = 481.0167) = . .814) = 969.105)− 4 41 Outflow 250 250 250 250 Factor .07 = −24.53 40.105   = 80 (6.5 percent.105)−26  = 110   .6707) = 650.18(.6707 NPV 226.30 167.28 PV0 = PV31 (1.50(.67 Note that you discount for four years because the annuity formula.9050 .75 185.38 18. This is the same as the end of year 4. which discounts payments at the end of each year from years 5 to 30 inclusive.105)−4 = 969.24 84.105)− Finally. will give you a present value which is at the beginning of year 5.28 + 24.05 16. discount the joint cash flows at 10.8190 . discount PV5 and PV31 to the present: PV0 = PV5 (1.18 Year 41 PV0 = F (1 + r)−n = 40 (1. you discount PV5 for four years. Year 1 2 3 4 Years 5–30  1 − (1 + r)− n  PV5 = 110   r    1 − (1.158 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.50(1.67 + 650. therefore. PV0 = 969.77 Therefore.105)−30 = 481.99 Inflow 10 20 25 60 NPV 9.20 + .99 + 84.25 204.

68 −80.90 (βL f ) + .9174 .45: B  βL = βU  1 + (1 − τc )  S  = . Project 1 Year 1 2 3 4 5–30 31–40 41 Cash Flow −120.09 −96. the new risk level of the firm’s equity is β = . (e) If the joint project is accepted.5 −95.7722 .0 −112.30 386.7084 (9.59375 2 2 We know by equation 12.5) + .095 = . The unlevered β of the joint project is 1 1 βU = (. project 1 would be accepted. if project 2 (the riskier project) can be considered separately.7118 .58 45.8929 .73 NPV of project 1 Therefore.8957) (.9 (.6 (1)) = .55 .10 (β p ) where βL p is the levered β of the joint project. since β ’s are additive.545 .0 20.6355 (7.08 −60. if it can be considered separately.97 7.4375) = .6355) (5. it should be rejected.0 40. the firm’s equity beta is the weighted average of the original β and the project’s β .0 PVIF .0754) (.45 + .5 −95.9290) (. Project 2 Year 1 2 3 4 5–30 31–40 41 Cash Flow −120.0 40.6502) (. The NPV of project 2 is obtained by discounting its cash flows at 12 percent.0 −112.0334) (.7084) (6.7972 .0 55.87 −67.19 −55.85 19.59375 (1 + .0 PVIF .4177) (.0 −115. the cash inflows and outflows of the joint project can be divided by 1/2.8417 .0 55.36 .0292) PV −110.80 −86.37 275. L β = .Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 159 (d) Discount project 1 at 9 percent.95) = .0 20.15 −91.0 −115.75) + (.10 (.0096) PV −107.95 Therefore.57 NPV of project 2 Therefore. Since both projects are of equal size.

5PA + .5PB PB − PA = . all that will happen is that a change in dividend policy will result in a change in clientele. the appropriate arbitrage condition would be that prices must adjust in a way which precludes any arbitrage profit from buying a security the day before it goes ex-dividend and selling it the day afterward.2D) + PA + τc (PB – PA) where π = the arbitrage profit PB = the price before the stock goes ex-dividend PA = the ex-dividend price τc = the corporate tax rate (for both capital gains and ordinary income) = 50 percent D = the dollar amount of the dividend payment In order for the arbitrage profit to be zero in equilibrium. we must have PB = . 3. but because improper use of dividend policy causes the firm to alter its investment decisions. . this will have no effect on the value of the firm.Chapter 16 Dividend Policy: Theory and Empirical Evidence 1. the value of the firm would be adversely affected. It does not depend on the risk aversion of shareholders.5 = 1. 2. If managers allow dividend policy to affect the investment decision. First. then the value of the firm will be affected.9 D . The potential arbitrage profit is π = −PB + D − τc (. Therefore. For example. As shown in Chapter 6. the market required rate of return depends on the market price of risk and the firm’s covariance risk. not because of dividend policy per se. If all market participants were corporations. Old shareholders will be replaced by new shareholders without a change in the value of the firm.8D This implies that the decline in price on the ex-dividend date must be 180 percent of the dividend payment. We are interested in a condition which might affect the value of the firm. at most. if a fixed dividend payout preempted the effort to invest in positive NPV projects. Even if a change in the firm’s dividend policy results in a change in the average risk aversion of its clientele.9D + . one must be careful to distinguish between the value of the firm and the value of equity.

If the price of common stock increases when a dividend increase is announced.500 It is possible to shelter any income.500 shares @ $10.000 16. Keough plans. by borrowing enough so that the interest payments reduce your taxable income to zero.500 shares @ $10 Insurance 25. or some other similar tax shelter.667 41.667 1.000 Capital Gains Sale of 2. The dividend per se has no effect no shareholders’ wealth. The relevant parameters on the ex-dividend data are: V = $1. 5.000 16.000 σ = .500 1.000 Taxable income 0 Nontaxable income 1.25 T = 4 years rf = 6% .167 Liabilities Loan 16.500 44.000 26.500 Less original basis 25. it is because the higher dividend payout is interpreted as an unambiguous message that future cash flows from investment are expected by management to be permanently higher.60 26.500 D = $1.667 44.667 Ordinary Income Dividends received $1.000 1.667 Net worth 25. Any portion of income which is consumed will also be taxed.17 (b) Once again we use the OPM to determine the market value of equity. not just dividend income. This has the effect of converting the income tax into a consumption tax.Chapter 16 Dividend Policy: Theory and Empirical Evidence 161 4. 6.500 shares @ $10. The example income statement and balance sheets used to support the Miller-Scholes argument are replicated below: Opening Assets 2. People do pay taxes.000 $1. but one might argue that they pay taxes only on that portion of income which they consume. (a) See the solution to Problem 15. Nontaxable income is provided by putting the borrowed funds into insurance.000 41. The problem is that the nontaxable income from these shelters remains tax free only until it is consumed.000 Less interest expense 1.167 Loan Accrued interest Net worth 16.667 Closing Assets 2.60 Accrued dividends Insurance Liabilities 26.

” The shareholder can reinvest his dividends in shares of the firm.4382 = .00) (1 + .24 + . EBIT1(1 − T) = EBIT0(1 − T) (1 + g) g = Kr = (.1) = $3.30 – 669.500/1.5) (.5 d 2 = d1 − σ T = 1. dividend decisions will not affect the stream of expected future investments.04 From the Table of Normal Areas.215. This is an increase of $23 over their original wealth position. The ex-dividend wealth of shareholders is $500 + $738 = $1. the new dividend policy is irrelevant.5 + . Whenever dividend payout is “too low” from the point of view of an individual shareholder. he can always sell a portion of his share holdings in order to consume.25 = 1.26 = 738.3508 = .20) = ..17.500)(.9382 N(d2) = . we have S = V N(d1 ) − De − rf T N (d2 ) = (1. the increased variability of dividend payout is irrelevant.000)e − (. The important fact is that the firm will continue to accept all profitable projects. It results from both the increased leverage and increased instantaneous variance after the dividend payout.25) 4 ln (1. After receiving $500 in dividends.238.8508 Substituting into the OPM.5)(. What counts is the riskiness of the anticipated cash flows from future investment and it will remain unchanged.06) (4) (. The opposite is true if dividends are “too high. equation 14.405465 + . 8. i.54 – (. Since the rate of return on investment is assumed to remain constant forever. 7. Fourth Edition The value of d1 will be d1 = = ln (V/D) + rf T σ T + (1/ 2)σ Τ + (.8508) = 1407. Finally. we can use the Gordon growth model.000) + (. the market value of equity is $738.5 + . to value the firm’s shares.54 .17b.25) (2) = 1. we determined that the market value of equity was $1. we have N(d1) = .06)(4) (. First we need the expected end-of-year EBIT1(1 − T).25) 4 = and the value of d2 is .1 EBIT1(1 − T) = (3.e.30 .162 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.04 From the solution to Problem 15.9382) − (1. Theoretically.

000 = = . .5 D + E 10. (ii) If the firm decides to issue $1.000 10. (c) The debt to equity ratio will increase on the ex-dividend date.000.000 6. the amount of the dividend payment.000 (d) Prior to the ex-dividend date the market value of the firm will be unchanged in a world without taxes.000 Predividend Ex-dividend D 5. P & E Total assets 2.05 9. we have V0 = EBIT1 (1 − T)(1 − K) D = 1 ρ − Kr ρ−g V0 = 3. (i) If the firm decides to pay a cash dividend (a) The systematic risk of its portfolio of assets will increase because a low risk asset (i. After the ex-dividend date it will fall by $2.000 6.00 .000 Debt Equity Total liabilities 6.000 8.000 3. the ex-dividend balance sheet will look like that below.Chapter 16 Dividend Policy: Theory and Empirical Evidence 163 Applying these results to the formula.000 of new debt and an equal amount of new equity in order to finance the dividend payment.000 Debt Equity Total liabilities 5. (b) If new debt is not subordinate to old debt.625 D + E 8.65 = = $33.000 = = .000 4.15 − .000 10..000 8.000 of cash) will be “spun off” to shareholders. then the market value of the outstanding bonds will decline because their claim on the assets of the firm must be shared with new bondholders.000 In anticipation of these changes the various impacts will be: (a) The systematic risk of the portfolio of assets will remain unchanged since the cash payment is raised from external funds. thereby leaving the assets side of the balance sheet unchanged. P & E Total Assets 0 2.000 D 5. the $2.000 2.10 .30(1 − . The ex-dividend balance sheet is shown below.5) 1. Cash Inventory P. (b) The market value of bondholders’ wealth will decline relative to the market value of equity because the debt-holders expect to have claim on fewer and riskier assets. Cash Inventory P.e.

VU since VL − Tc B = VU We also know that the value of the unlevered firm is VU = Therefore VL WACC EEBIT1 (1 − T) ρ= ρ Solving for V . However. Therefore. then pays out an equal amount in dividends. we have VL = VU + TcB We also know. (b) the wealth of original bondholders.E.000 is paid out the value of the firm returns to its original level.D.000 in new equity.6 the Modigliani-Miller valuation equation for a levered firm with no growth. from equation 15.6 D + E 10. Fourth Edition (c) The debt-equity ratio will obviously increase: D 6. From equation 15. (iv) Using cash to repurchase equity has the same effects (in a world without taxes) as paying a cash dividend to shareholders. (c) the debt to equity ratio. the exdividend balance sheet will be exactly the same as the pre-dividend balance sheet. the same answer used in i) applies here.000 = = . the weighted average cost of capital. we have ρ= ρ= VL WACC VL − TC B VL WACC . for ρ. when the $2. WACC Q. There will be no changes in (a) the systematic risk of the firm’s portfolio of assets. k0 = WACC. we have VL = E(EBIT1 )(1 − Tc ) .000 (d) Prior to the ex-dividend date the market value of the firm will increase by $2.164 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.000. 10. the value of the new debt and equity. or (d) the market value of the firm. (iii) If the firm issues $2. is B   k 0 = ρ  1 − Tc L  = WACC V   Solving.13. L E (EBIT1 )(1 − T) ρ .

the market value of debt decreases. ceteris paribus. Second. If the firm pays a cash dividend the market value of the firm goes down. if dividend payout carries no implications about cash flows from investment. This would leave them better off. . the ceteris paribus assumption is important. An increase in the retention rate does not necessarily affect the anticipated stream of investments at all.3) Retention is simply the difference between earnings and dividends. These results follow from the fact that the cash paid out is a relatively low risk asset.Chapter 16 Dividend Policy: Theory and Empirical Evidence 165 11. the greater risk of the firm’s portfolio of assets implies that the market value of equity increases if we think of equity as an option on the value of the firm. and the riskiness of the firm’s assets increases. the diminished asset base implies that bondholders have claim on less collateral in the event of bankruptcy. then we should observe the OPM implication—namely. we know that EBIT + mP = I + D (16. the market value of debt should fall when dividends are paid from cash.e. but it also increases the systematic risk of the firm’s portfolio of assets. This may lead to improper use of funds in projects with low rates of return.. If bondholders believe that higher dividend payout also implies greater future cash flows from investment. (EBIT – D). If the firm does not make use of external funding (i. ceteris paribus. The firm can retain any percentage of earnings it desires without affecting planned investment if it balances the sources and uses of funds by either selling new shares (if extra funds are needed for investment) or repurchasing shares (if excess funds remain after investment and dividends). From the point of view of bondholders. The option pricing model leads to the conclusion that. Not only does the cash payout diminish the assets which the firm holds. this is an undesirable effect for two reasons: First. a decline in the market value of debt. If the market value of equity increases. Referring back to the sources and uses of funds equation (for an all-equity firm). then a higher retention rate will imply more investment. However. However. mP = 0). 12. the market value of their collateral may actually increase in spite of the dividend payout.

15 (1 − . the formula must be modified as follows: NPV (to lessee) = I 0 − L (1 − τc ) − ∑ −∑ N N −1 L t (1 − τc ) t t =1 [1 + (1 − τ c ) k b ] τc dep t t t =1 [1 + (1 − τ c )k b ] Note that the information about the firm’s capital structure (50% debt to total assets) is irrelevant because leasing is a perfect substitute for debt.400 (3..3) PVIFa (4 yrs. Substituting the facts of the problem into the above equation. 7%) = 100.3872) − 6.1002) = −22.400 − 22.000) PVIFa (5 yrs.3) − 32.4) = 9% .. we have: NPV (to lessee) = 100. Equation 17. 2..000 − 32.874 The negative net present value of the lease contract clearly indicates that debt financing (i.3 (20. 7%) − .000 (4.4 shows the NPV of lease financing from the lessee’s point of view.000 (1 − .000 (1 − . owning) is preferred to leasing for this project. but it assumes that lease payments are made at the end of each year.e. If lease payments are made at the beginning of each year.Chapter 17 Applied Issues in Corporate Finance 1. (a) The NPV to the lessor is determined by using the following formula NPV (to lessor) = − I + ∑ L t (1 − τc ) + τcdep t (1 +WACCB )t t =1 N The weighted average cost of capital to the lessor is assumed to be the after-tax rate of return on the lease WACCB = kb (1 − τc ) WACCB = .000 − 22.

701 NPV (to lessee) = 5. The NPV of the lease for Reddi Roller Leasing is zero.000 – 147.577)(.889651) = −500.577 (3.577 t t = 1 (1 + . 15%)) = 500.4) + .. If the lease payments were larger.000) (3.9%)) = −500.000 + (147. we have NPV (to lessee) = −500. .09)t t =1 = −500.000 + 500. then the lessor would have a positive NPV.000 NPV (to lessor) = 0 Thus.546.000 + (88.000 + ∑ 147. 000) (1 + .000) ∑ 5 (PVIFa (5yrs.299 (c) It is usually the case that if the lessor is charging a competitive rate (i.577 (PVIFa (5 yrs. the lessor is charging a competitive lease rate.2 + 40.000 – 147. NPV (to lessee) = I − ∑ L t (1 − τc ) + τc dep t t t = 1 [1 + (1 − τ c ) k b ] N Substituting in the numbers from the problem.e. (b) The formula for the NPV of the lease to the lessee is rewritten below.577(1 − .352155) = 500. if he earns his weighted average cost of capital) and if the lessor has a higher rate than the lessee (which is always true when the lessee is a tax-free institution) then leasing is better than borrowing from the lessee’s point of view.6) + 40.000 + 147.15) 5 = 500.000 – 494.4(100...Chapter 17 Applied Issues in Corporate Finance 167 Substituting the facts of the problem into the formula we have NPV (to lessor) = −500.

This is a tax shield received by the lessee.542 ∆Dt — 14. (a) To answer this problem. • The after-tax salvage value is assumed to be received by the lessee (for a price of $1).52L 0 (5) (6) (7) τc Year 0 — 1 0 2 0 3 .843 + . The amount of displaced debt is initially $100. If we add the $100.236 + . The actual interest tax shield which is displaced by leasing (and which is added to cash flows in column 9) is equal to column 6 times column 2. Nevertheless.4561 Obviously.7305L − .8547 0 L . but it is reduced each year because annual (end-of-year) amortization payments are $31.e.000.256.715 ∆k b Dt −1 — 17.52L) + .680 19. it is useful to construct a table similar to Table S17.000 0 (4) (1 − τc)Lt L L L .48 4 .. • The opportunity cost of displaced debt.0000 0 L .000 14. • (1 − τc)Lt.236 + .000 − L − . This is an unreasonable assumption because the salvage value usually has greater uncertainty than the leasing (debt) cash flows.256 16. i.744 69.000 30..52L . Columns 6 plus 7 always add to equal $31.064 49. hence there is an immediate payment (in year 0) but no payment at the end of year 5. and should be discounted at a higher rate. the after-tax lease payments (from column 4).8547L − .5337 8.600) (8.6244 15. a positive cash inflow.256.456 (8. Column 6 shows the portion of each amortization payment which is interest and column 7 shows the reduction of principal.516 22.843 + .741 8. Table S17.52L (7. Note that the lease payments are made at the beginning of each year.52L 0 .7305 0 . $2180. The cash flows in column 9 are all outflows. the result will be the maximum lease fee which the lessee can afford to offer in negotiations NPV (lessee) = 0 = 89. It is the sum of: • τc dep.000 − $1) and pays capital gains taxes immediately.857 – 3. with the exception of the after-tax salvage value which is an inflow. To keep things simple we assumed the lessee reports a capital gain of ($10.576 11.48 D t −1 — 100. and solve for L. the depreciation tax shield (column 2 times column 3). given the complex pattern of cash flows.e.52L) − .1.000 salvage value (using sum-of-years’ digits over 4 years) has a present value (at 17%) of $7393 which amounts to a tax shield of $3549.833 26.423 4.168 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. an explanation of the cash flow column (column 9) is in order.000 10. using a 17% discount rate.5337 (8.000 investment outlay. ($3549). i. the NPV for the lessee is NPV (lessee) = 100. Column 9. we used 17% for convenience.48 5 .187412L L = $28. Note that the problem discounts the salvage value as if it had the same risk as the other cash flows. The sum is ($8969).548 26. which is saved by leasing.969) If we set the NPV (lessee) equal to zero.715 (8) (9) (10) After-tax t Salvage Value CFt (1 + kb)− — L 1. Future depreciation of the $10. the after-tax salvage value ($7600) and the present value of the depreciation tax shield. row 5 is the sum of: the displaced interest tax shield.6244 (15.191 . Fourth Edition 3.000 20.969) .1 (1) (2) Cash Outflows From The Lessee’s Point of View (3) dep — 40.000 85.

$0.200 + .000 30.48 (3) dep — 40.400 + .336 L = $26.7126 (4.48 . 4.2 (1) Year 0 1 2 3 4 5 Cash Inflows From the Lessor’s Point of View (2) τc — .800 + .52L + . The weighted average cost of capital for the lessor is WACCB = kb (1 − τC) = .52L) + .842 The results show that the lessor requires at least $26. Note that no capital gains are received by the lessor when the asset is sold because the sale price.200 + .52L) + .9188 .52L 9. Table S17. are essentially the same. therefore there is room to bargain.7756 .17 (1 − .600 + .400 + . then our cost of equity may be computed from the following formula (see Chapter 15.52L .Chapter 17 Applied Issues in Corporate Finance 169 (b) Next.6547 t Cash flows for the lessor are the sum of: • τC dep.000 0 (4) (1 − τc)Lt .2 shows the cash flows for the lessor. B/S = 9.18) k s = ρ + (ρ − k b )(1 − τc ) B S The information in the problem indicates that your tax rate is 40%.52L 0 (6) (1 + WACCB)− 1.0000 .84% interest rate.0884 Therefore. All you need to estimate is the before-tax rate of return on equity.52L 19.52L 0 (5) CFt .9188(19.52L . the discount factors in column 6 are based on a 8.52L) Setting this equal to zero and solving for L. therefore the NPV to the lessor is NPV (lessor) = −100. we have NPV (lessor) = 0 = 2. we need to analyze the same contract from the lessor’s point of view.600 + . and the book value. column 2 times column 3. the depreciation tax shield received. and • (1– τC)Lt.2106L – 59.7756 (9.800 + .8442 (14. Eq. column 4. The weighted average cost of capital for the lessor is WACCB = kb (1 – τC) . If we assume the Modigliani-Miller framework for our analysis.191.48) = .52L 14. the after-tax lease receipts.000 to acquire the asset and receives the cash flows in column 6.48 . The lessor spends $100. and kb = 14%.842 per payment while the lessee is willing to offer up to $28.000 + .52L) + .7126 .52L .48 .000 10.000 20. Any lease fee between these limits will provide a positive NPV to both parties. the lessor’s minimum acceptable lease payment.52L . 15. Table S17.8442 . $1.48 .52L 4. ρ.

600 + 5. Eq.36%. we have ks = . .0936 = ρ (1 − . the NPV is negative and the project should be rejected.86% This is the required rate of return on equity for your firm.48(.48 (.14) (1 − . 15.4) 9 ks = 29. therefore WACCB = .9)) ρ= . 5.4789% 1 − .000 (1 − .207163) = −100.)] = −100.5) + .164789 + (.5) = 17% Substituting in the information from the problem we have NPV = −100.0936 = 16.15 (1 − .500) (1.170 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.000) [PVIFa (17%.17)t t =1 8 = −100.4 (12. compute the NPV of the project to Mortar Bored Company using the usual NPV formula NPV = −I +∑ t =1 N E (∆EBITt ) (1 − τc ) + τc ∆dept (1 + WACC) t The WACC is computed as follows WACC = k b (1 − τc ) B S + ks B+ S B+ S WACC = . (a) First.600) (4.164789 − .000 + (18. Using the Modigliani-Miller definition of the weighted average cost of capital (Chapter 15.000 + (23.25 (.000 + 99.289 NPV = −711 Therefore. 8 yrs.4) (.18 (1 – .12) we have B   WACC = ρ  1 − τc B+ S   .000 + ∑ 31. Fourth Edition The rate paid on the lease contract is 18% and the lessor’s tax rate is 48%.48) = 9. if the company buys the project and finances it with debt.4) + .9) Substituting this back into the cost of equity formula.

)] = 100.000 – (17. . (c) If the project is leased. 8yrs. the project is acceptable if leased.534819) = 100.5.Chapter 17 Applied Issues in Corporate Finance 171 (b) The NPV of the lease is computed by using Eq. NPV (lessee) = I − ∑ L t (1 − τc ) + τc dep t [1 + (1 − τc )k b ]t t =1 N Substituting in the data from the problem we have NPV (lessee) = 100.500) [1 − (1 − .840) (5.400(.15]t t=1 8 = 100.000 − ∑ 21. Hence. the positive NPV of the lease contract will offset the negative NPV of the project.000 – 98.000 − (21. 17.259.259 Consequently.6) + 5.400 (1 − .741 NPV (lessee) = 1. the NPV of the lease contract is $1.4 (12.000) [PVIFa (9%.4) + .4).

50) = $7. Terms Paid (A for T) 9.43% Target (T) $100 100 $1 10x $10 $1.7% $20 100% 0. In a weak stock market.000 16.50 (new) $6. Total Market Value (old) 7. Premium to T 10.4(1.57% Combined $200 140 $1.010 (b) The basic generalization is that when a buying company has a higher P/E ratio than the target.43 +43% 100($71.4A/1 $1. the P/E ratio of the buyer may drop considerably. the weighted average growth will drop and/or risk of the buyer will increase. So as more low P/E companies are bought.Chapter 18 Acquisitions. P/E Ratio 5.010) = 71. the buyer will not have this artificial growth stimulus any more. For the buyer as its EPS grows.860/1.000)−1 = 43% (7. The problem was aggravated at Tyco because as the company seemed to prosper. EPS (new) 11. The relative P/E ratio is the one reflected in the deal terms.150/10.010) = 28. Proportions (old) 8. Number of Shares 3. Restructuring.000)−1 = 186% (2.50) = $10. the buyer will achieve accretion in earning per share while the target will suffer dilution in earnings per share. the manipulative atmosphere of the activity may provide a fertile soil for such behavior. Proportions (new) 140($71.000 83. Total Market Values (new) 13. and Corporate Governance 1. EPS (old) 4. EPS (% change) 12. These practices. Also this works great in a booming stock market. (a) Solution to Table S18. Market Values (% change) 14.5 Model of Differential P/E Ratios Acquirer (A) $100 100 $1 50x $50 $5. Its EPS growth and P/E ratio are likely to collapse. its P/E may actually increase.150 (7. Divestitures.3% 0.4(100)($71. A second problem is that when low P/E ratios companies are not available.572 −43% 0. are not inherent in playing the differential P/E game.43) = $0.860 (2. Price/Share 6.000 1. Some executives have been charged with stealing from the company. This facilitates doing other acquisitions with its growing P/E ratio. The target may end up with a higher market value per share received if the buyer has a really high P/E ratio as in the example. top executives received loans which were later forgiven. of course.50) = $2.860/10. Net Income 2.43 50x $71. .150/5. The basic fallacy in this process is that the higher P/E ratio company has either higher growth and/or lower risk than the low P/E ratio company. However.

We can use the OPM to determine the market value of equity.49733 = .30 = $2. From problem 18. shareholders would oppose the merger if there were no other merger effects.179 d1 = . Therefore.000) (.9924) S = 3989.2.179 = 2.693147 + .02.49240 = . only slightly more) than the value of equity after merger.24 = + .02.179) 4 + (1/ 2)(.000 Substituting into the formula.179) 4 (.99733 N(d2) = . and Corporate Governance 173 2. This is more (albeit. this amounts to $2.918 .179) 4 .32 – 1561. V = 4.276 − (.428 Referring to the Table of Normal Areas.000 D = 2.428. Doubling it. we have d1 = ln (4.428.510825 + .000 and the market value of the firm to $5.000)(.358 d 2 = 2.000/2.7856 − (.5 + .000 in debt. the face value of debt in the merged firm increases to $3.179) 4 = 2. S = V N(d1 ) − D e − rf T N(d 2 ) where d1 = = ln(V/D) + rf T + (1/ 2)σ T σ T ln(5.02 Prior to merger the value of equity in each firm was $1.06)(4) (2.99733) − e− (.000.5)(.7856 . we have $2.99240 and substituting into the Black-Scholes formula yields S = (4. using the parameters for the merged firm.24 + .179 = 2. Restructuring. we have N(d1) = .179) 4 rf = .5 + .43 per share.000 / 3. By taking on an additional $1.430.06)(4) + (.179)(2) = 1. 3.04. We can apply the Black-Scholes option pricing formula.215. Divestitures.06) (4) (.000) + (.276 .000) + (. With 1.Chapter 18 Acquisitions.000 shares outstanding.358 d 2 = d1 − σ T = 2.06 T=4 σ = . the market value of equity in the merged firm was $2.

we have S = (5.9 Using the Table of Normal Areas.693 + .70709 .000  + (.353545 = 1.093 + . 4.353545 = + .25/1.943 – 2294.353545 .899545 = 1.25 Price per share = 2648.000)(.5 d1 = .9 − .75 = 2648. N (d1 ) = .971 d2 = d1 − σ T = 1.70709 = 1.70709 ln(V/D) + rf T σ T + (1/ 2)σ T = 1. with an exercise price equal to the face value of the debt outstanding. then the value of the stock is given by S = VN(d1 ) − De− rf T N(d2 ) where d1 = and d2 = d1 − σ T.5 + .236) (.08)(5) ln  500    + (1/ 2) .06)(4) (. (a) If we regard the stock as a call option on the firm.000 = $2.31623)(2.883 . Fourth Edition From the Table of Normal Areas.19291 N (d 2 ) = .4886 = .31623)(2.10 . Substituting into d1.9724) = 4.546 + . the price per share increases by 22 cents.08)(5) + (.9886) − (3. we have N (d1) = .236) .65 Therefore.5)(.5 + .4 1.174 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. we have  1.9886 N (d2) = .000)e− (.9724 Substituting these values into the OPM.4724 = .10 5 = = ln 2 + (.

000 (.333 B = 2.9735) = 1985.5 N (d1) = .25)(.332.000 – 1. and Corporate Governance 175 Therefore. first calculate the new variance of returns.333 = $667 .6703) = 971 – 296 = $675 The value of the debt.5) (. V=S+B 1000 = 675 + B B = 325 The combined equities of Firms A and B are worth $1.9926) − e− (.54 = 1.25 = 2. SA + SB = 675 + 675 = $1350 To calculate B and S for the merged firm.2236)(2.883) = 971 – 500 (.9926 d2 = 2.236) 1.883) (.2 – 652. B.093 + .236) (.25)(.936 N (d2) = .5 = 1.9735) (.1) + (.000 (. the value of the equity for the merged firm is S = 2. S = 1.6703) = 1985. for each firm is $325.66 = 1.2236 Using the Black-Scholes formula once more. setting the covariance term equal to zero.6703) = 1985. Restructuring.9735 Therefore.08) (5) 1. 2 2 σ 2 = w2 σ 2 A + (1 − w) σ B = (. Divestitures.971) – e− (. d1 = = ln 2 + (.186 + .08) (5) + (.Chapter 18 Acquisitions.436 − .1) = .2236)(2.2 – 1.350.08)(5) 500 (.2 – (973.05 σ = .6703) = 971 – (441.5) (.000 (.000 (.25 = 2.5)(.436 .

176 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.8524 S′ AB = 2.8524) = 1.6485 = + .000) /1.560) + (. Company AB should issue new debt with a face value of $560.000(. 013(1.000). ′ B′ AB (D AB /D AB ) − BAB = L For the facts of this case.560 ln(2.547 − .5 N (d1 ) = .013 Test of results: ′ B′ AB (DAB /DAB ) − BAB = L 1. 000 /1.5 + .047 N (d2 ) = .25 = 1.5 + .333 – 1.08)(5) + (.350 = −17 So the merger transfers wealth from the equityholders to bond-holders.236) .6703)(.9390) − 1.3524 = .000 − 987 = 1.013(1.2236)(2. We can increase the face value of the debt to D′ AB to reverse and neutralize the effect of the merger such that. .560) = 649.560) − 667 = −18 ≅ −17 Note that BA + BB = 325 + 325 = 650.236) (.4390 = . Let L = the loss to the equityholders as a result of the merger: L = SAB – (SA + SB) = 1.333 = $17 The bondholders gain $17: BA + BB = 2 (325) = 650 667 − 650 = $17 (b) In order to restore the equity holders to their original position.5 = 1. The funds raised with the new debt are used to retire a portion of the equity (so that the value of the firm remains at $2. This is equivalent to ′ B′ AB (D AB /D AB ) = 1.5)(.2236)(2.000 /1. Fourth Edition Shareholders lose $17: $1350 – 1.000 D′ AB = 1.547 . The resulting market values of debt and equity are calculated below: VAB = 2. the merged firm increases the face value of debt D′AB.878 − 891 = 987 B′ AB = 2.9390 d1 = d2 = 1.560(.

it would imply that conglomerate firms were able to buy under-valued securities. If the firms’ performance were better. this indicates that diversification as such was not the dominant motive for conglomerate mergers. unless this synergy is firm-to-firm specific. One hypothesis of superior conglomerate performance is that synergy exists between the components of combining firms. However. 8. which is equal to the market value of the new debt: B′ AB BA + BB Market value of net debt S′ AB SA + SB 1. This is inconsistent with efficient markets.Chapter 18 Acquisitions. Since many conglomerates engaged in more than 15 mergers and still had betas well above 1 (and well above non-merging firms). H = high . M = medium. because persistent evidence that conglomerate firms were making bad choices (or overpaying) would result in a reassessment of their market values. Non-owner managers of the firm may be motivated to act like bondholders. 7. The managers would prefer to prevent bankruptcy by decreasing the probability of default on debt. Restructuring. Alternative Strategies for Growth (a) Solution to Table S18. if there are costs to losing their jobs. one would not expect the performance of conglomerate firms to be significantly worse than a broad market index. until the returns and risk of the conglomerate firms plotted on the security market line.013 – 650 363 + 987 1. and Corporate Governance 177 The wealth of the equityholders is now composed of the current value of equity (987) plus the amount of cash they have received.350 Market value of debt in merged firm Original market value of debt Cash received by equityholders Market value of equity in merged firm = 2 × 675 = Original market value of equity 5. 6.8 Multiple Strategies for Growth* Merger H M H H H L M JV M H H M M H M Alliance M M M L M H L Licensing H L L L H H H Investment M H H L L M M Internal Speed L Cost known L Add capabilities L Add products L Add markets L Avoid antitrust H Clarity H * L = low. A high percentage of the firms initiating merger activity in the early 60’s were in narrow lines of business whose growth outlook was highly uncertain or unfavorable. Thus. a competitive market among acquiring firms would cause any advantages to be priced according to their value. Hence both their risk measures and return measures would be changed. Studies have shown that betas usually approach 1 if investors diversify randomly among 15 or more stocks. They presumably sought a carry-over of some technical or general managerial abilities with the firms that they acquired. Similarly. manager-agents may behave in a manner more risk-averse than stockholders would wish. Divestitures. Over a long period of time one would not expect the risk-adjusted performance of conglomerate firms to be significantly different from the risk-adjusted performance of a broad market index.

The relative contributions and limitations of six major alternatives are depicted. and markets. but may create competitors. mergers have the advantage of speed in adding capabilities.9 illustrates alternative approaches to asset restructuring. 10.10 Changes in Ownership Structure (Financial Engineering) Leveraged Recap H M H H H H M LBO or MBO H H H H H H H Dual-Class Recap M M L H L H L Share Repurchase L L L M H M L Proxy Contest L M L H L L M Infusion of new capital Achieve a turnaround Tax benefits Takeover defense Leverage is increased Maintain control Management incentives (b) In Table S18. 9. high returns. Equity carve-outs raise funds as an initial step toward a spin off. Changes in Ownership Structure (a) Table S18. and lead to joint ventures and mergers. investment requirements. Fourth Edition (b) Table S18. Investments in other companies may provide new knowledge. A proxy contest represents an effort to change the management control group in a company. Dual class recapitalizations enable management. Asset Restructuring (a) Solution to Table S18. Alliances provide opportunities for gaining knowledge about new areas but the relationships have greater ambiguity. often a founding group. Compared with internal growth. financial engineering and changes in financial structure are evaluated by a different set of criteria.178 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. . Licensing quickly adds revenues.8 depicts multiple strategies for growth. Joint ventures reduce antitrust problems. and also offset dilution from the exercise of stock options. Spin offs may facilitate performance measurements and strengthen incentives. Share repurchases provide flexibility in returning cash to shareholders in a tax advantaged form. Tracking stocks achieve similar results but give the parent continuing control over the operations. The criteria in the first column reflect the important potential benefits of each alternative growth strategy.10. management is compensated by additional shares of stock which increase their ownership percentage. product.9 Asset Restructuring Divestiture H M L L H Equity Carve-Out H M H H H Spin-Off L H H H H Tracking Stock L H H H M Raise funds Improve efficiency by focus Measure performance better Tie compensation to performance Parent focus on core business (b) Table S18. to trade higher income to other shareholders in return for a class of stock with higher voting power. and risk. A leveraged recap maintains control by a capital infusion to pay a large dividend to existing shareholders. The strength of divestitures is in raising funds and enabling the parent to focus on its core business.

18 .4)(10)   .21) = $100 + 0 = $100 Value of Alber: L VA = 12(1 − .45 + 72 (.4)(10)   .4)  .15 .12  + 1. Divestitures.28 = $106.2232) = $50 + 20.09  .18 − .09  + 1(15)(1 − .15 $15.Chapter 18 Acquisitions.30 $12.11)  = $65.09 = $70. (a) Assets ($ millions) WATRO Total assets ALBER Total assets SABEN Total assets (b) After-tax r Before-tax r = r/(1−Tc) EBIT = before-tax r (TA) Watro .5733) = $65.00 Alber .73 Value of Saben: VSL = 10(1 − .15 − .09)  (14.12)  = $50 + 90 (.09(1.4)(10)   .25 $10. and Corporate Governance 179 11.12  . Restructuring.4)  .00 Saben .45 + 41.11  .12(1.21) for a levered firm. the value of the levered firm is: VL = Value of Watro: L VW = EBIT1 (1 − Tc ) r − WACC   + K[EBIT1 (1 − Tc )]N   WACC  WACC(1 + WACC)  (15)(1 − .00 $40 $40 ____ $100 Claims on Assets ($ millions) Debt $50 Equity 50 Total claims $100 Debt $20 Equity 20 Total claims $40 Debt $20 Equity 20 Total claims $40 (c) Using a form of Equation (14.5(10)(1 − .11(1.09 .09 .11  + 1(12)(1 − .4)  .09 − .

12  .63 + 198 (.4)(10)   . Fourth Edition (d) Value of Watro-Alber combination: L VWA = 30(1 − .11(1.07 = $156.04 Value of Watro-Saben combination: L VWS = 23(1 − .41 = $311.12(1.4)  .12)  = $115 + 138 (.4)  .11)  = $163.2976) = $115 + 41.4)(10)   .16 − .12  + 1(23)(1 − .07 Value of Watro-Alber = Less value of Alber = Less initial value of Watro = Increase in value Value of Watro-Saben = Less value of Saben = Less initial value of Watro = Increase (decrease) in value 12.63 + 147.11  .7445) = $163.2009 − . (a) Accounting Balance Sheets (millions of dollars) Debt Equity Total Assets (b) After tax r Before tax r = r/(1 − Tc) = r/.11  + 1.1(30)(1 − .6 EBIT = (before tax r) (TA) 6% 10% $16 12% 20% $16 15% 25% $20 Jordan $80 80 160 Konrad $40 40 80 Loomis $40 40 80 $311 107 $204 100 $104 $156 70 $86 100 ($14) million million million million million million million million million million .180 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.

8 + 108) $144 million = (28.2) βJK = 28.8 = . Total shares 6.8 115. Divestitures.6 $4.60 115.4 $.9474 = 1.2 4. New EPS (in dollars per share) 8.20 6X $7.00 108 $20 −4 16 –6.8 + 115. as calculated in part (c).60 15X $54.8 $1.2 βJL = (1.4) + (1.632 1.80 12X $57. number of acquired firm shares (millions) 3.48 .72 1.24 28.20 47% (e) Because β ’s are additive.8 108 (1.20 28.2) 136.8 7.4 9.5X 2 15 4 19 12 $. The total market value of equity for the proposed merged firm would be JK: JL: $136.8 136.2 or Loomis 8X 2 16 4 20 14. Dilution    row 8    $16 −8 8 –3. number of new shares 4. and Corporate Governance 181 (c) EBIT Less Interest (10% × debt) Income before taxes Taxes @ 40% Net Income EPS (earnings/shares) (in dollars) P/E Price = EPS (P/E) (in dollars) Total market value of equity (S) (d) Jordan Merges With Konrad share price of K or L 1.8 million = (28.8 $16 −4 12 –4. number of old J shares 5. share price of J 2.Chapter 18 Acquisitions.2 $3.2 = 1. Restructuring. J’s original EPS (in dollars per share)   row 8 − row 7   100  9.2947 + .4) + (1. we simply weight the β of each firm according to the market value of its equity.28 + 1.5) 144 144 = .20 40% 7. Net income of J + net income of K or L 7.

4)  (.24) = 122 = 12.10 = 10% (g) Using the Modigliani-Miller valuation equation (equation 14. In this instance.11 − .48) = 134 = 13.3%  120  144 WACCJL = (.028 + .8 WACCJK = (.4)  (.10)(1 − .8 JL $120 144 $264 WACC = k b (1 − Tc ) (B/V) + k s (S/V)  120  136.182 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.027 + .21) we have V0L = where: EBIT1 (1 − Tc ) r − WACC   + Tc B + K(EBIT1 )(1 − Tc )N   ku  WACC(1 + WACC)  EBIT1 (1 − Tc) = EBIT0 (1 + g) (1 − Tc) = the after tax end-of-period cash flows from operations B   WACC = k u  1 − Tc = the weighted average cost of capital B+S   (1 + g) = (1 + Kr) = the growth rate of the firm K = the investment rate (sometimes called the retention ratio) B = the market value of debt k u = the cost of capital for an all equity firm Tc = the marginal corporate tax rate N = the number of years of supernormal growth where r > WACC r = the rate of return anticipated on new investment .4% (f) Use the proposed merged firms’ market values of debt and equity to determine the weights for the cost-of-capital formula.10)(1 − .8  256. Fourth Edition The required return for equity of the proposed mergers can be found from the security market line: E(R j ) = R f + (E (R m ) − R f )β j k s (JK) = .8 $256.134) +  264  264 = . (in millions of dollars) Debt (D = B) Equity (S) Value (V) JK $120 136.06 + (. the market value of the debt is assumed to be equal to its book value.073 = .065 = .06) (1.093 = 9.11 − .06 + (.8 = .2% k s (JL) = .06) (1.122) +  256.

91 If Jordan merges with Konrad.093)  = 206.11 108.00 28.91 120. and Corporate Governance 183 Note that: VL = EBIT1 (1 − Tc ) EBIT1 (1 − Tc ) = + Tc B WACC ku Using this fact and the parameters of the problem we have V(JK) = 32(1 − . Restructuring.093(1 + .4)(10)   .00 million For JL.16 − .91 million figures in millions of dollars Post-merger value (V) Less: amount of debt (B) Value of equity (S) Less: J’s pre-merger market value of equity Gain in equity value Cost if acquired at market (market value of equity for K or L) Gain in value (loss) JK $333.5 = $333. A merger with Loomis results in a gain of $10.00 154.2 JL $274.13 − .0 $76. Divestitures.10(1 + .91 million in value.4)(10)   .8 184. .45 + 126. Jordan could pay a premium to Konrad shareholders and still have a positive NPV.4)  . the indicated net increase in value is $76.91 = $274.20 $10. The merger with Konrad is preferred and it also meets the capital budgeting requirement of having a substantial positive net value. we have V(JL) = 36(1 − .8 126.093  .00 213.4)  .91 28.093  + (1)(32)(1 − .2 million.2 115.10  + (1)(36)(1 − .Chapter 18 Acquisitions.10  .10)  = 216 + 58.00 120.

381 0.381p) = $105.S.381 − 10 0.0 = 5% per annum R p.1a. Dollar.0 = $100 × 10p/$ = 1000p Fp / $.38p/$.0 = 10p/$ Fp/$.05 = $105.381p/$ R$.0 10. The 1000p invested in the Mexican security carrying a 9% rate has an end of year value of 1090p.0) = 1000p × 1.0 − Sp / $.0 Peso $100 × Sp/$. The premium on the dollar is matched by the differential interest rate paid on the peso denominated security.09 = 1090p in $: 1090p × (1/Fp/$. (a) Parity relationships are illustrated in Table S19.09 − 0. One hundred dollars invested in a dollar security gives an end of year return of $105. Alternatively.04 = 10 1.0 = 10.03810 .0) = 1090p × ($1/10.0 1 + R$. so IRP obtains.0) = $100 × 1. we could convert the $100 dollars into pesos at the spot rate of 10p/$.05 = 10 1.0 − R$.0 Sp / $.1a Example of Interest Rate Parity (A) Equilibrium Mexican Peso and U.05 0.0 = R p.03810 = 0.0 = 9% per annum Year 0 Dollar $100 Year 1 $100 × (1 + R$. Table S19.05 0. Fp/$. Simultaneously we enter into a forward contract in which pesos are exchanged for dollars at an exchange rate 10.381p/$ Sp/$. This amount is exchanged at the forward rate shown in Table S19. It is also useful to analyze the adjustment processes for departures from IRP when the foreign exchange rate is expressed as Mexican pesos (foreign currency) per dollars.0 = 10.1a to obtain $105.Chapter 19 International Financial Management 1 Interest Rate Parity for FC/$: In the text example of covered interest arbitrage the foreign exchange rate was expressed as dollars per euros.0 in pesos: 1000p × (1 + Rp.

05) = 11.03810 = 0.5p/$ = 1102. discounted at the U.S.0) = 1000p × ($1/10p) = $100 Year 1 −1000p × (1 + Rp. Dollar.905p Net peso position PV net peso position at R$.5p/$ Sp/$. Hence.01190 .0 Sp / $. Invest (long) dollars at R$.5p/(1.905p which is equal to the forward premium of the dollar less the premium of the interest rate on the peso security over the interest rate of the dollar security.0 = $105.5 − 10 0. Repay at Rp.0 10.0 d= = 0p 0p Fp / $.0 × Fp/$.Chapter 19 International Financial Management 185 (b) Case B.05 = 0. capital will flow from pesos to dollars. representing a gain on the roundtrip of 12. its present value is 11.04 = − 10 1. The forward premium on the dollar is increased to 5%.05 = $105.05 − 0. Table S19.0 = 10p/$ Fp/$.1b reflects the condition that the forward rate for the peso is higher than the IRP rate holding everything else the same.5p/$.5p 12.0 1 + R$.0 − Rp. Buy peso forward. The analysis will show that the return from investing in a dollar security will be higher than investing in a peso security. Alternatively.0 −1000p in $: Sell pesos spot. Fp/$.5 0.5p.09 − 0. 1000p × (1/Sp/$.0 = 10.0) = −1000p × 1.5p/$ R$.05 0.5p = 12.0) = $100 × 1.09 = −1090p $100 × (1 + Rp.5p. rate. Table S19. The $100 is invested in the dollar denominated security to yield $105 at the end of the period.0 in pesos: $105.0 = 5% per annum R p.0 − Sp / $. An investment is made in pesos to provide a comparison return.0 − R$. A forward contract has been entered into in which at the end of the year dollars are converted into pesos at the forward rate of 10.0 = 9% per annum Year 0 Borrow (short) pesos.0 = 10.0.05 − 10 1. the pesos borrowed are converted into dollars at the spot exchange rate to obtain $100.0 × 10.5p −1090p + 1102. This yields 1102.1b Example Covered Interest Rate Arbitrage (B) Mexican Peso to U.S.

8 − 10 0. Dollar to Mexican Peso. the dollar is at a discount in the forward market.05 − 10 1.0) = 1000p × 1.03810 = −0.0 1 + R$.0 − Sp / $.8p/$ R$.05810 . Repay at R$.0p/(1.0 = 9.0 − R$.05) = 58.225.1c.8p/$ Sp/$. we convert the $100 into pesos at the spot exchange rate of 10p/$. Invest (long) peso at Rp. the investor owes $105.0 = 10p/$ Fp/$.0p 61. Table S19. The investor can close his short position with a gain of $6.0 + $111. At the end of the year.2 0. Alternatively.225 × 9.1c Example of Covered Interest Arbitrage (C) U. 0 d= = 0p Fp / $.05 −0.095p which is equal to the algebraic difference between the forward discount of the dollar and the discount on the dollar interest rate.186 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. The amount of dollars received is $111.8p/$.8p/$ = 61.02 − 0. simultaneously entering a forward contract in which pesos will be converted into dollars at the current forward rate of 9.0) = 1090p × ($1/9.225 $6.S.S.04 = − 10 1.0 = 9% per annum Borrow (short) $100.0) = −$100 × 1.0 = $6.09 = 1090p 1090p × (1/Fp/$. This amount converted into pesos at the forward rate and discounted at the U.05 = −$105. The reference investment will be to sell a dollar security short. interest rate gives 58. Fp/$.225 = $6. so outflows take place from dollars to pesos.095p − Rp.0 =$100 × 10p/$ = 1000p in $: Net dollar position Net peso position $0 0p −$105.225 × Fp/$.09 − 0.05 = −0.0 = 5% per annum R p. Fourth Edition (c) In Case C. Year 0 −$100 Year 1 −$100 × (1 + R$.225.0 in pesos: 1000p × (1 + Rp.225 $100 × Sp/$. Sell peso forward.0 = 9.0.8p) = $111.0 9.0 PV net peso position at R$. Table S19.0 Buy peso spot.0 Sp / $.

1a) is plotted as point A in the first quadrant with coordinates of 0. One is that β < 1 is evidence of a risk premium on foreign exchange. Dollars 2. Thus currencies trading at a forward discount tend to outperform those trading at a forward premium. (b) Investment banks had sought to develop models forecasting foreign exchange rates based on movements from short run departures. 2003. Another explanation is that the bias is evidence of expectational errors.05 on the dollar premium axis and 0.02). the forward rate bias persists. unexpected major improvements in economic activity could raise real interest rates. but the sum of the expected change in the exchange rate plus a risk premium. on average. p. This bias could be exploited by taking long positions in currencies that trade at a forward discount and short currencies that trade at a forward premium. Hence. Point C calculated from the data in Table S19. investments in dollar assets become more risky. but long run movements toward parity relationships. tend to weaken (rise) less than what is implied by the forward discount (premium). in the long run. Specifically. January 24. FX Weekly.1c plots below the IRP line in the area of capital outflows with coordinates (−0. Point B calculated from the data in Table S19. in the country with higher interest rates.Chapter 19 International Financial Management 187 (d) In Figure S19. −0. fiscal shock. 4–7). or real economy shock.0381 on the interest differential axis. The Deutsche Bank explains the forward rate bias as resulting from investors risk aversion to short term losses that can be substantial (Deutsche Bank.1 Interest Rate Parity Relations Pesos. When the dollar interest rate rises. Note also that it plots above the IRP line in the area where capital inflows take place from the peso to the dollar.0381. this arbitrage opportunity will disappear as investors make such trades. If investors are risk averse and foreign exchange risk is not fully diversifiable. Figure S19. the interest differential or the forward rate discount is no longer a pure estimate of the expected change in future exchange rates.0381. Hence.1b plots as 0. The adjustment processes move toward decreasing the dollar discount in the forward market and/or decreasing the interest rate differential of a dollar security versus a peso security. . currencies trading at a forward discount (premium). the IRP (Table S19. the higher nominal rates could represent lower future inflation rates since the higher real rates deducted from the nominal rates would represent lower inflation rates. But because of risk aversion.1. An example of expectational errors could be a counter monetary. In theory. (a) Two possible explanations for the bias are common in the literature.

i. vs. 3. They represent strong forces toward which foreign exchange rates will move.40 B imports FC15 A imports $1.5 Table S19.40.. For A: We have exports of 2Y and imports of 1X from B. Other prices can be calculated the same way.00 domestic FC6 domestic A imports $2.20 B imports FC5 Hence country A would import X and export Y. (a) Exchange rate uncertainties may be reduced by offsetting economic developments within the larger area. So FC12/W = 2($1). forecasters are guided by the parity conditions as relationships toward which exchange rates will move. whether to go long in the forwardrate bias trade or to close the position and do nothing. 5. Product Y vs. vs. Country B would export X and import Y. Fourth Edition The Deutsche Bank investment strategy is to go long in the three highest-yielding (highest forward discount) currencies and going short in the three lowest-yielding (highest forward premium) currencies. the price is FC12 × ($1/FC5) = $2. $3. This strategy is combined with a daily optimized technical moving-average model that yields a trading rule on when to move in and out of the forward-rate bias trades.188 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. The fixed exchange relationships within the union avoid speculative attacks. 4. The parity conditions therefore represent a useful framework for guiding economic policies. (c) Benefits are greater than costs if the underlying economies are highly integrated which implies a high level in trade in goods. In addition. (b) A country loses monetary and fiscal independence.00 domestic FC12 domestic $1. Labor mobility between countries in the union is high. This is called going long in the forward-rate bias trades. . The parity conditions are equilibrium relationships. Results are summarized in Table S19. the price is $3 × (FC5/$1) = FC15. A country loses the use of the exchange rate as a mechanism to facilitate adjustments to shocks among the nations within the union. Equilibrium would require for B: For one unit of X exported B could import 6Z of Y. Transaction costs of exchanging one currency for another are avoided. 12 = 6Z Z = 2 = number of units of imports of Y. If country A imports X. If country B imports X. Monetary and fiscal integration are successfully integrated.e.5 Domestic and Import Prices When FC5 = $1 Product X vs. W = FC6/$1 so FC6 = $1.

At FC6 = $1. Also.000 for the auto. The change in exchange rates in response to shifts in relative prices helps readjust the export–import imbalances. The A currency has four times as much gold content. suppose that an auto sold in the United States for $2. and reduce sales to the U. 6. A sale at $2. imports and exports in dollars rise. Conversely for the foreign country.S. IF the surplus country was not functioning at full employment. results in higher relative prices. lower prices in FC for both their imports and exports would stimulate purchases from the U. The value of the dollar falls.000 when the exchange rate was $1 to FC4. The demand for FC relative to dollars increases. the wheat can be sold at FC8. the export surplus increases its income and employment.S.S. The dollar price for exports could be increased and still represent substantially lower prices in the foreign currency. 8. Conversely. the demand for U. 7. causes the FC to rise in value relative to the $.S.S. But to do so at the new exchange rates. imports rise and its exports fall. Country A runs an export balance surplus. to be more expensive in $ and to represent a smaller number of FC to exporters at previous $ prices in the U. To illustrate the operation of the adjustment process let us assume an initial relationship of $1 to 4 FC: $1 = FC4. the U. at the old exchange rate. Let us assume. U. sold wheat at $4 per bushel.S. receiving FC16. each country’s exports and imports are equal in total value measured in units of its own currency.Chapter 19 International Financial Management 189 We then have: A imports 1X = $2 B exports 1X = FC12 and and exports 2Y = $2. that the new relationship is now: $1 = FC2.S.. The export deficit decreases income and employment in the deficit country. imports 2Y = FC12. But the higher value of the FC causes imports into the U. Or 1B is worth 0.000. Income and employment effects may also enter into the adjustment process as well as price changes. for purposes of illustration. Country A is an attractive market in which to increase sales from other countries and A’s imports increase. In the U. The flows of gold operate through prices to function as an adjustment mechanism for international balances of trade and payments as well as to regulate the price change relationships between countries. For example. Hence gold flows into Country A while it flows out of Country B. the increased relative demand for FC in relation to the $.S. Domestic prices in Country A rise. At the new exchange rates. At higher $ prices in the U.. A’s export surplus will be reduced or reversed until equilibrium between relative price relationships of the countries is restored. A formulation first set forth in the mid-eighteenth century is the price-gold-flow mechanism. Note the adjustment process.25A. Hence 1A will be equal to 4B. in order to receive $4 per bushel. goods in the foreign country will rise. the prices of exports in the foreign currency have fallen. import purchases would have to be made at higher prices and export sales could be made under more favorable conditions than before. Conversely. in the foreign country.S. Conversely. Let us now assume that the volume of imports in the United States exceeds its exports in relationship to countries whose currency is the FC. he must receive $4. in the adjustment process. A higher rate of inflation in the U. the prices of the U. . At the new exchange rate the auto exporter still seeks to receive FC8.S. the prices in Country B fall. At the new exchange rate.000 provided the auto exporter with FC8. Thus the prices of imports in dollars would rise substantially at the new exchange rates. A’s goods are more expensive in other countries so its export sales decrease. Similarly.S. employment may decline in A and increase in B. the demand for foreign goods will decline.000. while Country B runs a deficit.

092 = 3 1. and the foreign country inflation rate = 5%.04)(1. X0 = 3/1 is the beginning exchange ratio. and Xf is the forward rate.144 1.276 Xf = 2.144 Xf = 3. $10.000 $5. So we solve for X1: X1 1. Plus 2. a) b) c) d) e) f) Plus 1.2% nominal interest rate for the foreign country. 1 + Rn = (1.86 per $1.8636 The expected future spot rate is FC2.000 $5.. and f represents foreign.15 X1 = 2.10 1. Plus 1.000 $1 million $3.000 $2. The prices of internationally traded goods carry most of the adjustment process. (b) The Fisher Relation is P 1+ r = 0 1 + R n P1 where r = the real interest rate and Rn is the nominal interest rate.S..000 $500 10.040)(1. Minus 1. Xf 1.4% nominal interest rate for the U. So we solve for Xf.000 $2.000 $500 Minus 2. Fourth Edition An argument for the use of flexible exchange rates is that trade imbalances caused by changed relative prices of domestic and foreign goods are adjusted through shifts in exchange rates.190 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Plus 2. Minus 3.05) = 1. For the U.092 = 9.8636 . Minus 1. We know that the U. (a) The PPP: X1 Pf1 / Pf 0 = X 0 Pd1 / Pd0 In this example.S.1X1 = 3.000 $1 million $3. Plus 2.10) = 1. Plus 2. X0 is the current exchange rate. $10. Minus 3a.05 = 3 1. (c) The Interest Rate Parity relation is: Xf 1 + R f 0 = X 0 1 + R d0 where Rf0 is the nominal interest rate in the foreign country.S. the subscript d represents the U.S. 1 + Rn = (1. The Fisher Relation may be written as 1 + Rn = (1 + r)(P1/P0) Solving for the foreign country. inflation rate = 10%.S. 9. Minus 1.144 = 14. and Rd0 is the nominal rate in the U.

000 pesos 1.000 × 10 = Amount repaid in pesos = $1. The left-hand side is IRP = 11.10) R f 0 = 15.Chapter 19 International Financial Management 191 (d) The estimates are the same.5% If the interest rate in Mexico were below 15.5% Amount received in pesos = $1. 1 + R d Pd1 = 1+ r Pd0 (2) dividing (1) by (2) (and assuming the real rate of interest is equal across the two countries).000 pesos = 15.5 = 1.000.10 10 1 + R f 0 = 1. and equilibrium.000 10.05(1.000 Interest rate = = Pesos received 10.000 $ 100. (a) Loan amount Interest @ 10% $1. which was an effective rate of 15.000.5%.000 × 10. it would have been cheaper to borrow in pesos than to borrow in dollars at 10%.5 pesos/$1X 0 = 10 pesos/$ and R d0 = 10% 1 + R f 0 10. we have Pf1 1 + R f Pf 0 = 1 + R d Pd1 Pd0 1 + R f Xf = 1 + Rd X0 P /P X The right-hand side is the PPP = f1 f 0 = 1 Pd1 /Pd 0 X 0 The fundamental relations assumed are certainty.550.000 (b) Interest Rate Parity Theorem 1 + R f 0 Xf = 1 + R d0 X d X f = 10. . equal real rates of interest. (e) The Fisher relation for country f can be written 1 + R f Pf1 = 1+ r Pf 0 (1) and for country d.550.000 pesos 11.000.100.5 = Effective interest paid in pesos Interest paid in pesos 1.000.550.5%.

0.0025 = $0.491128/FC) = FC 20.07/1. 2.  n  1 + Rf 0    365   n  1 + R d0    365  $0.49(1.291 . Fourth Edition 12.6 million divided by the 6 month forward rate of FC per dollar (including commission costs): 4.000/($0.08/2 = 2.778. It buys FCs at the current forward rate for delivery in 180 days.0648379 FC/$ (d) The dollar price is FC 4. 13.49 (1 + 0.09/2  1.492356/1.09/2) E0 = Ef = $0.361.227.192 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.491128/FC (d) $10. (b) Use Interest Rate Parity relation  n  1 + Rf 0  Xf  365   = n X0   1 + R d0   365   Solving for the forward rate Xf 1 + 0. It buys dollars at the current forward rate for delivery in 180 days.08    1.08(0.07 FC/$ (c) FC are more expensive by the commission. so commission decreases the dollar value of the FC.9952) Xf = 2.000. (a) It goes long on dollar futures.004807692) E f = $0. so the net result is fewer FC/$.0648379 FC/$) = $2. (b) By IRP.040  = 2.000 FC/(2.08 1 + 0.08/2) = Ef (1 + 0.0025 = 2.600. (a) The company goes long on FC futures. to pay for the machinery at that time.045  = 2.492356/FC (c) Dollars are more expensive.

000 + 80. this means Xf (the forward exchange rate in FC/$) is lower (its value higher) than the expected future spot rate of FC/$.90/$) = $200. they could have covered their short position.90 = 191.000 X(1. If the speculator receives a risk premium for going long in FC.000 in 3 months.000 Substituting the correct numbers.000 which it uses to pay off the principal plus interest on the FC 384. Divide by the current exchange rate to convert this amount to dollars.000 Xf − 152.4) + − = 194. then: X[1 + Rn (n/365)(1 − τ)] = FC 380. Globalcorp would receive FC 380. The speculator would go long in FC.000 – 194. FC 364.4)] = FC 380. to determine the final net payment in dollars.000 $200. .939[1 + (0.683 loan. (a) First calculate the amount borrowed now in FC currency that will yield FC 380. X[1 + (0.000 380.4) 380.242 Xf X0 Xf 380.Chapter 19 International Financial Management 193 14.000 Xf = FC 2.758 This amount is equivalent to an “insurance premium” that Globalcorp is paying to limit the magnitude of unexpected foreign exchange rate losses during the next three months.242 = $5.28/4) (1 − 0. at the U.042) = 380.242 X f 114.242 Xf   X 0 X f    amount received tax shelter on loss F0 F0 F + (τUS ) − 0 (τUS ) = 194.683/1.000(0. Globalcorp realizes $194.242 X f = 228.S. Globalcorp receives FC 380.00/$ (d) Globalcorp would have sold FC forward short.4)] = 191. use the relationship: F0  F0 F0   +  −  τUS  = 194.242 Xf 1.000/(FC1. nominal interest rate.242 At the end of the three months. If the amount = X.939 Now compound the dollar amount for three months.012) = $194.000(0. 191. (b) If the payment were made immediately. When Globalcorp was paid in FC.90 Xf 380.683 This amount plus interest equals FC 380.000 X = 364.000 in three months.939(1.242.08/4)(1 – 0.000 = 194. (c) To determine the equivalent (non-arbitrage) forward rate.

i. (a) Let X be the amount of FC which.00/$) = $190.227 net cost (b) If payment were made immediately. is $192.5)(380.. .094 is invested in the foreign country today. including opportunity loss.4)] = FC 380. If the speculator receives a premium for his short position then Ef > E1.g.094/(FC 2.4) = $192. (d) The speculator would short the FC forward.505.000 = $2.5042/FC Transcorp would go long in FC forward to hedge their position.12/2)(1 − .000)(0.04)](185.000 E f = 268.093. E1 = 0.4)(380. (c) E f F0 + (E f − E 0 ) F0 (τUS ) = $192.227 380. when invested.e.000 X(1.194 Copeland/Shastri/Weston • Financial Theory and Corporate Policy..036)(185. it will yield FC 380.000 in 6 months.000 in six months.5042.000 = 192.227 380. Then.024) = FC 380.547) = (1. E.000/(FC 2.227 E f = $0.000) E f − (0.000 E f − 76. will equal FC 380.000 The difference between this amount and their actual cost. This represents an insurance premium against a rise in the FC rate while Transcorp delays six months in making the FC payment.08/2)(1 – 0.75 If FC 371.227 – 190.00/$) = $185. Fourth Edition 15. Transcorp would pay 380.S.547 investment is equal to the amount Transcorp could have earned if the dollars had been invested in the U.227. he sells at a higher forward value of FC relative to the future expected spot rate. and insure payment abroad in six months. Ef = 0. X = [1 + (0.000 X = FC 371.547 The net cost (or opportunity lost) that results from this $185. The amount that must be borrowed in dollars today to convert to FC is FC 371.227 532.547) = $192.000 E f + 152.000 E f + (0. X[1 + (0.

Sign up to vote on this title
UsefulNot useful