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

# 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

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

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

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

Borrowers originally chose levels of current consumption to the right of A. Figure S1. Figure S1. current wealth is the intercept of the capital market line with the C0 axis. 3. their utility has increased unambiguously from UB to U′ B. then it increases to W0′ . rate. The case for those who were originally lenders is ambiguous. (c) The amount of investment increases from I to I’. lower. Fourth Edition 2. The remaining lenders experience a decrease in utility. This implies a capital market line with a 45° slope (a slope of minus 1) as shown in Figure S1.2 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′.2 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Originally it is at W0.3 Market rate cannot fall below net rate from storage . 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). 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. from U L to U L′ .3. After the decrease in interest rate. Some individuals who were lenders become borrowers under the new.

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

consumption foregone along the C0 axis) and the extra consumption provided at the end of the investment period. This is done below.000. Figure S1.000.000 1.000 Sum of Outlays $3. In order to graph the production opportunity set.30 1. For example.6 Investment opportunity set 6.000 4.000 The production opportunity set plots the relationship between resources utilized today (i. Project D B A C One Plus the Rate of Return 1.5 shows a schedule of investments.000. The marginal rate of substitution between C0 and C1 is a constant. first order the investments by their rate of return and sum the total investment required to undertake the first through the ith project.000 7.e. Figure S1.5 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. if only project D were undertaken then $3 million in current . Figure S1.000 5.08 1.000.20 1.. Fourth Edition * 5. R .4 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.000.000.6.04 Outlay for the ith Project $3.000 1.000.000.000 2. all of which have the same rate of return.

The present value of this decision is W0 = + C0 1+ r 5. Figure S1. This is graphed below in Figure S1. Hence the optimal production decision is to undertake projects D and B. and Investment 5 consumption would be foregone in order to receive 1.1 and finding that it is tangent to point B.9 million in end-of-period consumption.7. Consumption.1 C1 Figure S1. The answer to part b of the question is found by drawing in a line with a slope of −1.3 × ($3 million) = $3.8 Production opportunity set .1 = + 3 = $7.7 If we aggregate all investment opportunities then $7 million in consumption could be foregone and the production opportunity set looks like Figure S1.Chapter 1 Introduction: Capital Markets.6364 million 1.8.

000) = 28.800 $7. interest and taxes Depreciation (straight-line) EBIT Taxes @ 40% Net income $22.000 (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 12.000 40.000 30.000 = 158.650)(28.000 10. 10 years) − I 0 N $140.000) − 100.000 10.200 .000 4. (a) Cash flows adjusted for the depreciation tax shelter Sales = cash inflows Operating costs = cash outflows Earnings before depreciation.4) + .000 + 10. 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 Alternatively.000 100.000 + (1 − .000 = (5. (a) Earnings before depreciation.200 − 100.000 − 100.000)(1 − .000 = 58.4(10. equation 2-13a can be used: CF = ∆NI + ∆dep + (1 − τ c )∆ k d D = 18.Chapter 2 Investment Decisions: The Certainty Case 1.4)(0) = 28.000 12.000 $18.200 2.

09 13.567 .000 (5/55)100.94 11.958.27 NPV = PV of inf lows − I 0 NPV = 100.712 .000 (4) (Revt − VCt)(1 − τc) + τcdep 13.000 22.91 13.000 (4/55)100.18 13.545.322 (6) PV 18.181.14 15.820 (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 .000 (1/55)100.45 13.281.000)(1 − .909.000 22.82 13.4) + .Chapter 2 Investment Decisions: The Certainty Case 7 Net present value using straight-line depreciation CF = (∆Rev − ∆VC)(1 − τc ) + τc ∆dep = (22.4(10.180 − 100.958.200 + 727.214.536.000 = 958.000 22.636 .797 .29 4.090.633.282.58 100.290.737.200 + 3.36 13.958.26 5.000 22.000 (7/55)100.893 .363.404 .12 13.200 NPV = ∑ CFt − I0 t + (1 WACC) t =1 = (annual cash flow) (present value annuity factor @ 12%.200 + 1.200 + 5. .64 13.000 22.650) − 100.200 + 5.000 22.72 13. 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 22.000 22.636.000 (9/55)100.540.818.000 (3) Dept (10/55)100.454.200(5.000 = 97.200 + 2.000 = −2.02 9.000 (3/55)100.484. 10 years) − I 0 = 17.272.000) = 17.200 + 6.507 .54 13.31 6.58 8.27 − 100.200 + 7.200 + 2.452 .000 22.27 Notice that using accelerated depreciation increases the depreciation tax shield enough to make the project acceptable.000 (6/55)100.200 + 4.361 .000 (2/55)100.000 (8/55)100.03 7.27 (5) PV Factor .000 22.

000 $18.968 $92.000 1.176 − 75.000 18. years 1–8 Savings from new investment Tax savings on depreciation $100.000 12.000 31.000 –25. Replacement Amount before Tax Outflows at t = 0 Cost of new equipment Inflows.00 $100.400 –2. Replacement with salvage value Amount before Tax Outflows at t = 0 Investment in new machine Salvage value of old Tax loss on sale Inflows.000 –15.8 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. buy the new machine.000 Amount after Tax $100.968 21.000 Net cash outlay = $75.000 12.000 1–8 1–8 1–8 8 Using the NPV rule the machine should be replaced.000 1–8 4.000 .4 dep = depreciation = 2.000 τc = the tax rate = .000 Year 0 PVIF @ 12% 1.176 Net present value = $109. 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 4.000 –15. There is no change in revenues.968 24.00 –15. 4.404 4.600 4.405 1–8 4.405 4.0 Present Value $100.000 = $34.936 .840 Present value of inflows = $117.000 1.245 − 100.000 12.848 Net cash inflows = $109. VC = cash savings from operations = −3.600 5.245 If the criterion of a positive NPV is used.000 –10.00 –10.859 4.968 92.000 Year 0 0 0 PVIF @ 12% Present Value 1.245 Net present value = $117.000 –5.176 $31. years 1–8 Savings from new machine Depreciation saving on new Depreciation lost on old Salvage value of new $100.968 –9.500 Amount after Tax $100.000 11.000 = $17. Fourth Edition 3. 5.

the project has a negative NPV. the annual net cash flows for years one through five are CF = 3. it should be rejected (even though the IRR is greater than the cost of capital). 6.83 302.600(2.991) = −2.1 graphs NPV versus the discount rate.4) + .000(1 − . The NPV at different positive rates of return is @ 0% 400 400 −1.000) = 2.78 −578. At an opportunity cost of capital of 10 percent.8 percent.66 1.75%. This is an interesting example which demonstrates another difficulty with the IRR technique. therefore.40 Therefore.70 32. Figure S2.1 The internal rate of return ignores the order of cash flows 1 There is a second IRR at −315.23 @ 20% 333.58 −751.223.600 The net present value of the project is NPV = −10. Note also that the function is undefined at IRR = −1.4(2. .64 330.83 344.000 + 2.33 277. that it does not consider the order of cash flows. namely.44 −7.000 −200 @ 10% 363. but it has no economic meaning.41 Figure S2.27 −657.46 297. the project should be rejected.10 1 Discounted Cash Flows @ 15% @ 16% 347. The IRR on this project is approximately 15.52 −640.32 −57.Chapter 2 Investment Decisions: The Certainty Case 9 Therefore.

3 years. By inspection. These rates are the Figure S2. 8. we see that the IRR is −200%.5(180) = 145 + 90 = 235 * Note: PVIFa: 10%.2 An IRR calculation internal rates of return for the project. 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 = 0.55 The project should be accepted.2 plots NPV against various discount rates for this particular set of cash flows. it would be “double counting” to include financing costs (or the tax changes which they create) in cash flows for capital budgeting purposes. the discount factor for a three year annuity paid in arrears (at 10%). 9.487) – 1.4(400) = 336 + 160 = 496 NPV = 496 (PVIFa: 10%.4) + . . Fourth Edition 7.10 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. We are told that the IRR for these cash flows is −200%. These are the cash flows for project A which was used as an example in section E of the chapter.5) + . 3 years)* – 1. Sources of financing. are included in the computation of the cost of capital.200 = 496 (2.200 = 33. The cash flows are: (∆Rev − ∆VC − ∆FCC − ∆dep)(1 − τ c ) + τ c ∆dep = (200 − (−360) − 0 − 0)(1 − . 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. First calculate cash flows for capital budgeting purposes: CFt = (∆Revt − ∆VCt )(1 − τ c ) + τ c ∆dep = (0 − (−290))(1 − . as well as their costs.

75 . this is not true. at the end of the first time period. NPV = $1.15 NPV(B + C) = 1. 11. On the other hand.751 A −1 0 2 −1 PV (A) −1. B + C is preferred.91 Project A has a two-year payback. Project (A + C) has a two-year payback.000 − 5.826 . Once Project C is combined with A or B.25 1. calculate the NPV: NPV = ∑ CFt − I0 t t =1 (1 + WACC) = (CFt ) (present value annuity factor @ 10%. in the first year the firm invests $5. NPV = $1. Using the method discussed in section F. Therefore.00 0 0 2.25 A+C −2 0 2 2 B+C −2 1 0 4 5 NPV(A + C) = 1.65 −. B was preferred. Payback violates value additivity. The net present values are calculated below: Year 0 1 2 3 PVIF 1.66 C −1 0 0 3 PV (C) −1. 5 years) − I 0 = 235(3. with payback. NPV does not. Project C has a three-year payback.3 of this chapter. Because C is an independent choice. Project B has a one-year payback.89 − 900.00 .91 0 . if projects A and B are mutually exclusive. the results change if we use the payback criterion. NPV does obey the value additivity principle.000(1 + IRR) During the second period the firm borrows from the project at the opportunity cost of capital.15. it should be irrelevant when considering a choice between A and B.000 .10 B −1 1 0 1 PV (B) −1. The amount borrowed is (10.00 = 890. we have 5. k. Therefore.000(1 + IRR)) .791) − 900.000 and expects to earn IRR. Now A + C is preferred.91.00 0 1. Therefore. Previously. 10.909 .00 = −9.75 −. However. Project (B + C) has a three-year payback.Chapter 2 Investment Decisions: The Certainty Case 11 Next. Its NPV is simply the sum of the NPV’s of B and C separately.12 The project should be rejected because it has negative net present value. project B would be preferable according to both capital budgeting techniques.

10) Solving for IRR.45% −5.000 at the end of the second time period: 3.000 − 1 = IRR = 45.000 = (10.000 − 5.12 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.000 1.000(1 + IRR)) (1.10 − 10.000 − 5.000(1 + IRR)) (1 + k) The firm then lends 3. Fourth Edition By the end of the second time period this is worth (10.000 . we have 3.

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

5(8.898.4967825 eln W = W e8.000) + .872.5ln(3.000.5ln(6.5(8.4967825 = $4. (b) The second gamble. Therefore.699515) = 8.000) + .14 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.5(8.517193) = 8.5(8.000) = .98 for sure. he would not buy insurance for $125.98 = W Now the individual would be willing to pay up to $127. (a) E[U(W)] = .1 Total utility of normally distributed returns for a risk lover Figure S3.006368) + .98 = W Therefore.2 Indifference curves of a risk lover 3.26178 eln W = e8.000 plus or minus $1. the individual would be indifferent between the gamble and $4.02. . This amounts to a risk premium of $101.26178 = $3.5ln(5. Fourth Edition Figure S3. Its expected utility is E[U(W)] = . is $4.5ln(4.000) = . Since insurance costs only $125.898. given his first loss.02 for insurance.29405) + . he will buy it.

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 U(Payoff) ln 1 = .693 ln 4 = 1. The expected payoff of the gamble is the sum of the expected payoffs (column four). 10.386 ln 8 = 2. they would be willing to pay an amount whose utility is equal to the expected utility of the gamble. 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. The reason is that people are usually risk averse.079 E U(Payoff) . we have ∑2 i=0 ∞ 1 i = 1 + 12 + 14 + 1 + ⋅⋅⋅ 8 = 1+ 1/ 2 =2 1 − 1/ 2 . Table S3.50 .173 . However. you will be indifferent.e. Below that level of wealth you will pay for the insurance while for higher levels of wealth you will not. a (.50 ln 2 = N ln 2 N N ln 2 N 1 = 0 2 + 2 The gamble has a .000. no one has ever paid an infinite amount to accept the gamble. and so on.5) probability of ending after the second flip (one head and one tail).173 . no heads). if your current level of wealth is $2.000 .1 shows the payoffs. expected payoffs. and utility of payoffs for n consecutive heads.50 . which is infinite.5 probability of ending after the first coin flip (i.000 ln 2 = .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) $. Table S3..Chapter 3 The Theory of Choice: Utility Theory Given Uncertainty 19 Therefore. Consequently.130 N (1/2) N+1 2 N . 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.50 .

000(.000(. since i 2 i=0 N ∑2 i=0 N i i = 2 If the expected utility of wealth is ln2.005) = $350 .20 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.01) + 10. (a) First calculate AVL from the insurer’s viewpoint. AVL1 ($30.005) = $250 AVL 2 ($40.000 insurance) = 0(.000(.005) + 40.005) + 50. 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.005) + 30.005) = $300 AVL 3 ($50.000(.01) + 10.000(. since the insurer sets the premiums. 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.000(. the corresponding level of wealth is U(W) = ln2 eln2 = W = $2 Therefore.000(.98) + 5. 11. 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.98) + 5.000(.000 insurance) = 0(. an individual with a logarithmic utility function will pay $2 for the gamble.01) + 10.000(. Fourth Edition Evaluating the second term.98) + 5.000 insurance) = 0(. we have E(U) = 1/ 2 ln2∑ i = ln 2.

000 loss (P = .8148 $10. Contingency Values Of Wealth And Utility of Wealth (Savings = $20.01) 5 + 2.) Finally.7608(.9598 With $40.9601 No loss (P = 1.005) 5 + 2.8091 $10.105 1.07) ≅ 7.14 0. E(U(W)) = ∑ Pi U(Wi ) i and choose the amount of insurance which yields the highest expected utility.98) 5 + 2(1.000 $50.5 = 6.2a With no Insurance: E(U(W)) = 1. given his utility function.98) + 1.000 loss (P = .11 1.035 − 1.11 With $40.0 = 2.6312 $20.005) 5 + 2.9615(.995) 5 + 2.10 Table S3.000 insurance 1.9582 With $30.07) ≅ 7.0280(1. Casadesus is $50.000 Premium 30 + 250 = $280 27 + 300 = $327 24 + 350 = $374 Next.005) 5 + 2.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 $30.9657 No loss (P = .9600 With $50.8931 $5.000’s) U(W) = ln W With no insurance 1.14 − .105 With $50.14 − . find the expected utility of wealth for each amount of insurance.9608 No loss (P = .000 insurance: E(U(W)) = 1.14 − 5.14 − 1.995) + 1.0374(1.9608(.995) 5 + 2.9601 Therefore.0 ≅ 6.000 insurance 1.8091(.14 With $30.000) Utility of End-of-Period Wealth Wealth (in $10.14 − .01) + 1.005) + 0.000 loss (P = .005) = 1. calculate the insuree’s ending wealth and utility of wealth in all contingencies (states).995) + 1. The utility of each ending wealth can be found from the utility function U(W) = ln W.0) 5 + 2. the optimal insurance for Mr.000 insurance: E(U(W)) = 1.07) ≅ 7.8148(.9615 No loss (P = .14 1.64 1.000 loss (P = .000.7608 $50.14 − . Assume he earns 7 percent on savings and that premiums are paid at the beginning of the year.6312(.14 − .000 insurance: E(U(W)) = 1.14 − .0 = 6.03 − 2 ≅ 5.005) 5 + 2.0327(1.8931(.2a.000 $40.000 insurance 1.005) = 1.07) = 7. . (See Table S3.9657(.000 loss (P = .005) = 1.

995) $20. so we can calculate the expected utilities for each amount of insurance directly.4) = –2.005(1.995(1.74 5 + 34. while a less wealthy individual may choose maximum coverage.07) ≅ 39. (c) The end-of-period wealth for all contingencies has been calculated in part a).047 = –2.3404 1.005(1.028(1.000 insurance No loss (P = .3543) + .000 insurance: E(U(W)) = .366162 = 1.000 insurance: E(U(W)) = .0 ≅ 38.366092 = 1.24 5 + 34.000 insurance No loss (P = 1.005U(21.00(1.3663) + .01(1.000 insurance: E(U(W)) = (1)1.0327(1.005) With $50.07) ≅ 39.98(1.005 (1.24 − .000) Utility of Wealth U(W) = ln W (Wealth in $100. (See Table S3.24 − .005(1.205 5 + 34.000 loss (P = .3662) + .005) With $40.035 − 1.000 loss (P = .3671 1.4) – .366071 With $50.000 insurance No loss (P = .24 − 1.24 5 + 34.3543 1.005) With $30.016 – .3140) .4)) + .98(U(71. only with $320.995(1.0 = 38.000 loss (P = .2308 1.21 5 + 34.21 5 + 34.0 = 34. the analysis illustrates that a relatively wealthy individual may choose no insurance.000 loss (P = .000 loss (P = .000’s) With no insurance No loss (P = .2b above for these calculations.3140 1.366038 With $40.98) $5.005) $50.3404) = 1.000.000’s) 1.4)) + .005(200/61.24 5 + 34.3661 End-of-Period Wealth (in $10.2308) = 1.3413 1.01(200/66. Although the numbers are close with logarithmic utility.24 − .98(200/71.24 − .0) 5 + 32.4) – .3671) + .3413) + .5 = 38.745 – .005(U(61.2b Contingency Values of Wealth and Utility of Wealth(Savings = $320.4) – .24 − .01(U(66.838 With $30.3662 1.000 in savings instead of $20.03 − 2 ≅ 37.20 (b) Follow the same procedure as in part a).4) = –.005(200/21.995) $10.01) $10.07) ≅ 39. Fourth Edition Table S3.0374(1.366092 The optimal amount of insurance in this case is no insurance at all.22 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.07) = 39.3663 1.) With no Insurance: E(U(W)) = .4)) + .030 – .205 5 + 34.24 − 5. With no insurance: E (U(W)) = .24 − .

Case 1 (a) Second order dominance—B dominates A because it has lower variance and the same mean.000. as his savings account is increased he prefers to bear greater and greater amounts of risk. the mean and variance are the only relevant parameters. It lies to the right of B. once his wealth is large enough. (b) First order dominance—There is no dominance because the cumulative probability functions cross.000 insurance: E(U(W)) = (1)( −200/71) = −2.Chapter 3 The Theory of Choice: Utility Theory Given Uncertainty 23 With $30.1)) = – .005(200/51. .1) = –2. 1 Properties of this utility function.000 insurance: E(U(W)) = .000.8169 Hence.8008 – .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. (b) First order dominance—A dominates B because its cumulative probability is less than that of B.05) = –2. 12. Mr.995(U(71.995(200/71.000W-2 > 0 MU′ W = −400.005(U(61. Eventually.1)) + .799 – .000W− : MUW = 200.1) – . U(W) = −200.0164 = –2. make his savings account = $400.005(200/61. he would prefer not to take out any insurance. Because returns are normally distributed.995 (200/71.995(U(71. Casadesus would renew his policy for $50.000 insurance: E(U(W)) = .020 = –2.05)) = – . To see this. Case 2 (a) Second order dominance—A dominates B because it has a higher mean while they both have the same variance.05) + .8172 With $50.05) – .005(U(51. with this utility function.819 With $40.

there is no second order dominance.010 2 X is clearly preferred by any risk averse individual whose utility function is based on mean and variance.888 .29) = 3.1 Y pi Yi . (b) First order dominance—Given normal distributions.784 .96) = 3.36) = 6.4 Xi − E(X) −16. 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.4 .69) = 2. it is not possible for B to dominate A according to the first order criterion.7 24.2 .272 var (X) = 37. Because Σ(F − G) is not less than (or greater than) zero for all outcomes.049 var(Y) = 66.3 .738 .896 .0 2.2 . Figure S3.6 pi(Xi − E(X)) .2(2.840 2 Prob Y .5 First order dominance not possible 13.6.1 .8 30 3.7 Yi − E(Y) −3.49) = 59.4 2 3 1.96) = .3(12.0 E(Y) = 5. . because X has a higher mean and a lower variance than Y.0 2.5 shows an example.5(7.89) = .578 . Figure S3.2(31.3 pi (Yi − E(Y)) .7 −2.6 5.0 3.5 .2 X −10 5 10 12 pi Xi −1. (a) Prob X .4 E(X) = 6.4(1.96) = 26.7 −1.645 .4 3.1(268. (b) Second order stochastic dominance may be tested as shown in Table S3.5 4 .3 on the following page.4 −1.1(590.2(13.24 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. as shown in Figure S3.

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.

0 . so the optimal investment is E.9 −2.2 −.7 −1.6 −.0 .7 .8 −.2 −.10.8 1.4 −.9 −1.0 .1 0 0 −. If the sum of cumulative differences crosses the horizontal axis.4 1. .0 1.6 1.2 −.6 1.0 .0 .0 1.4 2.4 −1.5 .0 1.1 0 0 −.8 −1.2 0 .9 .8 −1.4 .4 0 1.4 −.2 1.9 .2 0 .0 . Table S3.4 .1 0 .0 1.8 0 1.2 A>B SSD SSD (DA) (EA) 0 0 .0 .2 0 .2 −.6 0 1.8 0 .0 0 1.0 .2 B<E SSD (ED) 0 −.4 0 .7 1.2 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 −.5 −.1 .6 on the following page and graphed in Figure S3.9 .1 .1 −.4 0 .6 .0 .4 −.3 .0 −1. By SSD.8 .2 0 .6 Ret.1 −. they are also inferior by SSD.6 1.0 .5 0 1.2 1.8 0 1.0 1.0 0 .0 . The pairwise comparisons of the remaining four funds. E > A.6 .8 1. there is no second order stochastic dominance.9 First order stochastic dominance (b) The second widow is clearly risk averse.2 −.8 −.3 .1 −. E > B.1 −1.2 −.2 −.5 .7 0 .4 .0 −2.8 1.4 0 . −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 .0 .5 1. Since C and F are eliminated by FSD.0 −1. so second order stochastic dominance is the appropriate selection criterion.6 .4 0 1.5 .7 −.1 0 −.6 0 .6 −1.2 .5 . and E > D.2 0 0 0 0 0 A>D A<E SSD (DB) −.4 −1. as in the comparison of B and D.8 .3 −.4 0 1.0 .5 −1.1 0 0 .8 1.2 1.8 1.2 −.5 .4 0 .1 0 .Chapter 3 The Theory of Choice: Utility Theory Given Uncertainty 29 Figure S3.1 0 .3 1.0 1.6 −1.2 no 2nd order dominance SSD (EB) −.2 0 .4 0 1.2 1. Σ(F(X) − G(Y)) are presented in Table S3.

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

0 6. E E E Dominance Relationships A > C.80 2. D > F A > B.7 A B C D E F E(X) 6.89 (b) Mean-variance ranking and SSD both select trust E as optimal. B. However.36 0.2 6. D.Chapter 3 The Theory of Choice: Utility Theory Given Uncertainty 31 Table S3. C > B FSD SSD M-V . the rankings of suboptimal portfolios are not consistent across the two selection procedures.36 26.1 var(X) 1. D. Optimal A. C.00 28.2 6.0 6. F.2 6.96 26. A > D A > B.

20 = . 20P1 + 40P2 = 10 −[120P1 + 40P2 = 20] 100P1 = 10 P1 Substituting into the first equation. 30P1 + 10P2 = 5 20P1 + 40P2 = 10 Multiplying the first equation by 4 and subtracting from the second equation. (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.20 P1 = . 20P1 + 40P2 = 10 2 + 40P2 = 10 40P2 = 8 P2 = . B) Pj = price of pure security j (j = 1.10 P2 = .10 .Chapter 4 State Preference Theory 1. 2) Substituting the correct numbers.

4 S2 = War St. (a) The equations to determine the prices of pure securities.8 Substituting .5 and P2 = . solve two equations with two unknowns: 6P1 + 6P2 = St.8) = 22 12P1 = 22 – 16 P1 = 6/12 = . 12P1 + 20(. P1 and P2. From part a) we know the prices of pure securities. P1 = .5 (b) The price of security i. Pi. 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. and subtracting the second equation from the first. Thus. (a) The payoff table is: Nova Nutrients = j Galactic Steel = k S1 = Peace St. 6 St. P1 is the price of a pure security which pays $1 if state 1 occurs.8 for P2 in the first equation. can be determined by the payoff of i in states 1 and 2.00 3. 6 St. P1 and P2.5(6) + . 24P1 + 40P2 = 44 −[24P1 + 10P2 = 20] 30P2 = 24 P2 = 24 / 30 = . Pi = P1Qil + P2Qi2 = . and Pj is the price of security j. are given below: P1Qj1 + P2Qj2 = Pj P1Qk1 + P2Qk2 = Pk where Qj1 is the payoff of security j in state 1.8. 10 4P1 + 36P2 = St.8(10) =3+8 = $11. 36 To find the price of pure securities. 20 . and the prices of pure securities for states 1 and 2.Chapter 4 State Preference Theory 33 2.

and subtracting it from the second equation.000) 20 –2nj + 200 = –6nj nj = –50 shares of j (N.000/10 = 100 nk = W0/Pk = 1.25 6(1.N.S.800 St. 20 −[36P1 + 36P2 = St. Let W1 = his final wealth if peace prevails. he receives –njPj.000/20 = 50 If he buys only Nova Nutrients. which will return at least [(–njPj + W0)/Pk]Qk1. With this he can buy at most (–njPj + W0)/Pk shares of k.) .34 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.: W1 = njQj1 = 100(6) = 600 St. and W2 = his final wealth if war prevails.): If he sells –nj shares of j. 4P1 + 36P2 = St. W2 = njQj2 = 100(6) = 600 St. (c) For sales of j (N.N. If he buys G. he can buy 100 shares.25) + 6P2 = 10 P2 = . he must pay out at most –njQj1. W2 = nkQk2 = 50(36) = 1. Then nj = W0/Pj = 1.) and purchases of k (G.S. Fourth Edition Multiplying the first equation by six. and with his initial W0 he will have –njPj + W0.: W1 = nkQk1 = 50(4) = 200 St. the minimum –nj is determined by −n j Pj +W0 Pk Qk1 = −n jQ j1 4 = −6n j (−10n j + 1. he can buy 50 shares. 1. Therefore.N.4167 (b) Let nj = number of Nova Nutrients shares and nk = number of Galactic Steel shares. If he buys only Galactic Steel. 60] −32P1 = − 40 P1 = St. If he buys N.

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

1.000)(12/25) = $480 W1 = (4/3)480 = $640 .36 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. ∂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. 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.

n k = 0 = = 100 (a) Pj 10 Pk 12 If she buys j: W1 = njQj1 = 120(10) = $1. and nk the number she can buy if she buys only k. Let nj = the number of shares the investor can buy if she buys only j.200 final wealth in state 1 W2 = njQj2 = 120(12) = $1. 200 (8) = −12n j 12 − 20n j + 2. 4. nk = –5) is the optimum portfolio. in this case the investor buys 110 shares of Nova Nutrients and issues five shares of Galactic Steel.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. she must pay out at most –njQj2.200 nj = 0 = = 120. solve the two simultaneous equations W1 = njQj1 + nkQk1 W2 = njQj2 + nkQk2 Substituting the correct numbers.440 final wealth in state 2 If she buys k: W1 = nkQk1 = 100(20) = $2.Chapter 4 State Preference Theory 37 To find optimal portfolio. Therefore. she receives –njPj. the minimum –nj is determined by: −n j Pj +W0 Pk (Qk2 ) = − n jQ j2 −10n j + 1. Then W W 1.200 1. and with her initial wealth W0 she will have –njPj + W0. 400 = −36n j n j = −150 . 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.

5) = 10 P1 = . the minimum –nk is determined by: −n k Pk + W0 (Q j1 ) = −n k Q k1 Pj −12n k + 1.000 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) = 2.400 (c) To find the price of pure securities. 10P1 + 12(.50 Substituting .2.50 for P2 in equation 1.200 = = 150 Pi 8 .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. Fourth Edition For sales of k and purchases of j: If she sells –nk shares of k. Therefore.38 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. she must pay out at most –nkQk1.40)5 + (.40 (d) The price of security i is given by Pi = P1Qi1 + P2Qi2 = (. solve two equations for two unknowns as follows: 10P1 + 12P2 = 10 20P1 + 8P2 = 12 Multiplying the first equation by two. and subtracting the second equation from the first equation. the number of shares she can buy is given by ni = W0 1. with this she can buy at most (–nkPk + W0)/Pj shares of j.50)12 =2+6 =8 (e) (The state contingent payoffs of a portfolio invested exclusively in security i are plotted in Figure S4. which will return at least [(–nkPk + W0)/Pj]Qj1. and with her initial wealth W0 she will have –nkPk + W0. she receives –nkPk.) If the investor places all of her wealth in i. 20P1 + 24P2 = 20 −[20P1 + 8P2 = 12] 16P2 = 8 P2 = .

5 = 0.6 ) ∂W1 = 1.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. her wealth in state one will be zero.4W2−6 W1.875W2 .4 and substituting the correct numbers. and k In equilibrium: dW2/dW1 = P1/P2 1.5W2/W1 = . ∂W2 = (.4/.5W2 = 0. The W1 intercept is the portfolio of j shares sold to buy k. This portfolio plots as the W2 intercept in Figure S4.6 W2.4 ) ÷ (.2 State payoffs for securities i. resulting in zero wealth in state two.8W1 W1 = 1.800 If the investor sells k to purchase j. (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 = W1.5W2 / W1 Figure S4.2 on the following page.6W1−.4 W2. j.8 1.

5.800 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. S2 = $1. Substituting the correct numbers. Fourth Edition Wealth constraint: W0 = P1W1 + P2W2 1. we can use them to determine its equation.800 Since both of these payouts lie on the budget constraint (see Figure S4.800 = 10nj + 20nk W2 = 960 = 12nj + 8nk Solving equation one for nk in terms of nj.800 – 10nj) ÷ 20] 4.5W2 W2 = 1. W1 = 1.875(960) = 1.400. he can obtain $720 = 80 shares with payout $9 S1 = $2.800 – 10nj nk = (1.800 = 60nj + 3. If the individual uses 100 percent of his wealth to buy security j.800 – 10nj) ÷ 20 960 = 12nj + 8 [(1. S2 = $800 S1 = $900.3 on page 42). he can buy $720 = 90 shares with payout $8 If he spends $720 on security k.200 = 40nj nj = 30 nk = (1. and substituting this value into equation two: 20nk = 1. it will be possible to determine what an equal payout will be.200 = .875W2) + .4(1.600 – 20nj 1.800 – 10nj)/20 nk = 75 The investor should buy 30 shares of j and 75 shares of k.25 = 960 W1 = 1. the number of shares of each security to be purchased. (a) If we know the maximum payout in each state. The equation for the line is W2 = a + bW1 .200/1.40 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.

we have nj 20 + nk 60 = 2. the slope is − 1.500) − Therefore.440 nj20 + nk10 = 1.400 = $1. we want to find the point along the budget line where W2 = W1.400 − W2 = W1 2 W1 (the budget constraint) 3 (equal payout) 2 W1 3 Substituting the second equation into the first.Chapter 4 State Preference Theory 41 Substituting in the values of the two points. We now have two equations and two unknowns W2 = 2.400)] 1.400) = W1 = $3.880 −[n j 20 + n k 10 = 1.440 nk = 28.400 − A risk-free asset is one which has a constant payout. the payout of the risk-free asset is W1 = 2. your payout in states one and two will be State 1: State 2: nj10 + nk30 = 1. we obtain two equations with two unknowns. “a” and “b.440] n k 50 = 1.” 1.800 = a − 2 (900) 3 a = 2.6 .400 − W1 = 2. which we have already determined.500 3 2 and the intercept is 3 1. regardless of the state of nature which occurs.800 = a + b(900) –[800 = a + b(2.400. Therefore.400 The maximum wealth in state two is $2.000 = b(–1.600 0 = 2.000 2 = b =− 1.440 Multiplying the first equation by 2 and subtracting. The maximum wealth in state one is 2 W1 3 3/2(2.8 and nj = 57.440 = W2 5/ 3 If you buy nj shares of asset j and nk shares of k.

3) Take the partial derivative with respect to Q2 and set it equal to zero: −.2Q1 – .000 − .440 for sure.4) .2) or 1.6 2 + =0 50. No feasible solution exists.6Q1 − .8Q2 (4.3 The budget constraint (b) The risk-free portfolio contains 57.6Q1 – .4Q 2 3Q1 (4. In general. We to solve Max[log C + 2/3 log Q1 + 1/3 log Q2] subject to C + .2Q2 = 50.000 We can solve for C in (4.6Q1 – .4Q 2 3Q 2 or 1. 000 −. It costs $720 and returns $1.440 =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).2) and substitute for C in (4.4 1 + = 0 50. it is necessary to have at least as many assets as states of nature in order for complete capital markets to exist.6Q1 − .000 – .000 – 1.6 shares of asset j and 28.440 1 + rf 1. Fourth Edition Figure S4.1) (4. Any attempt to solve the problem would require solving for three unknowns with only two equations.42 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Max[log (50.000 – . the risk-free rate of return is 720 = 1 + rf = 1. Therefore.4Q2 (4.1).4Q2) + 2/3 log Q1 + 1/3 log Q2] Take the partial derivative with respect to Q1 and set it equal to zero: −.8Q1 = 100.8 shares of asset k.4Q2 = 50.6Q1 + . 6.

2).000 – $16.33 = $25.3333Q2 .2) and (4.000 Hence. 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).3) yields 2.333.333. C = 1.4) imply 1.666.3) and (4.777. but he buys more of pure security 1 (because its price per probability is lower) than of pure security 2. C = $50.67 – $8.000 Q2 = 20.8Q1 = 2. and 20.33 hence Q1 = 27.777.4).78 units of pure security 1 at $0.Chapter 4 State Preference Theory 43 Together.33 units of pure security 2 at $0.666.833. Substituting into (4.4Q2 = 50.40 each for a total of $8.67.833.78 (a) The risk-averse individual will purchase 27.000 also from (4.4Q2. (b) From (4.2Q2 = 25.60 each for a total of $16.33. (4. or Q1 = 1.

rx. var(y). y) σx σy requires the calculation of var(x). The equation for correlation between x and y.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.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 1. and cov(x. y). because the standard deviation must be positive σ y = bσ x and σ x σ y = bσ x 2 . while. 2. with limited liability. the probability distribution is bounded on the left by P = 0. Figure S5.1 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. Given y = a – bx.

045 –.003 . .25) = 1/2(. (a) Table S5.25 ri –.30) = .028 Range (r) = rmax – rmin = .002 .75 – r.016 0 .028 2 E(r) = Var(r) = ∑p r i i i = .10 .10 . 3.20 .30 –.15 .20 – (–.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 45 cov(x.000 .20 piri –.009 where r = P1 − P0 .13 . y) − b var(x) = = −1 σx σy b var(x) Thus.16 0 .020 .10 – (–.1 Prob. x and y are perfectly negatively correlated.009 i ∑ p (r .30 . P0 P0 = 50 pi(ri – E(r)) .014 . 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.009 ∑ = .16)) = . y) = cov(x.50 Semi-interquartile range = 1/2(r.050 ∑ = .E(r))2 = .

0) = 1. Semi-variance is important to risk-averse investors who are more concerned with downside risk (losses) than gains.05 ri –1.07 .20 . • Financial Theory and Corporate Policy. First.16 0 .38 Semi-interquartile range = 1/2(r.020 .004 .01 .009 Range = rmax – rmin = .05 . 4.13 The expected return and semi-interquartile range are the same.004 –.10 . because future possible prices are drawn from the same probability distribution as in part a).46 Copeland/Shastri/Weston. x i + = 0 if x i ≥ E(x) In the example above.30 .05 . y) = E(xy) – E(x)E(y) Therefore. has increased.021 . The range increases with n. y) = E[(xy – xE(y) – yE(x) + E(x)E(y)] and taking the expectation of the right-hand side we have cov(x. y) + E(x)E(y) .010 –.010 .15 .001 . y) = E(xy) – E(x)E(y) – E(y)E(x) + E(x)E(y) cov(x. Fourth Edition (b) Table S5. semi-variance = ∑ pi (ri− − E(r))2 = 0.010 .009 pi (ri− − E(r))2 (semivariance) .14 . we have cov(x.019 ∑ = .38 – (–1.023 E(r) = ∑piri = .2 Prob.25) = 1/2(.10 . y) = E[(x – E(x))(y – E(y))] Multiplying the factors in brackets on the right-hand side.20 –.00 –.005 . the number of observations.000 — — — — ∑ = . E(xy) = cov(x.20 .38 piri –.023 .23 –.75 – r.016 0 .02 .16)) = 0. Only n.10 – (–. cov(x. Semi-variance is given by E[(x i − − E(x)) 2 ] where xi − = xi if x i < E(x). recall the definition of covariance.30 –.003 .016 –.015 –. .

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. the variance is 2 var = σ2 x − 2 cov(x. Therefore. x) + b σ x The long position means 100% in X and the short position means –100% in X.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 47 5. The variance of a perfectly hedged portfolio is 2 2 var = a 2 σ2 x + 2ab cov(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. x) + σ x 2 2 σ2 x − 2σ x + σ x = 0 and the position is perfectly risk free. 6. Let σ2 x be the variance of the stock.78 3 3 3 . Using matrix notation. a=1 b = –1 Therefore.

0 –5.0 2.15 27.1] −24.2 .0 3.675 28.48 Copeland/Shastri/Weston.25) – .0 6.49 7.00 2 pi[Xi – E(X)] 16.2 Xi 18 5 12 4 6 piXi 3.298 4.048 The opportunity set is shown in Figure S5.0 –4.8(24.905 5.00 9.25 cov = [.1(26.4 .0 8.80) + .8 20.25 cov = [.466 7.370 5.0 6.875 50.10 25 32 12 −.25)] = –14.0 2.292 4.0 Yi – E(Y) –5 –8 10 7 –4 [Yi – E(Y)] 25 64 100 49 16 2 Xi – E(X) 9.8 1.0 pi[Yi – E(Y)] 5.100 29.8 2 Yi 0 –3 15 12 1 piYi 0 –.8 2 pi[Xi – E(X)][Yi – E(Y)] –9.1] −10 75 32 −.127 9.0 7.6 3.0 1.0 var(X) = 28.6 1.8 3.2 .2 E(Y) = 5.8 .2 on the following page.2 3.1 .0 9.2 .2 .2 (b) Mean = aE(X) + (1 – a)E(Y) 2 2 Variance = a 2 σ2 x + 2ab cov(X.475 19.2 var(Y) = 50.800 81.00 25. 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) 10.2 1. Y) = –1.4 cov(X.8 .1 .0 5.0 –7.0 12.0 9.8 5.0 var(Rp) 47.0 4.2 E(X) = 9.00 9.875 σp 6.000 18. • Financial Theory and Corporate Policy. Fourth Edition (b) The covariance is 24 −10 25 1. .0 2.00 16.4 6.1(27.0 –3.25 cov = [.0 [Xi – E(X)] 81. (a) pi .80 26.4 .

0) + .2 The opportunity set Thus. y). y) into equation 5.183 .8 − 2(−1. a* = = σ2 y − cov(x.36)2 (50.36(5.64(9. the minimum variance portfolio has 64 percent invested in asset x and 36 percent in asset y.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 49 (c) The minimum variance portfolio is given by equation 5.56 var(R p ) = (.0) + 2(.64) 2 (28. y) 2 σ2 x + σ y − 2 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.21.2) 28.64)(.21.8) = 11.584 = 17.64 Figure S5. y) = rxyσxσy and substituting cov(x.0) = 7. The portfolio mean and variance at this point are: E(R p ) = . y) 50.553 + 6.0 + 50. cov(x.36) ( −1.5 σp = 4.2) = .469 − .2) + (.8 − (−1.

and portfolio A is 28.1 = [.75 .36] = 17.5 17. (See Merton (1972).36] −1.0 −1.25 cov(A.2 .1 = [.25 20. A) = [.36 17.64 .75 6. this is not a coincidence. and portfolio B is 28.2 .0 −1.50 Copeland/Shastri/Weston.) (g) The variance of the global minimum variance portfolio is 28.64 .64 .36] −1. (a) Using the definition of expected value.488 = [.36] −1.25] 37. B) = [.25 cov(Vmin . It can be proven that the covariance between the global minimum variance portfolio and any portfolio along the efficient set is a constant.8 = .520 Again.1) + .36] = 17.8) = 14. • Financial Theory and Corporate Policy. 8.8 . Vmin.36] = 17.75 .64 . The covariance between the two is 28.25(37. B) = [.8 .5 37.0 −1. Fourth Edition (d) Portfolio A is 75 percent in X and 25 percent in Y.2 50.2 50. The variance of the minimum variance portfolio equals the covariance between the minimum variance portfolio and any other portfolio along the efficient set.75 6.025 (e) The covariance between the minimum variance portfolio.64 . portfolio B is 25 percent in X and 75 percent in Y.64 var(Vmin ) = [. 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) .8 . Vmin.0 −1.8 .2 .2 .2 50.75(6.7 = [.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).5 11.25] −1.75 cov(Vmin .8 The covariance between the minimum variance portfolio.64 .2 50.

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. y) + b cov(z. then σ(x + y) = σx + σy: var(x + y) = var(x) + var(y) + 2 cov(x. y) cov(x. y) (d) Using the definition of the covariance of x with itself. 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. 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. 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 .Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 51 (b) Using the definition of variance. cov[(ax + bz). y) = rxyσxσy Therefore. y) + b 2 var(y) (c) Using the definition of covariance. 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. 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. cov(x. if rxy = –1. if rxy = 1.

50% 3.50 Var(Rp) 7.33(.36 36. the portfolio’s standard deviation is σp min = [(.09 If we put two-thirds into asset 1.3 150 100 75 50 25 0 –50 Opportunity Set Computations % in Asset 2 –50 0 25 50 75 100 150 E(Rp) 0.33)(−.33)2 (.3 The opportunity set in mean-standard deviation space (b) Using equation 5.563 5.40 % in Asset 1 Figure S5. • Financial Theory and Corporate Policy.01) .18% .93% 14. line 4: E(Rp) = 5.000 13.250% 2.31 11.02 + .18 16.05) + 2(.063 1.75 8.08) = 4. 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.01)]1/ 2 = (.65% The minimum variance portfolio is plotted in Figure S5.67)2 (.01) .01 22.03) + .25 5.67 .3.50 6.00 10.67(.21.52 Copeland/Shastri/Weston.3.250 σ(Rp) 26.14 10.06 = = .05 − 2( −. Fourth Edition 9.00 4.67)(. y) then a * = .50% σ(Rp) = 11.02) + (. (a) Table S5. (c) From Table S5.05 − (− .01)1/ 2 = 10% E(R p min ) = .000 1.250 2.

5 − .0225) + .65) = .35) (.35) (1. we also need to know the mean and variance of the minimum variance portfolio.15) + .8225) = .1( −1.1(1.5 − . we have a* = = σ2 y − rxy σ x σ y 2 σ2 x + σ y − 2rxy σ x σ y .0225 .65 − . R 2 ) = .35) 2 + .0225) = .8(.450 Because the mean-variance efficient set is the upper half of the opportunity set.30) 2 = .8(−.2025 E(R 2 ) = .1(−1.5 − .0 − .2025 .35)2 = .3500 .15 − .2025) + .30) + .2025 σ(Rp) .1(.30 Var(R 2 ) = .8(.21.1(1.35)2 + .3375 .1(.4 % in 1 100 75 50 25 0 Computing The Opportunity Set % in 2 0 25 50 75 100 E(Rp) .8(.36 .0 − .0225) + .0225) + .0225) + .65 − .2025 − 2(.1(.0225) .1(.8(.1(.8(.3000 σ2(Rp) .2025) = .1(1.367 .8(.18 = = . Table S5.8225) + .0225 (b) The table on the following page gives the mean and standard deviation of returns for various portfolios.30) + .450 .2025 Cov(R1 .Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 53 10.5 .3250 .367 .1125 .30) = .1350 .3125 .1350 .1(−1.15) + .335 .5) + .5 − .30) 2 + .35) (.1(.1(.2025 − .1(.30) 2 + .15 − .2025 + .0) + .5) = .1(. (a) E(R1 ) = .8(.15 − .35 Var(R1 ) = .1(.15 − . Using equation 5.

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

i = 1. 1)dZ When an asset is risk-free. we have the result that when returns are normally distributed. we have Ψ = The first-order conditions are ∂Ψ = 2w i σii − λ = 0. the standard deviation of return is zero. 0. Forming a Lagrangian. 0.e. ∂w i ∂Ψ = ∂λ ∑w σ 2 i ii − λ ( ∑ w i − 1) . N equations. 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. the marginal rate of substitution between risk and return is dE − ∫ U ' (E + σZ)Zf(Z. Going back to Chapter 3 (Utility Theory). and since R = E + σZ . the optimal weight is 1/N. Proof is given by minimizing the variance subject to the constraint that the weights add to 100 percent.. the rate of return always equals its expected value. ∑ wi = 1.1)dZ = dσ ∫ U' (E + σZ)f(Z.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 55 Figure S5. we see that the minimum variance is achieved if all of the weights are equal. i. 12.6 Optimal portfolio composition this implies putting an equal amount of wealth in each security. If there are N securities.

0 –5.1 .15 . the slope of the capital market line must be positive at the y-axis.0 2. If we find the expected return represented by its intercept then we can determine the correct weights for the zero variance portfolio.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. • Financial Theory and Corporate Policy. No risk averter will hold all of his wealth in the risk-free asset.2 .56 Copeland/Shastri/Weston. The linear equation for the opportunity set is E(Rp) = a + bσ(RP) We know the coordinates of the points X and Y.0 4.8.2 . therefore the slope is b = E(X) − E(Y) 10 − 8 = = .0 E(X) = 10.1206 σX − σY 20.0 6. this implies that the slope of the risk-averter’s indifference curve is always zero at the y-axis.2(10) = 8 VAR(Y) = (.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 Xi 30 20 15 10 –50 piXi 3.2 Also. a tangency between a risk-averter’s indifference curve and the CML at the y-axis is impossible.7 No risk-averse investor will hold 100% of his wealth in the risk-free asset 13. the marginal rate of substitution between expected return and risk.04(430) = 17.2 = 4. As shown in the figure below. Fourth Edition we have the result that Z = 0. the standard deviations of X and Y are σX = 430 = 20. Therefore. is zero.1 1.74 σY = 17. pi . Therefore. First compute the expected returns and standard deviations for asset X.4 . The opportunity set is graphed in Figure S5.2) 2 VAR(X) = . dE/dσ. Figure S5.74 − 4. E(Y) = 6 + .2E(X) = 6 + . As long as the ex-ante market rate of return is greater than the risk-free rate. we can immediately write the expected value and variance of asset Y.

1206(20.74) a = 7.15) + (1.0.0)(20.74)(4.5 The zero variance portfolio has an expected return of 7. wi = Vi .79 + 26.2) = 26.88 ≈ 0 14. 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. Vi.25)2 430 + 2( −. (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 .1206σX 10.25)(1.5%. therefore each weight is defined as the value of the ith asset or liability. we should sell short 25% of our wealth in asset X and go long 125% in asset Y.88 − 53.5 = α(10) + (1 − α)8 α = − .8 Opportunity set for two perfectly correlated assets Using the coordinates for X and the slope we have E(X) = a + . 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) 7.0 = a + .25)(1. 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) = ( −.25 Therefore. S. divided by the market value of equity.25)2 (17.Chapter 5 Objects of Choice: Mean-Variance Portfolio Theory 57 Figure S5.

00040 .07) 0 . we have N=− N=− VUS rUS.340 .03) .04) 700(.09(.2(.03) w' Σ w = [1.02) 2 0 0 2 0 .00036 .07) (.0 − .TBσD PTB σTB Expanding this expression to fit the current problem.00040 0 0 .9)(.3(.0 2.00020 −.33003 = .000 − 3.08) .0 w US 2.03) 0 Σ = 0 .07) 2 0 .09(.574482 (b) The position in T-bond futures contracts is taken in order to minimize the standard deviation of the shareholders’ claim.03) 0 0 (.000420 0 0 .2(. 403 + 1.8(.5 − 8.07) −850(.3(.04) (. Fourth Edition The variance of the shareholders’ position is Var(R S ) = w' Σ w where: w' = [ w STA = [1.8(.TB σL PTB σTB − 200(.5 − 8.5 w' Σ w .0 − w STL − 5.000900 0 1.04) (.5)(.0 −w D ] − 8.04) .5] .00399 w' Σ w = .5] 0 0 0 0 .TB σi PTB σTB −VD rD.000360 0 .0 2. Equation (5.TBσ US PTB σTB − VL rL.5] (.00042 0 .07) 0 (.08) .0 7. • Financial Theory and Corporate Policy.0 2.0 wL 7.33) 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.00160 .5 −8.0 7.33003 The standard deviation is the square root of the variance σ= .0 − .00040 0 .00224 0 .03) (.04) (.07) (.01582 = [1.08) N = −1.58 Copeland/Shastri/Weston.0 −.04) (.063 N = −3.00490 0 .09(. N = −∑ i Vi ri.0 7.03521 −.00224 .02) 2 0 0 0 0 2 (.3)(.03) − − .

00640 1.00615 .09 million) = = −3.00280 0 .0 −.340 futures contracts. RS = wSTARSTA + wUSRUS + wLRL – wSTLRSTL – wDRD + wTBRTB The market value weight of the futures position is wF = NPTB −3340(.00040 0 0 0 0 0 0 .0 .00040 .00020 −.5 − 8.02681 w' Σ w = [1. .5 − 8.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 3.0] −.0 7.00280 0 .000360 .000720 .5 −3.253605 The new standard deviation is .0 2.00040 0 0 .0 7. The new shareholders’ position is augmented with futures as follows.00042 0 .000900 .5 − 3.00224 0 .00490 0 .0] .00072 .00036 .00160 .0 2.000420 .00288 .00004 w' Σ w = .00224 0 . The T-bond futures position will rise in value when interest rates fall.0 − .0 − .0 7.5 − 3.503592.5 −8.00288 0 0 0 0 . and this will offset the declines in the value of the bank’s loan portfolio and government bond holdings.0 2.0 S 100 million and the variance of the revised shareholder position is w' Σ w = [1.00718 .

3 . Rm) = .2(.02150 βj = cov(R j .03 .05 .15 . 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 –.10 = .15 .25) = .15 .15 .05 .00100 .02450 σ2 j = .00100 .1 .01 (a) E(Rm) = 10% (b) σ2 σm = 10% m = 1%.00675 .4(.45 –.10 – .0100 pi (R j − R j )(R m − R m ) –.1(–.30 .01125 .15% (e) E(R j ) = R f + [E(R m ) − R f ]β j E(Rj) = .35 .05 .015 .05) = .35)(.00700 cov (Rj.015 .00625 .R m ) σ 2 m E(R m ) = .040 .02025 .05 .05) = .06 + [.25 –.0525 Rm pi R m Rm − Rm –.060 .00200 σ2 m = .06]β j .20 –.05)(.4 .3(–.10 Rj − Rj pi (R j − R j )2 .2 –. namely (in addition to those already assumed) • • • • 2.08 .45)(–.15)(–.10) = .20 .00 .00225 .00 . Rm) = 2.15 pi R j –.0215 = 2.00075 .10 pi (R m − R m )2 . (c) E (R* j ) = 15% (d) cov(Rj.00100 . The conditions are those necessary for two-fund separation.Chapter 6 Market Equilibrium: CAPM and APT 1.50 E(R* j ) = .

06] = . E(R* j ) = 15% (see c above).09 E(R m ) = E(R j ) = ∑R N N m = = .6%.27 ∑ R m = .12 –.10 − .05 .05 –.01 = .0300 .15 (R m − R m )2 .30 ∑ R j = . Rm) = 2.20 .06 + [.0225 0 . Now we are dealing with sample data.06 + .05 –.10 . cov(X.8%.12 –. Covariance calculations also are divided by N – 1.0215 .03 .06 + [.Chapter 6 Market Equilibrium: CAPM and APT 61 (f) The required rate of return on Donovan Company is E(R j ) = .1% .6% The expected return. we might expect its price to increase so that in equilibrium its expected rate of return equals 14.018 5 ∑R j . R m ) = N −1 6 = (a) E(Rm) = 12% (b) var(Rm) = 1.15 –. 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 – 1.04(2.15 .Y) = Year Rm ∑ (X i − E(X))(Yi − E(Y)) N −1 Rj .12 6 var(R m ) = ∑ (R m − R m )2 N −1 = .15 0 –.03 .42% (c) E(Rj) = 10% (d) cov(Rj. therefore.05 –.0300 ∑ = .25 .0225 1978 1977 1976 1975 1974 1973 . we can use statistics formulae for sample mean and variance.021 5 σ m = 13. 3.15 0 –.6 Rj − Rj (R j − R j )(R m − R m ) Rm − Rm .105 = . We subtract one because we lose one degree of freedom in estimating the mean.R m ) σ2 m .0225 .0225 0 .0225 0 .10 cov(R j .15 –. . Because Donovan earns more than is required.72 .72 = . the number of observations minus one.09 = .105 ∑(R m − R)2 = .0225 0 .04] cov(R j .6 ∑ (R j − R j )(Rm − Rm ) = .15) = 14.20 .15 .27 .

8318 . and given that we have rational. expected-end-of-period-utility-of-wealth maximizers. and all assets are held.2644 = .0221 = .0801 = . we would expect the Milliken company to decline in value.0089 10 − 1 9 cov(R j . If every individual holds an efficient portfolio. 5) = 1.066 + [. 5) = . ∑ (col.1022 (a) R j = 10 10 (b) var(R j ) = ∑ (col. 7) = .0107 10 − 1 9 (c) σ(R m ) = .0964 = .1034 5. Panel B. all individuals will perceive the same efficient set and all assets will be held in equilibrium. risk-averse. then the market . 7) = .0107 6. Fourth Edition (e) E(Rj) = Rf + [E(R m ) − R f ] cov(R j .1. Given the two above assumptions.7796 ≅. ∑ (col.Rm ) var(R m ) (f) Required return (Rj) = . R m ) var(R m ) = . Therefore.e.9% and the expected (i. See table S6. the key assumptions are that 1) investors have homogeneous expectations and 2) all assets are perfectly divisible and marketable. It is not necessary that there be risk-free assets.1667) = 12.62 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.0107 = .0089 = . 4.054] (1. Panel A.078 (a) R m = 10 10 (b) var(R m ) = ∑ (col. Other than the usual perfect market assumptions. See table S6.1..9% The market requires 12. 8) = .0294 10 − 1 9 (c) cov(R j .R m ) = (d) β j = ∑ (col. anticipated) rate of return is only 10%.

1376 .0087 –.0221 (6) –.05 Variance of Returns (R jt − R j )2 Covariance with Market (R jt − R j )(R mt − R m ) (62) (7) .044 Panel B: GM Year Price (t) Pt % Change in Price Pt −1 Pt −1 (3) .22 (8) .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) .1471 –.0007 .0301 .0043 –.0137 –.0356 –.05 .0208 .0136 .1494 (4) .0029 .1136 .05 .1201 .1111 .04 .0020 .0864 –.0358 .4731 .0298 .0964 R f = .1947 .27 62.0324 .0736 –.1384 –.7796 (6) .0738 .0836 –.06 .0004 .38 69.0237 –.0337 .1212 .0190 .0971 –.0269 .84 83.05 .0534 .0587 .0201 .1640 .0322 .04 .0050 –.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) .1866 –.Chapter 6 Market Equilibrium: CAPM and APT 63 Table S 6.0811 –.05 .87 81.04 .05 .1030 .05 .03 .0250 .0320 .4231 .07 .0317 .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) .2162 .03 .03 .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) 55.85 66.17 85.1556 –.0770 .1102 .0340 .06 .0385 –.05 .0264 –.93 98.1043 .70 97.05 .0708 .0300 .37 88.0782 .0383 (5) .0010 .1993 –.0314 .0330 .1833 –.3709 .0192 .26 91.0010 .0013 .05 .0543 –.0070 .44 /10 = .1167 .0885 –.0242 .0307 .0004 .06 .0263 –.2644 (6 × 6 [Panel A]) (8) –.0010 .0054 .1333 –.0153 –.0029 .0397 .0059 .0191 .1286 –.1034 .0113 .0002 .0041 1.1063 .2662 .0165 .1518 .1646 .0106 .0612 .0066 .0274 .2164 –.

and if you can observe any one of the following: 1) the percent of the investor’s portfolio held in the market portfolio. Algebraically. a. Because the risk-free asset has no variance. 7. The capital market line is made up of a linear efficient set as in Figure 6. 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.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. as proven in problem 6. σp = βp σm . a = β p. The percentage. or 3) the expected return on the investor’s portfolio. Fourth Edition portfolio must also be efficient because it is merely the sum of all efficient portfolios held by all individuals. β is the measure of systematic risk. βp = aβm and since β m = 1. two-fund separation obtains. With risk-free borrowing and lending. Finally. of the market portfolio held in the investor’s total portfolio.7. To see this. Therefore.. a. 8. Thus. you can tell how risk averse he is. βp = βf (1 − a) + βm a where β f = 0 is the beta of the risk-free asset. all efficient portfolios are linear combinations of the risk-free asset and the market portfolio. total variance in every efficient portfolio is contributed solely by the market portfolio. These three measures are equivalent.0 βp = a The greater the percentage. Therefore. Yes. the less risk averse the investor is. i.R f = a 2 σ2 m +0+0 where a represents the percentage of the efficient portfolio invested in the market portfolio.e. Thus.64 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. a portfolio’s variance that can be explained in terms of market variance. 2) the β of the investor’s portfolio. 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. Systematic risk is defined as market risk. is equal to the portfolio’s β p. . all risk is systematic.

8β j β j = .10 = .09 or 9% The definition of correlation (from Chapter 5.14 − .05 = 2.06 + (. We can use it to find Rapid Rolling’s unsystematic risk.0 . The total risk (equation 6.2) = .2)2 + σε 2 σ2 ε = . σ2 ε 2 2 2 σ2 k = β k σ m + σε 2 .0)2 (. all assets must be priced so that they fall along the security market line.2) = .2 − .2) = cov(Rj.Rm ) σ jσ m cov(R j .04 = . 10. Rm). E(R j ) − R f + [E(R m ) − R f ]β j . σ2 ε . . β k is its systematic risk.15 − . i.3 or 30%. that all portfolios lie on the security market line. Rm) = .52 − 4(.25 − . Therefore.5(.5 (.05 We know that efficient portfolios have no unsystematic risk. σ j = β jsm = 1.52 = (2. (a) We know by the CAPM: R j = R f + [E(R m ) − R f ]β j .06 Since rjm = 1.Chapter 6 Market Equilibrium: CAPM and APT 65 9. equation 17) is rjm = To find cov (Rj. use the defination of β j bj = 2 cov(R j .36 or σε = .5 . we again have the result that efficient portfolios are perfectly correlated with the market (and with each other).0.R m ) σ2 m 1.5 The CAPM assumes that the market is in equilibrium and that investors hold efficient portfolios.e. Assuming equilibrium.05 Total risk is defined by equation 6.05 = β j = 1. E ( R k ) = R f + [E(R m ) − R f ]βk βk = .12) is 2 2 2 σ2 j = β j σ m + σε and since the unsystematic risk of an efficient portfolio. we can find the β k for Rapid Rolling.12..06)β j .15 − . is zero. σ2 j = .6 11.

the one period rate of return.08P0 + . we have $100 − P0 = .10 − .08 Therefore.66 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Rj.2P0 1. the covariance between two portfolios is cov = W1'Σ W2 . 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.e.125 13. Also. (a) A zero-beta portfolio has zero covariance with the market portfolio. Figure S6. Fourth Edition (b) Let “a” be the percent invested in the risk-free asset. Using the CAPM.11) The security market line 12.14 = .08 + [.08]2.1 (Problem 6. 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. (1 – a) = 50%. i..28P0 = $100 P0 = $78.0 P0 $100 − P0 = .18 − . Assuming that the company pays no dividends. the individual should put 50% of his portfolio into the market portfolio. using matrix notation. we have E(R j ) = Substituting in the appropriate numbers and solving for P0.5 a= −.

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

06) . 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 = .14W2 = .20 If the expected future price is still $45.98 . P1.19 –1 (σ jm )0 σ2 m σ2 m where (σ jm )0 = original cov(E(Rj). 13 percent.08)1.75 = 1 W1 = . the expected future price.07 = βj .08W2 = .06 W2 = .20 but the covariance with the market has doubled.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. E(Rm)) where 2(σ jm )0 = subsequent cov(E(Rj).20/1. was P0(1 + r) = P1 40(1.19 = 19% At the original rate of return.5 = .13 − . E(Rm)) 2(σ jm ) 0 = $37. 17.13 = .07 + (.75 W1 + .06W1 + .08 β j = .08) β j .75 If the rate of return covariance with the market portfolio doubles.13) = $45.06W1 + .06 and subtracting from the first gives . so that the required return is 19 percent.07 + (. and given a present price of P0 = $40.12 −(. We know from the CAPM E(Rj) = Rf + [E(Rm) – Rf] β j Substituting the correct numbers for β j . Multiplying the second equation by .Chapter 6 Market Equilibrium: CAPM and APT 71 We have two equations with two unknowns.06W2 = . then the present price will be P0 = P1(1 + r) = 45.

05 − .2.72 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. To find the percentage change in the security’s price that would result in a return of 10 percent * instead of 12 percent.08 + [R m − . investment in the security would be advisable. the market unexpectedly fell.10 where βj = σ jm σ 2 m = .12 = = 1.08]4 + .01 is E(R j ) = R f + [E(R m ) − R f ]β j = . P* is a 1.2)2 Since the expected rate of return is greater than the required rate of return.08 + (.08). the expected rate of return and the estimated risk could have been overestimated. The second possibility is that after the fact.01818 P0 1. (b) There are two possibilities for the low ex post return.08]. solve for P in terms of P0: P0 (1.01 = .16 − .05 = .08 = −. E(Rm)) = . If the security’s price increases more than 1.25 (.04 The ex post security market line is depicted in Figure S6.2 The ex post security market line . Fourth Edition 18. Figure S6.1 Thus.82 percent increase in price over P0. (a) The required rate of return on a security with cov(E(Rj).25 R m = [.10) P* 1. First. reverse the investment decision.25 E(R j ) = .12) = P*(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 .82 percent.

2) the portfolio of risky marketable assets.5] . E(R j ) = R f + λ[Vm cov(R j .R I ) = [1 0] .006075 = [.Chapter 6 Market Equilibrium: CAPM and APT 73 19. and 3) a hedge portfolio chosen to hedge against unforeseen changes in the future risk-free rate.R J ) = [.0025 0 .0025 0 . where the three funds are 1) the risk-free asset.25] .0081 COV(R A .0081) = .00405 = [.002650 .5(.R m ) + cov(R j .5] .75 .0001563 = . then we can use Merton’s (1973) continuous time model as shown in equation 6.006075 The variance of investor I’s index is 0 .5 . (b) If the risk-free rate is nonstochastic.R I ) .28 E(R j ) = rf + γ1[E(R m ) − rf ] + γ 2 [E(R N ) − rf ] Once again. the equilibrium pricing equation is given by Mayers (1972) in equation 6. 20.0047126 Repeating the exercise for investor J we have 0 .0025 0 .0025 0 . (a) Given that some assets are nonmarketable.5 .5 .00125 = .75 . This may be interpreted as a form of three-fund separation.75 .000625 = .0081 COV (R A .26. Separation is still valid in the sense that the marginal rate of substitution between risk and return is independent of individuals’ utility preferences.0045563 + .75 .002025 + . three-fund separation obtains.0081 COV(R J . investor I computes a β of βI = COV (R A .0081 COV (R I .25] .006075 = = 1.5 = .5 .R I ) . Every investor will hold one of three funds: 1) the riskfree asset.75(.R H )] This implies that the individual will take into consideration not only the covariance of an asset with the market portfolio.25 .R I ) = [. but also its covariance with the portfolio of nonmarketable assets.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. and 3) the portfolio of risky nonmarketable assets. 2) the market portfolio.289 COV (R I .5 .R J ) = [1 0] .0047126 Therefore.00405 0 . (a) The covariance between investor I’s index and asset A is 0 .000625 = .0081) = .

000625Wz2 = 0 Wz1 + Wz2 = 1 Solving. R J ) .2750 − . even though the investors perceive different security market lines and different betas. as shown below .1147(20%) = 18. 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 = . we have one equation and two unknowns . Since all portfolios lie on the minimum variance opportunity set in this two-asset world.25 Solving.1885 + [. (c) The key to understanding this problem is Roll’s critique. The zero-beta portfolio is the minimum variance portfolio which has zero covariance with the index portfolio. They must.1147.75 .0025)Wz2 = 0 However.R J ) . They have homogenous expectations therefore they perceive the same risk-return combination. They must be because this (over-simplified) world has only two risky assets. then we require that COV(R z . Consequently.74 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.R I ) = Wz' Σ WI = 0 The appropriate numbers for investor I are: COV(R z . SML for I .0081)Wz1 + .85% Therefore. we also know that the weights must add to one Wz1 + Wz2 = 1 This gives two equations and two unknowns. they will perceive the same expected rate of return on any portfolio. we have Wz1 = −. The expected return on asset A is 30%.002650 (b) Both investors will require the same rate of return on asset A. In order to numerically demonstrate this result the zero-beta portfolios and security market lines for each investor are computed below. If we define Rz as the return on the zero-beta portfolio and Wz as its vector of weights.0081 Wz2 ] =0 .528 COV(R J .1885]βI .00405 = = 1. for investor I the expected return on his zero-beta portfolio is E(R z ) = Wz1E(R A ) + Wz2 (R B ) = −. we need be concerned only with the zero covariance condition.006075Wz1 + .1147 Thus. Wz2 = 1.0025 0 . Any combination of them will lie on the mean variance opportunity set.25(. Fourth Edition And investor J estimates a β of bJ = COV(R A . The two investors choose different index portfolios.R I ) = [Wz1 0 .1147(30%) + 1.75(. but both portfolios are efficient.

2500 − . (a) Using the CAPM.11 . Investor J’s zero-beta portfolio will be COV(R z .05)(−.1885]1. 21. Wz2 = 1.4464(20%) = 15.1554]β j . we can now verify that investor I expects a 30% rate of return on asset A.0 = .50 The expected rate of return on investor J’s zero-beta portfolio is E(R z ) = −.05](−.08 + [.528 for asset A. The only difference is that investor J uses a different index portfolio. we see that his expected return on asset A is E(R A ) = . we have E(R i ) = .142 = .0081)Wz1 + .5) + [.R J ) = [Wz1 The two simultaneous equations are .1554]1.4464 0 . recalling that investor J estimated a β of 1.4464(30%) + 1.142 − . Finally.08 + [.11]βi2 .4464.142 Using the fact that E(Ri) = 14.1554 + [.Chapter 6 Market Equilibrium: CAPM and APT 75 Recalling that investor I estimated a β of 1.289 = 30% Next.2500 − .289 for asset A.0081 Wz2 ] .11]βi2 .2750 − .R j ) = Wz' Σ WJ = 0 COV(R z .50(.50 .2% for her portfolio and using the APT equation.142 = .05]βi1 + [.54% Investor J’s security market line equation is E(R j ) = .08 + [.50(.08 − (. SML for J.7909 = βi2 .1885 + [. E(R A ) = .1554 + [. we repeat the same procedure for investor J.062]1.528 = 30% This numerical example demonstrates that although investors I and J estimate different betas.5) = βi2 . they will have the same expected return for asset A.0025 0 .0025)Wz2 = 0 Wz1 + Wz2 = 1 and their solution is Wz1 = −. the expected return on her portfolio is E(R i ) = R f + [E(R M ) − R f ]βi = .

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

5 d2 = –. we have P0 = C0 − S0 + X − rf T e . 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 d1.5 – .5) + .1829) Solving for P0. we have N(d1) = .5( .3434 N(d2) = .40335 .5)( .5 −.40335 – ( .5) = –.970446)(.1829) C = 9.36 pricing European calls.52 2. and converting the formula to continuous rather than discrete compounding.25 = –.5) (. we have d1 = d1 = ln(28/40) + .1566 = . we have C = 28(.5)( .06(. the present value of a put. 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.90335 Using the Table for Normal Areas.6152 – 40(. according to equation 7.5 – .06(.5) .1829 Substituting these values back into the Black-Scholes formula. We can use the Black-Scholes formula (equation 7.356675 + .03 + .Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 1.8 C0 – P0 = (1 + rf )S0 − X (1 + rf ) –.5 .3434) – 40e = $2.3171 = .

62172) – e–(.3064 and d2 is equal to d2 = d1 – .08) (.6 . Fourth Edition First. calculate the value of the corresponding call option.216 = $2. we have d1 = ln(20 / 20) + (.6)( .5) = –16.434 – 8.1179 From the Table of Normal Areas.08)(.78 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.744 – 20 + (20)e–(.12172 = .744 Solving for the value of a put.4243 = –.690 = 3.3064 – .5 – . according to the Black-Scholes formula (equation 7. P = 3.45224) = 12.62172 N(d2) = .6)( .5) + (1/ 2)(.36) C = SN(d1) – e − rf T XN(d2) where: d1 = Solving for d1. we have N(d1) = .256 + 19.08)(.96 .5)(20)(.5 = .45224 Substituting these values into the Black-Scholes formula yields C = 20(.5) (.04776 = .5 + .5) ln(S/X) + rf T + (1/ 2)σ T σ T = .

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

375 / 45) + (.4920) = 25. T. C0 + a – (P0 + a) = C0 + P0.6538) = .4847 = $3. Because portfolio A has a value either greater than or equal to the value of portfolio B in every possible state of nature.18 d2 = . The payoffs are given in Table S7. X1) > P(S. we can determine N(. equation 7. P(S. Using the correct values from the problem.375(.4274) + (1/ 2)(.31)(. with the same time to maturity.32. given inside information of an increase in the instantaneous variance.07)(. 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. 4.80 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. C = 44.6538) (. the strategy of this portfolio is neither advantageous nor disadvantageous.0961)( 156 / 365) ( .9704)(.18) and N(–. and the second is a European put option written on the same stock. T.375N(d1) – 45e First calculate d1 and d2: d1 = = – (. Fourth Edition The return from this portfolio (C – P) remains unchanged by an increase in the instantaneous variance. but with exercise price X2 < X1. since.07)(156 / 365) + (1/ 2)( .02).02 Substituting these values into the Black-Scholes Formula. In the algebra of part (a) above.1.18) – 45e N(–.01398 + (.87 5. Therefore.18 – ( . X1) b) P(S.0961) (.5714) – 45(.03 Using the Table of Normal Areas. X2) Comparative Value of A and B S < X2 X1 – S X2 – S VA > VB if X2 < X1. C = 44.31)(.375N(.6538) = –. X2) Table S7. T. The value of the call option can be calculated directly by the Black-Scholes formula for a European call.02) –. T.3559 – 21. Compare the payoffs at maturity of two portfolios.07) (156/365) N(d2).0961)( 156 / 365) −. the put with a higher exercise price is more valuable. Substituting these values into the formula yields C = 44. The first is a European put option with exercise price X1. X1 ≤ S 0 0 VA = VB State Contingent Payoffs of Put Portfolios X2 ≤ S < X1 X1 – S 0 VA > VB .1 Portfolio a) P(S. by put-call parity. ln(44.

6065)(.000(.000.665 d2 = d1 – ( .000.399 = $892.44) – 4. the time to maturity.05.000e = 4.571 . ten years.05)(10) N(–.601.Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 81 6. is the time until the firm’s debt matures.107.000.5)( 10) ( .236 = –.399 The value of the debt is $5.000 – 4. and subtract this from $5MM to find B. the market value of the debt.2231) + (.000 (.05)(10) + (1/ 2 ( .5) + 1.5 – . we have d1 = = ln(5MM / 4MM) + (.118 2. For rf = .951 = $4.350 – 517.5 + .625.5 )( 10 ) (. $5MM.236 = 1. 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) + (.236 = –.000N(1.000.118 = 1.2865) ( .5)( 10) –(.000.795) = 5. d1 = (.107.44 – 2.5)( 10) = 1. For Rf = .795 S = 5.236 + 1.1)(10) + (1/ 2) ( .42507) – 4. The total value of the firm.10.44 d 2 = d1 – σ T = 1.5 )( 10 ) Note that T.665 – 2.2231 + 1) 2. 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).

110 The value of the debt decreases to –(. since stockholders in this example hold a call option on the value of the firm.5 – .342.095 .000.16 .050 .000.0 .045 . and the ∂rf T in Months 2 4 6 8 10 12 14 16 18 Table S7.3).9 .188 . (S = 1).118 .039 .571) = 5. the risk-free rate (in Table S7.089 .123 .028 .2.005 .21603) $5.452) – 4. The three tables below show the call price as a percentage of the stock price.342.107 S/X = 1.049 S/X = 1.034 .02 .024 . T = 6 months. and S7.153 . rf = 6%. σ = .016 .041 .110 .000.890 = $4.101 .091 S/X = 1.037 .071 .178 .131 . In order to standardize the results we assume that the stock price is equal to 1.038 .4.198 ∂C > 0. S7.077 .2 S/X = .82 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.9 .167 .028 .142 S/X = 1.890 Therefore.061 .18 . This is what we would expect. 7.000 – 417.3679)(. Table S7.000.160 .031 .056 .175 rf . the rise in rf benefits the stockholders.146 .000 – 4.167 .000 (.06 .1 .060 . and the standard deviation (in Table S7.083 .0 .142 .098 .10 .138 .121 .000e = 4.14 .2 S/X = .2).072 .104 .110 = $657. σ = .04 .3.118 . Fourth Edition S = 5.08 .4).156 .132 .2 C/S as a Function of T Assume: S = 1. The relationships given by the tables are graphed in Figures S7.000(.072 .665) – 4.3 C/S as a Function of rf Assume: S = 1.066 .1)(10) N(–.760. plotted against the time to maturity (in Table S7.131 .081 .085 .025 .000N(1.1 .000.12 .5 + .

T = 6 months.030 .141 .30 .124 .3 C/S as a function of rf .9 .045 .054 . rf = 6% S/X = .10 .028 .099 .085 .35 .175 Figure S7.25 .082 S/X = 1.005 .4 σ .0 .40 C/S as a Function of σ Assume: S = 1.131 .1 .000 .15 .20 .000 .05 .152 .126 S/X = 1.112 .058 .2 C/S as a function of T Figure S7.119 .033 .Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 83 Table S7.163 .072 .041 .015 .01 .118 .118 .068 .

t . there is no difference between having only common stock or having subordinated debt and common stock outstanding. it pays to exercise the put option early. then the Figure S7. This fact limits the maximum profit from holding a put. 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. Therefore. 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. Alternatively. As far as the senior debt is concerned.5 The early exercise of an american put option future value if you exercise now is (X – S) e rf T − t . Therefore. 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 condition for early exercise is (X − S) e rf (T − t) > X. However. BS.4 C/S as a function of σ 8. If the date of exercise is t and the time of maturity is T. Fourth Edition Figure S7.84 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. 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. if you hold the option to maturity it is worth X dollars at most. In words. 9. At any point in time the put can be exercised for X – S dollars. the price cannot fall below zero. The owner of a put option profits when the stock price falls. the value of senior debt.

are out of line relative to each other. Therefore. 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.09 while the actual market price was $0. there is no difference between senior or junior * debt.5 percent. From the point of view of the common stockholders. The Black-Scholes model estimated the price of the August 45’s options to be $0. both written on the same asset. This implies either that (1) the August 45’s are overpriced by the market or (2) the August 35s. It really doesn’t matter which price is out of line.15. all that matters is the exercise price they must pay to repurchase the firm at t . the correct ratio of all contracts per share of stock is [1/N(d1)]. all we need to do is write 1/( ∂ C1/ ∂ S) August 45s and buy 1/( ∂ C2/ ∂ S) August 35’s. the option which can be exercised at any time is more valuable than the corresponding option which may be exercised only at maturity. All that matters is that two options. 11. Therefore. 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. 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. the value of the subordinated debt must be . If these two positions are combined. 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.8. This is always equal to the face value of all the debt DS + DJ. which provided the estimate of implicit variance of 51.Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 85 where T is the remaining time to maturity. there are conditions where it pays to exercise an American put early. the stock holdings net out. If the August 45’s are overpriced relative to the August 35’s. 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. 10. are undervalued by the market. As shown in problem 7.

0683619 = .02 of the August 45 calls.1367239 = . we will have created the correct hedge.2614 + . d1 = ln(S/ X) + rf T + (1/ 2)σ T σ T ln(44.9793 Therefore.5 + .5 – . the hedged portfolio will be to write 1/( ∂ C1/ ∂ S) = 1/N(d1) = 1/.515)( 30 / 365) .31)( 71/ 365) . Given the facts of the problem.e.74 of the August 35 calls and buy ' 1/( ∂ C2/ ∂ S) = 1/N( d1 ) = 1/.515)( 30 / 365) = −.0683619 . Fourth Edition Using the facts applicable to Krespy Kreme August 45s we have ln(34.0126438 + .0171)(30 / 365) (.046 = 21.981559 = . we can solve for N(d1).2815 + .75/ 40) + (.0014 + .65 / 45) + (.112212 + .065)(71/ 365) = + (1/ 2(.9793 = 1.515)( 30 / 365) (. if we write [1/N(d1)] call contracts.31)( 71/ 365) (. 12.4793 = .04 .515)( 30 / 365) = and from the Table of Normal Areas ' N( d1 ) = .7603 + . (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.75/ 35) + (.454 = .9131966 + .86 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.0171)(30 / 365) d1 = + (1/ 2)(.0738 = –1.1477 = –1.0738 = 2.6865 and from the Table of Normal Areas N(d1) = . 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. we have ' d1 = ln(46.0014 + . sell) [1/( ∂C / ∂ S)] call options.0738 . we have ∂C = N(d1) ∂S Therefore.1476 + (1/ 2)(..046 Repeating this exercise for the KK August 35s.

the solution is only approximate. we have N(d1) = . you would buy 1/(1 – N(d1)) put options.Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 87 From the Table of Normal Areas.47 call options. 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. the put-call parity formula in continuous time. 1/(∂C/∂S) = 1/N(d1) = 1. Since N(d1) = .837.837 The inverse of N(d1) is the number of call options to write against each share.1947 Therefore.135 put options 1 − N(d 1 ) .337 = .815 shares. 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. we can obtain an exact solution to the problem by using put-call parity. you would buy .10.815 shares = 5 puts . 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. that is.5 + . From the equation for a put.135 puts. (b) If we assume that Honeybear options are European options. for each share of stock you hold.135 puts 1 shares = 1 put 6. If we write equation 7. If the option is an American option. we can find ∂S ∂P = –(1 – N(d1)) ∂S Therefore.163 shares = 1 put . a change in the stock price is offset by an equal and opposite ∂P : change in the value of the put options held. 1 1 = = 6.135 . if you hold 100 shares of stock you should write 119. if you owned one share you would buy 6. 1 share = 6.

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

Chapter 7 Pricing Contingent Claims: Option Pricing Theory and Evidence 89 If we assume discrete compounding. 80.000 = and the exercise price is X = $199. the put-call parity relationship is S0 + P0 – C0 = The amount lent is $80.000. Thus. therefore S0 + P0 – C0 = $80.60 X (1.065.000. .2) 5 X (1 + r ) 5 .

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) is valid. the short-term rate will decrease (increase) towards the long-run mean. impliea a negative (positive) drift. implies that.15)3 1. as compared to the long-run mean.5209 = = 1.5): B(t. T)r(t) . ν is the speed of adjustment of the short-term rate to the long-term mean and σ is the instantaneous standard deviation. the implied forward rate for the third year is 2 3 f = 17. 2. which.14) 2 1.02%. T) = A(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 ) (1.7796 = = 1.93%. 8. There is mean reversion in the process specified in equation (8. an average. The price at time t of a zero-coupon bond that pays $1 at time T is then given by (Eq. µ. 8.2996 The two-year rate of interest is 36.3693).1703 = 1 + 2 f 3 (1.2996 Therefore.Chapter 8 The Term Structure of Interest Rates.155)4 1.5 –1 = 17. 8.4) since a higher (lower) current short-term rate.4): dr = ν(µ – r)dt + σdz where µ is the long-term mean of the short-term rate. Vasicek (1977) assumes the following mean-reverting process for the short-term rate (Eq. (a) Assuming that the unbiased expectations hypothesis (as given by Eq.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.3693 (1. T)e –b(t. r.140) 2 1. and Futures 1. in turn. The average one-year rate is (1. Forward Contracts.

9208 0.3767 1. the short-term interest rate process is Eq.8): B(t.T)e − b(t. the instantaneous standard deviation.8572 Interest Rate [tRT] 2. Cox. In the model.8242 1. and the instantaneous standard deviation. the current short-term rate.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.7470 1. σ. In this question.T )−T + t) ν µ− 2 2 2 b (t. we get: t RT = 1 [b(t. is 0. 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.1617 A(t.T) = A(t. the long-run mean.58% 2. 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. ν.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. µ.87% 3.00% 3. T) = e− (t RT )(T− t) .9443 0. T) = 1 − e− v(T − t) v 2 σ2 ( b(t.9464 0. ν.8848 0.T)r(t) . where tRT is the interest rate at time t for a term of (T – t). T) 0.9139 0.9953 2. σ are specified.08% b(t. µ.T)r(t) − A(t. downward-sloping or humped depending on the values of the various parameters. T) = e Since B(t.9707 0. r(t) is 2%.4.Chapter 8 The Term Structure of Interest Rates. and Futures 91 where r(t) = the short rate at t b(t.9745 0. is 6%. Forward Contracts. (8.9907 0. T)] 0. the speed of adjustment. is 10% and the rate of adjustment. T) 0.T)] T−t Equation (8.8951 3.06% 3. The term structure can be upward-sloping.T) σ − 4ν ν2 A (t.

9373 0.3705 1. tRT. T) and A(t.9279 0.1284 A(t. T) = 2 γ + ( ν + γ ) (e γ (T− t) − 1) γ = ν 2 + 2σ 2 The interest rate from t to T.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.70% 3. 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.24% 3.8447 4.8874 0. would be given by Equation (8.97% 4. the instantaneous standard deviation is 10% and the rate of adjustment is 0. T) 0. T)] 0.9734 0. Ho and Lee propose a binomial model for bond prices of all following form: .8963 0. T) 0.65% 3. In this question.9): t RT = 1 [b(t.9895 0.T)] T−t where b(t. Fourth Edition where b(t.T)r(t) − A(t. the long-run mean is 6%.4. T) are as defined above.9634 0. the current short-term rate is 2%.8231 1.23% b(t.92 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.8531 0.8095 Interest Rate [tRT] 2.7326 1.

105 1.9.2) h(1) = B(1.5 and 0. respectively. In this question.14): h( τ) = δτ 1 and h*( τ ) = π + (1 − π)δτ π + (1 − π)δ τ for some constant 0 ≤ δ ≤ 1. These inputs would result in the following values for h(T). In addition.12) B(t. T) h*(T − t) The no-arbitrage condition implies that (Eq. 8. 5 and 6%. 1.i.i.t) B(t.11): B(t.9615 0. 0.9021 in one period.0.2)]B(0.9496 and B(1.T)]B(t.10a. 8.9021. 0.i + 1. we get: B(0. 8.T) = h(T − t) B(t − 1.i.053 1.T) = [πB(t + 1.8396 Applying the no-arbitrage condition to B(0. Combining the no arbitrage condition with the perturbation equations yields the following constraint on the perturbation function (Eq.2) + 0. 8.i. i.i. 0.t) h(0) = h* (0) = 1 The above equations when divided imply that (Eq.T) B(t.2) h*(1) Substituting values from the table into the above two equations yields B(1.5B(1. The evolution of bond prices is based on perturbation functions h(τ) and h*(τ) such that (Eqs.i. and Futures 93 where B(t.947 0. Note that B(0. t + 1) where π is the probability associated with the perturbation h(τ).i + 1. h*(T) and B (0.i. T) is the price of a bond at time t and state i that pays $1 at time T.i. 2) = [0. 8. the value of a bond that pays $1 in 2 periods is going to be 0.T) + (1 − π)B(t + 1. h(τ) and h*(τ) have to satisfy the following conditions(Eq.i + 1. π and δ to be 0.10b.9496 or 0. This implies that with a perturbation of h or h*. 0. T) 0.T) h(T − t) = B(t.1.157 h*(T) 0.10c) B(t − 1. 8.T) = B(t − 1. Forward Contracts. 0. . 8.2) = 0. 2).0. T): Maturity (T) 1 2 3 h(T) 1. respectively.0.13): 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.Chapter 8 The Term Structure of Interest Rates.T) h * (T − t) B(t − 1. T) is the initially observed term structure of bond prices.843 B(0.895 0.1) The perturbation conditions imply that: B(1.5B(1. 0. 2 and 3-year rates are at 4.1.1.9070 0.i. 2) = 0.

Fourth Edition Applying the same logic and process at different nodes yields the following tree of bond prices: 5. using the insight provided by Samuelson in Eq.749. 6.125) 90/360 (1.029884) (1.06) 77/360 1. (1. As a result.000(1.000. A second. it costs PV = $1.83 –167/360 = $1.95674983) .94 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. there may be no market for long-lived contracts because their prices would not fluctuate enough to create any trading volume.000.30). First.000(.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. and related. we note that the variance of futures contracts declines for longer-lived contracts. 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.10) = $956. answer is that hedgers may have no need to protect themselves against price fluctuations more than 18 months in the future. (8. (b) To make a riskless arbitrage profit we need to borrow enough to purchase one futures contract that will deliver a $1. (a) The first step is to compare the product of short-term rates against the longer-term rate.045205 > (1.000 face value 90-day T-bill on March 22.045205 > 1.012541) 1.000. There are several possible answers to this question.10) 167/360 > (1.

00 – $30. Therefore.5) 77/360 = $956. with the same time to maturity and exercise price.05 7.07 = $10.00 – $31. our profit (on March 22) is Cash received $970. 8.80 in order to have zero cash outlay. and the variance of the futures price will decrease as we get closer to maturity.54 Loan payment 968.5) = $42. the autoregressive coefficient.97098354) = $970. we deliver our $1. and Futures 95 We will borrow this amount at 6 percent for 77 days.000.748.748. (a) We can create a synthetic forward contract by buying a European call.012541) = 30 – 35(.000(.983.749. in Eq.000 face value T-bill that now has 90-days to maturity.49 Simultaneously.00 – S0e rf T = $42.07(. we receive the following amount of cash PV = $1. See Section D. In return. P. C0 – P0 = S0 – 0 FT e− rf T C0 – P0 = 30 – 35e –.79619 = –3.Chapter 8 The Term Structure of Interest Rates. on the underlying asset with maturity T and exercise price X equal to the forward price 0 FT.2 (b) Trends in spot prices can affect the variance of futures prices.00e.000 face value.00(1. The put-call parity equation is C0 – P0 = S0 – X e − rf T and the exercise price must equal the forward price. you profit at delivery will be Profit = ST – 0FT = $42.06) = $968.80 Consequently. 8.000.965605) = 30 – 33.03562) = $42.125)–90/360 = $1.21 is greater than one.83(1.54 After repaying the loan.00 – $30. a. to cover our short position. Since spot prices are expected to rise from February through May. (a) Trends or seasonality in spot prices have no effect on futures prices. on March 22 we will repay FV = $956.000(1. On March 22. (b) If you believe the stock price will be $42 six months from now.235.000.000.749. C.93 .83(1.07(. it would be necessary to borrow $3. therefore.983. Forward Contracts. and simultaneously writing a European put. we will sell short a T-bill futures contract with a $1.49 Pre-tax profit $ 2.

610 = St (1.08 mm) .3 + 421. Hence the carrying costs and convenience yields should be the same for your expected 3-month holding period. Note that this is not a hedged position. and set the result equal to zero. you can sell short stock index futures contracts. losses in your portfolio will be offset by gains in your stock index position (and vice versa).884. This is a cross-hedging problem. but would be offset by gains from a short position in T-bonds. From Chapter 5. An effective strategy.4 contracts Therefore. Since T-bonds are positively correlated with the market value of loans. If the yield spread narrows you will lose money. If the market index falls.02)(−900 mm) − .08(.020201) $. 10.084970 Consequently. the optimal hedge is to go short 1. you will make a profit.08) – Ste (3/12)(.08) $6. the difference in the values of the two contracts is t F3/88 – tF3/87 = Ste (15/12)(.08(. take the derivative with respect to the number of futures contracts.684 per ounce.306. if the spot price of silver is greater than $4. the market value of fixed rate loans falls.96 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Fourth Edition 9.9 = –1. (8. If the two contracts are highly correlated. we have N=– (. This effect is counterbalanced by the positive correlation between T-bonds and demand deposits (a liability).684 . The general idea is to minimize shareholders’ risk by augmenting their current position with T-bond contracts. .4 T-bond futures contracts. write out the variance of the shareholders’ position. a short position is required.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.06)(820 mm) (. 12. The risk of changes in expected convenience yields and carrying costs beyond the March 1987 contract expiration date are assumed to be small. you bear little risk until March 1987. (b) The profitability of your position depends on the spot price of silver. 11.398 = St = $4.F i 'i F !F where: N = the optimal number of futures contracts ρi.008 – $5. As interest rates rise. According to Eq.15)(. If so.30)(.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. To alleviate timing or market risk. you are planning to close out your positions in March of 1987 when the shorter-lived contract expires.08 mm) = –2.105171 – 1.884. (a) Presumably. then close out your position after the yield spread widens. similar to what was covered in Chapter 5 (Portfolio Theory). To find the optimal number of contracts. the result is N = −∑ − i =1 N ρ i. would be to sell long-term T-bill futures and buy shortterm T-bill futures. called a time spread.

16 1 1 = = 0.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 1.86 u 1.16 Figure S9. what is the project’s NPV without the option to expand? NPV cannot capture the flexibility of the project.1A Present value event tree .15 1 = 1. Without the option to expand the NPV of the project is 10. (a) First. The up and down movement for the binomial tree are estimated as follows: Annual Volatility (σ) = 15% u = eσ d= t = e 0.

The ROA value of the project is $11. we must estimate the NPV of the two mutually exclusive projects.98 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. The real option to expand turns out to be valuable because the project can be optimally expanded under certain scenarios. (a) First. which is higher than its NPV of $10 Figure S9. The NPV of both projects is $20 Figure S9.53. Fourth Edition (b) What is the project’s ROA value with the option to expand? (Prove your answer).2A NPVS of the projects and PV event trees . Present Value Event trees and objective probabilities for each project are given below.1B ROA decision tree 2.

Note that DTA discounts using the WACC even though risk has changed. the option to contract should be executed next year if their present values continuously decline. 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.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 99 Using a Decision Tree Analysis (DTA).2B Decision tree for the projects Using Decision Tree Analysis (DTA). 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. Figure S9.2C Optimal execution using DTA . With Project 2. the option should also be executed at the end of the second year if the present value is at its current level Figure S9.

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). which project should the company select? As the ROA value of Project 2 is higher and positive. it should be selected for execution Figure S9.2E Real options NPVs .100 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Fourth Edition Using a Decision Tree Analysis (DTA). 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.

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

Fourth Edition (c) Do the DTA and ROA valuation suggest the same optimal execution for the options. 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. and with ROA we correctly select Project 2.2H Project 1 and Project 2 valuations 3. Two companies are developing a 50/50 joint venture with an NPV of $25 million. One of the companies wants to buy the right from the other to acquire its 50% share in a year for $15 million. Figure S9.3A Two period per year event tree for 50% of the joint venture . The risk-free rate is 5%.102 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. The annual volatility of the venture is 20%. With DTA we wrongly select Project 1. Figure S9.

46 million Figure S9.3C Three period per year event tree for 50% present value of the joint venture . but the NPV remains the same.). Figure S9. the estimate of u and d changes. what is the maximum price the company should be ready to pay for the option? Valued with two period per year.3B PV of the right to acquire 50% of the JV (b) Given 3 periods per year.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 103 (a) Using a two period model (six months per period. the real option to acquire half of the joint venture in a year is worth 0.

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

000 tons) at the rate of 4. we have made the following assumptions: • If the mine is open to begin with.7A 8.000 tons per year. . If closed (at a cost of $30. This is a switching option. contract or keep your options open. there is no production a time zero (inventory is not reduced) and the closing cost is paid immediately.000) there is no production. • If the mine should be closed at time zero.7A shows the calculations. For example. Figure S9. Fourth Edition 7. To solve the problem.000) the mine depletes the inventory (12. so that if the mine stays open it will require 2 time periods to be exhausted.000 tons immediately.108 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. at node D you can expand. the values of these alternatives are respectively $2545. If open (at a cost of $20. Therefore. Figure S9. $1600 and $2554. the best action at node D is simply to keep your option open. it produces 4.

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. Of course. we need to construct the present value event tree for the project in operation.8B.8A. assuming the mine is open. all cash flows (FCFC) and the present value (PVC) are zero. Figure S9. if the project is not in operation.8A Price and cash flow event tree Since all decisions are based on value. as shown in Figure S9.8B Value event tree (open mine) . Figure S9.

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. 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.110 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Fourth Edition The Solution Logic is the following 1. Identify optimal action for each node on the event tree entering in “Open” and “Closed” mode. starting from the back of the tree 2.8C Present value assuming the mine starts in a “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. 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.

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.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. changes the optimum (enter open) then works back to period 1—using a “backward forward” algorithm Figure S9.

112 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.8F. assuming a current price of $26 per ton.8G Value tree . Figure S9.8F 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. 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. If the project is not in operation. Figure S9.

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

Fourth Edition The higher price per ton changes the optimal starting mode of operation to “open” and raises the value of the project to $67.8J New solution.114 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. give price of $26 .370 Figure S9.

7% Figure S9.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 115 9.9B Equity valued as on option on the levered firm . This is an example of a simultaneous compound option.9A.9B values the equity. The binomial event tree for the underlying unlevered form is given in Figure S9. Figure S9. therefore the present value of the debt is $516 and its yield to maturity is 15. The value of the equity is $484. Figure S9.9A Binomial event tree for the unlevered firm The equity of the levered firm is a call option on the firm. Its exercise price is the $800 face value of the zero-coupon debt and its maturity date is 3 years.

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

128) = 305 305 =1 m= 305 mu3 V + (1 + rt )B = 633 1.433 − 800 = 633 mu 2 dV + (1 + rt )B = 1.433 − 1. Because the option is worth $29 more if it is not exercised. no investment will be made at node E.433 −833 = = −771 1.08)B = 633 B= 633 − 1.271) − 771 = 500 at node E. if we use the replicating portfolio method.10b shows the valuation of the second option.08 1. the option is worth m u 2 V0 + B = 1(1. is $1.Chapter 9 Multiperiod Capital Budgeting under Uncertainty: Real Options Analysis 117 Figure S9. we have mu3 V + (1 + rt )B = 1. The option value at nodes F and G can be calculated in the same way.08 Figure S9.10b Valuing the Second Option Therefore. m = 0.89% and B = −583. for example.128 − 800 = 328 m(1. .271 − $800 = $471. The value if the option is exercised. At node H.433 + (1. At node E.

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

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

1: V(η) = ∑ q (m) max ∑ p (e|m)U(a. e) is a2 a1 e1 .8 q(m2) = .1) (0) = − .18 e 2 −1. price information that enables you to forecast future prices with better-than-random accuracy. all relevant publicly available information has not been instantaneously incorporated into stock prices.9 e 2 .18) + (. strong form efficiency is refuted but semi-strong and weak efficiency are unchallenged. benefits from insider information are accommodated under this hypothesis. In the example given. e) − V(η0 ) a e where m1 = favorable credit check m2 = unfavorable credit check q(m1) = .e) 1 1 = (. of not granting a loan when it would have been paid.5 If the message is m1 : for a1: for a2: ∑ p(e|m )U(a .1.162 .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 .062 ∑ p(e|m )U(a 1 2 . we need to find optimal actions given a specific message.120 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. (d) The semi-strong form of efficient capital markets assumes publicly available information is instantaneously impounded into prices. 3.18 −. e) a m2 .00 0. –18 percent.18) + (. In order to find the correct values for the last sum in equation 10.e) = (.5 . Fourth Edition (c) Semi-strong (as well as weak) capital market efficiency is contradicted. Thus.1 The benefit function. The value of credit check information can be calculated according to equation 10. You have discovered a trading rule based on past.00 Notice the benefit function includes the opportunity cost. which are calculated according to the formula max ∑ p( e|m) U (a. U(a. thus.9) (. nearly costless.9) ( − .1) ( − 1. where apparent abnormal profits were available to insiders.00) = .

6(–10) + .5) (0) = −.18) + (. in the absence of information.6(0) + . and –. If we decide to test market the Kidwich product we can determine the optimal action given the outcome of the test marketing experience.3(0) + . Finally.1(0) + 0(0) = 0 V (nationwide|no acceptance) = .2 (–. If the message is m2 : for a1: for a2: ∑ p(e|m ∑ p(e|m 2 2 )U(a1 .2 percent.18) + (.16 – (–3. with an expected payoff of $5 million. and the value of the optimal actions with information.5) ( − .18) (0) = –14.24% value of a2 = (.00) = −. the optimal action is a1 (grant the loan) for an expected return of 6. we need to estimate the value of the optimal action without any test market. for an expected loss of 3.062) + .16% Since the cost of information = 5 percent.9(–10) + .18) (–1.76% Therefore. 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 = . we know ∑ p(e|m) U(a.8 (.3(10) + .5) (. Recall that the loan program has been losing money. e) is . 4. Substituting these values into equation 10.4 percent.09 given m2. Without information value of a1 = (.41 )U(a 2 .24 percent. This represents the optimal action given that the firm does not use a test market.9(0) + .1 as follows: V(η) = . E(R) = –9 percent.82) (.062 given m1. If the experiment predicts no acceptance for Kidwich the optimal action is to cancel the product. given m1.18) + (.82) (–. Of these two actions it is best to go nationwide.Chapter 10 Efficient Capital Markets: Theory 121 Therefore.09 Therefore.1(80) = 5.00) = –3. calculate the difference between the value of the optimal action without information. the credit check should be implemented. First.e) = (.1(10) + 0(80) = –8 .18) + (. and the value of information.e) = (. = 3. (a) If the firm is risk neutral it will maximize expected pro-profits.5) ( − 1.24) = 6. With information From previous calculations. The expected payoff from going nationwide (without any test market) is V(nationwide) = . the bank would choose always to grant the loan (a1). the optimal action given m2 is a2 (refuse the loan) with an expected opportunity loss. V (η) – V(η0).1(0) = 0. The calculations are shown below: V (stop| no acceptance) = .09) = 3.

3(10) + .398 0 2.511) = 1. The value of the optimal action with no information at all was $5 million.6(0) + .3(3.511 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.3(3.1(0) + .398 3. 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 the experiment predicts success the optimal action is to go nationwide.3(2.3(22) + .1(0) + .398) + .1(0) + .3(0) + .1(4. expressed in utiles. it is interesting to see how risk aversion might affect the value of information. Finally.045) + . at least for this simple example. V (stop|marginal) = .045 2. Fourth Edition If the test market indicates marginal acceptance.398 U (nationwide|success) = .511) = 3.122 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.6(80) = 50.6. Given that U(W) = ln (W + 11) the payoff matrix.” you are correct.398 4. Nevertheless. therefore the value of the test market information is V(η1) – V ( η0) = 11. The cost of the test market is $5 million therefore the net value of the information is $1.6 million.6(0) + .6(2.6 million. .398) + . V (stop|success) = .7(0) + .2(80) = 22.365. U (stop|success) = . The major difference is that the payoff matrix must first be converted into utiles so that expected utility can be computed.7(10) + . U (stop) = .398) + .398) = 2.1(50) = 11.620. the optimal action is to go nationwide.398) = 2. as shown below.1(2.3(2.2(0) = 0 V (nationwide|marginal) = . (b) If you said to yourself “There’s no such thing as a utility function for a firm. The value of information is defined as V(η) = ∑ q(m) MAX ∑ p(e|m) V (a.1(2.1(–10) + .398) + .045) + . On the other hand. the optimal decision is to go nationwide. if there is a test market and if it predicts success.6(0) = 0 V (nationwide|success) = .6(2. is Action Cancel Go Nationwide 2.6 – 5 = $6.6(4.398 U (nationwide) = .1(–10) + .

over a period of time. Costs include the cost of gathering information and a fair rate of return on physical and human capital.Chapter 10 Efficient Capital Markets: Theory 123 If the experiment predicts marginal success the best action is to go nationwide. on the average.3(3.. There is nothing in the efficient markets hypothesis which says that arbitrageurs cannot make profits. However.045) + 0(4.398) + 0(2. U (stop|marginal) = . more individuals would enter the arbitrage business until it became true.2(4.7(2. 6.620) = 2.398) = 2. U (stop|no acceptance) = . If options are expected to expire unexercised 80 percent of the time.7(3. Because the test market costs $5 million. 5. It would not be undertaken in the risk average case because its net value is $4. And if the experiment predicts no acceptance it is best to cancel the product.711.04 million.045) + .1(0) = .9(0) + .1(3. If this were not true.305.398 U (nationwide|marginal) = .6(2.e. The corresponding level of wealth is V(η2) = $4. at the margin.04 million.398) + .398) + . 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) = . The options market would be a martingale if expected returns were equal to zero.96 million. 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.04 – $5 = $-.04 – 0 = $4.398) = 2.398) + .398) + . cancelling the product) has a value of zero. revenue from arbitrage just equals costs. (b) The options market is a fair game so long as investors’ expected returns are equal to their long-run average returns. And since the optimal action given no information (i. If our expect to lose 83 percent of the time and if. the game has been fair. One would expect that.034. (a) A fair game means that.1(2. you do in fact lose 83 percent of the time.2(2.511) = 3.034) + . outcomes conform to expectations. Martingale: [Prob (loss)] ⋅ [Expected value (loss)] + [Prob (gain)] ⋅ [Expected value (gain)] = 0 .1(2. V ( η0)=0. it is important to look at their net profits after costs.1(3.511) = .9(2. the value of information is V(η2) – V(η0) = 4.398 U (nationwide|no acceptance) = .

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. Assuming that the market portfolio is known to all investors so that systematic risk can be appropriately measured. In an efficient securities market. 9. then every investor will sell B and buy A. It reflects one of the “weaknesses” of equilibrium models like the CAPM or the APM. If we assume a is some positive amount. Fourth Edition In this example. but we cannot determine whether or not the value of the firm reflects excess profits above a competitive return in the product market. The second-order stochastic dominance criterion assumes that all individuals have positive but decreasing marginal utility. i. This has the effect of altering the expected distributions of returns until there is no second-order stochastic dominance in equilibrium. a martingale holds if –(. if the assumptions of the CAPM or the APM are a reasonable approximation of reality.2a > 0 This result is a submartingale with E(R) > 0.2) L G = 4L for L>0 If the expected NPV of a gain were more than four times the value of an investment loss.e. The net result is NPV = 0. 8. namely that the starting market value of the firm reflected no excess profits (or monopoly rents). it cannot tell us whether or not the initial market value of a firm reflects abnormal returns. But the higher price for A (and lower price for B) implies lower expected returns for A (and higher expected returns for B). since all securities are priced according to their perceived appropriate risk.2) (G) = 0 G = (. then the options market would follow a submartingale. driving up the price of A relative to B. 7. that although the model may be used to test for abnormal returns in an efficient capital market. whether or not product markets are efficient.2)(4L + a) – (.. the NPV of any security is zero. This is just another way of saying that you get what you pay for. they are greedy and risk averse. Thus. then it is possible to measure ex post excess returns via the CAPM or the APM.124 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. we can test for excess profits to investors in the capital market. Given a world with investors of this type. until the market price of A relative to B is high enough to make the marginal investor indifferent between A and B. namely.8) (L) + (. if asset A dominates asset B and if they have the same present value. In other words. .8)(L) = . However. then E (R) = (. If we make a second assumption.8/. the second assumption is unrealistic. This will continue. and NPV’s are discounted according to (the same) appropriate risk.

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. Roll’s critique (1977) is based on the assumption that capital markets are in equilibrium. (b) Information is material if it has an impact on securities prices when it becomes publicly available for the first time. then sells at the closing price. Before the disequilibrium situation as well as afterward. Should an individual—for example. (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. However. Ibbotson’s results can be interpreted either as 1) selection of an inefficient index. If it has no impact on prices. In this way. 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. 2. This is not a deficiency in the securities market. but rather a deficiency in the market for information. it is largely irrelevant. 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. All that is required for efficient securities markets is that prices fully reflect all available information. Therefore. . he can earn a risk-adjusted abnormal return after transactions costs. (a) As shown by the Ball and Brown study (1968). Given such a situation. However. although it may cause portfolio adjustments that leave prices unchanged.Chapter 11 Efficient Capital Markets: Evidence 1.56 percent trading rule of Dann. Mayers and Raab) where he purchases shares at the block price. 3. Or suppose that a new security is introduced into the marketplace. a corporate insider—have monopoly access to valuable information. If an individual can employ a strategy (using the –4. then the market cannot reflect the information because it is not publicly available. The empirical evidence on block trading is consistent with the semi-strong form of the efficient markets hypothesis. 4. Roll’s critique applies and we cannot expect to observe any abnormal performance relative to an efficient index. (a) The securities market can be efficient even though the market for information is not. because once the block trade becomes public information. it is not possible to earn an abnormal return. the adjustment process itself can be used to detect abnormal performance relative to the market index prior to equilibrium. or 2) detection of abnormal performance during disequilibrium. the abnormal performance of an asset can be measured by the arbitrage profits available as its price is adjusted to a new market equilibrium. almost all of the “information” in an annual report has already been discounted into the security price before the annual report is released. The evidence may or may not be inconsistent with the strong form of the efficient markets hypothesis. the “abnormal return” is defined relative to the risk-adjusted return which would be predicted by the capital asset pricing model. it is unlikely that the latest copy of the annual report will allow an investor to earn an abnormal return. as was the case of new issues studied in the Ibbotson (1975) paper.

If X were the amount extorted by the founders of the Ponzi fund. If there is a 50–50 chance of “winning” in the market during a year’s time. his first dividend would be paid out of the money he put in.000977 10 If there are 10 million investors in the market.000. It would operate as follows: If the first participant put in $1. by the end of the third month there remains a group of 10. and $20 per month was paid as “dividends. as with block trades.000 investors who have received three successive correct “predictions. 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. If no one else joined the “fund. Masulis and Titman [1984]).000(1 + Rf) T where T = time until declaration of the Ponzi Fund’s bankruptcy. A should not purchase the service. Mr. And finally. splits per se can be interpreted as good news about the future prospects of the firm. Thus. Ninety thousand are told the market will move up by more than 10 points. he would have $1. 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. Brennan and Copeland [1987] model splits as costly signals to show that they can convey information in a rational expectations equilibrium. However. 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 in the risk-free asset. If he purchased it. The investor’s loss can be determined as follows: If he had originally invested his $1. the probability of wining every year is (1/2) ≅ . then over a ten year interval. Consider the following scheme.000 − X ≤ 50 20 If the investor participated in the Ponzi Fund. then approximately 9. a large new issue is also a large block.” the first individual would be paid $20 per month out of his original $1. This would take 50 months if the perpetrators of the scheme kept their portion of the capital equal to zero. ninety thousnd are told it will stay within a 10 point range.000.770 of them will have beat the market every year for ten years. 5. 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. (d) As indicated by the Ibbotson (1975) study. After all.” the time until bankruptcy would be T= 1. 6.” The plan described has paid out 2 percent per month for 18 months.000 investors. A solicitation is mailed out to 270. until the money was gone and the fund was declared officially bankrupt.” 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. Ponzi frauds follow the adage “you can fool some of the people some of the time. his earnings would be 20(1 + Rf) T–1 + 20(1 + Rf) T–2 + · · · + 20 .126 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. he could be the victim of fraud. and the remainder are told it will fall by more than 10 points. After the first month the group of ninety thousand which received the correct “forecast” is split into three groups.

then one would expect it to have higher returns also. have not failed. of course. 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. Mutual funds do. If investors are rational. (d) The evidence is not creditable because the model has not been tested on independent date. Firms continuously listed on Compustat tapes from 1953 to 1973 are firms that. his total loss would be –$1. 1971. Fourth. T. April. then. the chart does not say whether or not the returns recorded there include dividends as well as capital gains. 20 = 20(1 + T–T Rf) . is reported in The Journal of Business. in fact. The major problem is that the portfolios’ performances in Figure Q11. It simply doesn’t make any difference. since the records of bankrupt firms are completely deleted from the tape. If portfolio 1 is riskier than the other portfolios. it would quickly go bankrupt. The flows of dollars which they provide to mutual funds will not affect the way in which security prices adjust to new information. any differential effects of taxation are unaccounted for. (c) The results of this research provide evidence against capital market efficiency because past. Therefore. If the information is predictive. Thus. In this instance. publicly available data provides predictive power for future price changes. it could go on indefinitely. But it isn’t.Chapter 11 Efficient Capital Markets: Evidence 127 where the first term represents the earnings from his first dividend paid. at any point in time their choice of a mutual fund should be a matter of indifference. and the last term. (a) A selection bias discredits any results implied by this study. It would be surprising if their preferences can be used to predict rate of return differences. brokers can beat the market. What should be reported is risk-adjusted abnormal rates of return. (b) It is not surprising that stock brokers’ preferences are correlated with rate of return differences across industries.8 have not been adjusted for risk. then they should know that last year’s performance is typically unrelated to this year’s performance. if enough new people joined the Ponzi Fund over time. by Niederhoffer. Either way. An individual investor who could accurately predict the time until bankruptcy. 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. 9.000(1 + Rf) + T ∑ t =1 T 20(1 + Rf) < 0 T–t However. The relevant test would be in the model’s predictive ability. When the scheme could not attract new participants. 8. their behavior is not inconsistent with efficient capital markets. attract more customers in a year which follows abnormal performance. the risk involved is his ability to predict the fraud’s demise. capital gains plus dividends. . 7. Third.e. by definition. A study of this nature. the fund’s investors lose money because no investments in real assets are being made. net of taxes and transactions costs). A second problem is that returns are reported without subtracting the transactions costs involved in readjusting the portfolios each week. Any model can be worked over to fit a set of data if enough variables are allowed. The mystery is why. In the aggregate.. represents his last dividend. on average.

2)(150) + (0. the minimum cost of false signalling should be 100.2)(150 − C) = 80 − 0.2)(250) + (0. greater than bad firms with value Vb = 150. The compensation of the manager of a bad firm is given by (0.36 1 + r 1.1 V1a 250 = = 227. Based on the signal (debt level) chosen by the manager of a good firm. If there is no uncertainty in the market and pricing is risk neutral. investors perceive the firm to be good and if D ≤ D*. good firms have a value.2)(150) + (0. the time 0 values of the two types of firms would be given by V0a = and V0b = V1b 150 = = 136.2)(150) = 60 if D ≤ D * The manager of a bad firm will have the incentive to signal correctly if 80 – 0. her compensation would be (0.2C if D* < D ≤ 250 B M (D) = (0. γ1 = 0.27 1 + r 1.Chapter 12 Information Asymmetry and Agency Theory 1.2)(250) = 100 if D* < D ≤ 250 A M (D) = (0. In this question. investors perceive the firm to be bad. Therefore. 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.2. at time 1.2)(250) + (0.2C < 60 or if C > 100. Va = 250. she will give the correct signal. since the marginal payoff from telling the truth is greater than that from a lie.2.1 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. that is 100 > 80. Further assume that if D > D*.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. Assume that D* is the maximum amount of debt a bad firm can carry without going into bankruptcy. In this case. . V1 is the value of the firm at time 1.

V(D) schedules are D*(t) = (0. D*(t).2 A = – 1.70 and V[D*(t)] = (0.5 1. The equilibrium dividend.2 A = – + 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 1.4 0.25) 2 (1.8) 0.3) 2 (1.5)(0.4) 0.84 If the shortfall cost is 70%.25) 2 (1. and value response. and value response.8412)(500) = 420.4) = 0.4 1 + (0.59 and V[D*(t)] = (0.25 1.25 1.25 1. D*(t).7 1. V(D) schedules are D*(t) = (0.7494)(500) = 374.8 .3 + (0. and value response.2) 2 The equilibrium dividend. V(D) schedules are D*(t) = (0. 5 The equilibrium dividend.3 1.4 The equilibrium dividend. β = 50% and t = 500.2 0.5)(1. τP = 25%.25 + (0. (0.7494 (0.8 0.7)(0.4 + A = – 1+ = 0.4) 0.5 1. Therefore.7494))(374.25 1.7869))(393.7347 (0.7869)(500) = 393.25 1.5)(1.4 0.2) 2 0.45 and V[D*(t)] = (0.5)(0.25 + (0.05 If the tax rate is 30% (0.5 1.4) 2 0.7 1. D*(t).8412 0 .8412))(420.4 + 0. 0. and value response.23 If the discount rate is 40% (0. β is the shortfall penalty and t is the upper bound on the cash flows.45) = 272.5 1.2 0.Chapter 12 Information Asymmetry and Agency Theory 129 2.2) 2 = 0.7)(1.2 A = – + 1+ = 0.2 0.7869 (0.59) = 282.4 1+ (0.5)(1.25) 2 (1. In this case.70) = 290. r is the after tax rate of interest. r = 20%.4 0.3 1.5 1. D*(t).

Now consider the situation in which a medium firm issues straight debt. medium and bad). consider a situation in which the bad firm issues straight debt. For example. The firm has three financing options at time 0—straight debt that matures at time 2. it issues debt with a true value of qI + (1 – q) bL. The expected costs of financial distress are (1 – p)c. the medium firm bears the cost of underpricing if it issues equity but gets no offsetting gain. and value response. 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. Therefore.5)(0. If it does so. that is. the value of the probability q is updated to either 0 or p.7347))(367.79 3. that is. To see this. .25 + (0. it issues debt with a true value of pI + (1 – p)bL.130 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Since the bad firm raises an amount I from the debt issue. the firms should not want to − p)bL mimic each other. the debt is overpriced by I – [qI + (1 – q) bL] = (1 – q) (1 – bL). Fourth Edition The equilibrium dividend. Equity issuance makes sense for a bad firm since the equity issue helps it avoid the costs of financial distress. the firm can force conversion with certainty and is never left with any debt. 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.7347)(500) = 367. bL < K < I and is I convertible to pbH + (1 of equity. bad firms issue a fraction qbH + (1 − q)bL of equity and medium firms issue convertible bonds with face value F > bL. this implies that (1 – q)c > (1 – q) (1 – bL). Each firm has access to a project with required investment at time 0 of I and expected net present value of B. convertible debt that matures at time 2 but can be called to force conversion at time 1 at a predetermined conversion ratio and equity. Stein considers a three-date model (time 0. At time 1. The discount rate is assumed to be zero. the debt is overpriced by I – [pI + (1 – p) bL] = (1 – p) (I – bL).33 and V[D*(t)] = (0. it does not pay for a bad firm to mimic a good firm. the firm type is revealed and for bad firms. For this to be a separating equilibrium. Firm types are private information at time 0 and the true value of bad firms is volatile between time 0 and time 1. The costs of financial distress are such that c > (I – bL). On the other hand. Finally. note than the medium firm knows that the conversion value of the convertible bond will always be I at time 1. the expected cost of financial distress is greater than the underpricing. If it does so. The expected costs of financial distress are (1 – q)c. Since it has been assumed that c > (I – bL). V(D) schedules are D*(t) = (0. the expected cost of financial distress is greater than the underpricing. consider the situation in which a medium firm chooses to issue equity. a call price K. medium types receiving bH with probability p and bad types receiving bH with probability q (q < p < 1). Firms differ in the ex ante probability of receiving bH with good types receiving bH with certainty. The probability of deterioration is assumed to be z and for consistency q = (1—z)p. D*(t). this implies that (1 – p) c > (1 – p) (I – bL). Debt financing is associated with a potential for costly financial distress where a deadweight cost of c is imposed on the owner-manager. it does not pay for a medium firm to mimic a good firm. 1 and 2) with three types of firms (good. Since this conversion value always exceeds the call price K. Since it has been assumed that c > (1 – bL).33) = 226. all agents are risk-neutral. Since the medium firm raises an amount I from the debt issue. Under these circumstances. there is a separating equilibrium I in which good firms issued debt with face value I. Each firm receives a cash flow from the investment of either bL or bH at time 2 (bL < I < bH). Therefore.

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

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

V is the agent’s utility function. λ is ∂f(s | a) a constant. f(s | a) 1 fa (s | a) γ (δ0 + δ1c) = λ + µ f(s | a) . c is the contract between the principal and agent. the principal is risk neutral. 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. f( s | a ) is the density function for the outcome conditional on the effort. ’ e e − (s −c)/α p − c/α A α α = λ or e A s + c =λ α αAαp where α = αA + αP. s is the final outcome. 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. a is the agent’s action. V(c) = [1/(1 – γ)](δ0 + δ1c) In this case. W ∂W In this question. p is a performance measure. V (c) = δ1/(δ0 + δ1c)γ and the sharing rule can be written as: ’ (1–γ) and fa (s | a) is linear in s. 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. It is linear if γ is one and concave if γ is larger than 1. if 0 < γ < 1.

Given a tenure of T and an opportunity cost of i.44 37. Fourth Edition 6.91 25.29 25 . the manger will use a hurdle rate of r* for project accept/reject decisions. Manager Tenure (T years) 2 5 10 15 20 40 Project Hurdle Rate (r* percent) 69.01 25.18 28. 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).132 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.

) Table S13.000.000 $0 $300 $300 $300 — (a) Bottleneck 10% $10.000 0 — — — $11.2 Data for Q13. (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).000 as found in early printings and the corresponding sale price is $11. and Incentive Design 1. 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. Assume that the cash flows of the firm are a constant perpetuity.2 Base Case 10% $10.2. therefore the economic profit approach is isomorphic to the DCF approach (before or after taxes).000 400 500 500 500 — (b) Cut Costs 10% $10. not $1. (Note that I0 = $10. and that E(ROICt) ICt −1 = E(EBITt).Chapter 13 The Role of the CFO.000 0 350 350 350 — (c) Sell for 10%Premium 10% $10. 2.000 WACC I0 ∆I0 ∆CF1 ∆CF2 ∆CF3 Sale Price .000. 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. Let’s take a look at the data provided in Table S13. Performance Measurement.

134 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.000 as after tax proceeds).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. therefore the NPV is NPV(cost cutting) = $50 $50 $50 + + = 45 + 41 + 38 = 124 > 0 1. Note that this is a lower business unit ROIC but higher NPV than debottlenecking. rises to only 3.000 = 3% in the base case and $500/$10. traditional NPV would cause us to accept the project although the ROIC.12 1.000 . Its net present value is NPV(debottlenecking) = −400 + 200 200 200 + + 1. By way of contrast. (c) If the sale takes place immediately and is not taxed (or if we interpret $11. then this is by far the best alternative with NPV of $11. if the cash flows of the other two alternatives were perpetuities their NPVs would be NPV (debottlenecking perpetuity) = −400 + NPV (base case perpetuity) = 500 = 4.1 350 NPV (cost cutting perpetuity) = = 3.1 1.000 (remember that the basic value of $10.13 Again.12 1. It will require an incremental investment of $400 and will bring in additional cash flows of $200 for three years.1 NPV (sale) = $11. (b) The second alternative is cost cutting. originally at 3%.000 .000 is a sunk cost).400 = 4. Fourth Edition (a) The first alternative is debottlenecking.5%. The incremental investment is zero and incremental cash flows increase by $50 per year.600 .8% after the debottlenecking.1 300 = 3.1 1.500 .

. Let’s assume that ROIC is after tax.e. yet ROIC declines every year because incremental return on new capital is less than the cost of capital (assumed to be 10%). 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.5 to $4.5 million. The alternative is to announce that Project B will not be taken.5 million $1.3% per year.063 548 –71 477 –238 239 3221 8.0 million $13.2% 7. invested capital and the cost of capital.5 million 1. 5.0% 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 4.331 –839 492 –43 449 –224 225 2662 9.5 million $10. they also have a very low correlation with TRS. Table S13..8% 620 4 1.8% 3.100 –675 425 –20 405 –202 203 2200 9. Project B should be taken.3 ROIC falls while the Top and Bottom Lines both Grow 1 1. The facts are summarized in Table S13. its stock price will fall because the expected total EBIT will fall from $13. Table S13.0 million $9.464 –951 513 –56 457 –228 229 2928 8. In this alternative the shareholders would be presumed to earn the opportunity cost of capital. Consequently. net income grows at an average of 3.4% 6.0% — 200 2 1.3 provides an example where revenue growth averages 8.5 to $10. If so.5.611 –1.4%.6% 862 5 1.0 million 3.5 million 1. Any new investment should be accepted when its expected return in greater than the cost of capital. New investment that earns less than the cost of capital destroys value. but omit expectations. Table S13.000 –600 400 –10 390 –195 195 2000 10.5% 6.3% 400 3 1.5 million $12.7% 6. the stock price would fall even more because expected earnings would fall from $13.5 million 12.9% 1128 6 1.5 million $4.5 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 $1.5 million. Therefore.0 million $13.210 –755 455 –31 425 –212 213 2420 9. 10%.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.Chapter 13 The Role of the CFO. i. and Incentive Design 135 3. EVA® and the change in EVA® contain information about earnings. Performance Measurement.

6 Tax considerations are not discussed in the answer to this problem. Fourth Edition 6.136 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. However. Figure S13. For tax policy see Chapter 14. The 3 alternative types of compensation are illustrated in Figure S13. Fixed salary provides a floor and non-salary incentives all provide gains that increase with performance. The structure of all three incentives is roughly the same.6 below. the graphed relationships break down if the measure of performance is not highly correlated with the stock price. . Panel C is the only one that stipulated performance is based on expectations and that will be correlated with TRS.

Therefore. and they earn a positive spread that creates value for shareholders. These liabilities provide (insured) liquidity service to customers of the bank who demand less than their opportunity cost. (a) Banks are different because of their franchise to provide demand deposits and certificates of deposit.” In this case you are assuming that the excess cash is used in a share repurchase program. However. future amounts are not because they are assumed to have zero NPV. (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. Although the current amount of excess cash and marketable securities is added to the DCF of operating cash flows. The value of equity is then derived by subtracting the market value of debt from the enterprise value.9% of current assets and by 2010 it is 7.7). This entity approach assumes that the act of issuing debt (or not issuing it) is independent of the forecast of operating free cash flows. 2. it is assumed that (assuming trivial transactions costs) the firm receives $1 of cash for each $1 of debt that it issues. One is to forecast growing dividend payments. the equity approach is also appropriate for valuing insurance companies.165 = 45. is combined with long term debt in Table 14. called unscheduled debt. This stream of free cash flow to equity is then discounted at the cost of equity.8 and subtract an amount from excess cash each year and add the same amount to “decrease in common and treasury stock. (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.Chapter 14 Valuation and Tax Policy 1. For this reason we recommend an equity cash flow approach for valuing banks. Free cash flow to equity include after-tax interest income and expense. The standard industrial valuation model (see Chapter 14) discounts free cash flows at a weighted average cost of capital to value the enterprise. (a) In 2003 excess cash is 472/6. Consequently these liabilities are an important part of the operations of the bank. Simply. and increases in deposit liabilities. 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.736/17. that has already been done in Table 14. . and that (other than providing an interest tax shield) no shareholder value results from issuing debt.1%. The second approach is to go to financing flows in Table 14. (b) There are two ways to bring excess cash into line.6 because the forecast has assumed that dividends are 65% of net income.830 = 6.

1% for the next 10 years then g = 5% thereafter.553.680—a small difference. The spreadsheet assumed r = 30% forever.083 + .33) = .138 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.2% whereas the spreadsheet assumes g = 12. not large enough to explain why the spreadsheet indicates that V0 = $128.52(1 − .35 for N = 10 years.8 in 2000 to 18. (a) Equation 14.265 = 31.000(1 − . Table S14.52 (1 – .837 = 45.3a Multiple Calculations Entity Value EBIT Multiple Table S14.3b 2000 $128.716 + 2.083) .716 + 2. .33)] + 10 .7 Averages for 2000–2010 NOPLAT Growth ROIC 12.1%/year 34.21 is the finite supernormal growth model: V0 = EBIT1 (1 − Tc ) r − WACC + Tc B + K[EBIT1 (1 − Tc )]N ku WACC(1 + WACC) (14. but more than halves from then to infinity.33)(. Fourth Edition 3.91 3. What creates the difference are the following: 1) First.000) + .3b compares the forecast Table S14.090 = 31.0845 .5 billion. These assumptions are consistent with a multiple that declines from 34. K = g/r therefore g = rK and g = . we get an estimate of continuing value as of year 10 V0 = 4. (b) The spreadsheet value estimate used EBIT (1 − T) = $2.15[4.80 10.7 in 2010.487 in 2000 whereas the formula above assumes EBIT (1 − T) = $2.1% during the explicit forecast.9 billion and the finite supernormal growth formula has V0 = $45.33(6.083(1.8 Assumptions Perpetuity Assumptions g r 5%/year 30% 2010 $194. Note that the NOPLAT growth rate is 12. 4.44 18.15) = 5.3a provides the calculations of the multiples and Table S14.52 (1 – . 2) Second the finite supernormal growth model assumes ROIC = .000 + 11.33) . The ROIC also comes down—but it is assumed to remain well above the cost of capital.4% assumptions during 2000–2010 with those of the perpetuity (years 11 and on).000 + 4.020 . The WACC assumptions are the same.000(1 − .33) − .21) Substituting in the suggested parameters.71 34. then declines to equal WACC (on new capital invested).

15 = 100% and the new valuation becomes infeasible because WACC – g = . Because we have NOPLAT in the 10th year. 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 ).g.1)(1 − .169 + 4. that ROIC goes up to achieve the higher growth that K does not change.667(1 − Tc ).040. in this case.15[4.33) + .8 = = = 9. so let’s use the finite supernormal growth model assuming that N = 4 years.000 + 14. the firm cannot grow at 15% forever.2% to 6.040(1 − .083 4.2% and assuming that K remains at 15%.10 . the new retention rate will be K = g/r K = .083(1.040(1 − .0845 .667(1 + .10 / .12 − .667(1.33)]10 .12(1.02 . If so r = g/k r = .21.12) = 9. . While possible for some companies to have negative working capital (e.4)]4 .4%.138 .1)(1 − .15/. and ROIC stays at 12%.1)(1 − .12 = 1.033 + 2.10 and r = .322 = 13.906 = 48. let’s assume that an extra revenue growth of 1% translates into 1% EBIT growth.00 for 4 years then r = WACC thereafter. Tc = 0. Growth in NOPLAT rises from 5.25 = 125% This implies that the company can grow with negative capital requirements. Thus.167) 182.15[1. we have V0 = ..02 If the growth in NOPLAT increases from 10% to 15%.15 < 0 in the denominator.6)(.12 . N stays at 10 years. EBIT.15/.667(1. Note that it is not appropriate to compare the result of the value driver formula (that assumes g = .000) + . 5.138.413 − .12 – .000 = 1. 10. magazine publishers who receive payment before delivering the goods) it is more likely. This is an increase of $3.4) 1.12 forever) and the result of the finite supernormal growth model (that assumes r = 1.386 million or 7.3% after tax. ROIC rises from 35% to r = g/k = .Chapter 14 Valuation and Tax Policy 139 (c) To simplify. becomes $4. Clearly.062/15 = 41. 14.00 − .4 = 1.12) 1. V10 = V10 = EBIT11 (1 − Tc ) r − WACC + K[EBIT11 (1 − Tc )]N WACC WACC(1 + WACC) 1.4)(1 − .834(.939.33(6. Substituting these into Eq.083) = 32.12 + .

3. 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 . (a) Following an approach similar to the Modigliani-Miller derivation of equation 15. and with the same assumptions. the total cash flows paid by the firm to shareholders. 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. 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.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 1.

k0.06 .30 = ρ + (ρ − .5 . k s = R f + [E(R m ) − R f ]β = . its unlevered β and its 100 percent equity cost of capital will be the same as that of the project. the weighted average cost of capital.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. For Carternut (assuming it has risk-free debt).30) 9ρ = + ρ − . The object is to find the appropriate weighted average cost of capital for the project assuming that it has 50 percent debt.06 = ρ = 28.18) in order to obtain the 100 percent equity cost of capital for Carternut.06) (1 − .0 = 30% P P+B+S Next.4 + . 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. since 9 9(. If so. 2.5) . B S 1 .06 + [.06]2.74% 19 k s = ρ + (ρ − k b ) (1 − τc ) B 1 = B + S 10 . X = 1 and the third term becomes τc If X = 0. Assume that Carternut has the same operating risk as the project.18 − . use the MM definition of the cost of equity (equation 15. the cost of equity is given by the CAPM.5 5. it is equal to zero.

if a more reasonable debt capacity were assumed for the project.09 = = .5 (.000 Therefore.5)) = 21.1125 1 − .000 = = . and if the project has the same operating risk as the firm but higher debt capacity.09 .8 B 1.667)) = . in order to solve for the weighted average cost of capital if the project has 50 percent debt. We need to know the appropriate weighted average cost of capital for the project.5) . Consequently. the WACC is 9 percent. . Fourth Edition Next. No project has 100 percent debt capacity.4(. accept the project since it earns 25 percent while the opportunity cost of capital.55 percent.667 B + S 1.0825 Because the project earns more than the new WACC. One should be immediately suspect that 100 percent debt financing is unrealistic. assuming the MM theorems apply. Before the project.000.13).2874 (1 − .1125(1 − .55% Therefore. B WACCτ = ρ 1 − τc B + S = .142 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. it should be accepted. the cost of equity if the firm were all equity would be B WACC = ρ 1 − τc B + S (15. then it is reasonable to expect the project to have a lower WACC than the firm. is only 21. WACCτ. Therefore.13) ρ= The new capital structure is .000 of debt to the firm’s capital structure.4(. (Note: There are at least three other ways to obtain the same answer. However. The treasurer’s reason for accepting the project is highly suspect. use the MM definition of the weighted average cost of capital (equation 15.500. a better approach is to calculate the new weighted average cost of capital for the firm after the addition of $500.) 3. the new WACC is WACCτ = .

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. as shown in Figure S15.24) If the corporate tax rate decreases. both before and after the change in the corporate tax rate. . Hence. If the reduction in the personal tax on dividend income exactly offsets the increase in capital gains tax for every individual. (a) According to Miller (1977). then in equilibrium the gain from leverage will be zero. Therefore.1.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 143 4. Even if the tax effects are not exactly offsetting. the gain from leverage is (1 − τc )(1 − τPS ) G = B 1 − (1 − τPB ) (15. ceteris paribus. (a) Using the Miller (1977) argument.1 Changes in the Aggregate Amount of Debt Given a Decrease in the Corporate Tax Rate. 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. Therefore. there is no optimal capital structure is either case. corporations will repurchase debt and the aggregate amount of debt will decrease. the gain from leverage will decrease. Equilibrium will be such that the gain from leverage will be zero both before and after the change. τPS. τc 5. there can be no systematic change in optimal capital structure from the point of view of a single firm. Figure S15. (b) Given the assumptions of Miller (1977). there will be no optimal capital structure from the firm’s point of view.

Of course.06)1. in fact. assuming a market in equilibrium. Fourth Edition (b) Figure S15.06) (1) = . one might expect the reduction in the dividend tax to be overriding. For firm C: ks = . Figure S15. WACC. is the more likely effect of the proposed tax changes because the capital gains tax can be deferred.5) = .12 because B is an all-equity firm. or offset.03 .12 − . an increase in * aggregate debt from B to B2.144 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.06 + (. then we might expect the aggregate demand for debt to decline because equity capital is relatively more attractive.06 + (.06 (. (a) The cost of capital. is given by equation 15. where ks = Rf + (E(Rm) – Rf)β For firm B: ks = . 6.15 (1 − τc) kb = Rf (1 − τc) = .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.2 shows the supply and demand curves for aggregate debt.12 WACC (B) = . Hence. The result would be to * reduce the aggregate amount of debt in equilibrium from B to B1.5 = .2 Supply and Demand Curves for Aggregate Debt If the effective tax on income from shares decreases. This. we would observe the opposite result.12 − . namely.

5) = 90 Therefore.15) = .6)(.5 (.1: VU = where E (EBIT1 (1 − τc )) = 180 (1 − . VU = 90 . The income stream should be capitalized at the rate of the weighted average cost of capital for firm C. Of course.0144) = = .09.5 (.12 E (EBIT1 )(1 − τc ) ρ = 750 For firm C. the equation for the value of a perpetuity can be used. For project 1. B/S. VL = = E (EBIT1 ) (1 − τc ) WACC 90 . rmj. transform the target debt-equity ratio. u we can calculate the β appropriate to the project and use this to determine the project’s cost of equity and WACC. The value of firm B can be computed using equation 15. βu j = cov (E (R j ). to B/(B + S): B =1 S B=S B = . E (R m )) var (R m ) rjm σ jσ m σ2 m (.09 = 1.5 (.12) = .000 (b) Since we know each project’s correlation between its unlevered free cash flows and the market. since E (EBIT1 (1 – τc)) is a perpetual stream. the WACC is the required rate of return on the project. WACC (C) = .5 B+S Then.03) + .5) = 180 (.10)(.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 145 To find the WACC.

Project 2 is rejected by firm B. the required return on project 1 is equal to the expected return (.03 The expected return for project 1 is E(R j ) = = E (EBITj (1 − τc )) cost j 9 = .0675 Therefore the expected return.06] .146 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. taxes a leveraged firm will always have a lower required rate of return on projects than an unleveraged firm.12 − . In order to compute the cost of equity given that firm C uses 50% debt financing.1 and S15. but no personal.09 100 Therefore.5 [1 + (1 − . the unlevered firm. is higher than the required return.5 = .06 + [.06 + (.06).105 The required rate of return for debt financing on project 1 is Rf (1 − τc ) = .12 − .5)1] = .5) B + S B + S = . . 3. the total required return on project 1 is S B ks + kb = .03(. we need to L estimate the levered βe for the project then use the CAPM. We know that B u βL e = β 1 + (1 − τc ) S = . but accepted by firm C with 50% leverage.75 Using the CAPM to compute the cost of the levered equity in firm C. and 4 are summarized in Tables S15.09 This is also the WACC for firm B.5) + . Similar calculations for projects 2. we have L k s = R f + [E(R m ) − R f ] βe = .2.09. This is because the differential tax structure in effect subsidizes debt.75 = . . for firm B (all-equity).105(. Fourth Edition The required rate of return given 100% equity financing of project 1 is Required E(Rj) = Rf + (E (Rm) – Rf)β j = . and the project is profitable. and the project is marginally acceptable. There is a gain from leverage. In a world with corporate. For firm C.09).

B+S WACC .1950 Required Return on Debt Financing .15 Decision marginally acceptable reject accept reject Project Evaluation for Firm C B LEVERED : B + S = . Proceeding on the basis of the second assumption.1125 . is B 500.4(.2 2.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 147 Table S15. It will still have $500.4 percent.108 .09 .1050 .1320 .0916 .6416 .5.13 1 − . There are two possible assumptions about the appropriate capital structure: a) that it cannot support any debt. βL j .9624 1.1178 .250.5 Project j 1 2 3 4 βL j .75 . ρ. The second assumption is more reasonable. B WACC = ρ 1 − τc B + S Substituting in the appropriate numbers and rearranging.0675 . we have ρ= .1.03 Required Return Given B = . from equation 15. we can compute the all-equity cost of capital.03 .12 Project j 1 2 3 4 Table S15. Project Evaluation for Firm B (All-Equity) Expected Rate of Return on the Project Ñ j − τc = cost j .4 B + S 1.09 .03 .0916 . we see that after the firm takes on the project.000 . or b) that since it does not change the operating risk of the firm.000 in debt.000 = = .0739 .12 Decision accept accept accept accept 7. then this would be the required rate of return.0985 .104 = .5) If we had adopted the first assumption. namely.250. the new debt to equity ratio.13. We also assume that the MM model in a world with only corporate taxes is appropriate. which is relevant for valuing the project. If the firm’s current WACC is 10.2.0810 . that the project cannot support any debt.5 Required Rate on Equity Financing Equals the WACC .25 Required Return on Equity Financing . Therefore.5 .09 . it can support a debt to total assets ratio of 50 percent.1125 .03 .1125 Expected Return: E (EBITj (1 − τc )) cost j .000 in assets. it will have $1.8 1.

Reject the project.) E(R) = Rf + [E(Rm) – Rf] β f . is less than the required rate of 18 percent. using the CAPM for the project.05 + (.1092 8. the appropriate valuation model is VL = E (EBIT(1 − T)) (1 − τ c ) k D + τc d ρ kb since.3β f since [E(Rm) – .13 (1 – . the expected rate of return on the project.10) (1. The investment decisions of the firm remain unchanged while the financing decision increases the firm’s weighted average cost of capital.148 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. assume away the effect of growth.18 = 15% 12 12 Next. WACC = . In a MM world with only corporate taxes VL = VU + τc B Reducing leverage reduces the value of the debt tax shield. β i = 1.4)) = . The value of the firm will decline if there is any gain to leverage.05] β f = 10% Next.09 + . 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. For a numerical answer.3).05] β f [E(Rm) – . use the CAPM and solve for variables E(Rm) and β f. 9. 17 percent. = . (β f = the beta for the firm.05 + [E(Rm) – Rf] (β f 1. we have (in a world without taxes) WACC = k b B S + ks B+ S B+ S 4 8 = . Using the standard equation for the weighted average cost of capital. = .05] β f = .3). It is the rate of return which the project must earn in order to be acceptable. Fourth Edition The new WACC can be computed from equation 15.13.10 VL = where E (EBIT(1 − T)) + τc B ρ E (EBIT (1 – T)) = E(EBIT (1 – T)) (1 – τc ) B= kdD kb . =18% Therefore.05 + [E(Rm) – . Varhard’s recommendation is obviously wrong.15 = .4(. When growth is zero.

000 NPV –500. the appropriate discount rate is the weighted average cost of capital.000 45.9 B+S Then. VL = 100(1 − .48) = .Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 149 Substituting in the facts of the problem.0275) (1 – τc) = (.000 50.08 Therefore.08 ρ= 50 = 5% 1.250 − 500 750 = = $37. Since the new branch is an extension of Community Bank’s business.09) (1 – . the price/share will fall from $50 to $37. we can solve for ρ. the value of the firm has fallen to $1.08(500) + . the cost of equity for the unlevered firm.000 .05312 ∼ 5% where kb is the total marginal cost of debt financing. kb = (. we can determine the value of the levered firm after the swap. The per share price after the swap will be VL − B S = .0625 + .823 . 10(50) + 1.952 . The new price-earnings ratio will be 37.000 = 100(1 − .200 .11 (.000 35.5) .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% 1.50 10.035 39. where n = number of shares outstanding n n 1.000 25. S B + kb WACC = k s B + S B + S = .9) = .05 .5 = 1.000 23.5 ρ .864 .50 = 25 1.1) + .000 45. The correct weights of debt and equity to use in the WACC formula are from the target debt/equity ratio: B = .800 31.784 CF –500.000) +.50.250 .880 37.50 20 20 Therefore. and thus is in the same risk class.5) .08(1.0468 (.250.907 .745 38.000 Using this fact.

which is identical to the value of the all-equity firm. Fourth Edition For the years after 5.784) = 784.000. The current IRS code allows individuals to deduct donations to not-for-profit organizations from their gross income before arriving at taxable income. Assets liquid assets property. Presumably. CF 50.000 The NPV of the branch is –500. the correct answer is difficult to determine. 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. plant and equipment β al β a2 Liabilities debt βB equity βS Let Wai = the percent in asset i. 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 r . However. and Wj = the percent in liability j.035 + 39.000 = 454. The project has positive NPV. it will be rejected by shareholders of the levered firm. with a levered firm.000 (. WACC = Rf + [E (Rm) – Rf] (walβ al + Wa2β a2) Once taxes are introduced into our set of assumptions. this lowers the after-tax opportunity cost of capital to such individuals.200 + 784. If the decrease is large enough to offset the small positive NPV of the project. The answer implicitly assumes that the firm does not change its leverage. 12. the lower variance of return on the assets of the firm will decrease the value of the call.880 + 37. yet the project is 100% equity financed. use the perpetuity formula. the branch should be opened. 11.000. This assumption is valid if the bank intends to maintain its long-term capital structure at a target of 90% debt. In a world without taxes.05 and discount five years to the present: 1. 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.150 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. shareholders’ equity may be thought of as a call option on the value of the firm. Consequently.000 + 23.800 + 31. If the firm is not levered the project will be accepted because it increases shareholders’ wealth (although marginally).660 Because the NPV > 0.745 + 38.000 = = 1. the systematic risk of the firm’s portfolio of assets must equal the systematic risk of its portfolio of liabilities. even though a not-for-profit organization has no debt or equity. .

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

11 million.975 − 5.3 1 = .3) 1 + (. we have S = (25) (.06) (1) (. 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.54) = 22. such as dividend payments.3) 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. however.06 + .040 = . Therefore.419 = .5 + .1 From the Table of Normal Areas.65 = 1.4 − 1.919 N(d2) = .152 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.4 1. even if the value of equity increases at the expense of an equal and offsetting decrease in the value of debt.06) 1 (1.3 d 2 = 1. then two-fund separation implies (among their things) that everyone holds the market portfolio of risky assets. 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. 15. during the year. If we live in a world where markets are perfect.086 = 17. we have N(d1) = . B=V−S = $25 − 17. the market value of equity is $17.89 = $7. we have d1 = ln (25/10) + (. Fourth Edition 14.5 + .540 and substituting into the option pricing formula.919) − 10e− (.89 million and the market value of debt is $7. The market portfolio is composed of proportions of every asset held according to its market value weight. every individual holds the market value fractions of the debt and equity of every firm.5)(1. every individual in the economy will be indifferent because his wealth is unaffected. Therefore.89 Therefore. Any action which leaves the market value of the firm unchanged will not affect the value of the market portfolio. There are.91629 + .11 .

face value of the debt) is zero.919) (1.06 + [.06] (1. we have ks = .48 for the cost of equity capital. E (Rj) = Rf + [E (Rm) − Rf] β j For equity. the equity) will equal the value of the underlying asset (here.919) (1.1758 and for debt we have kb = . we can use the CAPM to determine the cost of equity and debt. From equations 15. Thus S = V N(d1 ) − D e − rf T N(d2 ) becomes S=V The OPM was used to obtain equation 15. because the probability that the value of the option (here. we have 25 βS = (.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.93) = .43) = . The CAPM applies to any risky asset.06] (. If we assume an all-equity firm.5) = .11 ∂S V βV . (a) True.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.93 17. k s = R f + N (d1 ) (R v − R f ) V S .89 25 βB = (1 − .12 − .06 + [.0858 16. the firm) is a virtual certainty.42 and 15.43 7. then N(d1) = 1 in the OPM. ∂V S βΒ = ∂B V βV ∂V B And finally.5) = 1. The second term of the option pricing model drops out since the exercise price (here.12 − .

24 + . Fourth Edition substituting N(d1) = 1. since equity is a call option written on the value of the firm.06) (4) (1. by definition.2 = 2.533 − (.5)(.000)(.5 + .06) (4) (. N(d1) = .215.4 d2 = 2.786628) (983.66437 = $1. Given the Black-Scholes option pricing model.533 .99434) − e− (.99434 N(d2) = .000) (.02 .133 Using the Table of Normal Areas.68 – 773.693147 + .2) 4 d1 = .68 – (. 100 percent equity.52) S = 1988. this reduces to ks = Rv + (Rv − Rf) (0) ks = RV (b) If N(d1) = 1 in the OPM.5 + .000/1.98352 Substituting these values into the call formula gives us the market value of equity. we have d1 = ln (2. Thus structure is.98352) S = 1988. S = (2. 17.49434 = .2) 4 + (. For an all-equity firm.000) + (. we have S = V N(d1 ) − e − rf T D N(d2 ) ∂S = 1.154 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. then S = V as shown in (a) above. 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.2) 2 = 2.48352 = . 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.

Changes in the value of the firm may be attributed to investment policy.3) 4 = 1.968 d2 = 1. or differences in risk? Finally. d1 = ln (2.000) (. at least to the point where the marginal gain from debt is offset by the marginal loss from potential bankruptcy.98 18.06) (4) (. Consequently.25525 N(d2) = .000) (.5 + . will increase the value of the firm by $100 (NPV = $100). .000) + (.468 = .968) − e− (. For example. or expected rates of return.3) 4 + (1/ 2)(. the market value of equity has increased and the market value of debt has decreased.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 155 The market value of debt is B=V–S = 2. If the firm is above its optimal amount of debt. the project should be accepted. Only the aftertax operating cash flows provided by the project are relevant. the project should be financed with approximately 75 percent debt (since B/(B + S) .06) (4) (1.65 1215 19.000 – 1215. In addition to this fundamental problem.5 + .3)(2) N(d1) = . Each of the three reasons given by the president may be criticized. If the firm is already at its optimal capital structure. 20. financial policy.895) S = 1. This weakens the power of the statistical significance tests.395 = .3 = 1. or both. then the project should be financed with a higher percentage of equity. The new debt-equity ratio is 768 = . how can one measure future growth in the firm. when discounted at the appropriate risk-adjusted rate.75). Therefore. 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.936 − 704 = $1.000/1. then the value of the firm will increase by taking on more debt. these cash flows.24 + .693147 + . there are problems of measurement error. We can apply the Black-Scholes formula.232 Therefore.62 1232 The old debt/equity ratio was 785 = . the effects are indistinguishable. (a) If we assume that there really is a gain from leverage. As given in the problem.85525 (.85525 − (. (b) The earnings per share ratio has nothing to do with the economic value of the firm.895 S = (2.02 = $784. the empirical work frequently uses cross-section regressions which are likely to have highly correlated residuals.3) 4 d1 = .

If anyone knew that equities were undervalued in the market. First.4375 = .18 − .4% .06] . Knowledge of the risk-free rate. Unless the firm’s president has all of his wealth in the common stock of the firm.06 + [. then equity should be issued to finance investments. The statement that “equity markets are currently depressed” has no meaning.06) (1 − . then we solve for the appropriate discount rate assuming it has the same leverage as the firm. D+E (a) Project 1.1125 + (.06 + . therefore. he probably does not really believe that the equity of the firm is undervalued.4375 ρ = E (Rj) = Rf + [E (Rm) – Rf]β j = . we find ρ for project 1.5.4)1 = .5 = 6% 125 Since corporate debt is assumed to be risk-free. the expected return on the market.18) = . then that individual would take advantage of his nonpublic information by placing his wealth in equity securities while prices were artificially low. unlevered. Because Southwestern will finance the new project with 50 percent debt.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. The cash flows are operating cash flows and are.156 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. It is important to recognize that the market equilibrium rate should not be adjusted by the tax rate. 21.0525 = 11.0 D ratio of 50 percent and a beta of . Fourth Edition (c) If the project requires equity financing in order to bring the firm’s capital structure back to its target leverage ratio. E (R m ) − R f = slope = The firm has a 18% − 6% = 12% 1. the market equilibrium risk-free rate is also 6 percent.0315 = 14. and the market beta allows us to determine the market price of risk (the slope of the security market line). βU j (the levered β ) is .1125 + . the before-tax cost of debt is k b = Rf = 7.1125 − .

15 − .12) (.5) + .06] (.5) = .5) ks = 12% Therefore.5 (.5) = .06 (1 − .06) (1 − . WACC = k b (1 − τc ) B S + ks B+ S B+ S = . ρ = Rf + [E (Rm) – Rf] β j = .5) + . the required rate of return is simply their weighted average.018 + .054 = 20.18 − . k s = E (R j ) = ρ + (ρ − R f ) (1 − τc ) = .4) (.09) + .5 (.072 = 9% Repeat the same procedure for project 2.4) (.4) (.75.12 (.06 + .06 + (.5) + .144 (. the weighted average cost of capital for the levered firm is B S WACC = k b (1 − τc ) + ks B + S B + S = . WACC (joint project) = .09 = 15% Now find the required rate of return on equity if the project has the same leverage as the firm as a whole.75 = .102 = 12% Since the projects are of equal size.4% The correct discount rate is WACC = k b (1 − τc ) B S + ks B+ S B+ S B S = .18 − .5) = .06 + (. ks = E (Rj) = Rf + [E (Rm) – Rf]β j = .8% .06 (1 − .06) . which has an unlevered βU j = .15 + (.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.204 (.06 (1 − .5% (b) The security market line gives us the required rate of return on the equity of the firm.06 + [.4) 1 = .018 + .06 = 7.12) = 10.018 + .15 + .5) = .

105)−10 PV31 = 80 . This is the same as the end of year 4.99 Inflow 10 20 25 60 NPV 9. which discounts payments at the end of each year from years 5 to 30 inclusive. PV0 = 969.50(1.38 18.105)− 4 41 Outflow 250 250 250 250 Factor .05 16.18 Year 41 PV0 = F (1 + r)−n = 40 (1.105)− Finally.25 204.07 = −24.0167) = .0500) = 24.105 = 110 (8.28 PV0 = PV31 (1.5 percent.67 Note that you discount for four years because the annuity formula.6707 NPV 226.18(. Interest expenses should not be included as part of the outflows because the discount rate already includes the interest factor.9050 .50(.07 Add all of the present values to obtain the final NPV: NPV = −783. Fourth Edition (c) In order to determine the acceptability of the joint project. do not accept the joint project.105)−4 = 969. .105)−26 = 110 . Year 1 2 3 4 Years 5–30 1 − (1 + r)− n PV5 = 110 r 1 − (1.99 + 84. will give you a present value which is at the beginning of year 5.158 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.69 783.20 = 40 (. discount PV5 and PV31 to the present: PV0 = PV5 (1.50 Years 31–40 1 − (1.015) = 481. you discount PV5 for four years.30 167. therefore.7412 .8190 .53 40.814) = 969.28 + 24.105)−30 = 481.67 + 650.6707) = 650.77 Therefore.20 + . discount the joint cash flows at 10.24 84.75 185.105 = 80 (6.

5 −95.7722 .45: B βL = βU 1 + (1 − τc ) S = .7118 .75) + (.8417 .0 PVIF .0 55. (e) If the joint project is accepted.6355) (5.80 −86.59375 (1 + .19 −55.95 Therefore. if it can be considered separately.7084 (9. L β = .0 −115.0334) (.4177) (.68 −80. project 1 would be accepted.59375 2 2 We know by equation 12.0754) (.0292) PV −110.0 −115. Project 2 Year 1 2 3 4 5–30 31–40 41 Cash Flow −120.6355 (7.0 PVIF .095 = . the firm’s equity beta is the weighted average of the original β and the project’s β . The NPV of project 2 is obtained by discounting its cash flows at 12 percent.10 (β p ) where βL p is the levered β of the joint project.6 (1)) = .7972 . Since both projects are of equal size.5 −95.08 −60.0 40.30 386.15 −91. if project 2 (the riskier project) can be considered separately.55 . it should be rejected. since β ’s are additive.Chapter 15 Capital Structure and the Cost of Capital: Theory and Evidence 159 (d) Discount project 1 at 9 percent.0 20. Project 1 Year 1 2 3 4 5–30 31–40 41 Cash Flow −120.57 NPV of project 2 Therefore.10 (.85 19.8957) (. the cash inflows and outflows of the joint project can be divided by 1/2.9290) (.90 (βL f ) + .45 + .5) + .97 7.0 40.0 −112.58 45.9174 . The unlevered β of the joint project is 1 1 βU = (.0 −112.0 20.6502) (.9 (.0096) PV −107.09 −96.87 −67.8929 .545 .37 275.95) = .73 NPV of project 1 Therefore.36 .4375) = .7084) (6.0 55. the new risk level of the firm’s equity is β = .

For example. Even if a change in the firm’s dividend policy results in a change in the average risk aversion of its clientele. . If all market participants were corporations. Old shareholders will be replaced by new shareholders without a change in the value of the firm. at most. one must be careful to distinguish between the value of the firm and the value of equity.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. First. If managers allow dividend policy to affect the investment decision. It does not depend on the risk aversion of shareholders. As shown in Chapter 6. 2.5 = 1.Chapter 16 Dividend Policy: Theory and Empirical Evidence 1. this will have no effect on the value of the firm. if a fixed dividend payout preempted the effort to invest in positive NPV projects.5PA + . the value of the firm would be adversely affected. the market required rate of return depends on the market price of risk and the firm’s covariance risk. The potential arbitrage profit is π = −PB + D − τc (.9 D . all that will happen is that a change in dividend policy will result in a change in clientele. then the value of the firm will be affected. We are interested in a condition which might affect the value of the firm. Therefore.9D + . 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. we must have PB = .5PB PB − PA = .8D This implies that the decline in price on the ex-dividend date must be 180 percent of the dividend payment. 3. but because improper use of dividend policy causes the firm to alter its investment decisions. not because of dividend policy per se.

Nontaxable income is provided by putting the borrowed funds into insurance.000 26. The relevant parameters on the ex-dividend data are: V = $1.25 T = 4 years rf = 6% .000 σ = .17 (b) Once again we use the OPM to determine the market value of equity.000 Capital Gains Sale of 2.667 Net worth 25. (a) See the solution to Problem 15. but one might argue that they pay taxes only on that portion of income which they consume. by borrowing enough so that the interest payments reduce your taxable income to zero.167 Liabilities Loan 16. People do pay taxes. Keough plans.500 1. or some other similar tax shelter. 5. not just dividend income.000 16.667 Ordinary Income Dividends received $1. 6.667 Closing Assets 2. If the price of common stock increases when a dividend increase is announced.500 shares @ $10 Insurance 25.000 1.000 41. The dividend per se has no effect no shareholders’ wealth.500 It is possible to shelter any income.167 Loan Accrued interest Net worth 16.000 Taxable income 0 Nontaxable income 1. The example income statement and balance sheets used to support the Miller-Scholes argument are replicated below: Opening Assets 2.000 Less interest expense 1.Chapter 16 Dividend Policy: Theory and Empirical Evidence 161 4. This has the effect of converting the income tax into a consumption tax. Any portion of income which is consumed will also be taxed.667 1.667 41.500 44.500 shares @ $10.000 16.500 shares @ $10. 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. The problem is that the nontaxable income from these shelters remains tax free only until it is consumed.667 44.500 Less original basis 25.500 D = $1.60 26.000 $1.60 Accrued dividends Insurance Liabilities 26.

24 + .06)(4) (. the market value of equity is $738. Fourth Edition The value of d1 will be d1 = = ln (V/D) + rf T σ T + (1/ 2)σ Τ + (.3508 = . Whenever dividend payout is “too low” from the point of view of an individual shareholder.5 + .238.17.4382 = .000) + (.25) (2) = 1.5)(. i.20) = . we can use the Gordon growth model.04 From the solution to Problem 15. First we need the expected end-of-year EBIT1(1 − T). equation 14. The important fact is that the firm will continue to accept all profitable projects.04 From the Table of Normal Areas. Finally. we determined that the market value of equity was $1.9382 N(d2) = .30 .8508 Substituting into the OPM.5 + . 8.000)e − (.215.500/1.” The shareholder can reinvest his dividends in shares of the firm.405465 + .06) (4) (.8508) = 1407. he can always sell a portion of his share holdings in order to consume. the increased variability of dividend payout is irrelevant. The ex-dividend wealth of shareholders is $500 + $738 = $1.162 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. This is an increase of $23 over their original wealth position.500)(. After receiving $500 in dividends.25 = 1. EBIT1(1 − T) = EBIT0(1 − T) (1 + g) g = Kr = (. Since the rate of return on investment is assumed to remain constant forever. The opposite is true if dividends are “too high.5 d 2 = d1 − σ T = 1.54 . to value the firm’s shares. the new dividend policy is irrelevant.54 – (.e.30 – 669.17b.9382) − (1.25) 4 ln (1. we have N(d1) = . 7. we have S = V N(d1 ) − De − rf T N (d2 ) = (1.1 EBIT1(1 − T) = (3.25) 4 = and the value of d2 is .1) = $3. What counts is the riskiness of the anticipated cash flows from future investment and it will remain unchanged.00) (1 + . dividend decisions will not affect the stream of expected future investments. It results from both the increased leverage and increased instantaneous variance after the dividend payout. Theoretically.26 = 738..5) (.

30(1 − .000 8.15 − . thereby leaving the assets side of the balance sheet unchanged. (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.000 2.000 10. (b) If new debt is not subordinate to old debt. .000 of new debt and an equal amount of new equity in order to finance the dividend payment.000 = = . Cash Inventory P.000 8.000 Debt Equity Total liabilities 6.5) 1.000 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.000 (d) Prior to the ex-dividend date the market value of the firm will be unchanged in a world without taxes.00 . The ex-dividend balance sheet is shown below..000 Predividend Ex-dividend D 5. we have V0 = EBIT1 (1 − T)(1 − K) D = 1 ρ − Kr ρ−g V0 = 3.000 of cash) will be “spun off” to shareholders.000 4. (ii) If the firm decides to issue $1.Chapter 16 Dividend Policy: Theory and Empirical Evidence 163 Applying these results to the formula.05 9.000 3. the amount of the dividend payment. P & E Total Assets 0 2.000 6.e. the $2.000.10 . (c) The debt to equity ratio will increase on the ex-dividend date. 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 D 5. P & E Total assets 2.625 D + E 8. the ex-dividend balance sheet will look like that below. After the ex-dividend date it will fall by $2.000 Debt Equity Total liabilities 5.65 = = $33.5 D + E 10.000 = = .000 6.000 10. (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. Cash Inventory P.

D. the same answer used in i) applies here. There will be no changes in (a) the systematic risk of the firm’s portfolio of assets.E. the exdividend balance sheet will be exactly the same as the pre-dividend balance sheet. L E (EBIT1 )(1 − T) ρ .000. we have VL = E(EBIT1 )(1 − Tc ) . k0 = WACC. for ρ. when the $2.6 D + E 10. the weighted average cost of capital. (iii) If the firm issues $2. 10. 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 .000 (d) Prior to the ex-dividend date the market value of the firm will increase by $2.000 is paid out the value of the firm returns to its original level.6 the Modigliani-Miller valuation equation for a levered firm with no growth.13. Therefore.000 = = . the value of the new debt and equity.000 in new equity. from equation 15. or (d) the market value of the firm. (iv) Using cash to repurchase equity has the same effects (in a world without taxes) as paying a cash dividend to shareholders. then pays out an equal amount in dividends. we have ρ= ρ= VL WACC VL − TC B VL WACC . (b) the wealth of original bondholders. (c) the debt to equity ratio. is B k 0 = ρ 1 − Tc L = WACC V Solving. Fourth Edition (c) The debt-equity ratio will obviously increase: D 6. However. we have VL = VU + TcB We also know. WACC Q.164 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. From equation 15.

the market value of debt decreases. a decline in the market value of debt. then we should observe the OPM implication—namely. If the market value of equity increases. 12. ceteris paribus. Not only does the cash payout diminish the assets which the firm holds. If the firm does not make use of external funding (i. If the firm pays a cash dividend the market value of the firm goes down. (EBIT – D).. The option pricing model leads to the conclusion that. but it also increases the systematic risk of the firm’s portfolio of assets. 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). This may lead to improper use of funds in projects with low rates of return. Second. However. the market value of debt should fall when dividends are paid from cash.3) Retention is simply the difference between earnings and dividends. This would leave them better off. if dividend payout carries no implications about cash flows from investment. From the point of view of bondholders. An increase in the retention rate does not necessarily affect the anticipated stream of investments at all. . ceteris paribus. However. and the riskiness of the firm’s assets increases. this is an undesirable effect for two reasons: First.Chapter 16 Dividend Policy: Theory and Empirical Evidence 165 11. These results follow from the fact that the cash paid out is a relatively low risk asset. If bondholders believe that higher dividend payout also implies greater future cash flows from investment. mP = 0).e. we know that EBIT + mP = I + D (16. 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. Referring back to the sources and uses of funds equation (for an all-equity firm). then a higher retention rate will imply more investment. the market value of their collateral may actually increase in spite of the dividend payout. the ceteris paribus assumption is important. the diminished asset base implies that bondholders have claim on less collateral in the event of bankruptcy.

400 − 22.3) − 32. (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 = . owning) is preferred to leasing for this project.000 (4. 2. 7%) − .1002) = −22.. 7%) = 100.4) = 9% .400 (3. Equation 17..3872) − 6.000 − 22.. but it assumes that lease payments are made at the end of each year.000 (1 − .874 The negative net present value of the lease contract clearly indicates that debt financing (i. 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. we have: NPV (to lessee) = 100.3 (20. Substituting the facts of the problem into the above equation.000 (1 − . If lease payments are made at the beginning of each year.15 (1 − .3) PVIFa (4 yrs.000 − 32.Chapter 17 Applied Issues in Corporate Finance 1.000) PVIFa (5 yrs.4 shows the NPV of lease financing from the lessee’s point of view.e.

If the lease payments were larger. then the lessor would have a positive NPV.000) ∑ 5 (PVIFa (5yrs.Chapter 17 Applied Issues in Corporate Finance 167 Substituting the facts of the problem into the formula we have NPV (to lessor) = −500.577 (PVIFa (5 yrs.6) + 40.000 – 147.000 + ∑ 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..352155) = 500. we have NPV (to lessee) = −500. (b) The formula for the NPV of the lease to the lessee is rewritten below.4) + ..000 + (88.4(100.299 (c) It is usually the case that if the lessor is charging a competitive rate (i.000 – 494.15) 5 = 500.577(1 − .000 + 500.701 NPV (to lessee) = 5.9%)) = −500.577)(.577 t t = 1 (1 + .000 NPV (to lessor) = 0 Thus. 15%)) = 500.577 (3. . 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.09)t t =1 = −500.000 + 147.889651) = −500.000 – 147.2 + 40. The NPV of the lease for Reddi Roller Leasing is zero.000 + (147.546. the lessor is charging a competitive lease rate. 000) (1 + ..000) (3.e.

256.236 + . hence there is an immediate payment (in year 0) but no payment at the end of year 5.456 (8. • The after-tax salvage value is assumed to be received by the lessee (for a price of $1).548 26. Note that the lease payments are made at the beginning of each year. Note that the problem discounts the salvage value as if it had the same risk as the other cash flows.969) If we set the NPV (lessee) equal to zero.715 (8) (9) (10) After-tax t Salvage Value CFt (1 + kb)− — L 1.744 69. ($3549).576 11.000 85. This is an unreasonable assumption because the salvage value usually has greater uncertainty than the leasing (debt) cash flows. To keep things simple we assumed the lessee reports a capital gain of ($10. Fourth Edition 3.000 − $1) and pays capital gains taxes immediately.48 5 .8547 0 L .542 ∆Dt — 14.000 14..52L 0 (5) (6) (7) τc Year 0 — 1 0 2 0 3 . given the complex pattern of cash flows. Column 9.000 − L − . i. it is useful to construct a table similar to Table S17.064 49. It is the sum of: • τc dep.000 20.5337 8.680 19. Columns 6 plus 7 always add to equal $31.52L 0 .52L .423 4. which is saved by leasing.6244 (15.969) . 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. The cash flows in column 9 are all outflows. The sum is ($8969).600) (8. • (1 − τc)Lt. and solve for L. Nevertheless.000.715 ∆k b Dt −1 — 17. Future depreciation of the $10.843 + . If we add the $100.236 + . the NPV for the lessee is NPV (lessee) = 100.0000 0 L .000 0 (4) (1 − τc)Lt L L L . This is a tax shield received by the lessee.857 – 3.8547L − . $2180.191 .000 10.000 investment outlay.256 16. a positive cash inflow.000 30.168 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. Column 6 shows the portion of each amortization payment which is interest and column 7 shows the reduction of principal. with the exception of the after-tax salvage value which is an inflow.52L) + . (a) To answer this problem. the after-tax salvage value ($7600) and the present value of the depreciation tax shield.1 (1) (2) Cash Outflows From The Lessee’s Point of View (3) dep — 40.833 26.843 + .7305 0 . Table S17.e.741 8. the depreciation tax shield (column 2 times column 3).48 4 .256. using a 17% discount rate. and should be discounted at a higher rate. but it is reduced each year because annual (end-of-year) amortization payments are $31.7305L − .516 22.187412L L = $28.52L (7. row 5 is the sum of: the displaced interest tax shield.1. The amount of displaced debt is initially $100.6244 15. i..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. we used 17% for convenience.52L) − . • The opportunity cost of displaced debt.48 D t −1 — 100.5337 (8.4561 Obviously.e. the result will be the maximum lease fee which the lessee can afford to offer in negotiations NPV (lessee) = 0 = 89. an explanation of the cash flow column (column 9) is in order. the after-tax lease payments (from column 4).

8442 (14.18) k s = ρ + (ρ − k b )(1 − τc ) B S The information in the problem indicates that your tax rate is 40%. Note that no capital gains are received by the lessor when the asset is sold because the sale price.000 30. 4.52L . column 2 times column 3.84% interest rate.336 L = $26.52L . and kb = 14%.000 10.191.17 (1 − .2 (1) Year 0 1 2 3 4 5 Cash Inflows From the Lessor’s Point of View (2) τc — . 15.48 (3) dep — 40.52L 0 (5) CFt .000 to acquire the asset and receives the cash flows in column 6.000 20.Chapter 17 Applied Issues in Corporate Finance 169 (b) Next. we have NPV (lessor) = 0 = 2.2 shows the cash flows for the lessor.52L 14.52L) + .9188 . If we assume the Modigliani-Miller framework for our analysis. Eq. therefore there is room to bargain.48 .000 + . column 4. B/S = 9.52L . we need to analyze the same contract from the lessor’s point of view.7756 (9.2106L – 59.52L 0 (6) (1 + WACCB)− 1.400 + .7126 .800 + . and the book value.9188(19. are essentially the same.0884 Therefore.52L) + . The lessor spends $100.200 + .7126 (4. The weighted average cost of capital for the lessor is WACCB = kb (1 − τC) = .48 . then our cost of equity may be computed from the following formula (see Chapter 15. Table S17.0000 . therefore the NPV to the lessor is NPV (lessor) = −100.842 The results show that the lessor requires at least $26. the lessor’s minimum acceptable lease payment.400 + .48) = .48 .52L) Setting this equal to zero and solving for L.600 + . The weighted average cost of capital for the lessor is WACCB = kb (1 – τC) . All you need to estimate is the before-tax rate of return on equity. the discount factors in column 6 are based on a 8. ρ.7756 .48 .8442 .000 0 (4) (1 − τc)Lt . the after-tax lease receipts. $0. Table S17. the depreciation tax shield received.52L .52L 9. Any lease fee between these limits will provide a positive NPV to both parties. and • (1– τC)Lt.52L 4.600 + .52L + .52L 19.200 + .800 + .842 per payment while the lessee is willing to offer up to $28.52L) + . $1.6547 t Cash flows for the lessor are the sum of: • τC dep.

9) Substituting this back into the cost of equity formula.)] = −100.000 + (18.36%.15 (1 − .48(.500) (1. if the company buys the project and finances it with debt.14) (1 − .170 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. .9)) ρ= .0936 = 16.4) + .000 + ∑ 31.600 + 5. 15.4) (.4 (12. the NPV is negative and the project should be rejected. (a) First.000) [PVIFa (17%.000 + (23.207163) = −100.164789 + (. 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 = .25 (. 5.12) we have B WACC = ρ 1 − τc B+ S .0936 = ρ (1 − .48 (.4789% 1 − .164789 − .17)t t =1 8 = −100. therefore WACCB = . Fourth Edition The rate paid on the lease contract is 18% and the lessor’s tax rate is 48%. 8 yrs.5) = 17% Substituting in the information from the problem we have NPV = −100.48) = 9. we have ks = .5) + .86% This is the required rate of return on equity for your firm.000 + 99.4) 9 ks = 29.18 (1 – .600) (4. Using the Modigliani-Miller definition of the weighted average cost of capital (Chapter 15. Eq.289 NPV = −711 Therefore.000 (1 − .

(c) If the project is leased. Hence.15]t t=1 8 = 100.000 − (21. 17.)] = 100.400(.4 (12. the NPV of the lease contract is $1. the positive NPV of the lease contract will offset the negative NPV of the project.534819) = 100. . the project is acceptable if leased.840) (5.4) + . 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) = 100.400 (1 − .741 NPV (lessee) = 1.6) + 5.4).000 – (17.000 − ∑ 21.259.5.000 – 98. 8yrs.Chapter 17 Applied Issues in Corporate Finance 171 (b) The NPV of the lease is computed by using Eq.259 Consequently.500) [1 − (1 − .000) [PVIFa (9%.

Premium to T 10.150/5. the manipulative atmosphere of the activity may provide a fertile soil for such behavior. Also this works great in a booming stock market.010) = 28.010) = 71. Total Market Value (old) 7. Some executives have been charged with stealing from the company. For the buyer as its EPS grows.150 (7. Terms Paid (A for T) 9. Net Income 2.5 Model of Differential P/E Ratios Acquirer (A) $100 100 $1 50x $50 $5.43 +43% 100($71. Proportions (old) 8.7% $20 100% 0. The problem was aggravated at Tyco because as the company seemed to prosper. P/E Ratio 5. However. 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. the weighted average growth will drop and/or risk of the buyer will increase.010 (b) The basic generalization is that when a buying company has a higher P/E ratio than the target. Divestitures. Proportions (new) 140($71. of course. the buyer will achieve accretion in earning per share while the target will suffer dilution in earnings per share.57% Combined $200 140 $1. A second problem is that when low P/E ratios companies are not available. Number of Shares 3.50 (new) $6.50) = $10.4(1.43) = $0.000 16.43% Target (T) $100 100 $1 10x $10 $1.50) = $2. 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. Its EPS growth and P/E ratio are likely to collapse. . the buyer will not have this artificial growth stimulus any more. Total Market Values (new) 13. top executives received loans which were later forgiven.Chapter 18 Acquisitions.3% 0.860/1.860 (2. EPS (% change) 12.4A/1 $1.860/10. and Corporate Governance 1. its P/E may actually increase. EPS (new) 11. This facilitates doing other acquisitions with its growing P/E ratio. In a weak stock market.000 83.000)−1 = 43% (7. Restructuring. The relative P/E ratio is the one reflected in the deal terms.000 1. Price/Share 6. These practices. the P/E ratio of the buyer may drop considerably. Market Values (% change) 14.150/10.4(100)($71. are not inherent in playing the differential P/E game. EPS (old) 4.000)−1 = 186% (2.50) = $7. So as more low P/E companies are bought.43 50x $71.572 −43% 0. (a) Solution to Table S18.

06) (4) (.2.000) + (.000 shares outstanding.000) + (.99733) − e− (.179)(2) = 1.7856 − (. we have N(d1) = .02. We can use the OPM to determine the market value of equity.179) 4 = 2.99240 and substituting into the Black-Scholes formula yields S = (4.30 = $2. only slightly more) than the value of equity after merger.918 .179) 4 . 3.49240 = .02.428.000 Substituting into the formula. the market value of equity in the merged firm was $2.04.430. and Corporate Governance 173 2. With 1.000 / 3. the face value of debt in the merged firm increases to $3.428.000 in debt. Divestitures.358 d 2 = d1 − σ T = 2.000 and the market value of the firm to $5. Therefore.06)(4) (2. we have $2. we have d1 = ln (4.276 .43 per share.179 = 2.428 Referring to the Table of Normal Areas.06 T=4 σ = .358 d 2 = 2.693147 + .179) 4 + (1/ 2)(.5 + .000)(.000. V = 4.32 – 1561. Restructuring.179 = 2. shareholders would oppose the merger if there were no other merger effects. This is more (albeit.9924) S = 3989.24 + . Doubling it. 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(5.510825 + .000/2.000) (. From problem 18.7856 .06)(4) + (. using the parameters for the merged firm. this amounts to $2. By taking on an additional $1.5)(. We can apply the Black-Scholes option pricing formula.179) 4 rf = .99733 N(d2) = .179 d1 = .276 − (.179) 4 (.49733 = .Chapter 18 Acquisitions.000 D = 2.02 Prior to merger the value of equity in each firm was $1.24 = + .5 + .215.

353545 = + .899545 = 1. then the value of the stock is given by S = VN(d1 ) − De− rf T N(d2 ) where d1 = and d2 = d1 − σ T. Substituting into d1.4 1.9 Using the Table of Normal Areas.943 – 2294.5 + . we have S = (5.000 + (.4724 = .75 = 2648. (a) If we regard the stock as a call option on the firm.174 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.9724) = 4.5 d1 = . N (d1 ) = .9886) − (3.10 .9886 N (d2) = .236) (.9724 Substituting these values into the OPM.4886 = .000)(.06)(4) (.353545 = 1.31623)(2.000 = $2. the price per share increases by 22 cents.31623)(2.236) .5 + .5)(. 4.70709 .19291 N (d 2 ) = .10 5 = = ln 2 + (.093 + .883 .546 + . we have N (d1) = .70709 = 1.08)(5) ln 500 + (1/ 2) .971 d2 = d1 − σ T = 1.353545 .693 + .65 Therefore. we have 1. Fourth Edition From the Table of Normal Areas.70709 ln(V/D) + rf T σ T + (1/ 2)σ T = 1.25 Price per share = 2648.08)(5) + (.9 − .25/1. with an exercise price equal to the face value of the debt outstanding.000)e− (.

for each firm is $325.25)(. Divestitures. B.05 σ = .333 = $667 .2 – 1.000 (.08)(5) 500 (. the value of the equity for the merged firm is S = 2.936 N (d2) = . Restructuring.436 − .08) (5) + (.000 (. V=S+B 1000 = 675 + B B = 325 The combined equities of Firms A and B are worth $1.2 – (973. S = 1.5 N (d1) = .Chapter 18 Acquisitions.000 – 1.350.5) (.6703) = 1985.9926) − e− (.9735 Therefore.2236)(2.000 (. setting the covariance term equal to zero. d1 = = ln 2 + (.6703) = 971 – 296 = $675 The value of the debt.2236 Using the Black-Scholes formula once more.5 = 1.236) (.66 = 1.9926 d2 = 2.08) (5) 1.1) = .2236)(2. 2 2 σ 2 = w2 σ 2 A + (1 − w) σ B = (.6703) = 971 – (441.1) + (.6703) = 1985.236) 1.093 + .333 B = 2.436 .332.25 = 2. SA + SB = 675 + 675 = $1350 To calculate B and S for the merged firm.971) – e− (. and Corporate Governance 175 Therefore.883) (.186 + .9735) (.2 – 652.25)(.54 = 1.883) = 971 – 500 (.25 = 2. first calculate the new variance of returns.9735) = 1985.5) (.5)(.000 (.

The resulting market values of debt and equity are calculated below: VAB = 2.5 + .2236)(2.560 ln(2.5 + .350 = −17 So the merger transfers wealth from the equityholders to bond-holders. 013(1.236) (.000) /1.5 N (d1 ) = .9390) − 1.8524) = 1.560(.5)(. 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 $2.560) = 649.236) .08)(5) + (.25 = 1.547 .6703)(.176 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.000(. ′ B′ AB (D AB /D AB ) − BAB = L For the facts of this case.560) + (. Let L = the loss to the equityholders as a result of the merger: L = SAB – (SA + SB) = 1.560) − 667 = −18 ≅ −17 Note that BA + BB = 325 + 325 = 650. We can increase the face value of the debt to D′ AB to reverse and neutralize the effect of the merger such that. . Company AB should issue new debt with a face value of $560.000 − 987 = 1.6485 = + .2236)(2.000 D′ AB = 1. 000 /1.4390 = .333 – 1.3524 = .013 Test of results: ′ B′ AB (DAB /DAB ) − BAB = L 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.9390 d1 = d2 = 1. This is equivalent to ′ B′ AB (D AB /D AB ) = 1.547 − .878 − 891 = 987 B′ AB = 2.000).5 = 1.013(1. Fourth Edition Shareholders lose $17: $1350 – 1.000 /1.047 N (d2 ) = .8524 S′ AB = 2. the merged firm increases the face value of debt D′AB.

Divestitures. Hence both their risk measures and return measures would be changed. one would not expect the performance of conglomerate firms to be significantly worse than a broad market index. 8. 6. They presumably sought a carry-over of some technical or general managerial abilities with the firms that they acquired. M = medium. If the firms’ performance were better. a competitive market among acquiring firms would cause any advantages to be priced according to their value. Thus.013 – 650 363 + 987 1. Alternative Strategies for Growth (a) Solution to Table S18. 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. manager-agents may behave in a manner more risk-averse than stockholders would wish. Studies have shown that betas usually approach 1 if investors diversify randomly among 15 or more stocks. Restructuring.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). this indicates that diversification as such was not the dominant motive for conglomerate mergers. H = high .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. 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. Similarly. until the returns and risk of the conglomerate firms plotted on the security market line. The managers would prefer to prevent bankruptcy by decreasing the probability of default on debt. if there are costs to losing their jobs. 7. One hypothesis of superior conglomerate performance is that synergy exists between the components of combining firms. because persistent evidence that conglomerate firms were making bad choices (or overpaying) would result in a reassessment of their market values. unless this synergy is firm-to-firm specific. However. This is inconsistent with efficient markets. it would imply that conglomerate firms were able to buy under-valued securities. 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. Non-owner managers of the firm may be motivated to act like bondholders.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. 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.

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

11 .4) .4)(10) .15 .15 − .21) for a levered firm.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 (14.09 .4)(10) .09 − .12 .4) .11) = $65.09 = $70.4)(10) .28 = $106.11(1.45 + 72 (.11 + 1(12)(1 − . Restructuring.5733) = $65.09(1. and Corporate Governance 179 11.12(1. (a) Assets ($ 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 .30 $12. Divestitures.21) = $100 + 0 = $100 Value of Alber: L VA = 12(1 − .73 Value of Saben: VSL = 10(1 − .09 + 1(15)(1 − .12 + 1.Chapter 18 Acquisitions.09 .2232) = $50 + 20.15 $15.25 $10.18 − .45 + 41.12) = $50 + 90 (.00 Alber .5(10)(1 − .4) .09 .00 Saben .18 . 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 − .09) (14.

180 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.2009 − .4) .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 12.4)(10) . (a) Accounting Balance Sheets (millions of dollars) Debt Equity Total Assets (b) After tax r Before tax r = r/(1 − Tc) = r/.04 Value of Watro-Saben combination: L VWS = 23(1 − .12(1.4)(10) .63 + 147.11 .2976) = $115 + 41.12 + 1(23)(1 − .41 = $311.11(1.11) = $163. Fourth Edition (d) Value of Watro-Alber combination: L VWA = 30(1 − .12 .12) = $115 + 138 (.16 − .11 + 1.63 + 198 (.7445) = $163.1(30)(1 − .07 = $156.4) .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 .

Total shares 6.28 + 1.2) 136.632 1. Net income of J + net income of K or L 7.5) 144 144 = .20 28.60 115. J’s original EPS (in dollars per share) row 8 − row 7 100 9.20 6X $7.8 $16 −4 12 –4.8 $1. number of old J shares 5.4) + (1.5X 2 15 4 19 12 $.00 108 $20 −4 16 –6.4 9.8 108 (1.60 15X $54.20 47% (e) Because β ’s are additive.48 . number of new shares 4.8 million = (28.8 136. share price of J 2.4) + (1. we simply weight the β of each firm according to the market value of its equity.8 7. Dilution row 8 $16 −8 8 –3.2 βJL = (1.8 = . as calculated in part (c).2) βJK = 28.8 + 115.72 1.6 $4.2 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 1.2 or Loomis 8X 2 16 4 20 14. New EPS (in dollars per share) 8. number of acquired firm shares (millions) 3. Restructuring.4 $.2947 + .8 115. The total market value of equity for the proposed merged firm would be JK: JL: $136. Divestitures.2 = 1.2 $3.24 28.20 40% 7.8 + 108) $144 million = (28.80 12X $57.9474 = 1.Chapter 18 Acquisitions.

134) + 264 264 = .028 + .027 + .48) = 134 = 13.06 + (.4% (f) Use the proposed merged firms’ market values of debt and equity to determine the weights for the cost-of-capital formula.4) (. 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) = .06 + (.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 .10)(1 − .4) (.8 256.06) (1.8 JL $120 144 $264 WACC = k b (1 − Tc ) (B/V) + k s (S/V) 120 136.065 = .10)(1 − .182 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.122) + 256.093 = 9.11 − .8 = .06) (1.11 − .073 = .3% 120 144 WACCJL = (.10 = 10% (g) Using the Modigliani-Miller valuation equation (equation 14. the market value of the debt is assumed to be equal to its book value.24) = 122 = 12.8 WACCJK = (.2% k s (JL) = . (in millions of dollars) Debt (D = B) Equity (S) Value (V) JK $120 136.8 $256. In this instance.

4) .00 213.20 $10.91 = $274. Restructuring. The merger with Konrad is preferred and it also meets the capital budgeting requirement of having a substantial positive net value.91 million in value.093) = 206. Divestitures.0 $76.91 28.91 If Jordan merges with Konrad.10) = 216 + 58.4)(10) .2 JL $274. .093 + (1)(32)(1 − . Jordan could pay a premium to Konrad shareholders and still have a positive NPV.4)(10) .093(1 + .00 120.91 120.4) .00 28.45 + 126. we have V(JL) = 36(1 − .10 .00 million For JL.8 126.8 184.10 + (1)(36)(1 − . the indicated net increase in value is $76.11 108.2 115.2 million.13 − . 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 − .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 $333.16 − . A merger with Loomis results in a gain of $10.093 .Chapter 18 Acquisitions.5 = $333.10(1 + .00 154.

Simultaneously we enter into a forward contract in which pesos are exchanged for dollars at an exchange rate 10.Chapter 19 International Financial Management 1 Interest Rate Parity for FC/$: In the text example of covered interest arbitrage the foreign exchange rate was expressed as dollars per euros.0 = 10.0 1 + R$.S. The 1000p invested in the Mexican security carrying a 9% rate has an end of year value of 1090p. so IRP obtains. The premium on the dollar is matched by the differential interest rate paid on the peso denominated security. Table S19.05 = 10 1. Fp/$. 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.1a Example of Interest Rate Parity (A) Equilibrium Mexican Peso and U.03810 .09 − 0. (a) Parity relationships are illustrated in Table S19.0 − Sp / $.0) = $100 × 1.0 10.03810 = 0.0 − R$.05 0.0 = 10.381 − 10 0.0 Sp / $.05 = $105.09 = 1090p in $: 1090p × (1/Fp/$.05 0.0) = 1090p × ($1/10.0 in pesos: 1000p × (1 + Rp. Dollar.38p/$.0) = 1000p × 1.0 = 5% per annum R p.381 0.0 = R p. we could convert the $100 dollars into pesos at the spot rate of 10p/$.0 = 9% per annum Year 0 Dollar $100 Year 1 $100 × (1 + R$. Alternatively.381p/$ Sp/$. This amount is exchanged at the forward rate shown in Table S19.0 = 10p/$ Fp/$.381p/$ R$. One hundred dollars invested in a dollar security gives an end of year return of $105.04 = 10 1.0 Peso $100 × Sp/$.381p) = $105.1a to obtain $105.0 = $100 × 10p/$ = 1000p Fp / $.1a.

5p/$ R$.0.5p/$ Sp/$.05) = 11.0 = 5% per annum R p.Chapter 19 International Financial Management 185 (b) Case B. The analysis will show that the return from investing in a dollar security will be higher than investing in a peso security.S. discounted at the U.0 = 9% per annum Year 0 Borrow (short) pesos.5p.1b Example Covered Interest Rate Arbitrage (B) Mexican Peso to U. Repay at Rp. Invest (long) dollars at R$. rate. An investment is made in pesos to provide a comparison return. Alternatively.0 − R$. Dollar.04 = − 10 1.01190 .0 = 10.09 − 0.0 Sp / $.5p −1090p + 1102.09 = −1090p $100 × (1 + Rp.0 in pesos: $105.0 1 + R$. its present value is 11. 1000p × (1/Sp/$.05 = 0. the pesos borrowed are converted into dollars at the spot exchange rate to obtain $100. Fp/$.05 = $105.0 −1000p in $: Sell pesos spot. The $100 is invested in the dollar denominated security to yield $105 at the end of the period.0 − Sp / $. The forward premium on the dollar is increased to 5%.S.5p = 12.05 − 10 1.0 × 10.905p which is equal to the forward premium of the dollar less the premium of the interest rate on the peso security over the interest rate of the dollar security.5p/$.0) = $100 × 1.5p/(1.5 − 10 0. Hence. 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 10.5 0. Table S19. representing a gain on the roundtrip of 12.05 0.03810 = 0. Table S19.05 − 0.0) = 1000p × ($1/10p) = $100 Year 1 −1000p × (1 + Rp.0 − Rp. Buy peso forward.0 = 10.0 10.0) = −1000p × 1.1b reflects the condition that the forward rate for the peso is higher than the IRP rate holding everything else the same.5p/$ = 1102. This yields 1102.0 × Fp/$.5p.0 d= = 0p 0p Fp / $.5p 12.905p Net peso position PV net peso position at R$.0 = 10p/$ Fp/$.0 = $105. capital will flow from pesos to dollars.

This amount converted into pesos at the forward rate and discounted at the U.0) = −$100 × 1.0 in pesos: 1000p × (1 + Rp.0 − R$.225 $100 × Sp/$. simultaneously entering a forward contract in which pesos will be converted into dollars at the current forward rate of 9. The reference investment will be to sell a dollar security short.0.0 = 9.8p) = $111.03810 = −0.186 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.02 − 0.8p/$ = 61.05) = 58.0 1 + R$. interest rate gives 58.225.0p/(1. The amount of dollars received is $111.04 = − 10 1.0 =$100 × 10p/$ = 1000p in $: Net dollar position Net peso position $0 0p −$105. Invest (long) peso at Rp.8 − 10 0.225 × 9.2 0.0 9.225 = $6.8p/$ R$.225.S.0 = 5% per annum R p. 0 d= = 0p Fp / $.0 Buy peso spot.0) = 1000p × 1. At the end of the year. we convert the $100 into pesos at the spot exchange rate of 10p/$.05 −0.225 $6.0 − Sp / $.09 − 0. the dollar is at a discount in the forward market.0p 61.05 = −$105. Sell peso forward. Fp/$. Fourth Edition (c) In Case C. Table S19.S.09 = 1090p 1090p × (1/Fp/$. the investor owes $105.0 Sp / $. Repay at R$.0 = 10p/$ Fp/$.05 − 10 1. Dollar to Mexican Peso.1c.0 = 9% per annum Borrow (short) $100. Alternatively.095p which is equal to the algebraic difference between the forward discount of the dollar and the discount on the dollar interest rate.1c Example of Covered Interest Arbitrage (C) U. Year 0 −$100 Year 1 −$100 × (1 + R$.0 = $6. Table S19.225 × Fp/$.8p/$ Sp/$.05810 . so outflows take place from dollars to pesos.0 = 9.0 + $111.095p − Rp. The investor can close his short position with a gain of $6.0) = 1090p × ($1/9.0 PV net peso position at R$.8p/$.05 = −0.

unexpected major improvements in economic activity could raise real interest rates.02). Dollars 2. Note also that it plots above the IRP line in the area where capital inflows take place from the peso to the dollar. Point B calculated from the data in Table S19. 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. January 24.1c plots below the IRP line in the area of capital outflows with coordinates (−0. 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. FX Weekly.0381. Another explanation is that the bias is evidence of expectational errors. Point C calculated from the data in Table S19. One is that β < 1 is evidence of a risk premium on foreign exchange. on average. −0.1a) is plotted as point A in the first quadrant with coordinates of 0. Figure S19.0381 on the interest differential axis. or real economy shock. currencies trading at a forward discount (premium). 4–7). Hence. . If investors are risk averse and foreign exchange risk is not fully diversifiable. this arbitrage opportunity will disappear as investors make such trades. the forward rate bias persists. but long run movements toward parity relationships. the interest differential or the forward rate discount is no longer a pure estimate of the expected change in future exchange rates. in the country with higher interest rates. fiscal shock. investments in dollar assets become more risky. in the long run.1.Chapter 19 International Financial Management 187 (d) In Figure S19. (a) Two possible explanations for the bias are common in the literature. 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.1b plots as 0. the IRP (Table S19. tend to weaken (rise) less than what is implied by the forward discount (premium). Hence. Specifically.05 on the dollar premium axis and 0. 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. p. 2003. but the sum of the expected change in the exchange rate plus a risk premium. An example of expectational errors could be a counter monetary.1 Interest Rate Parity Relations Pesos.0381. The Deutsche Bank explains the forward rate bias as resulting from investors risk aversion to short term losses that can be substantial (Deutsche Bank. But because of risk aversion. When the dollar interest rate rises. In theory.

188 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. W = FC6/$1 so FC6 = $1. The parity conditions are equilibrium relationships. (a) Exchange rate uncertainties may be reduced by offsetting economic developments within the larger area. the price is $3 × (FC5/$1) = FC15.20 B imports FC5 Hence country A would import X and export Y. Transaction costs of exchanging one currency for another are avoided. If country A imports X. vs. whether to go long in the forwardrate bias trade or to close the position and do nothing. 5. They represent strong forces toward which foreign exchange rates will move. The fixed exchange relationships within the union avoid speculative attacks. This is called going long in the forward-rate bias trades.40 B imports FC15 A imports $1.. Product Y vs. the price is FC12 × ($1/FC5) = $2.5 Domestic and Import Prices When FC5 = $1 Product X vs. i.e. 12 = 6Z Z = 2 = number of units of imports of Y. If country B imports X. Monetary and fiscal integration are successfully integrated. 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. 4. $3. Other prices can be calculated the same way. In addition. Country B would export X and import Y. So FC12/W = 2($1). (c) Benefits are greater than costs if the underlying economies are highly integrated which implies a high level in trade in goods.00 domestic FC6 domestic A imports $2. Labor mobility between countries in the union is high.5 Table S19. (b) A country loses monetary and fiscal independence. 3. forecasters are guided by the parity conditions as relationships toward which exchange rates will move. vs. For A: We have exports of 2Y and imports of 1X from B. 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. A country loses the use of the exchange rate as a mechanism to facilitate adjustments to shocks among the nations within the union.00 domestic FC12 domestic $1. Equilibrium would require for B: For one unit of X exported B could import 6Z of Y. Results are summarized in Table S19. . The parity conditions therefore represent a useful framework for guiding economic policies.40.

000. 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.S. Conversely. Country A is an attractive market in which to increase sales from other countries and A’s imports increase. At higher $ prices in the U. Country A runs an export balance surplus. the demand for U.S. U. Domestic prices in Country A rise. at the old exchange rate. A formulation first set forth in the mid-eighteenth century is the price-gold-flow mechanism.000 when the exchange rate was $1 to FC4. At the new exchange rate the auto exporter still seeks to receive FC8. Let us assume. in order to receive $4 per bushel.S.000 provided the auto exporter with FC8. in the foreign country. Note the adjustment process. Conversely. To illustrate the operation of the adjustment process let us assume an initial relationship of $1 to 4 FC: $1 = FC4. 6. The demand for FC relative to dollars increases. goods in the foreign country will rise.S.S. 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. the wheat can be sold at FC8.. the export surplus increases its income and employment. At the new exchange rate. Similarly. lower prices in FC for both their imports and exports would stimulate purchases from the U. In the U. he must receive $4. Or 1B is worth 0.000 for the auto. The export deficit decreases income and employment in the deficit country. imports and exports in dollars rise. Hence gold flows into Country A while it flows out of Country B. At the new exchange rates. The value of the dollar falls. employment may decline in A and increase in B. The change in exchange rates in response to shifts in relative prices helps readjust the export–import imbalances.Chapter 19 International Financial Management 189 We then have: A imports 1X = $2 B exports 1X = FC12 and and exports 2Y = $2. the increased relative demand for FC in relation to the $. . the demand for foreign goods will decline. causes the FC to rise in value relative to the $. sold wheat at $4 per bushel. But the higher value of the FC causes imports into the U. suppose that an auto sold in the United States for $2. 8. to be more expensive in $ and to represent a smaller number of FC to exporters at previous $ prices in the U. in the adjustment process. Also. imports rise and its exports fall. Hence 1A will be equal to 4B. IF the surplus country was not functioning at full employment. Thus the prices of imports in dollars would rise substantially at the new exchange rates.S. results in higher relative prices. each country’s exports and imports are equal in total value measured in units of its own currency. the prices of the U. A sale at $2.. Conversely for the foreign country. Conversely. A’s goods are more expensive in other countries so its export sales decrease. for purposes of illustration. imports 2Y = FC12. A higher rate of inflation in the U. At FC6 = $1. The A currency has four times as much gold content. 7.25A. while Country B runs a deficit.S. The dollar price for exports could be increased and still represent substantially lower prices in the foreign currency.S. receiving FC16.S. But to do so at the new exchange rates. the U. that the new relationship is now: $1 = FC2. 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. and reduce sales to the U.S. For example.000. A’s export surplus will be reduced or reversed until equilibrium between relative price relationships of the countries is restored.S. the prices in Country B fall. Income and employment effects may also enter into the adjustment process as well as price changes.

the subscript d represents the U. So we solve for X1: X1 1.S. (a) The PPP: X1 Pf1 / Pf 0 = X 0 Pd1 / Pd0 In this example.10) = 1. 1 + Rn = (1.86 per $1. and Xf is the forward rate. Plus 2.04)(1.S. Plus 2. and f represents foreign.S.144 Xf = 3.190 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.2% nominal interest rate for the foreign country.144 = 14. (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. Plus 2. Xf 1. Minus 3a.S. The prices of internationally traded goods carry most of the adjustment process. X0 = 3/1 is the beginning exchange ratio. $10.. We know that the U.05 = 3 1.000 $5. a) b) c) d) e) f) Plus 1. $10.000 $5.000 $500 Minus 2..276 Xf = 2.000 $1 million $3. Minus 1.000 $2. For the U.4% nominal interest rate for the U.144 1. So we solve for Xf. Minus 1. Plus 2.S.1X1 = 3. The Fisher Relation may be written as 1 + Rn = (1 + r)(P1/P0) Solving for the foreign country.000 $500 10.10 1. (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.000 $1 million $3. Minus 1.8636 .092 = 3 1. inflation rate = 10%.040)(1.8636 The expected future spot rate is FC2.000 $2. and the foreign country inflation rate = 5%.092 = 9.05) = 1. Plus 1. and Rd0 is the nominal rate in the U. Minus 3.15 X1 = 2. 1 + Rn = (1. X0 is the current exchange rate. 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. 9.

which was an effective rate of 15.Chapter 19 International Financial Management 191 (d) The estimates are the same.000 pesos 11.100. (e) The Fisher relation for country f can be written 1 + R f Pf1 = 1+ r Pf 0 (1) and for country d.000. (a) Loan amount Interest @ 10% $1.550. .5 = 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).5% If the interest rate in Mexico were below 15. and equilibrium.10) R f 0 = 15.10 10 1 + R f 0 = 1.000 (b) Interest Rate Parity Theorem 1 + R f 0 Xf = 1 + R d0 X d X f = 10.000 pesos = 15.5 pesos/$1X 0 = 10 pesos/$ and R d0 = 10% 1 + R f 0 10.000 × 10.000.000. it would have been cheaper to borrow in pesos than to borrow in dollars at 10%.05(1. The left-hand side is IRP = 11.5%.550.5 = Effective interest paid in pesos Interest paid in pesos 1.5%.550.5% Amount received in pesos = $1. equal real rates of interest.000 10.000 × 10 = Amount repaid in pesos = $1.000 Interest rate = = Pesos received 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.000.000 pesos 1.000 $ 100.

49 (1 + 0.09/2 1. It buys FCs at the current forward rate for delivery in 180 days. (a) The company goes long on FC futures.07 FC/$ (c) FC are more expensive by the commission. It buys dollars at the current forward rate for delivery in 180 days. to pay for the machinery at that time.778.49(1.08 1. Fourth Edition 12. (b) By IRP.491128/FC (d) $10.000/($0.07/1.08 1 + 0.045 = 2.192 Copeland/Shastri/Weston • Financial Theory and Corporate Policy. n 1 + Rf 0 365 n 1 + R d0 365 $0.291 . (a) It goes long on dollar futures. so commission decreases the dollar value of the FC.08/2) = Ef (1 + 0.040 = 2.08(0. so the net result is fewer FC/$. 0.0025 = 2.600.491128/FC) = FC 20.000 FC/(2.08/2 = 2. 2.004807692) E f = $0.000.9952) Xf = 2.09/2) E0 = Ef = $0. 13.227. (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 + 0.0025 = $0.492356/1.6 million divided by the 6 month forward rate of FC per dollar (including commission costs): 4.0648379 FC/$ (d) The dollar price is FC 4.0648379 FC/$) = $2.492356/FC (c) Dollars are more expensive.361.

90 Xf 380.012) = $194. The speculator would go long in FC.683/1.28/4) (1 − 0. 191. Divide by the current exchange rate to convert this amount to dollars. they could have covered their short position.242.90 = 191.000(0. use the relationship: F0 F0 F0 + − τUS = 194. Globalcorp receives FC 380. FC 364.08/4)(1 – 0.042) = 380.000 X = 364. If the amount = X.4)] = FC 380. this means Xf (the forward exchange rate in FC/$) is lower (its value higher) than the expected future spot rate of FC/$.242 X f 114.242 = $5.939 Now compound the dollar amount for three months.939(1.Chapter 19 International Financial Management 193 14. Globalcorp realizes $194.000 which it uses to pay off the principal plus interest on the FC 384. nominal interest rate.4) 380.000 + 80. X[1 + (0. (a) First calculate the amount borrowed now in FC currency that will yield FC 380. to determine the final net payment in dollars. When Globalcorp was paid in FC.000 = 194.242 Xf 1.000/(FC1.000 in three months.90/$) = $200. If the speculator receives a risk premium for going long in FC.242 Xf X0 Xf 380.000 Substituting the correct numbers.00/$ (d) Globalcorp would have sold FC forward short. Globalcorp would receive FC 380. then: X[1 + Rn (n/365)(1 − τ)] = FC 380.000 Xf − 152.242 X f = 228.242 Xf X 0 X f amount received tax shelter on loss F0 F0 F + (τUS ) − 0 (τUS ) = 194. (c) To determine the equivalent (non-arbitrage) forward rate.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.000 in 3 months.4)] = 191.000 X(1.683 loan.683 This amount plus interest equals FC 380.000 $200.000 380.000 Xf = FC 2.4) + − = 194. .242 At the end of the three months.000 – 194.939[1 + (0.000(0.S. at the U. (b) If the payment were made immediately.

i.S.547) = (1.547 investment is equal to the amount Transcorp could have earned if the dollars had been invested in the U.12/2)(1 − . (a) Let X be the amount of FC which.000 = 192.00/$) = $190.000 E f + (0. X = [1 + (0..227 380. Transcorp would pay 380.04)](185.000)(0. The amount that must be borrowed in dollars today to convert to FC is FC 371.093.4) = $192.4)(380.094/(FC 2. (d) The speculator would short the FC forward.227 E f = $0.000 E f + 152.505.e. E1 = 0. it will yield FC 380.227 532.5042.5)(380. Ef = 0.227. including opportunity loss. This represents an insurance premium against a rise in the FC rate while Transcorp delays six months in making the FC payment.000/(FC 2.08/2)(1 – 0.00/$) = $185. . is $192.000 X = FC 371.000 E f = 268.000 in 6 months..000 E f − 76.000) E f − (0.000 X(1.547 The net cost (or opportunity lost) that results from this $185. Then. Fourth Edition 15.000 = $2.227 380. E. and insure payment abroad in six months.5042/FC Transcorp would go long in FC forward to hedge their position. If the speculator receives a premium for his short position then Ef > E1.194 Copeland/Shastri/Weston • Financial Theory and Corporate Policy.547) = $192.4)] = FC 380. he sells at a higher forward value of FC relative to the future expected spot rate.094 is invested in the foreign country today.75 If FC 371.024) = FC 380.000 The difference between this amount and their actual cost. when invested.036)(185.000 in six months. X[1 + (0. (c) E f F0 + (E f − E 0 ) F0 (τUS ) = $192. will equal FC 380.227 – 190.227 net cost (b) If payment were made immediately.g.