9

www.downloadslide.com

C H A P T E R

Risk Analysis, Real Options,

and Capital Budgeting
Chapters 7 and 8 covered the basic principles of capital budgeting, with particular emphasis
EXECUTIVE SUMMARY

on the net present value (NPV) approach. However, this is not the end of the story. Real-world
practitioners often wonder how much confidence they should place in NPV calculations. This
chapter examines sensitivity analysis, scenario analysis, break-even analysis, and Monte Carlo
simulations, all of which recognize that because it is based on estimates, NPV is really a dis-
tribution, not a single number. These techniques help the practitioner determine the degree of
confidence to be placed in a capital budgeting calculation.
Information is uncovered as a project unfolds, allowing a manager to make sequential
decisions over the life of the project. This chapter covers decision trees and real options,
capital budgeting techniques that specifically take the sequential nature of decision making
into account.

9.1  DECISION TREES

There is usually a sequence of decisions in NPV project analysis. This section intro-
duces the device of decision trees for identifying these sequential decisions.
Imagine you are the treasurer of the Solar Electronics Corporation (SEC), and
the engineering group has recently developed the technology for solar-powered jet
engines. The jet engine is to be used with 150-passenger commercial airplanes. The
marketing staff has proposed that SEC develop some prototypes and conduct test mar-
keting of the engine. A corporate planning group, including representatives from pro-
duction, marketing, and engineering, estimates that this preliminary phase will take a
year and will cost $100 million. Furthermore, the group believes there is a 75 percent
chance that the marketing tests will prove successful.
If the initial marketing tests are successful, SEC can go ahead with full-scale pro-
duction. This investment phase will cost $1,500 million. Production and sales will
occur over the next five years. The preliminary cash flow projection appears in
Table 9.1. If SEC goes ahead with investment and production on the jet engine, the NPV
at a discount rate of 15 percent (in millions) will be
5
NPV = −$1,500 + ​∑    ​​ ​  $900t 
 ______  ​​  
t=1 (1.15​)​​
= −$1,500 + $900 × ​A​  50.15   ​ 

= $1,517
 www.downloadslide.com
262 Part 2    Value and Capital Budgeting

TA B L E 9 . 1
Cash Flow Forecasts for SEC’s Jet Engine Base Case (in $ millions)*

Investment Year 1 Years 2–6

Revenues $ 6,000
Variable costs (3,000)
Fixed costs (1,791)
Depreciation    (300)
Pre-tax profit 909
Tax (Tc = 0.34)    (309)
Net profit $   600
Cash flow (EBITDA - Tax) $   900
Initial investment costs -$1,500

* Assumptions: (1) Investment is depreciated in years 2 through 6 using the straight-line method for simplicity; (2) tax rate is 34
percent; (3) the company receives no tax benefits on initial development costs.

Note that the NPV is calculated as of date 1, the date at which the investment of $1,500
million is made. Later we bring this number back to date 0.
If the initial marketing tests are unsuccessful, SEC’s $1,500 million investment has
an NPV of −$3,611 million. This figure is also calculated as of date 1. (To save space,
we will not provide the raw numbers leading to this calculation.)
Figure 9.1 displays the jet engine problem as a decision tree. If SEC decides to
conduct test marketing, there is a 75 percent probability that the test marketing will
be successful. If the tests are successful, the firm faces a second decision: whether to
invest $1,500 million in a project that yields $1,517 million NPV or to stop. If the tests
are unsuccessful, the firm faces a different decision: whether to invest $1,500 million
in a project that yields −$3,611 million NPV or to stop.

FIGURE 9.1
Decision Tree (in $ millions) for SEC
1RZ <HDU 
'HYHORSPHQW ,QLWLDO LQYHVWPHQW
DQG WHVW -
139 = 
,QYHVW

6XFFHVV
 SUREDELOLW\
'R QRW LQYHVW 139 = 
7HVW UHVXOWV
UHYHDOHG 139 = -
7HVW ,QYHVW

)DLOXUH
 SUREDELOLW\

1R WHVW 'R QRW LQYHVW 139 = 

6TXDUHV UHSUHVHQW GHFLVLRQ SRLQWV 7KH FLUFOH UHSUHVHQWV UHFHLSW RI LQIRUPDWLRQ

6(& PXVW PDNH WZR GHFLVLRQV
 :KHWKHU WR GHYHORS DQG WHVW WKH HQJLQH
 :KHWKHU WR LQYHVW LQ IXOOVFDOH SURGXFWLRQ
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 263

To review, SEC has the following two decisions to make:

1. Whether to develop and test the solar-powered jet engine.
2. Whether to invest in full-scale production following the results of the test.
One makes decisions in reverse order with decision trees. Thus we analyze the
second-stage investment of $1,500 million first. If the tests are successful, should SEC
make the second-stage investment? The answer is obviously yes, since $1,517 million
is greater than zero. If the tests are unsuccessful, should the second-stage investment
be made? Just as obviously, the answer is no, since −$3,611 million is less than zero.
Now we move back to the first stage, where the decision boils down to the ques-
tion, should SEC invest $100 million now to obtain a 75 percent chance of $1,517 mil-
lion one year later? The expected payoff evaluated at date 1 (in millions) is
Expected
   
​
payoff
= ​      
 ​  
  ​
of success( 
Probability     
× ​ 
 ​ 
 
Payoff if
successful
 
  )  ( 
Probability    
​  ​ + ​      
​
of failure
 ​ 
 
Payoff if
×​
failure
 ​  
 
​ ) 
=    (0.75        ×   $1,517)   +    (0.25      ×    $0)
= $1,138
The NPV of testing computed at date 0 (in millions) is
$1,138
NPV = −$100 + ______
​   ​ 
 
1.15
= $890
Since the NPV is a positive number, the firm should test the market for solar-powered
jet engines.1

CONCEPT • What is a decision tree?

QUESTIONS • How do decision trees handle sequential decisions?

9.2  S
 ENSITIVITY ANALYSIS, SCENARIO ANALYSIS,
AND BREAK-EVEN ANALYSIS
One thrust of this book is that NPV analysis is a superior capital budgeting technique.
In fact, because the NPV approach uses cash flows rather than profits, uses all the cash
flows, and discounts the cash flows properly, it is hard to find any theoretical fault with
it. However, in our conversations with practical business people, we hear the phrase
“a false sense of security” frequently. These people point out that the documentation
for capital budgeting proposals is often quite impressive. Cash flows are projected
down to the last thousand dollars (or even the last dollar) for each year (or even each
month). Opportunity costs and side effects are handled quite properly. Sunk costs are
ignored—also quite properly. When a high NPV appears at the bottom, one’s tempta-
tion is to say yes immediately. Nevertheless, the projected cash flow often goes unmet
in practice, and the firm ends up with a money loser.

Sensitivity Analysis and Scenario Analysis

How can the firm get the NPV technique to live up to its potential? One approach is
sensitivity analysis (a.k.a. what-if analysis and BOP analysis2), which examines how

1
We have used a discount rate of 15 percent for both the testing and the investment decisions. Per-
haps a higher discount rate should have been used for the initial test-marketing decision, which is
likely to be riskier than the investment decision.
2
BOP stands for Best, Optimistic, Pessimistic.
 www.downloadslide.com
264 Part 2    Value and Capital Budgeting

sensitive a particular NPV calculation is to changes in underlying assumptions. We

illustrate the technique with SEC’s solar-powered jet engine from the previous section.
As pointed out earlier, the cash flow forecasts for this project appear in Table 9.1. We
begin by considering the assumptions underlying revenues, costs, and after-tax cash
flows shown in the table.
Revenues  Sales projections for the proposed jet engine have been estimated by the
marketing department as
Number of jet engines sold = Market share × Size of jet engine market
3,000 = 0.30 × 10,000
Sales revenues = Number of jet engines sold × Price per engine
$6,000 million = 3,000 × $2 million
Thus, it turns out that the revenue estimates depend on three assumptions:
1. Market share
2. Size of jet engine market
3. Price per engine

Costs  Financial analysts frequently divide costs into two types: variable costs and
fixed costs. Variable costs change as the output changes, and they are zero when pro-
duction is zero. Costs of direct labour and raw materials are usually variable. It is
common to assume that a variable cost is constant per unit of output, implying that
total variable costs are proportional to the level of production. For example, if direct
labour is variable and one unit of final output requires $10 of direct labour, then 100
units of final output should require $1,000 of direct labour.
Fixed costs do not depend on the amount of goods or services produced during
the period. Fixed costs are usually measured as costs per unit of time, such as rent per
month or salaries per year. Naturally, fixed costs are not fixed forever. They are only
fixed over a predetermined time period.
The engineering department has estimated variable costs to be $1 million per
engine. Fixed costs are $1,791 million per year. The cost breakdowns are
Variable cost = Variable cost per unit × Number of jet engines sold
$3,000 million = $1 million × 3,000
Total cost before taxes = Variable cost + Fixed cost
$4,791 million = $3,000 million + $1,791 million
The above estimates for market size, market share, price, variable cost, and fixed
cost, as well as the estimate of initial investment, are presented in the middle column
of Table 9.2. These figures represent the firm’s expectations or best estimates of the
different parameters. For comparison, the firm’s analysts prepared both optimistic and
pessimistic forecasts for the different variables. These are also provided in the table.

TA B L E 9 . 2
Different Estimates for SEC’s Solar Jet Engine

Variable Pessimistic Expected or best Optimistic

Market size (per year) 5,000 10,000 20,000
Market share 20% 30% 50%
Price (per engine)* $1.9 $2 $2.2
Variable cost (per plane)* $1.2 $1 $0.8
Fixed cost (per year)* $1,891 $1,791 $1,741
Investment* $1,900 $1,500 $1,000

* In $ millions.
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 265

Standard sensitivity analysis calls for an NPV calculation for all three possibili-
ties of a single variable, along with the expected forecast for all other variables. This
procedure is illustrated in Table 9.3. For example, consider the NPV calculation of
$8,154 million provided in the upper right-hand corner of this table. This occurs when
the optimistic forecast of 20,000 units per year is used for market size. However,
the expected forecasts from Table 9.2 are employed for all other variables when the
$8,154 million figure is generated. Note that the same number—$1,517 million—appears
in each row of the middle column of Table 9.3. This occurs because the expected fore-
cast is used for the variable that was singled out, as well as for all other variables.

TA B L E 9 . 3
Net Present Value Calculations as of Date 1 (in $ millions) for the
Solar Jet Engine Using Sensitivity Analysis

Pessimistic Expected or best Optimistic

Market size -$1,802* $1,517 $8,154
Market share -696* 1,517 5,942
Price 853 1,517 2,844
Variable cost 189 1,517 2,844
Fixed cost 1,295 1,517 1,628
Investment 1,117 1,517 2,017

Under sensitivity analysis, one input is varied while all other inputs are assumed to meet their expectation. For example, an NPV of
-$1,802 occurs when the pessimistic forecast of 5,000 is used for market size. However, the expected forecasts from Table 9.2 are used
for all other variables when -$1,802 is generated.
* We assume that the other divisions of the firm are profitable, implying that a loss on this project can offset income elsewhere in the
firm, thereby reducing the overall taxes of the firm.

Table 9.3 can be used for a number of purposes. First, taken as a whole, the table can
indicate whether NPV analysis should be trusted. In other words, it reduces the false
sense of security we spoke of earlier. Suppose that NPV is positive when the expected
forecast for each variable is used. However, further suppose that every number in the
pessimistic column is highly negative and every number in the optimistic column is
highly positive. Even a single error in this forecast greatly alters the estimate, making
one leery of the NPV approach. A conservative manager might well scrap the entire
NPV analysis in this situation. Fortunately, this does not seem to be the case in Table
9.3, because all but two of the numbers are positive. Managers viewing the table will
likely consider NPV analysis to be useful for the solar-powered jet engine.
Second, sensitivity analysis shows where more information is needed. For exam-
ple, error in the estimate of investment appears to be relatively unimportant because
even under the pessimistic scenario, the NPV of $1,117 million is still highly posi-
tive. By contrast, the pessimistic forecast for market share leads to a negative NPV of
−$696 million, and a pessimistic forecast for market size leads to a substantially nega-
tive NPV of −$1,802 million. Since the effect of incorrect estimates on revenues is so
much greater than the effect of incorrect estimates on costs, more information on the
factors determining revenues might be needed.
Because of these advantages, sensitivity analysis is widely used in practice. Gra-
ham and Harvey report that slightly over 50 percent of the 392 firms in their sample
subject their capital budgeting calculations to sensitivity analysis.3 This number is
particularly large when one considers that only about 75 percent of the firms in their
sample use NPV analysis.

3
See Figure 2 of John Graham and Campbell Harvey, “The Theory and Practice of Corporate
Finance: Evidence from the Field,” Journal of Financial Economics (May/June 2001).
 www.downloadslide.com
266 Part 2    Value and Capital Budgeting

Unfortunately, sensitivity analysis also suffers from some drawbacks. For example,
sensitivity analysis may unwittingly increase the false sense of security among manag-
ers. Suppose all pessimistic forecasts yield positive NPVs. A manager might feel that
there is no way the project can lose money. Of course, the forecasters may simply have
an optimistic view of a pessimistic forecast. To combat this, some companies do not
treat optimistic and pessimistic forecasts subjectively. Rather, their pessimistic fore-
casts are always, say, 20 percent less than expected. Unfortunately, the cure in this case
may be worse than the disease, because a deviation of a fixed percentage ignores the
fact that some variables are easier to forecast than others.
In addition, sensitivity analysis treats each variable in isolation when, in reality,
the different variables are likely to be related. For example, if ineffective management
allows costs to get out of control, it is likely that variable costs, fixed costs, and invest-
ment will all rise above expectation at the same time. If the market is not receptive to
a solar-powered jet engine, both market share and price should decline together.
Managers frequently perform scenario analysis, a variant of sensitivity analysis,
to minimize this problem. Simply put, this approach examines a number of different
likely scenarios, where each scenario involves a confluence of factors. As a simple
example, consider the effect of a few airline crashes. These crashes are likely to reduce
flying in total, thereby limiting the demand for any new engines. Furthermore, even
if the crashes did not involve solar-powered aircraft, the public could become more
averse to any innovative and controversial technologies. Hence, SEC’s market share
might fall as well. Perhaps the cash flow calculations would look like those in Table
9.4 under the scenario of a plane crash. Given the calculations in the table, the NPV (in
millions) would be
−$2,023 = −$1,500 − $156 × ​A​  50 .15​ 
A series of scenarios like this might illuminate issues concerning the project better
than the standard application of sensitivity analysis would.

TA B L E 9 . 4
Cash Flow Forecast (in $ millions) under the Scenario of a Plane Crash*

Year 1 Years 2–6

Revenues $2,800
Variable costs -1,400
Fixed costs -1,791
Depreciation   −300
Pre-tax profit -691
Tax (Tc = 0.34)†     235
Net profit -456
Cash flow -156
Initial investment cost -$1,500

* Assumptions are
Market size 7,000 (70 percent of expectation)
Market share    20% (2/3 of expectation)
Forecasts for all other variables are the expected forecasts as given in Table 9.2.
†
Tax loss offsets income elsewhere in the firm.

Break-Even Analysis
Our discussion of sensitivity analysis and scenario analysis suggests that there are
many ways to examine variability in forecasts. We now present another approach,
break-even analysis. As its name implies, this approach determines the sales needed
to break even. The approach is a useful complement to sensitivity analysis, because
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 267

it also sheds light on the severity of incorrect forecasts. We calculate the break-even
point in terms of both accounting profit and present value (PV).
Accounting Profit  Net profit under four different sales forecasts is as follows:
Unit sales Net profit (in $ millions)
0 -$1,380
1,000 -720
3,000 600
10,000 5,220
A more complete presentation of costs and revenues appears in Table 9.5.
We plot the revenues, costs, and profits under the different assumptions about
sales in Figure 9.2. The revenue and cost curves cross at 2,091 jet engines. This is the
break-even point—the point where the project generates no profits or losses. As long as
sales are above 2,091 jet engines, the project will make a profit.

TA B L E 9 . 5
Revenues and Costs of Project under Different Sales Assumptions
(in $ millions, except unit sales)

Year 1 Years 2–6

Net
present
Annual Taxes* Operating value
Initial unit Variable Fixed (Tc = Net cash (evaluated
investment sales Revenues costs costs Depreciation 0.34) profit flows date 1)
$1,500 0 $     0 $      0 -$1,791 -$300 $   711 -$1,380 -$1,080 -$5,120
1,500 1,000 2,000 -1,000 -1,791 -300 371 -720 -420 -2,908
1,500 3,000 6,000 -3,000 -1,791 -300 -309 600 900 1,517
1,500 10,000 20,000 -10,000 -1,791 -300 -2,689 5,220 5,520 17,004

* Loss is incurred in the first two rows. For tax purposes, this loss offsets income elsewhere in the firm.

FIGURE 9.2
Break-Even Point Using Accounting Numbers
0LOOLRQV 
 7RWDO UHYHQXHV

7RWDO FRVWV

9DULDEOH FRVWV

 ‡
)L[HG FRVWV  LQFOXGLQJ GHSUHFLDWLRQ
 2XWSXW LQ WHUPV RI VDOHV XQLWV

This break-even point can be calculated very easily. Because the sale price is
$2 million per engine and the variable cost is $1 million per engine,4 the after-tax dif-
ference per engine is
(Sale price - Variable cost) × (1 - Tc) = ($2 million - $1 million) × (1 - 0.34)
= $0.66 million

4
Though the previous section considered both optimistic and pessimistic forecasts for sale price
and variable cost, break-even analysis uses just the expected or best estimates of these variables.
 www.downloadslide.com
268 Part 2    Value and Capital Budgeting

where Tc is the corporate tax rate of 34 percent. This after-tax difference is called the
contribution margin because each additional engine contributes this amount to after-
tax profit.
Fixed costs are $1,791 million and depreciation is $300 million, implying that the
after-tax sum of these costs is
(Fixed costs + Depreciation) × (1 - Tc) = ($1,791 million + $300 million) × (1 - 0.34)
= $1,380 million
That is, the firm incurs costs of $1,380 million, regardless of the number of sales.
Because each engine contributes $0.66 million, sales must reach the following level to
offset the above costs:
Accounting Profit Break-Even Point:
(Fixed costs + Depreciation) × (1 − ​T​c​) _____________
$1,380 million
​ ___________________________________
     ​ =   
     ​     ​ = 2,091 (9.1)
(Sale price − Variable costs) × (1 − ​Tc​​) $0.66 million
Thus, 2,091 engines is the break-even point required for an accounting profit.
Present Value  As we stated many times in the text, we are more interested in PV
than we are in net profits. Therefore, we must calculate the PV of the cash flows. Given
a discount rate of 15 percent, we have
Unit sales Net present value (in $ millions)
0 -5,120
1,000 -2,908
3,000 1,517
10,000 17,004
These NPV calculations are reproduced in the last column of Table 9.5. We can see
that the NPV is negative if SEC produces 1,000 jet engines and positive if it produces
3,000 jet engines. Obviously, the zero NPV point occurs between 1,000 and 3,000 jet
engines.
The PV break-even point can be calculated very easily. The firm originally
invested $1,500 million. This initial investment can be expressed as a five-year equiva-
lent annual cost (EAC), determined by dividing the initial investment by the appropri-
ate five-year annuity factor:

   ​  Initial investment

EAC = __________________________
    ​  Initial investment
 ​ = ________________
    ​
 
5-year annuity factor at 15% ​A​  50.15
  ​ 

$1,500 million
= _____________
  
​   ​   = $447.5 million
3.3522
Note that the EAC of $447.5 million is greater than the yearly depreciation of
$300 million. This must occur since the calculation of EAC implicitly assumes that the
$1,500 million investment could have been invested at 15 percent.
After-tax costs, regardless of output, can be viewed as
$1,528 million =
 $447.5 million + $1,791 million ×    0.66    - $300 million × 0.34
=       EAC        +   Fixed costs   × (1 - Tc) - Depreciation ×   Tc
That is, in addition to the initial investment’s EAC of $447.5 million, the firm pays fixed
costs each year and receives a depreciation tax shield each year. The depreciation tax
shield is written as a negative number since it offsets the costs in the equation. Because
each engine contributes $0.66 million to after-tax profit, it will take the following sales
to offset the above costs:
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 269

Present Value Break-Even Point:

EAC + Fixed costs × (1 − ​Tc​​) − Depreciation × ​Tc​​ _____________
$1,528 million
​ _____________________________________________
          ​     ​ = 2,315 (9.2)
 ​ =   
(Sales price − Variable costs) × (1 − ​Tc​​) $0.66 million
Thus, 2,315 engines is the break-even point from the perspective of PV.
Why is the accounting break-even point (equation (9.1)) different from the finan-
cial break-even point in equation (9.2)? When we use accounting profit as the basis
for the break-even calculation, we subtract depreciation. Depreciation for the solar jet
engines project is $300 million. If 2,091 solar jet engines are sold, SEC will generate
sufficient revenues to cover the $300 million depreciation expense plus other costs.
Unfortunately, at this level of sales, SEC will not cover the economic opportunity costs
of the $1,500 million laid out for the investment. If we take into account that the $1,500
million could have been invested at 15 percent, the true annual cost of the invest-
ment is $447.5 million and not $300 million. Depreciation understates the true costs
of recovering the initial investment. Thus, companies that break even on an account-
ing basis are really losing money. They are losing the opportunity cost of the initial
investment.

CONCEPT • What is sensitivity analysis?

QUESTIONS • Why is it important to perform a sensitivity analysis?
• What is break-even analysis?
• Describe how sensitivity analysis interacts with break-even analysis.

Break-Even Analysis, Equivalent Annual Cost,

and Capital Cost Allowance
So far, in this chapter and the previous one, all our examples with EAC and break-
even analysis have featured straight-line depreciation. While this had the advantage
of simplifying the examples, it lacks realism for Canada. Here, Example 9.1 shows
how to incorporate capital cost allowance (CCA) into break-even analysis using the
EAC approach.

EXAMPLE 13.1

Nack Trucks Inc. is expanding into the business of buying stripped-down truck platforms,
which it plans to customize to client specifications and resell. The company will rent its manu-
facturing facility for $6,000 per month. Truck platforms cost $20,000 each, and the typical
finished product sells for $42,000.
This new business line would require $60,000 in new equipment. This equipment falls
into class 8 with a CCA rate of 20 percent and would be worth about $5,000 after four years.
Nack’s tax rate is 43.5 percent, and the cost of capital is 20 percent.
There is only one major competitor, and Nack’s sales staff estimate that they could
achieve annual sales of 12 units. In order to determine if this sales level will be profitable, we
must find the NPV break-even point.
This problem is very similar to the one we solved for SEC in equation (9.2) above. The
key difference is that the depreciation here is based on CCA, so we have to replace the term
Depreciation × Tc in (9.2) with the EAC of the PV of the CCA tax shield (PVCCATS): EACPVCCATS.
This requires first calculating PVCCATS and then converting to an EAC:
 www.downloadslide.com
270 Part 2    Value and Capital Budgeting

1 + 0.5(0.20)
$60,000(0.20)(0.435) _____________
PVCCATS = ____________________
  
​      ​ ×   
​   ​  
0.20 + 0.20 1 + 0.20
$5,000(0.20)(0.435) _______
    ​ × ​  1  4 ​ 
− ___________________
​    = $11,438
0.20 + 0.20 (1.20​)​ ​
$11,438 ________ $11,438
​
EA​CPVCCATS​ = ________
​   ​ 
= ​ 
   ​ 
= $4,418.38
 
​A​  40 .20​  2.5887

Next we find the EAC of the investment as

[$60,000 - $5,000/(1.20)4]/2.5887 = $22,245.90
Note that your EAC of the investment may be slightly different due to rounding. Fixed costs
after tax are the rent of $6,000 × 12 = $72,000 per year times (1 − 0.435), or $40,680,
and the after-tax contribution margin is ($42,000 − $20,000)(1 − 0.435) = $12,430.
We can now rewrite equation (9.2) and fill in the numbers:
Present Value Break-Even Point with Capital Cost Allowance:
EAC + Fixed costs × (1 − ​Tc​​) − EA​CPVCCATS
​ ​ $22,245.90 + $40,680 − $4,418.38
​ _____________________________________
    
      ​ = __________________________________
​ 
         ​
(Sale price − Variable costs) × (1 − ​Tc​​) $12,430
= 4.71 or 5 trucks

Our calculations show that the break-even point is 5 trucks. This is well below targeted sales
of 12 trucks, so the expansion looks promising.

9.3  MONTE CARLO SIMULATION

Both sensitivity analysis and scenario analysis attempt to answer the question, “What
if?” However, while both analyses are frequently used in the real world, each has its
own limitations. Sensitivity analysis allows only one variable to change at a time. By
contrast, many variables are likely to move at the same time in the real world. Scenario
analysis follows specific scenarios, such as changes in inflation, government regula-
tion, or the number of competitors. While this methodology is often quite helpful, it
cannot cover all sources of variability. In fact, projects are likely to exhibit a lot of vari-
ability under just one economic scenario.
Monte Carlo simulation is a further attempt to model real-world uncertainty.
This approach takes its name from the famous European casino, because it analyzes
projects the way one might analyze gambling strategies. Imagine a serious blackjack
player who wonders if he should take a third card whenever his first two cards total 16.
Most likely, a formal mathematical model would be too complex to be practical here.
However, he could play thousands of hands in a casino, sometimes drawing a third
card when his first two cards add to 16 and sometimes not drawing that third card. He
could compare his winnings (or losses) under the two strategies to determine which
was better. Of course, since he would probably lose a lot of money performing this test
in a real casino, simulating the results from the two strategies on a computer might be
cheaper. Monte Carlo simulation of capital budgeting projects is in this spirit.
Imagine that Backyard Barbecues Inc. (BBI), a manufacturer of both charcoal and
gas grills, has the blueprint for a new grill that cooks with compressed hydrogen.
The CFO, Ellen H. Comiskey, being dissatisfied with simpler capital budgeting tech-
niques, wants a Monte Carlo simulation for this new grill. A consultant specializing
in the Monte Carlo approach, Lester Mauney, takes her through the five basic steps
of the method.
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 271

Step 1: Specify the Basic Model

Lester breaks up cash flow into three components: annual revenue, annual costs, and
initial investment. The revenue in any year is viewed as
        Market share of BBI’s
Number of grills sold           Price per
     
  
​    ​ ×    
​ 
    ​ ×   
​ 
   ​ (9.3)
by entire industry hydrogen grill (in percent) hydrogen grill
The cost in any year is viewed as
Fixed
        Variable Marketing    
Selling
​ 
  
   ​ +        
​ 
  
   ​ +     
​ +​
 ​ 
   ​ 
 
manufacturing costs manufacturing costs costs costs
Initial investment is viewed as
Cost of patent + Test-marketing costs + Cost of production facility

Step 2: Specify a Distribution for Each Variable in the Model

Here comes the hard part. Let’s start with revenue, which has three components in
(9.3). The consultant first models overall market size, that is, the number of grills sold
by the entire industry. The trade publication Outdoor Food (OF ) reported that 10 mil-
lion grills of all types were sold in North America last year and it forecasts sales of
10.5 million next year. Lester, using OF ’s forecast and his own intuition, creates the
following distribution for next year’s sales of grills by the entire industry:

Probability 20% 60% 20%

Next year’s industrywide unit sales 10 million 10.5 million 11 million

The tight distribution here reflects the slow but steady historical growth in the grill
market. This probability distribution is graphed in Panel A of Figure 9.3.
Lester realizes that estimating the market share of BBI’s hydrogen grill is more dif-
ficult. Nevertheless, after a great deal of analysis, he determines the distribution of next
year’s market share to be as follows:

Probability 10% 20% 30% 25% 10% 5%

Market share of BBI’s hydrogen grill next year 1% 2% 3% 4% 5% 8%

While the consultant assumed a symmetrical distribution for industrywide unit sales,
he believes a skewed distribution makes more sense for the project’s market share.
In his mind, there is always the small possibility that sales of the hydrogen grill will
really take off. This probability distribution is graphed in Panel B of Figure 9.3.
The above forecasts assume that unit sales for the overall industry are unrelated
to the project’s market share. In other words, the two variables are independent of each
other. Lester reasons that while an economic boom might increase industrywide grill
sales and a recession might decrease them, the project’s market share is unlikely to be
related to economic conditions.
Now Lester must determine the distribution of price per grill. Ellen, the CFO,
informs him that the price will be in the area of $200 per grill, given what competitors
are charging. However, the consultant believes that the price per hydrogen grill will
almost certainly depend on the size of the overall market for grills. As in any business,
you can usually charge more if demand is high.
After rejecting a number of complex models for price, Lester settles on the follow-
ing specification:
Next year’s price Industrywide unit sales
​       
  ​ = $190 + $1 ×         
   ​
        ​ ± $3 (9.4)
per hydrogen grill (in millions)
 www.downloadslide.com
272 Part 2    Value and Capital Budgeting

FIGURE 9.3
Probability Distributions for Industrywide Unit Sales, Market Share
of BBI’s Hydrogen Grill, and Price of Hydrogen Grill
3DQHO $ 3DQHO %


 



3UREDELOLW\

3UREDELOLW\
 


 
  



        
1H[W \HDU¶V LQGXVWU\ZLGH 7KH PDUNHW VKDUH RI %%,¶V
XQLW VDOHV LQ PLOOLRQV K\GURJHQ JULOO QH[W \HDU

3DQHO &
+LJK GHPDQG
 SUREDELOLW\

1H[W \HDU¶V SULFH SHU K\GURJHQ JULOO

([SHFWHG


/RZ GHPDQG
  SUREDELOLW\






   
1H[W \HDU¶V LQGXVWU\ZLGH
XQLW VDOHV LQ PLOOLRQV

)RU HDFK RI WKH WKUHH YDULDEOHV D GUDZLQJ LV JHQHUDWHG E\ FRPSXWHU VLPXODWLRQ ,Q DGGLWLRQ
SULFH SHU JULOO GHSHQGV RQ LQGXVWU\ZLGH XQLW VDOHV

The grill price in (9.4) depends on the unit sales of the industry. In addition, random
variation is modelled via the term “±$3,” where a drawing of +$3 and a drawing of −$3
each occur 50 percent of the time. For example, if industrywide unit sales are 11 mil-
lion, the price per grill will be either
$190 + $11 + $3 = $204   (50% probability)
or
$190 + $11 - $3 = $198   (50% probability)
The relationship between the price of a hydrogen grill and industrywide unit sales
is graphed in Panel C of Figure 9.3.
The consultant now has distributions for each of the three components of next
year’s revenue. However, he needs distributions for future years as well. Using forecasts
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 273

from OF and other publications, Lester forecasts the distribution of growth rates for
the entire industry over the second year to be as follows:

Probability 20% 60% 20%

Growth rate of industrywide unit sales in second year 1% 3% 5%

Given both the distribution of next year’s industrywide unit sales and the distribution
of growth rates for this variable over the second year, we can generate the distribution
of industrywide unit sales for the second year. A similar extension should give Lester a
distribution for later years as well, though we won’t go into the details here. And, just
as the consultant extended the first component of revenue (industrywide unit sales) to
later years, he would want to do the same thing for market share and unit price.
The above discussion shows how the three components of revenue can be mod-
elled. Step 2 is complete once the components of cost and of investment are modelled
in a similar way. Special attention must be paid to the interactions between variables
here, since ineffective management will likely allow the different cost components to
rise together. However, since you are probably getting the idea now, we will skip the
rest of this step.

Step 3: The Computer Draws One Outcome

As we said above, next year’s revenue in our model is the product of three components.
Imagine that the computer randomly picks industrywide unit sales of 10 million, a
market share for BBI’s hydrogen grill of 2 percent, and a +$3 random price variation.
Given these drawings, next year’s price per hydrogen grill will be
$190 + $10 + $3 = $203
and next year’s revenue for BBI’s hydrogen grill will be
10 million × 0.02 × $203 = $40.6 million
Of course, we are not done with the entire outcome yet. We would have to per-
form drawings for revenue in each future year. In addition, we would perform draw-
ings for costs in each future year. Finally, a drawing for initial investment would have
to be made as well. In this way, a single outcome would generate a cash flow from the
project in each future year.
How likely is it that the specific outcome above would be drawn? We can answer
this because we know the probability of each component. Since industry sales of
$10 million has a 20 percent probability, a market share of 2 percent also has a 20 per-
cent probability, and a random price variation of +$3 has a 50 percent probability, the
probability of these three drawings together in the same outcome is
0.02 = 0.20 × 0.20 × 0.50 (9.5)
Of course, the probability would get even smaller once drawings for future revenues,
future costs, and the initial investment were included in the outcome.
This step generates the cash flow for each year from a single outcome. What we
are ultimately interested in is the distribution of cash flow each year across many
outcomes. We ask the computer to randomly draw over and over again to give us this
distribution, which is just what is done in the next step.

Step 4: Repeat the Procedure

While the above three steps generate one outcome, the essence of Monte Carlo simula-
tion is repeated outcomes. Depending on the situation, the computer may be called on
to generate thousands or even millions of outcomes. The result of all these drawings is
a distribution of cash flow for each future year. This distribution is the basic output of
Monte Carlo simulation.
 www.downloadslide.com
274 Part 2    Value and Capital Budgeting

Consider Figure 9.4. Here, repeated drawings have produced the simulated distri-
bution of the third year’s cash flow. There would be, of course, a distribution like the
one in this figure for each future year. This leaves us with just one more step.

FIGURE 9.4
Simulated Distribution of the Third Year’s Cash Flow for
BBI’s New Hydrogen Grill
1XPEHU RI RXWFRPHV

 &DVK I ORZ
,Q 0RQWH &DUOR VLPXODWLRQV UHSHDWHG VDPSOLQJ RI DOO WKH YDULDEOHV IURP D
VSHFLILF PRGHO JHQHUDWHV D VWDWLVWLFDO GLVWULEXWLRQ

Step 5: Calculate Net Present Value

Given the distribution of cash flow for the third year in Figure 9.4, one can determine
the expected cash flow for this year. In a similar manner, one can also determine the
expected cash flow for each future year and then calculate the NPV of the project by
discounting these expected cash flows at an appropriate rate.
Monte Carlo simulation is often viewed as a step beyond either sensitivity analy-
sis or scenario analysis. Interactions between the variables are explicitly specified in
Monte Carlo, so, at least in theory, this methodology provides a more complete analy-
sis. And, as a by-product, having to build a precise model deepens the forecaster’s
understanding of the project.
Since Monte Carlo simulations have been around for at least 35 years, you might
think that most firms would be performing them by now. Surprisingly, this does not
seem to be the case. In our experience, executives are frequently skeptical of all the
complexity. It is difficult to model either the distributions of all variables or the inter-
actions between variables. For example, a company producing both oil and natural gas
in western Canada would require an estimate of the correlation between the prices of
oil and natural gas in order to implement simulation. While these prices were highly
correlated historically, they have decoupled in recent years as oil prices have gone
higher and gas prices have stagnated. While using a historical estimate might seem
reasonable, in this case it would have led to serious error. In addition, the computer
output is often devoid of economic intuition. Thus, while Monte Carlo simulations are
used in certain real-world situations,5 the approach is not likely to be the wave of the
future. In fact, Graham and Harvey report that only about 15 percent of the firms in
their sample use capital budgeting simulations.6

5
More than perhaps any other, the pharmaceutical industry has pioneered applications of this
methodology. For example, see Nancy A. Nichols, “Scientific Management at Merck: An Interview
with CFO Judy Lewent,” Harvard Business Review (January/February 1994).
6
See Figure 2 of Graham and Harvey, op. cit.
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 275

9.4  REAL OPTIONS

In Chapter 7, we stressed the superiority of NPV analysis over other approaches when
valuing capital budgeting projects. However, both scholars and practitioners have
pointed out problems with NPV. The basic idea here is that NPV analysis, as well as all
the other approaches in Chapter 7, ignores the adjustments that a firm can make after
a project is accepted and uncertainty surrounding cash flow estimates is resolved.
These adjustments are called real options.7 In this respect, NPV underestimates the
true value of a project, particularly in industries such as mining, oil and gas, phar-
maceuticals, and biotechnology, where investments are large and uncertainty about
future outcomes makes the flexibility in real options valuable.8 NPV’s conservatism
here is best explained through a series of examples.

The Option to Expand

Connie Willig, an entrepreneur, recently learned of a chemical treatment that causes
water to freeze at 30°C, rather than 0°C. Of all the many practical applications for
this treatment, Connie liked more than anything else the idea of hotels made of ice.
Connie estimated the annual cash flows from a single ice hotel to be $2 million, based
on an initial investment of $12 million. She felt that 20 percent was an appropriate
discount rate, given the risk of this new venture. Assuming that the cash flows were
perpetual, Connie determined the NPV of the project to be
−$12,000,000 + $2,000,000/0.20 = −$2 million
Most entrepreneurs would have rejected this venture, given its negative NPV. But
Connie is not your typical entrepreneur. She reasoned that NPV analysis missed a hid-
den source of value. While she was pretty sure that the initial investment would cost
$12 million, there was some uncertainty concerning annual cash flows. Her cash flow
estimate of $2 million per year actually reflected her belief that there was a 50 percent
probability that annual cash flows would be $3 million and a 50 percent probability
that annual cash flows would be $1 million.
The NPV calculations for the two forecasts are as follows:
Optimistic forecast: -$12 million + $3 million/0.20 = $3 million
Pessimistic forecast: -$12 million + $1 million/0.20 = -$7 million
On the surface, this new calculation doesn’t seem to help Connie very much since an
average of the two forecasts yields an NPV for the project of
50% × $3 million + 50% × (-$7 million) = -$2 million
(which is just the value she calculated in the first place).
However, if the optimistic forecast turned out to be correct, Connie would want
to expand. If she believed that there were, say, 10 locations in the country that could
support an ice hotel, the true NPV of the venture would be
50% × 10 × $3 million + 50% × (-$7 million) = $11.5 million
The idea here, which is represented in Figure 9.5, is both basic and universal. The
entrepreneur has the option to expand if the pilot location is successful. For example,
think of all the people who start restaurants, most of them ultimately failing. These

7
To obtain precise estimates of real option value, it is necessary to use the binomial option pric-
ing model. This is covered in depth in Chapter 24. In introducing the topic here, we take a simpler
approach based on the probabilities of different scenarios.
8
H. K. Baker, Shantanu Dutta, and Samir Saadi. “Management Views on Real Options in Capital
Budgeting,” Journal of Applied Finance 21 (1) (2011).
 www.downloadslide.com
276 Part 2    Value and Capital Budgeting

individuals are not necessarily overly optimistic. They may realize the likelihood of
failure but go ahead anyway because of the small chance of starting the next McDon-
ald’s or Burger King.

FIGURE 9.5
Decision Tree for Ice Hotel
,I VXFFHVVIXO
([SDQG

%XLOG ILUVW
LFH KRWHO

'R QRW H[SDQG

,I XQVXFFHVVIXO

The Option to Abandon

Managers also have the option to abandon existing projects. While abandonment may
seem cowardly, it can often save companies a great deal of money. Because of this, the
option to abandon increases the value of any potential project.
The above example on ice hotels, which illustrated the option to expand, can
also illustrate the option to abandon. To see this, imagine that Connie now believes
that there is a 50 percent probability that annual cash flows will be $6 million and a
50 percent probability that annual cash flows will be −$2 million. The NPV calcula-
tions under the two forecasts become
Optimistic forecast: -$12 million + $6 million/0.2 = $18 million
Pessimistic forecast: -$12 million - $2 million/0.2 = -$22 million
yielding an NPV for the project of
50% × $18 million + 50% × (−$22 million) = −$2 million (9.6)
Furthermore, now imagine that Connie wants to own, at most, just one ice hotel, imply-
ing that there is no option to expand. Since the NPV in (9.6) is negative, it looks as if
she will not build the hotel.
But things change when we consider the abandonment option. As of date 1, the
entrepreneur will know which forecast has come true. If cash flows equal those under
the optimistic forecast, Connie will keep the project alive. If, however, cash flows equal
those under the pessimistic forecast, she will abandon the hotel. Knowing these pos-
sibilities ahead of time, the NPV of the project becomes
50% × $18 million + 50% × (-$12 million - $2 million/1.20) = $2.17 million
Since Connie abandons after experiencing the cash flow of −$2 million at date 1, she
does not have to endure this outflow in any of the later years. Because the NPV is now
positive, Connie will accept the project.
The example here is clearly a stylized one. While many years may pass before a
project is abandoned in the real world, our ice hotel was abandoned after just one year.
Further, in practice, instead of abandoning a project, a firm can exercise an option to
contract by scaling it down. And, while salvage values generally accompany aban-
donment, we assumed no salvage value for the ice hotel. Nevertheless, abandonment
options are pervasive in the real world.
For example, consider the moviemaking industry. As shown in Figure 9.6, mov-
ies begin with either the purchase or the development of a script. A completed script
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 277

might cost a movie studio a few million dollars and potentially lead to actual produc-
tion. However, the great majority of scripts (perhaps well in excess of 80 percent) are
abandoned. Why would studios abandon scripts that they had commissioned in the
first place? While the studios know ahead of time that only a few scripts will be prom-
ising, they don’t know which ones. Thus, they cast a wide net, commissioning many
scripts to get a few good ones. And the studios must be ruthless with the bad scripts,
since the expenditure here pales in comparison to the huge losses from producing a
bad movie.

FIGURE 9.6
The Abandonment Option in the Movie Industry
+LJK ER[
RIILFH UHYHQXH 0RUH
DGYHUWLVLQJ

&RVW
FRQWDLQPHQW 5HOHDVH XSRQ
FRPSOHWLRQ

*RRG VFULSW 3URGXFH 1R IXUWKHU

PRYLH /RZ ER[ DGYHUWLVLQJ
RIILFH UHYHQXH

&RPPLVVLRQ $EDQGRQ
VFULSW /DUJH SULRU WR
FRVW RYHUUXQ FRPSOHWLRQ

$EDQGRQ
%DG VFULSW
0RYLH VWXGLRV KDYH DEDQGRQPHQW RSWLRQV WKURXJKRXW WKH SURGXFWLRQ RI D PRYLH

The few lucky scripts will then move into production, where costs might be bud-
geted in the tens of millions of dollars, if not much more. At this stage, the dreaded
phrase is that on-location production gets “bogged down,” creating cost overruns. But
the studios are equally ruthless here. Should these overruns become excessive, produc-
tion is likely to be abandoned in midstream. Interestingly, abandonment almost always
occurs due to high costs, not due to the fear that the movie won’t be able to find an
audience. Little information on that score will be obtained until the movie is actually
released.
Release of the movie is accompanied by significant advertising expenditures, per-
haps in the range of $10 to $20 million. Box office success in the first few weeks is
likely to lead to further advertising expenditures. Again, the studio has the option,
but not the obligation, to increase advertising here. It also has an option to produce a
sequel.
Moviemaking is one of the riskiest businesses around, with studios receiving hun-
dreds of millions of dollars in a matter of weeks from a blockbuster while receiving
practically nothing during this period from a flop. The above abandonment options
contain costs that might otherwise bankrupt the industry.

Timing Options
One often finds urban land that has been vacant for many years. Yet this land is bought
and sold from time to time. Why would anyone pay a positive price for land that has
no source of revenue? Certainly one could not arrive at this positive value through NPV
analysis. However, the paradox can easily be explained in terms of real options.
 www.downloadslide.com
278 Part 2    Value and Capital Budgeting

Suppose that the land’s highest and best use is as an office building. Total con-
struction costs for the building are estimated to be $1 million. Currently, net rents
(after all costs) are estimated to be $90,000 per year in perpetuity and the discount rate
is 10 percent. The NPV of this proposed building would be
-$1 million + $90,000/0.10 = -$100,000
Since this NPV is negative, one would not currently want to build. In addition, it
appears as if the land is worthless. However, suppose that the government is planning
a bid for the Summer Olympics. Office rents will likely increase if the bid succeeds.
In this case, the property’s owner might want to erect the office building after all. Con-
versely, office rents will remain the same, or even fall, if the bid fails. The owner will
not build in this case.
We say that the property owner has a timing option. While he does not currently
want to build, he will want to build in the future should rents in the area rise sub-
stantially. This timing option explains why vacant land often has value. While there
are costs, such as taxes, from holding raw land, the value of an office building after a
substantial rise in rents may more than offset these holding costs. Of course, the exact
value of the vacant land depends on both the probability of success in the Olympic bid
and the extent of the rent increase. Figure 9.7 illustrates this timing option.

FIGURE 9.7
Decision Tree for Vacant Land
5HQWV ULVH VXEVWDQWLDOO\
(UHFW RIILFH EXLOGLQJ

'R QRW EXLOG \HW VLQFH

UHQWV DUH WRR ORZ

5HQWV HLWKHU VWD\

WKH VDPH RU IDOO
'R QRW EXLOG \HW
9DFDQW ODQG PD\ KDYH YDOXH WRGD\ VLQFH WKH RZQHU FDQ HUHFW D SURILWDEOH RIILFH
EXLOGLQJ LI UHQWV ULVH

Mining operations almost always provide timing options as well. Suppose you
own a copper mine where the cost of mining each tonne of copper exceeds the sales
revenue. It’s a no-brainer to say that you would not want to mine the copper currently.
And since there are costs of ownership such as property taxes, insurance, and security,
you might actually want to pay someone to take the mine off your hands. However, we
would caution you not to do so hastily. Copper prices in the future might very well
increase enough so that production is profitable. Given that possibility, you could
likely find someone to pay a positive price for the property today.

Real Options in the Real World

As mentioned previously, common real options include the option to expand, the
option to abandon, the option to contract, and the option to delay a project. Although
we have explained the value real options add to the capital budgeting decision when
faced with uncertainty, it is surprising to see that real options are not readily utilized
in practice. In a survey of Fortune 1000 companies conducted in 2007, it was found that
out of the 279 responses received, only 40 managers used real options. The primary
reasons given for not using real options were a lack of top management support and
the complexity and required expertise needed. These results were duplicated using
847 Canadian firms listed on the Toronto Stock Exchange. In addition to its infrequent
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 279

usage, it was found that large firms utilized real options more frequently than small
firms because they had access to the expertise required to use real options.9
For example, Air Canada utilized real options as it emerged from its bankruptcy
protection in 2004 and underwent major restructuring. As part of its business restruc-
turing, the airline decided to expand its overseas routes, which required investments
in 96 new jetliners. However, due to the riskiness of such investments, instead of imme-
diately buying the entire fleet of planes from Boeing, Air Canada purchased a frac-
tion of the 96 jetliners immediately, followed by an option to purchase the remaining
planes in the future, contingent on the success of its initial investments. In this case,
real options provided value in facing uncertainty and helped to mitigate risk. Such
flexibility is a value that is completely missed by traditional capital budgeting meth-
ods like NPV.

CONCEPT • What are the different types of real options?

QUESTIONS • Why does traditional NPV analysis tend to underestimate the true value of a capital
project?

9.5
SUMMARY AND CONCLUSIONS
This chapter discusses a number of practical applications of capital budgeting.
1. Though net present value (NPV) is the best capital budgeting approach conceptually,
it has been criticized in practice for providing managers with a false sense of security.
Sensitivity analysis shows NPV under varying assumptions, giving managers a better
feel for the project’s risks. Unfortunately, sensitivity analysis modifies only one variable
at a time, while many variables are likely to vary together in the real world. Scenario
analysis examines a project’s performance under different scenarios (e.g., war break-
ing out or oil prices skyrocketing). Finally, managers want to know how bad forecasts
must be before a project loses money. Break-even analysis calculates the sales figure at
which the project breaks even. Though break-even analysis is frequently performed on
an accounting profit basis, we suggest that an NPV basis is more appropriate.
2. Monte Carlo simulation begins with a model of the firm’s cash flows, based on both
the interactions between different variables and the movement of each individual vari-
able over time. Random sampling generates a distribution of these cash flows for each
period, leading to an NPV calculation.
3. We analyze the hidden options in capital budgeting, such as the option to expand, the
option to abandon, the option to contract, and timing options. We return to this topic
in Chapter 23.

KEY TERMS Break-even analysis 266 Fixed costs 264 Scenario analysis 266
Contribution margin 268 Monte Carlo simulation 270 Sensitivity analysis 263
Decision trees 261 Real options 275 Variable costs 264

9
H. K. Baker, Shantanu Dutta, and Samir Saadi, op. cit.
 www.downloadslide.com
280 Part 2    Value and Capital Budgeting

QUESTIONS & Sensitivity Analysis and Break-Even Point

PROBLEMS 9.1 Lockhart Homebuilders is evaluating a project that costs $724,000, has an eight-year
life, and has no salvage value. Assume that depreciation is straight-line to zero over
the life of the project. Sales are projected at 75,000 units per year. Price per unit is
$39, variable cost per unit is $23, and fixed costs are $850,000 per year. The tax rate
is 35 percent, and Lockhart requires a 15 percent return on this project.
a. Calculate the accounting break-even point.
b. Calculate the base-case cash flow and NPV. What is the sensitivity of NPV to
changes in the sales figure? Explain what your answer tells you about a 500-unit
decrease in projected sales.
c. What is the sensitivity of operating cash flow to changes in the variable cost
figure? Explain what your answer tells you about a $1 decrease in estimated vari-
able costs.
Scenario Analysis
9.2 In Problem 9.1, suppose the projections given for price, quantity, variable costs, and
fixed costs are all accurate to within ±10 percent. Calculate the best-case and worst-
case NPV figures.
Calculating Break-Even
9.3 In each of the following cases, find the unknown variable. Ignore taxes.
Accounting Unit variable
break-even Unit price cost Fixed costs Depreciation
95,300 $41 $30 $  820,000 ?
143,806 ? 56 2,750,000 $1,150,000
7,835 97 ? 160,000 105,000

Option to Wait
9.4 Your company is deciding whether to invest in a new machine. The new machine
will increase cash flow by $340,000 per year. You believe the technology used in
the machine has a 10-year life; in other words, no matter when you purchase the
machine, it will be obsolete 10 years from today. The machine is currently priced at
$1,800,000. The cost of the machine will decline by $130,000 per year until it reaches
$1,150,000, where it will remain. If your required return is 12 percent, should you
purchase the machine? If so, when should you purchase it?
Decision Trees
9.5 Ang Electronics Inc. has developed a new DVD-R. If the DVD-R is successful, the
PV of the payoff (when the product is brought to market) is $22 million. If the
DVD-R fails, the PV of the payoff is $9 million. If the product goes directly to
market, there is a 50 percent chance of success. Alternatively, Ang can delay the
launch by one year and spend $1.5 million to test market the DVD-R. Test market-
ing would allow the firm to improve the product and increase the probability of
success to 80 percent. The appropriate discount rate is 11 percent. Should the firm
conduct test marketing?
9.6 The manager for a growing firm is considering launching a new product. If the
product goes directly to market, there is a 50 percent chance of success. For
$175,000, the manager can conduct a focus group that will increase the product’s
chance of success to 65 percent. Alternatively, the manager has the option to pay a
consulting firm $390,000 to research the market and refine the product. The con-
sulting firm successfully launches new products 80 percent of the time. If the firm
successfully launches the product, the payoff will be $1.9 million. If the product is
a failure, the NPV is zero. Which action will result in the highest expected payoff
to the firm?
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 281

9.7 B&B has a new baby powder ready to market. If the firm goes directly to the market
with the product, there is only a 55 percent chance of success. However, the firm can
conduct customer segment research, which will take a year and cost $1.2 million. By
going through research, B&B will be able to better target potential customers and will
increase the probability of success to 70 percent. If successful, the baby powder will
bring a PV profit (at time of initial selling) of $19 million. If unsuccessful, the PV payoff
is only $6 million. Should the firm conduct customer segment research or go directly
to market? The appropriate discount rate is 15 percent.
Sensitivity Analysis
9.8 Consider a four-year project with the following information: initial fixed asset
investment = $480,000; straight-line depreciation to zero over the four-year life;
zero salvage value; price = $37; variable costs = $23; fixed costs = $195,000; quantity
sold = 90,000 units; tax rate = 34 percent. How sensitive is operating cash flow to
changes in quantity sold?
Project Analysis
9.9 You are considering a new product launch. The project will cost $820,000, have a
four-year life, and have no salvage value; depreciation is at a CCA rate of 20%. Sales
are projected at 450 units per year, price per unit will be $18,000, variable cost per
unit will be $15,400, and fixed costs will be $610,000 per year. The required return
on the project is 15 percent, and the relevant tax rate is 35 percent. Assets will
remain in the CCA class after the end of the project.
a. Based on your experience, you think the unit sales, variable cost, and fixed cost
projections given here are probably accurate to within ±10 percent. What are
the upper and lower bounds for these projections? What is the base-case NPV?
What are the best-case and worst-case scenarios?
b. Evaluate the sensitivity of your base-case NPV to changes in fixed costs.
c. What is the accounting break-even level of output for this project?
9.10 McGilla Golf has decided to sell a new line of golf clubs. The clubs will sell for $875
per set and have a variable cost of $430 per set. The company has spent $150,000 for
a marketing study that determined the company will sell 60,000 sets per year for
seven years. The marketing study also determined that the company will lose sales
of 12,000 sets of its high-priced clubs. The high-priced clubs sell at $1,100 and have
variable costs of $620. The company will also increase sales of its cheap clubs by
15,000 sets. The cheap clubs sell for $400 and have variable costs of $210 per set. The
fixed costs each year will be $9,300,000. The company has also spent $1,000,000 on
research and development for the new clubs. The plant and equipment required will
cost $29,400,000 and qualify for depreciation at a CCA rate of 20 percent. The new
clubs will also require an increase in net working capital of $1,400,000 that will be
returned at the end of the project. The tax rate is 40 percent, and the cost of capital
is 14 percent. Calculate the payback period, the NPV, and the IRR. Assets will remain
in the CCA class after the end of the project.
Sensitivity Analysis
9.11 Refer to Problem 9.10. McGilla Golf would like to know the sensitivity of NPV to
changes in the price of the new clubs and the quantity of new clubs sold. What is the
sensitivity of the NPV to each of these variables?
Abandonment Value
9.12 Your company is examining a new project. It expects to sell 9,000 units per year at
$35 net cash flow apiece for the next 10 years. In other words, the annual operating
cash flow is projected to be $35 × 9,000 = $315,000. The relevant discount rate is
16 percent, and the initial investment required is $1,350,000.
 www.downloadslide.com
282 Part 2    Value and Capital Budgeting

a. What is the base-case NPV?

b. After the first year, the project can be dismantled and sold for $950,000. If
expected sales are revised based on the first year’s performance, when would it
make sense to abandon the investment? In other words, at what level of expected
sales would it make sense to abandon the project?
c. Explain how the $950,000 abandonment value can be viewed as the opportunity
cost of keeping the project in one year.
Abandonment
9.13 In Problem 9.12, suppose you think it is likely that expected sales will be revised
upward to 11,000 units if the first year is a success and revised downward to 4,000
units if the first year is not a success.
a. If success and failure are equally likely, what is the NPV of the project? Consider
the possibility of abandonment in answering.
b. What is the value of the option to abandon?
Abandonment and Expansion
9.14 In Problem 9.12, suppose the scale of the project can be doubled in one year in the
sense that twice as many units can be produced and sold. Naturally, expansion
would be desirable only if the project were a success. This implies that if the project
is a success, projected sales after expansion will be 22,000. Again, assuming that
success and failure are equally likely, what is the NPV of the project? Note that aban-
donment is still an option if the project is a failure. What is the value of the option
to expand?
Break-Even Analysis
9.15 Your buddy comes to you with a surefire way to make some quick money and help
pay off your student loans. His idea is to sell T-shirts with the words “I get” on them.
“You get it?” He says, “You see all those bumper stickers and T-shirts that say ‘got
milk’ or ‘got surf.’ So this says, ‘I get.’ It’s funny! All we have to do is buy a used silk
screen press for $5,600 and we are in business!” Assume there are no fixed costs,
and you depreciate the $5,600 in the first period. Taxes are 30 percent. What is the
accounting break-even point if each shirt costs $4.50 to make and you can sell them
for $10 apiece?
Decision Trees
9.16 Young screenwriter Carl Draper has just finished his first script. It has action, drama,
and humour, and he thinks it will be a blockbuster. He takes the script to every
movie studio in town and tries to sell it, but to no avail. Finally, ACME studios offers
to buy the script for either (a) $12,000 or (b) 1 percent of the movie’s profits. The stu-
dio will have to make two decisions. First is to decide whether the script is good or
bad, and second is to decide whether the movie is good or bad. There is a 90 percent
chance that the script is bad. If it is bad, the studio does nothing more and throws
the script out. If the script is good, they will shoot the movie. After the movie is shot,
the studio will review it, and there is a 70 percent chance that the movie is bad. If the
movie is bad, the movie will not be promoted and will not turn a profit. If the movie
is good, the studio will promote heavily; the average profit for this type of movie is
$20 million. Carl rejects the $12,000 and says he wants the 1 percent of profits. Was
this a good decision by Carl?
Option to Wait
9.17 Hickock Mining is evaluating when to open a gold mine. The mine has 48,000 ounces
of gold left that can be mined, and mining operations will produce 6,000 ounces per
year. The required return on the gold mine is 12 percent, and it will cost $34 mil-
lion to open the mine. When the mine is opened, the company will sign a contract
that will guarantee the price of gold for the remaining life of the mine. If the mine
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 283

is opened today, each ounce of gold will generate an after-tax cash flow of $1,400 per
ounce. If the company waits one year, there is a 60 percent probability that the con-
tract price will generate an after-tax cash flow of $1,600 per ounce and a 40 percent
probability that the after-tax cash flow will be $1,300 per ounce. What is the value of
the option to wait?
Abandonment Decisions
9.18 Allied Products Inc. is considering a new product launch. The firm expects to have
an annual operating cash flow of $10.5 million for the next 10 years. Allied Products
uses a discount rate of 13 percent for new product launches. The initial investment is
$51 million. Assume that the project has no salvage value at the end of its economic
life.
a. What is the NPV of the new product?
b. After the first year, the project can be dismantled and sold for $31 million. If the
estimates of remaining cash flows are revised based on the first year’s experi-
ence, at what level of expected cash flows does it make sense to abandon the
project?
Expansion Decisions
9.19 Applied Nanotech is thinking about introducing a new surface-cleaning machine.
The marketing department has estimated that Applied Nanotech can sell 15 units
per year at $305,000 net cash flow per unit for the next five years. The engineering
department has estimated that developing the machine will take a $15 million initial
investment. The finance department has estimated that a 16 percent discount rate
should be used.
a. What is the base-case NPV?
b. If unsuccessful, after the first year the project can be dismantled and will have
an after-tax salvage value of $11 million. Also, after the first year, expected cash
flows will be revised up to 20 units per year or to 0 units, with equal probability.
What is the revised NPV?
Scenario Analysis
9.20 You are the financial analyst for a tennis racquet manufacturer. The company is con-
sidering using a graphite-like material in its tennis racquets. The company has esti-
mated the information in the following table about the market for a racquet with the
new material. The company expects to sell the racquet for six years. The equipment
required for the project has no salvage value. The required return for projects of this
type is 13 percent, and the company has a 40 percent tax rate. Should you recom-
mend the project?
Pessimistic Expected Optimistic
Market size 130,000 150,000 165,000
Market share 21% 25% 28%
Selling price $      140 $      145 $      150
Variable costs per unit $      102 $       98 $       94
Fixed costs per year $1,015,000 $  950,000 $  900,000
Initial investment $2,200,000 $2,100,000 $2,000,000

9.21 Consider a project to supply Hamilton with 35,000 tonnes of machine screws annu-
ally for automobile production. You will need an initial $2,900,000 investment in
threading equipment to get the project started; the project will last for five years. The
accounting department estimates that annual fixed costs will be $495,000 and that
variable costs should be $285 per tonne; accounting will depreciate the initial fixed
asset investment at a CCA rate of 30 percent over the five-year project life. It also
estimates a salvage value of $300,000 after dismantling costs. The marketing depart-
ment estimates that the automakers will let you have the contract at a selling price
 www.downloadslide.com
284 Part 2    Value and Capital Budgeting

of $345 per tonne. The engineering department estimates you will need an initial net
working capital investment of $450,000. You require a 13 percent return and face a
marginal tax rate of 38 percent on this project. Assets will remain in the CCA class
after the end of the project.
a. What is the estimated operating cash flow for this project? The NPV? Should
you pursue this project?
b. Suppose you believe that the accounting department’s initial cost and salvage
value projections are accurate only to within ±15 percent, the marketing
department’s price estimate is accurate only to within ±10 percent, and the
engineering department’s net working capital estimate is accurate only to within
±5 percent. What is your worst-case scenario for this project? Your best-case
scenario? Do you still want to pursue the project?
Sensitivity Analysis
9.22 In Problem 9.21, suppose you’re confident about your own projections, but you’re a
little unsure about Hamilton’s actual machine screw requirements. What is the sen-
sitivity of the project operating cash flow to changes in the quantity supplied? What
about the sensitivity of NPV to changes in quantity supplied? Given the sensitivity
number you calculated, is there some minimum level of output below which you
wouldn’t want to operate? Why?
Abandonment Decisions
9.23 Consider the following project for Hand Clapper Inc. The company is considering a
four-year project to manufacture clap-command garage door openers. This project
requires an initial investment of $8 million that will be depreciated straight-line to
zero over the project’s life. An initial investment in net working capital of $950,000 is
required to support spare parts inventory; this cost is fully recoverable whenever the
project ends. The company believes it can generate $6.85 million in pre-tax revenues
with $2.8 million in total pre-tax operating costs. The tax rate is 38 percent, and the
discount rate is 16 percent. The market value of the equipment over the life of the
project is as follows:
Year Market value (in $ millions)
1 $5.1
2 3.8
3 3.2
4 0.0
a. Assuming Hand Clapper operates this project for four years, what is the NPV?
b. Now compute the project NPVs assuming the project is abandoned after only
one year, after two years, and after three years. What economic life for this proj-
ect maximizes its value to the firm? What does this problem tell you about not
considering abandonment possibilities when evaluating projects?
9.24 M.V.P. Games Inc. has hired you to perform a feasibility study of a new video game
that requires a $7 million initial investment. M.V.P. expects a total annual operating
cash flow of $1.3 million for the next 10 years. The relevant discount rate is 10 per-
cent. Cash flows occur at year-end.
a. What is the NPV of the new video game?
b. After one year, the estimate of remaining annual cash flows will be revised
either upward to $2.2 million or downward to $285,000. Each revision has an
equal probability of occurring. At that time, the video game project can be sold
for $2,600,000. What is the revised NPV given that the firm can abandon the
project after one year?
 www.downloadslide.com
Chapter 9    Risk Analysis, Real Options, and Capital Budgeting 285

MINICASE

Bunyan Lumber LLC

Bunyan Lumber LLC harvests timber and delivers logs The company expects to lose 5 percent of the timber
to timber mills for sale. The company was founded it cuts due to defects and breakage.
70  years ago by Pete Bunyan. The current CEO is The forest will be clear-cut when the company
Paula Bunyan, the granddaughter of the founder. The harvests the timber. This method of harvesting
company is currently evaluating a 5,000-hectare for- allows for faster growth of replanted trees. All of the
est it owns in central British Columbia. Paula has harvesting, processing, replanting, and transporta-
asked Steve Boles, the company’s finance officer, to tion are to be handled by subcontractors hired by
evaluate the project. Paula’s concern is when the com- Bunyan Lumber. The cost of the logging is expected
pany should harvest the timber. to be $140 per MBF. A road system has to be con-
Lumber is sold by the company for its “pond value.” structed and is expected to cost $50 per MBF on
Pond value is the amount a mill will pay for a log average. Sales preparation and administrative costs,
delivered to the mill location. The price paid for logs excluding office overhead costs, are expected to be
delivered to a mill is quoted in dollars per thousands $18 per MBF.
of board feet (MBF), and the price depends on the As soon as the harvesting is complete, the com-
grade of the logs. The forest Bunyan Lumber is evalu- pany will reforest the land. Reforesting costs
ating was planted by the company 20 years ago and is include the following:
made up entirely of Douglas fir trees. The table here
Cost per hectare
shows the current price per MBF for the three grades
of timber the company feels will come from the stand: Excavator piling $175
Broadcast burning 310
Timber grade Price per MBF
Site preparation 150
1P $620 Planting costs 225
2P 605
All costs are expected to increase at the inflation rate.
3P 595
Assume all cash flows occur at the year of har-
Steve believes that the pond value of lumber will vest. For example, if the company begins harvesting
increase at the inflation rate. The company is plan- the timber 20 years from today, the cash flow from
ning to thin the forest today, and it expects to realize the harvest will be received 20 years from today.
a positive cash flow of $1,000 per hectare from thin- When the company logs the land, it will immedi-
ning. The thinning is done to increase the growth ately replant the land with new saplings. The harvest
rate of the remaining trees, and it is always done period chosen will be repeated for the foreseeable
20 years after a planting. future. The company’s nominal required return is
The major decision the company faces is when to 10 percent, and the inflation rate is expected to be
log the forest. When the company logs the forest, it 3.7 percent per year. Bunyan Lumber has a 35 per-
will immediately replant saplings, which will allow cent tax rate.
for a future harvest. The longer the forest is allowed Clear-cutting is a controversial method of forest
to grow, the larger the harvest becomes per hectare. management. To obtain the necessary permits, Bun-
Additionally, an older forest has a higher grade of tim- yan Lumber has agreed to contribute to a conserva-
ber. Steve has compiled the following table with the tion fund every time it harvests the lumber. If the
expected harvest per hectare in thousands of board company harvested the forest today, the required
feet, along with the breakdown of the timber grades: contribution would be $250,000. The company has
Years from today Harvest (MBF) Timber grade agreed that the required contribution will grow by
to begin harvest per hectare 1P 2P 3P 3.2 percent per year. When should the company
20 14.1 16% 36% 48% harvest the forest?
25 16.4 20 40 40
30 17.3 22 43 35
35 18.1 24 45 31

For more information on the resources available from McGraw-Hill Ryerson,

go to www.mheducation.ca/he/solutions.