Chapter 7
Chapter 7
Chapter 7
In these spreadsheets, you will learn how to use the following Excel fu
Naming cells
Scenario Manager
One-way Data Table
Goal Seek
RAND
Frequency distribution
Frequency distribution charts
Shapes and lines
NOTE: For this workbook, any changes in input values will not change
alues will not change the output unless you hit the F9 key.
eadsheets:
equire that
Chapter 7 - Section 1
Sensitivity Analysis, Scenario Analysis, and Break-Even Analysis
Scenario Analysis
Scenario analysis is used to determine the range of possible outcomes for a project. Typically, the base-case, best-ca
when doing scenario analysis. Because of the repetitive nature of the calculations, spreadsheets are an excellent too
solar-powered jet engine project presented in the example in the textbook:
Base case
Unit sales 3,000
Price per unit $ 2,000,000
Variable costs per unit $ 1,000,000
Fixed costs per year $ 1,940,000,000
With these values, we need to calculate the base-case, best-case, and worst-case NPVs and IRRs. First, we want to c
OCF $ 900,400,000
NPV $ 1,518,280,450
Notice, in this case we calculated the NPV using the PV function rather than the NPV function. When the cash flows
To calculate the best case and worst case, we will use Scenario Manager, which is described below. The values for th
Scenario Manager is a powerful tool that allows you to evaluate different scenarios and is useful in cases such as thi
will be changing, in this case cells D9 through D12, and D14. Next, go to the Data tab, click What-If Analysis, Scenario
When you click on Add, another box comes up that will allow you to enter the scenario name. After entering the nam
Notice that the values are changed to the best case values. This is because the image was captured after we had cha
clicked Add, then added the worst-case scenario. When the values for both scenarios are entered, click OK. This brin
already added.
Now that all the scenarios are entered, we can click on Summary, which brings up the final box. This box allows us to
(NPV) as the final result we wanted Scenario Manager to calculate, then clicked OK. The results are shown on the ne
Sensitivity Analysis
In contrast to scenario analysis, sensitivity analysis holds all variables except one constant. This allows us to see how
we will perform sensitivity analysis using fixed costs, although all other variables could be similarly examined. Using
data table. Below, you will see a table with the NPV for different levels of fixed costs:
Graphically, the relationship between fixed costs and NPV looks like this:
$1
$-
0 0 0 0 0 0 0 0 0 0 0 0
,0 ,0 ,0 ,0 ,0 ,0
00 00 00 00 00 00 0 0
,0 ,0 ,0 ,0 ,0 ,0 ,0
91 91 91 91 91 91 91
,2 ,3 ,4 ,5 ,6 ,7 ,8
Net Present
$-
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
0 0 0 0 0 0 0
,0 ,0 ,0 ,0 ,0 ,0 ,0
91 91 91 91 91 91 91
,2 ,3 ,4 ,5 ,6 ,7 ,8
$1 $1 $1 $1 $1 $1 $1
Fixed Costs
As you can see, there is a negative relationship between fixed costs and project NPV. We would expect this: As costs
To set up a one-way data table, we need to first enter the inputs we want to use in the calculations in a column (or r
one cell above where the input values begin, we need to make the cell equal to the final value we want the data tab
this cell is C126. However, to make the data table look better, we have hidden this row. To unhide this row, select b
first step is to highlight the entire column with the numbers we want used in the calculation, as well as the final calc
"Data" tab, then "What-If Analysis," and "Data Table." Finally, enter the original cell that contains the variable we w
for fixed costs.
We should note that when you create a data table, you can change the input cells in which you entered the new val
table.
Of course, in our sensitivity analysis, we could be interested in how the NPV changes when two input variables chan
related since a higher cost would likely result in fewer units sold. In this case, we can use a two-way data table to co
data tables were introduced in Chapter 4.) The sensitivity analysis for price and units sold looks like this:
Price pe
$ 1,700,000 $ 1,800,000
2,750
2,800
2,850
2,900
Units Sold
2,950
3,000
3,050
3,100
3,150
Un
3,250
3,350
As you can see, when the price drops below $1,800,000, the project has a negative NPV for nearly all units sold exam
In the end, we are ultimately concerned with how sensitive the NPV is to changes in the inputs to the project. One w
the same percentage change in the inputs. Below, we have constructed one-way data tables for each of the inputs t
case values as inputs in these tables. The reason is that if we reference the original cells (D9 to D12), the calculation
% Change from
Base Case Units sales NPV
$ 1,518,280,450
-30% 2,100
-20% 2,400
-10% 2,700
0% 3,000
10% 3,300
20% 3,600
30% 3,900
To compare changes in each of the variables, we will graph the NPV for each of the sensitivity tables. Since the colum
hold down the CTRL and ALT keys, then use the cursor to select the four columns. The sensitivity of the NPV to perce
$-
-30% -20% -10% 0% 10% 20%
As you can see, the line with the steepest slope is the price per unit, followed closely by the variable cost per unit an
per unit, we should concentrate our efforts in determining whether or not our estimate for this variable is accurate.
Break-Even Analysis
In looking at break-even analysis, we can start with revenues, costs, and NPVs under different sales assumptions, wh
Investment $ 1,500,000,000
As with many other sets of data, a graph can help us better examine exactly what is happening. Given that, we will s
- 1,000 3,000
Annual revenues $ - $ 2,000,000,000 $ 6,000,000,000
Variable costs $ - $ 1,000,000,000 $ 3,000,000,000
Fixed costs
(incl. dep.) $ 2,240,000,000 $ 2,240,000,000 $ 2,240,000,000
Total costs $ 2,240,000,000 $ 3,240,000,000 $ 5,240,000,000
$18,000,000,000
$16,000,000,000
$14,000,000,000
$12,000,000,000
$10,000,000,000
$8,000,000,000
$6,000,000,000
$4,000,000,000
$2,000,000,000
$-
- 2,000 4,000 6,000
Although we could set up and equation to find the accounting break-even sales level, we will use Goal Seek instead.
The financial break-even is the point at which the NPV of the project is zero. Again, we used Goal Seek. Try it on you
2,427 engines (2,426.68 to be more precise).
ven Analysis
a project. Typically, the base-case, best-case (optimistic), and worst-case (pessimistic) values are calculated
lations, spreadsheets are an excellent tool for doing the analysis. Consider the Sonar Electronics Corporation
k:
st-case NPVs and IRRs. First, we want to calculate the NPV and IRR with the base-case projections, which are:
n the NPV function. When the cash flows are the same for each year, we find this calculation easier.
to name the cells. Click on the input cell for the units sold (D9), and look to the left of the formula bar in the
n the name bar to name the input in this cell. Whenever we want to use the input from this cell later, we can
le, if you look at the sales calculation, you will see that the formula we used in this cell is Unit_sales * Price.
In addition, Excel does not allow spaces in the variable name, so we used an underscore instead of the space
which is described below. The values for the pessimistic, expected or best, and worst cases are:
Optimistic
20,000
50%
10,000
$ 2,200,000
$ 800,000
$ 1,740,000,000
$ 1,000,000,000
cenarios and is useful in cases such as this. To use Scenario Manager, we first need to select the cells that we
e Data tab, click What-If Analysis, Scenario Manager. This will bring up a box that looks like this:
the scenario name. After entering the name, hit Add and another box will come up that looks like this:
the image was captured after we had changed the values. After we entered the values for the best-case, we
h scenarios are entered, click OK. This brings us to another box with the scenario names which we have
ings up the final box. This box allows us to save the results in a separate spreadsheet. We entered cell C32
cked OK. The results are shown on the next tab.
pt one constant. This allows us to see how changes in one variable affects the NPV of a project. In this case,
iables could be similarly examined. Using Excel, sensitivity analysis is most easily completed using a one-way
fixed costs:
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 ,0 ,0
0 00 0 00 0 00 0 00 0 00 0 00 0 00 0 00 0 00
9 1, 9 1, 9 1, 9 1, 9 1, 9 1, 9 1, 9 1, 9 1,
,4 ,5 ,6 ,7 ,8 ,9 ,0 ,1 ,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 ,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 0 0 0 0
1, 1, 1, 1, 1, 1, 1, 1, 1,
49 59 69 79 89 99 09 19 29
$ 1, $ 1, $ 1, $ 1, $ 1, $ 1, $ 2, $ 2, $ 2,
Fixed Costs
oject NPV. We would expect this: As costs increase, the value of the project should decrease.
to use in the calculations in a column (or row). Since we have used a column here, one cell to the right and
ual to the final value we want the data table to calculate, or in this case, the NPV. Notice that in our data table,
den this row. To unhide this row, select both rows 125 and 127, right-click, and then select "Unhide." This
in the calculation, as well as the final calculation cell at the top of the adjacent column. Next, select the
ginal cell that contains the variable we want to use to calculate the values in the data table, which is cell D12
ut cells in which you entered the new values to analyze, but you cannot change the size or layout of the data
V changes when two input variables change. Price and quantity sold are two variables that would seem to be
se, we can use a two-way data table to compute the NPV for changes in both of these variables. (Two-way
and units sold looks like this:
hanges in the inputs to the project. One way we can examine this is to determine how sensitive the NPV is to
e-way data tables for each of the inputs to our project that we believe will vary. Notice that we have the base
original cells (D9 to D12), the calculation of the ranges will create a loop.
% Change from
Base Case Price per unit NPV
$ 1,518,280,450
-30% $ 1,400,000
-20% $ 1,600,000
-10% $ 1,800,000
0% $ 2,000,000
10% $ 2,200,000
20% $ 2,400,000
30% $ 2,600,000
ch of the sensitivity tables. Since the columns we wish to graph are separated, to select the four NPV columns,
lumns. The sensitivity of the NPV to percentage changes in the base case values looks like this:
Unit sales
Unit sales
Price per unit
Variable cost per unit
Fixed costs
wed closely by the variable cost per unit and unit sales. Since the NPV is most sensitive to changes in the price
our estimate for this variable is accurate.
10,000
$ 20,000,000,000
10,000,000,000
1,940,000,000
300,000,000
$ 7,760,000,000
1,629,600,000
$ 6,130,400,000
$ 6,430,400,000
$ 20,055,698,142
y what is happening. Given that, we will set up the following table for our graph:
10,000
$ 20,000,000,000
$ 10,000,000,000
$ 2,240,000,000
$ 12,240,000,000
Annual revenues
Variable costs
Fixed costs
(incl. dep.)
Total costs
sales level, we will use Goal Seek instead. Remember, at the accounting break-even, net income is zero.
for, in this case, the net income cell. Go to the Data tab, click on What-If Analysis, Goal Seek. The Goal Seek
ng cell C251, the units per year. Once you have entered these values into Goal Seek, click OK and Excel will
general, Goal Seek and Solver can be used interchangeably. Goal Seek tends to be a little quicker and easier to
e your results. Notice, we could have used any of the income statements we set up for this problem.
o. Again, we used Goal Seek. Try it on your own and see if you don't agree that the financial break-even is
Scenario Summary
Current Values: Best case Worst case
Changing Cells:
Unit_sales 3,000 10,000 1,000
Price $ 2,000,000 $ 2,200,000 $ 1,900,000
Variable_cost_per_unit $ 1,000,000 $ 800,000 $ 1,200,000
Fixed_costs $ 1,940,000,000 $ 1,740,000,000 $ 2,000,000,000
$D$13
Investment $ 1,500,000,000 $ 1,000,000,000 $ 1,500,000,000
Result Cells:
NPV $ 2,312,741,208 $ 31,607,753,500 $ (1,302,055,241)
Notes: Current Values column represents values of changing cells at
time Scenario Summary Report was created. Changing cells for each
scenario are highlighted in gray.
Chapter 7 - Section 2
Monte Carlo Simulation
Excel will allow you to create a Monte Carlo simulation. In the following Monte Carlo simulation, we will use the ass
Barbeques, Inc. (BBI). The probabilities for the different inputs are:
We did not follow the textbook assumption of different market sizes and market shares for each succeeding year of
calculations become more cumbersome, and therefore more difficult to follow. Also, the example in the textbook is
investment, variable cost per unit, fixed costs, the project life, corporate tax rate, and required return. Below you w
assumed these inputs are constant, each input could also have its own probability distribution.
To create a number to assign for each probability, we will use the function RAND. The RAND function generates a ra
less than one.
For the compressed hydrogen grill project, we need three random numbers, one for the industry sales, one for BBI's
(+/- $3). While we could generate one random number and use it for all three inputs, this will result in only 30 differ
we should create three random numbers. Below you will find a set of random numbers and the resulting outputs fo
You will note that if you changed an input cell in this worksheet, the resulting values did not change. If you go to the
see that Manual is checked. The default setting for Excel is Automatic. We changed the setting in this workbook bec
RAND function, the random number is changed. As you will see below, we use the RAND function a lot in further cal
spreadsheet. If you want to refresh numbers in this worksheet, hit the F9 key, or change the setting on the Calculati
Once the random numbers are set up, we can calculate the NPV for the project for each set of random numbers. Be
While we could certainly have done more than 500 simulations, this is enough for our purposes here. To better exam
S134 to X169 and have a frequency distribution chart below.
Below, you can see the function arguments we used to create this frequency distribution.
Notice that beside the frequency distribution, we created another frequency distribution with ranges. We created th
we could graph a frequency distribution using the bins, the legend will not be as descriptive. We will use the ranges
below. If you look at the frequency distribution graph and hit F9, the graph will change with each new set of random
see if there is anything odd. (Hint: There is never an NPV in the ranges from -$10 million to $10 million, $30 million t
$130 million.)
80
70
60
Number of Observations
50
40
30
20
10
0
$-40 to $-30
$-30 to $-20
$-20 to $-10
$10 to $20
$20 to $30
$30 to $40
$40 to $50
$50 to $60
$70 to $80
$80 to $90
$110 to $120
$120 to $130
$150 to $160
$-50 to $-40
$-10 to $0
$0 to $10
$60 to $70
$90 to $100
$100 to $110
$130 to $140
$140 to $150
$160 to $170
NPV (in millions)
To create this frequency distribution, we selected the data we wanted to graph (X134:X169) and went to the Insert
the data for the horizontal axis and input the legends as normal. Generally, when Excel draws a frequency distributi
between the columns. You can change this width by right-clicking on a column and selecting Format Data Series. In t
will allow you to change the gap between the columns.
To change the color of the bars, click on any bar, go to the Format tab, select Shape Fill, and change the color of the
done with the bars representing negative NPVs, click twice on the bar you want to have a different the color. After y
select the Format tab, select Shape Fill, and change the color. Double-clicking on the same bar will allow you to chan
histogram.
So how do we interpret the Monte Carlo simulation? One way is to do a cumulative probability graph that shows the
have a positive NPV or a negative NPV. In cells Y134:Y169 we have calculated this probability. The cumulative densit
calculations (F9), the probability of a positive NPV for this project ranges from about 85 percent to 95 percent.
90%
80%
70%
Probability of Equaling or Exceeding NPV
60%
50%
40%
30%
20%
10%
0%
40 30 $0 $ 1 0 $ 2 0 $ 3 0 $ 4 0 $ 5 0 $ 6 0 $ 7 0 $ 8 0 $ 9 0 1 0 0 1 10 1 20 1 30 1 4 0 1 5 0 1 60 1
20 10
$- $- $- $- to $ $ $ $ $ $ $ $
to to to to 10 0 to 0 to 0 to 0 to 0 to 0 to 0 to 0 to 0 to to to to to to to to to
5 0 4 0 3 0 2 0 $ - $ 1 2 3 4 5 6 7 8 0 0 0 0 0 0 0 0
$- $- $- $- $ $ $ $ $ $ $ $ $9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1
$ $ $ $ $ $ $ $
25% 10% 5%
4% 5% 8%
ares for each succeeding year of the project. While we certainly could do this, the resulting
, the example in the textbook is missing several major assumptions including the initial
d required return. Below you will find the assumptions we made for these inputs. While we
stribution.
he RAND function generates a random number between greater than or equal to zero and
unction is RAND(). This will return a random number greater than or equal to zero and less
the industry sales, one for BBI's market share, and one for the randomness in the sales price
s, this will result in only 30 different NPVs. Since all of the random variables are independent,
ers and the resulting outputs for these three variables.
did not change. If you go to the Formulas tab and select the Calculations Options, you will
the setting in this workbook because every time a change is made in a worksheet with the
AND function a lot in further calculations. The resulting recalculation slows down the
ange the setting on the Calculation Options back to Automatic.
each set of random numbers. Below, in rows 134 to 633, we have done just this 500 times.
ur purposes here. To better examine the output, we have created a frequency table in cells
use the FREQUENCY function is somewhat complicated, we will walk through the process
ution.
ution with ranges. We created the ranges by concatenating the bins we created earlier. While
criptive. We will use the ranges for graphing the frequency distribution, which you will see
ge with each new set of random numbers. Do this a couple of times and watch the graph to
lion to $10 million, $30 million to $50 million, $80 million to $90 million, and $120 million to
$120 to $130
$150 to $160
$190 to $200
$200 to $210
$220 to $230
$230 to $240
$290 to $300
$90 to $100
$100 to $110
$130 to $140
$140 to $150
$160 to $170
$170 to $180
$180 to $190
$210 to $220
$240 to $250
$250 to $260
$260 to $270
$270 to $280
$280 to $290
34:X169) and went to the Insert tab, Column chart, 2-D, Clustered Column. We then selected
cel draws a frequency distribution as we have done here, there is a large amount of space
electing Format Data Series. In the box this brings up, there is a Series Option selection that
Fill, and change the color of the bar. To highlight a specific bar, or certain bars as we have
ave a different the color. After you click on the bar twice, use the same procedure, that is,
e same bar will allow you to change a specific bar without changing the other bars in the
probability graph that shows the probability that the compressed hydrogen grill project will
obability. The cumulative density function for the project is shown below. If you refresh the
85 percent to 95 percent.
0 0 1 10 1 20 1 30 1 4 0 1 5 0 1 60 1 70 1 80 1 9 0 2 00 2 10 2 20 2 3 0 2 4 0 2 50 2 6 0 2 70 2 8 0 2 90 3 00
$1 $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $
to to to to to to to to to to to to to to to to to to to to
0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0
$1 $1 $1 $ 1 $ 1 $1 $1 $1 $ 1 $ 1 $2 $2 $ 2 $ 2 $2 $ 2 $2 $ 2 $2 $2
0.00%
0.00%
8.60%
9.80%
9.80%
85.60%
81.80%
66.40%
64.80%
64.80%
64.80%
61.40%
48.40%
34.20%
34.20%
34.20%
32.20%
23.80%
14.80%
11.00%
10.00%
10.00%
5.60%
2.00%
0.60%
0.60%
0.60%
0.60%
0.60%
0.60%
0.60%
0.60%
0.60%
0.60%
0.60%
Chapter 7 - Section 4
Decision Trees
We can use Excel to draw decision trees and calculate the value of a project. Consider the decision tree for Solar Ele
Test results
revealed
Test
No Test
NPV = $1,518
Invest
Success
75%
Do not invest
NPV = $0
Test results
revealed
NPV = ($3,611)
Invest
Failure
NPV = $0
o "Insert" and select "Text Box." Text boxes allow you to enter text that is "linked" to the
xt boxes in this case allow the text to go across multiple rows and/or columns, which can
h we will show later. To copy and/or move a text box, click on the box. This will bring up a
sor to where you would like to copy the Text Box, right click again, and select Paste. To draw
e are numerous shapes and lines available on this menu.
Chapter 7 - Master It!
Dahlia Simmons, CFO of Ulrich Enterprises, is analyzing a new project to sell solar powered batteries for cell phones
for the variables in the project:
The unit price depends on the industry demand since a greater demand will result in a higher price. Dahlia determin
The random "+/-$2" term represents an increase or decrease in price according to the following distribution:
The length of the project, tax rate, and required return are:
a.
Create a Monte Carlo simulation for the project with at least 500 iterations of the calculation. Calculate the IRR
an error if the IRR of the project is too low. For example, what is the IRR if both the initial cash flow and the op
percent. This is not a problem when you are calculating the IRR one time since you can see the IRR is too low, b
problem trying to summarize the results. Because of this, you should create an IF statement that tests if the op
investment is less than .1. If this is the case, the cell will return an IRR of -99.99 percent, else the cell will calcul
b. Create a graph of the distribution of the IRRs from the Monte Carlo simulation for different ranges of IRR.
c. Create a graph for the cumulative probability function for the IRR distribution.
wered batteries for cell phones. Dahlia has estimated the following probability distributions
20% 10%
5% 6%
a higher price. Dahlia determines that the price per unit will be given by the equation:
e following distribution:
e calculation. Calculate the IRR for each iteration. NOTE: The IRR function in Excel will return
he initial cash flow and the operating cash flows are negative? The IRR is less than -100
ou can see the IRR is too low, but when you are running 500 or more iterations it can create a
F statement that tests if the operating cash flow divided by the absolute value of the initial
ercent, else the cell will calculate the IRR.
or different ranges of IRR.
Master It! Solution
Probability:
Industry demand:
Probability:
Ulrich market share:
Probability:
Initial cost:
Probability:
VC per unit
Probability:
Fixed costs:
Probability:
Price randomness: