Sensitivity Analysis Using SensIt

Add-In for Microsoft Excel

Contents

How to Install SensIt..............................................2

How to Uninstall or Delete SensIt........................2

Overview.................................................................2

Example Problem...................................................3

Plot...........................................................................4

Spider......................................................................6

Tornado...................................................................9

Tornado Tips ........................................................11

Sensit is a sensitivity analysis add-in for Microsoft Excel 5, 95, 97, 98, 2000, and 2001 for
Windows and Macintosh. It was written by Mike Middleton of the University of San Francisco
and Jim Smith of Duke University.

How to Install SensIt

Here are three ways to install SensIt:

(1) Start Excel, and use Excel’s File | Open command to open the SensIt.xla file from floppy or
hard drive.

(2) Copy the SensIt.xla file to the Excel | Library subdirectory of your hard drive. Start Excel, and
use Excel’s Tools | Add-Ins command to load and unload SensIt as needed.

(3) Copy the SensIt.xla file to the Excel | Startup subdirectory of your hard drive, in which case
the file will be opened every time you start Excel.

All of SensIt’s functionality, including its built-in help, is a part of the SensIt.xla file. There is no
separate setup file or help file.

How to Uninstall or Delete SensIt

(A) First, use your file manager to locate SensIt.xla, and delete the file from your hard drive.

(B1) If SensIt is listed under Excel's add-in manager and the box is checked, when you start Excel
you'll see "Cannot find ..." Click OK. Choose Tools | Add-Ins, uncheck the box for SensIt; you'll
see "Cannot find ... Delete from list?" Click Yes.

(B2) If SensIt is listed under Excel's add-in manager and the box is not checked, start Excel and
choose Tools | Add-Ins. Check the box for SensIt; you'll see "Cannot find ... Delete from list?"
Click Yes.

Overview

To run SensIt, start Excel and open the SensIt.xla file. Alternatively, install SensIt using one of
the methods described above. SensIt adds a Sensitivity Analysis command to the Tools menu.
The Sensitivity Analysis command has four subcommands: Plot, Spider, Tornado, and Help.

Before using the SensIt options, you must have a spreadsheet model with one or more inputs and
an output. All three SensIt options make it easy for you to see how sensitive the output is to
changes in the inputs.

Use SensIt’s Plot option to see how your model’s output depends on changes in a single input
variable.

2
Use SensIt’s Spider option to see how your model’s output depends on the same percentage
changes for each of the model’s input variables.

Use SensIt’s Tornado option to see how your model’s output depends on ranges you specify for
each of the model’s input variables.

Example Problem
Model Display
A B C
1 Spreadsheet Model For Eagle Airlines
2
3 Input Variables Input Cells
4 Charter Price/Hour $325
5 Ticket Price/Hour $100
6 Hours Flown 800
7 Capacity of Scheduled Flights 50%
8 Proportion of Chartered Flights 0.5
9 Operating Cost/Hour $245
10 Insurance $20,000
11
12 Intermediate Calculations
13 Total Revenue $230,000
14 Total Cost $216,000
15
16 Performance Measure
17 Annual Profit $14,000
18
19 Adapted from Bob Clemen's textbook,
20 Making Hard Decisions, 2nd ed., Duxbury (1996).

Model Formulas
A B
11
12 Intermediate Calculations
13 Total Revenue =(B8*B6*B4)+((1-B8)*B6*B5*B7*5)
14 Total Cost =(B6*B9)+B10
15
16 Performance Measure
17 Annual Profit =B13-B14
18

3
Plot

Use SensIt’s Plot option to see how your model’s output depends on changes in a single input
variable.

PLOT INPUT VARIABLE

Plot Input Variable’s Cells: Option: In the Label edit box, type a cell reference, or point to the cell
containing a text label and click. Required: In the Cell edit box, type a cell reference, or point to
the cell containing a numeric value that’s an input to your model.

PLOT OUTPUT VARIABLE

Plot Output Variable’s Cells: Option: In the Label edit box, type a cell reference, or point to the
cell containing a text label and click. Required: In the Cell edit box, type a cell reference, or point
to the cell containing a formula that’s the output of your model.

PLOT INPUT VALUES

Plot Input Values: Type numbers in the Start, Step, and Stop edit boxes to specify values to be
used in the input variable’s cell. Cell references are not allowed.

Send Output To: Select the destination for the output table and chart. If you send output to This
Worksheet, enter a Cell reference for the top left corner of the output. Output options are not
available on the Macintosh; output is always sent to a new worksheet.

Click OK: SensIt Plot uses the Start, Step, and Stop values to prepare a table of values. Each
value is copied to the input variable cell, the worksheet is recalculated, and the value of the output
variable cell is copied to the table. (You could do this manually using the Edit | Fill | Series and
Data | Table commands.) SensIt Plot uses the input and output values to prepare an XY (Scatter)
chart; optionally, the text in the label cells you identified are used as the chart’s axis labels. (You
could do this manually using the ChartWizard.)

4
SensIt Plot Dialog Box

SensIt Plot Numerical and Chart Output

Sensitivity Analysis ...

Input Output
Capacity of Scheduled Flights Annual Profit SensIt - Sensitivity Analysis - Plot
40% -$6,000
42% -$2,000
44% $2,000 $30,000
Annual Profit

46% $6,000
$20,000
48% $10,000
50% $14,000 $10,000
52% $18,000
$0
54% $22,000
56% $26,000 -$10,000
58% $30,000 40% 45% 50% 55% 60%
60% $34,000
Capacity of Scheduled Flights

5
Spider

Use SensIt’s Spider option to see how your model’s output depends on the same percentage
changes for each of the model’s input variables. Before using Spider, arrange your model input
cells in adjacent cells in a single column, arrange corresponding labels in adjacent cells in a single
column, and be sure your model’s input cells contain base case values.

For example, if your model has five inputs, the names of the five inputs could be text in A1:A5.
The input cells of your model could be numbers in B1:B5; when you change a number in one of
these cells, the output of your model changes; enter base case values in the input cells B1:B5
before using Spider.

SPIDER INPUT VARIABLES

Spider Input Variables’ Ranges: Labels edit box: Type a range reference, or point to the range
(click and drag) containing text labels. Cells edit box: Type a range reference, or point to the
range containing numeric values that are inputs to your model. Each range must be adjacent cells
in a single column.

SPIDER OUTPUT VARIABLE

Spider Output Variable’s Cells: Label edit box: Type a cell reference, or point to the cell
containing a text label and click. Cell edit box: Type a cell reference, or point to the cell
containing a formula that’s the output of your model.

SPIDER INPUT CHANGES

Spider Input Changes (%): Type numbers in the Start (%), Step (%), and Stop (%) edit boxes to
define the percents that will be multiplied times the current value in each input variable’s cell.
Cell references are not allowed.

Send Output To: Select the destination for the output table and chart. If you send output to This
Worksheet, enter a Cell reference for the top left corner of the output. Output options are not
available on the Macintosh; output is always sent to a new worksheet.

Click OK: SensIt Spider uses the Start (%), Step (%), and Stop (%) values and the original (base
case) numeric value in each input variable cell to prepare a table of percentage change input
values. For each input variable, all other input values are set at their base case values, each
percentage change input value is copied to the input variable cell, the worksheet is recalculated,
and the value of the output variable cell is copied to the table. The output variable values are also
expressed as percentage change of the base case output value. SensIt Spider prepares two XY
(Scatter) charts; the horizontal axis is percentage change of input variables; the vertical axis is
model output value on one chart and percentage change of model output value on the other; the
input variables’ labels are used for chart legends.

6
SensIt Spider Dialog Box

SensIt Spider Numerical Output

A B C D E F
1 Input Variables Values
2 80% 90% 100% 110% 120%
3 Charter Price/Hour $260 $293 $325 $358 $390
4 Ticket Price/Hour $80 $90 $100 $110 $120
5 Hours Flown 640 720 800 880 960
6 Capacity of Scheduled Flights 40% 45% 50% 55% 60%
7 Proportion of Chartered Flights 0.4 0.45 0.5 0.55 0.6
8 Operating Cost/Hour $196 $221 $245 $270 $294
9 Insurance $16,000 $18,000 $20,000 $22,000 $24,000
10
11
12 Output Variable Values (Annual Profit)
13 80% 90% 100% 110% 120%
14 Charter Price/Hour -$12,000 $1,000 $14,000 $27,000 $40,000
15 Ticket Price/Hour -$6,000 $4,000 $14,000 $24,000 $34,000
16 Hours Flown $7,200 $10,600 $14,000 $17,400 $20,800
17 Scheduled Capacity -$6,000 $4,000 $14,000 $24,000 $34,000
18 Chartered Proportion $8,000 $11,000 $14,000 $17,000 $20,000
19 Operating Cost/Hour $53,200 $33,600 $14,000 -$5,600 -$25,200
20 Insurance $18,000 $16,000 $14,000 $12,000 $10,000

7
SensIt Spider Chart Output
Charter Price/Hour
Sensit - Sensitivity Analysis - Spider
Ticket Price/Hour
Annual Profit Value

$60,000
$50,000
$40,000 Hours Flown
$30,000
$20,000
$10,000
$0 Capacity of
-$10,000 Scheduled Flights
-$20,000 76808488929610101011111212
-$30,000 % % % % % % 0 4 8 2 6 0 4 Proportion of
%% % % % % % Chartered Flights
% Change in Input Value
Operating
Cost/Hour

SensIt Spider Chart Output After Formatting

Sensit - Sensitivity Analysis - Spider

$60,000

$50,000

$40,000
Charter Price/Hour
Annual Profit Value

$30,000 Ticket Price/Hour

Hours Flown
$20,000
Scheduled Capacity
$10,000
Chartered Proportion
$0 Operating Cost/Hour
75% 80% 85% 90% 95% 100% 105% 110% 115% 120% 125% Insurance
-$10,000

-$20,000

-$30,000
% Change in Input Value

8
Tornado

Use SensIt’s Tornado option to see how your model’s output depends on ranges you specify for
each of the model’s input variables. Before using Tornado, arrange your model input cells in
adjacent cells in a single column, arrange corresponding labels in adjacent cells in a single
column, and arrange Low, Base, and High input values for each input variable in three separate
columns. Alternatively, the three columns containing input values can be worst case, likely case,
and best case.

For example, if your model has five inputs, the names of the five inputs could be text in A1:A5.
The input cells of your model could be numbers in B1:B5; when you change a number in one of
these cells, the output of your model changes. The Low input values could be numbers in D1:D5,
chosen as the min imum possible value you think each input variable could be. The Base input
values could be numbers in E1:E5, chosen as the most likely value for each input; you might also
have these same numbers in B1:B5 as current inputs to your model. The High input values could
be numbers in F1:F5, chosen as the maximum possible value you think each input variable could
be.

TORNADO INPUT VARIABLES

Tornado Input Variables’ Ranges: Labels edit box: Type a range reference, or point to the range
(click and drag) containing text labels. Cells edit box: Type a range reference, or point to the
range containing numeric values that are inputs to your model. Each range must be adjacent cells
in a single column.

TORNADO OUTPUT VARIABLE

Tornado Output Variable’s Cells: Label edit box: Type a cell reference, or point to the cell
containing a text label and click. Cell edit box: Type a cell reference, or point to the cell
containing a formula that’s the output of your model.

TORNADO INPUT VALUES

Tornado Input Values’ Ranges: In the Low, Base, and High edit boxes, type a range reference, or
point to the range (click and drag) containing numeric values for each of your model’s inputs.

Send Output To: Select the destination for the output table and chart. If you send output to This
Worksheet, enter a Cell reference for the top left corner of the output. Output options are not
available on the Macintosh; output is always sent to a new worksheet.

Click OK: For each input variable, SensIt Tornado sets all other input values at their Base case
values, copies the Low input value to the input variable cell, recalculates the worksheet, and
copies the value of the output variable cell to the table; the steps are repeated using each High
input value. For each input variable, SensIt Tornado computes the range of the output variable
values, sorts the table from largest range down to smallest range, and prepares a bar chart.

9
Example with Lower and Upper Bounds
A B C D E F
1 Spreadsheet Model For Eagle Airlines
2
3 Input Variables Input Cells Lower Bound Base Value Upper Bound
4 Charter Price/Hour $325 $300 $325 $350
5 Ticket Price/Hour $100 $95 $100 $108
6 Hours Flown 800 500 800 1000
7 Capacity of Scheduled Flights 50% 40% 50% 60%
8 Proportion of Chartered Flights 0.5 0.45 0.5 0.7
9 Operating Cost/Hour $245 $230 $245 $260
10 Insurance $20,000 $18,000 $20,000 $25,000
11
12 Intermediate Calculations
13 Total Revenue $230,000
14 Total Cost $216,000
15
16 Performance Measure
17 Annual Profit $14,000
18
19 Adapted from Bob Clemen's textbook,
20 Making Hard Decisions, 2nd ed., Duxbury (1996).

SensIt Tornado Dialog Box

10
SensIt Tornado Numerical and Chart Output
A B C D E F G H I J K
1 Tornado Analysis ...
2
3
4 Input Values Output Values (Annual Profit) Percent
5 Input Variable Low Base High Low Base High Swing Variance
6 Capacity of Scheduled Flights 40% 50% 60% -$6,000 $14,000 $34,000 $40,000 46.1%
7 Operating Cost/Hour $230 $245 $260 $26,000 $14,000 $2,000 $24,000 16.6%
8 Hours Flown 500 800 1000 $1,250 $14,000 $22,500 $21,250 13.0%
9 Charter Price/Hour $300 $325 $350 $4,000 $14,000 $24,000 $20,000 11.5%
10 Proportion of Chartered Flights 0.45 0.5 0.7 $11,000 $14,000 $26,000 $15,000 6.5%
11 Ticket Price/Hour $95 $100 $108 $9,000 $14,000 $22,000 $13,000 4.9%
12 Insurance $18,000 $20,000 $25,000 $16,000 $14,000 $9,000 $7,000 1.4%
13
14
15
16 SensIt - Sensitivity Analysis - Tornado
17
18
Capacity of Scheduled Flights 40% 60%
19
20 Operating Cost/Hour $260 $230
21
22 Hours Flown 500 1000
23 Charter Price/Hour $300 $350
24
25 Proportion of Chartered Flights 0.45 0.7
26
27 Ticket Price/Hour $95 $108
28 Insurance $25,000 $18,000
29
30 -$15,000 -$5,000 $5,000 $15,000 $25,000 $35,000
31
32 Annual Profit
33

Tornado Tips

When defining the high and low cases for each variable, it is important to be consistent so that the
"high" cases are all equally high and the "low" cases are equally low. For example, you might
take all of the base case values to be estimates of the mean of the input variable, take low cases to
be values such there is a 1-in-10 chance of the variable being below this amount, and take the
high cases to be values such that there is a 1-in-10 chance of the variable being above this
amount. Alternatively, you may specify low and high values that are the absolute lowest and
highest possible values.

When you click OK, SensIt sets all of the input variables to their base-case values and records the
output value. Then SensIt goes through each of the input variables one at a time, plugs the low-
case value into the input cell, and records the value in the output cell. It then repeats the process
for the high case. For each substitution, all input values are kept at their base-case values except
for the single input value that is setn at it low or high value. SensIt then produces a spreadsheet
that lists the numerical results as shown in columns F, G, and H above.

In the worksheet, the variables are sorted by their "swing" -- the absolute value of the difference
between the output values in the low and high cases. "Swing" serves as a rough measure of the
impact of each input variable. The rows of numerical output are sorted from highest swing at the
top down to lowest swing at the bottom. Then SensIt creates a bar chart of the sorted data.

11
"Percent variance" is a standardized measure of impact: it squares each swing, sums them up to
get a "Total Variance", and reports the percentage of the "total variance" attributed to each input
variable.

In general, you should focus your modeling efforts on those variables with the greatest impact on
the value measure.

If your model has input variables that are discrete or categorical, you should create multiple
tornado charts using different base case values of that input variable. For example, if your model
has an input variable "Government Regulation" that has possible values 0 (zero) or 1, the low and
high values will be 0 and 1, but you should run one tornado chart with base case = 0 and another
tornado chart with base case = 1.

Copyright  2001 by Michael R. Middleton Michael R. Middleton, Ph.D.

Professor of Decision Sciences
School of Business and Management
Decision Support Services University of San Francisco
2105 Buchanan Street, #1 2130 Fulton Street
San Francisco, CA 94115-2339 San Francisco, CA 94117-1045
Email: [email protected] Email: [email protected]
Web Site: http://www.treeplan.com Web Site: http://www.usfca.edu/~middleton

12