Business

Simulation
through Excel
Topics Covered
Monte Carlo Simulation
 Simulation
 Definition from Acting out or mimicking an
actual or probable real life condition, event, or
situation to find a cause of a past occurrence
(such as an accident), or to forecast future
effects (outcomes) of assumed circumstances or
factors.
 A simulation may be performed through
 (1) solving a set of equations (a mathematical
model),
 (2) constructing a physical (scale) model,
 (3) staged rehearsal,
 (4) game (such as wargames), or a computer
graphics model (such as an animated flowchart).
 Whereas simulations are very useful tools that
allow experimentation without exposure to risk,
they are gross simplifications of the reality
because they include only a few of the real-
world factors, and are only as good as their
underlying assumptions.

http://www.businessdictionary.com/definition/simulation.html
 A simulation is the execution of a model,
represented by a computer program that
gives information about the system being
investigated.
 The simulation approach of analyzing a
model is opposed to the analytical approach,
where the method of analyzing the system is
purely theoretical.
 As this approach is more reliable, the
simulation approach gives more flexibility
and convenience.
Why to use models?
 To fly a simulator is safer and cheaper than
the real airplane.
 For precisely this reason, models are used in
industry commerce and military: it is very
costly, dangerous and often impossible to
make experiments with real systems.
Provided that models are adequate
descriptions of reality (they are valid),
experimenting with them can save money,
suffering and even time.
When to use
simulations?
 Systems that change with time, such as a gas
station where cars come and go (called
dynamic systems) and involve randomness.
 Nobody can guess at exactly which time the
next car should arrive at the station, are
good candidates for simulation.
 Modeling complex dynamic systems
theoretically need too many simplifications
and the emerging models may not be
therefore valid.
 Simulation does not require that many
simplifying assumptions, making it the only
tool even in absence of randomness.
How to simulate?
 Suppose we are interested in a gas station.
We may describe the behavior of this system
graphically by plotting the number of cars in
the station; the state of the system.
 Every time a car arrives the graph increases
by one unit while a departing car causes the
graph to drop one unit. This graph (called
sample path), could be obtained from
observation of a real station, but could also
be artificially constructed.
 Such artificial construction and the analysis
of the resulting sample path (or more sample
paths in more complex cases) consists of the
simulation.
Types of simulations:
 Discrete event. The above sample path
consisted of only horizontal and vertical
lines, as car arrivals and departures occurred
at distinct points of time, what we refer to as
events. Between two consecutive events,
nothing happens - the graph is horizontal.
When the number of events are finite, we call
the simulation "discrete event."
 In some systems the state changes all the
time, not just at the time of some discrete
events. For example, the water level in a
reservoir with given in and outflows may
change all the time. In such cases
"continuous simulation" is more appropriate,
although discrete event simulation can serve
as an approximation.
 Further consideration of discrete event
simulations.
Steps to Simulation

 The steps are:

1. Formulate the problem.

2. Collect Data.

3. Analyze Data.

4. Formulate a model.

5. Write and test the programme.

6. Validate the model.

7. Analyze the results.

 Car Wash
The Number of cars arriving per hour at Pratap's Car
Wash during the past 200 hours of operation is
observed to be the following
                 

No. of Cars
Arriving Frequency   OR Frequency        

3 or Fewer 0     0        

4 20     10        

5 30     15        

6 50     25        

7 50     25        

8 40     20        

9 or more 10     5        

Total 200     100        

                 

(a) set up a probability and cumulative probability

distribution for the variable of Car arrivals
(b) Simulate 15 hours of car arrival and compute the
average number of car arrival.
 Mr. Magic Car Wash
Simulate 15 hours of car
arrival and compute the
average number of car
arrival.

Monte Carlo simulation Vlookup

Random Numbers

Probability
Distribution

Range Naming Techniques

The Number of cars arriving per hour at Mr. Magic Car

Wash during the past 200 hours of operation is
observed
 No. of Cars Freque
Arriving ncy
3 or Fewer 0
4 20
5 30
6 50
7 50
8 40
9 or more 10
Simulate 15 hours of car arrival and
compute Total 200of car arrival.
the average number
• Probability Distribution

No. of cars
arrived Frequency Probability
3 0 0
4 20 0.1
5 30 0.15
6 50 0.25
7 50 0.25
8 40 0.2
9 10 0.05
Total 200 1
Probability
class
 

Lower class Upper class # cars

limit limit arrived Freq Probability
0 0 3 0 0
0 0.09 4 20 0.1
0.1 0.24 5 30 0.15
0.25 0.49 6 50 0.25
0.5 0.74 7 50 0.25
0.75 0.94 8 40 0.2
0.95 0.99 9 10 0.05
Hours Random No. No. of cars arriving
1 Total
0.52 200 1
2 0.6
3 0.5
4 0.88
5 0.53
6 0.3
7 0.1
8 0.47
9 0.99
10 0.37
11 0.66
12 0.91
13 0.35
14 0.32
15 0
 Probability
class 
Lower class Upper # cars
limit class limit arrived Freq Probability
0 0 3 0 0
Note the
0.01 0.1 4 20 0.1
changes in 0.11 0.25 5 30 0.15
specifying
class limits
0.26 0.5 6 50 0.25
0.51 0.75 7 50 0.25
0.76 0.95 8 40 0.2
0.96 0.99 9 10 0.05
Total 200 1

Hours Random No. # cars arriving

1 0.52
2 0.6
3 0.5 Vlookup
4 0.88
5 0.53 VLOOKUP(lookup_
6 0.3 value, table_array,
7 0.1 col_index_num,
[range_lookup])
8 0.47
9 0.99
10 0.37
11 0.66
12 0.91
13 0.35
14 0.32
15 0
Vlookup
 VLOOKUP(lookup_value, table_array,
col_index_num, [range_lookup])
 range_lookup

Data Must be Sorted

TRUE
An exact or approximate match is
returned.
If an exact match is not found, the
next largest value that is less than
lookup_value is returned.
False

An exact match is searched.

If there are two or more values in the
first column of table_array then, the
first value found is used.
If an exact match is not found, the
error value #N/A is returned.
Simulation:
Staffing problem
 Bargain store is reviewing staffing levels in
one of its dept. The store operates from 9:00
to 18:00. Management is concerned that
customers may be leaving without purchase
because they are not served. They wonder if
this situation is due to the fact that there are
only two assistants in the department.
Management has record of customer arrival
and purchase data of the day’s 9 hrs divided
in fifty-four 10 min intervals.
 Bargain store would like to know how many
sales staff should be assigned to dept in
order to maximize profits, if it cost $5 per
hour per sales person.
No. of 0 1 2 3 4 5 6
Customers
No. of 5 6 8 10 12 9 4
intervals
Value of Zero 10 20 30 40 50
Purchases ($)
No. of purchases 14 22 30 53 33 17
Simulation: Monte Carlo Method
Monte Carlo method is technique based on selection of random
number ranging from 00 to 99. the series of 100 numbers can be
assigned to represent a probability.

 Uncertainty:
◦ No pattern of customer arrival : Queuing
models can not be used
◦ No pattern of purchasing habits.
 From the 0data1we can
No. of 2
find
3
probability
4 5
: 6
Customers

No. of 5 6 8 10 12 9 4
intervals

Probability 0.09 0.11 0.15 0.19 0.22 0.17 0.07

Value of Zero 10 20 30 40 50
Purchases ($)

No. of purchases 14 22 30 53 33 17

Probability 0.08 0.13 0.18 0.31 0.20 0.10

Simulation: Monte Carlo Method

 Useful in analyzing the problems containing elements of

uncertainty.
 Simulation model uses random number to imitate
uncertainty.

 Task: Monte Carlo method is technique based on

selection of random number ranging from 00 to 99. the
series of 100 numbers can be assigned to represent a
probability.

 How random number shall be used for data input related to

◦ How many customer arrived during a particular 10 minute
interval?
◦ How much purchase he did?
 Decide on the basis of cost-benefit analysis
◦ Number of sales staff
◦ Actual sales
◦ Lost sales
◦ Idle time (hrs) = No. of Idle 10 minutes intervals x
corresponding number of assistants)/6
Purchase totals = 400 480 320 220 150 50
Idle Intervals = 3 8 15 23 36 51

No. of Sales staff = 2 3 4 5 6

Actual Sales = 880 1200 1420 1570 1620
Lost sales = 740 420 200 50 50
Idle time (hrs) 1 4 10 19.16667 36
Simulation Model Formulae used:
F9: Recalculate
Cell Formula Copied to
C6 F6

B7 C6 B8:B12
C7 C6+F7 C8:C12
F6 E6/E$13 F7:F12
I6 L6

H7 I6 H8:H11
I7 I6+L7 I8:I11
L6 K6/K$12 L7:L11
D13 SUMPRODUCT(D6:D12,E6:E12)

E13 SUM(E6:E12)

K12 SUM(K6:K11)

C17 RAND() C18:C34

D17 VLOOKUP(C17,$B$6:$D$12,3) D18:D34
F17 IF(F$16>$D17,””,VLOOKUP(RAND F17:K34
(),$H$6:$j$11,3))

F35 SUM(F17:F34) G35:K35

F36 COUNTBLANK($F17:F34) G36:K36
Risk Analysis: Simulating Risky
investment
 We can use simulation to approximate the
NPV, or Expected Return and its dispersion
about the expected value.
 Simulation model proposed by Hertz1
considers the following factors in analyzing
risky investment projects

Market Investment Operating

Analysis Cost Analysis /Fixed Cost

Investment Operating
Market Size
Required Cost

Residual value
Selling Price Fixed Cost
of Investment

Market Useful life of

Growth Rate facilities.

David B Hertz. “Risk Analysis in Capital

Investment” Harvard Business Review,
42 (Jan-Feb 1964) pp 95-106
 Start
Develop Probability Distributions
of Key Factors

Market Size Selling Price Fixed Cost

Mkt Growth Residual Value

Investment
Rate of Investment

Useful life of
Mkt Share Operating Cost
facility

In the Simulation table: create sufficient values of these

variables using probability distributions identified
above. Calculate NPV using these random values

Analyze the NPV values and obtain its probability

distribution

Make decisions on the basis of this probability

distribution
What Is Central Limit
Theorem?
 if a random sample of size n (say, n larger
than 30) from an infinite population, finite
standard deviation , then the standardized
sample mean converges to a standard
normal distribution or, equivalently, the
sample mean approaches a normal
distribution with mean equal to the
population mean and standard deviation 
What Is a Least Squares
Model?
 Many problems in analyzing data involve
describing how variables are related. The
simplest of all models describing the
relationship between two variables is a
linear, or straight-line, model. The simplest
method of fitting a linear model is to "eye-
ball'' a line through the data on a plot. A
more elegant, and conventional method is
that of "least squares", which finds the line
minimizing the sum of distances between
observed points and the fitted line.
ANOVA: Analysis of
Variance
 The tests we have learned up to this point
allow us to test hypotheses that examine the
difference between only two means. Analysis
of Variance or ANOVA will allow us to test
the difference between 2 or more means
Statistics for Correlated
Data
 We concern ourselves with n realizations that
are related to time, that is having n correlated
observations; the estimate of the mean is
given by

 Let

 where the sum is over j = 1 to m, then

the estimated variance is:
 [1 + 2A ] S2 / n
 Where
 S2   = the usual variance estimate 
j,x = the jth coefficient of autocorrelation 
m    = the maximum time lag for which
autocorrelations are computed, such that j =
1, 2, 3, ..., m
Sample Size
Determination:
 We can calculate the minimum sample size
required by

 Application: A pilot run was made of a model,

observations numbered 150, the mean was
205.74 minutes and the variance S2 = 101,
921.54, estimate of the lag coefficients were
computed as: 1,x = 0.3301 2,x = 0.2993,
and 3,x = 0.1987. Calculate the minimum
sample size to assure the estimate lies
within + = 10% of the true mean with a = 0.05.
 n = [(1.96)2 (101,921.54) {1 + 2 [(1-1/4) 0.3301
+ (1 - 2/4) 0.2993 + (1- 3/4) 0.1987]}] /
(0.1)2 (205.74)2 

 » 1757