MonteCarlito v1 10-2

MonteCarlito 1.
10
M. Auer, 2004-2012
Distributed under terms of GNU General Public License
www.montecarlito.com
This tutorial explains how MonteCarlito can be used to run Monte Carlo
simulations in Excel.
Monte Carlo simulations are often used to compute mathematical functions
where no fast, analytical solution exists (e.g., some financial functions,
or integrals), or to look at the statistical behaviour of complex systems.
Example 1: Basics
Basically, in a Monte Carlo simulation you analyze a random function, often to
get its average result (its so-called "expected value"). Say you are offered
to play the following game: you throw a dice, and get 1$ it you throw 1,
4$ if 2, 9$ if 3, 16$ if 4, 25$ if 5, and 36$ if 6.
But you also have to pay 14$ upfront to participate. Would you play that game?
To answer that question, you determine the expected profit/loss; an easy way
to do this is to simulate the game many times and average the individual outcomes.
To simulate this in Excel, first you need to create a random number from 0-1 with RAND():
Random number:
0.21995765
(press F9 several times to see how this number chan
From this, derive a dice using a mathematical function:

Dice:
(again, press F9 several times to see how this numbe
To play the game once, determine the profit of the dice throw, and subtract the costs:
Game profit:
Costs:
4
14
One game's profit/loss:
-10
Pressing F9 repeatedly reveals that sometimes you win, sometimes you lose money.
If you repeat this many times and analyze the outcomes, you can determine if you should play.
MonteCarlito is a tool to help you with that.
Just put the result of one (random) outcome in any cell, and write "1000" next to it:
1000
-10
(check with F9 it the outcome still changes)
Now, select the whole light blue area (2x2 cells) and press Ctrl+W.
MonteCarlito runs the simulation 1000 times, and gives you the average result. If it's positive, you should pla
al outcomes.
m 0-1 with RAND():
ral times to see how this number changes.)
9 several times to see how this number changes.)
tract the costs:
ou lose money.
rmine if you should play.
0" next to it:
it the outcome still changes)
ge result. If it's positive, you should play!
Example 2: Area of the circle

Say you forgot the number PI.
Monte Carlo simulation is one way to approximate it (and more often, more difficult integrals).
You could throw a dart at a 4cm x 4cm sized rectangle with a circle painted on it,
repeat this 1000 times, and count the number of throws landing in the circle area (call it C) and
the number of throws landing on the rectangle, in- or outside the circle (call this number R).
The ratio of C/R should be fairly similar to the ratio of the circle area and the rectangle area (which is 4 x 4 =
C/R = circle area/16
The circle area is thus 16xC/R.
But it is also r^2 x PI = 2 x 2 x PI
Thus PI = 4 x C/R
To simulate one dart throw, first create two random numbers (the x and y coordinate) between -2 and 2:
Throw to coordinates:
x
y
-1.60165752 1.62428126
Check if it is inside the circle:
(press F9 a few times to make sure th

(1=inside circle, 0 outside)
Now, use MonteCarlito with 1000 throws (ie, R=1000):

1000
Mean
0
0.805
805 is the number of circle hits C

1000 is the number of throws R
Estimated PI = 4 x C/R:
3.220
(select blue area and press Ctrl+W)
ore difficult integrals).
rcle area (call it C) and

call this number R).
the rectangle area (which is 4 x 4 = 16):
y coordinate) between -2 and 2:
press F9 a few times to make sure that the throw is between -2 and 2 in both axes)
1=inside circle, 0 outside)
select blue area and press Ctrl+W)
Example 2: Project management

Project planning tools assign effort estimates to subtasks.
Total project effort is the sum of the individual efforts.
It's more realistic to treat the subtasks as having random efforts,
the total effort is the sum of those random variables.
MonteCarlito can not only answer the expected total effort, but
also its distribution, and thus answer questions like "what are the 10% of outcomes that are worst?"
A very similar issue arises when computing financial measure like the value-at-risk (VAR).
Subtask 1, between 20 and 30 days:

Subtask 2, usually 15 days, but sometimes much longer:
23.2608186
18.6542744
Total effort:
41.915093
47.9586665
0.32193268
45.5857
10.1804053
103.640651
1000
Mean
Standard error
Median
Standard deviation
Variance
Select the blue area and press Ctrl+W.

The expected total effort is:
47.9586665
The standard deviation*1.96 is:
19.9535943
This means roughly that in 95% of all outcomes, the effort is between
and
47.9586665
47.9586665
omes that are worst?"
plus
minus
19.9535943
19.9535943
MonteCarlito
Instructions:
1. Put all N formulas you want to simulate next to each other, preceded by the number of trials you want to run.
2. Select the N+1 cells and the 7x(N+1) cells beneath (indicated by frame).
3. Run macro "simulate" or press Ctrl+W to run simulation. Result:
1000
Mean
Standard error
Median
Standard deviati
Variance
Skewness
Kurtosis
0.0301
48.6044
0.5036
2198.8597
0.0094
135.6636
0.4878
186.1393
0.2986
4290.0602
0.0891 18404616.8190
0.0039
2.4879
1.7135
8.7563
0.0691
-0.0131
0.0309
-0.0174
0.9761
0.9529
0.0473
2.9563
Note:
- If the number of trials is negative, simulation is run in high-speed mode with minimized windows.
- If the cell with the number of trials has bold font type, histograms are created at end of simulation run.
- If you select less than 7 rows of cells beneath formulas, you get less output.
ls you want to run.
Examples:
a) Select range I11:K13; press Ctrl+W.
100
0.2966
-0.9944
b) Select range I18:J21; press Ctrl+W (a new workbook will be created with histogram).
1000
0.0825861673
h histogram).

MonteCarlito v1 10-2

Uploaded by

Copyright:

Available Formats

MonteCarlito v1 10-2

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MonteCarlito v1 10-2

Uploaded by

Copyright:

Available Formats

What is a Monte Carlo simulation and what is it used for?

What is a Monte Carlo simulation and what is it used for?

How can Monte Carlo simulations be used to estimate the value of pi?

How can Monte Carlo simulations be used to estimate the value of pi?

MonteCarlito 1.

(press F9 several times to see how this number chan

From this, derive a dice using a mathematical function:

(again, press F9 several times to see how this numbe

One game's profit/loss:

(check with F9 it the outcome still changes)

m 0-1 with RAND():

ral times to see how this number changes.)

9 several times to see how this number changes.)

tract the costs:

0" next to it:

it the outcome still changes)

ge result. If it's positive, you should play!

Example 2: Area of the circle

Check if it is inside the circle:

(press F9 a few times to make sure th

Now, use MonteCarlito with 1000 throws (ie, R=1000):

805 is the number of circle hits C

(select blue area and press Ctrl+W)

ore difficult integrals).

rcle area (call it C) and

y coordinate) between -2 and 2:

1=inside circle, 0 outside)

select blue area and press Ctrl+W)

Example 2: Project management

Subtask 1, between 20 and 30 days:

Select the blue area and press Ctrl+W.

The standard deviation*1.96 is:

omes that are worst?"

ls you want to run.

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.