0% found this document useful (0 votes)
10 views27 pages

Understanding Monte Carlo Simulation

This document provides an overview of Monte Carlo simulation, a method for predicting outcomes in uncertain decision-making situations by using probability distributions. It outlines the mechanics of simulation, including constructing deterministic models, applying distributions, and running iterations to analyze risk profiles. Additionally, it discusses the application of simulation in various scenarios, emphasizing its advantages over decision trees in complex situations with multiple uncertainties.

Uploaded by

ucheuke
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
10 views27 pages

Understanding Monte Carlo Simulation

This document provides an overview of Monte Carlo simulation, a method for predicting outcomes in uncertain decision-making situations by using probability distributions. It outlines the mechanics of simulation, including constructing deterministic models, applying distributions, and running iterations to analyze risk profiles. Additionally, it discusses the application of simulation in various scenarios, emphasizing its advantages over decision trees in complex situations with multiple uncertainties.

Uploaded by

ucheuke
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Overview of Chapter 11

In this chapter, we will cover Monte Carlo simulation:


A different approach than those already discussed to
dealing with uncertainty in a decision situation
Constructing a model that captures all of the relevant
aspects of the uncertainty in the decision problem
Then translating this model into a form that a computer
can use to run a simulation
The role of simulation in creating a requisite decision-
analysis model

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Monte Carlo Simulation

What Is a Monte Carlo Simulation?


• A Monte Carlo simulation is a model used to predict the
probability of a variety of outcomes when the potential for
random variables is present.
• Monte Carlo simulations help to explain the impact of risk and
uncertainty in prediction and forecasting models.
• A Monte Carlo simulation requires assigning multiple values to an
uncertain variable to achieve multiple results and then averaging
the results to obtain an estimate.
• A Monte Carlo simulation is used to tackle a range of problems in
many fields including investing, business, physics, and
engineering.
© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Monte Carlo Simulation

The Leah Sanchez calendar sales example is simulated


using the DecisionTools software.
We will be going over examples in class made using
MS Excel.
Use the tool(s) you prefer or some other spreadsheet
software such as Google Sheets.
The example spreadsheets are located in
Blackboard. Go to “Lecture Presentations”, “Chapter
11 Materials”.

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Mechanics of Simulation
A simulation model is a mathematical model in
which a probability distribution is used to represent
the possible values of an uncertain variable.
• Similar to decision trees and influence diagrams
• Allow for continuous as well as discrete distributions
Before the development of any model…

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Mechanics of Simulation
Summary of steps in running a simulation:
1. Construct a deterministic model
• A Deterministic Model allows you to calculate a future
event exactly, without using randomness.
• If something is deterministic, you have all of the data
necessary to predict (determine) the outcome with
certainty.
• No probability distributions are in this model.

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Mechanics of Simulation
Summary of steps in running a simulation:
2. Apply distributions to the values in the
deterministic model where you expect variation or
uncertainty
• You now have a probabilistic or stochastic model.
• A stochastic model is a method for predicting possible
outcomes by using for random distributions for one or
more values.
• These distributions may be assumed or based on
beliefs.
© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Mechanics of Simulation
3. Randomly draw (sample) values from the
distributions to apply to the model for recalculation
• With each new draw, you are running different
combinations of your model thru 1,000s of iterations.
• Each iteration is a single sample from the distribution.
4. Plot the outcomes of the iterations of the model
• This gives you the distribution of the uncertainty of
interest – risk profile
• You can now factor probabilities into your decisions.

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Mechanics of Simulation
Iteration – a recalculation of the model
For every iteration, a new value is chosen for each
uncertainty according to the corresponding probability
distribution, and this value is used in the calculations for
that particular iteration.
Increasing the number of iterations results in sampled
values more closely aligned with the distribution
At a minimum run 1,000 iterations; 10,000s is not
unusual

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Simulation Models

Since influence diagrams and decision trees also deal


with uncertainties, they can be helpful in constructing
simulation models.
Use them to identify which variables are decisions,
computations, and uncertain
Also they can help to determine the relationships
among the variables

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Simulation Models

Good modeling techniques:


1. Organize the model in a coherent and natural way
2. Put each constant in its own cell location
• Often constants are grouped in a “data area” in the
upper left corner of the spreadsheet.
3. Never enter constants directly into formulas
• Any formula should contain only cell references.

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Simulation Models
Good modeling techniques:
4. Break formulas down into logical pieces, with each
piece having a separate cell location
• Avoid long formulas – they can easily lead to errors!
5. Build in quality checks such as redundant
calculations
6. Verify the model
7. Fully document the model

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Simulating the Model
Next step: model each uncertain variable with its
chosen probability distribution.
• Substitute probability distributions for uncertain variables
• This makes the deterministic model into a probabilistic or
stochastic model, i.e., a model whose values are
indeterminate and governed by probability distributions.
• Run the simulation through many iterations to get a
distribution or risk profile of possible values
• Gain realistic insights into the decision situation
• Can also compare the results of using different input values
and find the best one
© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Simulation, Decision Trees, and Influence Diagrams

Monte Carlo simulations are powerful and easy to


use decision-making tools.
However, a few cautions are warranted:
• Consider whether or not your decision model needs to
incorporate dependence. If so, don’t use simulations
since they require independence.
• Only put in your simulation model variables that are
truly explanatory, i.e., those that provide insight.
• Too easy to throw into the simulation every possible source.

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Simulation, Decision Trees, and Influence Diagrams

“When should I use simulation, and when should I use


decision trees?”
• In many cases both approaches work fine.
• However, there are two key issues to consider:
• First, if your decision situation involves a large number of
uncertainties, the necessarily large decision tree can be very
clumsy to work with. Use a simulation approach.
• Second, if your decision situation involves future or
“downstream” decisions then a decision tree might be easier
to work with. Don’t use a simulation approach.

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
Summary

In this chapter, you have learned about Monte Carlo


simulation. Specifically, we have discussed:
Choosing values for uncertain variables according
to specified probability distributions
Calculating the consequences based on the values
chosen by running multiple iterations
Tracking the consequences to understand how they
are distributed, i.e., the risk profile

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
LAC LEMAN FESTIVAL DE LA MUSIQUE (Synopsis)

The organizers of a music festival may use video from the Friday
concert to create a DVD to sell to those who come to the Saturday
concert. Attendance on Saturday is uncertain, as is the percentage
of those who attend on Saturday that will buy the DVD.
Is this a good idea?
If so, how many DVDs should be burned early Saturday morning and
offered for sale at that evening’s performance?
By that time, Friday attendance is known, along with whether it rained
on Friday, and there is a forecast for whether it will rain on Saturday.
Historical information on these variables may help us to predict
Saturday attendance; along with the results of a marketing survey,
such analysis will help us make better purchasing decisions.

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
LAC LEMAN FESTIVAL DE LA MUSIQUE (Synopsis)

Written Submission
Part A: Provide a complete forecast for attendance at Saturday
night’s performance using case Exhibit 1 only. Note: a complete
forecast acknowledges the uncertainty with a probability
distribution.

Part B: If 4,500 DVDs are produced on Saturday for the concert


later Saturday evening, what is the risk profile of profit
(including all costs related to the DVD project)? Also, make a
recommendation for how many DVDs to order from the
production company. Provide a risk profile for the profit that will
result.

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
SPRIGG LANE (Synopsis)
Tom Dingledine, the president of a natural-resources exploration company, has
to decide
whether to invest in a new gas-well-drilling opportunity. He already has a
spreadsheet that projects the most likely scenario for the well and calculates
the net present value (NPV) and internal rate of return (IRR).
Dingledine, however, has discussed six uncertainties with another investor; he
now needs to incorporate them into the analysis.
He has prepared a spreadsheet for two downside scenarios: in the first, gas
cannot be produced after the well is drilled; in the second, gas can be
produced but all other uncertainties are at their one-percentile worst possible
values.
A potential investor, Henry Ostberg, wants to know what the chances are that
the gas well willresult in a loss of value.

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
SPRIGG LANE (Synopsis)

Written Submission
Based on the base-case scenario and the two alternative
downside possibilities, is this investment economically
attractive? Also, what benefit can Monte Carlo simulation add to
Dingledine’s understanding of the economic benefits of the
Bailey prospect?

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
APPSHOP, INC. (Synopsis)

A regional director of a consulting firm must decide how to compete for a


major consulting contract.
Appshop can take a level-payment contract, a lower-level payment with
a prospective bonus for high performance, or bid on an RFP where a
significant reward is given contingent on the client’s savings.

Written Submission
Appraise the risk of the alternatives, and recommend what Eric Clark
should choose. Suppose Clark’s specific secondary incentive is to keep
blended revenue per hour above 150. How does that affect your view of
the risk.

© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
CALAMBRA OLIVE OIL (Synopsis)

Frank Lockfeld, president and founder of Calambra Olive Oil, is


discussing over dinner with friends the order-quantity decision he is
facing. Lockfeld’s newly launched venture—to bring high quality,
vintage-dated California olive oil to market—has been selling product for
only three months, and has sold only 24 cases to date. Nevertheless,
even though the first-year marketing experiment has barely begun,
Lockfeld must decide by month’s end how many gallons of oil to order
for the next (1994 vintage) year.

Written Submission
How many gallons should Lockfeld order? Then, how would you
describe the situation Lockfeld is facing? Can you graphically depict it?
How and why?
© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.
[Link] (Synopsis)

A potential investor in a new on-line sheet-music business is ready to


deny funding based on simple expected monetary value. Further
reflection identifies potential downstream options. Two options are
important in the case of partial, but incomplete, success: to abandon the
business idea and sell the technology, or to switch to new technology
and keep the Web site.

Written Submission
Provide a description of any other contingent opportunities (generic) that
would add value to this business? Should Bernard invest in the
business? Does Bernard have reasonable cutoff levels to trigger action
on each of the ideas in the initial discussion question? What kinds of
assessment tools would you use to support your answers?
© 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part, except for use as permitted in a license distributed with a certain product or service or otherwise on a password-
protected website for classroom use.

Common questions

Powered by AI

Application of probability distributions transforms a deterministic model into a stochastic model by introducing variability and randomness into the model. In a deterministic model, outcomes are computed using fixed input values, without any consideration of risk or uncertainty . By substituting these fixed inputs with probability distributions, we account for the randomness inherent in real-world scenarios, allowing the model to simulate a range of possible outcomes and assess associated risks .

Using historical data and simulations in predicting attendance and DVD sales presents both risks and benefits. The benefits include obtaining a probabilistic forecast of attendance, which helps in decision-making related to production and sales strategies, reducing potential losses from unsold inventory . However, risks include reliance on past data which may not accurately reflect current consumer behavior or market conditions, and potential errors in estimating probability distributions that could lead to incorrect forecasts affecting profitability .

Simulation models and decision trees both address uncertainties but differ in approach and application. Simulation models use probability distributions to explore a range of outcomes, accommodating many uncertainties efficiently via multiple iterations, making them suitable for complex scenarios with numerous uncertain variables . Decision trees, however, map out sequential decisions and their potential consequences in a structured way, which is more intuitive when dealing with scenarios involving future decisions or fewer uncertainties, though they may become cumbersome if faced with extensive uncertainties .

When including variables in a Monte Carlo simulation model, caution should be exercised to ensure that only explanatory variables that provide meaningful insights are included. Overloading the model with every possible source of uncertainty can complicate the analysis and dilute the model's predictive power. It's crucial to determine if relationships among variables require dependence, as simulations assume independence. Thus, inclusion decisions should focus on essential variables that have a significant impact on the outcome .

A Monte Carlo simulation would be preferred over a deterministic model in scenarios where there is significant uncertainty in input variables. Deterministic models do not account for randomness and assume a single set of known inputs, which limits their applicability when predicting outcomes in uncertain environments. Monte Carlo simulations, on the other hand, allow for the exploration of multiple possible outcomes by applying probability distributions to uncertain variables, providing a more comprehensive risk assessment and probabilistic forecasts .

To construct an effective simulation model that handles uncertainties, begin by building a deterministic model, providing a baseline to understand outcomes without randomness . Next, apply probability distributions to values where variation or uncertainty is expected. This transition creates a stochastic model, representing variables with random distributions. Finally, run simulations with multiple iterations, drawing random samples for each variable from its respective distribution to observe varied outcomes, developing a risk profile for the decision situation .

Monte Carlo simulation is effective for businesses like Calambra Olive Oil facing uncertain demand because it enables them to analyze various demand scenarios by assigning probability distributions to uncertain variables such as market demand and sales. This helps in estimating potential sales outcomes and associated risks. By running numerous iterations, the business can develop a comprehensive risk profile, guide inventory decisions, and optimize order quantities to minimize the risk of overproduction or stockouts .

Influence diagrams and decision trees play a crucial role in constructing and refining simulation models by helping identify key variables involved in decisions, computations, and uncertainties as well as the relationships among them. These tools structure the decision-making process and simplify the identification of areas where uncertainty is critical, aiding in the formulation of models that appropriately represent real-world complexities and guide the setup of probability distributions for simulation .

Monte Carlo simulation aids in understanding risk and uncertainty by enabling predictions of a variety of outcomes when random variables are present. By assigning multiple values to an uncertain variable, Monte Carlo simulations generate multiple results, which are averaged to produce an estimate. This process can help illustrate the impact of risk and uncertainty on forecasting models, providing insights into possible outcomes and their probabilities .

Monte Carlo simulations closely resemble real-world business scenarios by modeling complex decisions involving uncertainty, as seen in Sprigg Lane's investment decision. They allow for multiple iterations to evaluate the impact of varying conditions like fluctuating gas prices or production costs, simulating different outcomes to assess risks and rewards. For example, simulations can predict the probability of different NPV or IRR outcomes for Sprigg Lane, aiding investors in understanding potential losses or gains and making well-informed decisions .

You might also like