Download UNIVERSITY OF SOUTHERN CALIFORNIA

IOM 427: Designing Spreadsheet-Based Business Models (Fall 2010)
Marshall School of Business, University of Southern California
Instructor: Hao Zhang
Office: Bridge Hall 401G
Office hours: 11am-12pm, Tu & Thur
Phone: (213) 821-2279
Email: [email protected]
Grader: TBA
Email: TBA
Textbook
Management Science & Decision Technology, by Jeffrey D. Camm and James R. Evans
(South-West College Publishing, 2000).
Course Objective
Spreadsheets are convenient and widely available platforms for organizing information
and performing “what if” analyses. Excel therefore, has become an indispensable tool for
business analysis. This course will focus on structuring, analyzing and solving
managerial decision problems through Excel spreadsheets.
The goal of the course is to help you acquire the skills of logical reasoning with formal
models and become an effective modeler who can build sound models to solve real-world
business problems. The course is about modeling, not about becoming an Excel expert
per se.
We will study five broad classes of managerial problems:
1. Data Analysis: How to summarize available data into useful information. The
cost of collecting data has declined dramatically and most firms now have a fair
amount of data. The first few, perhaps the most useful, steps in understanding
and structuring a business decision is to find out what data is available and
organizing it to support decision making.
2. Resource Allocation: How to optimally allocate a limited pool of resources
among available opportunities. This is the most common managerial problem,
occurring in every functional area. Examples in finance include constructing an
optimal risk-return portfolio and capital budgeting. Examples in marketing
include media planning and sales force territory planning. In operations
management, resource allocation problems arise in capacity, logistics, and
operations planning.
3. Estimation and Forecasting: How to find important information and
implications from historical data and how to extrapolate past data into the future.
We will explore a handful of models and techniques for estimation and
forecasting.
4. Decision Analysis/ Contingent Decisions: How to synthesize a sequence of
decisions involving uncertainty. An intuitive approach to handling uncertainty is
to explore the possibility of deferring a decision until some uncertainty is resolved,
especially when the stakes are high. If we can we should make sequence of
decisions instead of one big decision. Business examples where such decision
techniques are used include dynamic portfolio management, new product
development, and capacity expansion planning.
5. Risk Analysis: How to incorporate uncertainty in problem parameters. Almost
always managerial decisions are based on anticipated states of the business
environment. Clearly as the decision horizon becomes longer there is an increase
in uncertainty. Managers have to carefully consider different potential scenarios
while making decisions. In this part of the course we will learn how to explicitly
incorporate uncertainty into business models.
Excel Skills and Software Usage
Previous knowledge of Excel is not required. Knowing how to enter formulae involving
relative and absolute cell addresses and how to illustrate using chart wizard is sufficient.
We will learn Pivot Table, Filters, etc. for processing data, Solver for finding optimal
decisions, Time series models for forecasting, Treeplan for drawing decision trees, and
Crystal Ball for analyzing risk and uncertainty.
Grading
Grades will be curved and will be based on four homework assignments, two exams, and
class participation according to the following weights:
Homework assignments
Midterm exam
Final exam
Class participation
30%
30%
35%
5%
Exams will be in-class and open book/open notes. No make up exam will be given. Late
homework submission will not be graded.
Blackboard Usage
Blackboard will serve as the information center for the course. Handouts, Powerpoint
slides, example files, assignments, solutions, and supplementary reading materials will all
be posted on Blackboard.
Notice on Academic Integrity
The use of unauthorized material, communication with fellow students during an
examination, attempting to benefit from the work of another student, and similar behavior
that defeats the intent of an examination or other class work is unacceptable to the
University. It is often difficult to distinguish between a culpable act and inadvertent
behavior resulting from the nervous tensions accompanying examinations. Where a clear
violation has occurred, however, the instructor may disqualify the student's work as
unacceptable and assign a failing mark on the paper.
For Students with Disabilities
Any student requesting academic accommodations based on a disability is required to
register with Disability Services and Programs (DSP) each semester. A letter of
verification for approved accommodations can be obtained from DSP. Please be sure the
letter is delivered to the office as early in the semester as possible. DSP is located in
STU 301 and is open 8:30 a.m. - 5:00 p.m., Monday through Friday. The phone number
for DSP is (213) 740-0776.
Time Table
Dates
Topics
Aug. Course Introduction
24
Objective, outline, skill sets, textbook, expectations
Introduction to Modeling
Definition of modeling, types of models, functional area
applications, modeling steps;
A simple profit model in math and in Excel;
Card game simulation and analysis
26 Excel Basics
Workbook/sheet navigation, window split/freeze, column/row
operations, sequences, absolute/relative cell references, range
names, auditing tools, menu items, basic functions (Min, Max,
Average, Count, etc.)
31 Excel Basics
Intermediate functions (If, And, Or, Sumproduct, Vlookup,
Index, Match, Pmt, etc.)
Data Analysis
Data searching, editing, sorting, filtering, tabulating (pivot
tables)
Sept. Data Analysis
2
Data importing (from file and Internet), “data table” analysis
tool (with one or two changing variables);
Practice: database management for global manufacturer
Hanover, Inc.
7 Data Analysis
Data visualization
Review of data analysis and Hanover example
Example: Santa Cruz MicroProducts
Building Excel model, finding the best product mix by data
table and by Solver
Exercises
Read: Ch 1, pp. 213.
Exercise: Excel
tutorial (self
practice).
Read: Ch 2, pp.
49-71.
Exercise: selfpractice set 1.
Read: Ch 2, pp.
49-71.
Exercise: selfpractice set 2.
Read: Ch 1, pp.
13-15.
9 Introduction to Optimization
Optimization model components, types of optimization models,
types of linear constraints, Excel model layouts and guidelines,
solution possibilities
Linear Programming
Example: product mix (Santa Cruz MicroProducts)
Math formulation, Solver setup
Example: investment allocation (Kathryn)
Alternative Excel models, universal linear programming
template, intuitive solution
14 Linear Programming
Example: allocation of marketing effort (Phillips, Inc.)
Choosing decision variables, handling upper/lower bounds
efficiently, modeling and Solver tips
Example: diet problem (Colorado Cattle Company)
Standard linear programming form, bottleneck constraints
16 Linear Programming
Example: multiperiod production planning (Suzie’s Sweatshirts)
Network diagram, inventory balance equation, alternative
models, solution intuition
Example: working capital management (Vohio, Inc.)
Network diagram, flow balance equation, solution diagram,
solution intuition
21 Linear Programming
Example: transportation problem (Hanover, Inc.)
Connection with Hanover data analysis example, network
diagram, special (rectangle) Excel layout
Sensitivity Analysis
Changing right-hand sides, bottleneck constraints, shadow
prices; changing objective coefficients, basic (non-zero)
variables; Solver sensitivity report
Homework 1 Q&A
23 Homework 1 Student Presentation
Integer Programming
Introduction: integer and binary variables, Solver setup
Example: production planning (Queen City, Inc.)
Difference between linear and integer solutions
Example: project selection (Burke Construction)
Describing logical relationships by binary variables and linear
constraints, binary variables in objective function
28 Integer Programming
Example: production planning with setups (Chemco, Inc.)
Inventory model with and without setups, network diagram,
constraints linking binary variables and continuous variables
Example: staff scheduling (Airport Services, Inc.)
Choosing decision variables, binary coverage table
30 Integer Programming
Read: Ch 4, pp.
141-150.
Exercise: selfpractice set 3.
Read: Ch 4, pp.
150-154.
HW1 (due 9/23):
linear
programming.
Read: Ch 4, pp.
155-160.
Read: Ch 4, pp.
160-177.
Read: Ch 5, pp.
200-206, 212-216.
Read: Ch 5, pp.
207-208, handout.
Extra-credit
exercise: what if
each worker works
any five days (not
necessarily
consecutive) in a
week?
Read: Ch 5, pp.
Oct.
5
7
12
14
19
21
26
Example: location problem (Sun Bank) I
Choosing decision variables, modeling adjacency
relationships, linking principal place of business with
surrounding branches, creating coverage table
Example: location problem (Sun Bank) II
Preventing a branch from double counting
Nonlinear Programming
Introduction: nonlinear functions, “nice” vs. “nasty” functions,
global vs. local optimal solutions, nonlinear constraints,
Solver setup
Example: advertising budget allocation (Phillips, Inc.)
Determining nonlinear objective function (relationship
between profit and advertising money) through regression
Example: facility location (Jack’s Job Shop)
Determining nonlinear objective function through geometry
Nonlinear Programming
Example: Markowitz portfolio optimization model
Expressing expected portfolio return from allocation variables,
expressing portfolio variance from allocation variables,
variance-covariance matrix, 3-stock Excel model, 30-stock
Excel model, sensitivity analysis, risk-return frontier
Homework 2 Q&A
Homework 2 Student Presentation
Midterm Review I
Practice Problem 1: traffic network throughput
Practice Problem 2: baseball player selection
Midterm Review II
Practice Problem 3: emergency vehicle locations (I & II)
Practice Problem 4: landfill locations
Discussion: sample midterm exam solutions
Midterm Exam (4:00pm-5:50pm)
Forecasting
Introduction:
Time series components and decomposition, forecasting
model classification, forecast errors (mean absolute deviation,
root mean square error, mean absolute percentage error)
Simple moving average model (data with no trend or
seasonality):
Selecting window size based on errors
Simple exponential smoothing model (data with no tread or
seasonality):
Selecting the smoothing constant through data table or Solver
Double moving average model (data with trend):
Linear trend equation, level and trend
Midterm Exam Discussion
Forecasting
Double exponential smoothing model (data with trend):
Smoothing constants initialization, selecting smoothing
constants through data table or Solver
Additive seasonality model (data with seasonality)
209-211, 216-218.
HW2 (due 10/12):
integer &
nonlinear
programming.
Read: Ch 5, pp.
218-221, 223-224,
226-228, 229-231.
Read: Ch 5, pp.
221-223, 228-229.
Exercise: sample
midterm exam.
Exercise: sample
midterm exam.
Read: Ch 3, pp.
97-108.
Read: Ch 3, pp.
109-113.
28 Forecasting
Multiplicative seasonality model (data with seasonality)
Holt-Winters additive model (data with trend and seasonality)
Holt-Winters multiplicative model (data with trend and
seasonality)
Forecasting with regression:
Simple linear regression models, Excel “add trendline” tool
Nov. Decision Analysis
2
Introduction:
Decision alternatives, outcomes, payoff table, decision criteria
(average-payoff, aggressive, conservative, opportunity-loss,
and expected-value criteria)
Decision tree basics:
Diagram, decision nodes, event nodes, branches
Example: experimental drug development
Drawing decision tree, Treeplan Excel add-in, rollback
procedure (two methods)
4 Decision Analysis
Example: new product introduction
Drawing decision tree, Treeplan tips, sensitivity analysis by
one- and two-variable data tables
Example: who wants to be a millionaire?
Solving decision tree by Treeplan
9 No Lecture (self practice)
11 Decision Analysis
Example: DriveTek subcontracting problem
Solving decision tree by Treeplan, changing cost of
mechanical method (one-variable sensitivity analysis),
changing success probabilities of electronic and magnetic
methods (two-variable sensitivity analysis)
Homework 3 Q&A
16 Homework 3 Student Presentation
Monte Carlo Simulation
Introduction:
Simulation and risk, random numbers, histogram, probability
distributions (two points, binomial, Poisson, uniform,
triangular, normal), simulation modeling process
Practice:
Generating random numbers by Rand(), drawing distribution
by histogram data analysis tool, rolling one or two dice and
viewing distribution (dice roller simulator I & II)
Example: profit model of a firm
Creating base model, generating discrete random variables by
Rand(), defining random variables in Crystal Ball, defining
output cell in Crystal Ball, analyzing simulation result
18 Monte Carlo Simulation
Example: newsvendor problem
Key trade-off in newsvendor problems, profit formula, base
model with fixed order quantity, Crystal Ball simulation,
finding best order quantity via data table and decision table
Read: Ch 3, pp.
113-117, 121-123.
HW3 (due 11/16):
forecasting &
decision analysis.
Read: Ch 6, pp.
247-253, 254-257.
Read: Ch 6, pp.
253-254, 266-268.
Read: Ch 7, pp.
286-301.
Read: Ch 7, 304309.
HW4 (due 12/2):
simulation.
23
25
30
Dec.
2
9
tools
Example: pricing stock options
Daily or weekly stock price formula, types of stock options,
strike price and expiration date, fair price of an option, Crystal
Ball simulation
Crystal Ball tips: Crystal Ball functions and decision table tool
Monte Carlo Simulation
Example: bidding for contract (Miller Construction Company)
Modeling competitors’ bids with and without correlation,
choosing Miller’s bid through decision table tool
Crystal Ball tip: correlation matrix tool
Example: Craps game I
Casino games, rules of Craps game, model layout, formula for
win or loss in the “come-out” phase, modeling the “point”
phase with an arbitrary number of rolls, formula for win or
loss of the overall game, finding winning probability through
Crystal Ball
No Lecture (Thanksgiving Day)
Monte Carlo Simulation
Example: Craps game I
Introducing an alternative model
Example: Craps game II
Multiple rounds with winning probability from part I,
terminating the game after player losing all money or reaching
round 100
Crystal Ball tip: tornado chart tool
Homework 4 Q&A
Homework 4 Student Presentation
Monte Carlo Simulation
Example: NCAA basketball tournament
Modeling games with random outcomes according to
Sagarin’s ratings, managing multiple rounds of games from 64
teams down to the final two and the champion
Final Review
Final Exam (4:30-6:30pm)
Read: Ch7, pp.
321-323.
Read: Ch7, pp.
310-311.
Exercise: sample
final exam.
Exercise: sample
final exam.