MGT 6400 Forecasting Project
Instructions:
Complete and summarize your answers on this page in the blocks provided. The specifics of your solution may be
saved as separate worksheets on this spreadsheet and referred to in your summary. Turn in your projec
Interpreting Forecasts
How useful are our forecasts? The usefulness of a forecast can be judged (with some fuzziness)
by looking at the confidence interval in light of how you intend to use the forecast and by looking
at the probability of certain outcome
TheMovingAverageProcess:
IdentificationandForecasting
The MA(1) Process
The MA(1) process is given by the function
In words, an MA(1) process for Y states that Y is equal to a mean plus a constant times the
white noise value from the prior period plus the
Eviews
Retail Sales Data
Book
Interpreting Forecasts
Calculating confidence intervals in EViews:
In the forecast screen of EViews add a name to S.E. (optional).
This is the estimated standard error of the forecasts. Note that they
change with each observa
Eviews
Housing Price Data
1
Checking the Validity of The Identified Model
We have established that for many series the appropriate model is not
obvious. When that occurs we try to determine a set of models that we
hope includes the appropriate model to us
We want to find E(Y). To do so let us start with n=1 (a
Bernoulli experiment). The probability function is
x
P(x)
1
P
0
1-p
Rule 4: Expectation of a Product
If X and Y are independent, then E(XY) = E(X)E(Y).
[This is just the definition of independence.]
Random Variable
A random variable is the assignment of a number to the
outcome of a random process.
-Example: A coin is tossed. The outcomes are either a head or a
tail (formally, we say that the sample space S = cfw_H, T).
We might assign the number 1 if
2
?
?
Let ? ? (? , )
?
we start by transforming the above into a probability statement about a standard normal random
variable, or
?
?1 ?
? ?
?2 ?
?
? ?
< ?
<?
? ?
? ?
? ?
= ? ?
?1 <
?
? ?
< ?2 ?
?
? ?
c*s are constants found in a table of standard normal
Professor Jerry
Thursby?
Business
Forecasting?
Forecasting Elements
Our intention is to forecast some random variable Y. We
are at time t and we want to forecast Y at time t+h.
The elements used in the forecast are
The information set (data) includes
Cur
State of the world
Decision
Null true
Null false
Reject null
Type I error
correct
Accept null
correct
Type II error
The probability of a Type I error is the chosen significance level .
The probability of a Type II error is unknown
It depends on the true v
EXCEL
Find probability for a t(v) random variable
P(Yy)
T.DIST(y,v,TRUE)
Forrighttailprobability
T.DIST.RT(y,v)
Forinverse function (youprovideprobability&EXCELgivesy)
T.INV(probability,v)
Measures of Association: Covariance and Correlation
Themostcommon
StatisticsNotes
BusinessForecasting
ProfessorJerryThursby
ErnestScheller,Jr.Chair
ErnestScheller,Jr.CollegeofBusiness
GeorgiaInstituteofTechnology
Statistics Review
Statistics and probability are closely linked. At the heart of both areas is the probabili
Why is the normal distribution so important?
Under very general conditions, if we sum many values
of some random variable, then the sum is
approximately normally distributed.
The approximation improves as the number of values
in the sum increases.
This
ARMA(p, q) process
The autocorrelation and partial autocorrelations
functions will appear to be a mix of the ones we
expect for pure MA and pure AR processes
Eviews example:
ARMA11_example.xlsx
1
Identification of the MA(q)
Only the first q autocorrelatio
ARMA Models and White Noise
All models will involve (in some way) what is known as a
white noise process
Such a process is completely unpredictable. This is the random
rather than systematic part of what we wish to predict.
A white noise process is charac
Instructions:
The data on the next sheet represents weekly sales for a pack of bathroom tiles at a reta
store in Atlanta, GA. The stocking policy at the store is to replenish and hold 250 units i
stock at the beginning of each week.
Your task:
1. Use simp
Instructions:
1. Use a double exponential smoothing forecasting method to forecast the average CO2
levels for the years 2014 - 2016.
2. To initiate the double exponential smoothig method, you need to find initial values
for Smoothing Level and Smoothing T
MGT 6400 Forecasting Project
Instructions:
Complete and summarize your answers on this page in the blocks provided. The specifics of your solution may be
saved as separate worksheets on this spreadsheet and referred to in your summary. Turn in your projec
(Customized Pricing Problem) ABC Corporation makes and sells medical testing equipment to laboratories at
hospitals, clinics, and universities across North America. One popular item is a gas chromatograph refill cartridge
that has a list price of $11.85.
(Customized Pricing Problem) ABC Corporation makes and sells medical testing equipment to laboratories at
hospitals, clinics, and universities across North America. One popular item is a gas chromatograph refill cartridge that
has a list price of $11.85.
b
Single Price
Demand General
Demand Student
Student Demand
Total Demand
TOTAL REVENUE
20000
0
0 <=
20000 <=
Price
100
10000
20000
2000000
Different Prices
Demand General
Demand Student
16923.077
3076.92306
Total Demand
Student Demand
Price money
TOTAL RE
FINANCIALACCOUNTING
3. Rolling Stones Supplies has the following stock outstanding reported on its balance sheet:
5% Cumulative preferred stock, par $20
Common stock, par $5
$100,000
$200,000
In 2006, Rolling Stones Supplies paid $28,000 in dividends. No
Financial and Managerial Accounting
1.
Thebalancesheetshows:
a. dividendsdistributedtostockholders.
b. operatingexpensesfortheperiod.
c. earningspersharefortheperiod.
d. claimsownershaveagainstassets.
e. salesrevenue.
2.
AbalancesheetdatedDecember3l,2005,
DO NOT TURN THE PAGE AND START
INSTRUCTIONS
This exam will require that you complete TWO Scantron Sheets. One Scantron Sheet will be for all 2
Point Questions; the 2nd Scantron Sheet Will Be for all 8 Point Questions (The Problems). If While taking
the Ex
EXCEL Example:
Estimation/Prediction Sample and
Forecasting
AR(1)Process
(InClass)
1
In our example we observe prediction errors et,h = yt+h ft,h
Can we judge how bad is a particular prediction error?
We do so with loss functions
A loss function L(et,h) i
FORECASTING WITH AUTOREGRESSIVE (AR) MO
A Variety of Time Series Cycles
500
(in thousands)
450
400
350
300
250
200
90 91 92 93 94 95 96 97 98 99 00 01
Home sales in California (16 counties)
1
Cycle: A cycle is a time series pattern of periodic fluctuation