Chapter 5

Demand
Estimation and
Forecasting

Chapter Outline

Regression analysis
Limitation of regression analysis
The importance of business forecasting
Prerequisites of a good forecast
g techniques
q
Forecasting

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-2

Learning Objectives
Understand the importance of forecasting in
business
Know how to specify and interpret a
regression
i
model
d l
Describe the major forecasting techniques
used in business and their limitations
Explain basic smoothing methods of
forecasting such as the moving average
forecasting,
and exponential smoothing

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-3

Data Collection
Statistical analyses are only as good as the
accuracy and
d appropriateness off the
h sample
l
of information that is used.
Several
S
l sources off data
d t for
f business
b i
analysis:

buy from data providers (e

(e.g.
g ACNielsen,
ACNielsen IRI)
perform a consumer survey
focus groups
technology: point-of-sale data sources

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-4

Regression Analysis
Regression analysis: a procedure
commonly
l used
db
by economists to estimate
consumer demand with available data
Two types of regression:
cross-sectional:
cross sectional: analyze several variables for a
single period of time
time series data: analyze
y a single
g variable over
multiple periods of time

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-5

Regression Analysis
Regression equation: linear, additive
eg:

Y = a + b1X1 + b2X2 + b3X3 + b4X4

Y: dependent variable
a: constant value, y-intercept
Xn: independent variables, used to explain Y
bn: regression coefficients (measure impact
of independent variables)

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-6

Regression Analysis
Interpreting the regression results:
Coefficients:
negative coefficient shows that as the
independent variable (Xn) changes,
changes the variable
(Y) changes in the opposite direction
positive coefficient shows that as the
p
independent variable (Xn) changes, the
dependent variable (Y) changes in the same
direction
The regression coefficients are used to compute
the elasticity for each variable
Copyright 2014 Pearson Education, Inc. All rights reserved.

5-7

Regression Analysis
Statistical evaluation of regression results:
t-test: test of statistical significance of each
estimated
ti
t d coefficient
ffi i t (whether
( h th the
th coefficient
ffi i t is
i
significantly different from zero)

b
t=
SE b
b
=
=
Seb
coefficient
ffi i t

estimated coefficient
standard error of estimated

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-8

Regression Analysis
Statistical evaluation of regression results:
rule of 2: if absolute value of t is greater than
2 estimated
2,
ti
t d coefficient
ffi i t is
i significant
i ifi
t att the
th 5%
level (for large samples-for small samples, need
to use a t table))
if coefficient passes t-test, the variable has a
significant
i ifi
t iimpactt on d
demand
d

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-9

Regression Analysis
Statistical evaluation of regression results
R2 (coefficient of determination): percentage of
variation
i ti
iin th
the variable
i bl (Y) accounted
t d ffor by
b
variation in all explanatory variables (Xn)
R2 value ranges
g from 0.0 to 1.0
The closer to 1.0, the greater the explanatory power of
the regression.

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-10

Regression Analysis
Statistical evaluation of regression results
F-test: measures statistical significance of the
entire
ti regression
i
as a whole
h l (not
( t each
h
coefficient)

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-11

Regression Analysis
Steps for analyzing regression results
check coefficient signs and magnitudes
compute elasticity coefficient
determine statistical significance

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-12

Regression Analysis
Textbook example: Management lessons
f
from
estimating demand
d
d for
f pizza
demand for pizza affected by
1. price of pizza
1
2. price of complement (soda)

managers can expect price decreases to lead to

lower revenue
tuition and location are not significant

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-13

Regression Analysis
Challenges
Identification
Multicollinearity
Autocorrelation
A t
l ti

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-14

Regression Analysis
Challenge 1: Identification problem:
The estimation of demand may produce biased
results due to simultaneous shifting of supply
and demand curves.
curves
Solution: use of advanced correction techniques,
such as two-stage least squares and indirect
least squares may compensate for the bias

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-15

Regression Analysis
Challenge 2: Multicollinearity problem
Two or more independent variables are highly
correlated, thus it is difficult to separate the
effect each has on the dependent variable.
variable
Solution: a standard remedy is to drop one of
the closely related independent variables from
the regression

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-16

Regression Analysis
Challenge 3: Autocorrelation problem
Also known as serial correlation, occurs when the
dependent variable relates to the Y variable
according to a certain pattern
Note: possible causes include omitted variables,
or non-linearity; Durbin-Watson statistic is used
to identify autocorrelation
Solution: to correct autocorrelation consider
transforming the data into a different order of
magnitude or introducing leading or lagging data

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-17

Forecasting
Forecasting is very difficult, especially into
the
h future.
f

Common subjects of business forecasts:

gross domestic product (GDP)
components of GDP
examples: consumption expenditure
expenditure, producer durable
equipment expenditure, residential construction

industry forecasts
example: sales of products across an industry

sales of a specific product

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-18

Forecasting
A good forecast should:
be consistent with other parts of the business
be based on knowledge of the relevant past
consider
id th
the economic
i and
d political
liti l environment
i
t
as well as changes
be timely

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-19

Forecasting Techniques
Factors in choosing the right forecasting
technique:
h
item to be forecast
interaction
i t
ti
off the
th situation
it ti
with
ith the
th forecasting
f
ti
methodology--the value and costs
amount of historical data available
time allowed to prepare forecast

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-20

Forecasting Techniques
Six forecasting techniques

expert opinion
opinion polls and market research
surveys off spending
di
plans
l
economic indicators
projections
econometric models

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-21

Forecasting Techniques
Approaches to forecasting
qualitative forecasting is based on judgments
expressed
d by
b individuals
i di id l or group
quantitative forecasting utilizes significant
amounts of data and equations

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-22

Forecasting Techniques
Approaches to quantitative forecasting:
nave forecasting projects past data without
explaining
l i i
ffuture
t
ttrends
d
causal (or explanatory) forecasting attempts to
explain the functional relationships between the
dependent variable and the independent
variables
i bl

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-23

Forecasting Techniques
Expert opinion techniques
Jury of executive opinion: forecasts generated
b a group off corporate
by
t executives
ti
assembled
bl d
together

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-24

Forecasting Techniques
Expert opinion techniques
The Delphi method: a form of expert opinion
f
forecasting
ti
th t uses a series
that
i off questions
ti
and
d
answers to obtain a consensus forecast, where
p
do not meet
experts

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-25

Forecasting Techniques
Opinion polls: sample populations are
surveyed
d to d
determine consumption trends
d
may identify changes in trends
choice of sample is important
questions must be simple and clear

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-26

Forecasting Techniques
Market research: is closely related to
opinion polling
ll
and
d will
ll indicate
d
not only
l
why the consumer is (or is not) buying, but
also
who the consumer is
how he or she is using the product
characteristics the consumer thinks are most
important in the purchasing decision

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-27

Forecasting Techniques
Surveys of spending plans: yields
information
f
about
b
macro-type

data
d
relating
l
to the economy, especially:
consumer intentions
examples: Survey of Consumers (University of
Michigan), Consumer Confidence Survey (Conference
Board)

inventories and sales expectations

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-28

Forecasting Techniques
Economic indicators: a barometric
method
h d off forecasting
f
designed
d
d to alert
l
business to changes in conditions
leading, coincident, and lagging indicators
composite index: one indicator alone may not be
very
y reliable,, but a mix of leading
g indicators may
y
be effective

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-29

Forecasting Techniques
Leading indicators predict future economic
activity
average hours, manufacturing
initial claims for unemployment insurance
manufacturers new orders for consumer goods
and materials
vendor p
performance,, slower deliveries diffusion
index
manufacturers new orders, nondefense capital
goods
Copyright 2014 Pearson Education, Inc. All rights reserved.

5-30

Forecasting Techniques
Additional leading indicators to predict
future economic activity

building permits, new private housing units

stock prices, 500 common stocks
money supply,
supply M2
interest rate spread, 10-year Treasury bonds
minus federal funds
index of consumer expectations

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-31

Forecasting Techniques
Coincident indicators identify trends in
current economic activity

employees on nonagricultural payrolls

personal income less transfer payments
industrial production
manufacturing and trade sales

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-32

Forecasting Techniques
Lagging indicators confirm swings in past
economic activity
average duration of unemployment, weeks
ratio, manufacturing and trade inventories to
sales
change in labor cost per unit of output,
manufacturing (%)

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-33

Forecasting Techniques
Additional lagging indicators confirm swings
in past economic activity
average prime rate charged by banks
commercial and industrial loans outstanding
ratio,
ratio consumer installment credit outstanding to
personal income
change
g in consumer p
price index for services

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-34

Forecasting Techniques
Economic indicators: drawbacks
leading indicator index has forecast a recession
when
h
none ensued
d
a change in the index does not indicate the
precise size of the decline or increase
the data are subject to revision in the ensuing
months

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-35

Forecasting Techniques
Trend projections: a form of nave
f
forecasting
that
h projects trends
d from
f
past
data without taking into consideration
reasons for the change
compound growth rate
visual time series projections
least squares time series projection

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-36

Forecasting Techniques
Compound growth rate: forecasting by
projecting the
h average growth
h rate off the
h
past into the future
provides a relatively simple and timely forecast
appropriate when the variable to be predicted
increases at a constant percentage

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-37

Forecasting Techniques
General compound growth rate formula:
E/B = (1+i)n
E
n
B
i

=
=
=
=

final value
years in the series
beginning
b i i
value
l
constant growth rate

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-38

Forecasting Techniques
Visual time series projections: plotting
observations on a graph and viewing the shape of
the data and any trends

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-39

Forecasting Techniques
An Example in
Which
h h the
h
Constant
Compound
Growth Rate
Approach Would
Be Misleading

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-40

Forecasting Techniques
Time series analysis: a nave method of
f
forecasting
from
f
past data
d
by
b using least
l
squares statistical methods to identify
trends cycles,
trends,
cycles seasonality,
seasonality and irregular
movements

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-41

Forecasting Techniques
Time series analysis:
easy to calculate
does not require much judgment or analytical
skill
describes the best possible fit for past data
usually reasonably reliable in the short run

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-42

Forecasting Techniques
Time series data can be represented as:
Yt = f(Tt, Ct, St, Rt)
Yt
Tt
Ct
St
Rt

=
=
=
=
=

actual value of the data at time t

trend component at t
cyclical component at t
seasonall componentt att t
random component at t

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-43

Forecasting Techniques
Time series components: seasonality
need to identify and remove seasonal factors,
using
i
moving
i
averages to
t isolate
i l t those
th
factors
f t
remove seasonality by dividing data by seasonal
factor

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-44

Forecasting Techniques
Time series components: trend
to remove trend line, use least squares method
possible best-fit line styles:
straight line:
exponential line:
quadratic line:

Y = a + b(t)
Y = abt
Y = a + b(t) + c(t)2

choose one with best R2

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-45

Forecasting Techniques
Time series components: cyclical, random
isolate cyclical components by smoothing with a
moving
i
average
random factors cannot be predicted and should
be ignored

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-46

Forecasting Techniques
Smoothing techniques
Moving average
The larger the number of observations in the average,
the greater the smoothing effect.

Exponential
p
smoothing
g
Allows for the decreasing importance of information in
the more distant past.

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-47

Forecasting Techniques
Moving average: average of actual past results used
to forecast one period ahead
Et+1
= ((Xt + Xtt-11 + + Xtt-N+1
)/N
N+1)
Et+1
= forecast for next period
= actual values at their respective times
Xt, Xt-1
N
= number of observations included in average

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-48

Forecasting Techniques
Exponential smoothing: allows for decreasing
importance of information in the more distant past,
past
through geometric progression
Et+1 = wXt + (1-w) Et
w

= weight assigned to an actual

observation at period t

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-49

Forecasting Techniques
Econometric models: causal or
explanatory
l
models
d l off fforecasting
regression analysis
multiple
lti l equation
ti
systems
t
endogenous variables: dependent variables that may
influence other dependent variables
exogenous variables: from outside the system, truly
independent variables

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-50

Forecasting Techniques
Illustration of a model of the relationship between
the domestic currency and a foreign currency:
Et = a + bIt + cRt + dGt
where E = Exchange rate of a foreign currency in terms of the
domestic currency
I = Domestic inflation rate minus foreign inflation rate
R = Domestic nominal interest rate minus foreign nominal
interest rate
G = Domestic g
growth rate of GDP minus the g
growth rate of
foreign GDP
t = Time period
a, b, c, d Regression coefficients

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-51

Summary
Regression analysis is a primary tool used by
businesses to understand demand.
demand
Reliable input data and proper estimation and
evaluation are needed.
Forecasting is an important activity in many
organizations. In business, forecasting is a
necessity.
it
This chapter summarized and discussed six of the
ajo forecasting
o ecast g techniques
tec
ques used by businesses.
bus esses
major

Copyright 2014 Pearson Education, Inc. All rights reserved.

5-52