Chapter 5

Demand Estimation
and Forecasting

Chapter Five Copyright 2009 Pearson Education, Inc. 1

Publishing as Prentice Hall.
Overview
Regression analysis
Hazards with use of regression
analysis
Subjects of forecasts
Prerequisites of a good forecast
Forecasting techniques

Chapter Five Copyright 2009 Pearson Education, Inc. 2

Publishing as Prentice Hall.
Learning objectives

understand importance of forecasting in

business

describe six different forecasting

techniques

know how to specify and interpret a

regression

Chapter Five Copyright 2009 Pearson Education, Inc. 3

Publishing as Prentice Hall.
Learning objectives

 recognize limitations of consumer data

 use seasonal and smoothing methods

Chapter Five Copyright 2009 Pearson Education, Inc. 4

Publishing as Prentice Hall.
Data collection
 Data for studies pertaining to countries,
regions, or industries are readily available

Data for analysis of specific product

categories may be more difficult to obtain
 buy from data providers (e.g.
ACNielsen, IRI)
 perform a consumer survey
 focus groups
 technology: point-of-sale, bar codes
Chapter Five Copyright 2009 Pearson Education, Inc. 5
Publishing as Prentice Hall.
Regression analysis
 Regression analysis: a procedure
commonly used by economists to estimate
consumer demand with available data

Two types of regression:

 cross-sectional: analyze several
variables for a single period of time
 time series data: analyze a single
variable over multiple periods of time
Chapter Five Copyright 2009 Pearson Education, Inc. 6
Publishing as Prentice Hall.
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)

Chapter Five Copyright 2009 Pearson Education, Inc. 7

Publishing as Prentice Hall.
Regression analysis
 Interpreting the regression results:

coefficients:
 negative coefficient shows that as the
independent variable (Xn) changes, the variable
(Y) changes in the opposite direction
 positive coefficient shows that as the
independent variable (Xn) changes, the
dependent variable (Y) changes in the same
direction
 magnitude of regression coefficients is a
measure of elasticity of each variable
Chapter Five Copyright 2009 Pearson Education, Inc. 8
Publishing as Prentice Hall.
Regression analysis
 Statistical evaluation of regression results:

 t-test: test of statistical significance of

each estimated coefficient
b̂
t
SE b̂

b = estimated coefficient
SEb = standard error of estimated coefficient

Chapter Five Copyright 2009 Pearson Education, Inc. 9

Publishing as Prentice Hall.
Regression analysis
 Statistical evaluation of regression results:

 ‘rule of 2’: if absolute value of t is

greater than 2, estimated coefficient is
significant at the 5% level

 if coefficient passes t-test, the

variable has a true impact on demand

Chapter Five Copyright 2009 Pearson Education, Inc. 10

Publishing as Prentice Hall.
Regression analysis
 Statistical evaluation of regression results

 R2 (coefficient of determination):
percentage of variation in the variable
(Y) accounted for by variation in all
explanatory variables (Xn)
 R2 value ranges from 0.0 to 1.0
 the closer to 1.0, the greater the
explanatory power of the regression
Chapter Five Copyright 2009 Pearson Education, Inc. 11
Publishing as Prentice Hall.
Regression analysis
 Statistical evaluation of regression results

 F-test: measures statistical significance

of the entire regression as a whole (not
each coefficient)

Chapter Five Copyright 2009 Pearson Education, Inc. 12

Publishing as Prentice Hall.
Regression results

 Steps for analyzing regression results

 check coefficient signs and magnitudes

 compute implied elasticities

 determine statistical significance

Chapter Five Copyright 2009 Pearson Education, Inc. 13

Publishing as Prentice Hall.
Regression results

 Example: estimating demand for pizza

 demand for pizza affected by

1. price of pizza
2. price of complement (soda)
 managers can expect price decreases to
lead to lower revenue
 tuition and location are not significant

Chapter Five Copyright 2009 Pearson Education, Inc. 14

Publishing as Prentice Hall.
Regression problems
 Identification problem: the estimation
of demand may produce biased results
due to simultaneous shifting of supply and
demand curves

 solution: use advanced correction

techniques, such as two-stage least
squares and indirect least squares

Chapter Five Copyright 2009 Pearson Education, Inc. 15

Publishing as Prentice Hall.
Regression problems
 Multicollinearity problem: two or more
independent variables are highly
correlated, thus it is difficult to separate
the effect each has on the dependent
variable

 solution: a standard remedy is to drop

one of the closely related independent
variables from the regression

Chapter Five Copyright 2009 Pearson Education, Inc. 16

Publishing as Prentice Hall.
Forecasting
 Examples: common subjects of business
forecasts:
• gross domestic product (GDP)
• components of GDP
eg consumption expenditure, producer
durable equipment expenditure,
residential construction
• industry forecasts
eg sales of products across an industry
• sales of a specific product
Chapter Five Copyright 2009 Pearson Education, Inc. 17
Publishing as Prentice Hall.
Forecasting

 A good forecast should:

 be consistent with other parts of the business

 be based on knowledge of the relevant past
 consider the economic and political
environment as well as changes
 be timely

Chapter Five Copyright 2009 Pearson Education, Inc. 18

Publishing as Prentice Hall.
Forecasting techniques

 Factors in choosing the right forecasting

technique:

 item to be forecast
 interaction of the situation with the forecasting
methodology
 amount of historical data available
 time allowed to prepare forecast

Chapter Five Copyright 2009 Pearson Education, Inc. 19

Publishing as Prentice Hall.
Forecasting techniques

 Approaches to forecasting

 qualitative forecasting is based on

judgments expressed by individuals or
group

 quantitative forecasting utilizes

significant amounts of data and
equations

Chapter Five Copyright 2009 Pearson Education, Inc. 20

Publishing as Prentice Hall.
Forecasting techniques

 Approaches to forecasting

 naïve forecasting projects past data

without explaining future trends

 causal (or explanatory) forecasting

attempts to explain the functional
relationships between the dependent
variable and the independent variables

Chapter Five Copyright 2009 Pearson Education, Inc. 21

Publishing as Prentice Hall.
Forecasting techniques
 Six forecasting techniques

 expert opinion
 opinion polls and market research
 surveys of spending plans
 economic indicators
 projections
 econometric models

Chapter Five Copyright 2009 Pearson Education, Inc. 22

Publishing as Prentice Hall.
 Forecasting techniques
 Market research: is closely related to
opinion polling and will indicate not only
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
Chapter Five Copyright 2009 Pearson Education, Inc. 23
Publishing as Prentice Hall.
Forecasting techniques
 Surveys of spending plans: yields
information about ‘macro-type’ data
relating to the economy, especially:

 consumer intentions
Examples: Survey of Consumers (University of
Michigan); Consumer Confidence Survey
(Conference Board)

 inventories and sales expectations

Chapter Five Copyright 2009 Pearson Education, Inc. 24
Publishing as Prentice Hall.
Forecasting techniques
 Economic indicators: a barometric
method of forecasting designed to alert
business to changes in conditions

 leading, coincident, and lagging indicators

 composite index: one indicator alone may

not be very reliable, but a mix of leading
indicators may be effective

Chapter Five Copyright 2009 Pearson Education, Inc. 25

Publishing as Prentice Hall.
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 performance, slower deliveries
diffusion index
 manufacturers’ new orders, nondefense
capital goods
Chapter Five Copyright 2009 Pearson Education, Inc. 26
Publishing as Prentice Hall.
Forecasting techniques
 Leading indicators predict future economic
activity

 building permits, new private housing

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

Chapter Five Copyright 2009 Pearson Education, Inc. 27

Publishing as Prentice Hall.
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

Chapter Five Copyright 2009 Pearson Education, Inc. 28

Publishing as Prentice Hall.
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 (%)

Chapter Five Copyright 2009 Pearson Education, Inc. 29

Publishing as Prentice Hall.
Forecasting techniques
 Economic indicators: drawbacks

 leading indicator index has forecast a

recession when none ensued
 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
Chapter Five Copyright 2009 Pearson Education, Inc. 30
Publishing as Prentice Hall.
Forecasting techniques

 Trend projections: a form of naïve

forecasting that projects trends from past
data without taking into consideration
reasons for the change

 compound growth rate

 visual time series projections
 least squares time series projection

Chapter Five Copyright 2009 Pearson Education, Inc. 31

Publishing as Prentice Hall.
Forecasting techniques
 Compound growth rate: forecasting by
projecting the average growth rate of the
past into the future

 provides a relatively simple and timely

forecast

 appropriate when the variable to be

predicted increases at a constant %
Chapter Five Copyright 2009 Pearson Education, Inc. 32
Publishing as Prentice Hall.
Forecasting techniques
 General compound growth rate formula:

E = B(1+i)n

E = final value
n = years in the series
B = beginning value
i = constant growth rate

Chapter Five Copyright 2009 Pearson Education, Inc. 33

Publishing as Prentice Hall.
Forecasting techniques

 Visual time series projections: plotting

observations on a graph and viewing the
shape of the data and any trends

Chapter Five Copyright 2009 Pearson Education, Inc. 34

Publishing as Prentice Hall.
Forecasting techniques

 Time series analysis: a naïve method of

forecasting from past data by using least
squares statistical methods to identify
trends, cycles, seasonality and irregular
movements

Chapter Five Copyright 2009 Pearson Education, Inc. 35

Publishing as Prentice Hall.
Forecasting techniques
 Time series analysis:

Advantages:
 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
Chapter Five Copyright 2009 Pearson Education, Inc. 36
Publishing as Prentice Hall.
Forecasting techniques
 Time series data can be represented as:

Yt = f(Tt, Ct, St, Rt)

Yt = actual value of the data at time t

Tt = trend component at t
Ct = cyclical component at t
St = seasonal component at t
Rt = random component at t
Chapter Five Copyright 2009 Pearson Education, Inc. 37
Publishing as Prentice Hall.
Forecasting techniques
 Econometric models: causal or
explanatory models of forecasting
 regression analysis
 multiple equation systems
 endogenous variables: dependent
variables that may influence other
dependent variables
 exogenous variables: from outside the
system, truly independent variables

Chapter Five Copyright 2009 Pearson Education, Inc. 38

Publishing as Prentice Hall.
Forecasting techniques
 Example: econometric model
 Suits (1958) forecast demand for new
automobiles
∆R = a0 + a1 ∆Y + a2 ∆P/M + a3 ∆S + a4 ∆X
R = retail sales
Y = real disposable income
P = real retail price of cars
M = average credit terms
S = existing stock
X= dummy variable
Chapter Five Copyright 2009 Pearson Education, Inc. 39
Publishing as Prentice Hall.
Global application

 Example: forecasting exchange rates

 GDP
 interest rates
 inflation rates
 balance of payments

Chapter Five Copyright 2009 Pearson Education, Inc. 40

Publishing as Prentice Hall.