You are on page 1of 34

Demand Analysis and Forecasting

P C Padhan
pcpadhan@xlri.ac.in
Demand Analysis
⚫ The Demand Analysis is a process whereby the management
makes decisions with respect to the production, cost allocation,
advertising, inventory holding, pricing, etc.

⚫ How much a firm produces depends on its production capacity


but,

⚫ How much it must attempt to produce depends on the potential


demand for its product.

⚫ Low potential demand=>low production=> lower inventory and


vice versa.
Demand for a Commodity: Meaning

Demand is the willingness and capacity


Manufacturer of a customer to purchase the product
(s) or service(s) at each possible price
during a given period of time.

Price is a tool by which the market


coordinates individual desires.

Customer
Consumer
Demand for a Commodity: Meaning

The Quantity demand is the amount of a product that


people are willing and able to purchase at one specific
price.

The Law of demand – Inverse relationship between P


Price of a Commodity and Quantity Demanded (Q) of it,
assuming other thing held constant (ceteris paribus)
D
A rational consumer maximizes satisfaction by
reorganizing consumption until the marginal utility in
each good per price(rupees) is equal. Q

Two Reasons for Law of Demand


1. Income Effect
2. Substitution Effect
Features/Characteristics of Demand
The demand is
⚫ the specific quantity that a consumer is willing to purchase i.e.
expressed in numbeRs
⚫ is expressed per unit of time, per month, per week, per day.
⚫ is always at a price, e. any change in the price of a commodity
will bring about a certain change in its quantity demanded.
⚫ is always in a market, a place where a set of buyers and sellers
meet.

So demand is an important contributor to firm risk arises from


sudden shifts in demand for the product or service.
Demand Analysis: Objectives
⚫ Demand analysis serves four managerial objectives/purpose:

1. It provides the insights necessary for effective


management/manipulating of demand, and

2. it aids in forecasting sales and revenues.

3. Appraising salesmen’s performance for setting their sales


quotas, and

4. Watching the trend of the company’s competitive position.


How to Determine the Demand Function?

Based on the definition of demand:


a. Willingness to pay is determined by
⚫ Buyer’s tastes or needs

⚫ Types of goods/products:

▪ Necessity or luxurious
▪ Inferior or superior
▪ Durable or perishable
▪ Conspicuous or griffin etc.
⚫ Prices of the product

⚫ Substitutes goods

⚫ Complementary goods

⚫ Other factors

b. Capacity to pay is determined by


⚫ Income and wealth
Individual Consumer’s Demand Function

An equation representing the demand curve

QdX = f(PX, I, PY, T….)


QdX =quantity demanded of commodity X by an individual per time
period

PX =price per unit of commodity X

I =consumer’s income

PY =Price of related (substitute or complementary) commodity

T =tastes of the consumer


Individual Demand Table and Demand Curve

The demand curve is the graphic representation of the law of demand which slopes
down ward to the right

A Demand Table A Demand Curve


Rs 6.00

Price per DVDs (in Rs.)


Price per DVD rentals 5.00
cassette demanded per
week 4.00 E
3.50 G
A Rs 0.50 9 D
3.00
B 1.00 8
C 2.00 6 C Demand
2.00 for DVDs
D 3.00 4
E 4.00 2 F B
1.00
A
.50
0
1 2 3 4 5 6 7 8 9 10111213
Quantity of DVDs demanded (per week)
◼The demand table assumes other things remaining the same
Market Demand Function

QDX = f(PX, N, I, PY, T, D, Ms…..)


QDx=quantity demanded of commodity X
Px=price per unit of commodity X
N=number of consumers on the market
I=consumer income
PY=price of related (substitute or complementary)
commodity
T=consumer tastes
D= Demographic distribution
Ms= Market Saturation
From Individual Demand to a Market Demand Curve

Market Demand curve is the horizontal summation of individual demand curve

(1) (2) (3) (2) (3) Rs4.00


Price per Alice’s Bruce’s Cathy’s Market G
cassette demand demand demand demand 3.50

Price per cassette (in dollars)


3.00 F
A Rs 0.50 9 6 1 16 E
B 1.00 8 5 1 14 2.50
D
C 1.50 7 4 0 11 2.00
D 2.00 6 3 0 9 C
1.50
E 2.50 5 2 0 7 B
F 3.00 4 1 0 5 1.00
A
G 3.50 3 0 0 3 0.50
H 4.00 2 0 0 2 Cathy Bruce Alice Market demand
0
2 4 6 8 10 12 14 16
Quantity of cassettes demanded per week
A Sample Demand Curve

Price (per unit)

PA A

D
0
QA
Quantity demanded (per unit of time)
Assumption of Law of Demand: Other
Things Constant

⚫ Other things constant places a limitation on the


application of the law of demand.

⚫ All other factors that affect quantity demanded are


assumed to remain constant, whether they actually
remain constant or not.

⚫ These factors may include changing tastes, prices of


other goods, income, weather etc.
Changes in Demand Versus Changes in Quantity
Demand
A. Changes in Quantity Demanded
A movement along a demand curve is the graphical representation of the effect of a change in price on the
quantity demanded. In other words it refers to Change in Quantity Demanded - movement along the same
demand curve in response to a price change.

B
Rs2
Price (per unit)

Change in quantity demanded


(a movement along the curve)

A
Rs1

D1
0
100 200
Quantity demanded (per unit of time)
B. Change/Shift in Demand
•Change in Demand - shift in entire
demand curve in response to a
change in a determinant of demand
(a ceteris paribus variable)

Change in demand
(a shift of the curve)
Rs2
Price (per unit)

B A
Rs1

D0

D1
100 200 250
Quantity demanded (per unit of time)
Determinants and Shift Factors of Demand

Society’s
Income

Taxes or subsidies Prices of related goods


to consumers

Tastes

Number of buyers Expectations

•Shift factors of demand are factors that cause shifts in the demand curve:
Shift Factors of Demand: Math Notations

QdX = f(PX, I, PY, T)


1. Price(Px): QdX/PX < 0
2. Income(I): QdX/I > 0 if a good is normal, Ex. Rice, wheat, tooth paste etc.
QdX/I < 0 if a good is inferior, Ex. Corn, bread.
An increase in income will increase demand for normal goods.
An increase in income will decrease demand for inferior goods.
3. Price of Related Goods(Py):
QdX/PY > 0 if X and Y are substitutes Ex. BMW and Mercedes Benz
QdX/PY < 0 if X and Y are complements, Ex. Car and Petrol
When the price of a substitute good falls, demand falls for the good whose price has
not changed.
When the price of a complement good falls, demand rises for the good whose price
has not changed.

4. Taste(T): QdX/taste > 0 for Good taste/choice


QdX/taste < 0 for Bad taste /choice
Factors Affecting Demand for a commodity
(Internal vs External Factors)

❑ Product life-cycle o Income


management o Prices of Substitutes
o Prices of Complements
❑ Planned price changes o Expectations,
o Changing customer Tastes and
❑ Changes in the sales force preferences
o Random fluctuation
❑ Resource constraints o Seasonality
o Competition
❑ Marketing and sales o New customers
promotion o Plans of major customers
o Government policies
❑ Advertising o Regulatory concerns
o Economic conditions/cycles
❑ Product substitution
o Environmental issues
o Weather conditions
o Global and local trends
Forecasting ?

“Any Astronomer can Predict just where every Star will be at half past
eleven tonight; he can make no such prediction about his daughter”
James Truslow Adams

“The goal of forecasting is not to predict the future but to tell you what
you need to know to take meaningful action in the present”
Paulo Saffo, HBR
What is Forecasting ?

Forecasting is the art and science of predicting future


events .
A forecast is a statement or inference about the future, usually
based on historical information.
A forecast is an educated guess.

You’ll never get it right. But, you can always get it less wrong.

Forecasting consists of a Prediction consists of a


variety of processes for variety of processes for
identifying what possible identifying what possible
futures could happen futures may happen
Why Forecast?

To reduce avoidable uncertainty.

To stabilize production and employment over a period

To reduces the need for slack resources (inventory,


staffing) to meet uncertain demand
Who Forecasts?

⚫ Any one in the organisation.


⚫ Marketing
⚫ Finance,
⚫ Production and Operations
⚫ HR etc.

⚫ Mostly
⚫ Forecasting Team ( Consisting of all
professionals)
Major Areas of Forecasting?

Economic
Technology
Forecasting Demand
Forecasting
Forecasting
Predicts what
Predict the
the general
probability and /or
business
possible future
conditions will Predicts the
developments in
be in the future quantity and
technology
Ex. Inflation timing of
Ex. Competitive
rates, exchange demand for a
advantage or firm’s
rate, housing firm’s
competitors
starts, GDP, tax,
incorporate into products
level of
their products and
employment
process
etc.)
Major Areas of Forecasting?
Demand Forecasting

⚫ It is the scientific and analytical estimation of demand for a


product( service) for a particular period of time.

⚫ It is the process of determining how much of what products is


needed when and where.

▪ Sales is a function of demand.


▪ Cost of production also depends upon demand

⚫ Why Demand Forecasting?


➢ To plan the level of production and make arrangements for the
resource to be consumed.
How Forecasting?
Framework of a Forecast System?

Objecti
ve
Types / Kinds of Forecast?
External
Environment
Internal

Macro
Dimension
Micro
Kinds
Long
Time
Short

Qualitative
Levels of Forecast Methods
• at Firms Level
Quantitative
• at Industry Level
• at Total Market Level
Kinds of Forecast: Forecasting Methods
a. Qualitative
• Judgmental models
⚫ Jury of executive opinion
Predict how others will behave
⚫ Delphi method
⚫ Judgmental bootstrapping
⚫ Intentions Rely on intuitive judgments, opinions, and
⚫ Survey and Market Tests probabilities.
⚫ Indirect Methods
b. Quantitative (Uni-variate vs Multivariate Series)
⚫ Technological forecasting models Predict own behavior
⚫ Curve fitting
⚫ Analogous data
⚫ Time series extrapolation models Appropriate for very new technologies and
⚫ Simple Averages very long-range forecasting.
⚫ Moving averages
⚫ Exponential smoothing
⚫ Seasonal Decomposition models
⚫ Box-Jenkins
Based only on past data. Use patterns,
⚫ ARCH, GARCH etc.
changes, disturbances, etc. in the data to
⚫ Causal models
forecast the future.
⚫ Regression analysis, Logit, Probit
⚫ VAR, ECM
⚫ Input-Output Models
Based on relationship between predictable
⚫ Barometric Techniques
factors and outcomes.
⚫ Leading, Lagging and Coincident Economic Indicators
⚫ Diffusion and Composite Indexes
Approach to Demand Forecasting
1. Bottom Up Approach: Here forecaster first prepares forecast for each and
every SKU in each and every region( for each and every customer) and then
aggregate them to arrive at category and aggregate level forecasts.
2. Top Down Approach: Here the forecast first prepares the overall forecasts and
then disaggregates it into regions, categories and SKUs.
Approach to Demand Forecasting
Top-Down vs. Bottom Up Approach:.
Parameters for Selecting Forecasting Tools

⚫ Purpose of Forecast ( Ex. Forecasting Horizon)


⚫ Timing
⚫ Cost
⚫ Data
⚫ Patterns in the Data Series
⚫ Forecasting flexibility
(amenability of the model to revision; quite often)

⚫ Accuracy
# Patterns in Data Series and Forecasting Tools

Pattern in Forecasting Tool


Series-
Stationary Naïve Methods; Simple Averaging Models
Series Moving Averages; Exponential Smoothing;
ARIMA
Trending Moving Averages; Holt’s Linear Exponential
Series Smoothing
Simple Regression; ARIMA
Seasonal Winter’s Exponential Smoothing; Classical
Series Decomposition; ARIMA; Census X12 ARIMA
Time Series Multiple Regression
Cyclical Classical Decomposition; ARIMA
Series Economic Indicators; Multiple Regression
Econometric Models
# Statistics for Accuracy of Forecast:
The Forecast Error
n
1
1.Mean Absolute Deviation (MAD)
n
 |A − F |
t =1
Average deviation

n
1
2. Mean Squared Error (MSE) :
n
(A − F)
t =1
2
Penalises large error

• Low MPE model is


 A− F 
3. Mean Percent Error (MPE):   A  unbiased or else.

100 • Large –ve MPE model is


N over estimating.

A− F • Large +ve MPE model is


4. Mean Absolute P Error (MAPE)  A
underestimating

100
N Comparing models

A− F
5. Weighted Mean Absolute
Percent Error ( WMAPE)
 A
100  A

A
Different Forecasting Error Measurement
Thank You All

You might also like