FORECASTING

TECHNIQUES
Production and Operations in Management
 HIGHLIGHTS

Introduction of forecasting Steps in the Forecasting

Process

Forecasting Operations
Types of Forecasting Methods
 WHAT IS FORECASTING/FORECASTS?

• Forecasting is the art and science of predicting future events

• Forecasts are basic input in the decision processes of operations
management since they provide information on future demand
IMPORTANCE OF FORECASTING TO:
 PLANNING AND CONTROL

• Forecasts
a. Play an important role in the planning process - enable
managers to anticipate the future
b. Are input to all types of business planning and control

• Forecasting-deals with what we think WILL happen in the future

Planning- deals with what we think SHOULD happen in the future
 MARKETING

• Marketing uses forecasts to plan product promotion and pricing

 FINANCE

• It uses forecasting as an input to financial planning.

 OPERATIONS FUNCTION

• It serves as an input for decisions on process design, capacity

planning and inventory.
PRIMARY GOAL OF FORECASTING:
 FORECASTING OPERATIONS

A.Characterizing Demand
• Demand forecasting- refers to the prediction of what will happen
to your company’s existing product sales.

Factors that affect the demand:

1. Rapid change in consumer preference
2. Events affecting the geographical region (like earthquakes or
other natural disasters, major sports game)
3. Income of consumers
 FORECASTING OPERATIONS

B. Purpose of Forecasting Demand

•Helps a firm to arrange for the supplies of necessary inputs
without any wastage of materials and time (optimization)
•Helps a firm to diversify its output to stabilize its income
overtime
 ACCORDING TO THE PURPOSE OF
SHORT TERM FORECASTING

• It can be undertaken by affirm for the following purpose:

*Appropriate scheduling of production to avoid problems of
over production and under-production
*Proper management of inventories
*Evolving suitable price strategy to maintain consistent sales
*Formulating a suitable sale strategy in accordance with the
changing pattern .of demand and extent of competition
among the firms
*Forecasting financial requirements for the short period
 ACCORDING TO THE PURPOSE OF LONG
TERM FORECASTING

• Planning for a new project, expansion and modernization of an

existing unit, diversification and technological up gradation
• Assessing long term financial needs. It takes time to raise financial
resources
• Arranging suitable manpower. It can help a firm to arrange for a
specialized labor force and personnel
• Evolving a suitable strategy for changing pattern of consumption
 FORECASTING TECHNIQUES

Types of Forecasting/
Forecasts

Demand Technological Economic

forecasts forecasts forecasts
 ECONOMIC FORECASTS

• These predict a variety of economic indicators, like money supply,

inflation rates, interest rates, etc.
 TECHNOLOGICAL FORECASTS

• These predict rates of technological progress and innovation.

 DEMAND FORECASTS

• These predict the future demand for a company’s products or

services.
 EXTRAPOLATION TECHNIQUES OF
DEMAND FORECASTING

Time Series Analysis

Qualitative Forecasting Causal

Simulation
Forecasting
 TIME SERIES ANALYSIS

• Based on the assumption that the item forecasted follows a

similar pattern over time.
 QUALITATIVE FORECASTING

• Consists of gathering opinions from a variety of people, then

applying their own judgment

• Best used when there is insufficient historical data

 CAUSAL

• Refers to the application of leading indicators to create forecast

Example: Mortgage rates affect the purchase of new homes

 SIMULATION FORECASTING

• Combines the causal and the time series methods; often used in
“what-ifs” scenarios”
 STEPS IN THE FORECASTING
TECHNIQUES

Determine the Establish a time

purpose of the horizon that the Select a forecasting
forecast and when forecast must technique.
it will be needed cover.

Gather and analyze

the appropriate Monitor the
data & then prepare forecast.
the forecast.
 TYPES OF FORECASTING METHODS:

1. Qualitative Techniques -based on judgments, opinions,

intuition, emotions, or personal experiences and are
subjective in nature.

2. Quantitative Techniques - based on mathematical

(quantitative) models, and are objective in nature
 Types of Qualitative Techniques:

Opinions Subjective
Jury of The Bayesian Scenario
of the Consumer’s Approach Executive
Executive Delphi Decision Writing
Opinion
sales Expectations
Method Theory Method method
Opinions
person
 JURY OF EXECUTIVE OPINION

• Approach in which a group of managers meet and collectively

develop a forecast
 OPINIONS OF THE SALES
PERSON/SALES FORCE MEMBERS
• Approach in which each salesperson estimates future sales in his
or her region
 CONSUMER’S EXPECTATIONS

• Involves a survey of the customers (via questionnaires, researches

and other tools) as to their future needs

• The surveys are used to judge preferences of customer and to

assess demand.
 THE DELPHI METHOD

• Approach in which consensus agreement is reached among a

group of experts

• Developed by Rank Corporation in 1969 for forecasting military

events
 BAYESIAN DECISION THEORY

• Uses a network diagram and probability must be estimated for

each event over the network
 SCENARIO WRITING METHOD

• The forecaster generates several different future scenarios

(corresponding to different sets of assumptions).
 SUBJECTIVE APPROACH METHOD

• Allows individuals participating in the forecasting decision to

arrive at a forecast based on their feelings, ideas, and personal
experiences
 EXECUTIVE OPINIONS

• Usually involve a small group of upper- level managers

(marketing, operations and finance)

• Often used as a part of long-range planning and new product

development
 TYPES OF QUANTITATIVE
FORECASTING METHODS

1. Time Series Method - look at past patterns of data and

attempt to predict the future based upon the other
variables

2. Causal Method / Associative Models - relies on the use of

several variables and their “cause-and-effect” relationships
 Time Series Method

Simple Weighted Simple

Naïve Exponential Trend Line Seasonal
Moving Inventory Moving Linear
Forecast Smoothing Forecast Indexes
Average Average Regression
 NAÏVE FORECAST

 Uses last period’s actual value as a forecast; applied to a series

that exhibits seasonality or trend

Example: If the demand last week was 200 units, the naïve
forecast for the upcoming week is 200 units.
 SIMPLE MOVING AVERAGE

 Useful if we can assume that market demands will stay fairly

steady over time

 Formula: Moving Average = ∑Demand in previous n periods

n
*Where: n – is the number of periods in the moving
average
 SIMPLE MOVING AVERAGE
Example: Compute a three-period moving average forecast given the following
demand for cars for the last five periods.
Demand Supply
1 70
2 80
3 65
4 90
5 85

• Solution: The forecast for period 6 should be:

Moving Average Forecast = 65 + 90 + 85 = 80 cars
3
If the actual demand in period 6 turns out to be 95, the moving average
forecast for the period 7 would be:

Moving Average Forecast = 90 + 85 + 95 = 90 cars

3
 WEIGHTED MOVING AVERAGE

 Usesan average of a specified number of the most recent

observations, with each observation receiving a different emphasis
(weight) when there is a trend or pattern

 Formula: Weighted Moving Average = ∑[(Weight for period n) (demand in period n)]
∑Weights
 WEIGHTED MOVING AVERAGE

• Example: Compute a three- period weighted moving average

forecast given the following demand for cars the last five periods;
with an assigned weight of 1,2,3.

Demand Supply
1 70
2 80
3 65
4 90
5 85
 WEIGHTED MOVING AVERAGE
Solution: The forecast for period 6 would be:

WMA= 65(1)+90(2)+85(3) = 65+180+245 = 490/6 = 81.67 or 82 cars

(1+2+3) 6

If the actual demand in period 6 turns out to be 95, the weighted

Moving average forecast for period 7 would be:

WMA = 90(1)+85(2)+95(3) = 90+170+85 = 545/6 = 90.83 or 91cars

6 6
 EXPONENTIAL SMOOTHING
 Used to forecast sales when there is no trend in the demand for goods
or services

Formula: Ft = Ft-1 + ∝ [ At-1 – Ft-1]

*Where: Ft = new forecast or forecast for period

Ft-1 = previous forecast
∝ = smoothing constant; represents percentage of the
forecast error
At-1 = actual demand or sales for period t-1
 EXPONENTIAL SMOOTHING

• Example 1: A car dealer • Solution:

predicted a January demand
for 550 Honda V-tech cars. Ft = Ft-1 + ∝ [ At-1 – Ft-1]
Actual January demand was = 550 + 0.10 [680-550]
680 Honda V-tech cars and
∝ = 0.10. Forecast the = 550 + 0.10 [130]
demand for January, using = 550 + 13
the exponential smoothing Ft = 563
model.
 EXPONENTIAL SMOOTHING

• Example 2: Use exponential smoothing model to develop a series

of forecast for the following data and compute:
[Actual - Forecast] = Error for each period
Use a smoothing factor of .10
Use smoothing factor of .40
Plot the actual data and both sets of forecast on a
single graph.
 EXPONENTIAL SMOOTHING

Period Actual Demand

1 50
2 52
3 48
4 51
5 50
6 54
7 52
8 50
9 55
10 53
11
 EXPONENTIAL SMOOTHING

Solution:
a. ∝1 = 0.10 b. ∝2 = 0.40

PERIOD ACTUAL FORECAST Forecast FORECAST Forecast

DEMAND ERROR ERROR
1 50 - - - -
2 52 50 2 50 2
3 48 50.20 -2.2 50.80 -2.8
4 51 49.98 1.02 49.68 1.32
5 50 50.08 -0.80 50.21 -0.21
6 54 50.07 3.93 50.13 3.87
7 52 50.46 1.54 51.68 0.32
8 50 50.61 -0.61 51.81 -1.81
9 55 50.55 4.45 51.09 3.91
10 53 51 2 52.65 0.35
11 51.20 52.79
EXPONENTIAL SMOOTHING
 TREND LINE FORECAST

Yt = a + bt

*Where: t = specified number of time periods from t=0

Yt = forecast for period t
a = value of Yt at t=0
b = slope of the line

*The coefficient of line a and b can be computed using two

equations:
b = n∑ty - ∑t ∑y OR a = ∑y - b∑t
n∑t2 – (∑t)2 n

*Where n = number of periods; y = value of the time series

 TREND LINE FORECAST
• Example: The total sales of television sets of Manila-based firm
over the last 10 weeks is shown in the following table. Plot the
data, and visually check if a linear trend line would be appropriate.
Then determine the equation of the line and predict the sale for
weeks 11 and 12.
WEEK SALES
1 800
2 810
3 830
4 820
5 850
6 810
7 825
8 840
9 805
10 830
 TREND LINE FORECAST
• Solution:
a. The plot that a linear trend would be appropriate
 TREND LINE FORECAST
b.
Week (t) Unit Sales (y) ty t2

1 800 800 1
2 810 1,620 4
3 830 2,490 9
4 820 3,280 16
5 850 4,250 25
6 810 4,860 36
7 825 5,775 49
8 840 6,720 64
9 805 7,245 81
10 830 8,300 100
∑t = 55 ∑y = 8,220 ∑ty = 45,340 ∑t2 = 385
 TREND LINE FORECAST
b = n∑ty - ∑t∑y = 10(45,340) – 55(8,220)
n∑t2 – (∑t)2 10(385) – (55)2

= 453,400 – 452,100
3,850 – 3,025

= 1,300
825
b = 1.58

• a = ∑y - b∑t = 8,220 – 1.58 (55) = 8,220 – 86.9 = 8,133.10 = 813.31

n 10 10 10
 TREND LINE FORECAST
c.
When t = 11

Yt = a + bt
Y11= 813.31 + 1.58 (11)
= 813.31 + 17.38
Y11= 830.69

When t = 12

Y12 = 813.31 + 1.58(12)

 SIMPLE LINEAR REGRESSION

 The simplest and most widely used form of regression involves a linear
relationship between two variables.

• Formula: Yt= a + bX
*where:
Yt = Predicted (dependent) variable
X = Predictor (independent) variable
b = slope of the line
a = value of Yt when X=0
n = number of period observations
 SIMPLE LINEAR REGRESSION

 The coefficients a and b of the line are computed using these

two equations:

b = n(∑xy) – (∑x)(∑y) n(∑x2)-( ∑x)2

a = ∑y – b ∑x or a = y – bx
n
 SIMPLE LINEAR REGRESSION
Examples:
JR Hamburgers has a chain of 10 stores in Metro Manila. Sales
figures and profiles for the stores are giving in the following table.
Obtain a regression for the data, and predict profit for a store
assuming sales of 30 million.
Sales, x (Millions) Profits, y (Millions)
15 8
17 9
21 13
18 10
19 11
22 14
16 8.5
17 10
25 15
20 13
 SIMPLE LINEAR REGRESSION
Solutions:

Sales, x (Millions) Profits, y (Millions) xy x2

15 8 120 225
17 9 153 289
21 13 273 441
18 10 180 324
19 11 209 361
22 14 308 484
16 8.5 136 256
17 10 170 289
25 15 375 625
20 13 260 400
∑x=190 ∑y=111.5 ∑xy=2184 ∑x=3694
 SIMPLE LINEAR REGRESSION
b = 10(2184) – (190)(111.5)
10(3694) – (190)2
b = 21,840 – 21,185 = 655
36,940 – 36,100 840
when x = P30 million
b = 0.78
Yt = a + bX
a = ∑y – b ∑x Y30 = -3.67 + 0.78(30)
n Y30 = -3.67 + 23.4
111.5−0.78 (190) Y30 = 19.73 million
a=
10
−36.7
=
10

a = -3.67
 INVENTORY
(ECONOMIC ORDER QUANTITY)
 Ensures maintenance of an adequate inventory on hand at the
lowest total cost to the organization

2(annual inventory)(cost per code/order)

 Formula: EOQ = √
(percentage of carrying cost)(cost per unit)
 INVENTORY
(ECONOMIC ORDER QUANTITY)

Example: Suppose that R and C Beverage Company has a

beverage product that has a constant annual demand rate of
7,200 cases. A case of softdrink costs R and C P288. Ordering
cost is P200 per order and inventory carrying cost is charged
at 25% of the cost per unit. Solve for the EOQ
 INVENTORY
(ECONOMIC ORDER QUANTITY
Solution:

Given EOQ = √
2 7,2000 (200)
Annual inventory – 7,200 cases .25 (288)
Cost per case of softdrinks – P288 2,880,000
Cost per order – P200
= √
72
Percentage of carrying Cost – 25% = 40,000
EOQ = 200 cases
 SEASONAL INDEXES

Amechanism for adjusting the forecast to accommodate any

seasonal patterns inherent in the data.
 SEASONAL INDEXES
Example:
Col. 1 Col. 2 Col. 3 Col. 4 Col. 5 Col. 6
Year Q1 Q2 Q3 Q4 Annual
Demand
1 62 94 113 41 310
2 73 110 130 52 365
3 79 118 140 58 395
4 83 124 146 62 415
5 89 135 161 65 450
6 94 139 162 70 465
Avg. (62+73+ (94+110+ (113+130+ (41+52+
Demand 79+83+ 118+124+ 140+146+ 58+62+
Per Qtr. 89+94) 135+139) 161+162) 65+70)
÷ 6 = 80 ÷ 6 = 120 ÷ 6 = 142 ÷ 6 = 58
 SEASONAL INDEXES

• This would result in the following alternate seasonal index values:

Year Q1 Q2 Q3 Q4
Seasonal Index 80/100 = .80 120/100 = 142/100 = 58/100 = .58
1.20 1.42
 FOUR COMPONENTS OF TIME
SERIES MODELS:
1. Trend component – refers to the pattern of the demand (past, present
and future)

2. Cyclical Component – any recurring sequence of points lying above

or below the trend line that last for more than a year

3. Seasonal Component – it captures the variability in the data due to

seasonal fluctuations

4. Irregular Component – random variations in time series (caused by

short –term, unanticipated and nonrecurring factors that affect the
time series)
 ASSOCIATIVE FORECASTING
METHODS OR CAUSAL MODELS
 ASSOCIATIVE FORECASTING MODELS

Example: A distributor of drywall in a local community has

historical demand data for the past eight years as well as data on
the number of permits that have been issued for new home
construction. These data are displayed in the following table:
# of new home Demand for
Year construction 4’x8’ sheets of
permits drywall
2004 400 60,000
2005 320 46,000
2006 290 45,000
2007 360 54,000
2008 380 60,000
2009 320 48,000
2010 430 65,000
2011 420 62,000
 ASSOCIATIVE FORECASTING
METHODS
The independent variable (X) is the number of construction permits. The
dependent variable (Y) is the demand for drywall.

Application of regression formulas yields the following forecasting

model:

Y = 250 + 150X

If the company plans finds from public records that 350 construction
permits have been issued for the year 2012, then a reasonable estimate
of drywall demand for 2012 would be:

Y = 250 + 150(350) = 250 + 52,500 = 52,750

(which means next year’s forecasted demand is 52,750 sheets of drywall)
 ASSOCIATIVE FORECASTING
METHODS

Demand vs. Time

70000

65000

60000

55000

50000

45000

40000

35000

30000

• 2003 2004 2005 2006 2007 2008 2009 2010

 ASSOCIATIVE FORECASTING
METHODS
Demand vs . Construction Permits
70000

60000

50000

40000

30000

20000

10000

250 300 350 400 450

Construction Permits
THANK YOU FOR LISTENING