CHAPTER 3

FORECASTING
Course Contents
Meaning and use of forecasting
Forecasting Techniques

04/11/2024 1
Introduction :
Forecasting is the art and science of predicting future events.

Forecasts are a basic input in the decision processes of operations

management because they provide information on future demand.
The purpose of forecasting is to use the best available information to guide
to future activities toward organizational goals. using the demand
estimates, input resources like raw materials, manpower, capital and
machines can be planned.
Two aspects of forecasts are important. One is the expected level
of demand; the other is the degree of accuracy that can be
assigned to a forecast (i.e., the potential size of forecast error).
 The expected level of demand can be a function of some
structural variation, such as a trend or seasonal variation.

04/11/2024 2
Forecast accuracy is a function of the ability of
forecasters to correctly model demand, random variation,
and sometimes unforeseen events.
Forecasting is the basis for budgeting, planning capacity,
sales, production and inventory, personnel, purchasing,
and more.

 What is difference between forecasting and planning?

 Forecasting: what we think will happen
 Planning: what we think should happen

Forecasting is just about: future demand & supply

variation

04/11/2024 3
 Forecasting Time Horizons
Quantitative
1. Short-range forecast methods
• Generally less than 3 months Detailed
use of
• Purchasing, job scheduling,
system
workforce levels, job assignments, production
levels
2. Medium-range forecast
• 3 months to 2 years
• Sales and production planning, budgeting
3. Long-range forecast Design
• 2+ years of system
• New product planning, facility location, Qualitative
research and development Methods

04/11/2024 4
 Forecasting Based on Time Series Data :
A time series is a time-ordered sequence of observations taken at
regular intervals (e.g., hourly, daily, weekly, monthly, quarterly,
annually). The data may be measurements of demand, earnings,
profits, shipments, accidents, output, precipitation, productivity,
or the consumer price index.
Forecasting techniques based on time-series data are made on the
assumption that future values of the series can be estimated from
past values.
The underlying behavior of the series can often be accomplished
by plotting the data. patterns like: trends, seasonal variations,
cycles, or variations around an average might appear. Random
and perhaps irregular variation behaviors can be described as
follows:

04/11/2024 5
1. Trend:- refers to a long-term upward or
downward movement in the data. Population
shifts, changing incomes, and cultural changes
often account for such movements.

2. Seasonality refers to short-term, fairly regular

variations generally related to factors such as the
calendar or time of day. Restaurants, supermarkets,
and theaters experience weekly and even daily
“seasonal” variations.

04/11/2024 6
3. Cycles are wavelike variations of more than one year’s
duration. These are often related to a variety of economic,
political, and even agricultural conditions.
4. Irregular variations are due to unusual
circumstances such as severe weather conditions, strikes,
or a major change in a product or service. They do not
reflect typical behavior, and their inclusion in the series
can distort the overall picture. Whenever possible, these
should be identified and removed from the data.
5. Random variations are residual variations that remain
after all other behaviors have been accounted for.

04/11/2024 7
 Components of Demand

Trend
component
Demand for product or service

Seasonal peaks

Actual demand
line

Average demand
over 4 years

Random variation
| | | |
1 2 3 4
Time (years)

04/11/2024 8
Steps of Forecasting
1. Decide what needs to be forecasted.
 Level of detail, units of analysis & time horizon
required.
2. Evaluate and analyze appropriate data.
 Identify needed data & whether it’s available.
3. Select and test the forecasting model.
 Cost, ease of use & accuracy.
4. Generate the forecast.
5. Monitor forecast accuracy over time.

04/11/2024 9
Forecasting Techniques
Forecasting depends on having enough historical data
to be able to describe the record in statistical terms.
There are basically two broad categories of forecasting
techniques.
1)Qualitative methods
2)Quantitative methods

04/11/2024 10
 Qualitative methods
 Qualitative methods of forecasting techniques are based
upon judgment, especially when sufficient information and
data is not available.
Qualitative Forecasts: Is done when:-
 If management must have a forecast quickly, there may not be
enough time to gather and analyze quantitative data.
 Political and economic conditions are changing, available data
may be obsolete and more up-to-date information might not yet
be available.
 The introduction of new products and the redesign of existing
products or packaging suffer from the absence of historical
data that would be useful in forecasting
04/11/2024 11
Quantitative Methods
 These techniques use statistical analysis and other

mathematical models to predict future events, primarily

based upon past activities.

04/11/2024 12
 Overview of Quantitative Approaches

1. Naive approach
2. Moving averages
3. Exponential smoothing
4. Trend projection
5. Linear regression

04/11/2024 13
 1. Naive Approach

• Demand in next period is the same as demand

in most recent period
 May sales = 48 → June forecast = 48

• Usually not good

04/11/2024 14
 2. Moving Average Method
• Random variations demand can be smoothed
out by moving averages which preserve the
general pattern of the data.
• A moving average is the average of values
centered on the period in question.

∑ demand in previous n periods

Moving average =
n

04/11/2024 15
 2 a. Simple Moving Average
Assumes an average is a good estimator of future
behavior
Used if little or no trend
Used for smoothing
A 𝐷
DDtt t++A𝐷 +++ADD
tt-1t−1 ++...+
t - 2t t−2 ... + ADDtt-t−n−nn1+1+1
+...+
FFFt t+11+1== −1 −2
nnn

Ft+1 = Forecast for the upcoming period, t+1

n = Number of moving averaged
Dt = Actual Demand in period t

04/11/2024 16
 DDt ++ 𝐷𝐷t −1 ++ DDt −2 +...+
+...+DDt − n +1
FFt +1 == t t −1 t −2 t − n +1
Example Simple Moving Average t +1 nn

You’re manager in Amazon’s electronics

department. You want to forecast ipod sales for
months 4-6 using a 3-period moving average.

Sales
Month (000)
1 4
2 6
3 5
4 ?
5 ?
6
04/11/2024 ? 17
 AD + A 𝐷 +++ADD t -t2−2++...+ D 1
... + A
FFtt+11 == Dtt t + 𝐷tt-1−1
t −1 t −2 +...+ Dt -−tn−nn+1
+1
t +1 nnn
2a. Simple Moving Average
You’re manager in Amazon’s electronics
department. You want to forecast iPod sales for
months 4-6 using a 3-period moving average.
Sales Moving Average
Month (000) (n=3)
1 4 NA
2 6 NA
3 5 NA
4 ? (4+6+5)/3=5
5 ?
04/11/20246 ? 18
2a. Simple Moving Average

Sales Moving Average

Month (000) (n=3)
1 4 NA
2 6 NA
3 5 NA
4 5 5
5 ?
04/11/2024
6 ? 19
2a. Simple Moving Average

Forecast for Month 5?

Sales Moving Average

Month (000) (n=3)
1 4 NA
2 6 NA
3 5 NA
4 5 5
5 ? (6+5+5)/3=5.33
04/11/2024
6 ? 20
2a. Simple Moving Average

Actual Demand for Month 5 = 7

Sales Moving Average

Month (000) (n=3)
1 4 NA
2 6 NA
3 5 NA
4 5 5
5 5.33 5.33
04/11/2024
6 ? 21
2a. Simple Moving Average

Forecast for Month 6?

Sales Moving Average

Month (000) (n=3)
1 4 NA
2 6 NA
3 5 NA
4 5 5
5 5.33 5.33
04/11/2024
6 ? (5+5+5.33)/3=5 22
2a. Simple Moving Average

Forecast for Month 6?

Sales Moving Average

Month (000) (n=3)
1 4 NA
2 6 NA
3 5 NA
4 5 5
5 5.33 5.33
04/11/2024
6 ? (5+5+5.33)/3=5 23
2b. Weighted Moving Average
Gives more emphasis to recent data

FFtt+1+1== ww11DDtt++ ww22DDtt−1−1++ww33DDtt−2−2+...+w

+...+wnnDDtt−n+1
−n+1

Weights
Decrease for older data
Sum to 1.0
Simple
Simplemoving
moving
average
averagemodels
models
weight
weightall
allprevious
previous
periods
periodsequally
equally

04/11/2024 24
2b. Weighted Moving Average:
, 3/6

Month Sales Weighted

.
(000) Moving
Average
1 4 NA
2 6 NA
3 5 NA
4 ? 31/6 = 5.167
5 ?
6 ?

04/11/2024 25
 FFt +1 =
t +1 = w
w 1 D
D
1 tt +
+ w
w 2 D
D t −1
2 t −1
+
+ w
w 3 D
D t −2
3 t −2
+...+w
+...+w n D
n Dt t−n+1
−n+1
2b. Weighted Moving Average:
, 3/6

Month Sales Weighted

(000) Moving
Average
1 4 NA
2 6 NA
3 5 NA
4 3 31/6 = 5.167
5 7 25/6 = 4.167
6 32/6 = 5.333

04/11/2024 26
 3. Exponential Smoothing
• Form of weighted moving average
Weights decline exponentially
•
• Most recent data weighted most
• Requires smoothing constant ()
• Ranges from 0 to 1
• Subjectively chosen
• Involves little record keeping of past data
How to choose α
 depends on the emphasis you want to place on
the most recent data
Increasing α makes forecast more sensitive to recent data

04/11/2024 27
 Exponential Smoothing
New forecast = Last period’s forecast
+ a (Last period’s actual
demand – Last period’s
forecast)

Ft + 1 = Ft + a(Dt - Ft )
where Ft = previous forecast
Ft +1 = new forecast
a = smoothing (or weighting)
constant (0 ≤ a ≤ 1)

04/11/2024 28
3 a. Exponential Smoothing

Assumes the most recent observations have

the highest predictive value
gives more weight to recent time periods

FFt+1
t+1
=
= F
F tt
+
+ a(D
a(D tt
-
- F
Ft)
t)
et

Ft+1 = Forecast value for time t+1 Need

Needinitial
initial
Dt = Actual value at time t forecast
forecastFFt t
totostart.
 = Smoothing constant start.

04/11/2024 29
3 a. Exponential Smoothing – Example 1

FFt+1
t+1
=
= F
F tt
+
+ a(D
a(D tt
-
- F
Ft)
t)
i Di
Week Demand
1 820 Given
Giventhe
theweekly
weeklydemand
demand
2 775 data
datawhat
whatare
arethe
theexponential
exponential
3 680 smoothing
smoothingforecasts
forecastsfor
for
4 655 periods
periods2-10 usinga=0.10?
2-10using a=0.10?
5 750
6 802 Assume
AssumeFF1=D
1=D1 1
7 798
8 689
9 775
10
04/11/2024 30
3 a. Exponential Smoothing – Example 1

FFt+1
t+1
=
= F
F tt
+
+ a(D
a(D tt
-
- F
Ft)
t)
i Ai Fi
Week Demand a = 0.1 0.6
1 820 820.00 820.00
2 775 820.00 820.00
3 680 = F1+ a(D1–F
F2815.50 1)
793.00 =820+.1(820–820)
4 655 801.95 725.20 =820
5 750 787.26 683.08
6 802 783.53 723.23
7 798 785.38 770.49
8 689 786.64 787.00
9 775 776.88 728.20
10 776.69 756.28
04/11/2024 31
3 a. Exponential Smoothing – Example 1

FFt+1
t+1
=
= F
F tt
+
+ a(D
a(D tt
-
- F
Ft)
t)
i Di Fi
Week Demand a = 0.1 0.6
1 820 820.00 820.00
2 775 820.00 820.00
3 680 815.50 793.00
F3 = F2+ a(D2–F2) =820+.1(775–820)
4 655 801.95 725.20
5 750 787.26 683.08 =815.5
6 802 783.53 723.23
7 798 785.38 770.49
8 689 786.64 787.00
9 775 776.88 728.20
10 776.69 756.28
04/11/2024 32
3 a. Exponential Smoothing – Example 1

FFt+1
t+1
=
= F
F tt
+
+ a(D
a(D tt
-
- F
Ft)
t)
i Di Fi
Week Demand a = 0.1 0.6
1 820 820.00 820.00
2 775 820.00 820.00
3 680 815.50 793.00
4 655 801.95 725.20
5 750 787.26 683.08
6 802 783.53 723.23 This process
7 798 785.38 770.49 continues
8 689 786.64 787.00through week 10
9 775 776.88 728.20
10 776.69 756.28
04/11/2024 33
3 a. Exponential Smoothing – Example 1

FFt+1
t+1
=
= F
F tt
+
+ a(D
a(D tt
-
- F
Ft)
t)
i Di Fi
Week Demand a = 0.1 a = 0.6
1 820 820.00 820.00
2 775 820.00 820.00
3 680 815.50 793.00
4 655 801.95 725.20
5 750 787.26 683.08 What if the
6 802 783.53 723.23 a constant
7 798 785.38 770.49 equals 0.6
8 689 786.64 787.00
9 775 776.88 728.20
10 776.69 756.28
04/11/2024 34
 Class work
 Forecasts based on averages given the following data:
Weeks 1 2 3 4 5 6 7 8

Demands 60 65 55 58 64 70 68 67

Prepare a forecast for period 8 and 9 using each of these

approaches:
1. The appropriate naive approach.

2. A three-period moving average.

3. A weighted average using weights of 0.50, 0.30, and 0.20.

4. Exponential smoothing with a smoothing constant of 0.40.

04/11/2024 35
Measuring Forecast Error
 Forecasts are never perfect
 Need to know how much we should rely on our
chosen forecasting method.
 Measuring forecast error:

E t  A t  Ft
 Note that:
 Over-forecasts = negative errors
 Under-forecasts = positive errors.
 Large values of negative or positive errors shows
there is bias in the forecast.

04/11/2024 36
 Measures of Forecast Error
n


t =1
A t - Ft
MAD =
 Mean Absolute Deviation (MAD) n
 measures the total error in a forecast without regard to sign

 Cumulative Forecast Error (CFE) CFE   actual  forecast 

 Also called running sum of forecast error (RSFE)
 Measures any bias in the forecast n
RSFE  (At  Ft )
i 1

 Mean Square Error (MSE) n

 Penalizes larger errors 
 t t
A - F 2

t =1
MSE =
n
RMSE = MSE

 Ideal values =0 (i.e., no forecasting error)

37
Measuring Accuracy: Tracking signal

 The tracking signal is a measure of how often our

estimations have been above or below the actual
value. It is used to decide when to re-evaluate using a
model.
n
RSFE
RSFE  (At  Ft ) TS 
i1 MAD
 Positive tracking signal: most of the time actual values are
above our forecasted values
 Negative tracking signal: most of the time actual values are
below our forecasted values

04/11/2024 3
Examples
 A company is comparing the accuracy of two forecasting methods.

Forecasts using both methods are shown below along with the actual values
for January through May. The company also uses a tracking signal with ±4
limits to decide when a forecast should be reviewed. Which forecasting
method is best?

04/11/2024 39