Summer Class | Management Science

FORECASTTING TECHNIQUES

Forecasting

Forecasting is the process of making predictions of the future based

on past and present data. This is most commonly by analysis of
trends. Prediction is a similar, but more general term.

Sales Forecast

A sales forecast which is the estimate of how much the company will
actually sell. It is an estimate of expected sales revenue within a spe -
cific time frame, such as quarterly, monthly, or yearly.

Forecasting Horizons

 Long term forecasting tends to be completed at high levels in Seasonal

the organization.
 Medium term forecasting tends to be several months up to 2 Seasonality refers to regular fluctuations in demand that occur annu-
years into the future and is referred to as intermediate term. ally for a large number of products. Seasonality is quite important
 Short term forecasting is daily up to months in the future. when it comes to trends and outdoor items.

Forecasting Methods

 Qualitative forecasting techniques are subjective, based on the

opinion and judgment of consumers and experts.
 Quantitative forecasting models are used to forecast future data
as a function of past data.

Time Series Analysis

 A specific way of analyzing a sequence of data points collected

over an interval of time.
 Time is a crucial variable because it shows how the data adjusts
over the course of the data points as well as the final results. Irregular Variations

Demand Chart Events or a sequence of events that are not anticipated to occur again
can frequently have an impact on demand.
Demand Chart is a graphical representation of historical demand data
for a given time frame. It helps forecast future demand by visualizing Random Variations
trends, seasonal patterns, and possible cyclical behavior in the data.
The remaining unexplained changes in demand after all other factors
Elements of Demand Chart have been taken into account are known as random variations. This is
commonly referred to as noise.
Trends
Naive
It is a constant shift of the demand either up or down. This could be
connected to the life cycle of the product.  The simplest forecasting method.
 The forecast for the next period is set at the actual demand for
the previous period.

Formula

Y t =Y t −1

Example

Month Actual
Jan (1) 200
Feb (2) 300
Mar (3) 200
Apr (4) 400
May (5) 500
Jun (6) 600
Cycle Jul (7) -

A data pattern that appears frequently and lasts longer than a year is Moving Average
called a cycle. The political situation, consumer confidence, interest
rates, and other market conditions are frequently linked to them. A technique that averages a number of recent actual values, updated
as new values become available.

Formula

A t−n +… At −2+ A t−1

F t=MA n =
n
Note: n is the number of periods

Example
 Summer Class | Management Science

Week Demand Double Exponential Smoothing

1 125
2 175
 This method is also called exponential smoothing. Holt's trend
3 150
corrected or second-order exponential smoothing.
4 150
5 10  This method is used for forecasting the time series when the data
has a linear trend and no seasonal pattern. The primary idea be-
Solution hind double exponential smoothing is to introduce a term to take
into account the possibility of a series showing some form of
 Moving Average Example trend. This slope component is itself updated through exponen-
tial smoothing.
 Assume n=2

Triple Exponential Smoothing

Week Demand
1 125
2 175 In this method, exponential smoothing applied three times. This
3 150 (125 + 175)/2 = 150 method is used for forecasting the time series when the data has both
4 150 (175 + 150)/2 = 162.5 linear trend and seasonal pattern. This method is also called Holt-
5 10 (150 + 150)/2 = 150 Winters exponential smoothing.
(150 + 160)/2 = 155
Exponential Smoothing
Weighted Moving Average
This method uses a combination of the last actual value and the last
More recent values in a series are given more weight in computing forecast to produce the forecast for the next period.
the forecast.
Exponential Smoothing Formula
Formula
There are two versions of the same formula for calculating the expo-
t
nential smoothing:
∑ ( X i∗W i )
i=t −n
WMA= n  F t=F t + α ( A t−F t )
∑Wi  F t=α ( At ) +(1−α ) Ft
i =1

Where:
 t = total number of observations  α = Alpha/Smoothing constant
 i=t−n = series starts a total – window size  F t = Forecast value
 X i = value of ith previous day  At = Actual value
 W i = weight assigned to ith trailing days’ price
 i=t = the sum of all weights Sample Problem

Example Exponential Smoothing α =0.2

At Ft
 Assume n=2, W 1=0.7, W 2 =0.3
Week Sales Forecast F t=α ( At ) +(1−α ) Ft
Week Demand 1 39 39.00 F t+ 1=0.2 ( A t ) + 0.8(F ¿¿ t)¿
1 125 2 44 39.00 F t=0.2 ( 39 ) +0.8 (39.00)
2 175
3 150 3 40
4 150 4 45
5 10 5 38
6 43
Solution 7 39

 Assume n=2, W 1=0.7, W 2 =0.3 Exponential Smoothing α =0.2

At Ft
Week Demand Week Sales Forecast F t=α ( At ) +(1−α ) Ft
1 125
2 175 1 39 F t+ 1=0.2 ( A t ) + 0.8(F ¿¿ t)¿
3 150 (0.7)(175) + (0.3)(125) = 160
4 150 (0.7)(150) + (0.3)(175) = 157.5
2 44 39.00 F 2=A t
5 10 (0.7)(150) + (0.3)(150) = 150 3 40 40.00 F 3=0.2 ( 44 ) +0.8(39.00)
(0.7)(160) + (0.3)(10) = 157
4 45 40.00 F 4=0.2 ( 40 ) +0.8(40.00)
Exponential Smoothing 5 38 41.00 F 5=0.2 ( 45 ) +0.8 (40.00)
6 43 40.40 F 6=0.2 ( 38 ) +0.8(41.00)
There are three main methods to estimate exponential smoothing.
They are: 7 39 40.92 F 7=0.2 ( 43 )+ 0.8(40.40)
8 40.54 F 8=0.2 ( 39 ) +0.8(40.92)
 Simple or single exponential smoothing
 Double exponential smoothing
Forecasting Accuracy and Evaluation
 Triple exponential smoothing
Forecast accuracy is the measure of how accurately a given forecast
Simple or Single Exponential Smoothing matches actual sales. Forecast bias describes how much the forecast
is consistently over or under the actual sales. Common metrics used
 If the data has no trend and no seasonal pattern, then this method of to evaluate forecast accuracy include Mean Absolute Percentage Er-
forecasting the time series is essentially used. This method uses ror (MAPE), Mean squared error (MSE) and Mean Absolute Devia-
weighted moving averages with exponentially decreasing weights. tion (MAD).
 This forecasting method is most widely used of all forecasting tech-
niques. It requires little computation. Importance:
Summer Class | Management Science

Forecasting accuracy is crucial because it underpins efficient opera-

tions, informed decision-making, and strategic growth initiatives. By
improving the reliability of predictions about the future, businesses
can enhance their competitive advantage and adapt more effectively
to changing market conditions.

Period

1
Actual De-
mand

63
Fore-
cast

68
et |(e t )| et
2
[ ]
|( e t )|
Dt
×100 %

2 59 65
3 54 61
4 65 59

Here are what need to do:

 Step 1: Calculate the error as e t =D t −F t (the difference be-

tween the actual demand and the forecast) for any period t and
enter the values in the table above.
 Step 2: Calculate the absolute value of the errors calculated in
step 1 [i.e., |(e t )|], and enter the values in the table above.
2
 Step 3: Calculate the squared error (i.e., e t ) for each period and
enter the values in the table above.
 Step 4: Calculate ¿ for each period and enter the value under its
column in the table above.

Period

1
Actual De-
mand

63
Fore-
cast

68
et

-5
|(e t )|
5
et
25
2
[ ]
|( e t )|
Dt
7.94%
×100 %

2 59 65 -6 6 36 10.17%
3 54 61 -7 7 49 12.96%
4 65 59 6 6 36 9.23%

Solution:

Calculation for Accuracy measure:

 MAD = The average of what we calculated in step 2 (i.e., the av-

erage of all the absolute error values) =
(5+6 +7+6) 24
= =6
4 4
 MSE = The average of what we calculated in step 3 (i.e., the av-
erage of all the squared error values) =
(25+36 +49+36) 146
= =36.5
4 4
 MAPE = The average of what we calculated in step 4 = ¿ ¿