Chapter 3-

FORECASTING
 Learning Objectives:
• List the elements of a good forecasting
• Outline the steps in a good forecasting
system
FORECASTING • Describe at least 3 qualitative forecasting
Stevenson : 10th Edition
MCGRAW-HILL International techniques and the advantage and
Edition disadvantage of each
• Compare and Contrast Qualitative and
Quantitative Approaches to Forecasting
• Briefly describe averaging techniques, trend
and seasonal techniques and regression
analysis and solve typical problems
• Describe 3 measures of forecast accuracy
• Describe 2 ways of evaluating and
controlling forecast
• Identify the major factors to consider when
choosing a forecasting technique
 FORECAST:
• A statement about the future value of a variable of
interest such as demand.
• Forecasts affect decisions and activities throughout
an organization
• Accounting, finance
• Human resources
• Marketing
• MIS
• Operations
• Product / service design
Uses of Forecasts
 Features Common to all Forecasts

• Forecast techniques generally assume that the same

underlying causal system that existed in the past will
continue to exist in the future
• Forecasts are not perfect
• Forecast for groups of items tend to be more
accurate than forecasts for individual items
• Forecasts accuracy decreases as the time period
covered by the forecast
Elements of a good forecast

Timely

Accurate Reliable

Meaningfu Simple to
Cost-
l In Writing Understan
Effective
Units d and Use
STEPS in FORECASTING

“The forecast”

Step 6 Monitor the forecast

Step 5 Prepare the forecast
Step 4 Gather and analyze data
Step 3 Select a forecasting technique

Step 2 Establish a time horizon

Step 1 Determine purpose of forecast
 Forecast Accuracy

•Significant factor when deciding

among forecasting alternatives.
Accuracy is based on the
historical error performance of a
forecast.
 Forecast Accuracy
• Minimizing error
• Error - difference between actual value and
predicted value
• Error = Actual – Forecast
• et = At – Ft
• Positive errors result when the forecast is too low ,
negative errors when the forecast is to high
Ex. If actual demand for a week is 100 units and
forecast demand was 90 units the forecast was too
low : 100-90 = +10
 3 Commonly used measures for
Summarizing Historical Errors

• Mean Absolute Deviation (MAD)

• Average absolute error

• Mean Squared Error (MSE)

• Average of squared error
• Mean Absolute Percent Error (MAPE)
• Average absolute percent error
MAD, MSE, MAPE
Example:
Period Actual Forecast Error (A-F) |Error| Error2 [|Error|/Actual] * 100
1 217 215 2 2 4 0.92 % = (2/217)*100
2 213 216 -3 3 9 1.41
3 216 215 1 1 1 0.46
4 210 214 -4 4 16 1.90
5 213 211 2 2 4 0.94
6 219 214 5 5 25 2.28
7 216 217 -1 1 1 0.46
8 212 216 -4 4 16 1.89
-2 22 76 10.27 %

MAD = 22 = MSE = 76 = MAPE = 10.27% =

2.75 10.86 1.28 %
8 8-1 8
Example 2: MAD

Suppose you have the following actual sales data for a product over 5 months and the
corresponding forecasted sales:

Month Actual Sales Forecasted Sales

1 120 130

2 150 140

3 180 160

4 200 210

5 170 180
Steps to Calculate MAD:
1.Calculate the absolute deviation for each month (i.e., ∣Actual−Forecast∣).

| Month | Actual Sales | Forecasted Sales | Absolute Deviation (|Actual -

Forecast|) |

| 1 | 120 | 130 | 10 | | 2 | 150 | 140 | 10 | | 3 | 180 | 160 | 20 | | 4 | 200 | 210 |

10 | | 5 | 170 | 180 | 10 |

2.Sum up all the absolute deviations:

10+10+20+10+10=60

3.Divide the total by the number of months (n = 5):

MAD=60/5=12

So, the MAD in this example is 12.

Let’s use the same data as in Example 2 from earlier, where you have actual and forecasted sales data:

Month Actual Sales Forecasted Sales

1 120 130
2 150 140
3 180 160
4 200 210
5 170 180
Example 3: MSE
Squared
Forecasted Difference
Month Actual Sales
Sales (Actual−Foreca
st)2

1 120 130 (120−130)2=100

2 150 140 (150−140)2=100

3 180 160 (180−160)2=400

4 200 210 (200−210)2=100

5 170 180 (170−180)2=100

Steps to Calculate MSE:

1.Calculate the squared difference between the

actual and forecasted values:

2. Sum of the squared differences:

100+100+400+100+100=800

3. Divide by n−1, where n=5:

MSE=800/5−1=800/4=200

So, the MSE in this example is 200.

 Different Approaches to Forecasting

• Qualitative – Subjective (opinion) inputs

• Quantitative – mainly analyzing objective
or hard data
- avoid personal biases
 Forecasting Techniques

• Judgmental - uses subjective inputs

(Qualitative)
• Time series - uses historical data assuming
the future will be like the past (Quantitative)
• Associative models - uses explanatory
variables to predict the future (Quantitative)
 Judgmental Forecasts

• Executive opinions
• Sales force opinions
• Consumer surveys
• Outside opinion
• Delphi method
• Opinions of managers and staff
• Achieves a consensus forecast
 Time Series Forecasts
• Trend - long-term
movement in data
• Seasonality - short-term
regular variations in
data
• Cycle – wavelike
variations of more than
one year’s duration
• Irregular variations -
caused by unusual
circumstances
• Random variations -
caused by chance
 Associative Models

• Naïve Methods
• Moving Average Forecast
• Weighted Average Forecast TECHNIQUES
FOR
• Exponential Smoothing AVERAGING

• Simple linear trend smoothing

TECHNIQUES FOR
TREND
 Naive Forecasts – forecast for the next period is
equal to the forecast on the previous period
• Simple to use
Assume we have monthly sales data for a
• Virtually no cost company over five months:

• Quick and easy to •

•
Month 1: 100 units
Month 2: 120 units
prepare
• Month 3: 130 units
• Data analysis is • Month 4: 140 units
nonexistent • Month 5: 150 units

• Easily understandable We want to forecast the sales for Month

6.

• Cannot provide high Naive Method

accuracy Forecast for Month 6: The naive method

uses the most recent observation.
• Can be a standard for • Forecast: 150 units (since Month 5 had
accuracy 150 units).
TECHNIQUES FOR AVERAGING

Moving average – A technique that averages a number of

recent actual values, updated as new values become available.
- Average latest data

Period Demand Compute for a 3 period moving

1 42 average forecast given demand
for shopping carts for the last five
2 40
periods.
3 43
=
F6 = 43 +40+41
4 40 3 41.33
5 41
6 ?
 TECHNIQUES FOR AVERAGING
Weighted moving average – More recent values in a
series are given more weight in computing the forecast.
Example :
a.Compute a weighted Period Demand

average forecast using a 1 42 a. F6 = .10(40) + .20(43)

weight of .40 for the most 2 40 + .30(40) +.40 (41)
recent period, .30 for the = 41.0
3 43
most next recent
period , .20 for the next and 4 40 b. F7 = .10(43) + .20(40)
.10 for the next. 5 41 + .30(41) +.40(39)
b.If the actual demand for 6 39 = 40.2
period 6 is 39, forecast 7 ?
demand for period 7 using
the same
Weight distribution :

3 4 5
1/10 1/15
1/6
2/10 2/15
2/6 3/15
3/10
3/6
4/10 4/15
5/15
TECHNIQUES FOR AVERAGING
Exponential Smoothing – weighted averaging method
based on previous forecast plus a percentage of the forecast error
Next Forecast = Previous Forecast +  (Actual –
Previous Forecast)
Ft = Ft-1 + (At-1 - Ft-1)
Period request Actual F2 = 25 +0.5 (30-25)
1 25 30 = 25 + 0.5 (5)
2 =25+2.5
3
= 27.5

 =
0.5
 TECHNIQUES FOR TREND
Linear Trend Equation
Ft = a + Time (t) Sales (y) – in t2 ty
million peso

bt 1 (2013) 150 1 150

b= 2(2014) 157 4 314

3(2015) 162 9 486
2 2
n 4(2016) 166 16 664

a= 5(2017)
= 15
177
812
25
55
885
2499

b= 5 (2499) – = 315 = a= 812- = 717.5 =

15(812) 2 6.3 6.3(15) 143.5
n 5(55) - 15 50 5 5
Period Substituting formula Forecast
t = time
y = sales 6(2018) Ft = 143.5 + 6.3 (6) 181.3
n = number of
7(2019) Ft = 143.5 + 6.3 (7) 187.6
items
End of the presentation