Forecasting Techniques

• Managers require good forecasts of future events.

• Business Analysts may choose from a wide range of
forecasting techniques to support decision making.
• Three major categories of forecasting approaches:
1. Qualitative and judgmental techniques
2. Statistical time-series models
3. Explanatory/causal models

Copyright © 2021 Pearson Education Ltd. Slide - 1

Qualitative and Judgmental
Forecasting
• Qualitative and Judgmental techniques rely on
experience and intuition.
• They are necessary when historical data is not
available or when predictions are needed far into the
future.
• The historical analogy approach obtains a forecast
through comparative analysis with prior situations.
• The Delphi method questions an anonymous panel
of experts 2-3 times in order to reach a convergence
of opinion on the forecasted variable.
Copyright © 2021 Pearson Education Ltd. Slide - 2
Indicators and Indexes
• Indicators are measures that are believed to
influence the behavior of a variable we wish to
forecast.
• Indicators are often combined quantitatively into an
index, a single measure that weights multiple
indicators, thus providing a measure of overall
expectation.
– Example: Dow Jones Industrial Average (DJIA)

Copyright © 2021 Pearson Education Ltd. Slide - 3

Statistical Forecasting Models
• Time Series - a stream of historical data, such as weekly sales

– T = number of periods, t  1, 2, ..., T

• Time series generally have components such as:

– random behavior
– trends (upward or downward)
– seasonal effects
– cyclical effects
• Stationary time series have only random behavior.
• A trend is a gradual upward or downward movement of a time series.

Copyright © 2021 Pearson Education Ltd. Slide - 4

Example 9.4: Identifying Trends in a
Time Series
• Total Energy Production & Consumption
– General upward trend with some short downward trends;
the time series is composed of several different short trends.

Copyright © 2021 Pearson Education Ltd. Slide - 5

Seasonal Effects
• A seasonal effect is one that repeats at fixed intervals of
time, typically a year, month, week, or day.

Copyright © 2021 Pearson Education Ltd. Slide - 6

Cyclical Effects
• Cyclical effects describe ups and downs over a much
longer time frame, such as several years.

Copyright © 2021 Pearson Education Ltd. Slide - 7

Forecasting Models for Stationary
Time Series
• Moving average model
• Exponential smoothing model
– These are useful over short time periods when trend,
seasonal, or cyclical effects are not significant.

Copyright © 2021 Pearson Education Ltd. Slide - 8

Moving Average Models
• The simple moving average method is a smoothing method
based on the idea of averaging random fluctuations in the time
series to identify the underlying direction in which the time
series is changing.
• The simple moving average forecast for the next period is
computed as the average of the most recent k observations.

At  At 1   At  k 1
Ft 1  (9.1)
k

– Larger values of k result in smoother forecast models since

extreme values have less impact.

Copyright © 2021 Pearson Education Ltd. Slide - 9

Example 9.5: Moving Average
Forecasting
• The Tablet Computer Sales file contains the number of units
sold over the past 17 weeks.

• Three-period moving average forecast for week 18:

 A17  A16  A15  82  71  50
F18    67.67
3 3
Copyright © 2021 Pearson Education Ltd. Slide - 10
Spreadsheet Implementation of
Moving Average Forecasting

Copyright © 2021 Pearson Education Ltd. Slide - 11

Example 9.6: Using Excel’s Moving
Average Tool
• Data Analysis options

We do not recommend
using the chart or error
options because the
forecasts generated by
this tool are not properly
aligned with the data.

Copyright © 2021 Pearson Education Ltd. Slide - 12

Error Metrics and Forecast Accuarcy
n
• Mean absolute deviation  At  Ft
(MAD) MAD  t 1  9.2 
n
 At  Ft 
n 2

• Mean square error (MSE)
MSE  t 1  9.3
n

 At  Ft 2
n
• Root mean square error 
(RMSE) RMSE  t 1
 9.4 
n
n At  Ft
• Mean absolute percentage 
t 1 At
error (MAPE) MAPE  100  9.5 
n
Copyright © 2021 Pearson Education Ltd. Slide - 13
Example 9.7: Using Error Metrics to
Compare Moving Average Forecasts
• Tablet Computer Sales data
• 2-, 3-, and 4-period moving average models
• 2-period model has lowest error metric values.

Copyright © 2021 Pearson Education Ltd. Slide - 14

Exponential Smoothing Models
• Simple exponential smoothing model:
Ft 1  1    Ft   At
 Ft    At  Ft   9.6 
where Ft 1 is the forecast for time period t  1, Ft is the
forecast for period t, At is the observed value in period t,
and α is a constant between 0 and 1 called the smoothing
constant.

• To begin, set F1 and F2 equal to the actual observation in

period 1, A1.
Copyright © 2021 Pearson Education Ltd. Slide - 15
Example 9.8: Using Exponential Smoothing
to Forecast Tablet Computer Sales

Forecast for week 3 when   0.7 : 1  0.7  88    0.7  44   57.2

Copyright © 2021 Pearson Education Ltd. Slide - 16
Example 9.9: Finding the Best Exponential
Smoothing Model for Tablet Computer Sales

The forecast using α =0.6 provides the lowest error metrics.

Copyright © 2021 Pearson Education Ltd. Slide - 17
Example 9.10: Using Excel’s
Exponential Smoothing Tool (1 of 2)
• Select Data Analysis from the Analysis group and then choose
Exponential Smoothing.
• Note that Damping factor = 1−α
• The first cell of the Output Range should be adjacent to the first data
point.

Copyright © 2021 Pearson Education Ltd. Slide - 18

Example 9.10: Using Excel’s
Exponential Smoothing Tool (2 of 2)
• Exponential Smoothing tool results

Copyright © 2021 Pearson Education Ltd. Slide - 19

Forecasting Models for Time Series
with a Linear Trend
• Double moving average and double exponential
smoothing
• Based on the linear trend equation
Ft  k  at  bt k  9.7 
• The forecast for k periods into the future is a function of
the level at and the trend bt .

• The models differ in their computations of at and bt .

Copyright © 2021 Pearson Education Ltd. Slide - 20

Double Exponential Smoothing
• Estimates of the parameters are obtained from the
following equations:
at   Ft  1     at 1  bt 1 
bt    at  at 1   1    bt 1 (9.8)

• Initial values are chosen for a1 as A1 and b1 as A2  A1.

Equation (9.8) must then be used to compute at and bt
for the entire time series to be able to generate forecasts
into the future. The forecast for k periods beyond the last
period (period T) is
FT  k  aT  bT (k ) (9.9)
Copyright © 2021 Pearson Education Ltd. Slide - 21
Example 9.11: Double Exponential
Smoothing
• First ten years of data in the Excel file Coal Production
• Choose α = 0.6 and β = 0.4
– See text for computational details.

Copyright © 2021 Pearson Education Ltd. Slide - 22

Regression-Based Forecasting for
Time Series with a Linear Trend
• Simple linear regression can be applied to forecasting
using time as the independent variable.

Copyright © 2021 Pearson Education Ltd. Slide - 23

Example 9.12: Forecasting Using
Trendlines
• Coal Production data with a linear trendline

Note that the

linear model
does not
adequately
predict the
recent drop in
production after
2008.

Copyright © 2021 Pearson Education Ltd. Slide - 24

Forecasting Time Series with
Seasonality
• When time series exhibit seasonality, different techniques
provide better forecasts than the ones we have described:
– Multiple regression models with categorical variables
for the seasonal components
– Holt-Winters models, similar to exponential smoothing
models in that smoothing constants are used to
smooth out variations in the level and seasonal
patterns over time.

Copyright © 2021 Pearson Education Ltd. Slide - 25

Example 9.13: Regression-Based
Forecasting for Natural Gas Usage (1 of 3)
• Gas & Electric Excel file
• Use a seasonal categorical
variable with k = 12 levels.
• Construct the regression
model using k − 1 dummy
variables, with January being
the reference month.
gas usage = 0  1 time   2 February  3 March
  4 April  5 May   6 June   7 July
 8 August  9 September  10 October
 11 November  12 December

Copyright © 2021 Pearson Education Ltd. Slide - 26

Example 9.13: Regression-Based
Forecasting for Natural Gas Usage (2 of 3)

• Data matrix

Copyright © 2021 Pearson Education Ltd. Slide - 27

Example 9.13: Regression-Based
Forecasting for Natural Gas Usage (3 of 3)
• Final regression
results (time and
February were
insignificant)

gas usage  236.75  36.75 March  99.25 April

 192.25 May  203.25 June  208.25 July
 209.75 August  208.25 September
 196.75 October  149.75 November
 43.25 December
Copyright © 2021 Pearson Education Ltd. Slide - 28
Holt-Winters Models for Forecasting Time
Series with Seasonality and No Trend

• The Holt-Winters additive seasonality model with no trend applies

to time series with relatively stable seasonality and is based on the
equation
Ft  k  at  St  s  k (9.10)

• The Holt-Winters multiplicative seasonality model with no trend

applies to time series whose amplitude increases or decreases over
time and is
Ft  k  at St  s  k (9.11)

A chart of the time series should be viewed first to identify the

appropriate type of model to use.

Copyright © 2021 Pearson Education Ltd. Slide - 29

Holt-Winters Additive Seasonality
Model with No Trend (1 of 2)
• The level and seasonal factors are estimated as

Level component : at   ( At  St  s )  (1   )at 1

Seasonal component : St   ( At  at )  (1   ) St  s  9.12 

• The forecast for the next period is

Ft 1  at  St  s 1.

Copyright © 2021 Pearson Education Ltd. Slide - 30

Regression Forecasting with Causal
Variables
• In many forecasting applications, other independent
variables besides time, such as economic indexes or
demographic factors, may influence the time series.
• Explanatory/causal models, often called econometric
models, seek to identify factors that explain
statistically the patterns observed in the variable being
forecast, usually with regression analysis.

Copyright © 2021 Pearson Education Ltd. Slide - 31

Example 9.16: Forecasting Gasoline
Sales Using Simple Linear Regression
• Excel file Gasoline Sales
• Simple trendline using week as the independent variable

Predicted sales for week 11 = 812.99(11)  4790.1  13, 733 gallons

Copyright © 2021 Pearson Education Ltd. Slide - 32
Example 9.17: Incorporating Causal Variables in
a Regression-Based Forecasting Model (1 of 2)

• The average price per gallon changes each week, and this
may influence consumer sales. Average price per gallon is
a causal variable.
• Develop a multiple linear regression model to predict
gasoline sales using both time and price per gallon.

Copyright © 2021 Pearson Education Ltd. Slide - 33

Example 9.17: Incorporating Causal Variables in
a Regression-Based Forecasting Model (2 of 2)

• Multiple regression model sales=β0 +β1 week+β2 price/gallon

Predicted sales for week 11

 72, 333  508.7(11)  16, 463(3.80)  15, 368 gallons

Copyright © 2021 Pearson Education Ltd. Slide - 34

The Practice of Forecasting
• Judgmental and qualitative methods are used for
forecasting sales of product lines and broad company
and industry forecasts.
• Simple time-series models are used for short- and
medium-range forecasts.
• Regression methods are typically used for long-term
forecasts.

Copyright © 2021 Pearson Education Ltd. Slide - 35