Class 04: Forecasting – Basic Methods

 Chapter 9

Copyright © 2016 Pearson Education, Inc. 9-1

 Chapter Objectives
Be able to:
• Differentiate between demand forecasts, supply forecasts, and price forecasts.
• Describe the four laws of forecasting
• Select the most appropriate type of forecasting approach, given different
forecasting situations
• Describe four qualitative forecasting techniques
• Understand the components of a Time Series
• Be able to calculate the following forecasts:
• Last Period
• Moving Average
9-2
• Weighted Moving Average
 Introduction to Forecasting
 Forecast – An estimate of the future level of some variable.
 Underlying basis of all business decisions.
 Marketing: promotions and sales
 Supply Chain: purchasing, capacity, production, inventory
 Finance: Cash flow projections, profits
 Human Resources: hiring/firing
 MIS: user base size, technology development
 Challenge: The ACTUAL future level of the variable is likely to be either
higher or lower than the prediction, often by a significant amount. 9-3
 Forecast Types
• Demand forecasts
• Considering Overall Market Demand and
Firm-Level Demand Companies that are able to
develop better Supply Chain
• Supply forecasts forecasts experience:
 Predict material availability based upon • Lower inventories
suppliers, trends, risk • Reduced stockouts
• Smoother production plans
• Price forecasts • Reduced costs
• Improved customer service
• Forward buying, futures contracts, buying
frequency

• Economic forecasts
• Inflation rates, borrowing rates 9-4
 Laws of Forecasting
• Law 1: Forecasts are almost always wrong (but still useful)
• Law 2: Short term forecasts tend to be more accurate than longer term
forecasts.
• Law 3: Forecasts for Groups (categories) of Products or Services tend
to be more accurate than forecasts for specific products or services.
• Law 4: Forecasts are not a substitute for Calculated Values. Only use
forecasting when a more reliable method is not available.

4-5
 Steps in Forecasting
1. Determine how the forecast will be used
2. Select the values to forecast
3. Determine the planning time horizon of the forecast
4. Select potential forecasting model(s)
5. Gather historical data from which to forecast
6. Calculate forecasts using forecasting model(s)
7. Evaluate forecast accuracy & choose a forecasting model
8. Make future predictions based upon the forecasting model.

4-6
 Forecast Planning Time Horizons
• Long-range forecast (Asset Acquisition)
• Yearly planning bucket
• 3-10 years planning horizon
• New product planning, facility construction, technology

• Medium-range forecast (Asset Utilization)

• Monthly/Quarterly planning bucket
• 3 months to 2 years planning horizon
• Seasonal production, inventory, employment, budgeting

• Short-range forecast (Asset Execution)

• Weekly/Monthly planning bucket
• 1-26 week planning horizon
• Job scheduling, worker assignments, inventory stocking
 Forecasting Approaches
Qualitative Methods Quantitative Methods
¨ Subjective – Opinion Based ¨ Objective – Calculation Based
¨ Used when there is limited ¨ Used when quantitative historical
quantitative data available. data is available.
¨ Used when the relationship ¨ Used when the relationship
between the past and the between the past and the future is
future is uncertain. predictable.
¨ New or trendy products ¨ Existing or stable products
¨ New technology ¨ Current technology

¨ Involves intuition, experience ¨ Involves mathematical techniques

¨ e.g., forecasting sales of a new ¨ e.g., forecasting sales of products within a
product stable market
4-8
 Qualitative Forecasting Methods
 Market surveys
 Structured questionnaires or market research panels

 Panel consensus forecasting

 Experts meet together to develop forecasts

 Delphi method
 Experts develop forecasts separately & then revise

 Life-cycle analogy method

 Modeling growth and decline based upon similar products

 Build-up forecasts
 Market Segment experts develop forecasts thatPearson
Copyright © 2016 are Education,
added Inc.
together 9-9
 Key to Quantitative Forecasting:
Identify the underlying demand pattern
 Demand Patterns
 Level or Constant – Average value is relatively constant over time.
 Trend – Long-term movement up or down in a time series.
 Seasonality – A repeated pattern of spikes or drops associated with certain times of
the year.
 Cyclical – Long-term cycles of demand – often over several years.
 For instance: Product Life Cycles, Presidential Election impact on markets
 Randomness – Unpredictable movement from one time period to the next. This
component often serves to hide the underlying demand pattern. Random variation
is what makes forecasts especially difficult.
 Measured statistically by “variance” or “standard deviation” 9-10
 Quantitative Forecasting Approaches
Time Series Models: demand follows a trend and/or pattern over time.
• Last Period or Naïve Forecast
• Moving Average
• Weighted Moving Average
• Exponential Smoothing
• Adjusted Exponential Smoothing*
• Linear Regression*
Causal Models: demand is predicted by observing environmental factors
such as economic indicators
• Linear Regression*
• Multiple Regression*
4-11
 Quantitative Time Series
Forecasting Models
 Time Series
 A set of periodic observations arranged in chronological order
 A “period” is the regular frequency with which measurements are
plotted.
 Assumption
 The past is a good predictor of the future
 Risk
 The underlying demand pattern may change over time
 Trade-Off
 Forecast Responsiveness
 Forecast Stability 9-12
 Last Period or Naïve Approach
Current demand becomes the next period’s forecast.

Ft+1 = Dt
where Ft+1= forecast for the next period, t+1
and Dt = demand for the current period, t
This is the simplest time series model.
Very responsive to demand changes in the short-term, but
unstable for long-range planning
Major weakness is that it does not take into consideration
historical trends and patterns over time.
4-13
 Moving Average Model
Forecast averages the n most recent demand values.

• Smooths data randomness to illuminate data pattern

• Can be adjusted for the type of data being measured
• More Responsive …. Smaller n values (for data with frequent data pattern changes)
• More Stability …. Larger n values (for data with infrequent data pattern changes)
• Does not consider trend or seasonality
• NOTE: n=1 is the same as the Last Period forecast 4-14
Moving Average Model

Copyright © 2016 Pearson Education, Inc. 9-15

 Moving Average Forecasting Examples
2-Period and 4-Period
Moving Average
Forecasts
Example: 4-Period
F5 = (84+81+89+90)/4 = 86

Example: 2-Period F16

= (101 + 109)/2 =105

Which do you prefer?

Table 9.4
9-16
 Moving Average Example
Period Demand n=1 n=2 n=5
1 3 - - -
2 4 3 - -
3 2 4 3.5 -
4 3 2 3.0 -
5 2 3 2.5 -
6 4 2 2.5 2.8
7 1 4 3.0 3.0
8 3 1 2.5 2.4
9 2 3 2.0 2.6
10 4 2 2.5 2.4
11 ? 4 3.0 2.8
4-17
 Weighted Moving Average Method
A form of the moving average that applies varying weights to past observations.

• Weights are decimal values between 0 and 1

• Weights are listed, the sum of the weights must equal 1
• The first weight listed is applied to the most recent historical demand
• The last weight listed is applied to the n-th most recent historical demand
• Highest weights usually placed on more recent past demand
• Cyclical patterns can be modeled by adjusting weight values
• Responsiveness determined by weight values.
4-18
 Weighted Moving Average Forecasting
Example – Flavio’s Pizza revisited
 Flavio’s Pizza recorded the following sales history for the past five Friday nights.
 Prepare a sales forecast for week 6 using a 3-period moving average and a 3-period
weighted moving average using weights 0.4, 0.35, and 0.25.
Week Sales
MA 3-period forecast is:
1 62
F6 = (73 + 60 + 55)/3 = 188/3 = 62.67 pizzas
2 45
3 55
WMA 3-period forecast is:
4 60
5 73
F6 = . 4*73 + .35*60 + .25*55
= 29.2 + 21.0 + 13.75 = 63.95 pizzas
Why is this information important to know?
For this data set, which model do you prefer? Why?
9-19
 Weighted Moving Average
0.5, 0.2, 0.2, 0.1
Period Demand Calculation Forecast
1 30 -
2 40 -
3 20 -
4 30 -
5 20 0.5(30) + 0.2(20) + 0.2(40) + 0.1(30) 30.0
6 40 0.5(20) + 0.2(30) + 0.2(20) + 0.1(40) 24.0
7 10 0.5(40) + 0.2(20) + 0.2(30) + 0.1(20) 32.0
8 30 0.5(10) + 0.2(40) + 0.2(20) + 0.1(30) 20.0
9 20 0.5(30) + 0.2(10) + 0.2(40) + 0.1(20) 27.0
10 40 0.5(20) + 0.2(30) + 0.2(10) + 0.1(40) 22.0
11 ? 0.5(40) + 0.2(20) + 0.2(30) + 0.1(10) 31.0
Copyright © 2016 Pearson Education, Inc. 9-20
Copyright © 2016 Pearson Education, Inc. 9-21