Chapter Three

Forecasting
 Topics to be covered
 Introduction
 Importance of Forecasting
 Forecasting Range
 Techniques of Forecasting
 Qualitative
 Quantitative

1
Introduction to Forecasting
 A statement about the future value of a variable of

interest such as demand and a tool used for predicting

future value (demand) based on past information.
 Process of predicting a future event underlying basis of

all business decisions:

 Production

 Inventory

 Personnel

 Facilities
2
Importance of Forecasting
 Marketing managers:
 Use sales forecasts to determine optimal sales force
allocations.
 Set sales goals.
 Plan promotions and advertising.
 Planning for capital investments:
 Predictions about future economic activity.
 Estimating cash inflows accruing from the investment.
 The personnel department:
 Planning for human resources.
3
Importance of Forecasting
 Managers of nonprofit institutions:
 Forecasts for budgeting purposes.
 Universities:
 Forecast student enrollments.
 Cost of operations.
 The bank has to forecast:
 Demands of various loans and deposits
 Money and credit conditions so that it can determine
the cost of money it lends.

4
Importance of Forecasting
 Manufacturers:
 Worker absenteeism
 Machine availability
 Material costs
 Transportation and production lead times, etc.
 Service providers:
 Forecasts of population
 Demographic variables
 Weather, etc.

5
Principles of Forecasting
 Many types of forecasting models that differ in complexity

and amount of data & way they generate forecasts:

 Forecasts rarely perfect because of randomness.

 Forecasts more accurate for groups than individuals.

 Forecast accuracy decreases as time horizon increases.

6
Forecasting Range
 Short-range forecasts: Usually less than 3 months.

 Concerned with the daily operations of a business firm.

 Daily demand or resource requirements, Job scheduling and

worker assignment.

 A medium-range forecast:

 Encompasses anywhere from 3 months to 2 years.

 Yearly production and Sales planning and reflect peaks and

valleys in demand and the necessity to secure additional

resources for the upcoming year.

7
Forecasting Range
 A long-range forecast:

 Encompasses a period longer than 1 or 2 years.

 Long-range forecasts are related to management's

attempt to:
 Plan new products for changing markets.
 Build new facilities.
 Secure long-term financing.
 In general, the further into the future one seeks to predict,

the more difficult forecasting becomes.

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.

9
Forecasting Techniques
 Qualitative Methods
 Used when situation is vague and little data exist.
 New products
 New technology
 Innovative products
 Involves intuition, experience.
 Forecasts generated subjectively by the
forecaster.
 Educated guesses.
Forecasting Techniques
 Quantitative Methods
 Used when situation is ‘stable’ and historical data
exist.
 Existing products
 Current technology
 Involves mathematical techniques or mathematical
modeling.
 e.g., forecasting sales of color televisions.
 Commodity products that are sold every day.
Forecasting Techniques

12
Qualitative Forecasting Methods
 Individual Expert:

 Individual market experts can be hired to watch for industry

trends, to estimate future demand for products.

 Executive Opinions/Group Consensus:

 The subjective views of executives or experts from sales,

production, finance, purchasing, and administration are

averaged to generate a forecast about future sales.

13
Qualitative Forecasting Methods
 Delphi Method:

 Based on sequential questionnaires.

 Requires one person to administer and coordinate the

process and poll the team members (respondents) through a

series of sequential questionnaires.

 Consumer Surveys:

 Surveys regarding specific consumer purchases.

 Surveys may consist of telephone contacts, personal

interviews, or questionnaires as a means of obtaining data.

14
Quantitative Forecasting Techniques
 Quantitative analysis typically involves two approaches:

 Causal models
 Time-series methods

 Causal/Regression Methods:

 Causal models establish a quantitative link between

observable or known variable (like advertising

expenditures) with the demand for some product.

15
Causal Models
 Causal models establish a cause-and-effect relationship

between independent and dependent variables.

 A common tool of causal modeling is linear regression:

 Additional related variables may require multiple

regression modeling.
Y a  bX
 Y- Dependent Variable

 X- Independent Variable

16
Linear Regression
 Identify dependent (Y) and
independent (X) variables
 Solve for the slope of the
 XY  X  Y 
b
 X 2  X  X 
line:
b
 XY  n XY
2
 X  nX
2

 Solve for the y intercept:

a Y  b X
 Develop your equation for
the trend line:
 Y=a + bX
17
Linear Regression- Example
 A company has been tracking the relationship between
sales and advertising dollars. Use linear regression to find
out what sales might be if the company invested $53 in
advertising next year using the following previous data.
Sales $ Adv.$
(Y) (X)
1 130 32

2 151 52

3 150 50
4 158 55

18
Solution
Sales $
(Y)
Adv.$
(X)
XY X^2 Y^2
b
 XY  n X Y
2
1 130 32 4160 2304 16,900  X  nX 2

2 151 52 7852 2704 22,801

28202  447.25 147.25 
b 1.15
3 150 50 7500 2500 22,500 9253  447.25 
2

4 158 55 8690 3025 24964

a Y  b X 147.25  1.1547.25 
5 ?? 53 a 92.9
Tot 589 189 28202 9253 87165 Y a  bX 92.9  1.15X
Avg 147.25 47.25
Y 92.9  1.1553  153.85

1
Correlation Coefficient
 Correlation coefficient (r) measures the direction and

strength of the linear relationship between two variables.

The closer the r value is to 1.0 the better the regression line
fits the data points.
n  XY   X  Y 
r
 X   X  * n  Y  Y 
2 2
2 2
n
428,202  189589
r 2
.982
4(9253) - (189) * 487,165  589
2

r 2 .982  .964
2
Quantitative Forecasting Techniques
 Time Series Forecasting Methods
 Time series forecasting methods are:

 Based on analysis of historical data.

 Set of evenly spaced numerical data:

 Obtained by observing response variable at regular time

periods.
 Forecast based only on past values:

 Assumes that factors influencing past and present will

continue influence in future.

21
Time Series Patterns
 Historic data may exhibit one of the following pattern:

 Level (long-term average) – data fluctuates around a

constant mean.
 Trend – data exhibits an increasing or decreasing pattern.

 Seasonality – effects are similar variations occurring

during corresponding periods, can be quarterly, monthly,
weekly, daily, or even hourly indexes.
 Cycle – are the long-term swings about the trend line.

22
Time Series Patterns

23
Time Series Models
 Time Series : a set of observations measured at
successive times or over successive periods.
 Naïve or Projection
 Simple Moving Average
 Weighted Moving Average
 Exponential Smoothing

24
Time Series Models
 Naïve or Projection

 The forecast for the period t, Ft, is simply a projection of

previous period t-1 demand, At-1.

 Ft = At-1

 E.g. If the actual demand of period t is 120, then the forecast

of the period t+1 is 120.

 This method, although easy to use, doesn’t make use of

data that is easily available to most managers; thus, using

more of the historical data should improve the forecast.
25
Time Series Model
 Simple Moving Average (MA)

 An n-period moving average uses the last n periods of

demand as a forecast for next periods demand:

 Where n = total number of periods in the average,

 Ft= Forecast for period t,

 At-1, At-2,….At-n = Actual value for periods 1, 2,…, n.

26
Time Series Model
 Simple Moving Average (MA)

 To determine the length of n:

 Higher value of n - greater smoothing, lower

responsiveness.
 Lower value of n - less smoothing, more responsiveness.

 A large value of n is appropriate if the underlying pattern of

demand is stable.
 A smaller value of n is appropriate if the underlying pattern

is changing or if it is important to identify short-term

fluctuations.
27
Time Series Model
 Example: A company sells storage shed, Determine the forecast

of January using 3 month simple moving average.

28
Time Series Model
 Weighted Moving Average:

 A weighted moving average is a moving average where each

historical demand may be weighted differently.

 This runs counter to ones intuition that the most recent data is

the most relevant.

 Thus, the weighted moving average allows for more emphasis

to be placed on the most recent data. This forecast is:

29
Time Series Model
 Weighted Moving Average:

 Where wt-1 is the weight applied to the actual demand

incurred during period t-1, and so on.

 Intuitively, the expectation would be that the more recent

demand data should be weighted more heavily than older

data; so, generally, one would expect the weights to follow
the relationship wt ≥ wt-1 ≥ wt-2 ≥ ….

 The sum of the weights is one.

30
Time Series Model
 Weighted Moving Average: Consider the weights 3/6, 2/6,
1/6 for periods t-1, t-2 and t-3 respectively which are added to
one. Determine the forecast of January.

31
Time Series Model
 Exponential smoothing:

 Nice properties of a weighted moving average would be

one where the weights not only decrease as older and

older data are used, but one where the differences
between the weights are “smooth”.
 Obviously the desire would be for the weight on the most

recent data to be the largest.

32
Time Series Model
 Exponential smoothing:

 The weights should then get progressively smaller the

more periods one considers into the past.

 The exponentially decreasing weights of exponential
smoothing forecast fit this bill nicely.

33
Time Series Model
 Why use exponential smoothing?

 Uses less storage space for data

 More accurate

 Easy to understand

 Little calculation complexity

 The smoothing constant 𝜶 expresses how much our

forecast will react to observed differences.

 If 𝜶 is low: there is little reaction to differences.

 If 𝜶 is high: there is a lot of reaction to differences.

34
Time Series Model
 Selecting Smoothening Constant (α):

 The appropriate value of the smoothing constant, 𝜶,

however, can make the difference between an accurate

forecast and an inaccurate forecast.
 In picking a value for the smoothing constant, the

objective is to obtain the most accurate forecast.

 Several values of the smoothing constant may be tried,

and the one with the lowest MAD could be selected.

35
Time Series Model
 Exponential smoothing: Example

36
Time Series Model
 Exponential smoothing: Selecting smoothing constant.

 The smoothing constant with less MAD should be selected, thus α =

0.1
37
Selecting the Right Forecasting Model
 Selecting the right forecasting methods depends
on:
1. The amount & type of available data
 Some methods require more data than others

2. Degree of accuracy required

 Increasing accuracy means more data

3. Length of forecast horizon

 Different models for 3 month vs. 10 years

4. Presence of data patterns

38
Selecting the Right Forecasting Model
 Forecasting during product life cycle

39
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.

40
Measures of Forecast Error
 Mean Absolute Deviation (MAD)
 Measures the total error in a forecast without regard to sign

 Cumulative Forecast Error (CFE)

 Also called running sum of forecast error (RSFE)

 Measures any bias in the forecast

 Mean Square Error (MSE)

 Penalizes larger errors
n nn


AA
n
t -- F
Ft t  A - F 22
 At t - Ft t 
MAD
MAD =
=
t =1
t =1
t
MSE =
MSE =
t=
t =11 RMSE
RMSE== MSE
MSE
nn nn
n

CFE  actual  forecast  RSFE  (At  Ft )

i 1
 Ideal values = 0 (i.e., no forecasting error)
41
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.

RSFE
TS 
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

Usually 3 ≤ TS ≤ 8, out of this range investigate!

42
Measuring Forecast Accuracy and Error

Weighted (n=3,
Simple t-1=0.45, Exponential Exponential Exponential
S.N Actual Naïve
(n=3) t-2=0.35, (α=0.1) (α=0.5) (α=0.8)
t-3=0.2)
1 110 105 105 105
2 100 110.0 105.5 107.5 109.0
3 120 100.0 105.0 103.8 101.8
4 140 120.0 110.0 108.5 106.5 111.9 116.4
5 170 140.0 120.0 115.0 109.8 125.9 135.3
6 150 170.0 143.3 137.0 115.8 148.0 163.1
7 160 150.0 153.3 152.5 119.2 149.0 152.6
8 190 160.0 160.0 161.0 123.3 154.5 158.5
9 200 190.0 166.7 161.5 130.0 172.2 183.7
10 190 200.0 183.3 178.5 137.0 186.1 196.7
11 190.0 193.3 193.5 142.3 188.1 191.3
MAD 17.8 23.3 26.6 38.4 18.1 16.6
CFE 80.0 163.3 186.0 372.9 166.1 107.9
RMSE 58.3 74.5 81.8 141.5 72.5 61.1
TS 4.50 7.00 7.00 9.71 9.17 6.52

43
Quiz 2
Calculate forecasted
demand for week 10
using
A. naïve projection
B. four weeks simple
moving average
C. Waited moving
average using weight of
0.4, 0.3,0.2 and 0.1