Forecasting

Introduction
Weather forecasts are one of the many types of forecasts used
by some business organizations. Although some businesses
simply rely on publicly available weather forecasts, others turn
to firms that specialize in weather-related forecasts. For
example, Home Depot, Gap, and JCPenney use such firms to
help them take weather factors into account for estimating
demand.
Introduction
Every day, managers at Disney make decisions without
knowing what will happen in the future. They order inventory
without knowing what sales will be, purchase new equipment
despite uncertainty about demand for products, and make
investments without knowing what profits will be. Managers
are always trying to make better estimates of what will
happen in the future in the face of uncertainty. Making good
estimates is the main purpose of forecasting.
Example
The Walt Disney World forecasting department has 20
employees who formulate forecasts on volume and revenue for
the theme parks, water parks, resort hotels, as well as
merchandise, food, and beverage revenue by location.
Uses of forecasts in business
organizations
Accounting: New product/process cost estimates, profit projections, cash management.
Finance: Equipment/equipment replacement needs, timing and amount of funding/borrowing
needs.
Human resources: Hiring activities, including recruitment, interviewing, and training;
layoff planning, including outplacement counseling.
Marketing: Pricing and promotion, e-business strategies, global competition strategies.
MIS: New/revised information systems, Internet services.
Operations: Schedules, capacity planning, work assignments and workloads, inventory
planning, make-or-buy decisions, outsourcing, project management.
Product/service design: Revision of current features, design of new products or services.

In most of these uses of forecasts, decisions in one area have consequences in other areas.
Therefore, it is very important for all affected areas to agree on a common forecast.
Forecasting

“A statement “The art and

about the future science of
value of a variable predicting future
of interest.” events.”
Forecasting
Forecasting may involve taking historical data (such as past
sales) and projecting them into the future with a mathematical
model. It may be a subjective or an intuitive prediction (e.g.,
“this is a great new product and will sell 20% more than the old
one”). It may be based on demand-driven data, such as
customer plans to purchase, and projecting them into the future.
Or the forecast may involve a combination of these, that is, a
mathematical model adjusted by a manager’s good judgment.
Why forecasting is important

PREVENT MISMATCH BETWEEN TO MAXIMIZE PROFIT

DEMAND AND SUPPLY
Features of forecasting

Forecasts more
Forecasts rarely Forecast accuracy
Assumes Past accurate for
perfect because of decreases as time
reflects the Future groups vs.
randomness horizon increases
individuals
Elements of a Good Forecast

TIMELY RELIABLE ACCURATE MEANINGFUL EASY TO USE

Forecasting Time Horizons

Medium-
Short-range Long-range
range
forecast forecast
forecast
Short Range
Forecast
This forecast has a time span of up to
1 year but is generally less than 3
months. It is used for planning
purchasing, job scheduling, workforce
levels, job assignments, and
production levels
Medium-range
forecast
A medium-range, or intermediate,
forecast generally spans from 3
months to 3 years. It is useful in sales
planning, production planning and
budgeting, cash budgeting, and
analysis of various operating plans.
Long-range
Forecast
Generally 3 years or more in time span,
long-range forecasts are used in
planning for new products, capital
expenditures, facility location or
expansion, and research and
development.
Medium & Long Range Forecast vs Short
range Forecast
Medium- and long-range forecasts are distinguished from
short-range forecasts by three features:
First, intermediate and long-range forecasts deal with more
comprehensive issues supporting management decisions
regarding planning and products, plants, and processes.
Implementing some facility decisions, such as GM’s decision to
open a new Brazilian manufacturing plant, can take 5 to 8
years from inception to completion.
Medium & Long Range Forecast vs Short
range Forecast
Second, short-term forecasting usually employs different methodologies
than longer-term forecasting. Mathematical techniques, such as moving
averages, exponential smoothing, and trend extrapolation (all of which
we shall examine shortly), are common to short run projections.
Finally, short-range forecasts tend to be more accurate than longer
range forecasts. Factors that influence demand change every day. Thus,
as the time horizon lengthens, it is likely that forecast accuracy will
diminish. It almost goes without saying, then, that sales forecasts must
be updated regularly to maintain their value and integrity. After each
sales period, forecasts should be reviewed and revised.
THE BULLWHIP EFFECT
When businesses are situated further back in the supply chain,
inventory swings occur in larger waves in response to
customer demand, so that the largest impact of the whip hits
suppliers of raw materials, who feel the greatest demand
variation in response to customer demand variation. The result
is that supply chain participants build and maintain buffer or
safety stocks to compensate for swings in orders.
THE BULLWHIP EFFECT
As an example, the actual demand for a particular consumer
product is forecasted at 20 units per week. Even so, a retailer
might order 25 units from the distributor to protect against a
stock-out. In turn, the distributor might order 30 units from the
manufacturer to prevent future unfilled orders from the
retailer. The manufacturer receives the order for the 30 units
and orders enough raw materials to build 40 units, as an
added safety margin.
THE BULLWHIP EFFECT
Forty units will now be manufactured when there was only
customer demand for twenty units, creating a potential
oversupply of twenty units. Repeating this process puts
unnecessary demands on raw material suppliers, who are
furthest back in the supply chain, and might eventually force
manufacturers, distributors, and retailers to collaborate on
increasing demand by reducing prices and employing other
marketing techniques to sell excess inventory. This example
uses only one retailer; the effect when orders from multiple
retailers are exaggerated back through the supply chain could
be enormous.
Types of Forecast
Economic forecasts (Medium to long range forecast) address the
business cycle by predicting inflation rates, money supplies, housing
starts, and other planning indicators.
Technological forecasts (Long term forecast) are concerned with rates of
technological progress, which can result in the birth of exciting new
products, requiring new plants and equipment.
Demand forecasts(Projection of a company’s sales for each time period
in the planning horizon) are projections of demand for a company’s
products or services. Forecasts drive decisions, so managers need
immediate and accurate information about real demand.
Seven Steps in Forecasting
1. Determine the use of the forecast
2. Select the items to be forecasted
3. Determine the time horizon of the forecast
4. Select the forecasting model(s)
5. Gather the data
6. Make the forecast
7. Validate and implement results
Example
1. Determine the use of the forecast: Disney uses park attendance forecasts
to drive decisions about staffing, opening times, ride availability, and food
supplies.
2. Select the items to be forecasted: For Disney World, there are six main
parks. A forecast of daily attendance at each is the main number that
determines labor, maintenance, and scheduling.
3. Determine the time horizon of the forecast: Is it short, medium, or long
term? Disney develops daily, weekly, monthly, annual, and 5-year forecasts.
4. Select the forecasting model(s): Disney uses a variety of statistical models
that we shall discuss, including moving averages, econometrics, and
regression analysis. It also employs judgmental, or nonquantitative,
models.
Example
5. Gather the data needed to make the forecast: Disney’s forecasting
team employs 35 analysts and 70 field personnel to survey 1 million
people/businesses every year. Disney also uses a firm called Global
Insights for travel industry forecasts and gathers data on exchange
rates, arrivals into the U.S., airline specials, Wall Street trends, and
school vacation schedules.
6. Make the forecast.
7. Validate and implement the results: At Disney, forecasts are reviewed
daily at the highest levels to make sure that the model, assumptions,
and data are valid. Error measures are applied; then the forecasts are
used to schedule personnel down to 15-minute intervals.
Forecasting Approaches
There are two general approaches to forecasting, just as there are two
ways to tackle all decision modeling.
One is a quantitative approach; the other is a qualitative approach.
Quantitative forecasts use a variety of mathematical models that rely on
historical data and/or associative variables to forecast demand.
Subjective or qualitative forecasts incorporate such factors as the decision
maker’s intuition, emotions, personal experiences, and value system in
reaching a forecast. Some firms use one approach and some use the other.

In practice, a combination of the two is usually most effective.

Forecasting Approaches
Quantitative Qualitative
Based on mathematics, quantitative in Based on Human Judgement/ Opinion;
nature Subjective and nonmathematical
Consistent and objective; can process Can incorporate changes in the
many information and data at one time environment
Often enough reliable data are not Can cause bias and reduce the accuracy
available

26
Qualitative METHODS

Executive Opinion Market Research Delphi Method

Quantitative methods

Time Series Models Associative Models

Qualitative
Method
Qualitative Forecasting Techniques
Jury of Executive opinion
Delphi Method
Sales force composite
Market Survey
Jury of executive opinion
Under this method, the opinions of a group of high-level
experts or managers, often in combination with statistical
models, are pooled to arrive at a group estimate of demand.
It is good for forecasting of a new product launching, but here
one person's opinion can dominate the whole forecast.
Delphi Method
A forecasting technique using a group process that allows
experts to make forecasts. There are three different types of
participants in the Delphi method; decision makers, staff
personnel, and respondents. Decision makers usually consist
of a group of 5 to 10 experts who will be making the actual
forecast. It is an excellent method for long term forecasting.
However, it is very time consuming to develop.
Sales force composite
A forecasting technique based on salespersons’ estimates of
expected sales. In this approach, each salesperson estimates
what sales will be in his or her region. These forecasts are then
reviewed to ensure that they are realistic. Then they are
combined at the district and national levels to reach an overall
forecast. A variation of this approach occurs at Lexus, where
every quarter Lexus dealers have a “make meeting.” At this
meeting, they talk about what is selling, in what colors, and
with what options, so the factory knows what to build.
Market survey
A forecasting method that solicits input from customers or
potential customers regarding future purchasing plans. . It can
help not only in preparing a forecast but also in improving
product design and planning for new products. The consumer
market survey and sales force composite methods can,
however, suffer from overly optimistic forecasts that arise
from customer input. Also, it can also be difficult to develop
an effective questionnaire.
Quantitative
Method
Quantitative Forecasting Techniques

Associative
Time Series
Model

Naive Moving Exponential Trend Linear

approach averages smoothing projection regression
Time-Series Forecasting
Time-series models predict on the assumption that the future
is a function of the past. In other words, they look at what has
happened over a period of time and use a series of past data
to make a forecast.
These forecasts are based on a sequence of evenly spaced
(weekly, monthly, quarterly, and so on) data points.
Forecaster looks for data patterns as
Data = historic pattern + random variation
 Time-Series Forecasting
Day No. of Packets of Year Population (in
milk sold Million)
Monday 90 1921 251
From example 1 it is clear
Tuesday 88 1931 279 that the sale of milk
Wednes 85
1941 319
packets decrease from
day
Monday to Friday ,then
Thursda 75
y
1951 361 again its start to increase.
Friday 72 1961 439 In example 2 the
Saturda 90 1971 548
population is continuously
y increasing.
Sunday 102 1981 685
Decomposition of a Time Series
Analyzing time series means breaking down past data into
components and then projecting them forward. A time series
has four components:
1. Trend
2. Seasonality
3. Cycle
4. Random Variations
Trend
Trend is the gradual upward or downward movement of the data
over time. Changes in income, population, age distribution, or
cultural views may account for movement in trend.
population increases over a period of time, price increases over a
period of years, production of goods on the capital market of the
country increases over a period of years, These are the examples
of upward trend.
The sales of a commodity may decrease over a period of time
because of better products coming to the market. This is an
example of declining trend or downward.
Seasonality
PERIOD LENGTH “SEASON” NUMBER OF
Seasonality is a LENGTH “SEASONS” IN
data pattern that PATTERN
repeats itself after a Week Day 7
period of days,
weeks, months, or Month Week 4 – 4.5
quarters. There are Month Day 28 – 31
six common Year Quarter 4
seasonality Year Month 12
patterns.
Year Week 52
Cycle
Cycles are patterns in the data
that occur every several years.
They are usually tied into the
business cycle and are of major
importance in short-term
business analysis and planning.
Predicting business cycles is
difficult because they may be
affected by political events or by 0 5 10 15 20
international turmoil
Seasonality vs Cycle
Many people confuse cyclic behavior with seasonal
behavior, but they are really quite different. If the
fluctuations are not of fixed period then they are cyclic; if
the period is unchanging and associated with some aspect
of the calendar, then the pattern is seasonal.
In general, the average length of cycles is longer than the
length of a seasonal pattern, and the magnitude of cycles
tends to be more variable than the magnitude of seasonal
patterns.
 Random Variation
Random variations are “blips” in
the data caused by chance and
unusual situations. They follow no
discernible pattern, so they
cannot be predicted.
• Erratic, unsystematic,
‘residual’ fluctuations
• Due to random variation or
unforeseen events M T W T F

• Short duration
and nonrepeating
 Components of Demand

Trend
component
Demand for product or service

Seasonal peaks

Actual demand line

Average demand over

4 years

Random variation
| | | |
1 2 3 4
Time (years)
Time series method

Data = historic pattern + random variation

• Historic pattern to be forecasted:

• Level (long-term average) – data
fluctuates around a constant mean
• Trend – data exhibits an increasing or
decreasing pattern
• Seasonality – any pattern that regularly
repeats itself and is of a constant length
• Cycle – patterns created by economic
fluctuations
• Random Variation cannot be predicted

46
 Forecast Variations
Irregular
variation

Trend

Cycles

90
89
88
Seasonal variations
Three important Concepts…
1. Responsiveness (How close the forecasted data is with
actual data)
2. Smoothness (Randomness Causes the smoothness to drop)
3. Lag (Forecasting always lags behind the actual data)
 Naive Forecasts

Uh, give me a minute....

We sold 250 wheels last
week.... Now, next week
we should sell....

The forecast for any period

equals the previous period’s
actual value.
Naive Approach

Virtually no Quick and easy

Simple to use
cost to prepare

Data analysis is Easily Cannot provide

nonexistent understandable high accuracy
 Naive Forecasting
Month Demand Naive Month Demand Naive
January 110 January 35
Feb 98 110 Feb 36 35
March 89 98 March 34 36
April 100 89 April 33 34
May 112 100 May 35 33
June 89 112 June 36 35
July 90 89 July 37 36
August 90 August 37

Forecast Error = Actual - Forecast

Moving Averages
A forecasting method that uses an average of the n most recent
periods of data to forecast the next period.
A moving-average forecast uses a number of historical actual data
values to generate a forecast. Moving averages are useful if we can
assume that market demands will stay fairly steady over time . A
4-month moving average is found by simply summing the demand
during the past 4 months and dividing by 4. With each passing
month, the most recent month’s data are added to the sum of the
previous 3 months’ data, and the earliest month is dropped. This
practice tends to smooth out short-term irregularities in the data
series.
Moving Average
Mathematically, the simple moving average (which serves as
an estimate of the next period’s demand) is expressed as:
𝑆𝑢𝑚 𝑜𝑓 𝐷𝑒𝑚𝑎𝑛𝑑 𝑖𝑛 𝑝𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝑁 𝑝𝑒𝑟𝑖𝑜𝑑𝑠
𝑀𝑜𝑣𝑖𝑛𝑔 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 =
𝑁
where N is the number of periods in the moving average—for
example, 4, 5, or 6 months, respectively, for a 4-, 5-, or 6-
period moving average.
 Moving Average
MONTH ACTUAL SHED SALES 3-MONTH MOVING AVERAGE
January 10
February 12
March 13
April 16 (10 + 12 + 13)/3 = 11.667
May 19 (12 + 13 + 16)/3 = 13.667
June 23
July 26
August 30
September 28
October 18
November 16
December 14
 Moving Average
MONTH ACTUAL SHED SALES 3-MONTH MOVING AVERAGE
January 10
February 12
March 13
April 16 (10 + 12 + 13)/3 = 11.667
May 19 (12 + 13 + 16)/3 = 13.667
June 23 (13 + 16 + 19)/3 = 16
July 26 (16 + 19 + 23)/3 = 19.33
August 30 (19 + 23 + 26)/3 = 22.67
September 28 (23 + 26 + 30)/3 = 26.33
October 18 (26 + 30 + 28)/3 = 28
November 16 (30 + 28 + 18)/3 = 25.33
December 14 (28 + 18 + 16)/3 = 20.67
Weighted moving Average
When a detectable trend or pattern is present, weights can be
used to place more emphasis on recent values. This practice
makes forecasting techniques more responsive to changes
because more recent periods may be more heavily weighted.
Choice of weights is somewhat arbitrary because there is no
set formula to determine them. Therefore, deciding which
weights to use requires some experience.
Weighted Moving Average
► Used when some trend might be present
► Older data usually less important
► Weights based on experience and intuition
 Weighted Moving Average
MONTH ACTUAL SHED SALES 3-MONTH WEIGHTED AVERAGE
January 10
February 12
March 13
April 16 (1x10 + 12x2 + 13x3)/6 = 12.167
May 19 (12x1 + 13x2 + 16x3)/6 = 14.3333
June 23 (13x1 + 16x2+ 19x3)/6 = 17
July 26 (16x1 + 19x2 + 23x3)/6 = 20.5
August 30 (19x1 + 23x2 + 26x3)/6 = 23.8333
September 28 (23x1 + 26x2 + 30x3)/6 = 27.5
October 18 (26x1 + 30x2 + 28x3)/6 = 28.333
November 16 (30x1 + 28x2 + 18x3)/6 = 23.333
December 14 (28x1 + 18x2 + 16x3)/6 = 18.6667
Potential Problems With Moving Average

Increasing n smooths
the forecast but Does not forecast Requires extensive
makes it less trends well historical data
sensitive to changes
Solved Problems
Problem 1: Sales of hair dryers at the Walgreens stores in Youngstown, Ohio,
over the past 4 months have been 100, 110, 120, and 130 units (with 130 being
the most recent sales).
Develop a moving-average forecast for next month, using these three
techniques:
a) 3-month moving average.
b) 4-month moving average.
c) Weighted 4-month moving average with the most recent month weighted 4,
the preceding month 3, then 2, and the oldest month weighted 1.
d) If next month’s sales turn out to be 140 units, forecast the following month’s
sales (months) using a 4-month moving average.
Solved Problems
a) 3-month moving average = (110 + 120 + 130)/3 = 120 dryers
b) 4-month moving average = (100 + 110 + 120 + 130)/ 4 = 115
dryers
c) Weighted moving average = (4x130+3x120+2x110+1x100)/10
= 120 dryers
d) Now the four most recent sales are 110, 120, 130, and 140. 4
month moving average= (110 + 120 + 130 + 140)/4 =125 dryers
 Solved Problems
Problem 2: The following gives the number of pints of type B blood
used at Woodlawn Hospital in the past 6 weeks:
a) Forecast the demand for the week of October 12 using a 3-week
moving average.
Week of Pints used
b) Use a 3-week weighted moving August 31 360
average, with weights of .1, .3, and .6, September 7 389
September 14 410
using .6 for the most recent week.
September 21 381
Forecast demand for the week of September 28 368

October 12. October 5 374

Solved Problems
a) (374 +368 +381)/3=374.33
b) (381x.1 + 368x.3 + 374x.6) = 372.9
Solved Problems
Problem 3: The Carbondale Hospital is a) Forecast the mileage
considering the purchase of a new for next year (6th
ambulance. The decision will rest partly on year) using a 2-year
the anticipated mileage to be driven next moving average.
year. The miles driven during the past 5 b) Use a weighted 2-
years are as follows: year moving average
Year Mileage with weights of .4
1 3000 and .6 to forecast
2 4000
next year’s mileage.
3 3400
4 3800
5 3700
Solved Problems
a) Forecast for 6th Year = 3800+3700/2 = 3750
b) Forecast for 6th Year = (3800*4+3700*.6)/2 = 3740
Solved Problems
Month Sales
January 20 The monthly sales for Yazici Batteries, Inc.,
February 21 were as follows
March 15
April 14 Forecast January sales using each of the
May 13 following:
June 16 i) Naive method.
July 17 ii) A 3-month moving average.
August 18
iii) A 6-month weighted average using .1, .1,
September 20
.1, .2, .2, and .3, with the heaviest
October 20
weights applied to the most recent
November 21
months.
December 23
Solved Problems
i) Naive forecasting for the coming January = December = 23
ii) 3-month moving (20 + 21 + 23)/3 = 21.33
iii) 6-month weighted
[(0.1 x 17) + (.1 x 18)+ (0.1 x 20) + (0.2 x 20) + (0.2 x 21) + (0.3
x 23)]/1.0 = 20.6
Solved Problems
The actual demand for the patients at Omaha Emergency Medical Clinic
for the first 6 weeks of this year follows:
Week Actual no of Clinic administrator Marc Schniederjans wants you to forecast patient
Patients demand at the clinic for week 7 by using this data. You decide to use
a weighted moving average method to find this forecast. Your
1 65 method uses four actual demand levels, with weights of 0.333 on the
2 62 present period, 0.25 one period ago, 0.25 two periods ago, and 0.167
three periods ago.
3 70
a) What is the value of your forecast?
4 48 b) If instead the weights were 20, 15, 15, and 10, respectively, how
would the forecast change? Explain why.
5 63
c) What if the weights were 0.40, 0.30, 0.20, and 0.10, respectively?
6 52 Now what is the forecast for week 7?
Solved Problems
a) 52*0.333+63*0.25+48*.25+70*0.167/0.333+.25+.25+.167
=(17.36+15.75+12+11.69/1=56.8
b) 52*20+63*15+48*15+70*10/20+15+15+10
=(1040+945+720+700/60=56.75
c) 52*.40+63*0.30+48*0.20+70*0.10/0.40+.30+.20+0.10
=(20.8+18.9+9.6+7/1=56.3
Exponential Smoothing
Exponential smoothing is another weighted-moving-average forecasting method. It
involves very little record keeping of past data and is fairly easy to use. The basic
exponential smoothing formula can be shown as follows:
New forecast = Last period’s forecast + α(Last period’s actual demand − Last period’s
forecast)
where α is a weight, or smoothing constant , chosen by the forecaster, that has a value
greater than or equal to 0 and less than or equal to 1. It can also be written
mathematically
as:
Ft = Ft-1 + α (At-1 – Ft-1)
Summary
Exponential smoothing
A weighted-moving-average forecasting technique in which data points are
weighted by an exponential function.
Smoothing constant
The weighting factor used in an exponential smoothing forecast, a number
greater than or equal to 0 and less than or equal to 1.
Equation
Ft = Ft-1 + α (At-1 – Ft-1)
where F t = new forecast
Ft-1 = previous period’s forecast
α = smoothing (or weighting) constant (0 ≤ α ≥1)
At-1 = previous period’s actual demand
Smoothing Constant
The smoothing constant , α , is generally in the range from .05 to .50 for
business applications.
It can be changed to give more weight to recent data (when α is high) or
more weight to past data (when α is low). When α reaches the extreme
of 1.0, then in Equation (4-4) , F t = 1.0 A t -1 . All the older values drop
out, and the forecast becomes identical to the naive model mentioned
earlier in this chapter. That is, the forecast for the next period is just the
same as this period’s demand.
Smoothing Constant
Ft = Ft-1 + α (At-1 – Ft-1)
Ft = Ft-1 + 1 (At-1 – Ft-1)
Ft = At-1

As α Increases, Forecast becomes more responsive

As α decreases, Forecast appears smoother
 Solved Problems
The following gives the number of
Week of Pints used
pints of type B blood used at
August 31 360
Woodlawn Hospital in the past 6
September 7 389
weeks:
September 14 410
a) Compute the forecast for the September 21 381
week of October 12 using September 28 368
exponential smoothing with a October 5 374
forecast for August 31 of 360
and a = .2
Solved Problems
Week of Pints Forecast
August 31 360 360
September 7 389 360
September 14 410 365.8
September 21 381 374.64
September 28 368 375.91
October 5 374 374.32
October 12 374.26
Solved Problems
Sales of Volkswagen’s popular Year Sales Forecast
Beetle have grown steadily at auto 1 450 410
dealerships in Nevada during the 2 495
past 5 years (see table below). The 3 518
sales manager had predicted before
4 563
the new model was introduced that
5 584
first year sales would be 410 VWs.
Using exponential smoothing with a 6 ??

weight of a α=.30, develop forecasts

for years 2 through 6
Solved Problems
Year Sales Forecast
1 450 410
2 495 410+0.3x(450-410)=422
3 518 422+0.3x(495-422)=443.9
4 563 443.0+0.3x(518-443.9)=466.1
5 584 495.2=466.1+0.3x(563-
466.1)=495.2
6 ?? 495.2+0.3x(584-495.2)=521.8
Solved Problems
Consider the following actual and forecast
demand levels for Big Mac hamburgers at a local
Day Actual Forecast Demand
McDonald’s restaurant:
Demand
The forecast for Monday was derived by Monday 88 88
observing Monday’s demand level and setting
Monday’s forecast level equal to this demand Tuesday 72
level. Subsequent forecasts were derived by using Wednesday 68
exponential smoothing with a smoothing
constant of 0.25. Using this exponential Thursday 48
smoothing method, what is the forecast for Big Friday
Mac demand for Friday?
 Solved Problems

Day Actual Demand Forecast Demand

Monday 88 88
Tuesday 72 F2 = 88 + .25(88 – 88) = 88 + 0 = 88
Wednesday 68 F3 = 88 + .25(72 – 88) = 88 – 4 = 84
Thursday 48 F4 = 84 + .25(68 – 84) = 84 – 4 = 80
Friday F5 = 80 + .25(48 – 80) = 80 – 8 = 72
Measuring Forecast Error
The overall accuracy of any forecasting model—moving
average, exponential smoothing, or other—can be determined
by comparing the forecasted values with the actual or
observed values.
If Ft denotes the forecast in period t , and A t denotes the
actual demand in period t , the forecast error (or deviation) is
defined as:
Forecast error = Actual demand - Forecast value
= At - Ft
Measuring Forecast Error
Several measures are used in practice to calculate the overall
forecast error. These measures can be used to compare
different forecasting models, as well as to monitor forecasts to
ensure
they are performing well. Three of the most popular measures
are
1. Mean absolute deviation (MAD),
2. Mean squared error (MSE),
3. Mean absolute percent error (MAPE).
Common Measures of Error
Mean Absolute Deviation (MAD)

σ |𝐴𝑐𝑡𝑢𝑎𝑙 − 𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡|
𝑀𝐴𝐷 =
𝑛
Solved Problems
Quarter Actual Tonnage Forecast
unloaded
During the past 8 quarters, the 1 180 175
Port of Baltimore has unloaded
large quantities of grain from 2 168
ships. The port’s operations 3 159
manager wants to test the use of
4 175
exponential smoothing to see
how well the technique works in 5 190
predicting tonnage unloaded. He 6 205
guesses that the forecast of
grain unloaded in the first 7 180
quarter was 175 tons. Two 8 182
values of a are to be examined: 9 ????
α = .10 and α = .50.
Solved Problems
Quarter Actual Tonnage unloaded Forecast
1 180 175

2 168 175.5

3 159 174.75

4 175 173.18

5 190 173.36

6 205 175.02

7 180 178.02
8 182 178.22
9 ???? 178.59
Solved Problems

ACTUAL ABSOLUTE ABSOLUTE

TONNAGE FORECAST WITH DEVIATION FOR FORECAST DEVIATION FOR a =
QUARTER UNLOADED α = .10 α = .10 WITH α = .50 .5
1 180 175 5.00 175 5.00
2 168 175.50 7.50 177.50 9.50
3 159 174.75 15.75 172.75 13.75
4 175 173.18 1.82 165.88 9.12
5 190 173.36 16.64 170.44 19.56
6 205 175.02 29.98 180.22 24.78
7 180 178.02 1.98 192.61 12.61
8 182 178.22 3.78 186.30 4.30
Sum 82.45 98.62
MAD 10.31 12.33
On the basis of this comparison of the two MADs, a smoothing constant of α = .10 is
preferred to α = .50 because its MAD is smaller.
Common Measures of Error
Mean Squared Error (MSE)

σ(𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝐸𝑟𝑟𝑜𝑟)2
𝑀𝑆𝐸 =
𝑛
Solved Problems
Quarter Actual Tonnage unloaded Forecast
1 180 175

2 168 175.5

3 159 174.75

4 175 173.18

5 190 173.36

6 205 175.02

7 180 178.02
8 182 178.22
9 ???? 178.59
Solved Problems

σ(𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝐸𝑟𝑟𝑜𝑟)2
𝑀𝑆𝐸 = =1526.52/8=190.8
𝑛
Solved Problems
Is this MSE = 190.8 good or bad? It all depends on the MSEs
for other forecasting approaches. A low MSE is better because
we want to minimize MSE. MSE exaggerates errors because it
squares them.
LEARNING EXERCISE Find the MSE for α = .50
Answer: MSE = 195.24. The result indicates that α = .10 is a
better choice because we seek a lower MSE. Coincidentally,
this is the same conclusion we reached using MAD.
 Common Measures of Error
Mean Absolute Percent Error (MAPE)

σ𝑛𝑖=0 100 𝐴𝑐𝑡𝑢𝑎𝑙𝑖 − 𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡𝑖 /𝐴𝑐𝑡𝑢𝑎𝑙𝑖

𝑀𝐴𝑃𝐸 =
𝑛
 ACTUAL
TONNAGE FORECAST FOR ABSOLUTE % ERROR
QUARTER UNLOADED α = .10 100(ERROR/ACTUAL)
1 180 175.00 100(5/180) = 2.78%
2 168 175.50 100(7.5/168) = 4.46%
3 159 174.75 100(15.75/159) = 9.90%
4 175 173.18 100(1.82/175) = 1.05%
5 190 173.36 100(16.64/190) = 8.76%
6 205 175.02 100(29.98/205) = 14.62%
7 180 178.02 100(1.98/180) = 1.10%
8 182 178.22 100(3.78/182) = 2.08%
Sum of % errors = 44.75%

𝑆𝑢𝑚 𝑜𝑓 𝐴𝑏𝑠𝑜𝑙𝑢𝑡𝑒 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒 𝑒𝑟𝑟𝑜𝑟

𝑀𝐴𝑃𝐸 = =44.75%/8=5.59%
𝑛
Solved Problems
INSIGHT : MAPE expresses the error as a percent of the actual
values, undistorted by a single large value.

LEARNING EXERCISE : What is MAPE when α is .50?

[Answer: MAPE = 6.75%. As was the case with MAD and MSE,
the α = .1 was preferable for this series of data.]
Summary
α=0.1 α=0.5
MAD 10.31 12.33
MSE 190.92 195.24
MAPE 5.59% 6.76%
Solved Problems
As you can see in the following table, demand for
heart transplant surgery at Washington General
Hospital has increased steadily in the past few years. Year Heart Transparent
The director of medical services predicted 6 years ago 1 45
that demand in year 1 would be 41 surgeries.
2 50
a) Use exponential smoothing, first with a smoothing
constant of .6 and then with one of .9, to develop 3 52
forecasts for years 2 through 6. 4 56
b) Use a 3-year moving average to forecast demand 5 58
in years 4, 5, and 6. 6
c)With MAD as the criterion, which of the three
forecasting methods is best
 Solved Problems
Exponential Absolute Exponential Absolute
Year Demand Smoothing  = 0.6 Deviation Year Demand Smoothing  = 0.9 Deviation

1 45 41 4.0 1 45 41 4.0

2 50 41.0 + 0.6(45–41) = 43.4 6.6 2 50 41.0 + 0.9(45–41) = 44.6 5.4

3 52 43.4 + 0.6(50–43.4) = 47.4 4.6 3 52 44.6 + 0.9(50–44.6 ) = 49.5 2.5

4 56 47.4 + 0.6(52–47.4) = 50.2 5.8 4 56 49.5 + 0.9(52–49.5) = 51.8 4.2

5 58 50.2 + 0.6(56–50.2) = 53.7 4.3 5 58 51.8 + 0.9(56–51.8) = 55.6 2.4

6 ? 53.7 + 0.6(58–53.7) = 56.3 6 ? 55.6 + 0.9(58–55.6) = 57.8

Sum= 25.3 Sum= 18.5

MAD = 5.06 MAD = 3.7
Solved Problems
Three-Year Absolute
Year Demand Moving Average Deviation
1 45 - -
2 50 - -
3 52 - -
4 56 (45 + 50 + 52)/3 = 49 7
5 58 (50 + 52 + 56)/3 = 52.7 5.3
6 ? (52 + 56 + 58)/3 = 55.3
Sum= 12.3
MAD = 6.2
Month Order Month Order Naïve Abs Month Order 3MA abs Month Order 3MWM abs
ERROR ERROR A ERROR
Jan 121 Jan 121 Jan 121 Jan 121
Feb 95 Feb 95 121 26 Feb 95 Feb 95
Mar 100 Mar 100 95 5 Mar 100 Mar 100
Apr 90 Apr 90 100 10 Apr 90 105.33 15.33 Apr 90 101.92 12.7
May 125 May 125 90 35 May 125 95 30 May 125 94.15 31
Jun 100 Jun 100 125 25 Jun 100 105 5 Jun 100 109.2 9.5
Jul 88 Jul 88 100 12 Jul 88 105 17 Jul 88 106.55 17.5
Aug 121 Aug 121 88 33 Aug 121 104.33 16.67 Aug 121 98.25 22
Sep 102 Sep 102 121 19 Sep 102 103 1 Sep 102 106.54 4.9
Oct 96 Oct 96 102 6 Oct 96 103.67 7.67 Oct 96 105.89 8.9
Nov 100 Nov 100 96 4 Nov 100 106.33 6.33 Nov 100 102.23 2.8
Dec 122 Dec 122 100 22 Dec 122 99.33 22.67 Dec 122 99.02 22.8
Weights are 0.50,
Naïve (MAD) 3MA (MAD) 3 Month Weighted Moving(MAD)
0.33, and 0.17
17.91 13.52 14.65 respectively
Note:

Comparisons of these forecast summary measures for alternative forecasting methods using different
transformations of the data are not permissible.
For example:
1. If a researcher were comparing two forecasts, one generated from a model using the actual data and
another from a model in which the data have been transformed to logarithms, then a simple comparisons
of the forecast summary statistics would not be correct.
2. Comparison of forecast summary statistics for variable expressed in different frequencies (e.g., monthly
versus quarterly ) is generally not appropriate. How ever the forecast measure MAPE a unit free and could
be used to make such comparisons.
Thank You