MODULE 3

Forecasting
Prediction is very difficult,
especially if it's about the future.
Nils Bohr

What is forecasting all about?

We try to predict the

future by looking back
at the past

Demand for Mercedes E Class

Time
Jan Feb Mar Apr May Jun Jul Aug

Actual demand (past sales)

Predicted demand

Predicted
demand
looking
back six
months

Forecasting as a Planning Tool.

Forecasting is the first step of planning. Forecasting is estimation of future demand for
products and services and the resources necessary to produce these outputs.
Estimates of the future demand for products and services is commonly called Sales
Forecast. This is the starting point for all other planning in OM.
Important to estimate TIMING and MAGNITUDE of the demand.

Importance of Forecasting in OM:

1) New Facility Planning: To make allowances for lead time required to build factories, install
and implement new production processes.
2) Production Planning: To make allowances for lead time required to provide the production
capacity to produce these variable monthly demands.
3) Workforce Scheduling: To make allowances for lead time required to provide workforce
changes to produce the weekly demands. Workforce must be scaled up or down to meet the
demands by using reassignment, overtime, layoff's or hiring.
4) Financial Planning:

What is a forecast?
A statement about future values of a variable, in other words forecasts are prediction
of future, Something that can be predicted in advance.
Example of forecast
1) Weather forecast
2) Forecasting the demand of a product before it occurs
3) Manufacture according to the predicted demand

Companies that does demand forecasting: Wal-Mart, JCPenney, Gap, P & G etc.
Forecasting helps managers by reducing some of the uncertainty, thereby allowing
them to develop meaningful plans, how?

- Anticipating what buyers want

- Reasonable approximation

Forecasts are the basis for Budgeting, Planning, Capacity decisions, Sales, Production and Inventory, Personnel,
Purchasing etc. Forecasts affects decisions in all the departments in an organization
Accounting new product estimated cost, profit, projections.

Finance replacement of equipment, amount of funding/borrowing needs.

Human Resources hiring activities, layoff planning.
Marketing pricing and promotions etc.
MIS new/revised information systems.
Operations schedules, capacity planning, work assignments, inventory planning, make-or-buy decisions,
outsourcing etc.
Product/service design timeline to design a new product etc.
Forecasting is not an exact science, it is a blend of art and science.
It requires a 1) lot of experience.

2) Judgement.
3) Technical expertise.

Some general characteristics of forecasts

Forecasts are always wrong
Forecasts are more accurate for groups or families of
items
Forecasts are more accurate for shorter time periods
Every forecast should include an error estimate
Forecasts are no substitute for calculated demand.

Two uses of forecast.

1) Plan the productive system long term planning : Products and service to
offer, equipment, locations etc.

2) Plan the use of system short term and intermediate range planning.
Inventory, work force levels, purchasing, budgeting,scheduling etc.

Forecasting Time Horizons

Short-range forecasts:
Generally less than 3 months: could be up to a year
Used for purchase planning, work force scheduling and job
assignment

Medium-range forecasts
Varies from 1 year to 3 years
Used for sales planning, production planning, budgeting and
operating plan analysis.
Long-range forecasts:
Varies from 3 years and up
Used for new products planning, capital expenditures facility
location, research and development

Sources of data.
Sources of Data, used for Forecasting:
1) Salesforce Estimates:
2) Point of Sales (POS) Data Systems:
3) Forecasts from Supply Chain partners:
4) Trade/Industry Association Journals: For long term forecasting decisions.
5) B2B Portals/Market Places:

6) Subjective Knowledge:

Key issues in forecasting

1. A forecast is only as good as the information included

in the forecast (past data)
2. History is not a perfect predictor of the future (i.e.:
there is no such thing as a perfect forecast)

REMEMBER: Forecasting is based on the

assumption that the past predicts the future!
When forecasting, think carefully whether or not
the past is strongly related to what you expect to
see in the future

Types of Forecasting Models:

Qualitative Model:
a) Consumers Survey Method. (Complete Enumeration Survey, Sample Survey and Test
Marketing, End-use Method).

b) Salesforce Opinion Method Also known as Collective Opinion Method/Grass Roots

Approach.
c) Delphi Method Also known as Expert Opinion Method.
d) Past Analogy.

Quantitative Model:
a) Time Series Models:
- Naive Approach.
- Moving Averages Method ( Simple and Weighted ).
- Exponential Smoothing Method.
b) Trend Projection.
c) Linear Regression Analysis.

What should we consider when looking at

past demand data?

Trends
Seasonality
Cyclical elements
Random variation

Time Series: Moving average

The moving average model uses the last t periods in

order to predict demand in period t+1.
There can be two types of moving average models:
simple moving average and weighted moving average

The moving average model assumption is that the most

accurate prediction of future demand is a simple (linear)
combination of past demand.

Time series: simple moving average

In the simple moving average models the forecast value

is
At + At-1 + + At-n
Ft+1 =
n

is the current period.

Ft+1 is the forecast for next period

n is the forecasting horizon (how far back we
look),
A

is the actual sales figure from each period.

Example: forecasting sales at Kroger

Kroger sells (among other stuff) bottled spring water

Month

Bottles

Jan

1,325

Feb

1,353

Mar

1,305

Apr

1,275

May

1,210

Jun

1,195

Jul

What will
the sales
be for
July?

What if we use a 3-month simple moving average?

FJul =

AJun + AMay + AApr

= 1,227

What if we use a 5-month simple moving average?

FJul =

AJun + AMay + AApr + AMar + AFeb

5

= 1,268

Time series: weighted moving average

We may want to give more importance to some of the

data

Ft+1 = wt At + wt-1 At-1 + + wt-n Atn

wt + wt-1 + + wt-n (should be = 1)

t

is the current period.

Ft+1 is the forecast for next period

n
is the forecasting horizon (how far back we
look),
A

is the actual sales figure from each period.

w is the importance (weight) we give to each

period

Why do we need the WMA models?

Because of the ability to give more importance to what

happened recently, without losing the impact of the past.
Demand for Mercedes E-class

Actual demand (past sales)

Prediction when using 6-month SMA

Prediction when using 6-months WMA

Time
Jan Feb Mar Apr May Jun Jul Aug

For a 6-month
SMA, attributing
equal weights to all
past data we miss
the downward trend

Example: Kroger sales of bottled water

Month

Bottles

Jan

1,325

Feb

1,353

Mar

1,305

Apr

1,275

May

1,210

Jun

1,195

Jul

What will
be the
sales for
July?

6-month simple moving average

FJul

AJun + AMay + AApr + AMar + AFeb +

= AJan
6

= 1,277

In other words, because we used equal weights, a slight

downward trend that actually exists is not observed

What if we use a weighted moving average?

Make the weights for the last three months more than the
first three months

July
Forecast

6-month
SMA

WMA
40% / 60%

WMA
30% / 70%

WMA
20% / 80%

1,277

1,267

1,257

1,247

The higher the importance we give to recent data, the

more we pick up the declining trend in our forecast.

How do we choose weights?

1. Depending on the importance that we feel past data

has
2. Depending on known seasonality (weights of past
data can also be zero).

WMA is better than SMA

because of the ability to
vary the weights!

Time Series: Exponential Smoothing (ES)

Main idea: The prediction of the future depends mostly on

the most recent observation, and on the error for the latest
forecast.

Smoothin
g
constant
alpha

Denotes the
importance of the past
data smooths or
dampens older data

Why use exponential smoothing?

1. Uses less storage space for data

2. Extremely accurate

3. Easy to understand
4. Little calculation complexity
5. There are simple accuracy tests

Exponential smoothing: the method

Assume that we are currently in period t. We calculated

the forecast for the last period (Ft-1) and we know the
actual demand last period (At-1)

F
F

(A

F
)
t
t
1
t
1
t
1
The smoothing constant expresses how much weightage
is given to latest historical data

If is close to zero: very little weightage to latest historical

data.
If is one: More weightage given to latest historical data.

Example: bottled water at Kroger

Month

Actual

Forecasted

Jan

1,325

1,370

Feb

1,353

1,361

Mar

1,305

1,359

Apr

1,275

1,349

May

1,210

1,334

Jun

1,309

= 0.2

Example: bottled water at Kroger

Month

Actual

Forecasted

Jan

1,325

1,370

Feb

1,353

1,334

Mar

1,305

1,349

Apr

1,275

1,314

May

1,210

1,283

Jun

1,225

= 0.8

Impact of the smoothing constant

1380
1360
1340
1320
1300
1280
1260
1240
1220
1200

Actual
a = 0.2
a = 0.8

Trend..
What do you think will happen to a moving average or
exponential smoothing model when there is a trend in the
data?

Impact of trend

Sales
Actual
Data
Forecast

Regular exponential
smoothing will always
lag behind the trend.
Can we include trend
analysis in exponential
smoothing?

Month

Exponential smoothing with trend

FIT
t F
t T
t
Ft = Alpha(At-1) + (1 Alpha) (Ft-1 + Tt-1)
Tt = Beta(Ft Ft-1) + (1 - Beta) (Tt - 1)
FIT: Forecast including trend
Alpha: Forecast smoothing constant

Beta: Trend smoothing constant

The idea is that the two effects are decoupled,

(F is the forecast without trend and T is the trend

component)

Example: bottled water at Kroger

Alpha = 0.8

At

Ft

Tt

FITt
Beta = 0.5

Jan

1325

1380

-10

1370

Feb

1353

1334

-28

1306

Mar

1305

1344

-9

1334

Apr

1275

1311

-21

1290

May

1210

1278

-27

1251

1218

-43

1175

Jun

Linear regression in forecasting

Linear regression is based on

1. Fitting a straight line to data
2. Explaining the change in one variable through changes
in other variables.

dependent variable = a + b (independent variable)

By using linear regression, we are trying to explore which

independent variables affect the dependent variable

Least Squares Method of Linear Regression

YabX

a y bx
xy
nxy

b
x nx

FORECAST ERRORS:
Forecast Error is the numeric difference between forecasted demand and actual demand.

Errors can be classified as:

- Bias: When a consistent error is made. Eg: Failing to include the right variables,
Using the wrong relationship among variables, Mistakenly
shifting the
seasonal demand from where it normally
occurs etc.
- Random: Those that cannot be explained by the forecast model being used.
MEASUREMENT OF ERRORS:
1) Mean Absolute Deviation (MAD) : A forecast error measure that is the average
forecast error without regard to direction; calculated as:

MAD =

Sum of the absolute value of forecast error for all periods

----------------------------------------------------------------------------Number of Periods
n
= Summation | (Actual Demand Forecasted Demand) |
i=1
--------------------------------------------------------n

BIAS or MEAN FORECAST ERROR (MFE): Average forecast error with regard to
direction.
Unlike MAD, Bias indicates the directional tendency of forecast errors. If the forecast
repeatedly over-estimates actual demand, Bias will have a positive value, whereas,
Consistent under-estimation will be indicated by a negative value.

MFE

= Sum of the value of forecast error for all periods

----------------------------------------------------------------------------Number of Periods
n
= Summation (Actual Demand Forecasted Demand)
i=1
-------------------------------------------------------n

MEAN SQUARE ERROR (MSE): Is a measure of Forecast Accuracy. Here, the Mean of
the squares of deviations of forecast values from actual result is calculated.
n
MSE = Summation (Actual Demand Forecasted Demand)2
i=1 -------------------------------------------------------n
Large errors are penalized more than the small ones because of squaring.

MEAN ABSOLUTE PERCENTAGE ERROR (MAPE):

n
= Summation | Actual Demand Forecasted Demand |
I=1
------------------------------------------------------- * 100
Actual Demand
--------------------------------------------------------------------N

n
= Summation | At Ft|
i=1
------------ * 100
At
-----------------------N

Monitoring & Controlling Forecasting Models:

The appropriateness of a forecasting model depends on the nature of data
available. Since the nature of this data changes constantly, forecasts should be
reviewed and revised periodically.
There are 2 methods through which the appropriateness of a forecasting model
can be monitored:
1) By Comparison of Actual data with the forecasted values.

2) Tracking Signals : This checks whether the forecasting model is overestimating or under-estimating the forecasted value. A tracking signal is a
measurement of how well the forecast is predicting the actual values.
Why is it important to monitor the Forecasting models and reduce the
Forecast Errors??
- Helps Operations Managers in reducing the cost of the forecast errors.
- To allow an organization to plan its activities better.
- When all the departments of an organization base their work on the
same
forecast, their efforts are mutually supportive.

Tracking Signal =

RSFE
--------MAD

n
RSFE = Summation(Actual Demand Forecasted Demand)
i=1

Positive tracking signals indicate that the demand is greater than forecast. Negative
tracking signal means that demand is less than forecast.
A good tracking signal should center closely around zero.
When tracking signals are evaluated, they are compared with predetermined control
limits. When a tracking signal exceeds an upper or lower limit, there is some problem
with the forecasting method.