FORECASTING

By
Alok Kumar Singh

A.K.Singh, IMI, Delhi

 Introduction
• An essential aspect of managing any organization is
planning for the future.
• A forecast is an estimate of future event achieved by
systematically combining and casting forward in
predetermined way the past data.
• Organizations employ forecasting techniques to determine
future inventory, costs, capacities, interest rate changes
and for many such applications.
• There are two basic approaches to forecasting:
-Qualitative
-Quantitative
A.K.Singh, IMI, Delhi
 Forecasting Time Horizons
1. Short-range forecast
► Up to 1 year, generally less than 3 months

► Purchasing, job scheduling, workforce levels,

job assignments, production levels

2. Medium-range forecast
► 3 months to 3 years

► Sales and production planning, budgeting

3. Long-range forecast
► 3+ years

► New product planning, facility location, R & D

A.K.Singh, IMI, Delhi

Strategic Importance of Forecasting
► Supply-Chain Management – Good
supplier relations, advantages in product
innovation, cost and speed to market
► Human Resources – Hiring, training,
laying off workers
► Capacity – Capacity shortages can result in
undependable delivery, loss of customers,
loss of market share

A.K.Singh, IMI, Delhi

 Forecasting Approaches
Qualitative Methods
► Used when situation is vague and
little data exist
► New products
► New technology
► Involves intuition, experience
► e.g., forecasting sales on Internet

A.K.Singh, IMI, Delhi

 Forecasting Approaches
Quantitative Methods
► Used when situation is ‘stable’ and
historical data exist
► Existing products
► Current technology
► Involves mathematical techniques
► e.g., forecasting sales of LED TV’s

A.K.Singh, IMI, Delhi

 Qualitative Approaches to Forecasting
• Executive Committee Consensus
– Knowledgeable experts of different functions
– Consider all functions like capacity, labor , sales etc.
• Delphi Approach
– A panel of experts, physically separated and anonymous,
– respond to a sequential series of questionnaires.
– the responses are tabulated and the information and opinions
of the entire group are made known to each of the other
members with new appropriate questions.
– The process continues until some degree of consensus is
achieved.
– At least 3 to 6 rounds required

A.K.Singh, IMI, Delhi

Qualitative Approaches (continued)
• Sales Force Survey
– Good estimate but chances of Bullwhip
• Survey of customers
– Good but expensive
• Historical analogue
– Cheap and easy but need to look into contemporary
factors
• Market research
– New products or existing product in new market
A.K.Singh, IMI, Delhi
 Quantitative Approaches
• Quantitative methods are based on an analysis of
historical data concerning one or more time series.
• A time series is a set of Set of evenly spaced numerical
data
• If the historical data used are restricted to past values of
the series that we are trying to forecast, the procedure is
called a time series method. Assumes that factors
influencing past and present will continue influence in
future
• If the historical data used involve other time series that
are believed to be related to the time series that we are
trying to forecast, the procedure is called a causal
method. (Regression Analysis)
A.K.Singh, IMI, Delhi
Time-Series Components

Trend Cyclical

Seasonal Random

A.K.Singh, IMI, Delhi

 Components of a Time Series
• The trend component accounts long term persistent
movement/tendency/change in data.
• There could be periodic, repetitive variations in time-series
which occur because of buying or consuming patterns and
social habits, during different times of a year. This is called
Seasonal component of the series.
• The variations in time series which arise out of the
phenomenon of business cycles is called Cyclical
Component. The duration and the level of resulting demand
variation due to business cycles are quite difficult to predict.
• The Random or irregular component of the series is caused
by short-term, unanticipated and non-recurring factors. For
E.g. Bharat Band, Truck Strike etc.

A.K.Singh, IMI, Delhi

 Components of Demand
Trend
component

Seasonal peaks
Demand for product or service

Actual demand
line

Average demand
over 4 years

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

A.K.Singh, IMI, Delhi

 Components of a Time Series
• The historical time series, as obtained from the past records, contains all
the four factors described earlier.
• One of the major tasks is to isolate each of the components, as elegantly as
possible. This process of desegregating the time series is called
decomposition.
• The main objective here is to isolate the trend in time series by eliminating
the other components. The trend line can then be used for projecting into
the future.
• In most short-term forecasting situations the elimination of the cyclical
component is not attempted.
• Also, it is assumed that the irregular variations are small and tend to
cancel each other out over time.
• Thus, the major objective, in most cases, is to seek the removal of seasonal
variations from the time series. This process is known as deseasonalization
of the time series data.
A.K.Singh, IMI, Delhi
 Scatter Plot
Month Sales
1 3 Sales
10
2 5
8
3 4
6 Sales
4 2
4
5 3
2
6 7
0
7 6 0 2 4 6 8 10 12
8 4
9 8
10 5

A.K.Singh, IMI, Delhi

 Naive Forecasting Models
• These are the simplest models.
• It assumes that the more recent time period of
data represents the best prediction or forecast
for future outcomes.
• The naive model says that the forecast for the
next period is the actual value of the last
period.

Ft  X t 1

A.K.Singh, IMI, Delhi

 Averaging Models
• Averaging models are computed by averaging data from
several time periods and use the average as the forecast
for the next time period.
• Simple Average : It is the average of the periods values
starting from a given point.
• The date is decided by organization in consultation with
experts.

Ft  ( X t 1  X t  2  X t 3  .......... .......  X t  n ) / n
Ft 1  ( X t  X t 1  X t  2  X t 3  .......... .......  X t  n ) / n  1

A.K.Singh, IMI, Delhi

 Moving Average
• Here the forecast is based on the data for a particular period,
which keeps on changing depending upon the time period taken.
• For e.g. 4 month moving average, 5 month moving average
• The number of periods considered is done in consultation with
experts.
• Sometime forecasting is done considering various periods and
error of estimates are compared for selecting the desired period.

Ft  ( X t 1  X t  2  X t 3  X t  4 ) / 4

Ft  ( X t 1  X t  2  X t 3  X t  4  X t 5 ) / 5
A.K.Singh, IMI, Delhi
 Weighted Moving Average
• Here certain weights are assigned to the values of
previous periods.
• The assigning of weight is done in consultation
with experts.
• Forecasting could be done with various weights &
periods and using the error in estimates these
weights & period could be finalized.

Ft  (a * X t 1  b * X t  2  c * X t 3  d * X t  4 ) / a  b  c  d

A.K.Singh, IMI, Delhi

 Exponential Smoothing
• This method is used to weight data from
previous period with exponentially decreasing
importance.
Ft  X t 1  (1   ) Ft 1
Ft  Ft 1   ( X t 1  Ft 1 )

Ft  X t 1  (1   )[X t  2  (1   ) Ft  2 ]
Ft  X t 1   (1   ) X t  2  (1   ) 2 Ft  2

A.K.Singh, IMI, Delhi

 Error in Forecasting
• Error in forecast in the difference between the
actual value and the forecasted value.
• Mean Absolute Deviation (MAD)
• Mean Square Error (MSE)
• Mean Absolute Percentage Error (MAPE)
n n
MAD  (  X t  Ft ) / n MSE  ( 
t 1
( X t  Ft ) 2 ) / n
t 1

100 Actual
i 1
i  Forecast i / Actuali
MAPE 
n

A.K.Singh, IMI, Delhi

 Numerical Exercise 1
Month Shipment
• The table shows the Jan 1056
shipments for electric Feb 1345
Mar 1381
lighting and wiring Apr 1191
equipments over a 12 May 1259
June 1361
month period. Forecast
July 1110
using Naive, Simple Aug 1334
Average method. Sep 1416
Oct 1282
Nov 1341
Dec 1382

A.K.Singh, IMI, Delhi

 Solution (Naive)
Month Shipment Forecast Error
Jan 1056 -
Feb 1345 1056 289
Mar 1381 1345 36
Apr 1191 1381 190
May 1259 1191 68
June 1361 1259 102
July 1110 1361 251
Aug 1334 1110 224
Sep 1416 1334 82
Oct 1282 1416 134
Nov 1341 1282 59
Dec 1382 1341 41
Average Error 134.1818182

A.K.Singh, IMI, Delhi

 Solution (Simple Average)
Month Shipment Simple Average Error
Jan 1056 -
Feb 1345 1056.00 289.00
Mar 1381 1200.50 180.50
Apr 1191 1260.67 69.67
May 1259 1243.25 15.75
June 1361 1246.40 114.60
July 1110 1265.50 155.50
Aug 1334 1243.29 90.71
Sep 1416 1254.63 161.38
Oct 1282 1272.56 9.44
Nov 1341 1273.50 67.50
Dec 1382 1279.64 102.36
Average Error 114.2194576

A.K.Singh, IMI, Delhi

 4 month Moving Average
Month Shipment 4 month Moving Error
Average

Jan 1056
Feb 1345
Mar 1381
Apr 1191
May 1259 1243.25 15.75
June 1361 1294 67
July 1110 1298 188
Aug 1334 1230.25 103.75
Sep 1416 1266 150
Oct 1282 1305.25 23.25
Nov 1341 1285.5 55.5
Dec 1382 1343.25 38.75
Average error 80.25

A.K.Singh, IMI, Delhi

 5 month Moving Average
Month Shipment 5 month moving Error
Average
Jan 1056
Feb 1345
Mar 1381
Apr 1191
May 1259
June 1361 1246.4 114.6
July 1110 1307.4 197.4
Aug 1334 1260.4 73.6
Sep 1416 1251 165
Oct 1282 1296 14
Nov 1341 1300.6 40.4
Dec 1382 1296.6 85.4
Average Error 98.62857

A.K.Singh, IMI, Delhi

4 Month Weighted Moving Average (4:2:1:1)
Month Shipment 4 month Weighted Error
Moving Average
Jan 1056
Feb 1345
Mar 1381
Apr 1191
May 1259 1240.875 18.125
June 1361 1268 93
July 1110 1316.75 206.75
Aug 1334 1201.5 132.5
Sep 1416 1272 144
Oct 1282 1350.375 68.375
Nov 1341 1300.5 40.5
Dec 1382 1334.75 47.25
Average Error 93.8125

A.K.Singh, IMI, Delhi

 Exponential Smoothing (alpha = 0.2)
Month Shipment Forecasted Value
(alpha = 0.2) Error
Jan 1056 1100 44
Feb 1345 1091.20 253.8
Mar 1381 1141.96 239.04
Apr 1191 1189.77 1.232
May 1259 1190.01 68.9856
June 1361 1203.81 157.1885
July 1110 1235.25 125.2492
Aug 1334 1210.20 123.8006
Sep 1416 1234.96 181.0405
Oct 1282 1271.17 10.8324
Nov 1341 1273.33 67.66592
Dec 1382 1286.87 95.13274
Average Error 110.3306

A.K.Singh, IMI, Delhi

 Exponential Smoothing (alpha = 0.5)
Month Shipment Error
Exponential
Smoothing (alpha =
0.5)
Jan 1056 1100 44
Feb 1345 1078.00 267
Mar 1381 1211.50 169.5
Apr 1191 1296.25 105.25
May 1259 1243.63 15.375
June 1361 1251.31 109.6875
July 1110 1306.16 196.1563
Aug 1334 1208.08 125.9219
Sep 1416 1271.04 144.9609
Oct 1282 1343.52 61.51953
Nov 1341 1312.76 28.24023
Dec 1382 1326.88 55.12012
Average Error 106.561
A.K.Singh, IMI, Delhi
A.K.Singh, IMI, Delhi
A.K.Singh, IMI, Delhi