LOG 221 - INVENTORY
MANAGEMENT AND CONTROL
INVENTORY FORECASTING
INVENTORY FORECASTING
• What is Forecasting?
Forecasting is the process of organizing and analyzing information in
a way that makes it possible to estimate future demand
(production/sales/inventory).
Forecasting involves prediction, projection, or estimation of the level
or occurrences of uncertain future events.
Forecasting is an estimate or prediction of a future based on the past
and the present outlook in an organization.
Major Uses of Forecasting
► Production:-Production scheduling, inventory control etc
► Purchasing:-Determination of procurement requirements, scheduling of
purchases ,to get favorable prices [Link]
► Marketing:-Formulation of marketing strategies for products, setting of
sales quotas, scheduling of advertising expenditures and sales promotions
etc
► Personnel:-Planning of manpower requirements
► Finance:-Establishing of operating budgets, cash flow planning, capital
budget / expenditure decisions etc
► Top management:- overall planning and control of operations of the
company.
Demand Forecasting Models
Demand Forecasting Models
• Qualitative Forecasting Model
► Rely on Opinion; Subjective Judgment; Experience; and Expertise than on
historical data
► Good for new products with little or no demand history
► Soliciting Opinions:-
► Market Survey/ Surveys of buyer intentions:- Such as questionnaires, telephone
polls, and consumer interviews
► Delphi technique:- A body of experts, consulted separately, is asked to
arrive at a consensus opinion
► Sales force composite:-Based on the combined estimates of experienced
sales personnel.
Demand Forecasting Models
• Quantitative Forecasting Model
► Rely on historical data or associations among variables to develop
forecasts
► Good for established products with demand history
► Time Series; Economic Indicators; Econometric Models
Demand Forecasting Models
• Factors Determining the Preference of Model
► Cost
► Data Availability
► Staff Skills
► Availability of Hardware and Software
► Desired Accuracy
► Demand Pattern
• Stable pattern
• Unstable Pattern
Time Series Forecasting
Time series forecasting occurs when you make scientific predictions
based on historical time stamped data. It involves building models
through historical analysis and using them to make observations and
drive future strategic decision-making.
• Time series forecasting is the process of analyzing time series data using
statistics and modeling to make predictions and inform strategic
decision-making.
• Time series forecasting means to forecast or to predict the future value
over a period of time. It entails developing models based on previous
data and applying them to make observations and guide future strategic
decisions
Time Series Forecasting Models
1) Naïve Approach
2) Simple Moving Average
3) Weighted Moving Average
4) Exponentially Weighted Moving Average ( EWMA)
5) Regression Analysis
Time Series Forecasting Models
• Naïve Approach
► A simple way to forecast that assume demand in the next period
will be equal to demand in the most recent period.
► Example: If the demand of January for item X is 50 units, also we
can forecast the demand for February will be also 50 units.
Time Series Forecasting Models
Simple Moving Average
Average several periods of data
Dampen, smooth out changes
Use when demand is stable with no trend or seasonal pattern
n
Di
D̂t
i l n
Where:
Di = Demand in period
n = Number of period in moving average
Moving Average Example
Find the forecast of April-sept using 3 –moving average and May-sept
using 4-moving average for item X as shown in table below:
MONTH JAN FEB MARCH APRIL MAY JUNE JULY AUG SEPT
ACTUAL 10 12 13 14 16 17 18 19 20
STOCK
• SOLUTION
MONTH ACTUAL STOCK 3 MONTH MOVING 4 MONTH MOVING
AVERAGE FORECAST AVERAGE FORECAST
JAN 10
FEB 12
MARCH 13
APR 14 (10+12+13)/3=11.67
MAY 16 (12+13+14)/3= (10+12+13+14)/4=
JUNE 17 (13+14+16)/3= (12+13+14+16)/4=
JULY 18 (14+16+17)/3=
Weighted Moving Average
WiDi
WMA
i l W
Where:
Di = Demand in period i
Wi = The weight for period i
Weighted Moving Average
Example
• Find the forecast of April-sept using 3 month weighted moving average for item X as
shown in table below
MONTH JAN FEB MAR APR MAY JUNE JULY AUG SEPT
ACTUAL 10 12 13 14 16 17 18 19 20
STOCK
• Weights Applied Period
3 Last month
2 Two months ago
1 Three months ago
TOTAL6 Sum of weights
solution
month Actual stock 3 month weighted moving average
Jan 10
Feb 12
March 13
April 14 [(3x13)+(2x12)+(1x10)]/6=
May 16
June 17
July 18
► Ex2: For the data provided in the table below, determine the
forecast of demand for the April to July by using three moving
average period if weighted moving average approach is used with
0.2, 0.3, and 0.5 assigned as weights to the first, second and third
periods respectively
month Jan Feb March April May June July Aug Sept
Actual 37 40 41 37 48 38 36
stock
Exponentially Weighted Moving Average(EWMA)
Weights most recent data more strongly and reacts more to recent changes,
Relatively Accurate method
Ft+1 = a At + (1- a)Ft
Ft+1 = forecast for next period
At = actual demand for present period
Ft = previously determined forecast for present period
= weighting factor, smoothing constant, percentage of forecast error
example
week 1 2 3 4 5 6 7 8 9
demand 39 44 40 45 38 43 39 40 46
Given the weekly demand data
i) what are the exponential smoothing forecasts for
Periods(week 3-9 using a(weighting factor )=0.20
ii) Calculate mean square error
Ft+1 = a At + (1-
a)Ft
week demand a =0.2 a =0.6
1 39 NA
2 44 39
3 40 0.2(44)+(1-0.2)39=
4 45 0.2(40)+(1-0.2) _=
5 38
6 43
7 39
• Calculating mean square error with exponential smoothing
week demand forecast error error2
1 39
2 44 39 44-39=5 25
3 40
4 45
5 38
6 43
7 39
PRINCIPLES OF FORECASTING
• When planning a forecast, there are certain principles that should be considered:
1. The forecast will be more accurate for groups. Total units are easier to forecast than specific products.
The forecast that a company will sell a total of 2500 laths next year might be useful for cash flow
information, but there must be more details for manufacturing planning
2. The forecast will be more accurate for the short term. The farther out you go, the less accurate you are.
Every effort should be made to reduce the cumulative lead time required for the planned item. The
cumulative lead time is the total planned length of time to produce and distribute an item. It is the
longest combination of events, the critical path, necessary for product availability.
3. The forecast will be wrong. Although there will be a forecast error, it is most important to have an
estimate of that error. Through mathematical techniques, it is possible to estimate the probability of
error.
4. The forecast should be tested before using. There are many models to use for forecasting and it is
recommended to test the various techniques based on the same past history. The technique or model
that worked best in the past will most likely work best in the future.
5. The forecast is no substitute for actual demand. As there will be a degree of error in any forecast,
reducing lead time as much as possible, so as to allow actual demand history to have a greater impact
on the forecast, is most desirable.
EXERCISE
month Order per month
Jan 120
Feb 90
march 100
April 75
May 110
June 50
July
• Given the data above calculate forecast for July by using
• A)naïve approach
• B)Simple moving average
• C) Weigthed moving average with addition of weight of 1,2,3 respectively
• D)exponential moving average from Feb-July and mean square error where by constant is 0.5