pt session 4. demand forecasting
Post on 08-Apr-2018
226 Views
Preview:
TRANSCRIPT
-
8/7/2019 PT Session 4. Demand Forecasting
1/35
1
IV. DEMAND FORECASTINGSource: Geetika, Salvatore
-
8/7/2019 PT Session 4. Demand Forecasting
2/35
Meaning
A tool to scientifically predict the likelydemand for a product for a particular
period of time in futureCategorization can be:By level of forecasting - firm, industry, andeconomy(macro) levelBy time period- short and long termBy nature of goods
2
-
8/7/2019 PT Session 4. Demand Forecasting
3/35
3
Demand Forecasting
To get an overview of the market and actproactively
To adjust production and avoid over production and under productionEssential for production scheduling,purchase of raw materials, arrangingfinance and advertising
-
8/7/2019 PT Session 4. Demand Forecasting
4/35
Q ualitative/ Subjective Methods
Used for short term forecasts when dataare not available
Also for supplementing quantitativeforecasts
4
-
8/7/2019 PT Session 4. Demand Forecasting
5/35
Q ualitative/ Subjective Methods
1. Consumers Opinion Surveys2. Sales Force Composite Method
3 . Expert Opinion- Group discussion andDelphi methods
4. Market Simulation
5. Test Marketing
5
-
8/7/2019 PT Session 4. Demand Forecasting
6/35
6
Q ualitative/ Subjective Methods
Surveys on economic intentions canreveal and can be used to forecast future
purchase of capital equipment, inventorychanges and major consumer expendituresComplete enumeration (census) vssampleQ uestionnaire, interview
-
8/7/2019 PT Session 4. Demand Forecasting
7/35
7
Q ualitative/ Subjective Methods
Buying Intentions of consumers has limited usebecause
- Consumer may not be able to clearly foresee the
choice,- Wishful thinking-Answers tailored to impress interviewer - New alternatives may emerge,-Passive because it does not measure variables
which are under management control-Intention may not translate into actual buying
-
8/7/2019 PT Session 4. Demand Forecasting
8/35
Q ualitative/ Subjective Methods
Sales Force Survey- Easy, cost effective,reliable, ideal for short term forecasts
ButMay be biasesSales force may not have knowledge of
macro factors
8
-
8/7/2019 PT Session 4. Demand Forecasting
9/35
9
Q ualitative/ Subjective Methods
Expert Opinion Method :Group Discussion- Market consultants,
industry analysts with knowledge of product and the market conditions- P ersonal insights can be subjective
- To avoid the problem of dominantpersonality, Delphi method
-
8/7/2019 PT Session 4. Demand Forecasting
10/35
10
Q ualitative/ Subjective Methods
Delphi method: Developed by Rand Corporation in19 40s- arriving at consensus
- Anonymity
-Wide expertise-Effective when there is no urgencybut
-Difficulty in getting panelists-Requires understanding, skill and knowledge for
conceptualising, stimulating discussion andmaking inferences by researcher .
-
8/7/2019 PT Session 4. Demand Forecasting
11/35
-
8/7/2019 PT Session 4. Demand Forecasting
12/35
-
8/7/2019 PT Session 4. Demand Forecasting
13/35
Q ualitative/ Subjective Methods
But very costly as it requires actualproduction of the product
Time consumingMay be difficult to extrapolate in a largecountry like India with heterogeneouspopulation
13
-
8/7/2019 PT Session 4. Demand Forecasting
14/35
14
Q UANTITATIVE METHODS
1. Time Series Method- Nave forecastingVariables change with timeSources of variation in Time series :
Secular TrendSeasonal Changes
Cyclical FluctuationsRandom or irregular fluctuations
Total variation , say in sales, is the result of all four
factors operating together .
-
8/7/2019 PT Session 4. Demand Forecasting
15/35
Time Series Method
Trend ProjectionSimplest- projecting the past trend by
fitting a straight line to the data either visually or more precisely throughregression
15
-
8/7/2019 PT Session 4. Demand Forecasting
16/35
16
L east Squares Method
L east Squares Method- Most widely used timeseries method
Linear Equation of a straight line isY = a + bX
where Y is the demand and X is the time period(no of years), a and b are constants depicting
intercept and slope of the line. Calculation of Yfor any value of X requires the values of a and b,for which 2 normal equations are prepared:
-
8/7/2019 PT Session 4. Demand Forecasting
17/35
17
L east Squares Method
Y = na + b X
XY=a X + b X2With values of a and b, straight line equation
is obtained and forecast is made for Y for
given value of X.
-
8/7/2019 PT Session 4. Demand Forecasting
18/35
18
Time Series Method
Smoothing Techniques:These predict values of a time series on the basis
of some average of its past values.
Useful when time series exhibit little trend or seasonal variations but a great deal of irregular or random variations
-
8/7/2019 PT Session 4. Demand Forecasting
19/35
19
Time Series Method
Moving Average Method:In this method, 3 (or 5) yearly / 3 (or 5) monthly/ 3 (or 5) day
moving averages of the required series is obtained. Themethod is used in progression by ignoring the first entryand incorporating the next one. A verage value of the last 3 (or 5) entries becomes theforecast for the next period
The best forecast is chosen with the help of Root MeanSquare Error (RMSE).
RMSE is an Exponential smoothing method .
-
8/7/2019 PT Session 4. Demand Forecasting
20/35
Time Series Method
Simple moving average gives equal weightto all observations, even though morerecent observations are likely to be moreimportant.Exponential smoothing overcomes thisproblem.
20
-
8/7/2019 PT Session 4. Demand Forecasting
21/35
21
B arometric ForecastB arometric Forecast :- When data indicates cyclical fluctuations, this
method alerts businesses to changes in overall
economic conditions- To predict short term changes in economic
activity or turning points
- Barometric forecasting is done by NBER andthe Conference Board
-
8/7/2019 PT Session 4. Demand Forecasting
22/35
B arometric Forecast
Related variables are categorised into 3 groups:
- L eading variables : Those that changebefore the actual change
- Coincident Variables : Change along withvariable
- L ag variables : Follow the event- If leading variables are identified, easy to
predict actual variables
22
-
8/7/2019 PT Session 4. Demand Forecasting
23/35
23
Barometric Forecast
PEAK
Time
C. Lagging
variable
TroughPEAK
-
8/7/2019 PT Session 4. Demand Forecasting
24/35
24
B arometric Forecast
L eading indicators :Building permits, new private housing unitsNumber of loan applicationsNew orders for durable goods for their components and raw materialsIndex of consumer expectations
Stock prices(Population growth for demand for schooling,Original purchase of durable goods for
replacement demand)
-
8/7/2019 PT Session 4. Demand Forecasting
25/35
25
B arometric Forecast
Coincident indicators :Rate of unemployment
GDPIndustrial productionManufacturing and trade sales
-
8/7/2019 PT Session 4. Demand Forecasting
26/35
26
B arometric Forecast
L agging Indicators :Commercial and industrial loans
outstandingChange in consumer price index for services
-
8/7/2019 PT Session 4. Demand Forecasting
27/35
B arometric Forecast
Merits:Barometric forecasting is 80 to 9 0% successfulin forecasting turning points in economic activity.Method aligns macro with micro economicsLess costly as it mostly depends on past data
27
-
8/7/2019 PT Session 4. Demand Forecasting
28/35
B arometric Forecast
Problems:Can Only be used for short term forecasting
It is not always possible to identify or get data ona leading variable for the variable to be forecastVariability in lead time may be considerableThe method cannot predict the magnitude of the
changes.;Therefore, barometric forecasting should be used
in conjunction with other methods
28
-
8/7/2019 PT Session 4. Demand Forecasting
29/35
29
Econometric Modelling for
ForecastingIdentifying and measuring the relationshipCan be single variable or multivariate regressionmodelSingle equation models for a firms demand;Large multiple equation models for the entireeconomy
Identify relevant determinants ( explanatory/independent variables) of the variable to beforecastGet data and fit to equation to get coefficients
-
8/7/2019 PT Session 4. Demand Forecasting
30/35
3 0
D= a 0+ a 1Y+ a 2 Px+a 3 Py+a 4 A+Not subjective like the qualitative methods
Based on causal relationships andproduces accurate resultsMethod is consistent
Forecasts both direction and magnitude of change
-
8/7/2019 PT Session 4. Demand Forecasting
31/35
3 1
Uses complex calculationsCostly and time consuming
-
8/7/2019 PT Session 4. Demand Forecasting
32/35
Problems in Demand
ForecastingInadequate analysis of the market- includeall potential users of a product
Unforeseen eventsChanging fashions and tastesConsumer Psychology
Lack of expertiseLack of adequate data
3 2
-
8/7/2019 PT Session 4. Demand Forecasting
33/35
33
Sums in Forecasting
Data for demand for watches for 5 years isgiven, estimate demand for 201 3 :
Year 2004 2005 2006
2007
2008No 120 1 3 0 150 140 1 6 0
-
8/7/2019 PT Session 4. Demand Forecasting
34/35
3 4
Sum in Forecasting
Year X Y X 2 Y2 XY
2004 1 120 1 14400 120
2005 2 1 3 0 4 1 69 00 2 6 0200 6 3 150 9 22500 450
200 7 4 140 1 6 196 00 5 6 0
2008 5 1 6 0 25 25 6 00 800
Total 15( X)n= 5
700 ( Y)
55( X2 )
99 000 2190 ( XY)
-
8/7/2019 PT Session 4. Demand Forecasting
35/35
3 5
Sums in ForecastingNormal equation: Y= a + bX (i)
Y= na + b X (ii) XY =a X + b X2 (iii)7 00= 5a + 15b (iv) 21 9 0=15a + 55b(v)
Solving (iv) and (v) we get,10 b= 9 0 b =9; Substituting the value of b in (iv)
7 00=5a+15* 9 5a=5 6 5 a= 113
Y=113
+9
X. For year 2013
, X will be 10.Y 201 3 = 11 3 + 9 * 10= 20 3 watches
top related